The Routledge Encyclopedia of the Chinese Language

SAVE OFFLINE

The Routledge Encyclopedia of the Chinese Language Edited by Professor Chan Sin-wai Assisted by James Minett and Florence Li Wing Yee

The Routledge Encyclopedia of the Chinese Language will truly meet the academic, linguistic and pedagogical needs of those who are interested in the Chinese language in different capacities and for different reasons, such as Sinologists, Chinese linguists, and teachers and learners of Chinese as a second language. The Encyclopedia includes research on the changing landscape of the Chinese language by over 40 eminent academics in the field, from research on oracle bone and bronze inscriptions, to Chinese language acquisition, to the language of the mass media. This reference will be the most up-to-date and authoritative on the market; it will offer a guide to shifts over time in thinking about the Chinese language as well as providing an overview of contemporary themes, debates and research interests. The editors and contributors are assisted by an editorial board comprised of the best and most experienced Sinologists world-wide. The reference includes an introduction, written by the editor, which places the assembled texts in their historical and intellectual context. The Routledge Encyclopedia of the Chinese Language is destined to be valued by scholars and students as a vital research resource. Chan Sin-wai is Professor at the School of Humanities and Social Science, The Chinese University of Hong Kong (Shenzhen), China.

This page intentionally left blank

The Routledge Encyclopedia of the Chinese Language

Edited by Chan Sin-wai

Assisted by James Minett and Florence Li Wing Yee

First published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2016 selection and editorial matter, Chan Sin-wai; individual chapters, the contributors The right of Chan Sin-wai to be identified as author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data The Routledge encyclopedia of the Chinese language / edited by Chan Sin-wai; assisted by Florence Li Wing Yee and James Minett.   pages cm   Includes bibliographical references and index.   1.  Chinese language–Encyclopedias.  2.  Chinese philology–Encyclopedias.  3.  Chinese characters– Encyclopedias.  4.  Chinese language–Acquisition.  I.  Chan, Sin-wai, editor.  II.  Li Wing Yee, Florence, editor.  III.  Minett, James W., editor.   PL1031.R68 2015   495.1′3–dc23 2015017474 ISBN: 978-0-415-53970-8 (hbk) ISBN: 978-1-315-67554-1 (ebk) Typeset in Times New Roman by Graphicraft Limited, Hong Kong

Contents

List of Tables ix List of Figures xiii List of Consultant Editors xv List of Contributors xvi The Chinese Language: The Global, Historical, and Linguistic Aspects xxix Chan Sin-wai Acknowledgements xxxviii Routledge Encyclopedia of the Chinese Language   1 Ancient Chinese Alain Peyraube

1

 2 Cantonese Cheng Siu-Pong and Tang Sze-Wing

18

  3 Cantonese Romanization Cheng Siu-Pong and Tang Sze-Wing

35

  4 Chinese Characters John Jing-hua Yin

51

  5 Chinese Idioms Jiao Liwei

64

  6 Chinese Language and National Identity Andrew Simpson

90

  7 Chinese Language Education: Teacher Training Jane Orton v

104

Contents

  8 Chinese Language in a Global Context Liu Jin and Tao Hongyin

119

  9 Chinese Language Pedagogy Wu Weiping

137

10 Chinese Linguistics William S.-Y. Wang

152

11 Chinese Linguistics: Pragmatics Yan Jiang

184

12 Chinese Linguistics: Semantics Hsieh Shu-Kai

203

13 Chinese Morphology Jerome L. Packard

215

14 Chinese Neologisms: Word-formation Strategies in Chinese Antonella Ceccagno

227

15 Chinese: Parts of Speech Candice Chi-Hang Cheung

242

16 Chinese Proverbs and Popular Sayings Larry Herzberg

295

17 Chinese Psycholinguistics Jerome L. Packard

315

18 Chinese Rhetoric Andy Kirkpatrick

328

19 Chinese Slang Robert L. Moore

342

20 Chinese Syntax Walter Bisang

354

21 Chinese Taboo Amy He Yun

378

22 Chinese Xiehouyu 395 Grace Zhang 23 Classical Chinese Chris Wen-chao Li

408 vi

Contents

24 Computational Linguistics Xiaoheng Zhang

420

25 Corpus-based Study of Chinese Tony McEnery and Richard Xiao

438

26 Elastic Words in Chinese San Duanmu and Yan Dong

452

27 Hakka Dialect Lau Chun Fat

469

28 Hanyu Pinyin Lilly Lee Chen

484

29 Hong Kong Sign Language Gladys Tang

505

30 Lexicography Cheung Kam-Siu

531

31 Lexicography in the Contemporary Period Huang Chu-Ren, Li Lan, and Su Xinchun

545

32 Loanwords Miao Ruiqin

563

33 Mandarin Shi Dingxu

579

34 Mandarin and Other Sinitic Languages Hilary Chappell and Li Lan

605

35 Metaphor in Chinese: Cognition, Culture, and Society Zhuo Jing-Schmidt

629

36 Modern Chinese: Written Chinese Feng Shengli

645

37 Poetic Prosody Feng Shengli

664

38 Psycholinguistics: Reading Chinese Liang Tao and Alice F. Healy

685

39 Semantic Change in Chinese Janet Zhiqun Xing

706 vii

Contents

40 Standard Chinese Shi Dingxu

723

41 Syntax Li Yafei

736

42 Wade–Giles Romanization System Karen Steffen Chung

756

Index 777

viii

List of tables

1.1 37 Old Chinese initials 2 1.2 The six main vowels 2 Old Chinese codas 2 1.3 1.4 Middle Chinese initials 3 Middle Chinese main vowels 3 1.5 1.6 Middle Chinese codas 3 Initial consonants 47 3.1 3.2 Final consonants 47 48 3.3 Vowels 3.4 Tones 48 Ancient writing period 54 4.1 4.2 Modern writing period 54 Categories of Chinese character construction 56 4.3 4.4 Number of words and characters at each proficiency level 58 Chinese character strokes 60 4.5 5.1 Competitive idioms that mean ‘the difference/gap is large’ 83 7.1 ACTFL Program Standards for the Preparation of Foreign Language Teachers 108 9.1 ACTFL levels with sample linguistic functions 144 9.2 Summary of pragmatic points in the system of language use 146 10.1 Song of the Yue Boatman 156 10.2 Simplified chronology of Chinese history 161 10.3 Guan Ju, first poem of Shijing 165 10.4 Comparison of kanji, zhengti and jianti 171 172 10.5 Comparison of traditional and simplified sinograms 10.6 Comparison of three Romanization systems 173 174 10.7 Pinyin and some PTH obstruents 10.8 The vowel Spelled ‘i’ in Pinyin 175 175 10.9 Shapes of the Putonghua syllable 14.1 期 qī as a compound constituent and as a metacompound constituent 231 15.1 The list of prepositions that can be followed by -zhe 着 261 ix

List of tables

15.2 Prepositions functioning as prepositions and as verbs 263 15.3 Localizers in Chinese 280 403 22.1 Xiehouyu and others 24.1 Chinese font sizes in numbers, points, and mm 426 433 24.2 Rule-based English–Chinese machine translation with syntactic analysis 25.1 HSK graded lists and words and characters in Chinese 440 440 25.2 Coverage of top N words 25.3 A summary of the roles of corpora in lexicography in Chinese 442 452 26.1 Examples of elastic words in Chinese 26.2 Monosyllabic–trisyllabic pair 453 453 26.3 Either part of the long form can be the short form 26.4 More than one long form 453 453 26.5 An ordinary Chinese sentence and its English translation 26.6 Object required in a Chinese answer 453 454 26.7 Object not required in an English answer 26.8 Questions to be discussed with regard to elastic words 454 455 26.9 Semantic difference between short and long forms 26.10 Defining elastic words in Chinese 455 455 26.11 Three interchangeable expressions for ‘skilled worker’ 26.12 Mutual dictionary annotation of 煤 ‘coal’ and 煤炭 ‘coal-charcoal’ 456 456 26.13 Word pairs that do or do not form elastic words 26.14 Quantitative studies on the percentage of elastic words in Chinese 456 457 26.15 POS and length properties of Chinese morphemes 26.16 Example: monosyllabic word without elastic length 457 458 26.17 Morphologically and semantically related word pairs in English 26.18 Theories on why Chinese has so many elastic words 458 459 26.19 Predictions of the homophony theory and the prosody theory 26.20 Minimal word requirement in Chinese when there is no ambiguity 460 26.21 Homophone density and the percentage of elastic words among noun morphemes in Chinese. Nouns that are polysyllabic only are excluded. 462 A homophone density of 1 means a word has no homophone (but itself ) 26.22 Statistics of the data in Table 26.21, which show no correlation between 463 homophone density and the percentage of elastic words. 26.23 In [N N], 1+2 is disfavored 463 463 26.24 In [V O], 2+1 is disfavored 26.25 Phonological requirements (boldface indicates stress) 463 463 26.26 Analysis of [N N] 26.27 Analysis of [V O] 463 26.28 Word length patterns for [N N] in written Chinese (token frequencies) 464 26.29 Word length patterns for [V O] in written Chinese (token frequencies) 464 465 26.30 A comparison of how many things each theory can correctly account for 26.31 Elastic words created by truncation 465 466 26.32 Repetitive words (indicated by parentheses) in two Chinese sentences 26.33 A poem with extra positions to fill (three missing syllables per line) 466 466 26.34 Elastic words can fill extra positions 27.1 Monosyllabic Hakka words which can be traced back to Middle or even 473 Ancient Chinese 27.2 Onsets (18) 478 x

List of tables

27.3 Rimes (75) 478 27.4 Tones 479 480 27.5 Some lexical differences between Meixian and Hong Kong Hakka 27.6 Tones of the Bendihua group 480 Some examples of correspondence between the Gutian and Meixian rime 27.7  systems 482 482 27.8 The Gutian and Meixian tone systems compared 28.1 First-grade textbook 485 486 28.2 Xīzì Qíjì 28.3 Chinese syllable structure 491 492 28.4 Table of phonetic letters 28.5 Initial consonants 493 493 28.6 Finals 28.7 Tone marks 493 494 28.8 Tone mark placement 28.9 Vowel chart 494 494 28.10 Consonant chart 28.11 Four tones 495 497–498 28.12 Pinyin Chart 28.13 Comparison of retroflexs and alveolars in different systems 500 501 28.14 The use of the letter h in Wade–Giles and Pinyin 28.15 Mandarin homophonous words qishi and their correspondences 501 in Southern Min 28.16 Homophonous characters for qi [tɕ’i] 502 554 31.1 A list of Chinese corpora 31.2 The first Chinese–foreign language bilingual dictionaries 557 570 32.1 [Phonemic, semantic] hybrid loans 32.2 Mandarin substitution patterns of English plosives 572 573 32.3 Lack of stress-to-tone mapping 32.4 Native-likeness hierarchy of loanwords 575 575 32.5 Regional variations 33.1 Consonants of Mandarin 580 582 33.2 Glides of Mandarin 33.3 Vowels of Mandarin 582 584 33.4 Actual syllables with a glide as the onset 33.5 Initials of Pinyin 587 587 33.6 Finals of Pinyin 33.7 Mandarin pronouns 590 607 34.1 Sinitic languages of China 34.2 The eight dialect subgroups of Mandarin 609 610 34.3 Standard Mandarin tones 34.4 Pronouns of Standard Mandarin 610 34.5 Tone system in the Huojia dialect of Jin 获嘉方言 612 34.6 Pronoun paradigm for the Shanyin dialect of Jin (Shaanxi) 612 613 34.7 Changsha Xiang tones 34.8 Pronouns of Changsha Xiang 613 34.9 Pronominal paradigm: Longhui Xiang 隆回话 613 34.10 Tone system of Nanchang Gan 614 xi

List of tables

34.11 Pronouns of Nanchang Gan (Jiangxi province) 34.12 Singular personal pronouns in the Gan dialect of Qianshan 34.13 Tone system of the Jixi dialect of Hui (Anhui) 34.14 Jixi pronominal paradigm (Anhui) 34.15 Tone system of Shanghainese Wu 34.16 Pronouns of Shanghainese 34.17 Southern Min tone system (Taiwan) 34.18 Pronouns of Southern Min (Taiwan) 34.19 Tone system of Sung Him Tong Hakka (New Territories, Hong Kong) 34.20 Hakka pronominal paradigm (Bao’an, Huizhou subgroup) 34.21 Tone system of Hong Kong Cantonese 34.22 Examples of Cantonese tonal syllables 34.23 Cantonese pronominal paradigm (Hong Kong) 34.24 Tone system of the Tingzicun 亭子村 dialect of Southern Pinghua 34.25 Nanning pronominal paradigm 39.1 Summary of the polysemies of 23 lexemes 40.1 Shengmu 40.2 Yunmu 42.1 Some of Morrison’s spellings and descriptions of the phonetic values

xii

615 615 616 616 617 618 619 620 621 621 622 622 623 624 624 714 728 728 761

List of figures

Character component writing sequence 61–62 9.1 Major steps in CSL teaching syllabi focusing on language structure (LS) 138 9.2 Major steps in CSL teaching syllabi focusing on language use (LU) 143 9.3 Guiding structure in teaching materials preparation 147 Austronesian languages in Taiwan 157 10.1 10.2 The ‘initial China’ 160 Oracle bone inscription concerning Fuhao’s childbirth 162 10.3 10.4 A page from a rhyme book of the Song dynasty 168 Major dialects of Hanyu 170 10.5 12.1 The pyramid structure model 207 Syntagmatic relations from Chinese Word Sketch Engine 209 12.2 Chinese WordNet 210 12.3 12.4 Morpho-semantic linkage space 211 21.1 Pyramid hierarchy of words for death of people from different social classes in historical Chinese 386 24.1 Chinese character frequency of Hong Kong, mainland China, and Taiwan 431 A sample of a concordance for keyword ‘中文’ (Chinese) 432 24.2 25.1 Coverage of top N characters 441 Distribution and classification of Hakka 470 27.1 28.1 Beijing street sign 484 508 29.1a BEE (variation 1) 29.1b BEE (variation 2) 509 509 29.1c BEE (variation 3) 29.1d BEE (variation 4) 509 29.2a PERSON 人 510 29.2b FAIR 公平 510 29.2c INTRODUCE 介绍 510 29.3a BREAK_OFF_A_RELATION 511 29.3b INCOMPATIBLE 511 29.4a PIG 512 512 29.4b STUBBORN xiii

List of figures

29.5a BE_DETERMINED 512 29.5b SET_OFF 512 513 29.6 THANK_YOU 29.7a scold+CL_sem:human_entity 513 513 29.7b SCOLD 29.8a TEN 514 514 29.8b DIVIDE_EVENLY 29.8c DECREASE 514 515 29.8d INCREASE 29.9a FAR (X-plane) 515 516 29.9b CLOUD (Y-plane) 29.9c SIGN LANGUAGE (Z-plane) 516 516 29.9d FACE-PRETTY 29.9e HEALTHY 516 517 29.10 LOVE-SOMETHING-ARDENTLY 29.11a STOP 518 518 29.11b TRY 29.11c DOCTOR 518 518 29.11d SHARE 29.12a THINK and OLD (citation form) 519 519 29.12b THINK^OLD (assimilated thumb) 29.13 LEARN^HOUSE 520 520 29.14 PIG^MEAT 29.15 ROCKET 521 522 29.16a NOT_FEEL-ANNOYED 29.16b UNFAIR 523 29.17a stand+CL_sem:human 524 29.17b lie+CL_sem:human 524 Shuowen jiezi 532 30.1 30.2 Guangyun 536 Zihui 537 30.3 30.4 Kangxi zidian 538 The first tone 585 33.1 33.2 The second tone 585 The third tone 585 33.3 33.4 The fourth tone 585 Diagram of formal and informal Chinese 646 36.1 36.2 Modern formal Chinese 651 Tridimensional system of stylistic-register grammar 654 36.3 36.4 Interrelationships among elevated, formal and casual forms 655 Evidence for tridimensional register grammar 655 36.5 36.6 660 Stylistic wave (of formal Chinese) Visual differences between a Chinese and an English sentence 687 38.1

xiv

List of Consultant Editors

William S.-Y. Wang Department of Chinese and Bilingual Studies Hong Kong Polytechnic University, Hong Kong, China Huang Chu-Ren Department of Chinese and Bilingual Studies Hong Kong Polytechnic University, Hong Kong, China Feng Shengli Department of Chinese The Chinese University of Hong Kong, Hong Kong, China Cao Guangshun Institute of Linguistics Chinese Academy of Social Sciences, China

xv

List of contributors

Bisang, Walter Walter Bisang Dr. phil. I (Zürich) has been Professor of General and Comparative Linguistics at the University of Mainz (Germany) since 1992. He was the Director of a Collaborative Research Center on ‘Cultural and Linguistic Contact’ from 1999 to 2008 in Mainz. His research interests focus on linguistic typology, grammaticalization, language contact/areal typology, and the comparison of different theoretical approaches to language. His languages of interest are East and mainland Southeast Asian languages, Caucasian languages (Georgian and others), Austronesian languages (Bahasa Indonesia, Tagalog, Yabêm, Paiwan), and Yoruba (together with Remi Sonaiya). Ceccagno, Antonella Antonella Ceccagno is associate professor of Chinese Linguistics at the University of Bologna, Italy, where she also teaches Sociology of East Asia. In the field of Chinese Linguistics her interests revolve around Chinese morphology. More recently, she is also focusing analytical attention on the interactions between language and society. Her publications on Chinese linguistics and sociology of language include: ‘Farewell to tuhao, welcome to tuhao: Language and Society in China as they Emerge from the Buzzwords of the Last Decade’ 2014, in Magda Abbiati and Federico Greselin (eds) Il Liuto e i libri, Venezia, Cafoscari, 193–204; ‘Sino-Tibetan: Mandarin Chinese’ in Rochelle Lieber and Pal Stekauer (eds) The Oxford Book of Compounding, Oxford: Oxford University Press, 2009, 478–90, (with Bianca Basciano); ‘Metacompounds in Chinese’, Lingue e Linguaggio 8(2), 2009: 195–212; ‘The Chinese Language and Some Notions from Western Linguistics’, Lingue e Linguaggio, 8(1), 2009: 105–35 (with Bianca Basciano); Shuobuchulai, la formazione delle parole in cinese (‘Shuobuchulai (Word Formation in Chinese)’), Bologna: Serendipità, 2009 (with Bianca Basciano); ‘Compound Headedness in Chinese: An Analysis of Neologisms’, Morphology, n 17, 2007: 207–31 (with Bianca Basciano); ‘Classification, Structure and Headedness of Chinese Compounds’ in Lingue e linguaggio 2, 2006: 233–60 (with Sergio Scalise). Chan, Sin-wai Chan Sin-wai is Professor at the School of Humanities and Social Science, The Chinese University of Hong Kong, Shenzhen. His teaching and research interests lie mainly in the areas of translation studies, translation technology, and bilingual lexicography. He is the xvi

List of contributors

Chief Editor of Journal of Translation Technology, published by The Chinese University of Hong Kong. He has published more than 44 books in 54 volumes, mainly dictionaries and scholarly monographs, and translated works in different fields. He edited An Encyclopaedia of Translation and The Routledge Encyclopedia of Translation Technology (2015), revised Longman Dictionary of English Language and Culture (bilingual edition), and authored A Dictionary of Translation Technology and A Chinese–English Dictionary of the Human Body. His book translations from Chinese into English include An Exposition of Benevolence, Palaces of the Forbidden City, Letters of Prominent Figures in Modern China, Paintings and Calligraphy of Jao Tsung-I, Stories by Gao Yang, An Illustrated History of Printing in Ancient China, Famous Chinese Sayings Quoted by Wen Jiabao, and Selected Works of Cheng Siwei: Economic Reforms and Development in China, Vol. 2. He also translated My Son Yo Yo from English into Chinese. His most recent co-edited books include Style, Wit and Word-Play (2012) and The Dancer and the Dance (2013), both published by Cambridge Scholars Publishing. Chappell, Hilary Hilary Chappell is currently Chair Professor in Linguistic Typology of East Asian Languages at the Ecole des Hautes Etudes en Sciences Sociales in Paris, an appointment she took up in 2005 after teaching in the Linguistics Department at La Trobe University in Melbourne, Australia, for 18 years. She was originally awarded her doctoral degree in 1984 by the Australian National University in Canberra for her thesis entitled ‘A Semantic Analysis of Passive, Causative and Dative Constructions in Standard Chinese’ and has over 60 publications on Chinese linguistics and typology, including four edited books and a volume on Hakka, co-authored with Christine Lamarre. Her main research is on the typology of Sinitic languages (Chinese dialects) and the extent of their linguistic diversity, for which she directed an ERC Advanced Grant project from 2009 to 2013. She has also held two projects with Professor Alain Peyraube (CNRSEHESS) to investigate the diachronic syntax of Southern Min (Hokkien), using a corpus of late sixteenth and early seventeenth century materials. In 2010, she was elected member of the Academia Europaea. Chen, Lilly Lee Lilly Lee Chen received her doctoral and master’s degrees in linguistics from the University of Illinois, Urbana-Champaign, and her bachelor’s degree from the Department of Foreign Languages and Literatures at National Taiwan University. She taught for 30 years in the Department of Linguistics and the Center for the Study of Languages at Rice University. In the course of her teaching career at Rice she established a variety of language oriented new programs and taught linguistics and Chinese language and culture, including synchronic and diachronic syntax, structure and symbolism of the classical novel, Chinese cinema, and Taiwanese language, culture, and linguistics from a cognitive perspective. She has served as President of the Linguistic Association of Canada and the United States (2013), President of the Chinese American Educational Research and Development Association (2000), and Founding President of the Chinese Language Teacher Association of the South (2006). She was also the Founding Director of the Institute for Chinese Language Teacher Training (2007–2010). Cheng, Siu-Pong Cheng Siu-Pong is Research Assistant at T. T. Ng Chinese Language Research Centre of the Institute of Chinese Studies, The Chinese University of Hong Kong. His research interests focus on Cantonese discontinuous constructions, the cartographic approach, and empty categories. xvii

List of contributors

Cheung, Chi-Hang Candice Candice Chi-Hang Cheung is Research Assistant Professor in the Department of Chinese and Bilingual Studies at the Hong Kong Polytechnic University. Her research interests lie in formal syntax, the syntax-semantics interface, the syntax-information structure interface, and parametric syntax with special focus on Cantonese, Mandarin and other varieties of Chinese. She has published articles in a number of linguistic journals, such as the Journal of East Asian Linguistics and Natural Language and Linguistic Theory. Cheung, Kam-Siu Cheung Kam-Siu received his PhD in Chinese Language and Literature from The Chinese University of Hong Kong in 2007. He is Associate Professor in the Department of Chinese Language and Literature and Honorary Research Associate of D. C. Lau Research Centre for Chinese Ancient Texts at The Chinese University of Hong Kong. His research interests focus on Chinese philology, ancient Chinese texts studies, and lexicography. He has recently published two books on the research of the prominent Qing semanticist and textual critic Wang Niansun (1744–1832), including On Wang Niansun’s Collated Edition of Ancient Texts (Shanghai Ancient Books Publishing House, 2014), and Collected Annotations on Wang Niansun’s Collated Edition of the Hanfeizi (The Chinese University Press, 2014). He is also an editorial board member of Newsletter of Chinese Language. Chung, Karen Steffen Karen Steffen Chung (史嘉琳 Shǐ Jiālín), originally from St. Paul, Minnesota, USA, has taught English, linguistics, and phonetics in the Department of Foreign Languages and Literatures of National Taiwan University since 1990. She is currently an Associate Professor at National Taiwan University. She received her BA in East Asian Languages from the University of Minnesota in 1976, her MA in East Asian Studies from Princeton University in 1981, and obtained her PhD in Linguistics from Leiden University in 2004. The title of her doctoral dissertation is: ‘Mandarin Compound Verbs’. Dong, Yan Yan Dong received her MA in Linguistics and Applied Linguistics from Renmin University of China in 2008 and joined the PhD program in Linguistics at the University of Michigan in the same year. In July 2015 she recently defended her dissertation, which is entitled ‘The prosody and morphology of elastic words in Chinese: annotations and analyses’. Duanmu, San San Duanmu is Professor of Linguistics, University of Michigan. He received his PhD in Linguistics from MIT in 1990 and has held teaching posts at Fudan University, Shanghai (1981–1986) and the University of Michigan, Ann Arbor (1991–present). He is the author of The Phonology of Standard Chinese (2nd ed., Oxford, 2007) and Syllable Structure: The Limits of Variation (Oxford, 2008). Feng, Shengli Feng Shengli began his academic career as an undergraduate student in the History Department of Beijing Normal University in the class of 1977. He was accepted by the Graduate Program in the Department of Language and Literature at the Beijing Normal University in 1979, majoring in Ancient Chinese Exegesis under the supervision of Professor Lu Zongda. After completing his MA degree, he taught Classical Chinese and Language and Literature in the xviii

List of contributors

Department of Language and Literature at Beijing Normal University. He received his PhD in Linguistics from the University of Pennsylvania in 1995. He taught Chinese language, Chinese linguistics, and Chinese culture in the Department of East Asian Languages and Cultures at the University of Kansas from 1994, and in 1999 he was promoted to Associate Professor. He was appointed in 2003 as Professor of the Practice and Director of the Chinese Language Program at Harvard University and Director of Harvard-Beijing Academy. In 2005 he was appointed as Yangtze Scholar Adjunct Professor (长江学者讲座教授) at Beijing Language and Culture University. He is the author of The Prosody, Morphology, and Syntax of Chinese 漢語的韻律、詞法與句法 (1997), Prosodically Constrained Syntactic Changes in Early Archaic Chinese (1996), The Prosodic Syntax of Chinese (2002), and A Preliminary Theory of Chinese Poetic Prosody 漢語的韻律詩體學論稿 (2015). He has edited Written Chinese: the Present and the Past (2013), New Exploration on Chinese Prosodic Grammar (2015). He currently serves as associate editor for Language and Linguistics. He, Yun Amy Amy He Yun is a visiting research fellow in the Centre for Intercultural Politeness Research at the University of Huddersfield, the United Kingdom, where she has previously worked as a postdoctoral research associate. She obtained her PhD degree from Loughborough University for studying politeness in Mandarin Chinese, co-funded by an Overseas Research Studentship Award (Higher Education Funding Council for England), University Research Studentship, and Great Britain–China Educational Trust: Chinese Student Award. Her recent interests are in (im)politeness and face, but her research cuts across pragmatics, discourse analysis, sociolinguistics, intercultural communication, and translation studies. She has completed five research projects and has published widely in the aforementioned areas of research, including ‘Different generations, different face?’ (Journal of Politeness Research 8(1), 2012) and a chapter on identity in Pragmatic explorations in identity and communication (ed. Chen, X. Beijing: Higher Education Press, 2013). She is currently working on a book with the provisional title Generational Variation of Politeness in Contemporary Chinese. Healy, Alice F. Alice F. Healy is College Professor of Distinction and Director of the Center for Research on Training at the University of Colorado Boulder. She received her doctorate from the Rockefeller University in 1973, and she was on the faculty of Yale University from 1973 to 1981, when she joined the faculty of the University of Colorado. Dr. Healy has served as Editor of the journal Memory and Cognition, Chair of the Psychology Section of the American Association for the Advancement of Science, President of the Rocky Mountain Psychological Association, President of the Division of Experimental Psychology (Division 3) of the American Psychological Association (APA), and Chair of the Society of Experimental Psychologists. She has published over 250 journal articles and book chapters, is co-author of Cognitive Processes (Prentice Hall, 1986) and Train Your Mind for Peak Performance (APA, 2014), is editor of the volume Experimental Cognitive Psychology and its Applications (APA, 2005), and is senior editor of the two-volume Essays in Honor of William K. Estes (Erlbaum, 1992), Learning and Memory of Knowledge and Skills (Sage, 1995), Foreign Language Learning (Erlbaum, 1998), the Experimental Psychology volume of the Handbook of Psychology (Wiley, 2003 and 2013), and Training Cognition (Psychology Press, 2012). Herzberg, Larry Larry Herzberg has been a professor of Chinese at Calvin College in Grand Rapids, Michigan for the past 30 years, where he is also Director of Asian Studies. After studying Chinese for xix

List of contributors

five years at Vanderbilt University, he did his PhD work in Chinese language and literature at Indiana University. Together with his Chinese wife, Xue Qin, he has published three books, including The Chinese Survival Guide, Basic Patterns of Chinese Grammar, and Chinese Proverbs and Popular Sayings. He and his wife have also produced three films on China, including ‘China Today: Issues that Trouble Americans at the Start of the 21st Century’, ‘The China Threat: Perception versus Reality’, and ‘Chinese Dialects and Their Influence on the Pronunciation of Mandarin’. In 2011 he received the Presidential Award for Exemplary Teaching, the highest honor bestowed on faculty members by Calvin College. Hsieh, Shu-Kai Hsieh Shu-Kai is an associate professor in the Graduate Institute of Linguistics at National Taiwan University, Taiwan. He received his PhD in Computational Linguistics from the University of Tübingen in Germany. He has chaired and served on several program committees for linguistics at international conferences, and has given a number of invited talks on computational lexical semantics, corpus, and ontology at different universities. His primary research interests include computational lexical semantics and ontology. His publications include two books on computational treatment of Chinese characters and translation of the Generative Lexicon, and more than 30 refereed research papers published in journals, books, conferences, and workshops. He is a member of the editorial board and the linguistic section chair of IJCLCLP, an international journal of computational linguistics and Chinese language processing. He is a member of the council of the Linguistics Society of Taiwan and Global WordNet Association. Huang, Chu-Ren Huang Chu-Ren is a Chair Professor of Applied Chinese Language Studies at the Hong Kong Polytechnic University. He is Fellow and President of the Hong Kong Academy of the Humanities, a permanent member of the International Committee on Computational Linguistics, and a former President of the Asian Association of Lexicography. His main research areas are ontology, corpus and computational linguistics, lexical semantics, Chinese grammar, language resources, and digital humanities. The conference series in which he plays advisory and/or organizing roles include ALR, ASIALEX, CLSW, CogALex, COLING, IsCLL, LAW, OntoLex, PACLIC, ROCLING, and SIGHAN. His research output includes 12 licensable language resources, ten searchable online language databases, 20 books or edited volumes, over 130 journal articles or book chapters, and over 380 refereed conference papers. He is currently Chief Editor of the Journal Lingua Sinica, and the book series: Studies in Natural Language Processing and The Humanities in Asia. He received a Doctor honoris causa from Aix-Marseille University in 2013. Jiang, Yan Yan Jiang received his BA in English Language and Literature and his MA in modern English from Fudan University, Shanghai. He obtained his PhD in linguistics from London University. He is now a lecturer in linguistics and the languages of China at the Department of Linguistics, School of Oriental and African Studies, University of London. He has published studies on the Chinese adverb of scope ‘dou’, abduction and pragmatic inference, dexis and anaphora, Chinese counterfactuals, and quantification in Chinese. His book-length publications are listed below: Jiang Yan and Pan Haihua (1998, 2005) Introduction to Formal Semantics, China Social Science Press; Jiang Yan (2008) Translation of Dan Sperber and Deirdre Wilson: Relevance: Communication and Cognition, China Social Science xx

List of contributors

Press; and Jiang Yan (2011) (ed.) Approaching Formal Pragmatics, Shanghai Educational Publishing. Jiao, Liwei Jiao Liwei has a PhD in Chinese Linguistics from Nankai University, China. Since 1998 he has taught at Renmin University of China, the University of Durham, Princeton in Beijing, and the Duke Study in China Program. He is currently a Lecturer in Foreign Languages at the University of Pennsylvania. Among his many publications are 500 Common Chinese Idioms (co-authored, 2010), 500 Common Chinese Proverbs and Colloquial Expressions (co-authored, 2014), and A Cultural Dictionary of the Chinese Language: 500 Proverbs, Idioms and Maxims (forthcoming, 2017), all with Routledge, as well as Chinese Phonetics in the 20th Century (co-authored, 2004) with Shuhai Press, China. He is the Chinese translator of Professor Yuen Ren Chao’s oral history: Chinese Linguist, Phonologist, Composer and Author. His research fields include lexicology, relationships between language and culture, and Chinese phonetics. Kirkpatrick, Andy Andy Kirkpatrick is Professor in the Department of Languages and Linguistics, at Griffith University in Brisbane. He has worked in tertiary institutions across Asia including six years at the Hong Kong Institute of Education as Chair Professor. He has been a Visiting Scholar at Peking University and Bei Hang University in Beijing. His translation and commentary of Chen Kui’s Wenze won the International Society for the History of Rhetoric’s inaugural prize for best article. He is the editor-in-chief of the Journal of Multilingual Education and of a book series (with Springer) of the same name. His most recent book is Chinese Rhetoric and Writing: An Introduction for Teachers of Writing, co-authored with Xu Zhichang. Lau, Chun Fat Lau Chun Fat is Professor at the Department of Chinese Language and Literature at Xiamen University, China. He was born in the city but was brought up in a small Hakka village in the middle of the New Territories of Hong Kong. He studied Biology at The Chinese University of Hong Kong and received his Dr. rer. nat. degree at the Free University of Berlin. After spending more than a decade in the laboratory, he changed his career in the mid-1990s, when he found that practically no children were speaking Hakka around him. He received his master’s and PhD in Linguistics from the Polytechnic University of Hong Kong, and went to teach at Xiamen University in 2002. He has published more than 80 papers and authored eight books. His main interests are sociolinguistics, historical linguistics, Chinese dialectology, and documentation of endangered Chinese dialects around Hong Kong, as the whole area has shifted to Cantonese in the past three decades. He is also chairman of the Association of Conservation of Hong Kong Indigenous Languages. Li, Lan (Chapter 31) Li Lan is a Fellow of the Chartered Institute of Linguists, UK, with MPhil and PhD degrees in Applied Linguistics from the University of Exeter. She works as an Associate Professor at the Department of English, Hong Kong Polytechnic University. She is also a committee member of ASIALEX. She has been teaching and conducting research in semantics, lexicology, and bilingual lexicography. Her research interests and publications cover lexicology, lexicography, metaphor studies, computer-mediated professional communication, sociolinguistics, and corpus linguistics. xxi

List of contributors

Li, Lan (Chapter 34) After gaining his PhD in 1995 from the Chinese Academy of Social Sciences (CASS) in Beijing on the subject of ‘Variation in the Tones and Initials of Southwestern Mandarin Dialects’, Li Lan was offered a position at the CASS, where he continues to work as a Senior Research Fellow in the Institute of Linguistics and head of the Dialect Research Section. Apart from many articles on dialectology and typology of Sinitic languages, which have been published in the top-ranking linguistics journals in China such as Zhongguo Yuwen of the Chinese Language) and Fangyan《方言》(Dialects), Li Lan 《中国语文》(Studies   has also published two books on The Guiyang Dialect (1997) and on The Miao (Hmong) Language of Chengbu, Hunan (2004). He won the Wang Li prize at Peking University in 2005 for this latter monograph and has been the recipient of a large number of projects on linguistics funded by the Social Sciences Foundation of China. Li Lan has also undertaken research visits to the Hong Kong University of Science and Technology and to the Centre de Recherches Linguistiques sur l’Asie Orientale in Paris. Li, Chris Wen-chao Chris Wen-Chao Li is Professor of Linguistics at San Francisco State University, where he teaches linguistics, news writing, and translation-interpretation. He received his doctoral degree in General Linguistics and Comparative Philology from Oxford University, where his research focused on Mandarin sound change and Chinese phonology. He is the author of A Diachronically-Motivated Segmental Phonology of Mandarin Chinese (New York: Peter Lang, 1999) and Media Chinese (Taipei: Lucky Books, 2005). His translations and scholarship have appeared in Renditions, The Chinese Pen, Target, Language and Communication, and the Journal of the American Oriental Society. His current research interests include sound change, language contact, diglossia, standardization, phonological translation, Chinese Romanization systems, and the phonological description of Mandarin Chinese varieties. Li, Yafei Li Yafei is Professor of Linguistics at the University of Wisconsin-Madison, USA, and Siyuan Chair Professor (part-time) at Nanjing University, China. Born in Baoding, China, he received a BA in 1982 and an MA in 1985 from Shandong University and a PhD in Linguistics from MIT in 1990. Specializing in syntax and morpho-syntax, he has taught at Brandeis University, Cornell University, University of Wisconsin-Madison, and Nanjing University. His publications include articles (in Language, Linguistic Inquiry, Natural Language and Linguistic Theory, The Linguistic Review, Journal of East-Asian Linguistics,《中国语文》(Studies of the Chinese Language) and《当代语言学》(Contemporary Linguistics)) and two books, X 0: A Theory of the Morphology-Syntax Interface from MIT Press and the co-authored The Syntax of Chinese from Cambridge University Press. Currently, he is working on his third book, The Fact-first Principle, which evaluates the empirical foundation of the Chomskyan model of syntax and articulates novel analyses of facts from multiple languages that have not been accurately or adequately dealt with in the mainstream literature. Liu, Jin Liu Jin is Associate Professor of Chinese at Georgia Institute of Technology. She received her PhD in East Asian Literature and Culture from Cornell University and her MA and BA in Chinese Linguistics from Peking University. Her interdisciplinary research studies contemporary Chinese popular culture and media culture from the perspective of language, sound, voice, and music. She is the author of the book Signifying the Local: Media Productions xxii

List of contributors

Rendered in Local Languages in Mainland China in the New Millennium (Brill, 2013). She co-edited and contributed to the volume Chinese Under Globalization: Emerging Trends in Language Use in China (World Scientific, 2012). She has published articles in journals including Positions: Asia Critique, Modern Chinese Literature and Culture, Journal of Modern Literature in Chinese, Chinese Language and Discourse, and Harvard Asia Pacific Review. McEnery, Tony Tony McEnery is Distinguished Professor of English Language and Linguistics at Lancaster University. He has published a number of books on corpus linguistics including Corpus Linguistics (with Andrew Wilson, 1996), Corpus-Based Language Studies (with Richard Xiao and Yukio Tono, 2006), and Corpus Linguistics, Method Theory and Practice (with Andrew Hardie, 2011). A key feature of his research is that he has used the corpus approach to investigate a wide range of languages including, in collaboration with Richard Xiao, Mandarin Chinese. Tony is a Fellow of the Academy of Social Sciences and is currently Director of the ESRC Centre for Corpus Approaches to Social Science at Lancaster University. Previously he was Director of Research at the UK’s Arts and Humanities Research Council. Miao, Ruiqin Miao Ruiqin is currently Associate Professor of Linguistics in the School of Foreign Languages at Shanghai Jiao Tong University, China. She received her BA in English Language and Literature from Hunan University (1993), MA in English Linguistics and Applied Linguistics from the same university (1996), and PhD in Linguistics from the State University of New York at Stony Brook (2005). Her research focuses on second language acquisition, phonetics/ phonology, language contact, sociolinguistics, and language pedagogy. She has published in refereed journals and edited volumes including Journal of Asian Pacific Communication, ITL-International Journal of Applied Linguistics, and Developments in Applied Linguistics/ Series on Western Research in the Humanities and Social Sciences (Jianguo Ji and Nan Jiang, eds. Beijing: China Renmin University Press, 2007). She has written Foreign Language Listening Comprehension: Theory and Practice (co-authored with Liu Longgen, Beijing: Foreign Language Teaching and Research Press, 2011). She has also edited several English textbooks used in China and presented papers at various international conferences. Moore, Robert L. Robert L. Moore is a professor of anthropology at Rollins College and Director of Inter­ national Affairs for the college’s Hamilton Holt School. He has a BA in anthropology from Tulane University and a PhD from the University of California at Riverside. In 1993–1994 he was a visiting professor on the faculty of Qingdao University in China. He has undertaken research in Hong Kong, Qingdao, and Beijing and published a number of academic articles on linguistics, Chinese culture, and romantic love in non-Western societies. His current research focuses on contemporary youth cultures and family relationships in China. Orton, Jane Jane Orton, PhD, is Director of the Chinese Teacher Training Centre (CTTC) at the University of Melbourne, Australia, where she coordinated modern languages education for 20 years. Jane has an honours degree in Chinese language and literature from the University of Melbourne. She taught at Capital Normal University in Beijing from 1981 to 1983 and has been a frequent guest lecturer, conference speaker, and researcher in China over the ensuing 30 years. The CTTC undertakes research and professional development into Chinese language programs, teacher education, and innovative, research-grounded resource development. Jane’s xxiii

List of contributors

particular research interests are the learning demands of Chinese as a Second Language, especially in oral skill development. Jane is a board member of the association Chinese as a Second Language Research (CASLAR) and a member of the editorial board of the CASLAR Journal. Her recent publications include ‘Comparing teachers’ judgments of learners’ speech in Chinese as a foreign language’, Foreign Language Annals, 47(3), 2014: 507–26.〈澳大利亚 中小学汉语教学的复杂性〉(The Complexities of Teaching Chinese Language in Australian Primary and Secondary Schools)《云南师范大学学报(对外汉语教学与研究版)》(Journal of Yunnan Normal University (Teaching and Research of Chinese as a Foreign Language)) 4, 2014: 1–5; ‘Developing Chinese Oral Skills – A Research Base for Practice’, in Istvan Kesckes (ed.) Research in Chinese as a Second Language, Berlin: Mouton de Gruyter, 2013: 3–26. Packard, Jerome L. Jerome L. Packard is Professor of Chinese, Linguistics and Educational Psychology, and a faculty affiliate at the Beckman Institute for Advanced Science and Technology at the University of Illinois, USA. He received his BA in Chinese Language and Literature from the University of Massachusetts in 1976 and his PhD in Linguistics from Cornell University in 1984. He specializes in Chinese word structure, Chinese psycholinguistics, and Chinese language teaching and learning. He has taught Chinese at the University of Massachusetts, Cornell University, the University of Pennsylvania, and the University of Illinois, and was a Fulbright dissertation scholar in China. He has performed research on Chinese aphasia and dyslexia, the first language acquisition of Chinese, and the acquisition of Chinese as a second language. His current research interests include Chinese sentence processing, sentiment analysis of Chinese texts, and how children in China learn to read and write. He has received several grants and is the author of numerous publications relevant to Chinese language, including the books New Approaches to Chinese Word Formation, The Morphology of Chinese, Chinese Children’s Reading Acquisition (with W. Li and J. Gaffney), and Processing and Producing Head-final Structures (with H. Yamashita and Y. Hirose). Peyraube, Alain Alain Peyraube is currently emeritus Director of Research at the Centre National de la Recherche Scientifique (CNRS, Paris, France), Chair Professor of Chinese Linguistics at the Ecole des Hautes Etudes en Sciences Sociales (EHESS, Paris), and Director of the Institute of Advanced Studies ‘Collegium de Lyon’. He was Adjunct Professor at the University of Science and Technology of Hong Kong (2005–11) and Founding Member of the Scientific Council of the European Research Council (ERC) (2005–13), and he has been Honorary Professor at the Peking University since 2007, and Distinguished Member of the Chinese Academy of Social Sciences since 2009. He became a member of the European Academy (Academia Europaea) in 2006 and Chair of its Linguistics Section in 2013. As a specialist in Chinese historical syntax and linguistic typology of Sinitic languages, he has authored five books and more than a hundred articles on Chinese studies, mainly with respect to Chinese linguistics. His latest research has been done within a broadly functional and cognitive framework from a cross-linguistic perspective. He was conferred the title of Chevalier de la Légion d’honneur in 2008. Shi, Dingxu Shi Dingxu works in the Department of Chinese and Bilingual Studies of the Hong Kong Polytechnic University. He received his MA in applied linguistics from the University of xxiv

List of contributors

Pittsburgh, and MA and PhD in linguistics from the University of Southern California. His research interests are on syntax, typology, language change induced by contact, interface between syntax and semantics, and pedagogical grammar of Chinese. His publications cover a wide range of topics on Chinese in these fields. Simpson, Andrew Andrew Simpson is Professor of Linguistics and East Asian Languages and Cultures at the University of Southern California. His research focuses on the comparative grammar of East and Southeast Asian languages, and issues of language planning. He is the editor of Language and National Identity in Asia and Language and National Identity in Africa (Oxford University Press), and is the joint general editor of the Journal of East Asia Linguistics. Su, Xinchun Su Xinchun is the Head of the College of Humanities and Communications College of Jiageng College of Xiamen University, and also the Head of Language Education and Teaching Material Branch of National Language Resources Monitoring and Research Center (CNLR). His major teaching and research interests cover Chinese lexicology, ancient and modern Chinese vocabulary, cultural lexicology, metrology of lexicology, dictionary and corpus linguistics, terminology, and so on. He has published over 20 books and more than 100 articles on Chinese lexicology studies. Tang, Gladys Gladys Tang is Professor of Linguistics at the Department of Linguistics and Modern Languages, The Chinese University of Hong Kong. She is also Director of The Centre for Sign Linguistics and Deaf Studies. She has been researching sign language and how deaf children acquire language, signed and spoken. In recent years, she has also embarked on developing a model of education that benefits both deaf and hearing children. The sign bilingualism and co-enrollment approach promotes the early learning of signed language and spoken language to support children’s linguistic, cognitive and socio-emotional development. Tang, Sze-Wing Tang Sze-Wing received his BA and MPhil from The Chinese University of Hong Kong and completed his doctoral dissertation at the University of California, Irvine. He is Associate Professor and Vice-Chairman of the Department of Chinese Language and Literature, The Chinese University of Hong Kong, and concurrently serving as Director of T. T. Ng Chinese Language Research Centre of the Institute of Chinese Studies, The Chinese University of Hong Kong, and Chief Editor of Studies in Chinese Linguistics and Newsletter of Chinese Language. His research interests lie primarily in Chinese syntax, theoretical approaches to the study of Chinese dialects, and comparative grammar. Tao, Hongyin Tao Hongyin is Professor of Chinese Linguistics and Applied Linguistics at the University of California, Los Angeles. He has published a dozen books/edited volumes/special journal issues and over 90 articles in journals and edited volumes on Chinese discourse and grammar, applied linguistics, corpus linguistics, sociolinguistics, and English discourse grammar. Some of his books include Units in Mandarin Conversation: Prosody, Discourse, and Grammar (1996) and Current Trends in Sociolinguistics (1997/2004, co-edited). His most recent books are Working with Spoken Chinese (2011) and Chinese under Globalization: Emerging Trends xxv

List of contributors

in Language Use in China (2012, co-edited). In addition to being the president of the Chinese Language Teachers Association, USA (2014), he is on over a dozen editorial boards and is the Executive Editor of the journal Chinese Language and Discourse and its companion book series, Studies in Chinese Language and Discourse, both with John Benjamins. Tao, Liang Liang Tao is associate professor of linguistics and coordinator of the Chinese Language Program in the Department of Linguistics at Ohio University. She received her doctorate from the University of Colorado Boulder in 1993, and was a post-doctoral research associate in the Healy-Bourne Research Laboratory from 1993 to 1996, when she joined the faculty of Ohio University. Dr. Tao has broad research interests in the areas of theoretical linguistics, psycholinguistics, and applied linguistics, focusing on discourse-functional approaches in her research. She pioneered the study of repair in Chinese conversations, proposed the hierarchical structure of language processes against a popular iconic proposal of topic continuity, and proposed the interface of phonology and syntax (phono-syntactic conspiracy) in Mandarin grammaticalization. With Dr. Alice Healy she examined basic native and foreign language processes, using findings from experimental studies to propose, for instance, transfer of cognitive strategies in processing native and non-native languages, the developmental progression of the unitization effect, and the varied role of words in reading processes. She also conducted research on language development and cognitive approaches to Chinese pedagogy. Dr. Tao has published extensively in journal articles and book chapters, and has served as a co-editor of the volume Current Issues in Chinese Linguistics. Wang, William S.-Y. William S.-Y. Wang is Chair Professor at the Hong Kong Polytechnic University, based in its Department of Chinese and Bilingual Studies. He is also Professor Emeritus of the University of California at Berkeley (where he was Professor of Linguistics for 30 years), and Honorary Professor at Peking University and, Beijing Language and Culture University, and the Chinese University of Hong Kong. He is Editor of the Journal of Chinese Linguistics, which he founded in 1973. In 1992, he was elected President of the International Association of Chinese Linguistics at its formation. In the same year he was elected to the Academia Sinica in Taiwan. His central interest is in language within an evolutionary perspective. He has published some 200 papers and ten books in diverse areas of theoretical and applied linguistics. These have appeared in general magazines, such as American Scientist, Nature, Proceedings of the National Academy of Sciences (USA), Scientific American, etc., in specialized journals, such as Brain and Language, Diachronica, Language, Lingua, Language and Cognitive Processes, Neuropsychologia, Journal of Phonetics, etc., and in various encyclopedias. His writings have been translated into many languages. He has lectured widely in America, Asia, and Europe. Wu, Weiping Wu Weiping is Director of the Yale-China Chinese Language Center at The Chinese University of Hong Kong (CUHK) and a guest professor in four universities in China. He obtained his PhD in Linguistics from Georgetown University, and worked as a Research Associate and Director of the Chinese Language Testing Program at the Center for Applied Linguistics in Washington DC before moving to CUHK. His academic profile includes teaching, research, and publication in linguistics, teaching Chinese as a Second Language (CSL), and assessment. Among his contributions are articles in referred journals such as Forensic Linguistics, xxvi

List of contributors

Language Teaching and Research, TCSOL Studies, book and book chapters in CSL and sociolinguistics, the Computerized Oral Proficiency Assessment (COPA) in Putonghua and Cantonese for learners with background in English, Japanese, and Korean, and volumes of textbooks under the Pragmatic Framework. He is founding President of the CSL Teaching and Research Society (CSLTARS), a member of the editorial committee of several core journals in CSL, and series editor of Linguistics and Teaching Chinese as a Second Language. Xiao, Richard Richard Xiao is Reader in Corpus Linguistics and Chinese Linguistics in the Department of Linguistics and English Language at Lancaster University. His major research interests cover corpus linguistics, Chinese linguistics, contrastive linguistics, and translation studies. He has published numerous books in these research areas, including Aspect in Mandarin Chinese: A Corpus-based Study (2004), Corpus-based Language Studies: An Advanced Resource Book (2006), A Frequency Dictionary of Mandarin Chinese: Core Vocabulary for Learners (2009), Corpus-Based Contrastive Studies of English and Chinese (2010), Using Corpora in Contrastive and Translation Studies (2010), Corpus-Based Studies of Translational Chinese in English-Chinese Translation (2012), Translation and Contrastive Linguistic Studies at the Interface of English and Chinese (2014), and Corpus Pragmatics in Translation and Contrastive Studies (2015). He is a member of editorial boards of a range of international journals, such as Chinese Language and Discourse, Corpora, Corpus Linguistics, Foreign Language Learning: Theory and Practice, Glossa, Global Chinese, International Journal of Corpus Linguistics, and Languages in Contrast. Xing, Zhiqun Janet Janet Zhiqun Xing graduated from the University of Michigan in 1993, and is a Professor of Chinese and Linguistics at Western Washington University. Her research interests are in grammaticalization, semantic change, historical linguistics, discourse pragmatics, and Chinese language pedagogy. Her representative publications include two edited volumes – Newest Trends in the Study of Grammaticalization and Lexicalization (Mouton de Gruyter, 2012) and Studies of Chinese Linguistics – Functional Approaches (Hong Kong University Press, 2008); two monographs – A Contrastive Model of Teaching Chinese to Native English Speakers (Peking University Press, 2011) and Teaching and Learning Chinese as a Foreign Language: A Pedagogical Grammar (Hong Kong University Press, 2006); and several dozens of refereed journal articles related to grammaticalization, semantic change, and Chinese language pedagogy. Yin, Jing-hua John John Jing-hua Yin graduated from Beijing Capital Normal University in 1982, and he received his MEd and PhD in Foreign Language Education from the State University of New York at Buffalo in 1984 and 1995 respectively. He then began teaching Chinese at the Chinese Summer School at Middlebury College, the University of Oregon, and the College of William and Mary. Since 1997, he has been teaching Chinese language and literature at the University of Vermont. He is Professor of Chinese and Chair of the Department of Asian Languages and Literatures. His book Fundamentals of Chinese Characters was published by Yale University Press in 2006, and his book Practical Rhythmic Chinese was published by Foreign Language Teaching and Research Press in 2008. He has also published articles on his studies and the teaching of Chinese characters to college students in the USA. xxvii

List of contributors

Zhang, Grace Grace Zhang is an Associate Professor at Curtin University in Australia, where she teaches and supervises postgraduate students in Chinese linguistics. She has 30 years of tertiary teaching experience in Chinese language at various universities in Australia, China, New Zealand, and the United Kingdom. Her research interests include Chinese linguistics and pragmatics. She was awarded a PhD in linguistics by the University of Edinburgh in 1996, and has published extensively on linguistics. Her recently published books include: Using Chinese Synonyms (2010) by Cambridge University Press; Request Strategies: A Comparative Study in Mandarin Chinese and Korean (2008, co-author) by John Benjamins; Critical Perspectives on Language Education: Australia and Asia Pacific (2014, co-editor) by Springer; Elastic Languages: How and Why We Stretch Our Words (2015) by Cambridge University Press; Communicating through Vague Language: A Comparative Study of L1 and L2 Speakers (2015, co-author) by Palgrave Macmillan. Of specific relevance to this article is her co-authored journal article on the translation of Chinese xiehouyu: Liu, Chiung-wen and Grace Zhang (2006) ‘Translation of Chinese xiehouyu (sayings) and Relevance Theory’, Across Languages and Cultures 7(1): 49–76. Zhang, Xiaoheng Dr. Xiaoheng Zhang has a PhD degree in computer science from Oxford Brookes University, UK, an MSc degree in computer science from Hunan University, China, and a BA degree in English, Hunan University. His research interests include Modern Chinese writing systems, computational linguistics, computer-assisted language learning, and computer-assisted translation. He has published three books and over 60 research papers in these areas. Dr. Zhang is an assistant professor in the Department of Chinese and Bilingual Studies, Hong Kong Polytechnic University. The subjects he teaches at the undergraduate and post-graduate levels include Interface between Chinese and Information Technology, Computer-Assisted Chinese Language Teaching, Computer Tools for the Language Professionals, and Modern Chinese Characters and Information Technology. He is a member of the Editorial Committee of the Journal of Modernization of Chinese Language Education and a visiting professor of Hunan University. Zhuo, Jing-Schmidt Zhuo Jing-Schmidt received her BA and MA in German language and literature from Peking University, MA in Germanic Linguistics from the University of California, Los Angeles, and PhD in General Linguistics from the University of Cologne, Germany. She is associate professor of Chinese Linguistics at the University of Oregon. Her research is concerned with how language structure and its use are shaped by culture, society, and human psychology. The inquiry into the role played by human emotion in linguistic pragmatics constitutes an abiding focus of her work. Her corpus-based research methods have contributed to increased rigor in Chinese pragmatics and historical linguistics research. With scholarly experiences in three continents, she researches and publishes in English, Chinese, and German. She is the author of Dramatized Discourse: The Mandarin Chinese Ba-Construction (John Benjamins, 2005), and editor of Increased Empiricism: Recent Advances in Chinese Linguistics (John Benjamins, 2013), and has published numerous articles in leading linguistics journals including Cognitive Linguistics and Journal of Pragmatics.

xxviii

The Chinese language: The global, historical, and linguistic aspects Chan Sin-wai The Chinese University of Hong Kong, Hong Kong, China

Introduction This is an encyclopedia that will, we believe, meet the academic, linguistic, and pedagogical needs of millions of people in the world who are interested in the Chinese language in different capacities and for different reasons, such as Sinologists, Chinese linguists, and teachers and learners of Chinese as a second language. In the following, we will explain the contents of this encyclopedia of the Chinese language from a global, historical, and linguistic perspective.

The global aspects of the Chinese language Chinese, as we all know, is a language spoken by about one fifth of the world’s population (over a billion people) as China is a vast country with a population currently approaching 1.4 billion. According to statistics, the number of Chinese language speakers exceeds 1.4 billion, including 50 million people who have Chinese as their second language. As of 2014, there were 480 Confucius Institutes spread across the six continents, the number of Chinese language learners was over 40 million worldwide, and more than 3,000 higher education institutions in 109 countries offered Chinese language programs. All these figures show that the target users of this encyclopedia could be the countless language learners and linguists of Chinese worldwide, amateurs of Chinese literature and culture, and scholars in Chinese studies. It is not surprising then that the twenty-first century has sometimes been hailed as the century of the Chinese language. This encyclopedia contains a chapter on ‘Chinese Language in a Global Context’, co-authored by Liu Jin of Georgia Tech and Tao Hongyin of the University of California, Los Angeles. This chapter explores the relationship between globalization and the Chinese language, and discusses some major issues such as the contact of Chinese with other languages in historical times, the ongoing impact of English on Chinese and the perceived crisis involving Chinese, the rapidly ascending status of the Chinese language worldwide, and the impact of nationalization and modernization on regional varieties. Finally, a series of major changes in contemporary Chinese is analyzed, as is the emergence of notable new xxix

Chan Sin-wai

linguistic forms and their theoretical implications. Also related is the issue of the ‘Chinese Language and National Identity’, a chapter by Andrew Simpson of the University of Southern California. This chapter provides an overview of how Chinese has affected the development of national identity, and how the terms ‘nation’, ‘nationalism’, and ‘national identity’ apply in the context of China. It is shown that the relation of ‘Chinese’ to ‘national identity’ varies depending on the perspective of the Chinese nation that is adopted (ethnic vs. political), and that different emphases have been placed on the role of Chinese in the development of national identity at different times in modern history. The globalization of the Chinese language is closely related to the teaching of it as a second language and the training of Chinese teachers. Wu Weiping of The Chinese University of Hong Kong contributed a chapter on ‘Chinese Language Pedagogy’ in which he discusses the topic both diachronically and synchronically. While the former approach explores issues related to the major components of the language, namely phonology, vocabulary, grammar, and the writing system with characters; the latter focuses on the shift from language structure to language use in teaching activities. A systematic approach to implement the Pragmatic Framework in CSL teaching, covering assessment, curriculum design, teaching materials preparation, and teacher training, is also provided as part of this chapter. Jane Orton of the University of Melbourne focuses on ‘Chinese Language Education: Teacher Training’. She claims that certified schoolteachers of Chinese in the West generally graduate from generic courses in education studies and language teaching. They overwhelmingly comprise L1 users who need special training to manage the educational culture and learning styles of Western classrooms. The few dedicated Chinese teacher training programs available concentrate mainly on linguistic features. Analysis of Chinese as an object of learning, and the demands its very particular features make on speakers of European languages, are still to be incorporated in Chinese teacher training.

The historical aspects of the Chinese language The Chinese language is usually divided into Ancient (or Classical) Chinese and Modern Chinese, which are drastically different in several aspects. This encyclopedia includes a chapter on ‘Ancient Chinese’ written by Alain Peyraube of the Ecole des Hautes Etudes en Sciences Sociales (EHESS), Paris. Ancient Chinese, according to Peyraube, is simply defined by Wang Li as ‘the language of the documents of the past’. It covers a very long period, from the first Chinese inscriptions known to us, dated from the fourteenth century bce, until the nineteenth century ce. Three basic stages are usually distinguished for Ancient Chinese: (i) the Archaic period; (ii) the Medieval period; and (iii) the Modern period. It was during the Archaic period, between the fifth and second centuries bce, that what is known today as Classical Chinese par excellence was established, and it is this language, also called Late Archaic Chinese, that is discussed in this chapter. Covering the same period, but with a different approach, is the chapter on ‘Classical Chinese’ by Chris Li Wen-Chao of San Francisco State University. This chapter surveys the development of Classical Chinese from the language of early philosophical discourses to later imitations that have given rise to a written register distinct from the vernacular. Salient linguistic features of the classical language are explained, including the monosyllabicity of lexis, the fluidity of form class, and pronominal case grammar. The author also describes the shift in written language from Classical Chinese to the modern vernacular in the early twentieth century, viewed through the lens of diglossia. The language that we use today is known as ‘Standard Chinese’, about which Shi Dingxu of the Hong Kong Polytechnic University has written a chapter. According to Shi, Standard xxx

The Chinese language: The global, historical, and linguistic aspects

Spoken Chinese used to mean Guanhua ‘official language’ which was typically the dialect of the capital city. Guoyu ‘national language’ was created in the 1920s to take the role of the Standard Spoken Chinese and was succeeded by Putonghua ‘common language’ in the 1950s. The two versions were accompanied by Zhuyin Fuhao ‘phonetic symbols’ and Hanyu Pinyin ‘Chinese phonetic alphabets’ respectively. Another chapter on modern Chinese is written by Feng Shengli of The Chinese University of Hong Kong. This chapter, entitled ‘Modern Chinese: Written Chinese’, shows how written Chinese is different from spoken Chinese in modern times, how the formal register system of written Chinese has newly developed after the May Fourth Movement, and finally, what principles the formal register grammar must observe. The answer lies first on the argument that a formal style after the destruction of literary Chinese by the May Fourth Movement is established by creating a sense of expressive distance from everyday speech. That is, when linguistic expressions are used, the more distant the expressions are kept from everyday speech, the more formal sense they can create in faceto-face conversation. The demand for vernacular Chinese to function formally after the May Fourth Movement motivated and still motivates speakers and writers to use some classical forms to satisfy their urgent need for stylistic effect. This is the essential reason why the separation of writing (formal) from speaking (informal) has been resurrected in modern times. It is then argued that the formal features of written Chinese mainly consist of (i) Monosyllabic Words Used in Disyllabic Templates, (ii) Disyllabic Words Used in Disyllabic Copulates, and (iii) Formal Phrasal Patterns. Although these must be used following strict principles, such as Stylistic Coherence, Auditory Intelligibility (PAI), and Shaping by Prosody (PSP), they must also be mixed with some colloquial features in order to make the language natural. As a result, an amalgamation principle which modulates literary dictions with colloquial expressions is proposed here. During the analyses of the issues outlined above, the syllabic system of Prosodic Grammar and the Tripartite System of Stylistic-Register Grammar are introduced and their empirical applications are also discussed, such as the statistical measurement methods recently developed for degree of formality measurement, composition testing, readability scaling, style gradation, textbook compilation, L2 learning, literacy acquisition, etc. Finally, a discussion is held on what principles the formal register grammar must observe.

The linguistic aspects of the Chinese language The subjects under this section are numerous. They include characters, dialects, areas of linguistics, grammar, and Romanization systems. However, before diving into the chapters on these individual subjects, an overall review of the study of Chinese language, entitled ‘Chinese Linguistics’, has been written by William S. Y. Wang of The Chinese University of Hong Kong. His chapter gives a synoptic view on the languages and peoples of China from a multidisciplinary perspective. It describes the geographical setting of the country, as well as the prehistoric and historic background of the current multi-ethnic inhabitants. It draws upon knowledge gained in anthropology and genetics in addition to linguistics. While the main focus is on varieties of Mandarin, the dominant speech in China for over 1,000 years, the other major dialects are also discussed. This chapter serves as an introduction to other aspects of the Chinese language.

Chinese characters This encyclopedia has a chapter on ‘Chinese Characters’ by John Yin Jinghua of the University of Vermont, who introduces the history and development of Chinese characters from xxxi

Chan Sin-wai

ancient times to the present. It also explains the formation and structure of Chinese characters and presents the features that are unique to Chinese characters.

Dialects The Chinese language, which, as mentioned before, is spoken by the largest number of people on earth as their mother tongue has hundreds of different dialects. It is estimated that other than Mandarin, dialects in China are spoken by 1.2 billion people. The ten most popular dialects are: Mandarin, Min, Wu, Jin, Yue, Gan, Kejia (Hakka), Xiang, Pinghua, and Hui. This encyclopedia comprises a number of chapters that deal with dialects. ‘Mandarin and Other Sinitic Languages’, by Hilary Chappell of the Ecole des Hautes Etudes en Sciences Sociales in Paris and Li Lan of the Chinese Academy of Social Sciences in Beijing, provides a description of the ten main dialect groups which belong to the Sinitic or Chinese branch of the Sino-Tibetan language family. Beginning with the large Mandarin group which enjoys a wide geographical distribution in China, each branch of Sinitic is briefly characterized in turn in terms of its historical relation to Middle Chinese with respect to phonology (syllable structure), its tone system, and its personal pronoun paradigm, in addition to several interesting grammatical features with respect to a chosen representative dialect. Mandarin, as the most widely spoken dialect, is also discussed in detail in ‘Mandarin’ by Shi Dingxu of the Hong Kong Polytechnic University, who holds that it is the dialect spoken in Beijing in a narrow sense and the Northern dialect spoken in North China in a broader sense. The phonetic and phonologic system of Putonghua (‘common language’) is based on that of Mandarin. The description of Mandarin presented here covers its sound inventory, tonal system, words and morphology, phrases, clauses, sentences, aspect system, negation, comparison, and information package constructions. Two other dialects are also discussed in depth. First, ‘Cantonese’, a chapter written by Cheng Siu-Pong and Tang Sze-Wing of The Chinese University of Hong Kong; second, ‘The Hakka Dialect’ by Lau Chun Fat of Xiamen University.

Dictionaries Dictionaries are a necessary evil for language learners, translators, and language professionals, and lexicography has a particularly long history in China. According to Cheung Kam-Siu of The Chinese University of Hong Kong, who has contributed the chapter on ‘Lexicography’, this activity has evolved and developed in China for more than two millennia. This chapter gives an overview of various aspects of diachronic studies of Chinese lexicography, including dictionary compilation and theorization in China from the ancient time to the modern and contemporary periods, and the development of Chinese monolingual and bilingual dictionaries. With detailed illustrations, some of the classics in the history of Chinese lexicography are specifically introduced and analyzed. Lexicography has changed vastly from the ancient period to the present age. Huang Chu-Ren and Li Lan of the Hong Kong Polytechnic University, and Su Xinchun of Xiamen University, in ‘Lexicography in the Contemporary Period’, express the belief that the Chinese language enjoys a long history of dictionaries. The presentation of word meaning, pronunciation, grammar, and usage has developed in tandem with modern lexicography. The application of language mega-corpora has contributed to the compilation of Chinese dictionaries, either in paper or electronic form, for both native and non-native language learners and users. The dichotomy, however, is the definition of lexical units; there are no clear rules to segment words in data. Due to this reason, both character dictionaries and word dictionaries exist in the Chinese language. xxxii

The Chinese language: The global, historical, and linguistic aspects

Grammar Whether Chinese has the same parts of speech as the Indo-European languages has been the subject of much debate. This encyclopedia has a chapter on ‘Chinese: Parts of Speech’ by Candice Cheung Chi-Hang of the City University of Hong Kong that offers a comprehensive survey of the major parts of speech in Chinese, aiming to establish the parts of speech that are found both in Chinese and in the Indo-European languages, and those found only in Chinese. A significant implication of this study is that whereas some parts of speech may be viewed as universal, others are language specific.

Idioms Idiomatic expressions in Chinese include idioms, end-clippers, proverbs, slangs, and taboos. All these topics are covered adequately in this encyclopedia. Jiao Liwei of the University of Pennsylvania focuses on ‘Idioms’. According to Jiao, most Chinese idioms (chengyu) consist of four characters, and typically their meaning is more than the sum of the parts. Based on idiomatic salience, Chinese idioms can be divided into five categories and 15 types. After Written Vernacular Chinese replaced Classical Chinese in the 1910s, idioms have become an elite part of the Chinese language and received tremendous attention from people from all walks of life ever since. The chapter on ‘Chinese Proverbs and Popular Sayings’ by Lawrence Herzberg of Calvin College in Michigan takes a different approach and provides an overview of the rich treasury of Chinese proverbs and popular sayings. Over 400 sayings are presented in the chapter, including both the ancient sayings of Confucius and Lao Zi as well as the homespun truths of everyday life. All of these pithy proverbs offer insights into the Chinese language and culture. ‘Chinese Slang’ is discussed by Robert Moore of Rollins College. Slang is a category of informal speech, characterized by attributes related to playfulness or identification with a specific social group. In Mandarin, the term liyu is commonly translated as ‘slang’, though there are some differences between Chinese liyu and English slang. The main difference is that liyu was traditionally thought to be linked to local dialects. However, a new version of liyu has emerged on the Internet which is not dialect-based but is closely associated with youth culture. Xiehouyu, or end-clippers, are covered in ‘Chinese Xiehouyu’ by Grace Zhang of Curtin University in Australia. This chapter provides a general review of Chinese xiehouyu, including its definition, origin, types, content, structure, features, functions, and translation. Xiehouyu refers to Chinese metaphorical folk sayings which follow a two-part pattern of a metaphor. These two parts are separated by a pause in speech or a comma or dash in writing, providing time for or drawing attention to the intended meaning and enhancing the effect of humour or satire. Finally, Amy He Yun of the University of Huddersfield in United Kingdom, on the other hand, contributes a chapter on ‘Chinese Taboo’. This chapter presents an introduction to Chinese taboo, focusing particularly on its distinctive characteristics such as the hierarchical nature of death taboos in ancient China. Starting by sketching the milestone stages of historical development, the chapter provides detailed analyses of taboo words/expressions and their euphemistic or dysphemistic substitutes in five categories. Apart from the motivations and functions of taboo and prosecution of transgressors, she also discusses the diachronic and synchronic variation as well as punishment avoidance strategies with illustrative examples. xxxiii

Chan Sin-wai

Linguistic areas This encyclopedia covers the main areas of linguistics, such as semantics and pragmatics; a number of sub-areas, such as corpus linguistics, psycholinguistics, computational linguistics, and sign language; as well as other related fields, such as rhetoric and syntax. Semantics is discussed in ‘Chinese Linguistics: Semantics’ by Hsieh Shu-Kai from the National Taiwan University in Taiwan. According to Hsieh, meaning is the key to the understanding of communication, human cognition, and culture. As a subfield of linguistics, contemporary studies of meaning are characterized by various formal modelling, abstract representation, and empirical evidence that can be verified or disproved in a scientific manner. This chapter aims to give a brief and non-technical introduction to modern linguistic semantics, giving priority to the core issues surrounding lexical meanings in Chinese, ranging from morpho-semantics to lexical semantics. Janet Xing of Western Washington University in the United States writes on ‘Semantic Change in Chinese’. Through analysis of diachronic data, she demonstrates how new meanings are developed at the lexical, sentential, and discourse levels. The result of her study suggests that three mechanisms, namely meta[phoricalization], meto[nymization], and semantic reanalysis, are commonly present in the process of semantic change in Chinese, rather than the mere two (meta and meto) recognized in Indo-European languages. Furthermore, she argues that Chinese lexemes tend to develop polysemies accretively due primarily to their isolating-analytical characteristics. Meanings carried by several types of words are indeed noteworthy. Zhuo Jing-Schmidt writes on metaphors under the chapter heading of ‘Metaphor in Chinese: Cognition, Culture, and Society’, which focuses on the sociopolitical function of metaphor in a sociohistorical context by providing a close examination of two Chinese metaphors that have shaped China’s collective unconsciousness. Neologisms are discussed in ‘Chinese Neologisms: Word-formation Strategies in Chinese’ by Antonella Ceccagno of the University of Bologna. This chapter presents some central word-formation strategies in Chinese by focusing on the word-formation patterns that emerge as a result of the crucial role played by compounding in Chinese. The word-formation strategies that are discussed include: metacompounding, reanalysis of syllables as morphemes, neo­logisms formed by selecting a new meaning for the constituents of existing compounds, and the creation of new morphemes or new meanings for existing morphemes. Most of these word-forming patterns have been singled out only recently, and their implications are explored. In the history of the Chinese language, words from various languages, including both languages of foreign countries and non-Chinese languages within China, have been borrowed. In ‘Loanwords’, Miao Ruiqin of Shanghai Jiao Tong University claims that loanwords are adapted into Chinese through one of the following four approaches: (i) phonemic transliteration, (ii) meaning translation, (iii) graphic borrowing, or (iv) the hybrid method. Linguistic processes to nativize loanwords into Chinese mainly involve phonetic, phonological, and orthographic transformations, whereas semantic, morphological, and grammatical changes are much less common. The use of loanwords in modern Chinese demonstrates certain sociolinguistic features in terms of stylistic and regional variations. In addition, existence of multiple adaptations present challenges for language standardization and language policy. Elastic words, whose length can vary from monosyllabic to disyllabic, without a change in meaning, are discussed by Duanmu San and Dong Yan, both of Michigan University. Elastic words have been known since Karlgren (1918), but some theoretical and empirical questions still remain. A precise definition of elastic words is provided, along with an estimate xxxiv

The Chinese language: The global, historical, and linguistic aspects

of the percentage of elastic words in modern Chinese. Finally, theories of why these words are created in Chinese are compared. Pragmatics is about the use of language in different contexts. Jiang Yan of the Hong Kong Polytechnic University writes on Chinese pragmatics. This chapter starts with a survey of some general concepts and their exemplifications in Chinese. It then moves on to a short summary of relevance theory, followed by reports on findings from three case studies in Chinese pragmatics. The first concerns the Chinese adverb bai, which triggers a presupposition that resembles the Principle of Relevance. The second studies ‘list gem’ as a unique figure of speech, which is thought to have a special role to play in the current debate between contextualism and semantic minimalism in the sense that each side needs to supply extra accounts to accommodate such a case. The third provides a new analysis of Chinese counterfactuals, which takes most Chinese counterfactual conditionals as falsifying contingent counterfactuals. Related to the core areas of linguistics there are some subfields, including corpus linguistics, psycholinguistics, computational linguistics, and sign language. ‘Corpus-based Study of Chinese’ by Tony McEnery and Richard Xiao, both of Lancaster University, explores the state of the art in using corpora in Chinese linguistic investigations by focusing on areas where corpus linguistics has made its greatest impact, namely lexical study, grammatical study, and Chinese interlanguage research. ‘Chinese Psycholinguistics’, on the other hand, is a chapter written by Jerome L. Packard of the University of Illinois. This chapter offers an overview of the field of Chinese psycholinguistics. The chapter begins by describing the psychology of Chinese script processing, including the role of strokes and radicals, followed by psycholinguistic aspects of Chinese text reading. The psycholinguistics of lexical access in Chinese is then discussed, including the issue of phonetic and semantic activation during word formation and retrieval. A summary of sentence processing is then presented, followed by a synopsis of Chinese speech perception and production. An additional chapter on psycholinguistics, entitled ‘Psycholinguistics: Reading Chinese’ is provided by Liang Tao of Ohio University and Alice F. Healy of the University of Colorado. This chapter presents findings from experimental studies on reading Chinese, including examinations of processes in English and Chinese, two typologically different languages in their respective orthography, syntactic structures, and discourse grammar. This cross-linguistic comparison illustrates reading processes that reflect (i) universal cognitive behavior and (ii) unique characteristics in Chinese reading. The theoretical issues supported in the former include the unitization hypothesis and the Stroop effect. In the latter, Chinese orthographic features invoke special cognitive processes in word identification and in discourse processes. Computational linguistics is discussed in a chapter written by Zhang Xiaoheng of the Hong Kong Polytechnic University. This chapter focuses on Chinese, and especially on its differences with other languages, such as English. Chinese computational linguistics is often referred to as Chinese information processing in China. This chapter introduces the important areas of Chinese character information processing, word segmentation, natural language understanding and generation, corpus linguistics, and machine translation. ‘Sign language’ forms another chapter in this area, introduced by Gladys Tang of The Chinese University of Hong Kong. According to Tang and her fellow researchers, linguistic study of Hong Kong Sign Language (HKSL) was begun in the 1990s by James Woodward (1993). Since then, research on the linguistic properties of HKSL at different linguistic levels has been flourishing. In this chapter, she first provides an account of the historical development of HKSL, and then focuses on the descriptions of HKSL at the phonological, morphological, and syntactical levels. xxxv

Chan Sin-wai

Other areas related to linguistics covered in this encyclopedia are morphology, syntax, and rhetoric. ‘Chinese Morphology’ is discussed by Jerome L. Packard of the University of Illinois. This chapter describes the properties of words in Mandarin Chinese. It begins with Chinese typological characteristics, and defines word as a syntactically free form. The four Chinese morpheme types (free content morpheme, free function morpheme, bound root morpheme, and affix) are then described, including how the processes compounding, composition, and affixation compose the word types compound, bound root word, derived word, and grammatical word. The chapter concludes with a discussion of borrowing, reduplication, abbreviation, and neologism in Chinese. ‘Syntax’ is covered in a chapter by Li Yafei of the University of Wisconsin-Madison. Chinese syntax during the past 30 years has uncovered a wide range of new facts, explored various novel analyses, and come up with a non-trivial number of previously unknown questions. It has benefited from the general theory of linguistics as well as contributed to it. Both the similarities and differences between Chinese and other well-studied languages are becoming increasingly better understood. A question is how to improve the ways in which Chinese syntax helps shape the theory of the human languages. Another chapter on syntax, entitled ‘Chinese Syntax’, is written by Walter Bisang. This chapter starts out from a critical discussion of the extent to which there is syntax in Chinese and the question of the inter­ action between syntax and pragmatics. For that purpose, it briefly highlights some claims concerning subject–object asymmetry in coordinate-clause constructions and in Chinese relative clauses. Later on, it addresses the argument structure, nominal expressions, verbal structures, anaphora, information structure (focus and topic), the formation of questions, and issues of word-order typology (SVO vs. SOV). The presentation of these phenomena includes well-known syntactic constructions of Chinese like the ba-construction, the bei-construction, and numeral classifiers. ‘Chinese Rhetoric’, on the other hand, is a chapter written by Andy Kirkpatrick of Griffith University in Brisbane, Australia. His chapter provides an overview of the development of Chinese rhetoric from the earliest times to the present. It argues that China needs to rediscover its rich rhetorical tradition so that it may be adapted for public discourse in contemporary China. It includes some comparison and contrast with developments in Western rhetoric. Related to this area is a chapter on ‘Poetic Prosody’ by Feng Shengli of The Chinese University of Hong Kong, which deals with poetic prosody in Classical Chinese. It is shown that Archaic Chinese prosody had undergone a change from moraic foot structure (a foot consisted with two moras) to syllabic foot structure (a foot consisted with two syllables). More specifically, disyllabic Foot Formation developed during the Spring and Autumn period (722–481 bc). Only later did trisyllabic super-foot structure develop, requiring the maturity of disyllabic Prosodic Word Formation. It took even longer for Prosodic Word Compounding in the form of four-syllable structures to fully develop. Observing the development of the prosodic system from Old Chinese up to Middle Chinese, it would be expected that the development of pentasyllabic lines would have to wait for the maturity of super-foot formation and that heptasyllabic lines would have to wait for the maturity of PrWd Compounding. The former is seen in the trisyllabic compound formation of the Eastern Han dynasty (25–220), while the latter appears in tetrasyllabic verb formation during the Southern and Northern dynasties (420–589). According to the theory and facts presented here, the prosodically conditioned poetic effects might provide a basis for motivating a theory of prosodic stylistics in future studies.

xxxvi

The Chinese language: The global, historical, and linguistic aspects

Romanization systems Romanization consists of the spelling of Mandarin in Roman letters. It is related to syllable initial and final in traditional Chinese phonology. There are over 20 Romanization systems for the Chinese language. This encyclopedia contains chapters on the Romanization of Cantonese, the Hanyu Pinyin system, and the Wade–Giles Romanization system. In the chapter on ‘Cantonese Romanization’, Cheng Siu-Pong and Tang Sze-Wing of The Chinese University of Hong Kong hold that the diversity of Cantonese Romanization is a reflection of different backgrounds and purposes in rendering Cantonese into Roman letters. Some forms of Romanization are marked for their historical significance and influence on later proposals; whereas others are note­ worthy for their current popularity, as they are often found in books and other materials. Individual frameworks are discussed in detail, and the concluding tables summarize these discussions. It follows that such diversity compensates for the lack of a predominant form of Cantonese Romanization. Hanyu Pinyin, which is the most popular Romanization system in China and possibly in the world, is discussed in detail by Lilly Chen of Rice University. This chapter begins with an introduction to the official system for spelling Mandarin in Roman letters, and how it emerged from among many competing spelling systems, the earliest of which were devised by foreign missionaries. The structure of Pinyin is described in detail, including also the concepts of syllable initial and final in traditional Chinese phonology. The historical development from a morpho-syllabic to a phonemic system of sound representation is discussed, along with a functional evaluation of Pinyin. Another system that is popular in Taiwan is the ‘Wade–Giles Romanization System’, explained by Karen Steffen Chung of National Taiwan University. The earliest efforts to Latinize Chinese began in the sixteenth century with the work of Italian Jesuit priests Michele Ruggieri and Matteo Ricci. Their system was modified over the centuries by other Western missionaries, most notably Robert Morrison (1782–1834). Sir Thomas Francis Wade (1818–1895) further refined Morrison’s scheme, and Herbert Allen Giles (1845–1935) later popularized Wade’s system mainly through his 1,415-page A Chinese–English Dictionary. The Wade–Giles system was soon widely adopted, until its eclipse in the 1970s by Hanyu Pinyin, which is now the undisputed worldwide standard.

Conclusion It is hoped that these chapters on the global, historical, and linguistic aspects of the Chinese language, written by 61 scholars and experts from different parts of the world, will provide a comprehensive background and an authoritative reference to learners and researchers of the Chinese language.

xxxvii

Acknowledgements

I would like to express my gratitude to my former colleague Professor David Pollard, who served as Chair Professor of Translation at The Chinese University of Hong Kong for some years, for offering me a chance to work on this very meaningful project, and to Andrea Hartill, Senior Publisher of Routledge, for initiating this encyclopedia. Without her support, the present volume would never have been completed and published. My thanks also go to Camille Burns, Editorial Assistant, and Geraldine Martin, production editor, both of Routledge, for their assistance in the editing and production of this volume. I am grateful to all the Consultant Editors for their advice and support, to all contributors for their time and effort in writing articles for this volume, to Florence and James, assistant editors of this encyclopedia, for their help and hard work, and to Florence and Wei Yuxiang for compiling the Index. Chan Sin-wai The editor and publishers would like to thank the following copyright holders for permission to reproduce the following material: Tables and material excerpted from ‘Fundamentals of Chinese Characters’. © 2006 John Jing-hua Yin. Reproduced with kind permission of Yale University Press. ‘ACTFL/CAEP 2012 Standards’. Reproduced with kind permission of American Council on the Teaching of Foreign Languages. Figure ‘Austronesian Languages in Taiwan’, courtesy of Professor Paul Li. Figure ‘235’ from ‘The Archaeology of Ancient China’ by Kwang-chih Chang. © 1987 Yale University Press. Reproduced with kind permission. Figures, Tables, and Excerpts from ‘Love and War in Ancient China: Voices from the shijing’ pp. xxii–xxxiii and p. 38 Figure 8b, © 2013 City University of Hong Kong. Reproduced with kind permission. Figures, Tables, and Excerpts from ‘On Modern Written Chinese’ 37: 145– 61, © 2009 Journal of Chinese Linguistics. Reproduced with kind permission. Table (1) from ‘Shoubuchulai, La formazione delle parole in cinese’ p. 128, © 2009 Serendipita Editrice. Reproduced with kind permission. Figures, Tables, and Excerpts from ‘A Prosodic Explanation for Chinese Poetic Evolution’ 2: 223–57, © 2011 Tsing Hua Journal of Chinese Studies. Reproduced with kind permission. Material excerpted from ‘Chinese Proverbs and Popular Sayings: With Observations on Culture and Language’. © 2012 Qin Xue Herzberg and Larry Herzberg. Reproduced with kind permission of Stone Bridge Press, Inc. While the publishers have made every effort to contact copyright holders of material used in this volume, they would be grateful to hear from any they were unable to contact. xxxviii

1 Ancient Chinese Alain Peyraube Centre National de la Recherche Scientifique, Ecole des Hautes Etudes en Sciences Sociales, Paris, France

1. Introduction ‘Chinese is only one of a very few languages whose history is documented in an unbroken tradition extending back to the second millennium bc’ (Norman 1988: ix). Chinese is usually divided into Ancient Chinese (gǔdài hànyǔ 古代汉语) and Contemporary Chinese (xiàndài hànyǔ 现代汉语). Ancient Chinese is simply defined as ‘the language of the writings of the past’ (Wang 1979: 1). It covers a very long period, from the oracle bone inscriptions (OBI, the first Chinese inscriptions known to us, dated from the fourteenth century bce) until the nineteenth century. Three basic stages are generally distinguished for Ancient Chinese: (i) the Archaic period (shànggǔ 上古), until the second century bce; (ii) the Middle or Medieval period (zhōnggǔ 中古), from the first century bce to the middle of the thirteenth century ce; (iii) the Modern period ( jìndài 近代), from the middle of the thirteenth century to the middle of the nineteenth century. It was during the Archaic period that what is known today as Classical Chinese (wényán 文言) was standardized. This language, playing a role like Latin in Europe, remained the main written language used in literary texts until the beginning of the twentieth century. The period for Classical Chinese par excellence refers more precisely to the language used by the philosophers and scholars of the Warring States period (475–221 bce) and it was probably not very different from the educated speech of the period. The gap between the written and the spoken language began to develop in Han times or Pre-Medieval Chinese (206 bce – ce 220) and increased considerably with time. It is essentially Classical Chinese, also called Late Archaic Chinese (hòu shànggǔ hànyǔ 后上古汉语), that will be discussed in this chapter. I will nevertheless make several digressions regarding the period prior to the fifth century bce, i.e. what is known as pre-Classical Chinese or Early Archaic Chinese (qián shànggǔ hànyǔ 前上古汉语), and about the Medieval period, especially in the phonological section.

2. Phonology The history of Chinese phonology is usually divided into four periods based on the two important rhyme dictionaries, the Qiēyùn《切韵》of 601 and the Zhōngyuán yīnyùn《中原 音韵》(Rhymes according to the pronunciation of the Central Plains) of 1324: Old Chinese, Middle Chinese, Old Mandarin, and Modern Chinese. Old Chinese refers to the period before 1

Alain Peyraube

the Qiēyùn and is the earliest period; the Qiēyùn represents the beginning of Middle Chinese; the Zhōngyuán yīnyùn represents the beginning of Old Mandarin; and Modern Mandarin is the period leading up to present time, with no fixed dates. See Peyraube and Shen (forthcoming).

2.1.  Old Chinese phonology The source materials for Old Chinese phonological reconstruction, mainly based on rhyming materials (largely drawn from the Shī jīng《诗经》(The Book of odes), eleventh–sixth centuries bce), do not allow us to clearly identify the reconstructed phonological forms in space and time. The syllable in Old Chinese is analyzed as being composed of an initial and a final. Following Baxter (1992: 7), whose inventory of phonetic segments is given below, (i) the initial contains a pre-initial (treated as a prefix in Baxter and Sagart 1998) and an initial (Table1.1), and (ii) the final contains a medial, a main vowel (Table 1.2), a coda, and a postcoda (Table 1.3). Table 1.1  37 Old Chinese initials (Baxter 1992: 177) p t

ph th

b d

m n

hm hn

ts k kw Ɂ Ɂw

tsh kh kwh x

dz g gw ɦ

ŋ ŋw

hŋ hŋw

w l r j z

hw hl hr hj s

Table 1.2  The six main vowels (Baxter 1992: 180) i e

ɨ ɑ

u o

Table 1.3  Old Chinese codas (Baxter 1992: 181) zero j w

k t wk p

ŋ n m

Three medial elements have been reconstructed: *-r-, *-j- (though the reconstruction of the medial *-j- has now been replaced by a contrast of vowel length), and, marginally, *-l-. The two post-codas are *-ʔ and *-s, which are the respective sources of the rising tone (shǎngshēng 上声) and of the departing tone (qùshēng 去声) in Middle Chinese. 2

Ancient Chinese

This reconstructed Old Chinese phonemic system is far from being universally accepted. In fact, several specialists continue to consider that we are unable to do more than to approach a reconstruction of Old Chinese.

2.2.  Middle Chinese phonology The reconstruction of Middle Chinese phonology, in a more accurate sense, means the reconstruction of phonetic values for the existing categories, which are usually assumed to be phonemic. Table 1.4  Middle Chinese initials (Baxter 1992: 45) Labials Dentals/ Alveolars Lateral Retroflex stops Dental sibilants Retroflex sibilants Palatals Velars Laryngeals

p t

ph th

b d

m n

tr ts tsr tsy k ʔ

trh tsh tsrh tsyh kh

dr dz dzr dzy g

nr ny ŋ

l s sr sy

z zr zy

x

h

y

The basic medials are the glides -j- and -w-. The reconstruction of main vowels has changed several times. Efforts have been made to reduce the number of main vowels and to make the vowel system more natural in relation to linguistic universals. Table 1.5  Middle Chinese main vowels (Baxter 1992: 61) i e ɛ æ

u o

These main vowels may be followed by commonly reconstructed codas, which are either stop consonants or approximants (Table 1.6). Table 1.6  Middle Chinese codas (Peyraube and Shen, forthcoming) zero ŋ k

w

m p

j n t

Middle Chinese has also a system of four tones which were first identified and named in the fifth century. These tones are called píng 平 ‘level’, shǎng 上 ‘rising’, qù 去 ‘departing or falling’, and rù 入 ‘entering and short’. Each Chinese syllable belongs to one of these four tonal categories. 3

Alain Peyraube

3. Morphology It is often said that Ancient Chinese, being a monosyllabic and monomorphemic language, is a language with an impoverished morphology, a language in which the grammatical processes are almost totally syntactic, thus a language of that type called analytic or isolating. However, Ancient Chinese did indeed possess morphological processes, although none of them was fully productive. They actually may represent a vestige of older stages in which such a process was considerably more productive. These word-formation processes are of the same type as those found in Contemporary Chinese: compounding, reduplication (total reduplication such as wēiwēi 巍巍 ‘tall and grand’, or partial reduplication such as tángláng 螳螂 ‘praying mantis’), and even affixation, which was not at all an unproductive process in Ancient Chinese. Having derivational and even inflectional affixes, Ancient Chinese could be thought of as more synthetic than later stages of the language. Several prefixes, suffixes, and infixes have now been reconstructed for Ancient Chinese. These are derivational morphemes changing the meaning or part of the speech of the words to which they are attached (the ensuing discussion closely follows the treatment of Baxter and Sagart 1998). A good example of a prefix that has been reconstructed is the prefix *N- (causing a following voiceless obstruent to become voiced in Medieval/Middle Chinese) which, when attached to a verb (or even a noun in some cases), seems to produce an intransitive verb or adjective: thus, kens 见 ‘to see’ > *N-kens ‘to appear’. Another prefix *k-, added to verbs, could refer to more concrete actions taking place in a limited time frame: *ljuk 鞠 ‘to nourish’ > *k-ljuk ‘to breast feed’. See Sagart (1999). Another prefix *s- derives causative verbs from non-causative verbs or even from nouns (denominative suffix): *m-lun-s 顺 ‘obedient’ > *s-lun ‘to make (a horse) obedient’. Concerning the suffixes, there is in Old Chinese a reconstructed suffix *-s, which is the source of the departing tone of Middle Chinese, which, when added to adjectives or verbs produces derived nouns: for example, *tsrek 责 ‘to demand payment’ > *tsrek-s ‘a debt’. See Mei (1989). Some gradable adjectives also have corresponding noun forms in which the suffix *-s functions like English -th, occurring in pairs such as ‘deep/depth’, ‘wide/width’ (see Downer 1959; Mei 1980). Two infixes have also been reconstructed in Old Chinese by Sagart (1993). The exact function of the first one, *-j-, is difficult to establish, but forms with and without *-j- do appear to be semantically related. The second infix, *-r-, is said to produce forms that are plural or collective in the case of nouns, and iterative, durative, or indicating effort in the case of verbs. There are also some bi-morphemic monosyllabic words in Classical Chinese, which result from the fusion of two morphemes. One of the most cited examples of this fusion phenomenon is the negative fú 弗, which is analyzed as a combination of the negative bù 不 ‘not’ and the third-person pronoun zhī 之 ‘him, her, it’. Among other contractions of this kind in Classical Chinese are to be found: zhī 之 ‘third personal pronoun’ + yú 于 ‘to, at’ > zhū 诸; wú 毋 ‘not’ + zhī 之 ‘third personal pronoun’ > wù 勿; yú 于 ‘at, to, from’ + zhī 之 ‘third personal pronoun’ > yān 焉; hú 胡 ‘why’ + 不 bù ‘not’ > hé 盍. Examples: (1) 乞诸其邻而与之(《论语》) Qi zhu qi lin er yu zhi ask-for 3sg+to his neighbour and give 3sg ‘(He) asked his neighbour for it and gave (it to) him’ (Confucian Analects, fifth century bce) 4

Ancient Chinese

(2) 得之则生,弗得则死(《孟子》) de zhi ze sheng fu de ze si neg+3sg get then die get 3sg then live ‘(If he) gets it (he will) live, (if he) does not get it, (he will) die’ (Mencius, fourth century bce) (3) 百亩之田,勿夺其时(《孟子》) bai mu zhi tian wu duo qi shi hundred mu lig field neg+3sg deprive their time ‘The hundred mu of fields, do not deprive them of their time (of cultivation)’ (Mencius) (4) 学焉而后臣之(《孟子》) xue yan er hou chen zhi learn from+him and later make-subject 3sg ‘(He) learned from him and later made him (his subject)’ (Mencius) In fact, starting in the Pre-Medieval (Han) period, ca. first century bce, all the above examples of fusion disappeared and a fission process came into being, which indicates undoubtedly a move from syntheticity to analyticity: zhū 诸 > zhī 之 + yú 于; fú 弗 > bù 不 + zhī 之; wù 勿 > wú 毋 + zhī 之; yān 焉 > yú 于 + zhī 之; hé 盍 > hú 胡 + bù 不.

4. Syntax In Ancient Chinese, the subject (when expressed, as many sentences are without any subject) precedes the predicate, the verb precedes its object, and the modifiers precede the words they modify. Thus, Ancient Chinese is an SVO language (where S = subject, V = verb, and O = object).

4.1.  Sentence types The predicate might be complex, i.e. composed of more than one verb. Such cases involve serial verb constructions of the type V1  .  .  .  V2  .  .  .  (V3).  .  .  .  The semantic relationship between verbs in series is varied. In the following example, it involves an implication of purpose, where yǐ ‘in order to’ links the two verb phrases: (5)  楚人伐宋以救郑(《左传》) Chu ren fa Song yi jiu Zheng Chu people raid Song in-order-to save Zheng ‘The people of Chu raided Song in order to save Zheng’ (The Tradition of Zuo, fifth century bce) Complex predicates may also involve a ‘pivotal construction’, in which the noun phrase object of the first verb is the subject of the second verb: (6) 请君讨之(《左传》) qing jun tao zhi ask Prince attack 3pl ‘[I] ask [you] the Prince to attack them’ (The Tradition of Zuo) 5

Alain Peyraube

An interesting characteristic of the subject–predicate construction in Classical Chinese is that the whole construction can be nominalized by inserting the particle (ligature) zhī 之 between the two constituents of the construction:  (7) 人之爱人求利之也(《左传》) ren zhi ai ren qiu li zhi ye person lig love person pursue profit 3sg fin-part ‘One person loving another person [would] pursue profit [for] him’ (The Tradition of Zuo) A transitive verb can take one object (usually a noun phrase) or two, an indirect object (IO) and a direct object (DO) in that order (V+IO+DO). The double-object construction is restricted to those verbs with the semantic feature [+ give], [+ say] or [+ teach]: Apart from the pattern V+IO+DO, example (8), two other common orders, involving the prepositions yǐ (DO marker) or yú (‘to’, introducing the IO), are possible for the dative construction: (i) yǐ + DO+V+IO (9), and V+DO+yú+IO (10). See Peyraube (1987).  (8) 公赐之食(《左传》) gong ci zhi shi prince offer 3sg food ‘The prince offered him food’ (The Tradition of Zuo)  (9) 尧以天下与舜(《孟子》) Yao yi tianxia yu Shun Yao om Empire give Shun ‘Yao gave the Empire to Shun’ (Mencius) (10) 尧让天下于许由(《庄子》) Yao rang tianxia yu Xu You Yao leave Empire to Xu You ‘Yao left the Empire to Xu You’ (Zhuangzi, fourth century

bce)

Verbs may be followed by complements (bǔyǔ 补语), a term used for adjuncts when they follow the verb. When adjuncts precede the verb, they are denoted as adverbials (zhuàngyǔ 状语). Complements are divided into two types, depending upon whether or not they are introduced by a prepositional marker which forms a prepositional phrase (PP). Notable examples of PPs are locative complements, which are usually introduced by the preposition yú 于/於 ‘at, to’, as in: (11) 北学于中国(《孟子》) bei xue yu zhong guo north learn at central state ‘He went to the north to learn [it] in the Central States’ (Mencius) The PPs can also be adverbials, especially when they are formed with the preposition yǐ 以, as in: (12) 以羊易牛(《孟子》) yi yang yi niu with sheep change ox ‘Change the ox for a sheep’ (Mencius) 6

Ancient Chinese

Sentences are also customarily divided into simple and complex types and can be further classified as declarative, interrogative, imperative, and exclamatory sentences. There are three basic types of interrogative sentences: (i) yes/no questions; (ii) WH-questions; and (iii) rhetorical questions. The first type is formed with final question particles (hū 乎, yú 欤, yé 耶), which essentially transform statements into questions. The second type (WH) contains a question word (one of the interrogative pronouns listed below), generally without any final particle. The third type is more complex. Some rhetorical questions are formed with a final particle (hū, yú, or yé), but also with an adverb of negation placed before the verb, implying an affirmative answer. Others are formed with the modal particle qǐ 岂 ‘how could’ requiring an affirmative answer. The final particles hū and zāi 哉 are also generally used, though they may be omitted: (13) 予岂好辩哉?(《孟子》) yu qi hao bian zai 1sg how-could like debate fin-part ‘How could I [be one who] loves debating?’ (Mencius) Imperatives are not syntactically marked as such in Classical Chinese. The subject is usually deleted, but this in itself is not a sufficient diagnostic of the imperative sentence. However, when the imperative is intended to be understood as a request, and not as an order or a prohibition, the verbs yuàn 愿 ‘wish’ or qǐng 请 ‘beg’ are used: (14) 王请度之(《孟子》) wang qing du zhi prince beg measure it ‘[My] Prince, please measure it’ (Mencius) The final particle zāi 哉 or sometimes yǐ 矣 is the usual marker of the exclamatory sentence. It can be added either to a declarative or to an interrogative. The subject–predicate order is usually inverted in exclamatory sentences. Example: (15) 死矣盆成括!(《孟子》) si yi Pencheng Kuo! dead fin-part Pencheng Kuo ‘He is dead, Pencheng Kuo!’ (Mencius) A special kind of sentence in Classical Chinese is the copular sentence. If one defines the copula as an overt word which, when used in equational sentences, links the subject to a nominal predicate, and expresses (i) an equivalence meaning or (ii) a property or classificatory meaning, then one can identify the presence of copulas in Classical Chinese, even if they are not strictly necessary. The most common way of creating copular sentences is to add the final particle yě 也 at the end of a sentence (see Peyraube and Wiebusch 1995 for a detailed account of the history of copulas in Ancient Chinese, and especially for a discussion of the status of yě as a copula): (16) 彼丈夫也,我丈夫也(《孟子》) bi zhangfu ye wo zhangfu ye that reliable-man fin-part 1sg reliable-man fin-part ‘They were reliable men, I am a reliable man [too]’ (Mencius) 7

Alain Peyraube

In addition to yě, other copulas are attested in Classical Chinese. Thus, the negative copula fēi 非 ‘to not be’ is required in all negative nominal predicate sentences. In affirmative copular sentences, the verb wéi 为, which also means ‘to do, to regulate, to act, to consider as’, and so forth also acts regularly as a copula, and finally the copular verb shì 是 ‘to be’, still used today, and which comes from the demonstrative pronoun shì ‘this’ through a process of grammaticalization, is already attested no later than the Qin dynasty (second century bce). Concerning the passive sentences, there are semantic passives in Classical Chinese, i.e. passives expressing passivity without any overt morphological marker: a transitive verb can be made passive by placing its object (the patient) in subject position. However, there are also passive structures marked with some formant, such as a preposition, or an auxiliary verb. The most common way to form a passive construction in Classical Chinese is to use the preposition yú 于/于 ‘by’ to introduce the agent (V + yú + Agent), as in: (17) 治于人者食人治人者食于人(《孟子》) zhi yu ren zhe si ren zhi ren zhe si yu ren rule by other the-one-who feed other rule other the-one-who feed by other ‘Those who are ruled by others feed others, those who rule are fed by others’ (Mencius) Four other passive structures are commonly attested: wéi 为 + V, jiàn 见 + V (where wéi and jiàn are best considered to be auxiliary verbs, and where no agent is expressed), and wéi + Agent + V and wéi + Agent + suǒ 所 + V, which became the most common passive forms from the second century bce on. Examples: (18) 盆成括见杀(《孟子》) Pencheng Kuo jian sha Pencheng Kuo passive kill ‘Pencheng Kuo was killed’ (Mencius) (19) 后则为人所治(《史记》) hou ze wei ren suo zhi late then passive other passive control ‘[If I react] late, [I] will then be controlled by others’ (Records of the Historian, first century bce) Yet another passive form appears at the end of the Classical period: bèi 被 + V, where bèi is a verb meaning ‘to suffer; to be affected’. It will first become an auxiliary verb expressing passivity, like wéi and jiàn without any agent, and several centuries later, around the Early Medieval period, a preposition introducing a noun phrase agent, after being grammaticalized. For a more detailed analysis of the passive forms in Ancient Chinese, see Peyraube (1989). Example of bèi being a passive preposition: (20) 亮子被苏峻害(《世说新语》) Liangzi bei Su Jun hai Liangzi passive Su Jun kill ‘Liangzi was killed by Su Jun’ (Shi shuo xin yu, fifth century 8

ce)

Ancient Chinese

Complex sentences are composed of two or more clauses joined through coordination or subordination. The joining of clauses can be accomplished without any overt marking, as in example (21) or with a connective marker ér 而 ‘and, but’ or yì 亦 ‘also’, as in (22): (21) 老者安之,朋友信之,少者怀之(《论语》) lao zhe an zhi pengyou xin zhi shao zhe huai zhi trust 3pl young 3pl care 3pl old the-one-who soothe 3pl friend ‘As for the old, soothe them; as for friends, trust them; as for the young, care for them’ (Confucian Analects) (22) 人民少而禽兽众(《韩非子》) renmin shao er qin shou zhong people few but bird beast numerous ‘People are few but [wild] animals are numerous’ (Han Feizi, third century

bce)

Subordination may be indicated by subordinating conjunctions or particles, which can occur in the first clause, in the second, or in both. In the case of conditional sentences, the conjunctions rú 如, ruò 若 or gǒu 苟 ‘if ’ may appear in the first clause (if-clause), and the markers zé 则 or sī 斯 ‘then’ in the main clause (for an exhaustive analysis of the conditionals in Classical Chinese, see Harbsmeier 1981: 229–87). In concessive sentences, the most commonly used conjunction of concession is suī 虽 ‘although, even if ’. In the main clause one often finds ér 而, which then has its adversative meaning ‘yet’. In sentences expressing cause, the ‘because’ clause may be introduced by the preposition yǐ 以, and the main clause may contain the connective gù 故 ‘so, therefore’: (23) 以其不争故天下莫能与之争(《老子》) yi qi bu zheng gu tian xia mo neng yu zhi zheng because 3sg neg compete therefore Heaven under nobody can with 3sg compete ‘Because he does not compete, nobody can compete with him under Heaven’ (Laozi, fourth century bce)

4.2.  Word order Saying that Classical Chinese is an SVO language (see above) is not uncontroversial. The problem of word order and word order change in Chinese has been much debated since the 1970s. The discussions have essentially surrounded the hypothesis of Li and Thompson (1974), according to which Archaic Chinese was a SOV language, which might have changed to SVO, before shifting back again to SOV, the last stage still being underway: a) SOV > SVO; b) SVO > SOV. The hypothesis b) SVO > SOV has been criticized, and it has been shown that, synchronically, Chinese still is and remains a SVO language and that the OV order is a marked word order coding contrastiveness. The first change a) SOV > SVO has been challenged less because the SOV order for Archaic Chinese seems plausible, insofar as this order is found in Classical Chinese under special conditions. Consequently, it has been claimed that ProtoChinese must have been SOV and, therefore, Proto-Sino-Tibetan too, since almost all the Tibeto-Burman languages have a verb-final order (see LaPolla 1993; Mei 1997). Contrary to this viewpoint, the first attested documents of Pre-Archaic Chinese, i.e. the Oracle Bone inscriptions (OBI) (fourteenth–eleventh centuries bce), show a regular order of 9

Alain Peyraube

SVO, and this stage of the language is indeed more SVO than later stages of Chinese, such as Early or Late Archaic. To claim, then, that in a more ancient stage, before the oracle bone inscriptions, the basic order could have been SOV, is a pure surmise, and cannot be empirically proved (see Shen 1992; Peyraube 1997a, 1997b; and Djamouri 2001), who gives the following figures: 93.8% of the sentences in the OBI are (S)VO vs. 6.2% (S)OV). There are also several clear indications for SVO order being the more basic and dominant word order than SOV in Early and Late Archaic, i.e. in Classical Chinese. When the object is a full lexical noun phrase (NP), the basic order is almost exceptionlessly VO, as in: (24) 君必矢国(《左传》) Jun bi shi guo Prince certainly lose state ‘The prince (will) certainly lose the State’ (The Tradition of Zuo) There are some cases in which the NP-object is in a preverbal position, but these cases are marginal and the OV order is then marked [+ contrastive]. The same situation applies when the noun object is followed by a preverbal marker, usually shì 是 or zhī 之, as in (6): (25) 今吴是惧(《左传》) jin Wu shi ju now Wu om afraid ‘Now [they] are afraid of [the state of] Wu’ (The Tradition of Zuo) However, there are also cases of OV order in Archaic Chinese, not found in Contemporary Chinese, in which the object is a pronoun, being either (i) an interrogative pronoun (26), (ii) the demonstrative pronoun shì 是 ‘this’ or (iii) a pronoun in a negative sentence (27): (26) 吾谁欺?欺天乎?(《论语》) wu shei qi? qi tian hu? I who deceive deceive Heaven int-part ‘Whom should I deceive? Should I deceive Heaven?’ (Confucian Analects) (27) 不吾知也(《论语》) Bu wu zhi ye neg 1sg understand fin-part ‘(You) don’t understand me’ (Confucian Analects) What can be said about these various OV orders involving pronouns? There are first some statistics that show that the OV order has always been non-dominant, since it is only wellattested in the case of pronoun objects. In a corpus of 2,767 VO or OV sentences drawn from the bronze inscriptions, 88.56% of objects (O) are nouns, only 3.3% are pronouns (see Guan 1981: 88). The ratio of pronoun objects is certainly higher in other later documents of Classical Chinese, but it never exceeded 15% of the entire body of VO and OV constructions. It is also known that in many languages, the position of pronouns is different from that of noun phrases, and that ‘unstressed constituents, such as clitic pronouns, are often, crosslinguistically, subject to special positioning rules only loosely, if at all, relating to their grammatical relation, so sentences with pronouns can be discounted in favor of those with full noun phrases’ (Comrie 1989: 89). 10

Ancient Chinese

4.3.  Full words and empty words Classical Chinese words are traditionally divided into two categories: shízì 实字 ‘full words’ and xūzì 虚字 ‘empty words’. The former are content words (which carry semantic content) and form an open class; included in this category are nouns, verbs, and adjectives. The latter are function words or grammatical words, used to express grammatical relationships. They include pronouns, adverbs, prepositions, conjunctions, and particles (see Peyraube 2004). Words can be used in functions customarily reserved for other words. This does not imply, however, as some scholars have assumed, that there are no parts of speech in Classical Chinese, and that words can be used indifferently in any grammatical category.

4.3.1.  Full words Chinese nouns typically function as subjects or objects. However, under certain conditions, they may function like verbs, such as the predicates in (28) and (29), or like adverbs, such as the adverbials in (30): (28) 君君臣臣父父子子(《论语》) jun jun chen chen fu fu zi zi ruler ruler minister minister father father son son ‘The ruler acts as a ruler, the minister as a minister, the father acts as a father, and the son as a son’ (Confucian Analects) (29) 物物而不物于物(《庄子》) wu wu er bu wu yu wu thing thing and neg thing passive thing ‘Treat things as things and do not be treated as things by things’ (Zhuangzi, fourth century bce) (30)

豕人立(《左传》) shi  ren   li pig man stand-up ‘The pig, like a man, stood up’ (The Tradition of Zuo)

Chinese verbs are fundamentally predicative in nature. Unlike nouns, which are negated by the adverb of negation fēi 非 ‘to not be’, verbs are negated by the simple adverb bù 不 ‘not’. One particular use of intransitive verbs in Classical Chinese is in a causative function: thus, huó 活 ‘to live’: ‘to make (people) live’; xíng 行 ‘to go’: ‘to put into motion’; yǐn 饮 ‘to give to drink’: ‘to give something to drink’; dòu 斗(鬭)  ‘to fight’: ‘to make (people) fight’. Example: (31) 不如先斗秦赵(《史记》) buru xian dou Qin Zhao better first (make-)fight Qin Zhao ‘It is better to make the Qin and Zhao fight first of all’ (Records of the Historian) Another verbal subclass is composed of modal auxiliary verbs (qíngtài zhùdòngcí 情态助动词), also sometimes called ‘can-wish verbs’ (néngyuàn dòngcí 能愿动词). They are verbs that take other verbs as their objects and express the modality of the following 11

Alain Peyraube

verb phrase. This modality (ability, possibility, probability, certainty, obligation, volition, etc.) can be characterized as epistemic, deontic, or dynamic. Auxiliary verbs form a closed list and can be classified into the following four semantic groups: (i) verbs expressing mainly possibility and permission, including kě 可, néng 能, zú 足, dé 得, huò 获, kěyǐ 可以, and zúyǐ 足以 (32); (ii) the four verbs of volition, gǎn 敢, kěn 肯, yù 欲, and yuàn 愿 (33); (iii) the two auxiliaries of necessity, obligation and requirement (certainty and obligation), yí 宜 and dāng 当; and (iv) the passive auxiliaries jiàn 见, wéi 为, and bèi 被. For a detailed analysis of mood and modality in Chinese, see Chappell and Peyraube (forthcoming). (32) 天子不能以天下与人(《孟子》) Tian zi bu neng yi tianxia yu ren om Empire give other Heaven son neg can ‘The Emperor cannot give the Empire (to) others’ (Mencius) (33) 子欲居九夷(《论语》) Zi yu ju Jiu Yi Master intend-to live Jiu Yi ‘The Master intends to live in Jiu Yi’ (Confucian Analects) Adjectives can also be considered as a subcategory of verbs. Indeed, they are intransitive verbs of quality, being negated by the adverb bù 不: (34) 名不正则言不顺(《论语》) ming bu zheng ze yan bu shun name neg correct then word not justified ‘If names are not correct, then words cannot be justified’ (Confucian Analects) Like intransitive verbs, adjectives can also have a causative use: (35) 王请大之(《孟子》) Wang qing da zhi King beg great it ‘Your Majesty, [I] beg [you] to make it great’ (Mencius) In addition, adjectives are also typically found as noun phrase modifiers, as in bái mǎ 白马 ‘white horse’, or as verb phrase modifiers, for instance jí zǒu 急走 (lit. rapid-run) ‘run rapidly’. Finally, one can consider that numerals constitute a subclass of the category of adjectives. They do indeed behave syntactically like adjectives; thus, they can form predicates and are negated by the adverb bù. Most commonly, however, they function as modifiers of nouns: (36) 年已七十矣(《孟子》) nian yi qi shi yi age already seven ten fin-part ‘[He] is already 70 years old’ (Mencius) (37) 吾何爱一牛?(《孟子》) wu he ai yi niu I why begrudge one ox ‘Why (should) I begrudge one ox?’ (Mencius) 12

Ancient Chinese

4.3.2.  Empty words Pronouns are usually analysed as empty words. Several types of pronouns can be identified: personal, demonstrative, interrogative, and indefinite. Personal pronouns characteristically occur in different forms. The most common ones, with no distinction being made between singular and plural are: (i) for the first person, wú 吾, wǒ 我, and yú 余(予); (ii) for the second person, rú 汝, ruò 若, and nǎi 乃; (iii) for the third person, zhī 之, jué 厥 and qí 其. Several scholars have tried to characterize their different usages according to case (nominative, accusative, or genitive), but dialectal variation also is a factor: in most instances, the different usages of the pronouns depend on the different texts in which they occur. The most common demonstrative pronouns are: (i) shì 是, cǐ 此, sī 斯, zhī 之, and zī 兹 ‘this, these, here’; and (ii) bǐ 彼, fū 夫, and qí 其 ‘that, those, there’. Here too, it is difficult to explain formal differences without considering dialectal variation. All of these demonstratives can be used as adjectivals (modifying the following nouns or noun phrases), or as subjects or objects. The interrogative pronouns are divided into two categories: (i) those that replace subjects or objects (which are usually nouns), shéi 谁 ‘who’, shú 孰 ‘which, who’, hé 何 ‘what’; and (ii) those that replace predicative verbs or adverbs, hú 胡 ‘why, how’, xī 奚 ‘why’, hé 曷 ‘how, why’, ān 安 ‘where, how’, yān 焉 ‘how, where’. See Peyraube and Wu (2005). The class of indefinite pronouns includes huò 或 ‘some, someone, something’, mò 莫 ‘none, no one, nothing’, and mǒu 某 ‘some, a certain one’. Adverbs are also considered as empty words. Usually positioned in preverbal position, they typically modify the predicate of the sentence. One can distinguish several types: (i) adverbs of degree ( jí 极 ‘extremely’, zuì 最 ‘most’, shǎo 少 ‘little’, shén 甚 ‘very’, etc.); (ii) adverbs of quantification and restriction ( jiē 皆 ‘all’, gě 各 ‘each’, dú 独 ‘only’, etc.); (iii) adverbs of time or aspect ( yǐ 已 ‘already’, cháng 尝 ‘once’, jiāng 将 ‘be going to’, nǎi 乃 ‘then’, fāng 方 ‘just’, etc.); (iv) adverbs of negation (bù, fú, fēi, wú, wèi 未, etc.). See above for the processes of fusion for some of these adverbs of negation. Chinese prepositions are all verbal in origin (i.e. they arise from verbs through a process of grammaticalization). There are two commonly occurring prepositions in Classical Chinese: (i) yú 于/於 ‘at, to, in, from, towards, than, by’, etc.; and (ii) yǐ 以 ‘with, by means of, in order to, because’, etc. The first of these, yú, can be used to form locative, ablative, dative, comparative, or passive constructions; the second, yǐ, primarily instrumental, also expresses purpose and several other grammatical relationships. An important characteristic of yǐ is that it can also introduce the direct object of a double-object construction (32). Additional prepositions are yòng 用 ‘with’, wéi 为 ‘for, on behalf of, for the sake of, because’, yǔ 与 ‘with’, zì 自 ‘from’, among still others. Generally, simple juxtaposition is sufficient to coordinate nouns or noun phrases, as in fù mǔ 父母 (lit. father mother) ‘father and mother’; or verbs and verb phrases in serial verb constructions. However, some coordinative conjunctions also occur, such as jí 及 and yǔ 与 ‘and’ for coordinating noun phrases, or ér 而 ‘and’ and qiě 且 ‘and, moreover’ for coordinating verb phrases or clauses. Liu and Peyraube (1994) have argued that the conjunctions jí and yǔ did not directly develop from verbs, but from prepositions, which were themselves derived from verbs. In other words, two processes of grammaticalization have occurred sequentially: verb > preposition > conjunction (Chinese conjunctions are thus more grammaticalized than prepositions). Finally, the category of particles (zhùcí 助词) is also obviously included in empty words. This category is usually divided into structural particles (zhī 之, suǒ 所, zhě 者) and modal particles that can occupy the initial, the medial, and, in most cases, the final position of 13

Alain Peyraube

a sentence. These final particles can be divided according to the sentence types in which they occur – declarative, interrogative, exclamatory, and so on. In declarative sentences, one often finds yǐ 矣 (a particle of the perfect aspect), yě 也 (transforming a statement into an assertion, a judgment), ěr 耳, and yān 焉. Hū 乎, yú 欤, yě 也, and sometimes zhě 者, are more typically used in interrogative sentences. Zāi 哉 occurs in exclamatory sentences.

5. Lexicon The overall lexicon of Ancient Chinese is quite different from that of Contemporary Chinese. The former is composed of: (i) words that are still attested in the contemporary language, like shān 山 ‘mountain’ or shuǐ 水 ‘water’; (ii) words that only exist in the ancient language and have disappeared from the modern language, such as yuē 曰 ‘say’; (iii) words that are still used today, but with different meanings, like zǒu 走 ‘run’ (Ancient Chinese) > ‘walk’ (Contemporary Chinese). Of the three types, the second are rare and the last are numerous (see He and Jiang 1980: 3).

5.1.  Historical development of the lexicon The Ancient Chinese lexicon has changed considerably since the Pre-Archaic period. From the vocabulary of everyday life (lexemes for food, clothing, housing), we find only 15 words in the oracle bone inscriptions (fourteenth–eleventh centuries bce) that are still used today, 71 in the bronze inscriptions (tenth–sixth centuries bce), and 297 in Shuō wén jiě zì 说文解字 (Explain the figures [single characters] and interpret the characters [compound characters], second century ce). On the varying richness of Chinese vocabulary in different periods, see He and Jiang (1980: 9). According to these two authors (1980: 136–7), Classical Chinese has an identifiable basic vocabulary of about 2,000 full words, of which 1,100 occur quite commonly. From four major works of the Late Archaic period, including Lún yǔ《论语》(Confucian Analects) and Mèngzǐ《孟子》(Mencius), they have isolated 4,466 distinct words, estimating that about half are proper nouns such as personal names or place names. There is no implication that the vocabulary of Classical Chinese is impoverished compared with that of Con­temporary Chinese – simply different. For example, there is only a single verb meaning ‘to wash’ in Contemporary Chinese (xǐ 洗), whereas there are five in Classical Chinese: mù 沐 ‘to wash (the hair)’, yù 浴 ‘to wash (the body)’; hui 沬 ‘to wash (the face)’; zǎo 澡 ‘to wash (the hands)’; xǐ 洗 ‘to wash (the feet)’. During the long history of Ancient Chinese, several different processes have led to changes in the lexicon. The major processes of internal development include: (i) compounding, a highly productive process beginning in the first century ce (Han period); (ii) semantic extension (e.g. zú 卒 ‘foot soldier’ > zú 卒 ‘all sorts of soldiers’); and (iii) semantic reduction (e.g. zǐ 子 ‘child’ (boy or girl) > zǐ 子 ‘son’). In addition, the Ancient Chinese lexicon was enlarged by borrowing words from other languages.

5.2.  Inherited elements and loanwords There has been a strong tendency in the past to view the Ancient Chinese lexicon as a monolithic linguistic entity, resistant to influences from all surrounding foreign languages. This is certainly a fallacy. Without going as far as Norman (1988: 17), who states: ‘the fact

14

Ancient Chinese

that only a relatively few Chinese words have been shown to be Sino-Tibetan may indicate that a considerable proportion of the Chinese lexicon is of foreign origin’, we can undoubtledly rightly assert that the Ancient Chinese lexicon contains numerous loanwords. Nevertheless, the identification of such words and their sources is often uncertain. Below we mention a few uncontroversial examples of loanwords. There are two common words for ‘dog’ in Ancient Chinese: quǎn 犬, which is probably the native Chinese word, and gǒu 狗, which appears at the end of the Warring States period. Gǒu 狗 is a loanword from a language ancestral to the Modern Miao-Yao languages (Hmong-Mien). The word 虎 hǔ ‘tiger’ might have been borrowed from an Austronesian language in prehistoric times (see Norman 1988: 17–20). Other words of non-Chinese origin are xiàng 象 ‘elephant’ (borrowed from a Tai language?), pútáo 葡萄 ‘grape’ (from Old-Iranian?), mòlì 茉莉 ‘jasmine’ (from Sanskrit), luòtuó 骆驼 ‘camel’ (possibly from an Altaic language).

6. Conclusion The above chapter on syntax and lexicon is concerned chiefly with the Classical Language, as represented by the Warring States period, i.e. the Late Archaic period (fifth–second centuries bce). From the time of the Early Medieval period (second–sixth centuries ce), one can consider that the vernacular language is actually distinct from the literary, deserving a separate description of its own. For example, certain important grammatical structures which did not exist in Classical Chinese are found to have developed in this later vernacular form, prior to the sixth century ce. For a detailed review of the developments in the language between these two stages, on the birth of the disposal construction (NP1-Agent + OM [object marker: qǔ 取, chí 持, zhuō 捉, jiāng 将, bǎ 把] + NP2-patient + VP), on the locative PPs, on the resultative construction, or on the birth and development of the classifier system, etc., see Peyraube (1996).

Acknowledgements This research is part of a project entitled ‘Typologie des processus synchronique et diachronique en Min-Sud (langue sinitique)’ funded by the French National Research Agency’s (ANR) ‘blue sky’ programme blanc: TYSOMIN n° 11-ISH2-001-01, 2012–14, projet bilatéral France (ANR) et Taiwan (National Research Council), 2013–15. I would also like to thank Hilary Chappell for her comments and questions.

List of abbreviations 1sg = 1st singular personal pronoun 3sg = 3rd singular personal pronoun 3pl = 3rd plural personal pronoun fin-part = final particle int-part = interrogative particle lig = ligature neg = negation om = object marker passive = passive marker

15

Alain Peyraube

References Baxter, William H. (1992) A Handbook of Old Chinese Phonology, Berlin: Mouton de Gruyter. Baxter, William H. and Laurent Sagart (1998) ‘Word Formation in Old Chinese’, in Jerome Packard (ed) New Approaches to Chinese Word Formation: Morphology, Phonology and the Lexicon in Modern and Ancient Chinese, Berlin: Mouton de Gruyter, 35–76. Chappell, Hilary and Alain Peyraube (forthcoming) ‘Mood and Modality in Sinitic Languages’, in J. van der Auwera and Jan Nuyts (eds) The Oxford Handbook of Mood and Modality, Oxford: Oxford University Press. Comrie, Bernard (1989) Language Universals and Linguistic Typology (2nd edition), Oxford: Basic Blackwell. Djamouri, Redouane (2001) ‘Markers of Predication in Shàng Bone Inscriptions’, in Hilary Chappell (ed) Sinitic Grammar – Synchronic and Diachronic Perspectives, Oxford: Oxford University Press, 143–71. Downer, Gordon (1959) ‘Derivation by Tone-change in Classical Chinese’, Bulletin of the School of Oriental and African Studies 22(2): 258–90. Guan, Xiechu (1981) Xi-Zhou jinwen yufa yanjiu《西周金文语法研究》(Research on the grammar of the bronzes of the Western Zhou), Beijing: Shangwu yinshuguan. Harbsmeier, Christoph (1981) Aspect of Classical Chinese Syntax, London: Curzon Press. He, Jiuying and Jiang Shaoyu (1980) Gu hanyu cihui jianghua《古汉语词汇讲话》(Talks on the lexicon of Ancient Chinese), Beijing: Beijing chubanshe. LaPolla, Randy (1993) ‘On the Change to Verb-Medial Word Order in Proto-Chinese: Evidence from Tibeto-Burman’, Paper presented at the 26th International Conference on Sino-Tibetan Languages and Linguistics, Osaka. Li, Charles N. and Sandra A. Thompson (1974) ‘An Explanation of Word Order Change: SVO > SOV’, Foundations of Language 12(2): 201–14. Liu, Jian and Alain Peyraube (1994) ‘History of Some Coordinative Conjunctions in Chinese’, Journal of Chinese Linguistics 22(2): 179–201. Mei, Tsu-lin (1980) ‘Sisheng bieyi zhong de shijian cengci’〈四声别义中的时间层次〉(‘Chronological Strata in Derivation by Tone-change’), Zhongguo yuwen《中国语文》(Studies of the Chinese Language) 6: 427–43. Mei, Tsu-lin (1989) ‘The Causative and Denominative Functions of the *s- Prefix in Old Chinese’, in Proceedings of the Second International Conference on Sinology, Taipei: Academia Sinica, 31–51. Mei, Tsu-lin (1997) ‘Typological Changes in the History of the Chinese language’, Chinese Languages and Linguistics 4: 85–104. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Peyraube, Alain (1987) ‘The Double-object Construction in Lunyu and Mengzi’, in Wang Li Memorial Volumes, Hong Kong: Joint Publishing, 331–58. Peyraube, Alain (1989) ‘History of the Passive Construction in Chinese until the 10th Century’, Journal of Chinese Linguistics 17(2): 335–72. Peyraube, Alain (1996) ‘Recent Issues in Chinese Historical Syntax’, in James C. T. Huang and Audrey Y. Li (eds) New Horizons in Chinese Linguistics, Dordrecht: Kluwer Academic Press, 161–213. Peyraube Alain (1997a) ‘On Word Order and Word Order Change in Pre-Archaic Chinese’, Chinese Languages and Linguistics IV: 105–24. Peyraube Alain (1997b) ‘On Word Order in Archaic Chinese’, Cahiers de Linguistique – Asie Orientale 26(1): 3–20. Peyraube, Alain (2004) ‘Ancient Chinese’, in Roger D. Woodard (ed) The Cambridge Encyclopedia of the World’s Ancient Languages, Cambridge: Cambridge University Press, Chapter 41, 988–1,014. Peyraube, Alain and Shen Zhongwei (forthcoming) History of the Chinese Language, Cambridge: Cambridge University Press. Peyraube, Alain and Thekla Wiebusch (1995) ‘Problems Relating to the History of Different Copulas in Ancient Chinese’, in Matthew Y. Chen and Ovid J. L. Tzeng (eds) Linguistic Essays in Honor of William S.-Y. Wang, Taipei: Pyramid Press, 383–404. Peyraube, Alain and Wu Fuxiang (2005) ‘Origin and Evolution of Question-words in Archaic Chinese: A Cognitive Approach’, Cahiers de linguistique – Asie Orientale 34(1): 3–24.

16

Ancient Chinese Sagart, Laurent (1993) ‘L’infixe –r– en chinois archaique’, Bulletin de la Société Linguistique de Paris 88(1): 261–93. Sagart, Laurent (1999) The Roots of Old Chinese, Amsterdam and Philadelphia: John Benjamins Pubishing Company. Shen, Pei (1992) Jiagu buci yuxu yanjiu《甲骨卜辞语序研究》(Word order in the Oracle Bone Inscriptions), Taipei: Wenjin chubanshe. Wang, Li (1979) Gudai hanyu changshi《古代汉语常识》(Elements of Ancient Chinese), Beijing: Renmin jiaoyu chubanshe.

17

2 Cantonese Cheng Siu-Pong and Tang Sze-Wing The Chinese University of Hong Kong, Hong Kong, China

The old language–dialect debate is of particular relevance to Cantonese, or Yue, given the ambiguous status it holds in places like Hong Kong. This leads to an inquiry into its distinctive features. They might be connected to its origin, which relates to its divergence from other Chinese varieties. Whereas the subclassification of Yue reflects the intra-group variation, the comparison between Cantonese and Mandarin reveals not only their differences in sounds and words, but also in the lesser-known aspects of grammar.

1.  Cantonese: dialect or language? In English, the term ‘Cantonese’ either refers to the speech of Canton, modern-day Guangzhou city of Guangdong province, or the entire Yue group of similar tongues, or just any variety within the group. It is often used as a convenient term for the mainstream Cantonese spoken in Hong Kong. The native speakers of Cantonese (in the broad and narrow sense) are overwhelmingly of Han Chinese ethnicity.1 Depending on the communities, they refer to what they speak as gwongzauwaa / gwongfuwaa ‘Guangzhou speech’, gwongdungwaa ‘Guangdong speech’, jyutjyu ‘Yue language’, baakwaa ‘colloquial speech’, tongwaa ‘Tang (dynasty) speech’, or simply as zungman ‘Chinese language’.2 Like any other varieties of Chinese, the definition of Cantonese is tied to age-old disputes over whether it should be classified as a language or a dialect. Linguistically, it is not at all difficult to define ‘dialect’: a dialect is a subdivision of language. At issue is how the relationship between such terms as ‘Chinese’ and ‘Cantonese’ is incorporated into these notions. If ‘language’ is defined as a set of mutually intelligible varieties, many Chinese varieties, Cantonese included, can be conceived as separate languages. There is a very low degree of mutual intelligibility between Mandarin and Cantonese. In fact, the major ‘dialect groups’ of Chinese are all unintelligible to one another. This interpretation of ‘language’ implies that the term ‘Chinese’, when encompassing Cantonese (as well as Mandarin, etc.), should be treated as a language family rather than a single language. Thus, Cantonese is said to be a member of the Chinese or Sinitic languages. In most cases where Chinese is so defined, Cantonese is also deemed a language. None­ theless, there is a need for this Cantonese language to exclude a subset of in-group varieties. This is because the group of Yue varieties contains a number of subdivisions whose mutual intelligibility is at best marginal. In this regard, the kind of Cantonese spoken by the great 18

Cantonese

majority of Hong Kong people is a single language. The mainstream Cantonese as used in Guangzhou, Hong Kong, and Macao can also be grouped as one language, given their close similarities, but outside this subgroup which they form (to be discussed below), the inclusion of other varieties depends on one’s judgment on how similar they should be in order to be considered ‘intelligible’. The status of Cantonese also relates to the standard remark that ‘a language is a dialect with an army and a navy’, a quote attributed to Max Weinreich. Spoken in China, which now includes the two subnational entities of Hong Kong and Macao, Cantonese is also widely spoken by the Chinese diaspora in various parts of southeast Asia (Malaysia in particular) and the Western world (notably in Vancouver, Canada). Cantonese is not named as a national language in any of these countries, and no national or subnational governments explicate in their constitutional document that ‘Cantonese’ (the exact wording) serves as an official language. While the constitution of the People’s Republic of China does not touch on the issue of national or official languages, the Law on the Standard Spoken and Written Chinese Language, effective 2001, states that ‘the standard spoken and written Chinese language means Putonghua (a common speech with pronunciation based on the Beijing dialect) [.  .  .]’. Although Hong Kong is part of China, this law, together with the vast majority of other national laws, does not apply in the territory. Instead, the constitutional Basic Law states that ‘[i]n addition to the Chinese language, English may also be used as an official language by the executive authorities, legislature and judiciary of the Hong Kong Special Administrative Region’. A similar article is found in Macao’s Basic Law, with ‘Portuguese’ stated in place of ‘English’. The ambiguity of using ‘Chinese’ in the Basic Laws as well as other local laws (e.g. Hong Kong’s Official Languages Ordinance) raises questions over whether Cantonese per se is official. There is little doubt that in Hong Kong and Macao, Cantonese is the de facto official spoken variety of Chinese. Cantonese is spoken in courts and legislative council meetings almost whenever ‘Chinese’ is used there. Even so, the answer to the question of whether Cantonese by itself attains de jure status is not clear. There is one complicating factor. In formal contexts, government and business documents published in Chinese are given in standard written Chinese, which is based on Mandarin and comprehensible to all literate Chinese speakers worldwide. Written Cantonese, i.e. the written form of Cantonese expressed in Chinese characters, is mostly used in informal register. At first glance, this helps prove the subordinate status of Cantonese. However, since the Chinese script is a logographic writing system, the characters for standard written Chinese can be read with their Cantonese pronunciations. It is easier to come up with the view that it is rather the literary counterpart of the colloquial speech (known as shumianyu ‘literary language’), and its use does not undermine the status of Cantonese. In sociolinguistic terms, Cantonese is obviously an abstand language with respect to another Chinese variety, e.g. Mandarin; the variation between Cantonese and Mandarin in sounds and vocabulary is so great that they are on par with such European languages as French and Spanish. At issue is to what degree Cantonese is ‘autonomous’ enough to be an ausbau language. In mainland China, the answer is simple: Cantonese is considered by the Chinese government as a ‘local dialect’. In the best scenario, the official policies tolerate (but never promote) the use of Cantonese in broadcasting and, to a lesser degree, in schools. In Hong Kong and Macao, Cantonese is used in a variety of domains, ranging from educational to official settings. Nonetheless, given the unresolved issues surrounding the meaning of ‘Chinese’, the Cantonese variety spoken there should be viewed as more an abstand than an ausbau language. The Weinreichian notion of ‘language’ does not accord well with the reality of Cantonese regardless of where it is spoken. A final remark is on the term ‘dialect’ itself. While most Chinese-language publications indicate that 19

Cheng Siu-Pong and Tang Sze-Wing

Cantonese is a dialect, the original term they use in Chinese is fangyan, which literally means ‘regional speech’. It is construed as a notion relative to (and often subordinate to) gongtongyu ‘common language’. The possible semantic discrepancies between ‘dialect’ and fangyan had led DeFrancis (1984) to coin and adopt the term regionlect to capture the sense of the original word.3 This is another way to get over the dialect–language dichotomy. In the following sections, this dichotomy is set aside, and Yue and Cantonese are variably called ‘dialect’ or ‘language’ in a non-committal manner.

2.  Origin and distribution Cantonese, or Yue, belongs to the Sino-Tibetan language family. While all natural languages are descended from some proto-language, the discussions of the origin here focus on the emergence of Yue dialects in present-day southeast China where they are chiefly spoken. A vast area of southern China was known to be occupied by the Bai Yue ‘One Hundred Yue’ tribes. There is still no consensus as to which languages these people spoke in the absence of written records. They were speakers of Austroasiatic-, Kam-Tai-, Hmong-Mien-, or possibly Austronesian-related languages (Norman and Mei 1976; Blust 1984–1985; J. Li 1990; LaPolla 2001), or there was an assortment of speakers from some or all of these linguistic groups. Historical records show that Han Chinese migrated from their homeland in the north to this area starting from the third century bce. While most scholars agree that Modern Cantonese is descended from an earlier form of Chinese, it is still a matter of great debate as to what extent the Chinese variety first spoken in this area bears a relation to Modern Cantonese. The divergence of the earliest Yue variety from those of other parts of China is argued to have taken place from as early as the third century bce when substantial numbers of Han Chinese arrived (J. Li 1990) to as late as the thirteenth century ce (Lau 2000). In those accounts that posit a later divergence time, it is argued that another wave of Han Chinese migration was vital to its formation. Historical studies are also concerned about the influence of non-Han languages on the early Yue. Evidence based on some colloquial words has been provided to show that they form Tai or Hmong substrata (see Yue-Hashimoto 1976, 1991; Bauer 1987; J. Li 1990, etc.). The substratum influence of non-Han languages can be expressed in more neutral terms: Yue has experienced a prolonged period of language contact vis-à-vis non-Han languages. In modern times, the main area for the Yue dialect group is the central and western parts of Guangdong province and the southeastern part of Guangxi Zhuang Autonomous Region. This area also includes Hong Kong and Macao, which are now the Special Administrative Regions of the People’s Republic of China. The Yue dialects are spoken by around 4 to 5% of China’s 1.3 billion people. Language Atlas of China (2012) gives a figure of approximately 59.58 million speakers. In southeast Asia, there are considerable populations of Yue speakers in Malaysia, Vietnam, Singapore, and Indonesia. In particular, around 5% of Malaysia’s population are from the Cantonese speech group (Tan 2005), and an even larger proportion of Malaysians can speak Cantonese with varying degrees of fluency. Outside Asia, the Yue speakers amount to around 1% or more in Canada, the United States, Australia, and New Zealand. According to Language Atlas (1987), there were approximately 11 million overseas speakers in total during the time surveyed. Language Atlas (2012) states that there are over 68 million Yue speakers worldwide. Major works on Yue or Chinese dialects in general, such as Zhan (1991, 2002), Yuan et al. (1960/2001), X. Li (1994), and Language Atlas (1987, 2012), differ on the subclassification of Yue dialects. The following classification includes the four major Yue branches of: (i) Yuehai; (ii) Siyi; (iii) Gao-Yang; and (iv) Guinan; this 20

Cantonese

classification is ‘eclectic’ in nature given that these four branches are not subordinate to any others in the above works, while other branches, e.g. Goulou Branch, are subsumed elsewhere into one of them.4

2.1.  Yuehai Branch Also called the Guangfu Branch, it is mainly spoken in the Pearl River (Zhujiang) Delta. This area includes Guangzhou, the capital of Guangdong province, together with Hong Kong and Macao. For this reason, this branch is named after Guangzhou, otherwise known as Yuehai or Guangfu. Yuehai Branch is the most well-known and widely spoken branch of Yue dialects. Language Atlas (2012) gives a figure of about 28.34 million speakers (excluding the speakers of the Goulou Branch; to be discussed below). The same source indicates that approximately 20.72 million of them live in Guangdong. In addition, around 7 million people from Hong Kong and Macao can speak this variety. Depending on the extension of this subgroup, its area also covers the central and northern parts of Guangdong, and scattered parts of Guangxi such as downtown Hezhou and Wuzhou. As a result of migration from this core area, the Yuehai Branch is the predominant form of Yue spoken in overseas Chinese communities. In Malaysia, it is a major Chinese variety spoken in Kuala Lumpur, Ipoh, Kuantan, Seremban, and Sandakan. In most Chinatowns in North America, the Yuehai Branch has been the major Chinese variety (amid a growing number of Mandarin-speaking immigrants), often achieving its prominent status at the expense of the Siyi Branch. Particularly in Metro Vancouver, Canada, its use has been scaled up to the city level. The city of Richmond is an example par excellence. Accentual difference is found between the varieties spoken in downtown Guangzhou and the outskirt areas of the city, especially those districts newly (re-)attached to Guangzhou around this millennium, e.g. Panyu. A variety closely similar to the one used in downtown Guangzhou is also predominantly spoken in Hong Kong. Guangzhou and Hong Kong Cantonese are mutually intelligible to a very high degree, despite some lexical differences. However, it should be noted that the emergence and rise of Hong Kong Cantonese is resulted from the combined factors of immigration and language shift. Right before the beginning of the colonial time in the nineteenth century, a different Yue dialect, i.e. the Guan-Bao variety, was chiefly spoken. A gradual shift to Guangzhou-style Cantonese also took place in Macao. The rise of Cantonese to greater prominence in the latter half of the last century was mainly due to the social and economic development of Hong Kong. The extended influence of Hong Kong’s trade and entertainment industries has promoted the use of Cantonese in a greater area of Guangdong and overseas Chinese communities. As a result, many people in traditionally non-Cantonese-speaking regions, such as the eastern part of Guangdong, came to understand it and often speak it as a second language.

Guan-Bao Branch Zhan (1991) classifies it as a separate branch of Yue dialects. It is traditionally spoken in the eastern part of the Pearl River Delta. Precisely, it is the indigenous Yue variety of Hong Kong and Shenzhen, formerly parts of Xin’an (later named Bao’an), and of the neighboring Dongguan, hence the name ‘Guan-Bao’. In Hong Kong, it is known as Punti or waitauwaa. The number of its speakers is diminishing to the point that it has become endangered, and its use is now restricted to the older generation in the more rural parts of Hong Kong. A decline in use is also observed in the rapidly developing Shenzhen and Dongguan. The 21

Cheng Siu-Pong and Tang Sze-Wing

mutual intelligibility between Guan-Bao and Guangzhou Cantonese is less than ideal. Although downtown Guangzhou and Hong Kong are about 130 kilometers apart, the mainstream Cantonese in Hong Kong is much closer to Guangzhou Cantonese than it is to Guan-Bao.

Xiangshan Branch Another independent branch of Yue given in Zhan (1991), it is spoken in the western part of the Pearl River Delta, including Zhuhai and Zhongshan, which had previously been called Xiangshan before it was named after Sun Yat-sen (Sun Zhongshan). Formerly, the Xiangshan Branch was also spoken in Macao.

Goulou Branch First proposed around the time when the first edition of Language Atlas (1987) was published, it is spoken in the peripheral areas outside the Pearl River Delta, to the north and west of the heartland of Yuehai Branch. This area extends into Guangxi, and occupies the eastern portion of the region. It is named after the mountain range of Goulou in southeastern Guangxi bordering Guangdong. According to Language Atlas (2012), the number of Goulou speakers amounts to approximately 10.30 million.

2.2.  Siyi Branch Siyi ‘four counties’ refers to the region southwest of Guangzhou that lies to the west of the Pearl River Delta. The ‘four counties’ include Taishan, Kaiping, Enping, and Xinhui. In modern times, they are all under the jurisdiction of the city of Jiangmen, which also includes Heshan, so together they are called Wuyi ‘five counties’. The population of Siyi speakers in Guangdong is approximately 3.88 million, according to Language Atlas (2012). The main Siyi-speaking area includes modern-day Jiangmen and the western part of Zhuhai. Migration of Siyi speakers out of this region contributed to its prominence elsewhere. In Hong Kong after the Second World War and the subsequent Chinese Civil War, the influx of Siyi immigrants led to a time when the Siyi Branch was more spoken before it was largely displaced by Hong Kong Cantonese a few decades later. Besides, the Siyi Branch was once the most influential Chinese variety in the American Chinese communities. Up until half a century ago, it was spoken by the great majority of Chinese immigrants to the United States (Bauer and Benedict 1997). Within the Siyi Branch, the Taishan subdialect was the most prominent variety spoken overseas. Although its status was largely replaced by the Yuehai Branch and Mandarin, the use of Siyi remains significant in some American Chinatowns, especially among the earlier migrants.

2.3.  Gao-Yang Branch Also called the Gao-Lei Branch, it is the predominant Chinese variety spoken in the southwestern part of Guangdong, which includes Gaozhou and Yangjiang. The core areas of Gao-Yang, however, exclude the Leizhou Peninsula, together with some linguistic exclaves nearby, where the Gao-Yang Branch might be spoken but Hakka or Min dialects are the main speech. Language Atlas (2012) states that the Gao-Yang Branch is spoken by approximately 6.43 million people. The Wu-Hua Branch, considered a separate branch in Language Atlas, has about 1.28 million speakers. It is classified as part of the Gao-Yang-speaking region in the other accounts. 22

Cantonese

2.4.  Guinan Branch Guinan literally means ‘southern Guangxi’. The first edition of The Encyclopedia of China uses it as an umbrella term to cover all subdivisions of Yue dialects spoken in Guangxi. According to Language Atlas (2012), approximately 16.86 million Guangxi people speak the Yue dialects. The Yue varieties in Guangxi are further divided into the branches of: (i) Yuehai; (ii) Goulou; (iii) Yong-Xun; and (iv) Qin-Lian. The Goulou Branch is spoken in a continuous area across Guangdong and Guangxi, and the predominant use of Yuehai Branch in Guangxi is found in the linguistic enclaves, or ‘dialect islands’. Only the Yong-Xun and Qin-Lian Branches are specific to Guangxi.

Yong-Xun Branch It is used as the majority variety in scattered places along the rivers of Yongjiang and Xunjiang. Most of these places have served as inland river ports. Language Atlas (2012) gives a figure of approximately 4.69 million speakers.

Qin-Lian Branch It is spoken in the traditional areas of Qinzhou and Lianzhou (modern-day Hepu) along the southern (and the only) coast of Guangxi. Situated to the west of Guangdong coastal towns, this region was part of Guangdong until after the founding of the People’s Republic of China. According to Language Atlas (2012), the number of speakers of this branch amounts to approximately 3.90 million.

3.  Differences between Cantonese and Mandarin It has been well documented that Cantonese differs from Mandarin in terms of sounds and vocabulary. In general, the sound system of Yue dialects is comparatively more complicated by having more rhymes and tones. Bauer and Benedict (1997) identify 58 rhymes in Cantonese including two syllabic nasals. By contrast, a standard Mandarin rhyme table includes 35 rhymes. A major reason for the difference lies in the fact that in Cantonese /p, t, k, m/ are found in the syllable-final position, yet these codas are absent in Mandarin. Regardless of tones, a regular Mandarin dictionary contains a number of 410 syllables, which is less than the 584 Cantonese syllables given in Bauer and Benedict (1997) for the standard Chinese characters, not to mention others solely used for colloquial speech. The number of tonal categories in Yue dialects varies from six to ten (Zhan 1991). As for Hong Kong Cantonese (as well as other Yue varieties), the category of Entering tones (tones for syllables with a coda from /p, t, k/) might be subsumed into the other categories, depending on the categorization criteria, and the Entering tones are viewed as the short or extra-short variants of the other tones (Zee 1999). Even so, it contains a total of six tones. This contrasts with Mandarin, which has only four. Historically, tonal split and merger have given rise to varying numbers of tones across different Chinese varieties. As in other modern dialects, Yue experienced the split of tonal categories primarily as a result of the loss of voicing distinction of initials. Subsequent tone mergers, however, were more restrained in the case of Yue dialects. In modern-day Hong Kong Cantonese, there are 19 consonants, including plosives /p, ph, t, th, k, kh, kw, kwh/, nasals /m, n, ŋ/, affricates /ts, tsʰ/, fricatives /f, s, h/, and ‘semi-vowels’ as well as other approximants /j, w, l/. Vowels include both front and back, 23

Cheng Siu-Pong and Tang Sze-Wing

and high and low monothongs: /i, e, y, u, o, ɛ, ɔ, œ, ɵ, ɐ, a/. In addition to these monothongs, there are also ten gliding vowels (or diphthongs): /iw, ɵʮ, uʮ, ej, ɔj, ow, ɐj, ɐw, aj, aw/. The maximal syllable takes the form of ‘initial consonant + vowel + final consonant + tone’, or CiVCf /T (Bauer and Benedict 1997). In other words, a consonant cluster is not found within an indigenous syllable. Both the initial and final consonants might be absent, and a final consonant (from /m, n, ŋ, p, t, k/) is disallowed when the vowel is a gliding one. A syllable always contains a vowel, with the exception of two syllabic nasals: /m, ŋ/. Assuming that the three Entering tones, namely Upper Dark Entering /5/, Lower Dark Entering /3/, and Light Entering /2/, are respectively subsumed into /55/, /33/, and /22/, the modern-day Hong Kong variety of Cantonese contains a total of six tones, which are /55/, /25/, /33/, /21/, /23/, and /22/.5 In terms of vocabulary, Hou (2002) indicates that about two thirds of basic words in Yue dialects are common to both Yue and Standard Chinese, or, in his wording, the ‘common national language’. Out of the other one third or so of words that are different, over 90% of them were arisen as a result of the innovation of Yue speakers, whereas some others are old Chinese cognates that are no longer in use in modern Standard Chinese. Many scholars also note that some of the colloquial words owe their origin to non-Han languages including Tai or Hmong. The Yue dialects of major ports, notably Hong Kong Cantonese, stand out from other Chinese varieties by having more words borrowed from a European language, English in particular. Apart from English, Japanese is another major loan language. These words entered the Cantonese vocabulary in the form of semantic translation, phonetic borrowing, or in the case of most Japanese loanwords, the direct use of Japanese-derived words with the Chinese characters (kanji) read in Cantonese pronunciation. Structurally, the relative order of morphemes in a disyllabic word in Mandarin is sometimes reversed in Cantonese. Examples include ganjiu ‘important’ and siuje ‘night-time snack’. In Mandarin, the morpheme order is reversed and they are rendered respectively as yaojin and yexiao. Another well-quoted example is gaigung ‘rooster’. Although gunggai ‘male + chicken’ is also used, gaigung ‘chicken + male’ is indigenous. The morphemic ordering of this indigenous word in Yue and also in other Southern dialects has drawn speculation that the influence of non-Han languages was at work (see Yuan et al. 1960; Ouyang 1991, etc.). Other morphological variation includes the higher tendency for Cantonese, as compared to Mandarin, to have monosyllabic words, and the lower tendency for it to use certain suffixes, e.g. zi. It is often assumed that all Chinese varieties share the same grammar. Despite the general perception of close similarities in grammar among them, there are significant differences that also set them apart. Functional words are important contributors to the syntactic variation. The following parts are based on Tang and Cheng (2014) and focus on the distinctive grammatical features of Cantonese that are less salient or cannot be found at all in Mandarin. The variables involved include (i) structural particles; (ii) directional verbs; (iii) aspect markers; (iv) definiteness; (v) double-object construction; (vi) comparative sentences; (vii) passives; and (viii) post-verbal elements.

3.1.  Structural particles Homophonous particles in Mandarin might be audibly differentiated when translated into Cantonese. Zhu (1961) classifies three types of de in Mandarin: de1 follows an adverb to form an adverbial unit; de2 follows an adjective to form an adjectival phrase; de3 follows different phrases to form a nominal unit. 24

Cantonese

(1) Ta zhengtian buting-de shuo. he all-day non-stop-DE speak ‘He speaks non-stop all day’

de1-adverbial

honghong-de. (2) Zhe duo hua this Cl flower red-DE ‘This flower is a bit red’

de2-adjectival

(3) Xinxian-de ye you. fresh-DE also have ‘There is also something fresh’

de3-nominal

Zhu (1980) further suggests that the three subcategories of de correspond to gam, dei, and ge in Cantonese. The three sentences can be translated into Cantonese as follows. (4) Keoi sengjat mting-gam gong. he all-day non-stop-GAM speak ‘He speaks non-stop all day’

gam-adverbial

(5) Ni do faa hunghung-dei. this Cl flower red-DEI ‘This flower is a bit red’

dei-adjectival

(6) Sansin-ge dou jau. fresh-GE also have ‘There is also something fresh’

ge-nominal

In Cantonese, gam, dei, and ge are clearly separate morphemes as they differ in their pronunciation. With this in view, Zhu argues for the three subsets of de in Mandarin. As far as these structural particles are concerned, Cantonese is more ‘marked’ in their division of labor.

3.2.  Directional verbs Yuan et al. (1960) note that Mandarin tends to use shang ‘go up’ or dao ‘arrive’ as the main verbs, and qu ‘go’ and lai ‘come’ usually follow the object, as they do in (7) and (8). In Cantonese, on the other hand, heoi ‘go’ and lai ‘come’ are mostly used directly as the main verb, as in (9) and (10). (7) Wo shang Beijing qu. I go-up Beijing go ‘I go to Beijing’

(Mandarin)

(8) Wo jiejie shi zuotian dao Guangzhou lai de. I sister be yesterday arrive Guangzhou come DE ‘My elder sister came to Guangzhou yesterday’ (9) Ngo heoi Bakging. I go Beijing ‘I go to Beijing’

(Cantonese)

25

Cheng Siu-Pong and Tang Sze-Wing

(10) Ngo gaaze hai camjat lei Gwongzau ge. I sister be yesterday come Guangzhou GE ‘My elder sister came to Guangzhou yesterday’ According to Yuan et al., the use of qu ‘go’ and lai ‘come’ as the main verbs is also found in Mandarin, but this is obviously under the influence of the Southern dialects.6

3.3.  Aspect markers C. N. Li and Thompson (1989) identify four Mandarin aspect markers, le, zai, zhe, and guo: the perfective aspect is marked by le, the progressive aspect by zai, the durative aspect by zhe, and the experiential aspect by guo. Cheung (1972/2007) identifies six Cantonese aspect markers, in addition to a zero marker. They are perfective zo, experiential gwo, progressive gan, habitual hoi, durative zyu, and inchoative hei(soeng)lai. Matthews and Yip (1994/2011) and X. Li et al. (1995) added two more: delimitative haa and successive lokheoi. A linkage can be drawn between Mandarin and Cantonese perfective, experiential, and progressive / durative markers. Nonetheless, it is worth noting the distinction between Mandarin progressive / durative zai / zhe and Cantonese progressive / durative gan / zyu. Previous studies indicate that activity verbs take zai and stative verbs take zhe (Paris 1981; C. N. Li and Thompson 1989). Smith (1991/1997) also notes that in some Mandarin varieties the use of zhe has gradually become more extended and can convey both the progressive and durative aspect. In contrast, neither gan nor zyu has taken over the function of one another in Cantonese. While gan is a typical progressive maker, zyu has a static meaning and is used as a marker for durative aspect, as in (11) and (12). bo. (11) Zoengsaam  daa-gan Zoengsaam  play-Prog ball ‘Zoengsaam is playing ball’ (12) Zoengsaam  daai-zyu ngaangeng. Zoengsaam  wear-Dur eyeglasses ‘Zoengsaam is wearing eyeglasses’ While Mandarin zhe can be used in the equivalents of both sentences above, Cantonese progressive gan and durative zyu are unambiguously distinctive and cannot be used interchangeably.

3.4. Definiteness Like Mandarin classifiers, Cantonese classifiers precede nouns. In both varieties, the [classifier + noun] sequence follows a numeral, which itself might be preceded by a demonstrative, as in the following examples. (13) zhe san ben shu this three Cl book ‘these three books’

(Mandarin)

(14) ni saam bun syu this three Cl book ‘these three books’

(Cantonese)

26

Cantonese

The addition of demonstratives and numerals is not compulsory. The bare [classifier + noun] form is possible in both Mandarin and Cantonese. As indicated by Cheng and Sybesma (1999), the difference between Cantonese and Mandarin in this nominal phrase is a contrast in the interpretations. In Cantonese, a nominal phrase of this form can either have a definite or indefinite reading. In Mandarin, only the indefinite reading is available. Therefore, the [classifier + noun] form is found in the subject position of Cantonese, but not Mandarin, given that this position normally limits the interpretation of nominals to a definite one. (15) Wo xiang mai ben shu. I want buy Cl book ‘I want to buy a book’

(Mandarin)

(16) *Ben shu hen hou. *Cl book very thick (17) Ngo soeng maai bun syu. I want buy Cl book ‘I want to buy a book’

(Cantonese)

(18) Bun syu hou hau. Cl book very thick ‘The book is thick’ On the other hand, Mandarin bare nouns can receive definite, indefinite, or generic readings. Bare nouns in Cantonese can also have indefinite or generic readings, but the definite interpretation is largely prohibitive. For the definite reading, it is better for a bare noun to go with a classifier.7 (19) Gou yao guo malu. dog want cross road ‘The dog wants to cross the road’

(Mandarin)

(20) Zek gau jiu gwo maalou. Cl dog want cross road ‘The dog wants to cross the road’

(Cantonese)

3.5.  Double-object construction It has been noted that Cantonese has a double-object construction different from the one found in Mandarin (Yuan et al. 1960; Gao 1980). An animate object precedes an inanimate one in Mandarin, but in Cantonese the order appears to be in reverse. (21) Ta gei-le wo yi-ben shu. he give-Perf I one-Cl book ‘He gave me a book’

(Mandarin)

(22) Keoi bei-zo jat-bun syu ngo. he give-Perf one-Cl book I ‘He gave a book to me’

(Cantonese)

27

Cheng Siu-Pong and Tang Sze-Wing

In most cases, the inanimate and animate objects are respectively analyzed as the direct object (DO) and indirect object (IO). According to Cheung (1972), Cantonese places the DO before the IO most prevalently when the verb is bei ‘give’. Note that bei can be used in two different positions in the same sentence. The same is true for Mandarin gei. (23) Ta gei-le yi-ben shu gei wo. he give-Perf one-Cl book to I ‘He gave a book to me’

(Mandarin)

jat-bun syu bei ngo. (24) Keoi bei-zo he give-Perf one-Cl book to I ‘He gave a book to me’

(Cantonese)

While the first gei / bei ‘give’ is a verb, the second gei / bei ‘to’ is regarded as prepositional, as it often co-occurs with many other verbs to indicate IO.8 Traditionally, the form of doubleobject construction as illustrated in (22) is known to have been ‘inverted’, since the order found in Mandarin is reversed. On the other hand, more recent studies (Tang 1998a, 2003; Liu 2001) show that the ‘DO + IO’ order in Cantonese actually originated from the ‘DO + bei + IO’ construction through the deletion of bei ‘to’. That is to say, the original form of (22) is (24). Therefore, what distinguishes Mandarin and Cantonese is not the word order of objects, but the deletability of the preposition.9

3.6.  Comparative sentences Generally speaking, Cantonese and Mandarin use different comparison words, which results in different structures for comparative sentences. (25) Zhangsan  bi Lisi  gao. Zhangsan  compare Lisi  tall ‘Zhangsan is taller than Lisi’

(Mandarin)

(26) Zoengsaam  gou-gwo Leisei. Zoengsaam  tall-pass Leisei ‘Zoengsaam is taller than Leisei’

(Cantonese)

In the Cantonese example, the comparison word is gwo ‘pass’, which follows the adjectival predicate gou ‘tall’. The Cantonese comparative sentences are in the form of ‘X + predicate + comparison word + Y’. In the Mandarin example, the comparison word bi ‘compare’ precedes both Y and the adjective; the comparative sentences in Mandarin are in the form of ‘X + comparison word + Y + predicate’. This construction, however, is not impossible in Cantonese. As stated in Matthews and Yip (1994), there is an alternative maker in Cantonese, i.e. bei ‘compare’. It is based on the bi construction in Mandarin, and ‘has the formal flavor of Mandarin-based syntax’, as shown in their example (1994: 167). (27) Aawan  bei keoi muimui leng. Aawan compare she sister pretty ‘Aawan is prettier than her younger sister’ 28

Cantonese

It shows that the order of ‘X + comparison word + Y + predicate’ is also found in Cantonese, although the comparative construction with gwo ‘pass’ is arguably more common and ‘authentically’ Cantonese.10

3.7. Passives Following the distinction drawn by Ting (1995, 1998), passive sentences in Mandarin are classified into what are known as ‘long passives’ and ‘short passives’. A passive morpheme, bei, is used in both categories. Their difference is that an identified agent follows bei in the long passives, but such an agent is absent in the short ones. (28) Zhangsan  bei Lisi  piping-le. Zhangsan  BEI Lisi  criticize-Perf ‘Zhangsan was criticized by Lisi’

(Long Passive)

(29) Zhangsan  bei piping-le. Zhangsan  BEI criticize-Perf ‘Zhangsan was criticized’

(Short Passive)

Long and short passives are not equally acceptable in Cantonese. Unlike its counterpart in Mandarin, the indigenous passive marker bei2 (in high rising tone /25/ ) must contain an agent. Otherwise, the sentence will become ungrammatical. (30) Zoengsaam  bei2 *(ngo) naau. Zoengsaam  BEI I criticize ‘Zoengsaam was scolded (by me)’ Cantonese passives without an identified agent will be acceptable only if the passive morpheme is given in a different form, bei6, with the low level tone /22/. (31) Zoengsaam  bei6 zizaak. Zoengsaam  BEI censure ‘Zoengsaam was censured’ Passives with bei6 are normally used in formal speech and in literary contexts. While bei2 is the native form, bei6 is assumed to be borrowed from Mandarin. As a result, its usage is highly restricted. In everyday Cantonese, the use of long passives, rather than the short ones, is predominant. Even if the agent needs to be ‘hidden’ for some reason, it is more natural to use a long passive containing an unspecified agent, such as jan ‘person’, than to use a short passive with bei6.11

3.8.  Post-verbal elements Cantonese is rich in post-verbal elements in comparison to Mandarin. There are a large number of suffixes and sentence-final particles, and also a high degree of co-occurrence of post-verbal elements. Cheung (1972) identifies 15 predicative suffixes. Matthews and Yip (1994) include seven major aspect particles and 29 verbal particles. The number of sentencefinal particles is even higher. As noted by Fung (2000), previous studies show that it may range from 30 to as many as 90 (Kwok 1984; Ouyang 1990; Leung 2005). Furthermore, 29

Cheng Siu-Pong and Tang Sze-Wing

multiple particles often co-occur in the sentence-final position, as in the following example (Leung 2005: 87). (32)

Nei waa zeoido to do go leng jyut tim ge zaa aamaa? you say at-most delay more Cl more month TIM GE ZAA AAMAA ‘You said (I can) delay for one or more months at most, right?’

As for their usage, Kwok (1984) connects them with sentence functions. The particles are categorized as the signals of speaker’s attitude. Luke (1990) discusses three particles (laa, lo, and wo), which manifest different interactional goals. The particle laa is used to establish common grounds in conversation; lo is associated with interactional achievements; and wo is used to signal expectation and noteworthiness. Fung (2000) groups the majority of sentence-final particles into three main categories, i.e. Z-, L-, and G- families. Particles sharing the same initial belong to the same family, the distinctive core features of which are based on both their semantic and contextual meanings. Z-particles, including ze, zek, zaa, etc., have the core semantic feature of ‘restriction’; the core feature of L-particles, which include lei, laa, etc., is to signal temporal and aspectual meanings; the core function of G-particles, including ge and gaa, is to mark a situation given in a context.12 Cantonese is rich in both pre-verbal and post-verbal elements, and several pre-verbal and post-verbal elements share similar semantic properties. Take caa’mdo ‘almost’ and gamzai ‘almost’ as examples. They can co-occur in the same sentence. (33) Ngo caa’mdo gong-jyun gamzai. I almost speak-finish almost ‘I almost finished talking’ Tang (2006a, 2006b, 2007) suggests that post-verbal elements like gamzai ‘almost’ and pre-verbal ones like caa’mdo ‘almost’ form a constituent known as the ‘discontinuous construction’. He argues that in such a construction, the combination of pre-verbal and postverbal elements ‘appears to be redundant’. That means the apparent redundancy of the pre-verbal and post-verbal elements should be understood as their being close in meaning. So far, over 40 different pre- and post-elements have been identified to belong to this kind of construction. In comparison, the discontinuous constructions are rare in Mandarin.13

4. Implications Many scholars contend that Chinese varieties in essence only differ in sounds and vocabulary (Yuan et al. 1960; Chao 1968; Lü 1982). Since it is easier to observe the phonetic and lexical differences, many previous studies on Chinese dialects focus on these two aspects. Comparatively speaking, the outcomes of syntactic variation studies are less fruitful. On the surface, there are more similarities than differences in the grammar of different Chinese varieties. However, some ‘similarities’ are in fact more apparent than real. A number of obvious grammatical variables have been described, but there is still plenty of room for further research. Given the data from Cantonese as well as other dialects, we can discover many linguistic phenomena that are absent in Mandarin. This allows us to have a better understanding of Chinese as a whole. As shown in Zhu (1980), and Cheng and Sybesma (1999), etc., the comparative studies between Cantonese and Mandarin can help identify the grammatical subtleties of one another. 30

Cantonese

Notes   1 Many of the non-Han speakers of Cantonese belong to the Zhuang ethnic group, which mostly lives in the Guangxi Zhuang Autonomous Region.   2 The Romanization of Cantonese in this chapter is based on Standard Cantonese, and follows the Linguistic Society of Hong Kong (LSHK) Cantonese Romanization Scheme, also known as jyutping.   3 See also Mair (1991) for his proposal of using topolect as ‘an exact, neutral translation of fangyan’.   4 Those varieties whose status is contestable are excluded, e.g. Danzhouhua and Tuhua.   5 Tone numerals are used. The non-Entering tones of /55/, /25/, /33/, /21/, /23/, and /22/ are respectively known as Dark Flat, Dark Rising, Dark Departing, Light Flat, Light Rising, and Light Departing tones by their traditional names. /25/, /21/ and /23/ might be given as /35/, /11/ and /13/. For simplicity, the description above follows Bauer and Benedict (1997).   6 See Yiu (2005) for the spatial characteristics and semantic extension of Cantonese directional verbs.   7 Simpson (2010) observes that some native speakers of Cantonese seem to be able to interpret bare nouns as definite, and gives the following example to illustrate this. Coeng hai mai saan-zo? window be not-be closed-Perf ‘Is the window closed?’ See Y.-H. A. Li (1997), Cheng and Sybesma (1999), Au Yeung (2005), Sio (2006), Wu and Bodomo (2009), and Tang (2011) for the syntactic analysis of definiteness in Cantonese and Mandarin.   8 See Huang and Ahrens (1999) for the contrasting view that the post-DO gei in Mandarin is verbal.   9 See Zhang (2011) for typological evidence to support the deletion approach along these lines. 10 See Mok (1998) for the syntactic analysis of gwo comparatives. 11 See Chin (2009) for the historical development of bei that connects its functions in passives and double-object constructions. 12 See S.-P. Law (1990), Tang (1998b), Y.-K. A. Law (2004), and B. Li (2006) for the proposals on the syntactic positioning of sentence-final particles. 13 See Tang (2009) for the syntax and hierarchical order of discontinuous constructions.

References Australian Academy of the Humanities and Chinese Academy of Social Sciences (1987) Language Atlas of China, Hong Kong: Longman. Au Yeung, Wai-Hoo (2005) ‘An Interface Program for Parameterization of Classifiers in Chinese’, Doctoral Dissertation, Hong Kong University of Science and Technology. Bauer, Robert S. (1987) ‘Kadai Loanwords in Southern Chinese Dialects’, Transactions of International Conference of Orientalists in Japan 32: 95–111. Bauer, Robert S. and Paul K. Benedict (1997) Modern Cantonese Phonology, Berlin: Mouton de Gruyter. Blust, Robert (1984–1985) ‘The Austronesian homeland: a linguistic perspective’, Asian Perspectives 26(1): 45–67. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley and Los Angeles: University of California Press. Cheng, Lai-Shan Lisa and Rint Sybesma (1999) ‘Bare and Not-so-bare Nouns and the Structure of NP’, Linguistic Inquiry 30(4): 509–42. Cheung, Hung-Nin Samuel (1972/2007) Xianggang Yueyu Yufa de Yanjiu (A Grammar of Cantonese as Spoken in Hong Kong), Hong Kong: The Chinese University Press. Chin, Chi-On (2009) ‘The Verb GIVE and the Double-object Construction in Cantonese in Synchronic, Diachronic and Typological Perspectives’, Doctoral Dissertation, University of Washington. DeFrancis, John (1984) The Chinese Language: Fact and Fantasy, Honolulu: University of Hawaii Press. Fung, Suk-Yee Roxana (2000) ‘Final Particles in Standard Cantonese: Semantic Extension and Pragmatic Inference’, Doctoral Dissertation, Ohio State University. Gao, Huanian (1980) Guangzhou Fangyan Yanjiu (A Study of the Cantonese Dialect). Hong Kong: The Commercial Press. Hou, Jingyi (2002) Xiandai Hanyu Fangyan Gailun (An Outline of Modern Chinese Dialects), Shanghai: Shanghai Jiaoyu Chubanshe.

31

Cheng Siu-Pong and Tang Sze-Wing Huang, Chu-Ren and Kathleen Ahrens (1999) ‘The Function and Category of Gei in Mandarin Ditransitive Constructions’, Journal of Chinese Linguistics 27(2): 1–26. Institute of Linguistics of the Chinese Academy of Social Sciences, Institute of Ethnology and Anthropology of the Chinese Academy of Social Sciences, and Language Information Sciences Research Centre of the City University of Hong Kong (2012) Zhongguo Yuyan Dituji (Language Atlas of China), 2nd ed., Beijing: The Commercial Press. Kwok, Helen (1984) Sentence Particles in Cantonese, Hong Kong: Center of Asian Studies, University of Hong Kong. LaPolla, Randy J. (2001) ‘The Role of Migration and Language Contact in the Development of the Sino-Tibetan Language Family’, in Alexandra Y. Aikhenvald and R. M. W. Dixon (eds) Areal Diffusion and Genetic Inheritance: Problems in Comparative Linguistics, New York: Oxford University Press, 225–54. Lau, Chun-Fat (2000) ‘Xiandai Yueyu yuanyu Songmo yimin shuo’ (‘A Theory that Modern Cantonese Derived from Late Song Dynasty Immigrants’), in Proceedings of 7th International Conference on Yue Dialects, Beijing: The Commercial Press. Law, Sam-Po (1990) ‘The Syntax and Phonology of Cantonese Sentence-final Particles’, Doctoral Dissertation, Boston University. Law, Yan-Kei Ann (2004) ‘Sentence-final Focus Particles in Cantonese’, Doctoral Dissertation, University College London. Leung, Chung-Sum (2005) Dangdai Xianggang Yueyu Yuzhuci de Yanjiu (A Study of the Utterance Particles in Cantonese as Spoken in Hong Kong), Hong Kong: Language Information Sciences Research Centre, City University of Hong Kong. Li, Boya (2006) ‘Chinese Final Particles and the Syntax of the Periphery’, Doctoral Dissertation, Leiden University. Li, Charles N. and Sandra A. Thompson (1989) Mandarin Chinese: A Functional Reference Grammar, Berkeley and Los Angeles: University of California Press. Li, Jingzhong (1990) ‘Yueyu shi Hanyu zuqun zhong de duli yuyan’ (‘Cantonese as an Independent Language in the Family of Chinese Languages’), in Proceedings of 2nd Conference on Yue Dialects, Guangzhou: Jinan University Publishers. Li, Xinkui (1994) Guangdong de Fangyan (Dialects of Guangdong), Guangzhou: Guangdong Renmin Chubanshe. Li, Xinkui, Huang Jiajiao, Shi Qisheng, Mai Yun, and Chen Dingfang (1995) Guangzhou Fangyan Yanjiu (A Study of the Guangzhou Dialect), Guangzhou: Guangdong Renmin Chubanshe. Li, Yen-Hui Audrey (1997) ‘Structures and Interpretations of Nominal Expressions’, Ms., University of Southern California. Liu, Danqing (2001) ‘Hanyu geiyulei shuangjiwu jiegou de leixingxue kaocha’ (‘A Topological Study on the Ditransitive Construction of Chinese Give Verbs’), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 5: 387–98. Lü, Shuxiang (ed) (1982) Yuwen Changtan (Elementary Knowledge of Language), Hong Kong: Joint Publishing. Luke, Kang-Kwong (1990) Utterance Particles in Cantonese Conversation, Amsterdam and Philadelphia: John Benjamins Publishing Company. Mair, Victor H. (1991) ‘What Is a Chinese ‘Dialect/Topolect’? Reflections on Some Key Sino-English Linguistic Terms’, in Sino-Platonic Papers 29: 1–31. Philadelphia: Department of East Asian Languages and Civilizations, University of Pennsylvania. Matthews, Stephen and Virginia Yip (1994/2011) Cantonese: A Comprehensive Grammar, London and New York: Routledge. Mok, Sui-Sang (1998) ‘Cantonese Exceed Comparatives’, Doctoral Dissertation, University of California, San Diego. Norman, Jerry and Mei Tsu-Lin (1976) ‘The Austroasiatics in Ancient South China: Some Lexical Evidence’, Monumenta Serica 32: 274–301. Ouyang, Jueya (1990) ‘Guangzhouhua de yuqi zhuci’ (‘The Modal Particles in Cantonese’), in Wang Li Xiansheng Jinian Lunwenji (Wang Li Memorial Volumes), Beijing: The Commercial Press, 464–76. Ouyang, Jueya (1991) ‘Yunyong diceng lilun yanjiu shaoshu minzu yuyan yu Hanyu de guanxi’ (‘Study of the Relationship between Minority Languages and Han Chinese with the Substratum Theory), Minzu Yuwen (Minority Languages of China) 6: 23–9.

32

Cantonese Paris, Marie-Claude (1981) Problèmes de Syntaxe et de Sémantique en Linguistique Chinoise, Paris: Collège de France. Rao, Bingcai (1988) ‘Yue fangyan’ (‘Yue Dialects’), in Zhongguo Dabaike Quanshu – Yuyan Wenzi (Encyclopedia of China – Language Volume), Beijing: Zhongguo Dabaike Quanshu Chubanshe. Simpson, Andrew (2010) ‘The Definite Bare Classifier Construction: A Cross-linguistic Study’, Paper presented at Workshop on Definiteness in Chinese, The Chinese University of Hong Kong. Sio, Ut-seong Joanna (2006) Modification and Reference in the Chinese Nominal, Utrecht: LOT. Smith, Carlota S. (1991/1997) The Parameter of Aspect, Dordrecht: Kluwer Academic Publishers. Tan, Chee-Beng (2005) ‘Chinese in Malaysia’, in Melvin Ember, Carol R. Ember, and Ian Skoggard (eds) Encyclopedia of Diasporas: Immigrant and Refugee Cultures Around the World, New York: Kluwer, 697–706. Tang, Sze-Wing (1998a) ‘On the “Inverted” Double Object Construction’, in Stephen Matthews (ed) Studies in Cantonese Linguistics, Hong Kong: The Linguistic Society of Hong Kong, 35–52. Tang, Sze-Wing (1998b) ‘Parametrization of Features in Syntax’, Doctoral Dissertation, University of California, Irvine. Tang, Sze-Wing (2003) Hanyu Fangyan Yufa de Canshu Lilun (A Parametric Theory of Chinese Dialectal Grammar), Beijing: Peking University Press. Tang, Sze-Wing (2006a) ‘Yueyu dakzai, matzai, gamzai shifou shuyu tongyige jiazu?’ (‘Do dakzai, matzai, and gamzai in Cantonese Belong to the Same Family?’), Studies in Chinese Linguistics 1: 1–11. Tang, Sze-Wing (2006b) ‘Yueyu yiwenju sin de jufa tedian’ (‘Syntactic Properties of Sin in Cantonese Interrogatives’), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 3: 225–32. Tang, Sze-Wing (2007) ‘Yueyu kuangshi xuci de jubuxing he duochongxing’ (‘Locality and Multiplicity of Function Words in the Discontinuous Construction in Cantonese’), in Proceedings of the 10th International Conference on Yue Dialects, Beijing: Zhongguo Shehui Kexue Chubanshe. Tang, Sze-Wing (2009) ‘The Syntax of Two Approximatives in Cantonese: Discontinuous Constructions Formed with zai6’, Journal of Chinese Linguistics 37(2): 227–56. Tang, Sze-Wing (2011) ‘On Gerundive Nominalization in Mandarin and Cantonese’, in Foong-Ha Yap, Karen Grunow-Harsta, and Janick Wrona (eds) Nominalization in Asian Languages: Diachronic and Typological Perspectives, Amsterdam and Philadelphia: John Benjamins Publishing Company, 147–60. Tang, Sze-Wing and Cheng Siu-Pong (2014) ‘Aspects of Cantonese Grammar’, in C.-T. James Huang, Y.-H. Audrey Li, and Andrew Simpson (eds) Handbook of Chinese Linguistics, Hoboken, New Jersey: Wiley-Blackwell, 601–28. Ting, Jen (1995) ‘A Non-uniform Analysis of the Passive Construction in Mandarin Chinese’, Doctoral Dissertation, University of Rochester. Ting, Jen (1998) ‘Deriving the Bei-construction in Mandarin Chinese’, Journal of East Asian Linguistics 7(4): 319–54. Wu, Yicheng and Adams Bodomo (2009) ‘Classifiers ≠ Determiners’, Linguistic Inquiry 40(3): 487–503. Yiu, Yuk-Man Carine (2005) ‘Spatial Extension: Directional Verbs in Cantonese’, Doctoral Dissertation, The Hong Kong University of Science and Technology. Yuan, Jiahua, et al. (eds) (1960) Hanyu Fangyan Gaiyao (An Outline of the Chinese Dialects), Beijing: Wenzi Gaige Chubanshe. Yuan, Jiahua, et al. (eds) (2001) Hanyu Fangyan Gaiyao (An Outline of the Chinese Dialects), 2nd ed., Beijing: Yuwen Chubanshe. Yue-Hashimoto, Anne (1976) ‘South Chinese Dialects – the Tai Connection’, Computational Analyses of Asian and African Languages 6: 1–9. Yue-Hashimoto, Anne (1991) ‘The Yue Dialects’, in William Shi-Yuan Wang (ed) Languages and Dialects of China (Journal of Chinese Linguistics Monograph 3), Berkeley: Project on Linguistic Analysis, 237–93. Zee, Yun-Yang Eric (1999) ‘Chinese (Hong Kong Cantonese)’, in International Phonetic Association (ed) Handbook of the International Phonetic Association, Cambridge: Cambridge University Press, 58–60. Zhan, Bohui (ed) (1991) Hanyu Fangyan ji Fangyan Diaocha (Chinese Dialects and the Dialect Investigation), Wuhan: Hubei Jiaoyu Chubanshe.

33

Cheng Siu-Pong and Tang Sze-Wing Zhan, Bohui (ed) (2002) Guangdong Yue Fangyan Gaiyao (An Outline of Yue Dialects in Guangdong), Guangzhou: Jinan University Press. Zhang, Min (2011) ‘Hanyu fangyan shuangjiwu jiegou nan-bei chayi de chengyin: leixingxue yanjiu yinfa de xin wenti’ (‘Revisiting the Alignment Typology of Ditransitive Constructions in Chinese Dialects’), Bulletin of Chinese Linguistics 4(2): 87–270. Zhu, Dexi (1961) ‘Shuo de’ (‘On De’), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 12: 1–15. Zhu, Dexi (1980) ‘Beijinghua, Guangzhouhua, Wenshuihua he Fuzhouhua li de de zi’ (‘De in the Beijing, Guangzhou, Wenshui, and Fuzhou Dialects’), Fangyan (Dialects) 3: 160–5.

34

3 Cantonese Romanizaton Cheng Siu-Pong and Tang Sze-Wing The Chinese University of Hong Kong, Hong Kong, China

The diversity of Cantonese Romanization is a reflection of different backgrounds and purposes in rendering Cantonese into Roman letters. Some forms of Romanization are marked for their historical significance and influence on later proposals; others are noteworthy for their current popularity, as they are often found in books and other materials. Individual frameworks are discussed in detail, and the concluding tables summarize these discussions. It follows that such diversity compensates for the lack of a predominant form of Cantonese Romanization.

1. Introduction Cantonese Romanization methods are as varied as they are numerous. Up to the end of the last century, over 30 varieties have existed (Wu 1997; Lam 2009). None of them, however, is in a predominant position as pinyin is to Mandarin Chinese. Lack of government-sponsored standardization programs is part of the reason. The general public in Cantonese-speaking regions also cannot reach a consensus on how a Cantonese word should be Romanized. Unlike Mandarin or Vietnamese, the use of Chinese characters to write Cantonese has not been subjected to challenge by the proponents of the Roman alphabet, and no serious attempt has ever been made to change the written form of Cantonese into an alphabetical system. The main function of Cantonese Romanization is to transcribe Cantonese for educational and research purposes. Another major function is to transliterate Cantonese terms into English or other European languages. Most Romanization systems aim at the former function, and different schemes have been introduced to cater for different target recipients. The following section first describes the sound system of Cantonese. After that, individual Romanization frameworks will be discussed. If looking for a shortcut, you can move to the Romanization tables in the concluding section.

2.  Cantonese phonology There are 19 consonants in modern-day standard Hong Kong Cantonese. These include plosives /p, ph, t, th, k, kh, kw, kwh/, nasals /m, n, ŋ/, affricates /ts, tsh/, fricatives /f, s, h/, and ‘semi-vowels’ as well as other approximants /j, w, l/. These sounds are considered phonemes 35

Cheng Siu-Pong and Tang Sze-Wing

because they change the meaning of a word when one consonant is replaced by another. Take the (un)aspirated bilabial stops as examples. Unlike English, aspirated /ph/ (like p in pie) and unaspirated /p/ (like p in spy) are two phonemes, and a minimal pair between them can be found, e.g. /pha55/ ‘crouch’ vs. /pa55/ ‘daddy’. Other plosives are the same and as a result there are two sets of plosives in Cantonese, one aspirated and another unaspirated. /ts, tsh/ may be realized as [tʃ, tʃ h], but they are either allophones (when the following vowel is rounded, especially when it is /y/) or simply free variants of /ts, tsh/, depending on the speaker’s accent. All of these consonants are syllable-initial. Nasals are also syllable-final, and syllable-final plosives /p, t, k/ are unreleased, expressed as [p2, t2, k2]. The 19 consonants reflect the standard pronunciation. Many native speakers do not articulate /kw, kwh, n, ŋ/ syllable-initially and /k, ŋ/ syllable-finally (except in rhymes /ek, ok, eŋ, oŋ/; see Bauer and Benedict 1997). This is of particular interest as far as Romanization is concerned, because many Cantonese dictionaries and textbooks consider the drop or replacement of these sounds to be non-standard or even ‘sloppy’ and retain the standard way of pronunciation.1 Vowels in Cantonese include both front and back, and high and low monothongs /i, e, y, u, o, ɛ, ɔ, œ, ɵ, ɐ, a/. Depending on which analysis is adopted, the use of the vowel length mark may be optional or conditional,2 and even the total number of vowels can be reduced as some of them occur in complementary phonetic contexts. Whether or not they are deemed phonemes depends on how these sounds are perceived. For example, /e/ may lose its phonemic status and be argued as an allophone of /i/, since they are not only similar but also complementary: /e/ before /k, ŋ/ and /i/ elsewhere. This is the same for the sound pair of /o, u/. In the standard, conservative accent, the sound pair of /œ, ɵ/ follows a similar distribution, although /œ/ can stand alone without a final consonant and monothong /ɵ/ is always followed by /t/ or /n/. This part being introductory, the academic debate associated is set aside. What concerns us is the fact that many Romanization schemes use a single letter for both sounds of the pair. In addition to these monothongs, there are also ten gliding vowels in Cantonese: /iw, ɵʮ, uʮ, ej, ɔj, ow, ɐj, ɐw, aj, aw/. In some analyses, the final glides are rendered as a vowel symbol. Notice that in the same syllable, a gliding vowel (or diphthong) is not followed by a consonant. Only a monothong precedes a consonant coda. For this reason, the final glide in a gliding vowel is often considered as a coda as well. The vowels above are all phonemic, with the apparent ‘exceptions’ of /ɵʮ, uʮ/. Minimal pairs can be found only when they are preceded by /k/, and also /kh/ if tones are disregarded. For the sound pairs of /ɐj, aj/ and /ɐw, aw/, they are clearly distinctive, and it is easy to find a minimal pair to differentiate the two members within. But given that they are similar in sound quality, these distinctive sounds, together with /ɵʮ, uʮ/, are sometimes represented by a single grapheme when they are rendered in Romanization. This will be further addressed in the following sections. In Cantonese, a syllable always contains a vowel, with the exception of two syllabic nasals /m, ŋ/. They can stand alone without a vowel to follow, in which case they are rendered as [m̩, ŋ̍]. Cantonese is a tonal language. There are six or nine tones in Cantonese, depending on one’s approach to categorize them. Traditionally, tones in Chinese are classified into Flat, Rising, Departing, and Entering. In Cantonese, the first three groups contain two tones (one Dark and one Light), and together with three Entering tones (Upper Dark, Lower Dark, Light), they amount to a total of nine tones. For the non-Entering tones, Cantonese has Dark Flat tone /55/, Light Flat tone /21/, Dark Rising tone /25/, Light Rising tone /23/, Dark Departing tone /33/, and Light Departing tone /22/.3 For the Dark Flat tone, there is also a variant which is high-falling /53/. The two Dark Flat tones are either not distinctive (Cheung 1986), or in fact, the falling variant might have been fading out altogether (Matthews and 36

Cantonese Romanizaton

Yip 1994). Therefore, the maximum number of tones in Cantonese is still said to be nine. For the Entering tones, they specifically refer to the tones of those syllables ended with an unreleased stop [p2, t2, k2], and the Upper Dark, Lower Dark, and Light Entering tones are given as /5/, /3/, and /2/, which respectively correspond to the Dark Flat, Dark Departing, and Light Departing tones. Given this, these Entering tones are often not viewed as separate tones, and as such, Cantonese is said to have six tones only. The Entering tones are only viewed as the short or extra-short variants of the other tones (Zee 1999). Different Romanization schemes variably adhere to the six-tone and nine-tone categorization, and this gives rise to a noticeable difference in how a tone is transcribed. It has also been observed that some native speakers merge some of these tones, especially the two Rising tones /25/ and /23/ (Bauer et al. 2003). This is again seldom reflected in dictionaries or textbooks with words transcribed in Cantonese Romanization. The term Cantonese itself, when used in the Hong Kong variety, is literally translated as ‘Yue language’ or ‘Guangdong speech’. Regardless of tone merger or any pronunciation deemed ‘non-standard’, the phonetic notation for the two words is given as /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ respectively.

3.  Various Romanization systems 3.1.  Earlier Romanizations The publication of Cantonese dictionaries in the early nineteenth century, intended for the Western readers, marked the beginning of a time when substantial work was done to Romanize Cantonese. Robert Morrison’s A Vocabulary of the Canton Dialect in 1828 was one of these earliest attempts to offer transcription for Cantonese words in Roman letters. Like many other dictionaries and glossaries that followed, every effort was made to ensure that the Romanization given resembled the native language of the target readers. For example, /i/ and /u/ are transcribed as ee and oo, written in a way in conformity with the English orthography. Letter u is used for /ɐ/, which only occurs in closed syllables. For instance, /sɐm/ ‘heart’ is rendered as sum. At this stage, there was no one-to-one mapping between letters and sounds. For the same letter u, when it is used in open syllables, it refers to /y/. For example, /tsy/ ‘pig’ is given as chu. But when /y/ occurs in a closed syllable, it is rendered as ue, as in uet /jyt/ ‘moon’. Besides, Morrison’s Romanization omits the indication of tones. Therefore, even though /thin55/ ‘sky’ and /thin21/ ‘field’ are in different tones, they are all Romanized in the same way as teen. Aspiration is not addressed either. For instance, unaspirated /t/ is indistinguishable from aspirated /th/, so teen may be used for /thin/ as well as /tin/, as in /tin22/ ‘lightning’. Readers cannot know whether it refers to sky or field or lightning simply through this Romanization. Several Cantonese dictionaries and textbooks emerged after Morrison’s work. These include E. C. Bridgman’s Chinese Chrestomathy in the Canton Dialect (1841), S. W. Williams’ A Tonic Dictionary of the Chinese Language in the Canton Dialect (1856), J. Chalmers’ An English and Cantonese Pocket-dictionary (1859), E. J. Eitel’s A Chinese Dictionary in the Cantonese Dialect (1877), and J. D. Ball’s Cantonese Made Easy (1883). Compared with Morrison’s Romanization, correspondence between letters and phonemes was improved in later schemes. In Bridgman’s dictionary, for example, distinction was made between aspirated and unaspirated stops and affricates by using inverted apostrophes. Therefore, /t/ and /th/ are represented by t and t‘, and the Romanized forms for /thin/ ‘sky’ and /tin/ ‘lightning’ came to be distinctive. Another feature is the use of semicircular marks to represent tones. This is not an invention by Western lexicographers, but adoption of the Chinese traditions. The Flat, 37

Cheng Siu-Pong and Tang Sze-Wing

Rising, Departing, and Entering tones are indicated by a semicircular mark placed at four different corners: lower left for Flat, upper left for Rising, upper right for Departing, and lower right for Entering. The Dark and Light tones are differentiated by marking the Light tones with a line underneath the semicircle. For instance, /thin55/ ‘sky’ is rendered as ˓t‘ín, whereas for /thin21/ ‘field’, the tone mark is further underlined. Ball’s Cantonese Made Easy followed the same methods to mark aspiration and tones, and by the same token also achieved a higher degree of phonemic distinction. B. F. Meyer and T. F. Wempe’s The Student’s Cantonese–English Dictionary (1934) also has a Romanization system similar to Eitel’s and Ball’s. Comparatively speaking, Y. R. Chao’s Cantonese Primer (1947) was a more recent work, yet his Romanization scheme is worth mentioning for historical purposes. Chao’s system is fully alphabetical in that tones are also represented by letters. Different sets of initial consonant letters are used for Dark and Light tones and, for each set, the representation of the same rhyme changes with different tones. To illustrate, the Dark /pa55/, /pa25/ and /pa33/ are represented as pa, pax, and pah, and their Light counterparts are written as ba, bax, and bah. Under Chao’s system, /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ are expressed as Yutyux and Kwoagtong-wah* respectively. Notice that Chao’s system is particular about whether the tone is the base tone or changed from another one. Since /wa/ here is originally in the Light Departing tone /22/, it is so expressed in its original tone as wah, rather than as uax, with an asterisk added to mark its change to a rising tone. Chao’s Romanization is historically significant as it is arguably the most complex scheme ever proposed for Cantonese Romanization. It has fallen into oblivion though.

3.2.  Government Romanization In Hong Kong, the Government Romanization is the most recognizable means to Romanize Cantonese. It is mainly used for local place names and personal names. To many Cantonese speakers in Hong Kong, this is the only Cantonese Romanization they can understand, and perhaps provide if they are asked to. As previously suggested, it would be dubious to call it a ‘system’ (Kataoka and Lee 2008); it is rather a hodgepodge of various ways of Romanization (Kataoka 2014). For obvious reasons, this is not a Romanization scheme – there is a low degree of correspondence between letters and sounds. The same letter is often used for different sounds. This is likened to the English orthography, in which letter c represents /s/ in cede, but /k/ in candy. In this form of Romanization, the four pairs of Cantonese plosives /p, ph/, /t, th/, /k, kh/, /kw, kwh/ are rendered the same way as p, t, k, and kw, regardless of whether they are aspirated. Given that these sounds are phonemes, distinctive words such as /tin55/ ‘insane’ and /thin55/ ‘sky’ would become indistinguishable when they are Romanized as tin. The same also applies to affricates /ts, tsh/, but there is one more complication: while the two consonants are Romanized the same way, they are variably rendered as ts or ch. Thus, although the first character for the place names of Chai Wan /tshaj21 wan55/ and Tsuen Wan /tshyn21 wan55/ share the same initial consonant, they are written in different initial letters.4 Nasals /m, n, ŋ/, fricatives /f, s, h/, and approximants /w, l/ are straightforwardly Romanized as m, n, ng, f, s, h, w, l. As in English, y is used for /j/. The Hong Kong Government Romanization is a defective script, and in fact it is defective in both directions – devoid of one-to-one correspondence from grapheme to phoneme, as well as from phoneme to grapheme. This resembles the English orthography more than, for example, the French orthography. The one-to-many mapping from letters to sounds is illustrated in the case of (un)aspirated consonants. And phoneme-to-grapheme inconsistency 38

Cantonese Romanizaton

is shown in the affricates and more often in the case of some vowels. Take /i/ as an example. It can be Romanized differently in the place names, as i for yi in Tsing Yi, as ee for see in Sai See Street, as z for tsz in Tsz Wan Shan, and as ze for sze in Sze Tei Shan. Personal names transliterated from Cantonese show the same inconsistency. /tsy55/, a common surname, is rendered as Chu, but /tsyn/ would likely become Chuen or Tsuen when transliterated. On the other hand, Chun (as in the name of Chief Executive Leung Chun-Ying) is reserved for /ts(h)ɐn/. Added to this is the fact that not all letter combinations are legitimate, even though they may represent a legitimate syllable. This is again likened to the English orthography, in which both letters s and c can refer to /s/, but only see, rather than cee, is accepted as a valid written unit. Therefore, /tsɐn/ is expressed as Chun, but Yun for /jɐn/ is unlikely. The adding up of these instances of inconsistency implies that many who know this Romanization have had to learn it by rote, no matter whether they are native Cantonese speakers. Only generalizations can be made here: (i) correspondence between monothongs and single vowel letters: /a, ɐ/ as a; /ɛ/ as e; /i, e/ as i; /ɔ/ as o; /y, u, o, ɵ/ as u; and (ii) correspondence between gliding vowels and double vowel letters: /iw/ as iu; /ɵʮ, uʮ/ as ui; /ej/ as ei; /ɔj/ as oi; /ow/ as ou; /ɐj, aj/ as ai; /ɐw, aw/ as au. These generalizations are far from comprehensive since a monothong can also be (and is often) expressed in more than one letter, and that /ow/ is more commonly expressed in a single letter, i.e. o, not to mention the use of consonant letters (plus y) and other irregularities in vowel representation (see Cheng 2014). The Government Romanization is not meant to be a teaching tool for Cantonese learners. Its purpose is to translate Cantonese names in the form of transliteration. Therefore, it almost never goes with the tonal indication. Unlike other ways of Romanization, tone is out of concern under most circumstances. Besides, not all place and personal names in Hong Kong are Romanized from Cantonese. Some place names in English are literally translated, e.g. Green Island for /tsheŋ55 tsɐw55/, or they are named with no regard to how they are called in Cantonese, e.g. Aberdeen for /hœŋ55 kɔŋ25 tsɐj25/. Furthermore, not all place names are transliterated from standard Hong Kong Cantonese. The most notable example is the name Hong Kong itself. Hong is based on the pronunciation of the boat people, or Tankas, during the time when the British arrived at the territory in the nineteenth century. If transliterated from the standard variety, it will most likely be rendered as Heung or Heong. Personal names are not always transliterated from Cantonese either. This is especially true for those not of Cantonese descent. Surnames vary even greater, since a native Cantonese speaker might have inherited a non-Cantonese family name. Language change is another complicating factor. A place name was Romanized and subsequent sound change made it deviant from the current pronunciation. All these mean that it is easy to draw erroneous generalizations when one attempts to learn this Romanization through a list of local place names or personal names. Though unruly, the Government Romanization is very often the only Romanization known to the general populace, not only used in the administration to transliterate proper names, but also used by the general public for various purposes. For instance, the Cantonese ‘chat alphabet’ used in informal communication such as instant messaging is loosely based on the government’s way of Romanizing Cantonese. When /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ ‘Cantonese’ are Romanized, if needed, they are given as Yuet Yu(e) and Kwong Tung Wa(h).5

3.3.  Yale Romanization The origin of Yale Romanization for Cantonese is commonly attributed to Parker Po-Fei Huang and Gerald Kok for their collaborated textbook Speak Cantonese. First published in 39

Cheng Siu-Pong and Tang Sze-Wing

1958, it has been reprinted several times over a period of four decades. In its sequel Book II, Huang states that the Romanization system was developed by Kok. Based on the year of its first publication, it is reasonable to assert that the scheme came into being no later than the late fifties. Huang’s later work Cantonese Dictionary: Cantonese–English, English– Cantonese also adopted this system. Cantonese is one of the languages Romanized under a system created at Yale University. Other Yale Romanizations involve East Asian languages such as Mandarin, Korean, and Japanese. Since its introduction, the Yale system has been extensively used in English-language textbooks, dictionaries, and glossaries made for Cantonese learners. These include some of the most popular titles currently in circulation. Comparatively speaking, the use of the Yale system is less prevalent in Chinese-language publications. While the native speakers sometimes do need to know how to pronounce a particular Chinese character, they are often referred to another homophonous character. Even when Romanization is consulted (not very often), the Yale system is not the one needed. So far, it is chiefly used to study Cantonese as a foreign language, rather than offer a pronunciation guide for the native speakers. Thus, Cantonese learners using this system might encounter a situation where they know this Romanization much better than the native speakers.6 Under the Yale system, the 19 consonants are expressed in a way that makes a distinction between aspirated and unaspirated consonants. The plosives /p, ph, t, th, k, kh, kw, kwh/ are respectively rendered as b, p, d, t, g, k, gw, and kw. This contrasts with the Government Romanization. For example, in the Government version, p represents both /p/ and /ph/, but Yale Romanization uses two separate letters b and p. In this case, the Yale system has oneto-one mapping between letters and phonemes. To achieve this goal, the Yale system uses letters traditionally reserved for voiced consonants to represent the voiceless ones. Affricates /ts, tsh/ are given as j and ch. Similarly this distinction is not observed in the Government Romanization, and they are both expressed as ch or ts. Nasals /m, n, ŋ/, fricatives /f, s, h/, and /j, w, l/ are straightforwardly rendered as m, n, ng, f, s, h, y, w, l. With /j/ Romanized as y, the Yale system resembles the English orthography. Phoneme-to-grapheme consistency is also achieved in the vowel representation, although a single letter may in separate situations be used to represent distinct sounds. Letter a, for example, is used for both /a/ and /ɐ/. Yet /a/ is expressed in single letter a only when it stands alone in the rhyme. Gliding vowels with /a/ are given in double letters aa, and aa is also used when /a/ precedes a consonant coda. Since /ɐ/ (always given as a) never occurs syllable-finally, a distinction can still be made between the two vowels. Letter i is used for /i, e/, which is common to most Romanization schemes as they are in complementary distribution. The same applies to u, which represents both monothongs /u/ and /o/. It should be noted, however, that u is also used in yu as a grapheme representing /y/. The y in yu does not represent the semi-vowel (approximant), except in the case where the initial consonant is /j/. To illustrate, /sy/ is expressed as syu, whereas /jy/ is given as yu, instead of the incorrect yyu. The ‘shorthand’ rule takes advantage of the fact that /ju/ does not exist as an indigenous syllable in Cantonese, but problems may arise when loanwords are involved. /ju/, for example, can mean ‘university’ in daily speech. Apparently, the Yale system does not assign a separate Romanized form for this syllable. Letter u is also used for other vowels when it is associated with e. Letter e as a single letter represents /ɛ/, but when combined with u to form eu, it refers to both /œ/ and /ɵ/: /lœk/ is given as leuk and /lɵt/ is given as leut. Letter o is used for monothong /ɔ/. The same principles are generally extended to the gliding vowels, so /iw, ej, ɔj, ɐj, ɐw, aj, aw/ are expressed as iu, ei, oi, ai, au, aai, aau, with u representing the rounded glide and i for the unrounded one. However, /ɵʮ, uʮ/ 40

Cantonese Romanizaton

are given as eui and ui even though /ʮ/ is considered rounded. /ow/ is given as ou, although monothong /o/ is Romanized as letter u. Thus, /sow/ is expressed as sou, but /sok/ is expressed as suk. The tonal representation is what gives the Yale system a distinctive hue, as it makes use of both diacritics and letter h. It adheres to the six-tone categorization. In its original version, a distinction was made between the high and high-falling tones of the Dark Flat tones, so that /a55, a53/ are expressed as ā and à, with the vowel differently marked by macron and grave accents. The Dark Rising and Departing vowels are respectively marked by acute and no accents. Thus, /a25, a33/ are given as á and a. With considerations of the sound change factor (discussed above), subsequent variants may use a macron accent for all Dark Flat tones, whereas a few publications nowadays still follow the original design and use a grave accent. The Light counterparts are marked with the same diacritic mark (with Light Flat tone corresponding to the high-falling Dark Flat tone), with an h additionally attached to the vowel. Therefore, /a21, a23, a22/ are given as àh, áh and ah. As it follows the six-tone classification, a separate representation for Entering tones is unnecessary, and the three Entering tones are treated in the same way as the non-Entering ones: /at5, at3, at2/ are rendered as àat, aat and aaht. The tone letter h follows the vowel but not the whole rhyme; while /sɵʮ21/ is expressed as sèuih, /sɵn21/ is expressed as sèuhn. Taken all together, when /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ ‘Cantonese’ are Romanized in the Yale system, they become yuht yúh and gwóng dūng wá. Regarding the tone letter h, an additional rule was stipulated in the original system, which states that when a syllable begins with m, n or ng, the h will not be used as an indicator. This is due to the fact that these syllables are seldom expressed in the Dark tones. /ŋɔ23/ ‘me’, for example, should be rendered as ngó rather than as ngóh. Most later applications disregard this rule, with the possible exceptions of the two syllabic nasals. The peculiarity as a result of using tone letter h causes some later versions to discard it altogether,7 and replace the diacritics with tone number: /55/ (or /53/, /5/), /25/, /33/ (or /3/), /21/, /23/, and /22/ (or /2/) are numbered from 1 to 6. In this ‘simplified’ version, yuht yúh and gwóng dūng wá are given as yut6 yu5 and gwong2 dung1 wa2.

3.4.  LSHK Romanization The LSHK Romanization Scheme, also known as Jyutping, was proposed by the Linguistic Society of Hong Kong (LSHK) in 1993. The LSHK is a non-governmental organization comprising linguistic academics and professionals based in Hong Kong. This Romanization scheme was collectively designed by scholars specializing in various subfields of language studies. A conference had been held earlier in 1992 to gather thoughts on the future scheme before the finalized version was released at the Fourth International Conference on the Yue Dialects in December 1993. The scheme aims to be multipurpose, systematic, inclusive, and user-friendly and takes care of both the earlier traditions and word-processing technology of the time (Cheung 1994). Since its introduction, the LSHK Romanization has gained a strong foothold in the academia. Up to this day, a majority of English and Chinese-language academic papers published in Hong Kong and also in the West adopt this system for Romanization, apart from the direct use of phonetic symbols. The Government of Hong Kong does not have a fixed position on which scheme should be the default option. Having said that, the post-handover Education Bureau, in collaboration with the tertiary scholars, has published materials that use this system as the pronunciation guide. They abandoned the earlier scheme (‘ILE Romanization’; to be discussed) used for another guidebook published by the Education Department in the early nineties. As the scheme was invented by tertiary 41

Cheng Siu-Pong and Tang Sze-Wing

scholars, it is taught in local university courses that touch upon Cantonese pronunciation. Besides, its promotion has occasionally been extended to primary and secondary schools. But still most Hong Kong people, including university students, have never been required to learn Cantonese sound notation at school. Also due to its fairly recent date of emergence, the LSHK Romanization is still on its way to reaching a wider spectrum of people outside the academic community. The representation of consonants has given this system some noticeably distinctive features. Same as the Yale Romanization, plosives /p, ph, t, th, k, kh, kw, kwh/, nasals /m, n, ŋ/, fricatives /f, s, h/, and approximants /w, l/ are conveniently given as b, p, d, t, g, k, gw, kw, m, n, ng, f, s, h, w, l. What makes the LSHK system spectacular is how it renders affricates /ts, tsh/, which are given as z and c. It conforms to Mandarin pinyin, but deviates from the Government traditions in which single graphemes z and c are never used to represent consonants. In the Government Romanization, c is only used in ch for /ts, tsh/, and z is never used for consonants and only used in few syllables like tsz /ts(h)i/ or sze /si/. Another deviation from the commonplace Romanization is on how it expresses /j/. Letter j is used for /j/ in the LSHK system, but in Hong Kong’s common practice, it is instead used for /ts/, as is often found in the name of local companies. Note that these peculiarities were made ‘on purpose’, so that it takes care of both the pinyin (for z and c) and IPA traditions (for j), and in so doing ‘enhances user-friendliness’ (Cheung 1994). Regarding the Romanization of vowels, the LSHK system is arguably the most precise one in representing the distinctive sounds. One-to-one mapping from phoneme to grapheme is achieved in most vowels. And each rhyme corresponds to exactly one written unit. Monothongs /i, e, y, u, o, ɛ, ɔ, œ, ɵ, ɐ, a/ are respectively rendered as i, i, yu, u, u, e, o, oe, eo, a, and aa. /i, e/ and /u, o/ are expressed in one letter only (i and u) given that these sounds are in complementary distribution. Except them, each monothong matches one grapheme. To achieve phonemic distinction, the LSHK scheme resorts to the use of four digraphs yu, oe, eo, and aa to represent monothongs, in which yu is somewhat conspicuous as letter y is more often understood in English as a consonant letter. Unlike the Yale system, however, no ‘shorthand’ rule is added ( yyu becomes yu for /jy/); /jy/ is directly expressed as jyu. Digraphs are specially designed so that only the basic Roman alphabet (26 letters) is used. This is of particular significance during the time when keying in a diacritic was a hassle. As for /ɐ, a/, the distinction made by using aa and a is an inheritance of the earlier traditions. Another important feature is that it distinguishes /œ, ɵ/. They have similar sound quality and have traditionally been expressed in only one grapheme (in the Yale system, for example, they are Romanized as eu). In so doing, it manages to render some ‘non-indigenous’ rhymes and syllables arising as a result of the recent sound change. In standard, conservative accent, /œ, ɵ/ are complementary, but recent merger of codas /t, k/ and /n, ŋ/ into /t/ and /n/ has created new rhymes such as /œn, œt/. Thus, in this new variety, /œ, ɵ/ are contrastive and can be differentiated with minimal pairs. Assigning two graphemes oe and eo separately for /œ/ and /ɵ/ allows the system to become ‘inclusive’ by capturing this recent contrastive distinction. The same conversion rules are extended to gliding vowels: /iw, ɵʮ, uʮ, ɔj, ɐj, ɐw, aj, aw/ are given as iu, eoi, ui, oi, ai, au, aai, aau, with letter u representing /w/ and i representing /j/ or /ʮ/. Since the latter two glides are not contrastive, they are expressed in one single letter – a common practice in Cantonese Romanization. /ej, ow/ are treated as if the vowel were less high: monothongs /e, o/ are given as i and u, but as an element in the gliding vowels, they are rendered as e and o, and /ej, ow/ are given as ei and ou. The paradigm of the LSHK system allows the inclusion of new rhymes or syllables traditionally non-existent but recently created due to foreign-language influences. These include letter 42

Cantonese Romanizaton

pronunciation such as /ɛw55/ ‘L’ (eu1), /ɛm55/ ‘M’ (em1) and /ɛn55/ ‘N’ (en1), as well as transliterated loanwords, such as /ɛp5/ ‘app’ (ep1). Many other systems, on the other hand, do not specify how, or if at all, these syllables should be Romanized. For example, the use of eu for /œ, ɵ/ in the Yale system implies that /ɛw/ does not exist; otherwise, the two vowels will share the same grapheme eu, which is implausible. Under the LSHK scheme, all these rhymes are regarded as existent so that it achieves ‘phonological inclusiveness’ (Cheung 1994). The scheme is systematic and highly inclusive, but it also risks producing some Romanized forms unintelligible to people who are only familiar with the Government Romanization. Examples of this include jeon (/jɵn/) and zyu (/tsy/). They are commonly Romanized as yun and chu (tsu), quite dissimilar from what they are given in the LSHK format. The LSHK system adheres to the six-tone categorization, meaning that there is no separate notation required for the Entering tones. Tone number is used to mark the tones: number 1 for /55/ (or /5/), 2 for /25/, 3 for /33/ (or /3/), 4 for /21/, 5 for /23/, and 6 for /22/ (or /2/). The LSHK working group decided that tone number 1 does not necessarily refer to /55/ (Cheung 1994). If tone /53/ exists in anyone’s speech, it is still marked by the same number. The use of numbering is desirable because numerals are in the basic ASCII character set. Since in computing Chinese characters are coded with at least one ‘non-basic’ ASCII character, like those with diacritic marks, inputting diacritics for tones might mistakenly produce Chinese characters in the earlier versions of Chinese word-processing software. The ordering also reflects a more common practice at the time. An alternative ordering puts the Flat-Rising-Departing-Entering sequence in the first order and the Dark–Light distinction in the second. To write the tone number, some earlier schemes stipulate that it be written in superscript style. This is not required in the LSHK scheme. Taken all together, when /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ ‘Cantonese’ are Romanized in the LSHK system, they become jyut6 jyu5 and gwong2 dung1 waa2. Often, tonal indication is omitted in translated terms, such as the term Jyutping itself.8 If tone number is omitted, jyut6 jyu5 and gwong2 dung1 waa2 will be expressed as Jyutjyu and Gwongdungwaa. It is assumed that the same capitalization rules of the target language apply to those terms transliterated in LSHK Romanization.

3.5.  Rao’s Romanization It refers to the Romanization scheme used in the dictionaries of which Rao Bingcai is one of the editors. The official name of the scheme is called the Scheme of the Cantonese Phonetic Alphabet. The system attributed to Rao is a modified version of the said scheme, which was officially announced in 1960 by the executive branch for education of Guangdong province. The original scheme was part of a larger project encompassing four different Romanization schemes for Guangzhou dialect (Cantonese), Kejia dialect (Hakka), Hainan dialect, and Chaozhou dialect (Swatow). In theory, this is a government standard for Cantonese Romanization. However, the official scheme has yet to gain widespread adoption. It should be best understood as a government invention without much standardization effort to follow it up. Rao’s modified version, on the other hand, has become predominant in mainland China, with its use extended from his own publications to a variety of Cantonese textbooks and dictionaries written by others. It has largely replaced the original scheme, which has since then been only occasionally referred to in the footnotes of the pronunciation guide. The changes made in the modified scheme include the replacement of some vowel graphemes with another one and also the addition of two syllabic nasals.9 They will be highlighted below. 43

Cheng Siu-Pong and Tang Sze-Wing

Under this scheme, a phoneme may be represented by more than one grapheme, depending on the phonetic and orthographic conditions. When used syllable-initially as an onset, plosives /p, ph, t, th, k, kh/ are given as b, p, d, t, g, k. This agrees with their representation in most other systems. However, when /p, t, k/ are used syllable-finally as a coda, expressed as unreleased [p2, t2, k2], they are given as b, d, g. This differs from all major forms of Romanization, including Government, Yale, and LSHK, in which they are rendered as p, t, k. Thus, /phit/ is expressed as pid in this scheme, but is expressed as pit in others. As for the other two plosives /kw, kwh/, they are given as gu and ku. Letter u, instead of w as in some other Romanizations, is used to represent the labialization of velar stops. The original scheme clarifies this by stating that gu and ku are syllables when they stand alone, i.e. in the case of /ku, khu/, whereas they are initials when they are not used independently in the phonetic notation; as in the case of /kwa, kwha/, they are given as gua and kua. Nasals /m, n, ŋ/, fricatives /f, h/, and approximants /w, l/ are as usual Romanized as m, n, ng, f, h, w, l, and /j/ is given as y. The most complex way of conversion occurs to /ts, tsh, s/. They are variably expressed as z, c, s or j, q, x depending on the following vowel letter: as j, q, x when they precede vowel letters i or ü (shorthanded as u; to be discussed), and as z, c, s elsewhere. To illustrate, /si/ is given as xi, while /sa/ is expressed as sa. Some previous literature suggests that it indicates a distinction between /ts, tsh, s/ (as z, c, s) and /tʃ, tʃ h, ʃ/ (as j, q, x). This is more apparent than real. First, whether the two sets of affricates are in complementary distribution is debatable. Even if they are so, the post-alveolar /tʃ, tʃ h, ʃ/ usually precede round vowels, but j, q, x apply to both rounded (/y/) and unrounded (/i, e/) monothongs, so do z, c, s. Besides, monothong /e/ and gliding vowel /ej/ are differently expressed as i and éi, and different sets of letters are used for them. More importantly, the scheme explicates that there is no distinction between these two sets of consonants letters. This is totally out of orthographic concerns and is irrelevant to any ‘natural classes’ that show distinctive phonetic features. The orthographic considerations are in turn taken to resemble the Mandarin pinyin system. Such an arrangement is more pedagogically relevant than phonetically necessary. As far as the consonants are concerned, Rao’s modification is equivalent to the original scheme, except it additionally specifies that the two nasals can also be syllabic. In Rao’s Romanization, monothongs /i, e, y, u, o, ɛ, ɔ, œ, ɵ, ɐ, a/ are respectively rendered as i, i, ü, u, u, é, o, ê, ê, e, a. Complementary /i, e/ and /u, o/ are expressed in one letter, i.e. i and u. /œ, ɵ/ are both expressed as ê, a plausible arrangement for the standard accent (see the description above). Note that diacritics, instead of digraphs, are employed to render these sounds. Therefore, a monothong is always expressed by a single letter. Besides, the use of ü for /y/ allows it to conform to the pinyin system. Meanwhile, the pinyin orthographic rules are transferred to the Cantonese system: ü is written as yu for yu /jy/, yun /jyn/, and yud /jyt/; the two dots are also removed after j, q, x, e.g. ju /tsy/, qu /tshy/, xu /sy/. These are essentially the pinyin rules despite that in Mandarin /j/ is deemed non-existent, and letter y is added but represents no consonant. The ‘shorthand’ rule is possible only because /u/ does not follow /ts, tsh, s, j/ in the indigenous Cantonese phonological system. Borrowed syllable /sut/ (for ‘suit’) might be expressed as sud, making use of two sets of homophonous letters, z, c, s vs. j, q, x, but /ju/ (for letter ‘U’ or ‘university’) cannot be represented without conflicting with the indigenous syllable /jy/. Conformity with the pinyin system is also exhibited in the case of /ɐ/. In other systems, the /ɐ, a/ distinction is often made by using single letter a and double letters aa. Here, e is used for /ɐ/ and a is reserved for /a/. The more basic e is reserved for /ɐ/, whereas é and ê are for /ɛ/ and /œ, ɵ/. This is possibly because rhymes like en /ɐn/ and eng /ɐŋ/ resemble Mandarin en /ən/ and eng /əŋ/, and syllables such as zé, cé, sé /tsɛ, tshɛ, sɛ/ do sound quite differently 44

Cantonese Romanizaton

from their near-homographs in Mandarin, i.e. ze, ce, se /tsɤ, tshɤ, sɤ/. Gliding vowels basically follow the same conversion pattern: the vowel element in /iw, ɵʮ, uʮ, ɔj, ɐj, ɐw, aj, aw/ follows the same rules and so they are converted into iu, êu, ui, oi, ei, eo, ai, ao respectively. Apart from this, /ej, ow/ are given as éi and ou, treated as if they were the gliding variants of /ɛ, ɔ/. In this regard, the Yale and LSHK systems also share the same treatment. Note that the glides in these vowels are variably represented. /w/ is given as u for iu /iw/ and ou /ow/, but as o for eo /ɐw/ and ao /aw/; /ʮ/ in êu /ɵʮ/ is expressed as u, but is expressed as i in ui /uʮ/. The description above is for Rao’s modified scheme. In the original scheme, the four rhymes /ɐw, ow, ɔŋ, oŋ/ are given as ou, ô, ông, ong, rather than as eo, ou, ong, ung. Phonetically speaking, these changes improve phoneme-to-grapheme consistency. Besides the obvious problem of using e for /ɐ/ but ou for /ɐw/, the use of ô also undermines the coherence of the scheme. Letter ô is the only non-digraph for gliding vowels. Even if the sound it represents is said to be monothongized as /o/, ông should better represent /oŋ/ than /ɔŋ/, not the other way around as it does in the original scheme. It appears that the original assignment of ong for /oŋ/ is deliberate because this rhyme is closer to Mandarin ong /ʊŋ/ than /ɔŋ/ is. Both the original and Rao’s modified schemes adhere to the six-tone categorization. As such, Dark Flat (Upper Dark Entering), Dark Rising, Dark Departing (Lower Dark Entering), Light Flat, Light Rising, and Light Departing (Light Entering) tones are numbered from 1 to 6 respectively. Cheung (2003) notes that the original 1960 scheme has established the now-common tonal notation traditions, in which Entering tones are no longer independent and the tonal order is first determined by the Dark–Light distinction rather than by the traditional tonal categories. Later systems like the LSHK scheme follow this ordering. Again, Tone 1 refers to either /55/ or /53/, a distinction prevalent in the older-day Guangzhou Cantonese for which the scheme was invented. Taken all together, /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ ‘Cantonese’ are Romanized in Rao’s system as yud 6 yu5 and guong2 dung1 wa2. Unlike the LSHK scheme, a superscript style for tone number is required.

3.6.  Other systems Other influential systems include those proposed by S. L. Wong, Sidney Lau, and the Institute of Language in Education. In some areas, S. L. Wong’s scheme is perhaps more influential than the Yale and LSHK ones. It originated from his book A Chinese Syllabary Pronounced according to the Dialect of Canton (1941) and forms the basis of the phonetic symbols used in a majority of the Standard Chinese dictionaries published in Hong Kong. Yet Wong’s system and its variants may not fit the strict definition of ‘Romanization’ because phonetic symbols, rather than ordinary Roman letters, are used for transcription. Most dictionaries simply term it ‘International Phonetic Alphabet’. It is often seen as a ‘broad’ phonetic transcription for Cantonese, and as a matter of fact, some conventionalized symbols should be understood in order to interpret them correctly. Most notably, it uses voiced consonant symbols for unaspirated plosives and affricates. Thus, unaspirated /p, t, k, kw, ts/ are rendered as /b, d, g, gw, dz/. Beginning learners with prior knowledge of IPA might risk viewing them as voiced consonants.10 Other features that make it a ‘broad’ transcription include the use of /œ/ for both /œ/ and /ɵ/, and /i, u, y/ in gliding vowels, so that /iw, ɵʮ, uʮ, ej, ɔj, ow, ɐj, ɐw, aj, aw/ are expressed as /iu, œy, ui, ei, ɔi, ou, ɐi, ɐu, ai, au/. Note that the vowel length mark is not used and /ʮ/ is variably expressed as /y/ or /i/, as in /œy, ui/. This system is sufficient for native speakers to look up the pronunciation of a Chinese character, but nonnative learners using it might enunciate wrongly if not knowing the associated phonetic details. 45

Cheng Siu-Pong and Tang Sze-Wing

The Institute of Language in Education (ILE) Scheme, on the other hand, is an exemplar of Cantonese Romanization. It was proposed by the said institute, which had then been a subsidiary of the colonial Education Department and was later incorporated into the Hong Kong Institute of Education (HKIEd). This scheme is a modification of the one proposed by P. C. Yu in his book Rhyming Glossary (1971). Same as the LSHK scheme, only 26 letters and Arabic numerals are used for Romanization. This system once acquired some degree of authority through the publication of List of Cantonese Pronunciation of Commonly-used Chinese Characters, printed by the Government Printer in 1990. Later publications by the administration, however, have changed to the LSHK system. A slightly modified version of the ILE scheme is used in The Dictionary of Standard Cantonese Pronunciation edited by Zhan Bohui (2002). In vowel representation, the ILE scheme renders /a/ as a in the open syllables and as aa in the closed ones (same as the Yale Romanization). Yu’s original scheme has the same treatment, except that the letter a for /ɐ/ is always italicized, whereas the one for /a/, as well as all other letters, is not. Zhan’s modification uses aa for /a/ in all phonetic contexts. Finally, Sidney Lau’s Romanization is associated with a series of Cantonese textbooks written by him and published by the Government Printer starting from 1965. Lau has spent years in the Hong Kong Government teaching Cantonese and was once the Principal of the Government Language School. One of the main characteristics of this system is that it bears some resemblance to the English orthography. In addition to the use of oo for /u/, it also resorts to the use of final h to close the syllable with /ɔ, œ/ when they are not followed by a final consonant or glide, i.e. goh for /kɔ/ but gong for /kɔŋ/. Lau stated that his scheme was an adaptation of the Meyer-Wempe system, which already included these features. The former two systems, Wong and ILE, adhere to the nine-tone categorization. The Upper Dark, Lower Dark, and Light Entering tones are numbered from 7 to 9, rather than being merged into the non-Entering ones for tonal notation. Sidney Lau’s Romanization, on the other hand, adheres to the six-tone categorization.

4. Conclusion In the practice of Cantonese Romanization, it is not uncommon for writers and academics to develop a new scheme or improve on existing ones – a reason for a substantial and growing number of systems. The Romanization forms discussed above are those that have historical significance and/or stimulated adoption by others who are not the inventors or definite proponents, or their associates. Tables 3.1–3.4 summarize the most popular schemes over the past half century.11 Together with the Government Romanization, these Romanization forms are employed in various occasions and reach different groups of target audience. For instance, to trans­ literate a Cantonese term for a Hong Kong-based English bulletin, without much regard to phonemic precision, the Government Romanization allows the largest number of locals to understand what the transliterated term means. English-speaking Cantonese learners are often exposed to the Yale Romanization and might prefer a transcription in this system. Meanwhile, Mandarin-speaking learners of Cantonese who have prior knowledge of pinyin might find Rao’s scheme most accommodating. In Hong Kong’s academia, the LSHK Romanization has been becoming the de facto standard and is expected to be used by the faculty and students alike in academic papers. With such a diversity of uses and users of many backgrounds, the apparent chaos of Cantonese Romanization may well be thought of as a ‘blessing in disguise’. 46

Cantonese Romanizaton Table 3.1  Initial consonants IPA

Wong

Yale

Rao

ILE

LSHK

Lau

p ph m f t th n l ts tsh s j k kh ŋ h kw kwh w

b p m f d t n l dz ts s j g k ŋ h gw kw w

b p m f d t n l j ch s y g k ng h gw kw w

b p m f d t n l z, j c, q s, x y g k ng h gu ku w

b p m f d t n l dz ts s j g k ng h gw kw w

b p m f d t n l z c s j g k ng h gw kw w

b p m f d t n l j ch s y g k ng h gw kw w

Note Wong: Variant of S.L. Wong’s Scheme Used in Hong Kong’s Dictionaries. Yale: The General Scheme of Yale Romanization. Rao: Rao Bingcai’s Modification of the Scheme of the Cantonese Phonetic Alphabet. ILE: Institute of Language in Education Romanization. LSHK: Linguistic Society of Hong Kong Romanization. Lau: The General Scheme of Sidney Lau’s Romanization.

Table 3.2  Final consonants IPA

Wong

Yale

Rao

ILE

LSHK

Lau

m n ŋ p t k

m n ŋ p t k

m n ng p t k

m n ng b d g

m n ng p t k

m n ng p t k

m n ng p t k

47

Cheng Siu-Pong and Tang Sze-Wing Table 3.3  Vowels IPA

Wong

Yale

Rao

ILE

LSHK

Lau

i e y u o ɛ ɔ œ ɵ ɐ a iw ɵʮ uʮ ej ɔj ow ɐj ɐw aj aw

i i y u u ɛ ɔ œ œ ɐ a iu œy ui ei ɔi ou ɐi ɐu ai au

i i yu u u e o eu eu a a(a) iu eui ui ei oi ou ai au aai aau

i i ü u u é o ê ê e a iu êu ui éi oi ou ei eo ai ao

i i y u u e o oe oe a a(a) iu oey ui ei oi ou ai au aai aau

i i yu u u e o oe eo a aa iu eoi ui ei oi ou ai au aai aau

i i ue oo u e o(h) eu(h) u a a(a) iu ui ooi ei oi o ai au aai aau

Table 3.4 Tones IPA

Wong

Yale

Rao

ILE

LSHK

Lau

55 / 53 (5) 25 33 (3) 21 23 22 (2)

a1 (a7) a2 a3 (a8) a4 a5 a6 (a9)

ā/à á a àh áh ah

a1 a2 a3 a4 a5 a6

a1 (a7) a2 a3 (a8) a4 a5 a6 (a9)

a1 a2 a3 a4 a5 a6

a1º / a1 (a1º) a2 a3 a4 a5 a6

Notes   1 Since consonants are not obligatory in the syllable-initial position, a zero consonant is often suggested as a phonological unit (as onset).   2 Zee (1999) indicates that /i, y, u, ɛ, ɔ, œ, a/ are long in open syllables and a third shorter in syllables closed by a plosive or nasal. The vowel length mark is not used here for the latter syllables.   3 Tone numerals are used. Dark Rising, Light Flat and Light Rising tones might also be given as /35/, /11/ and /13/. For simplicity, the description above follows Bauer and Benedict (1997).   4 According to Kataoka and Lee (2008), the use of ch and ts makes historical sense since Cantonese used to show distinction between /ts(h)/ and /tʃ (h)/. The merger into /ts(h)/ has now become the norm.   5 Neighbouring Macau, a former Portuguese colony, also has its own way to Romanize Cantonese. Precisely, this is not Romanization to English, but to Portuguese, although its Romanization is transferred directly into English contexts. Certain features differentiate it from the Hong Kong

48

Cantonese Romanizaton version, possibly designed to adapt for the Portuguese orthography. Major differences include its using v for /w/ (though w is also used) and i (or no letter at all) for /j/. /jyt2 jy23/ and /kwɔŋ25 toŋ55 wa25/ may possibly be expressed as Ut U (or Iut Iu) and Kuong Tong Va in the Macau variety (see Cheng 2014 for a comparative study of Government Romanizations in Hong Kong and Macau).   6 For some Cantonese speakers, the Yale Romanization might be noticed through its association with the Google Cantonese Input Method; although it accepts a wide range of Romanization variants, the ‘suggestion’ given in the input box is based on the Yale system (in a ‘simplified’ version; to be discussed below).   7 The use of letter h to mark the Light tones also creates a peculiar sorting order – in alphabetizing the dictionaries entries, for example, characters with the same phonetic segment but different tones are not put in consecutive order. /sɵn25/ shares the same rhyme with /sɵn23/ but not with /sɵʮ23/. But when the order is determined by the alphabetic sequence of their Yale Romanization, the three syllables will be in the sequence of ‘/sɵn23/-/sɵʮ23/-/sɵn25/’ as they are Romanized as séuhn, séuih, and séun. This differs from the normal practice in which characters having the same segment but different tones are sequenced consecutively.   8 This is reminiscent of how Mandarin pinyin is employed in foreign-language publications and scholarly works. In the pinyin system, an apostrophe is used to reduce ambiguity by marking the end of one syllable and the beginning of another, when the second one starts with a vowel letter. Scholars using the LSHK scheme often follow the same practice and put an apostrophe before syllables that begin with a vowel as well as syllabic nasals, i.e., co’ng for co3 ng6 ‘error’ and cong for cong1 ‘tumor’.   9 The changes are said to have been made ‘in response to public opinion’. It is not sure whether this may violate the specification of the original scheme which states that ‘in principle no more new letters will be invented and the letter pronunciation will not be casually changed’ (Guangdongsheng sizhong fangyan pinyin fang’an 1960). 10 Wong did mention that /b, d, g, dz/ are in fact devoiced [b̥, d̥, ɡ̊, d̥z̥ ] (/gw/ is treated as /g/ plus /w/). These phonetic details are normally omitted in dictionaries adopting Wong’s system. 11 For easy reference, the details regarding the orthographic conventions of the original Yale scheme, the ‘simplified’ Yale version, the predecessor/successor of the ILE scheme, and the original schemes of Rao’s and Wong’s systems are not shown in the tables. They are addressed in the preceding paragraphs.

References Ball, James Dyer (1883) Cantonese Made Easy, Hong Kong: China Mail Office. Bauer, Robert S. and Paul K. Benedict (1997) Modern Cantonese Phonology, Berlin: Mouton de Gruyter. Bauer, Robert S., Cheung Kwan-Hin, and Cheung Pak-Man (2003) ‘Variation and Merger of the Rising Tones in Hong Kong Cantonese’, Language Variation and Change 15(2): 211–25. Bridgman, Elijah Coleman (1841) Chinese Chrestomathy in the Canton Dialect, Macau: S. W. Williams. Chalmers, John (1859) An English and Cantonese Pocket-dictionary, for the Use of Those Who Wish to Learn the Spoken Language of Canton Province, Hong Kong: The London Missionary Society’s Press. Chao, Yuen-Ren (1947) Cantonese Primer, Cambridge, Mass.: Harvard University Press. Cheng, Siu-Pong (2014) ‘Gang-Ao liang di de zhengfu Yueyu pinyin’ (‘The Government Romanization of Cantonese in Hong Kong and Macao’), Zhongguo yuwen tongxun (Newsletter of Chinese Language) 93(1): 27–38. Cheung, Kwan-Hin (1986) ‘The Phonology of Present-day Cantonese’, Doctoral Dissertation, University College London. Cheung, Kwan-Hin (1994) ‘Xianggang Yuyanxue Xuehui Yueyu Pinyin Fang’an de yuanqi, sheji yuanze he tedian’ (‘The Genesis, Design Principles, and Characteristics of the LSHK Cantonese Romanization Scheme’), in Yueyu Pinyin Zibiao (Guide to LSHK Cantonese Romanization of Chinese Characters), Hong Kong: Linguistic Society of Hong Kong, ix–xix. Cheung, Kwan-Hin (2003) ‘Guangzhouhua Pinyin Fang’an xilie de wanshanhua guocheng’ (‘A Series of Refinements to the Cantonese Romanization Scheme 1960’), in Tang Keng-Pan (ed) Proceeding of 6th International Conference on Yue Dialects, Macau: Associação de Literatura Chinesa de Macau, 13–19.

49

Cheng Siu-Pong and Tang Sze-Wing Eitel, Ernst John (1877) A Chinese Dictionary in the Cantonese Dialect, London: Trübner and Company. Guangdongsheng sizhong fangyan pinyin fang’an (Four schemes of dialectal phonetic symbols of the Guangdong Province) (1960) Yuwen jianshe (Language Planning) 15: 21–5. Huang, Parker Po-Fei (1962) Speak Cantonese: Book II, New Haven: Institute of Far Eastern Languages, Yale University. Huang, Parker Po-Fei (1970) Cantonese Dictionary: Cantonese–English, English–Cantonese, New Haven: Yale University Press. Huang, Parker Po-Fei and Gerald P. Kok (1958) Speak Cantonese, New Haven: Institute of Far Eastern Languages, Yale University. Institute of Language in Education (1990) Changyongzi Guangzhouhua duyinbiao (List of Cantonese Pronunciation of Commonly-used Chinese Characters), Hong Kong: Government Printer. Kataoka, Shin (2014) ‘Xianggang zhengfu pinyin: yi ge luanzhongyouxu de xitong’ (‘Finding Order in Disorder: An Investigation into the Hong Kong Government Cantonese Romanization’), Zhongguo yuwen tongxun (Newsletter of Chinese Language) 93(1): 9–25. Kataoka, Shin and Cream Lee (2008) ‘A System without a System: Cantonese Romanization Used in Hong Kong Place and Personal Names’, Hong Kong Journal of Applied Linguistics 11(1): 79–98. Lam, Yan-Yan (2009) ‘Xianggang Yueyu biaoyin de xianzhuang’ (‘The Current Situation of Hong Kong Cantonese Sound Notation’), Zhongguo yuwen yanjiu (Studies in Chinese Linguistics) 27(1): 59–66. Lau, Sidney (1965) ‘Kwang Tung Wah’: Cantonese by Radio, Intermediate. Listeners’ Guides to Accompany Lessons 1 to 117, Hong Kong: Government Printer. Matthews, Stephen and Virginia Yip (1994) Cantonese: A Comprehensive Grammar, London and New York: Routledge. Meyer, Bernard F. and Theodore F. Wempe (1934) The Student’s Cantonese–English Dictionary, Hong Kong: St. Louis Industrial School Printing Press. Morrison, Robert (1828) A Vocabulary of the Canton Dialect, Macau: The Honorable East India Company’s Press. Williams, S. Wells (1856) A Tonic Dictionary of the Chinese Language in the Canton Dialect, Guangzhou: The Office of the Chinese Repository. Wong, S. L. (1941) Yueyin yunhui (A Chinese Syllabary Pronounced according to the Dialect of Canton), Shanghai: Zhonghua. Wu, Man-Ying (1997) ‘Yueyu Luoma pinyin fang’an de bijiao’ (‘Comparison of Cantonese Romanization Schemes’), Master Thesis, The Hong Kong Polytechnic University. Yu, Ping-Chiu (1971) Tongyin zihui (Rhyming Glossary), Hong Kong: Kwong Wah. Zee, Yun-Yang Eric (1999) ‘Chinese (Hong Kong Cantonese)’, in International Phonetic Association (ed) Handbook of the International Phonetic Association, Cambridge: Cambridge University Press, 58–60. Zhan, Bohui (ed) (2002) Guangzhouhua zhengyin zidian (The Dictionary of Standard Cantonese Pronunciation), Guangzhou: Guangdong Renmin Chubanshe.

50

4 Chinese characters John Jing-hua Yin University of Vermont, USA

The Chinese language, which is spoken by the largest number of people on the earth as their mother tongue, has many different dialects in addition to Mandarin Chinese. Despite the different accents of these dialects, all Chinese language speakers share one writing system enabling them to communicate with each other in writing when dialects are too difficult to understand by listening. The Chinese writing system, unlike the writing system of alphabetical languages, is formed with no letters or combination of letters to represent the sounds of the language. It is formed with Chinese characters, which are symbols constructed and used to convey meanings.

History and development of Chinese characters Although it is still not known how long Chinese characters have been in existence, we do know that Chinese characters have been a highly developed writing system for at least 3,300 years. The long history of the Chinese characters can be divided into two major periods: the ancient period and the modern period. Associated with the two periods, there are six major writing styles, starting from the earliest systematic form of Chinese characters inscribed on oracle bones. During the ancient writing period, characters we can still see today were inscribed on tortoise shells and animal bones. These characters are often referred to as ‘oracle-bone inscriptions,’ and they date to the Shang dynasty (1711 bce–1066 bce) in China. Later on, Chinese characters were cast or inscribed on bronze bells and vessels, and these characters are often called ‘bronze inscriptions.’ The bronze inscriptions were common in the Zhou dynasty (1066 bce–256 bce). While the characters’ size, positioning, complexity, and textual format in the oracle-bone inscriptions are inconsistent, the size and the textual format in the bronze inscriptions are more fixed. In bronze inscriptions, the characters are more symmetrical and the strokes are simpler. Toward the end of the Zhou dynasty, its Qin state began to use a new script both to write on bamboo strips and pieces of silk and to inscribe onto rocks or precious stones. This script is called the ‘seal script,’ which is still used for inscribing names on seals nowadays. After the Qin state conquered the other six states of the Warring States period (475 bce– 221 bce), unified China, and established the Qin dynasty (221 bce–206 bce), the seal script 51

John Jing-hua Yin

was decreed as the standard of writing for the whole country. In the seal script, the positioning of characters and complexity of forms became consistent, and all the characters were roughly square in shape. However, the character strokes became lengthened, curved, and complicated. Because a large of number of documents had to be written or copied as part of the administrative work after the unification of China, the seal script with its lengthened and curved strokes was quite time-consuming and cumbersome. Some local officials at the lower levels of the government who dealt with relatively unimportant documents developed a new script to save time. This new script continued to be developed in the Han dynasty (206 bce–ce 220) and became the officially approved formal way of writing. This script is now referred to as the ‘clerical script.’ The clerical script is very different from the seal script because there are four basic simplifications and modifications for convenience and speed of writing as well as for tidiness. First, the curved strokes in the seal script became somewhat straighter. Second, the overall number of strokes was reduced. Third, some different components were merged into one. Fourth, some components were modified and simplified. The change from the seal script to the clerical script is often referred to as the ‘clerical change.’ This change, which is the largest transformation of Chinese character structure in history, dropped the pictographic appearance of Chinese characters almost completely and established the foundation of the structures for the Modern Chinese characters. Therefore, in the history of Chinese character development, the clerical script is viewed as the turning point that divides the ancient writing period and the modern writing period. The ancient writing period, from the earliest known oracle-bone inscriptions in the late Shang dynasty to the development of the seal script, lasted about 1,160 years, and the modern writing period, from the time of the clerical change to the present, has lasted more than 2,200 years. This period of time in the history of Chinese character development is still considered modern because the structures of Chinese characters have remained the same from the clerical change until today. Although the structures of Chinese characters have not changed since the clerical change, the strokes of Chinese characters have undergone two major changes: regularization and normalization. Toward the end of the Han dynasty, strokes with an undulate end, which was a common feature of characters in the clerical script, became smooth and straight in the regular script, and some thick curvy lines seen in the clerical script changed to hooks in the regular script. This change is known as ‘regularization.’ As the strokes were relatively smoother, straighter, and clearer in the regular script, it was easier to read and write. Regular script quickly became widespread. The invention of printing helped the regular script, which was used as one of the major printing types, become widely accepted as the standard form of Chinese writing. Many calligraphers, such as Zhong Yao in the Three Kindoms period (220–280), Wang Xizhi and Wang Xianzhi in the Eastern Jin dynasty (317–420), Ouyang Xun, Yan Zhenqing, and Liu Gonquan in the Tang dynasty (618–907), Su Shi, Huang Tingjian, and Mi Fu in the Song dynasty (960–1279), and Zhao Mengfu in the Yuan dynasty (1279–1368), among many others, also contributed to making the regular script the standard for people in China to admire, imitate, learn, and master. The regular script has been the standard of Chinese writing for more than 1,800 years, with the semi-cursive and cursive styles developing concurrently. However, semi-cursive and cursive styles, as there are many variations of them, have never become standardized printing types. Instead, they have been used for everyday communication and appreciated as a form of art. 52

Chinese characters

Although the regular script changed the appearance of strokes in the clerical script, many characters continued to be complicated. Some characters have variations resulting from many centuries of use, during which time character variants with fewer strokes and simpler structures were created for convenience. Other characters have variations arising from a lack of uniformity among printing types after the invention of printing. Although these alternative forms of characters were popularly used for years, they were never officially recognized or accepted as the ‘standard forms of characters’ up to the 1950s. In the first three and a half decades of the twentieth century, attempts to accept these ‘non-standard forms of characters’ officially and to continue to simplify more Chinese characters were made, but they failed for one reason or another. However, these efforts laid the foundation for the language reform that took place in the second half of the twentieth century. After World War II and the civil war in China in the late 1940s, the normalization of Chinese characters, which comprises systemization, simplification, and standardization, was finally on the agenda of the Chinese government. A special government organization, first called the Committee for Chinese Language Reform and later the National Language Commission, was established in 1954 to direct and oversee the normalization of Chinese characters. In 1955, to systemize Chinese characters, the ‘List of the First Group of Standardized Forms of Variant Characters’ was officially published. Among the 810 sets of characters, with two to six variants in each set, 1,027 variant characters were abolished, and the remaining characters have been referred to as the ‘standard forms of characters’ since then. In 1956, after discussion, revision, and consultation, the ‘Scheme for Simplifying Chinese Characters’ was officially published. In 1964, the ‘Complete List of Simplified Characters’ was officially published, and it was republished in 1986. The list has 2,235 simplified characters in total, and 1,116 of them are frequently used today in daily life. In fact, among the 2,235 simplified characters, only 482 of them are basic. The remaining 1,753 characters in the list are all derivatives of these 482 basic simplified characters. Among the 482 basic ones, about 20% were newly created in the 1950s, while the other 80% were created as long as several thousand years ago. In 1964, in an effort to standardize the printed forms of Chinese characters, the ‘List of Chinese Character Forms for General Printing’ was officially published. The list has provided the standardized printed forms for 6,196 generally used characters. In 1988, the ‘List of Frequently Used Characters in Modern Chinese,’ which includes 3,500 characters, was officially published, and later in the same year, the ‘List of Generally Used Characters in Modern Chinese’ was also officially published. This second list contains 7,000 characters, which include the 3,500 frequently used ones contained in the first list. In 1997, ‘Standard Stroke Order for Generally Used Characters in Modern Chinese’ was officially published. After 50 years of simplifying, systemizing, and standardizing work on Chinese characters, they are considered normalized. The ‘normalized forms of Chinese characters’ include two major types: (i) the inherited characters, which are the Chinese characters that have kept their structure ever since the ‘clerical change’ time, and (ii) the simplified characters, which are the Chinese characters that have been simplified and officially recognized. The original forms of the characters, before being simplified or replaced by today’s simple forms, are often referred to as ‘complex characters’ by people in mainland China. These complex characters together with the inherited characters constitute the ‘traditional forms of Chinese characters.’ ‘Modern Chinese characters’ is another term referring to the normalized forms of Chinese characters. They are officially recognized and used in mainland China, Singapore, and 53

John Jing-hua Yin Table 4.1  Ancient writing period (c. 1711 Major forms

Ancient form

Oracle-bone inscription

Most popular time

Examples

Shang dynasty (1711 bce–1066

person

tree

water

fire

cart

horse

bird

cloud

Zhou dynasty (1066 bce–256

Seal script

Late Zhou dynasty and the Qin dynasty (221 bce–206 bce)

Major forms Traditional form

bce)

Bronze inscription

Table 4.2  Modern writing period (c. 206

Normalized form

bce–256 bce)

Clerical script

bce)

bce–present)

Most popular time

Examples

Qin dynasty and Han dynasty (206 bce–ce 220)

From the late Han dynasty to 1955 Regular script From 1955 to the present

Malaysia, as well as at the United Nations. The traditional forms of Chinese characters are still in use in Taiwan, Hong Kong, and Macao, as well as among some Chinese communities outside China. Modern Chinese characters as included in the ‘List of Generally Used Characters in Modern Chinese’ consist of about 31% simplified characters and about 69% inherited characters. Tables 4.1 and 4.2 provide examples of the historical development of Chinese characters. The development of Chinese characters in terms of the evolution of forms over the past several thousand years has demonstrated three general trends in the process. First, Chinese characters used to be drawings depicting objects and have slowly become more abstract writing symbols. Second, Chinese characters started with the pictographic method as the major way of character construction and have ended up using more and more abstract 54

Chinese characters

symbols developed from pictographs as phonetic symbols in new characters. Third, Chinese characters used to have a lot of pictographic symbols that were complicated to write, but they have now become logographs that are much easier to write.

Formation and structure of Chinese characters In the earliest known stages of written Chinese, there were six kinds of Chinese characters: pictographic characters, indicative characters, associative characters, picto-phonetic characters, explanative characters, and phonetic loan characters. The first five kinds (i.e. pictographic, indicative, associative, picto-phonetic, and explanative characters) are categorized according to the ways characters were created, and the last one (i.e. phonetic loan characters) complemented the other methods by borrowing existing characters to refer to things that no characters had been created for. These six kinds of Chinese characters, which were summarized and exemplified by Xu Shen in his An Analysis and Explanation of Characters at the beginning of the first century ce, reflect the earliest stages of the creation and use of Chinese characters. 1. The pictographic method is the first major way of creating Chinese characters that came into use. The human body or body parts as well as things that can be observed in nature were depicted in simple drawings based on their most conspicuous and differentiated traits. 2. However, an abstract concept is not expressed easily by drawing a picture, so the indicative method, in which a symbol was added to a drawing to indicate the concept, was adopted. 3. When a meaning could be sensed or deduced by combining two existing pictographs, a combined form was produced to save the trouble of creating a new pictograph. This method of combining two existing pictographs is called the associative method. 4. When a meaning could not be expressed by creating a character with the associative method, then the picto-phonetic method was deployed. The picto-phonetic method forms a character with a pictograph as one part to indicate the major category of meaning and another part to indicate the pronunciation of the whole character. 5. When a pictographic component representing both the sound and the meaning was viewed as unable to indicate the meaning clearly, another pictographic component would be added to clarify or differentiate the meaning of the whole character. This method is the explanative method. The explanative method was also deployed when a character was borrowed (but never returned) to mean something else simply because this borrowed character had the same pronunciation as the new idea, which as yet had no character to express it. 6. Those borrowed characters are called phonetic loan characters. Among the above six methods, the picto-phonetic method has been the most productive in creating characters. Because the explanative method has not been clearly understood for many years and because a lot of work to distinguish explanative characters from pictophonetic characters has still not been satisfactorily done, many explanative characters are confused with picto-phonetic characters. Nevertheless, it is widely accepted that over 90% of the characters used today are picto-phonetic in a broad sense. Of course, because of developments and changes in the language over the past few thousand years, in Modern Chinese no more than 30% of the picto-phonetic characters contain a phonetic component that can accurately represent the pronunciation of the whole character. 55

John Jing-hua Yin

Table 4.3 shows examples of the six categories of Chinese character construction: Table 4.3  Categories of Chinese character construction Categories 1.

2.

3.

4.

5.

6.

Examples

Pictographic characters 人 person

木 tree

上 upper

本 root

tree + tree = woods

person + tree = rest

water + (sound) = shower

water + (sound) = wash

tree + branch = branch (sound & meaning)

water + spill = spill (sound & meaning)

六 six (originally ‘hut’)

北 north (originally ‘back’)

Indicative characters

Associative characters

Picto-phonetic characters

Explanative characters

Phonetic loan characters

Regardless of how a character was created, when created it almost always had the trait of a pictograph one way or another. Over thousands of years, the pictographic essence slowly faded and is no longer visually obvious in Modern Chinese characters. Now, formed with various kinds of lines, dots, and hooks, Chinese characters are highly symbolic. They mostly appear as logographs rather than pictographs. In other words, they look more like symbols than drawings.

Features of Chinese characters 1.  One character, one phonetic syllable A Chinese character is an independent symbol that expresses a meaning. What makes the pronunciation of a character easy is the fact that the pronunciation of any character in Chinese is monosyllabic: one character represents one phonetic syllable. A phonetic syllable is a segment of pronunciation that contains one vowel sound with or without one or more consonant sounds. Vowel sounds are loud and clear. When they are pronounced, the airflow

56

Chinese characters

from the lungs vibrates the vocal chords, and the shape of the mouth and the position of the tongue make the different sounds. In Chinese, it does not matter what a character means, and it does not matter whether the character is a word by itself or a morpheme (part of word), it is always pronounced with one syllable.

2.  One syllable, often many different characters Although it is convenient that each character is pronounced in one syllable, the tones make it a bit complicated. What makes the relationship between characters and the Chinese sound system more complicated are the homophones. In the Chinese sound system, there are only about 400 different syllables. Even with as many as four different tones for each syllable, the total number of syllables with different tones in use in Chinese is 1,196. On the other hand, as collected and recorded in the Giant Dictionary of the Chinese Language, published in 1990 in China, the total number of Chinese characters that have been created since ancient times is 54,678. If we take the 2,500 characters that were found to be the most frequently used and divide by 1,196, which is the total number of different syllables in Mandarin Chinese, we still get about two characters for each syllable on average. In reality, some syllables are represented by more than two of these 2,500 most frequently used characters. Take the syllable yi as an example. The first-tone syllable yī has four characters; the second-tone syllable yí has six characters; the third-tone syllable yǐ has six characters; and the fourth-tone syllable yì has 16 characters. It would be impossible to understand what someone means when he or she says yi in any of the four tones out of the context.

3.  Not all characters are in active use The number of Chinese characters has grown over the last few thousand years, as indicated by the major dictionaries that have been compiled during that time. An Analysis and Explanation of Characters is the earliest dictionary we know of. It was put together by Xu Shen around 121 ce with a collection of 9,353 characters. The Kangxi Dictionary, compiled by Zhang Yushu, Chen Tingjing, and others in 1716 under the rule of Emperor Kangxi in the Qing dynasty (1644–1911), includes 47,043 characters. The Giant Dictionary of the Chinese Language, compiled by Xu Zhongshu and others and published in 1990, records 54,678 characters. However, the number of characters in general use in modern China is about 7,000. Among the 7,000 characters, the general public now frequently uses only about 3,500, as found in the ‘List of Frequently Used Chinese Characters’ published by the National Language Commission of China in 1988. Furthermore, a test done in 1987 by Shanxi University, under the request of the National Language Commission, on randomly selected samples of printed materials containing 2 million characters, shows that 1) the 2,500 most frequently used characters in the first part of the list can account for 97.97% of the characters that appeared in the data; 2) the next 1,000 characters in the second part of the list can cover 1.51%; and 3) the total coverage rate of the two groups (with 3,500 characters in all) is 99.48%. This means that if someone could recognize the first 2,500 characters in the list, he or she would run into only two unrecognized characters in every 100 characters that appear in Chinese printed materials similar to the tested data. Having learned the remaining 1,000 characters in the list, he or she would be able to reduce the number of unrecognized characters to one in every one hundred characters seen in the same kind of printed materials.

57

John Jing-hua Yin

4.  Not all characters are words It is encouraging to realize that knowing 3,500 of the most frequently used characters enables one to recognize 99% of the characters that appear in newspapers, magazines, and books for general readers, but that knowledge does not guarantee readers’ understanding because characters are not always words in Modern Chinese. One must know the relationship between the characters in a word and the meaning of a word formed by two or more characters. In Classical Chinese, most words are only one character. However, in Modern Chinese, most words contain two characters; as a result, many of the characters that would be words in Classical Chinese are now morphemes that form words in Modern Chinese. In designing the Chinese Language Proficiency Test for learners of Chinese as a foreign language in 1992, the Testing Center of the Office for the National Chinese Language Proficiency Test Committee developed the ‘General Outline of the Chinese Vocabulary Levels and Graded Chinese Characters.’ It is based on 16 different lists of Chinese words (not characters) and includes 8,822 carefully selected words at four levels (refer to Table 4.4). The developers predicted that 3,000 of the most frequently used words would cover 86% of the language materials for general readers, 5,000 would cover 91%, and 8,000 would cover 95%. If we look at the number of characters that the selected 8,822 words contain and compare them with the 2,500 most frequently used characters in the ‘List of Frequently Used Characters’ mentioned earlier, we see that the developers’ prediction about the vocabulary coverage rate is believable. The 8,822 words in the ‘General Outline’ have been formed with 2,905 characters, of which 2,485 characters overlap with the 2,500 characters that appear in the first part of the ‘List of Frequently Used Chinese Characters.’ Although not all characters are words, knowing the meaning of characters will help one learn the meaning of the words they form. If the best way to increase one’s English vocabulary is to learn prefixes, suffixes, and roots, then similarly the most efficient way to increase one’s Chinese vocabulary is to learn Chinese characters, especially those most frequently used characters.

5.  The smallest units forming a character are strokes Chinese characters originated as pictographs, but through thousands of years of development and changes, the pictographic essence slowly faded and is not visually obvious in Modern Chinese characters, which now are highly symbolic and constructed with various kinds of lines, dots, and hooks. These lines, dots, and hooks are called ‘strokes.’ A stroke is a mark made by a writing instrument, such as a pen, that is put down on a writing object, such as paper, and moved until it is lifted from the writing object. These strokes are arranged in a special way to form characters. Some characters have two or more components, and these components are also formed by strokes. Strokes are the smallest units used to form characters. In Modern Chinese characters, there are six basic strokes: the horizontal stroke, the vertical stroke, the left-falling curved stroke, the right-falling curved stroke, the dot, and the rising stroke. Table 4.4  Number of words and characters at each proficiency level

Words Characters

Level 1

Level 2

Level 3

Level 4

Total

1,033    800

2,018    804

2,202    601

3,569    700

8,822 2,905

58

Chinese characters

Apart from the six basic strokes, there are also three dependent strokes: the bend, the turn, and the hook. Dependent strokes cannot be used independently; they have to be attached to at least one other kind of stroke to form a complex stroke. The six basic strokes and the three dependent strokes together form all of the 27 compound strokes used in Modern Chinese characters. Table 4.5 is a list of all the strokes in Modern Chinese characters.

6.  There are stroke order rules to follow when writing characters If there are more than two strokes in a Chinese character, there are rules for conventional stroke order that help writers to balance Chinese characters. The knowledge of stroke order also helps people appreciate handwritten forms of Chinese. Rule 1: From top to bottom. The strokes at the top of a character should be written before those at the bottom. For example: 三 (three): Rule 2: From left to right. The strokes on the left side of a character should be written before those on the right. For example: 八 (eight): Rule 3: The horizontal before the vertical or the left-falling. The horizontal stroke should be written before the vertical stroke or the left-falling stroke if they cross each other. For example: 十 (ten): Rule 4: The enclosing strokes first, then the enclosed, and finally the sealing horizontal stroke. The enclosing strokes, which form the sides and top of the frame, should be written first. Strokes inside the frame should be written next, and then the closing stroke, which is usually a horizontal stroke at the bottom of the frame. For example: 四 (four): Rule 5: Left-falling before right-falling. When the left-falling stroke and the rightfalling stroke come together in a character, write the left-falling stroke first and then write the right-falling stroke. For example: 木 (tree): Rule 6: Vertical stroke in the middle before the strokes on both sides. When a vertical stroke or the stroke containing a vertical segment in the prominent middle position does not cross other strokes, it should be written first. For example: 水 (water): These rules are not absolute, but relative and mutually conditioned. For example, Rule 3 requires us to write the horizontal stroke before the vertical when they are together in a character; however, if the horizontal stroke does not cross the vertical stroke and is on its right side, then the vertical stroke goes before the horizontal stroke. Therefore, in the . Sometimes, after lower part of the character 足 (foot), the stroke order is years of writing, people find that one way of writing is easier than another, even though it does not comply with the stroke order rules. Therefore, there are exceptions to the above rules. For example, if you follow Rule 3, there is no exception in writing the character . However, writing the character 女 (female) is an exception because the 大 (big) as conventional way of writing 女 does not start with the horizontal stroke. It goes like this: . You may find it easier to write out a well-balanced 女 this way than having the horizontal stroke first. Another example is the character 子 (baby). The horizontal stroke is . drawn last: 59

Table 4.5  Chinese character strokes Stroke

Basic strokes

Name in Chinese Example

Basic stroke variants

Variant 1

Variant 2

Variant 3

héng

shù

ー

ー

piě

nà

horizontal

vertical

leftfalling

rightfalling

short horizontal

short vertical

vertical leftfalling

leveled rightfalling

pointed vertical

small leftfalling

Dependent strokes diǎn

tí

wān

zhé

gōu

ヽ dot

rising

bend

turn

hook

left dot

big right dot

leveled leftfalling

Two strokes combined

Compound strokes

ー ー

plus plus

plus wān plus (bending) wò plus (crouching) xié plus (slanting)

Three strokes combined

plus plus plus

Four strokes combined

plus plus plus plus plus

Five strokes combined

plus

plus

 

Chinese characters

7.  Regardless of the number of strokes or components a character may consist of, it should be written consistently in the same size square box as the others In English, a word that has more letters is longer than a word that has fewer letters. Longer words occupy more space than shorter words when spelled out. In Chinese, even though associative, explanative, and picto-phonetic characters are formed with two or more character components, they still have to be written in the same size as those single-element characters. Therefore, the number of strokes and components does not determine the size of a character. All characters should be written consistently in the same size. The formation structures are determined by the way character components are put together. These components are also written out in a conventionally fixed sequence. See the following. The numbers given in the boxes below indicate the character component writing sequence. 1

2

  1. Top–Bottom (T–B) structure: 男, 分

2

1

  2. Left–Right (L–R) structure: 明, 好 1

2

3

  3. T–B structure embedded with an L–R on the top: 哭, 想

1

2

  4. T–B structure embedded with an L–R at the bottom: 罚, 霜

3

2 1 3

  5. L–R structure embedded with a T–B on the right: 解, 语 1 3

  6. L–R structure embedded with a T–B on the left: 剖, 数 1

2

3

  7. Left–Middle–Right structure: 湖, 街 1 2 3

  8. Top–Middle–Bottom structure: 意, 爱 61

2

John Jing-hua Yin

1 2 3

  9. Complete-Wrap structure: 困, 园 1

2

10. Top-Wrap structure: 风, 同 1 2

11. Bottom-Left-Wrap structure: 道, 近

1

2

12. Top-Right-Wrap structure: 包, 氧 1

2

13. Top-Left-Wrap structure: 病, 庆 1 2

14. Left-Wrap structure: 匣, 匹

3

1

15. Bottom-Wrap structure: 画, 凶

2

8.  Characters contain recurring side components, by which characters are arranged in character dictionaries and with which people learn to recognize more characters Multicomponent characters have many recurring side components, which are commonly referred to as ‘radicals.’ The entries in a dictionary are traditionally categorized by these radicals. People also learn to recognize more characters by knowing the meanings of these radicals first. Here are the characters commonly used as radicals1: 厂卜八人儿几刀力厶又士土工寸大弋小口山巾广门(門)尸己弓子女马(馬)幺王 韦(韋)木犬歹车(車)戈比瓦止日曰贝(貝)水见(見)牛手毛气片斤爪父月欠风(風) 殳文方火斗户心毋示石龙(龍)业(業)目田皿矢禾白瓜用鸟(鳥)立穴疋皮矛耒老耳 臣页(頁)虫缶舌竹臼自血舟衣羊米艮羽糸麦(麥)走赤豆酉辰豕卤(鹵)里足身采谷 豸角言辛青其雨齿(齒)黾(黽)隹金鱼(魚)革骨鬼食音鬥髟麻鹿黑鼠鼻 62

Chinese characters

There are also radicals that cannot be used alone as characters since they are either strokes or dependent components, including the reduced forms of characters: 勹匚冂廴廾刂亻亠冫冖讠(訁)卩  阝 2 阝 3 凵彐巛扌艹(艹)尢囗彳彡犭夕夂饣(飠) 丬(爿)氵忄宀辶纟(糹)屮攵攴灬礻肀罒钅(釒)疒衤覀虍

Notes 1 Those simplified characters used as radicals are followed by their traditional counterparts in parentheses. 2 A reduced form of the character 阜 (mound) used as a left side component in a character. 3 A reduced form of the character 邑 (town) used as a right side component in a character.

Bibliography Guojia Duiwai Hanyu Jiaoxue Lingdao Xiaozu Bangongshi Hanyu Shuiping Kaoshibu 国家对外汉语 教学领导小组办公室汉语水平考试部 ‘Chinese Language Proficiency Level Test Division of the National Office of the Leading Group for Teaching Chinese to Speakers of Other Languages’ (1992) Hanyu Shuiping Cihui Yu Hanzi Dengji Dagang《汉语水平词汇与汉字等级大纲》‘General Outline of the Chinese Vocabulary Levels and Graded Chinese Characters’, Beijing: Beijing Yuyan Xueyuan Chubanshe. Hanyu Dazidian Bianji Weiyuanhui 汉语大字典编辑委员会 ‘Editorial Committee of the Chinese Giant Dictionary’ (1995) Hanyu Dazidian (Suoyiben)《汉语大字典(缩印本)》‘Chinese Giant Dictionary (Compact Edition)’, Wuhan: Hubei Cishu Chubanshe and Chengdu: Sichuan Cishu Chubanshe. Li, Dasui 李大遂 (1993) Jianming Shiyong Hanzixue《简明实用汉字学》‘Concise and Practical Study of Chinese Characters’, Beijing: Beijing Daxue Chubanshe. Su, Peicheng 苏培成 (1994) Xiandai Hanzixue Gangyao《现代汉字学纲要》‘Essentials of the Study of Modern Chinese Characters’, Beijing: Beijing Daxue Chubanshe. Xinhua Dictionary with English Translation《英汉双解新华字典》(2000) Beijing: The Commercial Press. Xu, Shen 许慎 (1963) Shuo Wen Jie Zi《说文解字》‘An Analysis and Explanation of Characters’, Beijing: Zhonghua Shuju. Yin, Bingyong 尹斌庸 and J. S. Rohsenow (1994) Modern Chinese Characters《现代汉字》, Beijing: Sinolingua. Yin, John Jing-hua 印京华 (2006) Fundamentals of Chinese Characters, New Haven: Yale University Press. Yuyan Chubanshe 语言出版社 ‘Language Press’ (1997) Yuyan Wenzi Guifan Shouce《语言文字规范 手册》‘Language and Character Standardization Handbook’, Beijing: Yuyan Chubanshe. Zhang, Shuyan, Tiekun Wang, Qingmei Li, and Ning An 张书岩、王铁昆、李青梅、安宁 (eds) (1997) Jianhuazi Suyuan《简化字溯源》‘Origins of Simplified Characters’, Beijing: Yuwen Chubanshe. Zhao, Yuanren 赵元任 (1980) Yuyan Wenti《语言问题》‘Language Issues’, Beijing: The Commercial Press.

63

5 Chinese Idioms Jiao Liwei University of Pennsylvania, USA

1.  A brief introduction and three examples: 守株待兔 (shouzhu-daitu), 千方百计 (qianfang-baiji), and 当务之急 (dangwuzhiji) Chinese idioms (成语, chengyu) are one of four types of formulaic expressions (熟语, shuyu), which also include collocations (惯用语, guanyongyu), two-part allegorical sayings (歇后语, xiehouyu), and proverbs (谚语, yanyu; or 俗语, suyu) (D. Wen 2005: Ch. 2, Sec. 3). Conventionally, the term chengyu is translated into English as ‘idiom’, ‘set phrase’, or ‘four-character expression’. More than 90% of Chinese idioms consist of four characters, which can be parsed into groups of two when recited at normal speed. Typically, the meaning of idioms is more than the sum of the parts. In sentences, Chinese idioms function mainly as predicates, or often as attributives or adverbials, but seldom as subjects or objects. Chinese idioms make writing or speech more concise and more vivid, conveying a sharper impression to readers and listeners. Chinese idioms are an important marker of written language, but they also contribute to increasing the variety of colloquial expressions (L. Wang 1987: i). Below are three examples: ● 守株待兔 (shouzhu-daitu, guard-tree stump-wait for-hare, i.e. to wait passively for a windfall, wait for gains without pains, trust chance and windfalls). This idiom makes reference to a story in the Wu Du (five vermin) section in the Han Feizi (before 233 bc): ‘There was once a farmer who was ploughing land in the State of Song. One day a hare was dashing rapidly; all of a sudden, it ran into a wooden post in the ground, snapped its neck and died. The farmer was happy and therefore abandoned his plough and waited next to the wooden post, hoping to pick up hares again. But he did not get any more hares, and was mocked by the people in the State of Song.’ If one did not know the allusion, one could hardly understand the figurative meaning of this idiom, which goes far beyond the literal meaning of the four characters. ● 千方百计 (qianfang-baiji, thousand-method-hundred-plan, i.e. by every possible means, by hook or by crook). There are three grammatical or semantic linkages in this idiom: 千 (thousand) with 百 (hundred), 方 (method) with 计 (plan), 千方 (thousand methods) with 百计 (hundred plans), all of which contribute to the overall meaning. ● 当务之急 (dangwuzhiji, should-pursue-’s-urgency, i.e. a matter of great urgency). This idiom comes from a line in Mencius which nevertheless did not contribute to its qualification as an idiom. However, the nature of the four compositional characters in the written language is the prerequisite for this phrase to become an idiom. 64

Chinese idioms

2.  From a historical perspective: Song–Yuan dynasties, 1910s, and 1958 Nowadays idioms are highly esteemed by people from almost all walks of life in China, from middle school students to marketing managers of large pharmaceutical companies. As of July 15, 2013, a search of ‘成语’ (Chinese idioms) in book titles yielded 3,605 results in the Chinese literature collection of the National Library of China. It is extremely common to find Chinese idioms in newspapers. But it was not until the twentieth century that people began collecting and categorizing them (Fang 1943; Z. Zhou 1955; Ma 1962). The term 成语 was first used in the Tang dynasty. There were some collections of idioms in the Ming and Qing dynasties, such as《幼学须知》or《幼学琼林》(The Children’s Knowledge Treasury), compiled in the Ming dynasty,《陔余丛考》(The Notebook during Home Leisure Period, 1791),《恒言录》(The Record of Eternal Words, 1805),《常语寻源》(The Search for Origins of Common Sayings, 1876), or even as early as《释常谈》(Annotated Common Sayings) in the Song dynasty (see S. Wen and Wen 2009: 4–18; J. Zhou 2004). However, the entries, mixed with many other non-idioms, were intended mainly to remind the people that these expressions were quotations from earlier writers. The first true dictionary of idioms was《国语成语大全》(The Comprehensive Dictionary of Chinese Idioms), compiled by Houjue and published by Zhonghua Bookstore in 1926. It contains approximately 3,200 entries. Archiving and academic research on idioms emerged late, but the extensive use of idioms started early. The online Dictionary of Chinese Idioms《成语典》compiled by the Taiwan Ministry of Education has a nautilus fossil on its homepage to symbolize the fact that idioms are linguistic artifacts from the past – language fossils. However, though they are rooted in history, they are also widely used in modern times. Studies of the origin of idioms vary in their statistics. Xiao (1987), for example, claims that 3,128 of 4,600 idioms – roughly 68% – originated from before the Han dynasty, while only 8% dated from after the Song dynasty. Sun (1989: 90) has 15% of idioms dating from after the Yuan, while T. Zhang (1999) claims that of 6,593 idioms, 1,120 (14.52%) originate from before the Han dynasty, and 4,036 (55.56%) from after the Song dynasty. These differences reflect different ways of determining the date of an idiom (Sun 1989: 88–90). For example, 结草衔环 (jiecao-xianhuan) has two allusions: 结草 is from a story attested in 594 bc; 衔环 is from a story of a man who lived from ad 59 to 124. But the full form, 结草衔环, was not recorded until the Yuan dynasty (1271–1368). Almost all idioms consist of four characters, and more than 40.82% of idioms are symmetrical in structure (Y. Liu et al. 2010). According to Lu Zhiwei (1956), extensive use of juxtaposed quadrasyllabic expressions began as late as in the Yuan dynasty. From as early as the Song dynasty through to the beginning of the twentieth century, there was a major division between the formal/written language and the informal/vernacular language. That is to say, people used vernacular language when speaking, but Classical Chinese when writing, which is often considered a monosyllabic language. Poetry, prose, fiction, and drama, written mainly in Classical Chinese, were highly valued by the educated elite in history. Written Vernacular Chinese – modern written Mandarin – finally replaced Classical Chinese after the Written Vernacular Chinese Movement in the mid 1910s. Based on the major treebanks of Modern Chinese at University of Pennsylvania, Tsinghua University, Peking University, and Academia Sinica1 (see Wang Yuelong and Ji 2009), we estimate the average number of characters in a Chinese simple sentence is around 16. If a quadrasyllabic idiom appears in a 16-character sentence, it is salient in length, among other 65

Jiao Liwei

distinguished features including rhetorical and cultural ones. Since the average length of all non-idiomatical words in Chinese is 1.9 syllables/characters (H. Wang 2009), idioms of four syllables naturally leave a stronger impression upon readers and listeners. According to statistics by Chen (2011) and Zhang Jie (2007), there are 3,679 quadrasyllabic idioms of a total of 5,648 quadrasyllabic words and expressions in the authoritative Contemporary Chinese Dictionary (5th ed.); that is to say, about two thirds of all quadrasyllabic words or expressions are idioms. Idioms are strongly associated with quadrasyllables, which are longer than the Chinese standard foot of two or three syllables (Feng 1998). The above evidence shows that the rising status of Chinese idioms in language is a result of natural selection in language evolution and reform, and that the nature of idioms preserved certain forms of Classical Chinese and much content of Chinese culture. However, the dictionary《汉语成语小词典》(A Concise Dictionary of Chinese Idioms, 1958) itself changed the course of the spread of Chinese idioms among the people. On February 11, 1958, the Chinese Central Government promulgated the Scheme of the Chinese Phonetic Alphabet. In order to present a gift to the National Day on October 1, 17 linguistics students in the Department of Chinese at Peking University, under the supervision of the renowned linguists Wei Jiangong and Zhou Zumo, spent 20 days compiling a dictionary containing more than 2,700 Chinese idioms. This dictionary was the first of its kind to arrange entries according to the Latin alphabet, and was intended to target middle school students, cadres, and general readers with a middle school education. The dictionary became extremely popular. The first printing in 1958 was of 300,000 copies, and the first printing of the first revised edition in 1959 was of 800,000 copies. By 2012 this dictionary had had six editions and 165 printings, and had greatly popularized knowledge of idioms in mainland China. With the outburst of reality talent shows in China, two programs on idioms were broadcast on TV. The first one, named ‘Idioms Hero’ (成语英雄), was hosted by Henan TV and first aired in 2013. In the second season, the producers recruited a prominent writer, the former Minister of Culture, Wang Meng, to be guest commentator. The other program, ‘Chinese Idioms Conference’ (中国成语大会) was first broadcast in 2014 at prime time on China Central Television, and won the highest audience ratings among Sunday evening shows in its first season. Examinations play a crucial role in the Chinese formal education system and from October 1954, Chinese idioms have been included in the Chinese language guidelines of elementary schools. Chinese Language and Literature is compulsory part of the annual College Entrance Exams, which are vital to Chinese students’ futures. Since idioms are always tested, teachers and students pay great attention to them. Education institutions in Taiwan value Chinese idioms just as highly as those in mainland China. The Ministry of Education of Taiwan publishes an online dictionary of idioms, named 成语典 http://dict.idioms.moe.edu.tw/cydic/ index.htm. 成语典 and 成语词典 differ by one character only. 典 means ‘classic’ while 词典 means ‘dictionary’. The importance of idioms is stressed in the book title. At the turn of the twenty-first century and along with rise of the Internet and social media, the power to create new words and expressions shifted from the pens of famous writers to the keyboards and mouths of celebrities and occasionally to ordinary people. New words or expressions such as 神马都是浮云 (shenma doushi fuyun, celestial-horses-all-arepassing-clouds, i.e. nothing is eternal) seemingly prevail over idioms like 万事皆空 (wanshijiekong, all-things-are-empty, i.e. nothing is eternal). However, Chinese idioms will continue to be very important in the realm of Chinese words and expressions, owing to the fact that the use of idioms has remained stable throughout history and their prestige in the education of Chinese language and literature to young people. 66

Chinese idioms

3.  Quantity of Chinese idioms: 50,000, 10,000, and 500 In English there are about 25,000 idioms. As for Chinese idioms, some comprehensive dictionaries catalogue more than 20,000 entries. There are, for instance, 28,505 entries in 《成语典》(Dictionary   of Chinese Idioms) of the Ministry of Education of Taiwan, 45,000 in《汉大成语大词典》(‘Han Da’ Comprehensive Dictionary of Chinese Idioms), and 50,000 in《汉语成语源流大辞典》(Etymological Comprehensive Dictionary of Chinese Idioms) by Liu Jiexiu. For compact dictionaries,《汉语成语小词典》(A Portable Dictionary of Chinese Idioms) by the Commercial Press collects 4,600 common idioms, and the authoritative Contemporary Chinese Dictionary) has 3,400 entries. According to 《现代汉语词典》(The   one estimate, there are about 10,000 Chinese idioms excluding variants (T. Zhang 1999). Liu and Qin (2007) analyzed statistics from a corpus of 426 million Chinese characters collected from archives of 15 major newspapers in China over the course of a year and found a total of 8,637 different four-character idioms. Jiao et al. (2011) collected statistics from 60 million characters based on five corpora of different types and found 8,844 four-character idioms. It is a reasonable guess that the number of Chinese idioms in actual use is about 8,000 to 10,000. The frequency of occurrence of idioms naturally varies greatly. According to C. Liu and Qin (2007), cumulative occurrences of the 405 most common idioms account for 41.5% of occurrences of all idioms. Another statistic is the cumulative occurrence of the 500 most common idioms, which make up about 50% of the occurrences of all idioms (Jiao et al. 2011: vii). Therefore, we can infer that there are 400 to 500 very common idioms in the Chinese language. A total of 755 idioms appear in all the 24 volumes of Chinese textbooks in use in mainland China in 2009 (Jiao et al. 2011). As for single volume books, Dream of the Red Chamber (《红楼梦》), the greatest novel in the history of Chinese literature, uses 1,429 idioms in its total 730,000 characters. Among its 1,429 idioms, 1,044 are colloquial idioms which include proverbs and collocations (see J. Zhou and Shi 2012). This being said, there are only about 400 written-style idioms. In the most famous historical novel, Romance of the Three Kingdoms (《三国演义》), about 500 idioms were used. As for individuals in modern history, Mao Zedong was well known for his language skills, but there are only about 600 idioms and allusions in the five-volume Selected Works of Mao Zedong. There are 481 idioms in the three-volume Selected Works of Deng Xiaoping. One can conclude, therefore, that the number of Chinese idioms known to the average well-educated person is about 500.

4.  Definition of Chinese idioms: quadrasyllabic, conventional, and semantic duality On November 3, 2005, the name of the famous Western fairy tale The Three Little Pigs was included as an entry in the reference materials section in the online Dictionary of Chinese Idioms of Taiwan. In July of 2007 this matter was much hyped by the media, with the Minister of Education at that time asserting that 三只小猪 (san zhi xiaozhu, ‘three little pigs’) was, in fact, now a Chinese idiom. His opinion was ridiculed by the media and by linguistic professionals, both in Taiwan and in mainland China.2 But why was the phrase 三只小猪, consisting of four characters and with its apparent duality of meanings, not accepted as an idiom? What is the definition of a Chinese idiom? There are many definitions of Chinese idioms to date (see Xu 1997; Zeng 2011: 3–6). Some major features of Chinese idioms are listed below: 67

Jiao Liwei

Quadrasyllabic (四字格) Almost all idioms consist of four characters. (Z. Zhou 1955). Some researchers include other lengths, from disyllabic to as long as 16 syllables. For instance, 推敲 (tuiqiao, push-knock, to ponder intently) is derived from a story of a Tang dynasty monk and poet, Jia Dao, who tried to find the best word for a scenario. 各人自扫门前雪,莫管他人瓦上霜 (individualpeople-himself-sweep-door-front-snow, don’t-care-other-people-tile-top-frost, i.e. to mind one’s own business, to hoe one’s own row) is a maxim from the Song dynasty. The longest idiom might be 不是东风压倒西风,就是西风压倒东风 (if-not-east-wind-prevail-over-westwind, then-be-west-wind-prevail-over-east-wind), which is from a line in Dream of the Red Chamber.

2+2 (二二相承/骈体性) The four characters in an idiom can be parsed into two groups of two (Lü 1989; Sun 1989: 75; D. Wen 2005: 70, 291). On the phonetic level, all idioms, except 作壁上观 (read as zuo-bishangguan), 成一家言 (read as cheng-yijiayan), 居大不易 (read as ju-dabuyi) and possibly a few more, can be parsed into two groups of two, regardless of the internal grammatical structure, for example, the internal grammatical structures of 妙不可言 (miaobu­ keyan), 衣食住行 (yishizhuxing), and 一衣带水 (yiyidaishui) are 妙|不可言, 衣|食|住|行, and 一衣带|水.

Conventional and fixed (约定俗成的, 定型的) Generally, the compositional characters of an idiom cannot be changed to other characters. For example, the 伪 (wei) in 去伪存真 (quwei-cunzhen, eliminate the false and retrain the true) cannot be changed to 假 (jia) even though 伪 and 假 are synonyms (Z. Zhou 1955). Similarly, the 贫 (pin, impoverished) in 一贫如洗 (yipinruxi, in utter destitution) cannot be changed to 穷 (qiong, poor).

Integral (整体性) Regardless of the internal structure, most idioms usually function as a word in a sentence (Z. Zhou 1955; Ma 1962: 33). For example, the syntax of 愚公移山 (yugong-yishan, foolish-old man-move-mountain, freely translated as ‘where there’s a will, there’s a way’) is NP1 (愚公) + VP (V (移) + NP2 (山)), however, this idiom functions mainly as attributive, often followed by 精神 ( jingshen, spirit).

Semantic duality (表意的双层性) An idiom’s actual meaning is different from the sum of the literal meanings of each individual character; that is to say, an idiom has overtones or hidden meaning between the lines (Li and Liu 1975: 101; S. Liu 1982; Shi 1979: 12; Dictionary of Chinese Idioms Taiwan 2010). For example, 胸有成竹 (xiongyouchengzhu, chest-have-fully developed-bamboo) does not mean one has a bamboo in his chest, but ‘to have a well thought out plan, have a card up one’s sleeve’. Psychological experiments show that meaning integration does not depend on the intactness of structural information (Y. Liu et al. 2010). It is true that while a large number of idioms are used in a metaphorical (比喻), exaggerative/descriptive 68

Chinese idioms

(形容), referral (引申), or denotive (借代) way which could be easily identified in the annotation part of an idiom in any authentic Chinese dictionary of idioms, there are many idioms whose overall meaning is just the summarization of the four compositional characters, or else a very close approximation. For example, 全心全意 (quanxin-quanyi, complete-heartcomplete-intention), means ‘wholeheartedly’.

Classicality (经典性/历史性) Zhou Jian (1997) claims ‘Idioms have the features of primitive simplicity and elegance, i.e. classicality, and this characteristic of idioms was acquired in history’. Ma (1962: 15) and the Dictionary of Chinese Idioms (Taiwan) (2010) hold similar criterion. Needless to say, the idioms Zhou categorizes are accepted by all, professionals and nonprofessionals. However, as Shi (1979: 27) points out, many idioms taken from what were esteemed as classic works by later times were actually common sayings at the time. For instance, 亡羊补牢 (wangyang-bulao, lost-sheep-mend-fold, i.e. it’s never too late, better late than never) is from Stratagems of the Warring States. ‘The proverb goes, it’s not late to call a hunting dog after you have seen a hare; and it is not slow to mend the sheepfold after you have lost sheep.’

5.  Sources of Chinese idioms: from allusions, quotations, and folk sayings There are two primary sources of Chinese idioms: written literacy and folk sayings. Written literacy includes fables, historical events, and quotations (Z. Zhou 1955). The Dictionary of Chinese Idioms (Taiwan) (2010) collects idioms from three resources: 典出 (condensed from allusions and stories), 语出 (from direct quotations), and 语本 (adapted from a phrase or a line). For example: 狐假虎威 (hujia-huwei, fox-borrow-tiger-strength, i.e. to rely on powerful connection with one person to bully another person) is a fable from vol. 1 of Stratagems of Chu in Stratagems of the Warring States. 草木皆兵 (caomu-jiebing, grass-tree-all-soldier, i.e. frightened, panic-stricken) derives from a historical event in 383, the Battle of Fei River, in which the emperor of Former Qin was defeated and as he fled he thought all the bushes and trees on the mountain looked like enemy soldiers. 曲突徙薪 (qutu-xixin) and 焦头烂额 (jiaotou-lan’e) are from the line 曲突徙薪亡恩泽,焦头烂额为上客 (the man who suggested that the homeowner bend the chimney and remove the firewood from near the hearth to prevent a possible fire was not well treated, but the man who was badly hurt in fighting the fire was well treated) in the Biography of Huo Guang in the History of the Former Han. 怨天尤人 (yuantian-youren, complain-heaven-blame-others, i.e. blame everybody but oneself) is from a line in the Analects, 不怨天,不尤人 (no-complain-heaven, no-blame-others). Most researchers regard idioms that derive from literary resources highly, but Shi (1979: 30) claims the primary source of idioms is from common folk sayings. He even predicts that the next hundred years will bring a vast number of new idioms to the language (Shi 1979: 540). Recently statistics on idioms in specific books are increasingly rich. Sun (1989: 91–3) collects basic statistics of idioms distributed in the Confucian classics and major individual works of history, philosophy, and literature. For example, 101 idioms originate from Mencius, 91 from Zhuangzi, and 69 from works of Su Shi (1037–1101). The Dictionary of Chinese Idioms, Taiwan (2010) has more detailed statistics. The number of graduate and PhD theses on the subject has increased over the last 20 years. For example, the title of Zeng Xiangling’s 69

Jiao Liwei

PhD thesis is〈《诗经》成语研究〉(‘A Study of Chinese Idioms in the Book of Songs’) (Zeng 2011). Other theses examine idioms in The New Text Book of Documents, The Classic of Filial Piety, The Chronicle of Zuo, Stratagems of the Warring States, Records of the Great Historian, Book of Han, The Records of the Three Kingdoms, The Works of Yang Xiong, etc. Other books focus on the subject matter of idioms, such as the military, plant names, architecture, religion, etc. For example, a book named Buddhism Idioms has about 400 entries (R. Zhu 2006).

6.  Chinese idioms and rhetoric: figure of speech, idiomaticity, and idiomatic salience The use of idioms can serve to make speech or writing concise and vivid (Z. Zhou 1955; Ma 1962: 40–5). Shi (1979:124–5) describes 25 major figures of speech exploited in idioms. Harmony and symmetry in sound, structure, prosody, and semantics are extensively studied (Wu 1995; Z. Liu and Xing 2000, 2003; Z. Liu 2004). 非常聪明 (feichang congming, very smart) and 花钱浪费 (huaqian langfei, spend money lavishly) do not qualify as idioms, but 冰雪聪明 (bingxue-congming, ice-snow-smart, i.e. extremely intelligent) and 一掷千金 (yizhi-qianjin, one-throw-thousand-gold, i.e. to spend money recklessly or extravagantly) do qualify since 冰雪 is used metaphorically and 一掷千金 in a manner of exaggeration. Therefore, Chinese idioms are essentially figures of speech. Virtually no idioms appear in laws and regulations, nor on product manuals or specifications, because idioms cannot convey accurate meanings objectively. Eloquent lawyers, however, tend to have great command of idioms and proverbs. There is also an annoyingly large number of advertisements which abuse the use of Chinese idioms in Chinese media, in an attempt to impress their audiences. Once an idiom is used in a sentence, others are likely to follow.3 For example, after Steve Jobs, CEO of Apple Inc., died in October 2011, Al Gore, former vice president of the USA commended him in a TV interview. In the Chinese translation, there are five idioms in merely 200 characters, the first four of which convey praise for Jobs, 独一无二 (duyi-wu’er, unique, the only one of its kind), 举世无双 (jushi-wushuang, unrivalled in the whole world), 绝无 仅有 (juewu-jinyou, one and only), 非同凡响 (feitongfanxiang, uncommon).4 千变万化 (qianbian-wanhua) is a very common idiom which expresses the notion that changes are numerous. However, almost no one except lexicographers knows that this idiom is from Liezi and was later used in Zhuangzi and Records of the Grand Historian. When this idiom is mentioned, no one associates it with the above-mentioned famous works, although some people might associate it with the Monkey King. What qualifies this phrase as an idiom then? What is meant by the idiomaticity of Chinese idioms? The meaning of the above-mentioned idiom, 守株待兔 (shouzhu-daitu, waiting passively for a windfall), was enhanced by the fable 守株待兔, which describes the stupid peasant who abandoned farming to spend each day waiting for hares to crash into the same tree stump. The structure ‘千X万X’ intensifies the degree of ‘change’ which is embodied on 变 and 化 in the idiom 千变万化. The 冰雪 (ice and snow) is a metaphor to describe the smartness of a girl or woman. The meaning of ‘uniqueness’ of 独一无二 is intensified by the juxtaposition of two similar segments, 独一 (only one) and 无二 (no any other). The idiomaticity of 为所欲为 (weisuoyuwei) is embodied in three aspects, first, the two 为 are written Chinese and its corresponding colloquial word is 做; second, 所欲为 (what one wants to do) follows the structure of a Classical Chinese noun phrase; third, the phonetic structure 为所||欲为 differs from the grammatical structure 为||所欲为. Another example, 扬长 70

Chinese idioms

(yangchang), probably an erroneous form of 佯长, is a word of assonance, and it vividly describes the manner of 去 (to leave) in the idiom 扬长而去. The 而 in this idiom is a Classical Chinese function word. All the examples mentioned above show some structural, semantic, or other linguistic salience. Below are categories and types of Chinese idioms, and how their unique forms contribute to their idiomaticity.

7.  Types of Chinese idioms: five categories and 15 types Based on idiomatic salience, Chinese idioms can be divided into five categories and 15 types.

Category 1: Externally salient, for example, 狐假虎威 (hujia-huwei), 围魏救赵 (weiWei-jiuZhao), and 螳螂捕蝉,黄雀在后 (tanglang-buchan, huangque-zaihou) Type 1: originated from fables and myths, for example 狐假虎威 (hujia-huwei), 黔驴技穷 (Qianlü-jiqiong), and 八仙过海 (Baxian-guohai) Chinese idioms originated from fables are unanimously accepted by all people, professionals and common folk alike. Even though the literature of Chinese fables is not particularly rich, there are a considerable number of fables scattered in the literature of ancient philosophers, politicians, historians, and Buddhist masters. Idioms from fables in the pre-Qin period are most famous. Below are some samples with the works from which they originated. 愚公移山 (Yugong-yishan), Liezi; 揠苗助长 (yamiao-zhuzhang), Mencius; 朝三暮四 (zhaosan-musi), Zhuangzi; 刻舟求剑 (kezhou-qiujian), Lüshi Chunqiu; 滥竽充数 (lanyu-chongshu), Han Feizi; 塞翁失马 (Saiweng-shima), Huainanzi; 画蛇添足 (huashe-tianzu), Stratagems of the Warring States; 杯弓蛇影 (beigong-sheying), Fengsu Tongyi; 黔驴技穷 (Qianlü-jiqiong), Three Admonitions by Liu Zongyuan. In ancient times, culture was undeveloped and channels of information were largely obstructed, so fables became an important tool for analogical analysis. Idioms originated from fables are mainly from between the time of Mencius and Zhuangzi and the time of Stratagems of the Warring States. Later this kind of idiom became scarce, the last famous one perhaps being 黔驴技穷 of the Tang dynasty. This kind of idiom gives people a striking impression since they are associated with long and iconic stories. The four characters of an idiom cannot provide its true meaning since it comes from the associated fable. This is even truer for learners of Chinese. For example, the literal meaning of 黔驴技穷 is ‘Guizhou Province-donkey-skill-exhausted’, while its true meaning in English is ‘to exhaust one’s (limited) bag of tricks’. The English translation does not have strong negative connotations, but the idiom is highly derogatory in Chinese. The negative connotation does not come from 技穷 (tricks exhausted) but from the story of the whole fable. Textbook compilers in China highly value this kind of idiom. They have been incorporated in textbooks, and are sure to be passed down from generation to generation. Many books and dictionaries include almost all four-character fairy tales or myths as idioms, while actually only very few of them are qualified as such. Myths such as 女娲补天 (Nüwa butian), 夸父追日 (Kuafu zhuri), 后羿射日 (Houyi sheri), and 嫦娥奔月 (Chang’e benyue) are not idioms, while 精卫填海 (Jingwei-tianhai), 牛郎织女 (Niulang-Zhinü), 八仙过海 (Baxian-guohai), and a few dozen others are among the true idioms originating from fairy tales. The fairy tale of 嫦娥奔月 first appeared in Huainanzi. The author, Duke of Huainan (179–122 bc), used this myth to mock those politicians, such as Shen Buhai (385–337 bc), 71

Jiao Liwei

Shang Yang (395?–338 bc), and Han Fei (281–233 poorly. The myth goes like this:

bc)

who administered their countries

Yi pleaded and received the elixir of life from Xi Wangmu (Queen Mother of the West), but his wife, Chang E stole the elixir and used it for herself. Then she flew to the moon but once there, regretted the fact that she had no more elixir to fly back to the earth. She did not know how and where to obtain the elixir. In this sense, borrowing fire is not as good as learning how to drill wood to make fire, and getting water from your neighbor is not as good as digging a well. This fairy tale had first taken on the metaphorical meaning of a similar idiom, 舍本逐末 (sheben-zhumo, to attend to trifles/surface but neglect essentials). However, due to the fact that Chinese people worship the moon and its beauty, and Chinese writers have inexplicable zest for the sense of loneliness, the tale later evolved to be sympathetic to a lonely beauty and lost its original implication completely. Now 嫦娥奔月 is used only as a proper noun, not as a general word or an idiom.

Type 2: originated from historical events or stories, for example 负荆请罪 (fujing-qingzui), 胯下之辱 (kuaxiazhiru), and 望梅止渴 (wangmei-zhike) China has a long and rich history from which many important historical events and stories were later coined as idioms, such as 负荆请罪 (fujing-qingzui), 围魏救赵 (weiWei-jiuZhao), 纸上谈兵 (zhishangtanbing), 鸡鸣狗盗 (jiming-goudao), 四面楚歌 (simian-Chuge), 破釜沉舟 (pofu-chenzhou), 毛遂自荐 (Maosui-zijian), 乐不思蜀 (lebusiShu), 东山再起 (Dongshan-zaiqi), 草木皆兵 (caomu-jiebing), 请君入瓮 (qingjunruweng), and 东窗事发 (Dongchuang-shifa). These idioms are associated with certain historical figures and events in the minds of educated Chinese. For example, Chinese associate the idiom 负荆请罪 with two historical figures: General Lian Po (327–243 bc) and a politician named Lin Xiangru (329–259 bc), their conflicts and later reconciliation, and maybe the image of a powerful general, with twigs of chaste tree on his back, sincerely pleading guilty to a slim-figured politician. Many idioms originated from historical stories. Among historical figures, Han Xin, Xiang Yu, Zhuge Liang, Cao Cao, and Cao Zhi are associated with many idioms. Han Xin (231–196 bc) was a marshal and strategist. He suffered many humiliations before becoming famous, and his later life was full of twists and turns; eventually he came to an unexpected end. Idioms associated with Han Xin are quite popular among the Chinese people. Examples include: 胯下之辱 (kuaxiazhiru), 一饭千金 (yifan-qianjin), 暗渡陈仓 (andu-Chencang) and 多多益善 (duoduoyishan). The story of 胯下之辱 (hip-under-’s-humiliation) in Records of the Grand Historian goes like this: In the Qin dynasty a butcher in Huaiyin county insulted Han Xin by saying, ‘Even though you are tall and like to carry a sword with you, you are actually a coward.’ He publicly humiliated Han Xin and said further, ‘If you have guts, then stab me; if you don’t, crawl under my legs.’ Han Xin stared at him for a long while, then knelt down and crawled under the butcher’s legs. The crowd all laughed at Han Xin, judging him to be timid. Thus, the meaning of 胯下之辱, extreme and unbearable humiliation, is embodied in this shocking story. Some idioms stem from anecdotes about historical figures, such as 高山流水 (gaoshan-liushui), 望梅止渴 (wangmei-zhike), 三顾茅庐 (sangumaolu), 画龙点睛 (hualong-dianjing), 妙笔生花 (miaobi-shenghua), 凿壁偷光 (zaobi-touguang), 黄袍 加身 (huapao-jiashen), 屡败屡战 (lübai-lüzhan), etc. 72

Chinese idioms

Type 3: Eight-character idioms, for example: 螳螂捕蝉, 黄雀在后 (tanglang-buchan, huangque-zaihou); 近朱者赤, 近墨者黑 (jinzhuzhechi, jinmozhehei); and 翻手为云, 覆手为雨 (fanshou-weiyun, fushou-weiyu) There are scores of Chinese idioms composed of eight characters, such as 螳螂捕蝉,黄雀 在后; 鹬蚌相争,渔翁得利; 欲加之罪,何患无辞; 智者千虑,必有一失; 仁者见仁,智者 见智; and 一人得道,鸡犬升天, etc. All eight-character idioms can be parsed into two groups of four. The meanings of the two groups may be similar, for example, 无源之水,无本之木 (wuyuanzhishui, wubenzhimu, without-source-’s-water, without-root-’s-tree, i.e. things without solid foundation); or successive, for example, 一人得道,鸡犬升天 (yirendedao, jiquanshengtian, one-person-attain-Tao, chicken-dog-ascend-heaven, i.e. when a man gets to the top, all his relations and friends get there with him); or complementary, for example, 仁者见仁,智者见智 (renzhejianren, zhizhejianzhi, benevolent-man-see-benevolence, wise-man-see-wisdom, i.e. different people have different views); or contrastive, for example, 道高一尺,魔高一丈 (daogaoyichi, mogao­ yizhang, virtue-rise-one-foot, vice-rise-one-zhang (10 feet), i.e. the force of evil always manages to beat the force of law). The meaning of the whole idiom is enhanced by the semantic relations of its two parts. Sometimes only one part of the eight-character idioms is mentioned, such as 近朱者赤(近墨者黑) (jinzhuzhechi, jinmozhehei, approximate tocinnabar-person-red, approximate to-ink-person-black, i.e. different environments provide different influences), and 兼听则明(偏信则暗) (jiantingzeming, pianxinze’an, both-listenthen-enlightened, biased-heed-then-benighted, i.e. a clear head comes from an open mind). Furthermore, some eight-character idioms can be shortened into four characters, such as 见仁见智 rather than 仁者见仁,智者见智, and 翻云覆雨 rather than 翻手为云,覆手为雨 (turn up-hand-make-cloud, turn down-hand-make-rain, i.e. to play tricks). The three types of Chinese idioms described above can create a strong impression on learners, either because of their vivid semantics or because of their length and rhythm. Learners have to learn such idioms one by one, but once they are acquired, they are hard to forget.

Category 2: Structural salient or overall symmetrical, for example: 酸甜苦辣 (suantiankula), 千变万化 (qianbian-wanhua), and 根深柢固 (genshen-digu) Type 4: Juxtaposition of four characters, like 酸甜苦辣 (suantiankula), 之乎者也 (zhihuzheye), and 青红皂白 (qinghongzaobai) There are approximately two dozen idioms that juxtapose four examplars of a generic notion such as taste or color to create metaphorical meaning. Thus, the four tastes, 酸 (suan, sour), 甜 (tian, sweet), 苦 (ku, bitter), and 辣 (la, spicy) combine to form an idiom that signifies the ‘joys and sorrows of life’. Another example: 声 (sheng, sound), 色 (se, beauty), 犬 (quan, dog), and 马 (ma, horse) imply, individually, ‘going to the theatre, visiting brothels, keeping dogs, and riding horses’. Together, they form a phrase that signifies ‘sensual pleasure’. Another idiom consists of four common Classical Chinese particles: 之 (zhi, a possessive marker), 乎 (hu, an interrogative particle), 者 (zhe, a noun marker), and 也 (ye, an affirmative particle.) Together they signify ‘obscure literary jargon’. Other idioms of this kind include 悲欢离合, 博大精深, 古今中外, 生老病死, 喜怒哀乐, 青红皂白, 老弱病残, 光明正大, 是非曲直, etc. Juxtaposition of four similar, contrastive, or related words enhances their rhetorical effect. 73

Jiao Liwei

Type 5: With crossly corresponding structures, for example: 千变万化 (qianbian-wanhua), 惊天动地 (jingtian-dongdi), and 七嘴八舌 (qizui-bashe) In an idiom where the four compositional characters are represented as ABCD, characters AC or BD form a pair of synonyms, antonyms or related words, such as 变/化 in 千变万化 (thousand-transform-ten thousand-change, ever-changing), 惊/动 in 惊天动地 (shock-heavenshake-earth, earthshaking), 嘴/舌 in 七嘴八舌 (seven-mouth-eight-tongue, all talking in confusion). One pair serves as the foundation of the basic or transferred meaning of an idiom, while the other pair, for example, 千/万, 天/地, or 七/八, modifies the pair representing the basic meaning of the idiom. Take the example of 千变万化 which means ‘numerous changes’, 变 and 化 signify ‘changes’; ‘千X万X’ (literally ‘thousand X and ten thousand X’) signifies ‘a large number of’. There are about 60 idioms with the structure of ‘千X万X’, and almost all these idioms have the implication of ‘numerous in quality’ or ‘high in degree’. ‘天X地X’, combining with ‘X天X地’, is the most prolific structure for Chinese idioms. There are more than 180 idioms with this structure, and at least three dozen of them are very common. For example, 冰天雪地 (bingtian-xuedi, ice-heaven-snow-earth, i.e. bitterly cold), 欢天喜地 (huantian-xidi, merry-heaven-happy-earth, i.e. completely overjoyed), 天崩地裂 (tianbeng-dilie, heaven-fall-earth-rend, i.e. deafening (sound)), 天长地久 (tianchang-dijiu, heaven-long-earth-forever, i.e. everlasting), etc. Almost all of the idioms have implications of ‘extremely’ or ‘to a high degree’. Other structures have inscrutable connotations, for instance, ‘七X八X’ and ‘X七X八’ suggest some kind of ‘chaos’. There are a great number of fixed structures like ‘天X地X’ or ‘X七X八’ in Chinese. Zhang Weiguo (1989) describes about 70 structures. Jiao et al. (2011: 298–300) and Jiao (2012) describe 21 and 37 of the most common structures. Based on statistics of texts totaling 65 million characters, Jiao (2012) finds 8,844 four-character idioms among which 2,086 structures are based only on either the first and third characters, or the second and fourth characters. In the 8,844 idioms, 4,949 (55.92%) of them have at least one shared structure. The top 107 most common structures generate 1,291 idioms, about 14.46% of all idioms. This shows that Chinese idioms have a strong tendency to be paradigmatic. Most of the structures have semantic implications. Further observation, description, and explanation of the semantic features of these structures are highly valuable (Z. Zhou 1955). Combinations of numerals in idioms were thoroughly accounted (Han 1984). A distinguishing feature of fixed structures is that the two characters in a structure are either synonyms, near synonyms, antonyms, or closely related, and most form a word, for example, 天地 (heaven-earth), 千万 (thousand-ten thousand), 东西 (east-west), 头尾 (headtail), 云雾 (cloud-fog); meanwhile, the other two characters of an idiom are usually synonyms or antonyms, for example, 天翻地覆 (heaven-upturn-earth-overturn, extremely upheaval), 千军万马 (thousand-soldier-ten thousand-horse, a powerful army), 东奔西走 (east-dashwest-run, bustle about), 街头巷尾 (street-head-lane-tail, streets and lanes), and 腾云驾雾 (rise-cloud-ride-fog, travel through space, feel giddy). In this way, the two non-structural characters form the fundamental meaning of the idiom, while the structure modifies the fundamental meaning. For example, the fundamental meaning of 街头巷尾 derives from 街 (street) and 巷 (lane), and since the structure of 头 (head)/尾 (tail) represents ‘whole’ or ‘complete’, the whole idiom means ‘entire streets’ or ‘all streets and lanes’.

74

Chinese idioms

Type 6: With antithetical structure, for example: 根深柢固 (genshen-digu), 居安思危 (ju’an-siwei), and 应有尽有 (yingyou-jinyou) The defining characteristic of this structure is that characters A and B are either contrastive or synonymous to characters C and D. For example, in 根深柢固 (genshen-digu), 根深 literally means ‘secondary roots are deep’ and 柢固 literally means ‘primary root is stable’. The two parts are complementary, thus the meaning of the whole idiom is ‘the foundation is firm and unshakable’. Another example, in 居安思危 ( ju’an-siwei), 居安 means ‘to stay at a safe place’ and 思危 means ‘be vigilant against danger’. The two parts are contrastive and conjunctive, meaning ‘even if you stayed at a safe place, you still need to keep vigilant against danger’ when combined. Other idioms of this type include 聚精会神, 发号施令, 方兴未艾, 专心致志, 众叛亲离, 卖官鬻爵, 出类拔萃, 落井下石, and 感恩戴德 etc. There is a special form in this type of idiom: character A and C, or B and D are the same. For example, 我行我素 (woxing-wosu, I-do-I-usually do, i.e. stick to one’s own way of doing things), 顺风顺水 (shunfeng-shunshui, down-wind-down-stream, i.e. very smooth), 应有尽有 (should-have-all-have, i.e. lacking nothing, complete), and 心服口服 (xinfu-koufu, heart-convinced-mouth-convinced, i.e. be sincerely convinced), etc.

Category 3: Partially symmetric, for example: 林林总总 (linlin-zongzong), 比比皆是 (bibi-jieshi), and 未雨绸缪 (weiyu-choumou) The defining characteristic of idioms in this category is one part of the idiom is either phonetically or semantically salient.

Type 7: AABB idioms, for example: 林林总总 (linlin-zongzong), 风风雨雨 (fengfeng-yuyu), and 洋洋洒洒 (yangyang-sasa) As can be easily seen, this type of idiom consists of two parts, each of which repeats a character, for example, 林林||总总 (linlin-zongzong) and 风风||雨雨 (fengfeng-yuyu), etc. Some idioms of this type are the result of reduplication of a word AB. For example, 风风 雨雨 is the reduplication of 风雨, 浩浩荡荡 of 浩荡, 三三两两 of 三两, 朝朝暮暮 of 朝暮, 鬼鬼祟祟 of 鬼祟, etc. The basic meaning of 风风雨雨 is from 风雨 (wind-rain), which is a metaphor of ‘hardship’, and the meaning of the whole idiom is ‘many hardships’. Other idioms of this type do not result from reduplication of a word, but the combination of two words, AA and BB. For example, the meaning of 林林总总 is a combination of 林林 (numerous) and 总总 (numerous and chaos, large in quantity and various in category). This subtype of idiom also includes 浑浑噩噩, 洋洋洒洒, 熙熙攘攘, 兢兢业业, 战战兢兢, 卿卿我我, etc.

Type 8: AABC or ABCC idioms, for example: 比比皆是 (bibi-jieshi), 大名鼎鼎 (daming-dingding), and 蒸蒸日上 (zhengzheng-rishang) The defining characteristic of this type of idiom is that the first two or last two characters are the same, and are usually adjectives or adverbials. For example, in the idiom 比比皆是 (bibi-jieshi), 皆是 means ‘all are’, while 比比 means ‘one after another’, which enhances the meaning of 皆是. Another example, in 大名鼎鼎 (daming-dingding), 大名 means ‘big 75

Jiao Liwei

name’, while 鼎鼎 means ‘splendid, celebrated’, which enhances 大名. Other AABC idioms include 彬彬有礼, 沾沾自喜, and 蒸蒸日上, etc. Other ABCC idioms include 气喘吁吁, 千里迢迢, 风尘仆仆, and 生气勃勃, etc. Some ABCC idioms can be reversed to CCAB because character B is usually a noun. For example, 大名鼎鼎 and 鼎鼎大名 coexist, but ABCC is the more common form. AABC cannot be reversed as BCAA because AA is usually an adjective or adverbial which modifies the nucleus BC.

Type 9: Alliteration, assonance/rhyming, or juxtaposition of synonyms or antonyms, for example: 淋漓尽致 (linli-jinzhi), 未雨绸缪 (weiyu-choumou), 旗鼓相当 (qigu-xiangdang), and 名副其实 (mingfu-qishi) The defining characteristic of this type of idiom is that a part of the idiom is alliterative or assonant in sound, or synonymous or antonymous in meaning, so as to enhance the rhetorical result. For example, Alliteration: 踌躇满志 (chouchu), 光明磊落 (leiluo), 金碧辉煌 (huihuang) Assonance: 精神抖擞 (dousou), 五彩缤纷 (binfen), 逍遥法外 (xiaoyao) Juxtaposition of synonyms or antonyms: 德才兼备 (integrity-competence-both-possess), 不见经传 (not-seen-classic-commentary), 哭笑不得 (cry-laugh-not-appropriate), 乐极生悲 (joy-extreme-beget-sorrow). Alliteration or assonance has an immediate effect on the impression of the sound of an idiom, because the structure of Chinese syllables is relatively simple, consisting of only two parts, namely initial and final. If two consecutive syllables have half of their constituents alliterated or rhymed, the repetitive sound leaves a strong impression upon the hearer. Antithesis was a universal feature of Ancient Chinese verse, and remains prominent to this day. The coordination of two antithetical characters falls mainly into three groups. The first way is the use of antonyms, such as 朝夕 (morning-night) in 朝夕相处, and 名 (name) and 实 (fact) in 名副其实. The second way is the use of synonyms or near synonyms, such as 脍炙 (sliced meat-barbequed meat) in 脍炙人口. The third way is to use specific terms (co-hyponyms) to represent generic ones (hypernyms), such as 鸡犬 (chicken-dog) in 鸡犬 不宁, where 鸡 and 犬 are hyponyms of ‘livestock’. This subtype of idiom is categorized based on semantics, however, there are some functional words in the place of the third character of an idiom (XX[ ]X) which can serve as a marker of this type of idiom, for example: 相 (each other) in 旗鼓相当 (banner-drum) and 萍水相逢 (duckweed-water); 交 (simultaneously) in 饥寒交迫 (hungry-cold) and 心力交瘁 (heart-strength); 双 (both) in 名利双收 (fame-wealth); 两 (double) in 进退两难 (advance-retreat); 俱 (all) in 玉石俱焚 ( jade-stone); 皆 (all) in 啼笑皆非 (cry-laugh); 兼 (both) in 软硬兼施 (soft-hard); 并 (same) in 声情并茂 (voice-expression); 全 (all) in 文武全才 (literary-military); 共 (together) in 雅俗共赏 (elegant-vulgar); 齐 (together) in 金鼓齐鸣 (bell-drum). 都 is the most common adverb, meaning ‘both/all’, however, it does appear in this situation. It is worth noting that as the Chinese language evolved with time, its pronunciation probably underwent the greatest change. Some characters rhymed or alliterated in ancient times, but do not any more, and vice versa. Take 沆瀣 (hang-xie) in 沆瀣一气 for instance; they alliterated in ancient times, but do not in Modern Chinese. Another example, 身心 (shen-xin) in 身心交瘁 rhymes in Modern Chinese, but not in Ancient Chinese because 身 ended with ‘-n’ while 心 ended with ‘-m’. 76

Chinese idioms

Category 4: Partially marked/locally salient, for example: 视死如归 (shisi-rugui), 恍然大悟 (huangran-dawu), 刻不容缓 (keburonghuan), and 言而无信 (yan’erwuxin) Idioms in this category do not have salient allusions as in Category 1, salient parallel structures as in Category 2, or salient phonetic forms as in Category 3, but some locally salient words distinguish them from other idioms.

Type 10: Idioms with simile marker 犹, 若, 如, 似, for example: 记忆犹新 (jiyi-youxin), 呆若木鸡 (dairuomuji), and 如花似玉 (ruhua-siyu) Many idioms have figurative meanings, however, no idioms contain the most common simile marker 象/像 (xiang) ‘like, similar to’, which has been used to indicate similarity since the Ming dynasty (1368–). The most common simile markers in Chinese idioms include 犹 (you), 若 (ruo), 如 (ru), and 似 (si). Idioms with 如 include 如日中天 (like-sun-middle-sky, at the peak of one’s power or career), 易如反掌 (easy-as-turn-palm, piece of cake), and 视死如归 (view-death-as-returning home, face death with equanimity), etc. Idioms with 若 are not few in number, for example 呆若木鸡 (stunned-like-wood-chicken, dumbstruck), 口若悬河 (mouth-like-hanging-river, eloquent). Few idioms use 似: 繁花似锦 (flourishing-flower-like-brocade) and 归心似箭 (returning-heart-like-arrow) are the two most commonly used. Figurative meaning is vivid in this kind of idiom. If there are two simile markers in one idiom, the pattern is most likely to be ‘如X似X’, such as 如花似玉 (like-flower-like-jade), 如饥似渴 (as-hungrily-as-thirstily). ‘似X如X’ is scarce. There might only be one example, 似醉如痴 (as-drunk-as-obsessed); however, it takes a more common form in 如醉如痴. No idioms with the structures 若/似 or 如/若 have been found. Zhu Guanming (2000) describes all of the major monosyllabic simile markers in Ancient Chinese. The simile marker 犹 appeared early and disappeared early too, so there are few idioms that include it, for example 过犹不及 (excessive-as-not-enough, going beyond is as wrong as falling short), 记忆犹新 (memory-like-fresh, remember vividly), and 虽死犹生 (although-dead-like-live, live on in spirit). Since 如 and 似 coexisted for the most time, and 如 appeared earlier than 似, there are many idioms with the structure ‘如X似X’. These phenomena show that idioms were created roughly at the same pace that the Chinese language evolved in history.

Type 11: Idioms with 然, for example: 恍然大悟 (huangran-dawu), 井然有序 (jingran-youxu), and 春意盎然 (chunyi-angran) 然 can function as a suffix of certain adjectives and adverbs indicating a state of affairs. The character 然 and its preceding adjective or adverb make an idiom vibrant and lively. One form of this type is ‘X然XX’, for example, 恍然大悟 (huangran-dawu, sudden--ly-greatlyrealized, suddenly see the light) and 井然有序 (jingran-youxu, neat--ly-have-order, in good order). The third and fourth characters of this subtype of idiom are usually an adjectival phrase or verbal phrase, so the first character plus ‘然’ serves as attributive or adverbial. The other form of this type is ‘XXX然’, for example, 一目了然 (one-glanceunderstandable--ly, to understand fully at one glance) and 春意盎然 (spring-sign-abundant-ly, full of spring atmosphere). The first two characters of this subtype of idiom are most often nouns, and the second part of this subtype of idiom is a description of the nouns. 77

Jiao Liwei

Type 12: ‘X不XX’ and ‘X而XX’ idioms, for example: 刻不容缓 (keburonghuan), 言而无信 (yan’erwuxin), and 取而代之 (qu’erdaizhi) A large number of idioms belong to this category. The defining characteristic of this type of idiom is the discordance of 不 or 而 on syntactical and phonological levels. Syntactically the structure is ‘X||不XX’, ‘X||而XX’ or ‘X而X||X’; however, phonologically the structure is ‘X不||XX’ or ‘X而||XX’. This discordance makes such idioms very special, thus enhancing the effect. For example, 刻||不容缓 (moment-not-allow-delay), 马||不停蹄 (horse-not-stopgallop), 言||而无信 (speak-but-no-credit), 锲||而不舍 (carve-but-not-give up), and 取而代||之 (take-and-replace-it). Due to the special function of 不 and 而 in the formation of idioms, all expressions with the structure ‘不X而X’ tend to be classed as idioms. For example 不辞而别 (not-farewellbut-leave), 不寒而栗 (not-cold-but-shiver), 不约而同 (not-arrange-but-same), etc.

Category 5: Miscellaneous, for example: 不亦乐乎 (buyilehu), 惟利是图 (weilishitu), and 自告奋勇 (zigao-fenyong) Some idioms do not have any obvious marker, but there are unique reasons for their existence.

Type 13: From famous quotations, like 不亦乐乎 (buyilehu), 一尘不染 (yichen-buran), and 水落石出 (shuiluo-shichu) Zhou Jian (1997) claims that only idioms that originate from famous writers or famous works of literature are true idioms. To put it another way, they possess ‘classicality’. This is mostly true since in ancient times, only those famous works got much readership due to limited printing medium. The website of Chinese Idiom in Taiwan has statistics on the origins of idioms from major literary works, such as from the Confucian classics, history, philosophy, and literature. For example, 46 idioms originated from the Book of Odes, 113 from Records of the Grant Historian, 53 from Zhuangzi, and 24 from the works of Su Shi.5 Take the Analects for example. The Chinese Imperial Examinations existed for almost 1,300 years, from 607 to 1905. These examinations focus on the Confucian classics. The Analects is the first and foremost volume to be tested, so it was naturally the most popular of the classics among intellectuals and literati. Lines from the Analects were particularly prone to be coined as idioms. For example, a phrase in Confucius’s comment on Kong Wenzi, 不耻下问 (don’t feel ashamed to ask and learn from one’s subordinate) became an idiom. Confucius’s own ideology, such as 见义勇为 (to stand up for what is right), 当仁不让 (to not shirk responsibilities), and 过犹不及 (going beyond is as wrong as falling short) were revered as idioms. Even some very common expressions at the time, such as 不亦乐乎 (isn’t it a great pleasure?), had become revered idioms with a new meaning of ‘extremely, awfully’. Buddhism is a very influential religion in China, especially in secular culture. There are hundreds of idioms that originate from Buddhist works, for example 一尘不染 (one-dustnot-polluted, maintain one’s original pure character), 三生有幸 (three-incarnation-have-luck, stroke of luck), 天花乱坠 (sky-flower-chaos-drop, extravagantly colorful description). Buddhists believe that ‘color, sound, smell, taste, touch, object’ pollute people’s ‘eye, ear, nose, tongue, body, and mind’, so they are called ‘the six dusts’. 一尘不染 means ‘free from any speck of dust’, ‘spotless’ or ‘to maintain one’s original pure character’. 78

Chinese idioms

Ouyang Xiu (1007–1072) and Su Shi (1037–1101) were prominent writers in the Song dynasty, and their works were very popular at the time. Since 水落而石出者 and 水落石出 appeared in their prose An Account of the Old Toper’s Pavilion and Second Ode to The Red Cliff in 1046 and 1082 respectively, 水落石出 was coined as an idiom meaning ‘the truth is fully revealed’. It is worth noting that all the idioms mentioned above in this section have notable origins which helped them gain fixed form and popularity, but later generations, however, forgot the glorious origins of these idioms, and even transformed their meanings to be more close to literal meanings or new ones. Ordinary people do not know that 一尘不染 originates from Buddhism at all, or that 水落石出 was first written by Su Shi. They even changed the meaning of 空空如也 from ‘shallow (used in a self-deprecating way by Confucius)’ to the more literal ‘all empty’.

Type 14: With residue of Classical Chinese grammar or lexis, for example 唯利是图 (weilishitu), 不刊之论 (bukanzhilun), and 无能为力 (wunengweili) Some idioms have preserved the peculiar grammar of Classical Chinese, for example, the fronting of an object like 利 (profit) by putting it in front of the verb 图 (to seek) in the idiom 唯利是图 (only-profit-do-seek, to seek nothing but profit). Other idioms preserved flexible usage. For example, in 有口皆碑 (be universally acclaimed) 碑 is used as the verb ‘to commend’ rather than its original meaning of ‘stone tablet’. Some idioms preserved the etymological meaning of a character, for example, the original meaning of 刊 (to delete) is preserved in the idiom 不刊之论 (unalterable truth), while in Modern Chinese 刊 usually means ‘to publish’. Some characters have preserved their original meaning in idioms. For example, 别 is used mainly as the adjective ‘another’ (别的) or modal verb ‘don’t’ (不要) in Modern Chinese; however, its original meaning is ‘to differentiate, to dissect’, and it is partially preserved as ‘otherwise, differently’ in idioms such as 别出心裁 (come out with a different plan or concept), 别具一格 (have a unique or distinctive style), 别开生面 (start something new, break new ground), and 别有用心 (harbor ulterior motives). Some functional words are common in idioms, and these words help idioms maintain characteristics of Classical Chinese. For example, 莫名其妙 (its), 无能为力 (do, exert), 众望所归 (noun phrase marker), 一孔之见 (’s), 同归于尽 (at), 以卵击石 (with).

Type 15: Miscellaneous Fourteen types of idioms are elaborated above to show how their uniqueness in linguistic form or semantic connotation contributes to their idiomaticity or idiomatic salience. These types cover almost all common Chinese idioms, however, some idioms do not fit any of those patterns, so we categorize them as ‘miscellaneous’. For example: 引人入胜 (yinrenrusheng, lead-people-enter-wonderful, fascinating), 脚踏实地 (jiaota-shidi, foot-step on-solidground, conscientious and dependable, with honesty and dedication), 自告奋勇 (zigaofenyong, self-claim-effort-courage, offer to undertake a difficult or dangerous task), etc. It is obvious that some idioms belong to two or even more groups. For example, 南辕北辙 (nanyuan-beizhe, southbound-shaft-northbound-rut, i.e. to act in a way that defeats one’s purpose) belongs to both Type 1, as its original fable ‘南辕北辙’ appeared in vol. 4 of Stratagems of Wei in Stratagems of the Warring States, and Type 5, with its parallel structure ‘南X北X’. 79

Jiao Liwei

8.  Classification of Chinese idioms: by internal grammatical relationships, degree of lexicalization, and other linguistic forms Chinese idioms are counted by thousands, so it is necessary to classify them for further observation and analysis. Liu Shuxin (1990: §4.2) drew on his expertise in lexicology and divides idioms into eight types, namely those with (i) juxtaposed structures, e.g. 龙飞凤舞 (dragon-fly-phoenix-dance, elegant handwriting); (ii) modified structures, e.g. 众矢之的 (all-arrow-’s-target, a target of public criticism); (iii) complementary structures, e.g. 寄人篱下 (live-other-roof-under, depend on other people for a living); (iv) nominative structures, e.g. 拾人牙慧 (pick-others-toothdebris, plagiarize); (v) descriptive structures, e.g. 怒发冲冠 (furious-hair-spring-crown, bristle with anger); (vi) special structures, e.g. 开门见山 (open-door-see-mountain, come straight to the point), 引火烧身 (draw-fire-burn-oneself, ask for trouble); (vii) condensed structures, e.g. 种瓜得瓜 (sow-melon-get-melon, reap what one sows); and (viii) parataxis structures, e.g. 杯水车薪 (cup-water-cart-firewood, utterly inadequate). Sun (1989: 161–8) divides Chinese idioms into three types according to the degree of semantic fusion. Type 1: fused (融合性) idioms, e.g. 洛阳纸贵 (Luoyang-paper-expensive, become a bestseller) and 低三下四 (low-three-down-four, servile, humble). The meaning of this type of idiom is highly integral, and grammatical analysis is virtually impossible. Type 2: comprehensive (综合性) idioms, e.g. 按图索骥 (by-picture-seek-fine horse, do something mechanically) and 焦头烂额 (burned-head-scorched-forehead, in terrible shape, in trouble). The actual meaning of this type of idiom is extended to metaphorical interpenetration of the literal meaning. Type 3: compositional (组合性) idioms, e.g. 不期而遇 (not-expected-butmet, run into). The actual meaning of this type of idiom is basically the summarization of its constituents. This approach was developed so that five types were differentiated according to their degree of lexicalization (Qian and Yu 2003). For example, a) phrase-like idioms, e.g. 抱头鼠窜 (cover-head-mouse-flee, run helter-skelter); b) metaphorical idioms, e.g. 半途而废 (half-way-then-stop, give up halfway); c) asematic idioms, e.g. 班门弄斧 (Lu Ban-gate-play-axe, display one’s slight skill before an expert); d) agrammatical idioms, e.g. 白云苍狗 (white-cloud-black-dog, things change in an unpredictable manner); e) completely lexicalized idioms, e.g. 别风淮雨 (‘other’-wind-‘Huai’-rain, 别 and 淮 are ‘misspellings’ of 列 and 淫 which mean ‘fierce’ and ‘heavy’ respectively. The meaning of this idiom is to be ‘full of wrongly written characters’.) Tsou (2012) analyzes the internal linguistic features of idioms and enumerated eight types: a) hypernymy: 三五成群 (three-five-become-crowd, in small groups); b) classical language usage: 三年五载 (three-year-five-year, in a few years); c) culture bound: 三生有幸 (threeincarnation-have-luck, forever indebted); d) locus classicus: 朝三暮四 (morning-three-eveningfour, indecision); e) synonymy 说三道四 (say-three-call-four, mumbling insignificant things); f) word morphology 三长两短 (three-long-two-short, accident); g) homonymy (phonetic/ semantic replication or rhyme) 三三五五 (three-three-five-five, in small groups); h) antonymy 朝三暮四 (morning-three-evening-four, indecision). Most of the above classifications take into account linguistic structures; however, they are not exhaustive. The linguistic forms of idioms occur in obvious patterns. More than half of all idioms are either symmetric in semantics or harmonic in sound (Z. Liu 2004; Z. Liu and Xing 2000, 2003; C. Zhou 1988). Psychological experiments prove this. Evidence from eye-movements and event-related potentials (EPRs) show that gaze duration, total reading time, and the number of fixations of the third and fourth characters of an idiom are much shorter or less than those of the first two characters. The number of fixations of the last two 80

Chinese idioms

characters is only about 40% of that of the first two characters (Wang Yiwen et al. 2012). This proves that idioms are highly integral in semantics. Focus is not evenly distributed on the four constituent characters. This experiment also shows that gaze duration and total reading time of the second character are much longer that those of the other three characters. It shows that comprehension of an idiom is complete by the end of the second character. In other words, the first two characters lay the foundation or provide a strong hint as to the essential meaning of the whole. Supposing an educated native speaker of Chinese saw or heard 崇山 (towering-mountain), it is natural to expect the following part will be 峻岭 (lofty-ridge). The same for 夫唱 (husband-sing) and 妇随 (wife-follow), which in combination mean ‘wife being her husband’s echo’; 落井 (fall into-well) and 下石 (throw-stone) which in combination mean ‘to hit someone when he is down’; 天经 (heaven-normal) and 地义 (earth-righteousness) which in combination mean ‘entirely justified’. In rare cases where it is not possible to successfully guess the complete form of an idiom even after the second character is known, the third character will disclose everything. Statistics in the above paper show that the number of fixations of the fourth character is only about one sixth to one fifth of that of the first or the second character. During the first season of 中国成语大会 (China Idioms Conference) in 2014, the competitors usually did not need a hint of more than two characters to guess an idiom successfully. Any prerequisites for an idiom are not at all related to meaning. For instance, 惊天动地 (shock-heaven-shake-earth, earthshaking) is an idiom, so is 鸡毛蒜皮 (chicken-feathergarlic-skin, trivial); 凤毛麟角 (phoenix-feather-unicorn-horn) describes someone or something (precious and) rare, while 獐头鼠目 (river deer-head-mouse-eye) describes a repulsively ugly person. What matters most for an idiom is its rhetorical effect, which comes mainly from the salience reflected in linguistic forms. Therefore, classification of Chinese idioms should be based mainly on linguistic form, especially fixed structures or patterns.

9.  Evolution of idioms The evolution of idioms will be discussed here from four aspects, namely the shift of commentary connotation between complimentary and derogatory, competition among synonymous idioms, emergence of new idioms, and the spread of Chinese idioms to neighboring countries.

9.1.  Shift of commentary connotation between complimentary and derogatory Many idioms have commentary connotations. For example, 脱颖而出 (tuoying’erchu, talent revealing itself) has a complimentary connotation, whereas 别有用心 (bieyouyongxin, have an ulterior motive) has a derogatory connotation, and others are neutral, for example, 众所周知 (zhongsuozhouzhi, as everyone knows). Psychological experiments show that if a complimentary idiom is shown in the upper part of a screen, the mental processing time of the idiom is faster than when the idiom is shown in the lower part of the screen, and vice versa (Zhang Jijia et al. 2011). Several dozen common idioms have changed register over the time; however, the number is very small compared to the total number of idioms. Shi (1979: 222) points out that the main trend in the change of register was from complimentary to derogatory. The reason for this is that later generations were inclined to comprehend an idiom from its literal meaning without consulting its original context. For example, the origin of 难兄难弟 (nanxiong-nandi) is from Chapter 1 of Morality of Shishuo Xinyu (A New Account of the Tales of the World). 81

Jiao Liwei

A person commented on two brothers and said ‘Yuanfang 难为兄, and Jifang 难为弟’ which means the two brothers are equally excellent. The 难 here is obviously with the yangping tone (light level) and means ‘hard to be’. However, later generations take the character 难 (nan4, misfortune) at face value, and so the idiom is used to mean ‘fellow sufferers, birds of a kind’. Other examples include 浑浑噩噩, 老气横秋, 天花乱坠, 尔虞我诈, and 空空如也, etc.

9.2.  Competition among synonymous idioms Most idioms have close synonyms. If viewed from a historical perspective, this phenomenon is more obvious. Below are three pairs and one group of synonyms. In Modern Chinese, 千欢万喜 (qianhuan-wanxi, overjoyed) is not sayable though it could be well understood. However, in the PKUCCL (Peking University Center for Chinese Linguistics) corpus as of June 2013, there were 36 cases of 千欢万喜, and 74 cases of 欢天喜地 (huantian-xidi) in Ming dynasty documents. But there were only two cases of 千欢万喜 and 106 cases of 欢天喜地 in Qing dynasty documents. In the Ming dynasty short story series San Yan Er Pai (《三言二拍》), these two idioms coexisted. So this pair is an instance of what once coexisted but where one later prevailed and the other was eliminated. 班门弄斧 (Banmen-nongfu) means ‘to display one’s slight skill before an expert’ since Lu Ban was revered as the Grand Master of carpenters. This idiom originates from Liu Zongyuan’s line and Ouyang Xiu used it in its fixed form for the first time. However, there is a similar idiom 布鼓雷门 (bugu-Leimen, cloth-drum-thunder-gate, don’t hold a drum to pass by Gate Lei) which appeared in a much earlier document, from Biography of Wang Zun in the Book of Han. 雷门 is in Shaoxing, Zhejiang province. It is said that there was a giant drum on the city gate and when it was struck, the sound traveled as far as Luo Yang in Henan Province, 1,000km away. It seems that the figurative meaning of 布鼓雷门 was typical enough to qualify as an idiom. However, there was only one usage recorded in Chinese documents, which was in a drama by Wu Changling in the Yuan dynasty. Therefore, 班门弄斧 and 布鼓雷门 have the closest meaning but one was never the other’s match in prevalence. The Former Ode to the Red Cliff《前赤壁赋》, written by Su Shi (1037–1101), is one of the most famous pieces of prose in the literary history of China and all intellectuals could cite some parts of it since it includes many famous lines. One line is 寄蜉蝣于天地, 渺沧 海之一粟. 蜉蝣 means ‘mayfly’, which is an insect with a very short life span, ranging from several minutes to several days. 蜉蝣天地 (fuyou tiandi) is a perfect expression of ‘life is short’. However, there were only two usages of this idiom in ancient times, and no usage in Modern Chinese, while there emerged an idiom, 沧海一粟 (canghai-yisu) from the phrase directly next to 蜉蝣天地. 粟 means ‘millet’ which is clearly very small in comparison with the deep sea (沧海). It is a vivid expression to describe ‘how small human beings are compared to the universe’. We can see that even though two possible idioms were placed adjacent to one another, their fate was totally different. One became an idiom, and the other was kept in its original form. Besides the three pairs of competitive idioms mentioned above, we want to discuss a group of competitive idioms. There are a quite large number of idioms that mean ‘the difference/gap is large’ as seen from Table 5.1. Data is from PKU Chinese corpora. There are three idioms in the top five on the list of most common idioms in Ancient and Modern Chinese: 迥然不同, 大相径庭, and 千差万别. Two common idioms in Ancient Chinese, 天渊之别 and 不啻天渊, were replaced by 天壤之别 and 截然不同 in Modern 82

Chinese idioms Table 5.1  Competitive idioms that mean ‘the difference/gap is large’ Character to character translation 天冠地屦 天悬地隔 天差地远 天差地别 不啻天渊 天壤之判 天壤之别 天渊之别 霄壤之别 云泥之别 迥然不同 截然不同 千差万别 大相径庭

Number of cases in Ancient Chinese

heaven-hat-earth-shoe heaven-hang-earth-separate heaven-different-earth-apart heaven-different-earth-distinct not-different-heaven-lake heaven-earth-’s-distinguish heaven-earth-’s-distinct heaven-lake-’s-distinct sky-earth-’s-distinct cloud-mud-’s-distinct far--like-not-same clear cut-like-not-same thousand-different-ten thousand-distinct great-mutual-path-yard

 2  4  6  1 19  0 11 23  7  2 23  9 13 15

Number of cases in Modern Chinese 0 1 13 16 0 1 241 47 14 5 458 1,019 419 455

Chinese. The change was probably not because the character 渊 (swirling deep water) was rare and complex, since it survived in idioms such as 临渊羡鱼 and 为渊驱鱼. The structure 天/地 was probably popular at the time, so the contrast between 天 and 渊 was weakened. 啻 is more complex, so the elimination of 不啻天渊 was the result of natural selection. Although 天渊之别 was no longer common, a similar idiom 天壤之别 replaced it, so its sense is somewhat compensated. The 然 in 截然不同 helped the whole idiom deserve its idiomaticity. In brief, if an idiom, such as 蜉蝣天地, is defeated in competition against synonymous idioms in history, it is very likely that it cannot stage a comeback. Even some common Chinese idioms are facing challenges from newly emerged proverbs, for example, 趋炎附势 (quyan-fushi, to curry favor with powerful people) has been seriously challenged by 抱大腿 (bao datui, hug-big-thigh) in spoken Chinese.

9.3.  Emergence of new idioms New idioms emerge very slowly, mainly for the following two reasons. First, although new concepts emerge quickly due to the rapid development of science and technology, the need for new words focuses mainly on nouns. Chinese idioms are a kind of rhetorical device, and they are mainly applied to express Chinese people’s reflection of objects, not the objects themselves, so the emergence of new idioms lags far behind the emergence of proper nouns. For example, if one compares all the entries of the newest and the oldest editions of Cihai (辞海, lit. ‘sea of words’), one will find that there are many more entries starting with 电 (electricity or electronic) in newer editions; however, until now all the uses of 电 in idioms mean ‘lightning’, for example, in 电闪雷鸣, 风驰电掣, and 电光火石, not a single instance points to ‘electricity or electronic’. Second, there are already a large number of idioms, so demand for new idioms is weak. Given the limited memory of human beings, 8,000 general words may be able to satisfy 83

Jiao Liwei

ordinary people’s linguistic needs (词涯八千, Cheng 1998). As for the demand for idioms, perhaps a few hundred idioms would suffice. Now many idiomatic dictionaries are apt to collect more than 10,000 idioms, so these idioms have already been sufficient to satisfy the needs of ordinary people. But new idioms do emerge, albeit very slowly. Works of literature and film by famous writers and directors are the main inspiration behind new idioms. For example, 华山论剑 (Huashan-lunjian, Mount-Hua-compete-sword, discussion lead by top attendees) emerged from Louis Cha Jin Yong’s martial arts novel The Legend of the Condor Heroes (《射雕 英雄传》, 1957–9). This idiom is now widely used in summit meetings in many fields such as business, culture, and technology. 尘埃落定 (chen’ai-luoding, dust-fall-settled, to end up with) comes from the writer 阿来’s novel of the same name. 与狼共舞 (yulang-gongwu, with-wolf-together-dance, to venture) is from the famous film Dances with Wolves directed by Kevin Costner. 横空出世 (hengkong-chushi, across-sky-born, emerge out of the blue) is from a line in Mao Zedong’s poem Niannujiao Kunlun (念奴娇•昆仑) in 1935. There are many restrictions to the emergence of new idioms. Basically two conditions must be satisfied. The first is that the phrase must possess characteristics of idioms, and the second is there must be a large number of uses in a short period of time. In 2007 the incident of Zhou Zhenglong, who claimed to have taken a picture of a South China tiger but was later imprisoned for two years for swindling, was hyped by Chinese media, and many people predicted that 正龙拍虎 (Zhenglong-photograph-tiger) would become an idiom; however, it proved not to pass the test of time. 人艰不拆 (renjian-buchai, people/life-harddon’t-expose, i.e. life is already hard, don’t spoil the fun) appeared on the Internet from around 2013, and its suitability as an idiom is currently being tested.

9.4.  Borrowing and spread of Chinese idioms Idioms do not refer to objects, but are a way to creatively describe them. Simply put, idioms are a rhetorical device. Languages borrow far more ordinary words than idiomatic phrases from other languages. Chinese, for example, has borrowed hundreds of words from English and Japanese, but only a few of them are idioms. The few idioms borrowed from other languages include 以牙还牙, 以眼还眼 (yiyahuanya, yiyanhuanyan, a tooth for a tooth, an eye for an eye) and 象牙之塔 (xiangyazhita, ivory tower) from English, 火中取栗 (huozhongquli, fire-inside-pull-chestnut, i.e. be a cat’s paw, ‘tirer les marrons du feu’) from French, 一期一会 (yiqi-yihui, one-lifetime-one-chance, i.e. once in a lifetime) from Japanese. In return, English has borrowed about a few dozen words from Chinese, but none of them are idioms. Since ancient China had great influence over its neighboring countries in politics and culture throughout history, hundreds or perhaps even more than a thousand Chinese idioms were borrowed into Vietnamese, Japanese, and Korean, respectively. Vietnamese has 4,174 idioms, among them 1,760 entries are related to Chinese idioms, and 301 Chinese idioms were borrowed intact, for example ‘danh chinh ngon thuan’ (名正言顺, mingzheng-yanshun, name-proper-word-justified, i.e. perfectly justifiable) (Cai 2011). Japanese idioms are called ‘four-character compounds’, and the idiomatic youjijukugo were borrowed from Chinese, for example 単刀直入 たんとうちょくにゅう (单刀直入, dandao-zhiru, short hand-knifedirect-stab, i.e. come straight to the point). According to Moon (2005), Korean has 1,271 idioms borrowed from Chinese intact, for example 격화소양 (隔靴搔痒, gexue-saoyang, outside-boot-scratch-itch, i.e. attempt an ineffective solution). W. S.-Y. Wang (1991) studied language prefabs and habitual thought of the peoples in East Asia. 84

Chinese idioms

10.  Application of Chinese idioms Since Chinese idioms are such an important aspect of the Chinese language, they are naturally applied in many walks of lives. Below are some of their major applications.

10.1.  Idiom games The most common language game with idioms is 成语接龙 (idiom relay). The rule is a player gives an idiom, and the next player must come up with an idiom whose first character is same as the last character in the previous idiom. Then the first character of the third idiom must be the same as the last character of the second idiom, and so on. If the rules of this game are strictly enforced, the overlapping characters must be exactly the same; but if the rules are lax, the overlapping characters can be homonyms. The longest recorded idiom relay contains 1,792 idioms, starting from 胸有成竹 (xiongyouchengzhu, chest-have-conceivedbamboo, i.e. to have a well thought-out strategy) followed by 竹报平安 (zhubaoping’an, family letter-assure-safe and sound)  .  .  .  and ending with 成竹在胸 (chengzhu-zaixiong, conceived-bamboo-in-chest). However, relays with more idioms appear time and again. This game can be played without noting down the idioms and many people can join in, so it is very popular in China. Other idiom games can be found in Mo (2001: 242–55).

10.2.  Mottos of Chinese colleges When Chinese colleges were first established from the end of the nineteenth century, it happened that the status of idioms in Chinese language was rising at the same time. Therefore the mottos of many of the oldest colleges in China contain idioms which are also naturally inspiring. For example, one of the earliest mottos of Peking University was 兼容并包 (all-embrace-both-include, all-embracing); the motto of Tsinghua University contains 自强 不息 (self-improve-not-cease, to strive unremittingly); Nankai University 日新月异 (daynew-month-different, to make rapid progress); Tianjin University (originally ‘Imperial Tientsin University/Peiyang University’, the oldest university in China, founded in 1895) 实事求是 (to seek truth from facts). Many colleges chose idioms which are in accordance with the missions of their schools. For example, many teachers’ colleges set 为人师表 (to serve as a model) as their mottos. The motto of the Ocean University of China contains 海纳百川 (ocean-admit-hundred-river, the sea refuses no river), China University of Geosciences 艰苦朴素 (to live simply and work hard), China Agricultural University 博大精深 (abundantbig-refined-profound, broad and deep), China Youth University of Political Sciences 朝气蓬勃 (to be full of youthful energy). Other unique mottos include 德才兼备 (virtue-talent-bothpossess) of Beihang University and 饮水思源 (drink-water-think-source, do not forget where your happiness comes from) of Shanghai Jiao Tong University.

10.3.  Media titles Idioms are widely used in media such as titles for novels, films, TV series, albums, bands, and blogs. Not to mention those idioms with positive connotations such as 皆大欢喜 (jiedahuanxi, all-very-happy, i.e. to everyone’s delight and satisfaction), which has been used as the title of a Chinese movie, TV series, album, Pingju (a form of local Chinese opera), Chinese translation of Shakespeare’s play ‘As you like it’ and one of its film adaptations. Even the idiom 对牛弹琴 (duiniu-tanqin, to-cow-play-thither, i.e. to cast pearls before swine), 85

Jiao Liwei

which has a negative connotation, has been used as the title of two albums and a novel on the Internet. The second most common Chinese idiom 千方百计 (by hook or by crook) has been used as the title of a song, an album, and a TV series. From 1905 to 2000, there were 6,273 domestic and imported films shown in China, and among them 128 titles were idioms. For example, 一鸣惊人 (set the world on fire, 1954), 龙凤呈祥 (excellent good fortune, 1964), 归心似箭 (wish to wing one’s way home, 1979), 顾此失彼 (unable to manage two or more things at once, 1981), 时来运转 (have a change of luck, every dog has his day, 1985), 大惊小怪 (make a fuss about nothing, 2000). It may seem that 128 titles is only small fraction of the total number of titles over nearly 100 years; however, there were only about 30 titles in idioms before 1981, that is to say, from 1981 to 2000 there were nearly 100 movie titles that were Chinese idioms, five titles per year. The sensational martial arts movie 卧虎藏龙 (Crouching Tiger Hidden Dragon, 2000) and the title of its original novel were both Chinese idioms.

10.4.  Business and economic worlds In the business world Chinese idioms are most extensively used in advertisements in the form of homonyms. For example, a cough medicine company used 咳不容缓 which literally means ‘cough-not-allow-delay’ and is from 刻不容缓 (keburonghuan) which means ‘secondnot-allow-delay, be of great urgency’. 咳 and 刻 are homonyms of ‘ke’ but with different tones. Since these kinds of advertisements are so common on TV and in other media, parents and educators worry that this language phenomenon will have a negative impact on students’ learning. This had aroused such heated debate that some local governments in China passed regulations prohibiting the alteration of fixed forms of idioms, with a fine of as high as tens of thousands of RMB.6 Chinese idioms are used not only in business, but also in economics. There are about 315 Chinese idioms in the translation of Paul Samuelson’s Economy (18th ed.), occurring a total of 450 times. A Japanese economist, Atsushi Kajii published a book named『故事成语でわかる経済学のキーワード』which uses 30 idioms to exemplify 30 economic phenomena.

10.5.  Educational purposes Chinese idioms are a refined part of the language. They are not easy but quite necessary to grasp for all native speakers and learners of Chinese, including children. The very first Chinese idiom that children in mainland China learn is probably 五颜六色 (wuyan-liuse, five-color-six-color, i.e. be riotous with color). There are about 80 idioms in one of the most popular series of children’s books Les P’Tites Poules. In 2001, Sun Wenhua wrote a concise book about Chinese history and culture. The book was written totally in four-character phrases, more than 70% of which are Chinese idioms. According to a news report of October 2012, 27.8 billion RMB would be invested in a project named 中华成语文化园 (cultural park of Chinese idioms) at Handan, Heibei Province.7 Since 2001, a panel of economists, columnists, and intellectuals in South Korea have voted on one idiom each year as the idiom of the year. In 2001 it was 五里雾中 (utterly mystified) and in 2013 it was 倒行逆施 (go against the trend of the times). In April 2012, the Jiangsu Province Research Association of Chinese Idioms was founded.

86

Chinese idioms

Notes 1 An introduction and basic statistics of those Chinese treebanks are available at http://www.cis.upenn. edu/~chinese/ctb.html, http://cslt.riit.tsinghua.edu.cn/~qzhou/chs/Resources.htm, http://ccl.pku.edu. cn:8080/WebTreebank/WebTreebank_Readme.html, http://rocling.iis.sinica.edu.tw/CKIP/treebank.htm. 2 http://zh.wikipedia.org/zh-tw/三隻小豬. 3 But some linguists advocate not using idioms excessively. For instance, a prominent Chinese linguist Lü Shuxiang once made the joke: 成语词典害死人 (Dictionaries of Chinese idioms kill writers). See China Press and Publishing Journal 5, 2009: 8 http://www.chinaxwcb.com/xwcbpaper/html/20096/05/content_52748.htm. 4 See http://cn.wsj.com/gb/20111021/BOG144431.asp. 5 These statistics are different from those by Sun (1989). 6 See http://epaper.bjnews.com.cn/html/2011-03/18/content_212028.htm?div=-1 and http://www.jxnews. com.cn/jxrb/system/2010/12/07/011536612.shtml (Article 18). 7 http://news.xinhuanet.com/local/2012-10/06/c_113283723.htm.

References Cai, Xinjiao (2011)〈越汉成语对比研究〉(‘A Comparative Study of Vietnamese and Chinese Idioms’), Master Thesis, East China Normal University. Chen, Wen (2011)〈《现代汉语词典》中定型四字格的语义分析〉(‘Semantic Analysis of the Fixed Quadrasyllabics in the Contemporary Chinese Dictionary’), Master Thesis, Nanjing Normal University. Cheng, Chin-Chuan (1998)〈从计量理解语言认知〉(‘Quantification for Understanding Language Cognition’), in Benjamin K. T’sou, Tom B. Y. Lai, Samuel W. K. Chan, and William S.-Y. Wang (eds)《汉语计量与计算研究》(Quantitative and Computational Studies on the Chinese Language), Hong Kong: City University of Hong Kong, 15–30. Fang, Huisheng (1943)〈成语和成语的运用〉(‘Chinese Idioms and their Use’), Guowen Zazhi (Journal of National Language) 2(3): 21–6. Feng, Shengli (1998)〈论汉语的自然音步〉(‘On Natural Foot in Modern Chinese’), Zhongguo Yuwen 《中国语文》(Studies of the Chinese Language) 1: 40–7. Han, Chenqi (1984)〈论汉语成语中的数词〉(‘On Numerals in Chinese Idioms’), Jiaoxue yu Yanjiu (Teaching and Research) 3: 64–9. Jiao, Liwei, Cornelius C. Kubler, and Zhang Weiguo (2011) 500 Common Chinese Idioms: A Frequency Dictionary, London and New York: Routledge. Jiao, Liwei (2012)〈从固定结构的角度增进对汉语成语的理解和使用〉(‘Comprehension and Production of Chinese Idioms: A Pattern Approach’), in Xu Weimin and He Wenchao (eds)《国际汉语教学的 理念与教学实践研究》(Research on Concepts and Practice of Developing Chinese Language Teaching Materials), Hangzhou: Zhejiang University Press, 88–93. Li, Xingjian and Liu Shuxin (1975)《怎样使用词语: 汉语词汇学习》(How to Use Words: The Learning of Chinese Words), Tianjin: Tianjin People’s Press. Liu, Changzheng and Qin Peng (2007)〈基于中国主流报纸动态流通语料库(DCC)的成语使用情况 调查〉(‘An Investigation of Idioms in Chinese Newspapers Based on the Dynamic Circulation Corpus (DCC)’), Yuyan Wenzi Yingyong (Applied Linguistics) 3: 78–86. Liu, Jiexiu (1985)《成语》(Chinese Idioms), Beijing: The Commercial Press. Liu, Shuxin (1982)〈固定语及其类别〉(‘Fixed Expressions and Their Categories’), in Liu Shuxin (ed) Yuyan Yanjiu Luncong (Journal of Linguistic Studies) 2: 156–79. Liu, Shuxin (1990)《汉语描写词汇学》(Descriptive Lexicology of Chinese), Beijing: The Commercial Press. Liu, Youyi, Li Ping, Shu Hua, Zhang Qirui, and Chen Lang (2010) ‘Structure and Meaning in Chinese: An ERP Study of Idioms’, Journal of Neurolinguistics 23: 615–30. Liu, Zhenqian (2004)〈汉语四字格成语平仄搭配的对称性与认知〉(‘Symmetrical Features of Ping-ze Arrangement in Four-character Chinese Idioms and Cognition’), Shandong Daxue Xuebao (Journal of Shandong University) 4: 44–51. Liu, Zhenqian and Xing Meiping (2000)〈汉语四字格成语语义结构的对称性与认知〉(‘The Semantic Symmetrical Features of Four-character Idioms in Chinese and Their Effects on Cognition’), Shijie Hanyu Jiaoxue (Chinese Teaching in the World) 1: 77–81.

87

Jiao Liwei Liu, Zhenqian and Xing Meiping (2003)〈四字格成语的音韵对称与认知〉(‘Symmetry in Phonological Patterns in Four-character Idioms and Cognition’), Yuyan Jiaoxue yu Yanjiu (Language Teaching and Linguistics Studies) 3: 48–57. Lu, Zhiwei (1956)〈汉语的并立四字格〉(Chinese Quadrasyllabic Expressions), Yuyan Yanjiu (Studies in Language and Linguistics) 1: 45–82. Lü, Shuxiang (1989)〈《中国俗语大辞典》序〉(‘Preface to A Comprehensive Dictionary of Chinese Proverbs’), Shanghai: Shanghai Lexicographical Publishing House. Ma, Guofan (1962)《成语简论》(A Brief Study of Chinese Idioms), Shenyang: Liaoning People’s Press. Ministry of Education, R. O. C. (2010)《成语典》(Dictionary of Chinese Idioms), online dictionary, http://dict.idioms.moe.edu.tw/cydic/index.htm. Mo, Pengling (2001)《汉语成语与汉文化》(Chinese Idioms and Culture), Nanjing: Jiangsu Education Press. Moon, Mi-jin (2005)〈从语言结构看汉语成语对韩语与其成语的影响〉(‘A View of the Influence of Chinese Idioms on Korean Language and Korean Idioms from Language Contact’), Journal of Nanjing University of Aeronautics and Astronautics (Social Sciences), 7(4): 59–62. Qian, Yun and Yu Ge (2003)〈现代汉语四字格成语的词汇化研究〉(‘A Research in the Lexicalization of Four-character Idioms in Modern Chinese’), Yuyan Kexue (Linguistic Sciences) 2(6): 86–96. Shi, Shi (1979)《汉语成语研究》(Studies of Chinese Idioms), Chengdu: Sichuan People’s Press. Sun, Weizhang (1989)《汉语熟语学》(Chinese Phraseology), Changchun: Jilin Education Press. Tsou, Benjamin K. (2012) ‘Idiomaticity and Classical Traditions in Some East Asian Languages’, in Proceeding of the 26th Pacific Asia Conference on Language, Information and Computation, Bali, 8–10 November 2012, 39–55. Wang, Hui (2009)〈词义, 词长, 词频〉(‘Polysemous Words: Meaning, Length, and Frequency), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 2: 120–30. Wang, Li (1987)〈序言〉(‘Preface’), in Wang Tao et al. (eds)《中国成语大辞典》(A Comprehensive Dictionary of Chinese Idioms), Shanghai: Shanghai Lexicographical Publishing House. Wang, Tao et al. (eds) (1988)《中国成语大辞典》(A Comprehensive Dictionary of Chinese Idioms), Shanghai: Shanghai Lexicographical Publishing House. Wang, William S.-Y. (1991) ‘Language Prefabs and Habitual Thought’, in William S-Y. Wang Explorations in Language, Taipei: Pyramid Press, 397–412. Wang, Yiwen, Zheng Yuwei, Shen Deli, Cui Lei, Yan Guoli (2012)〈解读成语中的他人心理与互动 心理: 来自眼动和 EPR 的证据〉(Reading Another’s Mind and Reading the Interactive Mind from Chinese Idioms: Evidences from Eye-movements and ERPs), Acta Psychologica Sinica 44(1): 100–11. Wang, Yuelong and Ji Donghong (2009)〈汉语树库综述〉(A Review of Chinese Treebanks), Dangdai Yuyanxue (Contemporary Linguistics) 11(1): 47–55. Wen, Duanzheng (2005)《汉语语汇学》(Chinese Phraseology), Beijing: The Commercial Press. Wen, Shuobin and Wen Duanzheng (2009)《汉语语汇学研究》(A Study of the History of Chinese Phraseology), Beijing: The Commercial Press. Wu, Huiying (1995)〈四字格中的结构美〉(‘Structural Harmony in Quadrasyllabics’), Xiuci Xuexi (Rhetoric Learning) 1: 21–2. Xiao, Zhusheng (1987)〈四言成语的两项小统计〉(Two Statistics of Four-character Idioms), Zhongguo Yuwen Tiandi (Chinese Language World ) 5: 37–8. Xu, Yaomin (1997)〈成语的划界, 定型和释义问题〉(Chinese Idioms: Extension, Formation, and Annotation), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 1: 11–17. Zeng, Xiangling (2011)〈《诗经》成语研究〉(‘A Study of Chinese Idioms in the Book of Songs’), Doctoral thesis, National Taiwan Normal University. Zhang, Jie (2007)〈《现代汉语词典》与《现代汉语规范词典》成语比较研究〉(‘A Comparative Study of The Contemporary Chinese Dictionary and Standard Dictionary of Modern Chinese’), Master Thesis, Shandong Normal University. Zhang, Jijia, He Benxuan, and Chen Xuqian (2011)〈‘上下意象图示’对汉语成语感情色彩加工的 影响〉(‘The Up-down Image Schema Affects the Emotion Process of Chinese Idioms’), Xinlixue Xintan (Psychological Exploration) 31(2): 144–9. Zhang, Tiewen (1999)〈成语的数量及产生年代〉(Number of Idioms and the Time They Originated), Yuwen Jianshe (Language Planning) 5: 23–4. Zhang, Weiguo (1989)《四字语型及其应用》(Quadrasyllabic Patterns and Their Application), Beijing: China Logistics Publishing House.

88

Chinese idioms Zhou, Changji (1988)〈汉语成语的音律美〉(‘Harmonic Prosody of Chinese Idioms’), Xiucixue Yanjiu (Rhetoric Studies) 4: 311–19. Zhou, Jian (1997)〈论成语的经典性〉(‘On the Classicality of Chinese Idioms’), Naikai Xuebao (Journal of Nankai University) 2: 29–35, 51. Zhou, Jian (2004)〈四字组合论〉(On Four-character Combinations), Hanyu Xuebao (Chinese Linguistics) 1: 40–6. Zhou, Jinguo and Shi Min (2012)〈《红楼梦》成语的文化透视〉(‘Cultural Observation of the Idioms in the Dream of the Red Chamber), Changzhou Gongxueyuan Xuebao (Journal of Changzhou Institute of Technology, Social Sciences Edition) 6: 67–71. Zhou, Zumo (1955)〈谈‘成语’〉(‘On Chinese Idioms’), Yuwen Xuexi (Chinese Learning) 1: 31–4. Zhu, Guanming (2000)〈比喻词的历史更替〉(‘Evolution of Simile Words’), Xiuci Xuexi (Rhetoric Learning) 5–6: 74–5. Zhu, Ruiwen (2006)《佛教成语》(Buddhism Idioms), Shanghai: Publishing House of the Unabridged Chinese Dictionary.

89

6 Chinese language and national identity Andrew Simpson University of Southern California, USA

1.  Nations, nationalism, national identity, and Chinese This chapter provides an overview of how Chinese language has affected, and continues to affect, the development and maintenance of national identity, drawing on a body of works which have investigated related topics. In order to embark on this task, it is first necessary to discuss how the important terms ‘nation’, ‘nationalism’, and ‘national identity’ may be applied in the context of China both synchronically and over the course of the last 150 years, and also what the language label ‘Chinese’ is regularly used to refer to. It will be seen that the relation of ‘Chinese’ to ‘national identity’ in fact varies depending on the perspective of the Chinese nation that is adopted (ethnic vs. political), and that different emphases have been placed on the role of Chinese in the development of national identity at different times in modern history. Having discussed these key terms in section 1, section 2 describes the place of Chinese in the modernization of China and the development of the People’s Republic of China through to the end of the twentieth century, examining how the growth of Mandarin Chinese has affected the emergence of national identity among different Han Chinese groups. Section 3 turns to consider language and national identity issues specifically among the nonHan minorities in the People’s Republic of China, and section 4 looks at changing attitudes towards language, national identity, and nationalism in very recent times.

1.1.  Nations, nation-states, and nationalisms The term ‘nation’ is often used in two rather different ways in discussions relating to national identity. A majority of analysts of nationalism hold that the development of nations is a comparatively recent phenomenon, resulting from a change in the way that populations have organized themselves in modern, industrial times. Nations are prototypically suggested to have a number of properties and consist of populations which are made up of a single ethnic group with a common ancestry, history, culture, and language, living in a distinct territory, where the national population benefits from a uniform mode of political organization, citizenship, and equal rights. A second way that the term ‘nation’ is often applied is in reference to all politically independent states, regardless of whether such territories contain populations

90

Chinese language and national identity

connected by a common ancestry, language, history, and culture. In the literature focusing on nationalism, such potentially heterogeneous states are sometimes referred to as ‘official nations’ or ‘territorial nations’, whereas the term ‘ethnic nation’ is reserved for populations with a shared lineage and culture etc., and ‘nation-state’ for those ethnic nations which achieve political independence. Considering China, much twentieth-century political activity aimed at establishing China as a strong modern nation has been focused on building an official nation which includes not only the ethnic nation of the Han Chinese, but also other non-Han minority peoples living within the borders of China who do not share obvious ancestry and language with the Han majority. Safran (1998b) and others point out that the sizeable presence of such minority groups in China makes it extremely difficult to conceive of China as a nation-state populated by a single ethnic nation. Earlier, clear reference to properties of an ethnic Han nation were however prominent in the speech of Dr. Sun Yat-sen, who Guo (2004: 160) notes as having described the Chinese nation as potentially being defined by ‘blood kinship, common language, common livelihood, common religion, and common customs’. In assessing the relation of Chinese to national identity, it is therefore important to consider what kind of ‘Chinese nation’ is being conceived of in any characterization of national identity – a Han Chinese ethnic nation excluding minority groups, or a more inclusive but heterogeneous official nation made up of both Han and non-Han groups. The bifurcation between ethnic and official nations as (potentially) different kinds of population groupings also feeds into the occurrence of different types of nationalism – ‘cultural nationalism’ and ‘political nationalism’ (Hsiau 2000; Gladney 2004; Guo 2004; Wei 2002). The former type of nationalism is a movement focused on the maintenance and strengthening of a shared historical culture among a people with the properties of an ethnic nation, whereas the latter is a political movement which attempts to achieve and maintain independence and autonomy for a population as an official nation. As will be noted later in the chapter, both forms of nationalism have occurred in China at different times, with differing repercussions on issues relating to language and national identity.

1.2.  Language and national identity ‘National identity’ is commonly viewed as the self-conscious belief and subjective awareness that people have of belonging to a nation with certain objectively describable properties, such as a common history, language, ancestry, and culture, in ethnic nations, and membership of a territorially autonomous state with equal rights and a common future to invest in official nations, the latter occasionally also being referred to as ‘state identity’. In the cultivation of feelings of national identity, language has regularly been assumed to have an important role to play, as a symbolic marker of group identity and boundary device which can separate adjacent populations with different languages, and create sentiments of group self-interest and solidarity among those who speak the same language. Furthermore, the promotion of a standardized, common ‘national’ language throughout a population has the potential to even out socio-economic inequalities and stimulate the unification of a nation through the provision of better opportunities for advancement and future prosperity (Simpson 2007). With regards to China, Chinese language has consistently been seen to have served a major binding role among the Chinese nation, as it has emerged in modern times, as part of Chinese national identity. This will be discussed at some length in future sections. The primary

91

Andrew Simpson

components of this general (Han) Chinese national identity are suggested in Meissner (2006) to include and be predicated on the following: (i) the long history of China, (ii) the shared identity of the Han people as descendants of the Yellow Emperor, (iii) the notion of a Chinese Empire continuing through different dynasties, (iv) the uniqueness of the Chinese language, (v) shared traditions of religion and philosophy, (vi) Chinese literature, art and music, and (vii) Chinese achievements in the area of medicine, ceramics, and the invention of printing, papermaking, gunpowder, and the compass. Meissner (2006) and Smits (2013) both suggest that while the ready ingredients for a Chinese ‘national’ identity were indeed long present in previous centuries, clear feelings of nationalism and national identity only emerged and grew from the mid-nineteenth century onwards, when foreign incursion into China and military defeat triggered the beginnings of a new collective national consciousness and modernization, amid resistance to outside Western and Japanese forces. Such a national identity then developed further and in different ways following the course of major political changes in China during the twentieth century and the economic booming of the country in the early twenty-first century.

1.3. ‘Chinese’ The final, complex term which needs some clarification here before we consider the inter­ action of ‘Chinese’ and ‘national identity’ is the language label ‘Chinese’ itself, often discussed and disputed and of major importance to the notion of Chinese national identity. The critical question is whether there is any single language that can actually be referred to as ‘Chinese’, or whether there are many distinct languages which might all be classed as Chinese languages, as was suggested by Bloomfield (1933) in his statement that: ‘Chinese is not a single language but a family of languages made up of a variety of mutually unintelligible languages’. As noted by Bloomfield and many others, the problem is that, descriptively, there are many regional varieties of ‘Chinese’ that vary so significantly in their pronunciation that they cannot be understood by speakers of ‘Chinese’ from other parts of China. If the criteria of mutual intelligibility is applied as a linguistic means of distinguishing languages from dialects, the conclusion should be that northern/Mandarin Chinese and Min, Hakka, Xiang, Gan, Wu, Yue, Jin, and Hui varieties all constitute different languages. However, officially, all such regional forms have been classified as dialects of a single Chinese language, and subjectively, such a classification has largely been accepted as reasonable within China and has not been seriously challenged. What has significantly helped buttress the impression of a single Chinese language with regional dialect forms is the existence of two sets of connections among such forms. First, it has been noted that although the ‘dialects’ are pronounced in ways that are often extremely different from each other, they do nevertheless share much of a basic lexicon and grammar. Second, since perhaps the time of the Qin (221–206 bce) and Wei–Jin (ce 220–420) dynasties, the writing of Chinese has almost exclusively been carried out in all Chinese-speaking regions in a uniform way, making use of a parallel style, vocabulary, and orthographic means of representation, which for much of Chinese history has been substantively different from common forms of spoken Chinese. This special linking function of the written language, coupled with strong underlying lexical and grammatical similarities among regional forms of Chinese has been enough to maintain the image of a single Chinese language (with regional dialect forms), and allow it to serve as one of the several pillar components of Chinese cultural, ethnic, and later national identity, with common standardized forms of both written and spoken Chinese becoming ever more important during the course of the twentieth century, as will shortly be discussed. 92

Chinese language and national identity

2.  Language and the development of national identity in modern China 2.1.  The emergence of China as a nation, political nationalism, and language reform There is some divergence in opinion as to whether a Chinese ‘nation’ existed in pre-modern times. Certain scholars have suggested that China was indeed a nation from early times on, perhaps from as early as the Qin–Han unification (Townsend 1996: 25) with a population sharing a single culture, ancestry, and history, governed by a powerful, centrally administered state. However, the majority opinion is that nationalism, and the emergence of China as a nation, only occurred from the late nineteenth century onwards (Xu 2002; Guo 2004). What preceded this is viewed at most as being possibly a ‘collective consciousness’ (Guo 2004) of belonging to a larger state with a widespread, similar culture and historical continuity. Smits (2013) suggests that such a consciousness of being Chinese was furthermore largely confined to the elites of society, and that the vast majority of the population in the empire had no clear feelings of being Chinese. He adds that national consciousness in China remained minimal among non-elites through until the time of the first Sino-Japanese war in 1894–5. Ten years afterwards, however, it had started to become strong in all those areas in which foreigners had come to live and carry out business in China, and later on, in the 1920s, propaganda distributed by the Guomindang nationalists helped stimulate the wider growth of Chinese national identity among all sections of society. Subsequently, from 1949 onwards until the present, the Chinese Communist Party fueled this further into a much stronger sense of national consciousness among ordinary people, so that ‘Today, national consciousness in China is intense’ (Smits 2013). Considering language in pre-modern, pre-nationalist times, written and spoken forms of Chinese underwent significant divergence from the end of the Han dynasty (206 bce – ce 220) onwards, with regional, spoken forms of Chinese diverging both from each other and from the written language of classical Chinese, which itself remained unchanged and fully dominated the creation of all official written Chinese and high literature right until the early twentieth century. However, the acquisition of literacy skills in classical Chinese was both very costly and highly time-consuming and beyond reach for all but the advantaged classes in China. Consequently, when China entered the twentieth century, its general population still had no common spoken form of Chinese that could be widely used and understood, and little widespread knowledge of written Chinese, which had long become disconnected from all spoken forms of Chinese. In pre-modern China, language thus neither served as an effective linking device among most of the large population, nor helped facilitate the acquisition of literacy because of the differences between written and spoken Chinese and the effort needed to learn classical Chinese. At the beginning of the twentieth century, it was felt among emerging nationalist leaders that some kind of language change was necessary as part of the broader drive to modernize the country and stimulate a stronger sense of collective national identity among the people. It was argued that only when the population in general could communicate effectively, have easier linguistic access to education, and feel united as a single people would China be able to strengthen itself and successfully defend itself against foreign incursion and exploitation. The modernization that took place was influenced by changing orientations towards the value of Chinese traditions and culture in comparison to Western systems of knowledge, and had important consequences for the kinds of language reforms that were advocated and adopted. The language changes that occurred in turn reflected a particular view of the ways 93

Andrew Simpson

that language should best support the development of national identity and modernization in China. Meissner (2006) charts three distinct phases of attempts made at modernization prior to the founding of the People’s Republic of China in 1949, in which differing amounts of Western technology and culture were absorbed into China. In the first period, from the mid-nineteenth century to 1895, there was a selective adoption of Western scientific learning for purely practical purposes, and the Confucian system and other forms of Chinese traditional learning were maintained to regulate other social and cultural areas of life. In the second period, 1895–1911, it was claimed by certain intellectuals that the adoption of Western technology alone would not strengthen China sufficiently, and other aspects of Western learning were imported for use alongside traditional Chinese philosophies. In the third, most important period of modernization, 1911–49, four intellectual groups vied for dominance in the development of post-imperial China, with significantly different outlooks. The Confucians strongly defended Chinese traditions, and campaigned fervently against the increased adoption of Western ideas. The nationalists, led by Sun Yat-sen, attempted to blend Chinese traditional culture with Western political models. A vocal body of ‘liberal thinkers’ argued that only the full-scale import of Western learning would save China from the advances of its enemies. Finally, the Marxists, led by Mao Zedong, set out to destroy Chinese ‘feudal’ culture and Confucianism and replace this with a new socialist identity inspired by the writings of Marx and Lenin (Meissner 2006: 45–6). These different attitudes towards the optimal way forward for China competed with each other and affected the shape and success of proposals intended to ‘modernize’ Chinese and make literacy and a national language accessible to the masses. What was broadly agreed upon was the need for significant improvement in the ability of common people to read, write, and potentially communicate with other Chinese throughout the country. What caused dispute was the way this should practically be achieved, particularly in the area of the orthographic representation of Chinese either via traditional characters or some alternative mode of writing. Prior to the vigorous debates over orthography, however, important progress was made relatively swiftly in reforming the language used to write down Chinese. While the norms of classical Chinese had been used in all formal writing and high literature for two millennia, and had grown increasingly separated from spoken language, at the end of the nineteenth century a movement grew to fully replace classical Chinese style with a written style more closely connected to the vernacular. A form of vernacular writing known as baihua had in fact come into use from the time of the Tang dynasty (ce 618–907) onwards, but was highly restricted in its use, being reserved for certain ‘lower’ forms of literature and the writing of unofficial notes. As pressure for more effective written communication asserted itself at the turn of the century, a form of baihua based on northern varieties of Chinese came to be increasingly used in writing and in print, and was officially sanctioned by the government in place of the older written style. Chen (1999: 71) notes that by 1911 there were dozens of new newspapers and magazines published in baihua, over 1,500 baihua novels, and that rising generations soon came to learn to read and write via baihua following a 1920 decree of the Ministry of Education that baihua be used as the language of literacy in schools. The replacement of classical Chinese with a written style much closer to contemporary spoken forms of Chinese had the effect that the acquisition of literacy was made easier to achieve for the common population, a clearly positive result. However, it was argued that the use of thousands of complex Chinese characters to represent the language still created a major impediment to the rapid learning of written Chinese, and was likely to hold back progress towards mass literacy. Radical suggestions were subsequently made by many 94

Chinese language and national identity

leading liberal intellectuals to actually replace the traditional system of characters with an alphabetic system as a way to speed up the acquisition of written Chinese further, and though such proposals clearly threatened to eliminate one of the central symbolic components of Chinese tradition and identity, quite remarkably the motion to use a system of Romanization to write Chinese received very wide support as a desirable goal to work towards, being opposed only by the Confucianists. Chen (2007) highlights how surprising it may appear to be that major liberal thinkers and nationalists would advocate such a potentially drastic change to the distinctive writing system of Chinese, which had maintained a link among different Chinese dialect groups and functioned as a highly visible symbol of the Chinese throughout their long imperial history. Chen notes that prominent intellectuals of the time not only emphasized the simple practical value of introducing a less complicated alphabetic representation of Chinese for the learning of reading and writing, they also vigorously and emotionally depicted the traditional system of characters in strongly negative terms, referring to it as ‘backward’, ‘clumsy’, ‘abominable’, ‘primitive’, ‘uncivilized’, ‘coarse’, and having other highly undesirable qualities. Whereas aspects of inherited linguistic tradition have elsewhere often been made use of as important symbols of national unity by groups seeking to ignite popular nationalist sentiments (e.g. in Germany, Poland, France, Croatia), precisely the reverse appeared to happen with regard to written Chinese during the modernization of China in the twentieth century, and not only the classical Chinese style of writing but also the long-revered system of characters were targeted as negative inheritances from the past, responsible for impeding progress towards the attainment of a united, modern nation. Chen suggests that such an anti-traditionalist position successfully dominated discussions of language and modernization among early Chinese nationalists due to the special circumstances of China’s history and its language. He notes that China had long evolved as an independent polity, and Chinese nationalists therefore had no need to incite the population to rise up and establish China as a new territory. The potentially symbolic value of the Chinese language consequently did not need to be made use of as a force to trigger nationalist sentiments, unlike certain nation-building situations elsewhere in which a focus on inherited, shared language has been pivotal in helping amalgamate a previously un-united population in a new territorial space. Rather, what was needed in China in the view of nationalists and intellectuals was a purely utilitarian approach to language and the adaptation of spoken and written forms of Chinese in whatever way would best help the nation achieve its modern­ ization goals. If this meant the discarding of traditional forms in order to make progress and strengthen the country, so be it. China should not dwell on its past traditions as sacred in any way, but look to a new, modernized future in which China would secure its position and be able to defend itself against foreign agression. As things happened, a switch from the use of characters to an alphabetic representation of Chinese ultimately did not come about. Internal discord in China involving conflict between nationalist and the communist armies, followed by Japanese invasion, interrupted efforts at language reform, including plans to devise an alphabetic representation of Chinese, and a second major language project, the ‘creation’ of a spoken national language, guoyu, that could be used as a means of communication by all the population and unite the nation. When these two projects were resumed in the 1950s, the drive to replace characters with a form of Romanization was superseded by a different modernization strategy aimed at simplifying the learning of written Chinese, and the national language project also underwent change. These events played out as China was led in a new direction under a quite different nationalist philosophy, as the People’s Republic of China came into existence. 95

Andrew Simpson

2.2.  The People’s Republic of China, simplified characters, putonghua, and regional forms of Chinese The plans to reform spoken and written Chinese noted in 2.1 were made and effected during a time when the dominant approach to nation-building was utilitarian rather than symbolic and also significantly one of political nationalism, aimed at a strengthening of the Chinese state in general rather than being guided by ideas of any Han-centric ethnic nationalism. Certainly, it was northern Chinese that was earmarked to become promoted as the base of the national language, but under the circumstances this was simply a natural, practical choice given the very large numbers of speakers of mutually intelligible northern dialects. It was not supposed to symbolize any belief that Han Chinese in the north somehow represented the critical ‘essence’ of the nation, hence that their way of speaking should specially be adopted as the nation’s language. The selection of northern Chinese as the future expanded link language in China was purely motivated by the practicality of developing a successful lingua franca for the modernizing state. As the founding of the People’s Republic of China ushered in a fourth phase of attempted modernization up until the late 1970s (Meissner 2006), this new period in China’s history saw a significant rejection of both Western and Chinese ‘feudal’ culture, particularly during the Cultural Revolution, and the attempt to shape China according to Marxist–Leninist socialist principles. Concerning language and pre-People’s Republic of China plans to fully replace characters with an alphabetic representation of Chinese, this potentially dramatic change was not made by the new communist regime, and efforts at developing a Romanization schemata for Chinese were restricted to the adoption of the pinyin alphabet system as a supplementary way to help in the early learning of characters and writing. Chen (2007) suggests that the primary reason why characters were ultimately not abandoned was in fact a practical, linguistic one, specifically that the existence of large numbers of homophonous words in Chinese makes the use of an alphabetic system impractical for the writing of Chinese and would have resulted in much confusing ambiguity in the written language. A second interesting reason why the Romanization efforts may have been significantly reduced in the 1950s, hinted at in Chen (1999), is the fact that this might have opened up the way for the expanded application of alphabetic/phonetic writing to other, non-Mandarin dialects of Chinese in a politically undesired way. From the time of the Qing dynasty onwards through the nationalist era, there were common warnings that the development of ways to successfully write down southern dialects of Chinese was harmful to the unity of Chinese and the unity of the country, and similar concerns may have had an impact on Romanization language planning in the early years of the People’s Republic of China, when the leadership was heavily focused on establishing and maintaining a unified state. If the common character-based writing system were to be replaced and allow for multiple regional alphabets in different dialect areas, the unifying role of written Chinese would be lost, facilitating political division along regional lines and the possibility that China might even split into different linguistic ‘nations’. What did happen to the script, instead, is that it underwent a thorough simplification process in which many of the complex, traditional characters had their shapes converted into new simplified characters requiring fewer strokes, over 2,500 simplified characters being officially promulgated for use between 1956 and 1964. As this alteration of traditional characters into modern, new forms was mostly effected at a time that ‘feudal’ traditional culture was under common criticism, it did not cause the kind of defensive reaction from nationalists that was directed at similar simplification attempts in Japan during the 1930s, where it was argued that tampering with traditional characters was a direct attack on the 96

Chinese language and national identity

national identity of the country, heavily embodied in the writing system. Indeed, as Guo (2004) notes, the communist leadership in the People’s Republic of China aimed at revolutionizing China, and in the area of language this naturally translated into a jettisoning of the traditional language. The new simplified characters then came to be associated with modernization and the eradication of China’s ‘backward’ feudal past. As for spoken Chinese, the continuation of the goal to create a form of Chinese that could be learned and used by all of the population as a link language was re-initiated in the early years of the People’s Republic of China, taking the Beijing dialect as the base for most pronunciation norms and northern varieties of Chinese as the source of grammar and lexicon. Significantly, there was a renaming of this form of Chinese from the pre-People’s Republic of China term guoyu ‘national language’ to putonghua ‘common language’. In Chen (1999: 25), it is suggested that the adoption of the new designation ‘common language’ occurred so as to avoid the perception that northern Chinese was being presented as more representative of the nation than other varieties of Chinese, and signals a clear move away from overtly emphasizing any links between language and national identity as communicated by the term guoyu. In such a way, the non-political promotion of a national lingua franca in the People’s Republic of China contrasts with the symbolic use of language by ethnic nationalist movements in other states during periods of modernization and the strengthening of independence. The promotion of putonghua from the mid-1950s onwards was very successful, and putonghua is now very widely understood, used in education, the media, and also socially in many areas. Liu (2011) notes that the 2001 ‘Law of the People’s Republic of China on the Use of Chinese Language and Chinese Characters’ prescribes that putonghua and standard characters are to be the principal language and means of writing in broadcast media, movies, school education, and administration, and discourages the use of dialects in such domains. With the great success of putonghua, it can be said that though the initial motivation for spreading this form of Chinese was certainly functional, putonghua has now over time become a clear part of Chinese national identity for much of the population at least, in the sense of being a major part of everyday life in China for much of the population, speaking and ‘consuming’ putonghua in the media and in written Chinese. The regular daily use of putonghua therefore embeds a national link among the population and connects up people from around the very large area of the People’s Republic of China, instilling feelings of belonging to a single national entity. Although the government has strongly promoted the learning of putonghua as a shared means of communication in China, it does not forbid or even discourage the use of local dialects in non-official domains of everyday life (e.g. at home and with friends etc.). In the pre-People’s Republic of China nationalist times, there were also no initiatives to suppress dialect use, and this attitude has been largely maintained ever since. Dialects are seen as perfectly appropriate for use in the more informal domains of life, and at no point has the Chinese leadership pursued an agenda to fully replace other varieties of Chinese with putonghua. While there has thus not been any explicit suppression of non-Mandarin varieties of Chinese in domestic and social domains, the successful spread of putonghua has nevertheless had the effect that knowledge and use of regional forms of Chinese has decreased, significantly in the case of certain varieties. As putonghua has increasingly become associated with modernity and higher levels of education and socio-economic status, it has come to be used more frequently not just in those domains where it is officially promoted (education, media, and government administration) but also in informal interaction, and ever more so by urban younger generations. Despite the observation of such a general downward trend in use of 97

Andrew Simpson

regional varieties of Chinese as putonghua has strengthened, there are also some recent signs that certain dialects may now be enjoying new prestige among younger-generation speakers, and, for example, the use of Shanghai dialect is now viewed as ‘hip’ and trendy among the young, distinguishing them from their parents’ generation who may largely use just putonghua. This use of dialect is not confined to oral communication but has been making its way onto the Internet among the young, with dialect writing being used in online forums such as shanghaining.com where there is no strict emphasis on standard language. As Liu (2011: 62) observes, ‘Dialect writing on the Internet celebrates creativity, flexibility, multiplicity, heterogeneity, unpredictability, informality, freedom, and rebellion’. There is also an increased occurrence of the use of dialects on television in talk shows, docudramas, dialect-dubbed films, dialect television series, and sitcoms, despite the 2001 Chinese Language Law (Liu 2011). Yet in the absence of any well-accepted, standardized way of writing regional forms of Chinese, it is hard to see such forms posing a real challenge to putonghua and potentially replacing it in commerce, education, and officialdom, even if the government were to allow this. Hence the cohesion facilitated by putonghua in both its spoken and written modes is unlikely to be threatened by the existence of popular regional forms of Chinese. It may also be hoped by the government that the overarching Chinese state identity reinforced by daily use of a common language will serve as a counterbalance to new emphases on regional cultural differences among Han Chinese groups. Gladney (2004) remarks that with the rapid economic growth in the south of China, southerners such as the Cantonese, and others, have begun to assert their cultural differences. As part of such a process, there has been a reinterpretation of Chinese history, and southern scholars have argued that the early southern Kingdom of Chu spread its culture northwards to significantly influence the development of northern Chinese civilization, rather than the other way round as is commonly believed; hence there is a local re-evaluation of the importance of the south to China’s past and presently inherited identity. Furthermore, relating to the existence of an ethnic Han Chinese national identity, many southern groups are noted to have traditionally regarded themselves not as Han people descended from the Han Yellow Emperor, but as historical descendants of the Tang dynasty. The potential for cracks in an assumed all-encompassing Han identity is therefore present, with possible consequences for the broader unity of Chinese ‘state’ national identity. The connections strengthened throughout the People’s Republic of China from the increased use of putonghua may consequently be viewed as beneficial for national unity in shoring up signs of fragmentation caused by other factors.

3.  Non-Han minorities, language, and national cohesion Approximately 8% of the population in the People’s Republic of China is made up of non-Han minority groups, recognized as belonging to 55 different ‘nationalities’. All such minority nationalities are officially held to be Chinese, the People’s Republic of China proclaiming itself to be ‘a unitary multi-ethnic Chinese nation’ (Dwyer 2005, quoting a People’s Republic of China Embassy statement). The Chinese leadership, at least as far back as Mao and his slogan ‘nationalities unite’ (minzu tuanjie, quoted in Safran 1998b), therefore envisage non-Han groups as belonging to the Chinese nation along with the Han. A natural question is whether the minority nationalities themselves feel that they are genuine parts of such an announced multi-ethnic Chinese nation. To some considerable extent, this depends on how effectively a state national identity is stimulated among the minority nationalities through measures of inclusion, including equitable access to socio-economic resources, the stimulation of feelings of being welcome partners in the future of the nation, the enabling 98

Chinese language and national identity

of increased communication between Han and non-Han groups, and the non-devaluation of minority language and culture. With regards to language issues, government policy towards minority languages has gone through a number of different stages. In the first two decades of the People’s Republic of China, there was an assimilative drive to spread putonghua at the expense of the maintenance of minority languages. However, the 1970s and 1980s ushered in a new era of linguistic and cultural accommodation which has essentially been maintained until the present. The government initiated work to help develop minority languages, and the 1975 constitution confirmed the right of all nationalities to use their spoken and written languages (Dreyer 2003a: 369). Such rights were recognized again in the 1999 constitution (Article 4, quoted in Chen 2007: 162), which states, ‘The people of all nationalities have the freedom to use and develop their own spoken and written languages, and to preserve or reform their own ways and customs’. Since the 1970s, then, there have been no attempts to eliminate or repress the use of minority languages (and culture), and certain government support for the development of these languages (the provision of writing systems, educational materials, etc.) has been made available. Minority nationalities are relatively free to make use of their languages, and Han Chinese living in minority areas are also encouraged to learn local minority languages, though in practice inter-ethnic communication tends to be carried out in putonghua or other strong regional forms of Chinese, due to the economic pressures minorities feel to learn Chinese in order to access better employment. Partially as a result of such pressures and the ‘marketvalue’ of competency in putonghua, many minority groups have experienced a decline in the knowledge and use of their languages in the communities where they live, in a way similar to the increasing shift to putonghua among Han Chinese from non-Mandarin dialect groups, and paralleling global patterns of language shift to economically more powerful and prestigious languages. Rather more positively, however, putonghua does provide the means for minority nationalities to participate fully in the economic, political, and social activities of the People’s Republic of China, and so in theory develop feelings of attachment and belonging to the Chinese state where other negative conditions do not undermine the growth of such a broader, non-ethnic, ‘national’ identity. In other non-linguistic domains, the People’s Republic of China has indeed attempted to alleviate the poorer living conditions present in many minority areas, and introduced special preferential policies for minority nationalities in family planning, school and university admissions, hiring and promotion, the financing and taxation of businesses, and regional infrastructural support (Sautman 1998: 87). The government also provides considerable financial support for large regions inhabited by minority groups, 50% of Xinjiang’s annual budget being subsidized by the government, and much higher investment being made in Tibet (Wang 2002). In this connection, Gladney (2004: 18) notes that various of the minority nationalities who live near China’s borders are aware that they are economically better off in the People’s Republic of China than their ethnic kin living across the border in states such as Kazakhstan and Kyrgyzstan. Where there have been ethnic problems between minority nationalities and Han Chinese, these have typically not been caused by language issues, but by other factors (e.g. the broader issue of increased autonomy and religion in Tibet and Xinjiang province). Hence in sum it can be said that the non-repressive and generally inclusive language policies implemented in China since the 1970s have laid the groundwork for minorities to feel more included in the nation and not negatively discriminated against, potentially allowing for a sense of national identity to develop over time with new generations, as has occurred in other ethnically mixed states where an official state language has been spread as a lingua franca without any accompanying suppression of minority languages; for example, Indonesia and Tanzania. 99

Andrew Simpson

Finally, concerning the general issue of whether it may be possible to construct bona fide ‘nations’ that are ethnically heterogeneous with multiple minority groups, the mainstream conception of nations necessarily requiring a shared ethnic origin, language, history, and culture has been challenged by a number of scholars and political scientists who advocate an alternative model of genuinely multi-ethnic nations which potentially lends itself to the situation in the People’s Republic of China. The nineteenth-century French philosopher Ernest Renan, for example, has argued that modern nationhood is not based on a shared language, culture or ancestry, but on a shared sense of destiny and the desire of a population to live together (Dreyer 2003b: 393). Within China, Guo (2004: 13) notes that such thinking has been echoed by a prominent professor of philosophy at Zhongshan University, Li Zonggui, who argues that what is indispensable to a nation is a sense of belonging together, not necessarily a common language, culture or historical bloodlines. In such conceptions, then, it is perfectly feasible for the minority nationalities to become part of a modern Chinese nation oriented towards the future, and the adoption of a Chinese national identity perceived in such a way should not require the abandonment of other ethnic language and culture. A second, recent approach to conceiving of the Chinese nation in a way that is inclusive of minority nationalities has been referred to as ‘racial nationalism’ in Meissner (2006), and is quite different to Renan’s view of multi-ethnic nations, essentially denying that fundamental ethnic differences exist among minorities in the People’s Republic of China. Meissner notes that since the watershed events of Tiananmen in 1989, when the government saw it as increasingly important to stress the collective identity of the Chinese, official documents have attempted to emphasize a common ancestry of both the Han and the minorities. Han and Tibetans (and other minorities) are presented as parts of a common race, and the early cultures of such peoples are regarded as identical. In such a view, Tibetans (and quite possibly other minority groups) may belong to the Chinese nation through a genuine blood relationship, and there is no need to imagine the People’s Republic of China as a multiethnic nation. While the historical accuracy of such declarations may be questionable, it is nevertheless indicative of the leadership’s desire to integrate the non-Han into a closer national identity and may function to adjust traditional perceptions of such groups as being unconnected to the Han.

4.  Chinese and national identity in modern times Perhaps the most striking phenomenon relating to Chinese national identity in very recent times in the People’s Republic of China has been strident calls from many Han intellectuals for a return to Chinese traditions and Confucian values as a way to take the nation forward and save it from the spiritual decay perceived to be growing in China in the wake of an increasing loss of faith in Marxism and other aspects of Western thought which had fuelled China’s modernization during the twentieth century. For the new ‘cultural nationalists’, it is paramount that China reclaim its national identity from traditional sources and Confucianism to combat the social chaos (crime, corruption, materialism, gambling, prostitution, etc.) which has arisen since the 1980s (Guo 2004), and calls for a return to tradition as a means to reenergize nationalism have significantly not been censured by the political leadership, which is keen to see loyalty to the state stimulated at a time when its Marxist platform of support may be less inspiring than in the past. The Chinese national identity which is being brought into focus by the new movement of the cultural nationalists is one which is considerably Han-centric, based on aspects of Han Chinese tradition and culture, and seems to be primarily 100

Chinese language and national identity

aimed at strengthening and uniting the Han core of the People’s Republic of China. It is therefore an ethnic nationalism which is not preoccupied with the integration of non-Han minorities within the state as nation, and instead makes its first concern the rejuvenation of the morally-threatened Han population from its traditional roots. Organized into prominent bodies such as the International Confucius Foundation, the Chinese Yellow Emperor Association, the Chinese Culture Society at Beijing University, and the Beijing International Chinese Character Research Association, Guo (2004: 18) notes that the cultural nationalists take the Chinese language to particularize the Chinese way of thinking and enshrine the ‘spirit’ of the nation, and it is bemoaned that modern Chinese thought has become weakened and confused by the intrusion of Western influences. In new works of literature created by the cultural nationalists, Guo (1998: 172) reports that writers have accordingly set out to restore the purity of the national language corrupted by Western cultural influences and bring back ‘the Chinese essence’ through writing about Chinese traditions. Extending the idea that Chinese embodies the ‘spirit’ of the nation to include its special mode of orthography, a major linguistic focus of the cultural nationalists’ activities has been the defense of traditional characters, which in much of the twentieth century had been negatively portrayed as the direct cause of low levels of literacy among common people. The cultural nationalists argue that characters, and in particular traditional characters, should instead be recognized as instantiating ‘the root of Chinese culture, a symbol of the Chinese nation,  .  .  .  the single most important agent for national cohesiveness, the transmitter of China’s national spirit’ (Guo 2004: 105). They thus suggest that it is crucial for rising generations to learn traditional characters, as a means to access and appreciate the full meaning of Chinese literature and philosophy written in previous centuries, and also warn against any further attempts at character simplification or increased use of Romanization, which would cause more harm towards the national identity. Going on the offensive, the cultural nationalists have additionally made claims that, far from being detrimental to learning, Chinese characters actually help very considerably in the acquisition of literacy, and should therefore be valued and used as a possible mechanism to promote literacy in other languages of the world. Such an assertive defense of characters has led on to other eulogies of a range of ‘superior’ properties of the Chinese language which are taken to potentially qualify it for the role of a new world language (Lu 2008). These properties are well described in Guo (2004: 96) and include suggestions that: ● Chinese is the clearest and the most concise language. ● Chinese is the world’s easiest language to learn. ● The ideographs (characters) are ‘expressions of meanings’ and ‘symbols of ideas’ and store more information than linear writing. ● Chinese grammar is the closest to the grammar of mathematics, musical notation, and chemical symbols. Supporters of the potential spread of Chinese as an international language take heart from the growth in government-sponsored centers around the world (now over 200; Lu 2008: 269) which promote the international learning of Chinese language and culture; for example, the Confucius Institute in the USA. In this atmosphere of energetic assertion of the positive values/possible superiority of Chinese, it is also relevant to note that ‘Chinese’ is now increasingly being referred to as Hanyu, the language of the Hans, in everyday speech in China, in place of the ethnically more neutral putonghua ‘common language’ (Dwyer 1998: 80). 101

Andrew Simpson

This further points to the emergence of a Han ethnic nationalism at the heart of the resurgence in language pride and a perception of Mandarin Chinese as being principally anchored in the Han nationality. While such a movement to return to and re-exploit Chinese tradition has been strong amongst intellectuals in recent years, Guo (2004: 102–8) points out that there are also opponents to any increased use or learning of traditional characters, and it is maintained that the simplified script also has an important symbolic value for national identity, being the specially distinctive script of the People’s Republic of China. Guo notes that such proPeople’s Republic of China critics of cultural nationalism, termed ‘political state nationalists’, emphasize that ‘modern Chinese, simplified characters, and pinyin are the three pillars of the People’s Republic of China’s socialist culture’ (Guo 2004: 103), hence should not be displaced by the use of traditional characters, which are seen as symbols of capitalism, due to their continued use among overseas populations of Chinese living in non-socialist states. Guo concludes that this political symbolism of the simplified script will make it difficult for the government to allow for the reintroduction of traditional characters in any official way, though there has been frequent unofficial use of traditional characters in public signs such as restaurant and store names for several decades now, despite the imposition of fines for such displays (Chen 1999: 191). It therefore seems that it is unrealistic for the cultural nationalists to expect that simplified characters or pinyin will disappear and be replaced by full-form characters. However, in their campaigning for a return to the traditional past, the cultural nationalists have really put Chinese (Han) language back into focus as central to Chinese (Han) national identity, in a way that has not occurred so virulently before, during the twentieth century. It will be interesting to see how this engagement with national identity and the tensions between ethnic and political/state nationalism and language unfold further in the People’s Republic of China as China and its complex population confront new challenges through the twenty-first century.

References Bloomfield, Leonard (1933) Language, New York: Henry Holt. Chen, Ping (1999) Modern Chinese: History and Sociolinguistics, Cambridge: Cambridge University Press. Chen, Ping (2007) ‘China’, in Andrew Simpson (ed) Language and National Identity in Asia, Oxford: Oxford University Press, 141–67. Dreyer, June (2003a) ‘The Evolution of Language Policies in China’, in Michael Brown and Sumit Ganguly (eds) Fighting Words, Boston: MIT Press, 353–84. Dreyer, June (2003b) ‘The Evolution of Language Policies and National Identity in Taiwan’, in Michael Brown and Sumit Ganguly (eds) Fighting Words, Boston: MIT Press, 385–409. Dwyer, Arienne (1998) ‘The Texture of Tongues: Languages and Power in China’, in Safran (1998a), 68–85. Dwyer, Arienne (2005) The Xinjiang Conflict: Uyghur Identity, Language Policy, and Political Discourse, Washington: East-West Center. Gladney, Dru (2004) Dislocating China, Chicago: University of Chicago Press. Guo, Yingjie (1998) ‘Patriotic Villains and Patriotic Heroes: Chinese Literary Nationalism in the 1990s’, in Safran (1998a), 163–88. Guo, Yingjie (2004) Cultural Nationalism in Contemporary China, London and New York: Routledge. Hsiao, A-Chin (2000) Contemporary Taiwanese Cultural Nationalism, London: Routledge. Liu, Jin (2011) ‘Deviant Writing and Youth Identity: Representation of Dialects with Chinese Characters on the Internet’, Chinese Language and Discourse 2(1): 58–79. Lu, Dan (2008) ‘Pre-imperial Chinese: Its Hurdles towards Becoming a World Language’, Journal of Asian Pacific Communication 18(2): 268–79.

102

Chinese language and national identity Meissner, Werner (2006) ‘China’s Search for Cultural and National Identity from the Nineteenth Century to the Present’, China Perspectives 68: 41–54. Safran, William (ed) (1998a) Nationalism and Ethnoregional Identities in China, Portland: Frank Cass. Safran, William (1998b) ‘Introduction: Nation, Ethnie, Region, and Religion as Markers of Identity’, in William Safran (ed) (1998) Nationalism and Ethnoregional Identities in China, Portland: Frank Cass, 1–7. Sautman, Barry (1998) ‘Preferential Policies for Ethnic Minorities in China: The Case of Xinjiang’, in William Safran (ed) (1998) Nationalism and Ethnoregional Identities in China, Portland: Frank Cass, 86–118. Simpson, Andrew (2007) ‘Language and National Identity in Asia: a Thematic Introduction’, in Andrew Simpson (ed) Language and National Identity in Asia, Oxford: Oxford University Press, 1–29. Smits, Gregory (2013) ‘Topics in Modern East Asian History’, Online textbook at: http://www.personal. psu.edu/faculty/g/j/gjs4/textbooks/175/index.htm. Townsend, James (1996) Chinese Nationalism, New York: M. E. Sharpe. Wang, Lixiong (2002) ‘The “Tibetan Question”: Nation and Religion’, in Wei C. X. George and Liu Xiaoyuan (eds) (2002) Exploring Nationalisms of China: Themes and Conflicts, London: Greenwood Press, 151–72. Wei, C. X. George (2002) ‘A Cultural Search for National Identity: The Evolution of the Nationalism of Taiwan’, in Wei C. X. George and Liu Xiaoyuan (eds) (2002) Exploring Nationalisms of China: Themes and Conflicts, London: Greenwood Press, 123–50. Wei, C. X. George and Liu Xiaoyuan (eds) (2002) Exploring Nationalisms of China: Themes and Conflicts, London: Greenwood Press. Xu, Jilin (2002) ‘Intellectual Currents behind Contemporary Chinese Nationalism’, in Wei C. X. George and Liu Xiaoyuan (eds) (2002) Exploring Nationalisms of China: Themes and Conflicts, London: Greenwood Press, 27–39.

103

7 Chinese language education: Teacher training Jane Orton University of Melbourne, Australia

Provision Training paths The increasing global importance of China in the twenty-first century has led to growth in tertiary Chinese language programs and the quite widespread introduction of Chinese programs around the world in primary and secondary schools, and even kindergartens. This growth has brought a matching demand for teachers of Chinese at all levels of the educational system, as well as in commercial language programs. In Western countries, no formal training in teaching is required of those who teach languages at tertiary level, although some institutions offer professional development programs in teaching, and a few even make such programs compulsory. Primary and secondary school teachers, however, must be trained to teach their specific learning area, and the continuing demand for graduates trained to teach Chinese in schools has put considerable pressure on teacher education institutions. In the English-speaking world, and in many other countries, candidates obtain certification to teach school children by undertaking formal training as set out by the state, at national or provincial level. In many of these countries, certification is required of those who would teach in government schools, and most private schools also stipulate that their staff be certified. Typical routes to certification are a one-year Education course after obtaining a Bachelor’s degree, or a four-year Bachelor’s degree combining undergraduate studies and Education. In either case, for language teachers the undergraduate program will include major studies in the target language, literature, and culture, and the Education component will include studies of educational theories and systems, the nature of language and learning, curriculum design and assessment, pedagogy, and supervised classroom practice. There are less standard means of entering school teaching, including obtaining partial certification or provisional licensure. In Chinese language teaching these paths have opened over the past decade in some places due to the pressure to get the language going in the school system, and to the general shortage of qualified teachers to fill positions that are being 104

Chinese language education: Teacher training

created (Stewart and Wang 2005). For these reasons, too, some schools have asked first language (native) or heritage (Chinese background) speakers who have no training in language teaching, but are trained to teach other subjects in the curriculum, to shift into language teaching positions. In yet other cases, shortcuts to certification have come from innovative teacher training programs intentionally fast-tracking undergraduates into the classroom. These candidates receive some initial training and then follow their teacher education courses as they work, attending classes provided in an intensified format, often online.

Candidate profiles Prospective teachers of Chinese may be first language (L1) speakers of Chinese, who have grown up in a Chinese society, heritage speakers (HS), who have learned the language as a home language while living outside a Chinese society, or second language (L2) speakers, who have learned Chinese as a foreign language in a classroom. Where the student is already proficient in the language, they may be directed to follow undergraduate major studies largely in Chinese literature and culture. If they have already taken an undergraduate degree in the medium of the target language (in China, for example), students seeking to become teachers may even be exempted from the undergraduate language study requirement altogether, provided they pass a test showing that their culture and language knowledge and skills are at least equivalent to those of a graduating major in the language. Language studies offered to undergraduates in Arts faculties expect students to acquire proficiency in oracy and literacy, as well as a grounding in ancient and contemporary philosophy, history, and literature. These cultural studies make an excellent and, indeed, essential base for using the language well by providing deep, connected networks of meaning to the modern language being learned. Due to time pressures and the difficulty of catering to the very wide range of future areas of work of their diverse student body, undergraduate programs in language departments rarely offer professional language study as an option. An exception is the occasional document-reading course for law or business. As a result, while often able to recite classical poetry by heart, graduates of language departments who want to be teachers of Chinese have usually not acquired the vocabulary and phrases useful for running classroom activities and managing students of their own. Most tertiary teachers of Chinese are first language speakers and, like L1 speakers of all languages, without training in learning and teaching they often have little knowledge or understanding of the challenges their language poses for students. As well, having been educated in a Chinese system where students did not ask questions, and lessons of all kind consisted of teacher talk and home learning of the book by heart, the staff of most tertiary Chinese programs to date provide a narrow range of teaching methods and learning activities for their students, and have few strategies to pass on beyond being diligent and persevering. The American Foreign Service Institute in Washington DC estimates that it takes English speakers about 3.5 times longer to master Chinese than a European language, yet Chinese courses in schools and universities are scheduled in parallel with those of other languages, and all face restricted time limits. Given these factors, and the inherited devaluing of language development in favour of literature and social studies in Western universities, it is not surprising that L2 students of Chinese generally graduate with comparatively low proficiency in the language. Unless sustained by a lengthy time spent in country, spoken proficiency, in particular, is rarely of a standard desired as a model in teaching, or sufficient for running a classroom in the target language. There are some exceptions to this situation, but they are not yet the norm. 105

Jane Orton

Training programs Until very recently, almost all training in the teaching of Chinese as a foreign or second language in English-speaking countries was provided in a generic course on methods of teaching modern languages taught in a Faculty of Education as part of a broader pre-service secondary teacher education program. In some cases, casual staff were employed to add Chinese-specific matters to content originally designed for and staffed by teachers of European languages. In other cases, it was only when students went out for professional experience that they had contact with experienced teachers of their language. A dedicated method course in teaching Chinese language has been rare. By 2015 a number of Master degrees in the teaching of Chinese have appeared in Western countries. However, most such courses are taught in a department of Applied Linguistics and their content is predominantly, if not exclusively, centred on the nature of Chinese language. Few include pedagogy, and almost none combine linguistic knowledge with wider studies of learning and educational thinking or classroom management, and most have no period of supervised professional practice. Another strand of Chinese language teacher training comprises the Master of Chinese as a Foreign Language (MCFL) (对外汉语课) courses which have mushroomed in universities throughout China in recent years. The courses are offered by teacher education institutions such as normal universities, as well as by Applied Linguistics departments in faculties of Arts. In either case, they are strongly oriented to detailed studies of the language, with only some study of basic pedagogy, and no studies in educational thought. With at best only a handful of local institutions teaching Chinese as a Second Language available for their students to observe actual teaching of a class, usually at tertiary level, most graduates of these programs have not been exposed to learners of any age, let alone taught the language, and many have not even heard Chinese spoken by an L2 user. In an effort to strengthen their experience, the Chinese government’s Office of International Chinese, the Hanban, annually sends groups of MCFL students overseas for a year in one of a variety of countries, where they work as assistants to teachers of Chinese language. The training of primary teachers of Chinese is very varied in form. In some cases, L1 speakers who are not teachers are invited to conduct activities such as singing, counting, and storytelling with children while being overseen by a classroom teacher; in other cases, trained generalist primary teachers who have the language are invited to run such a program, or even to teach the language more formally; and, lastly, there are also certified primary teachers of the language, graduates of pre-service teacher education courses, who may exclusively teach Chinese, going from class to class in one or more schools, or who include language teaching as part of their general load. In addition to their Chinese language classes, generalist teachers of the language may largely teach curriculum in the local L1 language (e.g. English), or they may be fully occupied teaching in Chinese as part of a bilingual, immersion or intensive program of second language and content learning. There is little or no kindergarten or pre-school training for teaching Chinese as yet, and certified staff working with those age groups outside China have virtually all trained as early childhood educators in a Chinese society. As part of their training they will have undertaken child language development studies, although not studies in the development of Chinese as a foreign language.

Learner levels English-speaking countries have accepted migration for centuries, right up until the present. Chinese speakers have been a very large group among these immigrants, usually settling in 106

Chinese language education: Teacher training

concentrations within large cities. The children of these families speak their parents’ language, which may be a dialect, and many attend weekend classes to develop their literacy. With very mixed proficiency in standard Chinese, and very particular needs in terms of language development and cultural identity formation, their presence in classes in regular schools in numbers ranging from 30% to just one or two adds a challenging second level to L2 classes. As well, their proficiency is often a disincentive for L2 classroom learners to continue, especially in the competitive senior years. In the last years in some high schools there may also be international students from China preparing for university entrance. These are L1 speakers, who add a third level of Chinese curriculum, possibly to the one class, in which they follow studies in Chinese language and literature appropriate to their capabilities. There is little or no special teacher preparation for the two levels of background learners, and curricula and resources for them are oriented to the norms of education in a Chinese society, despite their being – or becoming – bilinguals outside Chinese society, with quite special needs.

Chinese teacher preparation Standards Teaching is a highly complex social practice that involves a practitioner drawing on integrated, articulated layers of different forms of knowledge in a great many areas. Like their colleagues in other professions, teacher educators continue to grapple with the tension between passing on the wisdom of experience and developing a new teacher’s own capacities. Programs are still commonly structured along a theory-to-practice line of progression, with candidates first instructed in formal knowledge of their subject matter and how to teach it, and then sent out into schools to apply this knowledge in real-life settings. Other programs, however, are based on perception of the task as a more iterative process, one in which the novice enters a cycle of knowledge and skill gathering, guided practice which tests the meaning of concepts and values comprehended so far and extends knowledge and understanding, reflection on practice, and the results of reflection fed back to inform and expand knowledge, understanding, and practice. While recognising that there are routines which can usefully be mastered by student teachers, professional educators aim to develop artistry in their novices: the ability to perceive and analyse situations which will inevitably arise that are novel, unexpected, or uncertain, and to design, test, and assess on the spot quite new action in response to them. It is this capacity that they nominate as the core attribute of a practitioner of high quality and an essential if a teacher’s best efforts are to be fruitful. Despite variations in perspective, three matters all teacher educators agree on are that ● deep knowledge of their subject is the absolute essential for a teacher to survive, to develop, and to innovate ● while learning to teach is a lifetime project, a certain breadth of factors must be involved in training from the start ● a practicum is an essential component of initial training. These elements are strongly represented in the professional standards for school teachers that have been developed in recent years in all English-speaking countries (e.g. United Kingdom Government Department for Education 2011). In the United States and Australia, dedicated standards for language teachers have also been established (American Council 107

Jane Orton

on the Teaching of Foreign Languages (ACTFL) 2002; Australian Federation of Modern Language Teachers Associations 2005), and the Hanban have published their own standards specifically for teachers of Chinese as a foreign language (2007). While each has its particularities, the lists all include ● ● ● ● ● ●

strong proficiency in the target language knowledge of the nature of language and culture knowledge of learning and learner development a repertoire of instructional strategies the capacity to plan a course of learning and to assess it the ability to reflect and improve their own practice.

Proficiency in the target language is the first area of knowledge required by all, and insistence on this reflects a reality in which the regulatory standard for admission to teaching candidature everywhere has most often simply been a pass in language courses. This has allowed for a very wide range of proficiency among those presenting as teacher candidates. Since publication of the standards, regulation testing of prospective language teachers to ensure they meet a high minimum spoken and written proficiency level in the standard form of the language has been increasing. Furthermore, as the teacher preparation standards in Table 7.1 show, development of this proficiency is seen as the shared responsibility of language departments and the Faculty of Education. In realising any of the sets of standards, thorough exploration of three educational factors present in the teaching of all language is especially critical in the sound preparation of Table 7.1  ACTFL Program Standards for the Preparation of Foreign Language Teachers The preparation of foreign language teachers is the joint responsibility of the faculty in foreign languages and education. In order for foreign language teacher candidates to attain the knowledge, skills, and dispositions described in the ACTFL Program Standards for the Preparation of Foreign Language Teachers, programs of foreign language teacher preparation must demonstrate that they include the components and characteristics described below. 1. 2. 3. 4. 5. 6. 7. 8.

The development of candidates’ foreign language proficiency in all areas of communication, with special emphasis on developing oral proficiency, in all language courses. Upper-level courses should be taught in the foreign language. An ongoing assessment of candidates’ oral proficiency and provision of diagnostic feedback to candidates concerning their progress in meeting required levels of proficiency. Language, linguistics, culture, and literature components. A methods course that deals specifically with the teaching of foreign languages, and that is taught by a qualified faculty member whose expertise is foreign language education and who is knowledgeable about current instructional approaches and issues. Field experiences prior to student teaching that include experiences in foreign language classrooms. Field experiences, including student teaching, that are supervised by a qualified foreign language educator who is knowledgeable about current instructional approaches and issues in the field of foreign language education. Opportunities for candidates to experience technology-enhanced instruction and to use technology in their own teaching. Opportunities for candidates to participate in a structured study-abroad program and/or intensive immersion experience in target language communities.

108

Chinese language education: Teacher training

Chinese teachers. The first is that, in addition to knowing the nature of language and student learning, and target language particularities, teachers need to understand the target language as an object of learning, that is, to perceive the demands that learning it will make on learners who are speakers of their particular language. This is the crucial link between knowledge of the teaching content and design of methods and resources which will successfully allow learners to master it. In the teaching of Chinese to speakers of Western languages, fine analysis of the learning task involved in the various aspects of the language identified as needing to be taught has been insufficient to date, and this is a major factor in the comparatively high failure rate in Chinese learning. A second factor that teachers must be taught to appreciate is the more general development that is achieved through language learning, what might be called its educational affordances. While this most normally is of central concern in school education, it is nonetheless always present and no less a value for college students or adult learners in later life. The third factor teachers in preparation must know and take account of is the culture of education in which the teaching will take place.

Educational factors Chinese as an object of learning There are a number of core features of Chinese language which make demands on learners from European language backgrounds that the expertise they developed in acquiring their first language is insufficient to meet, and hence these features require further foundations be laid. The demands lie principally in the fundamentals of the phonological and orthographic systems, and the features of Chinese vocabulary. There are also some very specific matters in the grammar and the culture which are a challenge to European learners. These very central aspects of Chinese need to be known by Chinese teachers, who must be prepared to use resources creatively if they are to assist learners to successfully meet and master them.

Oracy The most salient feature of the Chinese phonological system is that syllables are tonal, that is, they vary in pitch and contour and this variation changes the meaning. To obtain meaning, these shifts in pitch must be heard with the segmental sounds, so that mà and má are heard as being as different as ‘pan’ and ‘pen’ in English, that is, as different words, with different meanings. To perceive tone, L1 speakers take their cue from both the entry pitch of a syllable and the contour. Tone is a very new use of voice for speakers of non-tonal languages and most find tones difficult to perceive, and both entry pitch and contour hard to control. One fundamental reason for their difficulties in tone perception is that they simply forget to listen for it. Instead, as they are accustomed to doing in their own language, speakers of non-tonal languages naturally attend only to the segmental sounds. Thus they will say they heard, for example, the syllable shi and then wonder what tone ‘to put on it’. But there is no shi in Chinese, what they heard was either shī, shí, shǐ, or shì. A second difficulty is that tones are volatile, with only syllables that are being stressed in an utterance carrying full tonal value. Indeed, in modern rapid spoken Chinese, it is estimated that as much as 47% of syllables may be atonal (Triskova 2008). This causes perception difficulties, and in production, having diligently learned their vocabulary of tonal syllables, 109

Jane Orton

many learners have difficulty reducing tone on certain syllables when speaking, making their own speech overstressed and hard to understand. To speak intelligibly, learners need to master flows of syllables with varying tones and varying tone quality. The basic skill of judging pitch entry and the length of the contour takes work, and students can be unaware for quite some time that they are not achieving the sound they think they are making. On a positive note, studies show that work on tone perception and production are mutually beneficial and, over time, if the learning weaknesses of listening for tone and controlling pitch entry, contour, and tone quality are targeted, exercises can be quite effective in developing mastery (Miracle 1989; Wang et al. 2006; Orton 2013). Prosodically, rapid, intimate spoken Chinese is generally considered to be stress-timed, like English (but not French, for example), but Chinese word stress more often falls on the second syllable of a word, not, as in English, on the first. English speakers are thus often competent at chunking their utterances in Chinese but misplace stress. More difficult than rhythm and phrasing for these learners is restraining from using strong English intonational patterns to express attitude, which if used, very often interfere with the accurate production of tone. Use of the Chinese modal particles to carry some of the load intonation carries in their own language is a late acquisition for most learners of Chinese.

Literacy The densely compacted square of information that constitutes a Chinese character is a very different form of writing down human language from the linear strings of alphabetical combinations used in European and some other languages. Perceiving and obtaining meaning from characters demands learners of Chinese organise their reading eye in a quite new way, perceiving the square and noticing its components from left side to right side, and from top to bottom (Lin and Childs 2010). Although characters continue to be taught as a series of single strokes, the most efficient way to read a character is to perceive the, usually multistroke, components it is composed of, just as the most efficient way to read the English word London is to perceive the two starting strokes not separately as I and _ but together as L. Another significant difference involved in mastering reading in Chinese is that characters are written evenly spaced across the line, so that those representing the syllables of a multisyllabic word need to be gathered in one sweep of the eye, while other characters will be a single word in their own right. Characters do not represent sounds in the way that letters of an alphabet do, although some links with sound in about 40% of characters can be perceived once they are known. Thus learning a character requires learning both its sound and its meaning. As a way of recording the sound of a character, students are usually taught Pinyin, the orthodox form of Romanisation. Pinyin is a spelling system for the spoken language and, like all such systems, its sign-sound combinations are regular only to a point. Exceptions and anomalies exist and must be learned. Nonetheless, if mastered, Pinyin is the reliable way to know the accurate sound of a character or item of vocabulary. The most basic capacities in Chinese – to hear and utter, to read and write – require foreign learners to extend very basic motor skills and to change long ingrained habits in perception and production of spoken and written language. The skill base developed in learning their first language to a high degree of expertise and automaticity is simply insufficient to support European language speakers’ perception and production of the most basic Chinese. Their first attempts at writing characters also show starkly how true this is. To write an even passable hand requires the same laborious work gone through in elementary school. Some 110

Chinese language education: Teacher training

try to avoid the problem by using a word processor, but while the results look more pleasing than the childish script of the novice, studies are yet to be undertaken that could show whether the consequent reliance on the eye alone to identify a character, with no tactile support, and the reduction of time on basic task, will permit a student to achieve the same degree of reading competence and, especially, writing competence, once the volume of characters involved begins to mount.

Vocabulary Unlike any European language or even Japanese, Chinese vocabulary is special as an object of learning for speakers of European languages, first and foremost, for the sheer number of new words to be learned. There are virtually no cognates, so there can be no shortcuts as exist in the thousands of Latin- and Greek-derived words common to most European languages, such as education, commence, photograph, and in the everyday words shared among languages with a common history, such as the German and English Finger, Hand, Arm, Schulter, and finger, hand, arm, shoulder, or the many borrowings from English in modern Japanese, for example, pikunikku teiburu for ‘picnic table’. Chinese dictionaries list characters, which represent meaningful syllables, and are often also independent words. Ending only in a vowel or vowel + n or ng, to the European ear, Chinese syllables provide a restricted range of features to distinguish them and so are not easily perceived and retained. It is impossible to remember the learning process by which as babies we sorted out the sounds of our own language being made around us from all the possible sounds we could utter, but research shows we did all have to learn what variations matter and to recognise and disregard changes in sound that were due simply to phonic environment, or individual voice quality, or the emotion of a specific utterance (Gopnik et al. 2000). When confronted with very new sets of sounds that we are not in the habit of attending to and storing, we can perceive our bias towards those combinations found in our own language. Chinese presents the English-speaking learner with a large number of such sets. For example, the ratio of syllables starting with the letter ‘x’ in Chinese compared to English is in the order of 85 : 1, while the total of syllables starting with ‘x’, ‘y’, and ‘z’ make up some 22.4% of Chinese, but only about 0.6% of an English vocabulary (Orton 2014). The difficulty this presents becomes evident when students try to read characters they have only recently met: ‘Oh that’, they say, ‘that’s xu  .  .  .  xi  .  .  .  xue – something like that’. Even within one tone there are a great many homophones. The sound yì, for example, can be written in 50 different characters, that is, as 50 different syllables with 50 different meanings. While this is an extreme, there is considerable homophone load on a large number of syllables that appear commonly in everyday speech. The lack of cognates, the newness of many sounds, the limited range of sound combinations, and the heavy load of homophones make high demands on European learners of Chinese to develop their perceptive skills in ways never before required of them, perceptive skills they have largely left undisturbed since infancy and pre-school. At the same time, due to the high volume of new words that must learned, the burden on memory of learning Chinese vocabulary is substantially greater than that imposed in learning another European language. And this demand is made of a usually underdeveloped memory power, already overtaxed by having to remember a new character for every syllable. Unless they are assisted to extend their capacity to hear and have frequent opportunity to engage with vocabulary and characters in vital ways that involve mind and affect, learners will find new Chinese words difficult to internalise and have available for spontaneous use. 111

Jane Orton

Grammar Like all new languages, Chinese also presents its learners with new aspects of grammar, however most do not prove more challenging to master than, for example, those posed for a German speaker learning French, or a Spanish speaker learning English. Although a generally new class of word for learners, measure words and classifiers are not difficult to understand, but finding and remembering the correct one for each noun is a chore, as is linking verbs and their complements correctly. Having no articles or tenses, and no word for ‘Yes’ takes adjustment of perceptive and productive sensitivity to other terms being used for these functions, while the reverse clause order of complex sentences using de (的) segments and, especially, of a series of embedded de-segments, demands considerable cognitive rerouting in the learner if the relationships between clauses are to be correctly linked. Even when heard, the communicative significance of particles used to express aspect and mood takes a long time to impact the ear.

Culture A final aspect of Chinese as an object of learning is that of culture. Chinese are rightfully proud of their long and illustrious cultural heritage, and it is aspects of the traditional heritage which are taken to represent Chinese culture in most language textbooks and courses. Feast days such as Spring Festival and the Mid-Autumn Festival, historical events such as the founding of the People’s Republic, and world discoveries such as paper making and the compass, are described in detail, while lessons are provided in calligraphy and various folk arts and crafts such as paper cuts and seal making. In resources from China especially, there has been little about modern lived culture, nor any systematic introduction to the network of worldview and beliefs that still give meaning to much of the language in use today. Thus simply to say that Chinese has separate words for ‘older brother’ and ‘young brother’ is to leave the learner ignorant of the critical meaning of this: that age is the most significant relegating factor in Chinese society, one which people orient by constantly; that identity is realised in relationships organised by age; and that age relations will underpin expectations of the learner in social interactions with Chinese. At the same time, learners need to know that in the modern world, even such cultural basics as deference to age may be contested at times by the young owning new social power, such as money, technological knowledge, or fluency in a foreign language. While the exotic exemplified by dragons and grave sweeping can be enticing to young learners, their natural egocentrism may mean they find introduction to the new culture easier through the perspective of their own lives – what people eat for breakfast, what young people learn at school, how texting is done in Chinese – and only gradually develop interest in parts of the Chinese world different from their own. In any case, the essential in teaching about cultural matters is to help students see them as parts of a web of social meaning, linked to language, not just a set of singular, exotic customs.

Resources Knowledge and appreciation of the deep demands of Chinese as an object of learning set out above are still rare, and even the facts of the language and its culture are often not well known or understood by those who create resources, or the teachers who use them. Furthermore, to date few materials for teaching Chinese at any level are comprehensive in terms of 112

Chinese language education: Teacher training

the scope of interrelated factors they engage the learner with, and even within the content that is presented, many still repeat what has been standard information which modern research is widely known to have shown is not actually so. A case in point is the continual representation of a 3rd tone using a V-shaped mark, when most 3rd tones are actually pronounced with a ‘tub-shaped’ contour which has a very short drop, a lengthened, low, flat middle and a short rising finish (Cao 2002: 94). Despite bright colours and varied layout, a great many resources, including many produced on the Internet, remain not much more than compilations of exemplary grammatical patterns with vocabulary and character lists. There is often no development of the learners as language users, so succeeding chapters are identical in length and demand, with little to engage the mind in reflection on the nature of human language, or strategies for coping with the learning demands on memory. This gap between what is available and what is needed poses a challenge when seeking to help new teachers perceive the learning demands and build a repertoire of content and strategies that will enable their students to meet them. The gap becomes acute when considering teaching for real-life use, such as in intensive programs that use content and language integrated learning. One avenue offering remarkable potential in resource creation is information and communication technology (ICT). From flexibly scaffolded linguistic puzzles to trans-media adventure activities, to real-life contact on Skype, ICT can offer great support in dealing with the learning demands of Chinese. It is essential that beginning teachers steadily develop technical and pedagogical skills in using it effectively.

Educational affordances In most of Western culture, life is seen as ‘.  .  .  a series of situations in which one has to learn to think for oneself, to solve problems on one’s own, and even to discover new problems for which creative solutions are wanted’ (Gardner 1989: 5). Thus the education system in Western countries stresses individual identity and will, and aims to develop students to their fullest potential, able to live successfully in a society where the core value is to stand on one’s own feet. Within such a framework, language learning is valued for having particular potential to develop students cognitively and affectively (Adesope et al. 2010). As well, becoming bilingual has been shown to develop a tolerance for ambiguity and a flexible approach to problem solving, while accepting new ways of speaking and behaving allows appreciation of diversity and difference in thinking, valuing, and organising life. Learning a second language has proven to be influential in development of generic language skills, which can transfer to learners’ use of their first language (Demont 2001; Fortune 2012). This is especially the case in classrooms where metalinguistic knowledge is made available for discussion. Likewise, drawing learners’ awareness to similarities and contrasts between their two languages and cultures makes evident the nature of language and human society. From understanding cultural, historical, and geographical factors involved in producing the social phenomena they are studying, and examination of their own natural, emotional ethnocentricity, language learners can be assisted to develop a more rational acknowledgement of difference, and an inclusive, ethnorelative acceptance of it. These affordances are present in all language learning, but they have a particular significance in school learning, where the aim explicitly is to develop students as persons, not simply train them in a skill. In this regard, learning a language such as Chinese, which demands new development of the self cognitively and affectively, as well as presenting some challenging new ways of looking at the world, has particular potential value. 113

Jane Orton

The culture of education Learner-centred teaching Education is the quintessential cultural process by which a society passes on to the new generation its heritage of beliefs, values, and practices, and at the same time attempts to open these up to modification so students will have a viable life in an inevitably changed future society. The strong Western belief in an individual who has responsibility for his or her life leads to teaching–learning practices and ways of relating in the classroom aimed at fostering students’ ownership of their study and their development of well-reasoned personal opinions, their own stance on social and aesthetic matters, and the capacity to take initiative in responding creatively to novel practical situations. There is an underlying belief in the students’ capacity to discover knowledge for themselves and, indeed, this and the power to critically assess what they are discovering, are what schools set out to develop. Formal learning is undertaken by doing while in conversation with peers as well as teachers, and errors are seen as a natural outcome of experimenting, and treated as useful. From early childhood, living is presented as a series of causally linked intentional actions, a process in which the child is, or can become, an autonomous agent. Problem behaviour is handled by asking the child to reflect on the consequences of his or her actions, then left to choose a productive alternative. While generally accepted views are handed down, students are also expected to challenge propositions, and raising questions is encouraged, even admired. In contemporary education, the capacity to memorise matter is not highly regarded, with value instead placed on students being able to find knowledge when needed. Not having been required to learn much by heart beyond some spelling and the occasional poem, students grow up with very low powers of memorisation. They find the chore of learning the vocabulary and irregular grammatical patterns of any new language both arduous and irksome, and many find the burden of committing to memory the extensive vocabulary and characters of Chinese overwhelming, unless memory is assisted by frequent involvement with the items in mind engaging, puzzle-like activities. In teacher preparation, study of the philosophy and psychology of education is usually confined to specific courses, and in teaching method courses it is the learner-centred principles of practice which derive from these disciplines that are presented as rationale for the proposed teaching activities and learning strategies. There are some potential contradictions inherent in this approach when applied to modern languages education, and in the teaching of Chinese these are usually critical.

Contradictions in approach From the start of any language course, teacher and students are faced with the ambiguity of cultural reference point in their classroom: do they locate themselves in the local society of the school and only ‘visit’ the target language society through the learning resources, or do they locate themselves in a virtual target language society within their classroom? Or do they move between the two? Their position on this is expressed, primarily, in their choice of language for relating to one another and running the lesson: Do they begin with a greeting in English or the target language? Does the teacher become Ma Laoshi or remain Ms McIntyre even when speaking Chinese? Is there any type of activity, or point in the lesson, or place in the room, where only the target language may be used? These days there is not a great deal of difference in the culture of education and teaching between the various European 114

Chinese language education: Teacher training

and English-speaking societies, so that even where the classroom is transformed into a target language-only environment, the actual behaviour of those in the European language room is unlikely to be greatly modified from that being used in the mathematics class next door being run in, for example, English. In the Chinese language classroom this matter is more problematical. This is because the teacher is almost always a first language speaker of the language, who will have been educated in a culture of education and teaching which is different from that of Western societies, in several places to the point of contradiction. Furthermore, it will not only be in regard to relating to one another and organising activities in the room that these differences will appear, but also in their perception of the very nature of their task and the values they ascribe to the various means of working on it (e.g. Haley and Ferro 2011; Grant et al. 2013; Li 2014). Among key beliefs and values found commonly in teachers from China and Taiwan are that learning will be achieved by teachers displaying their own mastery of the subject matter through presenting aloud the information and explanations in the textbook; that language is to be first learned – by which is meant memorised – and only then used; the view that mistakes are bad and errors in language to be corrected at once by the teacher saying the right thing, while the error in the learner’s thinking or skills that led to the mistake is left to the learner to sort out; and the expectation that no questions will be raised, nor will assertions in the book or by the teacher be challenged. A heavy homework load is considered normal and learning is expected to come from committing chunks of language to memory using personally developed strategies and very high repetition of same and similar exercises. Inside and outside the classroom there is little personal contact between teacher and students beyond greetings or reprimands. Unless called on, students are expected to be silent, attentive, obedient, and diligent. Most first language speaker teacher candidates of Chinese approach their teaching in English-speaking countries not only armed with an experience of education throughout their life as described above, but also with the belief that, as the target is Chinese language, it is proper to expect students to learn it in a classroom that reflects Chinese educational culture. Most consider it natural to draw on their own learning experience as a child in which acquiring characters was paramount – 500 per year throughout primary school – and achieved through constant repetitive tracing and copying by strokes done in the correct order, and identifying individual characters on flashcards. As they were taught to as a child, many scorn the Pinyin writing system as babyish, and most have long forgotten it. Like L1 speakers everywhere, without special training, Chinese have difficulty identifying features of their own competently used language. Many cannot identify the tone of syllables they use and can say little that is useful to learners about the system of verb particles. To do so would require linguistic knowledge, deep understanding of the demands it makes on learners, and exceptionally good spoken command of the students’ own first language for explaining and discussing both. Knowledge of the features of the language can be taught and even where it is not part of formal study, most student teachers pick some of it up through textbooks and exposure to experienced teachers during their practicum. But even those who have graduated from Chinese as a Second Language courses in China, and arrive in classrooms overseas with knowledge of the language and its problem spots for foreign learners, have low understanding of the learning challenge as perceived by the learners themselves. Like their compatriots already abroad, most are also quite unprepared for relating personally to students in a firm yet nonauthoritarian way and for presenting their special knowledge so that it interests students and results in solid learning. In schools where Chinese is obligatory, they have few strategies for 115

Jane Orton

dealing with reluctant learners beyond bribing with treats or threatening with tests, and many give up, claiming the students are lazy and undisciplined. Even in primary school programs of only 30 minutes a week there is a relentless slide towards drilling the characters, which most young children eventually find tedious. Unused to the practice of speaking to students about their learning, and often fearful of being misunderstood or mocked for errors in their English, many miss out on the help in solving learning and behaviour problems that students can provide if asked. Outside the classroom, older teachers from China often find their perspectives not shared by local colleagues, so they retreat, participating in school life only where required, and rarely contributing in meetings except over formal matters. Younger, more recently educated L1 speakers are often more able to join in activities and relate to students with some warmth. However, with little prior experience of such an environment, these teachers often find themselves quickly losing control, unable to perceive the quite subtle boundaries for work and behaviour that their local colleagues set successfully. Although outwardly more inter­ active and varied in their teaching than their older compatriots, their view of the fundamental ends and means often remains traditional. As their unrestrained Chinese is not understood by students, most provide an immediate translation of anything they say, and many run their classrooms almost totally in the local first language, for example, English. Second language user teachers are commonly more confident in the social aspects of teaching and are experienced in a culture of education and learning that includes stimulating intellectual curiosity and independent discovery. Unlike their counterparts teaching a European L2 language, however, few have the language proficiency they need to teach in the target language, so while their approach might be advanced, their lessons also are often run largely in their own and the students’ first language.

The goal Teacher preparation in Chinese needs to assist L1 and L2 speaker teacher candidates to become aware of their deeply held beliefs, to perceive the differences and contradictions between them, and through dialogue to forge a platform for practice which they can both espouse. Doing this will mean recognising that educational policy in China has been considerably reformed in the past decade and that, in a break with tradition, it now promulgates an understanding of children’s capacity to learn for themselves and an appreciation that to be successful in the future, students must develop the intellectual curiosity, independence, and creativity cited by Gardner above. It will also mean understanding that the goal for students of the language outside China is not ‘to become Chinese’ but ‘becoming Sinophone’ – international users of the language with a developing bicultural identity (McDonald 2011). From these considerations, teacher educators and candidates will need to decide the question of which traditional Chinese teaching and learning practices are now obsolete, superseded by more creative modern strategies; and which remain cultural constants and hence essential learning experiences for students. Growth in demand for trained teachers of Chinese has strained existing providers and led in places to shortcut courses, a practice very much at odds with internationally accepted professional standards for language teachers. This is all the more risky given that teaching Chinese successfully requires teachers to develop extensive expertise to match the unique demands the language makes on speakers of European languages to develop quite new perceptive and productive capabilities, and manage a severe burden on memory. As well, to date the predominantly first language speaker cohort of teacher candidates and practising 116

Chinese language education: Teacher training

teachers of Chinese approach their task largely on the basis of beliefs and values drawn from their own experience in a very different educational environment from that in which they are teaching. In particular, their fundamental orientation to learning and student–teacher relationships contradict much proposed by Western educational institutions, while even those who would change lack experience of what it might mean. Second language speaker teachers rarely have sufficient language proficiency to teach as they might wish. To be effective, language teacher educators need a deep understanding of the challenges of the Chinese language and the professional skills to assist L1 and L2 teacher candidates to reflect on their understandings and develop dialogue over differences from which they may create a shared, bicultural perspective on their task and its goals, and jointly create the means to achieve them.

References Adesope, O. O., Lavin, T., Thompson, T., and Ungerleider, C. (2010) A systematic review and metaanalysis of the cognitive correlates of bilingualism. Review of Educational Research 80(2): 207–45. American Council on the Teaching of Foreign Languages (2002) Program Standards for the Preparation of Foreign Language Teachers. Alexandria, VA. Australian Federation of Modern Language Teachers Associations (2005) Professional Standards for Accomplished Teaching of Languages and Cultures. Retrieved 31 March, 2015, from http://pspl. afmlta.asn.au/doclib/Professional-Standards-for-Accomplished-Teaching-of-Languages-and-Cultures. pdf Cao, Wen (2002) 汉语语音教程 Hanyu yuyin jiaocheng (A course in Chinese pronunciation). Beijing: Beijing Yuyan Wenhua Daxue Chubanshe. Demont, E. (2001) Contribution of early 2nd-language learning to the development of linguistic awareness and learning to read/Contribution de l’apprentissage précoce d’une deuxième langue au développment de la conscience lingustique et à l’apprentissage de la lecture. International Journal of Psychology, 36(4): 274–85. Fortune, T. W. (2012) What the research says about immersion. In Asia Society (ed) Chinese language learning in the early grades: A handbook of resources and best practices for Mandarin immersion, 9–13. Retrieved 29 March, 2015 from http://asiasociety.org/education/chinese-language-initiatives/ chinese-language-learning-early-grades Gardner, H. (1989) To Open Minds. New York: Basic Books. Gopnik, A., Meltzoff, A., and Kuhl, P. (2000) How Babies Think: The Science of Childhood. London: Weidenfeld & Nicholson. Grant, L., Strong, J., Xu, X. X., and Popp, P. (2013) West Meets East: Best practices from expert teachers in the U.S. and China. Alexandria, VA: ASCD. Haley, M. and Ferro, M. (2011) Understanding the perceptions of Arabic and Chinese teachers toward transitioning into U.S. schools. Foreign Language Annals 44(2): 289–307. Hanban (2007) Standards for Teachers of Chinese to Speakers of Other Languages. Beijing: Hanban. Retrieved 31 March, 2015, from http://english.hanban.org/node_9906.htm Li, X. F. (2014) Reflection on pre-service science teacher education in China and North America. In Wen Ma (ed) East Meets West in Teacher Preparation: Crossing Chinese and American borders. New York and London: Teachers College Press. Lin, P. and Childs, R. (2010) An analysis of orthographic processing: non-Chinese and Chinese readers’ visual-spatial concept. In J. Chen, W. Wang, and J. Cai (eds) Teaching and Learning Chinese: Issues and Perspectives. Charlotte, NC: Information Age Publishing, 257–78. McDonald, E. (2011) The 中国通 or the ‘Sinophone’? China Heritage Quarterly 25. Retrieved 17 May, 2013, from http://www.chinaheritagequarterly.org/tien-hsia.php?issue.025 Miracle, W. C. (1989) Tone production of American students of Chinese: A preliminary acoustic study. Journal of Chinese Teachers Association 24(3): 49–65. Orton, J. (2013) Developing Chinese oral skills – a research base for practice. In I. Kecskes (ed) Research in Chinese as a Second Language. Mouton de Gruyter: Berlin, 9–31. Orton, J. (2014) Chinese as an Object of Learning. Paper presented at CASLAR’s Third International Conference on Chinese as a Second Language Research, 28–30 August 2014, Parma, Italy.

117

Jane Orton Stewart, V. and Wang, S. C. (2005) Take Both Short- and Long-Term Approaches to Create a Supply of Qualified Chinese-Language Teachers. Expanding Chinese Language Capacity in the United States. Washington: Asia Society. Retrieved 29 March, 2015, from http://www.asiasociety.org/files/ expandingchinese.pdf Triskova, Hana (2008) The sounds of Chinese and how to teach them. Archiv Orientalni 76(4): 509–44. Retrieved 29 March, 2015 from http://www.orient.cas.cz/miranda2/export/sitesavcr/data.avcr.cz/humansci/ orient/kontakty/pracovnici/publikace/Triskova/sounds.pdf United Kingdom Government Department for Education (2011) Teachers’ Standards. Retrieved 30 March, 2015, from http://webarchive.nationalarchives.gov.uk/20111218081624/http://tda.gov.uk/ teacher/developing-career/professional-standards-guidance/downloads.aspx Wang, Y., Jongman, A., and Sereno, J. (2006) L2 acquisition and processing of Mandarin tone. In P. Li, L. Tan, E. Bates, and O. J. L. Tzeng (eds) The Handbook of Asian Psycholinguistics. Cambridge: Cambridge University Press, 250–6.

118

8 Chinese language in a global context Liu Jin Georgia Institute of Technology, USA

Tao Hongyin University of California, Los Angeles, USA

1. Introduction As a language spoken by about one fifth of the world’s population (over a billion people), Chinese has always been in contact with other languages and undergoing changes. There is perhaps no better time than the beginning of the twenty-first century to witness the interlocking relation between Chinese and other languages, as the process unfolds in front of our eyes. In this chapter, we undertake the task of describing Chinese in a global context. We concentrate particularly on how globalization has impacted Chinese and how Chinese participates in the process of globalization. After a brief historical account of the contact of Chinese with other languages, we focus on the ongoing impact of English as a global language on Chinese and the perceived crisis involving Chinese. We then summarize a paradoxical phenomenon, namely the rapidly ascending status of the Chinese language worldwide. In the context of globalization, English is often perceived as the global hegemonic language, suppressing and marginalizing other national languages, including standard Mandarin, as local languages. Yet, we will show that the dynamics between the global and the local are complicated by the tension between the national language and the subnational, local dialects in this multilayered linguistic hierarchy. Next we explore the dominance of standard Mandarin and the impact of nationalization and modernization on the Chinese regional varieties. From this perspective, we show that the function of the nation-state seems more and more aligned with globalization and its concomitant homogenization and centralization. Next we illustrate that a variety of diverse, dynamic forces, including the impact of the English language, the increasing popularity of Internet-mediated communication, and the tremendous economic, social, and cultural changes China experiences in its modernization and integration with the world, all give impetus to a series of radical changes in the Chinese language. Finally, we document the emergence of the new linguistic forms and explore their theoretical implications along the axes of standardization and deviation, stability and mobility, and normativity and creativity. 119

Liu Jin and Tao Hongyin

2.  A historical perspective If history is of any guidance, the Chinese language is elastic and adaptive. In the long history of the Chinese language, there have been numerous times where rapid changes took place in the forms of expansion of vocabulary, development of new grammatical markers and syntactic structures, appearance of new phonological features, and emergence of new writing styles. Typically, these rapid and sometimes radical changes correlate with heightened interaction between indigenous and non-Chinese cultural elements. Thus, during the late Han period (second century ce) through the Tang dynasty (tenth century ce), the introduction of Hindu Buddhism brought to China not only Buddhist doctrines and practices, but also an expansion in vocabulary (such as loanwords from Persian, Sogdian, Sanskrit, Mongolian, Manchu, and other Asian languages, see M. Gao and Liu 1958, esp. Ch. 2) and novel syntactic features (as exemplified by the disposal construction, see Cao and Yu 2000). The best-known example, however, belongs to the period between the late nineteenth century and the early twentieth century, which led to the baihua vernacular movement in the May Fourth period. During this period, severance of ties with traditional Chinese culture (as represented by the Confucian tradition) facilitated by major influxes of Western ideas and thoughts resulted in a near total reconfiguration of the landscape of the Chinese language. Not only were writing styles changed, literary forms invented, and syntax increasingly Europeanized, even some basic vocabulary forms and a large number of neologisms were added (especially SinoJapanese-European loanwords) (Gunn 1991; L. Liu 1995). Into Mao Zedong’s era, partly following the former Soviet Union’s model in symbolic and ideological control, the Chinese Communist Party, after it won power in 1949, launched a massive revolutionary program of linguistic engineering, formalization, and orthodoxization, which culminated in the Cultural Revolution (Schoenhals 1992; Wagner 1999; Ji 2004; B. Wang 2011; and Link 2013). This seems an especially apt piece of evidence for Bourdieu’s prediction: Political crises are conducive to verbal explosion (Bourdieu 1977: 663). The beginning of the twenty-first century is remarkable in many ways: Now we are at the start of yet another period of historical significance. As China experiences an unprecedented level of economic and social transformation since its reform and opening-up policy in the late 1970s, and as the world becomes increasingly globalized and interconnected through the use of technology, freer flow of information, and the spread of cultural artifacts, language use in China has also undergone dazzling changes in the past few decades. At the global level is the clash between the ubiquitous global English and the Chinese national language. Within China, the struggle between the national language and regional dialects and ethnicminority languages at multiple levels plays out partly as a matter of nationalization versus the preservation of local identities and heritages. There is also the emergence of the Internet language and other new linguistic forms. Many of them have become creative sources for constructing a distinct identity for the Chinese youth. Moreover, it transcends locality and redefines key theoretical notions in sociocultural linguistics such as the speech community (Gumperz 1968) and linguistic landscape (Landry and Bourhis 1997).

3.  The global and the local: The English craze and the Chinese crisis While there are many major languages in the world with considerable political, cultural, as well geographical clout, English remains the only global language in every sense of the word (Crystal 2012). It is not surprising, then, that English has exhibited profound impact on the Chinese language and the society in which it is used. 120

Chinese language in a global context

Under the reform and opening-up policy, China witnessed a national obsession with learning English, a language denounced as bourgeois and imperialist yet manipulated to perpetuate Maoist worship during the Cultural Revolution. Today, English is viewed as a necessary tool in China’s modernization and integration with the world. For an extended period of time, the subject of English was listed as indispensable and mandatory by college and postgraduate entrance exams. Most elementary schools in urban cities have started to offer English classes. The English as a Second Language (ESL) industry has boomed in the past few decades. According to sources from the National Education Development Statistical Bulletin, the English-training industry in China had an estimated market value of 15 billion yuan (US$2.5 billion) in 2006, and the value doubled to 30 billion yuan (US$5 billion) in 2010 (He 2010). More and more affluent parents can afford to send their children to study abroad for college and increasingly for high school. According to the Institute of International Education, more than 194,000 Chinese students were enrolled in US colleges and universities in the 2011–12 academic year, far exceeding any other country outside the USA. The New Oriental School (Xindongfang xuexiao), which began in 1993 as a program to help Chinese students with the US-based TOEFL and GRE tests, has grown into the largest private language education enterprise in China. The school’s success even became the base storyline for a 2013 domestic hit film, Zhongguo hehuoren (American Dreams in China). The spread of the English language is both a resource for modernization and a source of contention. Chinese people, like others across the globe, have criticized the global dominance of English for suppressing other languages and monopolizing the world’s information. Chinese intellectuals, such as the writer Li Rui (2000), have expressed consternation that the Internet would encourage the global dominance of English and marginalize all other languages. The notion of a ‘Chinese Crisis’ even became the title of a book in 2008, in which Pan Wenguo, a linguist at East China Normal University, outlined five symptoms of Chinese in crisis, the top one being the overwhelming use of English on the Internet and the accelerated cultural globalization mediated in English (Pan 2008; see also Zhu 2005). Although the hegemony of English online has gradually weakened with the proliferation of Chinese-based sites and user bases, the debate between these academics and their opponents remains a hot topic in the media and public discourse. For example, Zhang Shuhua, a dean at the Chinese Academy of Social Sciences and a member of China’s top political advisor body, criticized that the excessive craze for learning English has wasted huge education resources, distracted students from their subject studies, jeopardized educational quality, and weakened students’ grasp of the Chinese language and culture. As he observed, many applicants have been denied admission to postgraduate programs solely due to their failed English tests; English is also an obstacle for working professionals to overcome if they are to get a promotion or salary raise, even though they barely use the language in their workplace. His remarks at a top national political meeting sparked heated debate among the public and triggered nearly 90,000 online comments within just two days (Xinhuanet 2013). Largely due to this resentment, the Chinese government has recently decided to downgrade English scores on the highly competitive national college entrance examination. Of course, this is not to suggest that the Chinese public is going to boycott English any time soon. On the contrary, many shrewd businessmen take the English frenzy as a window of opportunity and act upon it in clever ways. Li Yang, an engineer by training, cashed in on Chinese nationalism in his ‘Crazy English’ English education business, with his slogan ‘conquer English to make China stronger!’ (Woodward 2008, 2012). Other symbolic business changes abound. For example, it has become a common practice for landmark Chinese business entities to adopt an English identity, with one of the more famous cases being the 121

Liu Jin and Tao Hongyin

Tianjin-based restaurant chain that specializes in steamed stuffed buns, Gou Bu Li (lit., ‘dogs don’t bother’), officially adopting ‘Go Believe’ as the store’s English brand name (Meiri xinbao 2008).

4.  The global spread of Chinese While English is the undisputed world language, whether or not Chinese is rising to the status of a global language can be debated. However, there is no denying that the rising profile of China, backed by considerable economic, military, and political power, has popularized the Chinese language across the globe. As Lo Bianco (2012) puts it, Chinese has become a ‘gigantic up-and-comer’. This is of course in stark contrast with the perceived Chinese language crisis that was much discussed in the early 2000s. The numbers are quite striking. It is estimated that there are currently 30 million people around the world learning Chinese as a second language. According to statistics released by the Chinese Ministry of Education, the number of foreign students in China reached a record high of more than 260,000 in 2010 (Chen 2011). Based on the annual report by the US-based Institute of International Education (IIE), the number of Americans studying in China increased by 25% in the 2006–7 academic year over the previous year. Advanced Placement (AP) Chinese exams for US high school students were offered for the first time in 2007. The US government sees Chinese as one of its ‘critical languages’ for security reasons and has thus dedicated enormous resources to Chinese language learning. The Language Flagship Program is one of the federal government’s initiatives to produce professionally proficient speakers of languages critical to the national future. Of the 26 centers and programs, there are 12 Chinese programs, including three pilot Reserve Officers’ Training Corps (ROTC) Chinese programs launched in 2011. Accordingly, the enrollment in the Chinese programs (403 students) accounted for nearly half of the total enrollment in all of the flagship programs (932 students) in 2010. Similarly, the Chinese government sees the surge of interest in Chinese as a unique opportunity to promote Chinese culture and identity across the globe. In 1987, the Chinese government established the Office of Chinese Language Council International (Hanban), which is charged with the task of establishing Confucius Institutes and supporting Chinese language and cultural programs around the world. The institutes are modeled after governmentbacked programs such as France’s Alliance Française, the British Council, and Germany’s GöetheInstitut. While the first Confucius Institute opened in Seoul, South Korea, in 2004, as of December 2014, 475 Confucius Institutes had been established in 93 countries and regions. The rapidly ascending status of the Chinese language has brought nationalistic pride to many Chinese people. With the establishment of large numbers of nondegree and degree programs in teaching Chinese as a second language at Chinese universities and the mushrooming development of language-teacher training programs and private Chinese languagerelated enterprises, the twenty-first century has sometimes been hailed the century of the Chinese language. One of the most popular Chinese songs in 2007 was ‘Zhongguohua’ (‘The Chinese Language’), sung by the pop trio S. H. E. As the lyrics go: ‘The whole world is now learning Chinese; the language of Confucius is going global’. Nevertheless, in her critical reflection on Chineseness as a monolithic, culturally essentialized ethnic supplement to Western hegemony, Chow (2013a: 48) relates the problematic claim of a homogeneously unified, univocal China to the myth of ‘standard Chinese’, which has been affirmed in the pedagogical dissemination of the Chinese language(s) in the West. As she observes, the learning of the Chinese language has often been reduced to the learning of standard Mandarin in the United States, and those professionals who are non-Beijing 122

Chinese language in a global context

Mandarin native speakers, particularly Southern-dialect speakers, have often been discriminated against in the screening of Chinese teachers. The next section will take the multiple Chinese dialects and regional varieties into consideration. We explore the tension between the local and the national and the issues of Chinese language(s), Chineseness, and China in the context of globalization.

5.  Impacts of nationalization and standardization on dialects In linguistic discourse on contemporary mainland China, a distinction is often made between standard Mandarin, or Putonghua (lit. ‘common speech’), and nonstandard local languages and dialects, or fangyan (lit. ‘regional speech’). The classification of these local dialects is still debated, but seven major fangyan groups are conventionally recognized: Mandarin (guanhua or beifanghua), Wu, Min, Cantonese ( yue), Gan, Xiang, and Hakka (kejia); each can be further divided into various subvarieties. Chinese local languages and dialects have long been a fundamental feature of Chinese everyday life and popular culture. The essence of fangyan has been variously identified and characterized as the living, vernacular, or oral language; regional speech; one’s mother tongue; folk language; rural patois; the speech of the uneducated; vulgar slang; unofficial subcultural lingo, and so on. During the twentieth century, they were associated with and simultaneously dissociated from the historical project of building the Chinese nation-state. Although dialects were valued and promoted at various historical moments, building a unified, modern, national language remained the paramount and overarching concern for China in its quest for modernity. The central government has promoted standard Putonghua as the official national language and the principal language for mass media and school education. Consequently, the local varieties have been marginalized and subordinated during this ongoing process of building a modern nation-state, a national culture, and a national language. On October 31, 2000, the Beijing government promulgated its first law on language and writing, the Law of the People’s Republic of China on the Use of Chinese Languages and Chinese Characters (Zhonghua renmin gongheguo guojia tongyong yuyan wenzifa), effective as of January 1, 2001. The law prescribes Putonghua as the principal language for broadcast media, movies, school education, and administration, and it strongly discourages the public use of all other local languages and dialects (except in the non-Han Chinese regions). One of Bourdieu’s (1991) main theses is that the educational system as an institution plays a decisive role in the standardization, legitimization, and imposition of an official language. Kipnis (2012) also argues that educational standardization, including the school use of standard Mandarin by both teachers and students, is an important part of the nationbuilding in contemporary China. Indeed, through formal education, the mass media, and increasingly through administrative measures, the promotion of standard Mandarin has been largely successful. According to an official national survey result on language use in China released in 2004 (Zhongguo yuyan wenzi shiyong qingkuang diaocha), approximately 53% of the population is able to communicate in Putonghua, with urban areas more successful in implementation than rural areas (urban penetration is 66% and rural penetration is 45%) (Yuyan kexue 2005: 112). Even in Hong Kong, where Cantonese has remained predominant, according to the results of the population census in 2011 (14 years after the handover in 1997), about 46.5% of the population claimed to be able to speak Putonghua. While official documents attest to the importance of protecting regional linguistic and cultural heritages, the promotion of Putonghua has led to a sharp decline in the use of dialects, particularly among the urban, school-educated youth. For example, in a survey on 123

Liu Jin and Tao Hongyin

the use of the Suzhou Wu dialect among students in grades two through eleven – 90% of whose parents are native Suzhounese – more than 70% of the students claimed more fluency in Putonghua than in Wu (P. Wang 2003). In another survey on the use of the Zhejiang Jinhua Wu dialect among six- to 14-year-olds, as many as 52% said they could not speak their native dialect, though virtually all could speak Putonghua (jhnews.com 2003). In Shanghai, only about 60% of elementary and middle-school students who were born in Shanghai could speak Shanghai Wu in a survey conducted by the Shanghai Academy of Social Sciences in 2012 (Tian 2012). Forty years ago, this would have been unthinkable. Clearly, the longstanding tradition of diglossia, the use of more than one linguistic variety in a community, has tilted toward the dominance of one single language. For many children born in the 1990s and later, it is no longer their (grand)parents’ dialect but Putonghua which has become their mother tongue. Yet, as is the case with the reaction to the predominance of English, the national domination of Putonghua has not been imposed without challenge in China. Despite the 2001 language law, local media and literary genres produced in regional dialects have flourished in recent years. In television, dialect-based news talk shows that draw on the arts of local traditional story-telling are burgeoning; popular shows have also been localized, as in 2005 when the cartoon series Tom and Jerry was dubbed into around 20 dialects nationwide. In film, internationally renowned director Jia Zhangke’s Hometown Trilogy, largely in Shanxi Mandarin, ushered in a wave of underground and independent films employing dialects to represent the aesthetics of ‘the marginal and the unassimilated’. On the Internet, one can find rap music in dialects, tests on dialect competence mimicking the formats of official English exams, dialect texts parodying Chinese characters and the writing system, blogs employing local slang and expressions, and downloadable mobile ringtones recorded in dialects. In fiction, a number of established writers who previously adhered to standard Mandarin, such as Mo Yan, Jia Pingwa, Yan Lianke, and Zhang Wei, have begun to experiment with writing novels in their native dialects. In the post-socialist reform period, as the aesthetic, entertainment, and commercial value of regional dialects is (re)discovered, the tension between the state, capital, and art is intensified. The rise of regional television shows rendered in dialects, the proliferation of dialects on the Internet, and the commercial success of dialect cultural productions in the local market all attest to the urgency of reimagining a distinct local community that cannot be adequately represented by a single national language (J. Liu 2013). As the nation-state becomes increasingly homogenized and centralized, the use of dialects in mass media and in public spheres asserts the values of pluralism and diversity, and defies the characterization of China as a unified and homogeneous nation-state. The multiplicity of the Chinese dialects is directly involved in the recent academic trend of engagement with and rethinking of the issues of Chinese language, Chineseness, and China. The Chinese language is not a homogeneous entity. Numerous local dialects and varieties are subsumed under this single name and exhibit various degrees of diversity in different areas. While the speakers of the northern Mandarin varieties can largely communicate with each other despite the tonal and lexical differences, linguistic diversity (especially in phonology and lexicon) is much more pronounced in the south, especially in the Min-speaking areas. The multiplicity of Chinese local languages has led to much discussion over how the term fangyan should be translated into English. DeFrancis (1984) discusses the terminological dilemma created by the unique linguistic situation in China. He thinks neither ‘dialect’ nor ‘language’ can justifiably convey the ambiguities and obscurities attached to fangyan. Mair (1991) further examines the problem of the prevalent translation of fangyan as ‘dialect’, an issue he deems 124

Chinese language in a global context

as ‘extraordinarily sensitive’. He develops DeFrancis’s ‘regionalect’ and proposes the term ‘topolect’ as a literal translation of fangyan. Claiming a departure from the umbrella term ‘world/overseas literature in Chinese’ (shijie/haiwai huawen wenxue), the new discipline of Sinophone studies disputes the common practice of equating the Chinese language(s) with Mandarin (or more precisely, standard Putonghua/Guoyu/Mandarin) and defies overgeneralization and essentialism in the assessment of China and Chineseness. Taking Mair’s 1991 article as a major reference, Shih (2007: 189) claims that those southern fangyans that are most frequently the subject of Sinophone studies, including Cantonese, Hakka, and the Southern Min varieties such as Amoy (Xiamen) Min/Taiwanese, Zhangzhou Min, Quanzhou Min, and Teochew (Chaoshan) Min, are not dialects but ‘clearly separate languages from Putonghua’. In fact, the controversy over the relationship between dialect and language is a global and often politicized problem (Hobsbawm 1992: 51–63, 93–100; Haugen 1966). Linguistically speaking, the differences among Chinese fangyans are often conceived to be analogous to those that distinguish the Romance languages in Europe. However, the identification of the former as dialects and the latter as languages is never a purely linguistic issue, and cannot be separated from politics, nation, culture, ethnicity, and other nonlinguistic factors. At the grassroots level, efforts to promote local dialects and to construct a distinct local linguistic and cultural identity have emerged across regions in various forms. For instance, dialect training workshops in the major Wu-dialect metropolises such as Shanghai, Hangzhou, Ningbo, and Suzhou have all been well received. Local newspapers run articles on strategies to preserve readers’ mother tongues against the ‘over-dissemination’ of Putonghua. In Shanghai, a supplementary textbook designed for elementary students was revised in 2005 to include a reading which briefly introduces the history of Shanghai Wu and its vocabulary (Y. Wang 2005). In 2013, roughly 100 elementary and middle schools in Shanghai started to offer Shanghai Wu classes (Y. Liu 2013). Efforts such as these have spurred local legislators in various cities to launch municipal-level initiatives to ‘rescue and protect’ local dialects as ‘endangered cultural heritages’. In the summer of 2010, the pro-Cantonese protests and demonstrations against restrictions on Cantonese in the media, first in Guangzhou and soon echoed in Hong Kong, have been viewed not simply as a linguistic action (Waldron 2010). According to Eng (2010), ‘the size and fervor of the pro-Cantonese movement were derived from well-founded fears that Cantonese was losing ground, and also from legitimate sources of popular dissatisfaction beyond the realms of language policy: collective concerns about the progressive disappearance of Cantonese culture and social dislocations resulting from massive urban renewal and construction’. Just as the title of a popular rap song during the protest suggests, ‘Everything Is Being Dismantled; Cantonese Must Not Be Dismantled’ 乜都拆, 广州话唔可以拆 (‘Mat dou caak, gwongzauwaam hoji caak’, 2010). Nevertheless, the reaction to the protection and revival of regional dialects has been decidedly mixed. In contrast to Cantonese-, Min-, and Wu-speaking areas, which are largely well-developed coastal regions, localities in less-developed regions, which are mostly within or near the Mandarin-speaking sphere, are lukewarm to, if not totally detached from, the dialect crisis. Furthermore, the public use of dialect has always been controversial. A frequently debated topic is how to deal with the relationship between the use of dialect in the media and the overriding state policy of promoting a single national language. For instance, although Lu Chuan, the director of the hit studio film Xun Qiang (Missing Gun, 2002) that employs Guizhou Mandarin, highlights the aesthetic effect of dialect that can capture unique acoustic texture and modulation, a government official accused Jiang Wen, the leading actor of the film, of countering the national language policy and bringing to the 125

Liu Jin and Tao Hongyin

big screen the ‘dregs’ of the dialects (Yin 2002). In addition, the concern for education also stirs controversy concerning the use of dialect in public. From time to time, parents and teachers complain that the media use of ‘vulgar and uncultured’ dialects would be counterproductive to children’s learning Putonghua (in other words, speaking Putonghua is equated with being civilized). However, the dialect promoters often contend that it is equally important to pass on local cultural heritage transmitted in dialect down to the younger generation. The profusion of local languages in mass media has caused the authorities considerable concern. With increasing frequency, the State Administration of Radio, Film, and Television (SARFT) has issued or reiterated regulations to contain the media’s use of dialect and to promote a pure, normative Mandarin. However, this agenda has been continuously challenged by tides of new vocabulary and linguistic innovations that have swept through Chinese society, as the next section shows.

6.  Emergence of new linguistic forms 6.1.  Internet language and youth language Ever since its introduction to China in the early 1990s, the Internet has been developing very rapidly. As of January 2010, according to a state-sponsored study by the China Internet Network Information Center (CNNIC), there were 384 million netizens in mainland China. Various computer-mediated communication formats include emails, e-chatrooms, short message service (SMS) text messaging, BBS forums, social networking websites and smartphone-aided social media (such as renren, weibo [microblog], weixin [wechat]), and other electronic genres. The Internet is primarily a youth culture in China, as in many other countries. According to the same CNNIC study, more than 70% of the Internet users were young people under 35 years of age, and the age group between 18 and 24 years old had consistently accounted for a much higher portion (usually 35–42% between 2000 and 2009) of users than any other age group. Regarding their level of education, approximately half of China’s netizens had a college or associate degree. Roughly one third of China’s netizens were currently students. The rapid development of Internet communication has created a considerable impact on the Chinese language. Researchers have documented that computer-mediated communication (CMC) has helped give rise to a new variety of Chinese language, which generally terms as Chinese Internet language, also rather amusingly known as huoxingwen (language from Mars) (L. Gao 2008, 2012). Young people are the most active creators and trendsetters of the Internet language. The creativity of the Chinese Internet language is mainly manifested at the lexical level. Some common methods to coin new words and expressions include near homophone or phonological approximation (e.g. 神马 shenma for 什么 shenme ‘what’, 有木有 youmuyou for 有没有 youmeiyou ‘do you have  .  .  .  ?’), compounding (e.g. 给力 geili, lit. ‘give power’, now usually ‘cool, awesome, exciting’), phonological fusion (e.g. 酱紫 jiangzi from zheyangzi 这样子 ‘like this, this way’), semantic shift (e.g. 腐败 fubai originally ‘corruption’ and now ‘dine out’), emoticons (e.g. 囧 jiong for ‘embarrassing, in a bad mood’, an esoteric character originally meaning ‘bright [light]’), acronyms (e.g. LP for ‘wife’), script-mixing or script-switching (e.g. I服了U I fule U ‘You’re awesome’), letter and/or number homophones (e.g. 886 ‘bye-bye le’), word decomposition (e.g. 蓝介 lanjie from 尴尬 ganga for ‘awkward, embarrassing’, 竹本犬者 zhubenquanzhe from 笨猪 benzhu for

126

Chinese language in a global context

‘stupid [pig]’), regionalisms (e.g. 俺 an for the first-person pronoun ‘I’, from Shandong and Henan dialects), translation of English words into Chinese characters (e.g. 粉丝 fensi from ‘fans’, 闹太套 naotaitao from ‘not at all’), coinage of English translation for the newly coined Chinese words (e.g. ‘gelivable’ for geili, ‘shitizen’ for pimin [fart people, the common ordinary people]), and so forth. Similar to the Internet language elsewhere, the styles of Chinese Internet language are marked by: (i) a written–spoken hybrid style, where many sociolinguists regard the computermediated discourse as a hybrid register that integrates features of both oral and written discourse (Frehner 2008: 26, 31–5), or as the third genre on the spoken–written continuum (Crystal 2006: 52; Gibbon and Kul 2010: 76); (ii) brevity, which adheres to the Principle of Clarity and Principle of Least Effort, two major principles explaining the dominant economic strategies adopted in digital communication (Gibbon and Kul 2010: 80); (iii) bilingual or multilingual code, which is a distinctive feature of Internet communication; and (iv) a joking style, ‘which is usually achieved through using overly terse and straightforward language, unusual homophones, enigmatic folk similes, and other metaphorical devices’ (L. Gao 2012: 20). The playfulness can be subversive. The Chinese youth challenge and subvert the Chinese characters and the writing system by playing with the characters in transcribing their native dialects online. For example, in using 椅弯撵 yǐwānniǎn for the Shaanxi Mandarin version of 一万年 yíwànnián in Zhou Xingchi’s popular line in his film The Chinese Odyssey, the Chinese character is deprived of its semantic reference function and reduced to a phonetic symbol (J. Liu 2011). It is also noted that sometimes the youth intentionally write incorrect characters as a way to signify a distinct youth identity. Researchers show how misspellings or ‘spelling errors’ are employed symbolically to construct a rebellious youth identity against the mainstream and adult culture, and yield a tension between creativity and prescriptivism (Sebba 2003; Leung 2009). For example, the Chinese Internet youth deliberately write the word 悲剧 (bēijù ‘tragedy’) using the wrong characters 杯具 (bēijù ‘cup set’) in the widely circulated line 人生像茶几,上面摆满了杯具 (Life is like a tea table, all covered in cups [punning as ‘tragedies’]). Internet censorship in China, represented by the Great Firewall of China (or Jindun gongcheng, literally ‘the Golden Shield Project’), has fueled a surge of coded language. In 2009, as an impish protest against the anti-vulgarity campaign which the government launched to tighten political control of the entertainment industry and the Internet, the Chinese netizens coined an innocent, mythical animal ‘grass-mud horse’ (cǎonímǎ, a near homophone of càonǐmā, lit. ‘fuck your mother’), whose lives are threatened by the invasion of river crabs (héxiē, a near homophone of héxié ‘harmony’, a term directly derived from former President Hu Jintao’s regular exhortations for Chinese citizens to create a harmonious society) (Wines 2009). This highly subversive neologism has ever since become an icon of resistance to Internet censorship, and in 2010 the China Digital Times (CDT) launched a project called ‘Grass-Mud Horse Lexicon’ to compile the online resistance discourse. Young netizens’ subversion of the official discourse is very nuanced, packaged, and symbolic. The coded language and metaphors allow them to criticize and ridicule the government, vent out their resentment, and yet avoid outright censorship. Xiao Qiang and Perry Link (2013) argue that the coded language has powerful implications and significance. Taking guiguo (your [honorable] state) as an example, they argue that the netizens’ making distance between themselves and ‘your state’ entails a big question: ‘What do they identify with at the national level? What is it to be Chinese?’

127

Liu Jin and Tao Hongyin

6.2.  English’s impact on the Chinese language The English craze in the Chinese educational institution, the rapid development of the Internet and other high-tech electronic communications, as well as China’s pursuit of modernization and globalization, all contribute to the increasing impact of English, as the global language and the global lingua franca, on the Chinese language. Here are some new linguistic varieties exemplifying its impacts: ● zimuci (letter-word): foreign words (primarily English) and their abbreviations, and pinyin alphabet, written in alphabetic letters and not in Chinese characters, such as ATM, 三G手机 (third-generation mobile phone), and HSK (abbreviation of Hanyu shuiping kaoshi ‘Chinese language standard test’). Letter-words are arguably different from those loanwords that are transcribed into Chinese characters, such as 可口可乐 kekoukele ‘Coco-Cola’. The interference of the two topologically different languages, Chinese and English, is a central topic of interest for linguists. More specifically, the questions of how alphabetic letters integrate with the phonetics and syntax of the Chinese language; how to classify letter-words; what are the graphic and orthographic effects of letter-words on the ideographic Chinese writing; etc. (Kozha 2012: 110). At the discourse level, here are a couple of distinctive examples currently in vogue among the Internet-savvy youth: a) 你out了 ni out le ‘You’re left behind the times’; b) hold 住 hold zhu: ‘hold’ indicates ‘hold fast’, ‘hold your ground’, where the ending sound ‘d’ of ‘hold’ sounds and functions as the Chinese de in the potential complement construction, 住 means ‘live’, ‘stay’, and can be used as a result complement in the de construction, and the entire newly coined construction signifies perseverance, taking charge, being in command of a situation with self-confidence and conviction (Baidu Beat 2012); c) 我吃饭/大哭/ 开会/恋爱 ing wo chifan/daku/kaihui/lian’ai ing (the progressive aspect marker in English), usually in the present tense and meaning ‘I’m eating/crying/having a meeting/ in love’ (Kozha 2012: 118). ● Chinglish: Chinglish is usually defined as a depreciative usage of English by native speakers of Chinese. The increasing presence of public signs with English translations strongly suggests China’s eager engagement with the global tendency of advocating bilingual signs as a means of communication and a sign of social sophistication and cosmopolitanism (Radtke 2012: 151). However, due to the translators’ insufficient English competence, and increasingly due to the use of word-by-word machine translation programs, many public signs are ridiculously translated, and the errors range from simple typos to total gibberish. For example, 财务处处长 (caiwuchuchuzhang ‘chief financial officer’) was translated as ‘Financial affairs is everywhere long’; 残疾人厕所 (canjirencesuo ‘toilet for the handicapped’) was translated as ‘deformed man toilet’ (Hwang 2012: 7, 24). Chinglish may enjoy a slightly Orientalist reputation for the native speakers of English as an entertainment source. A 2011 Broadway play is entitled Chinglish, which was inspired by the playwright David Henry Hwang’s experience of visiting China. However, on the Chinese side, the official Commission for the Management of Language Use finds such mangled English humiliating and feels compelled to clean up English signage, especially prior to large-scale international events, such as the 2008 Beijing Olympic Games or the 2010 Shanghai EXPO (Jacobs, 2010). Global Times, a government organ, ran an article in 2010 titled ‘Chinglish Sentenced to Death’ (Fang 2010). Yet as Radtke (2012: 158–9) argues, for the majority of Chinese within China, the English words placed on billboards, road signs, stores, restaurants, menus, and 128

Chinese language in a global context

product labels may not be intended as a means of communication, but as a decoration for appearing bilingual to attract domestic consumers. In other words, these Chinglish signs are meant for ornamentation, not for information. A more intellectually informed analysis of Chinglish from the perspective of cultural studies can be found in Chow (2013b).

6.3.  Nonstandard Mandarin styles and varieties With the rise of the middle class and new elites in urban cities, the penetration of commercial and consumer-driven culture, and the influence of Hong Kong and Taiwanese popular culture on mainland China, innovative, nonstandard Mandarin linguistic styles and varieties have emerged to respond to the dramatic social and cultural changes and to mark distinctions. Zhang (2005) examines four linguistic variables used by the Beijing yuppies, those cultured cosmopolitans and enterprising professionals working for foreign businesses: rhotacization of the syllable rhyme, interdental realization of (ts), lenition of (sh), and full-tone realization of a neutral tone. She associates the first three local Beijing Mandarin variables with two well-known Beijing character types: the ‘Beijing smooth operator’ and the ‘alley saunterer’. The yuppies try to construct a distinctive cosmopolitan style from their counterparts in the state-owned enterprises, where professionals’ Mandarin carries a prominent Beijing accent. Zhang (2012) explores media hosts’ linguistic styles in consumer-oriented programming in the media, particularly the so-called shishang (trendy) television programs on fashion, consumption, and lifestyles. The hosts draw from linguistic sources, including sounds, words, discourse strategies, regional dialects, English, and other semiotic resources, including makeup, hairstyle, accessories, and clothing, to create a trendy and cosmopolitan persona. She argues that this new linguistic style plays a crucial role in the formation and dissemination of a new consumer culture and the education of the Chinese public to become consumers. Also on media language, Han (2012) examines metaphors used in contemporary enter­ tainment news. She identifies both the common source domains of metaphors (war, martial arts, fire, wind, food, etc.) and target domains (competition, conflict, celebrity, etc.). For example, the metaphor of ‘competition is war’ is illustrated by a news title about the TV entertainment show Happy Boy, quite similar to the American Idol show in format: ‘7 out of 9 candidates will go into the next round of Happy Boy. Today, the smoke of gunpowder is rising again’ (Kuaile Nansheng jiu jin qi xiaoyan jin zaiqi). She points out that metaphors serve multiple purposes: They cannot only embody abstract concepts but also serve as an ideological tool for describing and evaluating people and situations in discourse and influence the viewers’ perception of the world. Metaphors in use can thus serve as a window into the changing media ecology in contemporary China. Television, popular songs, and other media have played a key role in disseminating Hong Kong and Taiwanese popular culture to mainland China. The Cantonese- and Taiwaneseaccented Mandarin (gangtaiyin) that have penetrated the mainland through popular culture since the late 1970s are viewed as ‘cool’ and ‘trendy’ and are avidly imitated by young people in mainland China. Tang (2001) details the words and phrases from Taiwan and Hong Kong appearing in print in mainland Chinese publications. It demonstrates that during the 1980s and 1990s the vast majority of traffic was one-way, from Hong Kong and Taiwan to mainland China, and not vice versa. The soft- and tender-sounding gangtai Putonghua, perceived to be a perfect language for romance, has become the predominant speech style to imitate in mainland drama productions featuring youth idols. 129

Liu Jin and Tao Hongyin

6.4.  Socially embedded language Rapid social changes and network technologies have prompted new words, new expressions, and new structures to emerge literally every day to reflect the dramatically changing social settings in which the language is used in contemporary China. These changes can be outlined as follows. ● Semantic shift/extension: A considerable number of existing words have acquired new meanings. For example, both the addressing terms xiaojie (Ms.) and tongzhi (revolutionary comrade) became sensitive words in post-socialist China, with the former referring to prostitutes and the latter to homosexuals. In a similar vein, laoshi (teacher) has extended its function to now serve as a respectful marker after a person’s (sur)name, regardless of whether or not the person is a teaching professional. ● New ‘social words’: Neologisms are created to indicate social phenomena and events and to signify changing values, identities, and social meanings. To give a few examples, there are an increasing number of accomplished women who are well educated, well employed, but still single at a certain age, and they are called shengnü (leftover women). Fu’erdai and guan’erdai are terms designated for those young people who are the second generation of the rich and powerful. The frequent use of the terms, which are often associated with many negative stereotypes, indicates the growing gap between the rich and the poor and the growing disdain or even hatred of the super-rich and the powerful. Reflecting the social issue of forced property demolitions, the word dingzihu (nail resident) was coined to refer to those who refuse to move mainly because of the unsatisfactory compensation. It is so common to see the character of chai (demolish) painted on walls across the country, prompting some netizens to give the English word ‘China’ a new literal translation as chai na 拆那. ● Grammatical construction: Grammar is usually the least prone to change. So far the best example is the expanded use of the bei construction. bei is a preverbal passive marker for verbs that usually indicate suffering, misfortune, or adversity, and the construction conveys a usually negative, inflictive connotation. For example, wo che bei zhuang le (my car got crashed/got an accident). However, more and more nonpassive verbs, intransitive verbs, and even adjectives appear in this bei construction in contemporary Chinese. For instance, bei daibiao ([one’s opinions] being represented), bei hexie (being harmonized/censored), bei juankuan (being forced to donate), bei zengzhang ([income] being increased in statistics), bei jiuye (exaggerated report about graduates’ employment), bei zisha (a lawsuit settled by claiming to be a suicide case), and bei xingfu (being happiness-ified) are all commonly found on the Internet. Drawing on the construction’s ‘involuntarily passive, and compulsory’ implication, these expressions serve to reflect the absence of subjectivity and civil rights in contemporary China on the one hand and to express the frustration of being underprivileged and powerless in a very much top-down system found in China on the other.

7.  Theoretical implications of the new forms 7.1.  A reaction to the single, standard Putonghua The Beijing novelist Wang Shuo (S. Wang and Lao 2000: 209) points out that, in the contemporary post-socialist period, standard Putonghua in mainland China has evolved into an overly politicized language, featuring a hollow, exaggerated emotionalism, a harsh 130

Chinese language in a global context

and aggressive rhetoric, and a standard, authoritative ‘broadcasting intonation’. The linguist S. Robert Ramsey (1987: 47) aptly describes this ‘broadcasting intonation’ as: ‘All radio Peking announcers, both men and women, broadcast in a pitch range noticeably higher than that of their normal speaking voices. Each sentence begins high and shrill. Then pitch falls gradually, reaching a lower key by the end of the sentence. Pauses are exaggerated, and the normal rise of a nonconcluding clause becomes longer and more drawn out. The devices of this strident intonation may well be borrowed, in part, from traditional Chinese drama and opera; but their use in the media today seems intended to arouse in the audience an impression of struggle and determination.’ Similarly, the Yunnan poet Yu Jian (2004: 137) finds that when he speaks in Putonghua, he becomes a different person who ‘has no sense of humor and is self-abased, nervous, stuttering, and pretentiously serious’. He argues that from the perspective of sociolinguistics, Putonghua is not just a tool for nationwide communication, but rather a ‘social dialect’ that is highly politically charged. This ‘social dialect’ is best suited to mass mobilization; metaphysical spirituality; abstraction; central state-sanctioned ideological and literary orthodoxy; propagandistic eulogy in the public sphere; grandiose, heroic, and utopian narrative; formal diction and power; and revolutionary discourse. In this sense, the emergence of diverse linguistic styles, registers, nonstandard accents, dialects, and features outlined in the previous sections can be seen as a reaction to the single, official, standard Putonghua.

7.2.  The tension between mobility vs. stability and creativity vs. normativity Similar to the tension between the expanded use of dialect in the media and the tightened SARFT regulations, the emerging linguistic forms participate in a linguistic battle with the authorities and formal institutions as well. The trend of the popularity of gangtai accents among the mainland youth led to a SARFT regulation in 2005, stipulating that broadcasting hosts should always use standard Mandarin and should stop adopting Hong Kong or Taiwanese slang and accents. The language censors are also worried that the expanded use of English words and Roman letters is sullying the purity of the mother language, Chinese. In 2010, the General Administration of Press and Publication (GAPP), China’s regulator of news, print media, and Internet publications, stipulated that standardized Chinese should be the norm: The press should avoid foreign abbreviations and acronyms, as well as Chinglish. In his influential article entitled ‘Questioning the “standardization of modern Chinese language”,’ the linguist Qian Nairong (2004) argues about the constantly changing nature of language and the impossibility of language standardization and purification. The accusation against nonstandard usages represents a romanticist view of language purity which cannot be upheld in modern society and would make Putonghua a stagnant, minimalist, artificial, and dead language. He calls for an open and liberal mind to approach the emergent new linguistic varieties and to foster linguistic multiplicity and diversity. Zhang (forthcoming) incisively contends that underlying such regulations is a strategy of de-legitimization and ‘othering’ by the government and official institution, positioning ‘nonstandard’ accents and writings as inauthentic and thus ‘wrong’. In this way, Putonghua, based on Beijing Mandarin in the north, is construed as the center of the Chinese linguistic hierarchy, enjoying hegemony and prestige. So fundamentally this is a dynamic struggle between normativity and creativity, standardization and deviation, the center and the periphery, the official mainstream discourse and the unofficial popular discourse. The linguistic battle over enregisterment 131

Liu Jin and Tao Hongyin

reveals that what is really at stake is the location of the cultural center of modern Chineseness. The state’s project of a modern, globalizing China firmly locates the center on the mainland, whereas alternative visions of modern Chinese value transregionality, hybridity, and diversity. Nevertheless, language is in constant flux. Some of the new usages may be transient, while some may evolve into the core of the Chinese language and signify diachronic linguistic change in progress. As noted, at least three Internet words have been included in the 5th edition of Xiandai hanyu cidian (Dictionary of Modern Chinese, 2005), one of the most authoritative dictionaries on the modern Chinese language: yimeier ‘email’, ku ‘cool, awesome’, and heike ‘hacker’. The Internet buzzword in 2010, geili (lit. ‘giving power,’ meaning ‘cool, awesome, exciting, cooperative’), was granted the ‘official seal of approval’ by appearing in People’s Daily, the official paper of the Communist Party (Shanghai Daily 2010). In addition, the latest 2010 edition of the Oxford Chinese–English, English–Chinese Dictionary, the most widely used Chinese–English reference book, included trendy lingo such as shanzhai (‘cheap copy’, like a knock-off mobile phone), fangnu (mortgage slave), and shengnü (old maid) (The Independent 2010). In terms of grammar, the use of nouns as adjectives has been documented in Xiandai hanyu cidian (2005), for example, yangguang ‘sun’ as ‘healthy and optimistic’. In summary, most linguistic innovations can, at least partially, be attributed to mobile and network technology. While technology is not the only factor giving rise to the large volumes of linguistic innovation, it plays a vital and integral role in the process. A strong corollary of mobility is creativity: New technologies create new space and new communication modes which were not possible in the past and which foster user participation and fuel creativity. On the other hand, there is always the tension between creativity and normativity, where the forces to preserve regularity and tradition are at work. The interaction among these parameters, mobility and stability, creativity and normativity, and deviation and standardization, has resulted in a number of features characterizing the language of the new millennium and at the same time has created significant tensions involving linguistic identity, social commentary, and cultural politics.

8. Conclusions When the Chinese language is viewed in the context of globalization, we are presented with an extremely dynamic and complex landscape of language use and change. Out of the complexity, however, some major patterns emerge: 1. Language contact has been a constant phenomenon in the history of Chinese. What is really remarkable about contemporary Chinese under globalization is the extent of the contact with other languages (chiefly English) and the degree of impact that has been incurred on Chinese. 2. While Chinese often plays a receiver role in the global stage of language interaction, it simultaneously spreads its own influence behind a rising economic, political, and increasingly military superpower. This is just another reflection of the multitude of world powers that characterize the world scene in this age of globalization. 3. Language as a symbolic system is always shaped by factors both inside and outside itself. In the case of globalization, such extralinguistic factors as social–cultural identity, tradition and progress, conservation and creativity, technological advancement, and so forth, all prove to be key in defining what to change and how changes take place. 132

Chinese language in a global context

4. While we can perhaps never predict the end result, if there is ever such a thing as an end result, we can be certain that the Chinese language will continue to evolve as it interacts with the rest of the world languages and negotiates status among its subnational varieties. The key is to identify the underlying factors and understand how they impact the way the Chinese language evolves.

Note This chapter synthesizes our work on this topic with additional materials and observations. Most of the previous work can be found in our co-edited book, Chinese under Globalization: Emerging Trends in Language Use in China (J. Liu and Tao 2012). More specifically, Section 2 is from pages 1–2 and Sections 3–5 is the revised and updated version of pages 204–10 in the volume. We wish to thank Like Li for her meticulous editorial assistance with a draft copy of this chapter. However, we are solely responsible for the final shape of the chapter.

References Baidu Beat (2012) ‘Baidu’s Top 10 New Internet Phrases of 2011’, available at: http://beat.baidu. com/?p=3515. Bourdieu, Pierre (1977) ‘The Economics of Linguistic Exchanges’, Social Science Information 16(6): 645–68. Bourdieu, Pierre (1991) Language and Symbolic Power, edited by J. B. Thompson, Cambridge: Polity Press. Cao, Guangshun and Yu Hsiao-jung (2000) ‘The Influence of Translated Later Han Buddhist Sutras on the Development of the Chinese Disposal Construction’, Cahiers de Linguistique Asie Orientale 29(2): 151–77. Chen, Jia (2011) ‘Expat Student Numbers Rise’, China Daily, March 4. Chow, Rey (2013a) ‘On Chineseness as a Theoretical Problem’, in Shih Shu-mei, Tsai Chien-Hsin, and Brian Bernards (eds) Sinophone Studies: A Critical Reader, New York: Columbia University Press, 43–56. Chow, Rey (2013b) ‘The Subject of Chinglish’, Testo a Fronte: teoria e pratica della traduzione letteraria 48 (1 semestre 2013): 155–62, in Tim Parks and Edoardo Zuccato (eds) Special issue: Towards a Global Literature/Verso una Letteratura Globalizzata. Crystal, David (2006) Language and the Internet, 2nd ed., Cambridge: Cambridge University Press. Crystal, David (2012) English as a Global Language, 2nd ed., Cambridge: Cambridge University Press. DeFrancis, John (1984) The Chinese Language: Fact and Fantasy, Honolulu: University of Hawaii Press. Eng, Bob (2010) ‘Is Cantonese in Danger of Extinction? The Politics and Culture of Language Policy in China’, available at http://chinamusictech.blogspot.com/2010/08/is-cantonese-in-danger-ofextinction.html. Fang, Yunyu (2010) ‘Chinglish Sentenced to Death’, Global Times September 29. Frehner, Carmen (2008) Email, SMS, MMS: The Linguistic Creativity of Asynchronous Discourse in the New Media Age, Bern and New York: Peter Lang. Gao, Liwei (2008) Chinese Internet Language: A Study of Identity Constructions, Munich: Lincom Europa. Gao, Liwei (2012) ‘Synchronic Variation or Diachronic Change: A Sociolinguistic Study of Chinese Internet Language’, in Liu Jin and Tao Hongyin (eds) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific, 7–28. Gao, Mingkai 高明凯 and Liu Zhengtan 刘正埮 (1958) Xiandai hanyu wailaici yanjiu《现代汉语外 来词研究》(Studies of Loanwords in Modern Chinese), Beijing: Wenzi gaige chubanshe. Gibbon, Dafydd and Małgorzata Kul (2010) ‘Economy Strategies in Restricted Communication Channels: A Study of Polish Short Text Messages’, in J. Normann Jørgensen (ed) Vallah, Gurkensalat 4U and me! Current Perspectives in the Study of Youth Language, Frankfurt am Main: Peter Lang, 75–98. Gumperz, John (1968) ‘The Speech Community’, in David Sills (ed) International Encyclopedia of the Social Sciences, New York: Macmillan, 381–6.

133

Liu Jin and Tao Hongyin Gunn, Edward (1991) Rewriting Chinese: Style and Innovation in Twentieth-century Chinese Prose, Stanford: Stanford University Press. Han, Chong (2012) ‘The Metaphorical World of Chinese Online Entertainment News’, in Liu Jin and Tao Hongyin (eds) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific, 29–57. Haugen, Einar (1966) ‘Dialect, Language and Nation’, American Anthropologist 68(4): 922–35. He, Na (2010) ‘Rush to Learn English Fuels Quality Issues’, China Daily August 5. Hobsbawm, Eric J. (1992) Nations and Nationalism Since 1780: Programme, Myth, Reality, Cambridge: Cambridge University Press. Hwang, David Henry (2012) Chinglish: A Play, New York: Theatre Communications Group. Jacobs, Andrew (2010) ‘Shanghai is Trying to Untangle the Mangled English of Chinglish’, New York Times May 2. Jhnews.com (2003) ‘Jinhua fangyan, ping shenme bei paoqi’〈金华方言，凭什么被抛弃〉(‘Why the Jinhua Wu Dialect Has Been Abandoned’), available at: http://www.jhnews.com.cn/gb/content/ 2003-04/18/content_172194.htm. Ji, Fengyuan (2004) Linguistic Engineering: Language and Politics in Mao’s China, Honolulu: University of Hawaii Press. Kipnis, Andrew B. (2012) ‘Constructing Commonality: Standardization and Modernization in Chinese Nation-Building’, The Journal of Asian Studies 71(3): 731–55. Kozha, Ksenia (2012) ‘Chinese via English: A Case Study of “Lettered-Words” as a Way of Integration into Global Communication’, in Liu Jin and Tao Hongyin (eds) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific, 105–25. Landry, Rodrigue and Richard Bourhis (1997) ‘Linguistic Landscape and Ethnolinguistic Vitality: An Empirical Study’, Journal of Language and Social Psychology 16(1): 23–49. Leung, Genevieve Y. (2009) ‘Hong Kong University Students’ Language Use in Blogs: Tensions between Creativity and Prescriptivism (“omg my grammar is rubbish”)’, Sino-Platonic Papers 189: 30–62. Link, Perry (2013) An Anatomy of Chinese: Rhythm, Metaphor, Politics, Cambridge, MA: Harvard University Press. Li, Rui 李锐 (2000) ‘Wangluo shidai de fangyan’〈网络时代的方言〉(‘Dialect in the Age of the Internet’), Dushu《读书》(Reading) 4: 42–7. Liu, Jin (2011) ‘Deviant Writing and Youth Identity: Representation of Dialects with Chinese Characters on the Internet’, Chinese Language and Discourse 2(1): 58–79. Liu, Jin (2013) Signifying the Local: Media Productions Rendered in Local Languages in Mainland China in the New Millennium, Leiden: Brill. Liu, Jin and Tao Hongyin (eds) (2012) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific. Liu, Lydia H. (1995) Translingual Practice: Literature, National Culture, and Translated Modernity – China, 1900–1937, Stanford: Stanford University Press. Liu, Yilin 刘轶琳 (2013) ‘Benshi yiyou 100 duosuo zhongxiaoxue kaishe “Shanghaihua” ke’〈本市已有 100多所中小学开设《上海话》课〉(‘More Than 100 Elementary and Middle Schools in Shanghai Have Offered “Shanghai Wu” Classes’), available at: http://sh.eastday.com/m/20130524/u1a7412644.html. Lo Bianco, Joseph (2012) ‘Chinese: The Gigantic Up-and-Comer’, in Linda Tsung and Ken Cruickshank (eds) Teaching and Learning Chinese in Global Contexts, London and New York: Continuum, xiii–xiv. Mair, Victor (1991) ‘What is a Chinese “Dialect/Topolect”? Reflections on Some Key Sino-English Linguistic Terms’, Sino-Platonic Papers 29: 1–31. Meiri xinbao (2008) ‘Tianjin goubuli xuangua yingwen zhaopai Go Believe’〈天津狗不理悬挂英文 招牌 Go Believe〉(Tianjin Famous Chain Store Goubuli Puts on the English Brand Name Go Believe)’, available at: http://news.sina.com.cn/s/2008-01-10/180214714025.shtml. Pan, Wenguo 潘文国 (2008) Weijixia de Zhongwen《危机下的中文》(Chinese in Crisis). Shenyang: Liaoning Renmin Chubanshe. Qian, Nairong 钱乃荣 (2004) ‘Zhiyi “xiandai hanyu guifanhua” ’〈质疑《现代汉语规范化》〉(‘Questioning the “Standardization of Modern Chinese Language” ’), Shanghai wenxue《上海文学》(Shanghai Literature) 4: 78–83. Radtke, Oliver (2012) ‘More Than Errors and Embarrassment: New Approaches to Chinglish’, in Liu Jin and Tao Hongyin (eds) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific, 145–70.

134

Chinese language in a global context Ramsey, S. Robert (1987) The Languages of China, Princeton: Princeton University Press. Schoenhals, Michael (1992) Doing Things with Words in Chinese Politics, Center for Chinese Studies Research Monographs 41, Berkeley: Institute of East Asian Studies, University of California. Sebba, Mark (2003) ‘Spelling Rebellion’, in Jannis Androutsopoulos and Alexandra Georgakopoulou (eds) Discourse Constructions of Youth Identities, Amsterdam and Philadelphia: John Benjamins Publishing Company, 151–72. Shanghai Daily (2010) ‘The Rise of Net Buzzword “Gelivable” ’, available at: http://www.china.org. cn/china/2010-11/12/content_21327924.htm. Shih, Shu-mei (2007) Visuality and Identity: Sinophone Articulations across the Pacific, Berkeley: University of California Press. Tang, Zhixiang 汤志祥 (2001) Dangdai hanyu ciyu de gongshi zhuangkuang jiqi shanbian: 90 niandai Zhongguo dalu, Xianggang, Taiwan hanyu ciyu xianzhuang yanjiu《当代汉语词语的共时状况 及其嬗变: 90年代中国大陆,香港,台湾汉语词语现状研究》(The Synchronic state of contemporary Chinese vocabulary and its transformation: Studies in the situation of vocabulary in mainland China, Hong Kong, and Taiwan during the 90s), Shanghai: Fudan University Press. Tian, Xiaoling 田晓玲 (2012) ‘Zhongxiao xuesheng chengzhang qingkuang diaocha baogao fabu’ 〈中小学生成长情况调查报告发布〉   (‘The Survey of the Elementary and Middle School Students in Shanghai Is Released’), Wenhui bao May 11. The Independent (2010) ‘Dictionary Adds Chat-room Chinese Words That Are Simply “Niu” (awesome)’, available at: http://www.independent.co.uk/arts-entertainment/books/dictionary-adds-chatroomchinese-words-that-are-simply-niu-awesome-2074467.html Wagner, Rudolf 瓦格纳•鲁多夫 (1999) ‘Zhonggong 1940–1953 nian jianli zhengyu zhengwen de zhengce dalüe’〈中共 1940–1953 年建立正语、正文的政策大略〉(‘An Overview of CCP Policies in Establishing an Orthodoxical Language and Discourse in 1940–1953’), in Peng Xiaoyan 彭小妍 (ed) Wenyi lilun yu tongsu wenhua《文艺理论与通俗文化》(Literary Theory and Popular Culture), Taipei: Zhongyang yanjiuyuan Zhongguo wenzhe yanjiusuo choubeichu, 11–38. Waldron, Arthur (2010) ‘Will Linguistic Centralization Work? Protesters Demonstrate against Restrictions on Cantonese’, Jamestown Foundation, China Brief, 10(16), August 5. Wang, Ban (ed) (2011) Words and Their Stories: Essays on the Language of the Chinese Revolution, Leiden: Brill. Wang, Ping 汪平 (2003) ‘Putonghua he Suzhou Hua zai Suzhou de xiaozhang yanjiu’〈普通话和 苏州话在苏州的消长研究〉(‘A Study of the Rise of the Putonghua and the Decline of Suzhou Wu Dialect in Suzhou’), Yuyan jiaoxue yu yanjiu 1: 29–36. Wang, Shuo 王朔 and Lao Xia 老霞 (2000) Meiren zeng wo menghanyao《美人赠我蒙汗药》(A Beauty Presents Me with a Sleeping Potion), Wuhan: Changjiang wenyi chubanshe. Wang, Youjia 王有佳 (2005) ‘Shanghai: fangyan jieshao wenzhang “Nong xiaode fa” jin keben re zhengyi’〈上海:方言介绍文章《侬晓得伐》进课本惹争议〉(‘Shanghai: An Article Introducing Shanghai Wu ‘Do you know’ Evoked Controversy’), People’s Daily, September 6. Wines, Michael (2009) ‘A Dirty Pun Tweaks China’s Online Censors’, The New York Times March 12. Woodward, Amber R. (2008) ‘A Survey of Li Yang Crazy English’, Sino-Platonic Papers 180. Woodward, Amber R. (2012) ‘Learning Chinese to Promote Chinese: A Study of Li Yang’s Crazy English’, in Liu Jin and Tao Hongyin (eds) Chinese under Globalization: Emerging Trends in Language Use in China, Singapore: World Scientific, 127–43. Xiao, Qiang and Perry Link (2013) ‘In China’s Cyberspace, Dissent Speaks Code’, Wall Street Journal January 4. Xinhuanet (2013) ‘Chinese Craze for English Debated’, available at: http://news.xinhuanet.com/english/ china/2013-03/14/c_124460602.htm. Yin, Tingting 尹婷婷 (2002) ‘Guojia yuwei fuzeren piping Jiang Wen jiang yuyan ‘zaopo’ banshang yinmu’〈国家语委负责人批评姜文将语言《糟粕》搬上银幕〉(‘The Official from the State Language Commission Criticizes Jiang Wen for Bringing Linguistic Dregs to the Silver Screen’), Chengdu ribao《成都日报》(Chengdu Daily) September 2. Yu, Jian 于坚 (2004) ‘Shige zhi she de ying yu ran: shige yanjiu cao’an: guanyu dangdai shige de lianglei yuyan xiangdu’〈诗歌之舌的硬与软——诗歌研究草案:关于当代诗歌的两类语言向度〉 (‘The Hard and Soft of the Tongue of Poetry – A Draft of Poetry Study: On Two Different Directions in the Language of Contemporary Poetry’), in Yu Jian Jujue yinyu: Zongpi shouji, pinglun, fangtan 《拒绝隐喻:棕皮手记,评论,访谈》(Refusal   of Metaphor: Brown Notebook, Criticism, and Interviews), Kunming: Yunnan renmin chubanshe, 137–51.

135

Liu Jin and Tao Hongyin Yuyan kexue 语言科学 (Language Science) (2005) ‘Zhongguo yuyan wenzi shiyong qingkuang diaocha jieguo zai Beijing gongbu’〈中国语言文字使用情况调查结果在北京公布〉(‘The Result of the National Survey on Language Use in China Released in Beijing’), 3: 112. Zhang, Qing (2005) ‘A Chinese Yuppie in Beijing: Phonological Variation and the Construction of a New Professional Identity’, Language in Society 34(3): 431–66. Zhang, Qing (2012) ‘ “Carry Shopping Through to the End!”: Linguistic Innovation in a Chinese Television Program’, in Juan Manuel Hernandez-Campoy and Juan Antonio Cutillas-Espinosa (eds) Style-shifting in Public: New Perspectives on Stylistic Variation, Amsterdam and Philadelphia: John Benjamins Publishing Company, 205–24. Zhang, Qing (forthcoming) Sociolinguistic Variation in China: Undoing Commonness through Cosmopolitan Mandarin, New York and London: Routledge. Zhu Jing (ed) 朱竞 (2005) Hanyu de weiji《汉语的危机》(The Chinese Language Crisis), Beijing: Wenhua Yishu Chubanshe.

136

9 Chinese language pedagogy Wu Weiping Hong Kong polytechnic University, Hong Kong

Although Chinese language pedagogy (CLP) can be viewed from various perspectives (Xing 2006; Y. Wu 2012), a diachronic approach is adopted in this chapter so that the current situation in this field can be presented. Synchronically, some key stages of CLP will also be mentioned briefly to provide a vertical dimension for the current picture. Within the diachronic approach, the discussion here will center around two major trends of practices in the science and art of Chinese language teaching: one emphasizing language structure, while the other language use. Obviously, pedagogy will differ depending on whether we view Chinese as a Native Language (CNL) or Chinese as a Second Language (CSL). For discussion purposes, the assumption here is CSL if not otherwise noted, since this is a major area in most research and publications. There are of course similarities in pedagogy no matter whether it is for native speakers or for second language learners, such as the importance attached to certain salient features of the language (e.g. tone as a meaning-indicating feature in Chinese), but comparison between CNL and CSL approaches is not part of this chapter. In the course of the discussion, key issues related to the two trends in terms of the purpose, mode, scope, focus and assessment of CSL teaching will be explored.

1.  Chinese language pedagogy focusing on language structure (LS) Language teaching used to be viewed as any other teaching activities over the years, with the primary purpose of knowledge transmission from the teacher to the students. Learning a language, therefore, involves the study of the structure of that language. All pedagogical issues under such an approach naturally boil down to the three major components of a language in terms of its structure: namely, phonology, semantics and syntax. A syllabus focusing on language structure in teaching CSL may have a variety of forms and shapes, but Figure 9.1 is typical of most syllabi under such an approach. In Figure 9.1, we could further describe each step: 1. Presenting knowledge of language structure: Examples intending to show basic rules of the language in terms of its phonological, semantic and syntactic structure; 2. Explaining language structure illustrated by the examples, whether inductive or deductive in various attempts to provide explanations by teachers, with rules governing the structure of what we see and hear in communication by language as their final destination; 137

Wu Weiping Presentation of LS (be it sound, vocabulary or grammar) →Explanation of basic units in LS →Practices on selected units based on LS →Feedback on LS →Test on what is learned as part of LS

Figure 9.1  Major steps in CSL teaching syllabi focusing on language structure (LS)

3. Practicing by students, both in the classroom or in the form of homework, focusing on comprehension and correct use of language structure (e.g. sentence making with given vocabulary or key words in sentence patterns); 4. Checking, by teachers or their designated representatives who are assumed to possess the knowledge being transmitted, on the correctness of the language forms produced by the students; 5. Finding out what the students cannot do in an achievement test based on what has been taught, pointing out errors spotted and (occasionally) reminding students of contexts in language use. In step 1, basic units of the phonological features of the language (i.e. initials, finals and tones in Chinese), common vocabulary items (either based on structure of the characters that make up the words, or on various kinds of frequency counts) of the characters used and fundamental grammar rules (e.g. from simple to complex sentences in Chinese) are presented. By step 5, most people following such a structure-based syllabus, both teachers and students, would consider the job done because examinations are held and results obtained (W. Wu 2008).

1.1.  Language as a system of systems Nobody can talk about the way of doing something if that ‘something’ is not well-defined or clearly understood. While exploring various ways to teach Chinese focusing on the structure of the language, we tend to consider language as a system incorporating three subsystems. The sound system in Chinese, for example, consists of a definite number of initials, finals and tones (Norman 1988). This is a closed system from a diachronic perspective, meaning all the components of this system are relatively stable at any given time. The vocabulary, as a system to express meaning by words, is at the other end of the spectrum in terms of stability. While the majority of the words may have little change over a significant period of time, some words will become obsolete and eventually die and new words will pop up from time to time. Thus this system is viewed as an open system, in the sense that it will keep changing regardless of any conscious efforts from teachers, researchers or linguists. Grammar, or the rules for putting words together to make meaningful expressions, is something in between: not as stable as the sounds of language but with less change when compared with vocabulary. Typically, teachers deal with each of the subsystems in different ways pedagogically.

1.2.  Pedagogy in phonology Methods and skills in teaching the Chinese sounds may share many similarities with those in teaching other languages, but no pedagogical approach would be complete without considering tones as an integrated part of every syllable, a unique feature of the Chinese language. 138

Chinese language pedagogy

Chinese is often considered as a syllabic language in which each syllable can be decomposed into three components: initial, final and tone. Pictographically, it is usually represented by one character in the Chinese writing system, though some of the syllables used orally may not have their character representation. There are all together 23 initials, 34 finals and four tones (or five if ‘light tone’ is counted). Teaching the sound system is a multilayer task. In terms of comprehension, teachers need to make sure that students understand the rules of combining initials and finals together, the sound pitch and contour associated with each tone, as well as possible tone sandhi for certain combinations in connected speech. They also need to train their ears so that they can distinguish the sometimes subtle differences among certain minimum pairs, such as the retroflex /sh/ versus the non-retroflex initial /s/. In terms of production, teachers would try their best to cultivate accurate pronunciation among their students, usually by a combination of efforts including trial and error, imitation, error analysis and a formidable amount of exercises. Some of the more effective pedagogical tools are listed below: ● Use of visual aids such as charts and symbols, referring to articulating organs ● Audio-video assistance making use of modern technology ● Listen and imitate, the most basic and effective approach, may still be the best way in the words of many students (Xing 2006) To most learners, it is hard enough to distinguish the four tones and their numerical values (level 55, rising 25, dipping 214 and falling 52) (Chao 1968), let alone understanding the subtle and yet important differences between each minimum pair. Moreover, the step following comprehension and production, the ability to use them in connected speech for communication purposes, takes much more effort. Even though each and every student knows the value of the four tones, trying to get them right while using a random combination of them often proves to be too much.

1.3.  Various teaching approaches to vocabulary Unlike the sound system, vocabulary is an open system and only a selected number of items in this system are taught in any language program. Compared with the other two subsystems (sound and grammar), the teaching of vocabulary turns out to be most complicated because of many disputed concepts. To start with, people disagree on the line between a character and a word, which is largely responsible for the on-going debate over the character-based, or zibenwei (Bellassen 2002), versus word-based, or cibenwei, pedagogy (Liu 2010). The former insists that characters are the very basic units of all teaching activities, while the latter argues that words, which are often combinations of characters, should be such basic units. No matter which line they stick to, however, the common pedagogical approach is to teach a set of commonly used characters (or words) by the 3P model: Presentation of sample items by teachers, Practice of various kinds (individual, paired, group, teacher–students, student–students, etc.) and Production by students, which usually consists of using the newly learned vocabulary items in sentence-making exercises to show that they can use these words ‘correctly’ in a given language structure. As always, there are people who take the middle of the road and attempt to combine the strength from both sides. Then they argue about which set of vocabulary is most common, less common or uncommon, and by what criteria they are so labeled. The reason behind such argument is the belief 139

Wu Weiping

that the most common ones should be taught first. The problem is, there are quite a few vocabulary lists based on some kind of frequency count. Since the same item may occupy a different place in different lists, which list should be used becomes a hard choice in teaching-material preparation and in teaching-activity design for pedagogical purposes. To make matters worse, such frequency counts often come from written data, so they are far from satisfactory when it comes to courses and materials focusing on listening and speaking (Tao 2011). A large-scale corpus for spoken language in the natural setting has yet to be built and, without the support of such a corpus, frequency counts for spoken language are impossible. Obviously, what is frequently used in written materials may not be frequent at all in the spoken form. Thus relying on the ‘most frequent’ vocabulary in making pedagogical decisions at this stage can be rather misleading. Finally, scholars and practitioners, under the influence of different theories they believe in, tend to disagree on how the vocabulary should be presented when teaching them. Among the factors that affect pedagogical decisions is the major issue of the role of Romanization in vocabulary learning. Romanization is not a natural part of the Chinese language for native speakers, but an artificial tool created to facilitate the learning process. As such, it is often regarded as a crutch in most pedagogical approaches. Many teachers would start with it to help the students master the pronunciation of the vocabulary and make the link between sound and meaning, and let it fade out as the number of words remembered by the learner has increased to a considerable size (which again differs from program to program). The period in which such a crutch is used varies, from one or two weeks to several months. This explains why we often encounter the question: ‘When will characters be introduced in the program?’ At one extreme end of the spectrum, there are also programs in which Romanization is just pushed aside and the vocabulary learning process starts with making the link between meaning and shapes of the characters used to represent the words.

1.4.  Teaching grammar Grammar may mean totally different things in the eyes of a research scholar and a teacher. The former tries to find a system that can explain all linguistic phenomena of the language (research grammar), while the latter focuses on some basic rules that can guide the learners (pedagogical grammar) in their efforts to put words together to express meaning (Lu 2006). Teaching grammar from a pedagogical perspective, therefore, involves only the task of identifying a set of basic rules for teaching and how to make the students understand these rules. This may sound simple but is again rather complicated in reality. It is not hard at all to imagine that reaching an agreement in identifying the set of rules is mission impossible to start with. All language programs are built upon certain concepts about language and each individual teacher will always have his or her own view about grammar. A basic rule considered by Teacher A as part of the set for teaching may or may not be recognized by Teacher B in the same way. Unlike vocabulary where everybody agrees what a vocabulary item should be. The major debate usually centers on where such an item should stand in the frequency account and when is the best time to introduce it to the learners. The world of grammar is much more complicated, with parts of speech, prosodic features, normal SVO patterns, special syntactic structures like the ba-pattern (Feng 2013). Thus trying to select certain features to be included in teaching is already a formidable task in itself. When it comes to pedagogical approach in teaching grammar in CSL programs, the general agreement is to downplay issues related to parts of speech and tense, and stress issues related 140

Chinese language pedagogy

to basic and special sentence patterns and aspects. In addition to the introduction of general concepts such as the SVO order in sentences, including simple and some complex ones in which conjunctions are used, most programs will pay particular attention to special sentence patterns in Chinese, such as the ones with ba, le (or double-le), bei, the ‘left-branching’ structure with -de between the modifying phrases and clauses and the head noun, the differences in word-order between Chinese and Indo-European languages when time and place adverb or adverbial phrases are used, and other features such as the use of a designated modifier before most countable nouns. All these are proven to be difficult spots for most learners who are not familiar with the Chinese language and culture and tend to attract more attention pedagogically from teachers. Similar to the concept of character-based and word-based, there is also pedagogical approach advocating ‘sentence-based’, or jubenwei. Scholars and teachers holding this view argue that a sentence is the most basic unit in language because only a complete sentence can express meaning, however short it may be (Zhang 2006). In short, the ideal way to teach grammar is to have a ready set of rules recognized by all teachers so that every teacher is clear about what to teach. Such a set of rules, again ideally, will be used in all teaching materials. Moreover, the order of appearance for rules in this set will represent the accepted order of acquisition by most learners based on sound research in CSL acquisition. The reality in teaching CSL, however, is far from ideal due to the complexity of various issues related to Chinese grammar.

1.5.  Characters and reading in Chinese If the three subsystems discussed above, sound, vocabulary and grammar, are common to all languages, the writing system in Chinese is something unique and therefore requires a unique approach in teaching. In essence, the primary task in all language learning is to establish the link between sound and meaning in speaking and listening, and between the writing system and meaning in reading and writing. What makes CSL special is the lack of relationship between sound and the writing system for most learners. The huge number of Chinese characters, which developed from the tradition of meaning-expressing pictographs, is indeed overwhelming for both teachers and students. The learning process is greatly hindered in most regular programs because of the high threshold: around 3,000 individual characters to be memorized, recognized and used in reading and writing. No matter what pedagogical tricks experienced teachers can play – by meaning-expressing radicals, by soundexpressing components or by association, with flash cards, low-tech writing brushes to motivate interest, or a high-tech IT platform to woo the attention of younger learners – the amount of time that can be given in any program is much less than needed to master this basic set of characters, let alone the combination of individual characters in this basic set to form a much greater number of words and phrases. Apart from the built-in challenge of learning such a writing system, the linguistic and cultural backgrounds of the learners may also lead to differences in pedagogical approach in teaching. For Asian learners whose languages and cultures are somehow related to the use of characters, including Japan, Korea, Vietnam and Singapore (Cui and Chen 2000), the path of learning can be steep and the speed of learning can be fast, with reading and writing exercises at a very early stage of the program. For learners without any background in a writing system relying on ideographic and pictographic symbols, they need a much gentler approach, with a very limited number of characters in each lesson, a well-structured reappearance interval and a much slower speed in progress (An 1988; Jiang 2007). 141

Wu Weiping

As for the differences in the two varieties of some commonly used characters, namely the simplified and the traditional form, most discussions and approaches over the years are tinted by politics, as well as cultural constraints to a certain degree. Without going into any details, and therefore debates, we can simply state that the norm now accepted by many is to teach simplified characters in mainland China and to teach traditional ones in Taiwan. Where outside forces in politics and culture are not homogenous, powerful or dominant enough and teachers and programs can put academic choice first, such as in North America or Hong Kong, they do have the choice but with it comes the dilemma: meeting the needs of one group means running the risk of neglecting the wish of the other. It was under such circumstances that the pedagogical approach to ‘recognize the traditional but use the simplified’ came about, and practiced with varying degrees of success in many programs in these areas. Before we move to describing a different approach based on language use instead of language structure, we should at least mention the relationship between the approaches in teaching the four basic components in CSL as outlined above, and the training of four basic skills (listening, speaking, reading and writing) among learners of the language. Such a relationship is illustrated below: Phonology: courses focusing on listening and speaking Vocabulary: courses related to all the four basic skills Grammar: courses related to all the four skills Characters: all courses focusing on reading skills The realization of such a link is important, especially with increasing attention among scholars and practitioners to ‘pedagogical models’ in CSL teaching, as indicated in more and more discussions by key players in the field (Liu 2014).

2.  Chinese language pedagogy focusing on language use (LU) Language structure and language use are two different matters and, therefore, each requires a different pedagogical approach. Needless to say, one cannot have French fries if there is no potato. No matter how many potatoes you may have, on the other hand, they cannot become French fries if you don’t go through the required cooking process. Such a process also exists between the language structure you have and the language you use. You need to put the language structure you have learned in the appropriate setting so that the right meaning can be sent and received.

2.1.  The concept of Pragmatic Framework in CSL pedagogy If we really follow the pedagogical approach in which language use is the goal, we need to understand the process that makes what we hear and see meaningful. That means the significance of all the contextual factors in language use, namely the people who use the language, the context in which we use the language, and the timing or purpose related to the language being used (Bates 1976; Hymes 1972; Levinson 1983; Omaggio 1986). All these factors form what we call ‘Pragmatic Framework’ (W. Wu 2008), which should also be included as part of the pedagogical activities and be taught or acquired by language learners, in addition to the teaching of language structure that we are all familiar with. The importance of such a framework should prevent us from treating it as an add-on that we remind our students of at the end of the teaching process, and instead ensure that it is an integrated part of a pedagogical approach throughout the whole learning process, as explained below. 142

Chinese language pedagogy Identify factors involved in language use (starting point of the learning process) →Presentation of pragmatic points as the basic units in LU →Explanation of both structural and contextual factors in LU →Practices and feedback based on LU →Test focusing on LU

Figure 9.2  Major steps in CSL teaching syllabi focusing on language use (LU)

First of all, the presentation unit will be based on language use instead of language structure. Instead of the phonological features covering the sound system, selected semantic units as represented by words, and a representative set of sentence patterns based on the syntax, the unit representing the system of language use is what we call a ‘pragmatic point’ (W. Wu 2006), which is defined as a linguistic function in context. What we want the students to understand is that such a function includes all the structural information (sound, word and grammar), as well as the contextual information (who to whom, where and when) Then, the major stages of learning, as indicated in Figure 9.2, can each be handled pedagogically based on research findings in the CSL field. Each of the steps in Figure 9.2 can be described as follows: 1. Recognizing key factors in language use, in addition to language structure, especially subtleties related to who, where, when and how in a communication event, all of which are major components of the pragmatic framework; 2. Presenting authentic or semi-authentic materials in language use; 3. Explaining the components for the basic unit of language use, which consists of linguistic knowledge and pragmatic knowledge; 4. Practice in classroom, in the form of homework or in real-life contexts as part of the curriculum design, focusing on both language structure and language use, with frequent feedback on student performance, aiming at both appropriateness and correctness; 5. Finding out what the students can do by performance-based assessment tools. Now that the major steps in the teaching process are established, we can explore the pedagogical approach in each stage within the process. Such a process is here described as ‘counter-clockwise’ because assessment is put at the starting point of the learning process, rather than at the end as normally found in most curricula.

2.2.  Establish the link between purposes of teaching and assessment The approach of respecting and making use of the power of tests (Shohamy 2001) tries to establish the link between the purpose of learning and the assessment procedure. If we need to evaluate the actual ability of communication, then we need a direct test to gather information about the actual ability of the learner (Bachman and Palmer 1996; Byram 1997). In other words, ability is the aim of testing for ability-oriented programs, and knowledge is the aim of testing for knowledge-based approaches. So, a proficiency-oriented performance test would be the natural choice for programs focusing on language use, rather than an indirect test of language structure and, based on the result of such a test, ‘estimate’ the ability of the learner. One common practice in language programs is to start with teaching and end with a test to see if what has been taught has been learned. With the general tendency to avoid ‘teaching for testing’, most teachers would avoid placing too much emphasis on testing. Such 143

Wu Weiping

a precaution is quite understandable given the fact that most of the programs and teaching materials are structure-based. With a direct test aiming at the actual ability to use language (here we use spoken language and oral proficiency tests as examples for discussion), a task bank with real-life situations can be created. Since each task in the test is created as a real life, or similar to a real life task, the ability to perform such a task is deemed as close, or very close, to the real-life ability of the learner (Candlin 1987; Chi 1996). It is with such an assumption that a properly designed assessment tool becomes the start point rather than the end point of the learning process. That does not mean the first step for all learners is to take a test. A proficiency test, by definition, can be taken any time. The purpose of putting together a sizeable task bank is to use it as a guide for stages to follow. Just like a selected number of grammatical features are treated as a representation of the grammar system in terms of language structure, a group of real-life tasks, also known as pragmatic points, are treated as a representation of language use. Here we must reiterate the fact that, for many learners, they are learning in order to use the language, and the practice of presenting units of language use for learning purposes, instead of language structure, is not hard to accept. This group of tasks, each of them associated with a linguistic function, is then divided into categories and grouped into proficiency levels according to their relative degrees of difficulties. Each level serves as a subsystem of language use, similar in relation to the situation in which phonology is a subsystem of language structure. Borrowing the concept and the terminology used in the Speaking Proficiency Guidelines by the American Council on the Teaching of Foreign Languages (ACTFL 2012), each pragmatic point in the database belongs to one of the three proficiency levels: Intermediate, Advanced or Superior. A sample of linguistic functions commonly used in our daily life is provided in Table 9.1. In general, learners with higher proficiency in the language they learn will perform better than those with lower proficiency. The logic behind an assessment tool built on linguistic functions is the belief that: (i) using a linguistic function to perform a task in daily life will inevitably involve language structure, including sounds, utterances and discourse if we speak, and words, sentences and paragraphs when we write; and (ii) linguistic functions found in high proficiency levels (Superior and Advanced) tend to involve a larger variety of grammatical structures, vocabulary items and rhetorical devices, while those at the lower end of the proficiency levels will involve a smaller number of structural varieties, which also tend to be simpler in nature from the perspective of most users (Ellis 2003; Willis and Willis 2007).

Table 9.1  ACTFL levels with sample linguistic functions Level

Linguistic function

Context

Sample content

Novice

listing counting

Informal setting

Specific, memorized material

Intermediate

giving information asking questions

Informal to semi-formal

Specific, personal content ‘me’

Advanced

narrating a story summarizing an event

Semi-formal/ formal

Specific/abstract, concrete world

Superior

supporting an opinion making a speech in public

Formal and official

Abstract concepts, controversial issues

144

Chinese language pedagogy

Each task in the data bank, regardless of the level in which it is placed, has some specific factors from the pragmatic framework related to the linguistic function, including: 1. People/Interlocutor: factors such as age, gender, social status, one to one, one to many, one for one or one for many (on behalf of ) and so on; 2. Place/Setting: formal, informal, semi-formal, official, friendly, public, private, and so on; 3. Timing/Purpose: right or wrong moments for the interlocutor with their identity and social status, intended or not intended occasion for what is being discussed, and so on. With all these complex factors, as opposed to pure language structural ones, a lot more needs to be planned pedagogically in each of the following stages: curriculum design, teaching-materials preparation and teaching activities.

2.3.  Curriculum design based on expected outcome In terms of curriculum design, the focus is on the structure of the whole program and the variety of courses that will serve the ultimate goals of learning. Like all other language programs, knowledge of language is the foundation of any proficiency level that we expect our students to achieve (Krashen 1987; W. Wu 1993). That means learners will have to spend a reasonable amount of time trying to learn about rules and basic facts in pronunciation, vocabulary and grammar. The key difference in a program that claims to focus on language use is the emphasis on the context in which language is used. To better serve the needs in cultivating the communicative ability of the students, courses focusing on language use should be created and treated as an indispensable part of the curriculum. Since each and every task created for the data bank in assessment mentioned in the last section already has its built-in pragmatic clues in language use, the same database is used as the basis in designing the curriculum with all the courses. In other words, all the tasks, as summarized in Table 9.2, would somehow become the major focus of certain courses in the curriculum. Each category of linguistic functions in Table 9.2 has a certain number of specific tasks (between 5 and 50). Many of the tasks are used in this stage to form the skeleton of different courses at different levels in the curriculum. In the next stage, the content area found in each task will be used as a starting point in compiling teaching materials. While a detailed discussion of the whole curriculum is not feasible here, a brief description of one new course can be cited as an example to show how language use can become the focus in the learning process. In light of the need to remind students of their major goals in learning, a special course, to be conducted in real-life situations rather than the classroom, can be created. One feature of such a course is that students are required to participate in learning activities located in a Chinese speaking environment (e.g. cities like Beijing or Dalian where standard Chinese is widely spoken), with field trips and presentations, rather than classroom instructions and explanations. All the key elements of the course, from location of class and teaching materials to requirements for the students, are put together to give one message: learning is for using and ways to use can become ways to learn. The emphasis of the course is on cultivating learners’ sensitivity to contextual factors, which are always available in actual communication. Their sensitivity to appropriateness according to the situation in which they use the language, and their awareness that the learning process will not be complete until they are able to use the language in real life are also developed in the learning process. 145

Wu Weiping Table 9.2  Summary table of pragmatic points in the system of language use PL

Task categories

Tasks

Notes

I

 I1.  I2.  I3.  I4.  I5.  I6.  I7.  I8.  I9. I10.

50 10 10 10 40 10 15 15 25 10

All the 70 tasks under categories I1 to I3 are picture-based. Categories I4 to I10, with a total of 125 tasks, are I-level linguistic functions with basically informal contexts in various content areas.

 A1. 投诉/complaining  A2. 分析/analyzing  A3. 比较/comparing and contrasting  A4. 阐释/expounding  A5. 陈述/narrating  A6. 建议/suggesting  A7. 拒绝/refusing  A8. 推销/promoting  A9. 致歉/apologizing A10. 批评/criticizing A11. 介绍(半正式)/introduction (semi-formal)

15 50 50 35 20 30 20 10 10 15  5

All the tasks at the A-level categories, with 11 linguistic functions covering 260 scenarios with various content areas, are created with settings of a semi-formal nature, such as business or social encounters of a business nature.

  S1.   S2.   S3.   S4.   S5.   S6.   S7.   S8.   S9. S10.

35 40 10 10 10 10 10 10  5  5

Based on the 10 linguistic functions at the S-level, 145 tasks are created to cover the needs for language use in the formal setting, including some representative speech events of a public nature.

中 级

A 高 级

S 特 级

描述/describing 说明/explaining (picture-based) 指路/providing directions 留言/leaving messages 介绍(日常生活)/introduction (daily life) 感谢/expressing thanks 说明/explaining (word-based) 解释/illustrating 讨论/discussing 提供信息/offering information

发表见解/expressing an opinion 会议发言/speaking at meetings 公开致谢/expressing thanks in public 劝说/persuading 反驳/rebutting 祝贺/congratulating 演讲/delivering a speech 辩护/defending 号召/calling (to action) 介绍(正式)/introduction (formal)

(PL: Proficiency Level; I: Intermediate; A: Advanced; S: Superior)

2.4.  Teaching-materials preparation according to curriculum In a language program that emphasizes learner ability in oral communication, students would find it absurd to have their final grade based on a written test. Similarly, a program with focus on language use may not fare well with teaching materials built on language structure alone (Li 2004; Zhao 2007). Thus the principle in teaching-materials preparation designed to be used with such a pedagogical approach is to have the pragmatic framework as the top layer, with representative pragmatic points as the anchor tasks in the system of language use. Components of language structure, including phonology, vocabulary and grammar, are then used as subsystems attached to the top layer, as indicated in Figure 9.3. As seen from the structure in Figure 9.3, the top layer is the system of language use with pragmatic points as units of presentation; structural knowledge commonly found in traditional 146

Chinese language pedagogy

System of language use: as represented by real-life tasks

Phonology

Vocabulary

Grammar

Figure 9.3  Guiding structure in teaching materials preparation

textbooks are then organized as the second layer. The phonological component, as mentioned at the beginning of this discussion, is a closed system that includes all the sounds and basic tones of the Chinese language. All of them can be distributed in the initial lessons of the textbook in a contextualized situation. The semantic and grammatical components, on the other hand, consist respectively of a selected list based on frequency of usage, or structural complexity if applicable.

2.5.  Teaching training and teaching activities All pedagogical approaches are realized by teachers, even if we follow the student-centered approach. For it needs the teachers to make sure that the class is student-centered. It is from this perspective that we say teacher training plays a crucial role in all pedagogical activities. Teaching CSL has been around for many years, with ‘new’ theories replacing old ones from time to time. New pedagogical approaches would naturally come with new theories, but actual teaching practices are always a little bit behind. This is unavoidable because it takes time for teachers to first understand the theories, then persuade themselves to familiarize new types of activities and exercises with new pedagogical approaches, and finally spend time to either modify their current practice to implement new tactics, or start afresh, to learn new tricks and trial them out in their teaching practice. Due to the need for such a ‘transforming’ process, it takes time for most CSL teachers to cross the gap in pedagogy from the language structure side to the language use side. A systematic approach in teacher training to narrow that gap and make the crossing easier is, therefore, something we could not neglect at this stage (W. Wu 1995, 2014). Key issues in such a training program may be many, as illustrated by examples from one of the in-service training models outlined below. Current practices in teacher training for language instructors tend to give priority to what we call ‘microlevel training’, especially with basic skills and techniques used in the classroom. There is no guarantee, however, that a teacher who performs well in the classroom can also do well if learning happens beyond the classroom, which becomes more and more common in programs focusing on language use. Familiarization and explanation of fundamental concepts associated with new theories should form a major part of any training program. If teachers are expected to willingly conduct all pedagogical activities in programs focusing on language use, then they need to understand the theories and concepts behind such an approach. They need to re-examine their views on the nature of language, the nature of teaching and learning activities, and the aim of language learning for most students. 147

Wu Weiping

e.g. Macrolevel training focusing on fundamental concepts ● ● ● ● ● ●

Teaching Chinese as a Second Language: Concepts, trends and directions Speaking Chinese: A pragmatic approach Teacher training focusing on assessment Teacher training focusing on curriculum design Teacher training focusing on teaching-materials preparation Teacher training focusing on teaching activities

Together with macrolevel training in which the pedagogical approach is designed to improve the ability in language use, issues related to specific skills are still part of the training. In general, this kind of microlevel training can be divided into two categories. One will focus on the pedagogy related to the three language components (sound, word and grammar) and four skills (listening, speaking, reading and writing), as illustrated by the examples below. e.g. Microlevel workshops focusing on language components and language skills ● ● ● ● ● ● ●

Phonology: How to teach sounds and tones Vocabulary: Inter-language as indicated by vocabulary Grammar: How to handle difficulties in grammar under the Pragmatic Framework Listening: Varieties of listening comprehension in the IT era Speaking: Extended discourse for advanced learners Reading: Reading for speaking Writing and script: Stories and learning strategies for Chinese characters

The other category includes some basic techniques in teaching, which has been a major component in most training programs and is always welcome by new teachers. In addition to this, however, some sizeable program or institution also has specialized training related to their own program, known as Program Specific Teacher Training (PSTT). Without such a component, it would not be easy to maintain the characteristics, and therefore effectiveness and reputation of their program (Ji 2006). That is why quite a few of the long-standing and reputable summer CSL programs (e.g. Middlebury, Princeton in Beijing) conduct their own PSTT each year. e.g. Microlevel training focusing on basic techniques and program/course-specific features ● ● ● ● ● ● ● ●

Techniques: How to handle Q and A in class Techniques: Assignments, quizzes and tests Techniques: Student-centered classroom activities Techniques: Internet as a teaching platform PSTT: Characteristics of program focusing on language use PSTT: Special course for individualized training PSTT: Sector-specific language training PSTT: Summary on feedback from class auditing (as a means of quality control)

3.  Concluding remarks The on-going discussion of pedagogical models in the CSL field has witnessed many controversies, doubts and heated debates from time to time because of the lack of consensus 148

Chinese language pedagogy

about what a model should be in the first place (Brown 2001; Ma 2004). Some define a model from the learning environment, some from the procedure and framework for activities, while others would focus on key elements of such a model, thus leading to various models with four, five or six elements (Y. Wu 2012). Needless to say, each and every model may have its own strengths and weaknesses and, depending on the needs of the learning community, are either recommended or rejected by certain programs and institutions. Thus the first concluding remark we can draw is: There is no ideal pedagogical approach for all language programs. CSL in China in its native linguistic and cultural environment, for example, cannot rely on the same approach as CSL in places where Chinese is a foreign language. A CSL program designed to train scholars in Chinese linguistics, on the other hand, will not adopt any pedagogical model for programs focusing on cultivating the speaking ability of diplomats. A close look at various models also reveals the fact that certain features are indeed shared by almost all models no matter which school of theory they come from. So the second concluding remark we can draw is: There are key elements shared by many different approaches, although each model may have its own priority for these elements and each of the elements in any given model may differ in its relative importance within the model. With these two concluding remarks in mind, we will see that any search for the ideal pedagogical model will most likely be in vain. We may also realize that the best and most appropriate model for any CSL program will only become available once we clearly understand the specific traditions and objectives of our own program. Based on such understanding, the best option for us is to create our own model or to modify existing ones to meet our needs. Scholars in assessment once claimed that all tests are specific tests (Douglas 2000); we can also claim here that all good CSL programs are the ones with specific objectives and pedagogical approaches that serve such objectives. With such realization, our attention may then shift to two key questions in our quest for the best pedagogical model: 1. What are the shared elements in CSL pedagogy? 2. What are the factors that affect both the priority and relative importance of these elements in any given model? Based on what has been discussed, either in this chapter or beyond, among the shared elements in any CSL pedagogical model are issues related to what to teach (the materials, either in the form of a traditional textbook or an electronic platform), how to teach (teaching activities and methodology, including teacher training), as well as the criteria and procedures by which the effectiveness and results of teaching and learning are assessed. As for the second question, we can identify at least three factors that will affect what we teach, how we teach and why we assess our teaching and learning the way we do. The first is the primary purpose of the program. The second is the resources available, including time and availability of learning platform demanded by the pedagogical model. The third is not easy to pinpoint, but can be described as the degree to which the professional team can perform as expected, including the knowledge structure, expertise and dedication of the teachers in the team. Finally, we must say that various answers to these two questions can also be found in all the changes we have seen over some key stages in the development of CSL pedagogy. Regarding the mode of teaching, we saw the change from 3Ps focusing on language knowledge to task-based activities. In the scope of teaching, we witnessed the move from structure-only to structureplus contexts. We watched the focus of teaching shift from teacher-centered to 149

Wu Weiping

student-centered, and we have realized that, in assessment, we now rely more and more on information from proficiency tests, rather than just clues from course-based achievement tests. Instead of predicting the trends of development in our endless quest for the best pedagogical model, we can just stick to the hope that all CSL programs will always find the most appropriate model, as long as key players in the program are clear about what they want to do.

References ACTFL (2012) ‘ACTFL Proficiency Guidelines’. Available at http://actflproficiencyguidelines2012.org/ speaking. An, Z. (1988) ‘Chinese Characters Revisited’, in Collection of Papers from the 2nd International Conference on Chinese Language Teaching, Beijing: Beijing Language College Press. Bachman, L. and A. Palmer (1996) Language Testing in Practice, Oxford: Oxford University Press. Bates, E. (1976) Language and Context: The Acquisition of Pragmatics, New York: Academic. Bellassen, J. (ed) (2002) On Characters and Words, Beijing: Peking University Press. Brown, H. D. (2001) Teaching by Principles: An Interactive Approach to Language Pedagogy, Beijing: Foreign Language Teaching and Research Press. Byram, M. (1997) Teaching and Assessing Intercultural Communicative Competence, Clevedon: Mulitlingual Matters. Candlin, C. (1987) ‘Towards Task-based Language Learning’, in C. Candlin and D. Murphy (eds) Language Learning Tasks, Englewood Cliffs, NJ: Prentice Hall International, 5–21. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley, CA: University of California Press. Chi, T. R. (1996) ‘Toward a Communicative Model for Teaching and Learning Chinese as a Foreign Language: Exploring Some New Possibilities’, in S. McGinnis (ed) Chinese Pedagogy: An Emerging Field, Columbus, OH: Ohio State University Foreign Language Publications, 1–27. Cui, Y. and Chen X. (2000) ‘Analysis of Factors Affecting the Study of Chinese Characters by Learners without Character Background’, in Zhao L. and Huang G. (eds) The Application and Spread of Chinese Characters, Huayu Jiaoxue Press. Douglas, D. (2000) Assessing Languages for Specific Purposes, Cambridge: Cambridge University Press. Ellis, R. (2003) Task-based Language Learning and Teaching, Oxford: Oxford University Press. Feng, S. (2013) Chinese Prosodic Grammar, Beijing: The Commercial Press. Hymes, D. (1972) ‘On Communicative Competence’, in J. Pride and J. Holmes (eds) Sociolinguistics, Harmondsworth: Penguin, 269–93. Ji, C. (2006) ‘On the Construction of CSL Pedagogical Models: With Reference from the Teaching Practice in Middlebury’, Hanyuxuexi 4. Jiang, X. (2007) ‘Chinese Characters Teaching for Western Learners: Separation of Recognition and Writing, Read More and Write Less’, in A. Gu et al. (eds) Chinese Character Cognition and Teaching: Collection of Papers from International Conference on Character Learning by Western Learners, Beijing: Beijing Language and Culture University Press. Krashen, S. D. (1987) Second Language Acquisition and Second Language Learning, Oxford: Pergamon. Levinson, S. (1983) Pragmatics, Cambridge: Cambridge University Press. Li, Q. (2004) ‘CFL Teaching Materials: For Whom?’ Shijie Hanyu jiaoxue 2. Liu, S. (2010) ‘On the Relationship between Character-based and Word-based Pedagogy’, Huawen jiaoxue yu yanjiu 1. Liu, S. (2014) ‘Topic Studies on TCSOL Models’, TCSOL Studies 2. Lu, Q. (2006) A Practical Pedagogical Grammar in Teaching Chinese as a Foreign Language, Beijing: Peking University Press. Ma, J. (2004) ‘Research in Chinese Pedagogical Models’, Yuyan jiaoxue yu yanjiu 1. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Omaggio, A. C. (1986) Teaching Language in Context: Proficiency-oriented Instruction, 2nd ed., Boston: Heinle and Heinle Publishers, Inc. and Oxford University Press. Shohamy, E. (2001) The Power of Tests: A Critical Perspective on the Uses of Language Tests, London: Pearson Education.

150

Chinese language pedagogy Tao, H. (2011) Working with Spoken Chinese, State College, PA: Center for Advanced Language Proficiency Education and Research (CALPER) Publications. Willis, D. and J. Willis (2007) Doing Task-based Teaching, Oxford: Oxford University Press. Wu, W. (1993) ‘Towards a Theory of Teaching Chinese as a Second Language’, Springfield, VA: ERIC Document Reproduction Service, ED 366 216, Paper based on Presentation at Chinese Language Teachers Association (CLTA) Annual Meeting, San Antonio, Texas. Wu, Weiping (1995) ‘Education of Second Language Teachers: The Link between Linguistic Theory and Teaching Practice’, in Proceedings of the Georgetown University Round Table on Language and Linguistics, 480–97. Wu, Weiping (2006) ‘Pragmatic Points in Teaching Chinese: A Practical Approach’, Chinese Teaching in the World 1: 91–6. Wu, Weiping (2008) ‘Pragmatic Framework and its Role in Language Learning: With Special Reference to Chinese’, in W. Chan, K. Chin, M. Nagami, and T. Suthiwan (eds) Processes and Processorientation in Foreign Language Teaching and Learning, Germany: De Gruyter Mouton. (Reprinted 2011). Wu, Weiping (2014) ‘A Pragmatic Approach in Teacher Training: Ideals and Reality’, in Lee, S., C. Xie, and W. Wu (eds) Linguistics and CSL Teaching: Theories and Practices in Cultivating Pragmatic Ability, Hong Kong: The Commercial Press. Wu, Y. (ed) (2012) Pedagogy in Teaching Chinese as a Foreign Language, Hong Kong: The Commercial Press. Xing, J. (2006) Teaching and Learning Chinese as a Foreign Language: A Pedagogical Grammar, Hong Kong: Hong Kong University Press. Zhang, P. (2006) ‘ “Sentence” Is the Basic Unit of Language’, Hanzi Wenhua 3. Zhao, J. (2007) ‘New Pedagogical Model and Teaching Materials Preparation’, in Collection of Papers from the 8th International Conference on Teaching Chinese.

151

10 Chinese linguistics William S.-Y. Wang Hong Kong polytechnic University, Hong Kong, china

Preamble ‘All the world’s a stage,’ Shakespeare noted in As You Like It, ‘and all the men and women merely players; they have their exits and their entrances.’ The theme of this chapter is Chinese linguistics. We will need a stage, upon which the players are to speak their various languages. Therefore we begin our discussion with a description of the stage, which is the physical geography of China with its tall mountains and great rivers. We will then see that the earliest players to have entered the stage from Africa over 500,000 years ago, Homo erectus, were creatures quite distinct from the ones we see now. The modern players, Homo sapiens sapiens, also from Africa, had two major entrances, by a coastal route from the south, and by a northern route across the mountains of West Asia. This dual origin of the diverse Chinese peoples from ancient times is clearly reflected in the genes they carry and the many languages they speak. We have continuous specimens of the languages of China that reach back in time over 3,000 years, which is unique among the languages of the world. Concern with language may be traced to the philosophical observations of Kongzi1 (Confucius 孔子, 551–479 bce), who emphasized the pragmatic importance of words with the term 正名 ‘correct names’. In contrast to Shakespeare’s Juliet, who said famously, ‘What’s in a name? that which we call a rose by any other name would smell as sweet’, Kongzi saw words as the all-important foundation of correct behavior, a view he expressed in the following injunction: 名不正则言不顺; 言不顺,则事不成; ……2 ‘When names are not correct, what is said will not sound reasonable; when what is said does not sound reasonable, affairs will not culminate in success;  .  .  .’ There are many modern examples of Kongzi’s concern with using the right word. During a period of racial tension in the USA, it was critically important to distinguish the words ‘Negro’, ‘Black’, ‘Afro-American’. Sensitivity to women’s rights prompted the introduction of the title of ‘Ms.’ in preference to the traditional ‘Mrs.’ and ‘Miss’, which necessarily reveal marital status. Similarly, two new sinograms with the female radical, 妳 and 她, to 152

Chinese linguistics

complement the traditional 你 and 他, were invented in early twentieth-century China to encourage gender awareness, even though no new pronunciations were proposed. Kongzi’s discussion was a starting point for Xunzi’s (荀子, b. 313 bce) view that names, or words, are created by social conventions. It is a remarkable coincidence that Xunzi stated this view at approximately the same time as Plato (429–347 bce) did in Ancient Greece.3 Here are his words, expressed in three pithy sentences which appear in Section 22 of his anthology, titled〈正名篇〉to recall Kongzi’s injunction: 名无固宜,约之以命; 约定俗成谓之宜,异于约则谓之不宜。 名无固实,约之以命实,约定俗成,谓之实名。 名有固善,径易而不拂,谓至善名。 Roughly translated, they mean: Words have no intrinsic correctness; their correctness is established by convention. Words have no intrinsic content; their content is given by convention. Words do have intrinsic appropriateness; those which are direct and not misleading   are appropriate words. Since these words are the first expression of a fundamental property of words, and therefore of language, they are often taken to be the beginning of linguistics in China. The language of the ethnic group that has been dominant for well over two millennia is called Hanyu, named after the Han dynasty that flourished around the same time as the Roman Empire. The bulk of this chapter will be on Hanyu, its history and structure. But for now let us set the stage upon which the languages and peoples of China are to act out their destinies.

Setting the stage Some 50 million years ago, the continental plate on which India is now situated drifted northward and collided with the much larger continental plate we now call Eurasia. This collision forced an upward projection of the landmass, resulting in a mighty range of mountains, the Himalayas, a word of Sanskrit origin with a stem meaning of ‘snow’. The tallest peak in the range was marked on the maps compiled during the long reign of Emperor Kangxi (1661–1722) as Zhumulangmafeng 珠穆朗玛峰, based on the Tibetan word, Qomolangma,4 which may be translated as ‘Goddess-Mother’. The peak is almost 9,000m above sea level, and continues to rise. The physical stage upon which the peoples and languages of China play out their destinies was thus set in major outline a long time ago. It is marked by an ultra dry western part of high plateaus, since the moist air currents from the oceans down south are blocked by tall mountains. At the same time, the land has a pronounced west–east tilt, and the two great rivers which originate from these mountains, the Huanghe 黄河 ‘Yellow River’ and the Changjiang5 长江 ‘Long River’, flow eastward to the Pacific, giving life to the land in between. The Last Glacial Maximum peaked during 25,000 and 20,000 bp. Much of the Earth’s water was locked in huge glaciers; sea levels were low, exposing many passages which are now under water. The northeastern corner of Asia was joined by land to the Americas, thus providing passage for Asians to cross over Beringia and colonize the new continents. Recent 153

William S.-Y. Wang

investigations by anthropologists indicate that the Asians may have stopped for quite a while at Beringia before entering the New World and beginning the long trek down to the tip of South America.6 Linguists have begun exploring the genetic relations between Chinese and far-flung languages across continents.7 As sea levels were lowered by the Last Glacial Maximum, Taiwan was connected to the Asian mainland by land, allowing ancestral Austronesian speakers to reach Taiwan on foot. After sea levels rose again, many of these peoples sailed out to circle the earth: east to distant islands in the South Pacific, such as New Zealand and Samoa, and west to Madagascar in the Indian Ocean. This scenario is supported by the fact that Taiwan had the greatest diversity of Austronesian languages until recent decades, when many of these languages became extinct. Taiwan was the launching pad, so to speak, for the great diaspora of this great family of Austronesian languages.8 Similarly, the land bridges allowed the early speakers of Altaic languages to cross over from Korea to Japan, while leaving descendants in northern China who spoke languages that have evolved into modern Uighur, Mongolian, Manchu, Evenki, etc. These Altaic peoples have played important roles in China’s cultural history. They were especially prominent in two great dynasties: the Yuan9 元 (1206–1368) ruled by the Mongols, and the Qing 清 (1616–1911) ruled by the Manchus. A recent census (2010) lists their populations to be over 5 million for the Mongols and over 10 million for the Manchus. Altaic languages are distinguished from most of the languages in China in many ways. Syntactically the verb occurs at the end of the sentence, rather than in the middle. Morphologically the structure is agglutinative in having strings of suffixes after the verb stem. Phonetically Altaic languages are distinguished by vowel harmony, whereby there are restrictions on which vowels may co-occur in the same word. However, this feature is no longer so obvious in the Altaic languages spoken in China due to extensive lexical borrowings from neighboring languages. Returning to the present geography, the yellow, muddy color of the Huanghe is caused by the fact it flows out of the Loess Plateau, carrying with it large amounts of wind-borne clay dust. This sedimentation causes the river bed to build up quickly, leading to massive flooding when not properly controlled. Shortly after it descends from the Loess Plateau, the Huanghe makes three right-angle turns, first northeastward, shortly after Lanzhou 兰州, then eastward above 40 degrees north, then southward at around 110 degrees east, enclosing in this giant loop the Ordos region of Inner Mongolia. In its southward course, it separates two provinces with nearly homophonous names: Shaanxi 陕西 to its west and Shanxi 山西 to its east. Lastly, at around 35 degrees north, it turns sharply to resume its eastward journey, this time separating Shanxi to its north and Henan 河南 to its south. At this last turn, the Huanghe is joined by the Weishui 渭水, forming the bottom leg of an immense rectangle. This general region surrounding where the Weishui flows into the Huanghe and turns east is often referred to as Zhongyuan 中原 ‘Central Plains’. It is a region whose significance has been often stressed by historians with terms like ‘cradle of China’. Two cultural centers in this region have played special roles during the two millennia 1000 bce–1000 ce: Xi’an 西安 in the west and Luoyang 洛阳 in the east, though they had different names at different historical times. However, recognizing the importance of the Central Plains should not obscure the fact that the Chinese civilization of today is the result of at least 6,000 years of ethnic and linguistic amalgamation, with some sources quite distant from this region. From its headwaters in the western highlands of Qinghai 青海 and Xizang 西藏 (Tibet), the Changjiang flows southward in parallel with the Nujiang 怒江 and the Lancangjiang 澜沧江. Whereas the latter two rivers flow into Southeast Asia, to become the Salween and 154

Chinese linguistics

the Mekong respectively, the Changjiang is blocked in its path by some massive granite boulders in the vicinity of Shigu 石鼓 in Yunnan 云南. It makes some small hairpin curves and settles on an eastward journey, staying quite close to 30 degrees north. This latitude of 30 degrees north is of special interest in the cultural geography of China, being the location of several major cities, from Lhasa in Xizang, to Chongqing 重庆 in Sichuan 四川, to Shanghai on the Pacific coast. In addition to the two great rivers Huanghe and Changjiang, another important waterway should be mentioned: the Jinghang Da Yunhe 京杭大运河, also called the ‘Grand Canal’.10 Construction of this artificial river began in 605 ce. At over 1,700km, it joined together the regions of the two great rivers, Beijing in the north and Hangzhou 杭州 in the south. In its heyday, the Yunhe was the longest canal in the world, and won admiration from many international travelers, including Matteo Ricci11 (1552–1610) from Italy. Nowadays, however, only its southern half remains consistently navigable. Wide, fast-flowing rivers often created boundaries to separate peoples in early times, when they lacked the ability to cross them. At the same time, they also provided routes to travel along, whether by primitive rafts or boats or simply by foot along their banks, since water is essential for life. It has been speculated that many ethnic groups could have started from the western highlands and migrated along the two great rivers through the millennia. Such a scenario seems particularly relevant for the Austric hypothesis, which includes two major branches – Austronesian and Austro-Asiatic. While many scholars have argued for the historic unity of these two branches of languages, their modern forms are quite different. Austronesian languages typically begin their sentences with verbs, and have morphemes which are polysyllabic. The morphemes in Austro-Asiatic languages, on the other hand, are often monosyllabic and distinguished by lexical tones, much as in modern Chinese. In a pioneering paper of 1976, Norman and Mei proposed some lexical evidence for Austro-Asiatic in South China, particularly with respect to the names of the two great rivers. Although some of their evidence has been called into question recently (Zhang 1998), the reconstruction of language contact in prehistoric China is of obvious importance and should continue in a multidisciplinary perspective, in collaboration with archeology and genetics. This issue on Chinese river names is further discussed by Takashima (2012). The Bai Yue 百越 peoples mentioned in traditional history books were presumably largely Austro-Asiatic in their composition; they covered large regions of South and Southeastern China, extending into northern Vietnam. The Zhuangzu 壮族 of Guangxi 广西 descended from one of the Yue 粤 peoples. With a population over 16 million in 2010, it is the largest ethnic minority in China today, second in population size only to the Hans. There are some precious data on an early Austro-Asiatic language, preserved in the Yue Ren Ge 越人歌 ‘Song of the Yue Boatman’, transliterated with Chinese writing accompanied by a translation. The song was composed in the form of Chuci 楚辞, a genre of poetry from the first millennium bce. The Chinese linguist Zhengzhang Shangfang 郑张尚芳 (1991) analyzed this song with the phonology of Old Chinese (OC), which he reconstructed, and compared the verses with Written Thai; written language is typically more conservative than varieties of the spoken language. To give an idea of this work, let us consider the first line in the Yue Ren Ge, which appeared in the classical text as follows: 滥兮抃草滥. As Table 10.1 shows, pronouncing this line in Putonghua (PTH), the line would be: làn xī biàn cǎo làn. Neither the string of sinograms nor the Putonghua pronunciation makes any sense. However, Zhengzhang replaced the Putonghua pronunciations with his Old Chinese reconstructions for these sinograms, and got: OC *ɦgraams ɦee brons ts’uuɁ ɦgraams. It is 155

William S.-Y. Wang Table 10.1  Song of the Yue Boatman〈越人歌〉 Line 1 of Yue Ren Ge Putonghua Old Chinese Translation via Written Thai

滥 làn *ɦgraams evening

兮 xī ɦee particle

抃 biàn brons joy

草 cǎo ts’uuɁ meet

滥 làn ɦgraams evening

interesting to note from Table 10.1 that PTH làn and biàn both have a falling tone, which derives from historical tone III (Qu sheng 去声), and PTH cǎo has a low tone, which derives from historical tone II (Shang sheng 上声). Zhengzhang reconstructs tone III with a final *-s and tone II with a final *-Ɂ; both reconstructions are consistent with the views of other scholars in the field of historical Chinese phonology, such as those discussed in Baxter (1992). In addition to providing the OC reconstruction here, Zhengzhang made the contribution of comparing the reconstructions with phonetically similar syllables in Written Thai, which is ‘the most anciently attested form of Thai and other languages of the Tai group’ (1991: 160). By using these methods, he was able to meet the challenge presented by the classical text, verify that the song was in an Austro-Asiatic language, and translate the first line literally as: ‘evening, particle, joyful, to meet, evening’. It expressed the joy of the boatman to meet a Han prince that evening. The classical text indicated that the boatman was from the southern region of Chu 楚. This is reflected in the use of the particle xī as the second syllable in the line; this particle was abundantly used in Chuci, which are famously represented by the writings of Qu Yuan 屈原 (340–278). Zhengzhang concluded his investigation with this remarkable observation: ‘Although the words of the Yue People’s Song may be compared with Siamese, the verses would be easily understood by modem Thai speakers’ (1991: 167). Unfortunately, we have no ancient text for Austronesian comparable to the Yue Ren Ge for Austro-Asiatic. Many Austronesians must have settled in Taiwan many millennia ago, before the island became separated from mainland China by so wide a strait of water. The antiquity of their settlement is in part evidenced by how deeply divergent the Austronesian languages have become in Taiwan; a map of the historical distribution of these languages is shown in Figure 10.1.12 Many experts now believe that Taiwan was the last stop before the Austronesians went on to colonize far flung islands of the world, from Southeast Asia all the way eastward to Hawaii, and all the way westward to Madagascar. They were indeed the finest sailors of the ancient world.

Early ancestors in China There is general consensus among scientists in several disciplines that we, Homo sapiens sapiens (anatomically modern humans, AMH), are the only living representatives of the biological genus Homo; other species in this genus, such as the Neanderthals in Europe and Western Asia, the Denisovans in Siberia, and Homo floresiensis in Indonesia, all became extinct long ago. Such knowledge is based on the integration of a variety of evidence offered by many disciplines, including Comparative Anatomy, Physical Anthropology, Archeology, and especially Molecular Genetics. We also know that our species, Homo sapiens sapiens, emerged in Africa. Some members of this species left the homeland to colonize the world 156

Chinese linguistics

Basay

Luilang

Kulon

Northern

Qauqaut

Ts’ole’&Sqoleq

Southern

Ts’ole’

Ts’ole’

Atayal

Ts’ole’

Papora

Kavalan Trobiawan

Sqoleq

Saisiyat Taokas

Ketangalan

Pazih

Seediq Ts’ole’&Sqoleq

Babuza

Sakizaya

Ts’ole’

Thao

Northern

Northern

Tavalong-Vata’an

Hoanya

Duhtu Tfuea

Amis

Tsou

Tapangu

Central

Takupulan

Kanakanavu Taivoran Saaroa

Central

Formosan Languages Amis Atayal Babuza Basay Bunun Hoanya Kanakanavu Kavalan Kulon Paiwan Papora Pazih Puyuma Qauqaut Rukai Saaroa Saisiyat Siraya Taokas Thao Tsou Batanic(Philippines) Language Yami

Bunun

Southern

Southern

Siraya

Puyuma Rukai

Nanwang

Makatau

Lütao

Southern Amis

Paiwan Southern Amis Southern Amis

Yami Lanyü

Southern Amis

Figure 10.1  Austronesian languages in Taiwan, courtesy of Prof. Paul Li

over 100,000 years ago. Recent successes with ancient DNA research indicate there was a good deal of interbreeding between AMH and other ancient populations,13 though so far ancient fossils from China have not yet been analyzed for prehistoric mixing. The success of these early peoples to colonize the world is due to the emergence of language, which qualitatively enhanced their ability to think, to communicate, and to cooperate. The evolution of languages is always intricately interwoven with the evolution of the peoples who speak them, each enhancing the development of the other. Among the extinct species is Homo erectus, including the famous Peking Man 北京猿人, found in the caves near the village Zhoukoudian 周口店, on the southwest outskirts of 157

William S.-Y. Wang

Beijing. Zhoukoudian represents an extremely important archeological site in the study of human evolution because of ‘repeated human visitations, through 11 major depositional units that span a period of some 300,000 years, its abundance of archeological and faunal residues, and, of course, its unique sample of Homo erectus remains’.14 Peking Man lived some 600,000 years ago (Shen et al. 2009), and was among the earliest users of fire – a trait not found in any other animal. Since the discovery at Zhoukoudian in the 1920s, human fossils, stone tools, and bone artifacts have been unearthed at numerous other sites in China, reaching back well over a million years (Wu and Olsen 1985). A notable recent find is at the Zhirendong 智人洞 in Guangxi of a human mandible, which is the oldest fossil from Homo sapiens sapiens outside of Africa (Liu et al. 2010). Also worthy of mention here is the remarkable discovery of a set of flutes, made from crane’s leg bones, at Jiahu 贾湖 in Henan (Zhang et al. 1999). The authors call these ‘the oldest playable musical instruments’, and the audio file of a Chinese song played on these flutes can be accessed on the Internet via their publication. Also discovered at Jiahu are some potsherds with inscriptions on them which are highly suggestive of writing.15 These discoveries give us a glimpse of Neolithic culture in China, nearly 9,000 years ago, which was unexpectedly rich. According to current consensus, AMH emigrated from Africa in many waves, beginning over 100,000 years ago, settling first in Asia. From Asia, the human diaspora spread successively westward to Europe, southward to Australia, and across the Beringia to the Americas. For most of these 100,000 years, our ancestors lived as roving bands of hunters and gatherers, their movements driven by climatic conditions and the availability of fauna and flora for food. The little we know of their world is based on the material traces they left behind, primarily in the form of fossils and stone tools.16 Contrasting with the Out-of-Africa hypothesis, in which the emigrants from Africa completely replaced the earlier archaic Homo species, some scholars prefer an alternative scenario in which there was some interbreeding; cf. Abi-Rached et al. (2011). These latter voices favor a Multi-Regional hypothesis, in which archaic humans contributed to the gene pool alongside the AMH, who left Africa 100,000 years ago. Many decades back, Franz Weidenreich, who oversaw some of the first Zhoukoudian excavations in the early twentieth century, lectured on the anatomical continuities between ‘Peking Man’ and the modern Chinese.17 Wu Xinzhi 吴新智, a leading authority on Zhoukoudian fossils, continues to explore this line of thought.18 Interestingly, recent excavations in Zhoukoudian, at a site called Tianyuandong 田园洞, yielded fossils which are more compatible with the Multi-Regional hypothesis.19 Currently the evidence is accumulating that although our modern genes are mostly inherited from ancestors who left Africa 100,000 years ago, there have been admixtures with other archaic species within the Homo genus before they became extinct. Archeologically, of great interest is the very recent discovery of pottery remains in Jiangxi 江西20 which date back some 20,000 years, many millennia before agriculture was invented. After the onset of the Holocene period, the human condition changed dramatically with the advent of agriculture. With the raising of crops and domestication of animals around 10,000 years ago, ancient peoples settled down. With the systematic production of food, they became more numerous, and villages eventually grew into towns and cities. They started making pottery for keeping and transporting food and water, and began to mark the pottery with symbols to identify the maker or the owner. These marks are the early precursors of writing, which emerged several millennia later. There were numerous Neolithic villages all over China which developed relatively independently for many millennia.21 These include Dawenkou 大汶口 in Shandong 山东 and 158

Chinese linguistics

Zengpiyan 曾皮岩 in Guangxi, and many others. The best known of these many sites is the Banpo 半坡 village situated near modern Xi’an, where an on-site museum has been built to preserve and exhibit its remains. Banpo, dating back some 7,000 years, is an example of the Yangshao 仰韶 culture, known for its painted pottery, termed Caitao Wenhua 彩陶文化 ‘colored pottery culture’ in Chinese. ‘Yangshao’ is the name of a village in Henan, where the archeological site was first discovered in 1921 by the Swedish archaeologist/geologist Johan Gunnar Andersson22 (1874–1960). This discovery is regarded as the beginning of archeology in China. Shortly after the Yangshao discovery, Neolithic pottery of several millennia later was unearthed in Shandong, where the pottery was more delicate and painted black. This became known as the Longshan 龙山 culture, called Heitao Wenhua 黑陶文化 ‘black pottery culture’ in Chinese. It is interesting that an average linkage analysis of the cranial indices reported on nine fossil skulls unearthed at these sites revealed a major separation of north and south.23 This would suggest that the north/south division among the peoples of China has a very early origin. This division is supported by an extensive immunological study reported by Zhao and Lee (1989). Together with recent advances whereby DNA can be extracted from fossils to examine their genetic affinity,24 our understanding of the phylogeny of Asian populations has deepened significantly, thanks to the contributions from molecular genetics. As the scope of these villages increased over time, they met and interacted with greater regularity and frequency, eventually connecting together into one integrated cultural sphere some 6,000 years ago, when there began sustained contact and sharing of cultural innovations. The cultural sphere formed in this way was dubbed the ‘initial China’, by the late archeologist Chang Kwang-chih 张光直, whose authoritative volume The Archeology of Ancient China in its several editions guided the field and trained generations of students for several decades. Figure 10.2 is reproduced from p. 235 of the 4th edition of Chang’s classic book, published in 1986.25 Chang marked off nine Neolithic regions for 6,000 bp, and drew arrows to highlight the interaction among them. The identified sites going down the Pacific coast are: Hongshan 红山, Tuzhu 土珠, Dawenkou, Majiabang 马家浜, Hemudu 河姆渡, Tanshishan 昙石山, Fengbitou 凤鼻头, and Shixia 石峡. The remaining three sites are Shanbei 山背, Daxi 大溪, and Yangshao. Yangshao is presumably the best known of these cultures, as mentioned earlier, in part because it was the earliest discovered. In the words of Chang, it was 6,000 years ago26 that ‘these cultures became closely linked, and they share common archeological elements that bring them into a vast network within which the cultural similarities are quantitatively greater than without. By this time we see why these cultures are described together: not just because they are located within the borders of present-day China, but because they were the initial China’ (1986: 234; emphasis added). Later on, on p. 410, Chang stresses the organic whole of the Chinese culture thus united by the charming metaphor: ‘When the Weishui River valley sneezed, as it were, the Lake Taihu region caught cold’.

Shells and bones The English word ‘China’ derives from the name of the first imperial dynasty in China, Qin 秦, short-lived as it was. The man who accomplished the formidable task of uniting the warring kingdoms of the third century before the Common Era into one empire is called Qin Shi Huangdi 秦始皇帝 ‘Qin First Emperor’. The world knows of him through the fabulous terracotta army unearthed in his tomb near modern Xi’an. The name ‘China’, in its 159

William S.-Y. Wang

Hung-shan

Tu-zhu

Ta-wen-k’ou Yang-shao

Ma-chia-pang Ta-hsi

Ho-mu-tu Shan-pei

T’an-shih-shan Shih-hsia Feng-pi-t’ou −400

0 100 200 300 400 500 600 km

Figure 10.2  The ‘initial China’

various phonetic variants, is used in most European languages. A notable exception is the name for ‘China’ in Russian, Kitai, which derives from a people that ruled northern China during the Liao dynasty. Their name is now spelled Qidan 契丹 in China; also Khitan in older literature. Table 10.2 gives a thumbnail sketch of the history of China over some four millennia. There is still some controversy among specialists on the nature of the Xia dynasty, its location and its material remains. However, there is abundant archeological and textual evidence on the Shang dynasty, dating back over three millennia. Whereas books by earlier historians tell of this dynasty, the major breakthrough toward understanding it came only at the beginning of the twentieth century. It came when a scholar in Beijing named Wang Yirong 王懿荣 160

Chinese linguistics Table 10.2  Simplified chronology of Chinese history Xia Shang Zhou, Western _____ , Eastern Qin Han, Western _____ , Eastern Three Kingdoms Jin Nanbei Chao Sui Tang Wudai Shiguo Song, Northern _____ , Southern Liao Xixia (Western Xia) Jin Yuan Ming Qing Republic of China People’s Republic of China

21st c. to 16th c. bce 16th c. to 1046 bce 1046 to 771 bce 770 to 221 bce 221 to 206 bce 206 bce to 25 ce 25 to 220 220 to 280 265 to 420 420 to 581 581 to 618 618 to 907 907 to 979 960 to 1127 1127 to 1279 907 to 1125 1038 to 1227 1115 to 1234 1206 to 1368 1368 to 1644 1616 to 1911 1912 to 1949 1949 to present

夏 商 西周 东周 秦 西汉 东汉 三国 晋 南北朝 隋 唐 五代十国 北宋 南宋 辽 西夏 金 元 明 清 中华民国 中华人民共和国

bought some ‘dragon bones’ from local stores for medicinal purposes. These bones were primarily turtle shells and ox scapula, and some bore graphs on them. From his knowledge of ancient sinograms inscribed on bronze ware, he recognized that many of these graphs were ancient sinograms as well. For the study of the languages and peoples of China, this was a momentous event. Inscriptions on bronze ware, mostly on ritual vessels and weapons of war, are called Jinwen 金文; the ‘jin’ in the name means ‘gold’ in its narrow sense, but ‘metal’ in its broad sense. Inscriptions on shells and bones are called Jiaguwen 甲骨文, where the ‘jia’ refers to the turtle shells and the ‘gu’ refers to various bones. After Wang Yirong made the discovery, numerous specimen of Jiaguwen have been collected, mostly from archeological excavations in the city of Anyang 安阳 in the province of Henan. Anyang turned out to be the last capital of the Shang dynasty. Decades of research on Jiaguwen has given us much first-hand information on the languages and peoples during the Shang dynasty, which are highly valuable for understanding early societies in human civilization. The shells and bones were inscribed for the purposes of divination, not unlike the Delphic oracles in Ancient Greece. For this reason, Jiaguwen is also called Oracle Bone Inscription (OBI) in English. The inscriptions contained questions to be divined, regarding weather, warfare, childbirth, etc., and often recorded related events. From these inscriptions we get to know about the various ethnic groups that were active in the Central Plains area. We are also able to reconstruct a complete lineage of the many kings of the Shang dynasty, as well as of some of the important personages. Of the latter, particular mention should be made of a remarkable woman warrior, named Fuhao 妇好, consort of Wuding 武丁, one of Shang’s most successful kings. 161

William S.-Y. Wang

Figure 10.3  Oracle bone inscription concerning Fuhao’s childbirth, reproduced with permission from City University of Hong Kong Press

Space limitations do not permit much discussion of Jiaguwen here. But it is useful to see a sample rubbing in Figure 10.3, which is a divination on Fuhao’s childbirth.27 The sinograms were inscribed on a turtle plastron in this case, prepared beforehand for divination. After the question was inscribed, heat was applied to the plastron to induce cracks. The diviner, sometimes the king himself, would interpret these cracks as answers from heaven. The question here began in the rightmost column, and the columns move leftward in this case. Of special interest is the third sinogram down in the rightmost column, which looks like , but is now written 卜, pronounced bu in Putonghua but buk in Cantonese. This sinogram actually means ‘to divine’; it has the shape of cracks made on the shell, and a pronunciation that resembles the sound the shell makes when it cracks. Another sinogram to note here is the second one down in the leftmost column, which is made up of two parts. The left part resembles a stick figure kneeling, facing to the right; this part is now written 女, meaning ‘woman’. The right part of this sinogram is now written 子, meaning ‘child’. The modern sinogram made up of these two parts, ‘woman’ and ‘child’ is 好, meaning ‘good’, and is pronounced hao in Putonghua. In fact, it is the second sinogram in the name of the woman Fuhao, with whose childbirth the king is concerned in this divination. The main lesson from Jiaguwen for Chinese linguistics is that the sinograms and their principle of formation as well as the grammar of the Shang inscriptions can be clearly seen to be ancestral to modern Hanyu. Since these inscriptions represent the earliest texts in the language, the language itself is called Old Chinese,28 or Shanggu Hanyu 上古汉语.

Poems and reconstructions There are two sets of cultural treasures which shed light on the roots of Chinese civilization. One is Jiaguwen, which lay buried underground for three millennia before it was discovered 162

Chinese linguistics

accidentally. The other set is Shijing 诗经, a collection of songs, poems and odes, which also dates back to the Shang and Zhou dynasties. The collection is sometimes said to have been organized by Kongzi, though there is little evidence for this belief. On the other hand, he spoke of it often in his teachings, sometimes with reverence. For instance, he proclaimed: 《诗》   三百,一言以蔽之,曰: 思无邪 ‘The 300 poems in the Shijing can be covered by one phrase: have no improper thoughts.’ Here Kongzi referred to the Shijing with just the sinogram 诗 shi ‘poem’. The term now commonly used is《诗经》, which includes the sinogram jing 经 to show its high cultural status. This sinogram is also used in a religious context; for instance, the Bible is called Shengjing《圣经》, and the Koran is called Gulanjing《古兰经》. There are actually 305 poems in the commonly used version, ranging in length from just a few lines of a lover’s lament to several pages of a historical chronicle. These poems provide us with a precious window on the diverse cultures of Ancient China, not just from the viewpoint of affairs of state, as with Jiaguwen, but often with the everyday voices of common folk. The particularly valuable aspect of the Shijing for Chinese Linguistics is the information it provides us indirectly on how the language may have sounded 3,000 years ago. It does this by allowing us to pool together sinograms which rhyme with each other. We may assume that most of the time sinograms that rhyme were pronounced with a common tone and a common syllable final, i.e. the nuclear vowel of the syllable plus any consonant or consonants that may follow it. The systematic investigation of the pronunciation of the Shijing from a scientific viewpoint was pioneered by a remarkable scholar named Chen Di 陈第 (1540–1620). Whereas earlier scholars were at a loss to explain why the Shijing poems no longer rhymed in their pronunciation, Chen clarified the situation with these famous words: 盖时有古今,地有南北,字有更革,音有转移,亦势所必至。 Time, there is past and present. Space, there is south and north. Words change, pronunciations shift. All are in the nature of things. Chen went on to group the sinograms in the Shijing in terms of their phonetic relations, mindful of the type of evidence that is relevant according to the decisions made – whether it is internal to the Shijing itself, or external to it from other considerations. In blazing out a trail for generations of Chinese linguists to follow, the role he played in China is similar to that of William Jones (1746–94) in the West, who established the very successful paradigm of Indo-European linguistics, basing itself on the comparative method. It was Jones’s contribution, encapsulated in his Calcutta lecture of 1786, that launched the modern science of linguistics. The most significant limitation in Chen’s scholarship, and that of the brilliant Qing dynasty scholars who followed him, was restricting themselves to only textual data from ancient times, instead of availing themselves additionally of the live data of the numerous Hanyu dialects and minority languages which can shed light on the Shijing. Nonetheless, their contributions laid the foundation in discovering the abstract rhyme categories for later researchers to link up to modern data. One scholar who early brought insights from Western linguistics to the study of the Chinese language was Bernhard Karlgren (Gao Benhan 高本汉, 1889–1978) from Sweden, a younger colleague of the J. G. Andersson mentioned earlier. His PhD dissertation includes a pronouncing dictionary of many dialects, providing valuable data for the reconstruction of 163

William S.-Y. Wang

Old Chinese, and illustrating the importance of fieldwork. His early work Études sur la phonologie chinoise was translated into Chinese under the title《中国音韵学研究》by a trio of Chinese linguists, who were to exert a formative influence on the development of Chinese linguistics in the twentieth century: Y. R. Chao 赵元任 (1892–1982), F. K. Li 李方桂 (1902–87), and C. P. Luo 罗常培 (1899–1958). Other Western scholars who have contributed significantly to the reconstruction of Old Chinese and Middle Chinese include André-Georges Haudricourt from France and E. G. Pulleyblank from Canada. Chinese linguists who combine the training of traditional Chinese linguistics with the comparative method to work in this area include Zhengzhang Shangfang, mentioned earlier in connection with Yue Ren Ge, Tung Tungho 董同龢, Mei Tsulin 梅祖麟, Ting Pang-hsin 丁邦新, and Gong Hwangcherng 龚煌城, who was an outstanding scholar on the extinct Xixia language. US linguist William Baxter (Bai Yiping 白一平) has done the field of Old Chinese phonology a great service by publishing in 1992 a comprehensive handbook, putting the various contributions in a balanced historical perspective. However, Old Chinese phonology remains a dynamic area of scholarly explorations. Ho has provided an incise critique of the central issues in his review article of 2015. In the sample from the Shijing we will look at below, we will use Baxter’s reconstructions. The best known of the 305 poems in the Shijing is probably the first one, called Guan Ju〈关雎〉 . In his Analects VIII: 15, Kongzi praised this poem in these words: 师挚之始,〈关雎〉之乱,洋洋乎盈耳哉！ which has been translated as ‘When Zhi the Master Musician begins to play and when Guan Ju comes to its end, how the sound fills the ear!’29 The poem has a very simple theme, of a young man seized by the pangs of unrequited love. The ‘guan’ in the title is a syllable imitating the sound of bird cry; the ‘ju’ is the first syllable of the name of the bird jujiu 雎鸠. The setting is an idyllic one: an islet in a river, with water plants floating around it. Table 10.3 gives two authoritative translations. Arthur Waley was an English sinologist based in London who translated numerous classics from Chinese and Japanese into English. Yang was a professional translator of many Chinese classics into English; he and his collaborators were based in Beijing. The translations differ, as is to be expected when poetry is involved, but not by too much for this poem. On the other hand, there are poems for which experts differ significantly in their interpretations. Even though the literature on the Shijing is already extensive, there are numerous issues on it for which there is no consensus, simply because we do not know enough about Ancient China as yet. As we can see from Table 10.3, the poem is divided into five verses, each verse containing four lines, and each line containing four sinograms. The rhyme scheme in verses 1 and 3 is AAxA, with the rhyme on the last syllable. The rhyme scheme in verses 2, 4, and 5 is xAxA; however, since the last syllable in these lines is a grammatical particle, i.e. zhi, which probably was unstressed, the rhyme falls on the next-to-last syllable. I have italicized these unstressed syllables in the table to make clear this point. The rhyming syllables are presented in Table 10.3 in their Old Chinese pronunciations, preceded by *–, as these are reconstructed in Baxter’s Handbook. The Old Chinese reconstructions do not indicate lexical tones; experts are not agreed on whether Old Chinese made use of lexical tones. On the other hand, these syllables are marked in the Putonghua readings, according to the tone the sinogram had in Middle Chinese, the language reflected in a rhyme dictionary compiled around 601 ce during the Sui dynasty. 164

关关雎鸠 在河之洲 窈窕淑女 君子好逑

参差荇菜 左右流之 窈窕淑女 寤寐求之

求之不得 寤寐思服 悠哉悠哉 辗转反侧

参差荇菜 左右采之 窈窕淑女 琴瑟友之

参差荇菜 左右芼之 窈窕淑女 钟鼓乐之

1.1 1.2 1.3 1.4

2.1 2.2 2.3 2.4

3.1 3.2 3.3 3.4

4.1 4.2 4.3 4.4

5.1 5.2 5.3 5.4

〈关雎〉   Guan Ju

cen ci xing cai zuo you maoIII zhi *mawks yao tiao shu nv zhong gu leIII zhi *grawk

In patches grows the water mallow To left and right one must choose it Shy is this noble lady With gongs and drums we will gladden her

In patches grows the water mallow To left and right one must gather it Shy is this noble lady With great zither and little we hearten her

Sought her and could not get her Day and night he grieved Long thoughts, oh, long unhappy thoughts Now on his back, now tossing to his side

qiu zhi bu deIV *tik wu mei si fuIV *bjik you zai you zai zhan zhuan fan ceIV *tsrjik

cen ci xing cai zuo you caiII zhi *sriɁ yao tiao shu nv qin se youII zhi *wjiɁ

In patches grows the water mallow To left and right one must seek it Shy was this noble lady Day and night he sought her

‘Fair, fair’ cry the ospreys On the island in the river Lovely is this noble lady Fit bride for our lord

Translation by Arthur Waley

cen ci xing cai zuo you liuIv zhi *Crju yao tiao shu nv wu mei qiuIv zhi *grju

guan guan ju jiuIu *krju zai he zhi zhouIu *tju yao tiao shu nv jun zi hao qiuIv *grju

Table 10.3  Guan Ju, first poem of Shijing

Short and long the floating water plants Left and right you may collect them Gentle and graceful is the girl He’d like to marry her, bells and drums beating

Short and long the floating water plants Left and right you may gather them Gentle and graceful is the girl He’d like to wed her, the qin and se playing

When the courtship has failed Awake he thinks of her and in his dreams Filled with sorrowful thoughts He tosses about unable to sleep

Short and long the floating water plants Left and right you may pluck them Gentle and graceful is the girl Awake he longs for her and in his dreams

Merrily the ospreys cry On the islet in the stream Gentle and graceful is the girl A fit wife for the gentleman

Translation by Yang (1983)

William S.-Y. Wang

We will simply note here that the Iu and Iv tones in verses 1 and 2, have developed into tones 1 and 2 in Putonghua, pronounced high level and high rising respectively. Middle Chinese tone IV, as found in verse 3, occurred on syllables which end in –p, –t, or –k, as we see in the Old Chinese reconstructions. These endings are still preserved in some modern dialects, such as Cantonese, as in the pronunciation of the integers sap 十 ‘ten’, cat 七 ‘seven’, luk 六 ‘six’; but in Putonghua they have all become lost. The syllables which have lost these endings have merged into other tones; in this poem, de and fu have merged with Iv while ce has merged with III. Finally, note that the tone II in verse 4 has developed into Putonghua tone 3, a low tone, and tone III in verse 5 into Putonghua tone 4, a high-falling tone. As for the final –Ɂ in the Old Chinese form in verse 4, and the final –s in verse 5, these are both consistent with Zhengzhang’s reconstructions we saw earlier in Table 10.1. As one last observation on how Old Chinese was pronounced, we note the presence of consonant clusters in the reconstructed forms in this poem. In addition to the ending *–ks in line 5.2, we find initial consonant clusters which include the liquid *–r– in several lines. This *–r–, in fact, later develops into an l– initial. These Old Chinese consonant clusters have all essentially disappeared in modern dialects. Nonetheless, the fact is that liquid consonants, such as [l] and [r], frequently form initial clusters with obstruents in languages of the world, as in English play, clay, pry, cry, frill, shrill, slay, etc.

Phonograms and rhyme dictionaries The primary line of evidence for the existence of consonant clusters in Old Chinese comes from phonograms in the written language. Phonograms are constructed from two parts, typically with a radical indicating its meaning on one side and a phonetic on the other side. It turns out many sets of phonograms can be found in modern dialects which share a phonetic, but which differ in their initial consonant; furthermore, one of these initial consonants is an obstruent while the other is the liquid [l]. Thus the phonograms 剥 and 禄 share a phonetic, but the former is pronounced bo while the latter is pronounced lu. Similarly, 果 is pronounced guo while 裸 is luo; and 使 is pronounced shi while 吏 is li. As Old Chinese changed into Middle Chinese, its clusters reduced to single consonants in different ways in different phonograms, some keeping the obstruent and losing the liquid, while others kept the liquid and lost the obstruent. As can be seen from Table 10.2, starting with the Zhou, several major dynasties are divided into two periods according to when their capitals were moved, usually forced by military invasions. The two philosophers we mentioned earlier, Kongzi and Xunzi, were both active during Eastern Zhou. Also, note that while the Zhou and the Han moved their capitals eastward to avoid invaders from the west, the Song moved its capital southward due to military pressures from the north. Census figures from historical studies show it was shortly after the Song dynasty moved its capital to Lin’an 临安 (modern Hangzhou) that the population in South China came up approximately to the same level as that in the north. Previously, the bulk of the population centered around the Central Plains. It was during the Han dynasty that the first empirical studies in Chinese linguistics began; particularly the pioneering works of Yang Xiong 扬雄 on linguistic geography, and of Xu Shen 许慎 on lexicography. Yang compiled a precious list of words in a volume commonly called the Fangyan 方言, as these were pronounced in different regions of the Han empire. Xu produced the first comprehensive dictionary of Chinese sinograms, called the Shuowen Jiezi 说文解字; in the post-face to this dictionary, he proposed a scheme of classifying sinograms into six types, a scheme that continues to be used in traditional scholarship today. 166

Chinese linguistics

It is also a legacy of this great dynasty that the dominant ethnic group call themselves Hanzu 汉族, and call their language Hanyu 汉语. Another name also used by Chinese to refer to themselves comes from another great dynasty, the Tang dynasty. Thus the name Tangren 唐人, and many Chinatowns in Western countries are called Tangren Jie 唐人街. Yet another term of similar meaning is Hua 华, whose etymology is more obscure. In Singapore, for instance, the Chinese refer to themselves as Huaren 华人 and the language as Huayu 华语. While terms like Han, Tang, and Hua suggest influence from the dominant ethnic group, we should keep in mind that these ethnic labels did not acquire their current meaning until the Common Era. Previously, there were numerous ethnic groups, in the forms of minor kingdoms and nomadic tribes, which do not connect in simple ways with modern populations because of the extensive mixings over the many centuries. Furthermore, even in the simplified chronology of Table 10.2, there are many polities which were not Han and which did not speak Hanyu. These include periods when China was highly divided, such as during the Wudai Shiguo and Nanbei Chao, and when the rulers were obviously not Han, such as the Xixia, who were Tibeto-Burman, and Liao, Jin, Yuan, and Qing, when the rules were Altaic. So the peoples of China, their cultures and their languages, are truly heterogeneous, once we get past the superficial similarities.30 During the latter part of the Han dynasty, Buddhism began to enter China. With the immense task of translating its sutras into Chinese, China became aware of the Sanskrit language, its script, and its associated linguistic scholarship. Largely as a result of this strong influence, two major advances were soon made toward understanding how Chinese was pronounced. One was the explicit awareness of the nature of lexical tones. Four tones were recognized; they were named: Ping 平 ‘level’, Shang 上 ‘rising’, Qu 去 ‘departing’, and Ru 入 ‘entering’; we have represented these respectively as I, II, III, and IV. The other advancement was in the representation of pronunciations by a method called fanqie 反切 ‘reverse cut’. Before this method was introduced, the only way to help a reader with how a sinogram X is pronounced was by means of another sinogram Y, with the formula ‘Read X as Y’. However, if the reader does not know the pronunciation of Y either, then he receives no help. The fanqie method, on the other hand, generalizes the formula to ‘Read X as A+B’, with the understanding that X has the same initial consonant as A, and that it has the same tone and final as B. These two advancements paved the way to the structure of rhyme dictionaries and rhyme charts, and led to deeper knowledge in phonetics. The Sui dynasty, though short lived, is important because it was then that a rhyme dictionary of critical importance was compiled. This dictionary, the Qie Yun 切韵 ‘cut rhyme’, and its descendent dictionaries form the basis for projecting backward in time to reconstruct Old Chinese, and for projecting forward in time for connections to the modern dialects. The language inferred from this dictionary is aptly called Middle Chinese. This dictionary must have had an important influence on the greatest poets China ever produced, who composed their brilliant literature in rhymes during the great Tang dynasty which directly followed the Sui. By Qie Yun’s time, it was clear that each sinogram corresponds to a single syllable, which has a three-part phonetic structure: tone, initial, and final, respectively called shengdiao 声调 ‘tone’, shengmu 声母 ‘initial’, and yunmu 韵母 ‘final’. The rhyme dictionary was organized by first grouping the sinograms by the four tones, Ping, Shang, Qu, and Ru. Then it was next organized by groups of finals. Lastly, it was organized by grouping the sinograms by their initials. The sinograms which share all three parts of phonetic structure are homophones. The first sinogram, or the main sinogram, of each group of homophones is distinguished two ways: a little circle is placed above it to mark the beginning of the group, and a fanqie is provided at its end to notate the pronunciation. 167

William S.-Y. Wang

Figure 10.4  A page from a rhyme book of the Song dynasty, adapted from Wikipedia: http://en. wikipedia.org/wiki/Rime_dictionary#/media/File:Guangyun_Dong_Rhyme_1.jpg, under CC-BY-SA 3,0 (free to share in any format for any purpose, even commercially)

The Qie Yun itself, unfortunately, is no longer available. But many later versions based on the same organizing principles are available. Current knowledge about ancient pronunciations are largely based on these later versions. Figure 10.4 is a page taken from the Guang Yun 广韵, compiled in the Song dynasty as a successor to the Qie Yun. In the fourth column, counting from the right, we see the sinogram 同 tong, with a little circle above it to mark the beginning of a group of homophonous sinograms. 同 is then followed by various commentaries on this sinogram, written in vertical lines which are half-column wide. These commentaries end in the next column to the left in a string of sinograms, 徒红切四十五, before the next homophonous sinogram begins, which is 仝 tong. The third sinogram in the string, 切 qie, tells us that its two preceding sinograms are the fanqie of the main sinogram. It reminds us to use the formula, read X as A+B. In this case, the A sinogram is 徒 tu, with the initial t–, and the B sinogram is 红 hong, with the final and tone –óng. Putting A and B together, we get the pronunciation of tóng, which is correct for 同, even though the fanqie was proposed in the rhyme book Qie Yun well over a millennium ago. Another important rhyme dictionary was compiled during the Yuan dynasty, the Zhongyuan Yinyun 中原音韵, which reflects the pronunciation of yet another period of the language, usually called Early Mandarin. The English translation ‘Mandarin’ is based on the Chinese term Guanhua 官话 ‘Official Speech’. However, recent research by Shen Zhongwei31 shows 168

Chinese linguistics

that Mandarin with its defining features actually emerged during the Liao dynasty, several centuries earlier than the Zhongyuan Yinyun. This discovery was made by careful examination of phonetic information contained in texts written in Khitan 契丹 and Jurchen 女真 scripts.

Many peoples, many voices We began this chapter by setting the stage of what is now China, though the political boundaries of the nation have changed many times since the earliest dynasty, the Shang, which is attested both archeologically and textually. The home of a nation is marked by its mountains and rivers. We have discussed the actors on this stage, the Chinese peoples, who are heterogeneous genetically, culturally, and linguistically. The linguistic roots of China may be traced to the dual heritage of the Shang dynasty – Jiaguwen and the Shijing, expressed in the language reconstructed as Old Chinese. The early emergence of a complete writing system, in the form of sinograms, well over 3,000 years ago, is of inestimable importance in binding the various peoples together into one over-arching literate culture over great expanses of space and time. Another landmark in the evolution of the language is the rhyme dictionary of the Sui dynasty, reconstructed as Middle Chinese. The language reflected in the Qie Yun served as a critical midpoint for the backward reconstruction of Old Chinese on one hand, and for the projection to modern dialects on the other. By the time of the Zhongyuan Yinyun, another rhyme dictionary compiled in the Yuan dynasty, Mandarin has already taken root in Beijing, which has served as the nation’s capital for most of the past millennium. More than any other dialect, many features of Mandarin have spread far and wide across the land, from the northeast to the southwest. Similarly, the other major dialects have also been shaped from continuous southward migrations from the Central Plains in the north, two major waves taking place when the Han dynasty collapsed, and when the Song dynasty moved its capital to modern Hangzhou. The migrants mixed their Hanyu with the languages of the earlier settlers in China’s southeast, which were mostly Austric. The dialects which resulted have evolved sufficiently apart from each other over many centuries that they qualify as separate languages. Nonetheless, the term ‘dialect’ is the preferred usage because of a highly interlocked mosaic of cultures and because of the shared writing system. Figure 10.5 gives an approximate geography of the major dialects of Hanyu. Over the past half century, the government in Beijing has advocated the use of the variety of Mandarin that is called Putonghua 普通话 ‘common speech’, which is primarily based on the pronunciation of the educated in Beijing and which draws its vocabulary and phrases from other regions as well. Besides Mandarin, which is also simply called Beifanghua 北方话 ‘northern speech’, there are six major dialects which are well recognized, three in Central China: Xiang 湘, Gan 赣, and Wu 吴, and three along the southern coast: Min 闽, Kejia 客家, and Yue 粤. Kejia ‘guest family’ is labeled Hakka on the map, which is a southern pronunciation; whereas the other five major dialects are named with province names, the Kejia came south later and were thus referred to as ‘guest families’. The map also shows Jin in the north, Hui in Central China, and Ping in the south; these have not yet been investigated with as much depth as the six major ones.

Sinograms and alphabets We will now return for more discussion on the writing system. Alphabetic spelling,32 of which Hanyu Pinyin is the latest version, came to China only in the sixteenth century. The indigenous script that has long been in use and continues to be used in China consisting of thousands of graphic units is popularly called ‘Chinese characters’, which we here call 169

William S.-Y. Wang

Hohhot Beijing

JIN

丨

Ta iy u an

Lanzhou

.Xi'an

Naniinq Shanghai

MANDARIN W uhan

Chenadu

Hangzhou

HUI

Chongqing

XIANG

WU

Nanchang

GAN

Ihangsha

Fuzhou

HAKKA

K u nm in g

YUE

Nanning

PING

MIN

Taipei

Guangzhou Ho ng Kong

Figure 10.5  Major dialects of Hanyu, adapted from Wikipedia: http://en.wikipedia.org/wiki/List_of_ varieties_of_Chinese#/media/File:Map_of_sinitic_dialect_-_English_version.svg, under CC-BY 3,0 (free to share in any format for any purpose, even commercially)

‘sinograms’. In China, these units are called Hanzi 汉字, where the ‘Han’ again refers to the largest ethnic group and ‘zi’ refers to a unit of writing. The earliest specimens of sinograms currently available date back only some 3,000+ years. However, given the maturity of these specimens, as we saw in the inscriptions in Figure 10.3, it is almost certain that the earliest forms of this script date much further back; presumably such ancient specimens are buried deep somewhere, waiting to be discovered in some future excavation. The first sinograms invented were in all likelihood iconic in nature, in that their shapes are suggestive of some physical properties of the object represented – such as the head of a sheep 羊 yang, or the branching of a tree 木 mu, or the two sides of a door 門 men, now simplified to 门. A particularly interesting iconic sinogram was 且 qie, which is based on a phallic symbol, showing early awareness of the important role males play in reproduction, 170

Chinese linguistics

and of the father–child bond in addition to the mother–child bond.33 The emergence of this sinogram may be related to the early transition from matrilineal to patrilineal society in Ancient China. Before long, these simple sinograms were combined into more complex ones, such as putting two or three trees together 林 lin, 森 sen to mean ‘forest’, or combining 日 ri ‘sun’ and 月 yue ‘moon’ to mean ‘brightness’ 明 ming.34 且 qie became combined with another sinogram, 示 shi, which has to do with matters of the spirit; the complex sinogram which results is 祖 zu, meaning ‘ancestor’, a core concept in Chinese civilization. The 礻on the left side of 祖 is a compressed version of 示; such compression commonly occurs when simple sinograms are combined to form complex ones. However, by far the most productive principle in constructing sinograms is to combine a component having to do with sense with another component having to do with sound. This can be exemplified by 材 cai, where the left component is the ‘tree’ shown above, or more generally, ‘wood’; here it gives the approximate sense that the sinogram refers to the nature of the material some object is made from. The right component 才 is another simple sinogram, pronounced cai. Thus the complex sinogram is pronounced cai. Such sinograms are called ‘phonograms’ in English, or xingsheng zi 形声字 in Chinese, literally, ‘shape–sound writing’. We have already seen several phonograms in our earlier discussion of Old Chinese. As Chinese civilization progressed over the millennia, the number of sinograms accumulated, so that large dictionaries list over 50,000 of them, including some whose shapes are extremely complex in terms of the number of strokes they contain. No one of course knows all of these sinograms, let alone uses them on a daily basis. For ordinary usage, such as reading newspapers, some 3,000 or 4,000 would suffice. There have been several occasions when government stepped in to systematize the sinograms, to standardize usage, and to redesign some of them. An early occasion was during the time of the First Emperor, when the Qin conquered the other kingdoms and unified China into a single empire. The most recent occasion of standardization for writing Chinese was in the 1960s, when Hanyu Pinyin was officially adopted, and the first batch of simplified sinograms were introduced. This resulted in two systems of sinograms: the traditional sinograms called either zhengti zi 正体字 or fanti zi 繁体字, and the simplified sinograms called jianti zi 简体字. Mainland China and Singapore now use jianti zi; on the other hand, Hong Kong, Macao, and Taiwan continue to use zhengti zi. The sinograms in this chapter are typically in jianti zi. The sinograms used in Japan, called kanji, which is the Japanese pronunciation of the Chinese word Hanzi, underwent some changes after they were adopted, some of them well over a thousand years ago. Many of the kanji have the same shape as both the zhengti and the jianti, as illustrated in line 1 of Table 10.4. Or, all three sinograms may be different, as illustrated in line 2. In line 3, the kanji is the same as the zhengti, while in line 4, the kanji is the same as the jianti. Table 10.4  Comparison of kanji, zhengti and jianti

1. 2. 3. 4.

Kanji

Zhengti

Jianti

母 読 車 体

母 讀 車 體

母 读 车 体

171

meaning meaning meaning meaning

‘mother’ ‘to read’ ‘vehicle’ ‘body’

William S.-Y. Wang Table 10.5  Comparison of traditional and simplified sinograms Traditional and simplified sinograms Trad.

Simp.

Trad.

Simp.

Trad.

Simp.

十 土 水 我 的 家 金 笑

十 土 水 我 的 家 金 笑

滅 龍 夢 廣 護 農 葉 塵

灭 龙 梦 广 护 农 叶 尘

詳 讀 銀 鐵 紅 線 財 帳

详 读 银 铁 红 线 财 帐

All four possibilities occur. Nonetheless, Chinese and Japanese readers can usually read texts in any of the three systems of sinograms, with varying degrees of difficulty. Table 10.5 illustrates the differences between the zhengti zi and the jianti zi in terms of three categories. As can be seen by comparing the leftmost pair of columns, many of the high-frequency sinograms have not been affected – they are the same in both systems. In the second category in the middle pair of columns, however, the sinograms have been simplified quite a bit. The top sinogram 滅, for example, has 13 strokes while its simplied form 灭 has only five strokes. The reduction in the number of strokes in these sinograms is the most significant. For the third category, illustrated in the last pair of columns in Table 10.5, the number of sinograms affected is the greatest. This is because the simplification is applied to the radicals, and each simplified radical affects dozens of sinograms which contain that radical. For example, the radical 言 in the top sinogram 详 has seven strokes, while the simplified form of this radical in 详 has only two strokes. Thus this saving of five strokes applies to all sinograms which contain this radical. There are occasional exceptions to this generalization; examples are the 5th sinograms in the middle pair of columns. Here, the traditional form 護 ‘to protect’ with the 言 ‘speech’ radical has been simplified to 护, changing it into a phonogram with a new ‘hand’ radical 扌 derived from 手, and a phonetic 户, pronounced hu. Moreover, to make the discussion here more complete, note also that in the even numbered sinograms in this rightmost pair of columns, the right component of the sinograms have been additionally simplified as well. Sustained Western influence entered China around 1580 in the form of Christian missionaries; an early one who had a deep impact was the Italian Jesuit Matteo Ricci.35 Among Ricci’s contributions was his translation of Euclid’s Elements, which was published in 1607. Over the next centuries many others came, bringing with them pieces of Western knowledge in various disciplines, including astronomy, mathematics, and medicine. Ricci’s work on the Chinese language laid the foundation for a successor, the Flemish Jesuit Nicolas Trigault,36 to build upon and devise the first Romanization of Chinese, published in 1626. Such works provide valuable data for investigating the history of the language. But the system of Romanization for Mandarin that had the longest influence in Chinese studies was one that was devised by two Englishmen; started by Thomas Wade and completed by Herbert Giles at the end of the nineteenth century. The Wade-Giles system was by far the most used Romanization for Mandarin until 1958, when Hanyu Pinyin became officially 172

Chinese linguistics

adopted by the National People’s Congress in Beijing. One can often learn something about a Chinese from the way his surname is Romanized. For instance, the common surnames 张, 江, and 徐 would be spelled Zhāng, Jiāng, and Xú according to Hanyu Pinyin, but Chang1, Chiang1, and Hsu2 according to the Wade-Giles system. If a person’s surname is spelled according to Wade-Giles, it means either that the surname was Romanized before Hanyu Pinyin became official, or that he is from Taiwan, where Hanyu Pinyin was not officially adopted until 2009. In connection with Romanization, mention should also be made of an ingenious system mostly developed by the linguist Y. R. Chao called Gwoyeu Romatzyh 国语罗马字 or Guóyǔ Luómǎ zì in Hanyu Pinyin. In this system, tone distinctions do not require superscript numerals as used by Wade-Giles or diacritic marks as used by Hanyu Pinyin. Instead, the distinctions are spelled into the letters for vowels and consonants. For instance, the syllable guo would be written guo, gwo, guoo, and guoh, four different spellings for the four tones. Table 10.6 gives a quick comparison to illustrate the three systems we have just mentioned. As we saw earlier, Pinyin uses diacritics to mark tones while Wade-Giles uses superscript numerals. For unvoiced, unaspirated stops, Pinyin uses ‘b’ to distinguish from unvoiced aspirated ‘p’. For this distinction, Wade-Giles uses an apostrophe instead, which is closer to the notation of the International Phonetic Alphabet. Gwoyeu Romatzyh is unique in distinguishing the four tones of Putonghua by changing the vowels and consonants in the spelling. For instance, the vowel is doubled in rows 1, 2, 5, and 6 to indicate the longer syllables associated with the PTH tone 3. Similarly, it uses consonants like ‘q’ and ‘h’ to indicate the shorter syllables associated with the PTH tone 4. While Gwoyeu Romatzyh has some loyal users, who stress its greater effectiveness in learning and remembering tones, and Wade-Giles continues to be used in some contexts, Hanyu Pinyin has emerged as the one standard system; the advocacy of this system has been summarized in a document published in 2001 in the Journal of Chinese Linguistics jointly by Cheng Chin-Chuan 郑锦全, Ting Pang-hsin, William S.-Y. Wang 王士元, and Mei Tsulin. We will consider briefly some aspects of pronunciation that an English speaker should note about the Pinyin system. For a fuller discussion as well as a CD to illustrate the sounds, the reader may consult Lin (2007). While our focus is on Putonghua, many of these aspects pertain to other Chinese dialects as well. Following the traditional order, we will begin with tones, proceed to consonants, to vowels, and to syllable structure. Chinese is a tone language in the sense that the words are constructed not only with consonants and vowels, as in all the languages of the world, but also with tones, realized by distinctive pitch contours. Although we cannot be sure at present, it is quite likely that tones have been in the Chinese sound system since ancient times.37 In Putonghua these tones Table 10.6  Comparison of three Romanization systems

1. 2. 3. 4. 5. 6. 7.

Hanyu Pinyin

Wade–Giles

Gwoyeu Romatzyh

bang pang zai cai xi jin qiao

pang p’ang tsai ts’ai hsi chin ch’iao

bang pang tzai tsai shi jin chiau

173

barng parng tzair tsair shyi jyn chyau

baang paang tzae tsae shii jiin cheau

banq panq tzay tsay shih jinn chiaw

William S.-Y. Wang

are typically numbered as 1 = high, 2 = rising, 3 = low, and 4 = falling. Other dialects may differ in the number of tones they have and in how their tones are numbered. The tones in Putonghua may be illustrated with a set of words that differ only in tone, such as: 1 2 3 4

= = = =

mā, má, mǎ, mà,

妈 麻 马 骂

meaning meaning meaning meaning

‘mother’, or ‘hemp’, ‘horse’, ‘to scold’,

shī, shí, shǐ, shì,

诗 时 史 市

meaning meaning meaning meaning

‘poetry’ ‘time’ ‘history’ ‘city’

Oftentimes, the diacritic in tone 1 is left out, and a syllable not marked with tone diacritics is understood to carry tone 1.38 Since tone 1 occurs with the highest frequency among the four tones, not marking it gives the page a cleaner look. However, when tone diacritics are left out altogether, many ambiguities arise. An inconvenient ambiguity that arises in connection with Ancient China is between Zhōu 周, the name of a dynasty, and Zhòu 纣, the name of the last king of the Shang dynasty. Another ambiguity is between two names of provinces, Shānxi 山西 and Shǎnxi 陕西. The conventional solution here is to write them respectively as Shanxi and Shaanxi.39 There are cases of true homophony as well, such as the names of two ethnic groups in China, both pronounced Yí. They are written distinctly of course: 夷 versus 彝. In general the letters for consonants in Pinyin have similar values as they have in English. The letter ‘g’ before ‘e’, is always pronounced like in English ‘get’ and never like English ‘gem’. The English ‘r’ sound is always pronounced with the lips rounded, like in whistling. In contrast, the Chinese ‘r’ is pronounced with the lips rounded only before rounded vowels like ‘u’, like in ru 入 ‘enter’; otherwise the lips are not rounded, like in re 热 ‘hot’. The differences between Pinyin and English become greater in the case of the consonants arranged in Table 10.7 according to their phonetic values. After each Pinyin letter, I have also provided its closest English equivalent. The relevant letters in the English words are underlined. The consonants in the top two rows are called affricates; the aspirated ones are pronounced with a puff of air immediately following, while the unaspirated ones are not followed by such a puff of air. In the dental column, the sounds represented by Pinyin ‘z’ and ‘c’ do not occur as initial consonants in English. A good way to start practicing them is to put a word after them, such as ‘bids are’ and ‘bits are’, and then suppress the ‘bi’, much as one removes scaffolding after it has served its purpose. The consonants in the retroflex column are perhaps the most unfamiliar and challenging to speakers of other languages. Their Pinyin spelling is unusual in that a sequence of two letters is used to spell single sounds, like the use in English of ‘th’ for spelling the word ‘the’. They are called retroflex because they are pronounced with the tip of the tongue tilted slightly upward, as one can verify from x-ray photographs. We will see these consonants a lot in the pages to follow, since the name of both the Shang dynasty and the Zhou dynasty begin with retroflex consonants in Putonghua. Table 10.7  Pinyin and some PTH obstruents

Unaspirated Aspirated Fricative

Dental

Palatal

Retroflex

z = bids c = bits s = sew

j = gin q = chin x = shin

zh = drew ch = true sh = shrew

174

Chinese linguistics Table 10.8  The vowel spelled ‘i’ in Pinyin

Unaspirated Aspirated Fricative

Dental

Palatal

Retroflex

zi = 字 ‘character’ ci = 次 ‘second’ si = 四 ‘four’

ji = 记 ‘record’ qi = 气 ‘air’ xi = 细 ‘thin’

zhi = 志 ‘ambition’ chi = 翅 ‘wing’ shi = 市 ‘city’

The consonants clusters in the English words in Table 10.7 are similar in acoustic effect; nonetheless they are but an approximation to the Chinese retroflex consonants. However, it may be a comfort to some readers who are especially challenged here to know that many Chinese speakers coming from some dialect regions often speak Putonghua without these retroflexes. They are simply pronounced as dental consonants instead. It is useful to compare the Pinyin vowels with those in European languages where the spelling is more regular and simple, like Italian or Spanish. The situation with English is much more complicated because every vowel letter has several different pronunciations, depending on the word; for example the ‘i’ in ‘five’ and ‘fifth’; or the ‘e’ in ‘me’ and ‘met’. The only Pinyin vowel that the reader may find challenging is ‘i’. To discuss this vowel, in Table 10.8 we return to the consonants we have met, except now they are followed by the Pinyin vowel ‘i’. For the palatal column, there should be no problem: Pinyin ‘ji’ is like English ‘gee’, Pinyin ‘qi’ is like English ‘chee’, and Pinyin ‘xi’ is like English ‘she’, all pronounced with a falling tone. However, in the dental column, Pinyin ‘i’ is pronounced as an apical vowel; this is a sound most readers unfamiliar with Chinese would not have encountered. The sound has been described as a continuation of the preceding consonant but without the frication. The same description may be offered for pronouncing ‘i’ after the retroflex consonants, except there should be ‘r’-coloring added as well. Of course, no amount of paper discussion on pronunciation can substitute for the benefits of just listening to some examples by a friendly native speaker. After all, the eye is no substitute for the ear. The nine words in Table 10.8 are selected to illustrate the phonetics just discussed, specifically the apical vowels and retroflex consonants. As if to make up for the problems that the tones, the retroflexes, and the apical vowel may cause the learner of the language, the syllable structure in modern Chinese is elegantly simple. A syllable begins with at most one consonant, which may or may not be followed by a glide. A syllable may end with at most one consonant. In the Pinyin system, a glide which begins a syllable is spelled ‘w’ or ‘y’. However, when the glide is preceded by a consonant, it is spelled ‘u’ or ‘i’. A syllable without consonants at the end is called ‘open’; otherwise it is ‘closed’. All eight possible patterns are shown in Table 10.9, with illustration for each pattern, all in Table 10.9  Shapes of the Putonghua syllable Open syllable 1. 2. 3. 4.

V GV CV CGV

a wa ha hua

阿 蛙 哈 花

Closed syllable

‘prefix’ ‘frog’ ‘interjection’ ‘flower’

5. 6. 7. 8.

175

VC GVC CVC CGVC

an wan han huan

安 弯 憨 欢

‘peace’ ‘curve’ ‘honest’ ‘happy’

William S.-Y. Wang

tone 1. These eight patterns may be summarized by a simple formula, i.e. (C)(G)V(C), where C is consonant, G is glide, and V is vowel, using parentheses to indicate optional elements.40 Ambiguities may arise with certain sequences of consonants and vowels with respect to syllable boundaries. For such cases we use an apostrophe to indicate a syllable boundary. For instance, we write Xi’an for the disyllabic name of the famous city in the province of Shaanxi. Written without the apostrophe, xian is a single syllable 县 that may mean ‘county’. Xi’an means ‘western peace’. In olden times, as in the Tang dynasty, the city was called Chang’an 长安, meaning ‘eternal peace’. Note, however, without the apostrophe to mark the syllable boundary, the sequence could represent the two syllables chan and gan, each representing several sinograms. One final comment on the scope of the syllable in Chinese. As mentioned earlier, there are three dimensions in language that may be called the three S’s: namely, Sound, Script, and Sense. The syllable is a unit of Sound, but in Chinese it always corresponds to a unit in Script, which is the zi 字, or Chinese character. In turn, the Chinese character almost always corresponds to a unit in Sense, which is the morpheme.41 Thus in Chinese there is the simple equation relating sound to script to sense: One syllable = One sinogram = One morpheme Such an equation would not hold for an alphabetic language like English, where the unit in Script is the word, which may contain two or more syllables as well as two or more morphemes. In Chinese, words are not separated by spaces as in English. Here is an example to illustrate the difference. The Chinese sentence has four syllables, four characters, and four morphemes; the relations among the three dimensions are straightforward. In contrast, the corresponding English sentence with the same meaning has four syllables, the first syllable containing two morphemes, syllables 2 and 3 represent the third morpheme, and syllable 4 representing the fourth morpheme. The relations are therefore more complex. Chinese: Ni de fang ke.  你  的  访  客. English: You-r  visit-or.

Retrospect and prospect In addition to the topics discussed above, there are many others which merit inclusion in an introductory chapter such as this one, but that we must forego due to lack of space. The reader will find them treated in greater depth in various parts of this Encyclopedia, as well as in the recently published Oxford Handbook of Chinese Linguistics (Wang and Sun, ed. 2015). An excellent book-length overview of the field is Norman (1988). For deeper bibliographic coverage, there is Wang and Lyovin (1970) for earlier studies, and Wu (2009) for publications in the Journal of Chinese Linguistics since 1973. Toward presenting a more complete picture on Chinese Linguistics in this chapter, some references are suggested here. For a bird’s eye view of Chinese history, the series of atlases edited by Tan Qixiang 谭其骧 (1982) is a valuable reference. In relating language to folk culture to China, the Manzu 满族 scholar Luo Changpei (1989/1950) was an early pioneer; two recent books in this tradition are Zhou and You 周振鹤、游汝杰 (1986), and Deng Xiaohua 邓晓华 (1993). 176

Chinese linguistics

In the area of language contact, an important discussion was that of Hashimoto Mantaro 桥本万太郎, who published in Japanese in 1977 (Chinese translation 1985). A more recent investigation is the book by Chen Baoya 陈保亚 (1996), which offers a more fine-grained discussion of extended multilingualism in China’s Southwest, and the in-depth analysis of Hanyu and Bai by Wang Feng 汪锋 (2012). Of particular interest is the analysis of extensive mixing of Hanyu and Zangyu (Tibetan) in a community in Sichuan by the Zangzu 藏族 scholar Yeshes Vodgsal Atshogs 意西微萨•阿错 (2005), and of Hanyu and Zhuangyu in Guangdong by the Zhuangzu 壮族 scholar Li Jingzhong 李敬忠 (2006). Minority languages are finally getting the attention they have always deserved.42 In the area of historical linguistics, the textbook by Xu Tongqiang 徐通锵 (1991) has been influential for explaining the subject in Chinese in a wide perspective. The monograph by Pan Wuyun 潘悟云 (2000) presents a systematic account of Chinese phonology as it has developed from 3,000 years ago to the present. With modern phonetics, we are now in a much better position to examine with great precision and with large corpora of Hanyu as it is currently spoken, as illustrated by Peng (2006). With focus specifically on lexical tones, Cheng and Wang (1971) tabulate the categorical changes from Middle Chinese, while Chen (2000) explores the depth of how these tones sandhi from dialect to dialect. The basic units of any language are its words. Masini (1993) explores how the Chinese lexicon was formed, especially over the past two centuries. Cheng (1998) reports the very interesting finding that there is a constant upper bound in the number of sinograms used throughout the dynastic histories, even though the cumulative number inevitably increases over the two millennia. A recent discussion of the full Chinese lexicon is Yip (2000); a detailed analysis of how Chinese words are formed is Packard (2000), especially with respect to derivational morphology. Recently several volumes have been published on aspects of the lexicon which are useful both for research and for language teaching. These include an extensive compilation of synonyms by Zhang (2010), and collections of idioms (2011) and of proverbs (2014) by Jiao and colleagues. Although Chinese scholars have discussed grammatical categories within a traditional perspective for centuries, Sun (1996) offers a modern approach on how these categories may change. The volume by Xu (2006) discusses the major historical changes in Chinese syntax, especially their effects on typology. As for an overall synchronic grammar of the language, an early one in the indigenous framework is Lü Shuxiang 吕叔湘 (1957). The classic treatment, however, is that by Chao (1968), a distillation of many decades of acute observations on the language. It enjoys the singular distinction of having been translated into Chinese twice within a year of each other: in 1979 by Lü, then director of the Institute of Linguistics in Beijing, and in 1980 by Ting Pang-hsin, then director of the Institute of History of Philology43 in Taipei. A series of later grammars followed, each with its own orientation of applications and theories; these include: Li and Thompson (1981), Cheung et al. (1994), Tao (1996), Chen (1999), Chappell (2001), Ross and Ma (2006), Huang et al. (2009), etc. While there is no dearth of treatments on Mandarin, there is a clear lack of grammars of the other dialects, and especially of the minority languages. In recent decades, Chinese linguistics has broadened its horizons considerably. As Edward Sapir wrote cogently nearly a century ago when he discussed the status of linguistics as a science (1929: 214):44 ‘It is peculiarly important that linguists, who are often accused, and accused justly, of failure to look beyond pretty patterns of their subject matter, should become aware of what their science may mean for the interpretation of human conduct in general.’ The broadening is in part reflected in the title of the anthology compiled by Jing-Schmidt (2013): ‘Increased Empiricism’. Linguists have joined up with archeologists and geneticists 177

William S.-Y. Wang

to explore prehistory within a multidisciplinary perspective; see Jin et al. (2001). Increasing attention is being paid to the written language, and the associated issues of literacy, such as in the discussions by Wang et al. (2009) and in the anthology compiled by Feng and Branner (2011). Language is a behavior, ultimately the result of activities of the brain. It is particularly encouraging that scholars from psychology and neurosciences have brought in their expertise and are now collaborating with linguists to investigate the languages and dialects of China. Here is a sample of an expanding literature on both normal language as well as language disorders, including several very useful anthologies: Tzeng et al. (1977), Tzeng and Wang (1983), Chen and Tzeng (1992), Li et al. (2006), Law et al. (2008). A quarter century ago (1988), the linguist Jerry Norman produced a masterly synopsis of Chinese linguistics in a succinct 290 pages; see Coblin’s eulogy (2013). It is still the best single-volume introduction to the essentials of the field. However, so much has happened on many fronts over this quarter century that a comparable synopsis now would certainly require many more pages to do justice to the field. This can be seen in the massive Oxford Handbook of Chinese Linguistics published in 2015, edited by Wang and Sun, which brings together the expertise of 66 scholars in the field.

Notes   1 Chinese words are spelled in Hanyu Pinyin, and will be italicized when they first appear. They are usually written in simplified sinograms jianti zi 简体字; however traditional sinograms zhengti zi 正体字 are used when special needs arise.  2 Confucius, The Analects; translated by D. C. Lau (1992: 120), Chinese University of Hong Kong Press.   3 See: W. S.-Y. Wang (1989) ‘Language in China: a chapter in the history of linguistics’, Journal of Chinese Linguistics 17: 183–222; and Anthony Yu 余国藩 (2008) ‘Cratylus and Xunzi on Names’, Classical, Comparative, and Contemporary: Essays on Literature, Religion, and Other Topics, Columbia University Press.   4 The British named it Mount Everest, apparently not aware that the mountain already had a name which dates back several centuries. A Chinese name for the peak is Shengmufeng 圣母峰, which is an approximation to the Tibetan meaning.   5 Sections of the Changjiang also have other names: Jinshajiang 金沙江 for its initial section and Yangzijiang 扬子江 for its middle section. The spelling Yangtze is an older form of Yangzi.   6 Heather Pringle (2014) ‘Welcome to Beringia’, Science 343(6174): 961–3; John F. Hoffecker, Scott A. Elias, and Dennis H. O’Rourke (2014) ‘Out of Beringia?’ Science 343(6174): 979–80.   7 John D. Bengtson (1999) ‘Wider genetic affiliations of the Chinese language’, Journal of Chinese Linguistics 27: 1–12.   8 However, this view may be complicated by recent findings in molecular genetics; cf. s32 of The HUGO Pan-Asian SNP Consortium et al. (2009) ‘Mapping human genetic diversity in Asia’, Science 326(5959): 1541–5. P. Li (2015) is an up-to-date discussion of Austronesian linguistics in the light of the most recent archeological finds.  9 Dates for dynasties in this chapter follow those given in the Xinhua Zidian 新华字典, 11th ed. (2011). However, it is good to remember that the dynasties often overlap quite a bit, and many dates are still debated among specialists. 10 Ian Johnson (2013) ‘China’s ancient lifeline’, National Geographic, May. 11 Known in Chinese as 利玛窦 Lì Mǎdòu. 12 We thank Professor Paul J. K. Li of Academia Sinica for this map. 13 Elizabeth Pennisi (2013). ‘More genomes from Denisova Cave show mixing of early human groups’, Science 340(6134): 799. 14 Quote is from the US anthropologist F. Clark Howell in his introduction to the 1985 anthology edited by R. Wu and D. Olsen, p. xx. 15 Li, Xueqin, Garman Harbottle, Juzhong Zhang, and Changsui Wang (2003), ‘The earliest writing? Sign use in the seventh millennium BC at Jiahu, Henan Province, China’, Antiquity 77: 31–44.

178

Chinese linguistics 16 Population geneticist Cavalli-Sforza pioneered the multidisciplinary study of human prehistory, and gives a general account of current knowledge in this area in his 2000 book, Genes, Peoples, and Languages, London: Penguin Books. 吴一丰 等译 (2003) 追踪亞当夏娃。台北：远流出版社。 17 Weidenreich, Franz (1943) ‘The skull of Sinanthropus pekinensis: a comparative study on a primitive hominid skull’, Palaeontologia Sinica 10: 1–289. 18 Wu, Xinzhi and Frank E. Poirier (1995) Human Evolution in China, Oxford University Press. 19 Shang, Hong, Haowen Tong, Shuangquan Zhang, Fuyou Chen, and Erik Trinkaus (2007) ‘An early modern human from Tianyuan Cave, Zhoukoudian, China’, PNAS 104: 6,573–78. 20 Wu, Xiaohong, Chi Zhang, Paul Goldberg, David Cohen, Yan Pan, Trina Arpin, and Ofer Bar-Yosef (2012) ‘Early pottery 20,000 years ago in Xianrendong Cave, China’, Science 336: 1,696–70, June 29. 21 See Wu and Olsen 1985 for a fuller listing of these sites. 22 Another Swedish scholar whose work is of great importance was Bernhard Karlgren 高本汉 (1889– 1978), a younger colleague of Andersson. Karlgren introduced linguistics methods developed in Europe to China, and exemplified these methods with dialect studies as well as historical reconstruction. Andersson founded the famous Museum of Far Eastern Antiquities in Stockholm in 1926; Karlgren succeeded him as director of the Museum (1939–1959). 23 W. S.-Y. Wang 1998: Figure 1. 24 Green et al. 2010. 25 The names in Figure 10.2 are written in the Wade-Giles spelling, which was widely used in Western literature until it was replaced by Hanyu Pinyin. Archeology has progressed tremendously over the last several decades since Chang 1986. See Chang 1999. A good overview of the current state of knowledge is the elegant volume edited by Chang and Xu 2005. 26 This date is 1,000 years earlier than the traditional folk view in China that the civilization dates back 5,000 years. 27 This sample is discussed in greater detail in Wang (2013) Love and War in Ancient China. 28 Old Chinese is also called Archaic Chinese; Middle Chinese is also called Ancient Chinese. 29 Translation based on D. C. Lau 2002: 73. 30 See 王士元 (2013) 谁是中国人?《科学中国人》8: 38–43. 31 See Shen, Zhongwei (2011) ‘The origin of Mandarin’, Journal of Chinese Linguistics 39: 1–31. 32 For a comparative discussion of Chinese writing and alphabetic writing, see W. S.-Y. Wang and Tsai 2011. For a historical survey of literacy in China, see F. Wang et al. 2009. 33 Ancient writings often spoke of the offspring knowing the mother but not the father, 知母不知父, which is a natural outcome of primitive biological scenarios. 34 Text starting from this paragraph to the third paragraph below Table 10.5 (including Table 10.5) on p. 172 is originally published in Love and War in Ancient China: Voices from the Shijing, pp. xxix–xxxii, © 2013 City University of Hong Kong. Used by permission of City University of Hong Kong Press. All rights reserved. 35 A highly readable account of Matteo Ricci in China is: Spence, Jonathan D. (1984) The Memory Palace of Matteo Ricci, New York: Viking. 36 Known in Chinese as 金尼阁, Jīn Nígé. His Xiru Ermu Zi 西儒耳目资, contains the first Romanization system for writing Chinese. 37 For the historical background of Chinese tones, see W. S.-Y. Wang and Cheng 1987. It is very likely that Chinese had become a fully tonal language by the Han dynasty, though the tones were not explicitly discussed until several centuries later, presumably under the influence of Sanskrit writings. Text starting from this paragraph to the section end on p. 176 (including Tables 10.7, 10.8 and 10.9) was originally published in Love and War in Ancient China: Voices from the Shijing, pp. xxii–xxvii, © 2013 City University of Hong Kong. Used by permission of City University of Hong Kong Press. All rights reserved. 38 To be more precise, unstressed syllables without tones are also unmarked with diacritics; however they can usually be distinguished by context. 39 Shaanxi is a generally used artificial spelling where the ‘aa’ indicates the third tone, to distinguish it from the name of another province, Shanxi, which has the first tone. 40 In this formula, the V stands for vocalic nucleus, which includes simple vowels as well as diphthongs. 41 There are few exceptions here, such as 葡萄 ‘grape’, 玻璃 ‘glass’, 蟋蟀 ‘cricket’, etc. 42 This is a point I tried to highlight at the formation of the International Association of Chinese Linguistics; see W. S.-Y. Wang 1993.

179

William S.-Y. Wang 43 The Institute of History and Philology was one of the first research institutes established at the Academia Sinica, home of eminent linguists such as Y. R. Chao and F. K. Li. In 2004, some members formed a new Institute of Linguistics for further development of the field. 44 Sapir, Edward (1929) ‘The status of linguistics as a science’, Language 5: 207–14.

Selected references English Abi-Rached, Laurent, et al. (2011) ‘The Shaping of Modern Human Immune Systems by Multiregional Admixture with Archaic Humans’, Science 334: 89–94, 7 October. Baxter, W. H. (1992) A Handbook of Old Chinese Phonology, Berlin: Mouton de Gruyter. Chang, Kwang-chih (1986) The Archeology of Ancient China, 4th ed., New Haven: Yale University Press. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, University of California Press, Chinese translations by Lü Shuxiang 吕叔湘 in 1979, and by Ting Pang-hsin 丁邦新 in 1980. Chappell, Hilary (ed) (2001) Sinitic Grammar: Synchronic and Diachronic Perspectives, Oxford: Oxford University Press. Chen, Hsuan-chih and Ovid J. L. Tzeng (eds) (1992) Language Processing in Chinese, Amsterdam: Elsevier. Chen, M. Y. (2000) Tone Sandhi: Patterns across Chinese Dialects, Cambridge: Cambridge University Press. Chen, Ping (1999) Modern Chinese: History and Sociolinguistics, Cambridge: Cambridge University Press. Cheng, Chin-Chuan (1973) ‘A Quantitative Study of Chinese Tones’, Journal of Chinese Linguistics 1: 93–110. Cheng, Chin-Chuan (1968) ‘English Stresses and Chinese Tones in Chinese Sentences’, Phonetica 18: 77–88. Cheng, Chin-Chuan (1994) ‘DOC: lts Birth and Life’, in M. Y. Chen and Ovid J. L. Tzeng (eds) In Honor of William S.-Y. Wang: Interdisciplinary Studies on Language and Language Change, Taipei: Pyramid Press. Cheng, Chin-Chuan (1998) ‘Quantification for Understanding Language Cognition: Quantitative and Computational Studies on the Chinese Language,’ in B. K. T’sou, T. B. Y. Lai, S. W. K. Chan, and W. S.-Y. Wang (eds) Quantitative and Computational Studies on the Chinese Language, Hong Kong: City University of Hong Kong Press, 15–30. Cheng, Chin-Chuan and W. S.-Y. Wang (1971) ‘Phonological Change of Middle Chinese Initials’, Tsing Hua Journal of Chinese Studies 9: 216–70. Cheung, Samuel Hung-nin, Liu Sze-yun, and Shih Li-Lin (1994) A Practical Chinese Grammar, Hong Kong: The Chinese University Press. Coblin, W. South (2013) ‘Jerry Norman: Remembering the Man and His Perspectives on Chinese Linguistic History’, Journal of Chinese Linguistics 41: 219–45. Feng, Li and David Prager Branner (eds) (2011) Writing and Literacy in Early China, Seattle: University of Washington Press. Ho, Dah’an (2015) ‘Such Errors Could be Avoided. Review of “Old Chinese, a New Reconstruction by W.H. Baxter and Laurent Sagart” ’. In press with Journal of Chinese Linguistics. Huang, C-T. James, Y.-H. Audrey Li, and Li Yafei (2009) The Synrax of Chinese, Cambridge: Cambridge University Press. Jiao, Liwei, Cornelius C. Kubler, and Weiguo Zhang (2011) 500 Common Chinese Idioms, London and New York: Routledge. Jiao, Liwei and Benjamin M. Stone (2014) 500 Common Chinese Proverbs and Colloquial Expressions, London and New York: Routledge. Jin, Li, Mark Seielstad, and Chunjie Xiao (eds) (2001) Genetic, Linguistic and Archaeological Perspectives on Human Diversity in Southeast Asia, Singapore: World Scientific. Jing-Schmidt, Zhuo (ed) (2013) Increased Empiricism: Recent Adventures in Chinese Linguistics, Amsterdam and Philadelphia: John Benjamins. Karlgren, Bernhard (1939) Etudes sur la phonologie chinoise, Chao Y. R. 赵元任, Li Fangkuei 李方桂, and Luo Changpei 罗常培 (trans.)《中国音韵学研究》(A Study of Chinese Phonology), Beijing: The Commercial Press.

180

Chinese linguistics Karlgren, Bernhard (1950) The Book of Odes, Stockolm: Museum of Far Eastern Antiquities. Keightley, David N. (1983) The Origins of Chinese Civilization, Berkeley: University of California Press. Law, Sam-Po, Brendan Stuart Weekes, and Anita Mei-Yin Wong (eds) (2008) Language Disorders in Speakers of Chinese, Multilingual Matters / Channel View Publications. Li, Charles N. and Sandra A. Thompson (1981) Mandarin Chinese: A Functional Reference Grammar, Berkeley: University of California Press. Li, Paul Jenkuei (2015) ‘The Discovery of Liangdao Man and its Implications for the Pre-Austronesian Homeland’, Journal of Chinese Linguistics 43: 224–31. Li, Ping, Li Hai Tan, Elizabeth Bates, and Ovid J. L. Tzeng (eds) (2006) The Handbook of East Asian Psycholinguistics: Volume 1, Chinese, Cambridge: Cambridge University Press. Lin, Yen-Hwei (2007) The Sounds of Chinese, Cambridge: Cambridge University Press. Liu, Wu, Chang-Zhu Jin and Ying-Qi Zhang, Yan-Jun Cai, Song Xing, Xiu-Jie Wu, Hai Cheng, R. Lawrence Edwards, Wen-Shi Pan, Da-Gong Qin, Zhi-Sheng An, Erik Trinkaus, and Xin-Zhi Wu (2010) ‘Human Remains from Zhirendong, South China, and Modern Human Emergence in East Asia’, PNAS 107: 19,201–06. Masini, Federico (1993) The Formation of Modern Chinese Lexicon, Journal of Chinese Linguistics Monograph No. 6, Huang Heqing 黄河清 (trans.) (1997) Xiandai Hanyu Cihui de Xingcheng《现代 汉语词汇的形成》, Shanghai: Hanyu dacidian chubanshe 汉语大词典出版社. Mei, Tsulin (1977) ‘Tones and Tone Sandhi in 16th century Mandarin’, Journal of Chinese Linguistics 5: 237–60. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Norman, Jerry and Tsulin Mei (1976) ‘The Austroasiatics in Ancient South China: Some Lexical Evidence’, Monumenta Serica 32: 274–301. Packard, Jerome L. (2000) The Morphology of Chinese: A Linguistic and Cognitive Approach, Cambridge: Cambridge University Press. Peng, Gang (2006) ‘Temporal and Tonal Aspects of Chinese Syllables: A Corpus-based Comparative Study of Mandarina and Cantonese’, Journal of Chinese Linguistics 34: 134–54. Pulleyblank, E. G. (1993) ‘Old Chinese Phonology: A Review Article’, Journal of Chinese Linguistics 21: 337–80. Ross, Claudia and Jing-Heng Sheng Ma (2006) Modern Mandarin Chinese Grammar, London and New York: Routledge. Shen, Guanjun, Xing Gao, Bin Gao, and Darryl E. Granger (2009) ‘Age of Zhoukoudian Homo erectus Determined with 26Al/10Be Burial Dating’, Nature 458: 198–200. Simmons, Richard VanNess (1999) Chinese Dialect Classification: A Comparative Approach to Harngjou, Old Jintarn, and Common Northern Wu, Amsterdam and Philadelphia: John Benjamins. Sun, Chaofen (1996) Word-order Change and Grammaticalization in the History of Chinese, Stanford: Stanford University Press. Takashima, Ken-ichi (2012) ‘Etymology and Paleography of the Yellow River’, Journal of Chinese Linguistics 40: 269–306. Tao, Hongyin (1996) Units in Mandarin Conversation: Prosody, Discourse and Grammar, Amsterdam and Philadelphia: John Benjamins. T’sou, B. K. Y. (1981) ‘A Sociolinguistic Analysis of the Logographic Writing System of Chinese’, Journal of Chinese Linguistics 9: 1–19. Tzeng, Ovid J. L., D. L. Hung, and W. S.-Y. Wang (1977) ‘Speech Recoding in Reading Chinese Characters’, Journal of Experimental Psychology: Human Learning and Memory 3: 621–30. Tzeng, Ovid J. L. and W. S.-Y. Wang (1983) ‘The First Two R’s’, American Scientist 71: 238–43. Wang, Feng, Yaching Tsai, and W. S.-Y. Wang (2009) ‘Chinese Literacy’, D. Olson and N. Torrance (eds) Cambridge Handbook on Literacy, Cambridge: Cambridge University Press, 386–417. Wang, W. S.-Y. (1973) ‘The Chinese Language’, Scientific American 228: 53–62. February issue. Follow-up discussion in May 1973, 228: 8–9. Chinese translation by You Rujie 游汝杰 1987, in Yuyan yu Renlei Jiaoji《语言与人类交际》, Guangxi Education Press 广西教育出版社, 56–71. Wang, W. S.-Y. (1993) ‘An Association of Our Own’, Newsletter of the International Association of Chinese Linguistics 1(1): 1–2. Wang, W. S.-Y. (ed) (1995) The Ancestry of the Chinese Language, Journal of Chinese Linguistics Monograph No.8. Chinese translation by Li Baojia 李葆嘉 (2005), Hanyu de Zhuxian《汉语的  祖先》, Beijing: Zhonghua shuju 中华书局.

181

William S.-Y. Wang Wang, W. S.-Y. (ed) (1991) Languages and Dialects of China, Journal of Chinese Linguistics, Monograph No. 4. Wang, W. S.-Y. (1998) ‘Three Windows on the Past’, in Victor H. Mair (ed) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, Pennsylvania: University of Pennsylvania Museum Publications, 508–34. Wang, W. S.-Y. (2013) Love and War in Ancient China: The Voices of Shijing, Hong Kong: City University of Hong Kong Press. Wang, W. S.-Y (forthcoming) A Billion Voices: Languages and Peoples of China, in press with World Scientific Press. Wang, W. S.-Y. and A. Lyovin (1970) CLIBOC: Chinese Linguistics Bibliography on Computer, Cambridge: Cambridge University Press Wang, W. S.-Y. and Cheng Chin-Chuan (1987) ‘Middle Chinese Tones in Modern Dialects’, in Robert Channon and Linda Shockey (eds) In Honor of Ilse Lehiste, Dordrecht: Walter de Gruyter, 513–23. Wang, W. S.-Y. and Yaching Tsai (2011) ‘The Alphabet and the Sinogram: Setting the Stage for a Look across Orthographies’, in P. McCardle, J. R. Lee, B. Miller and Ovid J. L. Tzeng (eds) Dyslexia Across Languages: Orthography and the Brain–Gene–Behavior Link, Brookes Publishing, 1–16. Wang, W. S.-Y. and Chaofen Sun (eds) (2015) Oxford Handbook of Chinese Linguistics, Oxford: Oxford University Press. Wu, Rukang and John W. Olsen (eds) (1985) Palaeoanthropology and Palaeolithic Archaeology in the People’s Republic of China, Orlando: Academic Press. Wu, Yifeng, (ed) (2009) Cumulative Indexes and Abstracts to Journal of Chinese Linguistics, Journal of Chinese Linguistics, Monograph No. 23. Xing, Janet Zhiqun (ed) (2009) Studies of Chinese Linguistics: Functional Approaches, Hong Kong: Hong Kong University Press. Xu, Dan (2006) Typological Change in Chinese Syntax, Oxford: Oxford University Press. Yan, Margaret Mian (2007) Introduction to Chinese Dialectology, LINCOM Studies in Asian Linguistics. Yang, Xianyi (1983) Selection from the Book of Songs. Beijing: Panda Books. Yip, Po-Ching (2000) The Chinese Lexicon: A Comprehensive Survey, London and New York: Routledge. Zhao, Tongmao and Lee Tsungdao (1989) ‘Gm and Km Allotypes in 74 Chinese Populations: A Hypothesis of the Origin of the Chinese Nation’, Human Genetics 83: 101–10. Zhang, Grace Qiao (2010) Using Chinese Synonyms, Cambridge: Cambridge University Press. Zhang, Hongming (1998) ‘Chinese Etyma for “River” ’, Journal of Chinese Linguistics 26: 1–47. Zhengzhang, Shangfang (1991) ‘Decipherment of Yue-Ren-Ge’, Cahiers Linguistiques d’Asie Oriental 20(2): 159–68. Zhou, Minglang (2003) Multilingualism in China: The Politics of Writing Reforms for Minority Languages, 1949–2002, Berlin: Mouton de Gruyter.

Chinese Atshogs, Yeshes Vodgsal 意西微萨•阿错 (2005)〈语言深度接触机制与藏汉语言类型差异问题〉(A Deep Contact Mechanism for Language and the Issue of Typological Differences of Tibetan Languages), Journal of Chinese Linguistics 33: 1–33. Chen, Baoya 陈保亚 (1996)《论语言接触与语言联盟》(On Language Contact and Language Alliance), Beijing: Language and Culture Press 语文出版社. Cheng, Chin-Chuan 郑锦全, Ting Pang-hsin 丁邦新, William S.-Y. Wang 王士元, and Mei Tsulin 梅祖麟 (2001)〈我们对中文音译的看法 – 请以汉语拼音为中文音译的唯一标准〉(Our Views on the Transliteration of Chinese – Please Take Hanyu Pinyin as the Only Standard for Chinese Transliteration), Journal of Chinese Linguistics 29: 165–7. Chou, Fa-Kao 周法高 (1980)《中国语言学》(Chinese Linguistics), Hong Kong: The Chinese University Press 中文大学出版社. Deng, Xiaohua 邓晓华 (1993)《人类文化语言学》(Human Cultural Linguistics), Xiamen: Xiamen University Press 厦门大学出版社. Hashimoto, Mantaro 桥本万太郎 (1977/1985)《语言地理类型学》(Linguistic Geographical Typology), Yu Zhihong 余志鸿 (trans.), Beijing: Peking University Press 北京大学出版社. Li, Jingzhong 李敬忠 (1994/2007)《语言演变论》(Language Change), Guangzhou: Guangzhou Press 广州出版社.

182

Chinese linguistics Lü, Shuxiang 吕叔湘 (1957)《中国文法要略》(Concise Chinese Grammar), Beijing: The Commercial Press 商务印书馆. Luo, Changpai 罗常培 (1950/1989)《语言与文化》(Language and Culture), Beijing: Language and Culture Press 语文出版社. Pan, Wuyun 潘悟云 (2000)《汉语历史音韵学》(Historical Chinese Phonology), Shanghai: Shanghai Education Press 上海教育出版社. Tan, Qixiang 谭其骧 (1982)《中国历史地图集》(Maps of Chinese History), Beijing: SinoMaps Press 地图出版社. Wang, Feng 汪锋 (2012)《语言接触与语言比较 – 以白语为例》(Language Contact and Language Comparisons – The Bai Language as an Example), Beijing: The Commercial Press 商务印书馆. Xu, Tongqiang 徐通锵 (1991)《历史语言学》(Historical Linguistics), Beijing: The Commercial Press 商务印书馆.   and Zhou, Zhenhe 周振鹤 and You Rujie 游汝杰 (1986)《方言与中国文化(修订本)》(Dialects Chinese Culture (Rev.)), Shanghai: Shanghai People’s Publishing House 上海人民出版社.

183

11 Chinese Linguistics: Pragmatics Yan Jiang school of oriental and african studies, University of london

1.  Short definition Pragmatics can be succinctly defined as the study of meaning in context. Although it does not have a long history as an independent academic subject of inquiry, the field has always been vibrant with diverse approaches, cross-disciplinary interactions, and several ingenious and highly sophisticated theoretical frameworks. In Chinese linguistics and applied language studies, pragmatics has also aroused great interest, all the more so because the term ‘pragmatics’ is translated into Chinese as yǔyòngxué 语用学, meaning ‘the study of language use’, which can often be shortened into yǔyòng 语用 (language use). Thus, pragmatics in Chinese has acquired a broader interpretation, encompassing all aspects of language use, making it more liable to be recognized as a useful subject of study by Chinese speakers, whether or not they have been exposed to the more rigorously defined pragmatic studies.

2.  Pragmatic phenomena Although there are several strands of pragmatics, due to page limit, this chapter is only focused on linguistic pragmatics, which claims the longest tradition and has always been engaged with central aspects of pragmatic inquiry. Linguistic pragmatics evolved from the school of Ordinary Language Philosophy as represented by the works of Anglo-analytic philosophers J. L. Austin (who is the founder of Speech Act Theory), P. F. Strawson (who initiated studies on presupposition), and H. P. Grice (founder of the theory of Logic in Conversation). It became a self-contained discipline at the beginning of the 1980s, heralded by the publication of two territory-defining textbooks: Levinson (1983) and Leech (1983). It differs from linguistic semantics in that it goes beyond the analyses of word meaning and sentence meaning, treated respectively by lexical and compositional semantics, and studies how sentence meaning is developed and even twisted in discourse. Its focal object of study is utterance meaning, which is roughly equivalent to propositional meaning read off a sentence in a given use context. The initial meaning of a sentence, composing out of the meaning of the words contained therein, when uttered in a context, may need to be explicated, through disambiguation and enrichment, to reach its full-fledged literal meaning, also called explicature. In addition, it may trigger the perception of some hidden and separate proposition taken for granted by 184

Chinese linguistics: Pragmatics

the speaker, formally called presupposition. Furthermore, it may implicate some extra, inferred propositions that the speaker tries to convey through the use of the original sentence – what is termed implicature. On top of all these layers of pragmatic meaning, a sentence always conveys a higher-order sense: what the speaker wants to do with the proposition or the set of propositions he produces, be it for making a statement, performing a speech act such as declaring, requesting, naming or marrying, making commentaries such as in a live football show, giving directions such as in an iPad manual or in a recipe, expressing irony in that what is stated is to be taken as contrary to what is intended to convey.  .  .  .  These additional tones variously attached to the propositions expressed form kinds of propositional attitude or higher-order explicature. Last but not least, some words and expressions can only be interpreted with reference to the physical properties in the situation of utterance, thereby revealing the very essential referential nature of natural language: that of using language to point to things right on the scene or in the vicinity, with demonstrative or symbolic gestures – what is called the deictic or indexical use of language. The notions introduced so far form a cluster of pragmatic phenomena that have received extensive studies in the literature. Each of them is illustrated below with two examples in Chinese. [1] méi xiǎng dào nǐ hái tǐng néng hē, yě tǐng néng liáo 没想到你还挺能喝,也挺能聊。 ‘I was surprised to find that you can drink a lot, and can talk a lot too’ [Explicature: Explicating the sentence so as to supplement what is left unsaid: what you drink is alcoholic, not just any soft drinks or water.] [2] shìzhě sòng shàng jiān hǎo de niúpái, wǒ fēn fù guò tā, jiān dé lǎo diǎn, qiē kāi shí,   lǐmiàn háishì hóng hóng de xuèsī 侍者送上煎好的牛排,我吩咐过他,煎得老点,切开时,里面还是红红的血丝。 ‘The waiter served our steaks. Although I had told him to cook the steaks well-done, when they were carved up, there were still threads of red blood’ [Explicature: The word lǎo (old) is enriched to give the special interpretation ‘well-done’ in this context.] [3] hěnshǎo yǒu rén zhīdào, fāmíng zìxíngchē de shì déguó de yī gè kànlínrén,   míng jiào déláisī 很少有人知道,发明自行车的是德国的一个看林人,名叫德莱斯。 ‘Not many people know that the man who invented the bicycle was a German forester called Drais’ [Presupposition: The de-construction fāmíng zìxíngchē de (发明自行车的) serves as a presupposition trigger showing that the speaker takes it for granted that the bicycle has been invented. This information is presupposed rather than asserted because nowhere in the sentence is the information ‘the bicycle has been invented’ explicitly stated. What is explicitly stated is only the identity of the inventor of the bicycle.] [4] zhèxiē kòngdì, yàoshì yònglái zhòng zhuāngjià, guójiā jiù bú yòng huā nàme duō   wàihuì jìnkǒu liángshi le 这些空地,要是用来种庄稼,国家就不用花那么多外汇进口粮食了。 ‘If these abandoned lands are used to plant crops, our country will no longer have to spend so much foreign currency to import grains’ 185

Yan Jiang

[Presupposition: ‘bú yòng huā nàme duō wàihuì’ presupposes the fact that a lot has already been spent or has at least been committed to spend on purchasing something.] [5] ‘wǒ xiǎng nǐ yěxǔ bú shì qù sànbù de,’ tā túláo wúyì de pángqiāocèjī shuō ‘我想你也许不是去散步的’, 他徒劳无益地旁敲侧击说。 ‘ “I thought maybe you didn’t [go for a walk],” he said, beating about the bush in the most useless manner.’ (from Sister Carrie, Chapter XXIII, by Theodore Dreiser.) [Implicature: the speaker hinted at some other propositions, e.g. that he suspected that she was seeing someone else. Neither in form nor in meaning does the intended implicature bear any similarities with the original sentence.] [6] gāngcái chīfàn shí jiàn nǐ méi chī shénme dōngxi, xiǎngbì shì wèikǒu bu hǎo 刚才吃饭时见你没吃什么东西,想必是胃口不好。 ‘At the dinner just now, I saw that you didn’t eat much. Thought you must have had a rather bad appetite’ [Implicature: Although ‘méi chī shénme dōngxi’ explicitly asserts that not much was eaten, it implicates that something was eaten, even though the quantity was minimal. This is because, according to the theory of scalar implicature, ‘shénme dōu méi chī’ (什么没都吃) and ‘méi chī shénme’ (没吃什么) can be said to form an ordered set: The item to the left is semantically stronger than the one to the right, conveying information that is higher in degree, larger in number, or more informative than the weaker one. The mentioning of the item to the right implicates the unavailability of the item to the left, as the speaker wants to be informative to the right extent.] [7] zī rènmìng X xiānsheng wéi jiǔdiàn shìchǎng xiāoshòubù zǒngjiān 兹任命X先生为酒店市场销售部总监。 ‘Mr. X is hereby appointed as the director of the Marketing Sales Department of the hotel’ [Propositional attitude: a speech act of appointing.] [8] fēi xiàng guòhé bù shǒu lǐ, shuāngjiǎo wù tà duìmiàn wèi 飞象过河不守礼,双脚勿踏对面位。 ‘(Hong Kong Bus notice) Literally: It is against the rule (in Chinese chess) to march your elephant across the boundary river into the opponent’s half. Paraphrase: Do not put your feet on the opposite seat.’ [Propositional attitude: an indirect speech act, making a request in the guise of a statement.] [9] jīntiān wǒmen zhè’ér xiàyǔ le, bù zhīdào háiyǒu nǎxiē dìfang xiàyǔ le 今天我们这儿下雨了,不知道还有哪些地方下雨了。 ‘Today it rained here at our place. Don’t know which other places also rain’ [Deictic use of jīntiān (today), wǒmen (we) and zhè’ér (here), whose exact referents can only be worked out with reference to the parameters of the utterance act, such as information about the speaker, the place and the time of the utterance, etc.] 186

Chinese linguistics: Pragmatics

One common feature shared by the terms introduced above is that such kinds of pragmatic meaning are inferred rather than completely encoded. That is, pragmatic meaning is never directly gleaned from the encoded meaning of linguistic expressions alone, and can only be obtained by making contextualized inferences, using the linguistic expressions as the starting point of total meaning construction. Hence what matters is not just the identification and analysis of pragmatic phenomena per se, but also the characterization of the general inferential mechanism involved. This calls for the need of an explanatory theory of pragmatics that can give an adequate account of the inferential mechanism, has the potential of giving unified analyses of all the pragmatic phenomena, and is sympathetic and compatible with recent developments in related areas and disciplines, notably cognitive science, psycholinguistics, neurolinguistics, clinical linguistics, natural language logic, artificial intelligence and information technology.

3.  Pragmatic theory There have been several sophisticated theories of pragmatics, each competing with the others in trying to achieve explanatory adequacy. Only one of them, relevance theory, is presented here in detail and applied to the analysis of cases in Chinese.1 Aiming at providing a cognitive account of utterance interpretation, relevance theory tries to formulate a model of pragmatic inference with psychological reality, i.e. true to what people do in real-life communication. The theory starts with the underdeterminacy thesis, arguing that every sentence needs to have its encoded meaning developed through inference to reach the more complete meaning. Meaning construction in this sense is carried out in the cognitive context of the hearer, formed by the new assumptions in the utterance he receives and the old assumptions he retrieves from his memory in order to process the new ones. When old and new assumptions interact, cognitive effects can be generated, in the form of deductive conclusions or strengthened assumptions or updated ones that contradict and replace some older ones. But the cognitive context can be a vast pool of assumptions in which old assumptions can freely interact with the new ones, and the resulting cognitive effects will initiate even newer interactions, ad infinitum. The hearer has no time nor resources to process and evaluate every conceivable cognitive effect before deciding on the one he takes to be the intended meaning conveyed by the speaker. What he can do is to pick one cognitive effect he can get without exerting too much cognitive effort and take it to be the one intended by the speaker. This will not sound so arbitrary when several related factors are taken into consideration. First, the multitude of old assumptions in the cognitive context are not scattered in disarray, but are ranked and stacked, according to their different degrees of accessibility in the memory. As old assumptions are retrieved for the purpose of processing the new assumption, the most readily retrieved must be the most accessible, which is most likely to interact with the new assumptions, and most likely to yield enough cognitive effects without costing undue processing effort. Such an interpretation of the utterance is termed an optimally relevant interpretation. Second, the hearer is not working all by himself. He is also aided by the speaker in an unconscious, yet intuitively quasi-cooperative way. The speaker knows that, in order to get her meaning across, there is an optimally relevant way to express her thoughts, one that matches with the somewhat individualistic knowledge state of the hearer to the extent that this is made known to her. That should be the manner of expression she unconsciously chooses to employ. Third, according to relevance theory, communication is by nature a guess work, miscommunication does occur frequently, especially between parties 187

Yan Jiang

who do not make pertinent judgments of each other’s knowledge state. When communication occurs, the hearer can only presume that the speaker has something to inform that will let him derive enough cognitive effects without exerting unjustified extra processing effort. This is called the presumption of optimal relevance. The speaker chooses her wording with the knowledge that the hearer uses the presumption of optimal relevance as an expectation in communication. Hence her wording is likely to help the hearer to infer her meaning, and she can have some degree of confidence that her utterance will be properly comprehended. And the hearer will take his first interpretation as the intended meaning of the speaker, without bothering to entertain alternative interpretations. This intersubjective production and comprehension process makes it possible for communication to succeed with much efficiency, even though it does not guarantee success, as a presumption is the weakest assumption and often turns out not to be failsafe, especially when the communicative parties are not able to make the right judgments of each other’s knowledge state due to unfamiliarity, lapse of memory, or physical or mental exhaustion. Relevance theory summarizes the above considerations into the Principle of Relevance: Every utterance comes with a presumption of its own optimal relevance. Utterance comprehension can now be viewed as a cognitive act in search of relevance. The underdetermined meaning of an utterance is developed to the extent that optimal relevance is attained, which is also the first relevant interpretation that the hearer can read off the processed utterance. Hence comprehension, and communication in general, appear to be instantaneous and effortless. The relevance-theoretic comprehension mechanism can provide a programmatic characterization to the whole range of pragmatic phenomena introduced above. It also has the potential to give accounts to specific cases, with each account being a detailed case-based story. Overall, such a pragmatics starts with the encoded meaning of the utterance, taking the lexical and grammatical meanings as providing conceptual or procedural information. The former initiates other relevant assumptions, the latter instructs the hearer on how to establish relevance: whether to trigger a presupposition or to derive an implicature or to take the proposition with a special attitude or to obtain a specific deictic referent. Assumption introduction and deduction is guided by procedural cues, and is constrained by the principle of relevance, making it possible for pragmatic phenomena to be comprehended in no time, in spite of the underdeterminacy of explicitly encoded meaning.2 Relevance theory is particularly helpful in that it can be used to investigate topics that are not wholly linguistic in nature. In such cross-disciplinary studies, the theory can often provide a novel explanatory perspective. The next section examines a range of pragmatic cases in Chinese, showing how relevance theory can provide original and specific accounts as well as demonstrating how the study of Chinese pragmatics can make its unique contributions to linguistic pragmatics in both empirical and theoretical aspects.

4.  Case Study One: the presupposition of the adverb bái (白) that resembles the principle of relevance Textbook examples of presupposition triggers in English often appear straightforward and obvious, and textbook discussions on the topic often seem to create the impression that all the known types of presupposition triggers can be found in all languages. In reality, each language has very unique ways to encode its triggers. In Chinese, quite a number of them trigger complicated presuppositions and cannot be literally translated into English without 188

Chinese linguistics: Pragmatics

losing their triggering power. Among these, the adverb bái (白) [in vain/for free] provides an interesting case for discussion.3 Predicates modified by ‘bái’ yield two kinds of meaning: (i) doing something gratuitously, as shown in (10)–(12), and (ii) getting something (done) for free, as shown in (13)–(15): [10] wǒ gēge bái gàn yī nián, yī gè zǐér yě méi lào zhào, hái āi le yī dùn chòumà 我哥哥白干一年,一个子儿也没落着,还挨了一顿臭骂。 ‘My elder brother worked hard in vain for a whole year without getting a single penny. Instead, he got a heap of abuse.’ [11] zhēn xiǎng zhīdào wèishénme gàn le yībèizi, què méiyǒu le tuìxiūjīn, sānshí duō nián   dōu bái gàn le? qián dào nálǐ qù le nē? 真想知道为什么干了一辈子,却没有了退休金,三十多年都白干了？钱到哪里 去了呢？ ‘I really want to know why I worked for all my life, but ended up without the pension. Does more than 30 years of work come to nothing? Where did the money go?’ [12] chū gèrén zhuānjí shì hěn róngyì de shìqíng, dàn jīngbànzhě shì fēi yào zuàn nǐ yī   bǐ de, tā jué búhuì gěi nǐ bái gàn 出个人专集是很容易的事情,但经办者是非要赚你一笔的,他决不会给你白干。 ‘It is easy to publish a collection of your own works. But the publishing agent will certainly make you pay a lot for it. There is no way for him to work for you for nothing’ [13] xǐhuān xiàndài tōngsú yīnyuè de qīngniánrén shòubúliǎo xìqǔ de màn jiēzòu yǔ tīng   bú dǒng de chàngcí hé dàobái, qǐng tā bái kàn xì tā yě wèibì zuò dé zhù 喜欢现代通俗音乐的青年人受不了戏曲的慢节奏与听不懂的唱词和道白, 请他白看戏他也未必坐得住。 ‘Those young people who prefer modern pop music to traditional Chinese opera cannot appreciate the latter’s slow tempo and its incomprehensible song-lines and spoken parts. Even if they were treated to the performance for free, they would probably still fail to stay on in the theatre’ [14] miǎndiàn duì xxx guānjiàn shíkè zhǐyào zǒu chū yībù tuì mǎ de qí, shuāngfāng   jiù huì lìkè héqí; dàn xxx cǐkè jǐnzhāng zhī xià wù zǒuchū yībù jìn jiāng,   bèi yy bái dǎ yī pào 缅甸队 xxx关键时刻只要走出一步退马的棋,双方就会立刻和棋; 但xxx 此刻紧张 之下误走出一步进将,被yy白打一炮。 ‘At this critical juncture, Burmese player xxx only needed to make the move of retreating the horse, and the game would have ended with a draw. But xxx was so nervous that he advanced the general by mistake, letting player yy take his cannon for free’ [15] jiǎrú wǒ xiànzài è le, kěyǐ dào diǎnxīnpù bái ná xiē bōbō mā? 假如我现在饿了,可以到点心铺白拿些饽饽吗？ ‘If I were hungry now, could I go to the bakery and get some cakes for free?’ 189

Yan Jiang

This apparent ambiguity of ‘bái’ between the two opposite meanings of ‘loss’ and ‘gain’ is found to be two sides of the same coin: what one side loses, the other side gains, depending on the speaker’s perspectives. When the speaker talks about the subject as the receiver of a transaction, the ‘gain’ sense is conveyed; when the subject is viewed as the giver of a transaction, the ‘loss’ sense is conveyed. The kind of transaction presented by ‘bái’ is special in the sense that the presumed equilibrium between gains and losses is not reached. What ‘bái’ literally contributes as its lexical meaning is a sense of imbalance, not a negation on the transaction itself, but a claim that the transaction is not conducted according to a presupposition it triggers: that normal transactions are expected to be balanced in gains and losses. Negation of ‘bái’ will create a contrary effect, the maintenance or reinstatement of the balance-expecting presupposition. Looking at a broader range of examples involving ‘bái’, it can be concluded that this balance-expecting presupposition does not only apply to transactions involving financial gains and losses measured in terms of exertion of strength and labour. Criteria of evaluation on the presupposed balance may vary a lot, from the very concrete exemplified by [10]–[15] to the very subjective and abstract, as shown in [16]–[23]: [16] ‘bàn tiáo yú,’ tā shuō. ‘nǐ yuánlái shì tiáo wánzhěng de. wǒ hěn bàoqiàn, wǒ chūhǎi   tài yuǎn le. wǒ bǎ nǐ wǒ dōu huǐ le. búguò wǒmen shāsǐ le bù shǎo shāyú, nǐ gēn   wǒ yīqǐ, hái dǎshāng le hǎoduō tiáo. ni shāsǐ guò duōshǎo ā, hǎo yú? ni tóu   shàng zhǎng zhe nà zhī chángzuǐ, kě búshì bái zhǎng de ā.’ ‘半条鱼,’ 他说。‘你原来是条完整的。我很抱歉,我出海太远了。我把你我都 毁了。不过我们杀死了不少鲨鱼,你跟我一起,还打伤了好多条。你杀死过 多少啊,好鱼？你头上长着那只长嘴,可不是白长的啊。’ ‘ “Half fish,” he said. “Fish that you were. I am sorry that I went too far out. I ruined us both. But we have killed many sharks, you and I, and ruined many others. How many did you ever kill, old fish? You do not have that spear on your head for nothing.” ’ (Ernest Hemingway, The Old Man and the Sea) [17] gēn wǒ hǎo zěnme le? chīkuī mā? nándào wǒ jiù bái zhǎng zhème shuài le?   shuài jiùshì běnqián! 跟我好怎么了？吃亏吗？难道我就白长这么帅了？帅就是本钱！ ‘What’s wrong with dating me? Do you think I am not worthy enough? Do I own this handsome face for nothing? Being handsome is my investment capital!’ [18] nà zhǎng dé piàoliàng búshì bái zhǎng dé piàoliàng de, yào zhíqián de. 那长得漂亮不是白长得漂亮的,要值钱的。 ‘Being pretty is not for nothing. It should be rewarding’ [19] sūn jiàndōng de nèixīn yīrán bǐbó yè táo: diēmā xīnkǔ gòng tā niàn le shíwǔ   nián shū, què dé bú zháo tā yī fēn qián de hǎochù, ……bái zhǎng le yīfù   hǎo pínáng. 孙建冬的内心依然鄙薄叶陶: 爹妈辛苦供他念了十五年书,却得不着他一分钱的 好处,……白长了一副好皮囊。 ‘Inside his heart, Sun Jiandong still looked down on Ye Tao: his parents paid for his tuition for 15 years, yet did not get even one cent from him. Although Ye Tao had a handsome face, it was pretty useless’ 190

Chinese linguistics: Pragmatics

[20] wǒmen péiyǎng chūlái de réncái, dōu bú wèi zìjǐ de guójiā fúwù, wǒmen bú shì bái   péiyǎng le mā? wǒmen de tóunǎo nándào bù yīnggāi qīngxǐng yīxiē mā? 我们培养出来的人才,都不为自己的国家服务,我们不是白培养了吗？我们的 头脑难道不应该清醒一些吗？ ‘Of the talents we trained, no one served for our own country. Didn’t we train them gratuitously? Shouldn’t we be more sober on this point?’ [21] huí dào cèzhuāng yǐjīng kāishǐ màishōu le, shuǐkù píruǎn wúlì, zuìhòu de yīdiǎndiǎn   nítāng shuǐ kěliánbābā, bái kàn zhe hēicūn de rénmen yòu fēngshōu le yīhuí 回到册庄已经开始麦收了,水库疲软无力,最后的一点点泥汤水可怜巴巴, 白看着黑村的人们又丰收了一回。 ‘Back to Ce Village, wheat was harvesting. The reservoir did not have enough water. The last bits of muddy water looked miserable. People watched in vain, seeing that the Hei Village farmers had a bumper harvest again, without themselves being able to do the same.’ [22] dàjiā dōu kàn zhèngzhì, bù guǎn xuéshí, jíshǐ yǒu jiùguó de zhēnxīn, érqiě nádào   zhèngquán, yě shì shìdàolíntóu bái dèngyǎn! méiyǒu yīngfù de nénglì yǔ zhīshí 大家都看政治,不管学识,即使有救国的真心,而且拿到政权,也是事到临头 白瞪眼！没有应付的能力与知识。 ‘If we all just study politics and do not care for knowledge, even if we are genuinely dedicated to the salvation of our country and win the power, we still can only stare and do nothing in the end! That is because we will not have the ability and knowledge to run the country.’ [23] zhègè fāxiàn jiājù le wǒ de kǒngjù, yě chèdǐ dǎxiāo le wǒ dúzì yī rén xiāomiè yāoguài   de xióngxīn. shuí dōu zhīdào yīgè rén zhīshēn hé shīzǐ bódòu nà jiào bái gěi 这个发现加剧了我的恐惧,也彻底打消了我独自一人消灭妖怪的雄心。谁都知道 一个人只身和狮子搏斗那叫白给。 ‘This discovery intensified my horror, and completely took away my ambition of dealing with the monster single-handedly. Everyone knows that a single man is no match for a lion, if he wishes to grapple with it.’ In [16], the old man was prescribing his conception of the use of the spear mouth of the big fish, which was to help him kill sharks so that he could keep the big fish as his prey. If this usage was not put into action, the spear was considered to be gratuitously grown. In [17]–[19], a commonly conceived view was presupposed: good-looks should be rewarding, and rewarding in an unspecified sense, since the other end of the balancing scale is not explicated. [20] again presupposes a vague balancing scale: students who received free education are expected to render service to the country as a form of repayment. If not, the teachers will feel that their teaching work is not adequately rewarded, so are the officials who work for the student-aid system or for the education institutes in general. But the real balance is between the students who benefit from the free education and the country that provides the service. [21] presents a case in which people in village A watch people in village B harvesting, but A were not able to do the same. Hence they could only watch in vain. [22] talks about people not being able to rise to the occasion, but can only look around in vain, with no beneficiary side at all. In both [21] and [22], only one end of the balancing scale is clear, the other side is not directly involved. Finally, the ‘bái gěi’ (give in vain) in 191

Yan Jiang

[23] means being no match to the opponent, letting him take the upper hand without being able to put up a resistance. The emphasis is on ‘no match’ rather than on the balancing scale and its two ends. The above considerations lead to some modifications of the content of the presupposition triggered by ‘bái’, that is: events or processes, whether involving humans or not, are expected by humans to preserve a balance of energy flow, which may transform itself into some other forms in the process, so long as the energy is preserved in some way, i.e. labour at one end being compensated by money at the other end, good-looking having the potential of being rewarded by marrying a wealthy spouse, dedicated research being credited with good research indices, etc. On the other hand, since the criteria are wholly subjective, there is no guarantee for proper balance. The presupposition is only to expect that energy flow is preserved to the extent that the speaker does not feel the need to use ‘bái’, or he feels justified to negate ‘bái’ in his utterance. Otherwise, the balance is upset, and ‘bái’ can be used to describe the event. With this understanding, we can enrich the presupposition of ‘bái’ in the following way: Events or processes, whether involving humans or not, are expected by humans to preserve a balance of energy flow, which may transform itself into some other forms in the process, so long as the energy is preserved in some way, to the extent that the balance is still considered to hold. Upsetting the balance at either end will lead to the likelihood of the use of ‘bái’, and awareness of the maintenance of the balance provides the reason for the use of the negation of ‘bái’. The above presupposition, henceforth shortened to ‘bái-presupposition’, is probably a rather rare presupposition because it addresses the general tendency of human cognition. Put in an even stronger way, it is an embodiment of human cognitive principle, which explains most human endeavours throughout the whole human history and even the whole process of human evolution. It then comes as a no surprise that ‘bái-presupposition’ bears some resemblance to the communicative principle of relevance introduced above, which also professes to be a principle of human cognition. Applying ‘bái-presupposition’ to the case of utterance comprehension, the hearer expects that his comprehension effort will not come to nothing. That is tantamount to saying that he expects the utterance to give him some cognitive effects so that his cognitive comprehension effort will not be of no avail. And the speaker is also aware of the hearer’s expectation because her cognitive actions are also guided by the principle of relevance. So she will justify her verbal behavior by producing an utterance that will give the hearer some cognitive effects. Otherwise, even her own cognitive effort will be exerted gratuitously. In verbal communication, therefore, there is an expectation of cognitive effort preservation, to some extent, being a special case of the ‘báipresupposition’. As the concept of ‘relevance’ in relevance theory is a specially defined term, using the notions of cognitive effect and effort, which is different from the usual sense of ‘relevance’ in ordinary English language, we can envisage the possibility of using ‘bái’ to redefine the principle of relevance, making it sound more natural, at least in Chinese. Since the presupposition can be overtly revealed and maintained by using ‘bái’ with negation, we come up with the following definition: [24] The Not-bái Principle of Communication: Speaker: bú bái shuō [不白说] ‘Do not talk with no effects’ Hearer: bú bái fèi jīnglì [不白费精力] ‘Do not exert mental effort in vain’ 192

Chinese linguistics: Pragmatics

[24] follows from the more general [25], which could have been the shorter definition of the ‘bái-presupposition’, yet to do that is to make the definition circular, as it contains ‘bái’, which needs to be defined itself: [25] bú bái fèi lì [不白费力] ‘Do not exert energy in vain’ This coincidental resemblance of ‘bái-presupposition’ to the communicative principle of relevance comes as a happy surprise. But at a more deliberated level, it comes as no surprise at all. Both address one fundamental aspect of human cognition. If there are laws of human cognition to be found, relevance theory discovered the principle of relevance, and we find in ‘bái-presupposition’ a natural language embodiment of the principle in the form of a presupposition. Such a resemblance is not an impossibility.

5.  Case Study Two: the rhetorical figure ‘lièjˇın’ 列锦格 (list gem) as a touchstone in the debate between Contextualism and Semantic Minimalism As a figure of speech christened only for 30-odd years,4 list gem refers to an idiosyncratic structure in Chinese, which consists of several semantically unrelated NPs (single nouns or noun phrases) lined up in a row as one sentence, with no verb predicates nor adjectives or nouns used as predicates. The general understanding is that the figure of list gem should contain three NPs as its norm. A sentence containing only two NPs should best be categorized as couplet, as two single NPs forming a couplet is common in Chinese poetry. Besides, the interpretation of couplet differs a lot from list gem, as a couplet either forms an antithesis or expresses a sequence. On the other hand, in most cases, a sentence containing four or more stand-alone NPs is to be categorized as enumeration or escalation, which means the NPs contained therein are also related in meaning, again being very different from list gem. Classic examples of list gem are found in classical poetry, as shown by [26]–[27]. [26]

kūténg lǎoshù hūnyā xiǎoqiáo liúshuǐ rénjiā gǔdào xīfēng shòumǎ xīyáng xī xià duàncháng rén zài tiānyá



‘Dry-vine old-tree blind-crow, Small-bridge flowing-water human-home Old-road west-wind lean-horse Setting-sun falls west Heart-broken man at end of the world’

[27]

gūcūn luòrì cánxiá qīngyān lǎoshù hányā yī diǎn fēi hóng yǐng xià qīng shān lǜ shuǐ bái cǎo hóng yè huáng huā



‘Lone-village falling-sun patch-twilight, Thin-smoke old-tree jack-daw One spot of shadow cast down by the flying swan-goose Green-mountain olive-water White-grass red-leaf yellow-flower’

枯藤老树昏鸦, 小桥流水人家。 古道西风瘦马。 夕阳西下, 断肠人在天涯。

孤村落日残霞, 轻烟老树寒鸦, 一点飞鸿影下。 青山绿水, 白草红叶黄花。

193

[list gem 1] [list gem 2] [list gem 3] [subject predicate construction] [existential construction]5

[list gem 1] [list gem 2] [complex description] [couplet] [list gem 3]6

Yan Jiang

In [26], the first three lines are three cases of list gem instead of one, as they form three distinct meaning clusters. The remaining two lines do not contain list gems. Likewise in [27], even though only NPs can be found, not all lines are composed of list gems. Each of [28]–[31] are two-line couplets taken from whole poems, but each line in each couplet constitutes a list gem, which introduces further complications to the distinction between couplets and list gems. [28] jī shēng máo diàn yuè, rén jì bǎn qiáo shuāng 鸡声茅店月,人迹板桥霜。[two list gems] ‘Rooster-sound small-inn moon/Foot-print wood-bridge frost’ [29] táolǐ chūnfēng yībēi jiǔ, jiānghú yèyǔ shínián dēng 桃李春风一杯酒,江湖夜雨十年灯。[two list gems] ‘Peach-plum spring-wind a-cup-of-wine/River-lake night-rain ten-year-light’ [30] sān shí gōng míng chén yǔ tǔ, bā qiān lǐ lù yún hé yuè 三十功名尘与土,八千里路云和月。[two list gems] ‘Thirty-feat-fame dust and mud/Eight-thousand-li-road cloud and month’ [31] wèn rǔ píng shēng shì yè, huáng zhōu huì zhōu dān zhōu 问汝平生事业,黄州惠州儋州。[list gem in the second line] ‘Ask you life achievement/Huáng-Town Huì-Town Dān-Town’ The interpretation of list gems could be an interesting topic in cognitive–pragmatic studies of rhetoric, as there do not seem to exist similar structures in languages that have been well exposed to pragmatic studies. Since a list gem forms a sentence but does not have a predicate, it seems necessary to accommodate one through explicating, so as to obtain the complete explicature with a saturated argument–predicate structure. But the three NPs in the list gem are no more than keywords. Each one may be accommodated with a different predicate. And the thorny issue is that different language users may supply different predicates to the same NP, to the extent that they are confident enough in obtaining the explicature of the sentence. Even the poet himself may find it hard to supply the exact predicate, for there simply isn’t one that can capture the rich meaning left open by the non-presence of any predicates, which is the hallmark of implicit meaning. Perhaps the best an interpreter can achieve is some impressions created by the line-up of the three stand-alone NPs. He can try to make inferences to establish the links among the three, but he is always ill at ease in deciding on the one single literal meaning for the sentence. At the same time, he can obtain a rich array of weak implicatures, appreciating fully the poetic effects of list gem. The comprehension process is unavoidably guided by the principle of relevance. The interpreter understands list gem in the optimally relevant way, resulting in a very unclear explicature and a set of weak implicatures. But to the mind of an individual interpreter, some weak implicatures may be even clearer than the explicature. One important theoretical bearing of this case has to do with the recent debate between Contextualism and Semantic Minimalism in philosophy of language.7 Contextualists hold the view that the exact truth value cannot be determined at the sentence level, because meaning directly gleaned from the logical form of any sentence is unavoidably underdetermined and cannot be recovered without the use of free pragmatic inference. Hence truth-conditional semantics cannot be self-sufficient, as pragmatic processes always intrude into the determination 194

Chinese linguistics: Pragmatics

of literal meaning of a sentence, without which the truth value can never be ascertained. Truth can only be established at the level of truth-conditional pragmatics. On the other hand, Semantic Minimalists contend that truth-conditional semantics can still be self-contained and adequate. Although meaning is underdetermined at sentence level, the logical form of the sentence can be reanalyzed so that motivated indexicals can be posited. Thus all pragmatic processes aiming to derive perfect literal meaning are activated by indices, which are repre­ sented as free variables in the logical form and instantiated through pragmatic inference. Truth-conditional semantics is thus salvaged because semantic interpretation refers to all the items in the logical form, including interpretive rules for all the indices. It is the indices that refer outward to contextual information. Pragmatic intrusion is fended off at sentence level. Arguments contributing to this debate come in two kinds. From a theoretical point of view, it is important to propose claims and mechanisms that are strong in explanatory power. From an empirical point of view, it is also important that either side can take into consideration a wide range of structures and cases, including those in different languages that prove to be challenging to existing accounts. In such a testing ground, Contextualism and Semantic minimalism can compete with each other, and the results can best reveal the advantages or inadequacies of the either side. List gem appears to provide a strong case in support of Contextualism, and a knock-down case against Semantic Minimalism, as it is impossible to determine the explicature of list gem sentences: language users don’t bother to flesh out the complete meaning of list gems. If pressed, one may supply an ad hoc explicature, which is most likely to be very different from the one constructed for the second time when he is asked to explicate the very same sentence. When two people are asked to explicate the same list gem, they always differ a lot from each other in what each can produce. Hence, classic list gem examples never have standard explicatures. Contextualists can therefore use list gem to show that the truthconditional content cannot be established at sentence level, as there is no commonly accepted way to explicate a list gem. However, Semantic Minimalists can find an escape hatch by arguing that list gems do not have explicatures at all: the whole point of using list gems is to let them create impressions, which lead to the derivation of weak implicatures. In list gems, we find a rare case of language use which has a lot of implicatures but no agreed explicatures.

6.  Case Study Three: the reach of counterfactual meaning in Chinese – a unique logic? How counterfactual conditionals are expressed in Mandarin Chinese (and in Chinese dialects) has remained a mystery since the advent of modern linguistics.8 Unlike Indo-European languages which almost always use verb inflexion to mark counterfactuals, Chinese lacks explicit morphology to distinguish counterfactual conditionals from indicative ones. Although there are a few lexicalized constructions such as ‘yàobúshì’ [要不是] (if-not-be) and ‘zǎozhīdào’ [早知道] (early-know) whose presence necessarily lead to counterfactual interpretations, there are many others that are interpreted counterfactually in contexts of utterance that do not make any use of these constructions. As native speakers of Chinese do perceive counter­ factual conditionals and produce them as well, the challenge is to provide an adequate mechanism that can characterize the reach of counterfactual meaning in Chinese. Right now, there are two approaches. According to the first approach, as Chinese counterfactuals are largely unmarked, Chinese conditionals are either indicative or counterfactual, depending on the information gathered from the context. It is the contextual information that provides the 195

Yan Jiang

interpretive guidance. If the antecedent of a conditional makes a contrary-to-fact hypothesis in content (though not in form), counterfactual meaning is reached through pragmatic inference.9 This thesis can lead to the claim that every conditional is ambiguous between indicative and counterfactual. But native speakers do not seem to perceive conditionals in this way. At least, more specific claims are needed to explain how context can determine counterfactuality for a conditional. Another approach argues that some words and particles in Chinese serve as markers or weak features of counterfactuality.10 Although each of them may be too weak to make it happen, some such features can work together in some subtle ways to bring about counterfactuality. Such an approach also needs more clarifications. Weak features seem to be all optional in nature. Although sometimes, their presence makes the difference and counterfactuality is reached, some other times, similar counterfactual meaning can be reached without any of the weak features being present. Again, it is not clear how the mechanism works. The account we propose is different from either of the above. We start by examining the semantics of the explicit counterfactuals, i.e. the ‘yàobúshì’ [要不是] (If-not-be) conditionals, followed by a comparison between ‘yàobúshì’ and ‘me51hau3’ [蛮好] (Much-preferred) in Shanghainese, reaching the conclusion that they are by nature falsifying and truthifying counterfactuals. Then we try to generalize the findings to the majority of Chinese counterfactuals, claiming that they can be treated on a par with the ‘yàobúshì’ and ‘me51hau3’ counterfactuals. Some examples involving ‘yàobúshì’ are given below. [32] dāngchū, wǒ cóng diànyǐngyuàn cízhí láidào běijīng de shíhòu, jiālǐrén jí tóngshì,   péngyǒu jiù jílì fǎnduì. yàobúshì wǒ yìng xià yītiáo xīn, gēnběn jiù cí bú liǎo zhí,   gèng lái bú liǎo běijīng. (当初,我从电影院辞职来到北京的时候,家里人及同事、朋友就极力反对。要不 是我硬下一条心,根本就辞不了职,更来不了北京。) ‘Initially, when I quit my cinema job to come to Beijing, my family members, colleagues and friends were all dead against it. Had I not made up my mind, I would not have been able to resign, nor would I have been able to come and make it in Beijing.’ [33] 1:0, zhōngguó zúqiúduì yíng de gòu xuán de, bǐsài jìnxíng de yě gòu xié de.   yàobúshì Hǎo Hǎidōng nà yǒudiǎn yùnqì de jìnqiú, píngjú kěndìng huì   zhāo lái bùmǎn hé fēiyì. suǒyǐ, shuǎng bú qǐlái. (1:0, 中国足球队赢得够悬的,比赛进行得也够邪的。要不是郝海东那有点运气 的进球,平局肯定会招来不满和非议。所以,爽不起来。) ‘One: Nil. The Chinese football team won a very close match with real good luck. Had it not been for Hao Haidong to score that lucky goal, a tie would surely have incurred dissatisfaction and complaints. That is why nobody took it with a light heart.’ [34] xīlà dàibiǎotuán suī shì zuìhòu yīgè rùchǎng, dàn huòdé de zhǎngshēng shì zuì jiǔ,   zuì rèliè de. zhè bú qíguài, yīnwèi tāmen shì dōngdàozhǔ, guānzhòng méiyǒu   lǐyóu lìnxī zìjǐ de zhǎngshēng. zhídé yītí de shì, xīlà dàibiǎotuán cóng zhǔnbèi   rùchǎng dào ràochǎng yīzhōu duō hòu, gòng dédào le 18 fēn 2 miǎo zhōng de   zhǎngshēng, yàobúshì zhǔchírén dǎduàn, hěn nánshuō gǔzhǎng néng chíxù dào   shēnme shíhòu. 196

Chinese linguistics: Pragmatics

(希腊代表团虽是最后一个入场,但获得的掌声是最久、最热烈的。这不奇怪, 因为他们是东道主,观众没有理由吝惜自己的掌声。值得一提的是,希腊代表 团从准备入场到绕场一周多后,共得到了18分2秒钟的掌声,要不是主持人 打断,很难说鼓掌能持续到什么时候。) ‘Although the Greek delegation were the last to enter the stadium, they won the longest and loudest round of applause. This is hardly surprising, as they represented the host country and the spectators had no reason to save on their applauses. It is worth pointing out that the Greek delegation won a round of applause lasting for 18 minutes and 2 seconds, from the time they prepared to enter the stadium to the end of one lap’s march. Had it not been for the ceremony presenter who put the cheers to an end, it would have been hard to tell how long the applause would last.’ In the above examples, ‘yàobúshì’, though internally complex, behaves exactly like a conditional functor. It is used as a single lexical item, different from the compositional meaning of ‘yàoshì  .  .  .  bú’/if  .  .  .  not/‘要是  .  .  .  不’. ‘yàoshì  .  .  .  bú’ is not an explicit marker of counterfactual conditionals, because it can also form indicative conditionals. ‘yàoshì  .  .  .  bú’ resembles ‘if  .  .  .  not’ in English, but ‘yàobúshì’ is similar to the English ‘had it not been for  .  .  .’ in meaning. ‘Yàobúshì’ is internally complex for a special reason. It takes on a proposition P, which should be about a state or event that is both true and real, and returns a protasis P* which is counterfactual. Here, negation obviously plays a vital role, but negation is only a necessary condition, not a sufficient one. In the ‘yàoshì  .  .  .  bú’ construction, which also forms a protasis, negation works as well, but the resulting protasis does not have to be contrary to fact. What is more, the proposition led by ‘yàobúshì’ should not be some proposition about abstract, general, timeless states. This can be established through both introspection and corpus search. Try to create a ‘yàobúshì’ counterfactual with abstract ideas, the resulting conditional is bound to be weird. What also tend to be anomalous are ‘yàobúshì’ counterfactuals containing a proposition with an extra layer of negation. To say the least, they are difficult to process. To sharpen the above characterizations of ‘yàobúshì’, we look for more precise theoretical concepts in linguistic and philosophical studies. One viable notion to adopt is (subjective) veridicality, whose definition can be found in Giannakidou and Mari (2014): ‘truth judgments depending on what epistemic agents know or believe to be true, and other factors in the context relating to the epistemic status of individuals’. Aided by this new definition, ‘yàobúshì’ can be characterized as an operator taking a veridical proposition as its argument: what the agent knows or believes to be true. We can also call ‘yàobúshì’ a veridical operator. A similar operator in Mandarin is ‘zǎozhīdào  .  .  .’ (早知道) /Early know  .  .  .   /, which, however, is not a conditional connective, as it can follow ‘rúguǒ’ or ‘yàoshì’: [35] Tā xiàng xiāoxié tóngzhì kūsù shuō: ‘yàoshì zǎo zhīdào shàngdàng shòupiàn,   wǒ jiù bú gàn le. xiànzài nòng dé yīshēn máfan, xiōngdì jiěmèi dōu bù xiāng rèn,   hái jīyā le yīdàduī méi tuīxiāo chūqù de chǎnpǐn, wǒ qù zhǎo shuí a?’ (她向消协同志哭诉说: ‘要是早知道上当受骗,我就不干了。现在弄得一身麻烦, 兄弟姐妹都不相认,还积压了一大堆没推销出去的产品,我去找谁啊？’) ‘She told staff from Consumer’s Association in tears, “Had I known it was a fraud, I would not have got involved. Now I am all troubles. My brothers and sisters have all turned away from me. And I am stuck with loads of unsold goods. Who can I turn for help?” ’ 197

Yan Jiang

Subjective veridicality constitutes one necessary condition for the use of ‘yàobúshì’, but it still does not capture the intuition that ‘yàobúshì’ does not take on abstract propositions like science or math laws, even though they can be subjectively established as truth. In Armstrong (2002), non-abstract counterfactuals are given the name contingent counterfactuals: ‘Contingent counterfactual claims are often to be found in ordinary discourse, for instance, “If you had not put your foot on the brake so promptly just then, there would have been a nasty accident.” ’ Talks of historical contingencies are also frequently encountered: ‘What if there had been no American War of Independence? What if Ireland had never been divided? What if Britain had stayed out of the First World War? What if Hitler had invaded Britain or had defeated the Soviet Union? What if the Russians had won the Cold War? What if Kennedy had lived? What if there had been no Gorbachev?’ (Ferguson 1999). Another useful source is Rescher (2007), in which he distinguishes between falsifying and truthifying causal counterfactuals. Falsifying counterfactuals hypothesize what actually did happen had not happened, whereas truthifying counterfactuals hypothesize what did not actually happen had happened. According to Rescher, historical counterfactuals of the falsifying type ‘are in general retrospectively cause-determinative in nature’ and ‘generally address the preconditions for an actual occurrence’. Such counterfactuals are less speculative and more situation-bound. This rightly fits the characterization of ‘yàobúshì’ counterfactuals. So what follows ‘yàobúshì’ can now be re-characterized as a proposition which is both veridical and a falsifying contingent. And Armstrong’s term of contingent counterfactuals can now be elucidated as contingency-referenced falsifying counterfactuals. The whole protasis led by ‘yàobúshì’ can therefore be taken as an encoded falsifying counterfactual context. On the other hand, some structures can be encoded as counterfactual, but allowing either falsifying or truthifying possibilities. Jiang and Wang (forthcoming) characterize the behavior of the Shanghainese ‘me51hau3’ (蛮好……) /Much-preferred  .  .  .   / as a counter­ factual desiderative, which can give a truthifying reading: [36]

Mehau ganggang cen ditik qi nao!11 Much-preferred just-now take underground go EM-SMP 蛮好刚刚乘地铁启孬！ (EM-SMP = emotive sentence-final particle) ‘It would have been much better that we took the underground at the time’ 刚才要是坐地铁去就好了。

But the negation of P in such a context turns the whole sentence into a falsifying one: [37]

Mehau nong ganggang vyao gang bak yi tin nao Much-preferred you just-now do-not tell to him hear EM-SMP 蛮好侬刚刚勿要岗摆伊听孬！ ‘It would have been much better that you did not let him/her know (that) at the time’ 你刚才要是不告诉他就好了。

The proposition following ‘me51hau3’ is necessarily antiveridical. That is, the speaker presupposes that the truth-value of the proposition is false. But it can either take the positive form with a truthifying reading, as in [36], or adopt a negated form with a falsifying reading, as in [37]. Thus, ‘yàobúshì + S’ and ‘me51hau3 + S’ share the same feature of being encoded forms of counterfactual meaning, even though the former S is marked by [+veridical], and 198

Chinese linguistics: Pragmatics

the latter, by [+antiveridical]. This comparative study provides a clue to the treatment of implicit counterfactual conditionals in Mandarin. In addition, sentences involving ‘zǎozhīdào  .  .  .’ (早知道) /Early know  .  .  .   / should now be taken as encoded truthifying counterfactual constructions, as shown in [35]. Previous psycholinguistic studies have revealed that native speakers of Chinese are not used to processing unexpected, out of the blue, implicit counterfactual conditionals. But they are ready to weigh up different contingencies if enough contextual information is supplied.12 Relating this finding to our characterization about explicit counterfactuals, we can further hypothesize that even in the case of implicit counterfactuals, native speakers of Chinese still treat them as either falsifying or truthifying counterfactuals. The question then boils down to how a Chinese can tell the difference between an indicative conditional and a falsifying/ truthifying counterfactual conditional. The cues, we suggest, are the weak features, such as the aspect marker ‘le’ 了 (optional marker of perfect aspect), the temporal adverb ‘zǎo’ 早 (early), and the negation words such as ‘bú’ 不 (not) and ‘méi’ 没 (not-in-past). These words have varied meaning and functions, but when used in conditionals, they share the similar function of indicating a change of eventuality, presenting an alternative which falsifies the current case or truthifies what is not there by making things happen. Hence the counterfactual meaning. Related examples are given below: [38] yàoshì nǐ qù le, jiù bú huì bú zhùyì tā nà piàoliàng de tàitài de [truthifying] 要是你去了,就不会不注意他那漂亮的太太的。 ‘If you had been there, you would not have failed to notice his pretty wife’ [39] rúguǒ wǒ zǎo zhīdào tā bú lái, wǒ yě bú huì lái le [truthifying] 如果我早知道他不来,我也不会来了。 ‘If I knew it earlier that he would not come, I would not come either’ [40] yàoshì wǒ zhēnde zhòng le liùhécǎi, nà wǒ mǎshàng jiù huì cízhí [truthifying] 要是我真的中了六合彩,那我马上就会辞职。 ‘If I really won the Mark Six, I would quit my work immediately’ [41] yàoshì gāngcái nàgè rènyìqiú méi fá jìn, jiù huì tī jiāshísài le [falsifying] 要是刚才那个任意球没罚进,就会踢加时赛了。 ‘If that free kick had not been in, the match would have gone into extra-time’ [42] zhè chǎng huǒ, xìngkuī xiāofángduì lái dé zǎo. fǒuzé hòuguǒ búkān shèxiǎng [falsifying] 这场火,幸亏消防队来得早。否则后果不堪设想。 ‘This fire, lucky was it that firemen arrived in good time [to put it out]. Otherwise, it would have caused unimaginable damage.’ Having treated explicit and implicit counterfactual conditionals, the remaining counterfactuals are the conditionals that hypothesize a situation known to be blatantly false according to language users’ knowledge of the world: situations that are absurd, improbable or self-contradictory. Some such examples are given below: [43] yàoshì huàn le wǒ de huà, jiù búhuì duì tā zhème kèqì le 要是换了我的话,就不会对他这么客气了。 ‘If I were to deal with the case, I would not be so nice to him’ 199

Yan Jiang

[44] rúguǒ tàiyáng cóng xībiān chūlái, wǒ yīdìng jià gěi nǐ 如果太阳从西边出来,我一定嫁给你。 ‘If the sun comes out from the west, I will be your wife for sure’ [45] yàoshì tā shuō èr jiā èr děngyú wǔ – nàme hǎo de, èr jiā èr jiù děngyú wǔ 要是他说二加二等于五——那么好的,二加二就等于五。 ‘If he says that two and two are five – well, two and two are five’ [George Orwell: Looking Back on the Spanish War] [46] jiǎrú wǒ yǒu yīshuāng chìbǎng, wǒ xiǎng xiàng xiǎoniǎo yīyàng zìyóu de fēixiáng 假如我有一双翅膀,我想像小鸟一样自由地飞翔。 ‘If I had a pair of wings, I’d wish to fly freely like a little bird’ The above investigation may give the impression that only semantic factors are taken into consideration. But, gradually, we have outlined an inferential path for the comprehension of Chinese counterfactual conditionals, and the overall reasoning is a matter of pragmatic inference – in our present framework of analysis – a comprehension inference conducted under the guidance of the principle of relevance. Whether a sentence is interpreted counterfactually depends on the following factors: [47] The encoded meaning of ‘yàobúshì’ and a cluster of similar expressions [48] The identification of the contingent fact under discussion [49] The pragmatic function of the weak features [50] The identification of the absurd, the impossible, or the self-contradictory content Obviously, [48]–[50] are pragmatic factors. The typology of Chinese counterfactual conditionals is now summarized in the following graph: [51] 

Counterfactual Conditionals Explicit (encoded) Falsifying

Implicit (inferred) Truthifying Counterpossibles

7. Epilogue This is an extended introduction to Chinese theoretical pragmatics, or linguistic pragmatics for Chinese. After introducing some general concepts and phenomena, I present relevancetheoretic pragmatics and apply theoretical notions to three case studies. With these I hope to show how pragmatics can be put to work in linguistic analysis. It is important for students of Chinese grammar to seek insights in pragmatics so as to give novel accounts of some puzzling grammatical issues. It is equally important for workers of Chinese pragmatics to roll up their sleeves and work on actual topics, rather than always remain at the stage of reading, surveying and evaluating works and ideas from the West. 200

Chinese linguistics: Pragmatics

Notes   1 Cf. Blakemore (1992) and Clark (2013) for textbook introductions, and Sperber and Wilson (1995), Wilson and Sperber (2012) for more detailed discussions.   2 More studies can be found in Carston (1988, 2002) and Soria and Romero (2010) on explicature, and in Blakemore (1987) and Escandell-Vidal et al. (2011) on procedural meaning.   3 Substantive studies on this topic can be found in Zhang (2000).   4 Cf. Tan (1996) for its naming history.   5 Composed by Mǎ Zhìyuan 马致远 [c.1250–c.1321].   6 Composed by Bái Pǔ 白朴 [1226–c.1306].   7 Cf. Borg (2004), Carston (2004, 2010), and Stanley (2000).   8 Wang (2013) contains a survey on this topic.   9 Cf. Jiang (2000). 10 Cf. Chen (1988) and Wang (2013). 11 As tone sandhi in Shanghainese varies a lot in different phonetic contexts, it is customary for nonphonetic studies to omit tones in example sentences. 12 Cf. Bloom (1981).

References Armstrong, David. M. (2002) ‘Truths and Truthmakers’, in Richard Schantz (ed) What is Truth? Current Issues in Theoretical Philosophy vol. 1, Berlin and New York: Walter de Gruyter, 27–37. Blakemore, Diane (1987) Semantic Constraints on Relevance, Oxford: Blackwell. Blakemore, Diane (1992) Understanding Utterances, Oxford: Blackwell. Bloom, Alfred (1981) The Linguistic Shaping of Thought, Hillsdale, NJ: Lawrence Erlbaum. Borg, Emma (2004) Minimal Semantics, Oxford: Clarendon. Carston, Robyn (1988) ‘Implicature, Explicature, and Truth-theoretic Semantics’, in Ruth Kempson (ed) Mental Representations: The Interface between Language and Reality, Cambridge: Cambridge University Press, 155–81. Carston, Robyn (2002) Thoughts and Utterances: The Pragmatics of Explicit Communication, Oxford: Blackwell. Carston, Robyn (2004) ‘Explicature and Semantics’, in Steven Davis and Brendan Gillon (eds) Semantics: A Reader, Oxford: Oxford University Press, 817–45. Carston, Robyn (2010) ‘Explicit Communication and “Free” Pragmatic Enrichment’, in Belén Soria and Esther Romero (eds) Explicit Communication: Robyn Carston’s Pragmatics, Basingstoke: Palgrave Macmillan, 217–85. Chen, Guohua 陈国华 (1988)〈英汉假设条件句比较〉(‘A Comparison between English and Chinese Hypothetical Conditionals’),《外语教学与研究》(Foreign Language Teaching and Research) 73: 10–19. Clark, Billy (2013) Relevance Theory, Cambridge: Cambridge University Press. Escandell-Vidal, Victoria, Manuel Leonetti, and Aoife Ahern (eds) (2011) Procedural Meaning: Problems and Perspectives, Bingley: Emerald Group Publishing. Ferguson, Niall (1999) ‘Virtual History: Towards a “Chaotic” Theory of the Past’, in Niall Ferguson (ed) Virtual History: Alternatives and Counterfactuals, New York: Basic Books, 1–90. Giannakidou, Anastasia and Alda Mari (2014) ‘Future and Universal Epistemic Modals: Reasoning with Non-veridicality and Partial Knowledge’, Manuscript. Jiang, Yan 蒋严 (2000)〈汉语条件句的违实解释〉(‘On Counterfactual Interpretation of Chinese Conditionals’), in《语法研究和探索》(Grammatical Analysis and Investigation), vol. 10, Beijing: The Commercial Press, 257–79. Jiang, Yan and Wang Yuying 王宇婴 (forthcoming) ‘Counterfactual Subjunctive Assertions in Shanghai Dialect.’ To appear in Commemorative Essays on the 120 th Birthday of Professor Y. R. Chao. Levinson, Stephen. C. (1983) Pragmatics, Cambridge: Cambridge University Press. Leech, Geoffrey (1983) Principles of Pragmatics, London: Longman. Rescher, Nicholas (2007) Conditionals, Cambridge, MA: The MIT Press. Soria, Belén and Esther Romero (eds) (2010) Explicit Communication: Robyn Carston’s Pragmatics, Basingstoke: Palgrave Macmillan.

201

Yan Jiang Sperber, Dan and Deirdre Wilson (1986/1995) Relevance: Communication and Cognition, Oxford: Blackwell. Stanley, Jason (2000) ‘Context and Logical Form’, Linguistics and Philosophy 23(4): 391–434. Tan, Yongxiang 谭永祥 (1996)《修辞新格》(New Figures of Speech), Guangzhou: Jinan University Press. Wang, Yuying 王宇婴 (2013) The Ingredients of Counterfactuality in Mandarin Chinese, Beijing: China Social Science Press. Wilson, Deirdre and Dan Sperber (2012) Meaning and Relevance, Cambridge: Cambridge University Press. Zhang, Yisheng 张谊生 (2000)《现代汉语副词研究》(Studies on Adverbs in Modern Chinese), Shanghai: Xuelin Press.

202

12 Chinese Linguistics: Semantics Hsieh Shu-Kai National Taiwan University, Taiwan

1. Introduction Semantics talks about the meaning of language, and the role it plays in actual communication. In this subfield of linguistics, there have been studies focusing on different levels of units (morphemes, words, sentences, discourse) with different approaches (sense: lexical semantics; denotation: formal semantics; mental image: cognitive semantics, etc.). This chapter does not aim to achieve a comprehensive introduction of related works, but will try to convey more specialty in the semantic studies of Chinese.1 In singling out the relevant topics for inclusion in this chapter, the criteria I have used are firstly, giving priority to the particular properties of Chinese language, and secondly, skipping technical details for a smoother reading of the non-specialists who are curious about Chinese semantics. The intricate relation between language and writing system in Chinese has arguably led the studies of Chinese semantics much more lexically driven, necessitating much deeper focus on the issues of lexical meanings. This has motivated the organization and approaches of this chapter. In the following, Section 2 starts with addressing the basic ingredients of a semantic theory of Chinese, with the main focus on the interface of morphology and semantics, Section 3 moves to the topics of lexical semantics, and Section 4 summarizes the chapter.

2.  Chinese morpho-semantics Most (Chinese) linguistic researches assume the notion of wordhood. In the developmental history of linguistic theory, words have been supposed to be the core linguistic units for the sake of scientific investigation of languages.2 Among different versions of the definition of ‘word’, the most popular one, ‘the minimum (meaningful) free form’ (Bloomfield 1926: 155), claimed to be universally applicable to all languages. Morphologically, new words can be created via different ways of word formation, and result in different types of complex words. For example, you could coin the word payable, deriving it from the verb pay by adding the suffix –able, or you could form the compound fast food by combining two words to form a composite one. However, the semantics of word formation in Chinese is a field that is much less clear than its counterparts in other languages. This is mainly due to the controversial status of 203

Hsieh Shu-Kai

word, which has resulted from the intricate relationship between ideographic writing system and language. In the Chinese writing system, in principle, each character 字 (zi4) represents one syllable from the aspect of sound, and one morpheme which carries strong semantic weight. Word inventory in contemporary Chinese thus consists of: a single morpheme, or combination of two (and arguably three) morphemes. It was reported and can be empirically justified that in many cases, whether a single morpheme or the combination of morphemes constitute a word or not is open to debate, let alone the unstable notion of compound which is built on top of word. This confusion regarding what a word is in Chinese among lay speakers3 and disagreement among linguists has worsened due to the fact that the relative predominance of the monosyllabic word in Ancient Chinese has shifted to bisyllabic words in Modern Chinese. The huge amount of semantic information encoded by the characters has deeply ingrained the idea of character-centered lexicon in Chinese minds. It is noted that although it is astonishingly difficult to give a definition of ‘word’, it does not necessarily follow that it is impossible. Taking an understanding of the above-mentioned background as the point of departure, in the following section, we will take a closer look at the morpho-semantics of Chinese by way of its quantitative behavior and conceptual representation.

2.1.  Chinese morphemes, words and compounds Linguistically, words are made up of at least one meaningful piece, which is defined as a morpheme, the minimal meaningful unit that is used to form words. Morphemes can be further divided into free morphemes, those that can stand alone; and bound morphemes, those cannot stand alone and must be attached to other morphemes to form words. Bound morphemes come in different varieties: those that come before the base of the word are prefixes, and suffixes are bound morphemes that come after the base of the word (Lieber 2009). As remarked previously, we can take the characters for the orthographic unit representing morphemes in Chinese, but still encounter problems. For instance, 健康 (jian4-kang1, ‘health’) is a word consisting of two bound morphemes, while 财富 (cai2-fu4, ‘wealth’), though recognized as a word, consists of two morphemes where the boundedness (i.e. whether they can be used alone in Modern Chinese) is a question mark. The indeterminacy of morphological status then affects the decision of whether 财富 is one word composed of two bound morphemes (财 and 富), or a compound word composed of two words that are both free morphemes, respectively.4 To tackle the issue, Chinese linguists have proposed different kinds of wordhood test, such as insertion criteria (Lu 1957), exocentric structure (Chao 1968), conjunction reduction (J. Huang 1984), etc. However, none of the criteria seem to be always reliable to license a true word in every case, leaving the problem still unsolved. In addition, orthographically, the lack of inflectional markers and word delimiters such as space in texts, also worsens the achievement of consensus regarding the distinction between words, compounds and phrases. Interplayed with structural ambiguity, the automatic word segmentation has become a long-standing heated topic in Chinese Natural Language Processing (NLP). For instance, the phrase ‘已结婚的和尚未结婚的青年’ can be grammatically segmented into two structural readings: ‘已/结婚/的/和/尚未/结婚/的/青年’ and ‘已/结婚/的/ 和尚/未/结婚/的/青年’. This has led to diverse segmentation principles in the field of Chinese NLP, where a set of operational rules is provided to guide word segmentation for engineering purpose.5 In addition to the linguistic discussion, we can also observe the intricate morpheme behavior in a quantitative way, which falls under the rubric of morphological productivity. 204

Chinese linguistics: Semantics

The notion of morphological productivity in the field of Morphology refers to the process of lexeme formation that can be used by native speakers to form potential words, in particular the process of derivational affixation that involves a base word and an affix. In English for example, though the suffix –th might be recognized as a suffix such as in warmth, it is rarely used to make new words, compared to other suffixes such as –ity and –ness (Lieber 2009). That is, different morphemes have different potentials to form words. A number of ideas have been proposed to measure the degree of productivity. Baayen (1992) considered the hapax legomena (i.e. words that appear only once in a given corpus), and proposed a wellrecognized measure P = n1/N, where N is the number of tokens (e.g. of all nouns ending in –ness), n1 represents the hapax; while Nishimoto (2003) proposes a type-based measure instead, and reports the yielding results of productivity ranking of five Chinese suffixes ‘–men, –hua, –r, –zi, –tou’ where –men is the most productive and –tou is the least productive. For the sake of illustrative purposes, the author calculated the Baayen’s P values for four major suffixes in Chinese based on a sample of the Academia Sinica Corpus,6 with the results of ((noun plural –men 们: 0.53), (nominal derivational suffix –zi 子: 0.35), (noun-forming suffix –lu 率: 0.45), (occupational suffix –jia 家: 0.32), (experiential suffix –guo 过: 0.33), (experiential suffix –jian 见: 0.36)). Comparing with some sample P values from English suffixes (Sproat 2002) , such as ((–ness: 0.0044), (–ish: 0.0034), (–action: 0.0006), (–ity: 0.0007)), we can find a significant difference showing that Chinese suffixes might have higher morphological productivity in terms of Baayen’s P value. The experiment can give us a quick idea of how Chinese morphology, due to its ideographic characters (morphemes), differs essentially from other languages. In fact, the notion of suffix does not fit well with Chinese morphology either. A suffix is defined as a bound morpheme in linguistics, which is an issue in dispute in Chinese. To highlight the specialty, Packard (2000: 77–8) proposes that unlike ‘typical’ affixing languages, Chinese has a large class of morphemes – which he coined as bound roots – that possess certain affixal properties (i.e. they are bound and productive in forming words), but encode lexical rather than grammatical information.7 These morphemes may occur as either the left- or right-hand component of a word. For example, the morpheme 家 (jia1) is treated as an occupational suffix (-家) as mentioned in previous experiments, or prefix (as in 家-, ‘anything related to home’). However, as remarked repeatedly, the difficulty is exacerbated by the fact that the boundary between bound and free morpheme is mostly fuzzy. A native speaker can easily come up with a sentence where the morpheme 家 itself is a word used alone, as in 我家在那儿 (‘My house is over there’). The phenomenon leads many Chinese linguists to emphasize the role of the morphemic component (i.e. character) itself as the building block in the semantic compositional process of di- or trisyllabic words, regardless of its morphological status.8 To get a deeper understanding of the semantic compositional process, researchers have proposed various rules for deriving the meaning of a morpheme combination from the meaning of the morphemic components. Löbner (2013: 13) noted that these compositional rules are complicated by the fact that there are several interpretation rules for the same composition rule. Compounds are particularly intricate in how the meaning of the components are semantically related: a 书架 (shu1-jia4, ‘book shelf’) is a shelf on which books can be stored; a 书名 (shu1-ming2, ‘book title’) is a title that names a book; while a 书券 (shu1 quan4, ‘book coupon’) is a voucher entitling the holder to a discount of a particular book. Nevertheless, the inter­ pretation of compounds (morpheme combinations) is not just mere arbitrary guessing, for the native speaker obviously agrees on how to interpret regular compounds. Based on empirical corpus data, Sproat (2002) suggested some semantic relations such as ‘used for’ 205

Hsieh Shu-Kai

in 饭碗 (fan2wan3, ‘rice bowl’); ‘powered by’ in 电灯 (dian4deng, ‘electric light’), etc. All the suggestions indicate that there should be a conceptual knowledge system underlying the characters. The following section will lead the reader to the studies of how Chinese character-morphemes encode semantic-conceptual knowledge and how the structures can be explicitly specified.

2.2.  Ontology with and within Chinese characters Notwithstanding the long-standing belief that core conceptual information is implicitly encoded in Chinese characters, the writing system in general has marginal relevance in formal linguistic theories, as it is a conventionalized secondary representation, and the issue of how writing systems encode linguistic elements has thus rarely been studied in semantics. In this section, I will walk readers through how the current studies of ontology can reveal the missing part of Chinese linguistics. First of all, we start with three assumptions: (i) Concepts and lexical meanings (i.e. word sense) are different, (ii) Character-morphemes are relatively objective cues to reflect conceptual organization, and (iii) Character-morpheme networks constitute the interface of concepts and word senses. Figure 12.1 schematizes a three-layered representation of our proposal (C.-R. Huang et al. 2008). At the middle layer (the character level), we take an ‘atomic globule’ network view (Aitchison 2003), where the characters – realized as core concept units – which share similar conceptual information, cluster together. Down to the word level, we take the ‘cobweb’ view, as words – built up from a pool of characters – are connected to each other through different semantic relations. It has been further argued that radicals 部首 (bu4 shou3) – the composite units of characters – represent conventionalized knowledge system, and offer an opportunity to explore cross-language ontologies (Wong and Pala 2002). The radicals, a.k.a. 意符 (yi4 fu2, ‘semantic symbols’) in the Chinese lexicological tradition have been considered as essential semantic components within characters. The most classic lexicographical work – 说文解字 (ShuoWenJieZi) compiled by Xu Shen (121 ce) – is organized according to the semantics of the 540 radical forms. Recent studies (Chou and Huang 2007) show that the class of characters sharing the same radicals can be linked to the qualia structure proposed by Pustejovsky (1998). The qualia structure, whose original idea can be traced back to Aristotle’s aitiae (‘modes of explanation’) (Physics 194 b17–20), can be viewed as a set of semantic constraints by which a lexical item is conceptually understood. Aristotle stated that there are four modes of explanation that drive our basic understanding of an object or relation in the world: (1) Material causality: what constitutes the object in question, (2) Formal causality: what distinguishes the object from others, (3) Efficient causality: how the object in question came about, and (4) Final causality: what purpose the object has. Pustejovsky (1998) transforms the ‘modes of explanation’ into one of the lexical interpretive levels called qualia structure with four qualia: (1) Formal quale: what distinguishes it from others, (2) Constitutive quale: what constitutes it, (3) Telic quale: what purpose it has, and (4) Agentive quale: how it comes about. In applying the theory to Chinese radical semantic knowledge, C.-R. Huang et al. (2007) analyzed the Bovid Domain Ontology Conventionalized by Radical ‘羊’ and found that the deriving concepts, i.e. characters shared with the radical, can be explained by the qualia, e.g. formal quale: ‘羳’ (fan2, ‘bovid which has yellow belly’) involves both the constitutive part of a bovid’s belly and its visual attribute; constitutive quale: ‘群’ (qun2, ‘a group of’), and agentive quale: ‘羜’ (zhu4, ‘lamb of five months old’), etc.9 206

Chinese linguistics: Semantics

CONCEPT

Character level (HanziNet)

Word level (WordNet)

Figure 12.1  The pyramid structure model

3.  Chinese lexical semantics As we have seen so far, the lexical meanings are complex. This complexity can be explained by tracing back to its ontological status at the morpheme (character) level, it has also been tackled largely at the lexeme (word) level. Lexical semantics is the study of word meaning, by representing lexical knowledge in terms of phenomena such as argument linking, selectional restrictions, lexical relations, etc. As introduced in Section 2, the linguistic status of word in Chinese has not been commonly recognized, it is worth noting that despite such controversy, the establishment of the concept of word as a relatively stable linguistic unit has its advantages in mirroring the conceptualization of language on the one hand, and helping determine the course of rigorous compositional linguistic analysis on the other hand. In this section, I will sketch some discussions of lexical semantics in Chinese, as well as an implementation of a lexicon database called WordNet, which can serve as a bridge for cross-language comparison. 207

Hsieh Shu-Kai

3.1.  Word meaning (semantic properties of words) The meaning of words, a.k.a. word sense, can be defined as the instantiation of lexicalized concepts. In the context of lexical semantics, the sense of a word can be scrutinized through the lens of its inherent features and meaning relations with other words, rather than by its denotational relation to things in the world. Many intricate lexical phenomena have been distinguished and discussed. Polysemy and Homonymy are two typical examples for the variants of word meanings. Polysemy refers to the phenomenon where a lexeme has at least two interrelated meanings. Considering the verb example 开 (kai1), which means ‘move something (door or windows) so as to leave a space allowing access’ (e.g. in 开门 ‘kai1 men2’); ‘make available’ (e.g. in 开放 ‘kai1 fang4’) or unfold or be unfolded (e.g. in 开展 ‘kai1 chang3’), which are closely related.10 In contrast, Homonymy is concerned with the phenomenon where two lexemes have unrelated meanings, but happen to share the same form. For example, though 打 (da3) can have multiple meanings, at least two of them are totally unrelated – ‘punch’ and ‘classifier for counting objects’ – thus the lemma 打 represents two homonymous lexemes which share the same lemma form. The polysemous behavior of words results in the difficulties of computational semantic processing of human languages. A subfield of computational semantics called Word Sense Disambiguation (WSD) has emerged to handle this issue. Based on a sense inventory (e.g. from WordNet11 or HowNet12), a few machine-learning techniques have been employed to automatically pick out the most likely sense for a certain word/all words in a text. However, the assumption in WSD tasks, that is, one-sense-per-word in a certain context, will encounter meaning shifts in various contexts. Asher (2011) describes co-predication as one of the sorts of predication in which two or more predicators select the same argument with distinct yet systematically related senses. So in (a), the modifier delicious selects the [content] sense, while took forever selects [event] sense of lunch. Similarly, in (b), 种 (‘zhong4’) and 很好喝 (‘hen2 hao3 he1’) both predicate the [plant] and [liquid] sense of 茶 (‘cha2’), respectively.13 (a) The lunch was delicious but took forever. (b)

埔里种的茶很好喝。 Pu4li3 zhong4 de cha2 hen2 hao3 he1. Puli plant DE tea very good drink ‘The tea grown in Puli tastes delicious’

It is worth noting that the co-predication, where lexical meanings allow co-existing multiple readings in a given sentence, is different from widely recognized lexical ambiguity, where multiple readings are allowed, but can and need to be resolved given more contextual information. As can be seen in (c), where the ambiguous word 看病 (‘kan4 bing4’ ) can be disambiguated given that fact that the reference status of 他 (‘ta1’, he), i.e. either as ‘doctor’ or ‘patient’, is determined. (c)

他正在看病。 ta1 zheng4 zai4 kan4 bing4 he PROG seeing sickness ‘He is seeing the doctor/He is examining the patient’ 208

Chinese linguistics: Semantics

3.2.  Semantic relations Semantic relations also constitute the core part of our lexical knowledge. Saussure ([1916] 1983) emphasized that these differences are of two kinds: paradigmatic (concerning substitution) and syntagmatic (concerning positioning). Paradigmatic relations hold between words of which the meaning is related in some systematic ways. Often they belong to the same syntactic category, as for example 便笺, 便条, 条子 (note, post-it note, notelet). A few paradigmatic relations that receive the most attention in lexical semantics are semantic paradigmatic ones between word senses, where they share many semantic properties, but fail to share others. For example, synonymy constitutes the relation among words having the same meaning (e.g. 从不 cong2 bu4 / 未曾 wei4 ceng2; ‘never’); hypernymy-hyponymy defines a ‘type-of ’ meaning (歌 ge1 / 民谣 min2 yao2; ‘song/folk song’); meronymy-holonymy for the ‘part-of ’ meaning (手 shou3 / 手掌 shou3 zhang3; ‘hand/palm’), etc. A set of substitution tests as well as logical conditions for each relation is also defined in Tsai et al. (2002) and Huang et al. (2004), which are meant to be working criteria that facilitate identification of relation instances and promote consistency of decisions among linguists. In contrast, syntagmatic relations are based on the co-occurrence of words within a sentence, like collocations. Typically, they hold between words of different syntactic categories such as nouns and verbs, e.g. the relations between [吃]–[水果] (chi1–shui2guo3; ‘eat–fruit’) in the sentence 他吃了一盘水果 (He ate a dish of fruit). There has been no available lexical resource for explicit specification of syntagmatic relations in Chinese; however, interested readers can refer to a corpus-based query system called Chinese Word Sketch Engine14 developed at Academia Sinica. Figure 12.2 shows the sketch results of words syntagmatically related to the target word 吃.

3.3.  Chinese WordNet In addition to the description of word senses, semantic relations are at the core of WordNetlike architecture, and constitute the essential and integral part of linguistic and conceptual knowledge formalization. A lexical knowledge base is a general repository of knowledge about lexicalized concepts and their relationships. WordNet (Fellbaum 1998) and FrameNet (Baker et al. 1998) represent the advances of these two kinds of semantic relations respectively. In

Home Concordance Word Sketch Thesaurus Sketch-Diff sinica freq = 4660 SentQbject of365 6.1 Object 2616 3.0 Modifier 1375 2.8 Subject 2080 2.4 pp 在 96 37.92 7 1 36.94 46 39.92 36 26.06 M 36.56

75 36.32

20 23.01

7 21.47

10 27.61

25 33.89

20 22.84

12 20.69

18 24.92

7 1 32.98

108 22.39

29 20.53

6 19.99

143 32.74

13 21.94

7 18.87

16 19.16

39 32.56

50 21.1

37 18.07

8 18.91

27 30.06

70 20.41

48 17.58

Figure 12.2  Syntagmatic relations from Chinese Word Sketch Engine

209

9

0.5

5

22.06

183 0.2 30 28.07 7 15.13 5 13.96 5 8.09

Hsieh Shu-Kai

particular, the WordNet-like framework pioneered by the Princeton WordNet, and greatly enriched by EuroWordNet (Vossen 1998) and further extended to Global WordNet Grid, has thus become the de-facto standard for a lexical knowledge base enriched with lexical semantic relations (LSRs).15,16,17 Modeled on English WordNet, the Chinese WordNet18 project aims to provide a bilingual mirror for lexical knowledge representation. Following Princeton’s English WordNet, Chinese WordNet (CWN) is structured as a hierarchy of synsets (sets of lexical units that are synonyms), and each synset is intended to represent lexicalized concept. These lexicalized concepts are connected with different kinds of lexical semantic relations (such as hypernymy, meronymy, antonymy, etc.) Each synset is also described by a lexicographic definition and examples. For instance, 好 has 15 senses distinguished in CWN, which are assigned to 15 different synsets. One of these synsets is the set containing the overlapping sense of {好, 不错}, etc. Figure 12.3 shows a snapshot of the system. As defined in English WordNet, different semantic and lexical relations have also been implemented as a core part in the CWN. It is noted that the CWN developer team has posted many interesting issues reflecting the specific properties of Chinese language. For example, concerning the superordinate relations (a.k.a. sister terms) in WordNet, which are used to define those coordinate words that have the same hypernym, Huang et al. (2007, 2015) observed that not all coordinate terms are equal when detailed lexical analysis is done for a set of coordinate terms sharing the same hypernym. For instance, the first intuition for the concept of 季节 (ji4 jie2, ‘season in a year’) will be four seasons – spring, summer, autumn, and winter. Therefore, the term spring is a hyponym with respect to season, while spring, summer, autumn and winter are co-hyponyms, i.e. hyponyms of the same superordinate term. However, other terms for seasons, such as 干季 (dry season) and 湿季 (rainy season) are not thought of intuitively as parallel as the four seasons, although all of them share the same superordinate concept, ‘seasons in a year’. So the default relation of co-hyponym in WordNet

C h in e s e w o r d n e t

(ABOUT)

(DOCS)

(DOWNLOADS)

(STATS)

(Enter a word, e.g. (Query)

15 l.

Figure 12.3  Chinese WordNet

210

(QUERY)

(LICENCE)

Chinese linguistics: Semantics

is too general to cover the detailed relations for its set of hyponyms. These coordinate terms could be re-classified into conceptually salient groups, and the conventional ways used to define the co-hyponyms, that is, antonymy and near-synonymy, seem inadequate to account for such conceptual classification. In addition to varied lexical semantic relations connecting different word senses and lemma, it has been also found that interesting relations also exist among their morphological behavior. This can be pointed out in three aspects in general: (i) inflectional variations of words (e.g. work, working and worked, which share everything except their tense form), (ii) derivational variations (e.g. work and worker), and (iii) regular compounds, which comprise the modifier and the head part (e.g. flower shop) (Murphy 2003; Löbner 2013). As mentioned, the vast majority of Chinese characters represent the morpheme, and the intricate ways of combining morphemes to form a word or compound in Chinese has raised much discussion in previous literature. In order to enrich the lexical semantic relations in CWN, Hsieh and Chang (2014) proposed a simple computational method to automatically acquire semantic relations relying on the in situ morpho-semantic structure in Chinese, and the system has achieved great results. This method assumes a model called Morpho-Semantic Linkage (MSL) in Chinese. Because of its strong semantic weight, in most cases of characters (i.e. a word-to-be morpheme or word-used-to-be morpheme), there must be some relation existing among them (A-AB, B-AB, AB-BC, etc), and this ‘to-be labeled’ morpho-semantic relation is defined as Morpho-Semantic Linkage in Chinese. LSRs among concepts (in terms of CWN) is thus only a subset of MSLs (recall the qualia relations mentioned). Figure 12.4 shows the scheme. So the proposed method enriches the CWN with relations leveraged by operationalizing MSL, i.e. based on the position and semantic role of (word-to-be) morphemes in modification construction. It is noted that it is not concerned with semantic relations between nominals, as often studied in the field of Information Extraction, i.e. the relations between N1 and N2 in N1N2 compounds. It aims to operationalize linguistic theories for the automatic labeling task of lexical relations on word-pairs. For N1N2 compounds, pairs are labeled with hypernymy–hyponymy, and pairs are labeled with meronymy–holonymy. In some ways the MSL space parallels the lexical fields theory (Löbner 2013: 215). The notion of lexical field comes from the observations that most lexical items form groups with other lexemes, the senses of these lexemes have something in common, and they are interrelated by definable meaning relations. For example, we can define lexemes sharing the same morpheme (thus the same radical component) as forming a lexical field. Written in

MSL space LSRs

Figure 12.4  Morpho-semantic linkage space

211

Hsieh Shu-Kai

set notation, it can be a set including a list of lexemes in brackets {.  .  .}, such as {车, 车辆, 车门, 车窗, 车厢, 汽车, 火车, 机车,  .  .  .}. The lexical semantic relations among the lexemes can be seen in its subsets: (1) hypernymy–hyponymy: {车–汽车}; {车–火车};  .  .  . (2) meronymy–holonymy: {车–车门}; {车–车窗};  .  .  . (3) (near-) synonymy: {车–车辆};  .  .  . (4) co-hyponymy: {车辆–汽车};  .  .  . Further exploration can be done using radical information as clues to determine the finegrained classification of the retrieved relations.

4. Summary The study of meaning has attracted a number of disciplines, including linguistics, psychology, cognitive science, philosophy, as well as many others. As the subfield of linguistics, contemporary semantics is characterized by various formal modeling, abstract representation, and empirical evidence that can be verified or falsified in a scientific manner. This chapter sketches core issues surrounding lexical meanings in Chinese, ranging from morpho-semantics to lexical semantics. As has been remarked, there are clearly several other interesting topics such as formal semantics and cognitive semantics that could each deserve a chapter-length full treatment, but I have had to leave them out and focus on the parts bearing on lexical aspects. Hopefully, it still provides a well-rounded picture of the endeavor in the study of Chinese semantics.

Notes   1 Interested readers are referred to Löbner (2013) for general introduction.   2 There have been few alternative views proposed. For example, Davis (2001) aims to reorient linguistics by reminding that the concept of ‘word’ does not pre-exist for ‘ordinary’ speakers.   3 Sproat and Shih (1996) only measured 76% agreement among human judges.   4 For instance, in real sentences like ‘有土斯有[財]’ and ‘國家[富]起來是件好事’, 財 and 富 are used as free morphemes.   5 There have been at least three ‘word segmentation standards’ proposed by different groups which reflect different views of Chinese wordhood: (1) ROCLING standard (Huang et al. 1997; CNS14366, 1999), (2) U-Penn standard (Xia 2000), (3) Peking University (GB/T13715, 1992).   6 Calculation is based on Sproat’s data available at http://rws.xoba.com/exercises/.   7 An example of a bound root from English could be the –ceive (in receive, conceive, etc.).   8 For a detailed review on this topic, please refer to Zhu (2004).   9 The qualia structure also enables the compositional interpretation within the compounds. See Lee et al. (2010). 10 More sense distinctions can be found at http://lope.linguistics.ntu.edu.tw/cwn2/query/ 11 http://wordnet.princeton.edu/. 12 http://www.keenage.com/. 13 Note that the pairings like [content]–[event] or [information]–[physical object] are only some instances of what is coined as systematic polysemy. Refer to Pustejovsky (1998) for more discussion. 14 http://wordsketch.ling.sinica.edu.tw/. 15 http://wordnet.princeton.edu/. 16 http://framenet.icsi.berkeley.edu/fndrupal/. 17 http://globalwordnet.org/global-wordnet-grid/. 18 http://lope.linguistics.ntu.edu.tw/cwn2. This is different from HowNet, a Chinese–English bilingual online lexicon where the inter-conceptual relations and inter-attribute relations of concepts are assumed. See http://www.keenage.com.

212

Chinese linguistics: Semantics

Bibliography Aitchison, Jean (2003) Words in the Mind: An Introduction to the Mental Lexicon, Oxford: Blackwell Publishing. Asher, Nicolas (2011) Lexical Meaning in Context: A Web of Words, Cambridge: Cambridge University Press. Baayen, Harald (1992) ‘On Frequency, Transparency and Productivity’n Yearbook of Morphology, 181ook. Baker, Collin, Charles J. Fillmore, and John B. Lowe (1998) ‘The Berkeley FrameNet Project’, in Proceedings of the COLING-ACL 1998, 86–90. Bierwish, Manfred (1970) ‘Semantics’, in J. Lyons (ed) New Horizons in Linguistics, Harmondsworth: Penguin, 166–84. Bloomfield, Leonard (1926) ‘A Set of Postulates for the Science of Language’, Language 2(3): 153–64. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley: University of California Press. Chou, Ya-Min and Huang Chu-Ren (2007) ‘Hantology: An Ontology based on Conventionalized Conceptualization’, in Proceedings of the Fourth Onto Lex Workshop, Jeju, Korea. CNS14366 (1999)《中文信息处理分词规范》(Word-breaking Standards in Chinese Information Processing), Taipei: The Central Standardization Office of the Ministry of Economy of Taiwan. Cruse, Alan (1986) Lexical Semantics, Cambridge: Cambridge University Press. Davis, Hayley (2001) Words: An Integrational Approach, Richmond: Curzon. Dong, Zhendong and Dong Qiang (2006) HowNet and the Computation of Meaning, Hackensack, NJ: World Scientific. Fellbaum, Christina (1998) WordNet: An Electronic Lexical Database, Cambridge MA: MIT Press. GB/T13715 (1992)《信息处理用现代汉语分词规范》(Modern Chinese Word-breaking Standards in Information Processing), Beijing: State Standardization Management Committee. Huang, Chu-Ren, Chang, Ru-Yng, and Lee, Shiang-Bin (2004) ‘Sinica BOW (Bilingual Ontological Wordnet): Integration of Bilingual WordNet and SUMO’, in Proceedings of The 4th International Conference on Language Resources and Evaluation (LREC2004), Lisbon, Portugal. Huang, Chu-Ren, K. J. Chen, and L. L. Chang (2007) ‘Segmentation Standard for Chinese Natural Language Processing’. CKIP Technical Report 96-01. Taipei: Academia Sinica. Huang, Chu-Ren and Hsieh Shu-Kai (2015) ‘Chinese Lexical Semantics’, in W. Wang and C. Sun (eds) The Oxford Handbook of Chinese Linguistics, Oxford: Oxford University Press. Huang, Chu-Ren, Su I-Li, Hsiao Pei-Yi, and Ke Xiu-Ling (2007) ‘Paranyms, Co-hyponyms and Antonyms: Representing Semantic Fields with Lexical Semantic Relations’, in Proceedings of Chinese Lexical Semantics Workshop, 20–23 May, Hong Kong Polytechnic University, Hong Kong. Huang, Chu-Ren, Elanna I. J. Tseng, Dylan B. S. Tsai, and Brian Murphy (2003) ‘Cross-lingual Portability of Semantic Relations: Bootstrapping Chinese WordNet with English WordNet Relations’, Language and Linguistics 4(3): 509–32. Huang, Chu-Ren, Yang Y.-J., and Chen S.-Y. (2008) ‘An Ontology of Chinese Radicals: Concept Derivation and Knowledge Representation Based on the Semantic Symbols of Four Hoofedmammals’, in Proceedings of the 22nd Pacific Asia Conference on Language, Information and Computation, Cebu City, Philippines. Huang, Chu-Ren 黄居仁, Zhou Yamin 周亚民, and Hsieh Shu-Kai 谢舒凯 (2008)《语言、文字与 知识架构: 由汉字出发的知识本体研究》(Language, Words, and Knowledge Structure: Studies on Knowledge Ontology Based on Chinese Characters), Taipei: Academia Sinica. Huang, James (1984) ‘Phrase structure, Lexical Integrity, and Chinese Compounds’, Journal of the Chinese Language Teachers Association 19: 53–78. Hsieh, Shu-Kai (2011) ‘Sense Structure in Cube: Lexical Semantic Representation in Chinese WordNet’, International Journal of Computer Processing of Languages 23(3): 243–53. Hsieh, Shu-Kai and Chang Yu-Yun (2014) ‘Leveraging Morpho-semantics for the Discovery of Relations in Chinese WordNet’, in Proceedings of 7th International Conference on Global WordNet, Tartu, Estonia. Jian, Y. and Pan H. (2005) Introduction to Formal Semantics, Beijing: Chinese Social Sciences Publisher (in Chinese). Lee, Chi-Yao, Chang Chia-Hao, Hsu Wei-Chieh, and Hsieh Shu-Kai (2010) ‘Qualia Modification in Noun-Noun Compounds: A Cross-Language Survey’, in Proceedings of the 22nd Conference on Computational Linguistics and Speech Processing (ROCLING 2010), Chi-Nan University, Taiwan, 379–90.

213

Hsieh Shu-Kai Lieber, Rochelle (2009) Introducing Morphology, Cambridge: Cambridge University Press. Löbner, Sebastian (2013) Understanding Semantics, London and New York: Routledge. Lu, Zhiwei 陆志韦 (1957)《汉语的构词法》(Word Formation in Chinese), Beijing: Science Press. Maienborn, C., K. von Heusinger, and P. Portner (eds) (2011) Semantics: An International Handbook of Natural Language Meaning, Vol. 1:3, Amsterdam: De Gruyter Mouton. Miller, George and Christina Fellbaum (2003) ‘Morphosemantic Links in WordNet’, Journal of Traitement automatique de langue 44(2): 69–80. Murphy, Lynn (2003) Semantic Relations and the Lexicon, Cambridge: Cambridge University Press. Murphy, M. L. (2010) Lexical Meaning, Cambridge: Cambridge University Press. Nishimoto, Eiji (2003) ‘Measuring and Comparing the Productivity of Mandarin Chinese Suffixes’, Computational Linguistics and Chinese Language Processing 8(1): 49–76. Packard, Jerome L. (2000) Morphology of Chinese: A Linguistic and Cognitive Approach, Cambridge: Cambridge University Press. Pustejovsky, James (1998) The Generative Lexicon, Cambridge, MA: MIT Press. Ruppenhofer, J., M. Ellsworth, M. R. L. Petruck, C. R. Johnson, and J. Scheffczyk (2010) ‘FrameNet II: Extended Theory and Practice’, (https://framenet2.icsi.berkeley.edu/docs/r1.5/book.pdf.) Saussure, Ferdinand (1916/1983) Course in General Linguistics, trans. Roy Harris, London: Duckworth. Sproat, Richard (2002) ‘Corpus-based Methods in Chinese Morphology’, Tutorial slide at COLING 2002, Taipei, Taiwan (http://www.cslu.ogi.edu/~sproatr/newindex/coling.pdf ). Sproat, Richard and Shih Chilin (1996) ‘A Corpus-based Analysis of Mandarin Nominal Root Compounds’, Journal of East Asian Linguistics 4(1): 1–23. Su, Xinchun 苏新春 (2008)《汉语词义学》(Chinese Semantics), Beijing: Foreign Language Teaching and Research Press. Tsai, Dylan Bo-Sheng, Huang Chu-Ren, Tseng Shu-Chuan, Lin Jen-Yi, Chen Keh-Jiann, Chuang Yuan-Hsun (2002)〈中文词义关系的定义与判定原则〉(‘The Definition and Defining Principles of the Meanings of Chinese Terms’), Journal of Chinese Information Processing 16(4): 21–31. Vossen, Piek (1998) EuroWordNet: A Multilingual Database with Lexical Semantic Networks, Dordrecht: Kluwer Academic Publishers. Wong, K. F., Li W., Xu R., and Zhang Z. (2009) Introduction to Chinese Natural Language Processing, San Rafael, CA: Morgan and Claypool. Wong, Shun-Hua and Karel Pala (2002) ‘Chinese Characters and Top Ontology in EuroWordnet’, in Singh, U. N. (ed) Proceedings of the First Global WordNet Conference, Mysore, India. Xia, Fei (2000) ‘The Segmentation Guidelines for the Penn Chinese Treebank (3.0)’, (http://www.cis. upenn.edu/~chinese/segguide.3rd.ch.pdf ). Yip, P. C. (2000) The Chinese Lexicon. A Comprehensive Survey, London and New York: Routledge. Zhu, Yan 朱彦 (2004)《汉语复合词语义构词法研究》(Morpho-semantic Studies of Chinese Compounds), Beijing: Peking University Press.

214

13 Chinese Morphology Jerome L. Packard University of Illinois, USA

Introduction The Chinese language has significant morphological processes, forming complex words from word components and using affixation processes as one way of forming those complex words.1 Complex word formation dominates in a discussion of Chinese morphology however, because the number of complex, multimorphemic words in the language is far greater than the number of non-complex single-morpheme words. While Chinese does have affixation processes, those properties play a role that is less frequent and visible than other ways of forming complex words. As a first step in discussing morphology in Chinese, it is worthwhile to consider where the language fits when compared to the rest of the world’s languages from the perspective of morphological typology, i.e. whether Chinese is an isolating (or analytical ), agglutinative or inflecting language (cf. Sapir 1921/1949).

Morphological typology In terms of morphological typology, Chinese is usually cited as being a language of the isolating or analytical variety. But the extent to which Chinese is considered a language of the isolating type depends upon how ‘isolating’ is defined (Packard 2009). If the term is used to mean whether the words of the language are mostly monomorphemic – then Chinese cannot really be considered very isolating, because most words in Chinese are multimorphemic rather than monomorphemic (when counted in terms of word ‘types’ versus ‘tokens’2). However, if by ‘isolating’ we mean whether the morphemes of Chinese are easy to isolate and identify, then Chinese must be considered quite isolating indeed. The second definition of isolating given above – whether the morphemes of a language are clearly identifiable – is defined by three criteria: (i) whether morpheme boundaries in the language are sharply defined, (ii) whether there is a single distinct morphemic identity represented within a defined morphemic space, and (iii) whether allomorphy is a significant feature of the language, i.e. whether morphemes in the language have multiple phonological forms. Addressing the first criterion, morpheme boundaries in Chinese are extremely well defined, perhaps more clearly defined than in any of the world’s languages. In reading a Chinese text, 215

Jerome L. Packard

it is a simple task to identify the morphemes because the orthographic unit in Chinese – the character – lines up with Chinese morphemes on a virtually perfect one-to-one basis. English, in contrast, tends to have a one-to-one relation between the word and the orthographic unit. The second criterion for deciding whether Chinese morphemes are easy to identify – i.e. whether Chinese usually has only a single morphemic identity represented within a defined morphemic space – is answered resoundingly in the affirmative, as Chinese has virtually no morphemic forms that contain more than one morphemic element. To offer a contrasting example, in Latin the accusative singular form of the word ‘girl’ puella (nom. sing.) is puellam, with puell– serving as the root meaning ‘girl’ and the suffix –am representing both accusative case and singular number. The Latin morpheme –am has two morphemic identities – accusative and singular – and there is no way to isolate the part of –am that stands for singular number and the part that stands for accusative case. The term ‘overlapping exponence’ is used in linguistics to describe morphemic forms that contain more than one morphemic element, and in Chinese, there is no overlapping exponence whatsoever. In Chinese, there are no forms that combine two types of morphemic information into one defined morphemic space. The third criterion used to rate whether the morphemes of a language are easy to identify is the presence of allomorphy. In Chinese, there is very little allomorphy, i.e. there are very few instances of one morpheme having multiple phonological identities. Even the nominal suffix 儿 –er, which may appear to have multiple phonological identities when it suffixes to different nouns, in fact has only one, i.e. [er]. The apparent phonological changes of –er that occur when it suffixes to different nouns in fact are completely predictable phonologically, and actually cause changes to the phonological form of the noun to which it suffixes, and not changes to the –er suffix itself. In fact, what may be the sole example in Mandarin of what could arguably be termed allomorphy is the alternation of the negative morphemes 不 bu and 没 mei in different contexts. The distributional facts are clear: if negation occurs either in the context of completed aspect or in conjunction with the verb 有 you ‘be/have’ that marks existence or possession, the alternate mei is used, and in all other contexts negation is marked by the alternate bu. Aside from this very clear example of mei/bu alternation, allomorphy does not appear to exist in Mandarin. Returning now to the discussion of the morphemic structure of Chinese words, Chinese words are mostly complex, i.e. multimorphemic, entities. If the words in a selected portion of Chinese text are categorized according to whether they consist of one morpheme or more than one morpheme, the majority will invariably be found to be multimorphemic. Having discussed the identity of Chinese in terms of morphological typology, let us now examine the structure and formation of Chinese words. But before considering the structure of words and how they are formed, criteria must first be provided to identify and isolate the entity that we are calling ‘word’, that is, we must define what a word is in Chinese.

Defining the word Despite much discussion and speculation to the contrary, the definition of word in Chinese is relatively straightforward: a word in Chinese is an entity that may independently occupy a syntactic part-of-speech slot. This syntactic definition of word is driven by syntactic analysis, and is only one of several possible definitions (semantic, morphological, phonological, etc.; see Packard 2000) that could be used. It is, however, an ideal choice for defining ‘word’ because but it has eminent theoretical justification, given that words may be defined using the syntactic criterion in virtually all of the world’s languages. 216

Chinese morphology

This definition of ‘word’ based on syntactic form class criteria has the advantage of being independent of morphological structure, i.e. there are no morphological criteria per se that influence our decision regarding what is and is not to be considered a word. The arbitrariness that sometimes arises is in defining form class in a way that has sharp boundaries – a problem that occurs in virtually all languages. So even though part of speech identities such as Noun, Verb and Adjective (Stative Verb) sometimes overlap, they nonetheless generally lend themselves to clear definition, as do Prepositions (Coverbs), Adverbs, and Conjunctions. Some instances of potential arbitrariness do arise, as happens in any language. So, for example, the sentence-final aspect marker le that indicates change-of-state may seem difficult to assign to a form class category, until we consider the fact that there does exist in Chinese a sentence-final slot that functions to provide its occupant with a modification scope over the entire sentence. So such sentence-final particles as 了 le ‘aspect’, 啊 a ‘exclamation’, 吧 ba ‘suggestion’, 吗 ma ‘question’, 哇 wa ‘incredulity’, 呢 ne ‘nominal question’, 诶 ei ‘suspicion’ and 呀 ya ‘surprise’, each carry specific meanings, and those meanings have scope over the entire utterance by virtue of the fact that they occupy the sentence-final functional slot. In viewing the sentence-final position in Chinese defined in this way, all of these particles may be considered words. This syntactic definition of ‘word’ safely accounts for most of the data. Even the components of idiomatic four-morpheme phrases are accounted for in a relatively straightforward manner. For example, in modern usage 子 zi ‘son, child’ is a bound morpheme, because it cannot stand alone as a noun. But in the four-morpheme idiomatic phrase 望子成龙 wang-zi-cheng-long hope-child-become-dragon ‘have great expectations for one’s child’, zi appears free because it is has the identity of a noun in that phrase, and wang appears free because it is has the identity of a verb. In fact, zi and wang must still be considered bound because while they function as free elements in the limited, classical syntax of a fourmorpheme idiomatic phrase, according to the rules of modern syntax they are still unable to serve as a free words. In order to function as free words in modern syntax wang and zi must be further augmented with morphemic material. In modern syntax, in the case of zi, the proper free word is 子女 zinü son-daughter ‘children’ and in the case of wang the proper free word is the two-syllable 希望 xiwang hope-hope ‘hope’. Up to this point we have mostly discussed only simple Chinese words – those that are composed of single free morphemes. But in fact most words in Chinese are complex – that is, formed by combining morphemes – and those complex words may be conveniently categorized according to the nature of the components that compose them. So let us move on to a discussion of word component morphemes in Chinese.

Word component morphemes Having defined word in Chinese, it is important to consider the properties of Chinese word components before we consider how these components fit together to form the different word types.

Free content morphemes A free content morpheme is a morpheme that contains content rather than functional (grammatical) information, and is free, i.e. it may stand alone in a syntactic slot as a X0-level lexical item. Single-morpheme words are legion and non-mysterious in Chinese: as examples 217

Jerome L. Packard

we have the nouns 火 huo ‘fire’, 水 shui ‘water’ and 树 shu ‘tree’; the main verbs 跑 pao ‘run’, 开 kai ‘open’ and 打 da ‘hit’; the auxiliary verbs ken 肯 ‘willing to’, xiang 想 ‘think about’, 会 hui ‘can, will’, 能 neng ‘can’ and 要 yao ‘want to’; the adjectives (also known as stative verbs) 快 kuai ‘fast’, 高 gao ‘high’ and 瘦 shou ‘thin’; and the prepositions (also known as coverbs) 跟 gen ‘with’, 替 ti ‘for’ and 用 yong ‘with’.

Free function morphemes Free function morphemes are what are commonly known as function words. Some examples are 的 de ‘nominalization marker’, 才 cai ‘then’, 就 jiu ‘then’, 都 dou ‘all’, 还 hai ‘still’, 只 zhi ‘only’ and 也 ye ‘also’.

Bound content morphemes Bound roots Bound roots are morphemes that, like free content morphemes, contain content rather than functional (grammatical) information, but unlike free content morphemes are bound, and so must be augmented with additional morphemic information before they can occupy a syntactic slot. Some examples of bound roots in Chinese are 脑 nao ‘brain’, 房 fang ‘house’, 橡 xiang ‘rubber’, 镜 jing ‘mirror’, 裤 ku ‘pants’, 森 sen ‘forest’, 足 zu ‘foot’, 虎 hu ‘tiger’, 羽 yu ‘feather’ and 衣 yi ‘clothing’. Note that, unlike in some other languages, most bound roots in Chinese are quite flexible regarding which position they may occupy in a two-syllable word. In English, for example, bound roots are Greek and Latinate word stems that tend to be either the first or second morpheme in a word, as seen in examples such as pseudo– ‘false’, anti– ‘against’, neuro– ‘nerve’, oculo– ‘eye’, bio– ‘life’, –itis ‘inflammation’, –cide ‘killing’, –cyte ‘cell’, –ectomy ‘removal of’, and –gnosis ‘knowledge’. In Chinese, in contrast, a given bound root often can be either the first morpheme or the second morpheme in a word. So, for example, the bound root 脑 nao ‘brain’ is the first morpheme in the word 脑袋 naodai brain-bag ‘brain’, and it is the second morpheme in the word 电脑 diannao electric-brain ‘computer’, and the bound root 衣 yi ‘clothing’ is the first morpheme in the word 衣服 yifu clothingclothing ‘clothing’, and it is the second morpheme in the word 上衣 shangyi top-clothing ‘overcoat’.

Bound function morphemes A bound function morpheme is a morpheme that, like a function word, contains functional rather than content information but is bound, and so must be attached to another morpheme before it can appear in a syntactic form class slot. Bound function morphemes are also known as affixes, and come in two varieties: those that are not related to grammatical processes are derivational (or word-forming) affixes, and those that tend to signal grammatical processes are grammatical (or inflectional) affixes. This derivational–inflectional affix category distinction will be familiar to practitioners of linguistic morphology because of its broad traditional application to word formation processes across languages. The characteristic of being related to grammatical processes is only one of a cluster of properties that serves to distinguish derivational and inflectional affixes in Chinese. These are presented in the following discussion of the two affix types. 218

Chinese morphology

Derivational (word-forming) affixes Derivational affixes are bound function morphemes that have the following properties: ● ● ● ● ●

They can change the form class of morphemes or words to which they apply They apply selectively to only certain members of a category They have a relatively variable and unpredictable meaning across contexts They may attach either to free words or to bound roots They may attach to the right or left of a word or word component, i.e. they may be either prefixes or suffixes. ● They are unrelated to processes traditionally thought of as being ‘grammatical’, such as the marking of number, aspect or mood. The following are examples of Chinese derivational prefixes: 无 wu– ‘not’, 未 wei– ‘not’, 非 fei– ‘not’, 第 di– ‘ordinalizer’, 复 fu– ‘again’, 可 ke– ‘may’ and 再 zai– ‘again’. The following examples of derivational suffixes are the nominalizers 子 –zi, 儿 –er, 头 –tou, 性 –xing, and 度 –du; the verbalizing suffix 化 –hua, the adverbial suffix 然 –ran, and the agentive suffix 者 –zhe. While the difference in Chinese between derivational affixes and inflectional affixes (to be discussed below) is relatively straightforward and clear, the difference in the language between bound roots and derivational affixes is less clear. The difference hinges upon the distinction between content and function morphemes, which in the field of linguistic morphology is traditionally considered to involve a continuous rather than discrete difference, in other words a difference in degree rather than a difference in kind. For example, let us contrast several bound roots that generally entail the meaning ‘one who______’ (i.e. 客 –ke, 奴 –nu, 员 –yuan, 手 –shou, 民 –min, 霸 –ba, 精 –jing), with the derivational affix 者 –zhe which also means ‘one who______’, but is considered a derivational affix rather than a bound root because it is a productive suffix that marks grammatical case, i.e. it indicates agency. First, –zhe is more productive in the way seen in functional rather than content elements, able to suffix to virtually any verb (and some non-verbs as well: 第三 disan ordinal-three ‘the third’ > 第三者 disanzhe ordinal-three-AFF ‘the third party (in a relationship)’; 前 qian front ‘front’ > 前者 qianzhe front-AFF ‘the one in front’). The cited bound roots on the other hand generally co-occur with a more limited number of morphemes and so are more restricted in their productivity. Second, –zhe is usually restricted to simply mean ‘one who______’ (读者 duzhe read-one ‘reader’, 患者 huanzhe sick-one ‘patient’, 老者 laozhe old-one ‘one who is old’, and 编者 bianzhe edit-one ‘editor’). However, the bound roots (客 –ke, 奴 –nu, 员 –yuan, 手 –shou, 民 –min, 霸 –ba and 精 –jing) entail the meaning ‘one who______’, but also include an additional element of meaning, such as one who is adept at something, one who does something as a profession, or one who has achieved notoriety for doing something. As examples we have ‘one who (is adept at) ______’ (手 –shou: 射手 sheshou shoot-one ‘marksman’, 能手 nengshou able-one ‘expert’, 高手 gaoshou high-one ‘expert’, 水手 shuishou water-one ‘sailor’, and 鼓手 gushou drum-one ‘drummer’); ‘one who ______ (as a profession)’ (员 –yuan: 演员 yanyuan act-one ‘actor’, 教员 jiaoyuan teach-one ‘teacher’; 民 –min: 农民 nongmin farm-one ‘farmer’, 股民 gumin stock-one ‘investor’); ‘one who (is notorious for) ______’ (客 –ke: 政客 zhengke politics-one ‘political hack’, 刺客 cike stab-one ‘assassin’, 说客 shuoke speak-one ‘persuader’); ‘one who (is stuck doing, or is a “slave” to) ______’ (奴 –nu: 车奴 chenu car-one ‘one who is stuck making car payments’, 房奴 fangnu 219

Jerome L. Packard

house-one ‘one who is stuck making mortgage payments’); ‘one who (constantly does or is very good at) ______’ (霸 –ba: 学霸 xueba study-one ‘one who is very good at studying’, 面霸 mianba face-one ‘one who frequently goes on job interviews’, 投霸 touba submit-one ‘one who is a frequent job applicant’); and ‘one who (is infamous for) ______’ (精 –jing: 害人精 hairenjing harm-person-one ‘one who always harms others’, 马屁精 maipijing horserear-one ‘flatterer’ and 白骨精 baigujing white-bone-one ‘female who is successful, educated and wealthy’).

Inflectional (grammatical) affixes Inflectional affixes are bound function morphemes, like derivational affixes, but they have the following properties that distinguish them from the latter: ● They do not change the form class of morphemes or words to which they affix ● They do not apply selectively; that is, they apply more broadly to most members of a given category ● They have a relatively constant meaning across contexts ● They affix only to free words ● They affix only to the right edge of a word (or word component), i.e. in Chinese they may only be suffixes. ● They are related to processes considered to be ‘grammatical’, such as the marking of grammatical number, aspect or mood. The following are examples of inflectional suffixes in Chinese: 了 –le ‘completed aspect’, 过 –guo ‘experienced aspect’, 着 –zhe ‘continuing aspect’, 得/不 –de/–bu ‘irrealis potential’, 们 –men ‘human plural’, 得了/不了–deliao/–buliao ‘irrealis (in)ability’, and 得 –de ‘extent complementation’. These suffixes are general because they apply to virtually all members of their designated categories, they attach as suffixes to free words only, and they mark grammatical operations that contribute a relatively constant meaning.

Word types and morphological processes There are several word types in Chinese, composed using a variety of morphological processes. But as it turns out, listing the different word types in Chinese ends up being largely an exercise in taxonomy, because in Chinese a word’s type has little effect on its status as a word and on how it functions within the grammar. The fact that a word is what minimally occupies a form class slot is true whether it is simple or complex and regardless of the elements from which it is composed. So let us first examine the two categories of word, simple and complex, in Chinese, and the different morphological processes used to compose them.

Simple words A simple word is a word composed of a single morpheme. Since morphemes and syllables – as well as the orthographic units characters – are virtually coextensive in Chinese, simple words are usually single-syllable, monomorphemic one-character elements. Simple words can be nouns, main verbs, auxiliary verbs, adjectives, prepositions, or the various function words. 220

Chinese morphology

Complex words A complex word is a word composed of more than one morpheme. Complex words come in various forms, depending on the process used to form them and what the nature of their components is.

Compounding Compounding is the joining together of two preexisting words to make a new word, which is termed a compound. Here are some examples: 火山 huoshan fire-mountain ‘volcano’, 打破 dapo hit-break ‘break’, 冰河 binghe ice-river ‘glacier’, 马熊 maxiong horse-bear ‘brown bear’, 猫熊 maoxiong cat-bear ‘panda’, 冰山 bingshan ice-mountain ‘iceberg’, 马路 malu horse-road ‘street’, 走进 zoujin walk-in ‘walk in’, 进去 jinqu in-go ‘go in’, 动作 dongzuo move-do ‘activity’, 买办 manban buy-manage ‘comprador’, 买卖 maimai buy-sell, ‘business’, 操守 caoshou operate-preserve ‘integrity’, 传动 chuandong transmit-move ‘a (vehicle’s) transmission’. Chinese is erroneously thought to be a language in which the majority of complex words are compounds – in fact that distinction belongs to bound root words created via composition (see below).

Composition Composition is the joining of two word components, at least one of which is a bound root, to make a word. Therefore the type of word formed from the process of composition is called a bound root word. Bound root is the most common morpheme type in Chinese, and so it is not surprising that composition is the most common word-formation process in Chinese, and that the most common word type in Chinese is the bound root word.

Affixation Affixation involves the attachment of an affix. The two types of affix that exist in Chinese (see section 4.4 above) are derivational (or word-forming) affix and inflectional (or grammatical ) affix. Therefore, the two types of affixation that exist in Chinese are derivation and inflection, which yield the two types of affixed words derived words and grammatical words.

A.  Derived words Derived words are words formed by the addition of a word-forming affix to the left (prefix) or right boundary (suffix) of a word. The following examples include both prefixation and suffixation. 阿混 ahun AFF-muddle ‘muddle-head’, 插头 chatou insert-AFF ‘plug’ 电化 dianhua electricity-AFF ‘electrify’, 调子 diaozi tune-AFF ‘tune’, 房子 fangzi house-AFF ‘house’ 孵化 fuhua hatch-AFF ‘incubate’, 复查 fucha AFF-investigate ‘reinvestigate’, 复发 fufa AFF-occur ‘relapse’, 复活 fuhuo AFF-live ‘resurrect’, 复习 fuxi AFF-study ‘review’, 复现 fuxian AFF-appear ‘reappear’, 钢化 ganghua steel-AFF ‘temper’, 感化 ganhua feelAFF ‘sensitize’, 骨化 guhua bone-AFF ‘ossify’, 归化 guihua return-AFF ‘naturalize’, 开化 kaihua open-AFF ‘civilize’, 可体 keti AFF-body ‘fit’, 欧化 ouhua Europe-AFF ‘Europeanize’, 皮儿 pir skin-AFF ‘skin’, 汽化 qihua vapor-AFF ‘vaporize’, 劝化 quanhua persuade-AFF ‘convert’, 烧化 shaohua burn-AFF ‘burn’, 石化 shihua stone-AFF ‘petrify’, 退化 tuihua 221

Jerome L. Packard

retreat-AFF ‘degenerate’, 无视 wushi AFF-see ‘disregard’, 氧化 yanghua oxygen-AFF ‘oxidize’, 烟子 yanzi smoke-AFF ‘soot’, 羽化 yuhua feather-AFF ‘(of an adult insect) emerge’.

B. Grammatical words Grammatical (or inflected) words are words that are formed by the affixation of an inflectional suffix. Most grammatical suffixes in Chinese are considered word-external inflection because they attach to the right boundary of a simple or complex word. The 得/不 –de/–bu marker of ‘irrealis potential’ on the other hand is considered word-internal inflection because it attaches inside a word, on the right side of the first syllable (i.e. the head) of a compound result verb. Word-external inflection  Word-external inflection suffixes an inflectional affix to the right boundary of a simple or complex word. Some word-external inflectional affixes are: 了 –le ‘completed aspect’, 过 –guo ‘experienced aspect’, 着 –zhe ‘continuing aspect’, 们 –men ‘human plural’ and 得 –de ‘extent complementizer’. What these inflections have in common is that they all attach to the right edge of a word. To provide examples, the aspect marker 了 –le suffixes to the verb 吃 chi ‘eat’ to produce 吃了 chile ‘have eaten’; the marker of experience aspect 过 –guo suffixes to the verb 看 kan ‘see’ to produce 看过 kanguo ‘have seen before’; the marker of continuing aspect 着 –zhe suffixes to the verb 走 zou ‘walk’ to produce 走着 zouzhe ‘walking’. The grammatical marker of human plural 们 –men suffixes to the second-person pronoun 你 ni to produce 你们 nimen ‘you (pl)’. The 得 –de affix marking extent complementation suffixes to a verb, and serves to introduce a phrasal complement that indicates a measure of the extent of application of the verb to which it suffixes. So, for example, the active verb 走 zou ‘walk’ that takes the suffix –de yielding 走得 zoude may then be followed by an extent complement phrase such as 很快 hen kuai ‘very fast’ to give 走得很快 zoude hen kuai ‘walk very fast’. There is similarity between these ‘phrasal’ resultatives marked by a verb inflection and ‘lexical’ resultatives formed by word-internal inflection that is more apparent than real. This is discussed in the following section. Word-internal inflection In word-internal inflection, the ‘irrealis potential’ inflection markers 得 –de and 不 –bu suffix (some would say infix) internal to the word, to the right side of the first syllable (head) of a compound result verb, signaling the ability or inability of the potential result indicated by the verb to actually take place. So, for example, the compound result verb 找到 zhaodao seek-attain ‘find’ may take the –de potential inflection to yield 找得到 zhaodedao seek-AFF-attain ‘able to find’, or it may take the –bu potential inflection to yield 找不到 zhaobudao seek-AFF-attain ‘unable to find’. As mentioned above, ‘phrasal’ resultatives formed by a word-external verb inflection and ‘lexical’ resultatives formed by word-internal inflection may appear similar and in some cases may be phonologically and orthographically identical. They are, however, entirely distinct in terms of both their semantics and the process of their formation. For example, the simple verb 做 zuo means ‘do’, and may be suffixed with an extent complement (such as 很快 hen kuai ‘very fast’) as can any verb, as in 做得很快 zuode hen kuai do-AFF:extent-very fast ‘do quickly’. The verb zuo may also be suffixed with the extent complementizer 得 –de and the simple extent complement 好 hao ‘good, well’ to yield 做得好 zuodehao ‘do well’. 222

Chinese morphology

There also independently exists the compound result verb 做好 zuohao do-good ‘do well’ which, as a compound result verb, may affix the ‘irrealis potential’ inflection marker 得 – de to the right side of its first syllable to yield 做得好 zuodehao ‘able to do well’. Now, even though these two three-morpheme strings (one an internally-inflected word, and one the simple verb zuo inflected with the extent inflection –de followed by a complement of extent hao) that have the different meanings ‘do well’ and ‘able to do well’ are phonologically and orthographically identical (i.e. 做得好 zuodehao), they are differently derived and semantically distinct. This is best demonstrated by how the two different forms accommodate negation. The negative form of the lexical potential resultative 做得好 zuodehao is 做不好 zuobuhao, while the negative form of the phrasal extent resultative 做得好 is 做得不好 zuode bu hao. The meanings of the two forms are clearly different, with the negated phrasal extent resultative meaning ‘do poorly’ and the negated lexical resultative meaning ‘unable to do well’.

Reduplication Reduplication is the creation of a new word by copying all or part of another word. The word type formed from this process is called a reduplication, and the resulting meanings include diminution, superlative, intensification, enhanced description or adverbial modification. Verbs, adjectives, nouns and classifiers all may be reduplicated to form new words. Both simple and complex words may be reduplicated, and when it is complex words that are reduplicated, different parts of the word are duplicated depending on the part of speech and the meaning of the duplicated word.

A. Single-syllable reduplication When it is a simple verb that is reduplicated, the meaning is generally one of limitation or diminution of verb action, as in 走 zou walk ‘walk’ > 走走 zouzou walk-walk ‘walk a bit’, 看 kan look ‘look’ > 看看 kankan look-look ‘take a look’ and 吃 chi eat ‘eat’ > 吃吃 chichi eat-eat ‘eat a little’. Such simple verb reduplication often inserts the morpheme 一 yi ‘one’ between the reduplicated elements, as in e.g. 走 zou walk ‘walk’ > 走一走 zouyizou walkone-walk ‘walk a bit’. When it is a simple noun or classifier that is reduplicated, the meaning is one of pro­ liferation or enumeration, for example 人 ren person ‘person’ > 人人 renren person-person ‘every person’, 天 tian day ‘day’ > 天天 tiantian day-day ‘every day’ or 本 –ben one:volume ‘a volume’ > 本本 benben volume-volume ‘every volume’ or 个 –ge one:piece ‘a piece’ > 个个 gege piece-piece ‘every piece’. When it is an adjective (stative verb) that is reduplicated, it has the effect of intensification or manner adverbialization. When the reduplicated adjective indicates intensification, the nominal modifier 的 de usually occurs before the modified element, for example, 红 hong red ‘red’ > 红红的 honghongde red-red-MOD ‘really red’ and 高 gao high ‘high’ > 高高的 gaogaode high-high-MOD ‘really high’. When the reduplicated adjective indicates manner adverbialization, the adverbial suffix 地 –de is usually added to the reduplicated element, as in 快 kuai fast ‘fast’ > 快快地 kuaikuaide fast-fast-AFF ‘quickly’ or 好 hao good ‘good’ > 好好地 haohaode good-good-AFF ‘well’. In manner adverbial reduplication (especially in the Beijing dialect), the tone on the reduplicated syllable changes to high-level and a phonological [r] sound is added to the reduplicated syllable, as in 慢 màn slow ‘slow’ > 慢慢儿地 mànmānrde fast-fast-r-AFF ‘slowly’. 223

Jerome L. Packard

B. Bi-syllabic reduplication Stative verbs  Two-syllable stative verbs accomplish reduplication by first duplicating the first, then duplicating the second syllable in the form AB > AABB. The function of reduplication in these words is for the purpose of adverbial modification and intensification. For example, 高兴 gaoxing ‘happy’ > 高高兴兴 gaogaoxingxing ‘very happily’, 快活 kuaihuo ‘cheerful’ > 快快活活 kuaikuaihuohuo ‘very cheerfully’ and 无聊 wuliao ‘bored, boring’ > 无无聊聊 wuwuliaoliao ‘in a very boring manner’. Active verbs  Two-syllable active verbs accomplish reduplication by duplicating the entire two-syllable form as one unit, in the form AB > ABAB. The function of reduplication in these words is for diminution. For example, 研究 yanjiu ‘research’ > 研究研究 yanjiuyanjiu ‘research a bit’, 讨论 taolun ‘discuss’ 讨论讨论 > taoluntaolun ‘discuss a bit’, 商量 shangliang ‘consult’ > 商量商量 shangliangshangliang ‘consult a bit’.

Loan words Chinese has many loan words that are either phonetic loans or semantic loans (calques), or a combination of the two.

A. Phonetic loans The following are examples of phonetic loan words: 三明治 sanmingzhi ‘sandwich’, 保龄 baoling ‘bowling’, 克隆 kelong ‘clone’, 博客 boke ‘blog’, 卡 ka ‘card’, 卡通 katong ‘cartoon’, 可乐 kele ‘cola’, 吉他 jita ‘guitar’, 咖啡 kafei ‘coffee’, 土司 tusi ‘toast’, 坦克 tanke ‘tank’, 基因 jiyin ‘gene’, 尼龙 nilong ‘nylon’, 巧克力 qiaokeli ‘chocolate’, 布丁 buding ‘pudding’, 幽默 youmo ‘humor, funny’, 引擎 yinjing ‘engine’, 拷贝 kaobei ‘copy’, 朋克 pengke ‘punk’, 柠檬 ningmeng ‘lemon’, 沙发 shafa ‘sofa, upholstered chair’, 派 pai ‘pie’, 浪漫 langman ‘romantic’, 汉堡 hanbao ‘hamburger’, 爵士 jueshi ‘jazz’, 白兰地 bailandi ‘brandy’, 的士 dishi ‘taxi’, 维他命 weitamin ‘vitamin’, 荷尔蒙 heermeng ‘hormone’, 苏打 suda ‘soda’, 酷 ku ‘cool’, 雪茄 xuejia ‘cigar’, 雷达 leida ‘radar’, 马达 mada ‘motor’, 高尔夫 gaoerfu ‘golf ’, 麦克风 maikefeng ‘microphone’.

B. Semantic loans (calques) The following are examples of semantic loan words (calques): 黑板 heiban black-board ‘blackboard’, 电话 dianhua electric-talk ‘telephone’, 电视 dianshi electric-view ‘television’, 电脑 dianhua electric-brain ‘computer’, 键盘 jianpan key-plate ‘keyboard’, 软体 ruanti soft-body ‘software’, 蜜月 miyue honey-moon ‘honeymoon’, 硬体 yingti hard-body ‘(computer) hardware’, 硬盘 yingdie hard-saucer ‘hard drive’, 牛仔裤 niuzaiku cow-boy-pants ‘jeans’, 热狗 regou hot-dog ‘hotdog’, 摩天楼 motianlou brush-sky-building ‘skyscraper’, 忙音 mangyin busy-sound ‘busy signal’.

C. Phonetic-semantic loans (phonetic calques) The following are examples of words borrowed from other languages that are phonetically similar to the borrowed word and also bear a semantic resemblance to the borrowed word in its translated form: 播客 boke broadcast-guest ‘podcast’, 黑客 heike black-guest ‘hacker’, 224

Chinese morphology

拖拉机 tuolaji pull-pull-machine ‘tractor’, 博蒂 bodi abundant-fruit ‘birdie (golf )’, 脱口秀 tuokouxiu pull-mouth-display ‘talk show’, 利基 liji advantage-base ‘niche (market)’, 冰淇淋 bingqilin ice-cream-shower ‘ice cream’, 俱乐部 julebu utensil-happy-place ‘club’, 信用卡 xinyongka trust-use-card ‘credit card’.

Abbreviation An abbreviation is a shortened version of a longer word or phrase, with acronyms being a sub-type of abbreviation. The following are examples of abbreviation in Mandarin: 北大 beida north-big < 北京大学 Beijing Daxue north-capital big-study ‘Peking University’, 劳保 laobao work-protect < 劳动保险 laodong baoxian work-move protect-risk ‘labor insurance’, 文革 wenge culture-change < 文化大革命 wenhua da geming culture-script big change-mandate ‘Cultural Revolution’, 环保 huanbao surround-protect < 环境保护 huanjing baohu surround-border protect-protect ‘environmental protection’, 校车 xiaoche schoolvehicle < 学校公车 xuexiao gongche study-school public-vehicle ‘school bus’, 车险 chexian vehicle-risk < 汽车保险 qiche baoxian steam-vehicle protect-risk ‘car insurance’, 电大 dianda electric-big < 广播电视大学 guangbo dianshi daxue wide-broadcast electric-view big-study ‘TV University’. Abbreviations are not to be confused with contractions, which are phonologically shortened versions of spoken forms of words created by omission (and/or modification) of the individual sounds of words or phrases. For example, 别 bie ‘do not’ is a contraction of 不要 buyao not-want ‘do not’, 甭 beng ‘need not’ is a contraction of 不用 buyong not-use ‘need not’, and 孬 nao ‘bad’ is a contraction of 不好 buhao not-good ‘bad’. An acronym is an abbreviation that is formed by retaining only the initial morphemes of a word or phrase, for example 工农兵 gong-nong-bing work-farm-soldier < 工人, 农民, 兵员 gongren, nongmin, bingyuan work-person farm-person soldier-person ‘workers, peasants and soldiers’, 桌椅凳 zhuo-yi-deng table-chair-stool < 桌子, 椅子, 凳子 zhuozi, yizi, dengzi table-AFF chair-AFF stool-AFF ‘tables, chairs and stools’, 欧非美 Ou-Fei-Mei EuropeAfrica-America < 欧洲, 非洲, 美洲 Ouzhou, Feizhou, Meizhou Europe-continent Africacontinent, America-continent ‘Europe, Africa and America’.

Neologism A neologism is a new word in the language that is not a loan and not an abbreviation. The concept of ‘new word’ is always relative to a given time period – generally the neologisms cited below have been coined within the past 30 years: 八卦 bagua eight-symbol ‘rumored/ gossipy’, 富二代 fuerdai rich-two-generation ‘second-generation rich kids’, 飙车 biaoche storm-car ‘drag racing’, 爆料 baoliao explode-material ‘tip (news)’, 粗口 cukou crude-mouth ‘curse’, 发飚 fabiao emit-storm ‘have a fit, lose one’s temper’, 狗仔队 gouzaidui dogson-team ‘paparazzi’, 过山车 guoshanche pass-mountain-vehicle ‘roller coaster’, 挤提 jiti squeeze-take ‘run (on a bank)’, 闭卷 bijuan close-scroll ‘closed-book exam’, 冰毒 bingdu ice-poison ‘crystal meth’, 吊球 diaoqiu drop-ball ‘drop shot’, 火锅 huoguo fire-pot ‘slam-dunk’, 假唱 jiachang fake-sing ‘lip-synch’, 街舞 jiewu street-dance ‘hip-hop’.

Notes 1 I thank Yun Yao for help with several of the examples; I alone, however, am responsible for any errors of interpretation.

225

Jerome L. Packard 2 In terms of word ‘types’ versus ‘tokens’, meaning that if words are counted in a given page or paragraph, the total number of single-morpheme words may come close to the number of multi­ morphemic words, because many monomorphemic words (e.g. the modification marker de) have a very high frequency, a fact which serves to inflate the number of monomorphemic words if counted in terms of token frequency.

References Packard, Jerome L. (2000) The Morphology of Chinese, New York: Cambridge University Press. Packard, Jerome L. (2009) ‘Chinese as an Isolating Language’, in K. Brown and S. Ogilvey (eds) Concise Encyclopedia of Languages of the World, Oxford: Elsevier, 220–3. Sapir, Edward (1921/1949) Language, New York: Harcourt Brace Jovanovich (copyright renewed 1949).

226

14 Chinese neologisms: Word-formation strategies in Chinese Antonella Ceccagno University of Bologna, Italy

Introduction This paper presents some central word-formation strategies in Chinese by mainly focusing on compound words and word-formation patterns emerging as a result of the central role played by compounding in Chinese. In fact, while in some languages compounding is a peripheral word-formation strategy, in Chinese it is the most productive means of word formation. This emerges clearly from the corpus of neologisms developed over the last 30 years included in The Contemporary Chinese Dictionary (CCD) (2002) where out of 709 new words (those with no more than two syllables) almost 95% (672) are compounds, while only a little more than 2% (16) are derived words. Packard (1998) points out that a central change in the Chinese language has been the shift from monosyllabic to disyllabic words that started during the Zhou dynasty (1122–256 bc). The growing disyllabic nature of Chinese has engendered new processes of word formation, among which compounding stands as the most salient one. It has been shown that 70–80% of Chinese words are compounds (Zhou et al. 1999; Xing 2006). Compounding is a much discussed topic in linguistics as reflected in the publication of new works focusing on the questions posed by compounds in different languages (see for example the books edited by Lieber and Stekauer 2009; and by Scalise and Vogel 2010). Compounds can be described as ‘words that are composed of more than one meaningful element’ (Plag 2003). If the variety of constituents is taken into consideration, a compound can be described as a word formed by at least two elements – including phrases, roots, words, and stems. The formation of compound words in Chinese has been discussed recently by Huang (1998), Packard (2000), Ceccagno and Scalise (2006, 2007), Ceccagno (2007), and Ceccagno and Basciano (2007, 2009a, 2009b). After discussing the existing definitions for Chinese compounds, Ceccagno and Basciano (2009a: 67) argue that a compound word in Chinese is formed by at least two morphological words. With this definition, also words including bound roots – which are not syntactic words but are morphological words – are included as constituents of compound words. In fact, in Chinese the only difference between syntactic 227

Antonella Ceccagno

words and bound roots lies in the fact that bound roots cannot be used independently in a syntactic structure. The word-formation processes in Chinese discussed in this paper are (i) metacompounding, or compound words whose constituents are truncated forms of underlying compounds; (ii) neologisms coined from previous constituents of metacompounds; (iii) re-analysis of syllables from polysyllabic words, loanwords, calques and hybrids as morphemes, and the resulting new morphemes; (iv) monosyllabic neologisms coined as loanwords; (v) neologisms formed through the re-analysis of the semantics of the constituents of existing compound words, and (vi) neologisms coined on the Internet that intentionally aim at breaking the prevailing word-formation strategies.

Methodology Traditional linguistics used to analyze established words, which however often display various kinds of idiosyncrasies and therefore are not indicative of the speakers’ knowledge of the structure of possible words. This approach has also been adopted in the analysis of word formation in Chinese, where traditional forms were analyzed along with new ones (Huang 1998; Packard 2000; Ceccagno and Scalise 2006). Recently, however, linguists have highlighted the importance of focusing on neologisms in the analysis of the morphological competence of today’s speakers. Therefore neologisms are now more widely used as crucial data (Plag 2003 among others) because they also help to single out patterns in word formation in a given historical period. By focusing their research on Chinese neologisms, Ceccagno and Basciano (2007, 2009a, 2009b) have been able to unveil some outstanding word-formation patterns and to develop a new theoretical approach to the study of Chinese compound words. Accordingly, this chapter addresses some central word-formation patterns by focusing on neologisms. To do so it makes use of two different sets of data. The first source consists of 709 disyllabic new words presented in the CCD (2002) new words section. The dictionary does not provide information on when the new words were formed. However, some of the listed neologisms were already formed in the early 1980s, while most of them seem to have been coined afterwards. The second source is a list of 2,255 new words found on the website Shanghai Daily’s Well-Acclaimed Buzzwords (WABs) in 2013. Language change takes place through innovation and propagation. The new words from WABs (2013) are in the stage of innovation, that is their usage is still recent and therefore still subject to abandonment by the speaking community. In fact, new words or loans do not necessarily last for long. They can be abandoned and replaced by other neologisms. The success of the semantic innovation in fact largely depends on the social dignity of the newly coined word (for a sociolinguistic analysis, see Labov 1972 and Croft 2000). Instead, the new words listed in CCD (2002), exactly because they are listed in a dictionary, are in the stage of propagation – they are used by different groups of the linguistic community (on innovation and propagation, see Croft 2000).

Abbreviations The basic phonological unit – or ‘the standard foot’ in Feng’s terminology (1998) – in Modern Chinese is made up of two syllables. Given that in Classical Chinese lexical morphemes are mainly monosyllabic, the prosodic requirements are mainly met through the creation of disyllabic compound words1 (for a discussion see Feng 1998; Arcodia 2007). 228

Chinese neologisms: Word-formation strategies in Chinese

Abbreviation is quite a widespread pattern for coining new lexical items in Chinese. This is a neologism-forming strategy linked to the prosodic requirements of the language that also has morphologic and semantic implications (Ling 2000; Wang 2001; Yuan 2002; Ceccagno and Basciano 2009a; Basciano and Ceccagno 2009). The most common process of abbreviation of nouns occurs when two disyllabic words are abbreviated into one, with one syllable selected from the first word and one syllable from the second. When possible, abbreviations seek to maintain semantic transparency. To this end, in the abbreviation process, the selected lexical morphemes are those with the same meaning of the compound word of which they are the truncated form, or that are semantically close to it. Wang (2001) argues that the morphemes chosen for the abbreviated neologisms are those that express the core meaning, while Ceccagno and Basciano (2009a) define them as those that are ‘semantically relevant’. An example that illustrates the choice of semantically relevant constituents for the abbreviated compounds is 影评 yǐngpíng, the abbreviation for 电影评论 diànyǐng pínglùn ‘cinema + comment = film review’, (see Yuan 2002; XHSC 2002). The compound 评论 pínglùn and the morpheme 评 píng chosen as one of the two constituents for the abbreviated compound word share the meaning of ‘comment’, ‘to comment on’, while 论 lùn, which has been eliminated, means ‘to discuss’, ‘opinion’, ‘theory’. These meanings are less relevant from the semantic point of view. The same mechanism is evident in the choice of the morpheme 影 yǐng as the first constituent of the abbreviation. In fact, 影 yǐng shares with 电影 diànyǐng the meaning of ‘cinema’, while 电 diàn, meaning ‘electric’, ‘telegram’, ‘to send a telegram’ is not semantically relevant and therefore is not included in the abbreviation. Abbreviation according to semantic relevance is a word-forming strategy that, to a certain extent, taps into the diachronic evolution of the Chinese language and its constant tension between recursion and abbreviation (for a discussion, see Ceccagno and Basciano 2009a). In fact, the diachronic shift of Chinese towards a preference for disyllabic words instead of monosyllabic words has brought about the creation of new compounds with the same meaning of pre-existing monosyllabic words. This has created a dynamic relationship of cross-reference between monosyllabic lexical morphemes and compound words that makes it possible to abbreviate existing words by going back to the morphemes with the same meaning. Coordinate compounds2 whose constituents are in a relation of synonymy, such as, for example, 美丽 měilì ‘beautiful + beautiful = beautiful’ are the best examples of the crossreference between monosyllabic words and compounds. Therefore it could be argued that this subcategory, more than others, lends itself to the formation of new abbreviated compounds according to semantic relevance.

Metacompounds Only in a limited number of abbreviations by means of semantic relevance, however, is semantics entirely transparent. In many other cases, one or all the constituents of the new compound word, despite being selected as the most semantically relevant, do not have the exact meaning of the compound word of which they are an abbreviated form. This is the case of the example in (1). (1)  报纸摘要 bàozhǐ zhāiyào ‘newspaper + to make a summary = newspaper clippings’, abbreviation 报摘 bàozhāi 229

Antonella Ceccagno

In the abbreviated word 报摘 bàozhāi both the original form 报纸 bàozhǐ ‘report/newspaper + paper = newspaper’ and the constituent of the abbreviated compound 报 bào mean ‘newspaper’. Thus, this constituent of the new compound word is semantically transparent. However, 摘 zhāi means ‘to select’, ‘to pick (flowers)’ and therefore does not coincide with the meaning of the original form 摘要 zhāiyào ‘to summarize + main points = summary/make an abstract’. Here full semantic transparency is not achieved even though the two meanings are close. Besides, while 摘 zhāi is only a verb, 摘要 zhāiyào may also be a noun with the meaning of ‘abstract/summary’. Thus, this type of abbreviated compound can be fully understood only by referring back to the semantic and formal elements of the underlying compound. Ceccagno and Basciano (2007, 2009a) and Ceccagno (2009) argue that in the case of compounds in which at least one of the surface constituents is a truncated form of an underlying compound and for which the analysis of the surface form is problematic, the features of the underlying form are to be taken into consideration. They call these compounds ‘metacompounds’ and suggest analyzing them through a method called ‘metacompounding’. In examples (2) and (3), I present some compounds formed through abbreviation in which the semantics of the surface form can be clarified only by analyzing them as metacompounds. (2)

内退 nèituì ‘retire under the legal age (but be entitled to all privileges as a normal retired person)’ 退休 tuìxiū

内 nèi

In (2) the constituent 退 tuì of the metacompound has the meanings ‘to retreat’, ‘to move back’, ‘to withdraw’, ‘to cancel’, but does not have the explicit meaning of ‘retirement (from work)’, which is instead provided by the underlying compound 退休 tuìxiū. In (3) another example is presented in which the underlying compound clarifies the semantics of the metacompound. (3)

话网 huàwăng ‘telephone network’ [电话] [diànhuà] ‘telephone’

网 wăng ‘net’

The compound 话网 huàwǎng ‘words + net’ in (3) should be considered a metacompound because the semantics becomes transparent only if the constituent 话 huà is analyzed as a truncated form of the underlying compound 电话 diànhuà ‘electric + words = telephone’. In fact the CCD (2002) does not list ‘telephone’ as one of the meanings of 话 huà. Therefore, only after the constituent 话 huà is considered the truncated form of the underlying compound 电话 diànhuà, can the semantics of the entire metacompound be fully grasped (similar compounds which ‘inherit’ their semantics from the underlying forms are discussed in Myers 2006). 230

Chinese neologisms: Word-formation strategies in Chinese Table 14.1  期 qī as a compound constituent and as a metacompound constituent 期房

qīfáng

期价 [期货]价 期市 [期货]市

qījià [qīhuò] jià qīshì [qīhuò] shì

‘date/term + house = forward house; house to be completed and delivered for use at an agreed date on real estate market’ ‘date + price = futures price’ ‘futures + price’ ‘date + market = futures market’ ‘futures + market’

One morpheme can be just a constituent of a compound in some cases and the truncated form of an underlying compound in other cases, as shown in Table 14.1. In Table 14.1, 期房 qīfáng ‘date/term + house = forward house’ is a regular compound whose meaning is obtained compositionally by combining the two constituents 期 qī ‘date/ term’ and 房 fáng ‘house’. Instead, the other two neologisms are metacompounds. In fact, only by considering 期 qī as the truncated form of the underlying compound 期货 qīhuò ‘futures’, does the semantics of the entire compound become transparent. Therefore, a compound word is considered a metacompound when its constituents do not have among their meanings one meaning that compositionally contributes to the meaning of the surface compound and therefore it is only by analyzing the underlying form that the compound can be correctly interpreted. Ceccagno and Basciano (2007, 2009a) and Ceccagno (2009) also discuss examples in which the compound words present not only semantic opacity but also categorical opacity if they are not analyzed as truncated forms of underlying compounds. Both the constituents of the compounds in (4) are truncated forms of underlying compounds: (4)

[卫视]N wèishì ‘satellite TV’ [卫星]N wèixīng ‘satellite’ [卫]V wèi ‘defend’

[星]N xīng ‘star’

[电视]N diànshì ‘television’ [电]N [视]V diàn shì ‘electricity’ ‘to look at’

The surface compound 卫视 wèishì presents both semantic and categorial opacity, and it is not possible to identify the head. However, if the underlying compounds 卫星 wèixīng and 电视 diànshì are taken into consideration, the structure of the metacompound and its meaning become clear and the head can be identified. The input lexical categories are not [V+V] but rather [N+N]; the compound is not coordinate but attributive; the compositional meaning is ‘satellite + TV’ and not ‘defend + to look at’; and the head is on the right 231

Antonella Ceccagno

(on headedness, see Scalise and Fabregas 2010; on the positions of the head in Chinese compounds, see Ceccagno 2007, Ceccagno and Basciano 2007, 2009a). By adopting metacompounding as a method of analysis, all compound properties – semantics, lexical category and the head – can be identified. Therefore, a compound is considered a metacompound not only in case of semantic opacity but also when the compound structure and/or the position of the head are opaque, and only by analyzing the underlying form/s can they be disambiguated. Thus, a metacompound can be thought of as an iceberg with only its upper portion exposed, and whose morphology can only be explained by referring to the submerged parts. Out of 709 disyllabic neologisms listed in CCD (2002), 49 or almost 7% of the total are metacompounds. They are presented and analyzed in Ceccagno and Basciano (2009a). Metacompounding is only intended as a method for the synchronic analysis of opaque compounds that are the truncated forms of underlying compounds. Therefore, this method only focuses on the analyzed compounds and does not try to retrace the original forms. For instance, in the analysis of the compound 彩电 cǎidiàn ‘color television’ it will be enough to analyze the constituent 电 diàn as the truncated form of the underlying compound 电视 diànshì ‘electric + to look at = television’, without going back to the previous form 电视机 diànshìjī or to 电视接收机 diànshì jiēshōujī, which is the original one. As for the constituent 彩 cǎi, even though it is the abbreviated form of 彩色 cǎisè ‘color’, it should not be considered the truncated form of an underlying compound because one of the meanings of 彩 cǎi is ‘color’. Abbreviation is also used for another word-forming strategy. This is a neologism-forming process that involves loanwords, calques, and hybrids and it will be addressed.

Traditional word-formation patterns for loanwords and hybrids One much discussed feature of the Chinese lexicon is the very high correspondence between syllables, morphemes and written characters (Chao 1968, to name only one work on this topic). In Chinese the large majority of morphemes correspond to a single syllable. And conversely, the majority of syllables can be analyzed as morphemes, for example, 花 huā ‘flower’, 马 mǎ ‘horse’, 书 shū ‘book’. This peculiarity has an impact on what is perceived as a word by native speakers. Chung (2006) contends that both the morpheme and the syllable should be considered as the relevant morphological units playing a central role in the processes of word formation. DeFrancis (1990: 187) has coined the word morphosyllabic for describing the close correspondence between the phonologic syllabic unit and the morpheme in Chinese. Given that the large majority of Chinese morphemes have a lexical nature, Chinese polysyllabic words are mainly built starting from available monosyllabic units and in many cases they can be analyzed according to the meaning of each monosyllable, as in (5): (5) 生 shēng ‘to be born’, ‘to give birth’, ‘life’, ‘to grow’ 生物 shēng-wù ‘to be born + matter = living creature’ 生物层 shēngwù-céng ‘living creature + layer = biosphere’ Loanwords instead are often rendered by creating new polysyllabic but monomorphemic words (on loanwords in Chinese, see, for example, Shi 2000).

232

Chinese neologisms: Word-formation strategies in Chinese

Many loanwords have entered into the Chinese lexicon starting from the second half of the nineteenth century, when new concepts for which there was no Chinese equivalent were introduced in China (Masini 1993). However, some disyllabic morphemes, such as 葡萄 pútáo ‘grape’, 蝴蝶 húdié ‘butterfly’, and 玻璃 bōli ‘glass’ had been introduced in Chinese as loanwords much earlier. Loanwords are formed by dividing foreign words into syllables and associating the syllables to Chinese morphemes/characters that are phonologically close to the foreign word. For example ‘sofa’ is rendered in Chinese as 沙发 shāfā; the name ‘Alexander’ is rendered as 亚历山大 Yālìshāndà. In the specific context of loanwords – and only there – the morphemes/characters lose their original meaning. For instance, the morphemes/characters 亚历山大 yā + lì + shān + dà (‘Asia + to experience + mountain + large’) are used for rendering the word ‘Alexander’ only because they are phonologically close to the foreign word (but for an emerging different pattern see below). Even though the phonological form plays the central role in the formation of neologisms, the best loanwords are those where both sound and meaning remind of the foreign word. For instance, in the loanword 黑客 hēikè ‘hacker’ the two syllables provide the Chinese sound most similar to the sound in English, while the two morphemes/characters mean ‘black guest’, thus also contributing semantically to the neologism. In particular, the morpheme/ character 黑 hēi ‘black’ has been preferred to other homophones in order to strengthen the negative meaning of the loanword. A particular case of loanword formation can be observed in the word 咖啡 kāfēi ‘coffee’, where the strategy of creating new characters has been adopted. The radical 口 kǒu ‘mouth’, which provides vague semantic information, has been added to the pre-existing characters 加 jiā ‘add’ and 非 fēi ‘mistake’. The resulting new characters 咖 kā – which has a pronunciation similar to 加 jiā but not exactly the same – and 啡 fēi have being coined to form the new monomorphemic disyllabic word 咖啡 kāfēi ‘coffee’. Hybrids are coined by combining a loanword with a native element that provides semantic information on its taxonomical category. In Chinese the phonological part of the hybrid is tied to morphemes, such as 车 chē ‘vehicle’ and 酒 jiǔ ‘wine’, as shown in the examples in (6): (6) 摩托车 mótuō-chē ‘motor (loanword) + vehicle = motorcycle’ 酒吧 jiǔ-bā ‘wine + bar (loanword) = wine bar’

Re-analysis of syllables of loanwords and hybrids as morphemes As seen above, Chinese loanwords are rendered by creating polysyllabic and monomorphemic words. Ceccagno and Basciano (2009a, 2009b) highlight a neologism-formation strategy that consists of the re-analysis of syllables of loanwords as morphemes with the same meaning of the original polysyllabic morpheme. As a rule, these syllables should not be used independently, given the fact that they are not morphemes but syllables that in the context of the polysyllabic word are to be intended as devoid of meaning. In actual fact, syllables of loanwords are re-analyzed as morphemes – which often are bound – and become constituents of newly formed compound words. This takes place through a process of truncation. In (7), examples of compound words formed using this strategy are presented:

233

Antonella Ceccagno

 (7) 奶咖 nǎikā ‘milk + kā = caffèlatte’ 清咖 qīngkā ‘pure + kā = black coffee’ In the examples in (7), the syllable 咖 kā, of the morpheme 咖啡 kāfēi, is re-analyzed as a morpheme that encompasses the meaning of the polysyllabic morpheme. As such, 咖 kā is used as the constituent of new compound words with the meaning of ‘coffee’. Another example of a loanword syllable re-analyzed as a morpheme that takes on the meaning of the entire loanword is presented in (8):  (8) 红客  hóngkè ‘red + kè = red hacker’ Red hackers are: ‘Chinese hackers motivated by patriotism. They use their expertise to defend the security of domestic networks and fend off attacks’ (WABs). In this case the second constituent of the new word is the second syllable of the monomorphemic word 黑客 hēikè ‘hacker’, a loanword from English. This is therefore another example of a syllable from a polysyllabic loanword re-analyzed as a morpheme that represents the entire loanword and as such is used as a constituent in a new compound word. Here, as in the previous examples, the truncated form absorbs the meaning of the polysyllabic word and becomes a bound morpheme. In this case, the neologism plays with colors. In the newly formed word 红客 hóngkè ‘red hacker’, red as a color for ‘red China’ replaces the black color. This suggests that in this word-formation strategy also the syllable 黑 hēi of 黑客 hēikè ‘hacker’ is implicitly intended as a morpheme. In (9) another example of a syllable of a polysyllabic loanword analyzed as a morpheme is presented. 麦克风 màikèfēng is the Chinese loanword for the English word ‘microphone’. Among the clients of karaoke parties, the new compound word ‘microphone monopolizer’ has been coined in which 麦 mài, the first syllable of 麦克风 màikèfēng, is used as the first constituent of the compound with the meaning of ‘microphone’ (Shanghai Daily, 21/8/2006):  (9) 麦霸  màibà ‘mài + tyrant = microphone monopolizer’ The same pattern of re-analysis of syllables as bound morphemes presented above is used starting from hybrids, as in the example presented in (10): (10) 扎啤 zhāpí ‘to prick + pí = draft beer’ 淡啤 dànpí ‘light + pí = light beer’ In the neologisms 扎啤 zhāpí ‘draft beer’ and 淡啤 dànpí ‘light beer’, the syllable 啤 pí of the hybrid 啤酒 píjiǔ ‘beer’ is used as a morpheme that takes on the meaning of ‘beer’. The hybrid 酒吧 jiǔbā ‘alcohol + bā = bar’ has undergone a similar process. Starting from the 1990s, neologisms containing ‘bar’ only use 吧 bā (Cheng 2004), which in the original loanword provided the phonological part of the hybrid, as shown in (11): (11) 网吧 wǎngbā ‘net + bā = Internet bar’ 吧台 bātái ‘bā + platform = bar/pub counter’ He (2001, cited in Cheng 2004) lists 24 neologisms formed with 吧 bā. The strategy of forming neologisms through the re-analysis of syllables of loanwords and hybrids as morphemes and using them as constituents of new compounds seems to be 234

Chinese neologisms: Word-formation strategies in Chinese

linked to two main features of Chinese. The first is the strong tendency of Chinese towards disyllabism and the trend to abbreviate words in order to meet this requirement. The second is the central role played by the syllable in Chinese and the strong correspondence between the syllable/character and the morpheme, to the point that speakers of Chinese language tend to associate each syllable/character with a meaning. It could be argued that the re-analysis of syllables as morphemes is a word-forming strategy that tends to bring back the traditional correspondence between syllable, character and morpheme. In fact, previously existing morphemes that had been transformed into loanword syllables are given back the status of morphemes.

Breaking the established rules in neologism formation In some cases newly formed words and phrases seem to intentionally aim at breaking pre-existing word-formation rules. This is the case of the neologism in (12) coined as an homophone – but not a homograph – of the loanword 亚历山大 Yālìshāndà with the meaning of ‘Alexander’ presented above: (12) 压力山大  yālìshāndà ‘pressure + mountain + large = pressure heavy as a mountain’ 压力山大 yālìshāndà is ‘a new way for Chinese people to express the feeling of too much stress or pressure’ (Chinasmack 2013). A similar neologism forming process is shown in (13) where starting from the Chinese word for the famous Spanish painter Picasso, 毕加索 Bìjiāsuǒ, a new word is formed: (13) 闭家锁  bìjiāsuǒ ‘close + house + locked’ This neologism has been translated as ‘stay-at-home person’. It refers to young persons ‘staying at home to pursue obsessive interests, particularly Japanese video games, comic books and animation movies’ (WABs). In (12) and (13) the syllables are interpreted as morphemes/characters. What is more, the syllables are not connected with the corresponding characters/morphemes for ‘Alexander’ and ‘Picasso’ but with characters/morphemes that are homophones but allographs. The tones are the same. This new word-formation pattern has emerged in the written language and taps into one central feature of the Chinese language, i.e. the high number of homophones. The new words presented in (12) and (13) are by no means an isolated phenomenon. The neologism 谐音成语 xiéyīn chéngyǔ ‘homophonic idiom’ has been coined to describe the widespread use on the web and in advertising of characters that are homophones but written differently. Online it is trendy to write 女猪 nǚzhū ‘female + pig’ instead of and with the meaning of 女主 nǚzhǔ ‘female protagonist’ (WABs); 打铁 dǎtiě ‘hit + iron = to forge iron’ is often used instead of and with the meaning of 打贴 dǎtiē ‘writing a post’. In these last two examples, the character that replaces the correct one is not a homotone but only a homophone. This phenomenon may have a lasting impact on Chinese writing.

New morphemes and increased lexical ambiguity As shown above, Chinese exhibits a strong tendency towards abbreviation. As a result, some constituents of metacompounds that are the truncated form of an underlying compound may 235

Antonella Ceccagno

over time acquire the meaning of the underlying compound. 警 jǐng is one such word. Previously, it was an abbreviation for 警察 jǐngchá ‘police’ but now ‘police’ is one of its meanings. Therefore the compound in (14) should not be analyzed as a metacompound but as a regular compound given that there is no need to go back to underlying compounds when analyzing it. (14) 特警  tèjǐng ‘special + police = special police’ The same holds true for the morpheme 影 yǐng. Its original meanings were ‘shadow’, ‘image of an ancestor’, ‘to hide’, but after being used in the formation of compounds as the truncated form of the underlying compound 电影 diànyǐng ‘film’, ‘cinema’, it has absorbed the meaning of ‘cinema’.3 This is the reason why in the analysis of the compound 影评 yǐngpíng (above) the constituent 影 yǐng has not been considered the truncated form of the underlying compound 电影 diànyǐng. A diachronic analysis is out of the scope of this chapter. Here I only intend to highlight the fact that the process of abbreviation in the formation of metacompounds can diachronically assign to a constituent the same meaning of the underlying compounds of which it was the abbreviated form. This seems to be a widespread trend in Chinese, even though it is impossible to foresee which truncated forms will become lexical morphemes. As seen above, a word-forming strategy consists of re-analyzing syllables of polysyllabic but monomorphemic words as morphemes. This has more than one implication. First, syllables of polysyllabic words that till recently had only been used as syllables, as a result of the re-analysis strategy, become new bound morphemes. Compound neologisms such as those in (15) create previously nonexistent morphemes that can be used as constituents of new compound words: (15) 奶咖 nǎikā ‘milk + kā [咖啡 kāfēi ‘coffee’] = caffèlatte’ new morpheme: 咖 kā 葡糖 pútáng ‘pú [葡萄 pútáo ‘grape’] + sugar = glucose’ new morpheme: 葡 pú 淡啤 dànpí ‘light + pí [啤酒 píjiǔ ‘beer’] = light beer’ new morpheme: 啤 pí Second, in the case of loanwords whose syllables already existed as morphemes before the loanword was coined, the re-analysis strategy contributes to increase the lexical ambiguity of existing morphemes. The neologism 麦工 Màigōng ‘mài + job = McJob’ (WABs, 9/4/2007) is a calque from English, where the word ‘McJob’ has existed since the 1980s with the meaning of a job, such as the one offered in McDonald’s restaurants, that is low paying, often temporary, and offers minimal or no benefits and opportunity for promotion (NOAD 2005). In Chinese, ‘McDonald’s’ is rendered as 麦当劳 Màidāngláo, which is a polysyllabic and monomorphemic word. In the 麦工 Màigōng neologism, 麦 mài, the first syllable of the loanword, is re-analyzed as a morpheme with the meaning of ‘McDonald’s style’, with the negative connotation that ‘Mc’ for ‘McDonald’ has taken in English. Another example is the neologism 麦时尚 Màishíshàng ‘McFashion’, a calque from English that indicates those brands which offer designer-looking clothes for a small budget: ‘as fashion has begun to resemble fast food: fast and cheap’ (Shanghai Daily 9/4/2007). Therefore 麦 mài is a good example of the growing lexical ambiguity resulting from the word-forming strategies discussed in this chapter. The original meaning of the morpheme 麦 mài is ‘wheat’. With the re-analysis of the first syllable of a loanword as a new morpheme, presented in (9), 麦 mài also acquires the meaning of ‘microphone’; and the calque 236

Chinese neologisms: Word-formation strategies in Chinese

from English described above adds to 麦 mài the new meaning of ‘in the same style as McDonald’s’. Another word-forming pattern consists of coining monosyllabic loanwords. Given that existing morphemes are used for the new loanwords, this strategy also contributes to increased lexical ambiguity in Chinese. Examples of recently coined monosyllabic loanwords are presented in (16): (16) 秀 xiù ‘show’ (original meaning: ‘beautiful’, ‘elegant’) 酷 kù ‘cool’ (original meanings: ‘cruel’; ‘very’, ‘extremely’) 猫 māo ‘modem’ (original meaning: ‘cat’) This pattern also involves morphemes newly formed as a result of the re-analysis of syllables as morphemes. In (17) the newly formed bound morpheme 咖 kā has taken on the meaning of ‘cast’: (17) A咖  A kā ‘A + cast = leading actor’ According to WABs, the neologism A咖 refers to ‘a leading actor in a movie or drama’ and ‘frontline showbiz pop stars’. Therefore, as it happens with metacompounding, also the word-formation strategy that re-analyzes syllables as morphemes contributes to increased lexical ambiguity in Chinese. Lexical ambiguity is also the result of the creation of new monosyllabic loanwords.

Lexical ambiguity as a fishpond for new compound neologisms Another word formation pattern in Chinese consists of coining neologisms by selecting a different meaning for one or more constituents of existing compound words. As a general rule, in Chinese compounding helps reduce lexical ambiguity. In fact, the restrictions of meaning that the compound’s constituents impose on each other helps disambiguate the morphemes from the semantic point of view, exactly because only one of the possible meanings is chosen. Ceccagno and Basciano (2009a) have singled out a word-formation pattern in Chinese that taps into the lexical ambiguity of the constituents of existing compound words. Different meanings are chosen for the constituents and as a result a new compound word is created that is homonymous with the existing one, as shown in (18): (18) 透析  tòuxī  ‘dialysis’  new meaning: ‘penetrating analysis’ In the compound word 透析 tòuxī ‘dialysis’, the meaning ‘to pass through’ is chosen among the possible meanings of the first constituent 透 tòu, and the meaning ‘to separate, to divide’ is chosen for the second constituent 析 xī. But given that 透 tòu also has the meaning of ‘penetrating’ and 析 xī also means ‘to analyze’, a homonymous compound word has been coined, with the meaning of ‘penetrating analysis’. In both cases, 透析 tòuxī can be a noun and a verb. Another example of a compound neologism coined by tapping into the different meanings of the constituents is presented in (19): 超生  chāoshēng  (i) ‘reincarnation’; (ii) (fig.) ‘be lenient’  new meaning: (19)  ‘have more children than the family planning policy allows’, ‘have unplanned births’ 237

Antonella Ceccagno

In the previous Buddhist compound word 超生 chāoshēng, the compositional meaning of ‘reincarnation’ was obtained by selecting the meaning ‘to transcend’ for the first constituent 超 chāo4 and the meaning ‘life’ for the second, 生 shēng. In the newly formed compound, the meaning ‘to exceed’ is instead chosen for the first constituent 超 chāo and the meanings ‘to give birth/to be born’ for the second, 生 shēng. With these new meanings selected for the constituents, a new compound is formed with the meaning of ‘exceeding the number of births (allowed by the family planning policy)’. A recent neologism formed through the reinterpretation of the constituents of an existing compound is presented in (20): 月光  yuèguāng ‘moonlight’ new meaning: 月光 (族) yuèguāng(zú) (20)  ‘(people with) empty pockets at the end of the month’ In the previously existing compound 月光 yuèguāng the meaning ‘moonlight’ was formed compositionally with the first constituent 月 yuè meaning ‘moon’ and the second 光 guāng meaning ‘light/ray’. In the new compound word, instead, the meaning of ‘month’ is chosen for 月 yuè and the meaning of ‘empty’ is chosen for 光 guāng. 族 zú denotes a group of people who share common characteristics.5 Thus, 月光族 yuèguāngzú refers to people who have spent all their money by the end of the month (WABs). In (21) a case of semantic loan from English that engenders the reinterpretation of an existing compound word is presented: (21) 溜冰  liūbīng ‘slide + ice = skating’  new meaning: ‘to take drugs’ In English ‘ice’ is also the street name for ‘crystal methamphetamine hydrochloride’, which is a synthetic drug. The new meaning has emerged as a metonymy linked to the crystal form of the molecular structure of both ice and the drug. The Chinese language also adopts the English metonymy and thus adds a new meaning to the word 冰 bīng ‘ice’. With the reinterpretation of the second constituent of the compound according to the new meaning of ‘ice’ as a street name for a drug, the compound 溜冰 liūbīng takes on the new meaning of ‘to take ice’. This word-forming strategy seems to stem from and to build both on lexical ambiguity and on compounding as an ambiguity reducing word-forming strategy. In fact, on the one hand, morphemes’ lexical ambiguity facilitates the choice of new meanings for the constituents; on the other hand, the restrictions of meaning that the compound’s constituents impose on each other help disambiguate the morphemes from the semantic point of view. This makes it possible to coin new compound words with the same constituents of previously existing compounds. This word-forming strategy is different from the word-forming pattern through semantic change. Semantic change is used extensively as a word-forming strategy. For instance, in the new words listed in CCD (2002) about 50 new words or 7% of the total are formed through semantic change. In (22) a new word coined as a metaphor is presented: (22) 牛皮癣  niúpíxuǎn ‘psoriasis’ new meaning: ‘nagging problem, eyesore ads’ In this case, the previous compound word 牛皮癣 niúpíxuǎn ‘psoriasis’, a disease that affects the skin, has taken on the new meaning of ‘prolonged nagging problem’ and according to WABs (2013) this new meaning is by now also used to depict ‘the eyesore ads, such as 238

Chinese neologisms: Word-formation strategies in Chinese

illicit trashy ads posted or printed on walls, telecomm poles, door steps or even pavements, which are very difficult to get rid of ’. In the examples presented in (18), (19), (20), and (21), instead, the new meaning is chosen at the constituents level and, as argued above, it taps into the lexical ambiguity of the constituents and relies on compounding as a means for disambiguating the meaning of the constituents.

Conclusions This chapter has presented some crucial word-forming strategies in Chinese; namely, metacompounding, the re-analysis of syllables as morphemes, neologisms formed by selecting a new meaning for the constituents of existing compound words, and the creation of new morphemes or new meanings for existing morphemes. Most of these word-forming strategies have been singled out only recently. Moreover, neologisms coined on the Internet that intentionally aim at breaking the prevailing word-formation strategies have also been discussed. The examples that illustrate these strategies have been drawn from a rich corpus that includes (i) neologisms formed in the last decades, found in the neologism section of CCD (2002); (ii) recent and very recent neologisms found on the WABs website (2013). With the exception of monosyllabic loanwords, the strategies presented in this chapter tend to meet the strong preference of Chinese for disyllabism. This chapter has shown that both metacompounding and the re-analysis of syllables as morphemes are word-forming strategies that create new compound words through truncation. In metacompounds at least one constituent is the truncated form of an underlying compound word; in the re-analysis of syllables as morphemes a constituent of a new compound word is a loanword’s syllable transformed into a bound morpheme. In both cases, the prosodic requirements of the language cause constituents of compounds to acquire a new meaning from an underlying item. Most of the word-forming patterns discussed in this paper – i.e. metacompounding, the re-analysis of syllables as morphemes, the selection of new meanings for the constituents of existing compounds, and the coining of monosyllabic loanwords – seem to contribute to increased lexical ambiguity. This suggests that prosodic trends play a central role in word formation in Chinese, to the point that they tend to override semantic–pragmatic considerations (see also Arcodia and Montermini 2012; and Duanmu 2012). Instead, recent word-forming patterns for words coined on the Internet – and therefore presented in the written form – seem to intentionally violate each and every pre-existing word-forming strategy.

Notes 1 As pointed out by Arcodia (2007), disyllabification and compounding are two distinct phenomena, albeit related and interacting with each other. 2 Coordinate compounds are those whose constituents exhibit a semantic relation of synonymy, as in the case of 美丽 měilì, or a logical coordination (‘and’), as in 新锐 xīnruì ‘new + sharp = new and sharp’, or antonymy, as in 呼吸 hūxī ‘exhale + inhale = breath’, redundancy as in 苹果 píngguǒ ‘apple + fruit = apple’, or reduplication, as in 天天 tiāntiān ‘day + day = every day’. A classification scheme that identifies three macro-types in compounding (subordinate, attributive, and coordinate) has been argued for by Bisetto and Scalise (2005). On coordinate compounds, see also Arcodia et al. (2010). For the classification of Chinese compounds, see Ceccagno and Scalise (2007) and Ceccagno and Basciano (2007).

239

Antonella Ceccagno 3 The Mathews’ Chinese–English Dictionary (MCED 1972) does not list the meaning of ‘cinema’ for 影 yǐng. However, Le Dictionnaire Français de la Langue Chinoise (DFLC 1976) does list it with the meaning of ‘cinema’, as an abbreviation. 4 Recently 超 chāo has also been used with the meaning of ‘ultra’, as can be seen in the word 超声 波 chāoshēngbō ‘ultra + sound + wave = ultrasound’. 5 族 zú is an example of semantic change. It is a bound morpheme originally meaning ‘clan’, ‘race’, ‘nationality’, ‘ethnicity’ that through semantic change has taken on the meaning of ‘a group of people who share common characteristics’. Shen (1995) and Ceccagno and Basciano (2009a) consider 族 zú as a lexeme undergoing grammaticalization that is in the process of becoming an affix.

References Arcodia, Giorgio Francesco (2007) ‘Chinese: A Language of Compound Words?’, in Fabio Montermini, Gilles Boyé, and Nabil Hathout (eds) Selected Proceedings of the 5th Décembrettes: Morphology in Toulouse, Somerville, MA: Cascadilla Proceedings Project, 79–90. Available at: http://www. lingref.com/ (document #1617). Arcodia, Giorgio Francesco, Nicola Grandi, and Bernhard Wälchli (2010) ‘Coordination in Compounding’, in Sergio Scalise and Irene Vogel (eds) Cross-disciplinary Issues in Compounding, Amsterdam and Philadelphia: John Benjamins, 177–98. Arcodia, Giorgio Francesco and Fabio Montermini (2012) ‘Are Reduced Compounds Compounds? Morphological and Prosodic Properties of Reduced Compounds in Russian and Mandarin Chinese’, in Vincent Renner, François Maniez, and Pierre J. L. Arnaud (eds) Cross-disciplinary Perspectives on Lexical Blending, Berlin-New York: Mouton de Gruyter, 93–113. Basciano, Bianca and Antonella Ceccagno (2009) ‘The Chinese Language and Some Notions From Western Linguistics’, Lingue e Linguaggio 8(1): 105–35. Bisetto, Antonietta and Sergio Scalise (2005) ‘Classification of Compounds’, Lingue e Linguaggio 2: 319–32. CCD (2002)《现代汉语词典》(The Contemporary Chinese Dictionary), Beijing: Foreign Language Teaching and Research Press. Ceccagno, Antonella (2007) ‘Regolarità e peculiarità dei composti del cinese’, in Laura De Giorgi and Guido Samarani (eds) Percorsi della civiltà cinese tra passato e presente, Venezia: Cafoscarina, 149–66. Ceccagno, Antonella (2009) ‘Metacompounds in Chinese’, Lingue e Linguaggio 2: 195–212. Ceccagno, Antonella and Bianca Basciano (2007) ‘Compound Headedness in Chinese: An Analysis of Neologisms’, Morphology 2: 207–31. Ceccagno, Antonella and Bianca Basciano (2009a) Shuobuchulai. La formazione delle parole in cinese, Bologna: Serendipità. Ceccagno, Antonella and Bianca Basciano (2009b) ‘Sino-Tibetan: Mandarin Chinese’, in Rochelle Lieber and Pavol Stekauer (eds) The Oxford Handbook of Compounding, Oxford: Oxford University Press, 478–90. Ceccagno, Antonella and Sergio Scalise (2006) ‘Classification, Structure and Headedness of Chinese Compounds’, Lingue e Linguaggio 2: 233–60. Ceccagno, Antonella and Sergio Scalise (2007) ‘Composti del cinese: analisi delle strutture e identificazione della testa’, in Annamaria Palermo (ed) La Cina e l’altro, Napoli: Università degli studi di Napoli ‘L’Orientale’, 503–43. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley-Los Angeles: University of California Press. Cheng, Lixia (2004)〈语言接触,类推与形态化〉(‘Linguistic Contact, Analogy and Morphologization’), Waiyu yu waiyu jiaoxue 4: 53–6. Chinasmack, available at: http://advertising.chinasmack.com/2013/top-10-chinese-buzzwords-of-thelast-year.html. Chung, Karen (2006) Mandarin Compound Verbs, Taiwan Journal of Linguistics: Book Series in Chinese Linguistics, Taipei: Crane Publishing. Croft, William (2000) Explaining Language Change: An Evolutionary Approach, Harlow, Essex: Longman. DeFrancis, John (1984/1990) The Chinese Language: Facts and Fantasy, Honolulu: University of Hawaii Press; Taipei: Crane Publishing (reprinted).

240

Chinese neologisms: Word-formation strategies in Chinese DFLC (1976) Dictionnaire Français de la Langue Chinoise, Paris-Taipei: Insitut Ricci-Kuangchi Press. Duanmu, San (2012) ‘Word-length Preference in Chinese: A Corpus Study’, Journal of East Asian Linguists 21: 89–114. Feng, Shengli (1998) ‘Prosodic Structure and Compound Words in Classical Chinese’ in Jerome L. Packard (ed) New Approaches to Chinese Word Formation, Berlin-New York: Mouton de Gruyter, 196–259. Huang, Shuanfan (1998) ‘Chinese as a Headless Language’, in Jerome L. Packard (ed) New Approaches to Chinese Word Formation, Berlin-New York: Mouton de Gruyter, 261–83. Labov, William (1972) Sociolinguistic Patterns, Philadelphia, PA: University of Pennsylvania Press. Lieber, Rochelle and Pavol Stekauer (eds) (2009) The Oxford Handbook of Compounding, Oxford: Oxford University Press, 478–90. Ling, Yuanzheng (2000)《现代汉语缩略语》(Abbreviations in Contemporary Chinese), Beijing: Yuwen Chubanshe. Masini, Federico (1993) ‘The Formation of Modern Chinese Lexicon and its Evolution Towards a National Language: The Period from 1840 to 1898’ (English translation), Journal of Chinese Linguistics Monograph Series 6. MCED (1972) Mathews’ Chinese–English Dictionary, Cambridge, MA: Harvard University Press. Myers, James (2006) ‘Processing Chinese Compounds: A Survey of the Literature’, in Gary Libben and Gonia Jarema (eds) The Representation and Processing of Compound Words, Oxford: Oxford University Press, 169–96. NOAD (2005) The New Oxford American Dictionary (2nd ed.), Oxford: Oxford University Press. Packard, Jerome L. (1998) ‘Introduction’, in Jerome L. Packard (ed) New Approaches to Chinese Word Formation, Berlin-New York: Mouton de Gruyter, 3–34. Packard, Jerome L. (2000) The Morphology of Chinese, Cambridge: Cambridge University Press. Plag, Ingo (2003) Word-Formation in English, Cambridge: Cambridge University Press. Scalise, Sergio and Antonio Fabregas (2010) ‘The Head in Compounding’, in Sergio Scalise and Irene Vogel (eds) Cross-disciplinary Issues in Compounding, Amsterdam and Philadelphia: John Benjamins, 109–26. Scalise, Sergio and Irene Vogel (eds) (2010) Cross-disciplinary Issues in Compounding, Amsterdam and Philadelphia: John Benjamins. Shen, Mengying (1995)〈试论新词缀化的汉民族性〉(‘On the Chineseness of the New Affixation Trend’),《南京师大学报》(Journal of the Nanjing Normal University) 1: 35–41. Shi, Youwei (2000)《汉语外来词》(Loanwords in Chinese), Beijing: The Commercial Press. WABs, Shanghai Daily’s Well-acclaimed buzzwords, available at: http://buzzword.shanghaidaily.com/ Wang, Jihui (2001)《现代汉语缩略语词研究》(Research on Abbreviations in Contemporary Chinese), Tianjin: Tianjin Renmin Chubanshe. XHSC (2002)《现代汉语缩略语词典》(Dictionary of Abbreviations in Contemporary Chinese), Beijing: Yuwen chubanshe. Xing, Janet Zhiqun (2006) Teaching and Learning Chinese As a Foreign Language: A Pedagogical Grammar, Hong Kong: Hong Kong University Press. Yuan, Hui (2002) ‘Introduction’, in Hui Yuan and Ruan Xianzhong (eds)《汉语缩略语词典》(Dictionary of Abbreviations in Contemporary Chinese), Beijing: Yuwen Chubanshe. Zhou, Xiaolin, William Marslen-Wilson, Marcus Taft, and Shu Hua (1999) ‘Morphology, Ortography and Phonology in Reading Chinese Compound Words’, Language and Cognitive Processes 14(5–6): 525–65.

241

15 Chinese: Parts of Speech Candice Chi-Hang Cheung The Hong Kong Polytechnic University, Hong Kong, China

1. Introduction Whether Chinese has the same range of parts of speech (or categories) as the Indo-European languages has been of much discussion in the literature. In particular, while it is generally recognized that Chinese makes a distinction between nouns and verbs, scholars hold different opinions regarding the rest of the categories (see Chao 1968; C. N. Li and Thompson 1981; D. Zhu 1982; and Xing and Ma 1992, inter alia). The differences in opinions are partly due to the different theoretical backgrounds adopted by linguists, and partly due to the use of different terminological conventions. As a result, different criteria are employed for classifying words, and different terminological conventions are used to label the categories. In order to address the question of whether Chinese possesses the same range of categories as the Indo-European languages, I shall make use of the familiar categories of the IndoEuropean languages whenever possible. Before we proceed, it is important to mention that this chapter is intended to offer a comprehensive survey of the major categories in Chinese, aiming to establish the set of categories that is commonly found in Chinese and the Indo-European languages (section 2) and those that are available in Chinese only (section 3). In particular, in the discussion of the major categories shared by Chinese and the Indo-European langugages, I shall examine their characteristic features and discuss in what ways they are similar to and different from those in the Indo-European languages. Furthermore, I shall review the factors that contribute to the long-standing debate on the categorial status of adjectives, prepositions and localizers.

2.  Categories commonly found in Chinese and the Indo-European languages This section introduces the categories that are commonly found in Chinese and the IndoEuropean languages, including nouns, verbs, adjectives, adverbs, prepositions and conjunctions.

2.1. Nouns The existence of nouns as an independent category in Chinese is uncontroversial (Chao 1968, D. Zhu 1982, C.-T. J. Huang, Li and Li 2009: Ch. 1). One of the characteristics of nouns is that they can be preceded by a [Numeral + Classifier] sequence (Chao 1968, D. Zhu 1982), 242

Chinese: Parts of speech

as shown in the following examples.1 In this regard, Chinese nouns behave like mass nouns in the Indo-European languages, as they always require the presence of classifiers in enumeration (see Chierchia 1998 and Cheng and Sybesma 1999 for discussion of the availability of count-mass distinction in Chinese).2,3 (1) (a)

一个 人 yī-ge rén one-Cl person ‘one person’

(b) 三本 书 sān-běn shū three-Cl book ‘three books’

 

两杯 啤酒 liǎng-bēi píjiǔ two-glass beer ‘two glasses of beer’

(d) 四磅 猪肉 sì-bàng zhūròu four-pound pork ‘four pounds of pork’

(c)

Another characteristic of nouns is that they cannot be negated by bù 不 ‘not’ (Chao 1968, D. Zhu 1982 and C.-T. J. Huang, Li and Li 2009: Ch. 1) nor can they be modified by the intensifier hěn 很 ‘very’ (Chao 1968, D. Zhu 1982), as evidenced by the ill-formedness (indicated by ‘*’ below and throughout) of (2) and (3), respectively.4 (2) (a)

*不 *bù not

人 rén person

(b) *不 *bù not

书 shū book

 

*不 *bù not

啤酒 píjiǔ beer

(d) *不 *bù not

猪肉 zhūròu pork

(c)

(3) (a)

*很 *hěn very

人 rén person

(b) *很 *hěn very

书 shū book

 

*很 *hén very

啤酒 píjiǔ beer

(d) *很 *hěn very

猪肉 zhūròu pork

(c)

Apart from the fact that nouns in Chinese behave like mass nouns in the Indo-European languages, Chinese nouns differ from those in the Indo-European languages in that they are not inflected for number, gender and case. This is due to the fact that Chinese is an isolating language and is morphologically impoverished. Furthermore, unlike Indo-European languages, Chinese lacks definite and indefinite articles. Thus, the same form of a noun such as xiǎohái 小孩 ‘child’ may appear in the subject or object position. In terms of interpretational properties, it can be interpreted as singular or plural and generic, definite or indefinite, depending on the contexts (Chierchia 1998, Cheng and Sybesma 1999, Y.-H. A. Li 1999):5 (4) (a)

小孩 喜欢 猫。 xiǎohái xǐhuān māo child like cat (i)   ‘Children like cats’  (generic) (ii)  ‘The child/children like(s) cats’  (definite) 243

Candice Chi-Hang Cheung

(b) 他 会 去 找 小孩。 tā huì qù zháo xiǎohái he will go find child (i)  ‘He will go find a/some child/children’  (indefinite) (ii)  ‘He will go find the child/children’  (definite)

2.1.1.  Classification of nouns Based on the semantic properties of nouns and the types of classifiers being used, we can divide the class of nouns into four subclasses, namely, count nouns, mass nouns, collective nouns and abstract nouns (D. Zhu 1982; see also Chao 1968).

2.1.1.1.  Count nouns Count nouns refer to the type of noun that denotes tangible and discrete entities that are countable. Typical examples of count nouns include kèrén 客人 ‘guest’, shū 书 ‘book’, shé 蛇 ‘snake’, huā 花 ‘flower’, among many others. Unlike count nouns in the Indo-European languages, which can be directly combined with a numeral (e.g. one guest, three books, etc.), count nouns in Chinese require the use of sortal classifiers, such as wèi 位, běn 本, tiáo 条, etc., which encode some distinctive features like shape, as illustrated below (D. Zhu 1982; see also Chao 1968): (5) (a)

一位 客人 yī-wèi kèrén one-Cl guest ‘one guest’

(b) 三本 书 sān-běn shū three-Cl book ‘three books’

 

两条 蛇 liǎng-tiáo shé two-Cl snake ‘two snakes’

(d) 十朵 花 shí-duō huā ten-Cl flower ‘ten flowers’

(c)

For instance, the sortal classifier tiáo 条 in (5c) classifies long, cylindrical and flexible objects (H.-Y. J. Tai and Wang 1990), and thus it can serve as the classifier of shé 蛇 ‘snake’ and entities with similar characteristics: (6) (a)

一条 绳子 yī-tiáo shénzi one-Cl rope ‘one rope’

(b) 三条 鱼 sān-tiáo yú three-Cl fish ‘three fish’

Furthermore, a subset of count nouns designating human beings can be suffixed by the collective marker -men 们 to refer to a group of people that are known to the speakers and hearers (Iljic 1994, 2001, Y.-H. A. Li 1998, among others). This can be witnessed by the fact that a count noun suffixed with -men 们 is interpreted as definite, as in (7a), in contrast with the bare count noun, which can be interpreted as indefinite or definite, as in (7b):

244

Chinese: Parts of speech

(7) (a)

他 tā he ‘He

(b) 他 tā he (i)   (ii) 

会 去 找 孩子们。 huì qù zhǎo háizi-men. will go find child-MEN will go find the children’  (definite) 会 去 找 孩子。 huì qù zhǎo háizi. will go find child ‘He will go find a/some child/children’  (indefinite) ‘He will go find the child/children’  (definite)

However, unlike genuine plural markers in the Indo-European languages, count nouns suffixed with -men 们 cannot be preceded by a [Numeral + Classifier] sequence even when the numeral is greater than ‘one’ (Iljic 1994, 2001; Y.-H. A. Li 1998, among others), whereas a bare count noun can, as illustrated by the contrast between (8a) and (8b). (8) (a)

*他 *tā he

会 huì will

(b) 他 会 tā huì he will ‘He will go

去 qù go

找 zhǎo find

三个 sān-ge three-Cl

去 qù go find

找 三个 zhǎo sān-ge find three-Cl three children’

孩子们。 háizi-men child-MEN 孩子。 háizi. child

2.1.1.2.  Mass nouns In contrast with count nouns, mass nouns refer to the type of noun that denotes entities that are not countable. Typical examples of mass nouns include shuǐ 水 ‘water’, jiǔ 酒 ‘wine’, táng 糖 ‘sugar’, qìyóu 汽油 ‘petrol’, bù 布 ‘cloth’, among many others. As noted by Chao (1968), there are four ways of classifying mass nouns. One way is to make use of standard measure words such as chǐ 尺 ‘foot’, bàng 磅 ‘pound’, jiālún 加仑 ‘gallon’, gōngjīn 公斤 ‘kilo’, as illustrated below: (9) (a)

一两 yī-liǎng one-ounce ‘one ounce

 

两加仑 liǎng-jiālún two-gallon ‘two gallons

(c)

银子 yínzi silver of silver’ 汽油 qìyóu petrol of petrol’

(b) 三磅 sān-bàng three-pound ‘three pounds (d) 十公斤 shí-gōngjīn ten-kilo ‘ten kilos of

糖 táng sugar of sugar’

米 mǐ rice rice’

Another way is to make use of container measure words such as bēi 杯 ‘glass’, píng 瓶 ‘bottle’, etc. or temporary measure words such as shēn 身 ‘body’, wūzi 屋子 ‘room’, etc. that indicate the temporary location of the mass:

245

Candice Chi-Hang Cheung

(10)

(a)

一杯 yī-bēi one-glass ‘one glass

水 shuǐ water of water’

(b) 三瓶 sān-píng three-bottle ‘three bottles

(c)

一身 yī-shēn one-body ‘a body of

雪 xuě snow snow’

(d) 一屋子 yī-wūzi one-house ‘a houseful

酒 jiǔ wine of wine’

烟 yān smoke of smoke’

A third way is to employ indeterminate number/amount measure words such as yīdiǎn 一点 ‘a little’ and yīxiē 一些 ‘some’: (11)

(a)

一点 yīdiǎn a little ‘a little

(b) 一些 yīxiē some ‘some

水 shuǐ water water’

酒 jiǔ wine wine’

A fourth way is to employ measure words that describe the shape in which the mass can be gathered. Examples of this type of measure word include duī 堆 ‘heap’, tán 滩 ‘pool’, etc., as shown below: (12)

(a)

两堆 liǎng-duī two-heap ‘two heaps

泥 ní earth of earth’

(b) 一滩 yī-tán one-pool ‘one pool

水 shuǐ water of water’

2.1.1.3.  Collective nouns Collective nouns are inherently plural. Some illustrative examples are given below: (13)

(14)

(a)

父母 fù-mǔ father-mother ‘parents’

(b) 饭菜 fàn-cài rice-vegetable ‘food’

(c)

子女 zí-nǚ son-daughter ‘children’

(d) 亲友 qīn-yǒu relative-friend ‘friends and relatives’

(a)

纸张 zhǐzhāng paper ‘paper’

(b) 花朵 huāduǒ flower ‘flowers’

(c)

马匹 mápǐ horse ‘horses’

(d) 书本 shūběn book ‘books’ 246

Chinese: Parts of speech

The examples in (13) involve juxtaposition of two subordinate terms, whereas the examples in (14) are seemingly composed of a noun and a sortal classifier, as evidenced by the fact that the second element in the collective nouns in (14a–d) can be used as a sortal classifier, as shown below. The fact that the second member can be used as a sortal classifier in Modern Chinese is not surprising, as many sortal classifiers have their historical origins as nouns (Wang 1955, Liu Shiru 1965, among others): (15)

(a)

一张 纸 yī-zhāng zhǐ one-Cl paper ‘a piece of paper’

(b) 三朵 花 sān-duǒ huā three-Cl flower ‘three flowers’

(c)

两匹 马 liáng-pí mǎ two-Cl horse ‘two horses’

(d) 四本 书 sì-běn shū four-Cl book ‘four books’

Due to the fact that collective nouns are inherently plural, they cannot be preceded by a [Numeral + Classifier] sequence, as evidenced by the ill-formedness of the following examples (Chao 1968, D. Zhu 1982): (16)

(a)

*两个 *liǎng-ge two-Cl Intended:

(c)

*一张 纸张 *yī-zhāng zhǐzhāng one-Cl paper Intended: ‘one piece of paper’

父母 fù-mǔ father-mother ‘two parents’

(b) *三碟 *sān-dié three-Cl Intended:

饭菜 fàn-cài rice-vegetable ‘three plates of food’

(d) *四朵 *sì-duǒ four-Cl Intended:

花朵 huāduǒ flower ‘four flowers’

Instead, collective nouns can only take measure words concerned with grouping (e.g. duì 对 ‘pair’ and pī 批 ‘batch’), temporal measure words concerned with location (e.g. zhuō 桌 ‘table’) or indeterminate measure words such as yīxiē 一些 ‘some’, as demonstrated below: (17)

(a)

两对 liǎng-duì two-pair ‘two pairs

(c)

一批 yī-pī one-batch ‘a batch of

父母 fù-mǔ father-mother of parents’ 纸张 zhǐzhāng paper papers’

(b) 一桌 yī-zhuō one-table ‘a table of (d) 一些 yīxiē some ‘some

饭菜 fàn-cài rice-vegetable food’

花朵 huāduǒ flower flowers’

2.1.1.4. Abstract nouns Abstract nouns, as its name implies, do not refer to concrete entities. Rather, they are often used to denote abstract notions, such as complex or sophisticated situations, experiences, processes, qualities or phenomena in diverse areas. Some illustrative examples of abstract nouns are given below: 247

Candice Chi-Hang Cheung

(18)

(a)

真理 zhēnlǐ truth ‘truth’

(b) 病 bìng sickness ‘sickness’

(c)

矛盾 máodùn contradiction ‘contradiction’

(d) 效果 xiàoguǒ effect ‘effect’

Abstract nouns may be preceded by the general classifier ge 个, measure words indicating type, such as zhǒng 种 ‘type, kind, category’ or indeterminate measure words such as yīxiē 一些 ‘some’ and yīdiǎn 一点 ‘a little’ (Chao 1968, D. Zhu 1982), as demonstrated below: (19)

(a)

一个 真理 yī-ge zhēnlǐ one-Cl truth ‘a truth’

(b) 一种 yī-zhǒng one-kind ‘a kind of

(c)

一些 yīxiē some ‘some

(d) 一点 yīdiǎn a.little ‘a little

矛盾 máodùn contradiction contradictions’

病 bìng sickness sickness’

效果 xiàoguǒ effect effect’

2.2. Verbs Unlike nouns, verbs can be negated by bù 不 ‘not’ in Chinese (Chao 1968; C.-T. J. Huang, Li and Li 2009: Ch. 1): (20)

(a)

我 不 休息。 wǒ bù xiūxī. I not rest ‘I do not rest’

(b)

他 不 喜欢 数学。 tā bù xǐhuān shùxué. he not like mathematics ‘He does not like mathematics’

(c)

他 不 批评 别人。 tā bù pīpíng biérén. he not criticize other.person ‘He does not criticize others’

(d)

我 不 送 wǒ bú sòng I not give ‘I do not give

他 tā him him

礼物。 lǐwù. present (a) present(s)’

2.2.1.  Classification of verbs: action verbs vs. stative verbs Among the verbs in Chinese, they can be divided into two classes: action verbs and stative verbs. The former depict actions, whereas the latter express the states of being. A well-known characteristic of action verbs is that they are compatible with aspect markers, such as the progressive aspect marker zài 在, as in (21a-d) (see C. N. Li and Thompson 1981: Ch. 6 and Smith 1991 for discussion of the aspect markers in Chinese), while stative verbs can be modified by the intensifier hěn 很 although its presence is not mandatory, as shown in (22a–d) (Chao 1968, D. Zhu 1982, inter alia).6 248

Chinese: Parts of speech

(21)

(22)

(a)

他 在 休息。 tā zài xiūxī. he Prog rest-Perf ‘He is resting’

(b)

我 在 批评 他。 wǒ zài pīpíng tā. I Prog criticize he ‘I am criticizing him’

(c)

我 在 研究 这个 问题。 wǒ zài yánjiū zhè-ge wèntí. I Prog study this-Cl problem ‘I am studying this problem’

(d)

他 在 教 我 法语。 tā zài jiāo wǒ Fáyǔ. he Prog teach me French ‘He is teaching me French’

(a)

他 (很) tā (hén) he very ‘He likes

(b) 我 (很) wǒ (hěn) I very ‘I respect

(c)

他 (很) 羡慕 我。 tā (hěn) xiànmù wǒ. he very envy me ‘He envies me (a lot)’

喜欢 数学。 xǐhuān shùxué. like mathematics mathematics (a lot)’

尊敬 他。 zūnjìng tā. respect him him (a lot)’

(d) 我 (很) 想念 他。 wǒ (hén) xiǎngniàn tā. I very miss him ‘I miss him (a lot)’

2.2.2.  Classification of verbs: subcategorization properties Another way of classifying verbs is based on the number of arguments for which they subcategorize. Similar to the verbs in the Indo-European languages, verbs in Chinese can be classified into intransitive, transitive and ditransitive verbs.

2.2.2.1. Intransitive verbs Intransitive verbs are those that subcategorize for a single argument. Typical examples of intransitive verbs are lái 来 ‘come’, sǐ 死 ‘die’, shuì 睡 ‘sleep’, fāshēng 发生 ‘happen’, pǎo 跑 ‘run’, xiūxī 休息 ‘rest’, líkāi 离开 ‘leave’, etc. Some illustrative examples are given below:7 (23)

(a)

他 tā he ‘He

来了。 lái-le. come-Perf came’

(b) 他  死了。 tā  sǐ-le. he die-Perf ‘He died’

(c)

他 tā he ‘He

睡了。 shuì-le. sleep-Perf slept’

(d) 刚刚 发生了 gānggāng fāshēng-le just.now happen-Perf ‘A car accident happened

(e)

他 tā he ‘He

跑了。 pǎo-le. run-Perf ran’

(f)

车祸。 chē-huò. car-accident just now’

他  休息了。 tā  xiūxī-le. he rest-Perf ‘He rested’

Interestingly, many intransitive verbs in Chinese can be used transitively (see Chappell 1999, Xu 1999, 2001, J. Han 2000, Wen and Chen 2001, Shen et al. 2002, Sun and Wu 2003, S.-W. Tang 2004, Hole 2005, X. Zhu 2005, Pan and Han 2006, C.-T. J. Huang 2007, inter alia), as shown below: 249

Candice Chi-Hang Cheung

(24)

(a)

你 来 碗 牛肉 面 吧？ nǐ lái wǎn niúròu miàn ba? you come bowl beef noodle Prt ‘Have a bowl of beef noodles, will you?’

(b) 他 tā he ‘His (c)

死了 父亲。 sǐ-le fùqīn. die-Perf father father died’

他们 tāmen they ‘A car

(d) 他 tā he ‘He

昨天 发生了 zuótiān fāshēng-le yesterday happen-Perf accident happened to them

车祸。 chē-huò. car-accident yesterday’

跑了 两个 犯人。 pǎo-le liǎng-ge fànrén. run-Perf two-Cl criminal had two criminals ran away’

2.2.2.2. Transitive verbs As for transitive verbs, they subcategorize for two arguments: an external argument that corresponds to the subject and an internal argument that corresponds to the object. Both action and stative verbs can be used as transitive verbs. Typical examples of action verbs that are transitive include zuò 做 ‘do’, mǎi 买 ‘buy’, qù 去 ‘go’, pīpíng 批评 ‘criticize’, etc. and those of stative verbs that are transitive include xǐhuān 喜欢 ‘like’, zūnjìng 尊敬 ‘respect’, xiǎngniàn 想念 ‘miss’, xiànmù 羡慕 ‘envy’, etc. As mentioned earlier, action verbs can be followed by aspect markers, such as the perfective aspect marker -le 了, as shown in (25a–d), whereas stative verbs can be modified by the intensifier hěn 很, as in (26a–d). (25)

(26)

(a)

他 做了 很多 坏 事。 tā zuò-le hěnduō huài shì. he do-Perf many bad thing ‘He did many bad things’

(b) 他 买了 一本 书。 tā mǎi-le yī-běn shū. he buy-Perf one-Cl book ‘He bought a book’

(c)

他 去过 北京。 tā qù-guo Běijīng. he go-Exp Beijing ‘He has been to Beijing’

(d) 他 批评了 别人。 tā pīpíng-le biérén. he criticize-Perf other.people ‘He criticized others’

(a)

他 很 喜欢 数学。 tā hěn xǐhuān shùxué. he very like mathematics ‘He likes mathematics a lot’

(b) 我 很 尊敬 他。 wǒ hěn zūnjìng tā. I very respect him ‘I respect him a lot’

(c)

他 很 羡慕 我。 tā hěn xiànmù wǒ. he very envy me ‘He envies me a lot’

(d) 我 很 想念 他。 wǒ hén xiǎngniàn tā. I very miss him ‘I miss him a lot’ 250

Chinese: Parts of speech

2.2.2.3.  Ditransitive verbs Ditransitive verbs subcategorize for three arguments: an external argument that corresponds to the subject and two internal arguments, one corresponding to an indirect object and the other corresponding to a direct object. Typical examples of ditransitive verbs include sòng 送 ‘give’, jiāo 教 ‘teach’, dì 递 ‘hand’, jì 寄 ‘send’, huán 还 ‘return’, mài 卖 ‘sell’, jièshào 介绍 ‘introduce’, etc. Unlike English whose ditransitive verbs may appear in double object constructions (e.g. I gave him a book) and dative constructions (e.g. I gave a book to him), ditransitive verbs in Chinese may appear in three types of constructions: double object constructions in the form of [Verb + Noun Phrase + Noun Phrase], dative constructions in the form of [Verb + Noun Phrase + gěi 给 + Noun Phrase]8 and verb-gěi 给 double object constructions in the form of [Verb-gěi 给 + Noun Phrase + Noun Phrase], with the verb and gěi 给 forming a verbal complex (see D. Zhu 1979, Y.-H. A. Li 1990, and F.-H. Liu 2006 for more detailed discussion). As noted by F.-H. Liu (2006), not all ditransitive verbs can appear in the three constructions in Chinese. As shown below, while sòng 送 ‘give’ can appear in all three constructions (27a–c), jiāo 教 ‘teach’ can appear in double object constructions and verb-gěi double object constructions but not dative constructions (28a–c) and dì 递 ‘hand’ can appear in dative constructions and verb-gěi 给 double object constructions but not double object constructions (29a–c) (the examples in (28)–(29) are adapted from Liu 2006). (27)

(28)

(a)

他 tā he ‘He

送了 sòng-le give-Perf gave me a

我 一本 wǒ yī-běn me one-Cl book’

(b) 他 tā he ‘He

送了 一本 sòng-le yī-běn give-Perf one -Cl gave a book to me’

(c)

他 tā he ‘He

送给 sòng-géi give-to gave me a

(a)

他 tā he ‘He

教了 jiāo-le teach-Perf taught me a

书。 shū. book

书 shū book

我 一本 wǒ yī-běn me one -Cl book’ 我 一个 wǒ yī-ge me one-Cl method’

给 géi to

我。 wǒ. me

书。 shū. book

方法。 fāngfǎ. method

(b) *他 教了 一个 方法 给 *tā jiāo-le yī-ge fāngfǎ gěi he teach-Perf one-Cl method to Intended: ‘He taught a method to me’ (c)

他 tā he ‘He

教给 jiāo-gěi teach-to taught me

我 一个 wǒ yī-ge me one-Cl a method’

方法。 fāngfǎ. method

251

我。 wǒ. me

Candice Chi-Hang Cheung

(29)

(a)

*他 递了 我 *tā dì-le wǒ he hand-Perf me Intended: ‘He handed

一杯 啤酒。 yī-bēi píjiǔ. one-glass beer me a glass of beer’

(b) 他 tā he ‘He

递了 一杯 dì-le yī-bēi hand-Perf one-glass handed a glass of beer

(c)

递给 我 一杯 啤酒。 dì-gěi wǒ yī-bēi píjiǔ. hand-to me one-glass beer handed me a glass of beer’

他 tā he ‘He

啤酒 给 píjiǔ géi beer to to me’

我。 wǒ. me

2.3. Adjectives Adjectives and stative verbs share many common properties in Chinese. For this reason, there are scholars who hold the view that adjectives should be subsumed under stative verbs and that adjectives should not be regarded as an independent category (McCawley 1992; see also Chao 1968). Contrary to this view, many scholars argue that adjectives should be regarded as an independent category distinct from stative verbs (D. Zhu 1982; Paul 2006, 2010). Adopting the latter view, I begin by reviewing the common properties shared by adjectives and verbs (section 2.3.1), followed by a discussion of the arguments for treating adjectives as an independent category (section 2.3.2).

2.3.1.  Common properties of adjectives and verbs Unlike adjectives in the Indo-European languages, those in Chinese can function as predicates without the presence of the copular verb shì 是 ‘be’. Furthermore, they are typically accom­ panied by the intensifier hěn 很 in the predicate position, as demonstrated below.9 In this regard, adjectives behave like stative verbs in Chinese (D. Zhu 1982). (30)

(a)

他 很 高。 tā hěn gāo. he very tall ‘He is tall’

(b) 他 很 聪明。 tā hěn cōngmíng. he very smart ‘He is smart’

(c)

这本 书 很 贵。 zhè-běn shū hěn guì. this-Cl book very expensive ‘This book is expensive’

(d) 他 很 伤心。 tā hěn shāngxīn. he very sad ‘He is sad’

In addition, similar to verbs, adjectives in the predicate position can be negated by bù 不 ‘not’: (31)

(a)

他 不 高。 tā bù gāo. he not tall ‘He is not tall’

(b)

252

他 不 聪明。 tā bù cōngmíng. he not smart ‘He is not smart’

Chinese: Parts of speech

(c)

这本 书 不 贵。 zhè-běn shū bú guì. this-Cl book not expensive ‘This book is not expensive’

(d)

他 不 伤心。 tā bù shāngxīn. he not sad ‘He is not sad’

2.3.2.  Distinguishing adjectives from verbs In spite of these similarities, there is a need for distinguishing adjectives from verbs and treating adjectives as an independent category (D. Zhu 1982; C.-T. J. Huang, Li and Li 2009: Ch. 1; Paul 2010). Specifically, D. Zhu (1982) notes that adjectives crucially differ from verbs in that they cannot subcategorize for an internal argument or an object, as evidenced by the ill-formedness of the following examples. In this regard, adjectives behave on a par with intransitive verbs in being one-place predicates. (32)

(a)

*他 *tā he

高 gāo tall

(b)

*他 *tā he

聪明 cōngmíng smart

(c)

*这本 *zhè-běn this-Cl

(d)

*他 *tā he

那个 nà-ge that-Cl

书 shū book

伤心 shāngxīn sad

男孩。 nánhái. boy

那个 nà-ge that-Cl

学生。 xuésheng. student

贵 guì expensive 这件 zhè-jiàn this-Cl

那本 nà-běn that-Cl

书。 shū. book

事。 shì. matter

Despite the fact that adjectives are akin to intransitive verbs in not being able to take an object, adjectives should not be subsumed under intransitive verbs, since adjectives show different reduplication patterns from verbs in Chinese (D. Zhu 1956, 1980, 1982; C. N. Li and Thompson 1981: Ch. 3; Lü 1984; C.-T. J. Huang, Li and Li 2009: Ch. 1; Paul 2010). Specifically, while disyllabic adjectives of the AB form can undergo reduplication to give rise to the AABB pattern, as in (33a–d), disyllabic verbs of the AB form typically give rise to the ABAB pattern after reduplication, as demonstrated in (34a–d). (33) (a) gānjìng 干净 ‘clean’ → gāngānjìngjìng 干干净净 ‘thoroughly clean’ (b) lǎoshi 老实 ‘honest’ → láolǎoshishi 老老实实 ‘honest and frank’ (c) jiǎndān 简单 ‘simple’ → jiánjiǎndāndān 简简单单 ‘rather simple’ (d) hútu 胡涂 ‘muddleheaded’ → húhútutu 糊胡涂涂 ‘rather muddleheaded’ (34) (a) qǐngjiào 请教 ‘inquire’ → qǐngjiàoqǐngjiào 请教请教 ‘inquire a little’ (b) pīpíng 批评 ‘criticize’ → pīpíngpīpíng 批评批评 ‘criticize a little’ (c) yánjiū 研究 ‘research’ → yánjiūyánjiū 研究研究 ‘research a little’ (d) tǎolùn 讨论 ‘discuss’ → tǎolùntǎolùn 讨论讨论 ‘discuss a little’ Apart from the fact that adjectives and verbs show different reduplication patterns, they also differ in terms of interpretational properties (Paul 2010). Specifically, reduplication of adjectives often yields a higher degree of liveliness or intensity (D. Zhu 1956; Chao 1968: 209; 253

Candice Chi-Hang Cheung

Tang T.-C. 1988), while reduplication of verbs often gives rise to the so-called ‘tentative aspect’ according to Chao (1968: 204). The different reduplication patterns and interpretational properties thus constitute another piece of evidence for distinguishing adjectives from verbs. Additional evidence for analyzing adjectives as an independent category comes from the fact that adjectives are composed of predicative and non-predicative adjectives (Y. Li 1996; Aoun and Li 2003; Paul 2005, 2010) just like those in the Indo-European languages.10 If adjectives were subsumed under intransitive verbs, one would expect that all adjectives can be used predicatively on a par with intransitive verbs, contrary to fact. Some illustrative examples of non-predicative adjectives are given below (see Aoun and Li 2003; Paul 2005, 2010 for more examples of non-predicative adjectives): (35)

(a)

*这条 *zhè-tiáo this-Cl Intended:

(b)

*他们的 目标 很 共同。 *tāmen-de mùbiāo hěn gòngtóng. they-Poss goal very common Intended: ‘Their goals are common’

道路 很 主要。 dàolù hén zhǔyào. road very main ‘This road is main’

Finally, it is worth mentioning that both predicative and non-predicative adjectives can function as attributive adjectives when they precede de 的 and a noun, as illustrated below (Aoun and Li 2003; Paul 2005, 2010; see also C. C.-H. Cheung 2012): (36)

(a)

聪明 cōngmíng smart ‘intelligent

(37)

(a)

主要 的 zhǔyào de main DE ‘main roads’

的 学生 de xuésheng DE student students’ 道路 dàolù road

(b)

贵 guì expensive ‘expensive

(b)

共同 gòngtóng common ‘common

的 书 de shū DE book books’

的 目标 de mùbiāo DE goal goals’

2.4. Adverbs Adverbs typically occur between the subject and a verb or an adjective, as shown in the following examples. The presence of the verb or adjective is mandatory, as adverbs, unlike verbs and adjectives, cannot serve as predicates:11 (38)

(a)

他 tā he ‘He

刚 *(来)。 gāng *(lái). just arrive just arrived’

(b)

他 tā he ‘He

常常 *(跳舞)。 chángcháng *(tiàowǔ). often dance dances often’

(c)

他 tā he ‘He

非常 *(高)。 fēicháng *(gāo). extremely tall is extremely tall’

(d)

他 tā he ‘He

真 *(聪明)。 zhēn *(cōngmíng). truly smart is really smart’

254

Chinese: Parts of speech

2.4.1.  Classification of adverbs: sentence-level adverbs vs. VP-level adverbs Adverbs in Chinese can be roughly divided into two types: sentence-level adverbs and VP-level adverbs (C. N. Li and Thompson 1981: Ch. 8; Paul, forthcoming). The former may appear before or after the subject, whereas the latter can only appear after the subject.12

2.4.1.1. Sentence-level adverbs Among the sentence-level adverbs, they can be further divided into time adverbs and attitude adverbs, both of which can immediately precede or follow the subject (C. N. Li and Thompson 1981: Ch. 8), with only a few exceptions (Paul, forthcoming). 2.4.1.1.1. Time adverbs  Typical examples of time adverbs include jīntiān 今天 ‘today’, qùnián 去年 ‘last year’, zànshí 暂时 ‘temporarily’, jiānglái 将来 ‘(in the) future’, xiànzài 现在 ‘now’, xiàwǔ 下午 ‘(in the) afternoon’, among many others. Time adverbs can either occur in the sentence-initial position or immediately after the subject, as demonstrated below: (39)

(a)

今天 他 很 高兴。 jīntiān tā hěn gāoxìng. today he very happy ‘He is very happy today’

(b) 他 今天 很 高兴。 tā jīntiān hěn gāoxìng. he today very happy ‘He is very happy today’

(40)

(a)

去年 他 没有 回来。 qùnián tā méiyǒu huílai. last.year he not.have return ‘He did not return last year’

(b) 他 去年 没有 回来。 tā qùnián méiyǒu huílai. he last.year not.have return ‘He did not return last year’

(41)

(a)

暂时 他 住 在 这里。 zànshí tā zhù zài zhèli. temporarily he live at here ‘He lives here temporarily’

(b) 他 暂时 住 在 这里。 tā zànshí zhù zài zhèli. he temporarily live at here ‘He lives here temporarily’

The time adverbs in the above examples clearly function as sentential adverbs, as they signal the time at which or during which the event described by the sentence occurs. 2.4.1.1.2. Attitude adverbs  Attitude adverbs denote the speakers’ attitude toward or evaluation of the event described by the sentence. Typical examples of attitude adverbs include xiǎnrán 显然 ‘obviously’, yěxǔ 也许 ‘perhaps’, dàgài 大概 ‘probably’, xìngkuī 幸亏 ‘fortunately’, dāngrán 当然 ‘of course’, tūrán 突然 ‘suddenly’, qíshí 其实 ‘in fact’, among many others. The following examples show that they can occur in the sentential-initial position or immediately after the subject on a par with time adverbs: (42)

(a)

显然 他 不 喜欢 你。 xiǎnrán tā bù xǐhuān nǐ. obviously he not like you ‘Obviously, he does not like you’ 255

(b) 他 显然 不 喜欢 你。 tā xiǎnrán bù xǐhuān nǐ. he obviously not like you ‘Obviously, he does not like you’

Candice Chi-Hang Cheung

(43)

(a)

也许 他 yéxǔ tā perhaps he ‘Perhaps he

(44)

(a)

大概 他 dàgài tā probably he ‘Probably he

(b) 他 也许 tā yéxǔ he perhaps ‘Perhaps he

会 来。 huì lái. will come will come’ 是 shì be has

感冒 gǎnmào cold caught a

会 来。 huì lái. will come will come’

了。 (b) 他 大概 是 le. tā dàgài shì Prt he probably be cold’ ‘Probably he has

感冒 gǎnmào cold caught a

了。 le. Prt cold’

However, as noted by Paul (forthcoming), a few attitude adverbs, especially those containing shuō 说 ‘speak’, such as lǎoshí shuō 老实说 ‘frankly speaking’, tǎnbái shuō 坦白说 ‘honestly speaking’, huàn jù huà shuō 换句话说 ‘in other words’, among others, are confined to the sentence-initial position, as evidenced by the contrast between (45a) and (45b). (45)

(a)

老实说， 我 lǎoshí-shuō, wǒ frankly-speaking I ‘Frankly speaking, I do

(b)

*我 老实说 不 wǒ lǎoshí-shuō bù I frankly-speaking not Intended: ‘Frankly speaking,

不 喜欢 他。 bù xǐhuān tā. not like him not like him’ 喜欢 他。 xǐhuān tā. like him I do not like him’

2.4.1.2. VP-level adverbs As for VP-level adverbs, they can be subdivided into manner and non-manner adverbs, both of which occur between the subject and the verb (C. N. Li and Thompson 1981: Ch. 8). 2.4.1.2.1.  Manner adverbs  Manner adverbs serve to modify the verb phrase by describing the manner in which the action depicted by the verb phrase is carried out. The majority of the manner adverbs is derived from adjectives through the addition of the suffix -de 地, such as xìngfèn-de 兴奋地 ‘excitedly’, yánlì-de 严厉地 ‘sternly’, kuàilè-de 快乐地 ‘happily’, zǐxì-de 仔细地 ‘meticulously’, jiǎndān-de 简单地 ‘simply’, zhèndìng-de 鎭定地 ‘calmly’, etc. As mentioned earlier, they are confined to the position between the subject and the verb, as illustrated in (46a–c). (46)

(a)

他 tā he ‘He

兴奋地 跑 xīngfèn-de pǎo excitedly run ran in excitedly’

(b)

他 tā he ‘He

严厉地 责备了 yánlì-de zébèi-le sternly reproach-Perf sternly reproached me’

(c)

他们 快乐地 跳舞。 tāmen kuàilè-de tiàowǔ. they happily dance ‘They danced happily’

进来。 jìnlái. enter 我。 wǒ. me

256

Chinese: Parts of speech

For many monosyllabic and disyllabic adjectives, reduplication applies before they are suffixed by -de 地, such as mànmàn-de 慢慢地 ‘slowly’, jìngjìng-de 静静地 ‘quietly’, kuàikuài-de 快快地 ‘quickly’, qīngqīng-de 轻轻地 ‘softly’, rènrènzhēnzhēn-de 认认真真地 ‘seriously’, shūshūfufu-de 舒舒服服地 ‘comfortably’, gāogāoxìngxìng-de 高高兴兴地 ‘happily’, qīngqīngchǔchǔ-de 清清楚楚地 ‘clearly’, etc., as shown in (47a–c). In this respect, the Chinese suffix -de 地 behaves like the -ly suffix in English, which can turn an adjective into an adverb (e.g. slow → slowly, quiet → quietly, etc.). (47)

(a)

他 tā he ‘He

慢慢地 回答 我的 问题。 mànmàn-de huídá wǒ-de wèntí. slowly answer I-Poss question slowly answered my question(s)’

(b)

他 tā he ‘He

静静地 躺在 草地 jìngjìng-de tǎng-zài cǎodì quietly lie- at grass quietly lay on the grass’

(c)

他 tā he ‘He

会 认认真真地 huì rènrènzhēnzhēn-de will seriously will consider this problem

上。 shang. on

考虑 这个 kǎolǜ zhè-ge consider this-Cl seriously’

问题。 wèntí. problem

Furthermore, there is a subset of disyllabic manner adverbs, which can be optionally followed by -de 地 (D. Zhu 1982): (48)

(a)

他们 赶紧(地) 往 屋里 tāmen gánjǐn(-de) wǎng wū-li they rapidly toward house-inside ‘They rapidly ran inside the house’

(b)

树叶 在 shùyè zài leave at ‘The leaves

(c)

他们 认真(地) 讨论了 这个 tāmen rènzhēn(-de) tǎolùn-le zhè-ge they seriously consider-Perf this-Cl ‘They seriously considered this problem’

微 wēi light slowly

跑。 pǎo. run

风中 悠悠(地) 落 fēng-zhōng yōuyōu(-de) luò wind-in slowly drop fluttered down in the light wind’

下。 xia. down

问题。 wèntí. problem

2.4.1.2.2.  Non-manner adverbs  As for non-manner adverbs, they are mainly composed of VP-level temporal adverbs, such as yǐjīng 已经 ‘already’, zǎo 早 ‘long ago’, cháng 常 ‘often’, gāng 刚 ‘just’, mǎshàng 马上 ‘immediately’, among many others (see C. N. Li and Thompson 1981: Ch. 8; D. Zhu 1982; Yip and Rimmington 1997: Ch. 14, 2004: Ch. 9.1.2).13 Unlike sentence-level time adverbs, VP-level temporal adverbs are restricted to the position between the subject and the verb, as illustrated below:

257

Candice Chi-Hang Cheung

(49)

(a)

他 tā he ‘He

已经 离开了。 yǐjīng líkāi-le. already leave-Perf has already left’

(b)

他 tā he ‘He

早 zǎo long.ago knew that

(c)

他 tā he ‘He

常 来 这里。 cháng lái zhèli. often come here often comes here’

(d)

他 tā he ‘He

刚 回来。 gāng huílai. just return just returned’

知道了。 zhīdào-le. know-Perf long ago’

2.5. Prepositions Prepositions are one of the most poorly defined categories in Chinese partly due to the fact that many of the prepositions have their historical origins as verbs (C. N. Li and Thompson 1974a; D. Zhu 1982; Norman 1988; Heine et al. 1991; and Hopper and Traugott 1993) and partly due to the fact that Chinese is a morphologically impoverished language and there is no inflectional morphology to mark prepositions or verbs. Furthermore, some prepositions have characteristics of verbs (Chao 1968; C. N. Li and Thompson 1974a, 1974b, 1981; D. Zhu 1982). For these reasons, some scholars consider prepositions as a ‘mixed’ category containing verbs and prepositions (McCawley 1992), while others maintain that prepositions form an independent category distinct from verbs (C. N. Li and Thompson 1974a, 1974b; see also C.-T. J. Huang, Li and Li 2009). This section is not intended to resolve the debate as to whether prepositions are best regarded as a mixed or independent category. Instead, it will focus on the basic properties of prepositions and their categorization (section 2.5.1), followed by a discussion of the verbal properties of a subset of prepositions (section 2.5.2), seeking to provide a better understanding of the characteristic features of prepositions and the underlying factors that give rise the debate on the categorial status of prepositions in Chinese.

2.5.1.  Basic properties of prepositions and categorization In Chinese, prepositions (also referred to as ‘coverbs’ in C. N. Li and Thompson 1974a, b, 1981) are typically followed by a noun phrase and most of the prepositions and the following noun phrase appear between a subject and a verb phrase, as visualized in the generalized schema below (C. N. Li and Thompson 1974a, 1974b, 1981): (50) Subject + Preposition + Noun Phrase + Verb (+ Noun Phrase) According to C.-T. J. Huang, Li and Li (2009: Ch. 1), there are mainly four classes of words that are considered as prepositions in the literature: (51) (a) zhìyú 至于 ‘as regards’, guānyú 关于 ‘concerning’, etc. (b) cóng 从 ‘from’, duì 对 ‘toward’, etc. (c) gěi 给 ‘to/for’, zài 在 ‘at’, etc. (d) bǎ 把, bèi 被, jiào 叫, ràng 让, etc.14 All four classes of prepositions can be followed by a noun phrase, as shown in (52)–(55). However, as noted by C.-T. J. Huang, Li and Li (2009), the class of prepositions in (51a) requires the preposition and the following noun phrase to precede the subject (see also 258

Chinese: Parts of speech

C. N. Li and Thompson 1981), as demonstrated in (52a–b), unlike those in (51b–d), which can appear after the subject, as illustrated in (53)–(55). (52)

(53)

(54)

(55)

(a)

至于 zhìyú as.regards ‘As regards

(b)

关于 guānyú concerning ‘Concerning

(a)

他 tā he ‘He

从 北京 带 回来 很多 礼物。 cóng Běijīng dài huílai hěnduō lǐwù. from Beijing bring back many present brought back many presents from Beijing’

(b)

他 tā he ‘He

对 duì toward does not

(a)

他 tā he ‘He

给 学生 写过 不少 推荐 gěi xuésheng xiě-guo bùshǎo tuījiàn for student write-Exp not.few recommendation wrote quite a few recommendation letters for students’

(b)

他 tā he ‘He

在 zài at put a

(a)

他 tā he ‘He

把 花瓶 打破 bǎ huāpíng dǎ-pò BA vase hit-break broke the vase’

(b)

那个 警察 neì-ge jǐngchá that-Cl policeman ‘That policeman was

其他 问题, 我们 以后 qítā wèntí, wǒmen yǐhòu other question we later other questions, we will talk about

再 说。 zài shuō. again speak them later’

这件 事, 我 没有 意见。 zhè-jiàn shì, wǒ méiyǒu yìjiàn. this-Cl matter I not.have opinion this matter, I do not have an opinion’

这件 事 没有 意见。 zhè-jiàn shì méiyǒu yìjiàn. this-Cl matter not.have opinion have an opinion on this matter’

桌子 zhuōzi table book on

上 放了 shang fàng-le top put-Perf the table’

一本 yī-běn one-Cl

信。 xìn. letter

书。 shū. book

了。 le. Prt

被 流氓 打伤 了。 bèi liúmáng dǎ-shāng le. BEI hooligan hit-wound Prt wounded by some hooligans’

Interestingly, as noted by C.-T. J. Huang, Li and Li (2009), the classes of prepositions in (51b–c) and the following noun phrase can also appear in the pre-subject position like those in (51a), as in (56)–(57), while the class of prepositions in (51d) cannot, as evidenced by the ill-formedness of (58). (56)

(a)

从 cóng from ‘From

北京, 他 带 回来 很多 礼物。 Běijīng, tā dài huílai hěnduō lǐwù. Beijing he bring back many present Beijing, he brought back many presents’ 259

Candice Chi-Hang Cheung

(57)

(58)

(b)

对 duì toward ‘On this

(a)

给 gěi for ‘For

学生, xuésheng, student students, he

(b)

在 zài at ‘On

桌子 上, zhuōzi shang, table top the table, he put

(a)

*把 *bǎ BA

花瓶, huāpíng, vase

(b)

*被 *bèi BEI

流氓, liúmáng, hooligan

这件 zhè-jiàn this-Cl matter, he

事, 他 没有 意见。 shì, tā méiyǒu yìjiàn. matter he not.have opinion does not have an opinion’

他 写过 不少 推荐 tā xiě-guo bùshǎo tuījiàn he write-Exp not.few recommendation wrote quite a few recommendation letters’

他 tā he

他 放了 tā fàng-le he put-Perf a book’ 打破 dǎ-pò hit-break

那个 neì-ge that-Cl

一本 yī-běn one-Cl

信。 xìn. letter

书。 shū. book

了。 le. Prt

警察 jǐngchá policeman

打伤 dǎ-shāng hit-wound

了。 le. Prt

While the two classes of prepositions in (51b–c) behave alike in that the prepositions and the following noun phrase can either appear in the pre-subject or post-subject position, the two classes differ in that a subset of the prepositions in (51c) can form complex verbal predicates such as sòng-gěi 送给 ‘give-to’ and fàng-zài 放在 ‘put-at’, as in (59), whereas those in (51b) cannot (C.-T. J. Huang, Li and Li 2009). (59)

(a)

他 tā he ‘He

送给了 我 一件 sòng-gěi-le wǒ yī-jiàn give-to-Perf me one-Cl gave me a sweater’

(b)

他 tā he ‘He

把 bǎ BA put a

毛衣。 máoyī. sweater

书 放在 桌子 shū fàng-zài zhuōzi book put-at table book on the table’

上。 shang. top

In sum, we have seen that all four classes of prepositions must be followed by a noun phrase. In this respect, they behave like prepositions in the Indo-European languages. In addition, the preposition and the following noun phrase form a syntactic unit that serves to modify the verb phrase.

2.5.2.  Verbal nature of prepositions Despite the fact that prepositions in Chinese can take a noun phrase as its object on a par with those in the Indo-European languages, many prepositions share some properties of verbs due to their historical transition from verbs to prepositions. In the following, we shall consider two properties that show the verbal nature of prepositions. 260

Chinese: Parts of speech Table 15.1  The list of prepositions that can be followed by -zhe 着 àn 按 āi 挨 cháo 朝 chòng 冲 duì 对 kào 靠

píng 凭 shùn 顺 xiàng 向 wǎng 往 wèi 为 yán 沿

‘according to’ ‘adjacent to’ ‘facing’ ‘facing’ ‘toward’ ‘on the strength of ’

‘depend on, according to’ ‘along’ ‘facing’ ‘toward’ ‘for’ ‘along’

2.5.2.1.  Compatibility with aspect markers One of the properties shared by verbs and many prepositions in Chinese is that they are compatible with aspect markers. For instance, it has been noted that a subset of prepositions can be followed by the durative aspect marker -zhe 着 (see C. N. Li and Thompson 1981: Ch. 6.2 for detailed discussion of the durative aspect marker -zhe 着) similar to verbs (Chao 1968, and C. N. Li and Thompson 1974a, 1974b, 1981). The representative prepositions that can be followed by -zhe 着 are given in Table 15.1 (adapted from C. N. Li and Thompson 1981: 361). Some illustrative examples showing that the prepositions in Table 15.1 can be followed by -zhe 着 are given below. As noted by C. N. Li and Thompson (1981: Ch. 9), the presence of -zhe 着 does not contribute to the meaning of the sentence, as witnessed from the fact that the following examples have the same meanings with or without -zhe 着: (60)

(a)

你们 得 按(着) nǐmen déi àn(-zhe) you must according.to -Dur ‘You must speak in order’

(b)

他 tā he ‘He

朝(着) cháo(-zhe) toward-Dur walked over

(c)

他 tā he ‘He

对(着) 我 duì(-zhe) wǒ toward-Dur me nodded to me’

我 走 wó zǒu me walk to me’

次序 cìxù order

发言。 fāyán. speak

过来。 guòlái. come

点了 diǎn-le nod-Perf

点 diǎn nod

头。 tóu. head

The fact that -zhe 着 is devoid of durative meaning when it occurs with the above prepositions suggest that they have undergone grammaticalization into prepositions and are on the verge of losing their verbal properties. Additional support for this view comes from the fact that the list of prepositions that can take -zhe 着 is fairly restricted and must be memorized (C. N. Li and Thompson 1981: Ch. 9), as evidenced by the fact -zhe 着 cannot be added freely to any prepositions: (61)

(a)

*他 *tā he

从着 cóng-zhe from-Dur

北京 Běijīng Beijing

带 dài bring

回来 huílai back

261

很多 hěnduō many

礼物。 lǐwù. present

Candice Chi-Hang Cheung

(b)

*他 *tā he

给着 gěi-zhe for-Dur

学生 xuésheng student

写 xiě write

推荐 tuījiàn recommendation

(c)

*他 *tā he

在着 zài-zhe at-Dur

图书馆 túshūguǎn library

看 kàn read

书。 shū. book

信。 xìn. letter

In addition to -zhe 着, a few prepositions can be followed by -le 了 akin to verbs (Chao 1968, and C. N. Li and Thompson 1981): (62)

他 tā he ‘For

为了 你的 事 一 wèi-le nǐ-de shì yī for-Perf you-Poss matter one your matter, he didn’t sleep for the

整 晚 zhéng wǎn whole night whole night’

都 dōu all

没 méi not.have

睡。 shuì. sleep

Despite their apparent similarities, whether -le 了 still marks the perfective aspect when it occurs with a preposition is far from clear when compared with true verbs (cf. (25)). The fact that a limited set of prepositions is compatible with -le 了 should thus be regarded as evidence for their historical transition from verbs to prepositions.

2.5.2.2.  Prepositions functioning as verbs The verbal nature of prepositions can also be reflected from their dual categorial status: many prepositions in Chinese can be used as prepositions as well as verbs (Chao 1968, and C. N. Li and Thompson 1974a, b, 1981), as shown in Table 15.2 (adapted from C. N. Li and Thompson 1981: 368–9). The following examples illustrate that the prepositions in Table 15.2 can be used as prepositions (63a, 64a, 65a) as well as verbs (63b, 64b, 65b). (63)

(64)

(a)

我们 wǒmen we ‘We will

(b)

是不是 有 人 在 按 shì-bu-shì yǒu rén zài àn be-not-be have person Prog press ‘Is someone ringing the doorbell?’

(a)

他 tā he ‘He

(b)

他的 房子 tā-de fàngzi he-Poss house ‘His house faces

会 按 你的 意思 办 huì àn nǐ-de yìsī bàn will according.to you-Poss idea do do this matter according to your ideas’

朝 我 笑了 cháo wǒ xiào-le facing me smile-Perf smiled at me’

笑。 xiào. smile

朝 海。 cháo hái. face sea the sea’ 262

门铃？ mén-líng? door-bell

这件 zhè-jiàn this-Cl

事。 shì. matter

Chinese: Parts of speech Table 15.2  Prepositions functioning as prepositions and as verbs Prepositions

As prepositions

As verbs

āi 挨 àn 按 bǐ 比 chǎo 朝 chéng 乘 chòng 冲 chúle 除了 dàitì 代替 dāng 当 dào 到 duì 对 gěi 给 gēn 跟 jiě 解 kào 靠 nì 逆 píng 凭 shùn 顺 tì 替 wǎng 往 xiàng 向

‘next to’ ‘according to’ ‘compare’ ‘facing’ ‘take advantage of ’ ‘facing’ ‘except, besides’ ‘in place of ’ ‘in front of ’ ‘to’ ‘to’ ‘to, for’ ‘with’ ‘from’ ‘on the strength of ’ ‘against’ ‘depend on, according to’ ‘along’ ‘in place of ’ ‘facing’ ‘facing’

‘be next to’ ‘press’ ‘compare’ ‘face’ ‘ride on’ ‘face’ ‘remove’ ‘take the place of ’ ‘serve as’ ‘arrive’ ‘face’ ‘give’ ‘follow’ ‘untie, relieve’ ‘lean against, depend on’ ‘be opposed to’ ‘depend on’ ‘follow’ ‘substitute for’ ‘face’ ‘face’

(65)

(a)

他 tā he ‘He

(b)

春天 到了。 chūntiān dào-le. spring arrive-Perf ‘Spring has come’

常 到 cháng dào often to often goes to

北京 Běijīng Beijing Beijing’

去。 qù. go

While many prepositions have verbal counterparts, there are quite a few prepositions that lack verbal counterparts, such as zhìyú 至于 ‘as regards’, guānyú 关于 ‘concerning’, cóng 从 ‘from’, bǎ 把, bèi 被, among others. Given the fact that many prepositions show verbal properties and have dual categorial status, a natural question that arises is whether prepositions in Chinese are best analyzed as a ‘mixed’ category consisting of verbs and prepositions. This is a plausible view that has been explored in the literature (McCawley 1992). An alternative is to maintain that prepositions form an independent category distinct from verbs and attribute the verbal properties of a subset of prepositions to their historical development from verbs (C. N. Li and Thompson 1974a, 1974b). On this view, prepositions should be distinguished from their verbal counterparts, which are best analyzed as homophonous verbs. An advantage of this view is that it can provide a natural explanation for the verbal properties of a subset of prepositions while maintaining a clear distinction between prepositions and verbs. 263

Candice Chi-Hang Cheung

2.6. Conjunctions Similar to conjunctions in the Indo-European languages, conjunctions in Chinese are function words in that they form a closed class and have grammatical rather than content meaning. In Chinese, conjunctions can be divided into two major types: those that function to conjoin phrases (section 2.6.1) and those that serve to conjoin clauses (section 2.6.2).

2.6.1.  Phrasal conjunctions For conjunctions that serve to conjoin phrases, they can be further divided into three types, depending on the types of phrases they conjoin. They include (i) conjunctions that serve to conjoin nouns or noun phrases, (ii) conjunctions that serve to conjoin adjectives or adjectival expressions and (iii) conjunctions that serve to conjoin verbs or verb phrases.

2.6.1.1.  Conjunctions conjoining nouns In Chinese, there are four conjunctions that serve to conjoin nouns or noun phrases. They are hé 和 ‘and’ (Cheung et al. 1994: 46), gēn 跟 ‘and, with’, tóng 同 ‘and, with’ and yǔ 与 ‘and, with’ (Yip and Rimmington 2004: Ch. 22.1), as illustrated below: (66)

(a)

我的 学生 和 他的 学生 都 wǒ-de xuésheng hé tā-de xuésheng dōu I-Poss student and he-Poss student all ‘My students and his students are Shanghaiese’

(b) 北京 Běijīng Beijing ‘Beijing

跟 香港 都 gēn Xiānggǎng dōu and Hong.Kong all and Hong Kong are both

上海 Shànghǎi Shanghai

是 很 值得 去 看 shì hěn zhìdé qù kàn be very worth go see places worth seeing’

(c)

他 同 妹妹的 关系 很 tā tōng mèimei-de guānxi hěn he and sister-Poss relationship very ‘The relationship between him and his sister

(d)

他 tā he ‘He

对 法国的 duì Fǎguó-de to France-Poss is interested in the

是 shì be

人。 rén. person

的 地方。 de dìfang. DE place

不 错。 bú cuò. not bad is quite good’

文化 与 历史 都 wénhuà yǔ lìshǐ dōu culture and history all culture and history of France’

很 hén very

感 gǎn feel

兴趣。 xìngqù. interest

Despite the fact that all four conjunctions have the meaning of ‘and’, they differ in terms of register, i.e. some are more formal than the others (Lü 1980).

2.6.1.2.  Conjunctions conjoining adjectives As for conjunctions that function to conjoin adjectives or adjectival expressions, they include simplex conjunctions that involve a single word like ér 而 ‘also’ (Yip and Rimmington 2004: Ch. 22.1) and complex ones that involve more than one word, such as yòu 又  .  .  .  yòu 又 ‘both  .  .  .  and  .  .  .’ (Cheung et al. 1994: 427), etc. Both types of conjunctions can conjoin adjectives or adjectival expressions in the predicative (67a, 68a) as well as attributive positions (67b, 68b). 264

Chinese: Parts of speech

(67)

(68)

(a)

他 tā he ‘He

坚定 而 jiāndìng ér steadfast and is steadfast and

(b)

他 tā he ‘He

是个 认真 而 shì-ge rènzhēn ér be-Cl serious and is a serious and rigorous

(a)

这个 西瓜 又 大 又 zhè-ge xīguā yòu dà yòu this-Cl watermelon also large also ‘This watermelon is large and sweet’

(b)

他 tā he ‘He

是 shì be is a

个 ge Cl smart

勇敢。 yónggǎn. brave brave’ 严谨 的 yánjǐn de rigorous DE philosopher’

又 聪明 又 yòu cōngmíng yòu also smart also and diligent student’

哲学家。 zhéxuéjiā. philosopher

甜。 tián. sweet

勤奋 qínfèn diligent

的 de DE

学生。 xuésheng. student

2.6.1.3.  Conjunctions conjoining verbs We have seen that yòu 又  .  .  .  yòu 又 ‘both  .  .  .  and  .  .  .’ can be used to conjoin adjectives. In addition to adjectives, it can serve to conjoin verbs or verb phrases to emphasize the coexistence of the two situations depicted by the verb phrases (Cheung et al. 1994: 427), as demonstrated below: (69)

(a)

他们 对 他 又 打 tāmen duì tā yòu dǎ they to him also hit ‘They hit and scolded him’

(b)

他们 又 唱 又 tāmen yòu chàng yòu they also sing also ‘They sang and danced, and

又 yòu also

骂。 mà. scold

跳, 非常 高兴。 tiào, fēicháng gāoxìng. dance extremely happy were extremely happy’

Another conjunction used to conjoin verbs is bìng 并 ‘also’. Unlike yòu 又  .  .  .  yòu 又 ‘both  .  .  .  and’, which is often used to conjoin intransitive verbs, bìng 并 is used to conjoin transitive verbs that share the same object (Yip and Rimmington 2004: Ch. 22.1): (70)

(a)

他们 讨论 tāmen tǎolùn they discuss ‘They discussed

(b)

大家 dàjiā everyone ‘Everyone

并 通过了 这个 bìng tōngguò-le zhè-ge also pass-Perf this-Cl and passed this project’

都 同意 并 dōu tóngyì bìng all agree also agreed and supported

计划。 jìhuà. project

拥护 我的 yōnghù wǒ-de support I-Poss my proposal’

265

提议。 tíyì. proposal

Candice Chi-Hang Cheung

In (70a), zhè-ge jìhuà 这个计划 ‘this project’ is the shared object of tǎolùn 讨论 ‘discuss’ and tōngguò 通过 ‘pass’, both of which are transitive verbs. In (70b), the shared object is wǒ de tíyì 我的提议 ‘my proposal’.

2.6.2.  Clausal conjunctions Most of the conjunctions in Chinese are clausal conjunctions. They serve to conjoin two clauses and display three characteristic features. First, many of the clausal conjunctions occur in pairs and in some cases, one of the elements in the pair can be optionally omitted (C. N. Li and Thompson 1981: Ch. 23; Yip and Rimmington 2004: Ch. 22.2), as shown in (71). (71) (a) búdàn 不但  .  .  .  érqiě 而且 (b) jíshǐ 即使  .  .  .  yě 也  .  .  . (c) níngkě 宁可  .  .  .  yěbù 也不 (d) (rúguǒ) (如果)  .  .  .  jiù 就 (e) suīrán 虽然  .  .  .  dànshì 但是  .  .  . (f) wúlùn 无论  .  .  .  dōu 都  .  .  . (g) (yàoshì) (要是)  .  .  .  jiù 就  .  .  . (h) ( yīnwéi) (因为)  .  .  .  suǒyǐ 所以  .  .  . (i) zhǐyǒu 只有  .  .  .  cái 才  .  .  .

‘not only  .  .  .  but also  .  .  .’ ‘even though  .  .  .  still  .  .  .’ ‘would rather  .  .  .  and not  .  .  .’ ‘if  .  .  .  then  .  .  .’ ‘although  .  .  .  but  .  .  .’ ‘no matter what  .  .  .  still  .  .  .’ ‘if  .  .  .  then  .  .  .’ ‘because  .  .  .  therefore  .  .  .’ ‘only if  .  .  .  then  .  .  .’

There are also some clausal conjunctions that are used individually (C. N. Li and Thompson 1981: Ch. 23; Yip and Rimmington 2004: Ch. 22.2), as in (72). (72) (a) búguò 不过  .  .  . (b) .  .  .  de shíhòu 的时候 (c) kěshì 可是  .  .  . (d) yǐqián 以前  .  .  .

‘but  .  .  .’ ‘when  .  .  .’ ‘but  .  .  .’ ‘before  .  .  .’

Some illustrative examples of the two types of clausal conjunction are given below: (73)

(a)

他 tā he ‘He

不但 骂 búdàn mà not.only scold not only swore at

(b) 我 宁可 不 wǒ níngkě bù I rather not ‘I would rather not (c)

(如果) (rúguǒ) if ‘If he is

(d)

大家 dàjiā everyone ‘Although

人, rén, person people,

而且 还 érqiě hái but.also in.addition but also hit them’

打 dǎ hit

人 rén person

呢。 ne. Prt

喝 酒， 也 不 能 酒 后 驾 车。 hē jiǔ, yě bù néng jiǔ hòu jià chē. drink wine and not can wine after drive car drink than drive after drinking’

他 喝醉了, tā hē-zuì-le, he drink-drunk-Perf drunk, we will take him

我们 wǒmen we home’

虽然 很 累, suīrán hěn lèi, although very tired everyone was tired, their 266

就 jiù then

送 sòng take

他 tā he

但是 心情 都 dànshì xīnqíng dōu but mood all mood was cheerful’

回 huí return 很 hěn very

家。 jiā. home 愉快。 yúkuài. cheerful

Chinese: Parts of speech

(e)

无论 wúlùn no.matter ‘No matter

(f)

(要是) 明天 不 下雨, 我 (yàoshì) míngtiān bú xiàyǔ, wǒ if tomorrow not rain I ‘If it doesn’t rain tomorrow, I’ll definitely

(g) (因为) (yīnwéi) because ‘Because

(74)

你 ní you how

走得 多 快, zǒu-de duō kuài, go-Res how fast fast you go, he can still

他 腿部 受了 tā tuǐbù shòu-le he leg bear-Perf he had a leg injury, he

他 都 能 tā dōu néng he still can catch up’ 就 jiù then go’

一定 yídìng definitely

伤, 所以 shāng, suóyǐ wound therefore did not take part in

赶得 gǎn-de catch-Res

上。 shàng. up

去。 qù. go

没 参加 比赛。 méi cānjiā bǐsài. not.have join match the match’

(h)

只有 乐观, 你的 zhíyǒu lèguān, nǐ-de only.if optimistic you-Poss ‘Only by being optimistic could

(a)

他 tā he ‘He

(b)

我 回来 的 wǒ huílai de I return DE ‘When I returned

(c)

来, 可是 没 赶上 飞机。 我 本来 想 早   点 wǒ běnlái xiáng záo  diǎn lái, kěshì méi gǎn-shàng fēijī. I originally think early a.little come but not.have chase-ascend plane ‘I had originally intended to come earlier, but I didn’t catch the plane’

(d)

我 离开 以前, wǒ líkāi yǐqián, I leave before ‘Before I left home,

不 是 很 bú shì hěn not be very is not smart, but

病 才 能 恢复得 快。 bìng cái néng huīfù-de kuài. sickness then can recover-Res quick you be able to have a speedy recovery’

聪明, 不过 考上 大学 cōngmíng, búguò kǎo-shàng dàxué smart but exam-ascend university he passed the university entrance exam’

时候, shíhòu, time home, my

了。 le. Prt

妈妈 已经 睡着了。 māma yǐjīng shuì-zháo-le. mother already sleep-fall-Perf mother was already asleep’

弟弟 还 dìdì hái younger.brother still my younger brother still

没 回来。 méi huílai. not.have return had not returned’

Second, some of the clausal conjunctions have monosyllabic and disyllabic variants depending on the rhythmic requirement (Yip and Rimmington 2004: Ch. 22.2): (75) (a) dànshì 但是 ‘but’ (b) suīrán 虽然 ‘although’ (c) rúguǒ 如果 ‘if’ (d) yīnwéi 因为 ‘because’

→  →  →  → 

dàn 但 ‘but’ suī 虽 ‘although’ rú 如 ‘if’ yīn 因 ‘because’

267

Candice Chi-Hang Cheung

The following examples illustrate the use of the monosyllabic variants: (76)

(a)

你的 建议 很 nǐ-de jiànyì hén you-Poss suggestion very ‘Your suggestion is good, but

好, 但 时机 还 hǎo, dàn shíjī hái good but time still the time is still not ripe’

(b)

文章 wénzhāng article ‘Although

却 很 有 què hěn yǒu yet very have it is very forceful’

(c)

你 如 有 困难, 我 可以 帮助 nǐ rú yǒu kùnnán, wǒ kěyǐ bāngzhù you if have problem I can help ‘If you have a problem, I can help you’

(d) 因 yīn because ‘Because

虽 短, suī duǎn, although short the article is short,

年代 niándài age it was in

不 bù not

成熟。 chéngshú. ripe

力。 lì. force

久远, 这件 事 jiúyuǎn, zhè-jiàn shì remote.past this-Cl matter the remote past, there is no way

你。 nǐ. you 已经 yǐjīng already to check

无法 考察。 wúfǎ kǎochá. no.way check this matter’

Third, for clausal conjunctions that occur in pairs, the first member of the pair may appear before or after the subject (Yip and Rimmington 2004: Ch. 22.2; see also Y. Shi 1986: section 3.7; McCawley 1988; Gasde and Paul 1996; Tsao 1996; Zhou 2002: Ch. 5; N. N. Zhang 2009), as demonstrated below: (77)

(78)

(a)

要是 yàoshì if ‘If you

你 不 舒服, 就 别 nǐ bù shūfu, jiù bié you not well then not are not well, then don’t come’

来 lái come

了。 le. Prt

(b)

你 要是 不 舒服, 就 别 nǐ yàoshì bù shūfu, jiù bié you if not well then not ‘If you are not well, then don’t come’

来 lái come

了。 le. Prt

(a)

虽然 suīrán although ‘Although

他 身軆 不 tā shēntǐ bù he body not he wasn’t well, he

好, hǎo, good rarely

但是 很少 dànshì hěnshǎo but rarely requested leave’

请假。 qǐng-jià. request-leave

(b)

他 虽然 身軆 不 tā suīrán shēntǐ bù he although body not ‘Although he wasn’t well, he

好, hǎo, good rarely

但是 很少 dànshì hěnshǎo but rarely requested leave’

请假。 qǐng-jià. request-leave

Finally, it is worth mentioning that clauses can be conjoined without the presence of any explicit clausal conjunctions in Chinese (C. N. Li and Thompson 1981: Ch. 23; Yip and Rimmington 2004: Ch. 22.2), as shown below: 268

Chinese: Parts of speech

(79)

(a)

(如果) 你 (rúguǒ) nǐ if you ‘If you go, I’ll

去, 我 (就) qù, wǒ (jiù) go I then go with you’

(b)

他 刚 想 出去 (的时候), 忽然 下起 大雪 来 了。 tā gāng xiǎng chūqù (de shíhòu), hūrán xiàqǐ dà-xué lái le. he just want go.out DE time suddenly fall heavy-snow come Prt ‘When he was just thinking of going out, it suddenly began to snow heavily’

跟 gēn with

你 nǐ you

去。 qù. go

3.  Categories uniquely available in Chinese In this section, I turn to discuss the three categories that are available in Chinese, but not in the Indo-European languages, namely, classifiers, localizers and sentence-final particles.

3.1. Classifiers Classifiers (also referred to as ‘numeral classifiers’) are a salient feature of classifier languages like Chinese. They are regarded as a semi-lexical category due to the fact that many classifiers are historically derived from nouns. In our discussion of count and mass nouns in sections 2.1.1.1–2.1.1.2, I have implicitly assumed that sortal classifiers should be differentiated from measure words. However, whether sortal classifiers ought to be distinguished from measure words remains a point of contention in the literature. In fact, two traditional views can be identified in the literature: one tradition does not differentiate between sortal classifiers and measure words. For instance, Chao (1968: 584–620) regards classifiers as individual measures and subsumes them under the rubric of ‘measure words’. A similar view is held by C. N. Li and Thompson (1981: 106), who state that ‘any measure word can be a classifier’. One piece of evidence in support of this view comes from the fact that both sortal classifiers (e.g. wèi 位, běn 本, tiáo 条 and duǒ 朵) and measure words (e.g. chǐ 尺 ‘foot’, bàng 磅 ‘pound’, jiālún 加仑 ‘gallon’ and gōngjīn 公斤 ‘kilo’) can appear between a numeral and a noun: (80)

(81)

(a)

一位 客人 yī-wèi kèrén one-Cl guest ‘a guest’

(b)

三本 书 sān-běn shū three-Cl book ‘three books’

(c)

两条 蛇 liáng-tiáo shé two-Cl snake ‘two snakes’

(d)

十朵 花 shí-duǒ huā ten-Cl flower ‘ten flowers’

(a)

一尺 yī-chǐ one-foot ‘one foot

(b)

三磅 sān-bàng three-pound ‘three pounds

(c)

两加仑 liǎng-jiālún two-gallon ‘two gallons

(d)

十公斤 shí-gōngjīn ten-kilo ‘ten kilos of

布 bù cloth of cloth’ 汽油 qìyóu petrol of petrol’

269

糖 táng sugar of sugar’

米 mǐ rice rice’

Candice Chi-Hang Cheung

The other tradition holds the view that measure words are distinguishable from sortal classifiers. An observation that is indicative of such a distinction comes from the fact that sortal classifiers tend to resist the insertion of de 的, while measure words can always allow the insertion of de 的 (Tai and Chao 1994; Kuo 1998; Cheng and Sybesma 1999), as shown below:15 (82)

(83)

(a)

*一位 *yī-wèi one-Cl

(c)

*两条 *liǎng-tiáo two-Cl

(a)

一尺 yī-chǐ one-foot ‘one foot

(c)

两加仑 liǎng-jiālún two-gallon ‘two gallons

的 de DE

客人 kèrén guest

的 de DE

蛇 shé snake

的 布 de bù DE cloth of cloth’ 的 汽油 de qìyóu DE petrol of petrol’

(b)

*三本 *sān-běn three-Cl

的 de DE

书 shū book

(d)

*十朵 *shí-duǒ ten-Cl

的 de DE

花 huā flower

(b)

三磅 sān-bàng three-pound ‘three pounds

(d)

十公斤 shí-gōngjīn ten-kilo ‘ten kilos of

的 糖 de táng DE sugar of sugar’

的 米 de mǐ DE rice rice’

Additional support for the distinction between sortal classifiers and measure words comes from their different semantic functions. According to Tai (1990: 312), ‘[a] classifier categorizes a class of nouns by picking out some salient perceptual properties, whether physically or functionally based, which are permanently associated with the entities named by the class of nouns; a measure word does not categorize but denotes the quantity of the entity named by a noun’. The semantic difference becomes more transparent if we compare the sortal classifier tiáo 条 in yī-tiáo shé 一条蛇 ‘a snake’ with the measure word chǐ 尺 ‘foot’, as in yī-chǐ bù 一尺布 ‘a foot of cloth’: the former picks out the salient properties of shé 蛇 ‘snake’, i.e. it has a long and cylindrical body, whereas chǐ 尺 ‘foot’ does not pick out any salient properties of ‘cloth’ nor does it categorize a particular type of cloth. Rather, it denotes the quantity. Their different semantic functions can also be witnessed from their selectional restrictions. In particular, while sortal classifiers can only be used to classify a limited and specific set of nouns, e.g. count nouns (see section 2.1.1.1), measure words can be used with a wide variety of nouns (Tai 1990). For instance, container measure words such as xiāng 箱 ‘box’ and dài 袋 ‘bag’ can be used with count nouns as well as mass nouns: (84)

(a)

一箱 yī-xiāng one-box ‘one box

书 shū book of books’

(b)

三袋 sān-dài three-bag ‘three bags

苹果 píngguǒ apple of apples’

(85)

(a)

一箱 yī-xiāng one-box ‘one box

酒 jiǔ wine of wine’

(b)

三袋 sān-dài three-bag ‘three bags

米 mǐ rice of rice’

270

Chinese: Parts of speech

In light of the aforementioned arguments, I follow the second view and maintain that sortal classifiers should be distinguished from measure words. I shall discuss sortal classifiers and measure words in the following sections.

3.1.1.  Sortal classifiers Sortal classifiers often reflect the intrinsic features of the nouns they classify. Somewhat similar to gender in the Indo-European languages, the choice of classifiers is to some extent predictable from the meaning of the noun, especially when the classifiers are encoded with specific semantic information, such as shape, animacy, etc. Let us consider some common sortal classifiers that classify nouns in terms of shape. As shown below, sortal classifies denoting shape can be divided into three subclasses, namely, those denoting long and narrow entities (86), those denoting round and oval entities (87) and those denoting entities with a flat surface (88). (86) Sortal classifiers denoting long and narrow entities: (a) zhī 支: yī-zhī xuějiā 一支雪茄 ‘a cigar’, yī-zhī làzhú 一支蜡烛 ‘a candle’, yī-zhī qiāng 一支枪 ‘a gun’, yī-zhī jiàn 一支箭 ‘an arrow’, etc. (b) gēn 根: yī-gēn zhēn 一根针 ‘a needle’, yī-gēn xiàn 一根缐 ‘a thread’, yī-gēn tóufà 一根头发 ‘a hair’, yī-gēn huǒchái 一根火柴 ‘a match’, yī-gēn xiāngcháng 一根香肠 ‘a sausage’, yī-gēn zhúzi 一根竹子 ‘a bamboo’, etc. (c) tiáo 条: yī-tiáo chóng 一条虫 ‘a worm’, yī-tiáo shé 一条蛇 ‘a snake’, yī-tiáo yú 一条鱼 ‘a fish’, yī-tiáo shéngzi 一条绳子 ‘a rope’, yī-tiáo hé 一条河 ‘a river’, yī-tiáo tuǐ 一条腿 ‘a leg’, yī-tiáo huángguā 一条黄瓜 ‘a cucumber’, yī-tiáo chuán 一条船 ‘a boat’, yī-tiáo gǒu 一条狗 ‘a dog’, etc. (87) Sortal classifiers denoting round and oval entities: (a) kē 颗: yī-kē zhēnzhū 一颗珍珠 ‘a pearl’, yī-kē zhǒngzi 一颗种子 ‘a seed’, yī-kē zhàdàn 一颗炸弹 ‘a bomb’, yī-kē zǐdàn 一颗子弹 ‘a bullet’, yī-kē shǒuliúdàn 一颗手榴弹 ‘a hand grenade’, yī-kē xīn 一颗心 ‘a heart’, yī-kē yáchǐ 一颗牙齿 ‘a tooth’, yī-kē xīng 一颗星 ‘a star’, yī-kē tang 一颗糖 ‘a candy’, etc. (b) lì 粒: yī-lì mǐ 一粒米 ‘a grain of rice’, yī-lì dòuzi 一粒豆子 ‘a bean’, yī-lì huāshēng 一粒花生 ‘a peanut’, yī-lì pútáo 一粒葡萄 ‘a grape’, yī-lì shāzi 一粒沙子 ‘a grain of sand’, etc. (c) tuán 团: yī-tuán máoxiàn 一团毛缐 ‘a ball of wool’, yī-tuán miàn 一团面 ‘a ball of dough’, yī-tuán miánhuā 一团棉花 ‘a ball of cotton wool’, yī-tuán huǒ 一团火 ‘a fire’, yī-tuán yún 一团云 ‘a dense patch of cloud’, etc. (88) Sortal classifiers denoting entities with a flat surface: (a) zhāng 张: yī-zhāng zhǐ 一张纸 ‘a piece of paper’, yī-zhāng biǎo 一张表 ‘a chart’, yī-zhāng piào 一张票 ‘a ticket’, yī-zhāng yóupiào 一张邮票 ‘a stamp’, yī-zhāng chuáng 一张床 ‘a bed’, yī-zhāng zhuōzi 一张桌子 ‘a table’, etc. (b) piàn 片: yī-piàn yèzi 一片叶子 ‘a leaf’, yī-piàn miànbāo 一片面包 ‘a slice of bread’, yī-piàn ānmiányào 一片安眠药 ‘a sleeping tablet’, yī-piàn tǔdì 一片土地 ‘a stretch of land’, yī-piàn wāngyáng 一片汪洋 ‘a vast sheet of water’, etc. (c) miàn 面: yī-miàn jìngzi 一面镜子 ‘a mirror’, yī-miàn luó 一面锣 ‘a gong’, yī-miàn qí 一面旗 ‘a flag’, etc. 271

Candice Chi-Hang Cheung

Animacy also plays a role in the choice of classifier. For instance, animate beings, including mammals, birds and insects are commonly classified by the sortal classifier zhī 只, as in (89)–(91), whereas plants are generally classified by kē 棵, as shown in (92). (89)

(90)

(91)

(92)

(a)

一只 羊 yī-zhī yáng one-Cl sheep ‘a sheep’

(b)

两只 狼 liǎng-zhī láng two-Cl wolf ‘two wolves’

(c)

三只 老虎 sān-zhī láohǔ three-Cl tiger ‘three tigers’

(d)

六只 老鼠 liù-zhī láoshǔ six-Cl mouse ‘six mice’

(a)

一只 鸟 yī-zhī niǎo one-Cl bird ‘a bird’

(b)

四只 燕子 sì-zhī yànzi four-Cl swallow ‘four swallows’

(c)

三只 麻雀 sān-zhī máquè three-Cl sparrow ‘three sparrows’

(d)

十只 鸡 shí-zhī jī ten-Cl chicken ‘ten chickens’

(a)

一只 甲虫 yī-zhī jiǎchóng one-Cl beetle ‘a beetle’

(b)

两只 蜻蜓 liǎng-zhī qīngtíng two-Cl dragonfly ‘two dragonflies’

(c)

三只 蝴蝶 sān-zhī húdié three-Cl butterfly ‘three butterflies’

(d)

五只 毛虫 wǔ-zhī máochóng five-Cl caterpillar ‘five caterpillars’

(a)

一棵 树 yī-kē shù one-Cl tree ‘a tree’

(b)

两棵 葱 liǎng-kē cōng two-Cl spring.onion ‘two spring onions’

(c)

三棵 麦子 sān-kē màizi three-Cl wheat ‘three stalks of wheat’

(d)

四棵 草 sì-kē cǎo four-Cl grass ‘four tufts of grass’

The most versatile sortal classifier is ge 个, which can be used to classify a wide range of nouns, including nouns denoting humans (93), inanimate objects (94) and abstract concepts (95). (93)

(a)

一个 人 yī-ge rén one-Cl person ‘a person’

(b)

272

两个 朋友 liǎng-ge péngyǒu two-Cl friend ‘two friends’

Chinese: Parts of speech

(94)

(95)

(c)

三个 老师 sān-ge lǎoshī three-Cl teacher ‘three teachers’

(d)

六个 医生 liù-ge yīshēng six-Cl doctor ‘six doctors’

(a)

一个 手表 yī-ge shóubiǎo one-Cl watch ‘a watch’

(b)

三个 西瓜 sān-ge xīguā three-Cl watermelon ‘three watermelons’

(c)

两个 岛 liǎng-ge dǎo two-Cl island ‘two islands’

(d)

十个 城市 shí-ge chéngshì ten-Cl city ‘ten cities’

(a)

一个 主意 yī-ge zhǔyì one-Cl idea ‘an idea’

(b)

三个 目的 sān-ge mùdī three-Cl purpose ‘three purposes’

(c)

两个 愿望 liǎng-ge yuànwàng two-Cl wish ‘two wishes’

(d)

一个 机会 yī-ge jīhuí one-Cl opportunity ‘an opportunity’

In the literature, ge 个 has been viewed as a ‘general classifier’ (also referred to as a ‘neutral classifier’), since it can be used with a wide range of nouns of different semantic properties (C.-R. Huang and Ahrens 2003). More often, however, particular sets of nouns with common characteristics or belonging to the same type are classified by more specific sortal classifiers rather than ge 个 (see Ahrens 1994). Furthermore, many nouns may be classified by two or more alternative sortal classifiers, as shown below: (96)

(a)

一个 客人 yī-ge kèrén one-Cl guest ‘a guest’

(b)

一位 客人 yī-wèi kèrén one-Cl guest ‘a guest’

(97)

(a)

一个 窗 yī-ge chuāng one-Cl window ‘a window’

(b)

一扇 窗 yī-shàn chuāng one-Cl window ‘a window’

(98)

(a)

一条 金鱼 yī-tiáo jīnyú one-Cl goldfish ‘a goldfish’

(b)

一尾 金鱼 yī-wěi jīnyú one-Cl goldfish ‘a goldfish’

273

Candice Chi-Hang Cheung

 (99)

(a)

一支 步枪 yī-zhī bùqiāng one-Cl rifle ‘a rifle’

(b)

一杆 步枪 yī-gān bùqiāng one-Cl rifle ‘a rifle’

(100)

(a)

一支 毛笔 yī-zhī máobǐ one-Cl writing.brush ‘a writing brush’

(b)

一管 毛笔 yī-guǎn máobǐ one-Cl writing.brush ‘a writing brush’

In many cases, the choice of the classifiers is determined by the register, discourse type, and the age and dialectal background of the speakers.

3.1.2.  Measure words Measure words are employed to denote the quantity or amount of the entity named by a noun. In Chinese, they can be subdivided into four types, namely, standard measures, collective measures, container measures and generic measures (Chao 1968, C. N. Li and Thompson 1981: Ch. 4.2.1, D. Zhu 1982, Yip and Rimmington 2004: Ch. 2.8, among others).

3.1.2.1. Standard measures Standard measure words express nationally or internationally accepted standards of measurement, indicating standards for weight, quantity, volume or size. Typical examples include jīn 斤 ‘catty’, dǎ 打 ‘dozen’, liǎng 两 ‘ounce’, bàng 磅 ‘pound’, gōngjīn 公斤 ‘kilo’, gōngshēng 公升 ‘liter’, chǐ 尺 ‘foot’, jiālún 加仑 ‘gallon’, etc. Some illustrative examples are given below: (101)

(a)

一斤 yī-jīn one-catty ‘a catty of

(c)

三两 sān-liǎng three-ounce ‘three ounces

(e)

四公斤 sì-gōngjīn four-kilo ‘four kilos

白菜 báicài Chinese.cabbage Chinese cabbage’ 银子 yínzi silver of silver’

苹果 píngguǒ apple of apples’

(b)

两打 liáng-dǎ two-dozen ‘two dozens

鸡蛋 jīdàn egg of eggs’

(d)

五磅 wǔ-bàng five-pound ‘five pounds

茶叶 chá-yè tea-leaf of tea’

(f )

六公升 水 liù-gōngshēng shuǐ six-liter water ‘six liters of water’

3.1.2.2.  Collective measures Collective measure words resemble collective nouns in English. In most cases, specific collective measures are used for nouns denoting humans and inanimate beings, whereas nouns denoting animate beings such as mammals and insects typically occur with qún 群 ‘a group’, as illustrated below: 274

Chinese: Parts of speech

(102)

(103)

(104)

(a)

一班 yī-bān one-bunch ‘a bunch of

年轻 人 niánqīng rén young person young people’

(b)

一队 yī-duì one-file ‘a file of

(c)

一帮 yī-bāng one-group ‘a group of

孩子 háizi child children’

(d)

一股 yī-gǔ one-gang ‘a gang of

(a)

一束 yī-shù one-bunch ‘a bunch of

花 huā flower flowers’

(b)

一串 yī-chuàn one-bunch ‘a bunch of

(c)

一挂 yī-guà one-string ‘a string of

鞭炮 biānpào firecracker firecrackers’

(d)

一笔 yī-bǐ one-sum ‘a sum of

钱 qián money money’

(a)

一群 yī-qún one-flock ‘a flock of

羊 yáng sheep sheep’

(b)

一群 yī-qún one-herd ‘a herd of

牛 niú cow cows’

(c)

一群 yī-qún one-swarm ‘a swarm of

(d)

一群 yī-qún one-swarm ‘a swarm of

蜜蜂 mìfēng bee bees’

战士 zhànshì soldier soldiers’ 土匪 túfěi bandit bandits’ 钥匙 yàoshí key keys’

蚂蚁 máyǐ ant ants’

However, in some cases, nouns denoting humans, animate beings or inanimate beings may occur with the same collective measure words. Two typical examples are qún 群 and pī 批: the former denotes ‘a crowd’ or ‘a group’, which is compatible with nouns denoting humans or animate beings such as mammals, fish and insects, as shown in (105a–f ), whereas the latter denotes ‘a group’, ‘a batch’ or ‘a pile’ and is compatible with nouns denoting humans or inanimate beings, as in (106a–d). (105)

(a)

一群 yī-qún one-crowd ‘a crowd of

(c)

一群 yī-qún one-pack ‘a pack of

(e)

一群 yī-qún one-swarm ‘a swarm of

人 rén person people’

狼 láng wolf wolves’ 蜜蜂 mìfēng bee bees’

(b)

一群 yī-qún one-crowd ‘a crowd of

(d)

一群 yī-qún one-shoal ‘a shoal of

(f )

一群 yī-qún one-swarm ‘a swarm of

275

小孩 xiǎohái child children’

鱼 yú fish fish’ 蚂蚁 máyǐ ant ants’

Candice Chi-Hang Cheung

(106)

(a)

一批 yī-pī one-group ‘a group of

人 rén person people’

(b)

一批 yī-pī one-group ‘a group of

(c)

一批 yī-pī one-batch ‘a batch of

货 huò goods goods’

(d)

一批 yī-pī one-pile ‘a pile of

学生 xuésheng student students’

文件 wénjiàn document documents’

In addition, there are specific collective measures such as duì 对 and shuāng 双, both meaning ‘a pair’, that are used with nouns, which denote humans, animate or inanimate entities that come in pairs: (107)

(108)

(a)

一对 夫妇 yī-duì fū-fù one-pair husband-wife ‘a couple’

(b)

一对 yī-duì one-pair ‘a pair of

天鹅 tiān’é swan swans’

(c)

一对 yī-duì one-pair ‘a pair of

(d)

一对 yī-duì one-pair ‘a pair of

耳环 ěr-huán ear-ring ear-rings’

(a)

一双 手 yī-shuāng shǒu one-pair hand ‘a pair of hands’

(b)

一双 眼睛 yī-shuāng yǎnjing one-pair eye ‘a pair of eyes’

(c)

两双 liǎng-shuāng two-pair ‘two pairs of

(d)

三双 sān-shuāng three-pair ‘three pairs

枕头 zhěntóu pillow pillows’

鞋 xié shoe shoes’

袜子 wàzi sock of socks’

Note that even though there are nouns that are treated in pairs in English (e.g. a pair of trousers, a pair of scissors, a pair of spectacles, etc.), they are incompatible with duì 对 or shuāng 双 in Chinese, as evidenced by the ill-formedness of (109a, 110a, 111a). Instead, they are classified by sortal classifiers, as shown in (109b, 110b, 111b). (109)

(a)

*一对/*一双 *yī-duì/*yī-shuāng one-pair/one-pair Intended: ‘a pair of

裤子 kùzi trousers trousers’

(b)

一条 裤子 yī-tiáo kùzi one-Cl trousers ‘a pair of trousers’

(110)

(a)

*一对/*一双 *yī-duì/*yī-shuāng one-pair/one-pair Intended: ‘a pair of

剪刀 jiǎndāo scissors scissors’

(b)

一把 剪刀 yī-bǎ jiǎndāo one-Cl scissors ‘a pair of scissors’

276

Chinese: Parts of speech

(111)

(a)

*一对/*一双 *yī-duì/*yī-shuāng one-pair/one-pair Intended: ‘a pair of

眼镜 (b) yǎnjìng spectacles spectacles’

一副 眼镜 yī-fù yǎnjìng one-Cl spectacles ‘a pair of spectacles’

Collective measures may also express indeterminate numbers or amounts through the use of yīxiē 一些 ‘some’ and yīdiǎn 一点 ‘a little’: the former is applicable to count and mass nouns (112a–b), whereas the latter is applicable to mass nouns (113a–b) (see section 2.1.1.2). (112)

(a)

一些 yīxiē some ‘some

(113)

(a)

一点 yīdiǎn a.little ‘a little

朋友 péngyǒu friend friends’ 酒 jiǔ wine wine’

(b)

一些 yīxiē some ‘some

(b)

一点 yīdiǎn a.little ‘a little

酒 jiǔ wine wine’ 牛奶 niúnǎi milk milk’

3.1.2.3.  Container measures Container measure words are typically used with nouns denoting food and drinks. Typical examples include bēi 杯 ‘cup’, píng 瓶 ‘bottle’, hé 盒 ‘box’, guō 锅 ‘pot’, wǎn 碗 ‘bowl’, xiāng 箱 ‘box’, guàn 罐 ‘can’, hú 壶 ‘(tea) pot’, dài 袋 ‘bag’, bāo 包 ‘packet’, among many others (Chao 1968, C. N. Li and Thompson 1981: 109, D. Zhu 1982). Some illustrative examples are given below: (114)

(a)

一杯 yī-bēi one-cup ‘a cup of

(b)

两瓶 liǎng-píng two-bottle ‘two bottles

(c)

三盒 巧克力 sān-hé qiǎokèlì three-box chocolate ‘three boxes of chocolate’

(d)

一锅 yī-guō one-pot ‘a pot of

(e)

四碗 饭 sì-wǎn fàn four-bowl rice ‘four bowls of rice’

(f )

六箱 liù-xiāng six-box ‘six boxes

茶 chá tea tea’

酒 jiǔ wine of wine’

炖肉 dùnròu stew stew’ 书 shū book of books’

Note that these container measure words can be used as nouns in their own right, in which case they are classified by sortal classifiers and some of them must be followed by the suffix -zi 子 (see C. N. Li and Thompson 1981: 42–3 for detailed discussion of -zi 子; see also Cheng 2009), as shown below: 277

Candice Chi-Hang Cheung

(115)

(a)

一个 杯子 yī-ge bēizi one-Cl cup ‘a cup’

(b)

两个 瓶子 liǎng-ge píngzi two-Cl bottle ‘two bottles’

(c)

三个 盒子 sān-ge hézi three-Cl box ‘three boxes’

(d)

一个 yī-ge one-Cl ‘a pot’

(e)

四个 碗 sì-ge wǎn four-Cl bowl ‘four bowls’

(f )

六个 箱子 liù-ge xiāngzi six-Cl box ‘six boxes’

锅 guō pot

Another type of container measure may employ body parts or enclosed areas as temporary ‘containers’. They are typically preceded by the numeral yī 一 ‘one’ and can be optionally followed by de 的, as demonstrated below (Chao 1968, C. N. Li and Thompson 1981: 111–12): (116)

(a)

一脸 yī-liǎn one-face ‘a faceful

(的) 灰 (de) huī DE dust of dust’

(b)

一屋子 yī-wūzi one-house ‘a houseful

 

(c)

一头 yī-tóu one-head ‘a headful

发 (的) 白 (de) bái fà DE white hair of grey hair’

(d)

一地 yī-dì one-floor ‘a floorful

(的) 面粉 (de) miànfěn DE flour of flour’

 

(e)

一肚子 yī-dùzi one-stomach ‘a stomachful

(f )

一桌 yī-zhuō one-table ‘a tableful

(的) 客人 (de) kèrén DE guest of guests’

(的) 气 (de) qì DE anger of grievance’

(的) 烟 (de) yān DE smoke of smoke’

Since any nouns denoting body parts or enclosed areas can serve as temporary container measures, this type of container measure forms an open class. This derivation is akin to the suffixation of -ful in English, which is applicable to nouns to derive measure words (e.g. house → houseful, head → headful, etc.)

3.1.2.4. Generic measures Generic measure words, denoting types and kinds, can be used with count nouns, mass nouns and abstract nouns. Typical examples are zhǒng 种 ‘kind’, lèi 类 ‘category’ and yàng 样 ‘type’, as demonstrated below: (117)

(a)

一种 yī-zhǒng one-kind ‘a kind of

水果 shuíguǒ fruit fruit’

(b)

278

两种 liáng-zhǒng two-kind ‘two kinds of

蔬菜 shūcài vegetable vegetables’

Chinese: Parts of speech

 

(c)

四种 酒 sì-zhóng jiǔ four-kind wine ‘four kinds of wine’

(d)

一种 yī-zhǒng one-kind ‘a kind of

病 bìng illness illness’

(118)

(a)

这类 zhè-lèi this-species ‘this species

动物 dòngwù animal of animal’

(b)

这类 zhè-lèi this-kind ‘this kind

红酒 hóng-jiǔ red-wine of red wine’

 

(c)

这类 zhè-lèi this-genre ‘this genre

问题 wèntí problem of problems’

(d)

这类 zhè-lèi this-kind ‘this kind

角色 juésè role of role’

(119)

(a)

几样 jǐ-yàng several-kind ‘several kinds

东西 dōngxi thing of things’

(b)

各样 gè-yàng different-kind ‘several kinds

 

(c)

几样 jǐ-yàng several-kind ‘several kinds

饭菜 fàn-cài rice-vegetable of food’

(d)

四样 食物 sì-yàng shíwù four-kind food ‘four kinds of food’

商品 shāngpǐn commodity of commodities’

3.2. Localizers Localizers (also referred to as fāngwèicí 方位词 in Zhu 1982: 40) express the spatial location of the entities denoted by nouns. Most of the localizers are monosyllabic and they include shàng 上 ‘up’, xià 下 ‘down’, qián 前 ‘front’, hòu 后 ‘back’, lǐ 里 ‘inside’, wài 外 ‘outside’, páng 旁 ‘side’, nèi 内 ‘inside’, etc. (Chao 1968, D. Zhu 1982: Ch. 4.4, C. N. Li and Thompson 1981: Ch. 11, Chappell and Peyraube 2008). Most of the disyllabic localizers are derived from the monosyllabic ones through the addition of suffixes such as -bian 边, -mian 面 and -tou 头 (Chao 1968, Zhu 1982: Ch. 4.4, C. N. Li and Thompson 1981: Ch. 11, Chappell and Peyraube 2008). The representative localizers are listed in Table 15.3 (adapted from D. Zhu 1982: 40–1 and C. N. Li and Thompson 1981: 391). In terms of distribution, localizers are preceded by a noun phrase, as shown in the following generalized schema (Chao 1968; see also C. N. Li and Thompson 1981: Ch. 11): (120) Noun Phrase + Localizer Some illustrative examples are given below: (121)

(a)

他 tā he ‘He

 

(b)

车子 停 在 chēzi tíng zài car stop at ‘The car is parked

在 zài at fixed

房子 里 修理 fángzi li xiūlǐ room inside repair the television inside the

电视机。 diànshìjī. television room’

学校 外边。 xuéxiào wàibian. school outside outside the school’ 279

Candice Chi-Hang Cheung Table 15.3  Localizers in Chinese Localizers

Gloss

shàng 上 xià 下 lǐ 里 wài 外 qián 前 hòu 后 páng 旁 zhōngjiān 中间 dōngbù 东部 nánbù 南部 xībù 西部 běibù 北部 nèi 内

shàngbian 上边 xiàbian 下边 lǐbian 里边 wàibian 外边 qiánbian 前边 hòubian 后边 pángbian 旁边

shàngmian 上面 xiàmian 下面 lǐmian 里面 wàimian 外面 qiánmian 前面 hòumian 后面

zuǒbian 左边 yòubian 右边 dōngbian 东边 nánbian 南边 xībian 西边 běibian 北边

zuǒmian 左面 yòumian 右面 dōngmian 东面 nánmian 南面 xīmian 西面 běimian 背面

shàngtou 上头 xiàtou 下头 lǐtou 里头 wàitou 外头 qiántou 前头 hòutou 后头

‘on top of, above’ ‘under, below’ ‘in, inside’ ‘outside’ ‘in front of ’ ‘in back of, behind’ ‘beside’ ‘in the center of ’ ‘left of ’ ‘right of ’ ‘east of ’ ‘south of ’ ‘west of ’ ‘north of ’ ‘inside’

One of the important debates on localizers is their categorial status. As noted by Chappell and Peyraube (2008), in regard to the categorial status of localizers, almost every possibility has been entertained in the literature. For instance, they have been analyzed as adjectives (Ma 1898: Ch. 3), adverbs (Lü 1947, Jingxi Li and Liu 1955), nominal suffixes (Cartier 1972), postpositions forming a discontinuous constituent with the prepositions they frequently co-occur with (Hagège 1975: 220 ff., Peyraube 1980: 53 ff.; see also Tai 1973, Ernst 1988), spatial clitics (F.-H. Liu 1998, C. Sun 2008) and even pronouns (Rygaloff 1973: 143). However, one of the dominant views is that localizers are best analyzed as a subclass of nouns (D. Zhu 1982, Y.-H. A. Li 1990, C.-T. J. Huang, Li and Li 2009: Ch. 1). One piece of evidence comes from the fact that the noun phrase and the following localizer, similar to noun phrases, can function as the subject (122a) or object (122b) (C.-T. J. Huang, Li and Li 2009: Ch. 1; see also Y. Li 1983, 2003; Y.-H. A. Li 1990). (122)

(a)

城 外 很 chéng wài hén city outside very ‘The outside of the city

 

(b)

他 tā he ‘He

去过 qù-guo go-Exp has been

美丽。 měilì. beautiful is beautiful’

城 外。 chéng wài. city outside to the outside of the city’

Another piece of evidence is that the noun phrase plus the localizer, similar to noun phrases, can be preceded by the preposition zài 在 ‘at’ (C.-T. J. Huang, Li and Li 2009: Ch. 1), as demonstrated below: 280

Chinese: Parts of speech

(123)

(a)

他 tā he ‘He

在 zài at fixed

房 里 修理 电视机。 fáng li xiūlǐ diànshìjī. room inside repair television the television inside the room’

 

(b)

他 tā he ‘He

在 zài at fixed

这个 房子 zhè-ge fángzi this-Cl room the television in

修理 电视机。 xiūlǐ diànshìjī. repair television this room’

Despite these similiarities, there are reasons for distinguishing localizers from nouns. For one thing, unlike nouns that can be preceded by a [Numeral + Classifier] sequence (see section 2.1), localizers cannot. For another, if localizers were nouns, it would wrongly predict that they can always function as the subject or object on their own without the preceding noun phrase:16 (124)

(a)

*外 很 *wài hén outside very Intended: ‘The

美丽。 měilì. beautiful outside is beautiful’

(b)

*他 去过 *tā qù-guo he go-Exp Intended: ‘He

外。 wài. outside has been outside’

Based on these reasons, I adopt the view that localizers should be regarded as an autonomous category in line with Chongxing Li (1992) and Z. Chu (1997, 2006), among others. The fact that localizers, in particular the disyllabic ones, share similar properties with nouns may be attributed to the fact that they are historically derived from nouns (Chappell and Peyraube 2008).

3.3.  Sentence-final particles Sentence-final particles are one of the most distinctive features of Chinese. As its name implies, they are used in the sentence-final position, serving various semantic and pragmatic functions. There are roughly six sentence-final particles in Chinese, including le 了, ne 呢, ma 吗, ba 吧, ou 呕 and a/ya 啊/呀 (C. N. Li and Thompson 1981: Ch. 7; see also Chao 1968, D. Zhu 1982), which I shall discuss in detail in the following sections.

3.3.1. le 了 As noted by C. N. Li and Thompson (1981: Ch. 7), the sentence-final le 了 is different from other sentence-final particles in that it can be followed by other sentence-final particles, such as a 啊, ou 呕 and the yes–no question particle ma 吗, as illustrated below: (125)

Q:

他 tā he ‘Is

 

A:

他 tā he ‘Of

是 学生 shì xuésheng be student he a student?’ 当然 是 dāngrán shì of.course be course he is!’

吗？ ma? Q 了 le Prt

啊。 a. Prt

281

Candice Chi-Hang Cheung

(126)

我 告诉 他 那件 事 wǒ gàosù tā nà-jiàn shì I tell him that-Cl matter ‘I told him about that matter’

(127)

他 去 tā qù he go ‘Did he

买 东西 mǎi dōngxi buy thing go shopping?’

了 le Prt

了 le Prt

呕。 ou. Prt

吗？ ma? Q

In terms of its semantic and pragmatic functions, it has been noted that the sentence-final le 了 can mark a change of state or indicate what is expressed is contrary to expectation (Chao 1968, C. N. Li and Thompson 1981, Ross 1995, Sybesma 1999, Soh 2009). The change of state reading becomes more transparent when we compare sentences with and without the sentence-final le 了:17 (128)

(a)

他 tā he ‘He

当了 父亲。 dāng-le fùqīn. become-Perf father became a father’

 

(b)

他 tā he ‘He

当了 父亲 了。 dāng-le fùqīn le. become-Perf father Prt is now a father. (He wasn’t one before and now he has a child)’

(129)

(a)

他 tā he ‘He

不 怕 鬼。 bú pà guǐ. not fear ghost is not afraid of ghosts’

 

(b)

他 tā he ‘He

不 bú not is no

怕 鬼 pà guǐ fear ghost longer afraid

了。 le. Prt of ghosts (and he was before)’

The examples in (128a, 129a) are simple declarative sentences, while the addition of the sentence-final le 了 in (128b, 129b) conveys a change of state. More precisely, the addition of the sentence-final le 了 in (128b) signals that the person referred to as tā 他 ‘he’ was not a father before and has become a father. In (129b), the sentence-final le 了 signals a change of state from being afraid of ghosts to no longer being afraid of ghosts. As for the contrary to expectation reading, consider the following pair of sentences with and without the sentence-final le 了 (taken from Soh 2009: 632). (130)

(a)

这片 西瓜 很 甜。 不必 zhè-piàn xīguā hěn tián. Bú bì this-Cl watermelon very sweet not need ‘This watermelon is sweet. It is not necessary to 282

加 糖。 jiā táng. add sugar add sugar’

Chinese: Parts of speech

 

(b) 这片 西瓜 很 甜 了。 不必 加 糖 了。 zhè-piàn xīguā hěn tián le. Bú bì jiā táng le. this-Cl watermelon very sweet Prt not need add sugar Prt ‘This watermelon is sweet (contrary to what one may assume/expect). It is not necessary to add sugar (contrary to what one may assume/expect)’

(130a) is a simple declarative sentence asserting that the watermelon is sweet. In contrast, the speaker uttering (130b) not only makes the same assertion, but also implies that someone in the discourse context may think otherwise and the sentence is intended to correct this wrong assumption (see C. N. Li and Thompson 1981: 263–70).

3.3.2. ne 呢 The sentence-final ne 呢 can be used in declarative and interrogative sentences. In declarative sentences, it is mainly used to indicate that the sentence is the speaker’s response to some previous claim, expectation or belief on the part of the hearer. Pragmatically, it has the effect of drawing the hearer’s attention to the information conveyed by the sentence (C. N. Li and Thompson 1981: Ch. 7; see also Chao 1968, M. Hu 1981, C.-H. C. Chu 1984, 1985a, 1985b, 1998, C.-J. W. Lin 1984, King 1986, Shao 1989, G. Wu 2005). To highlight the function of ne 呢, let us consider the following pair of sentences: (131)

(a)

他们 有 五条 tāmen yóu wǔ-tiáo they have five-Cl ‘They have five cows’

牛。 niú. cow

 

(b)

他们 有 tāmen yóu they have ‘(Listen,) they

牛 呢。 niú ne. cow Prt cows’

五条 wǔ-tiáo five-Cl have five

(131a) is a simple declarative sentence asserting that the people referred to as ‘they’ have five cows. In contrast, the speaker uttering (131b) not only makes the same assertion, but also draws the hearer’s attention to this piece of information, which could be a perfect response if the hearer has claimed that the people referred to as ‘they’ do not have anything in the previous discourse. In such a context, (131b) could be construed as a response challenging the hearer’s expectation. Apart from declarative sentences, the sentence-final ne 呢 can be used in four types of interrogative sentences, including A-not-A questions (132), disjunctive questions (133), wh-questions (134) and truncated questions consisting of only one noun (135B) (C. N. Li and Thompson 1981: Ch. 7; see also M. Hu 1981, C.-H. C. Chu 1984, 1985a, 1985b, 1998, King 1986, Shao 1989, Jin 1996). (132)

他们 来不来 tāmen lái-bu-lái they come-not-come ‘Will they come?’

呢？ ne? Q

283

Candice Chi-Hang Cheung

(133)

你 nǐ you ‘Do

(134)

他们 tāmen they ‘When

(135)

A:

他 tā he ‘He

B:

你   呢？ nǐ   ne? you Q ‘How about you?’

想 喝 xiǎng hē want drink you want tea or 什么 shénme what will they

茶 还是 chá háishi tea or coffee?’

时候 shíhòu time come?’

明天 míngtiān tomorrow needs to go

来 lái come

咖啡 kāfēi coffee

呢？ ne? Q

呢？ ne? Q

要 去 学校。 yào qù xuéxiào. need go school to school tomorrow’

3.3.3. ma 吗 Unlike ne 呢 that can be used in A-not-A questions, disjunctive questions, wh-questions and truncated questions, ma 吗 is confined to yes–no questions (C. N. Li and Thompson 1981: Ch. 7, D. Zhu 1982), as illustrated below: (136)

你 nǐ you ‘Do

想 她 吗？ xiǎng tā ma? miss her Q you miss her?’

3.3.4. ba 吧 According to C. N. Li and Thompson (1981: Ch. 7), the sentence-final ba 吧 serves to signal the speaker’s desire to solicit the approval or agreement from the hearer with regard to the information conveyed by the sentence (Chao 1968, D. Zhu 1982: Ch. 16; see Y. Han 1995, C.-H. C. Chu 1998, Y.-H. A. Li 2006 for more detailed discussion and the different analyses of the sentence-final ba 吧). It can be used in different types of speech acts, such as commands (137) and advice (138) (the following examples are taken from C. N. Li and Thompson 1981: 307–8). (137)

我们 走 wǒmen zǒu we go ‘Let’s go!’

(138)

你 想一想 nǐ xiǎng-yi-xiǎng you think-one-think ‘Why don’t you think

吧！ ba! Prt 吧！ ba! Prt about it a little?’ 284

Chinese: Parts of speech

The sentence-final ba 吧 can also be used in declarative sentences to solicit agreement or confirmation from the hearer with respect to the information conveyed by the sentence: (139)

她 tā she ‘She

很 漂亮 吧！ hěn piàoliang ba! very pretty Prt is pretty, don’t you agree?’

For instance, the speaker uttering (139) is intending to solicit agreement from the hearer that the person referred to as ‘she’ is pretty. This function of ba 吧 is comparable to the function of tag questions in English (e.g. She is pretty, isn’t she?), which is intended to seek confirmation of a statement from the hearer. Finally, the sentence-final ba 吧 can also be used in interrogative sentences to seek confirmation from the hearer (D. Zhu 1982: Ch. 16): (140)

今天 jīntiān today ‘Today

是 星期天 吧？ shì Xīngqītiān ba? be Sunday Prt is Sunday, isn’t it?’

The most natural context in which (140) is used is when the speaker is fairly certain that today is Sunday and would like to seek further confirmation from the hearer.

3.3.5. ou 呕 As noted by C. N. Li and Thompson (1981: Ch. 7), the sentence-final ou 呕 is used to indicate that the sentence preceding ou 呕 is intended to be construed as a friendly warning showing concern and care on the part of the speaker (see also Chao 1968). In light of its semantic function, ou 呕 can be used to soften a command and convert it into a concerned warning: (141)

小心 呕！ xiǎoxīn ou! careful Prt ‘Be careful, okay?’

The sentence-final ou 呕 can also appear in conditionals when they are construed as concerned warnings (C. N. Li and Thompson 1981: Ch. 7): (142)

如果 你 吃 rúguǒ nǐ chī if you eat ‘Let me tell you, if

太多 就 会 肚子 疼 呕。 tàiduō jiù huì dùzi téng ou. too.much then will stomach hurt Prt you eat too much, you’ll have a stomachache’

Due to the semantic nature of ou 呕, it is commonly found in the speech of an adult addressing a child (C. N. Li and Thompson 1981: Ch. 7). For instance, it is perfectly natural for a mother to utter (142) to warn her child while expressing her concern that her child might have eaten too much. 285

Candice Chi-Hang Cheung

3.3.6. a/ya 啊/呀 According to C. N. Li and Thompson (1981: Ch. 7), the sentence-final a/ya 啊/呀 may function to reduce the forcefulness of the information conveyed by a sentence (Chao 1968, D. Zhu 1982, Dow 1983; see Shie 1991, C.-H. C. Chu 2002, R.-J. R. Wu 2004 for detailed discussion of the discourse functions of the sentence-final a 啊). When used in interrogative sentences such as A-not-A questions (143), wh-questions (144) and truncated questions with a wh-word alone (145), the addition of a/ya 啊/呀 has the effect of softening the query. (143)

你 nǐ you ‘Will

(144)

你 要 去 哪儿 啊/呀？ nǐ yào qù nǎr a/ya? you want go where Prt/Prt ‘Where do you want to go?’

(145)

谁 shéi who ‘Who

去不去 qù-bu-qù go-not-go you go?’

啊/呀？ a/ya? Prt/Prt

啊/呀？ a/ya? Prt/Prt is it?’

Apart from interrogative sentences, a/ya 啊/呀 can be used in imperative sentences, serving to reduce the forcefulness of a command (C. N. Li and Thompson 1981: Ch. 7). This becomes more transparent if we compare the following pair of sentences with and without a/ya 啊/呀: (146)

(a)

你 nǐ you ‘You

(b) 你 nǐ you ‘You

过来！ guòlái! come come here!’ 过来 啊/呀！ guòlái a/ya! come Prt/Prt come here’

(146a) is a simple imperative sentence expressing a command. With the addition of a/ya 啊/呀 in (146b), it is interpreted as a suggestion or an encouragement rather than a command.

4. Conclusion In this chapter, I have provided a comprehensive survey of the categories in Chinese. In particular, we have seen that Chinese has six major categories, including nouns, verbs, adjectives, adverbs, prepositions and conjunctions, which are also available in the Indo-European languages. In addition, I have shown that Chinese possesses three categories that are absent in the Indo-European languages, namely, classifiers, localizers and sentence-final particles. 286

Chinese: Parts of speech

These findings suggest that while some basic categories such as nouns, verbs, adjectives, adverbs, prepositions and conjunctions can be considered as universal, other categories such as classifiers, localizers and sentence-final particles are more language-specific and may be viewed as distinctive features of Sinitic languages. Among the three categories whose categorial statuses are debatable in the literature, I have argued that adjectives and localizers are best analyzed as distinct categories, refuting the alternative views, which take adjectives as a subclass of verbs and localizers as a subclass of nouns. As for prepositions, I have demonstrated that they may either be analyzed as a mixed category consisting of verbs and prepositions or an independent category consisting of prepositions only. Under the latter view, the verbal properties of prepositions can be traced back to their historical origin as verbs.

Acknowledgements I benefit greatly from discussion with Paul Law, Audrey Li, Stephan Matthews and SzeWing Tang. My special thanks go to Hoi-ki Jess Law, Haoze Li, Zheng Wei and Jiahui Yang for consolidating the Chinese data and providing useful comments and suggestions on earlier versions of this chapter. I also thank Zheng Wei for editorial assistance. The research leading to this chapter was sponsored by the CUHK Direct Grant for Research (#4051007) and the Research Fund for Comparative Syntax (#6903134). All the financial support is gratefully acknowledged. As usual, all remaining errors and shortcomings are my own responsibility.

Notes   1 The term ‘classifier’ here is used as a cover term equivalent to ‘numeral classifier’. As I shall discuss in section 3.1, classifiers can be subdivided into sortal classifiers and measure words.   2 The following abbreviations are used in glossing examples: Cl: classifier; Dur: durative aspect marker; Exp: experiential aspect marker; Perf: perfective aspect marker; Poss: possessive marker; Prog: progressive aspect marker; Prt: sentence-final particle; Q: question particle; Res: resultative aspect marker.   3 The Romanization system used for the Chinese data is pīnyīn, and the tones given in the examples are in compliance with the standard tone sandhi rules in Mandarin.  4 As noted by C.-T. Huang, Li and Li (2009: 10–11), there are examples such as bù-rén-bù-guǐ 不人不鬼 ‘not-human-not-ghost’ in Chinese that seemingly allow a noun to be negated by bù 不. I follow C.-T. J. Huang, Li and Li and regard these examples as idiomatic expressions rather than genuine counterexamples.   5 The opacity of bare nouns with respect to number is referred to as ‘transnumerality’ in Greenberg (1974: 25), who proposes that it is correlated with the obligatory presence of classifiers in numeral– classifier languages (see Bisang 1998, 1999).   6 Parentheses are used as a notational convention. Specifically, an expression between a pair of parentheses is regarded as optional. For instance, in (22a–d), the pair of parentheses is used to indicate that the presence of hěn 很 is optional.   7 Note that intransitive verbs can be further divided into unergative and unaccusative verbs. Interested readers are referred to Y.-H. A. Li (1985), Lü (1987) and C.-T. J. Huang (1989, 2007) for detailed discussion.   8 Note that the categorial status of gěi 给 in dative constructions is controversial in the literature: it has been analyzed as a coverb (C. N. Li and Thompson 1981), a preposition (C.-C. J. Tang 1990; S. Zhang 1990; Yang 1991) or a verb (Chao 1968; Y.-H. A. Li 1985, 1990; C.-R. Huang and Mo 1992; C.-R. Huang and Ahrens 1999).   9 As pointed out in Aoun and Li (2003: 251 fn.14), although hěn 很 is generally translated as ‘very’, it is a ‘bleached’ element in the sense that it does not carry much meaning when it precedes an adjective (see S.-Z. Huang 1997, 2006; C.-S. L. Liu 2010; He and Jiang 2011; Grano 2012 for detailed discussion of the function of the intensifier hěn 很).

287

Candice Chi-Hang Cheung 10 Note that although Chinese makes a distinction between predicative and non-predicative adjectives, the inventory of the so-called ‘non-predicative adjectives’ in Chinese is not identical to that in English. Interested readers are referred to Y. Li (1996) for fuller discussion of the properties of non-predicative adjectives in Chinese. 11 Parentheses are used as a notational convention. Specifically, an asterisk ‘*’ that immediately precedes the left parenthesis indicates that the presence of the expression inside the parentheses is mandatory and the deletion of the expression will result in ill-formedness. In (38a–d), this means that the presence of the verb or adjective inside the parentheses is obligatory. 12 There are many alternative ways to classify adverbs in Chinese. Interested readers are referred to Jinxi Li (1924); Wang (1934); Z. Zhang (1959); Y. Hu (1962); Lü (1979); Y. Liu et al. (1983); J. Zhang (1987); and Qian (1990) for alternative means of classifying adverbs and their different categorizations. 13 Another type of adverb that is commonly discussed in the literature includes yòu 又 ‘again’, zài 再 ‘again’, jiù 就 ‘only, immediately, emphatic, thereupon (then), zhǐ 只 ‘only’, cái 才 ‘just now, only then’, hái 還 ‘still’, yě 也 ‘also’ and dōu 都 ‘all’ (C. N. Li and Thompson 1981: Ch. 8; D. Zhu 1982, among others), many of which are associated with multiple meanings. Interested readers are referred to Hole (2004) for detailed discussion of jiù 就, cái 才, hái 還 and yě 也 in relation to information structure and Lee 1986; Cheng 1995; J. Li 1995; J.-W. Lin 1996, 1998; N. N. Zhang 1997: Section 7.5, 2008; Xiang 2008, inter alia for detailed discussion of dōu 都. 14 The absence of gloss here is due to the fact these words have grammaticalized into function words and lack descriptive content. Following C. N. Li and Thompson (1981), bǎ 把 could be regarded as an object marker (see F.-H. Liu 1997; Sybesma 1999; Y.-H. A. Li 2006 for fuller discussion of bǎ 把), whereas bèi 被, jiào 叫, ràng 讓 could be regarded as agent markers that serve to introduce the agent in passive sentences (D. Zhu 1982; see Feng 1995; Ting 1995, 1998; C.-T. J. Huang 1999 for detailed discussion of bèi 被). 15 Note that although sortal classifiers have a tendency to resist the insertion of de 的, as illustrated by the ill-formedness of (82a–d), this should not be taken to mean that sortal classifiers can never allow the insertion of de 的. As revealed by Y.-H. A. Li (2013) and D. Shi (2013), there are many counter-examples showing that de 的 can be inserted between a sortal classifier and the following noun. In particular, as noted in D. Shi (2013), insertion of de 的 becomes possible when the numeral+sortal classifier+de 的 is understood as a modifier of some kind, as shown in (i). (i) 一套 二十本 的 紅樓夢 yí-tào èrshí-běn de Hónglóumèng one-Cl twenty-Cl DE Dream.of.the.Red.Chamber ‘a series of Dream of the Red Chamber that consists of twenty volumes’. 16 Note that the sentences in (124a–b) will become well-formed when the disyllabic localizer wàimian 外面 ‘outside’ is used instead. However, this fact does not suffice to undermine our arguments for differentiating localizers from nouns, since localizers, whether monosyllabic or disyllabic, are always incompatible with [Numeral + Classifier] sequences. 17 As shown in (128b), the sentence-final le 了 may co-occur with the perfective aspect marker -le 了. Interested readers are referred to Chao (1968), Rohsenow (1978), C. N. Li and Thompson (1981), C. N. Li, Thompson and Thompson (1982), Ross (1995), Sybesma (1999), J.-W. Lin (2003), Soh (2009), among others for detailed discussion of the similarities and differences between the two and the different analyses of the sentence-final le 了.

References Ahrens, Kathleen (1994) ‘Classifier Production in Normals and Aphasics’, Journal of Chinese Linguistics 22(2): 202–47. Aoun, Joseph and Li Yen-hui Audrey (2003) Essays on the Representational and Derivational Nature of Grammar: The Diversity of Wh-constructions, Cambridge, MA: MIT Press. Bisang, Walter (1998) ‘Adverbiality: The view from the Far East’, in Johan van der Auwera and Dónall Ó Baoill (eds) Adverbial Constructions in the Languages of Europe, Berlin-New York: Mouton de Gruyter, 641–812.

288

Chinese: Parts of speech Bisang, Walter (1999) ‘Classifiers in East and Southeast Asian Languages: Counting and beyond’, in Jadranka Gvozdanović (ed) Numeral Types and Changes Worldwide, Berlin: Mouton de Gruyter, 113–85. Cartier, Alice (1972) ‘Les Indicateurs de Lieu en Chinois’, La linguistique 8(2): 83–101. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley: University of California Press. Chappell, Hilary (1999) ‘The Double Unaccusative Construction in Sinitic Languages’, in Doris L. Payne and Immanuel Barshi (eds) External Possession, Amsterdam and Philadelphia: John Benjamins, 195–228. Chappell, Hilary and Alain Peyraube (2008) ‘Chinese Localizers: Diachrony and Some Typological Considerations’, in Dan Xu (ed) Space in Languages of China: Cross-linguistic, Synchronic and Diachronic Perspectives, Dordrecht: Springer, 15–38. Cheng, Lisa Lai-Shen (1995) ‘On Dou-quantification’, Journal of East Asian Linguistics 4: 197–234. Cheng, Lisa Lai-Shen (2009) ‘Counting and Classifiers’, Handout distributed at Mass/Count Workshop, University of Toronto. Cheng, Lisa Lai-Shen and Rint Sybesma (1999) ‘Bare and Not-so-bare Nouns and the Structure of NP’, Linguistic Inquiry 30: 509–42. Cheung, Candice Chi-Hang (2012) ‘Adjectival Modification in Mandarin: Evidence for the Adjunction Analysis’, Studies in Chinese Linguistics 33(1): 41–62. Cheung, Hung-nin Samuel, in collaboration with Liu Sze-yun and Shih Li-lin (1994) A Practical Chinese Grammar, Hong Kong: Chinese University Press. Chierchia, Gennaro (1998) ‘Reference to Kinds across Languages’, Natural Language Semantics 6: 339–405. Chu, Cheng-Hsi Chauncey (1984) ‘Beef It Up With ne’, Journal of the Chinese Language Teachers Association 19(3): 87–91. Chu, Cheng-Hsi Chauncey (1985a) ‘How Would You Like Your ne Cooked?’ Journal of the Chinese Language Teachers Association 20(3): 71–8. Chu, Cheng-Hsi Chauncey (1985b) ‘Pragmatics and Teaching of Chinese’, Proceedings of the First International Conference on Teaching Chinese as a Foreign/Second Language, Taipei: World Chinese Language Association, 223–40. Chu, Cheng-Hsi Chauncey (1998) A Discourse Grammar of Mandarin Chinese, New York and Berne: Peter Lang. Chu, Cheng-Hsi Chauncey (2002) ‘Relevance Theory, Discourse Markers and the Mandarin Utterancefinal Particle a/ya’, Journal of the Chinese Language Teachers Association 37(1): 1–42. Chu, Zexiang (1997) ‘Xiàndài Hànyǔ de Mìngmíngxíng Chùsuǒcí’ (‘A Study of Nomenclative Place Words in Contemporary Chinese’), Zhōngguó Yǔwén (Studies of the Chinese Language) 5: 326–35. Chu, Zexiang (2006) ‘Hànyǔ Chùsuǒcí de Cílèi Dìwèi jíqí Lèixíngxué Yìyì’ (‘The Word-class status of Place Words in Chinese and Its Typological Significance’), Zhōngguó Yǔwén (Studies of the Chinese Language) 3: 216–24. Dow, Daoming Francis (1983) Hănyŭ Yŭfă (Chinese Grammar), Edinburgh: University of Edinburgh, Department of Chinese. Ernst, Thomas (1988) ‘Chinese Postpositions?–Again’, Journal of Chinese Linguistics 16(2): 219–45. Feng, Shengli (1995) ‘The Passive Construction in Chinese’, Studies in Chinese Linguistics 1: 1–28. Gasde, Horst-Dieter and Waltraud Paul (1996) ‘Functional Categories, Topic Prominence and Complex Sentences in Mandarin Chinese’, Linguistics 34(2): 263–94. Grano, Thomas (2012) ‘Mandarin hěn and Universal Markedness in Gradable Adjectives’, Natural Language and Linguistic Theory 30: 513–65. Greenberg, Joseph (1974) ‘Numeral Classifiers and Substantival Number: Problems in the Genesis of a Linguistic Type’, in Luigui Heilmann (ed) Proceedings of the 11th International Congress of Linguistics, Bologna: Mulino, 17–37. Hagège, Claude (1975) Le Problème Linguistique des Prépositions et la Solution Chinoise, Louvain: Peeters. Han, Jingquan (2000) ‘Lǐngyǒu Míngcí Tíshēng Yíwèi yū Gělǐlùn’ (‘A Study of Possessor Raising Movement in Relation to Case Theory’), Xiàndài Wàiyǔ (Modern Foreign Languages) 3: 261–72. Han, Yang (1995) ‘A Pragmatic Analysis of the ba Particle in Mandarin Chinese’, Journal of Chinese Linguistics 23(2): 99–127. He, Chuansheng and Yan Jiang (2011) ‘Type Shifting, Chinese hen+N Structure, and Implications for Semantic Parameters’, Lingua 121: 890–905.

289

Candice Chi-Hang Cheung Heine, Bernd, Claudi Ulrike, and Friederike Hünnemeyer (1991) Grammaticalization: A Conceptual Framework, Chicago: University of Chicago Press. Hole, Daniel (2004) Focus and Background Marking in Mandarin Chinese: System and Theory Behind cái, jiù, dōu and yě, London and New York: Routledge Curzon. Hole, Daniel (2005) ‘Extra Argumentality: A Binding Account of “Possessor Raising” in German, English and Mandarin’, in Ji-Yung Kim, Yury A. Lander and Babara Hall Partee (eds) Possessives and Beyond: Semantics and Syntax, Amherst, MA: GLSA Publications, 365–86. Hopper, Paul J. and Elizabeth Closs Traugott (1993) Grammaticalization, Cambridge: Cambridge University Press. Hu, Mingyang (1981) ‘Běijīnghuà de Yǔqì Zhùcí hé Tàncí’ (‘Mood-helping-words and Interjections in Beijing Dialect’) Zhōngguó Yǔwén (Studies of the Chinese Language) 5/6: 347–50, 416–23. Hu, Yushu (ed) (1962) Xiàndài Hànyǔ (Modern Chinese), Shanghai: Jiaoyu Chubanshe. Huang, Chu-Ren and Kathleen Ahrens (1999) ‘The Function and Category of gěi in Mandarin Ditransitive Constructions’, Journal of Chinese Linguistics 27(2): 1–26. Huang, Chu-Ren and Kathleen Ahrens (2003) ‘Individuals, Kinds and Events: Classifier Coercion of Nouns’, Language Sciences 25: 353–73. Huang, Chu-Ren and Ruo-ping Mo (1992) ‘Mandarin Ditransitive Constructions and the Category of gěi’, in Proceedings of the 18th Berkeley Linguistic Society, Berkeley: Berkeley Linguistic Society, 109–122. Huang, C.-T. James (1989) ‘Zhōngwén de Liǎngzhǒng Jíwù Dòngcí he Liǎngzhǒng bù Jíwù Dòngcí’ (‘Two Kinds of Transitive Verbs and Intransitive Verbs’), in Proceedings of the 2nd World Congress of Chinese Language Studies, Taipei: World Chinese Language Association, 39–59. Huang, C.-T. James (1999) ‘Chinese Passives in Comparative Perspective’, Tsing Hua Journal of Chinese Studies 29: 423–510. Huang, C.-T. James (2007) ‘Hànyǔ Dòngcí de Tíyuánjiégòu jíqí Jùfā Biǎoxiàn’ (‘Thematic Structures of Verbs in Chinese and their Syntactic Projections’), Yǔyán Kēxué (Linguistic Sciences) 4: 3–21. Huang, C.-T. James, Y.-H. Audrey Li and Yafei Li (2009) The Syntax of Chinese, Cambridge: Cambridge University Press. Huang, Shi-Zhe (1997) ‘Some Remarks on Adjectives in Mandarin Chinese’, Paper presented at the 6th Annual Conference of the International Association of Chinese Linguistics (IACL-6), Sinological Institute, Leiden University. Huang, Shi-Zhe (2006) ‘Property Theory, Adjectives and Modification in Chinese’, Journal of East Asian Linguistics 15: 343–69. Iljic, Robert (1994) ‘Quantification in Mandarin Chinese: Two Markers of Plurality’, Linguistics 32: 91–116. Iljic, Robert (2001) ‘The Origin of the Suffix -men in Chinese’, Bulletin of the School of Oriental and African Studies 64(1): 74–97. Jin, Lixin (1996) ‘Guānyú Yíwènjù zhōng de ne’ (‘On ne in Interrogatives’), Yǔyán Jiàoxué yǔ Yánjiū (Language Teaching and Linguistic Studies) 4: 104–9. King, Brian (1986) ‘Ne: A Discourse Approach’, Journal of the Chinese Language Teachers Association 21(1): 21–46. Kuo, Yi-jun (1998) ‘A Semantic and Contrastive Analysis of Mandarin and English Measure Words’, Master Thesis, National Chung Cheng University. Lee, Hun-Tak Thomas (1986) ‘Studies on Quantification in Chinese’, Doctoral Dissertation, University of California, Los Angeles. Li, Boya (2006) ‘Chinese Final Particles and the Syntax of the Periphery’, Doctoral Dissertation, Leiden University, Netherlands. Li, Charles N. and Sandra A. Thompson (1974a) ‘Co-verbs in Mandarin Chinese: Verbs or Prepositions?’ Journal of Chinese Linguistics 2(3): 257–78. Li, Charles N. and Sandra A. Thompson (1974b) ‘A Linguistic Discussion of the “Co-verb” in Chinese Grammar’, Journal of the Chinese Language Teachers Association 9(3): 109–19. Li, Charles N. and Sandra A. Thompson (1981) Mandarin Chinese: A Functional Reference Grammar, Berkeley: University of California Press. Li, Charles N., Sandra A. Thompson, and R. McMillan Thompson (1982) ‘The Discourse Motivation for the Perfect Aspect: the Mandarin Particle le’, in Paul Hopper (ed) Tense-aspect: Between Semantics and Pragmatics, Amsterdam and Philadelphia: John Benjamins, 19–44.

290

Chinese: Parts of speech Li, Chongxing (1992) ‘Chùsuǒcí Fāzhǎn Lìshǐ de Chūbù Kǎochá’ (‘A Preliminary Study on the Historical Development of Place Words’), in Hu Zhu’an, Yang Naisi, and Jiang Shaoyu (eds) Jìndài Hànyǔ Yánjiū (A Study on Modern Chinese), Beijing: Commercial Press 24–63. Li, Jie (1995) ‘Dōu and Wh-questions in Mandarin Chinese’, Journal of East Asian Linguistics 4: 313–23. Li, Jinxi (1924) Xīnzhù Guóyǔ Wénfǎ (The Grammar of Mandarin Chinese: New Edition), Beijing: Shangwu Chubanshe. Li, Jingxi and Shiru Liu (1955) Zhōngguó Yǔfǎ Jiàocái (Teaching Materials on Chinese Grammar), Beijing: Wushi nian chubanshe. Li, Yafei (1983) ‘Existential Sentences and the Category of the Locative Words’, Paper presented at the First Harbin Conference on Generative Grammar, Harbin: Heilongjiang University. Li, Yafei (2003) ‘Localizers in Chinese and the Cost of Computation’, Ms., University of Wisconsin, Madison. Li, Yen-Hui Audrey (1985) ‘Abstract Case in Mandarin Chinese’, Doctoral Dissertation, University of Southern California. Li, Yen-Hui Audrey (1990) Order and Constituency in Mandarin Chinese, Dordrecht: Kluwer. Li, Yen-Hui Audrey (1998) ‘Argument Determiner and Number Phrases’, Linguistic Inquiry 29: 693–702. Li, Yen-Hui Audrey (1999) ‘Plurality in a Classifier Language’, Journal of East Asian Linguistics 8: 75–99. Li, Yen-Hui Audrey (2006) ‘Chinese bǎ’, in Martin Everaert and Henk van Riernsdijk (eds) The Blackwell Companion to Syntax, vol. 1, Oxford: Blackwell, 374–468. Li, Yen-Hui Audrey (2013) ‘P-insertion and Ellipsis’, Paper presented at the Workshop on the Syntax of Ellipsis, The Chinese University of Hong Kong. Li, Yuming (1996) ‘Fēiwèi Xíngríngcí de Cílèi Dìwèi’ (‘The Word-class Status of Non-predicative Adjectives’), Zhōngguó Yǔwén (Studies of the Chinese Language) 1: 1–9. Lin, C.-J. William (1984) ‘What Does the Mandarin Particle ne Communicate?’ Cahiers de Linguistique Asie Orientale 13(2): 217–40. Lin, Jo-Wang (1996) ‘Polarity Licensing and Wh-phrase Quantification in Chinese’, Doctoral Dissertation, University of Massachusetts, Amherst. Lin, Jo-Wang (1998) ‘On Existential Polarity Wh-phrases in Chinese’, Journal of East Asian Linguistics 7: 219–55. Lin, Jo-Wang (2003) ‘Temporal Reference in Mandarin Chinese’, Journal of East Asian Linguistics 12: 259–311. Liu, Chen-Sheng Luther (2010) ‘The Positive Morpheme in Chinese and the Adjectival Structure’, Lingua 120: 1,010–56. Liu, Feng-Hsi (1997) ‘An Aspectual Analysis of bǎ’, Journal of East Asian Linguistics 6: 51–99. Liu, Feng-Hsi (1998) ‘A Clitic Analysis of Locative Particles’, Journal of Chinese Linguistics 26(1): 48–70. Liu, Feng-Hsi (2006) ‘Dative Constructions in Chinese’, Language and Linguistics 7(4): 863–904. Liu, Shiru (1965) Wèijìn Nánběicháo Liàngcí Yánjiū (Studies on Classifiers in Wei, Jin and Southern and Northern Dynasties), Beijing: Zhonghua Shuju. Liu, Yuehua, Wenyu Pan, and Wei Gu (1983) Shíyòng Xiàndài Hànyǔ Yǔfǎ (A Practical Grammar of Modern Chinese), Beijing: Waiyu Jiaoxue yu Yanjiu Chubanshe. Lü, Shuxiang (1947) Zhōngguó Wénfǎ Yàolǜe (Essence of Chinese Grammar), Shanghai: The Commercial Press. Lü, Shuxiang (1979) Hànyǔ Yǔfǎ Fēnxī Wèntí (Issues on the Analysis of Chinese Grammar), Beijing: The Commercial Press. Lü, Shuxiang (1980) Xiàndài Hànyǔ Bābǎi Cí (800 Major Words in Contemporary Chinese), Beijing: The Commercial Press. Lü, Shuxiang (1984) Hànyǔ Yǔfǎ Lùnwén Jí (Theses on Chinese Grammar), Beijing: The Commercial Press. Lü, Shuxiang (1987) ‘Shuō “shèng” hé “bài ” ’ (‘On “Win” and “Defeat” ’), Zhōngguó Yǔwén (Studies of the Chinese Language) 1: 1–5. Ma, Jianzhong (1898) Mǎshì Wén Tōng (Basic Principles for Writing Clearly and Coherently by Mister Ma), Beijing: The Commercial Press (1983 reprinted).

291

Candice Chi-Hang Cheung McCawley, James (1988) ‘The Comparative Conditional Constructions in English, German and Chinese’, in Proceedings of the 14th Annual Meeting of the Berkeley Linguistics Society, Berkeley: BLS, 176–87. McCawley, James (1992) ‘Justifying Part-of-speech Distinctions in Mandarin Chinese’, Journal of Chinese Linguistics 20(2): 211–46. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Pan, Haihua and Jingquan Han (2006) ‘Xūcí there de Jùfǎ Dìweì jí Xiāngguān Lǐlùn Wèntí’ (‘The Syntactic Status of Expletive there and Some Related Theoretical Issues’), Dāngdài Yǔyánxué (Contemporary Linguistics) 1: 17–35. Paul, Waltraud (2005) ‘Adjective Modification in Mandarin Chinese and Related Issues’, Linguistics 43(4): 757–93. Paul, Waltraud (2006) ‘Zhu Dexi’s Two Classes of Adjectives Revisited’, in Christoph Anderl and Halvor Eifring (eds) Studies in Chinese Language and Culture, Festschrift in Honour of Christoph Harbsmeier on the Occasion of his 60th Birthday, Oslo: Hermes Academic Publishing, 303–15. Paul, Waltraud (2010) ‘Adjectives in Mandarin Chinese: The Rehabilitation of a much Ostracized Category’, in Patricia Cabredo Hofherr and Ora Matushansky (eds) Adjectives: Formal Analyses in Syntax and Semantics, Amsterdam and Philadelphia: John Benjamins, 115–52. Paul, Waltraud (forthcoming) ‘Adverbs in Mandarin Chinese’, in Wolfgang Behr, Yueguo Gu, Zev Handel, C.-T. James Huang and Rint Sybesma (eds) The Encyclopedia of Chinese Language and Linguistics, Leiden: Brill. Peyraube, Alain (1980) Les Constructions Locatives en Chinois Moderne, Paris: Langages Croisés. Qian, Nairong (ed) (1990) Xiàndài Hànyǔ (Modern Chinese), Beijing: Gaodeng Jiaoyu Chubanshe. Rohsenow, John S. (1978) ‘Perfective le: Aspect and Relative Tense in Mandarin Chinese’, in Robert Cheng, Li Ying-Che, and Tang Ting-Chi (eds) Proceedings of Symposium on Chinese Linguistics, Taipei: Student Book Co., 269–91. Ross, Claudia (1995) ‘Temporal and Aspectual Reference in Mandarin Chinese’, Journal of Chinese Linguistics 23(1): 87–135. Rygaloff, Alexis (1973) Grammaire élémentaire du chinois, Paris: Presses Universitaires de France. Shao, Jingmin (1989) ‘Yǔqìcí “ne” zài Yíwènjù zhōng de Zuòyòng’ (‘The Function of the Expressive Particle “ne” in Interrogatives’), Zhōngguó Yǔwén (Studies of the Chinese Language) 3: 170–75. Shen, Yang, Yuanjian He and Yang Gu (2002) Shēngchéng Yǔfǎ Lǐlùn yǔ Hànyǔ Yǔfǎ Yánjiū (Generative Grammar Theory and Studies on Chinese Grammar), Harbin: Heilongjiang Education Press. Shi, Dingxu (2013) ‘Mass Nouns, Countable Nouns, Massifiers and Classifiers’, Paper presented at the Workshop on the Grammar of Measurement in Chinese, National Taiwan Normal University, Taiwan. Shi, Youwei (1986) ‘Hànyǔ Liáncí de Gōngnéng, Jièxiàn hé Wèizhì’ (The Functions, Criteria, and Positions of Conjunctions) Zhōngyāng Mínzú Xuéyuàn Xuébào (Journal of Minzu College of China) 3. Shie, Chi-chiang (1991) ‘A Discourse-functional Analysis of Mandarin Sentence-Final Particles’, Master Thesis, Taipei: National Chengchi University. Smith, Carlota S. (1991) The Parameter of Aspect, Dordrecht: Kluwer Academic Publishers. Soh, Hooi Ling (2009) ‘Speaker Presupposition and Mandarin Chinese Sentence Final-le: A Unified Analysis of the “Change of State” and the “Contrary to Expectation” Reading’, Natural Language and Linguistic Theory 27(3): 623–57. Sun, Chaofeng (2008) ‘Two Conditions and Grammaticalization of the Chinese Locatives’, in Dan Xu (ed) Space in Languages of China: Cross-linguistic, Synchronic and Diachronic Perspectives, Heidelberg, Germany: Springer Science, 199–227. Sun, Jinwen and Yaqing Wu (2003) ‘Zàilùn Lǐngyǒu Míngcí Tíshēng Yíwèi’ (A Reanalysis of Possessor Raising Movement), Yǔyán Kēxué (Linguistic Sciences) 6: 46–52. Sybesma, Rint (1999) The Mandarin VP, Dordrecht: Kluwer Academic Publishers. Tai, Hao-Yi James (1973) ‘A Derivational Constraint on Adverbial Placement in Mandarin Chinese’, Journal of Chinese Linguistics 1(3): 397–413. Tai, Hao-Yi James (1990) ‘Variation in Classifier Systems across Chinese Dialects: Towards a Cognitionbased Semantic Approach’, in Proceedings of the First International Symposium on Chinese Languages and Linguistics, Taipei: Academia Sinica, 308–22.

292

Chinese: Parts of speech Tai, Hao-Yi James and Fangyi Chao (1994) ‘A Semantic Study of the Classifier zhāng’, Journal of Chinese Language Teachers Association 29(3): 67–78. Tai, Hao-Yi James and Lian-Qing Wang (1990) ‘A Semantic Study of the Classifier tiáo’, Journal of the Chinese Language Teachers Association 25(1): 35–56. Tang, Chih-Chen Jane (1990) ‘Chinese Phrase Structure and the Extended X’-Theory’, Doctoral Dissertation, Cornell University. Tang, Sze-Wing (2004) ‘Zuògéhuà hé Hànyǔ Bèidòngjù’ (‘Ergativization and Chinese Passives’), Zhōngguó Yǔwén (Studies of the Chinese Language) 310(4): 291–301. Tang, Ting-Chi (1988) ‘Guóyǔ Xíngríngcí de Chóngdié Guīlǜ’ (‘Reduplication Rules for Adjectives in Mandarin Chinese’), in Tang Ting-Chi (ed) Hànyǔ Cífǎ Jùfǎ Lùnjí (Studies on Chinese Morphology and Syntax), Taipei: Student Book Co., 29–57. Ting, Jen (1995) ‘A Non-Uniform Analysis of the Passive Construction in Mandarin Chinese’, Doctoral Dissertation, University of Rochester. Ting, Jen (1998) ‘Deriving the bèi-construction in Mandarin Chinese’, Journal of East Asian Linguistics 7: 319–54. Tsao, Feng-fu (1996) ‘Hànyǔ de Tíshēng Dòngcí’ (Chinese Raising Verbs), Zhōngguó Yǔwén (Studies of the Chinese Language) 3: 172–82. Wang, Li (1934) Zhōngguó Xiàndài Yǔfǎ (Modern Chinese Grammar), Beijing: Shangwu Chubanshe (1985 reprinted). Wang, Li (1955) Zhōngguó Yǔfǎ Lǐlùn (Theory of Chinese Grammar), Beijing: Zhonghua Shuju. Wen, Binli and Zongli Chen (2001) ‘Lǐngyǒu Míncí Yíwèi: Jīyú MP de Fēnxī’ (‘An MP Approach to Possessor Raising in Chinese’), Xiàndài Wàiyǔ (Modern Foreign Languages) 4: 412–16. Wu, Guo (2005) ‘The Discourse Function of the Chinese Particle ne in Statements’, Journal of the Chinese Language Teachers Association 40(1): 47–82. Wu, Ruey-Jiuan Regina (2004) Stance in Talk: A Conversation Analysis of Mandarin Final Particles, Amsterdam and Philadelphia: John Benjamins. Xiang, Ming (2008) ‘Plurality, Maximality and Scalar Inferences: A Case Study of Mandarin dōu’, Journal of East Asian Linguistics 17: 227–45. Xing, Gongwan and Qingzhu Ma (eds) (1992) Xiàndài Hànyǔ Jiàochéng (A Course in Modern Mandarin Chinese), Tianjin: Nankai University Press. Xu, Jie (1999) ‘Liǎngzhǒng Bǎoliúbīnyǔ Jùshì yǔ Xiāngguān Yǔfǎ Lǐlùn’ (‘Some Theoretical Issues of the Two Types of Chinese Retained-Object Constructions’), Dǎngdài Yǔyánxué (Contemporary Linguistics) 1: 16–29. Xu, Jie (2001) Pǔbiànyǔfǎ Yuánzé yǔ Hànyǔ Yǔfǎ Xiànxiàng (Grammatical Principles and Grammatical Phenomena), Beijing: Peking University Press. Yang, Shu-Ying (1991) ‘Dative Alternation in Chinese and English’, Doctoral Dissertation, University of Connecticut. Yip, Po-Ching and Don Rimmington (1997) Chinese: An Essential Grammar, London and New York: Routledge. Yip, Po-Ching and Don Rimmington (2004) Chinese: A Comprehensive Grammar, London and New York: Routledge. Zhang, Jing (1987) Hànyǔ Yǔfǎ Wèntí (Grammatical Issues in Mandarin Chinese), Beijing: Zhonguo Shehui Kexue Chubanshe. Zhang, Ning Niina (1997) ‘Syntactic Dependencies in Mandarin Chinese’, PhD dissertation, University of Toronto. Zhang, Ning Niina (2008) ‘Encoding Exhaustivity’, UST Working Papers in Linguistics, National Tsing Hua University, Taiwan, 133–43. Zhang, Ning Niina (2009) Coordination in Syntax, Cambridge: Cambridge University Press. Zhang, Shi (1990) ‘Correlations between the Double-Object Construction and Preposition Stranding’, Linguistic Inquiry 21(2): 312–16. Zhang, Zhigong (ed) (1959) Hànyǔ Zhīshì (Knowledge on Mandarin Chinese), Beijing: Renmin Jiaoyu Chubanshe (1979 reprinted). Zhou, Gang (2002) Liáncí yǔ Xiāngguān Wèntí (Coordinators and the Relevant Issues), Hefei: Anhui Jiaoyu Press. Zhu, Dexi (1956) ‘Xiàndài Hànyǔ Xíngróngcí Yánjiū’ (‘Studies of Adjectives in Modern Chinese’), Yǔyán Yánjiū (Language Research) 1; (reprinted in Zhu Dexi (ed) (1980) Xiàndài Hànyǔ Yǔfǎ Yánjiū (Grammatical Studies of Modern Chinese), Beijing: The Commercial Press, 3–41.

293

Candice Chi-Hang Cheung Zhu, Dexi (1979) ‘Yǔ Dòngcí gěi Xiāngguān de Jùfǎ Wèntí’ (The Syntactic Issues Related to Verb gěi), Fāngyán (Dialects) 2: 81–7; (reprinted in Zhu Dexi (1999) Zhū Dēxī Wénjí (Collections of Zhu Dexi), vol. 2, Beijing: The Commercial Press, 231–48.) Zhu, Dexi (1980) Xiàndài Hànyǔ Yǔfǎ Yánjiū (Grammatical Studies of Modern Chinese), Beijing: The Commercial Press. Zhu, Dexi (1982) Yǔfǎ Jiǎngyì (Lecture Notes on Grammar), Beijing: The Commercial Press. Zhu, Xingfan (2005) ‘Qīngdòngcí hé Hànyǔ Bùjíwù Dòngcí Dàibīnyǔ Xiànxiàng’(‘Light Verb Syntax of Intransitives Containing Objects’), Xiàndài Wàiyǔ (Modern Foreign Languages) 3: 221–33.

294

16 Chinese Proverbs and Popular Sayings Larry Herzberg Calvin College, USA

Introduction The proverbs of any country or culture contain the wisdom accumulated by its people over countless generations. China is the world’s oldest continuous civilization, with over 3,000 years of history. It is, therefore, hardly surprising that the Chinese language is a particularly rich source of wisdom. Arguably no other language or culture has such a huge treasure trove of proverbs and popular sayings that comment on every aspect of the human experience. There are as many as 20,000 literary idioms and tens of thousands of popular maxims and sayings, in addition to the pithy and profound quotes to be gleaned from three millennia of Chinese philosophy and literature. In Western countries like the USA, proverbs are heard less and less. Since individuality and creativity are prized, quoting age-old sayings is considered trite and out of date. Few young people are heard to utter such maxims as ‘a stitch in time saves nine’, ‘an ounce of prevention is worth a pound of cure’ or ‘the early bird catches the worm’. The Chinese people, however, quote various sayings all the time in their daily speech. An ancient culture gives them a shared pool of thousands of phrases which allow the Chinese to sum up a situation in a matter of a few words understood by everyone. To reference some of the literary idioms displays an educated background and the use of the wide variety of sayings adds color and spice to a person’s speech. Without the liberal use of proverbs and maxims, a person’s speech is thought to have no flavor. In a group-centered society, the emphasis is not on originality so much as on enjoying a shared cultural background. Wit and originality are displayed by cleverly changing the proverb’s wording to say things in a fresh and humorous way. The Chinese themselves like to make clear distinctions between literary idioms known as 成语 (chéngyǔ), which are usually four words and have ancient stories behind them; five- and seven-word maxims called 谚语 (yányǔ), which are considered reasonably refined and often come in pairs that rhyme; and the popular sayings or 俗语 (súyǔ) passed down by the ‘common people’ through many generations and which have no set length or form. Many of the 成语 (chéngyǔ) come from famous Chinese philosophers such as Confucius, Laozi, and Zhuangzi. A great many others come from ancient historical records, especially those of the Warring States period (476–221 bce), a time when China often suffered civil war. The origins of 谚语 (yányǔ) and 俗语 (súyǔ), however, are lost in the mists of time. 295

Larry Herzberg

They have been passed down through many centuries and represent the insights obtained by ordinary Chinese people as they went about their daily lives. Rather than giving long lists of each of the three main kinds of Chinese proverbs and sayings, this chapter will present some of the ones that are most commonly used. They are grouped by the theme or topic of human experience on which they comment, but only introduce a fraction of the myriad topics covered by Chinese proverbs. Since the sayings of any culture reveal so much about the way the people of that culture view the world, a short introduction is given to the proverbs that speak to any given theme, to reflect on what these proverbs reveal about Chinese thought and culture. While discovering certain perspectives in these sayings that seem particularly Chinese, a great many will affirm that which the more enlightened among us have always known, namely that we are one race, the human race, with a commonality of shared experience as human beings. It is true, after all, what Confucius said nearly 2,500 years ago, that, ‘Within the four seas, all men are brothers’ (All human beings in the whole world are one family): Sì hǎi zhī nèi jiē xiōngdì yě 四海之内皆兄弟也

Learning No other society in the history of the world has for so long revered learning through study as have the Chinese. Political power and social status were completely determined by birth until the nineteenth century in the West or in Japan. By contrast, China chose all of its government officials, except for the emperor, exclusively through an examination system based on the study of classic books, starting as early the seventh century. Of course it was much more difficult for a boy not of the landed gentry to be able to afford the books and instruction necessary to pass the various levels of exams to become a governor or imperial advisor. However, throughout Chinese history there were numerous examples of young men from humble beginnings who rose to prominent positions through self-education. It is therefore only to be expected that Chinese culture has always greatly respected teachers, since learning has been the path to success for so long. The most revered teacher of all time in China is, of course, Confucius. To this day you will see temples all over the country dedicated to that ancient sage and to the realm of books and of study that he represents. Here are some of the most famous quotes related to teaching, learning, and knowledge. Whoever is your teacher even for a day consider as your father (to respect and care for) your whole life long Yírì wéi shī, zhōngshēn wéi fù 一日为师,终身为父  (Confucius) When we walk together with three people, one of them (at least) will have something to teach us (Whenever you encounter any group of people, whether 3 or 33, there is something to be learned from at least one of them) Sān rén xíng, bì yǒu wú shī 三人行,必有吾师  (Confucius) Never feel tired of studying and never tire of teaching others Xué ér bú yàn, huì rén bú juàn 学而不厌,诲人不倦  (Confucius) 296

Chinese proverbs and popular sayings

What you learn depends on whom you study with; if you study with a butcher, you’ll never become a cobbler. Gēn shénme rén xué shénme yàng, gēnzhe túfū xué bùchéng píjiàng 跟什么人学什么样,跟着屠夫学不成皮匠 Give a man a fish and he’ll have food for a day (three meals); Teach a man to fish and he’ll have a skill to use all his life. Shòu rén yǐ yú sān cān zhī xū，shòu rén yǐ yù zhōngshēn zhī yòng 授人以鱼三餐之需,授人以渔终身之用 Hearing about something a hundred times cannot compare to seeing it once for yourself (‘A picture is worth a thousand words’; ‘Seeing is believing’) Bǎi wén bùrú yíjiàn 百闻不如一见 Eat/suffer a fall into the pit, a gain in your wit (One often gains in wisdom after suffering (‘eating’) a misfortune) Chī yíqiàn, zhǎng yízhì 吃一堑,长一智 Without experiencing the cold of winter, one cannot appreciate the warmth of spring (You cannot truly appreciate the good times in your life if you’ve never suffered hardship.) Bùjīng dōng hán, bùzhī chūn nuǎn 不经冬寒,不知春暖 Receive one blow, and you’ll learn a lesson; receive ten blows, and you’ll end up a genius (literally you’ll become like Zhuge Liang, a legendary master strategist and clever man) Ái yì quán, dé yì zhāo, ái shí quán, biàn Zhūgé 挨一拳,得一招; 挨十拳,变诸葛 Indigo comes from blue but exceeds blue (in its beauty) [Expressing a teacher’s hope that their student will surpass them in knowledge in the future]: Qīng chū yú lán shèng yú lán 青出于蓝,胜于蓝 Teachers open the door; you enter by yourself [lit.: The master leads the student through the door, but perfecting one’s skill is up to the student] (A teacher can only expose students to knowledge; it’s then up to the student to work hard to learn what he or she has been taught) Shīfu lǐng jìn mén, xiūxíng zài gèrén 师傅领进门,修行在个人 Live ’til you’re old and study ’til you’re old, but there’s still 30% you’ll never learn! Huódào lǎo xuédào lǎo, hái yǒu sānfēn xuébudào 活到老学到老,还有三分学不到 297

Larry Herzberg

Learning has no boundaries  (Confucius) Xué wú zhǐ jìng 学无止境 Familiarity can engender skill (‘Practice makes perfect’) Shú néng shēng qiǎo 熟能生巧 Learning without thinking is ignorance; thinking without learning (study) is dangerous Xué ér bùsī zé wǎng, sī ér bù xué zé dài 学而不思则罔,思而不学则殆 If you know (something), to recognize that you know it; and if you don’t know (something), to realize that you don’t know it – that is (true) knowledge  (Confucius) Zhī zhī wéi zhī zhī bù zhī wéi bù zhī shì zhī yě 知之为知之不知为不知是知也 (Don’t be) a frog at the bottom of a well (This proverb is from a parable told by the fourth-century bce Daoist philosopher Zhuangzi. It refers to the story of a frog who lived in the bottom of a well. He thought that the well was the entire world, until one day a sea turtle fell in and informed him of a much larger world outside the narrow confines of that well. The Chinese, as well as the Japanese, use this fable to describe people with a provincial, narrow-minded view of things due to ignorance.) Jǐng dǐ zhī wā 井底之蛙

Patience and perseverance Traditional societies with long histories perhaps know better than societies with much shorter histories the value of patience. They understand that it often takes a long time to bring about the changes you desire. The first saying below derives from the fact that in ancient China the majority of people often didn’t have enough to eat. To become fat in traditional China, and indeed in almost every traditional society in the world, was a sign that you were prosperous enough to have an abundance of food. So being ‘fat’ was a good thing! The second proverb below refers to an ancient story about an old farmer who lived on one side of a huge mountain but whose fields were on the other side of the mountain. Every morning the old man and his sons had to go around the mountain to cultivate their crops. One day the old man convinced his sons to start removing the mountain with their shovels. Of course the man’s neighbors felt the old man daft for having such a foolish notion and told him so. The old man replied that although the mountain could not be ‘moved’ away in his lifetime or even that of his sons and grandsons, over time his family would eliminate this obstacle that made their lives difficult, if only they persevered. This realization of the need to be patient in achieving one’s goals is not unknown to us in the West. After all, we do on occasion hear the expressions ‘Rome wasn’t built in a day’ and ‘Good things take time’. But our tendency is to achieve quick results, to ‘learn Chinese in ten minutes a day’, to want things ‘done yesterday’. We Westerners too much resemble the farmer who wanted 298

Chinese proverbs and popular sayings

to help his crops grow faster by pulling on their stalks, which only ends up killing the plant! Older societies like China caution against such unreasonable expectations. A fat person didn’t get fat with just one mouthful (‘Rome wasn’t built in a day’) Pàngzi búshì yìkǒu chīde 胖子不是一口吃的 The foolish old man moved the mountain (Anything can be done if you work long and hard enough at it) Yú gōng yí shān 愚公移山 One who is burning with impatience can never eat hot porridge (Patience is necessary to accomplish anything, even when it comes to waiting for your rice gruel to be cool enough to eat) Xīn jí chībùdé rè zhōu 心急吃不得热粥 With enough work, an iron rod can be ground into a needle (Almost anything can be achieved if you put enough effort into it) Zhǐyào gōngfu shēn, tiě chǔ móchéng zhēn 只要功夫深,铁杵磨成针 A thousand mile journey starts under your feet (The longest journey begins with a single step) Qiān lǐ zhī lù, shǐ yú zú xià 千里之路,始于足下  (Lao Zi) Don’t fear going slow (making slow progress); just fear standing still Bú pà màn, jiù pà zhàn 不怕慢,就怕站 (Don’t) pull on seedlings to help them grow (It’s human nature to be impatient for good things to happen, but you can’t force good things to happen before the required time has elapsed) Bá miáo zhù zhǎng 拔苗助长

Humility and contentment Every traditional culture has stressed the need for humility. When the Church in Europe many centuries ago posited the Seven Deadly Sins, Pride was first among them. And yet East Asian societies have throughout their history laid an even bigger emphasis on the virtue of humility than most other societies. China for millennia has been one of the most densely populated countries in the world, as Japan and Korea came to be many centuries ago. That perhaps is what caused the Chinese and the other cultures of East Asia to emphasize the good of the group over a glorification of the individual, such as we see in Western societies. In a crowded society where people lived in large communities side by side with their neighbors, there was no place for anything that disturbed social harmony. The need for humility was stressed to promote that harmony between people would be disturbed by boastfulness and a self-centered attitude. 299

Larry Herzberg

The Chinese have argued throughout the ages that a humble attitude should encourage you to admit when you don’t know something. They have also stressed that people should not pridefully compare themselves with others, but to modestly accept the fact that there will always be others who are superior to you in some way. The proverbs and sayings below all emphasize the importance of being humble, dating back to the words of Confucius 2,500 years ago. Contentment can only be found in not envying others or comparing yourself to them, but to be satisfied with what you have. Compare yourself to those who are superior to you, and you’ll find yourself lacking; compare yourself to those who are inferior to you, and you’ll find yourself more than good enough. Bǐ shàng bùzú, bǐ xià yǒu yú 比上不足,比下有余  (Confucius) If you compare yourself to others, it’ll just get you angry Rén bǐ rén, qìsǐ rén 人比人,气死人 Looking from this mountain that yonder mountain seems higher (‘The grass is always greener on the other side of the fence’; don’t envy others as having it better than you do) Zhè shān wàngzhe nà shān gāo 这山望着那山高 They who know they have enough are always happy Zhī zú zhě cháng lè yě 知足者常乐也 The fool does not ask; he who asks is no fool (If you really want to learn, you have to be humble enough to ask questions and reveal your ignorance) Yú zhě bú wèn, wèn zhě bù yú 愚者不问,问者不愚  (Confucius) Beyond the sky is another sky; beyond the mountain is another mountain (No matter how great a person or a thing might be, there’s always someone or something out there that’s at least as great and perhaps even greater) Tiān wài yǒu tiān, shān wài yǒu shān 天外有天,山外有山 Wielding the ax in front of Ban’s gate (Trying to show off one’s carpentry skills in front of the legendary master carpenter, Ban; performing in front of someone who is much more skilled than you are) Bān mén nòng fǔ 班门弄斧 Old Wang praises his own melons while he sells them (‘Toot your own horn’; said teasingly of those who boast of their own accomplishments) Lǎo wáng mài guā, zì mài zì kuà 老王卖瓜,自卖自夸 300

Chinese proverbs and popular sayings

A half-filled bottle makes noise (when shaken), while a full bottle makes none (Those who know the least boast the most, while the superior person remains silent) Yīpíng bùxiǎng, bànpíng dīngdāng 一瓶不响,半瓶叮当 Those who retreated 50 paces (running away in battle) laugh at those who retreated 100 paces (‘The pot calling the kettle black’) Wǔshí bù xiào bǎi bù 五十步笑百步

Talking In a densely populated and group-oriented society like China, it has always been important to watch what one says so as not to risk offending one’s neighbors. More significantly, China has always been an autocratic society under the control of a ruler with absolute power. The country was first ruled by kings, including in the time of Confucius, then by emperors from 221 bce until ce 1911. Then followed the dictatorial rule of Chiang Kai-shek from the 1920s until after World War II, and in the past six decades China has been under the rule of the Communist Party. This has meant that it has always been extremely dangerous to speak one’s mind too freely, especially in challenging the people in power. It is not surprising, therefore, that there are a huge number of proverbs and sayings in Chinese that warn against expressing one’s thoughts too freely. Below is but a sampling of cautionary advice when it comes to talk and speech. Sickness enters through the mouth; misfortune comes out of it (Careless speech can land one in hot water) Bìng cóng kǒu rù, huò cóng kǒu chū 病从口入,祸从口出 The mouth and tongue have always been the roots of calamity Kǒu shé cónglái shì huò jī 口舌从来是祸基 The mouth is the door to disaster; the tongue is the knife that can kill you Kǒu shì huò zhī mén, shé wéi zhǎn shēn dāo 口是祸之门,舌为斩身刀 A closed mouth keeps flies from flying in (If you keep your mouth shut, you can’t get into trouble) Bìzhe de zuǐ fēibújìn cāngyīng 闭着的嘴飞不进苍蝇 When meeting people, only say 30% (of what you’re thinking); don’t toss out everything that’s in your mind Féng rén zhǐ shuō sān fēn huà, wèi kě quán pāo yípiàn xīn 逢人只说三分话,未可全抛一片心

301

Larry Herzberg

After something has passed through three mouths, even snakes are said to have grown legs (Every rumor becomes exaggerated to a ridiculous point and is not to be believed) Huà jīng sān zhāng zuǐ, cháng chóng yě zhǎng tuǐ 话经三张嘴,长虫也长腿 People have a face (just as) a tree has a bark (Never humiliate anyone or cause them to ‘lose face’, or they will feel vulnerable in the same way a tree becomes when it loses the bark that protects it) Rén yǒu liǎn, shù yǒu pí 人有脸,树有皮 Once a word is spoken, even a team of four horses cannot catch up to it (‘A promise is a promise’; the superior person keeps his word and never takes it back) Yìyán jì chū, sì mǎ nán zhuī 一言即出,驷马难追 Ice that’s three feet thick didn’t form in just one cold day Bīng dòng sān chǐ fēi yírì zhī hán 冰冻三尺非一日之寒 Never pull on your shoes in a melon patch; never adjust your cap under a plum tree (Avoid doing the slightest thing that might arouse suspicion that you’ve done something improper) Guā tián bú nà lǚ, lǐ xià bù zhěng guān 瓜田不纳履,李下不整冠 Each person should first clear away the snow from their own front door before worrying about the frost on their neighbor’s roof (Your own problems are enough to handle without you interfering in other people’s affairs) Gè rén zì sǎo mén qián xuě, mò guǎn tā rén wǎ shàng shuāng 各人自扫门前雪，莫管他人瓦上霜 Don’t fear a thousand-to-one chance something bad might happen; fear the one chance in a thousand (We usually guard against the dangers we expect; it’s the unlikely we must also guard against) Búpà yíwàn, zhǐ pà wànyī 不怕一万,只怕万一 Never harm another, but beware of others who may intend to harm you (Never harbor the intention to do harm to others, but be aware that others may harbor the intention to harm you) Hài rén zhī xīn bù kě yǒu, fáng rén zhī xīn bù kě wú 害人之心不可有,防人之心不可无 302

Chinese proverbs and popular sayings

Morality For the past 1,500 years or so, Buddhism has been the dominant religion of China. Like any of the world’s great religions, Buddhism preaches that one should not kill, steal, slander, or harm his fellow human beings in any way. It is a religion that emphasizes compassion and self-abnegation. The Eight-fold Path formulated by Buddha as the way people should live their lives has a lot in common with the Ten Commandments of the Bible. However, it has not been Buddhist morality that has had the largest influence on Chinese culture throughout the ages but rather the moral teachings of Confucius. In the Analects, the book that records his teachings, the ancient sage talked about his ideals for human relationships. He preached virtues such as righteousness, justice, respect, kindness, and reciprocity. Books based on his teachings became the canon of Chinese thought that was the basis of Chinese education for close to 2,000 years. It was these books that every scholar had to study in order to pass the imperial examinations to become an official. To this day, Confucian ideals and morality continue to influence not only the people of China, but those in Japan and Korea, as well. What you do not want done to you, do not do to other people (Do not do unto others what you would not wish done to you) Jǐ suǒ búyù wù shī yú rén 己所不欲勿施于人  (Confucius) A superior person knows what is moral; a petty person (only) knows what is profitable Jūnzǐ yù yú yì, xiǎo rén yù yú lì 君子喻于义,小人喻于利 Plant melons, reap melons; plant beans, reap beans (‘You reap what you sow’) Zhòng guā dé guā, zhòng dòu dé dòu 种瓜得瓜,种豆得豆 Better to be a broken piece of jade than an unbroken shard of clay (Better to die with honor than live with dishonor) Níng wéi yù suì, bù wéi wǎ quán 宁为玉碎,不为瓦全 If you do nothing shameful during the day, there’s no need to fear ghosts will come knocking at your door at night (If your behavior is moral and upright, you’ll sleep with a clear conscience) Báitiān búzuò kuīxīn shì, yèlǐ búpà guǐ qiāo mén 白天不做亏心事,夜里不怕鬼敲门 Heaven’s net is vast; it’s cast far and wide, but lets nothing through (No one escapes justice) Tiān wǎng huīhuī, shū ér búlòu 天网恢恢,疏而不漏

303

Larry Herzberg

You can’t conceal fire by wrapping it in paper (‘The truth will out’; no wrong-doing can be concealed for long) Zhǐ lǐ bāobuzhù huǒ 纸里包不住火 A monk may run away but he can’t escape the temple (You may try to escape punishment for your wrongdoing, but the authorities will eventually find and catch you by knowing your family and workplace) Pǎole héshàng pǎobùliǎo miào 跑了和尚跑不了庙 If you walk often by the river’s edge how can your shoes not get wet? (A person with the power to grant favors will likely give into the temptation to be bribed) Cháng zài hé biān zǒu, nǎ néng bù shī xié? 常在河边走,哪能不湿鞋?

Money Buddhism taught that only if people ceased to desire things like money and fame would they find relief from the suffering that is human existence. Confucius exhorted people to be highminded and moral rather than to seek petty profit for themselves. The Chinese, however, have always been a very pragmatic people. Regardless of what Chinese religion and philosophy warned about the pursuit of wealth, most Chinese through the ages have realized that, while money might not be able to buy happiness or love, it can buy pretty much everything else. For centuries up to the present day, at New Year, the most important of all Chinese holidays, people greet each other with the time-honored phrase 恭喜发财 (Gōngxǐ fācái), which means ‘Congratulations and may you become wealthy!’ Most of the popular sayings about money in Chinese culture are from the common people of previous centuries. They felt keenly the pain of poverty and realized only too well what the lack of money meant in their lives. Money allows you to speak with the gods (Money is power) Qián néng tōng shén 钱能通神 With money you can even get the devil to push the millstone for you (Money makes everything possible) Yǒu qián néng shǐ guǐ tuī mò 有钱能使鬼推磨 A poor man’s ambition stops short; a thin horse’s mane appears long (Poverty stifles ambition) Rén qióng zhì duǎn, mǎ shòu máo cháng 人穷志短,马瘦毛长 (You pay) One penny of money, (you get) one penny of merchandise (‘You get what you pay for’) Yīfēn qián yīfēn huò 一分钱一分货 304

Chinese proverbs and popular sayings

Expensive things aren’t really expensive; cheap things aren’t really cheap (‘Buy cheap, get cheap’; expensive things tend to last a long time and so are a better bargain in the end, while cheap things quickly become useless because they’re poorly made) Guìde búguì, jiànde bújiàn 贵的不贵,贱的不贱

Conformity In the West and particularly in the USA, individuality is glorified, as are competition and fame. The Chinese people, on the other hand, have always had a real mistrust and even fear of fame or of sticking out from the crowd. That is because calling attention to oneself too often resulted in misfortune. At the very least, being superior or different in any way can arouse jealousy among lesser people. At the worst, a famous or individualistic person can easily become the victim of malicious and harmful gossip. There are a great variety of Chinese sayings that warn against the danger of calling attention to oneself or of showing off. Asian societies have always tended to be conformist, group-oriented societies. They had no wilderness to carve out as self-reliant pioneers, which is the American heritage. China has been a settled, highly populated society for over two millennia. While countries like the USA have the most heterogeneous societies in the world, China has one of the more homogeneous ones. While Western nations have a fairly long history of democracy, China has always been hierarchical and ruled by autocratic authority. All of these factors have contributed to making Chinese society prize social harmony over the Western glorification of the rights of the individual. It is significant that another major Asian society, Japan, also has similar sayings to those of the Chinese that advise conformity rather than individuality. Most famous of these is ‘the nail that sticks up gets pounded down’. In English there are no equivalent sayings. Instead there is the expression ‘the squeaky wheel catches the grease’. The implication is the opposite of the Chinese sayings, namely that the person who calls attention to himself or herself is likely to succeed in getting what he or she wishes. In Western societies that glorify the individual and also protect the right to be different, many people seek the rewards that come with ‘sticking out’ from the crowd. When trees are big (tall), they invite the wind (to knock them down) Shù dà zhāo fēng 树大招风 The bird that sticks its head out is the first to get shot Qiāngdǎ chū tóu niǎo 枪打出头鸟 People fear getting famous; pigs fear growing fat Rén pà chūmíng zhū pà zhuàng 人怕出名猪怕壮 Only when the water in the big river is high will the small streams rise (An individual’s well-being depends on collective prosperity) Dà hé yǒu shuǐ xiǎo hé mǎn 大河有水小河满 305

Larry Herzberg

A single thread cannot make a cord; a lone tree cannot make a forest Dān sī bù chéng xiàn, gū mù bù chéng lín 单丝不成线,孤木不成林 If everyone is of one mind, yellow dirt can be made into gold (Almost anything can be accomplished if people will but work together) Dàjiā yītiáo xīn, huáng tǔ biànchéng jīn 大家一条心,黄土变成金

Age Unlike in contemporary Western society, where youth is glorified and old age is something to be lamented, traditional societies and particularly China revered the elderly. In ancient times, when there was little technological innovation or change from one generation to the next, the older you were, the more you knew about life, simply because you had experienced so much more than the young. Old age was revered to such an extent in ancient China that people wanted to be con­ sidered up in years as soon as possible. To that end, it has traditionally been true that the day you were born, you were said to be ‘one year old’. Each New Year you would gain another year of age. Therefore, if you happened to be born on the day before the Chinese New Year, usually sometime in February, you would be considered two years old when, by Western reckoning, you might only be two days old! While most of the proverbs and sayings below do reflect respect for the wisdom that comes with age, a few other popular maxims argue that the elderly are not always wiser for having lived longer. The largest number of sayings about age and aging, however, have nothing to do with wisdom or lack of it, but rather with our mortality. At three years old you can see the person when grown; at seven years old you can see the person when old (‘The child is father to the man’; our personalities can already be seen from childhood) Sān suì kàn dà, qī suì kàn lǎo 三岁看大,七岁看老 In one’s life how many times can you see the moon overhead? (Life is so brief, how many times can we experience good times?) Rén shēng jǐ jiàn yuè dāng tóu 人生几见月当头 The springtime of one’s life passes quickly and white hair is hard to avoid (Our youth passes quickly and old age comes all too soon) Qīngchūn yì guò, bái fǎ nán ráo 青春易过,白发难饶 Trees when old have more roots; people when old have more life experience Shù lǎo gēn duō, rén lǎo jiànshi duō 树老根多,人老见识多 We only sleep best right before dawn; people only learn wisdom in old age Shí dào tiān liàng fāng hǎo shuì, rén dào lǎo lái cái xué guāi 时到天亮方好睡,人到老来才学乖 306

Chinese proverbs and popular sayings

An old general who rides out on his horse (to battle) is worth any two others (A older person with experience is better than two young and inexperienced people in getting a job done) Lǎo jiàng chū mǎ, yígè dǐng liǎ 老将出马,一个顶俩 An old horse knows the way (Wisdom often comes with age) Lǎo mǎ shí tú 老马识途 People fear being poor in old age; rice fears a cold wind in late autumn Rén pà lǎo lái qióng, hé pà hán lù fēng 人怕老来穷,禾怕寒露风

Time Given our mortality, the older a person lives, the more he or she will come to realize that it is time, not money, that is our most precious treasure. The Chinese through the ages have emphasized that fact perhaps more than most other cultures. Here are some of the best statements of how fleeting life is and therefore all the more to be valued. If you compare an ‘inch’ (a small measure) of gold and an ‘inch’ of time, (you’ll find that) an inch of gold can’t buy an inch of time (Time is more valuable than money) Yícùn guāngyīn yícùn jīn, cùn jīn nán mǎi cùn guāngyīn 一寸光阴一寸金,寸金难买寸光阴 Gold lost may be found somewhere; time lost is nowhere to be found (You may recover money you’ve lost, but you can never recover lost time) Shīluò huángjīn yǒu chù zhǎo; shīluò guāngyīn wú chù xún 失落黄金有处找; 失落光阴无处寻 A hundred years of age is like a traveler passing by (Life is short and fleeting) Bǎi suì guāngyīn rú guò kè 百岁光阴如过客 One day not seeing (a dear one seems) like three autumns (One day apart from those we love seems three years; ‘Absence makes the heart grow fonder’) Yírì bújiàn rú gé sānqiū 一日不见如隔三秋 When seeing off a dear friend, there must finally be a farewell (‘All good things must come to an end’. Usually said in parting from a friend.) Sòng jūn qiān lǐ, zhōng xū yì bié 送君千里,终须一别 307

Larry Herzberg

Even the grandest feast must come to an end (All good things must come to an end; nothing lasts forever) Shèng yàn bì sàn 盛宴必散 No flower stays red for a hundred days (No flower blooms for long; ‘nothing gold can stay’) Huā wú bǎi rì hóng 花无百日红 An opportunity can’t be missed; if lost it will never come again (‘Opportunity knocks but once’) Jī bù kě shī, shī búzài lái 机不可失,失不再来 The whole year’s plan depends on a good start in spring; the whole day’s work depends on a good start at dawn (Do not procrastinate but use your time wisely from the beginning) Yīnián zhī jì zàiyú chūn; yírì zhī jì zàiyú chén 一年之计在于春; 一日之计在于晨

Friends Friendship is valued in every culture, but in few other cultures will you experience the depth of friendship that has always been prevalent in China. Confucius spoke of five sacred relationships, namely those between ruler and subject, husband and wife, parents and children, older and younger siblings, and between friends. The first four of these seminal relationships were hierarchical, with only the last being a relationship between equals. With the bonds of friendship made sacred by Confucius well over 2,000 years ago, it is no wonder that this relationship is particularly emphasized in Chinese culture. The Chinese poetry written in the Tang and Song dynasties, namely from the seventh century through the thirteenth century, is some of the greatest poetry in world literature. Much of it is concerned about the sorrow the poet feels at having to part from a close friend, because one of them has been assigned an official post in a far-off part of the empire. In Western culture, where romantic love has been glorified for so many centuries, the love between a man and a woman was considered the closest bond. In traditional China, however, where women were not generally given an education, it was the relationship between two male friends that was usually the stronger emotional bond. In this world if one has a friend who understands you, that friend seems near even in the farthest corners of the earth (The greatest distances cannot separate close friends, whom you carry in your heart) Hǎi nèi cún zhī jǐ, tiān yá ruò bǐ lín 海内存知己,天涯若比邻 When your hair is in disarray, look for a comb; When you heart is in disarray, look for a friend. Fǎ luàn zhǎo shūzi, xīn luàn zhǎo péngyǒu 发乱找梳子,心乱找朋友 308

Chinese proverbs and popular sayings

The friendship between two superior people is clear as shallow water; interaction between petty people is sweet as honey (True friendship is transparent and not muddied by selfishness; relationships between petty people are superficially honey-tongued but lack sincerity) Jūnzǐ zhī jiāo dàn rú shuǐ, xiǎo rén zhī jiāo tián rú mì 君子之交淡如水,小人之交甜如蜜 If water is too clear (and pure), you can’t raise fish (If you are too trivial or too inflexible about small matters, people will not hang around you) Shuǐ qīng bùyǎng yú 水清不养鱼 A thousand friends are too few, but one enemy is too many Péngyǒu qiānge shǎo, chóu rén yígè duō 朋友千个少,仇人一个多 At home rely on your parents; outside your door rely on your friends Zài jiā kào fùmǔ, chūmén kào péngyǒu 在家靠父母,出门靠朋友 If you hang around vermilion, you’ll turn red; If you hang around ink, you’ll turn black. (You are influenced by the company you keep) Jìn zhū zhě chì, jìn mò zhě hēi 近朱者赤,近墨者黑

Family The family is the basic unit of every society. But arguably few societies in the history of the world have placed such an emphasis on its importance and defined so carefully the nature of the relationships within it. For most of their history, the Chinese have lived in large extended families. When the sons of the family married, they would bring their bride to live with their parents and raise their children in their ancestral home. In the complex families that resulted, the relationships between the various family members, including parents and children, elder and younger brothers, husbands and wives, etc. were all governed by certain social protocol and there were distinct terms for each family member. The family and the relationships within it have been deemed so vital to the Chinese culture that there are no less than 460 separate terms for family members in the Chinese language, showing the great stress Chinese society has always placed on family. Children were expected to always be respectful and obedient to their parents, as well as to care for them in their old age. After all, children owed the gift of life to their father and mother, and this meant they must spend their lives repaying that gift. There is no adequate term in English to fully convey the weight of the Chinese concept that we translate into our language as ‘filial piety’, namely this obligation to respect and care for one’s parents not just in this life, but in the next world. The latter part of this responsibility has meant the burning of paper money and the saying of prayers for the repose of the souls of their dead parents that dates back more than several millennia. 309

Larry Herzberg

Having children to take care of you and to carry on the family line has always been considered the highest good in Chinese society. To this day, the word for ‘good/fine/well/OK’ is 好 (hǎo), the character for which shows a woman and a child together. For most of Chinese history the ideal was to have as many sons as possible, to be one’s security in old age. Here are but a few of the many sayings that focus on relationships within the family. If the family lives in harmony, everything will go well for them Jiā hé wàn shì xīng 家和万事兴 Every family’s story has pages that are difficult to read (Every family has its own problems and difficult issues to deal with) Jiā jiā yǒu běn nán niàn de jīng 家家有本难念的经 If you’re not members of the same family, you won’t enter by that family’s door (It’s not surprising that people in the same family should have the same opinions or habits as one another) Búshì yījiā rén bújìn yìjiā mén 不是一家人不进一家门 One raises children to provide against old age, just as one stores grain to provide against famine (Children are one’s security in old age) Yǎng ér fáng lǎo, jī gǔ fáng jī 养儿防老,积谷防饥 A son will always take after his father (‘Like father, like son’) Yǒu qí fù bì yǒu qí zǐ 有其父必有其子 A boatman’s child knows how to stay afloat on the water (You can’t help picking up something of your family’s trade) Chuánjiā háizi huì fú shuǐ 船家孩子会浮水 Other people’s wives always seem better than your own; (but) your own children always seem better than other people’s (In traditional China, you had no choice of a wife with arranged marriages, but every man has always favored his own children) Lǎopó shì biérén de hǎo, háizi shì zìjǐde hǎo 老婆是别人的好,孩子是自己的好 There are three types of unfilial behavior, the greatest of which is to have no descendants (The Confucian philosopher, Mencius, in the fourth century bce stated that the three major offenses against one’s parents were to not give them grandsons, to not support them when alive, and to not give them a decent burial when they died) Búxiào yǒu sān, wú hòu wéi dà 不孝有三,无后为大 310

Chinese proverbs and popular sayings

In the Yangtze River the waves in back push forward the waves in front of them; each generation is even stronger than the previous one (The older generation helps the younger, stronger one on its way and is replaced by it in an endless cycle) Chángjiāng hòu làng cuī qián làng; yídài gèng bǐ yídài qiáng 长江后浪催前浪; 一代更比一代强 One day together as husband and wife is like a hundred days of grace (The marriage relationship is a true blessing) Yírì fūqī bǎirì ēn 一日夫妻百日恩 If you are fated for one another, you will meet though separated by a thousand miles; If you are not fated for one another, you will not encounter each other even when face to face. Yǒu yuán qiān lǐ lái xiāng huì, wú yuán duì miàn bù xiāng féng 有缘千里来相会,无缘对面不相逢 A clever woman must often sleep next to a foolish man (A woman with brains often ends up married to a stupid man) Qiǎo fù cháng bàn zhuō fū mián 巧妇常伴拙夫眠

Women In traditional Chinese culture, as in almost every traditional society, the position of women was subservient to that of men. Women were taught to submit, first to their fathers, then to their husbands, and finally to their sons. They received no formal education and could not own property. Their only hope of achieving a position of respect was to marry and to live long enough to become the respected matriarch of their family. Although several women in Chinese history rose to become the effective rulers of the country, including the Empress Wu in the late seventh century and the Empress Dowager Ci Xi at the end of the nineteenth century, women were deemed inferior to men and treated accordingly. The many characters in the Chinese language in which the pictograph for ‘woman’ appears are indicative of how women were viewed in Chinese society. These characters include the words for ‘yielding’, ‘jealousy’, and ‘vanity’. Almost all the sayings below are ancient ones that reflect the traditional view of women. Only the last is the slogan that Mao Zedong was fond of quoting, that signaled a sea change in the position of women in China in the latter half of the twentieth century. Three women (together) make for a lively comedy (Whenever a group of women gather, there’s bound to be a lot of lively talk) Sānge nǚrén yìtái xì 三个女人一台戏 A daughter once married is like water poured out on the ground (Once a daughter marries, she is lost to her family forever, since by tradition she always has gone to live with her husband’s family) Jiàchūqù de nǚér, pōchūqù de shuǐ 嫁出去的女儿,泼出去的水 311

Larry Herzberg

Women can hold up half the sky (Women are capable of doing anything that men can do, and have a responsibility to do their share. This was a Communist slogan quoted often by Mao Zedong) Fùnǚ néng dǐng bàn biān tiān 妇女能顶半边天

Fate The Chinese have always been among the most fatalistic of people. In the West, we often say that our fate is determined by the Will of God. But we also claim that we are ‘captains of our own ship, masters of our own destiny’. Neither concept really exists in the way most Chinese view the world. Instead, the Chinese through the ages have expressed their sense that whatever man proposes, Fate disposes. All is in the hands not of a loving God but of Fate or Destiny, which is indifferent to us. One of the most common expressions in the contemporary Chinese language when trying to find a solution to one of life’s problems is 没有办法 (méiyǒu bànfǎ) ‘there’s nothing you can do about it’. Just as it is impossible for human beings to know what Fate has in store or to change their ‘luck’, it has always been considered equally impossible to challenge authority, whether political or within the family. We have therefore included some famous proverbs about the futility of trying to stand up to the powers that be, as well as the impossibility of forestalling destiny. One of the most famous proverbs is: ‘The old man on the border loses his horse; who knows if it isn’t good fortune (in disguise)’. Often this is all that is recited, since all Chinese know the rest of the proverb, namely: ‘The old man on the border gains a horse; who knows if it isn’t misfortune (in disguise)’. Equivalent English expressions include ‘Every cloud has a silver lining’ and ‘It might be a blessing in disguise’, by which we mean that when something unfortunate happens, it might turn out to be for the best. While the Chinese proverb acknowledges that, it also maintains that when something seemingly advantageous happens, it might turn out to be harmful to us in the end. Like all four-character idioms, this saying is based on an ancient tale. This one tells of an old peasant whose home is close to the border with Mongolia. One day his only horse, upon which he depends for plowing, runs away. When his neighbors all come to commiserate with him, he philosophically tells them that it might turn out to be for the best. Sure enough, several days later his horse returns, with a wild Mongolian pony trailing behind him. When all the neighbors come to congratulate the old man on this piece of unexpected good luck, the old man calmly philosophizes that this might turn out to be unfortunate in the end. As luck would have it, when the old man’s only son tries to break in the wild pony by riding it, he falls from the horse and breaks his leg. The old man shrugs off the sympathy from his neighbors by maintaining that perhaps it might turn out to be good fortune. He’s proven right when days later the army comes into the village to recruit young men to fight in the distant wars, with little likelihood they would return. But since the old man’s son has a broken leg, he is spared from the draft. No other story better sums up the resignation to fate that most Chinese have always felt, as well as their sense of powerlessness before forces greater than themselves. The third proverb below is taken from a story in the Liezi, a Daoist text of the fifth century bce. It tells of a man from the state of Qi who was always worrying that the sky would fall in. He would incessantly bother his friends by asking them what they would do should that happen. In spite of their reassurances that no such thing could happen, he never believed them and lived in a constant state of fear. This 成语 (chéngyǔ) is used to advise people to not worry needlessly about things that may never happen. 312

Chinese proverbs and popular sayings

The old man on the border loses a horse; who knows if it isn’t good fortune (in disguise)? [the old man on the border gains a horse; who knows if it isn’t bad fortune?] Sài wēng shī mǎ, yān zhī fēi fú 塞翁失马,焉知非福 Everything is controlled by fate and not in the slightest by people Wànshì jiē yóu mìng, bàn diǎn bù yóu rén 万事皆由命,半点不由人 The man from Qi worries about the sky (You shouldn’t worry about things you can’t control) Qǐ rén yōu tiān 杞人忧天 Blessings never come in pairs but misfortunes never come singly Fú wú shuāng zhì, huò bù dān xíng 福无双至,祸不单行 Just when you have a leaky roof you’ll meet with a prolonged rain storm; Just when your boat needs repairs you’ll meet with a head wind. (‘When it rains, it pours’; bad luck seems to strike us at the worst possible time) Wū lòu piān féng lián yīn yǔ; chuán pò piān yù dǐng tóu fēng 屋漏偏逢连阴雨; 船破偏遇顶头风 This four-character idiom is based on an ancient fable told by Zhuangzi of a praying mantis who felt himself so powerful that he tried to stand in the way of the king’s chariot when it was about to pass by. While the king was impressed with the courage of the tiny insect, he was also amused that so small a creature thought he could block a horse-drawn chariot going full speed.) Like a praying mantis trying to block a chariot with its little arms (‘You can’t fight City Hall’; it’s useless for the ‘common man’ to challenge the powers that be. Táng bì dàng chē 螳臂当车 Purposely plant flowers and the flowers won’t necessarily blossom; Mindlessly stick a willow sapling in the ground and it may grow to give you shade (‘The best laid plans of mice and men often go awry’ and we sometimes fail when trying too hard. On the other hand, sometimes we succeed with very little effort) Yǒu yì zhòng huā huā bù fā, wú xīn chā liǔ liǔ chéng yīn 有意种花花不发,无心插柳柳成荫 Even when there are no problems out there (in the world), foolish people create trouble for themselves. (‘Leave well enough alone’; don’t create problems where there aren’t any) Tiānxià běn wú shì, yōng rén zì rǎo zhī 天下本无事,庸人自扰之 313

Larry Herzberg

If at home you do not act morally, you’ll encounter a storm when you go out your door. (If in our daily lives we do immoral things, we’ll eventually meet with disaster out in the world.) Zài jiā bù xíng shàn, chū mén dà yǔ lín 在家不行善,出门大雨淋 When your cart comes to a mountain, there’s always a way (around it) Chē dào shān qián bì yǒu lù 车到山前必有路

314

17 Chinese Psycholinguistics Jerome L. Packard University of Illinois, USA

Introduction In the 1960s and 1970s, the field of Chinese psycholinguistics was driven by the psychology of Chinese script processing, due to intense popular and academic interest in the properties of the Chinese writing system. The decade of the 1970s showed a broadening beyond the Chinese character processing paradigm, expanding to include the psycholinguistics of Chinese as a spoken and aurally perceived language. This chapter therefore begins with a discussion of Chinese graph processing, followed by the psycholinguistics of lexical access, sentence processing, speech perception and speech production in Chinese.

Script processing Strokes Chinese characters are traditionally analyzed into the eight basic strokes that compose them. Strokes have no semantic or phonetic content – they are just the smallest functional units into which characters can be reduced. Characters are written in an imaginary box shape of fixed size, and are written stroke-by-stroke within the box from left-to-right, top-to-bottom. Because each character is written in the same size space regardless of the number of strokes, characters containing more strokes are generally more difficult to identify than those with fewer strokes (Huang and Hsu 2005). Y. P. Chen et al. (1996) found that stroke patterns, not strokes per se, are the smallest functional units that are relevant in Chinese character recognition. Tseng et al. (1965) found that when strokes are removed, subjects have difficulty identifying the character depending on the extent and nature of strokes removed. G. Yan et al. (2012) found results similar to Tseng et al., and found that the most important strokes tended to be located on the left side of the character, with the least important located in the bottom-right portion of the character. H. Wang et al. (2013) found that it was the deletion of segments – and not necessarily strokes – located on the left side of a character that caused identification difficulty.

Radicals Radicals are recurring character elements that have both a special formal and a special functional status within Chinese orthography. Their special formal status is that they have 315

Jerome L. Packard

been used in various Chinese character indexing and classification systems over the course of Chinese history. Their special functional status is that they carry specific information on the meaning and pronunciation of characters. Radicals are subparts of characters known as phonetic compounds, which comprise over 80% of all Chinese characters. Phonetic compounds have two subparts: a semantic radical – which usually occurs on the left – and a phonetic radical, which usually occurs on the right. As the names imply, semantic radicals contain information about the meaning of a character, and phonetic radicals contain information about a character’s pronunciation. Phonetic compound characters are composed from a set of about 200 semantic radicals and over 800 phonetic radicals. The processing of radicals as graphic units and not as bearers of phonetic or semantic information has been examined by several investigators. Feldman and Siok (1999) found that semantic radicals do have an effect in the course of character recognition, while Taft et al. (1999) and Ding et al. (2004) posit different levels of representation for radicals and complex characters, with the radical level of representation being activated by featural information, including positional features.

Basic properties of Chinese text reading The average number of words read per minute in Chinese is 386 as compared with 382 for English (Sun and Feng 1999). The mean fixation duration for Chinese reading does not differ substantially from that for English reading, in both cases being about 260ms (Sun and Feng 1999, Feng 2006). Given an average word length of 1.5 characters in Chinese and 6 letters for English, the average length of a saccade in skilled reading for both languages is about 1.5 words (2–2.5 characters for Chinese, and 7–9 letter spaces for English; Feng 2006). Feng (2006) gives the average recognition span for both languages at about 1.5 words, or approximately the length of an average saccade. Sun and Feng (1999) give average saccade length figures of 1.71 words for English and 1.75 words for Chinese. H. Yang and McConkie (1999) found no effect of the character (versus the inter-character space) in planning saccade landing sites during Chinese reading, meaning that the eyes are equally likely to land on a character or a space when reading. The authors did not find that words function as saccade targets in Chinese as they do in English, where readers rely on word-delineating spaces in parafoveal vision to plan saccades to the center of the following word. M. Yan et al. (2010), however, did find that words affect saccade planning in Chinese, demonstrating that Chinese readers target their saccade landing site at the center of words if they succeed in performing parafoveal word segmentation previewing. When they are not able to perform such parafoveal word segmentation, they send their eyes to the beginning rather than the center of the following word. Furthermore, Yen et al. (2012) offer evidence that readers of Chinese not only are able to perform word segmentation during parafoveal processing in planning saccades, they are also able to use statistical word-boundary information to plan those saccades. Thus it appears that even though Chinese does not explicitly mark word boundaries with spaces, readers of Chinese are able to perform word-segmentation analysis during parafoveal previewing, but that they are able to perform it less efficiently than in English due to the absence of word-delineating spaces.

316

Chinese psycholinguistics

Lexical access Phonetic and semantic activation in lexical access The topic of phonetic versus semantic processing of characters arose in earnest when Tzeng et al. (1977) found that phonetic recoding – activating the sound of a written form – occurs when reading Chinese, just as with any other writing system. The motivation for that research was in response to the long-held traditional belief that Chinese characters are ideographic – capable of directly representing meaning when read – with little, if any, involvement of a character’s pronunciation. This belief was held because Chinese characters do contain explicit semantic content in the form of pictographs, indicatives and semantic radicals, and it was popularly believed that the characters contained little, if any, systematic phonetic content. The Tzeng et al. findings gave rise to a spate of research in support of phonetic recoding in Chinese, offering evidence that phonology activation occurs at least as early as semantic activation in reading Chinese characters (S. Zhang and Perfetti 1993, Perfetti and Zhang 1995, Tan and Hoosain 1996, Weekes et al. 1998, Feng et al. 2001, Y. Liu et al. 2003, Kuo et al. 2004, Perfetti et al. 2005, and Lee et al. 2006). In response to the findings on phonetic recoding, other investigators offered competing results suggesting that phonological coding may be optional in reading Chinese and that it has no inherently privileged role, with semantic information in the lexicon activated at least as early and just as strongly as phonological information (Zhou and Marslen-Wilson (1999), H. C. Chen and Shu (2001), Meng et al. (2008), B. Liu et al. (2011), M. Yan et al. (2012). Neurophysiological research on this issue offers evidence of brain regions common to both logographic and alphabetic languages for the processing of orthographic, phonological, and semantic information in reading Chinese (Tan et al. 2001; and C. Wu et al. 2012). In addition, Lee et al. (2004) provide evidence that neuronal mechanisms for orthographic-tophonological conversion are universal across different writing systems, and that there are separate neural routes for addressed and assembled phonological reading processes. J. Yang et al. (2012b) found task by stimulus class interactions throughout the reading system, with stimulus selectivity directly related to task difficulty, suggesting that whether phonetic or semantic information receives priority coding during reading may in fact be dependent on the nature and difficulty of the task.

Words and morphemes in the lexicon The issue of lexical storage and retrieval is one of particular relevance in Chinese psycholinguistics due to the traditional emphasis on character orthography, which has often led to a corollary assumption that the character/morpheme serves as the basic lexical storage and retrieval processing unit in Chinese. Some of the earliest work on morpheme versus word processing in lexical access was done by Zhou and Marslen-Wilson (1994, 1995), who used auditory morpheme and word stimuli to test hypotheses regarding lexical access. The investigators found that two-morpheme words are stored and retrieved as complete, precompiled units rather than on an individual morpheme-by-morpheme basis, providing evidence that the word, not the morpheme, is the basic unit of lexical processing in Chinese. Results supporting this conclusion have been found by many investigators: J.-Y. Chen (1999a), Hsu and Huang (2000a), T. Chen and Chen (2006), Myers et al. (2006), Bai et al. (2008), Mok (2009), X. Li et al. (2012), J. Yang et al. (2012a) and X. Li et al. (2013). 317

Jerome L. Packard

Hsu and Huang (2000a), for example, found that word spacing reduced reading time and increased question response accuracy rate, leading the authors to conclude that Chinese text layouts using words as presentation units are superior to traditional layouts that do not mark word boundaries. X. Li et al. (2012) found that reading times are shorter when characters belonging to a single word are located on the same line rather than split between two adjacent lines, leading the authors to suggest that word boundaries should be taken into account when placing line breaks in Chinese texts. J. Yang et al. (2012a) found that processes of semantic integration in reading Chinese are performed at the word rather than the character level, arguing based on their results that word segmentation must take place very early in the course of Chinese script processing. X. Li et al. (2013) provide additional evidence of the advantage of word-based processing in Chinese reading. An alternative to the word-based processing view is provided by Jia et al. (2013), based on event-related potential (ERP) results involving the temporal characteristics of relational priming for two-character noun+noun words. The authors interpret their results as evidence for a serial, two-stage lexical processing model in which Chinese words are stored and accessed at stage one by individual morphemes, and in stage two by the whole word. Results supporting this view are also provided by Peng et al. (1999).

Sentence processing Sentence processing refers to the procedures by which the human language parser integrates words to derive the meaning of clausal and sentential propositions. Processing above the clause level is also examined in, e.g. syntactic dependencies such as anaphoric and other memory-based long-distance relations that occur between and within clauses and clause components. The investigation of Chinese has made significant contributions to the understanding of sentence processing. Investigating Mandarin verb selectional restrictions, P. Li et al. (2006) examined the behavioral and neurophysiological characteristics associated with different verb–object combinations, and found that the different degrees of fit between the verb and its object are an integral part of the mental representation of verbs that are accessed and used during on-line sentence processing. In experiments that used the competition model (e.g. MacWhinney and Bates, 1989) to examine the roles of word order, animacy and grammatical marker cues in sentence parsing, P. Li et al. (1992) found an overall cue strength order of passive marker bei > animacy > word order > object marker ba, interpreting their results as indicating that Chinese speakers prefer animacy over word order as a cue indicating agency, but that the cue strength of animacy may be overridden by the grammatical marker bei. Pronominal anaphora is the relation between an anaphor and its antecedent, which provides information about how the parser assigns reference to sentence constituents. In a series of experiments examining the reference of Mandarin pronominal expressions, C. Yang et al. (2003) found that syntactic prominence and lexical feature matching operate in concert to resolve pronoun reference, and that syntactic prominence is sensitive to structural relations in addition to grammatical functions (such as subject/object) and semantic roles (such as agent/patient) in determining the reference of pronouns. Anaphoric relations have a special status in Chinese, given that the reflexive pronoun ziji ‘oneself ’ may be either locally or long-distance bound, unlike in most other languages, in violation of Principle A of Chomsky’s (1981) binding theory. In an ERP study investigating whether this violation of Principle A incurs a processing cost, Li and Zhou (2010) found that ERP responses to reflexive ziji were significantly more positive in the long-distance reference 318

Chinese psycholinguistics

condition. The authors suggest that linking the reflexive with the distant rather than local antecedent requires more processing resources. Lexical ambiguity refers to a single word that has multiple meanings, with the parser resolving which of the meanings is intended based on context. The point of interest here is at what point the word is disambiguated and what types of information are used in the disambiguation. In Chinese the issue is particularly relevant because of the existence of so many single-syllable homophones, so Chinese strongly relies on sentential context to resolve the single-syllable ambiguity. Investigators have found that sentential context influences meaning access for ambiguous words at a very early stage in lexical processing (N. Wu and Shu 2002; P. Li et al. 2002), and that context and meaning dominance both have a significant influence on meaning access in the processing of ambiguous words. Like lexical ambiguity, syntactic ambiguity involves multiple candidates competing for prominence within an utterance, but in the case of syntax, the ambiguity results from a choice of candidate structures that arise from bracketing paradoxes, with the parser working to determine which bracketing is intended by the speaker. The parsing of a syntactic ambiguity often involves whether an incoming constituent is to be immediately integrated with the structure currently under construction, or whether the parser must wait for additional incoming information to make a parse. The resolution of a syntactic ambiguity often provides evidence in support of different parsing models, as in, for example, the choice between a parallel and serial processor. An example of competing parsing model evaluation occurs in the processing of Mandarin ‘Verb NP1 de NP2’ structures – which are temporarily ambiguous between a relative clause (RC; [[[øi Verb NP1] de] NP2i]) and a Verb (genitive)–Object (VO; [Verb]V [NP1 de NP2]O) structure. Both Y. Zhang et al. (2000) and Hsieh et al. (2009) found the RC interpretation of such sentences to be the default, but Hsieh et al. argue that although the RC analysis is preferred, Chinese speakers utilize all available sources of information to evaluate the structural alternatives, and do not commit to an interpretation until they have reached the disambiguation point. Hsieh et al. take this as evidence for a limited parallel processor that maintains multiple analyses if the structures are supported by the available constraints. The processing of object-gap versus subject-gap relative clauses has been a particularly controversial topic in Mandarin sentence processing. While research has shown that in English and many other languages, object-gap relative clauses are harder to process than subject-gap relative clauses (Frauenfelder et al. 1980, Frazier 1987, Schriefers et al. 1995, Traxler et al. 2002, Gibson et al. 2005, Kwon et al. 2010), the results with Mandarin are mixed, with most studies finding that subject-gap relative clauses are harder to process (Hsiao and Gibson 2003, Su et al. 2007, B. Chen et al. 2008, C. Yang et al. 2010, Packard et al. 2011, Lin and Garnsey 2011, Qiao et al. 2012, and Gibson and Wu 2013), but others offering evidence that object-gap clauses are harder to process (Lin 2008; Lin and Bever 2011; F. Wu et al. 2012 and Vasishth et al. 2013). Several factors affect relative clause gap site processing difficulty, all of which predicting that object relatives are harder to process in English but making mixed predictions for Mandarin: (1) accessibility of the relativized NP (Keenan and Comrie 1977), which predicts object relatives to be harder for Mandarin, (2) the preference for the role of agent to be expressed by the clause-initial NP (Diessel and Thomasello 2005), which predicts subject relatives to be harder for Mandarin, (3) canonical versus non-canonical word order within the relative clause, which predicts subject relatives to be harder in Mandarin, (4) frequency of occurrence, which predicts that object relatives will be harder in Mandarin, (5) linear 319

Jerome L. Packard

distance between gap and filler, which predicts subject relatives will be harder in Mandarin, and (6) structural distance (i.e. relative depth of embedding) separating gap and filler, which predicts that object relatives will be harder for Mandarin. The issue of distance between gap and filler in (5) and (6) arises because Chinese relative clauses occur to the left of the clause head, which is on the right. The distance hypothesis implicitly involves verbal working memory, because for the processor to retain gap information until the resolution of the identity of the filler is thought to be a short-term memory function. As seen above, of these six factors posited to explain relative clause gap site processing difficulty, all predict that object-gap relatives are more difficult to process in English, but they split the difference in predicting Mandarin processing difficulty. While it has proven a challenge to ferret out exactly which of the cited factors might be dominant in determining subject- and object-gap processing difficulty, the mixed results for Mandarin are not surprising given that the explanatory factors make mixed predictions. The relationship between syntactic and semantic processing has been hotly debated and investigated in Chinese sentence processing, mostly focusing on whether there is a syntaxindependent semantic compositional mechanism. Z. Ye et al. (2006) posit separate mechanisms, observing that syntactic processes appear earlier (at about 50ms) than semantic processes (at about 150ms) during the processing of Chinese ba sentences. Z. Ye et al. (2007) found verb semantic processing to be influenced by syntactic structure, and propose that sentential semantics might in fact be syntactically, rather than lexically, derived, as proposed by construction grammar (Goldberg 2005). In an experiment relying on eye-movement data, J. Yang et al. (2009) found that though semantic violations can be detected immediately by Chinese readers, nonetheless the processing of syntactic and semantic information is distinct in both first-pass and second-pass readings. Y. Zhang et al. (2010), however, argue that there is no functional primacy of syntactic over semantic processing in Chinese, and that semantic interpretation proceeds even in the absence of a well-formed syntactic parse. Y. Zhang et al. (2013) further demonstrate that semantic integration is attempted by the parser even in the absence of syntactic category or subcategorization information, and therefore that the processing of syntactic information is not a prerequisite for semantic integration in Chinese. Chow and Phillips (2013), in their examination of ERP responses to role-reversed sentences in Chinese, interpret their results as indicating that online semantic composition is entirely dependent upon surface syntax, and that there is no syntax-independent semantic compositional mechanism.

Speech perception Y. Wang et al. (2001), in a study of dichotic perception, found that Mandarin tones are predominantly processed in the left hemisphere by native Mandarin speakers, but bilaterally processed by non-tone language speakers, suggesting a left-hemisphere superiority for native Mandarin tone processing. Neuroanatomical evidence of such lateralization was provided by Klein et al. (2001), who used positron emission tomography (PET) to compare Mandarin tone perception in Mandarin speakers with that of non-tone language speakers, and found that the Mandarin speakers showed activation in frontal, parietal, and parietooccipital regions of the left hemisphere, and that only the non-tone-language group showed activity in the right inferior frontal cortex. In a similar finding, Nan and Friederici (2013) found that the neural network for linguistic tone processing includes portions of Broca’s area and the right superior temporal gyrus (STG), and that the right STG is more sensitive to musical tone than linguistic (Mandarin) tone. 320

Chinese psycholinguistics

Turning to studies examining the speech processing of tone and syllable information in Mandarin, Brown-Schmidt and Canseco-Gonzalez (2004) found hypothesized N400 ERP effects elicited by the three types of anomaly created by manipulating tone, syllable, or both tone and syllable, with the largest effect elicited by the tone-plus-syllable double anomaly. Zhao et al. (2011) also found larger N400 effects for mismatches involving the entire syllable than for partial mismatches in onset, rime, and tone. Both of these studies suggest that Chinese speech recognition might rely more on holistic syllable-based processing than phonemic segment-based processing. In studies that examined the relative importance of tonal versus segmental (i.e. CV) information, Malins and Joanisse (2010) conclude based on their eye-movement results that tonal and segmental information are accessed concurrently and play comparable roles in constraining lexical activation. In a contrary finding, Ye and Connine (1999) found differences in vowel and tone processing that the authors explain by positing separate representations for vowels and tones in the lexicon. Along those same lines, C. Li et al. (2013) found that tonal and segmental information both play roles in lexical retrieval, but that segmental information plays a greater role than tonal information, a result also found by Taft and Chen (1992).

Speech production Mandarin, in addition to being one of the most informationally dense spoken languages, with a language density factor of 0.94 (compared with a baseline value of 1.0 for Vietnamese, and values of 0.91 for English, 0.79 for German, 0.63 for Spanish and 0.49 for Japanese; Pellegrino et al. 2011), is also one of the slowest spoken languages, with a syllables/second speech rate of 5.18, compared with 5.22 for Vietnamese, 5.97 for German, 6.19 for English, 7.82 for Spanish and 7.84 for Japanese (Pellegrino et al. 2011). In terms of spoken words, Hsu and Huang (2000b) found that orally produced interword intervals were significantly longer than intercharacter intervals, providing evidence from speech production for the psychological reality of the word construct in Chinese. A multistage model is usually posited for speech production, beginning with a conceptual message stage followed by various functional (lexical, syntactic), positional and phonological output stages (e.g. Garrett 1988, Levelt 1989). Cai et al. (2012) present evidence from Chinese for an alternative, single-stage production model. Using a structural priming paradigm, these investigators found that, in addition to the usual syntactic priming effects, speakers of Mandarin also demonstrated priming effects in the mapping of conceptual elements to syntactic structure. To explain these results, Cai et al. argue against the traditional multistage production model and propose instead a one-stage model where the processor computes the function, structure and linear order of output constituents in a single stage. Some investigators have suggested that orthography activation occurs as part of the speech production process, at least in Chinese, but that it may be optional (Q. Zhang and Damian 2009), and that in any case it would occur between the conceptual activation (Q. Zhang and Damian 2009) and phonological activation (Q. Zhang and Weekes 2009) stages. J.-Y. Chen et al. (2002), in an implicit priming study examining the role of tone in the production of Mandarin disyllabic words, found that the same syllable but different tone produced priming, but that the same tone or onset alone did not. Based on these findings, the authors suggest that the syllable and not the individual components of the syllable is the basic planning unit in Chinese speech production, and that this is an intrinsic property of Chinese. T. Chen and Chen (2013) offer further support for this claim, finding syllable 321

Jerome L. Packard

priming but no segmental priming effects, regardless of whether the prompts were written, spoken, or pictorial in nature. Tang (2013) argues that the minimal building block of speech production is the word, based on the finding that when Mandarin speakers perform utterance repair for error correction or clarity enhancement, they most often suspend articulation at word closure. Tang further argues that it is the phrase, and not the clause or the sentence, that serves as the integral planning unit in speech production, based on the finding that when Mandarin speakers perform speech retracing during speech repair, they consistently return to the beginning of a phrasal boundary to restore the suspended articulation. In an investigation of speech errors involving tones in Chinese, J.-Y. Chen (1999b) found that tonal speech errors were rare compared with segmental errors, and that tones never participated in segmental movement errors. The author likened this finding to the relative paucity of English speech errors involving stress, arguing that tones in Mandarin are represented and processed similarly to lexical stress. L. Liu et al. (2006), in a study investigating whether the production of tone and CV units involves different neural mechanisms, found the answer to be in the affirmative, observing that there was more activation for tones than vowels in the right inferior frontal gyrus, and that tone production was less left-lateralized than vowel production, even though both showed left-hemisphere dominance. In a priming study that examined the role of morphological encoding in the production of Chinese disyllabic words, T. Chen and Chen (2006) found that morphological encoding is at best minimally involved in the production of Chinese disyllabic words, in contrast with morphological encoding results found for languages such as Dutch and English. This result comports with Packard (2000, 299–301), which posits that the only morphologically-complex words in Chinese formed on-line by rule are grammatical words, that is, words formed by the addition of a grammatical suffix.

Bibliography Bai, X., Yan, G., S. P. Liversedge, Zang, C., and K. Rayner (2008) ‘Reading Spaced and Unspaced Chinese Text: Evidence from Eye Movements’, Journal of Experimental Psychology: Human Perception and Performance 34: 1,277–88. Brown-Schmidt, S. and E. Canseco-Gonzalez (2004) ‘Who Do You Love, Your Mother or Your Horse? An Event-Related Brain Potential Analysis of Tone’, Journal of Psycholinguistic Research 33(2): 103–35. Cai, Z., M. Pickering, and H. Branigan (2012) ‘Mapping Concepts to Syntax: Evidence from Structural Priming in Mandarin Chinese’, Journal of Memory and Language 66(4): 833–49. Chen, B., Ning, A., Bi, H., and S. Dunlap (2008) ‘Chinese Subject-relative Clauses are More Difficult to Process than Object-relative Clauses’, Acta Psychologica 129: 61–5. Chen, H.-C. (ed) (1997) Cognitive Processing of Chinese and Related Asian Languages, Hong Kong: The Chinese University Press. Chen, H.-C. and Shu, H. (2001) ‘Lexical Activation during the Recognition of Chinese Characters: Evidence against Early Phonological Activation’, Psychonomic Bulletin and Review 8(3): 511–18. Chen, H.-C. and O. J.-L. Tzeng (eds) (1992) Language Processing in Chinese, Amsterdam: Elsevier. Chen, J.-Y. (1999a) ‘Word Recognition during the Reading of Chinese Sentences’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 239–56. Chen, J.-Y. (1999b) ‘The Representation and Processing of Tone in Mandarin Chinese: Evidence from Slips of the Tongue’, Applied Psycholinguistics 20(2): 289–301. Chen, J.-Y., Chen, T., and G. Dell (2002) ‘Word-form Encoding in Mandarin Chinese as Assessed by the Implicit Priming Task’, Journal of Memory and Language 46(4): 751–81. Chen, T. and Chen, J.-Y. (2006) ‘Morphological encoding in the production of compound words in Mandarin Chinese’, Journal of Memory and Language 54(4): 491–514.

322

Chinese psycholinguistics Chen, T. and Chen, J.-Y. (2013) ‘The Syllable as the Proximate Unit in Mandarin Chinese Word Production: An Intrinsic or Accidental Property of the Production System?’ Psychonomic Bulletin and Review 20(1): 154–62. Chen, X. and Kao, H. S. R. (2002) ‘Visual-Spatial Properties and Orthographic Processing of Chinese Characters’, in Kao, H. S. R., Leong, C.-K., and Gao, D.-G. (eds) Cognitive Neuroscience Studies of the Chinese Language, Hong Kong: Hong Kong University Press, 175–94. Chen, Y., Fu, S., S. Iversen, S. Smith, and P. Matthews (2002) ‘Testing for Dual Brain Processing Routes in Reading: A Direct Contrast of Chinese Character and Pinyin Reading Using fMRI’, Journal of Cognitive Neuroscience 14(7): 1,088–98. Chen, Y. P., D. A. Allport, and J. C. Marshall (1996) ‘What Are the Functional Orthographic Units in Chinese Word Recognition: The Stroke or the Stroke Pattern?’ Human Experimental Psychology: A Quarterly Journal of Experimental Psychology 49: 1,024–43. Chomsky, N. (1981) Lectures on Government and Binding, Dordrecht: Foris. Chow, W.-Y. and Phillips, C. (2013) ‘No semantic illusions in the “Semantic P600” phenomenon: ERP evidence from Mandarin Chinese’ Brain Research 1506: 76–93. Diessel, H. and M. Thomasello (2005) ‘A New Look at the Acquisition of Relative Clauses’, Language 81(4): 882–906. Ding, G., Peng, D., and M. Taft (2004) ‘The Nature of the Mental Representation of Radicals in Chinese: A Priming Study’, Journal of Experimental Psychology: Learning, Memory and Cognition 30(2): 530–9. Feldman, L. B. and W. W. T. Siok (1999) ‘Semantic Radicals Contribute to the Visual Identification of Chinese Characters’, Journal of Memory and Language 40(4): 559–76. Feng, G. (2006) ‘Eye Movements in Chinese Reading’, in Li, P., Tan, L. H., E. Bates, and Ovid J. L. Tzeng (eds) Handbook of East Asian Psycholinguistics (Vol. 1: Chinese), Cambridge: Cambridge University Press, 187–94. Feng, G., K. Miller, Shu, H., and Zhang, H. (2001) ‘Rowed to Recovery: The Use of Phonological and Orthographic Information in Reading Chinese and English’, Journal of Experimental Psychology: Learning, Memory, and Cognition 27(4): 1,079–100. Frazier, L. (1987) ‘Syntactic Processing: Evidence from Dutch’, Natural Language and Linguistic Theory 5: 519–59. Frauenfelder, U., J. Segui, and J. Mehler (1980) ‘Monitoring around the Relative Clause’, Journal of Verbal Learning and Verbal Behavior 19: 328–37. Garrett, Merrill F. (1988) ‘Processes in Language Production’, in Newmeyer, F. (ed) Linguistics: The Cambridge Survey: Vol. III, Psychological and Biological Aspects, Cambridge: Cambridge University Press, 69–96. Gibson, E., T. Desmet, D. Watson, D. Grodner, and Ko, K. (2005) ‘Reading Relative Clauses in English, Cognitive Linguistics 16(2): 313–53. Gibson, E. and Wu, I. (2013) ‘Processing Chinese Relative Clauses in Context’, Language and Cognitive Processes 28(1-2): 125–55. Goldberg, A. (2005) The Language of Constructions, Oxford: Oxford University Press. Hsiao, F. and E. Gibson (2003) ‘Processing Relative Clauses in Chinese’, Cognition 90: 3–27. Hsieh, Y., J. Boland, Zhang, Y., and Yan, M. (2009) ‘Limited Syntactic Parallelism in Chinese Ambiguity Resolution’, Language and Cognitive Processes 24(7–8): 1,227–64. Huang, K. and Hsu, S. (2005) ‘Effects of Numbers of Strokes on Chinese Character Recognition during a Normal Reading Condition’, Percept Mot Skills 101(3): 845–52. Hsu, S. and Huang, K. (2000a) ‘Effects of Word Spacing on Reading Chinese Text from a Video Display Terminal’, Percept Mot Skills 90(1): 81–92. Hsu, S. and Huang, K. (2000b) ‘Interword Spacing in Chinese Text Layout’, Percept Mot Skills 91(2): 355–65. Jia, X., Wang, S., Zhang, B., and Zhang, J. (2013) ‘Electrophysiological Evidence for Relation Information Activation in Chinese Compound Word Comprehension’, Neuropsychologia 51(7): 1,296–1,301. Kao, H. S. R., Leong, C.-K. and Gao, D.-G. (eds) (2002) Cognitive Neuroscience Studies of the Chinese Language, Hong Kong: Hong Kong University Press. Keenan, E. L., and B. Comrie (1977) ‘Noun Phrase Accessibility and Universal Grammar’, Linguistic Inquiry 8: 63–99. Klein, D., R. Zatorre, B. Milner, and Zhao, V. (2001) ‘A Cross-linguistic PET Study of Tone Perception in Mandarin Chinese and English Speakers’, NeuroImage 13(4): 646–53.

323

Jerome L. Packard Kuo, W., Yeh, T., Lee, J., Chen, L., Lee, P., Chen, S., Ho, L., Hung, D. L., Ovid J. L. Tzeng, and Hsieh, J. (2004) ‘Orthographic and Phonological Processing of Chinese Characters: An fMRI Study’, NeuroImage 21(4): 1,721–31. Kwon, N, Lee, Y., P. C. Gordon, R. Kluender, and M. Polinsky (2010) ‘Cognitive and Linguistic Factors Affecting Subject/Object Asymmetry: An Eye-tracking Study of Pre-nominal Relative Clauses in Korean’, Language 86(3): 546–82. Lee, C., Tsai, J., Huang, H., Hung, D. L. and Ovid, J. L. Tzeng (2006) ‘The temporal signatures of semantic and phonological activations for Chinese sublexical processing: An event-related potential study’, Brain Research 1121(1): 150–9. Lee, C.-Y., Tsai, J.-L., Kuo, W.-J., Yeh, T.-C., Wu, Y.-T., Ho, L.-T., Hung, D. L., Tzeng, O. J.-L., and Hsieh, J.-C. (2004) ‘Neuronal correlates of consistency and frequency effects on Chinese character naming: an event-related fMRI study’, NeuroImage 23(4): 1,235–45. Levelt, W. J. M. (1989) Speaking: From Intention to Articulation, Cambridge, MA: MIT Press. Li, C., Lin, C., Wang, M., and Jiang, N. (2013) ‘The Activation of Segmental and Tonal Information in Visual Word Recognition’, Psychonomic Bulletin and Review 20(4): 773–79. Li, P. (1998) ‘Crosslinguistic Variation and Sentence Processing: The Case of Chinese’, in D. Hillert (ed) Sentence Processing: A Cross-linguistic Perspective, San Diego: Academic Press. Li, P., E. Bates, Liu, H., and B. MacWhinney (1992) ‘Cues as Functional Constraints on Sentence Processing in Chinese’, in Chen, H.-C. and O. J.-L. Tzeng (eds) Language Processing in Chinese, Amsterdam: Elsevier, 207–34. Li, P., Shu, H., Yip, M., Zhang, Y., and Tang, Y. (2002) ‘Lexical Ambiguity in Sentence Processing: Evidence from Chinese’, in M. Nakayama (ed) Sentence Processing in East Asian Languages, Stanford, CA: Center for the Study of Language and Information Publications, 111–29. Li, P., Tan, L. H., E. Bates, and O. J.-L. Tzeng (eds) (2006) Handbook of East Asian Psycholinguistics (Vol. 1: Chinese), Cambridge: Cambridge University Press. Li, P. and Yip, M. (1998) ‘Context Effects and Processing of Spoken Homophones’, in C. K. Leong and K. Tamaoka (eds) Reading and Writing: An Interdisciplinary Journal 10: 223–43. Li, W., J. S. Gaffney, and J. L. Packard (eds) (2002) Chinese Children’s Reading Acquisition: Theoretical and Pedagogical Issues, Dordrecht: Kluwer Academic Publishers. Li, X., J. Gu, P. Liu, and K. Rayner (2013) ‘The Advantage of Word-based Processing in Chinese Reading: Evidence from Eye Movements’, Journal of Experimental Psychology: Learning, Memory, and Cognition 39(3): 879–89. Li, X. and Shen, W. (2013) ‘Joint Effect of Insertion of Spaces and Word Length in Saccade Target Selection in Chinese Reading’, Journal of Research in Reading 36(1): 64–77. Li, X., Shu, H., Liu, Y., and Li, P. (2006) ‘Mental Representation of Verb Meaning: Behavioral and Electrophysiological Evidence’, Journal of Cognitive Neuroscience 18(10): 1,774–87 Li, X., Zhao, W., and A. Pollatsek (2012) ‘Dividing Lines at the Word Boundary Position Helps Reading in Chinese’, Psychonomic Bulletin and Review 19(5): 929–34. Li, X. and Zhou, X. (2010) ‘Who is ziji? ERP responses to the Chinese reflexive pronoun during sentence comprehension’, Brain Research 1331: 96–104. Lin, C. (2008) ‘The Processing Foundation of Head-final Relative Clauses’, Language and Linguistics 94: 813–39. Lin, C. and T. Bever (2011) ‘Garden Path and the Comprehension of Head-final Relative Clauses’, in H. Yamashita, Y. Hirose, and J. Packard (eds) Processing and Producing Head-final Structures. Studies in Theoretical Psycholinguistics 38, Dordrecht: Springer Publishing Company, 277–97. Lin, Y. and S. Garnsey (2011) ‘Animacy and the Resolution of Temporary Ambiguity in Relative Clause Comprehension in Mandarin’, in H. Yamashita, Y. Hirose, and J. Packard (eds) Processing and Producing Head-final Structures. Studies in Theoretical Psycholinguistics 38, Dordrecht: Springer Publishing Company, 241–75. Liu, B., Jin, Z., Qing, Z., and Wang, Z. (2011) ‘The Processing of Phonological, Orthographical, and Lexical Information of Chinese Characters in Sentence Contexts: An ERP Study’, Brain Research 1372: 81–91. Liu, L., Peng, D., Ding, G., Jin, Z., Zhang, L., Li, K., and Chen, C. (2006) ‘Dissociation in the Neural Basis Underlying Chinese Tone and Vowel Production’, NeuroImage 29(2): 515–23. Liu, Y. and Peng, D. (1997) ‘Meaning Access of Chinese Compounds and its Time Course’, in Chen, H.-C. (ed) Cognitive Processing of Chinese and Related Asian Languages, Hong Kong: The Chinese University Press, 219–32.

324

Chinese psycholinguistics Liu, Y., C. Perfetti, and L. Hart (2003) ‘ERP Evidence for the Time Course of Graphic, Phonological, and Semantic Information in Chinese Meaning and Pronunciation Decisions’, Journal of Experimental Psychology. Learning, Memory and Cognition 29(6): 1,231–47. Liversedge, S., J. Hyönä, and K. Rayner (2013) ‘Eye Movements during Chinese Reading’, Journal of Research in Reading 36(1): 1–3. MacWhinney, B. and E. Bates (1989) The Crosslinguistic Study of Sentence Processing, New York: Cambridge University Press. McConkie, G. W. and K. Rayner (1975) ‘The Span of the Effective Stimulus during a Fixation in Reading’, Perception and Psychophysics 17: 578–86. Malins, J. and M. Joanisse (2010) ‘The Roles of Tonal and Segmental Information in Mandarin Spoken Word Recognition: An Eyetracking Study’, Journal of Memory and Language 62(4): 407–20. Meng, X., Jian, J., Shu, H., Tian, X., and Zhou, X. (2008) ‘ERP Correlates of the Development of Orthographical and Phonological Processing during Chinese Sentence Reading’, Brain Research 1219: 91–102. Miao, X. (1999) ‘Sentence Understanding in Chinese’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 279–95. Mok, L. (2009) ‘Word-superiority Effect as a Function of Semantic Transparency of Chinese Bimorphemic Compound Words’, Language and Cognitive Processes 24(7–8): 1,039–81. Myers, J., Huang, Y., and Wang, W. (2006) ‘Frequency Effects in the Processing of Chinese Inflection’, Journal of Memory and Language 54(3): 300–23. Nakayama, M. (ed) (2002) Sentence Processing in East Asian Languages, Stanford CA: Center for the Study of Language and Information Publications. Nan, Y. and A. Friederici (2013) ‘Differential Roles of Right Temporal Cortex and Broca’s Area in Pitch Processing: Evidence from Music and Mandarin’, Human Brain Mapping 34(9): 2,045– 54. Newmeyer, F. (ed) (1988) Linguistics: The Cambridge Survey: Vol. III, Psychological and Biological Aspects, Cambridge: Cambridge University Press. Packard, J. (2000) The Morphology of Chinese, Cambridge: Cambridge University Press. Packard, J., Ye, Z., and Zhou, X. (2011) ‘Filler-gap Processing in Mandarin Relative Clauses: Evidence from Event-related Potentials’, in H. Yamashita, Y. Hirose, and J. Packard (eds) Processing and Producing Head-final Structures. Studies in Theoretical Psycholinguistics 38, Dordrecht: Springer Publishing Company, 219–40. Pellegrino, F., C. Coupé, and M. Marisco (2011) ‘A Cross-language Perspective on Speech Information Rate’, Language 87(3): 539–58. Peng, D., Liu, Y., and Wang, C. (1999) ‘How is Access Representation Organized? The Relation of Polymorphemic Words and their Relation in Chinese’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 65–89. Perfetti, C., Liu, Y., and Tan, L. (2005) ‘The Lexical Constituency Model: Some Implications of Research on Chinese for General Theories of Reading’, Psychological Review 112(1): 43–59. Perfetti, C. and Tan, L. (1999) ‘The Constituency Model of Chinese Word Identification’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 115–34. Perfetti, C. and Zhang, S. (1995) ‘Very Early Phonological Activation in Chinese Reading’, Journal of Experimental Psychology, Learning, Memory and Cognition 21(1): 24–34. Qiao, X., Shen, L., and K. Forster (2012) ‘Relative Clause Processing in Mandarin: Evidence from the Maze Task’, Language and Cognitive Processes 27: 611–30. Schriefers, H., A. Friederici, and K. Kuhn (1995) ‘The Processing of Locally Ambiguous Relative Clauses in German’, Journal of Memory and Language 34: 499–520. Sheng, Y. and Wu, F. (2013) ‘Audience Design Affects Classifiers Positioning in Chinese Relative Clauses: Evidence from Spoken Corpus and Sentence-production Data’, in Proceedings of GLOW in Asia IX 2012. Su, Y., Lee, S., and Chung, Y. (2007) ‘Asyntactic Thematic Role Assignment by Mandarin Aphasics: A Test of the Trace-deletion Hypothesis and the Double Dependency Hypothesis’, Brain and Language 101: 1–18. Sun, F. and Feng, D. (1999) ‘Eye Movements in Reading Chinese and English Text’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 189–206.

325

Jerome L. Packard Taft, M. and H.-C. Chen (1992) ‘Judging Homophony in Chinese: The Influence of Tones’, in Chen, H.-C. and O. J.-L. Tzeng (eds) Language Processing in Chinese, Amsterdam: Elsevier, 151–72. Taft, M. and Zhu, X. (1997a) ‘Using Masked Priming to Examine Lexical Storage of Chinese Compound Words, in Chen, H.-C. (ed) (1997) Cognitive Processing of Chinese and Related Asian Languages, Hong Kong: The Chinese University Press, 233–41. Taft, M., Zhu, X., and Peng, D. (1999) ‘Positional Specificity of Radicals in Chinese Character Recognition’, Journal of Memory and Language 40(4): 498–519. Tan, L. and R. Hoosain (1996) ‘Activation of Phonological Codes before Access to Character Meaning in Written Chinese’, Journal of Experimental Psychology: Learning, Memory and Cognition 22(4): 865–83. Tan, L., Liu, H., C. Perfetti, J. Spinks, P. Fox, and Gao, J. (2001) ‘The Neural System Underlying Chinese Logograph Reading’, NeuroImage 13(5): 836–46. Tang, C. (2013) ‘Planning Units in Speech Production: Evidence from Anticipatory Retracing in Spoken Mandarin Chinese Narratives’, Chinese Language and Discourse 4(2): 253–75. Traxler, M. J., R. K. Morris, and R. E. Seely (2002) ‘Processing Subject and Object Relative Clauses: Evidence from Eye Movements’, Journal of Memory and Language 47: 69–90. Tseng, S. C., Chang, L. H., and Wang, C. C. (1965) ‘An informational analysis of the Chinese language: the reconstruction of the removed strokes of the ideograms in printed sentence-texts’ [in Chinese], Acta Psychologica Sinica 10: 299–306. Tzeng, O. J.-L. (2002) ‘Current Issues in Learning to Read Chinese’, in Li, W., J. S. Gaffney, and J. L. Packard (eds) Chinese Children’s Reading Acquisition: Theoretical and Pedagogical Issues, Dordrecht: Kluwer Academic Publishers, 3–15. Tzeng, O. J.-L., Hung, D. L., and Wang, W. S.-Y. (1977) ‘Speech Recoding in Reading Chinese Characters’, Journal of Experimental Psychology: Human Learning and Memory 3(6): 621–30. Tzeng, O. J.-L., Chen, S., and Hung, D. L. (1991) ‘The Classifier Problem in Chinese Aphasia’, Brain and Language 41: 184–202. Vasishth, S., Chen, Z., Li, Q., and Guo, G. (2013) ‘Processing Chinese Relative Clauses: Evidence for the Subject-relative Advantage’, PLoS One 8(10): e77006. Published online 2013 October 2. doi: 10.1371/journal.pone.0077006. Wang, H., E. Schotter, B. Angele, Yang, J., D. Simovici, M. Pomplun, and K. Rayner (2013) ‘Using Singular Value Decomposition to Investigate Degraded Chinese Character Recognition: Evidence from Eye Movements during Reading’, Journal of Research in Reading 36(1): 35–50. Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) (1999) Reading Chinese Script, Mahwah, NJ: Erlbaum. Wang, Y., A. Jongman, and J. Sereno (2001) ‘Dichotic Perception of Mandarin Tones by Chinese and American Listeners’, Brain and Language 78: 332–348. Weekes, B., Chen, M. J. and Lin, Y.-B. (1998) ‘Differential Effects of Phonological Priming on Chinese Character Recognition’, Reading and Writing 10(3–5): 201–21. Wong, A. and H.-C. Chen (2012) ‘Is Syntactic-category Processing Obligatory in Visual Word Recognition? Evidence from Chinese’, Language and Cognitive Processes 27(9): 1,334–60. Wu, C., Ho, M. and Chen, S. (2012) ‘A Meta-analysis of fMRI Studies on Chinese Orthographic, Phonological, and Semantic Processing’, NeuroImage 63(1): 381–91. Wu, F., E. Kaiser, and E. Andersen (2012) ‘Animacy Effects in Chinese Relative Clause Processing’, Language and Cognitive Processes 27(10): 1,489–524. Wu, J., Wang, B., Yan, T., Li, X., Bao, X., and Guo, Q. (2012) ‘Different Roles of the Posterior Inferior Frontal Gyrus in Chinese Character Form Judgment Differences between Literate and Illiterate Individuals’, Brain Research 1431: 69–76. Wu, N. and Shu, H. (2002) ‘Meaning Activation of Chinese Syntactic Category Ambiguous Words in Sentence Context’, Acta Psychologica Sinica (in Chinese) 34: 454–61. Yamashita, H., Y. Hirose, and J. Packard (eds) (2011) Processing and Producing Head-final Structures. Studies in Theoretical Psycholinguistics 38, Dordrecht: Springer Publishing Company. Yan, G., Bai, X., Zang, C., Bian, Q., Cui, L., Qi, W., K. Rayner, and S. Liversedge (2012) ‘Using Stroke Removal to Investigate Chinese Character Identification during Reading: Evidence from Eye Movements’, Reading and Writing 25: 951–79. Yan, M., R. Kliegl, E. M. Richter, A. Nuthmann, and Shu, H. (2010) ‘Flexible Saccade-target Selection in Chinese Reading, Quarterly Journal of Experimental Psychology 63: 705–25.

326

Chinese psycholinguistics Yan, M., Zhou, W., Shu, H., and R. Kliegl (2012) ‘Lexical and Sublexical Semantic Preview Benefits in Chinese Reading’, Journal of Experimental Psychology: Learning, Memory, and Cognition 38(4): 1,069–75. Yang, C., P. Gordon, R. Hendrick, and C. Hue (2003) ‘Constraining the Comprehension of Pronominal Expressions in Chinese’, Cognition 86: 283–315. Yang C., C. Perfetti, and Liu, Y. (2010) ‘Sentence Integration Processes: An ERP Study of Chinese Sentence Comprehension with Relative Clauses’, Brain and Language 112: 85–100. Yang, H. and G. McConkie (1999) ‘Reading Chinese: Some Basic Eye-movement Characteristics’, in Wang, J., Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script, Mahwah, NJ: Erlbaum, 207–22. Yang, J., Wang, S., Chen, H.-C., and K. Rayner (2009) ‘The Time Course of Semantic and Syntactic Processing in Chinese Sentence Comprehension: Evidence from Eye Movements’, Memory and Cognition 37(8): 1,164–76. Yang, J., A. Staub, Li, N., Wang, S., and K. Rayner (2012a) ‘Plausibility Effects When Reading Oneand Two-character Words in Chinese: Evidence from Eye Movements’, Journal of Experimental Psychology: Learning, Memory, and Cognition 38(6): 1,801–9. Yang, J., Wang, X., Shu, H., and J. Zevin (2012b) ‘Task by Stimulus Interactions in Brain Responses during Chinese Character Processing’, NeuroImage 60(2): 979–90. Ye, Y. and C. Connine (1999) ‘Processing Spoken Chinese: The Role of Tone Information’, Language and Cognitive Processes 14(5–6): 609–30. Ye, Z., Zhan, W., and Zhou, X. (2007) ‘The Semantic Processing of Syntactic Structure in Sentence Comprehension: An ERP Study’, Brain Research 1142: 135–45. Ye, Z., Luo, Y., A. D. Friederici, and Zhou, X. (2006) ‘Semantic Processing of Syntactic Processing in Chinese Sentence Comprehension: Evidence from Event-related Potentials’, Brain Research 1071: 186–96. Yen, M., R. Radach, Tzeng, O. J.-L., and Tsai, J. (2012) ‘Usage of Statistic Cues for Word Boundary in Reading Chinese Sentences’, Reading and Writing 25(5): 1,007–29. Zhao, J., Guo, J., Zhou, F., and Shu, H. (2011) ‘Time Course of Chinese Monosyllabic Spoken Word Recognition: Evidence from ERP Analyses’, Neuropsychologia 49(7): 1,761–70. Zhang, Biyin and D. Peng (1992) ‘Decomposed Storage in the Chinese Lexicon’, in Chen, H.-C. and O. J.-L. Tzeng (eds) (1992) Language Processing in Chinese, Amsterdam: Elsevier, 131–50. Zhang, Y., Li, P., Piao, Q., Liu, Y., Huang, Y., and Shu, H. (2013) ‘Syntax Does Not Necessarily Precede Semantics in Sentence Processing: ERP Evidence from Chinese’, Brain and Language 126(1): 8–19. Zhang, Y., Wu, N., and Yip, M. (2006) ‘Lexical Ambiguity Resolution in Chinese Sentence Processing’, in Li, P., Tan, L. H., E. Bates, and O. J.-L. Tzeng (eds) Handbook of East Asian Psycholinguistics (Vol. 1: Chinese), Cambridge: Cambridge University Press, 187–94. Zhang, Q, and M. Damian (2009) ‘The Time Course of Semantic and Orthographic Encoding in Chinese Word Production: An Event-related Potential Study’, Brain Research 1273: 92–105. Zhang, Q. and B. Weekes (2009) ‘Orthographic Facilitation Effects on Spoken Word Production: Evidence from Chinese’, Language and Cognitive Processes 24(7–8): 1,082–96. Zhang, S. and C. Perfetti (1993) ‘The Tongue-twister Effect in Reading Chinese’, Chinese Journal of Experimental Psychology. Learning, Memory and Cognition 19(5): 1,082–93. Zhang, Y., Yu, J., and J. Boland (2010) ‘Semantics Does Not Need a Processing License from Syntax in Reading Chinese’, Journal of Experimental Psychology: Learning, Memory, and Cognition 36(3): 765–81. Zhang, Y., Zhang, H., and Shu, H. (2000) ‘A Study on the Processing of Ambiguous Phrases in Chinese’, Chinese Journal of Psychology 32: 12–19. Zhao, J., Guo, J., Zhou, F., and Shu, H. (2011) ‘Time Course of Chinese Monosyllabic Spoken Word Recognition: Evidence from ERP Analyses’, Neuropsychologia 49(7): 1,761–70. Zhou, X. and W. Marslen-Wilson (1994) ‘Words, Morphemes and Syllables in the Chinese Mental Lexicon’, Language and Cognitive Processes 9(3): 393–422. Zhou, X. and W. Marslen-Wilson (1995) ‘Morphological Structure in the Chinese Mental Lexicon’, Language and Cognitive Processes 10(6): 545–600. Zhou, X. and W. Marslen-Wilson (1999) ‘Phonology, Orthography, and Semantic Activation in Reading Chinese’, Journal of Memory and Language 41(4): 579–606.

327

18 Chinese Rhetoric Andy Kirkpatrick Griffith University, Australia

Rhetoric has long been a discrete discipline in the Western tradition, but this was not the case in Classical China, where a variety of different terms all touching on some aspect of ‘rhetoric’ were used and rhetoric was not considered a separate discipline. This is one reason why it is difficult to be accurate about the origins of Chinese rhetoric. Different terms described different aspects of rhetoric. These included: yan (言), speech and the use of language; ci (辞), modes of speech, types of discourse, eloquence, style; and jian (谏), giving advice, persuasion (Lu 1998: 5). The current Chinese term for rhetoric is xiuci (修辞), which literally means something like ‘refining words’. Its first recorded use is in the Confucian Classic, The Book of Changes. As Lu points out in her book, Rhetoric in Ancient China, persuasion was known as shui (说), explanation ming (明), and argumentation bian (辩). Although there is overlap between these terms (and others), Lu argues that each word has a particular function in conceptualising and contextualising persuasive discourse. For example shui is associated with face-to-face persuasion and ming deals with the use of symbols in social and epistemological contexts. Lu suggests that the term mingbianxue (名辩学) is comparable to the Western study of rhetoric, with ming aiming to seek truth and justice and bian concerning the art of persuasion. This term also captures the contradiction inherent in the two key concepts of Western rhetoric, viz truth and/or persuasion. For the purpose of this entry, we will adopt Lu’s definition of rhetoric as ‘the art of persuasion, the artistic use of oral and written expressions, for the purpose of changing thought and action at social, political and individual levels’ (1998: 2). The Chinese rhetorical tradition was not monolithic, as frequently supposed, but represented by several schools. The Confucian school was much concerned with the moral impact of speech and the need to maintain social order. ‘Sophistry’ was frowned upon and the orthodox Confucian style stressed clarity and simplicity. The Confucian style has been summarised in the following way: ‘Explaining things plainly and simply is good enough’ (Pu and Wei 1983: 111). Confucius himself (b. 551 bce) was the Minister of Justice in the State of Lu, before departing, disappointed that his policies were not being adopted. He then hawked himself around several other states trying to win influence, but without success, only returning home when he was in his late sixties. A second school was Mohism. Mohist philosophers were concerned with developing a rational system of argumentation (Graham 1989). Garrett has suggested that the Mohists 328

Chinese rhetoric

were the only people who used hypothetico-deductive arguments and a possible reason for this was that they were artisans and engineers. But these deductive methods of argument ‘seem to have exerted remarkably little influence on their contemporaries and even less on later historical tradition’ (1983: 9). The Ming school was more influential. Their concern was with ‘probability, relativism and classification under the general umbrella of epistemology and social justice’ (Kirkpatrick and Xu 2012: 14). The best-known exponent of this school was Gong-sun Long (公孙龙) (325–250 bce) who is probably best known for the famous ‘white horse’ dialogue where debaters argue over the possible meanings of the statement ‘white horses are not horses’ (Graham 1989). This type of paradox is in the tradition of Hui Shi (惠施) (380–305 bce), whose aphorisms include ‘the sun at noon is the sun declining’ and ‘a creature born is a creature dying’ (Forke 1901: 2). He also contented that ‘there is a time when a swiftly flying arrow is neither moving nor at rest’, a conundrum similar to Zeno’s paradox of the arrow which stated: When the arrow is in a place of exactly its own size, it is at rest; in flight the arrow is always in a place exactly its own size; an arrow in flight is therefore at rest. Hui Shi was a member of the zongheng (纵横) school. The school’s founder was Gui Guzi (鬼谷子), who was more or less contemporary with Aristotle. His eponymous book, which, like the Confucian Analects, was almost certainly compiled by his disciples rather than by Gui Guzi himself, represents an important benchmark in the history of Chinese philosophy and rhetoric for reasons outlined below. Distinctions between Western and Chinese concepts of rhetoric have been frequently drawn (Harbsmeier and Harbsemeier 1999; Shankmann and Durrant 2002; Jullien 2004). The distinctions include the origins of rhetoric, with the law courts of Classical Greece providing the impetus for the use of forensic rhetoric while, in China, rhetoric was used to persuade rulers (Graham 1989). This leads to a further distinction of Classical Greek rhetoric, namely its use in Greece among equals, while Chinese rhetoric was more ‘bottom-up’. This, in turn, meant that the Chinese persuader had to be careful. Gui Guzi was clearly aware of this and the influence of the audience upon rhetorical style. He considered that persuasion from below ( yin 阴) to above ( yang 阳) to be a disturbance of the natural order of things. Yin persuasion from below to above or from an inferior to a superior required special effort. Persuading from above to below was yang and required less effort. ‘Yang  .  .  .  encourages straightforward speaking. Yin  .  .  .  encourages speaking in forked tongue’ (Tsao 1985: 103). That ‘speaking in forked tongue’ encouraged yin persuasion led Confucian philosophers to criticise the Gui Guzi school. Xunzi described the school’s disciples as ‘ingratiating courtiers’ who were ‘inadequate in uniting people domestically, inadequate in confronting enemies externally, unable to win the affinity of the people nor the trust of the nobles. But they were good at crafty persuasion and good at courting favour from the high ranks’ (Tsao 1985: 19). As will be suggested below, however, contemporary Chinese public and political discourse might benefit from an infusion of yin persuasion from the Chinese rhetorical tradition. A further distinction which developed as a consequence of the different Chinese and Western rhetorical traditions and the social and political contexts in which they operated is that Western rhetoric is considered more ‘agonistic’ while Chinese is considered more ‘irenic’ (Durrant 2002: 283). But, as Durrant also points out, Chinese could be agonistic, giving as examples Wang Chong 王充 (ce 27–110) and others who criticised the great historian Sima Qian 司马迁 (c. 145–90 bce). This is a salutary reminder that there was (and remains) 329

Andy Kirkpatrick

many different ways of ‘doing’ rhetoric in both these great traditions. In this context, while there is truth that Chinese rhetoric is more associated with writing than speech, while Western rhetoric more with speech than writing, it is certainly not true to say that this was ever at the expense of the other. The philosopher-persuaders mentioned above were operating at the time of the Warring States, a time which was unsettled and when several individual states were vying for power. Han Feizi 韩非子 (280–233 bce) was the founder of the legalist school. He was the author of the Han Feizi, a book aimed at strengthening and maintaining the power of the state and in which the overall philosophy is at complete odds with the Confucian tenet of the importance of ethical and moral behaviour and of rule by example. The key points of the Han Feizi have been summarised by Hu Shi, the leader of the language-reform movement of the late nineteenth and early twentieth centuries, to whom we shall return below. The key points (taken from Watson 1964) were: In governing a state, the wise ruler does not depend on the people’s becoming good for his sake, but on their necessity not to do evil. A wise man never expects to follow the ways of the ancients, nor does he set up any principle for all time. To be sure of anything without corroborating evidence is stupidity. To base one’s argument on anything which one cannot be sure is perjury. Therefore, those who openly base their argument on the authority of the sage-elders of antiquity, and who are dogmatically certain of the ages of Yao and Shun, are men either of stupidity or wilful perjury. (Hu 1923: 175–83) This basically accuses Confucius of being either stupid or a liar (if not both). The Han Feizi includes a chapter ‘On the Difficulties of Persuasion’〈说难〉and this clearly demonstrates the importance of adopting Gui Guzi’s yin approach to rhetoric and persuasion, as these quotes will illustrate. They also call to mind the importance of captatio benovolentiae, the securing of the goodwill of the audience as advised by Cicero. On the whole, the difficult thing about persuasion is to know the mind of the person one is trying to persuade and to be able to fit ones words to it. The important thing in persuasion is to learn how to play up the aspects that the person you are talking to is proud of, and play down the aspect he is ashamed of. The perils associated with attempting to persuade the all-powerful emperor are also acknowledged. Men who wish to present their remonstrances and expound their ideas must not fail to ascertain their ruler’s loves and hates before launching into speeches  .  .  .  if you gain the ruler’s love, your wisdom will be appreciated and you will enjoy favour as well. But, if he hates you, not only will your wisdom be rejected but you will be regarded as a criminal and thrust aside  .  .  .  The beast called the dragon can be tamed and trained to the point where you may ride on its back. But on the underside of its throat it has scales a foot in diameter that curl back from the body, anyone who chances to brush against them is sure to die. The ruler of men too has his bristling scales. Only if a speaker can avoid brushing against them will he have any hope of success. 330

Chinese rhetoric

It is perhaps not surprising that the unification of China with the establishment of the Qin Empire under the Emperor Qin Shi Huang saw the adoption of legalism as official policy. With such a powerful emperor, the importance of yin persuasion became even more important. To date we have argued that the origins of Chinese rhetoric are difficult to disentangle as rhetoric was not seen as a discrete discipline in Classical China. While arguing that the Chinese rhetorical tradition is not monolithic, we have identified two major schools, one, with a Confucian leaning that promulgated plain and simple speaking, and the other, which recognised the significance of audience upon rhetorical style. Above all, this school recognised the difficulties in persuading ‘up’ and advised a range of strategies for yin persuasion. I shall now turn to a brief discussion of what has often been referred to as China’s first account of rhetoric, the Wen Xin Diao Long《文心雕龙》(The Literary Mind and the Carving of Dragons) written by Liu Xie 刘勰 (ce 465–520). The Wen Xin Diao Long is probably more correctly classified as a book of literary criticism (D. Wang 1983: 50). By the time of its publication, the plain and simple Confucian style, known as guwen (古文), or classical language, had given way to a more florid and ornate style, known as pianwen (骈文) (Hightower 1965). This is often translated as ‘parallel prose’ (an early meaning of pian is of a carriage drawn by six horses, thus giving the idea of grandiose ornateness) (Kirkpatrick and Xu 2012: 37–8). Pianwen was a reaction to the traditional guwen style and represented the increasing ‘belletricisation of Chinese literary criticism and theory that paralleled the Buddhicisation of Chinese society’ (Mair 2001: 81). Instead of being examined on the Confucian Classics, candidates for the civil service were being judged on their skill in the use of pianwen style. This was a time of ‘art for art’s sake’ with some disadvantage to governance. ‘Ignoring the sages and favouring embellishment meant that the literary brush became more lush and the government more chaotic’ (Bol 1992: 91). The pianwen style occasioned ridicule among some guwen scholars, being described by one as a ‘boat of magnolia wood propelled by ostrich feather oars’ (S. Chen 1961: 6). The Wen Xin Diao Long was written at this time, and Liu Xie, while in favour of a return to the Confucian style, is not prescriptive, allowing the need for writers to adapt to the times, as this advice on the composition and organisation of texts demonstrates. The division into paragraphs and the construction of sentences conform to different tempos at different times. For these differences, there is no fixed rule, and one must adapt  .  .  .  to varying circumstances. (Shih 1959: 186) The balance between using the classics as models against the need for writers to be able to adapt to the times is further captured in the following verse. Against the background of the ten dynasties, literary trends have changed nine times Once initiated at the central pivot, the process of transformation circles endlessly Literary subject matter and the form in which it is treated are conditioned by the needs of the times. But whether a certain subject matter or a certain form is emphasised or overlooked depends on the choice made by the writers. Antiquity, however, remote, can be made to display itself before us like a human face. (Shih 1959: 224) 331

Andy Kirkpatrick

This tension between the use of the traditional guwen style with its respect for the classics and the use of the florid pianwen style is, along with the associated tension between a plain and simple style and a yin style of persuasion, a constant of the Chinese rhetorical tradition. It is thus part of the normal cyclical pattern that guwen became the favoured style during the Tang dynasty, with Han Yu 韩愈 (768–824) its best-known proponent. The simple ‘classical’ style of a later guwen proponent, Ouyang Xiu 欧阳修 (1007–1072), has been described in the following way: The works of Ouyang Xiu are lucid and fluent; his style is easy and unaffected. In his prose writings he showed his mastery by a continuous flow of thought and argumentation, with a significant content couched in clear and simple language. (Chai 1974: 46) While the Wen Xin Diao Long was obviously concerned with rhetoric and other texts also touched upon rhetoric, it is not until the Southern Song dynasty (1127–1279) that the book which many Chinese scholars refer to as China’s first systematic account of rhetoric was published (e.g. Liu 1988; Z. Zhou 1999). This was the Wenze《文则》by Chen Kui 陈葵. As its title, which I have translated as the Rules of Writing (Kirkpatrick 2005) indicates, its focus is on writing. Interestingly, its 1170 date makes it contemporaneous with the Ars Dictaminis treatises of Medieval Europe, where the focus was also on written texts at this time. Chen Kui’s aim was to summarise the rules and techniques of writing, using classical texts for his examples and source materials. Five main topics make up the book: genre, ‘negative’ rhetoric, ‘positive’ rhetoric, syntax and style (Liu 1988). Negative rhetoric deals with such aspects of rhetoric as text structure and argument sequencing. Positive rhetoric deals with rhetorical tropes. As a fervent advocate of the guwen style, Chen Kui identifies the general overriding principle that language should be simple, clear, succinct and contemporary (Kirkpatrick 2005: 115). Given that the publication of the Rules of Writing coincided with the Song dynasty’s decision to remove privilege and establish a meritocracy by ensuring that those chosen for the civil service passed the exams, it is tempting to propose that the Rules of Writing was aimed as a guide for those men now preparing for the civil service exams (Kirkpatrick and Xu 2012: 51). The advice given by the author to aspiring writers – especially those who need to master the art of writing composition for the civil service exam – seem both clearly linked to the guwen style, while, at the same time appearing remarkably modern, as some would not seem out of place in contemporary textbooks on English composition. These excerpts help give a flavour of this (Kirkpatrick and Xu 2012: 54ff.). To be good, things need to be simple and easy; to be appropriate, language needs to be simple and clear. If a musical performance is not harmonious, then music is unpleasant; if a text is not coherent, then it cannot be read  .  .  .  classical texts were natural and coherent and were without adornment and embellishment. At the same time, agreeing with Liu Xie, Chen Kui felt that writers must adapt to the times and write in ways that would be easily understood by their contemporaries. Although classical Texts used classical language, classical language cannot be fully understood by later generations, unless there are explanatory notes  .  .  .  The use of 332

Chinese rhetoric

language that was the common speech of one period will be found abstruse and difficult by people of later periods. Chen Kui also provided advice on the arrangement of arguments and again these seem strikingly familiar to students of Anglo-rhetoric. For example, he writes that there are three ways in which this can be done. Writers can first state the summary or overall point and then list the individual details. Or they can list individual details first and then summarise and explain. Finally, they can combine these two methods and provide the main point at the beginning, then list the individual details and conclude by repeating the overall point. What then of the civil service exams for which Chen Kui’s Rules of Writing may have providing guidance? The later history of these exams is closely linked to the history of an (in)famous Chinese text structure, the baguwen (八股文), known in English as the octopartite or eight-legged essay. The first civil service exams were held during the Sui dynasty (581–618) at a time when the pianwen style was in favour. The exam system was developed further during the Tang dynasty (618–907), although the exams provided a route to the civil service for only a small percentage of civil servants (6–16%) (Chafee 1985). It was at this period that Han Yu started to promote a return to the guwen style. The actual baguwen was introduced as the exam essay form in the Song dynasty by Wang Anshi 王安石 (1021–1086). From the beginning it was associated with orthodoxy and required taking the official line presented from above. This has led many modern scholars to view the baguwen with contempt, summarised in this quote from Qi et al. (1994: 1): The baguwen has been called stale and rotten, cliché-ridden, rigid and well past its use by date. It is despised and rejected and those who are against it have given it the epitaph of being the essence of evil. Nevertheless, the baguwen lasted several hundred years, up until the civil service exams were themselves abolished in 1905. Its form remained remarkably stable, with the only major alteration being the gradual increase of the length of a baguwen from 550 to 700 characters. Zhu identifies three key features of the baguwen (1990: 472ff.). The first two concern content: they had to be based on the Confucian canon and they had to take the neo-Confucian ‘Cheng-Zhu’ school as orthodox. The third feature was that they had to follow a regulated format, outlined below (see also Kirkpatrick and Xu 2012: 78ff.). 1. Poti (破题) (Opening the topic). Here two sentences were required to introduce the topic. 2. Chengti (承题) (Carrying the topic forward). This section provided further information about the topic and could contain three or four, or four or five sentences. 3. Qijiang (起讲) (Elaborating). A more profound discussion about the topic was provided here. The length of this section might vary considerably, from ‘a few sentences’ to ‘more than ten’. 4. Ruti (入题) (Revealing the topic) (Lingti 领题, Tiju 题举 or Rushou 入手). This section used either 1 or 2, or 4 or 5 sentences, and its function was to clarify ideas of an essay topic that was of some length. For example, an essay topic could be a substantial extract from one of the Confucian classics. Thus, this section was optional. In the example baguwen provided below, this section actually occurs after section five, the first of the parallel legs. 333

Andy Kirkpatrick

These first four sections, along with the conclusion, were written in a relatively free prose style. After these opening four sections, there followed the parallel legs from which the eight-legged essay derives its name. Unlike the first four sections and the concluding section, each of these sections required at least two sentences and they had to provide stylistic balance. The required style has been described ‘as one falls another one rises’ (Tang 1980: 27). It should be noted, however, that the form varied in some ways. First, as noted above, the Ruti section was optional. Second, while four sections of parallel legs are described here, the final parallel leg, the Shugu, was also optional. This meant, of course, that an essay that omitted the final Shugu might have only six legs. The third point of note, however, is that each parallel ‘leg’ might have more than two legs. Some baguwen had as many as 20 legs. The four customary parallel legs were: 5. Qigu (起股) (Opening legs) (Qibi 起比, Tibi 题比, Qiangu 前股, Tigu 题股) 6. Zhonggu (中股) (Middle Legs) (Zhongbi 中比) 7. Hougu (后股) (Latter legs) (Houbi 后比) 8. Shugu (束股) (Concluding legs) (Shubi 束比) After these parallel legs, the baguwen ended with a final section, the conclusion: 9. Dajie (大结) (Luoxia 落下) The baguwen is a fundamental part of the Chinese rhetorical tradition. It has been called the ‘bloodchild’ of guwen and pianwen and the ‘crystalisation’ of Chinese literature (Tang 1980: 28). Wu (1988: 217) sees it as not only the amalgam of guwen and pianwen, but also of the four-part structure of qi-cheng-zhuan-he (起承转合), to which I return below. Yet it is little known in mainland China, a fact mourned by the famous architect of Chinese language reform, Zhou Youguang 周有光, who, in an article written expressly to demonstrate its form for those who were unfamiliar with it (2004), adapted the form to write a piece critical of President Jiang Zemin’s regime at the time, thereby showing that the form could be used to criticise government as well as serve it. While the baguwen was traditionally a state-prescribed rhetorical form, the companion four-part qi-cheng-zhuan-he structure was adapted by people who wanted to voice indirect criticism of the state. The Dictionary of Chinese Grammar and Rhetoric (DCR) defines the rhetorical form of the structure as follows: A common logical belle-lettres structure and sequence but also the epitome of a common structural pattern for a variety of texts, both ancient and modern. Qi is the opening or beginning, cheng continues or links the opening to the next stage. Zhuan is the transition or turning point, he is the summary or conclusion. (Hanyu Yufa Xiuci Cidian 1985: 314) Its application across a variety of texts is evidenced by the fact that the DCR gives a famous four-line Tang dynasty poem by Li Bai 李白 as an example of the form. By the end of the Yuan dynasty (1368), it had been adopted as a rhetorical structure for the writing of prose (Wu 1988: 204) and lent itself to the art of ‘criticism by indirection’ (Kao 1985: 121) or yin persuasion. A well-known and comparatively modern example of the qi-cheng-zhuan-he structure providing the form for a piece of indirect criticism was written by the exiled scholar 334

Chinese rhetoric

Gong Zizhen 龚自珍 in 1839, shortly after he had been dismissed from his post. The piece, A Sanitarium for Sick Plumtrees, comprises four paragraphs. There is not space to include the entire piece here, but following Di’s analysis (1984), the first paragraph (qi) introduces the text. It reads: Longpan Mountain in Jiangning, Dengwei Mountain in Suzhou and the Western banks of West Lake in Hangzhou all have an abundance of plum trees. The second paragraph (cheng) describes the underhand schemes of scholars and artists in their suppression of the growth of these plum trees. (The scholars and artists represent the reactionary Qing dynasty ruling classes and the plum trees represent men of talent, of which the author no doubt saw himself as one). The third paragraph (zhuan) described how the author attempted to care for the sick plum trees and the fourth paragraph (he) concludes the piece. Ai! How I wish I had the free time and the idle land so that I could gather in the sick plum trees of Jiangning, Hangzhou and Suzhou, and within my lifetime cure them! To summarise to date, I have shown how the tensions between two literary styles, guwen and pianwen played out over centuries. Guwen was the classical orthodox style that favoured plain and simple writing and pianwen was the more ornate florid style, more concerned with art for art’s sake. I have also traced the origins and development of two iconic rhetorical writing styles, the baguwen and the qi-cheng-zhuan-he, showing how the former was prescribed by the state and the latter was used against the state to express indirect criticism using yin persuasion. Yet, the baguwen can be adapted to express yin persuasion, as exemplified by Zhou Youguang. I now turn to look at the establishment of rhetoric as a discrete discipline in China and then conclude with a review of the state of Chinese rhetoric today. As noted above the term currently used for rhetoric, xiuci (修辞), literally means something like ‘refining words’ and has its first mention in the Confucian Classic The Book of Changes. The term to mean rhetoric as an independent discipline is much more recent. Only at the collapse of the Qing empire in 1911 and the onset of the various reform movements did rhetoric as a discipline became established. This included radical language reform, as it was considered that the complexity of the traditional Chinese wenyan (文言) literary style was a major factor in the low levels of literacy across China and that this, in turn, was a major cause of China’s lack of progress towards modernisation. That Japan had modernised its language and developed syllabaries only added to the impetus for China to do the same. The first two decades of the twentieth century saw thousands of Chinese studying in Japan and many studying in the West. These included one of the major architects of language reform, Hu Shi 胡适. Hu had studied at Cornell and Columbia universities in the USA and, as we saw above, summarised the key points of Han Feizi’s philosophy. Below are the eight guidelines Hu (1917) penned promoting the vernacular baihua (白话) to replace wenyan as the medium of educated discourse. Many of these echo the advice given by Chen Kui several centuries earlier. 1. 2. 3. 4.

Language must have content. Do not (slavishly) imitate classical writers. Make sure you pay attention to grammar and structure. Do not complain if you are not ill – in other words, don’t overdo the emotion. 335

Andy Kirkpatrick

5. 6. 7. 8.

Cut out the use of hackneyed clichés. Don’t cite or rely on the classics. Don’t use parallelism. Embrace popular and vernacular language.

Chinese scholars also started to translate Western works in large numbers and write their own books about this ‘new’ rhetoric, heavily influenced by Western rhetoric. This gave added impetus to the establishment of rhetoric, xiuci, as a discrete discipline Tang Yue’s 唐钺 Rhetorical Style《修辞格》was published in 1923 and introduced AngloAmerican rhetoric to China. Chen Wangdao (陈望道), China’s most revered scholar of rhetoric, called it China’s first scientific book on rhetoric (Hui 2009). Chen, who had studied overseas in Japan, himself published the most influential modern text on Modern Chinese rhetoric, Xiucixue Fafan《修辞学发凡》, in 1932 (W. Chen 1932/1988). This remains the most influential book of its type, both because it managed to combine key concepts of Western and traditional Chinese rhetoric, and also because its author later became Mao’s appointee in 1952 as president of the prestigious Fudan University in Shanghai (Hui 2009), a position he held for some 25 years. Fudan remains a leading site for the study of rhetoric. In his book, Chen advises writers who want readers (and listeners) to be persuaded to alter their positions after reading what they have written to observe seven conditions (W. Chen 1988: 130). These seven conditions are: 1. 2. 3. 4. 5. 6. 7.

Do not use too much abstract language. Be tactful, mild and indirect Be serious, but not overly so. Do not over-elaborate. Make sure your choice of language suits the readers. Avoid monotony, use variety. Use a light (qing) to heavy (zhong) sequence.

These ‘conditions’ recall Hu Shi’s eight guidelines listed above and also echo Chen Kui’s advice. In addition, it is not hard to see the influence of Gui Guzi’s advice about yin persuasion here, especially in the condition that reminds writers to use language that suits the readers. This is also reflected in the condition advising writers to move in a ‘light’ to ‘heavy’ sequence. What this means is that writers should adopt a more inductive approach when their aim is to persuade by leading their readers to the main point, rather than presenting it at the outset. I would argue on the strength of this brief survey of Chinese rhetoric, that the Chinese rhetorical tradition, while characterised by cyclical shifts in favoured styles and influenced by ideas from outside China, nevertheless remained virtually unbroken to this time. The establishment of the People’s Republic of China in 1949 has, in stark contrast, led to a severing of ties with the rhetorical tradition in ways that will be briefly exemplified below. The first example is of a typically constructed sentence written by Mao with the translation following the ‘logical cause–effect’ sequence of the original Chinese. The sentence is taken from Chen Zhenqiu (1980: 120ff ). The lower middle peasants, because in economic status they are more or less similar to the lower middle peasants among the new middle peasants, as they were not well 336

Chinese rhetoric

off to start with and the interests of some were improperly encroached upon at the time of the agrarian reform, are generally interested in joining the cooperatives. Given this rather ponderous style, it is not surprising that Mao felt the need to publish his thoughts in a more user-friendly way in the Little Red Book. This Mao extract follows the traditional cause–effect order of Chinese, but his political writing was heavily influenced by translations of European philosophers, particularly, Marx and Engels. European influence on Chinese writing and rhetoric was significant. The wellknown Chinese linguist Wang Li 王力 has pointed out how Chinese has become Europeanised, caused by the many translations of Western works into Chinese (1955: 281). Traditionally, Chinese favours a modifier–modified principle of sequencing along with a principle of temporal sequence whereby events are presented in the order in which they occur (Tai 1985). Thus cause always precedes effect, to the extent that no explicit connectors are needed to signal the relationship between two arguments. The two propositions, ‘He fell over; he hurt his ankle’, must mean in Chinese ‘Because he fell over, he hurt his ankle’. The cause precedes the effect. One result of the influence of the translations was the appearance of reverse ordering of clauses, along with the use of explicit connectors, as these become essential in indicating the relationship between two propositions when ‘logical’ or chronological order is not followed. As a result, sentences like ‘He hurt his ankle because he fell over’ became increasingly common in Chinese (Xie 1989). A second consequence was the increased length of the sentences, but with the traditional sequence of modifier–modified segments preceding the main point. It is likely that this influence from the Western, particularly the Anglo-American, rhetorical tradition, coupled with increased dissension within China, both among members of the Communist Party themselves and between the Party and those who were opposed to it has led to an increasingly confrontational style, far removed from the traditionally ‘irenic’ style discussed earlier. As an example, here is a translation of Mao’s thoughts of General Peng Dehuai 彭德怀, who in earlier times had been a highly trusted colleague, but who had fallen into disfavour: I am hated by many, especially comrade Peng Dehuai, his hatred is so intense that he wished me dead. My policy with Peng Dehuai is such: you don’t touch me, I don’t touch you; you touch me, I touch you. Even though we were once like brothers, it doesn’t change a thing. This was recorded at the Lu Shan meetings of 1959, after which Mao ‘touched’ General Peng by labelling him the leader of an anti-Party clique and stripping him of all his positions. This confrontational style reached its apotheosis in the Cultural Revolution, but is still in evidence. In her study of the rhetoric of the cultural revolution, Lu quotes a person she interviewed: The language used to attack Falun Gong is exactly the same language as that used to attack ‘cow ghosts and snake spirits’ during the Cultural Revolution. (2004: 196) The confrontational style has been adopted by some dissidents. Lu quotes another of her interviewees: 337

Andy Kirkpatrick

There is definitely a trace of the cultural-revolutionary style, even in the writings of political dissidents  .  .  .  The language they use to attack the CCP is very similar to the Red Guard style. They use Mao’s style of verbal aggression to condemn Mao. (2004: 196) The confrontational style of ‘using Mao’s style of verbal aggression to condemn Mao’ is also in evidence in Charter 08, the document prepared and signed by a group of 303 authors, including the Nobel laureate Liu Xiaobo (Kirkpatrick and Xu 2012: 170ff.). Charter 08 sets forth 19 specific demands, but follows a rhetorical style suited to yang or top-down persuasion rather than yin or bottom-up persuasion. The tone of the Charter is one of command, marked by the use of hyperbole and metaphor typical of powerful discourse. Features typical of powerless discourse such as mitigation where, ‘We do tend to leave implicit all propositions that we believe to be known or deliverable by the recipients’ (Van Dijk 2008: 184–5), are omitted. The use of pronouns also reflects its confrontational style. The use of ‘we’ is confined to the authors who are associated with civic responsibility. ‘It’, on the other hand is used for the intended audience, the Chinese Communist Party. This excerpt from the Charter’s Preamble exemplifies these stylistic tropes.1 The ‘New China’ established in 1949 is a ‘people’s republic’ in name only. In fact, it is under the party’s dominion. The ruling power monopolizes all the political, economic and social resources. It created a string of human rights catastrophes such as the Anti-Rightist Campaign, the Great Leap Forward, the Cultural Revolution, June 4, and attacks on non-governmental religious activities and on the rights defense movement, causing tens of millions of deaths, and exacted a disastrous price on the people and the country. A frequent use of modals of obligation also heightens the yang or command tone of the document. To ensure human rights must be the foundation of the first objective of government and lawful public authority  .  .  .  The principle of equality before the law and citizen’s society must be implemented; the principle of equality of economic, cultural and political rights must be implemented. This confrontational style has become the default rhetorical style across much political and public discourse in mainland China today. But as the French sinologist Jullien has pointedly asked: ‘In the name of what, therefore, can the Chinese man of letters break free from the forces of power, affirm his positions, and thus speak openly?’ This is in response to his own contention that the traditional promotion of ‘obliquity’, or what we have termed yin persuasion, renders dissidence impossible (2004: 137). However, one can’t help but wonder whether a return to the Chinese rhetorical tradition and the adaptation of earlier techniques of indirect criticism and yin persuasion might not help create an environment in which constructive public engagement and criticism could develop. Perhaps Zhou has shown a possible way forward with his adoption of the baguwen to level criticism at the government. Traditional Chinese rhetoric may well have an important role to play in indicating that a yin style might well be more persuasive than a confrontational yang style. But for traditional Chinese rhetoric to make this contribution, it will need to be taught at universities. Sadly, this is not currently the case. A recent review of Chinese composition 338

Chinese rhetoric

textbooks currently being used across universities in mainland China showed that the major focus of such books was on the writing of yingyong (应用) or practical genres, such as administrative documents, public relations documents and finance documents (Kirkpatrick and Xu 2012: 200ff.). This focus suggests that university writing courses are designed with the needs of the contemporary bureaucracy in mind rather than the need for writers to be able to produce persuasive texts or discourses based on the Chinese rhetorical tradition. The situation appears even bleaker when it is learned that the only students in Chinese universities who study any form of Chinese rhetoric and composition are Chinese majors, who represent a tiny minority of the total number of students at Chinese universities. Ironically, a compulsory subject for all students is English composition, with the consequence that Chinese university students now receive far more instruction in the writing of English essays than they do in writing Chinese. This renders even less likely the re-emergence of the Chinese rhetorical tradition. In concluding this chapter, I cite a plea taken from the 2012 annual letter written by the mothers of those who died on June 4, 1989. The overall tone of this letter is also confrontational and yang, but it does include the following passage: If we are able to use dialogue to replace confrontation on the problem of ‘June Fourth’, it would benefit the whole country and be a blessing for all our people. The more dialogue we have, the more civility and law and order, and the less ignorance and tyranny. Dialogue does not lead society towards opposition and hatred, but rather towards tolerance and reconciliation. For there to be a chance of this, courses of rhetoric, including traditional Chinese rhetoric with its emphasis on irenic and yin persuasion, will need to be re-introduced to schools and universities across China. Until such time, it seems inevitable that a confrontational and highly agonistic yang style will remain, making dialogue, tolerance and reconciliation impossible. The re-discovery of Chinese rhetoric as practised for centuries and then formalised as a discipline in the early twentieth century is crucial if China is to develop civic discourse which will allow constructive criticism and debate.

Note 1 Translation by Human Rights in China (http://www.hrchina.org/public/index). The Chinese text can be seen at http://www.2008xianzhang.info/chinese.htm.

Bibliography Bol, Peter K. (1992) This Culture of Ours: Intellectual Transitions in T’ang and Song China, Stanford: Stanford University Press. Chaffee, John (1985) The Thorny Gates of Learning in Song China: A Social History of Examinations, Cambridge: Cambridge University Press. Chai, Winberg (ed) (1974) A Treasury of Chinese Literature, New York: Thomas Y. Crowell. Chen, Shou-yi (1961) Chinese Literature: A Historical Introduction, New York: Ronald Press. Chen, Wangdao 陈望道 (1932/1988)《修辞学发凡》(An Introduction to Rhetoric), Hong Kong: Guangming Publishing. Chen, Zhenqiu 陈真求 (1980)《翻译问题探索》(On the Problems of Translation), Beijing: The Commercial Press. Di, Chen (1984)〈篇章修辞初探〉(‘A Preliminary Discussion on the Rhetoric of Text’) in Wang Xi Jie (ed)《修辞学论集》(Collected Works on Rhetoric), Fuzhou: Fujian People’s Publishing, 289–306.

339

Andy Kirkpatrick Durrant, Steven (2002) ‘Creating Tradition: Sima Qian Agonistes’, in Stephen Shankman and Steven Durrant (eds) Early China Ancient Greece: Thinking Through Comparisons, New York: State University of New York Press, 283–311. Forke, Alfred (1901) ‘The Chinese Sophists’, Journal of the Royal Asiatic Society (North China Branch) XXXIV: 1–100. Garrett, Mary (1983) ‘The “Mo-Tzu” and “Lu-shih Ch’un-Ch’iu”: A Case Study in Classical Chinese Theory and Practice of Argument’, Doctoral Dissertation, University of California, Berkeley. Graham, Angus (1989) The Disputers of the Dao, La Salle, IL: Open Court. Hanyu Yufa Xiuci Cidian (1985)《汉语语法修辞词典》(A Dictionary of Chinese Grammar and Rhetoric), Jinan: Anhui Education Publishing. Harbsmeier, Christopher and Christoph Harbsmeier (1999) ‘Chinese Rhetoric’, T’oung Pao LXXXV: 114–26. Hightower, James R. (1965) Topics in Chinese Literature, Cambridge MA: Harvard University Press. Hu, Shi 胡适 (1917)〈文学改良初议〉(‘An Initial Discussion on Literary Innovation’) in《胡适文存》 (A Collection of Hu Shi’s Works), vol. 1, Taipei: Yuandong Tushu Gongsi, 5–16. Hu, Shi (1923) The Development of the Logical Method in Ancient China, Shanghai: Oriental Books. Hui, Wu (2009) ‘Lost and Found in Translation: Modern Conceptualization of Chinese Rhetoric’, Rhetoric Review 28(2): 148–66. Jullien, François (2004) Detour and Access, New York: Zone Books. Kao, Karl S. Y. (1985) ‘Chinese Rhetoric’ in William H. Nienhauser (ed), The Indiana Companion to Traditional Chinese Literature, Bloomington: Indiana University Press, 121–37. Kennedy, George (1980) Classical Rhetoric and its Christina and Secular Tradition, London: Croom Helm. Kirkpatrick, Andy (2005) ‘China’s First Systematic Account of Rhetoric: An Introduction to Chen Kui’s Wenze’, Rhetorica 23(2): 103–52. Kirkpatrick, Andy and Xu Zhichang (2012) Chinese Rhetoric and Writing: An Introduction for Language Teachers, Andersen SC: Parlor Press. Liu, Yancheng 刘彦成 (1988)《文则注译》(Notes and Translation of the Wenze), Beijing: Shumu Wenxian. Lu, Xing (1998) Rhetoric in Ancient China: Fifth to Third Century bce: A Comparison With Classical Greek Rhetoric, Columbia: University of South Carolina Press. Lu, Xing (2004) Rhetoric of the Chinese Cultural Revolution, Columbia: University of South Carolina Press. Mair, Victor (2001) ‘Buddhism in the Literary Mind and Ornate Rhetoric’, in Cai Zongqi (ed) A Chinese Literary Mind: Culture, Creativity and Rhetoric in the Wen Xin Diao Long, Stanford: Stanford University Press, 63–81. Pu, Kan 濮侃 and Wei Qun 蔚群 (1983)〈试论我国古代修辞研究的特点〉(‘A Study of the Characteristics of Classical Chinese Rhetoric’),《修辞研究》(Studies in Rhetoric) 2: 111–24. Qi, Gong 启功, Zhang Zhongxing 张中行, and Jin Kemu 金克木 (1994)《说八股》(Talking About the Bagu), Beijing: Zhongguo Shuju. Shankman, Stephen and Steven Durrant (eds) (2002) Early China Ancient Greece: Thinking Through Comparisons, New York: State University of New York Press. Shih, Vincent (1959) The Literary Mind and the Carving of Dragons, New York: Columbia University Press. Tai, James H-Y (1985) ‘Temporal Sequence and Word order in Chinese’, in John Haiman (ed) Iconicity in Syntax, Amsterdam and Philadelphia: John Benjamins, 49–72. Tang, Tao 唐弢 (1980)《文章修养》(Developing Texts), Hong Kong: Xianggang Wenxue Yanjiu Press. Tsao, Ding-ren (1985) ‘The Persuasion of Kuei Ku Tzu’, Doctoral Dissertation, University of Minnesota. van Dijk, Teun A. (2008) Discourse and Context. A Sociocognitive Approach. Cambridge: Cambridge University Press. Wang, Dechun 王德春 (1983)《修辞学探索》(An Exploration of Rhetoric), Beijing: Beijing University Press. Wang, Li 王力 (1955)《中国语法理论》(A Theory of Chinese Grammar), Beijing: Zhonghua Shuju. Watson, Burton (1964) Han Fei Tzu: Basic Writings, New York: Columbia University Press. Wu, Yingtian 吴应天 (1988)《文章结构学》(The Construction of Texts), Beijing: People’s University Press.

340

Chinese rhetoric Xie, Yaoji 谢耀基 (1989)《现代汉语欧化语法概论》(An Outline of Western Grammar in Contemporary Chinese), Hong Kong: Guangming Publishing. Zheng, Ziyu 郑子瑜 (1979)《中国修辞史稿》(A History of Chinese Rhetoric), Shanghai: Shanghai Education Press. Zhou, Youguang 周油光 (2004)〈八股文的默〉(‘A Description of the Baguwen’) , in《写修辞学习》 (Studies in Rhetoric) 2: 1–2. Zhou, Zhenfu 周振甫 (1999)《中国修辞史》(A History of Chinese Rhetoric), Beijing: Shang Wuyin Shuguan. Zhu, Binjie 禇斌杰 (1990)《中国古代问题概论》(An Introduction to Classical Chinese Genres), Beijing: Peking University Press. Zong, Tinghu 宗廷虎 and Li Jinling 李金苓 (1998)《中国修辞学通史隋唐五代宋金元卷》(A Complete History of Chinese Rhetoric, Volume 2: The Sui, Tang, Five Dynasties, Song, Jin, and Yuan Dynasties), Jilin: Jilin Education Press.

341

19 Chinese slang Robert L. Moore Rollins College, USA

Introduction Slang is a linguistic category that has long defied those who have sought to define it (Dumas and Lighter 1978; Adams 2009). The qualities that make it so difficult to define apply to both its English and Chinese versions.1 Part of the difficulty in identifying specific defining criteria stems from slang’s status as a folk category. Slang is, in a sense, what the speakers of a given language believe it to be, and different speakers have different understandings about what it is. The qualities most commonly attributed to it by those who define it are as follows: 1. 2. 3. 4. 5. 6.

Informality Playful or imaginative use of standard vocabulary items Humor Ephemerality Appropriateness in egalitarian relationships Association with contexts wherein the dignity entailed in high-status positions is not deferred to 7. Rebelliousness 8. Association with marginal and/or relatively powerless populations (such as youth, military personnel or criminal gangs) 9. Vulgarity For present purposes this somewhat loose collection of attributes will serve as a guideline and a beginning point for an approach to slang. Not all of these qualities can be considered equally salient, however. Informality is a necessary attribute (though not a sufficient one) of any expression that might be considered slang. Other features listed here, humor and vulgarity, for example, may characterize a given slang expression, but they are not necessarily associated with slang in all of its usages and contexts. Furthermore, these attributes can be clustered into related groups, each of which reflects a different aspect of slang. For example, those listed as numbers (2) through (4) can be seen as derived from slang’s typically playful quality, playfulness here encompassing both a nonserious attitude and the willingness of an inventive speaker to work his or her will on the 342

Chinese slang

words and phrases of ordinary discourse. The playful or inventive redesignation of standard vocabulary as slang is one of the factors that makes much slang ephemeral or fleeting. What is currently a clever linguistic invention can become a hackneyed phrase with the passage of time. But, as will be discussed below, the ephemerality of slang is not a constant and universal attribute. The features listed as numbers (5) through (9) refer to status relationships. Slang is most appropriately used as a register among people of roughly equal status and, conversely, is inappropriate where deference to authority figures and others of high status is demanded. Though people of differing status may use slang, such usage tends to minimize their status differences. Following from this is the common use of slang among marginal groups, those lacking power and/or status. Such groups are typically not committed to social hierarchies in the way that those who lead and organize such hierarchies are. Furthermore, these marginalized groups are often regarded as crude and vulgar by those of high status, the latter often using their refined speech and manners both as symbols of their status and devices for justifying their position. When a student uses slang in speaking to a teacher, or a defendant does so in addressing a judge, the effect is often for the speaker to appear both rebellious and crude, i.e. lacking in the deferential attitude and polished manners that more formal language entails.

Slang, informality and play The importance of playfulness and humor in slang is linked to its role in the promotion of social connections. Eble (1996) has noted the important part that slang plays in promoting sociability in her study of slang usage among US university students. Slang encourages speakers to display their sentiments in an egalitarian context where playfulness or humor may help promote social rapport. Humor and play often serve to diminish social barriers, and the wordplay and humor so common in slang expressions can be partly explained by the usefulness of these constructs in promoting sociability. Even when slang is not specifically humorous or particularly playful, it still entails a non-serious quality that distinguishes it from most standard language. Another widely recognized function of slang is the marking of social boundaries as when specific slang terms are used by in-group members who, unlike non-members, are familiar with them. This boundary-marking function is also sometimes characteristic of humor – particularly, though not exclusively, in the form of satire. Slang, then, like humor, can be said to both promote sociability among group members, particularly in egalitarian contexts, and to mark the boundary between in-group members and everyone else. Not every expression of slang is humorous or playful, but most can be said to conjure up an ethos that at least suggests these qualities by virtue of their hyper-casual aspect. Slang can also be used aggressively, of course, but aggression is not a central function of slang, as it is, for example, of swearwords (Moore 2012). The distinction between slang and swearwords, and their different prototypical functions, will be discussed further below.

Liyu In Mandarin, the term most commonly glossed as ‘slang’ is lǐyǔ (俚语), with the element lǐ (俚) meaning unpolished or crude, and yǔ (语) meaning language (Xu 1963). The Cihai, or the Encyclopedic Dictionary of Chinese, defines lǐyǔ as ‘a type of colloquial speech, often with dialectical characteristics’. This definition is followed by an eleventh-century citation in which a famous general is described as ‘illiterate and prone to the use of lǐyǔ’. 343

Robert L. Moore

Lǐyǔ, however, is not merely unpolished or crude speech; it is best regarded as a subcategory of informal speech. Precisely what the defining attributes of that subcategory are is a matter about which different Chinese speakers have different opinions (Moore et al. 2010). A popular volume found in Chinese bookstores goes by the English title New Slang of China, and is further identified on the cover with the Chinese characters《最新中国 俚语》(Zuixin Zhongguo Liyu). What this volume contains is not merely slang, but a number of expressions that might best be called common phrases (some of which can indeed be thought of as lǐyǔ proper) and common swearwords. The loose collection of expressions in a volume that purports to be strictly about slang bespeaks the difficulty of defining this category. Though lǐyǔ is widely used as a gloss for slang, the history and current status of lǐyǔ does not precisely match that of the corresponding English term. The most significant difference is that lǐyǔ has long been understood as describing informal expressions strongly associated with specific regions. These expressions are embedded in local dialects and often thought to be long-lived rather than ephemeral. Lǐyǔ, so conceived, is not a perfect match with the fly-by-night words and phrases that most English speakers regard as true slang. But Chinese lǐyǔ and English slang do match closely in a number of ways. Both serve to promote an egalitarian ethos by virtue of their informal, playful or humorous quality, both are largely oral, and both are inappropriate in formal contexts. For example, one is not likely to hear many lǐyǔ lexemes in the speech of CCTV news anchors, or to see them in school textbooks (Moore et al. 2010). Chinese discourse is distinctive, and different from that of many Western traditions, in its use of various kinds of common or traditional sayings. Among these various types are, for example, four-character sayings, often taken from classical texts or traditional stories, that are generally referred to as idioms or adages (chéngyǔ 成语). These are quite different from lǐyǔ phrases. However, another category of popular sayings, often referred to as súyǔ (俗语) is sometimes compared to lǐyǔ or even thought to be a version of lǐyǔ. Given the difficulty in drawing sharp boundaries around the concept of lǐyǔ, I won’t attempt to distinguish these two categories here on the basis of any absolute criteria, except to say that many súyǔ, in light of the respect they are sometimes given, and their widespread use in relatively formal contexts, do not qualify as lǐyǔ or slang.

Regional liyu The traditional idea that lǐyǔ is a kind of informal language pertaining to specific localities parallels Chinese linguistic variation in general. Chinese is famous for comprising not one, but a multitude of speech forms, all of them ultimately linked to the written characters whose ancestry is rooted in Shang dynasty artifacts of the second millennium bce. This collection of speech forms, which includes Mandarin, Cantonese, Shanghainese and others, comprises ‘Chinese’ in the broadest sense of the word. Furthermore, each of these major regional varieties or fāngyán (方言) of Chinese can be subdivided into minor dialects some of which are associated with speech communities so localized as to be centered on a single market town and its neighboring villages. An outstanding feature of Chinese, then, is the extraordinary array of speech forms that characterize it. Traditional understandings of lǐyǔ focus on this variety as a key defining feature. Standard Mandarin Chinese is distinct from this diverse collection of local dialects and is formally regarded as enjoying superior prestige. Since it was the only linguistic avenue to advancement for the scholar official class of the Qing dynasty and is currently the official 344

Chinese slang

language of both the People’s Republic of China and the Republic of China, Mandarin continues to be singularly important on both mainland China and Taiwan. However, local pride is not without significance, and local dialects, including their lǐyǔ expressions, do engender a kind of local patriotism for those who speak and understand them. A Cantonese speaker from, say, Guangzhou, will take a certain pride in Cantonese generally and perhaps in locally understood versions of lǐyǔ in particular. As further discussion will show, local and regional expressions are not the only form of slang typical of China. However, these have loomed large in traditional Chinese concepts of slang. An example of lǐyǔ from Beijing is gèsè (各色). In standard Mandarin this term refers to an array or an assortment, but in Beijing slang it metaphorically describes someone with an odd personality, particularly someone who is unappealing or objectionable. This illustrates a typical way of constructing a slang expression, i.e. by putting standard vocabulary to new uses. Another slang expression illustrating the same process originated in Beijing but is now more widely known: hùnqiúr (混球儿). In standard Mandarin the first character means ‘mixed up’ or ‘muddled’, while the second one means ‘ball’. Its slang meaning is ‘rascal’ or ‘bad boy’. The third element marks the retroflex –er (儿) ending typical of many Beijing Mandarin words and this ending gives this expression a distinctively ‘Beijing-style’ quality. Also featuring the Beijing –er is the phrase, dàlǎo yémen’er (大老爷们儿) which is slang for a ‘big shot’. Another expression typical of Beijing and the northeast is chědàn (扯淡) which means ‘nonsense’ or ‘bullshit’. It is typically applied to people who talk big but who don’t back up their talk with action. As these words and phrases suggest, slang is often descriptive of people or behaviors that are emotionally charged, a quality that is related to slang’s basic informality. For entities that are likely to conjure affective reactions in a speaker, slang can provide the vocabulary that will do these feelings justice. The inherent informality of slang helps convey those sentiments that more formal linguistic forms are designed to obscure. The local slang of Beijing is perhaps better known than that of any other region due to the city’s role as the home of the central government and as the historical center for modern vernacular Chinese. Furthermore, the standard Mandarin that is China’s national language is based on Beijing Mandarin. Other locales have their own local slang, of course, but these are less likely to catch on at the national level than is the slang of Beijing. A nationally known slang term is èr (二), which means ‘two’ in Mandarin, but which as a slang term means ‘stupid’. In Hangzhou’s local slang, however, this term is sometimes modified to liù èr (六二) or ‘six two’. The Cantonese-speaking south of China has its own slang both because its dominant language is not Mandarin and because the long colonial history of Hong Kong has nurtured a local culture distinctly different from that of the People’s Republic of China which has influenced the surrounding region (Bolton and Hutton 1997; Wright and Kelly-Holmes 1997). Slang in Cantonese is generally referred to as juhkyúh (俗语, the Cantonese term corresponding to Mandarin súyǔ) rather than léihyúh (the Cantonese term corresponding to lǐyǔ). A popular Cantonese expression meaning ‘crazy’ that has been in use for several decades is chìsin (痴线). The literal meaning of this phrase is something like ‘crazy wiring’ and it implies that the individual described has had the equivalent of a short circuit in the brain. As with Mandarin terms, Cantonese chìsin gains some of its lively expressiveness from a somewhat humorous image suggested by the comparison between a malfunctioning brain and a shorted-out appliance. An example of figurative Cantonese slang based on a standard phrase is hōichāan, which literally means ‘to dine’ or ‘to have a meal’, but has recently come to mean ‘to take drugs’. 345

Robert L. Moore

Similarly, sāanjaaih (山寨), which means ‘mountain stronghold’ in ordinary Cantonese, is a slang term for a cheap imitation of a quality product, often a pirated copy. The widely shared idea underlying this term is that mountain-dwelling people are poor and therefore forced to make do with inferior homemade goods. They are also thought to sometimes operate beyond the reach of the law. This expression is now widely used in both Cantonese and Mandarin (shānzhài) and it isn’t clear in which dialect it originated. This borrowing (either from Cantonese to Mandarin or Mandarin to Cantonese) across not only regional but significant dialectical borders, serves as a reminder that the linguistic boundaries between localities, though real, are often porous. One form of playfulness is the inserting of foreign words into Chinese discourse; slang terms are sometimes formed on the basis of non-Chinese lexemes. A Cantonese example of this is A A jai (A A制), where ‘A A’ is from the Latin alphabet and is pronounced as it would be in English. A A jai describes the custom that in English is known as ‘going Dutch’, i.e. where each individual pays for his or her own meal. This custom has become fairly common in China, though there was a time when it was rare among the Chinese and carried an antisocial connotation. Since gathering together for a meal has long been (and continues to be) an important social activity, and one that may even be required for sealing a business deal, the dividing of the check so that each individual could each pay for his or her own portion was contrary to the spirit of togetherness that diners traditionally sought. Of course, when one individual pays for the entire meal in the standard Chinese manner, it is common practice for the diners to fight over the check, sometimes with such heated language as to create the impression that a real fight was threatening to break out any minute. In any case, it was assumed in the past that one individual would pay for the entire meal and the Westerninfluenced pattern of A A jai (or ‘one-by-one style’) marks a new approach to dining in China, one that has given rise to a relatively new slang expression. Yāt gauh séui (一嚿水) or ‘one piece of water’ refers to HK$100, water being symbolic of money generally. Gáauchēut yàhn mehng (搞出人命) means ‘to kill someone’, but as a slang expression it means ‘to get a woman pregnant’, or, in English slang terms, ‘to knock her up’. Also, in the realm of love and sex, is galēi gāi (咖哩鸡, literally, ‘chicken curry’), a slang term for ‘hickey’ or ‘love bite’. These Cantonese slang expressions again point to the emotionally charged nature of those topics that attract slang – mental disabilities, shoddy merchandise, money, and sex, for example.

Slang vs. swearwords In popular culture, slang is often thought to include swearwords. These categories are, however, distinguishable, though they do overlap in function. In Chinese, swearwords are designated as zānghuà (脏话), that is, ‘dirty talk’. The most important difference between slang terms and swearwords is that the latter are strongly linked to stress and aggression. In fact, it can be said that the primary function of swearwords is to express serious and usually negative sentiments by referencing emotionally charged concepts using words that are taboo in most contexts. Swearwords typically comprise a vocabulary drawn from the semantic fields of sex, scatology and/or religion. In addition to this, swearwords are usually not ephemeral, but endure with their emotional charge intact for centuries. Furthermore, they are not associated with specific subgroups of a society, though less well-educated people are often thought of as using them more freely than do those of the middle and upper classes. In spite of these differences, swearwords are often thought of as slang because they share slang’s informality and its egalitarian spirit (Moore 2012). 346

Chinese slang

Swearing is not usually indulged in when one is in a formal setting (e.g. a courtroom or a religious ceremony) and its use pointedly contrasts with the attitude of deference that, for example, formal titles and other honorifics call for. In these regards it resembles slang. In fact, swearwords, though prototypically used to express pain, anger or similarly serious emotions, are often ‘borrowed’ for use in contexts where the playfulness of slang is the dominant ethos. They may serve, in this borrowed role, to promote sociability in the same way that slang typically does. It is the borrowing of swearwords for this sociable function along with the traits of informality and egalitarianism that make swearwords seem rather slang-like despite their prototypically distinct function. Slang and swearwords (in both English and Mandarin) can be seen as subcategories of a single category, that is, language whose affective tone is inappropriate in formal contexts – aggressively so for swearwords, but in a manner that is lightly jocular, lively or otherwise non-deferential for ordinary slang. Tāmāde (他妈的), a widely used Chinese expletive whose literal meaning is ‘His mother’s  .  .  .’, is typically used in moments of stress or aggression. It derives its obscene charge from its implied reference to sexuality, and it has been part of popular Chinese discourse for over a century. The twentieth-century writer Lu Xun wrote a well-known essay on tāmāde in which he declared it to be China’s ‘national swear’ guómà (国骂). Though it is a longstanding expression and one whose use is often not at all sociable or playful, many Chinese regard it as a kind of slang. Other swearwords can similarly be regarded as borderline cases, i.e. so slang-like as to be popularly regarded as slang. The word rì (日), for example, is a regional expression in northern China equivalent in meaning (and vulgarity) to the more widely known cào (操) meaning ‘fuck’. As a regional expression, it is sometimes thought of as a kind of slang. Nevertheless, it is undeniably a swearword. There are also expressions like niúbī (牛屄) and diǎosī (屌丝), both of which are obscene, and therefore classifiable as swearwords, yet whose sudden popularity among China’s younger generation and their common usage in non-aggressive contexts, makes them unmistakably slang. In other words, the categories of slang terms and swearwords overlap both in function and in terms of how certain borderline lexemes may be designated by virtue of their characteristics. Cào in fact, is one of the most vulgar and strongly tabooed words in Mandarin. It is widely used as a term of abuse in a particular construction, cào nǐ mā (操你妈), which means ‘fuck your mother’. A slangy euphemism that substitutes for cào is kào (靠). Kào is often used in the phrase Wǒ kào, which has the general quality of the English slang phrase ‘Screw it!’ Kào works well as a euphemism for the stronger expletive since they are roughly similar in pronunciation, even to the point of sharing the same tone. Furthermore kào, like cào, is a verb. Kào, whose literal meaning is ‘to lean on’ or ‘to depend on’, represents another example of standard language being recruited into slang vocabulary.

Internet liyu One contemporary arena in which a new kind of lǐyǔ has flourished is the Internet, particularly as this brave new world has been put to use by young Chinese (Yang 2009). Chinese campus life, especially in elite institutions, brings together students from widely varying regions and the heavy use of the Internet by young urban Chinese connects them to fellows of their generation from all over the country and beyond. In addition to the Internet per se is the cell phone, a device which, among young urban Chinese, has become the standard means of communicating with friends. Text messaging and the use of such sites as Weibo and Renren on the Internet parallel each other as major forums within which China’s post-1980 347

Robert L. Moore

and 90 generation are developing new ways of speaking. And this has had a transformative effect on the concept of lǐyǔ. When Mandarin speakers talk about lǐyǔ, they often think in terms of the regional expressions whose use is thought to particularly characterize the laobaixing (老百姓) or common people of provincial China. However, lǐyǔ today is not understood to be merely a localized, dialect-based or long-lived register of vivid, informal expressions. The Internet and cell-phone networks have been great enablers of China’s current youth culture and the language of the Internet, as used by young Chinese, is heavily infused with a new kind of lively lǐyǔ that is neither regional nor enduring. The informal language of the Internet (wǎngluò yǔyán 网络语言) is a subcategory of ‘fashionable speech’ (liúxíng yòngyǔ 流行用语). Fashionable speech itself is a designation for the ever-changing vocabulary of China’s youth culture. Where the slang or lǐyǔ of the Internet is concerned, several features that mark it are also attributes of English slang, particularly ephemerality and a propensity for wordplay. Of course, given that this new version of lǐyǔ is transmitted by Internet and text messages, it is also different from most types of slang in that it is written rather than oral. But being a written form has by no means dampened its creative and playful qualities (Moore et al. 2010). The use of letters to stand in for Chinese characters is a common feature of Internet slang. Two common Internet expressions are gg and mm, the former referring to an appealing guy, the latter to a cute girl. The logic of these abbreviations is that gēge (哥哥) is actually the word for ‘older brother’, a term which can itself be considered slang (after all, one’s older brother is not, in fact, a potential romantic partner). Similarly, the word for ‘younger sister’ mèimei (妹妹) is slang for attractive girl. From these two expressions, the Internet abbreviations gg and mm are derived. The appropriateness of the abbreviation mm is reinforced by the expression měiméi (美眉) ‘pretty eyebrow’, which is also a slang term for a pretty girl. That these simple, two-letter codes could be used effectively to convey their meanings is testament to the interest of contemporary young Chinese in romantic issues. One of slang’s functions, after all, is to provide a ready expression for a concept of widely shared, affectively charged significance. tmd is an Internet abbreviation for the swearword tāmāde (他妈的). As with gg and mm, the tmd abbreviation calls to mind its designated lexeme so readily in part because the word is so well known. Numbers are also used to represent certain widely used phrases: ‘88’ by virtue of its Mandarin pronunciation (bābā) is used to mean ‘bye-bye’ or ‘good-bye’. The Mandarin phrase yīshēng yīshì (一生一世) means ‘forever’. The numbers ‘1314’ are pronounced yīsān yīsì, and by virtue of their similarity in pronunciation to yīshēng yīshì, also stand for ‘forever’. Sometimes tones are helpful in identifying the meaning of Internet number codes. For example, ‘520’, pronounced wǔèrlíng, is commonly used to mean ‘I love you’, which is pronounced wǒ ài nǐ (我爱你) in Mandarin. Though the similarity between these two phrases may seem slight, their corresponding tones are quite close and render the 520 code easily interpretable. Chinese Internet slang sometimes incorporates English phrases such as ‘lol’ and ‘omg’. A Japanese word that is found both in Internet and in the spoken slang of young Chinese is kawayi, meaning ‘cute’. The similarity between this word and the Mandarin phrase kě’ài (可爱), which is very similar in meaning, has perhaps encouraged the adoption of the Japanese expression. Other slang terms popular on the Internet include zhuǎi (拽), which describes a person who is somewhat arrogant about his or her talent or accomplishments and nǎocán (脑残) meaning ‘brain damaged’ and descriptive of someone exhibiting foolish behavior. Dǎ zhuānkuài 348

Chinese slang

(打砖块, literally, ‘hitting bricks’) describes the playing of various online video games. This term may have originated with a popular game called Yí Zhuānkuài (移砖块 ‘moving bricks’). The Mandarin word for ‘thunder’ (léi 雷) is used to express amazement or embarrassment, more or less as the English expression ‘omg’ is used. The character 囧 (pronounced jiǒng in standard Mandarin) has occupied a prominent place in recent Internet slang. Its usage relies on a kind of wordplay that is only possible in a language whose written form employs ideographs, e.g. a language like Chinese. The character 囧 is thought to resemble a shocked or unhappy face, and it is used on the Internet to convey this kind of sentiment. Its use has been so popular in recent years that there might be said to have been a kind of 囧 fever in Internet communications in China. In American youth slang, there is perhaps no semantic field richer in slang terminology than that concerning alcohol and drugs. Alcohol use among high school and university students has resulted in a proliferation of English slang terms descriptive of drunkenness – smashed, plastered, three sheets into the wind – to name a very few. Then, since the emergence of the 1960s counterculture, a parallel series of drug-related terms was born: e.g. wasted, zonked, high, and stoned. Chinese youth culture has never focused on alcohol or drug use as the kind of context within which supposedly ‘liberating’ or sexually adventurous activities were likely to be found, and, no doubt because of this, Chinese slang lacks the rich and endlessly augmented vocabularies related to drunkenness and drug use in American youth culture. There are, of course, a few slangy ways to reference alcohol-related behavior, as, for example the traditional colloquial term for a heavy drinker, jiǔguǐ (酒鬼) meaning, literally, ‘alcohol ghost’.

Generational slang Slang expressions can follow a number of different career paths. They may emerge suddenly, become popular and widely used and then quickly disappear. This is what happened to the American expressions ‘groovy’ and ‘far out’, as the counterculture rose and then dispersed in the 1960s and 70s, for example. Alternatively, a slang expression may catch on and become part of the standard vocabulary as happened to ‘skyscraper’ in the mid-twentieth century. And, as expressions like Cantonese chìsin and English ‘cool’ illustrate, slang terms may endure for decades or longer. Ephemerality is one of the features commonly attributed to slang, but this quality is less constant than is often realized. Some expressions, particularly those of the young generation, are, in fact, more likely to come and go quickly than are others. This is partly due to the fact that young people are both more concerned with fashions and trends than are the middle-aged, and more amenable to change in general. Sometimes the changes that the young embrace are partly motivated by their desire to mark themselves as distinct from their (stuffy, out-of-it) elders. This happened in the USA twice in the twentieth century, each time in association with youth movements that transformed important social values. The youth rebellion of the 1920s was linked to the slang term ‘swell’, while that of the 1960s was linked to ‘cool’, and each of these slang terms caught on with a young generation and then endured for decades as that generation grew up (Moore 2004). In the 1990s, young Chinese embraced major cultural changes and their new, individualistic value system was linked to a new Chinese slang term: kù (酷). By the mid-1990s, China had become open in a way that contrasted sharply with the closed and tightly controlled society of the Mao era (1949–76). The reform era, which officially began in 1978 and which resulted in a gradual opening to the outside world, along with steadily increasing opportunities 349

Robert L. Moore

for entrepreneurship, utterly changed the way families and individuals thought about their lives and their prospects. Consequently, the generation that came of age in the 1990s embraced individualism in a way that their parents could not – given that their parents grew up in the 1960s and 70s when striving for individual benefits was regarded as immoral and was often harshly punished. It was at this point in history, just as individualism was taking hold of a young generation, that China opened itself to the Internet. Starting in 1994, universities and eventually Internet cafés began to make it possible for young people to link up with the World Wide Web. Kù was one of the first Internet slang terms to become widely popular, and as it did so, it came to be associated with the young generation of that era and with the individualism that they, in contrast to their elders, were happy to make their own (Moore 2005). Kù, is phonetically based on the English slang term ‘cool’, and expressed in writing as 酷. This character, in standard Mandarin, is pronounced kù and means something like ‘cruel’. As is so often the case an old word was put to a new and slangy use. In the case of kù, the word made its way to the People’s Republic of China mainly through the Internet via Hong Kong and Taiwan where the word was already being used by the 1990s. The semantics of kù include the image of the self-possessed, emotionally restrained male, that is, an image that matches quite closely the prototypical concept of cool in English. In addition to this, kù also calls to mind the flamboyant young rock star, male or female, whose outstanding trait is a style expressing an individualistic flair. Both of these ideals, the emotionally restrained and the flamboyant, are positively valued as were all things designated by the term kù in China’s youth culture of the 1990s. As happened with swell and cool in the West, the word kù in China evolved from a term designating a specific ideal to one whose referents could be anything of which the speaker approved. By about 2005, kù had become so pervasive in China that some declined to consider it a slang term, since, according to some Chinese speakers, lǐyǔ should refer only to expressions that are linked to a subgroup, regional or otherwise. Kù, once a slang term that only young Chinese recognized, has become widely known among generations old and young today (Moore 2005). Another recently emergent youth culture term is niúbī (牛屄) hich literally means ‘cow vagina’ or, more crudely, ‘cow cunt’. Neither of these English words quite captures the affective tone of niúbī, which is much slangier and more vulgar than the former, but not quite so vulgar as the latter (Moore et al. 2010). This expletive, which has been part of youth discourse since the 1990s, is used to express excitement and approval. It can be seen as a somewhat crude way to say, ‘Awesome!’ and it is widely used, for example at sporting events when one side sees its team make a winning move. The prominence of niúbī as a youth culture term resulted in Eveline Chao using it as the primary title of her 2009 book, Niúbī! The Real Chinese You Were Never Taught in School, the subtitle designating a category within which niúbī can certainly be counted. Those reluctant to use the term in light of its vulgarity, may simply say niú to the same general, slangy effect. Females, for example, are much less likely than males to utter the full expression. On the Internet there are also ways to avoid writing 牛屄, again, given the crudeness of its second element. For example, the letter B may be substituted (牛B) or, the commonly used letter X (which often stands in for obscene characters in Chinese) might also be substituted: 牛X. More recently some new slang terms have emerged and taken China’s younger generation by storm. The first of these is gāofùshuài (高富帅), a word descriptive of the perfect young boyfriend: tall, rich and handsome. Of course these qualities don’t really capture everything that every young woman looks for in a boyfriend or potential husband, but the expression 350

Chinese slang

gāofùshuài implies an image of young, male perfection. In a way it is comparable to the English phrase ‘tall, dark and handsome’ that was popularly applied to ideal romantic partners in the English-speaking world in the 1950s. Gāofùshuài is one of the most widely known current youth slang terms and has triggered the creation of a female counterpart, báifùměi (白富美), meaning light-complexioned, rich and beautiful. The rather sudden emergence of these twin terms is an example of slang, particularly the slang of adolescents and young adults, expressing an attitude that characterizes a generation and distinguishes the young from their elders. The terms gāofùshuài and báifùměi came along at a point when Chinese, particularly young Chinese, were feeling the positive effects of their impressively booming economy and when China itself had just confidently introduced itself to the twenty-first-century world via the 2008 Olympics. These terms may well be linked to an attitude of youthful optimism – an attitude, to be sure, much more characteristic of educated urban youth than of young peasants or migrant workers. It is also connected to the freedom with which young people now pursue romantic relationships, a freedom that has evolved only gradually since the beginning of the reform era. By 2012, gāofùshuài had produced another spinoff term: diǎosī (屌丝). Diǎosī is a somewhat vulgar expression whose literal meaning is ‘(male) pubic hair’. Despite its crudeness, diǎosī took off dramatically in 2012 as a youth slang expression meaning ‘loser’ or ‘slacker’. Originally it was applied as a derogatory term to people who were regarded as having no prospects in life, perhaps by virtue of their laziness or lack of ability. The prototypical diǎosī was a young man who had no job (or no decent job), who had few prospects of finding a decent job, and who wasted time playing video games rather than trying to improve his circumstances. Eventually young urbanites began to apply the term ironically to themselves when they perceived their situation as less than ideal. It became a kind of fad for young college graduates to call themselves diǎosī when they couldn’t get a job or their career prospects seemed otherwise less than promising. In 2012, just as the Chinese economy was slowing down and showing signs indicative of long-term trouble, posing as a diǎosī became all the rage among disappointed or frustrated young urbanites. Another trend supported by the Internet is anonymous messages that direct criticism at the government, or, in some cases, business enterprises that act irresponsibly. The government itself has provided fodder for the creation of slang expressions as, for example, in its use of the phrase héxié shèhuì (和谐社会), meaning ‘harmonious society’. The efforts of the authorities to ensure a ‘harmonious society’, have at times been so heavy-handed that Internet users began to use héxié (harmonious) as a verb meaning, essentially, ‘to suppress’. This is a bit of lǐyǔ in the form of satirical wordplay that grew quickly popular on the Internet until héxié itself was ‘harmonized’ or suppressed by the powers who control the Great Firewall of China. Websites that used héxié to mock government policy found themselves blocked. So, a new slang term emerged on the Internet: héxiè (河蟹) or ‘river crab’ (Clark 2012). This phrase is almost identical in pronunciation to ‘harmonious’ differing only slightly in tone. Héxiè is now one of the more popular Internet slang expressions. To support the idea of the river crab as an emblem expressing resentment against the government, a series of satirical images began to appear on Weibo including, for example, a depiction of China’s national flag with five crab images replacing the five stars in the upper-left-hand corner. Added to the satirical rhetoric of héxiè is that of sāngè dàibiǎo (三个代表) or ‘the three represents’. The three represents is a program promoted by the former president of China, Jiang Zemin, starting in 2002. In putting forward the sāngè dàibiǎo slogan, Jiang intended to encourage the party to represent the forces promoting advanced economic production, cultural development and reliance on the support of the majority of the people. However, 351

Robert L. Moore

in the hands of Internet rebels, sāngè dàibiǎo became sāngè biǎo (三个表) or ‘three wristwatches’. Again, pictorial images of three wristwatches soon began to supplement the written phrase, and before long, Internet artists came up with combined images, e.g. a river crab (héxiè) wearing three watches (sāngè biǎo). An equally popular Internet phrase with attendant imagery is cǎo ní mǎ (草泥马) meaning ‘grass mud horse’. This seemingly meaningless combination is actually a near homonym of the abusively obscene phrase discussed above, cào nǐ mā’ (操你妈). Cǎo ní mǎ has been widely used to mock government authorities in a way that is difficult to censor since ‘grass mud horse’ has no overt political meaning and includes no forbidden obscenities – it merely suggests them. Once again, verbal play has been supplemented by visual imagery as llamas with fuzzy coats or stuffed animals resembling llamas have appeared on the Internet as representatives of the ‘grass mud horse’. The phrases héxiè, sāngè biǎo, and cǎo ní mǎ are examples of slang being put to pointedly political use in a way that was not possible before the emergence of the Internet in China. Though their political significance makes them different in function from other slang terms, their liveliness, their humor and their use of old words for new purposes are all classic features of lǐyǔ.

Nicknames Nicknames can be seen as correlates of slang in that they reflect many of the same features typical of slang and they serve some of the same functions. Like slang, nicknames (Mandarin: wàihào 外号) are used in informal contexts wherein egalitarian relationships predominate. They are pointedly informal, they often rely on wordplay and/or humor for their non-serious effects, and their use sometimes defines social boundaries, i.e. distinguishing those who may from those who may not use a particular name. As with English speakers, Chinese who use nicknames are often those who are marginalized or relatively powerless, e.g. young people and members of criminal gangs (Moore 1993). If as Eble argues, one of the primary functions of slang is to promote sociability, then it is noteworthy that the same thing can be said of wàihào. Slang appears to be a universal feature of languages everywhere. It may well be that nicknames are similarly widespread and, if so, their existence, like that of slang, bespeaks a need for people to have access to a register or form of discourse that can set aside or overcome the barriers inherent in formality and status differences.

Conclusion Lǐyǔ is a register that was once disparaged as the crude and lively talk of uneducated provincials. But the Internet and cell phone have engendered a lively new style of communication, particularly favored by urban youth, that is informal and egalitarian. This is a new kind of lǐyǔ, one linked to the very same young people who are destined to be China’s leaders tomorrow. Given this, China’s slang may be said to have evolved to the point where it is now a bifurcated entity with one foot in traditional provincial speech and the other in the social media that are reshaping the world. Where future modes of communication and future generations will take lǐyǔ, of course, remains to be seen.

Acknowledgement I would like to thank Professor Li Wei of Rollins College for helpful comments offered on an early draft of this chapter. 352

Chinese slang

Note 1 In this chapter, pinyin is used to Romanize Mandarin words and the Yale system is used for Cantonese.

References Adams, Michael (2009) Slang: The People’s Poetry, New York: Oxford University Press. Bolton, Kingsley and Christopher Hutton (1997) ‘Bad Boys and Bad Language: Chou Hau and the Sociolinguistics of Swearwords in Hong Kong Cantonese’, in Grant Evans and Siumi Maria Tam (eds) Hong Kong: The Anthropology of a Chinese Metropolis, Honolulu: University of Hawaii Press. Chao, Eveline (2009) Niubi! The Real Chinese You Were Never Taught in School, New York: Penguin Books. Cihai (Encyclopedic Dictionary) (1979) Shanghai: Shanghai Cishu Publishing House. Clark, Paul (2012) Youth Culture in China: From Red Guards to Netizens, Cambridge: Cambridge University Press. Dumas, Bethany and Jonathan Lighter (1978) ‘Is Slang a Word for Linguists?’ American Speech 53(1): 5–17. Eble, Connie (1996) Slang and Sociability: In-group Language among College Students, Chapel Hill: University of North Carolina Press. Li, Shujuan and Ligang Yan (2006) Zuixin Zhongguo Lǐyǔ: Han Ying Duizhao (New Slang of China: Chinese–English), 3rd ed., Beijing: New World Press. Moore, Robert L. (1993) ‘Nicknames in Urban China: A Two-tiered Model’, Names 41(2): 67–86. Moore, Robert L. (2004) ‘We’re Cool, Mom and Dad Are Swell: Basic Slang and Generational Shifts in Values’, American Speech 79(1): 59–86. Moore, Robert L. (2005) ‘Generation Kù: Individualism and China’s Millennial Generation’, Ethnology 44(4): 357–76. Moore, Robert L. (2012) ‘On Swearwords and Slang’, American Speech 87(2): 170–89. Moore, Robert L., Eric Bindler, and David Pandich (2010) ‘Language with Attitude: American Slang and Chinese Lǐyǔ’, Journal of Sociolinguistics 14(4): 524–38. Wright, Susan and Helen Kelly-Holmes (eds) (1997) One Country, Two Systems, Three Languages: A Survey of Changing Language Use in Hong Kong, Clevedon: Multilingual Matters. Xu Shen (1963) [orig. ca. 100 ce] Shuo Wen Jie Zi (Comprehensive Dictionary), Beijing: Zhonghua Shuju. Yang, Guobin (2009) The Power of the Internet in China: Citizen Activism Online, New York: Columbia University Press.

353

20 Chinese syntax Walter Bisang University of Mainz, Germany

1. Introduction Chinese is often seen as a typological and a theoretical challenge to assumptions concerning the relevance of syntax in its interaction with semantics and pragmatics. The range of theoretical positions on the importance of syntax in Chinese reaches from its total irrelevance to its equal importance across the languages of the world. The former positions see no necessity of assuming a specific level of syntax, since everything can be accounted for in terms of semantics/cognition or in terms of pragmatics (LaPolla 1990, 2009), while the latter positions take the relevance of syntax for granted (e.g. C.-T. J. Huang et al. 2009), since it is based on Universal Grammar (UG) and the general assumption that syntax as an innate property of the human brain is the same across all human languages. Y. Huang (1994: 2) takes an intermediate position in his statement that ‘the extent to which syntax and pragmatics interact varies typologically’. In his view, languages like Chinese, Japanese and Korean are characterized by the fact that pragmatics is involved in linguistic domains that are governed by syntax in languages like English, French or German. This chapter will base its discussion of the relevance of syntax on the observation that Chinese grammar allows the omission of syntactically relevant marking to a larger extent than English (cf. hidden complexity, Bisang 2009). This favours situations in which grammatical information must be inferred from context and licenses the production of simple-looking surface structures with more than one syntactic analysis. In spite of this, the increased necessity of pragmatic inferencing does not automatically imply the absence of syntactic structure and principles. A look at current syntactic approaches to Chinese from this perspective reveals that syntax is clearly more relevant than is assumed by some approaches (cf. Paul 2015) but there are also certain limitations and problems. After a short introduction of hidden complexity (section 2), the following issues will be addressed: ● Topic-prominence and the category of subject (section 3), ● The domain of the verb (argument structure, passivization and the bèi-construction, the bǎ-construction; section 4), ● Nominal syntax and the determiner phrase (DP) (numeral classifiers, number marking and the structure of the DP; section 5), ● The anaphor zìjǐ ‘self’ (section 6). 354

Chinese syntax

2.  On hidden complexity and the relevance of syntax The concept of hidden complexity is seen as the pragmatics-based other side of complexity as it is generally discussed in typological literature (Sinnemäki 2011). It is based on the ‘articulatory bottleneck’ as discussed by Levinson (2000: 6, 27–30) and the observation that human speech encoding is by far the slowest part of speech production and comprehension. Other processes like prearticulation, parsing and pragmatic inference show a much higher transmission rate and thus allow the processing of much more information within a given time span. This creates an asymmetry between articulation and pragmatic inference in terms of costs at the level of each individual utterance – ‘[i]nference is cheap, articulation expensive’ (Levinson 2000: 29). There is a certain need of explicitness to guarantee mutual understanding between speaker and hearer on the one hand and there is the economic need of avoiding unnecessary costs on the other. In such a scenario, every utterance is the result of the two competing motivations of economy vs. explicitness (on competing motivations, cf. Haiman 2011). In the case of grammatical complexity, explicitness stimulates overt complexity. It supports the existence of grammatical markers which are obligatory1 in the sense that the speaker has to express a certain value of the relevant grammatical category in a given syntactic environment (e.g. the expression of past vs. non-past in a finite declarative clause of a language with a binary tense system). This type of explicitness-based complexity corresponds to overt complexity, i.e. that type that is generally discussed as ‘complexity’ in typological literature (McWhorter 2001; Sinnemäki 2011). Economy is responsible for hidden complexity and the pragmatic inference of grammatical structures. It favours grammars in which the speaker is not forced to overtly use grammatical markers. In such a context, hidden complexity applies if there is an expression format for a given grammatical category (e.g. markers of tense) and if that category can be left unmarked if the corresponding information can be pragmatically inferred.2 Grammars of individual languages differ with regard to the extent to which they are open to pragmatic inference. East and mainland Southeast Asian languages are characterized by their cross-linguistically high degree of hidden complexity (Bisang 2009, 2013). A good example of hidden complexity is radical pro-drop, i.e. the omission of verbal arguments without the existence of concomitant agreement morphology on the verb (Neeleman and Szendrői 2007; on Chinese, cf. C.-T. J. Huang 1984 and later). As can be seen from the second sentence in (1a), English is a non-pro-drop language since its subject position needs to be filled either by a noun or a pronoun. Other European languages like Italian can drop the subject but it remains retrievable from verb agreement (cf. Rizzi 1986). In the case of Chinese (1b), the subject can only be inferred from context, since there is neither an overt subject noun phrase nor an agreement marker on the verb. (1) (a) English: Did you buy a new car yesterday? – I bought one. / *ø bought one. (b) Chinese: 你昨天买了新车吗？— 买了。 Nǐ zuótiān mǎi-le xīn chēi ma? – ø mǎi-le you yesterday buy-PFV new car Q buy-PFV ‘Did you buy a new car yesterday? – [I] bought [one]’

355

øi.

Walter Bisang

In contrast to more radical stances concerning the irrelevance of syntax in Chinese, hidden complexity does not automatically imply the absence of syntax. But grammars characterized by a high degree of hidden complexity license effects that create challenges for assessing the relevance of syntax. An important effect are surface structures that can be associated with more than one syntactic analysis. This effect shows up in various central domains of grammar that have given rise to controversial discussions. The following issues will be addressed in this chapter: ● Topic-prominence and the existence of the syntactic category of subject (section 3): What are the criteria for distinguishing subjects from topics if NP1 in a simple surface structure of the type [NP1 V NP2] could be both? ● It is often difficult to decide whether a postverbal NP is an object or an adjunct. Thus, one may ask whether there are good independent tests for distinguishing arguments from non-arguments (section 4.1). ● Bare nouns may express different values for number (singular, plural) and definiteness (definite, indefinite, generic). Even the [Numeral Classifier Noun]-construction can either express quantification or indefiniteness (cf. section 5).3 In spite of this, there is evidence for a DP (cf. section 5). ● In certain types of complex clauses, the anaphor 自己 zìjǐ ‘self ’ can be bound to the local subject or to a remote subject. This implies that pragmatic inference is at least of some relevance for binding (section 6). A central issue of syntactic theory is subject/object asymmetry. The assumption that subjects are structurally higher than objects is one of the basic tenets of UG-based approaches. In Government and Binding Theory (Chomsky 1981) for instance, the subject takes the specifier position of the inflectional phrase (IP), while the object is a complement of the verb (V). This is illustrated by (2a), which contrasts with the flat structure in (2b), in which subject and object are at the same hierarchical level:

IP

(2) (a)  NPsubject

(b) 

I′ I

IP NPsubject

V

NPobject

VP V′ V

NPobject

The universality of subject/object asymmetry is rejected for reasons of empirical problems by many linguistic theories, among them Lexical–Functional Grammar (Bresnan 2001) and Role and Reference Grammar (Van Valin 2005). In these theories, only hierarchically flat structures of the type in (2b) are compatible with cross-linguistic variation. The phenomena discussed in this chapter will show a mixed picture. The subject/object asymmetry seems to be relevant to a certain extent but there are inconsistencies: 356

Chinese syntax

● Topic-extraction out of relative clauses is not fully governed by subject/object asymmetry (cf. the relevance of animacy and specificity in (10) vs. (11)). ● The assignment of agent-arguments and patient arguments to the positions of subject and object is hard to be fully captured by a thematic hierarchy. ● There is no diathesis that neatly follows the subject/object asymmetry (in contrast to the English passive). Neither the bèi-construction (section 4.2) nor the bǎ-construction (section 4.3) reflect this asymmetry. ● Anaphors are not exclusively subject-oriented. In some cases, we find object coreference (cf. (52) in section 6).

3.  Topic-prominence and the category of subject The topic is generally defined either in terms of aboutness or in terms of a frame. The following definition by Lambrecht (1994: 131) is an example of the aboutness-based definition: ‘[a] referent is interpreted as the topic of a proposition if in a given situation the proposition is construed as being about this referent, i.e. as expressing information which is relevant to and which increases the addressee’s knowledge of this referent’. In Chafe’s (1976: 51) definition, the topic provides ‘the frame within which the sentence holds’. Both functions are relevant to Chinese. The information that follows the topic is called the comment. In various descriptions, Chinese grammar is much more about the topic than about the subject. Chao (1968) describes Chinese basically as a topic–comment language in his extensive grammar rather than as a subject-predicate language. He states that ‘[t]he grammatical meaning of subject and predicate in a Chinese sentence is topic and comment’ (Chao 1968: 69) and goes on saying that ‘[t]he subject is literally the subject matter to talk about, and the predicate is what the speaker comments on when a subject is presented to talk about’ (Chao 1968: 70). Since C. N. Li and Thompson (1976), Chinese is often characterized as a topic-prominent language rather than as a subject-prominent language. In LaPolla’s (2009) view, the topic is sufficient, there is no need to posit syntactic categories. In this section, we will briefly look at the notions of topic in Chinese and the question of how relevant syntactic categories and the subject/object asymmetry are. As it will turn out, there are criteria for distinguishing subjects from topics and subject/object asymmetry works in certain contexts but not fully consistently. If Chinese is a topic-prominent language, the question is whether and how topics can be distinguished from subjects. Since Chinese is a radical pro-drop language, a single noun phrase in the preverbal position of a structure [NP1 V NP2] may be interpreted as a topic followed by a zero-subject but it may also be analysed simply as a subject or as a topic (in LaPolla’s 2009 approach). In the view of C. N. Li and Thompson (1981), the ‘subject’ does not necessarily have to be equivalent to a syntactic category, since they give primary importance to the topic-prominence of Chinese. In fact, they argue that ‘“subject” is not a structurally definable notion’ (C. N. Li and Thompson 1981: 19).4 Thus, NP1 could basically be both but its status as a topic is of greater importance in a topic-prominent language. The adequacy of this statement crucially depends on reliable tests of subjecthood, an issue that has been widely and controversially discussed. The following two tests provide some good evidence for the existence of a separate syntactic category of subject. Since the aboutness relation expressed by topics presupposes accessible concepts (Lambrecht 1994), only subjects can be indefinite. As can be seen from (3), the explicitly marked indefinite object yī ge bēizi ‘a cup’ in (3a) cannot be moved to the topic position in front 357

Walter Bisang

of the subject wǒ ‘I’ (3b). However, preverbal indefinite subjects are acceptable in thetic5 utterances like (3c) and (3d). In (3c), the indefinite NP is part of the passive construction (on bèi, cf. section 4.2). In (3d) it occurs with dǎ-suì ‘be broken’ in its function as an intransitive verb. In these constructions, the preverbal NP must be a subject, it cannot be a topic. (3) Lu et al. (2015):

(a) 我打碎了一个杯子。 wǒ dǎ-suì-le yī I hit-break-PFV one ‘I broke one cup’

(b) *一个杯子我打碎了。 *[yī ge bēizi] wǒ one CL cup I ‘One cup, I broke [it]’

ge bēizi. CL cup

dǎ-suì-le. hit-break-PFV



(c) 一个杯子被我打碎了。 yī ge bēizi bèi wǒ dǎ-suì-le. one CL cup PASS I hit-break-PFV ‘A cup was broken by me (and I was negatively affected)’

(d) 一个杯子打碎了。 yī ge bēizi dǎ-suì-le. one CL cup hit-break-PFV ‘A cup broke/was broken’ The above example shows that subjects differ syntactically from topics. The problem is that the above criterion depends on the overt presence of the relevant criterion. Without overt indefinite marking (hidden complexity), it will simply be impossible to come up with a clear-cut categorial decision on the status of NP1 in a given utterance of the type [NP1 V NP2]. Another test is relativization. As the following example from C.-T. J. Huang et al. (2009: 212–13) shows, only subjects can be relativized. The example is based on the verb fāshēng ‘happen’, which can be intransitive with a patient argument (4a) as well as transitive with an additional experiencer argument (4b): (4) C.-T. J. Huang et al. (2009: 212):6 (a)

意外发生了。 yìwài fāshēng-le. accident happen-PFV ‘An accident happened’



(b) 那些人发生了意外了。 nèixiē rén fāshēng-le yìwài those man happen-PFV accident ‘Those people had an accident’

358

le. PF

Chinese syntax

If the intransitive verb takes a topic as in (5a), that topic cannot be relativized (5b): (5) C.-T. J. Huang et al. (2009: 213):7

(a) 那些人,发生了意外了。 nèixiē rén, fāshēng-le yìwài le. those man happen-PFV accident PF ‘As for those people, an accident happened’

(b) *意外发生了的那些人 *[yìwài fāshēng-le de] nèixiē rén accident happen-PFV MOD those man ‘The people such that an accident happened’ Relativization is only possible if the noun phrase nèixiē rén ‘those people’ is the subject in the transitive construction shown in (4b): (5) (c) C.-T. J. Huang et al. (2009: 213): 发生了意外的那些人 [ fāshēng-le yìwài de] nèixiē happen-PFV accident MOD those ‘the people who had an accident’

rén man

To get a full picture of the topic in Chinese, it is necessary to distinguish two types of topic constructions. In the first type, the topic is syntactically related to the comment, in the second type it is only semantically or pragmatically related to the comment. The former type is sometimes called ‘English-style topic’, the latter ‘Chinese-style topic’ (Xu and Langendoen 1985). In the following example of an English-style topic, the topic Lǐ Lín ‘Li Lin’ is coindexed with a zero-object in the comment wǒ pīpíng-guo [I criticize-EXP]: (6) English-style topic: 李林,我批评过。 [Lǐ Lín]i, wǒ pīpíng-guo øi. Li Lin I criticize-EXP ‘Li Lin, I criticized him’ In Chinese-style topics, there is no corresponding extraction site in the comment. The relation between the topic and the remainder of the clause (comment) must be inferred (for another example, cf. (5a)): (7) Chinese-style topic: C. N. Li and Thompson (1981: 96): 那场火,幸亏消防队来得快。 [nèi chǎng huǒ], xìngkuī xiāofángduì lái de that CL fire, fortunately fire.brigade come CSC ‘That fire, fortunately the fire brigade came quickly’.

kuài. quick

In the case of Chinese-style topics, which are characteristic of topic-prominent languages like Chinese, the semantic relation between the topic and the subject is that between a whole (topic) and its part (subject). This definition includes semantic relations like set/member, abstract/instance, process/step, object/attribute or generalization/specific (cf. C. N. Li and Thompson 1976; Y. Huang 1994: 286). 359

Walter Bisang

C. N. Li and Thompson (1981) claim that movement is generally irrelevant for topics in Chinese. While this looks plausible for Chinese-style topics, which are not related to any element of the comment that follows them, it is problematic for English-style topics. In fact, examples like (6) above can be described in terms of movement. The NP Lǐ Lín ‘Li Lin’ has the function of an object in the corresponding clause wǒ pīpíng-guo Lǐ Lín [I criticize-EXP Li Lin] ‘I criticized Li Lin’ and is moved out of that position to the topic position. C.-T. J. Huang et al. (2009: 199–211) offer additional evidence for this analysis which cannot be discussed in detail here. Their findings clearly argue against C. N. Li and Thompson’s (1981) view that movement is generally irrelevant for Chinese topics. After having discussed some evidence for the subject/topic distinction, it remains to be seen to what extent there is subject/object asymmetry in English-style topics. The following example shows that the grammar allows situations in which the topic can be either the subject or the object. The noun phrase jī ‘chicken’ can either be the patient argument (8a) or the agent argument (8b). Possible contexts for inferring object coreference or subject coreference are given in examples (9a) and (9b), respectively:   (8)

Y. Huang (1994: 168): 鸡,吃完了。 jī, chī-wán-le. chicken eat-finish-PFV



(a) Object coreference: ‘[The chicken]i, we have eaten iti up’ (b) Subject coreference: ‘[The chicken]i, iti has eaten [it] up’

  (9) Y. Huang (1994: 168): (a) Object coreference: 鸡,吃完了,肉还有。 chī-wán-le øi, ròu hái yǒu. jīi , chicken eat-finish-PFV meat still there.is ‘Chicken, we have eaten all of it, but there is still meat’ (b) Subject coreference: 鸡,吃完了,要不要喂点食？ øi chī-wán-le, yào bú yào wèi diǎn shí? jīi chicken eat-finish-PFV want NEG want feed some food ‘The chicken, it has eaten. Does it want to have some more food again?’ While example (9) looks like a straightforward case of no subject/object asymmetry, there are instances that support the assumption of a subject/object asymmetry. An interesting case in point is the extraction out of a relative clause that modifies either the subject or the object of a clause. In this case, extraction out of a relative clause modifying an object is only possible if the zero-element in the relative clause is bound by a noun that is inanimate as in (10). If the extracted noun is animate (or specific), extraction is ungrammatical as illustrated by (11) (cf. Xu and Langendoen 1985): (10) Xu and Langendoen (1985: 15): 这个问题,我从来没遇到过能回答的人。 rén] [zhè ge wèntí]i, wǒ cónglái méi yùdào-guò [[néng huídá øi de] this CL question I always NEG meet-EXP   can answer MOD man ‘This question, I have never met a person who can answer’ 360

Chinese syntax

(11) C.-T. J. Huang and Li (1996: 82): *张三,我认识很多批评的人。 [[pīpíng *[Zhāngsān]i, wǒ rènshi hěn duō Zhangsan I know very many criticize ‘Zhangsan, I know many people who criticize him’

øi

de] MOD

rén]. man

Thus, there is subject/object asymmetry in English-type relative clauses of Chinese but it interacts with non-syntactic criteria like animacy and specificity in the case of extraction out of a relative clause.

4.  The domain of the verb 4.1.  Argument structure Arguments are parts of the semantic representation of the verb. They are linked to syntactic categories like subject or object. If a verb has two arguments, its agent (A) is linked to the subject position and its patient (P) is linked to the object position in an active clause (e.g. JohnA eats an appleP). Thus, the argument structure of a verb is crucial for linking semantics with syntax. In addition to arguments, there are also non-arguments which provide additional information that does not belong to the semantic representation of the verb and connot be linked to the object position (e.g. locatives as in the garden in John eats an apple in the garden). In Chinese, it is difficult to find clear-cut tests for determining the argument structure of a verb. A particular problem is the distinction between intransitive (one argument) and transitive verbs (two arguments) because it is quite hard to determine whether an unmarked postverbal noun phrase is an object or an adjunct of an intransitive verb. Generally speaking, the number of positions after the verb is much more limited than in the preverbal position. In most cases, there are only two postverbal constituents, which can be noun phrases, prepositional phrases or adverbs that quantify the duration or the occurrence of an event. In the following example, we have five preverbal constituents and two postverbal constituents: (12) Lu et al. (2015): 他上星期天在办公室用计算机认真的填了八个小时表格。 [tā] [shàngxīngqītiān] [zài bàngōngshì] [yòng jìsuànjī] [rènzhēnde] he last.Sunday at office with computer carefully tián-le [bā ge xiǎoshí] [biǎogé]. fill.in eight CL hours form ‘He filled in forms carefully on the computer in the office for eight hours last Sunday’ The following five examples will be used to illustrate the problem of deciding whether NP2 in a structure [NP1 V NP2] is a second argument that qualifies for the object function or a non-argument that can be linked to an adjunct position (cf. Lu et al. (2015)): (13) 他吃了一个苹果。 tā chī-le yī s/he eat-PFV one ‘S/he ate an apple’

ge píngguǒ CL apple

361

Walter Bisang

(14) 我听耳机。 wǒ tīng ěrjī. I hear earphone ‘I use an earphone to listen’ (15) 米煮了粥。 mǐ zhǔ-le zhōu. rice cook-PFV gruel ‘The rice has been cooked into gruel’ (16) 他出了房子。 tā chū-le fángzi. s/he go.out room ‘S/he goes out of the room’ (17) 他高三英寸。 tā gāo sān yīngcùn. s/he tall three inches ‘S/he is three inches tall’ As is suggested in Bisang (2006) and Lu et al. (2015), headless relative clauses can be used for testing the argument status of a noun phrase. Regular relative clauses precede their head noun and are marked by the modifier de (MOD) at their right edge (18a). In the case of subject and object coreference, the head noun is normally represented by a corresponding zero-element in the relative clause. Headless relative clauses are formed by the omission of the head noun (cf. (18a) vs. (18b)): (18) (a) 我们吃的苹果 [wǒmen chī øi de] we eat MOD ‘the apple we eat’

(b) 我们吃的 [wǒmen chī øi we eat ‘what we eat’

píngguǒi apple

de] MOD

If no arguments are expressed in the relative clause we get a minimal headless relative clause (MHRC), which only consists of the verb and the modification marker [V de]. From a purely pragmatic perspective, it should be possible to interpret the empty head in the MHRC as an argument as well as a non-argument. Thus, one would expect that a construction like chī de [eat MOD] may mean ‘the one who eats’ (agent coreference, subject), ‘the thing that is eaten’ (patient coreference, object) as well as ‘the place where X eats’ (locative non-argument, adjunct) or ‘the thing with which X eats’ (instrumental non-argument, adjunct). But this is not the case. The MHRC only allows coreference with arguments. Even in sentences in which the MHRC takes the subject position and is followed by a wh-word that asks for an adjunct, adjunct coreference is not possible. If the MHRC chī de [eat MOD] is followed by zài nǎr ‘where?’, as in (19), the sentence can mean ‘Where is [the thing] X eats?’ or ‘Where 362

Chinese syntax

is the one who eats?’ but not ‘Where is [the place] where X eats?’. Thus, (19) allows subject or object coreference (19a, b) but not coreference with an adjunct (19c): (19) 吃的在哪？ [chī de] zài nǎr? eat MOD where

(a) ‘Where is the thing to be eaten/food?’ (b) ‘Where is the one who eats?’ (c) *‘Where is the place for eating?’

Similarly, if the question word is shénme ‘what?’, chī de [eat MOD] can only be interpreted in terms of patient coreference ‘what s/he eats’ but not in terms of adjunct (instrumental) coreference: (20) 吃的是什么？ [chī de] shì shénme? eat MOD is what (a) ‘What is [s/he] eating?’ or ‘What is it that [s/he] is eating?’ (Lit.: ‘What [s/he] eats is what?’) (b) *‘What is s/he eating with?’ The question word must be a general question word. If it is too specific or if it contains a noun like fànguǎn ‘restaurant’ that triggers non-argument coreference, non-argument coreference is possible in a normal relative clause with its head (21a) as well as in the MHRC in (21b): (21) (a) 吃的饭馆 [chī de] fànguǎn eat MOD restaurant ‘the restaurant where we eat’ (b)

吃的是什么饭馆？ [chī de] shì shénme fànguǎn? eat MOD is what restaurant ‘What kind of restaurant is the place of eating?’

As is shown in Lu et al. (2015), the coreferential properties of MHRCs can be used as a test of argumenthood. Only argument coreference is allowed in MHRCs. In the above examples, the MHRC test only works with the verbs in (13) to (14). In the case of tīng ‘hear’ in (14), we get the following picture: (22) (a) Agent coreference: 听的是谁？ [tīng de] shì shéi? hear MOD be who ‘Who is the one who is listening?’ 363

(b) Instrument coreference: 听的是什么？ [tīng de] shì shénme? hear MOD be what ‘What was he listening with?’

Walter Bisang

In contrast, there is no way of asking for fángzi ‘room’ in (16) or sān yīngcùn ‘three inches’ in (17). Therefore, the verbs chū ‘go out’ and gāo ‘be tall’ are intransitive verbs according to the MHRC test. The argument roles of agent and patient combine a number of more specific semantic roles (cf. e.g. the concepts of Proto-agent vs. Proto-patient (Dowty 1991) or the MacroRoles of Actor and Undergoer in Role and Reference Grammar (Van Valin 2005)). In many approaches, the assignment of the individual arguments and their specific semantic roles to the syntactic categories of subject and object is defined in terms of hierarchies. If a verb has two arguments, the higher semantic role is linked to the subject, while the lower argument takes the object position. Chen (2004) has proposed the following hierarchy for Chinese: (23) Agent > Experiencer > Instrument > Time/Location > Goal/Source > Theme > Patient In example (14), the verb tīng ‘hear’ has two arguments with the semantic roles of experiencer (tā ‘s/he’) and instrument (ěrjī ‘earphone’). Since the experiencer role is ranked higher than the instrument, the experiencer is the agent argument and the instrument is the patient argument. This mechanism works very well for many verbs but there are configurations which are hard to capture. In the following example, xiǎobēi ‘small cup’ is the instrument and lǜchá ‘green tea’ is the patient. Under the assumption that the subject is preverbal and the object postverbal, example (24b) violates the hierarchy in (23) because the lower ranked patient argument lǜchá ‘green tea’ is in the subject position. (24) Lu et al. (2015): (a) 小杯喝绿茶。 xiǎobēi hē lǜchá. small.cup drink green.tea ‘Small cups are for drinking green tea’ / ‘Small cups are used for drinking green tea’ (b) 绿茶喝小杯。 lǜchá hē xiǎobēi. green.tea drink small.cup ‘Green tea is drunk with small cups’ / ‘For drinking green tea, the small cup is used’ The MHRC test basically confirms the existence of three types of argument structures, i.e. intransitive, transitive and ditransitive verbs. The class of intransitive verbs also includes state-denoting lexemes. Thus, what is separated as a specific word class of adjectives in English is subsumed under the word class of verbs in Chinese (cf. Schachter 1985 on adjectival-verb languages). Intransitive verbs can be divided into verbs whose single argument is a patient (unaccusative verbs) and verbs whose single argument is an agent (unergative verbs). The subject of unaccusative verbs can occur in the preverbal and in the postverbal position, while the subject of unergative verbs is limited to the preverbal position (Y.-H. A. Li 1990). There are quite a few ditransitive verbs in Chinese (for the transitive verbs, cf. the above discussion of examples (13) to (15)). They all occur in the neutral ditransitive construction [NPagent V NPrecipient/source NPpatient], which is characterized by two unmarked postverbal noun phrases. Depending on whether the indirect object is a recipient (25) or a source (26), there 364

Chinese syntax

are two major groups of ditransitive verbs.8 Both of these types can occur in the neutral ditransitive construction: (25) Neutral ditransitive construction with recipient: 我送了他一本书。 wǒ sòng-le tā yī běn shū. I give-PFV s/he one CL book ‘I gave her/him a book’ (26) Neutral ditransitive construction with source: 李林买了他一件雨衣。 Lǐ Lín mǎi-le tā yī jiàn yǔyī. Li Lin buy-PFV s/he one CL raincoat ‘Li Lin bought a raincoat from her/him’ Ditransitive verbs of the above types can also occur in other constructions. In the coverb construction, the recipient (or source) is introduced by a coverb/preposition9 (27a) in the second postverb position. The applicative construction10 is only possible with a reduced number of coverbs, among them gěi ‘give/to’. For that reason, it is ungrammatical with source arguments. As can be seen from (27b), the coverb is suffixed to the main verb, followed by a tense-aspect marker. Finally, the patient argument can occur with the bǎ-construction (cf. section 4.3) in the preverbal construction (27c): (27) (a) Coverb construction: 我送了一本书给他。 wǒ sòng-le yī běn I give-PFV one CL ‘I gave a book to him’

shū book

gěi tā. PREP s/he

(b) Applicative construction: 我送给了他一本书。 wǒ sòng-gěi-le tā yī běn shū. I give-to-PFV s/he one CL book ‘I gave him a book’ (The transfer is completed) (c) bǎ-construction: 我把一本书送给了他。 wǒ bǎ yī běn shū sòng-gěi-le I BA one CL book give-to-PFV ‘I gave a (specific) book to him’

tā. s/he

Some ditransitive verbs can only occur in the neutral ditransitive construction and the bǎ-construction, among them the verb 给 gěi ‘give’ and verbs of information transfer like 告诉 gàosù ‘tell’. Finally, there are a few verbs like 吃 chī ‘eat’ and 喝 hē ‘drink’, which can get an additional source noun phrase in the neutral ditransitive construction: (28) 他喝了我三瓶酒。 tā hē-le wǒ sān píng jiǔ. s/he drink-PFV I three bottle wine ‘S/he drank three bottles of my wine’ [Lit.: ‘S/he drank three bottles of wine off of me’] 365

Walter Bisang

4.2. The bèi-construction and passivization The 被 bèi-construction is often called a passive and there is indeed a certain similarity in its canonical form. As can be seen from the comparison of (29a) with (29b–c), the patient-argument of the transitive verb mà ‘scold, rebuke’ takes the subject position and the agent-argument follows bèi if there is one ((29b) vs. (29c)). (29) (a) 母亲骂了他。 mǔqīn mà-le tā. mother scold-PFV s/he ‘Mother scolded her/him’ (b)

他被母亲骂了。 Tā bèi mǔqīn mà-le. s/he PASS mother scold-PFV ‘S/He was scolded by mother’

(c)

他被骂了。 Tā bèi mà-le. s/he PASS scold-PFV ‘S/He was scolded’

In spite of this surface similarity, there are also significant differences to passives in languages like English. Thus, the bèi-construction occurs in a second non-canonical or indirect form, in which an additional experiencer-NP is introduced into the subject position, while the actor-NP follows bèi and the patient-argument takes the postverbal position (30): (30) Indirect passive (C.-T. J. Huang et al. 2009:140): 张三被土匪打死了爸爸。 Zhāngsān bèi tǔfěi dǎsǐ-le bàba. Zhangsan PASS bandit kill-PFV father ‘Zhangsan had his father killed by bandits’ The passive construction is less productive than the English passive. This has to do with its specific meaning that implies that the event it marks is perceived as negative or adversative to the agent argument. Even though the meaning of adversity has lost ground in more formal registers (government documents, news reports, science and technology writings), it still holds widely in other registers. Moreover, the bèi-construction does not apply to every verb with two arguments. It works with verbs like chī ‘eat’ in (13) or mà ‘scold’ in (29) but it is ungrammatical with verbs whose second argument is less affected by the event expressed by the verb. For that reason, it is ungrammatical with (14) and (15). The word bèi is analyzed either as a coverb/preposition or as a verb. The former analysis goes back to C.-T. P. Wang (1970), the latter to Hashimoto (1969). In Wang’s (1970) account, an underlying object moves into the subject position, while the agent is integrated into a prepositional phrase headed by bèi. Hashimoto (1969) takes bèi as a matrix verb with a complement clause whose object is deleted under identity. As C.-T. J. Huang et al. (2009: 112–52) point out, both approaches have their shortcomings. Their analysis includes both movement and complementation. The matrix verb bèi takes a noun phrase as its subject and a clause (IP) as its complement. Movement comes in because of a null operator (NOP) which is assumed to be adjoined to the complement clause. 366

Chinese syntax

4.3. The bǎ-construction The 把 bǎ-construction is prototypically described by examples like (31), in which the postverbal object of a transitive verb (31a) takes the preverbal position and is marked by bǎ (31b). The object noun phrase is preferably indefinite in (31a) and definite or specific in the bǎ-construction (31b): (31) The bǎ-construction: (a)

他吃了苹果。 tā chī-le píngguǒ. s/he eat apple ‘S/he ate (the) apple(s)’

(b)

他把苹果吃了。 tā bǎ píngguǒ chī-le. s/he BA apple eat-PFV ‘S/he ate the apple(s)’

In C. N. Li and Thompson’s (1981: 463–91) classical syntactic analysis, bǎ is a coverb. Definite objects can be moved out of their postverbal position into the preverbal position (SVOdef → S bǎ Odef V). This analysis is problematic in various ways. The object position of the bǎ-construction can be filled by another noun phrase that is in a part–whole relationship (and some other relationships) with the noun following bǎ (32). Moreover, bǎ acts as a causative marker that can select an additional cause argument in the subject position (33): (32) 李林把十个苹果吃了三个。 Lǐ Lín bǎ shí ge píngguǒ chī-le Li Lin BA ten CL apple eat-PFV ‘Li Lin ate three apples out of ten’

sān three

ge. CL

(33) 这件事把他哭累了。 zhè jiàn shì bǎ tā kū-lèi-le. this CL affair BA s/he cry-tired-PFV ‘This affair made her/him cry to her/his exhaustion’ Given these shortcomings of the classical coverb/prepositional analysis, most more recent accounts describe bǎ as a causative verb. Sybesma (1999) proposes that bǎ takes the position of little v0. In C.-T. J. Huang et al. (2009: 176–85) it is analysed as a separate functional category above vP with the subject taking the specifier position of the bǎ-phrase: (34) [baP Subject [ba‘ ba [vP NP [v’ v [VP XP]]]]]. In C. N. Li and Thompson’s (1981) account, the NP after bǎ is definite. As example (35) shows, it is also possible to have indefinite NPs in that position as long as their identifiability can somehow be presupposed. This is the case with selling an object. Example (35) would be ungrammatical with the verb mǎi ‘buy’ because it implies a larger set of potential objects than mài ‘sell’. The verb mài ‘sell’ presupposes the possession of an object and thus implies a certain familiarity of the subject NP with its object NP. 367

Walter Bisang

(35) Lu et al. (2015): 他把一辆车卖了。 tā bǎ yī liàng s/he BA one CL ‘S/he sold a car’

chē car

mài-le. sell-PFV

The semantics of the bǎ-construction are often described in terms of the two correlated notions of disposal and affectedness. The definition in terms of disposal11 goes back to L. Wang (1943) and takes the perspective of the subject, which disposes of the element marked by bǎ in the way described by the verb. In turn, the object is affected by the action expressed by the verb. Even though this definition accounts for many cases, it does not cover the whole functional range of the bǎ-construction. The event of ‘catching a glimpse’ in (36) neither implies disposal nor affectedness: (36) 李林把张三看了一眼。 yī Lǐ Lín bǎ Zhāngsān kàn-le Li Lin BA Zhangsan look-PFV one ‘Li Lin caught a glimpse of Zhangsan’

yǎn. glimpse

Reading the literature on bǎ, one may get the impression that linguists feel more con­ fident about its syntactic analysis than its semantic analysis. The following statement from C.-T. J. Huang et al. (2009: 195) may reflect this situation: ‘The traditional wisdom of “affectedness/disposal” seems to capture the function of the ba construction intuitively. However, only the core cases submit easily to the characterization. Our account did not address this difficult issue.’

5.  Nominal syntax and the DP As in many other languages of East and mainland Southeast Asia, bare nouns can express many functions which need obligatory marking in other languages. Thus, a bare noun can be interpreted as singular or plural. It can also be interpreted as generic, definite or indefinite. In the following example, lǎobǎn ‘boss’ is interpreted as definite and singular or plural, while chē ‘car’ is indefinite and singular or plural: (37)

老板买了车。 lǎobǎn mǎi-le chē. boss buy-PFV car ‘The boss(es) bought a car/cars’

The Chinese noun phrase is head-final. For that reason, the head noun consistently takes the final position of the noun phrase. It can be preceded by demonstratives (DEM), numerals (NUM), classifiers (CL) and relative clauses (RC).12 Numerals and demonstratives are obligatorily combined with a numeral classifier. If both of them cooccur in a noun phrase the demonstrative precedes the [NUM CL]-constituent, which cannot be interrupted by any other constituent: (38) DEM [NUM CL] N 368

Chinese syntax

(39)

这/那三只猫 zhè/nà sān zhī māo this/that three CL cat ‘these/those three cats’

Relative clauses either take the phrase-initial position or the position between the classifier and the head noun:13 (40) (a) RC DEM [NUM CL] N (b) DEM [NUM CL] RC N (41) (a) 我很喜欢的那三只猫 [wǒ hěn xǐhuan de] nà sān I very like MOD this/that three ‘those three cats that I like very much’

(b) 那三只我很喜欢的猫 nà sān zhī [wǒ hěn xǐhuan those three CL I very like ‘those three cats that I like very much’

zhī māo CL cat

de] MOD

māo cat

The remainder of this section will focus on numeral classifiers, the fine-structure of nominal syntax and the functional differences expressed by the two word-order options of the relative clause. As pointed out above, Chinese numerals need to be combined with a numeral classifier (42a). The omission of the classifier is ungrammatical (42b): (42) (a)

三封信 (b) sān fēng xìn three CL letter ‘three letters’

*三信 *sān xìn three letter intended meaning: ‘three letters’

Since Greenberg (1972), the use of numeral classifiers is frequently associated with the idea that the concept expressed by nouns in classifier languages like Chinese needs to be individuated before it can be quantified by a numeral (Paris 1981; Bisang 1999). If all Chinese nouns are mass nouns like English water or milk, the classifier serves to profile a unit that is intrinsic to a given nominal concept. By this process, the noun is made accessible to counting. From the perspective of formal semantics, Chierchia (1998) argues that nominal concepts have to be atomized for counting. In his analysis, Chinese nouns are all like English furniture, i.e. they include both the set of atoms of a nominal concept and the set of sums of atoms of that concept. If they have to be counted they have to be atomized, i.e. they have to be singled out as individual atoms by a numeral classifier. In the case of English furniture, the word piece is used for that purpose (cf. three pieces of furniture vs. *three furnitures). Even though the general analysis of Chinese nouns as mass nouns is widespread, it is not uncontroversial. Thus, Cheng and Sybesma (1999: 517) argue that ‘the fact that all nouns require a classifier does not mean that all nouns are mass nouns’. In their view, there is a count/mass distinction in Chinese but that distinction ‘is not visible at the noun level,  .  .  .  , it is reflected at the classifier level’ (Cheng and Sybesma 1999: 519). In some more recent 369

Walter Bisang

publications, classifiers are accounted for by two properties. The two properties are [±numerable] and [±countable] in Zhang’s (2013) view and [±counting] and [±measuring] in X. Li’s (2013) approach. Chinese has a rich inventory of classifiers that point out different semantic properties of the nouns they individuate or atomize. Some examples are 只 zhī, 口 kǒu ‘mouth’ and 头 tóu ‘head’ (for animals), 棵 kē ‘stalk, stem’ (for trees), 条 tiáo ‘branch’ and 支 zhī ‘twig’ (for one-dimensional objects), 张 zhāng (for two-dimensional objects), 块 kuài ‘lump, chunk’ (for three-dimensional objects), 件 jiàn (for clothes, goods/merchandise, weapons), 所 suǒ ‘place’ and 间 jiān ‘space, interval’ (for places, buildings and rooms), 本 běn (for books), 封 fēng ‘seal’ (for letters) and 尊 zūn (for statues, temples and pieces of artillery).14 In addition to these specific classifiers, there is the general classifier 个 ge. This classifier occurs with nouns denoting humans and with other nouns for which there is no specific classifier. Moreover, it can replace specific classifiers in spoken discourse (Erbaugh 1986). For a more fine-grained syntactic analysis of nominal syntax, it is necessary to look at the expression of plural and of the definiteness/indefiniteness function of [NUM CL N]. As is well known, Chinese has a suffix -men that is used for marking collective plural with nouns denoting humans. Even though it is not a strict marker of plurality it is incompatible with numerals: (43)

*三个朋友们 *sān ge péngyou-men. three CL friend-PL Intended meaning: ‘three friends’

Y.-H. A. Li (1999) argues that this is a syntactic fact rather than a semantic fact associated with the collective meaning of -men (also cf. C.-T. J. Huang et al. 2009: 311–15, 287–95). The basic idea is that there is a classifier phrase (ClP) between the numeral phrase (NumP) and the NP (cf. (44)). In such a tree structure, the plural marker cannot be realized in N because there is a ClP between NP and NumP (Head Movement Constraint). In other languages like English, there is no ClP. For that reason, the combination of numerals plus number marking on the noun in English is not blocked. DP

(44) Dem

NumP Spec

Num′ Num

CIP CI

NP

In addition to the NumP, Y.-H. A. Li (1999) also argues for a determiner phrase (DP) above NumP. An important reason for that analysis is the fact that a structure of the type [NUM CL N] can either be semantically limited to the expression of quantity as in example 370

Chinese syntax

(45), which is only about the quantity of ‘three students’, or to the expression of indefiniteness as in example (46), which is about two students whose identity is assumed to be unknown to the hearer: (45) C.-T. J. Huang et al. (2009: 289): 三个学生不够。 sān ge xuésheng bú gòu. three CL student NEG enough ‘Three students are not enough’ (46) C.-T. J. Huang et al. (2009: 290): 我叫两个学生回去把他们的车子开来。 chēzi kāi lái. wǒ jiào liǎng ge xuéshengi huíqù bǎ tāmeni de I ask two CL student return BA their MOD car drive over ‘I asked two students to go back and drive their car over’ There is an additional difference between the quantity-denoting function of [NUM CL N] in (45) and the individual-denoting function in (46) that is associated with the indefiniteness interpretation. As can be seen from (46), the individual-denoting interpretation allows pronominal coreference binding (tāmen ‘they is coreferent with liǎng ge xuésheng ‘two students’). This is not possible in the case of quantity-denoting expressions – it is ungrammatical to take up sān ge xuésheng ‘three students’ in (45) by the pronoun tāmen ‘they’ in ongoing text. The head D of a DP is generally associated with reference. If there is no reference available, coreference or binding relations are not possible. For that reason, Y.-H. A. Li (1999) claims that identity-denoting [NUM CL N]-constructions are DPs (cf. (44)) with an empty determiner in D. In contrast, quantity-denoting constructions are NumPs and thus lack a DP-layer. The two positions of the relative clause are associated with interesting differences concerning the referential status of a noun. One of them is the difference between specific and non-specific (Zhang 2006). If the relative clause takes the DP-initial position the DP is interpreted as specific (47a). If it occurs after the [NUM CL]-constituent, the DP is non-specific (47b): (47) X. Li and Bisang (2012: 347): (a) 我在找学英语的一个学生,我觉得他可能在图书馆。 yī ge xuéshēng]], wǒ zài zhǎo [[RC xué yīngyǔ de] I PROGR seek learn English MOD one CL student wǒ juéde tā kěnéng zài túshūguǎn. I think s/he maybe be.at library ‘I am looking for a student who learns English. I think that he might be in the library’

(b) 我在找一个学英语的学生。谁都可以,但是最好是大四的。 xuéshēng]. wǒ zài zhǎo [yī ge [RC xué yīngyǔ de] I PROGR seek one CL learn English MOD student Shéi dōu kěyǐ, dànshi zuìhǎo shì dàsì de. Who all possible but best be senior MOD ‘I am looking for a student who learns English. Anyone will do, but it is better to have a senior student’ 371

Walter Bisang

6.  The anaphor zìjǐ ‘self ’ The Chinese anaphor or reflexive 自己 zìjǐ ‘self ’ offers interesting insights into the interaction of syntax, semantics and pragmatics. In Chomsky’s (1981) classical Binding Theory, anaphors (zìjǐ ‘self ’), pronouns (tā ‘s/he’, etc.) and referential expressions (r-expressions, common nouns) show characteristic coreferential behaviour (binding) within the clause:15 (48) Binding principles (Chomsky 1981):

(A) An anaphor must be bound in its clause. (B) A pronoun must be free (not bound) in its clause. (C) An r-expression must be free.

The following examples illustrate principles A and B.16 In (49a), the anaphor zìjǐ ‘self ’ is coreferent with Li Lin. Since the subject Lǐ Lín ‘Li Lin’ is in the same clause as the anaphor, this behaviour is predicted by binding principle A. The binding properties of pronouns like tā ‘s/he’ is complementary to those of anaphors. For that reason, it is to be expected that tā ‘s/he’ cannot be coindexed with the subject of its clause in (49b). (49) (a)

李林批评了自己。 zìjǐi. Lǐ Líni pīpíng-le Li Lin criticize-PFV SELF ‘Li Lini criticized himselfi’

(b)

李林批评了他。 tā*i/j. Lǐ Líni pīpíng-le Li Lin criticize-PFV s/he ‘Li Lini criticized him*i/j’

As is well known, the Chinese reflexive zìjǐ ‘self’ violates the classical form of the binding principle in (48a). In a sentence like (50), zìjǐ ‘self ’ cannot only be coreferent with the subject of the same clause, it can also be coindexed with the remote subject of the matrix clause (Zhāngsān ‘Zhangsan’): (50) 张三认为李四恨自己。 Zhāngsāni rènwéi [Lǐsìj hèn zìjǐi/j]. Zhangsan think Lisi hate SELF ‘Zhangsani thinks that Lisij hates himselfj/himi’ Binding across the relevant clause (governing category, cf. fn. 15) is known under the term of long-distance binding. C.-T. J. Huang and Liu (2001) claim that this fact can be accounted for by logophoricity (Hagège 1974; Sells 1987). In logophoric systems, binding of an anaphor depends on a particular point of view or discourse perspective that can be determined by a source (the intentional agent of a communication), a self (the one whose mental state or attitude the content of the proposition describes) or a pivot (the one with respect to whose time-space location the content of the proposition is evaluated) (Sells 1987: 457). In such an analysis, zìjǐ ‘self’ in (50) must be seen from what Zhangsan really said in direct speech:

372

Chinese syntax

(50′)

张三认为李四恨我。 Zhāngsān rènwéi ‘Lǐsìj hèn wǒ’. Zhangsan think Lisi hate SELF ‘Zhangsan thinks “Lisi hates me” ’

The pronoun wǒ ‘I’ in (50′) refers to the speaker Zhangsan and this underlying discourse representation is reflected in the long-distance coreference in (50). Logophoricity is not only claimed to explain the properties of long-distance binding, it is also claimed to account for blocking effects in which coreference with the remote subject is blocked by intervening noun phrases. This effect can be seen if the intervening noun phrase is a pronoun of the first or second person. In (51), Lǐsì ‘Lisi’ is exchanged by wǒ ‘I’. (51) 张三认为我恨自己。 Zhāngsāni rènwéi [wǒj hèn zìjǐ*i/j]. Zhangsan think I hate SELF ‘Zhangsani thinks that Ij hate myselfj/*himi’ Seen from the perspective of the speaker’s report, the sentence in (51) was uttered as follows in direct speech: (51′) 张三认为我恨我。 Zhāngsān rènwéi [wǒ hèn wǒ]. Zhangsan think I hate I ‘Zhangsani thinks: “I hate myself ” ’ This interpretation is responsible for the blocking of long-distance binding. Thus, zìjǐ ‘self ’ in (51) must be coreferent with the local subject wǒ ‘I’. Based on the above explanations, C.-T. J. Huang et al. (2009: 350) conclude that there is an anaphoric use of zìjǐ ‘self ’ that shows syntactic effects of local binding and a logophoric use that exhibits semantic effects of logophoricity. In both cases, zìjǐ ‘self ’ is always coreferent with a subject (the local or the remote subject). This explanation covers a large amount of data but it also disregards certain contexts in which zìjǐ ‘self ’ is used. A general problem comes from the observation that one and the same sentence may allow both local and long-distance coreference. This is the case with example (50) above. The only way of deciding which of the two options is intended is pragmatic inference from context. Thus, pragmatics must have its relevance for a full understanding of the anaphoric properties of zìjǐ ‘self ’ (cf. Y. Huang 1994). This can also be seen from example (52), which additionally undermines the assumption that zìjǐ ‘self ’ is strictly subject-orientated. As Y. Huang (1994) argues, object coreference with Xiǎohuā ‘Xiaohua’ is the preferred interpretation but subject coreference is not excluded in the right context: (52) Object or subject coreference (Y. Huang 1994: 103): 小明问小花自己会不会英语。 huì bú huì Xiǎomíngi wèn Xiǎohuāj zìjǐj/i Xiaoming ask Xiaohua SELF know NEG know ‘Xiaomingi asks Xiaohuaj whether hej/i knows English’

373

yīngyǔ. English

Walter Bisang

In many cases, pragmatic inference is based on the meaning of the verbs involved. Some verbs express actions that are typically performed on somebody else (extroverted or otheroriented verbs), while other verbs denote actions that are performed on the agent him/herself (introverted or self-oriented verbs) (König and Vezzosi 2004). Self-oriented verbs like kàojìn ‘approach’ conform to coreference with the remote subject in (53), while self-oriented verbs like fǎnxǐng ‘(self-)examine’ trigger local coreference in (54): (53) With an other-oriented verb (Schumacher et al. 2011): 张三让李四靠近自己。 [Lǐsìj kàojìn zìjǐi/*j]. Zhāngsāni ràng Zhangsan let/ask Lisi approach SELF ‘Zhangsani asked Lisij to approach himi/*j’ (54) With a self-oriented verb (Schumacher et al. 2011): 李林让赵六反省自己。 [Zhàoliùj fǎnxǐng zìjǐ*i/j ]. Lǐ Líni ràng Li Lin let/ask Zhaoliu (self-)examine SELF ‘Li Lini asked Zhaoliuj to (self-)examine himself*i/j’ Finally, pragmatic inference may depend on highly specific semantic properties of individual verbs. The following example from Y. Huang (1994: 183) is based on the fact that Chinese has two different verbs for ‘marry’. The verb 娶 qǔ is used if the object is a woman, while 嫁 jià means that a man is married. For that reason, local binding in (55) is pragmatically excluded for reasons of world knowledge: (55) 徐小姐希望王先生娶自己。 [Xú xiǎojie]i xīwàng [Wáng xiānsheng]i qǔ Xu Miss hope Wang Mr marry(man) ‘Miss Xui hopes that Mr Wangj will marry heri / ???himselfj’

zìjǐi/???j. SELF

7. Conclusion The present chapter started out from the cross-linguistically high degree of hidden complexity in Chinese as the reason for controversial discussions concerning the importance of syntax in its interaction with semantics and pragmatics. As it turned out, there clearly is syntax in Chinese. There are syntactic reasons for distinguishing subjects from topics (section 3), there is a potential test of argumenthood and a distinction between objects and adjuncts in the postverbal position (section 4), there are reasons for assuming a hierarchical structure like the DP in the domain of the noun (section 5) and anaphors follow syntactic rules up to a certain extent (section 6). In spite of this, certain strongholds of syntax like the subject/object asymmetry do not operate with the same rigour in various domains of grammar (cf. section 1 and subsequent discussions on topic extraction out of relative clauses, the relevance of thematic hierarchies, the lack of diatheses and the option of coindexing anaphors with objects under certain circumstances). It is these problems at the boundaries between syntax, semantics and pragmatics that are of crucial importance for a deeper understanding of what makes syntax in a cross-linguistic perspective. For that reason, the study of Chinese syntax has enormous theoretical potential. 374

Chinese syntax

List of abbreviations BA CL CSC DEM EXP MOD NEG NUM PASS PF PFV PL PROGR Q RC SELF

Marker bǎ in the bǎ-construction Numeral classifier Complex stative construction Demonstrative Experiential aspect Modification marker (e.g. for relative clauses and possession) Negation Numeral Passive (bèi-passive) Perfect marker Perfective aspect Plural marker Progressive aspect Question marker Relative clause marker Reflexive anaphor

Notes  1 Another aspect of hidden complexity is the existence of multifunctional grammatical markers whose concrete meaning must be pragmatically inferred (Bisang 2009). This aspect will not be discussed here.   2 This definition does not include grammatical distinctions that are attested somewhere in the languages of the world but happen not to be expressed in a given language. For that reason, there is no hidden complexity with regard to evidentiality in English because evidentiality is not a grammatical category in that language.   3 In many Sinitic languages, among them Wu Chinese and Cantonese, bare classifier phrases of the structure [CL N] can express definiteness and indefiniteness (X. Li and Bisang 2012).   4 It can only be defined semantically as an element that ‘must always have a direct semantic relationship with the verb as the one that performs the action or exists in the state named by the verb’ (C. N. Li and Thompson 1981: 15).   5 Thetic utterances consist exclusively of information the speaker assumes to be inactivated in the hearer. Thus, there is no topic that provides an anchor for situating that information (Kuroda 1972). Otherwise, preverbal positions are strongly associated with definiteness in Chinese (C. N. Li and Thompson 1981 and many others).   6 C.-T. J. Huang et al. (2009: 212) use the pronoun tāmen ‘they’ in the subject position of (4). I use the noun phrase nèixiē rén ‘those people’, which will also be used as the head of the relative clause in (5b) and (5c).   7 I use again the noun phrase nèixiē rén ‘those people’ instead of the pronoun tāmen ‘they’ in (5a).   8 Some verbs whose indirect object is a recipient: 给 gěi ‘give’ [occurs only in the neutral ditransitive construction], 送 sòng ‘give, send’, 赠 zèng ‘give as a present’, 寄 jì ‘mail’, 赏 shǎng ‘bestow’, 赐 cì ‘bestow’, 教 jiāo ‘teach’, 传 chuán ‘pass’, 还 huán ‘return’. Some verbs whose indirect object is a source: 买 mǎi ‘buy’, 拿 ná ‘take’, 偷 tōu ‘steal’, 抢 qiǎng ‘rob, grab’, 收 shōu ‘receive’, 骗 piàn ‘cheat somebody out of something’, 削 xiāo ‘peel’. In a few verbs, the indirect object can be either a recipient or a source: 借 jiè ‘borrow, lend’, 租 zū ‘rent to/from’.   9 The term ‘coverb’ refers to verbs that have been grammaticalized into prepositions. 10 The term ‘applicative’ is not identical to what is generally understood as applicatives in linguistics. 11 L Wang (1943) uses the Chinese term 处置 chǔzhì, which is translated as ‘disposal’. 12 Since Chinese is an adjectival-verb language (cf. section 4.1), modification by adjectives is not discussed separately. It is roughly subsumed under the heading of relative clauses. 13 There is a less prominent third option in which the RC occurs between DEM and NUM. This option will not be discussed in this chapter.

375

Walter Bisang 14 A translation is added to the classifier if there is a synchronically transparent relation between the classifier and a nominal concept. 15 I use the term ‘clause’ for reasons of simplification. The classical definition of the relevant local domain is the Governing Category. 16 Principle C will not the discussed in this chapter.

References Bisang, Walter (1999) ‘Classifiers in East and Southeast Asian languages: Counting and beyond’, in Jadranka Gvozdanovic (ed) Numeral Types and Changes Worldwide, Berlin: Mouton de Gruyter, 113–85. Bisang, Walter (2006) ‘From Meaning to Syntax – Semantic Roles and Beyond’, in Ina Bornkessel, Matthias Schlesewsky, Bernard Comrie, and Angela D. Friederici (eds) Semantic Role Universals and Argument Linking, Berlin: Mouton de Gruyter, 191–236. Bisang, Walter (2009) ‘On the Evolution of Complexity – Sometimes Less Is More in East and Mainland Southeast Asia’, in Geoffrey Sampson, David Gil, and Peter Trudgill (eds) Language Complexity as an Evolving Variable, Oxford: Oxford University Press, 34–49. Bisang, Walter (2011) ‘Grammaticalization and Typology’, in Heiko Narrog and Bernd Heine (eds) Handbook of Grammaticalization, Oxford: Oxford University Press, 105–17. Bisang, Walter (2013) ‘Chinese from a Typological Perspective’, in Tseng Chiu-yu (ed) Papers from The Fourth International Conference on Sinology: Human Language Resources and Linguistic Typology, Taipei: Academia Sinica, 1–32. Bresnan, Joan (2001) Lexical–Functional Grammar, Oxford: Blackwell. Chafe, Wallace (1976) ‘Givenness, Contrastiveness, Definiteness, Subjects and Topics’, in Charles N. Li (ed) Subject and Topic, New York: Academic Press, 25–55. Chao, Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley and Los Angeles: University of California Press. Chen, Ping (2004) ‘Hanyu shuangxiang mingciju yu huati-chenshu jiegou’ (‘Double Noun Phrase Constructions and Topic–comment Articulation in Chinese’), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 303(6): 493–507. Cheng, Lai-Shen Lisa and Rint Sybesma (1999) ‘Bare and Not-so-bare Nouns and the Structure of NP’, Linguistic Inquiry 30: 509–42. Chierchia, Gennaro (1998) ‘Reference to Kinds across Languages’, Natural Language Semantics 6: 339–405. Chomsky, Noam (1981) Lectures on Government and Binding, Dordrecht: Foris. Dowty, David R. (1991) ‘Thematic Proto-roles and Argument Selection’, Language 67: 547–619. Erbaugh, Mary S. (1986) ‘Taking Stock: The Development of Chinese Noun Classifiers Historically and in Young Children’, in Colette Craig (ed) Noun Classes and Categorization, Amsterdam and Philadelphia: John Benjamins, 399–436. Greenberg, Joseph H. (1972) ‘Numeral Classifiers and Substantival Number: Problems in the Genesis of a Linguistic Type’, Working Papers on Language Universals 9: 2–39. Hagège, Claude (1974) ‘Les pronoms logophoriques’, Bulletin de la Société de Linguistique de Paris 69: 287–310. Haiman, John (2011) ‘Competing Motivations’ in Jae Jong Song (ed) The Oxford Handbook of Linguistic Typology, Oxford: Oxford University Press, 148–65. Hashimoto, Mantaro (1969) ‘Observations on the Passive Construction’, Unicorn 5: 59–71. Huang, C.-T. James (1984) ‘On the Distribution and Reference of Empty Pronouns’, Linguistic Inquiry 15: 531–74. Huang, C.-T. James and Li Y.-H. Audrey (eds) (1996) New Horizons in Chinese Linguistics, Dordrecht, Boston and London: Dordrecht Academic Publishers. Huang, C.-T. James, Li Y.-H. Audrey, and Li Yafei (2009) The Syntax of Chinese, Cambridge: Cambridge University Press. Huang, C.-T. James and Liu C.-S. Luther (2001) ‘Logophoricity Attitudes and Ziji at the Interface’ in Peter Cole, G. Hermon, and C.-T. James Huang (eds) Long Distance Reflexives, New York: Academic Press, 141–95. Huang, Yan (1994) The Syntax and Pragmatics of Anaphora, Cambridge: Cambridge University Press.

376

Chinese syntax König, Ekkehard and Letizia Vezzosi (2004) ‘The Role of Predicate Meaning in the Development of Reflexivity’ in Walter Bisang, Nikolaus P. Himmelmann, and Björn Wiemer (eds) What Makes Grammaticalization? A Look from its Fringes and its Components, Berlin: Mouton de Gruyter, 213–44. Kuroda, S.-Y. (1972) ‘The Categorical and the Thetic Judgment’, Foundations of Language 9: 153–85. Lambrecht, Knud (1994) Information Structure and Sentence Form, Cambridge: Cambridge University Press. LaPolla, Randy J. (1990) ‘Grammatical Relations in Chinese: Synchronic and Diachronic Considerations’, Doctoral Dissertation, University of California, Berkeley. LaPolla, Randy J. (2009) ‘Chinese as a Topic–Comment (Not Topic-Prominent and Not SVO) Language’, in Janet Xing (ed) Studies of Chinese Linguistics: Functional Approaches, Hong Kong: Hong Kong University Press, 9–22. Levinson, Stephen C. (2000) Presumptive Meanings: The Theory of Generalized Conversational Implicatures, Cambridge, MA: MIT Press. Li, Charles N. and Sandra A. Thompson (1976) ‘Subject and Topic: A New Typology of Language’ in Charles N. Li (ed) Subject and Topic, New York: Academic Press, 459–89. Li, Charles N. and Sandra A. Thompson (1981) Mandarin Chinese: A Functional Reference Grammar, Berkeley and Los Angeles: University of California Press. Li, Xuping (2013) Numeral Classifiers in Chinese, Berlin: Mouton de Gruyter. Li, Xuping and Walter Bisang (2012) ‘Classifiers in Sinitic Languages: From Individuation to Definiteness-marking’, Lingua 122: 335–55. Li, Y.-H. Audrey (1990) Order and Constituency in Mandarin Chinese, Dordrecht: Kluwer. Li, Y.-H. Audrey (1999) ‘Plurality in a Classifier Language’, Journal of East Asian Linguistics 8: 75–99. Lu, Bingfu, Guohua Zhang and Walter Bisang (2015) ‘Valency classes in Mandarin’ in Andrej Malchukov and Bernard Comrie (eds), Valency Classes in the World’s Languages, Berlin: Mouton de Gruyter, 709–64. McWhorter, John H. (2001) ‘The world’s simplest grammars are creole grammars’, Linguistic Typology 5: 125–66. Neeleman, Ad and Kriszta Szendrői (2007) ‘Radical pro drop and the morphology of pronouns’, Linguistic Inquiry 38: 671–714. Paris, Marie-Claude (1981) Problèmes de syntaxe et de sémantique en linguistique chinoise, Thèse, Université de Paris-VII. In: Mémoires de l’Institut des Hautes Études Chinoises XX. Collège de France, Paris. Paul, Waltraud (2015) New Perspectives on Chinese Syntax, Berlin: Mouton de Gruyter. Rizzi, Luigi (1986) ‘Null Objects in Italian and the Theory of Pro’, Linguistic Inquiry 17: 501–57. Schachter, Paul (1985) ‘Parts of speech systems’, in Timothy Shopen (ed) Language Typology and Syntactic Description, Vol. 1: Clause Structure, Cambridge: Cambridge University Press, 3–61. Schumacher, Petra B., Walter Bisang, and Sun Linlin (2011) ‘Perspective in the Processing of the Chinese Reflexive ziji: ERP Evidence’ in I. Hendrickx, A. Branco, S. Lalitha Devi, and R. Mitkov (eds) Anaphora Processing and Applications, Lecture Notes in Artificial Intelligence, Vol. 7099, Heidelberg: Springer, 119–31. Sells, Peter (1987) ‘Aspects of Logophoricity’, Linguistic Inquiry 18: 445–79. Sinnemäki, Kaius (2011) Language Universals and Linguistic Complexity, Helsinki: University of Helsinki. Sybesma, Rint (1999) The Mandarin VP, Dordrecht: Kluwer. Van Valin, Robert D. Jr. (2005) Exploring the Syntax-semantics Interface, Cambridge: Cambridge University Press. Wang, Li (1943) Zhongguo xiandai yufa (Modern Chinese Grammar), Beijing: The Commercial Press. Wang, C.-T. Peter (1970) ‘A Transformational Approach to Chinese ba and bei’, Doctoral Dissertation, University of Texas at Austin. Xu, Liejiong and Terence D. Langendoen (1985) ‘Topic Structures in Chinese’, Language 61: 1–27. Zhang, Niina Ning (2006) ‘Representing Specificity by the Internal Order of Indefiniteness’, Linguistics 44: 1–21. Zhang, Niina Ning (2013) Classifier Structures in Mandarin Chinese, Berlin: Mouton de Gruyter.

377

21 Chinese taboo Amy He Yun University of Huddersfield, United Kingdom

入竟(境)而问禁,入国而问俗,入门而问讳(《礼记》) Ask about the prohibition upon entering a land, the customs upon entering a country, and the names to be avoided upon entering a home. (Liji or Book of Rites)

Introduction Why are chopsticks called kuài in Chinese? Why do etiquette books advise us not to give clocks as presents in China? And why are tourism promotional slogans such as Wǒ-kàoChóngqìnɡ and Yī-zuò-jiào-chūn-de-chéngshì criticized for transgressing common decency? The key to answering these questions lies in the notion of taboo and related concepts such as euphemism. In principle, anything can be taboo, including persons (e.g. pregnant women), acts (e.g. incest) and things (e.g. blood), for example. This chapter, nevertheless, aims to provide an introduction to Chinese taboo, past and present, including taboo words and expressions, euphemisms, motivations, functions and means to avoid punishments and sanctions for taboo violations. In what follows, I first contextualize this article by delineating the key concepts, historicizing the milestones of Chinese taboo development and reviewing very briefly relevant literature, in a somewhat streamlined and simplified form. Then, after outlining the structure of this chapter, I proceed to present Chinese taboo in detail. Central to this chapter is the concept of taboo, which is the prohibition or restriction of an action based on the belief that such behavior is too sacred and consecrated, obscene and dirty, dangerous, accursed and unlucky or immoral for ordinary individuals to undertake. Linguistic taboo is accordingly defined as the total or partial prohibition or restriction of the use of certain words, expressions and topics on the basis of the aforementioned belief. Another core term is euphemism, referring to circumlocutionary usage of language in which an inoffensive or pleasant term is substituted for another that is thought to be coarse, painful, unpleasant, derogatory or unlucky (Hughes 2006; Rawson 1981). Euphemisms are in a constant state of flux: Some die away soon after being created while others last for centuries and develop into unconscious ones. Like taboos, they are variously motivated, for example, by fear, kindness, good taste, politeness, and political correction or commercial ‘deception’. 378

Chinese taboo

Then, to mention in passing, the opposite of euphemism is dysphemism, which, by and large, is tabooed. These definitions allow me to sketch the prominent stages of Chinese taboo development and to give a suggestive overview of the literature. Taboo is such an extremely ancient phenomenon in China that words tabooed at a certain point in history total 987 (Y. Wang 1997). Nonetheless, there is broad agreement that most basic ideas about it were shaped in the Zhou dynasty that ruled ancient China for almost a millennium (Chen 1997[1928]). Among other historical texts, taboo is discussed most insightfully in the Wujing, ‘five classics’, the main part of Confucian canon, for example, in the highly quoted passage from the Liji provided at the outset of this chapter. This is partly attributable to Confucius’s effort to zhèngmíng, ‘rectify names’, ultimately to restore the social hierarchy of the Zhou. In the Qin–Han empires, the hierarchical nature of taboo, particularly name taboos and taboo words for death, was reinforced with the dominance of Confucianism being cemented at the expense of the ‘Hundred Schools’. The decree of zhèn as the emperor’s unique and exclusive self-reference term marks an important step in institutionalizing Chinese taboo. In the succeeding period taboo is indeed recorded in texts of different genres (J. Wang 2002), but it did not appear very salient in social life. It came into prominence again in the united Sui–Tang dynasties because for the first time in history (name) taboo was legislated, blurring the boundary between taboos as customs and law (see 2.4). This is followed by the Song dynasty, which saw the ‘climax’ of Chinese taboo (J. Wang 2002; Chen 1997[1928]). After intervening centuries, Chinese taboo was recurrently exploited in the Ming and Qing dynasties, particularly the latter empire: Their political and cultural despotism was enforced by literary inquisitions in which many intellectuals were executed (see 2.4). With the fall of the last feudal empire and the advent of modern science and democracy in the early twentieth century came fundamental changes of Chinese views on taboo, particularly words for death, naming and addressing (see 2.3 and 2.4). In consistence with the significance of taboo in Chinese history and culture, a great deal of research has been carried out in this field (Wan 2005; Ren 2004; J. Wang 2002; Shen 1996). Unfortunately, I can just survey some key works selectively, aiming to suggest a general picture of the field. Scholarly studies date back to the turn of the twentieth century, which witnessed an upsurge of interest in the Chinese language and history. As taboos were found critical in understanding historical texts, first seminal contributions were made in historical and philological research. Most notably, in his pioneering book on shǐhuì, roughly translated as ‘name taboos in history’, Chen (1997[1928]) discusses systematically name taboos and substitutes, and many other issues such as the application of this knowledge in research. Chen’s approach is adopted in many subsequent studies, particularly those focusing on bìhuì, referring narrowly to name taboos, which is widely viewed as unique to China (Y. Wang 1997). Influenced by great works such as Freud (2004[1913]), Steiner (2013[1956]) and Frazer (2012[1922]), recent studies are much broader in scope and perspectives. To be brief, as suggested above, earlier works tended to limit their attention to name taboos in history and their research was principally taken as subsidiary to history studies (Chen 1997[1928]; Y. Wang 1997). Recent literature, however, not only covers taboos on names, sex, excretion, death, and so on in history but also discusses those current in society (Ren 2004; J. Wang 2002). While research completed decades ago was biased heavily towards taboos of the Han Chinese, recent literature (Wan 2005; Song and Xu 1999; Q. Wang 1990) has also explored fruitfully the taboo custom of minority ethnic groups. Theoretically, building on the classical philological approach, Chinese taboo is now examined from multidisciplinary perspectives, including sociolinguistics (He 2002; Chen 2001[1980]; Shen 1996), sociopsycholinguistics 379

Amy He Yun

(D. Wang et al. 1995), cultural linguistics (Qu 1996), folklore (Wan 2005), ethno-anthropology (Ren 2004) and sociology (Pan 1988). While every effort has been made to ensure that these research traditions are synthesized, the present chapter is broadly couched within traditional sociolinguistics. The remainder of this chapter is organized in three sections. Section 2 details Chinese taboo, highlighting features characteristic of Chinese culture such as the richness, strictness and hierarchy of name taboos and words for ‘death’ in history. As the major part of this chapter, it is presented in five subsections, devoted respectively to taboos in mutually exclusive taboo subjects. Section 3 is concerned with variation of taboos at diachronic and synchronic dimensions, focusing particularly on differences between dialects and/or regions and social groups that are largely historically grounded and socioculturally specific. Verbal means to avert punishment for taboo violations are also discussed briefly in this section. The chapter closes with Section 4, in which I provide a summary of Chinese taboo and a brief discussion of the pragmatics of taboo.

Taboo and the language of Chinese Except in some theoretical works such as Freud (2004[1913]) and Douglas (2003), most studies tend to examine linguistic taboo according to subjects. A notable example is Frazer’s (2012[1922]) Golden Bough, in which taboo words used as personal names, names of relations, names of the dead, names of kings and other sacred persons, and names of Gods are distinguished. However, a discernible weakness of this categorization is, among others, the failure to capture prohibited words for body wastes. Thus, drawing on recent studies such as Allan and Burridge (2006; hereafter A&B), this chapter describes Chinese taboo in five thematic categories, respectively designating (i) bodies and their effluvia; (ii) organs and acts of sex, and sexuality; (iii) diseases, mental and physical defects, and death; (iv) naming and addressing; and (v) other aspects of human existence. Moreover, as detailed below, this subject-based typology of taboo is also roughly consistent with the underlying motivations: They are, in a certain way, associated with uncleanness, filthiness and obscenity, privacy, fear, bad luck and respect.

Taboo, bodies and their effluvia The close association between dirt and bodily effluvia such as excreta, urine, menstrual blood and semen is obvious. Therefore, it is little wonder that all cultures and societies have a strong tendency to place a taboo on terms for these biological and physiological phenomena. Consequently, most of them are tabooed or used in restricted contexts such as to vent anger, frustration, and so on. This motivates a succession of replacements, giving rise to a range of euphemisms and dysphemisms. This subsection overviews Chinese taboo words and expressions for this type of human waste and related behaviors. To begin with, while biàn is a general medical term for ‘excreta’, words such as shǐ and fèn, both for ‘faeces’, are strictly taboo. Multiple ways can be used not to call a spade a spade. In historical Chinese it was often euphemized as huángjīn (yellow gold) and yèxiāng (nocturnal fragrance). Then, as expected, the job of collecting sanitary waste for the rich was referred to as dào-yèxiāng (emptying nocturnal fragrance) and the collector yèxiāng-fù (nocturnal fragrance woman). In today’s parenthood discourse baby’s poo has ‘achieved’ the status of bāba (pancakes), a euphemistic food metaphor. ‘To poo’ is then ē-bāba (discharge pancakes). These are classic examples of euphemism that are inflated in extreme size. As 380

Chinese taboo

many characters are composed of radicals that can stand alone as separate characters, an interesting way to avoid fèn, perhaps unique to Chinese, is to deconstruct it into its components mǐ (rice) and gòng (common) or mǐ, tián (field) and gòng if used in the classical form. In addition, dozens of euphemisms or slangs are substituted for lāshǐ (discharge faeces) or ēshǐ (discharge faeces). Jìngshǒu (clean hands) and chūgōng (exit respect), like dēngdōng (climb east) in historical texts, are very formal and hence may sound archaic or pedantic to some ears. Rather, one is more likely to say jiěshǒu (untie hand) or fāngbiàn (convenience) when one needs to use the toilet. Expressions such as chūqù-yīxià and chūqù-yīhuìr, literally ‘go out for a while’, appear vague but the speaker’s intention of ‘excuse me for a few minutes’ (A&B 2006: 51) cannot be misunderstood when used in the right context. In a similar way, it would be very polite to say nèijí (internally urgent) or nèipò (internally forced) when one needs to answer urgent calls of Nature. Needless to say, as universally practiced in all societies, using sign language is often an alternative strategy to effectively blot out taboo words altogether. This euphemizing mechanism is also true of ‘toilet’ language. In Modern Chinese cèsuǒ used to dominate, but these days xǐshǒujiān (hand-washing room), wèishēngjiān (sanitary room) and guànxǐshì (washroom) are apparently more popular. The substitutes give us the impression that the rooms are for washing only or provide better sanitation, which is not necessarily true. Similarly, unless in remote rural areas, máosi (straw-covered latrine) and máokēng (straw-covered pit) have virtually disappeared in Chinese ‘toilet’ discourse due probably to the backwardness they connote. In a similar vein, one is expected to censor their use of words for ‘break or breaking wind’. The noun pì and verb fàngpì would mostly be perceived as indecent, improper or shocking. Euphemisms do not appear to be many, so most people may just opt out or say it bluntly. If one has to talk about it, an elegant expression is chū-xūgōng, which is derived from chūgōng. Interestingly, due to its ‘bad reputation’, pì is frequently employed in abusive terms. For instance, pìhuà means ‘nonsense’ and fàngpìhuà ‘talking nonsense’. As in other cultures, the physiological phenomenon of menstruation is a taboo topic in China. Some studies (Ren 2004) contend that it has many taboo senses such as ‘bloodstain’, ‘dirtiness’, ‘filthiness’, ‘collapse’ and ‘cut’. More fundamentally, women are stereotyped as huòshuǐ (troublesome water), i.e. source of trouble, according to Wan (2005: 124), because of their blood, primarily menstruation blood. These sexist attitudes towards women give rise to many of the unpleasant connotations of yuèjīng (monthly experience). There is indeed a list of sweet-sounding or evasive alternatives. Some options portray menstruation as something good, including (gàn)-hǎoshì ((do) good thing) and hǎo-péngyou-(láile) (good friend (visits)). Other substitutes give emphasis to periodicity, for example, yuèshì (monthly matter), yuèshuǐ (monthly water) and yuèxìn (monthly message). As in English, several euphemistic expressions are based on the salient color hóng (red): hóngdēng-liàngle (red lights on) and xiǎohóng-láile (little red visits). In addition, dàyímā (mother’s elder sister) and gūmā (father’s elder sister), two euphemisms gaining currency, are said to have originated in a folktale from the Han empire. Finally, while tèshū-qíngkuàng (special circumstance) is used for an obvious reason, lìjià (regular leave) mirrors the social development: Women experiencing period pains are entitled to paid leave in New China. As the final point I would touch upon in this section, semen has long been viewed as a symbol of masculinity, which is evidenced in terms such as jīngyè (essential fluid), jīngzǐ (essential seed) or simply jīng (essence) and yuányáng (primordial masculinity). It is traditionally believed that ‘a drop of semen is worth ten drops of blood’. Loss of semen was 381

Amy He Yun

thought to be more debilitating and life-threatening than loss of blood, although this has already been proved to be an old wives’ tale. Cases of death from exhaustion of semen are often found in the ancient erotic literature such as The Plum in the Golden Vase.

Taboo, organs and acts of sex, and sexuality A substantial number of Chinese words and expressions pertaining to acts and organs of sex are four-letter words. The Chinese belief that wàn-è-yín-wéi-shǒu, i.e. ‘lust is the worst of all sins’, suffices to warrant an extended discussion of taboos in this category. Above all, words for sexual acts such as cào (fuck) should never be used in polite conversations because they are āngzānɡ (filthy) and xiàliú (obscene/pornographic). Due partly to this, ‘tāmāde (mother fucker)!’, an abbreviation for cào-tāmā-de-bī (fuck his mother’s cunt), was viewed by Lu Xun (2005) as ‘the national abusive term’. In writing, these taboos are variously euphemized, for example, by using a socially acceptable dash, cross and asterisk or by using pleasant-sounding homonyms or vague expressions. The fact that erotic literature previously banned is now easily accessible gives the impression that taboos under this category are on the wane. Many studies (Wan 2005; Pan 1988), however, show that sex remains one of the strongest taboo subjects. To talk about sex comfortably, therefore, it is advisable to learn about not only taboo words and expressions but also euphemisms, dysphemisms, and neutral or straightforward terms such as xìngjiāo (sexual intercourse). For example, connubial sex is euphemistically referred to as tóngfáng (share room), xíngfáng (perform room) and fángshì (room affairs) in polite situations. Zuòài or zàoài, most preferred in popular discourse, is an interlinear translation of ‘make love’, which itself is an English euphemism (Jay 2000: 126). Dating back to pre-Qin China, yúnyǔ (cloud rain) is perhaps the most poetic and ancient euphemism for xìngjiāo that prevails in romantic novels. New slang terms such as chǎofàn (fry rice) and gǔn-chuángdān (roll bed sheets) occur mainly in Chinese cyberspeak. A married person who has extramarital affairs is bùzhōng (unfaithful) and chūguǐ (derail). A promiscuous man is often depicted as fēngliú (wind stream), neutrally if not positively. In contrast, however, their female counterpart is dàngfù (dissolute woman), yínfù (wanton woman) or even pòxié (worn-out shoes), all intended to stigmatize females. Technology proves to have contributed to the creation of terms for sex, some of which may also become taboo. For example, most people may only conjure up the image of the piston movement when coming across huósāi-yùndòng. Surprisingly, however, Google’s results for a search on this term suggest that it is nowadays used overwhelmingly as a slang term for vaginal intercourse in cyberspace. Systematic research is needed to confirm this preliminary finding, though. Similarly, chēzhèn (car quiver) and chēzhènmén (car quivering door), perhaps coined by celebrity news writers, respectively refer to car sex and the ensuing scandal. In the film industry, sex scenes in movies, probably excluding pornographic ones, are normally referred to as chuángxì (bed play). Prostitution, ‘the world’s oldest profession’, is infamously described as màiyín (sell sex) or màishēn (sell body). It is, however, more positively known as shēngyì (business) among the insiders and sèqíng-fúwù (sex service) among the public. As regards words for brothels, the default term seems to be jìyuàn (prostitutes yard), but it was generally replaced by yáozi (kiln) and qīnglóu (green house). Visiting brothels is guàng-yáozi (visit kiln) or xúnhuā-wènliǔ (seek flowers and ask for willows). In contemporary Chinese hóngdēngqū (red-light area) is fast giving way to euphemisms such as fàláng (hair salon), xǐyù-zhōngxīn (spa centre) and sāngná-huìsuǒ (sauna club) (Pan 2000). 382

Chinese taboo

As in many other languages, jìnǚ (prostitute woman) and màiyín-nǚ (sell sex woman) are highly taboo. During feudal times prostitutes were often called fēngchén-nǚzǐ (wind-dust woman), qīnglóu-nǚ (green-house woman) or fēngyuè-nǚ (wind-moon woman). Xiǎojiě (young sister) was a polite term of address decades ago, but this term is no longer innocent now: It has undergone a process of semantic derogation over recent years, taking on the sense of ‘prostitute’. For example, saying zhǎo/yào-xiǎojiě (seek/need young sisters) would be unmistakably understood as asking for sex service. Chāng, chāngfù, chāngjì and jī (chick, rhyming homophonous with jì, ‘prostitute’) are all dysphemistic and offensive. Biǎozi is probably the most derogatory and insulting word for female prostitutes, comparable in force with the English ‘bitch’ or ‘slut’. Their clients, normally males, however, are positively stereotyped as fēngliú-cáizǐ (wind stream scholar) or fēngliú-tìtǎng, ‘talented in letters and unconventional in life’. UNAIDS (2012) noted that ànchāng (underground prostitute), xiǎojiě (young sister) and shīzú-fùnǚ (lost-foot woman) are discriminatory terms for female prostitutes: Nǚxìng-xìng-gōngzuò-zhě (female sex worker) should be substituted, and accordingly xìngfúwù-gòumǎi-zhě (sex service buyer) should be used instead of piáokè (whorehouse visitor). However, these ‘politically correct’ labels fail to gain ground in China probably because they do not appear as judgment-free as claimed – both legitimize prostitution as a profession. Male prostitutes were variously termed nòngchén, xiǎoguān and jīnán in ancient China (Li 2003). In the Ming and Qing dynasties they were replaced by xiànggong and xiànggū. For multiple reasons male prostitution appears to be on the rise in today’s China. There is evidence that wǔnán (male dancers) provide sex service, and hence sometimes function as nánjì (male prostitute), the counterpart of jìnǚ. According to Pan’s (1999) covered observation of red-light districts in South China, yāzi (duck) and èryé (second grandpa) are most frequently used for male prostitutes. Not surprisingly, female clients refer to their sexual activities as wányāzi (play with ducks) and bāo-èryé (contract second man). The most dangerous anatomical taboos across cultures may be sexual organs. To begin with males, jība (cock) is extremely vulgar, hence is rarely heard in daily life with notable exceptions, for example in swearing. Yīnjīng (hidden stem), along with nánxìng-shēngzhíqì (male sex organ), is a technical term for the penis. Vague references including xiàshēn (lower body), xiàtǐ (lower body) and sīchǔ (private part) are appropriate in most contexts. Colloquial and vague expressions such as dōngxī (stuff ), wányìr (stuff ) and nàhuàr (that thing) appear generic on the surface, but they are conventionally gender-specific, reserved for males only. A range of euphemistic metaphors are available. Some emphasize the penis’s symbolic function of masculinity: nángēn (man root), mìnggēn (life root) and yángjù (masculine tool). Bird metaphors include niǎo (bird), máquè (sparrow) and jījī (chick) (Li 2003). If weapon metaphors such as qiāng (gun), bàng (club) and biān (whip) appear too ‘belligerent’ or ‘aggressive’, then lǎoèr (second brother), dìdi (younger brother) and bǎobèi (treasure) convey affection. Regarding words for the female sexual organ, bī is one of the most strictly taboo, equivalent of the English ‘cunt’. This is in a way evidenced by the multitude of insulting terms such as chòubī (smelly cunt) and shǎbī (stupid cunt), which, according to Wan (2005), are sexist as no similar expressions are directed against males. In some romantic literature the female organ of sex is given poetic names such as mìtánɡ-guàn (honey pot) and xiāoyáo-dòng (fun cave), both comparing the vagina to a place (for men) to explore. In a word, the female sex organ is a xiūchù (embarrassing part) that needs to be covered with ‘linguistic fig leaves’. Since many Chinese characters are hieroglyphic and ideographic, some words for sex are taboo due to both obscenity and the ‘dirty’ visual images they ‘draw’. Most typically, many 383

Amy He Yun

Chinese would find it extremely embarrassing to tell how taboo words such as bī, cào and diǎo are structured. For example, the character bī is composed of two radicals, the upper shī denoting the body and the lower xué meaning a hole or cave. Similarly, the upper radical of cào is rù, which, symbolic of the penis, means ‘insert’, while ròu at the bottom means ‘meat’ and symbolizes the vagina. These words are strictly taboo and restricted in use, usually in insulting, swearing or toilet graffiti. Due to their shock value, these characters are not at all collected in popular dictionaries such as Xinhua-zidian and Xiandai-hanyu-cidian, nor can they be entered by using some common Chinese input methods. Interestingly, in writing, the homonym bī (force or compel) with a completely different denotative meaning is thus used euphemistically: niúbī (cool, stunning, arrogant) and shǎbī (stupid, idiot; stupid cunt). Since foreign languages sound finer (Harris et al. 2003; Rawson 1981), ‘hybrid words’ such as niúB (abbr. NB) and shǎB (abbr. SB) were also ‘invented’ by juxtaposing the second letter of the English alphabet and Chinese bī. In both cases, the dirty image the Chinese characters represent is rendered invisible. By tradition, erotic pictures are named chūngōng (spring palace), chūncè (spring book) and mìxìtú (backyard-game drawings). Sex activities in public are thus reported in the media as huó-chūngōng (live erotica). As Li (2003) argues, chūn (spring) has stood for human fertility since antiquity. According to the author, this is evidenced by the term chūncháo (spring tide), designated for the first menstruation. By the end of the Ming, the author maintains, the word had acquired the meaning of sex. At this point, an analysis of the second slogan mentioned at the beginning of this article is in place. It is controversial because the compound jiàochūn means the meowing and yowling of female animals, particularly a queen cat on heat to highlight her existence for mating purposes. In this light, the taboo sense becomes salient and the slogan is widely condemned as léirén (thundering people), i.e. shocking.

Taboo, diseases, mental and physical defects, and death Health is the biggest asset one can possess, all else being peripheral. Linguistically, any words for or reminiscent of illness, mental and physical defects, and death may be perceived as taboo. This necessitates employment of avoidance strategies if one does not hope to violate taboos and get punished. To begin with, bìng or jíbìng is the straightforward and general word for any illness. As noted in many studies (Wan 2005; D. Wang et al. 1995), for fear of diseases or ‘to reflect the cooperative desire not to impose one’s troubles on others and not to be seen to whinge’ (A&B 2006: 203), speakers of Chinese resort to a range of strategies. Understatements include qiàn’ān (unwell), yǒuyàng (indisposed), bùshì (unwell), and bùshūfu (uncomfortable), among others. Metaphors are also what the Chinese live by. The most notable example is perhaps bùxīnxiān (unfresh) (Wan 2005), comparing healthiness to fresh air or food. An equally interesting example is the use of chá (tea) to replace yào (medicine) as in hēchá (drink tea), hē-hǎochá (drink nice tea) or hē-xìchá (drink fine tea) (cf. Wan 2005). In respect of motivations, the metaphoric use of chá may indeed be related to the Tang dynasty’s view of tea as medicinal, it can arguably help distance users psychologically from the disease. The most dreaded of diseases typically go by a number of different labels. Xìngbìng, the sexually transmitted disease (STD), is a notable example. Known as nányán-zhīyǐn (unmentionable concern), STDs such as syphilis, gonorrhoea and AIDS (though not necessarily acquired through sex) are masqueraded ‘poetically’ as huāliǔbìng (flower-willow disease). Clearly, shèhuìbìng (social disease) is favored by those who hope to voice their view 384

Chinese taboo

that society at large should be held more accountable. Another equally unspeakable disease is male sexual impotence. The medical jargon bóqǐ-gōngnéng-zhàng’ài, ‘erectile dysfunction’, is circulated among health professionals. The colloquial term yángwěi (masculine wither) is also used in clinical contexts quite often, but its dysphemistic associations such as ‘unmanliness’ make the patient a convenient target for ridicule and disparaging humor. On the contrary, face-anointing references such as bùxíng (no good) would be appreciated. As well as taboos on diseases, talking about mental and physical defects or incapacity such as deafness, blindness, some form of disability or handicap is almost always inappropriate. A rule is dānɡzhe-ǎirén-biéshuō-duǎnhuà, literally, not to say any words that may suggest ‘short’ before someone who is apparently undersized. For example, words or expressions such as duǎn (short) and ǎi (short), mùdūn (wood stump) and Wǔdàlán, the name of a dwarf character in the fiction Outlaws of the Marsh, should all be used with great care in such a situation. In a similar vein, rather than using offensive words such as bǒ (lame) and qué (crippled), preference is given to understatements bùlíngbiàn (not mobile), bùdéjìn (not strong) and bùlìsuo (not nimble). For a similar reason, it is very likely that mentioning dēngpào (light bulb) or héshangtóu (monk-head) will be perceived by one who has gone bald as highly offensive unless intended as a ritual insult between intimate friends. The above ‘golden rule’ holds true in many other similar situations. For instance, zhòngtīng (heavy hearing) was used for lóngzi (deaf ) in the past whereas they are often replaced by shīcōng (lost hearing) and ěrduobèi (hard of hearing) in Modern Chinese. Finally, it is worth noting that xiāzi (the blind), lóngzi (the deaf ), and yǎba (dumb person) are frequently used as abusive epithets, which itself evidences the delicacy of these terms. Death, like diseases, is a typical fear-based taboo. The importance of knowing this taboo phenomenon is best summarized by Pound: ‘One of mankind’s gravest problems is to avoid a straightforward mention of dying or burial. Every ingenuity is practiced to find words which shroud the idea of death’ (1936: 195). In Chinese, according to D. Wang et al. (1995), there are more than 200 euphemistic words and expressions for death. In the classic novel A Dream of Red Chambers alone dozens of evasive or polite terms for death are identified (Zhuang 2008). Roughly speaking, familiar figurative expressions such as bùzàile (no longer exist), zǒule (left) and qùle (gone) do not make clear value judgments about the defunct and, therefore, can be used in most cases. Nevertheless, as Pound noted above, more often than not finding an appropriate way of expressing death can be very challenging. This is because, as illustrated below, multiple factors such as social status, faiths and beliefs, age, gender, marital status, and value judgment need to be taken into account when talking about death. First, as a reflection of the hierarchical social structure of ancient China, a distinction was made between words for death of people occupying different positions on the social ladder. This is noted succinctly in the Liji that 天子死曰 ‘崩’,诸侯死曰 ‘薨’,大夫死曰 ‘卒’,士曰 ‘不禄’,庶人曰 ‘死’: The death of the Son of Heaven is expressed by bēng; dukes, hōng; officials, zú; intellectuals, bùlù; and the grass roots, sǐ. Thus, as shown in Figure 21.1, sǐ (die), the straightforward word, is for the death of ordinary people at the bottom of the pyramid. Death of the elite classes, however, is further variously worded depending on their social positions. Precisely, the lowest ruling class called shì (intellectuals) bùlù (no longer salaried), whereas the lower-middle class known as dàfū (senior officials) zú (meet their end) when they die. Finally, the onomatopoeic words hōng and bēng are respectively for zhūhóu (dukes and princes) and tiānzǐ (Son of Heaven, king/ emperor), both of which emphasize the tremendous impact of their death as the collapse of massive mountains would have. Therefore, what matters most in expressing death in Ancient China was to ensure the correct match between the word and the status of the dead. This 385

Amy He Yun Words for death

Social classes tiānz (Son of Heave or emperor)

bēng (fall) hōng (crash)

zhūhóu (duke and prince)

zú (meet one’s end)

dàfū (senior official)

bùlù (no longer salaried)

shì (intellectual)

s (die)

shùrén (grass root)

Figure 21.1  Pyramid hierarchy of words for death of people from different social classes in historical Chinese

can be interpreted as part of Confucius’s effort to zhèngmíng, i.e. rectify names, in order to restore the social hierarchy and order of the slavery system of the Zhou dynasty. Moreover, as illustrated above, the social classificatory function of taboos, albeit not necessarily hierarchical, is also manifested in other areas. Second, faiths and beliefs also significantly impact the way death is expressed. A notable example is the use of nièpán, yuánjì, jìmiè, mièdù, or zuòhuà, different Chinese renderings of Buddhist nirvana, ‘blown out’ as in a candle. More interestingly, according to the Buddhist canon Dàzàngjīng (Tripitaka), a human being dies with one of the 12 ‘stages of achievement’ ranging from wúyúsǐ or nièpán (attaining the supreme goal) to jī’èsǐ (death from hunger) of the least achieved. Expressing death appropriately in Daoism can also be a challenge: As in Buddhism, shījiě (body disappears) and yǔhuà (feather disappears) or chántuì, as in the metamorphosis of a nymph into a cicada or butterfly, are used to show Daoist belief in the immortality of life. They are thus used to show respect for Daoists who made significant contributions. Christianity also finds its way into Chinese: jiàn-shàngdì (called to God) (He 2002; Chen 2001[1980]). In a similar way, the introduction of Marxism, basically a theory and practice of socialism, into China in the early twentieth century led to the invention of jiàn-Mǎkèsī (called to Marx), obviously modeled after the aforementioned Christian figurative expression. This term is also used, more or less humorously, in popular parlance. To mention in passing, traditional Chinese theists would use metaphors of departure such as positive guītiān (gone to Heaven) and negative jiàn-yánwang (called to King of Hell), whereas an atheist may simply use flippant expressions such as dēngjiǎo (kick legs) and qiàobiànzi (raise braids), rough equivalents of English ‘kick the bucket’. Third, many taboo words and expressions for death are value-loaded and cannot be used randomly. They are roughly of three types: positive, negative and neutral in terms of value judgement. First, words such as xīshēng (sacrifice life), yùsuì (break jade), jiùyì (be executed for championing a just cause) and xùnzhí (martyred at one’s post) are used appreciatively. In this group, shìshì (pass away) only appears in obituaries of and tributes to important people. Then, bìmìng (meet violent death), jìn-guāncai (enter the coffin), xiàdìyù (descend to Hell), sàngshēng (lose life) and sàngmìng (lose life) are dysphemistic. This distinction suggests that use of sǐ is well-founded on occasions. 386

Chinese taboo

Fourth, the choice of terms for ‘death’ is also determined partly by the age of the dead. Yāo and shāng, for instance, means ‘die prematurely’; xiōng, ‘die at the age of seven or eight’; and duǎn, ‘die before 12’. By contrast, xiānshì (gone with the immortal), zuògǔ (become history) and shòuzhōng-zhèngqǐn (die in bed of old age) refer to the natural death of old people. Fifth and finally, metaphors that objectify women such as xiāngxiāo-yùyǔn (fragrance disappears and jade breaks), zàngyù (bury jade) and máixiāng (bury fragrance) are exclusively reserved for women, predominantly in romantic literature. Fear and anxiety also give rise to many death-related taboo words and euphemisms (He 2002). Most typically, sāngshì (funeral affairs) is given the name báishì (white affairs). Further, as shown by the idiom hóngbái-xǐshì (red-white happy affairs), ‘funeral and wedding rites’, the funeral is presented as something pleasant. Similarly, guāncai (coffin/casket) is euphemized as shòucái (longevity material), and shòuyī (longevity clothing) has largely become an unconscious euphemism for grave-clothes. With adjunct materials about the departed rendered warmer by highlighting longevity of the deceased, hòurén (behind people) or the bereaved may feel more comfortable in coping with the departure of their beloved. The most ‘serious’ and ‘weird’ contamination of death taboo is on people’s superstitious reluctance to use words that sound similar to sǐ (death). The irrational fear of the number sì (four), known as tetraphobia, is most representative. In Chinese (also in other East Asian languages such as Korean and Japanese), the number is considered unlucky due to phonetic similarity. Some buildings, for example, have no fourth floor or room. Some customers reportedly complain about their vehicle registration number ended with sì or more unlucky digit configurations such as wǔsì (five four) that sounds like ‘I die’ (Guo 2004; Shen 1996). Recent studies (Fortin et al. 2010) show that tetraphobia even has significant effects on North American housing markets.

Taboo, naming and addressing The name of a person is an inalienable part of her identity. Hence, ‘an assault on one’s name is treated as comparable with, or even worse than, an assault on one’s body’ (A&B 2006: 125). This is believed to be the root reason for taboos on names (Frazer 2012[1922]; D. Wang et al. 1995). This applies in all societies, but what is known as huì (taboos on names) appears peculiar to Chinese culture. As a considerable number of the taboo names show deference to the name-bearers, they are called jìnghuì or honorific taboos. Consequently, any naming, name-calling and addressing that potentially stigmatizes a person is subject to taboo and censorship. It has been 3,000 years since taboo in Chinese culture was first institutionalized. Originating in the Zhou dynasty, the practice of name taboos was first formally institutionalized in Qin–Han China, culminating in the Tang and Song. It was then exploited to the full by the Qing emperors (Chen 1997[1928]; Y. Wang 1997; Shen 1996; Goodrich 1935). Name taboos are generally distinguished between guóhuì (national taboos) and jiāhuì (family taboos), both of which should be avoided, i.e. bìhuì (avoidance of taboo names) in Chinese nomenclature. To begin with, guóhuì proscribes using without modification the names of emperors, designations of reigns and sometimes their homonyms. Taboos of this type, as the term suggests, had to be avoided nationally and are hence also known as gōnghuì (public taboos). The number of nearly 1,000 name-related taboo characters collected in Y. Wang’s (1997) book is staggering. Perhaps the most striking case of taboo in Chinese history is that as many as 55 words phonetically identical or similar to gòu were ever banned merely to avoid the given name of Zhao Gou of the Song dynasty (Qu 1996). It is necessary to note that Chinese given 387

Amy He Yun

names are also commonly composed of two (or more) characters such as Shìmín of Emperor Taizong. In this case, avoidance of the constituent shì (world) or mín (people) was not compulsory if they occured separately, with the notable exception of Guānyīn for Guānshìyīn, i.e. ‘bodhisattva’ in Buddhism. In addition, some sacred or holy names also come under national taboos. This applies, in particular, to Confucius and Zhao Xuantan, whose importance was considered comparable with that of emperors. Therefore, in some historical texts, for example, the last stroke of Confucius’s given name was dropped to avoid the proper qiū. In others it is replaced by mǒu, which denotes an indefinite someone. Similarly, Zhao Xuantan, a folk deity in Daoism, has been venerated as ‘god of wealth’ in China since Qin. The taboo on this name is such that saying mǎi (buying) is considered profane when buying a picture of the god. Rather, yāoqǐng (inviting) should be used (Shen 1996). A distinctive feature of Chinese name taboo lies in its legislation, dating back to the Sui–Tang period. This is most clearly recorded in the Sui Code and Tang Code, in which breaking name taboos is named dàbùjìng-zuì (big disrespect crime), listed along with rebellion and treason as one of the ‘Ten Crimes’. Failure to abide by the law was criminal and would inflict severe punishment in some cases, including beheading. An extreme example is the infamous literary inquisition in Chinese history known as wénzìyù (word imprisonment), referring to the persecution of intellectuals for writings of whatever genre that were perceived by emperors, often abetted by his intimates, as transgressing taboos that were not necessarily the conventional ones. These ‘cultural disasters’ took place in almost every dynasty, the earliest case found in Han China (Shen 1996). Moreover, not only the perpetrator, but also his/her immediate and extended families, as well as others implicated, would be killed if the charge of transgressions was ‘proved’, often at the mercy of emperors. Most notorious cases of wénzìyù were ‘tried’ in the Qing dynasty, including the death of Xu Jun during the Kangxi reign for allegedly violating Manchu rulers’ taboo on míng (bright), the designation of the previous Han Chinese regime. Unlike guóhuì, which requires national obeisance, jiāhuì (family taboo) or sīhuì (private taboo) is restricted to the clan. It refers to taboo prohibition against straightforward use of the names of one’s parents and ancestors. In a patriarchal society where women were viewed as subordinate to men, only names of paternal ancestors were subject to taboo. This, as some scholars (Shen 1996) argue, is consistent with Confucian emphasis on the hierarchical social relation in China. For example, in Shiji (Historical Records), the first general history of its kind in China, tóng is substituted by Sima Qian for his father’s name tán. An entertaining example concerns Li He, a talented poet of the Tang. He was assessed as ineligible to take the highest imperial examination for jìnshì only because jìn 进 (enter), one of the characters for the academic title, sounds similar to his father’s name jìn 晋 (promote), hence was judged as violating his family taboo. His contemporary Han Yu, a precursor of neo-Confucianism, criticized this as ridiculous by writing the essay Bianhui (‘Critiquing taboo’), the most compelling argument of which is reproduced below: 父名晋肃,子不得举进士; 若父名仁,子不得为人呼？ The son is barred from sitting the examination for jìnshì only because of his father’s name Jìnsù. Then cannot he be rén (human) if his father’s name was rén (benevolence)? Taboos on naming and addressing have, in a large sense, been suggested by the national and family taboos: People were precluded from using names of their fathers, forefathers and 388

Chinese taboo

emperors, and also from addressing them by these names. This is no longer practiced today, but it is still considered inappropriate or rude to call the elderly including one’s relatives by their names. There are also many other specific taboos against naming and addressing. For example, Qin Shi Huang, the first emperor in Chinese history, decreed his exclusive use of the first-person singular pronoun zhèn, whereas his subjects were only allowed to use other self-reference terms such as wú. This hierarchically based taboo system was inherited by the succeeding dynasties and did not disappear until the fall of the Qing dynasty in the early 1900s when wǒ (I) was promoted as a universal self-reference term. There are linguistic prescriptions and proscriptions concerning names and naming as before, but many differ in nature from those in history (Wan 2005; Shen 1996). For example, traditionally characters associated with strength, wealth and status such as qiáng (strong) and fù (wealthy) appear too masculine for women. Conversely, plant and flower names such as jú (chrysanthemum), méi (plum) and huā (flower) are popular among females only. However, as already noted, Chinese speakers now have the freedom to choose names and terms of address. Anecdotally, for example, my eldest sister’s given name yǒng (brave) encodes my parents’ hope that she could be as ‘manly’ as men. Changes in addressing are especially fundamental in the way Chinese leaders are addressed. For example, Deng Xiaoping, the ‘architect of China’s reform’ has often been addressed as ‘Xiaoping’ since 1984. Similarly, the former Chinese Premier Wen Jiabao became bǎobao (baby), a term of endearment in the media (Zhang and Li 2009), as opposed to strict national taboos in feudal China.

Taboo, other aspects of human existence There is always some kind of linguistic prohibition against any aspect of human existence and behavior. This subsection is thus devoted to taboos that do not fit well into the above ‘classic’ taboo topics. Due to space limitations, my discussion focuses selectively on subjects such as gift-giving, occupations, and food. Let us start by addressing the first question I raised at the outset of this article. As demonstrated below, now-standard words such as kuài for ‘chopsticks’ entered the Chinese lexicon so long ago that they have become ‘unconscious euphemisms’ that people hardly remember the original motivation (Rawson 1981: 3): 民间俗讳,各处有之,而吴中为甚。如舟行讳住讳翻,以箸为快儿,幡布为抹布;  讳离散,以梨为圆果,伞为竖笠。(明•陆容《菽园杂记》) Taboos are ubiquitous, particularly in Wu. For example, the words zhù (stop) and fān (capsize) are taboo in sailing, hence zhù (chopsticks) is replaced by kuàir (fast), fānbù (sail) by mābù (dusting cloth); lí (parting) and sàn (break-up) are taboo, so lí (pear) is called yuánguǒ (round fruit) and sǎn (umbrella), shùlì (vertical bamboo hat). (Shuyuan Zaji by Lu Rong of the Ming dynasty) It becomes clear that zhù had been the word for chopsticks prior to kuàir (N.B. ‘r’ is a noun suffix). This was initiated by fishermen because they hoped to sail fast (kuài) and smoothly but not to stop (zhù) or get stranded during fishing expeditions. We also learn that other euphemisms including mābù for sail were taboo-induced and have long become unconscious and collective. Like fishing, many other pursuits that often put life in danger such as war, mining, and hunting are beset with taboos and euphemisms (Wan 2005). A war wound, for example, is euphemistically portrayed as guàcǎi (festooning) or dàihuā (bearing flowers). 389

Amy He Yun

Linguistic taboo impacts greatly on social behaviors. Most strikingly, we are often cautioned against giving certain presents lest the behavior violate certain word taboo. For example, clocks are widely deemed as inappropriate gifts because sòngzhōng (give clocks) sounds the same as a homonym that means ‘attend upon the dying’. Similarly, most Shanghai dialect speakers would choose not to visit patients in hospital with apples, which sound like bìnggù (die of disease) (Shen 1996). A number of food-based taboos have also entered Chinese daily parlance. They are basically of two types, associated with obscenity and bad luck. Firstly, believed to have erotic properties, foods in the first group, particularly the genitalia of male animals and poultry, are inextricably linked with aphrodisiacs and hence consumed to enhance fertility or to boost sexual performance. Hence, in consistence with taboo on sex and sex organs, some food names are considered taboo. For example, the testicles of cockerels, boars, goats, bulls and fish are served in some restaurants, but terms such as luǎn (egg), wán (pellet), zǐ (seed), yāo (kidney) and dàn (egg) are more likely than gāowán (testicles), shēngzhíqì (reproductive organ) and xìngqìguān (sex organ) to appear on the menu. Similarly, the penis of many animals such as donkeys and deer are also considered powerful aphrodisiacs. Yet, when used in Chinese medicine and cuisine, any other words or expressions except for biān (whip) should be used carefully. In some Chinese communities, such as Taiwan, cooked chicken tails are served as a local specialty. Instead of pìgu (chicken buttocks), it is given a ‘mouthwatering’ name qīlǐxiāng (seven miles aroma), possibly because the ‘default’ name might invoke sexual activities in which this body part is engaged or just because they want to make something that is intrinsically unappetizing or mundane sound better by giving it a nice name (A&B 2006). A particularly interesting taboo on food concerns eggs, especially those from chickens: Polysemous dàn (egg) refers simultaneously to testicles. Its tabooness and offensiveness is in a way evidenced by swear words or insults such as húndàn (bastard), bèndàn (scumbag/ dumb), wángbādàn (bastard). Consequently, jīdàn is disguised as báiguǒ (white fruit) or shuǎiɡuǒr (throwing fruit) in Beijing dialect, in which the egg is metaphorized into ginkgo seeds or suaveolens. Thus, visitors to the city are very likely to find on their hotel menu dishes such as chǎobáiguǒ, chǎomùxī and mùxītāng (suaveolens soup) instead of chǎodàn (fried egg) and jīdàntāng (egg soup). Human beings are no exception to this ‘crosscontamination’: An unlucky person is dǎoméidàn. The second type of taboo on food is against words for certain (cuts of ) meat with negative associative meanings. For example, contaminated by shé as in shéběn ((businesses) lose money), shé (tongue) is believed to pòcái (suffer financial loss). Lucky words including lì (profit) and zhāocái (draw money) are thus substituted. One would see zhūlì (pig profit) and zhūzhāocái (pig drawing money) rather than the straightforward zhūshé (pig tongue) in Canton’s and Sichuan’s markets. Another example concerns taboo words sǐ (die) and/or shā (kill): Whole fish and ducks ready for cooking are respectively called wényú (motionless fish) and wényā (motionless duck).

Variation of taboos and avoidance of punishment for taboo violations Variation of taboos is richly documented in Chinese texts. In addition to historical texts such as the Liji and Shuyuan Zaji, a wealth of current literature also demonstrates that taboo customs, verbal and non-verbal, vary across time and space (Wan 2005; Ren 2004; Chen 2001[1980]). This section sketches out how taboo words vary diachronically and synchronically. 390

Chinese taboo

Moreover, as it is not always easy to follow taboos, a brief discussion on ways to escape punishment for violations is also offered. To begin with, taboos vary diachronically (Wan 2005; Shen 1996). This was largely dealt with in Section 2. To highlight again, as an important aspect of diachronic variation, taboos on names and naming in ancient and modern China differ fundamentally in function and severity. For example, the main function of both national and family name taboos in history was to maintain social hierarchy. With the name bearers holding higher social status, the general public were living in a state of fear. This asymmetrical practice contrasts markedly with present-day emphasis on egalitarianism and reciprocity as evidenced by the increased first name-based addressing. Section 2 has also indicated clearly that generally name taboos in history no longer apply today. A further illustration is the diachronic change of cūn (village): 避‘村’名犹甚于避庙讳。盖中都人以外人为‘村’ ,故讳之。(宋•叶绍翁  《四朝闻见录》   ) The taboo on cūn (village) is stricter than that on the emperor’s (fore)fathers. It is so tabooed probably because city dwellers consider rural outsiders cūn (provincial). (Sichao wenjianlu by Ye Shaoweng of the Song dynasty) It thus appears that the taboo on cūn is a reflection of the rapid urbanization in the Song dynasty, during which villagers in the country became the object of contempt. The word acquired negative senses such as ‘foolish’, ‘gross’ and ‘rude’. However, cūn (village) is not at all a taboo word in Modern Chinese although some of these senses still remain as in cūnyě-bǐfū, i.e. ‘blackguard’. Synchronically, taboos vary considerably across occupations, places, dialects, and ethnic groups, past and present (Wan 2005; Ren 2004). Differences between dialects, social and geographical, are probably most studied. For example, dàn (egg) and qiú (ball) as discussed above are taboo mainly in the Beijing dialect. Similarly, in the Shanghai dialect é (goose), homophonous with wǒ (I), is euphemistically called bái-wūguī (white tortoise) so as to sidestep the confusion between shā’é (kill goose) shāwǒ (kill me). The last example is hǔ (tiger): It is alluded to as dàchóng (big bug) or māo (cat) in many places, this taboo is especially strict in Changsha due to recurrent tiger attacks in history. Another important dimension of taboo variation concerns ethnic groups, each of which has their own distinctive cultural heritage, ancestry, totem, faiths, beliefs and folklores (Wan 2005; Ren 2004; Song and Xu 1999). With 56 officially recognized ethnic groups in China, it is plausible to assume that some taboo words and expressions are group-specific. This is best illustrated with taboos from Muslim Hui and Tibetan communities. Islam, for example, proscribes pork; zhū (pig) thus becomes the strictest taboo word among Hui and other Muslim groups such as Uyghur in west China. The pig is believed to be dirty and execrable, and consequently various euphemisms like hēi-shēngkou (black beast), hēihuò (black goods), and hēi-chóngzi (black bug) arose (Gladney 1996). Further, some Hui with the surname Zhu, homophonous with zhū (pig), have even changed their surnames to Hēi (black). More interestingly, to observe the Buddhist taboo on taking life, conservative Tibetans do not even mention the word ròu (meat) reminiscent of shā (kill). Their solution is to use roundabout expressions such as dōngtiān-chīde (something to eat in the winter) (Wang 1990). It is forbidden to say shā among both Muslims and Tibetans, but diverging euphemisms zǎi (destroy) and zhuā (grab) are used in respective communities. These not only illustrate the rich variation of word taboos but also provide a window into colorful ethnic cultures. 391

Amy He Yun

Before concluding this section, it is crucial to outline verbal means to get away with punishments for taboo violations. Inherent in a taboo is the assumption that both accidental breach and intentional defiance of the taboo will be followed by some kind of trouble as observed below: Infractions of taboos can lead to illness or death, as well as to the lesser penalties of corporal punishment, incarceration, social ostracism or mere disapproval. Even an unintended contravention of taboo risks condemnation and censure. (A&B 2006: 1) Some of these punishments have already been illustrated or suggested. For example, as well as financial loss, the tourism authority was forced to issue an apology after one of its slogans (see Section 1), albeit eye-catching, was perceived by many as indecent in adverts (Zou 2012) because kào is interpretable as a taboo slang used to the effect of English ‘fuck’. In the Chinese cultural context, as noted by Ren (2004), the perpetrator can be punished immediately, in this life or the next. In some cases proscription is the only way to avoid this danger; in others it can be overcome through ritual, for example, cleansing after consecrating the dead. Verbally, however, it is believed that people who inadvertently violate a taboo can often purify themselves or be purified by uttering a mantra that is considered to possess mystical or spiritual efficacy. In Chinese five mantras are usually chanted to defuse the danger represented by the taboo depending on contexts. First, if the perpetrator is a child the guardian can say tóngyán-wújì (child’s speech no taboo) to dispel the malevolent spirit; if the violator is an adult, she can recite jiāngtàigōng-zàicǐ, bǎiwú-jìnjì (Jiangtaigong present, nothing taboo) (Wan 2005; D. Wang et al. 1995). Second, the first five characters from the Yijing, i.e. qián-yuán-hēng-lì-zhēn, are believed to be magic and can be uttered to avert the possible misfortune (Wan 2005; D. Wang et al. 1995). Third, when transgressing a taboo Tibetan Buddhists would be ready to recite the six-syllabled mantra Om-Ma-Ni-Pad-Me-Hum, which is believed to encapsulate all Buddhist teachings. Finally, non-Tibetan Buddhists would recite the sacred utterance Ēmítuó fó to wash off the sin committed.

Concluding remarks What has been presented above hopefully fulfills well my goal of providing a general introduction to Chinese taboo. To summarize briefly, it is at least four millennia since the earliest ideas about taboo were recorded in writing. As already shown, some major stages of the historical development can be discerned, each characterized by salient features (see Section 1). Topically, Chinese taboo is mainly concentrated in five subjects (see Section 2). Words and expressions are tabooed and euphemized or substituted for complex and various reasons. Generally they are inexplicably linked to a wide range of notions such as uncleanness, impurity, filth, obscenity, distaste, fear, danger, holiness, bad luck, respect or politeness. Moreover, taboo varies considerably diachronically and synchronically. Amongst many characteristics of Chinese taboo, a point that deserves highlighting is the social hierarchy and institutionalization that is manifested most obviously in taboo names and words for death in ancient China. It is also important to note that Chinese taboo, primordially the linguistic aspect of custom, was often exploited by dynastic rulers as a form of social control, which is typified by recurrent literary inquisitions spanning about 2,000 years. This political and cultural despotism contrasts sharply with the freedom and democracy we enjoy today in naming and addressing. 392

Chinese taboo

The last but absolutely not the least point I want to make is the pragmatics of taboo – nothing is taboo for all people, under all circumstances, at all places, for all time. As noted earlier, this article is essentially sociolinguistic in perspective while drawing on a range of (sub)disciplinary traditions (see Section 1). Consequently, the question of what counts as a taboo word or expression in specific interaction is left almost untouched. This falls squarely within the domain of pragmatics and recently (im)politeness studies such as Christie (2013), Culpeper (2011) and Jay and Janschewitz (2008) explore the interpersonal aspect of language taboos, particularly English swearwords. Basically, ‘there is no such thing as an absolute taboo that holds for all worlds, times and contexts’ (A&B 2006: 27). This is partly intended as a belated reminder for readers of this chapter. Moreover, to the best of my knowledge, comparable studies of Chinese taboo are still lacking. Attention is, therefore, needed to be accorded to this new research direction in order to shed fresh light on the age-old linguistic phenomenon under discussion.

References Allan, Keith and Kate Burridge (2006) Forbidden words. Cambridge: Cambridge University Press. Chen, Yuan (1997[1928]) Shihui juli (‘Historical taboos illustrated’). Shanghai: Shanghai Shudian chubanshe. Chen, Yuan (2001[1980]) Yuyan yu shehui shenghuo (‘Language and social life’). Taibei: Taiwan shangwu yinshu guan. Christie, Christine (2013) The relevance of taboo language. Journal of Pragmatics 58: 152–69. Culpeper, Jonathan (2011) Impoliteness. Cambridge: Cambridge University Press. Douglas, Mary (2003) Purity and danger. London: Routledge. Fortin, Nicole M., Andrew J. Hill and Jeff Huang (2010) Superstition in the housing market. Retrieved from http://faculty.arts.ubc.ca/nfortin/Superstition.pdf. Frazer, George (2012[1922]) The golden bough. New York: Cambridge University Press. Freud, Sigmund (2004[1913]) Totem and taboo, trans. J. Strachey. New York: W. W. Norton. Gladney, Dru (1996) Muslim Chinese. Cambridge, MA: Harvard University Asia Center. Goodrich, Carrington (1935) The literary inquisition of Ch’ien-Lung. Baltimore: Waverly Press. Guo, Xi (2004) Zhongguo shehui yuyanxue (zengding ben) (‘Chinese sociolinguistics (revised edition)’). Hangzhou: Zhejiang University Press. Harris, Catherine, Ayse Aycicegi and Jean Berko Gleason (2003) Taboo words and reprimands elicit greater autonomic reactivity in a first language than in a second language. Applied Psycholinguistics 24(4): 561–79. He, Yun (2002) Hanying yuyan jinji de shehui yuyanxue toushi (‘Chinese and English taboos from a sociolinguistic perspective’). Xueshu tansuo 2: 146–48. Hughes, Geoffrey (2006) An encyclopedia of swearing. New York: M. E. Sharpe. Jay, Timothy (2000) Why we curse: A neuro-psycho-social theory of speech. Amsterdam: John Benjamins. Jay, Timothy and Kristin Janschewitz (2008) The pragmatics of swearing. Journal of Politeness Research 4(2): 267–88. Li, Liyan (2003) Nanren nüren (‘Men and women’). Beijing: Zhongguo Shehui Chubanshe. Lu Xun (2005) Lu Xun quanji (‘The complete works of Lu Xun’). Beijing: Renmin Wenxue Chubanshe. Pan, Suiming (1988) Shenmi de shenghuo: Xing de shehui shi (‘The mystery of holy fire: A social history of sexuality’). Zhengzhou: Henan Renmin Chubanshe. Pan, Suiming (1999) Cunzai yu huangmiu: Zhongguo dixia xingchanye kaocha (‘Reality and absurdity: Investigating the underground sex industry in China’). Beijing: Qunyan Chubanshe. Pan, Suiming (2000) Shengcun yu tiyan: Dui yige hongdengqu de zhunzong kaochao (‘Survival and experience: The follow-up investigation of an underground ‘red light district’). Beijing: China Social Sciences Press. Pound, Louise (1936) American euphemisms for dying, death, and burial. American Speech 11: 195–202. Qu, Yanbin (1996) Zhongguo minsu yuyanxue (‘Chinese “folklore” linguistics’). Shanghai: Shanghai Wenyi Chubanshe.

393

Amy He Yun Rawson, Hugh (1981) A dictionary of euphemisms and other doubletalk. New York: Crown Publishers. Ren, Cheng (2004) Zhongguo minjian jinji (‘Folk taboos in China’). Beijing: China Social Sciences Press. Shen, Xilun (1996) Yuyan wenzi de bihui, jinji yu weiwan biaoxian (‘Avoidance, taboo and euphemism: The linguistic realizations’). Taipei: The Commercial Press. Song, Quan and Xu Wanbang (eds) (1999) Zhongguo shaoshu minzu minjian jinji (‘Taboo customs of the minority ethnic groups in China’). Beijing: Minzu University of China Press. Steiner, Franz (2013[1956]) Taboo. Abingdon: Routledge. UNAIDS (2012) Lianheguo aizibing guihuashu: Aaizibing xiangguan yongci zhinan (‘UNAIDS: Terminology guidelines’). Retrieved from http://unaids.org.cn/pics/20130104152113.pdf. Wan, Jianzhong (2005) Zhongguo minjian jinji fengsu (‘Taboo custom in China’). Beijing: Zhongguo Dianying Chubanshe. Wang, Dechun, Sun Rujian and Yao Yuan (1995) Shehui xinli yuyanxue (‘Socio-psycholinguistics’). Shanghai: Shanghai Jiaoyu Chubanshe. Wang, Jian (2002) Zhongguo gudai bihui shi (‘A history of word taboo in ancient China’). Guiyang: Guizhou Renmin Chubanshe. Wang, Qingshan (1990) Zangyu zhong de jinjiyu yu weiwanyu (‘Taboo and euphemism in Tibetan’), Qinghai Minzu Yanjiu 1: 59–65. Wang, Yankun (1997) Lidai bihuizi cidian (‘Dictionary of taboo words in history’). Zhengzhou: Zhongzhou Guji Chubanshe. Zhang, Shuo and Li Peng (2009) Wen Jiabao fangwen jianqiao xuesheng dachu ‘woai baobao’ hengfu (‘Chinese students at the University of Cambridge put up a slogan reading “I love baobao” to welcome visiting Premier Wen Jiabao’). Retrieved from http://news.xinhuanet.com/edu/2009-02/03/ content_10753678.htm. Zhuang, Weiguo (2008) Hong Lou Meng liang yingyiben siwan weiwanyu fanyi de wenhua bijiao (‘A contrastive study of words and euphemisms for death from two English versions of Dream of the Red Chamber’). Yancheng Shifan Xueyuan Xuebao 28(2): 87–90. Zou, Tongqian (2012) Dui dongfang lüyou mudidi yingxiao zhihui de fansi (‘Reflections on the wisdom of marketing Oriental tourist destinations’). Beijing di’er waiguoyu xueyuan xuebao 7: 81.

394

22 Chinese xiehouyu 1 Grace Zhang Curtin University, Australia

1. Introduction For any popular linguistic expression needing wit and a grass-root base, xiehouyu (歇后语) fits the bill. Xiehouyu are Chinese metaphorical folk sayings which follow a two-part pattern of metaphor (vehicle) + intended message (destination). The two parts are separated by a pause in speech or a comma/dash in writing, providing time for or drawing attention to the intended meaning to be inferred and enhancing the effect of humour or satire. While some xiehouyu do not follow the above pattern, they are limited in number. According to Wen’s 2004 dictionary, there are at least 85,000 xiehouyu. They are used when ‘one wants to display wit or humour, often disparagingly when describing a person, a thing, or a situation’ (Walls and Walls 2009: 133). Their expressive form and witty content enable xiehouyu to play a unique role as ‘part of the rich oral repertoire of the common people for centuries’ as well as appearing in literary Chinese writings (Rohsenow 1991: x). Metaphor is a figure of speech. ‘What distinguishes a metaphorical use of an expression is the relationship between its figurative meaning and its literal meaning. Metaphor involves a relation of resemblance or analogy, although this is not explicitly stated’ (Cruse 2006: 106). Metaphor is used in xiehouyu to transform the literal meaning presented in one part to the metaphorical meaning revealed in the other part. For example, an fang zhuayao—zhao ban (按方抓药—照办 ‘it is just like filling a prescription—simply following the instructions’; the intended meaning/message (IM): just doing the job, no questions asked) uses the metaphor of an fang zhuayao to crystallize the intended message, with vivid or satirical connotation. There are very few xiehouyu-like sayings in English. One example given in Walls and Walls (2009: 133) is, ‘these jeans fit like a cheap hotel—no ball room’, which ‘captures almost perfectly the form and the spirit’ of the Chinese xiehouyu. Another ‘near-equivalent’ English expression might be: ‘between a rock and a hard place—no way out’ (ibid.: 135). What is a xiehouyu? Xiandai Hanyu Cidian (《现代汉语词典》(A Modern Chinese Dictionary)) (2006: 1,505) defines it as a two-part saying. It forms a subcategory of folk sayings (熟语 shuyu), colloquialisms (俗语 suyu, 家常话 jiachanghua), and wisecracks (俏皮话 qiaopihua), also traditionally called proverbs (谚语 yanyu) (Wen 1985: 31). Xiehouyu is polysyllabic, metaphorical, and allegorical (Yip 2000: 229, 246). Cheung (1982: 52) names xiehouyu as ‘post-pause expressions’, because the second part is the focus of the entire 395

Grace Zhang

xiehouyu. He describes the structure of xiehouyu as clue phrase/message-pointer, (pause), and answer/message-carrier. Rohsenow (1991: ix) calls it ‘an enigmatic folk simile’, and a ‘special type of popular Chinese folk saying, extremely common but little studied’. Simile is a rhetorical device, typically using seems, like and to be to explicitly draw a similarity between different things and action; for example, ‘she is like a rose’. While xiehouyu may use a simile to make an analogy, they also employ many other mechanisms. As simile is not fixed and with no ‘intended meaning’ as such, simile and xiehouyu are not equivalent. Existing English translations of the term xiehouyu do not match precisely the particulars of xiehouyu contextualized in Chinese language and culture (Wang 1990). It is therefore better to retain the original xiehouyu term (Zeng 2013: 2 [preface]), as is done in this chapter. Xiehouyu, as a language form, has existed since the early Qin dynasty (221–206 bc). For example, wang yang bulao—wei wei chi ye (亡羊补牢—未为迟也 ‘mending the fold after the sheep have been stolen—still not too late’; IM: never too late to mend) appeared in Zhanguoce Chuce Si (《战国策•楚策四》, Chapter IV of the Chu Strategies in The Strategies of the Warring States). Interestingly, English says ‘shutting the stable door after the horses have gone’ and the intended message is ‘too late to mend’. Xiehouyu is thought to have developed from xiehouti (歇后体 ‘omitting the latter part of an expression’) (W. Chen 1997: 160–2). For example, in you yu xiongdi (友于兄弟 ‘brotherly love/friendship’), the last two characters can be omitted. The term xiehouti did not appear until around the Tang dynasty (ad 898), when it was used by Zheng Qing 郑綮 in chapter 179 of the Jiu Tangshu. Zheng Qi Liezhuan (《旧唐书•郑綮列传》(Biography of Zheng Qi in The Old Book of Tang) (Z. Sun and Wang 1982: 1 [preface]). In the Song dynasty (ad 960–1279), Yan Youyi 严有翼 was the first to explicitly use the term of xiehouyu in his Yiyuan Cihuang (《艺苑雌黄》(Art and Literature Review) to refer to xiehouti (Rohsenow 1991: xi). Xiehouti are considered an earlier form of present-day xiehouyu (Mao 1954: 95). W. Chen (1997: 160–2) discusses two types of xiehouyu as part of Chinese rhetoric: Old Style and New Style. The Old Style is called cangci (藏词 ‘hiding word’), similar to xiehouti. For example, you yu xiongdi (友于兄弟 ‘brotherly love/friendship’) can be used as youyu (友于 ‘brotherly love/friendship’), by ‘hiding’ the last two characters. The New Style is a two-part saying but with the second part omitted. For example, an bieren de jiaoma mai xie—sheng ban ying tao (按别人的脚码买鞋—生搬硬套 ‘buying a shoe according to someone else’s measurement—imitating from others mechanically’; IM: copy blindly), when only the part of an bieren de jiaoma mai xie is used, it is a xiehouyu in Chen’s term. The transformation from Old Style to New Style may represent a step towards the contemporary form, although Wen (1985: 9) considers that xiehouti (also a poetic style) and xiehouyu represent two different concepts: the former rhetorical, the latter lexicological. When both parts are present, Chen (1997: 14) calls it pijieyu (譬解语 ‘example-explanation saying’) instead of xiehouyu, because xiehou in his view literally means ‘to leave unspoken’ the ending of a phrase. His view is shared by others (e.g. Hu 1996: 263; Huang and Liao 2002: 321). Wen (1985: 20) prefers the term yinzhuyu (引注语 ‘lead-remark saying’), which highlights the second part as the focus, because the first part acts like a lead-in to trigger the explication of the second part. Most researchers agree that although xiehouyu may not be a perfect name, it has been conventionalized over the years, so retaining it is a convenient choice (e.g. Li 2001: 34–5). Collections of xiehouyu appeared around the Tang and Song eras (ad 618–1279) in works like Zazuan (《杂纂》‘Miscellany’) by Li Shangyin 李商隐 (Rohsenow 1991: xii). Written xiehouyu became popular in the Yuan period (ad 1271–1368), when the writers of Yuan Plays 396

Chinese xiehouyu

were mostly grass-roots intellectuals familiar with folk vernacular (Wen 1985: 27). In 1902, Smith published a collection of Chinese proverbs and common sayings including xiehouyu, praising xiehouyu as immeasurable treasures of the Chinese language. Along with the ‘Common Speech Movement’ in the May Fourth Period (1919 to the 1930s), xiehouyu as vernacular folk sayings were highlighted (Yu 2000). Rohsenow (1991) argues that the movement promoted the use of xiehouyu and the like in written literature, but xiehouyu did not come into widespread use until the Cultural Revolution led by Mao Zedong; Mao repeatedly used folksy xiehouyu in his speech and writing. Two examples are discussed by Rohsenow: in the 1940s, Mao said Jiang Taigong diaoyu—yuan zhe shang gou (姜太公钓鱼—愿者上钩 ‘Jiang Taigong fishes—whoever gets hooked does so of their own free will’; IM: a willing victim/ the willing rise to the bait), when talking about his taking US relief aid. In 1971, Mao used another xiehouyu: heshang da san—wu fa ( fa) wu tian (和尚打伞—无发(法)无天 ‘a Buddhist monk holding an umbrella—no hair, no sky’; IM: lawless) to indicate that he was ‘having [regard for] neither [earthly] law nor heaven [i.e. heavenly principle]’ (p. ix). It implies that during the Cultural Revolution Mao, disregarding the laws of man and the will of heaven, saw himself as bound by neither. Here fa (发 ‘hair’) is an aural pun on fa (法 ‘law’), and tian (天 ‘sky’) is a metaphor for ‘heaven’. There are three major phases of xiehouyu research: from 1930 to 1949 focus laid on the origins of xiehouyu; from 1949 to 1960 on their functions, and since then on the linguistic features and functions of xiehouyu within specific contexts (Wen and Zhou 2000). Chen Wangdao’s 陈望道 work in the 1930s on xiehouyu was pioneering, and remains influential in contemporary xiehouyu study (Wen and Zhou 2000). In recent years, there have been a number of xiehouyu dictionaries published; one of the most comprehensive and up-to-date is by Wen (2004).

2.  Content and structure Xiehouyu carry the civilization and wisdom of many generations, and are distinguished by their long history, wide range of sources, strong grass roots, and earthy flavour, unique humour, and satire (Zeng 2013: 2–3 [preface]). The relation between the two parts of a xiehouyu is primarily interpretative and inferential, especially when the second part is omitted as the intended meaning becomes implicit.

2.1. Content Xiehouyu originate from diverse literature and colloquial sources, including fables, historical events, beliefs, daily life, and experience. Some examples are as follows: ● Fable: Ye Gong hao long—kou shi xin fei (叶公好龙—口是心非 ‘Lord Ye’s love of dragons—say yes but mean no’; IM: duplicity). This allegorical xiehouyu comes from a folk tale: Lord Ye was known for his love for dragons; he put up pictures of dragons everywhere. A real dragon heard of this and dropped by one day to say hello. Lord Ye was terrified and ran away, so his love for dragons turned out, ironically, to be a sham. ● Historical story: Liu Bei jie Jingzhou—you jie wu huan (刘备借荆州—有借无还 ‘Liu Bei borrowed Jingzhou—borrowed but never returned’; IM: a dead loan, untrustworthy). Liu Bei was a well-known figure in the era of the Three Kingdoms (ad 220–280). It is said that he once ‘borrowed’ Jingzhou (a place) as a military base but never returned it. 397

Grace Zhang

● Experience in daily life: qiangniu de gua—bu tian (强扭的瓜—不甜 ‘an [unripe] melon wrenched off the vine—cannot be sweet’; IM: a forced result cannot be good). This xiehouyu usually refers to the situation of dating or marriage, conveying a life experience such as ‘forced love does not last’. It is similar to the English saying, ‘you can lead a horse to water, but you can’t make him drink’. Xiehouyu carry positive (complimentary), negative (pejorative), or neutral connotations. One of the common connotations is satirical: for example, Cai Shenye da guansi—you qian jiu youli (财神爷打官司—有钱就有理 ‘the God of Wealth goes to court—his money always prevails’; IM: money talks). The observation wryly notes that it does not matter who is guilty; those who have money always get their way. Xiehouyu have been created during different eras and in different regions, and are inevitably marked by time and locality (Wen 1985: 33–7). For example, titou tiaozi—yi tou re (剃头挑子—一头热 ‘a barber’s load on a carrying pole—warm on one end only’; IM: one’s own wishful thinking, unilateral willingness). In the old days, barbers in China walked through streets to do their business carrying their tools on a shoulder pole, including a bucket of hot water on one end. While this practice no longer exists, the xiehouyu has survived. Xiehouyu also show the traces of different places; for example Jinanfu de hama—gan gu du (济南府的蛤蟆—干鼓肚 ‘the frogs in Jinan—inflate their stomachs in vain’; IM: no way to vent one’s anger). The reason is that frogs in Jinan cannot croak like frogs in other places. Understanding this type of xiehouyu requires local knowledge. The form of xiehouyu can be varied, and alternative versions may be partially or completely reworded (Yip 2000: 230). For example, lanponiang de guojiaobu—you chang you chou (懒婆娘的裹脚布—又长又臭 ‘a slattern’s foot binding—long and smelly’), Pan Jinlian de guojiaobu—you chang you chou (潘金莲的裹脚布—又长又臭 ‘Pan Jinlian’s foot binding—long and smelly’), laotaipo de guojiaobu—you chang you chou (老太婆的 裹脚布—又长又臭 ‘an old woman’s foot binding—long and smelly’). The sayings refer to a speech or writing that is too long and very boring/stinking. In these variants, the first parts offer different metaphors, but all use the same words for the second part. In another group of variants, roubaozi da gou—you qu wu hui (肉包子打狗—有去无回 ‘a meat bun thrown at a dog—irretrievable’), roubaozi da gou—you qu wu huan (肉包子打狗—有去无还 ‘a meat bun thrown at a dog—irretrievable’), and roubaozi da gou—yi qu bu huitou (肉包子打狗— 一去不回头 ‘a meat bun thrown at a dog—gone, never to return’), the intended meaning for all three is ‘something gone forever’. In this set, the latter parts are in different forms (though with very similar meanings), but use the same words in the first part. The multiform phenomenon could be attributed to factors such as time, regional, and dialectal variations, and mutations occurring in oral traditions particularly in the folksy sayings. The creation and development of new xiehouyu reflect new phenomena and experience. For example, chashui ‘fayan’—erci ganran (茶水 ‘发炎’—二次感染 ‘ “inflamed” tea— infected/sick twice’; IM: both parties are guilty) is based on events in recent years (Ding 2012: 37–8). In 2007, suspecting unethical practices by money-grabbing doctors, a journalist in China faked a urine test by supplying a sample of tea. The test result was positive. The act was copied in 2012, and the result was again positive. Neither the journalists nor the doctors are considered innocent, as the former faked the test samples, and the doctors were also culpable. Xiehouyu are closely attached to Chinese folk customs, and are culturally loaded. For example, daguniang sheng haizi—feili bu taohao (大姑娘生孩子—费力不讨好 ‘an unmarried girl gives birth to a baby—her hard labour and pain will not be appreciated’; IM: a thankless 398

Chinese xiehouyu

job) draws on a traditional attitude towards unmarried mothers, and without cultural awareness it would be difficult to follow. Factors such as this make xiehouyu appear somewhat mysterious; they have been described as ‘enigmatic’ (Rohsenow 1991: ix) and ‘cryptic’ (Matisoff 2011: 275). While they can be culturally specific, they also share universalities with other cultures. For example, na zhe jimao dang lingjian—xiao ti da zuo (拿着鸡毛当令箭— 小题大做 ‘holding a cock feather as if it were an ancient token of military authority— making a big deal of something trivial’; IM: overkill) refers to treating a casual remark as an order and make a big fuss about it. This is similar to ‘making a mountain out of a molehill’ or ‘creating a storm in a teacup’ in English.

2.2. Structure The formation of xiehouyu is diverse. For example, tie gongji—yi mao bu ba (铁公鸡— 一毛不拔 ‘an iron cock—no hair to pluck out’; IM: a miser, too stingy to give a penny) has a noun phrase in the first half and a subject–predicate structure in the second part. Gou na haozi—duo guan xianshi (狗拿耗子—多管闲事 ‘a dog chasing mice—meddling in cats’ business’; IM: doing something uncalled for, poking one’s nose into other people’s business) has a subject–predicate–object structure followed by a predicate–object structure. Syntactically, xiehouyu are often used as the predicate or object of a sentence, with is (是 shi) or like (像 xiang). For example, ta shi zhulan da shui—yi chang kong (他是竹篮打水—一场空 ‘He is like “drawing water with a bamboo basket—all in vain” ’). They may also be a subordinate component, as in bie zuo naxie zhulan da shui yi chang kong de shi (别做那些 竹篮打水一场空的事 ‘Do not do things that are like “drawing water with a bamboo basket”, achieving nothing’). Xiehouyu are most distinctive in their two-part form. The first part is usually a concise metaphorical description in an inexplicit way, so the meaning is often implied rather than stated. That is perhaps why xiehouyu are sometimes misunderstood. For example, in his 1971 report on an interview with Mao Zedong, US writer Edgar Snow misinterpreted heshang da san—wu fa ( fa) wu tian (和尚打伞—无发(法)无天 ‘a Buddhist monk holding an umbrella—no hair, no sky’; IM: lawless). He said that Mao described himself as ‘a lone monk walking in the world with a leaky umbrella’, Mao actually used it to imply that he was the one behind the Cultural Revolution which overturned laws and principles in the world (cited in Rohsenow 1991: ix). The first part of a xiehouyu offers a clue, leading people to search for the intended answer which may be presented in the second part. Walls and Walls (2009: 133) call the second part ‘the punch line’. In their view, xiehouyu is structurally similar to a riddle; both have a clue then the answer. For example, naima bao haizi (奶妈抱孩子 ‘the child a wet nurse is holding’), this first part is a metaphor, having more than one possible implication (e.g. ‘time to feed’ or ‘not your own’). The intended meaning revealed by the second part renjia de (人家的 ‘someone else’s’) is actually ‘it belongs to someone else’. This characteristic of inference makes xiehouyu both fascinating and demanding. He (2011: 360) looks at the two-part structure in a different way, arguing that the two parts actually construct a simile in a pattern of ‘part 2 is like part 1’ or ‘part 2 is as  .  .  .  as part 1’. For example, mao ku laoshu (part 1)—jia cibei (part 2) (猫哭老鼠—假慈悲 ‘a cat weeps over a mouse’s misfortune—pretence of being sympathetic’; IM: shed crocodile tears). Instead of the usual explanation of metaphor + intended meaning, she suggests that the structure should be ‘jia cibei (part 2) is like mao ku laoshu (part 1)’, making the whole xiehouyu a simile. 399

Grace Zhang

There is debate on whether or not the omission of the second part is a common practice. Some claim it is (e.g. Xiandai Hanyu Cidian 2006: 1,505). However, when Wen (1985: 2) conducted a survey of 4,893 xiehouyu in literary works, he found only about 10% omitted the second part. While few omissions occur in literary contexts, they may be more frequent in spoken Chinese, to encourage people to use their imagination and reasoning powers (Chen 1997: 160–2). Walls and Walls (2009: 133) state that if the second part is well known, it may not be verbalized, a view supporting Cheung’s (1982: 53) observation that through time xiehouyu have become ‘so conventionalized that a listener would need no more than the cue itself  .  .  .   to arrive at the intended meaning’. For example, for the popular huangshulang gei ji bainian—mei an haoxin (黄鼠狼给鸡拜年—没安好心 ‘the weasel wishes the chicken a Happy New Year—without good intentions’, IM: with ulterior motive), the omission may be preferred as it can reduce redundancy. However, if a xiehouyu is not well-known, the omission may cause ambiguity and confusion. Some xiehouyu have second parts that cannot be omitted, because they would mean something else. For example, chutou de chuanzi—xian lan (出头 的椽子—先烂) means ‘exposed rafters—the first to rot’; IM: people in the limelight bear the brunt of attack, don’t stick your neck out. Without the second part, it means instead ‘a rafter sticking out’; IM: the leader (with a pejorative connotation).

3.  Punning: sound and meaning Pun is ‘a form of word-play in which two or more meanings of an expression are activated at the same time’ (Cruse 2006: 148). Yip (2000: 247) defines punning as ‘exploit[ing] the double sense of the word or the sense of any of its homophonic counterparts in the lexicon’. Punning manipulates the inexplicitness of multiple meanings of words or of similar-sounding words for a deliberate meaning and effect. The Chinese language is full of homophones, so the extent to which puns are employed seems natural. Punning is a common mechanism in xiehouyu as a ‘word game’ to produce humorous and satirical effects. It consists of two major types: punning through sound (exploiting similar-sounding words); and punning through meaning (exploiting multiple meanings of a word).

3.1.  Sound punning This type of xiehouyu works through homophone or homonymy, playing on two or more words of the same or nearly the same sound with different meanings. For example, sanjiu de luobo—dong (dong) le xin (三九的萝卜—冻(动)了心 ‘a mid-winter turnip—the heart is frozen’; IM: get stirred up/become keen on something). Dong (冻 ‘frozen’) is homonymic with dong (动 ‘get stirred up’), although written in different characters. This xiehouyu plays on the homonymy, and allows the literal meaning of dong (冻) to imply the metaphorical meaning of dong (动). Due to the complexity of sound punning, it may be demanding to work out exactly what sound and meaning mean. This type of punning brings excitement and exoticism to the Chinese language, with a humorous touch.

3.2.  Meaning punning Xiehouyu’s meaning punning plays on a word with two or more meanings. ‘Almost every Chinese character combines with itself many meanings totally distinct from each other, a circumstance which gives occasion for a multitude of puns in which a word is used in one sense to suggest another sense of the same word’ (Smith 1902: 195). Lefevere (1992: 51–2) 400

Chinese xiehouyu

explains that readers get ‘both the obvious usual meaning of a word and the frequently less obvious, more unusual meaning the author intended. They get the “norm” and the “deviation” from that norm simultaneously. The clash between the two heightens the pun’s illocutionary power’. For example, mawei shuan doufu—ti bu qilai (马尾栓豆腐—提不 起来): ‘tying bean-curd with a horse-hair—cannot lift it’ (obvious meaning), punning of ‘too inferior or shameful to be mentioned’ (less obvious meaning). The character ti (提) has two meanings: ‘to lift’ (obvious meaning) and ‘to mention’ (less obvious meaning); the play on the two meanings creates vivid illocutionary power. There are many other punning devices which enhance further the expressive power of xiehouyu, such as the rhetorical devices of personification, hyperbole, contrast, rhetorical question, antithesis, and repetition. Onomatopoeia is another technique used in xiehouyu, as in gou nian yazi—guagua jiao (狗撵鸭子—呱呱叫 ‘a dog chases a duck—[the duck] cries out: “Quack! Quack!” ’ IM: splendid). Guagua jiao is an onomatopoeic expression imitating the sound of a duck’s call; it also means ‘excellent’.

4. Functions Xiehouyu are quirky, whose creative metaphors encourage the addressee to explore contextual implications. ‘The wider the range of potential implicatures and the greater the hearer’s responsibility for constructing them, the more creative the metaphor is’, argue Sperber and Wilson (1986: 168), who maintain that metaphors are an integral part of language, rather than serving a merely decorative function.

4.1.  Vivid and satirical power Xiehouyu are a manifestation of the optimistic, distinctive humour and imagination of the Chinese people (Wang 1990). It is ‘the wit of Chinese language’, and owing to its humorous doubling of meaning or sound, ‘the humorous touch is the most important feature of xiehouyu’, particularly because the apodosis either comes from an unexpected angle or carries a punning sense (C. Sun 1981: vii). For example, gou zhang jijiao—yang ( yang) shi de (狗长犄角—羊(洋)式的 ‘a dog grows horns—in a sheep’s style’; IM: foreign-style). ‘In the style of a sheep’ is twisted to mean ‘foreign-style’ through the sound punning between yang (羊 ‘sheep’) and yang (洋 ‘foreign’), that is a dog with sheep’s horns appears to be foreign. Such unexpected wordplay is unique and effective (Liu and Zhang 2006: 49). Xiehouyu feature vividness and satire, and are ‘ubiquitous, vividly employing the full spectrum of traditional Chinese history and imagery, from the lofty to the mundane’, and have played an indispensable role in communication for centuries (Rohsenow 1991: ix). They are also called witticisms/wisecracks (俏皮话, qiaopihua, Zeng 2013: 2 [preface]), showing the common perception that xiehouyu make people laugh. They are a valuable tool for drawing characters (people) and increasing artistic charm in literary works (Cheng and Li 1963: 35). Xiehouyu are made of bright and unexpected metaphors and clever puns that enliven communication while conveying the rich, centuries-long heritage of Chinese civilization. The need to interpret implications arouses curiosity, and listeners enjoy the challenge of working out the intended meaning. For example, laoshu la muqiao—datou zai houtou (老鼠 拉木锹—大头在后头 ‘a mouse dragging a wooden shovel—the bigger end comes behind’; IM: the bigger one is yet to come; the best is saved for last’) was used by someone presenting a paper towards the end of a conference and won a round of applause from the audience (Wen 2011: 14 [preface]). 401

Grace Zhang

Another attractive function of xiehouyu is satirical. For example, zhubizi cha dacong— zhuangxiang (xiang) (猪鼻子插大葱—装象(相) ‘a pig with scallions in nostrils—pretending to be an elephant’; IM: put on a show) uses sound punning from zhuangxiang (装象 ‘pretend to be an elephant’) to zhuangxiang (装相 ‘put on an act’), conveying a sharp, almost pictorial image of someone who pretends. Such satire may be applied to anything and anyone from commoners to authority, and even to saints. Xiehouyu about Confucius are mostly sarcastic, as in Kong Fuzi ban jia—jing shi shu (shu) (孔夫子搬家—净是书(输) ‘Confucius moving his house—nothing but books’; IM: always lose). Here, shu (书 ‘book’) is a pun for shu (输 ‘loss’). A xiehouyu employing words like dog, rat, or turtle is generally expected to be disparaging or derisive, as these words usually represent something unpleasant. This can be seen in na zhe goushi dang mahuar—xiang chou bu fen (拿着狗屎当麻花儿—香臭不分 ‘mistaking dog dung for a fried dough twist—unsure about what’s appetizing and what’s stinking’; IM: cannot tell good from bad).

4.2.  Metaphorical manifestation of thought and experience Metaphors are fundamental to xiehouyu, as the whole operation is based on metaphorical analogy. Mao (1954) regards xiehouyu as a language game or wordplay, although Lakoff and Johnson (1980) consider that metaphors are not merely rhetorical and stylistic, but are something cognitively important that we cannot live without. That is, metaphors are much more than just figurative language; they represent how we think. For example, the metaphor ‘time is money’ is a common concept ingrained in our thoughts and everyday lives. Based on the view that how we talk about things determines how we think about them, Lakoff and Johnson challenge the claim that the meaning has very little to do with what people find substantive in their lives. An important function of xiehouyu, overlooked in the existing literature, is that it serves as a manifestation of our thoughts and day-to-day experience, as metaphors not only influence how we communicate ideas but affect how we think. Particularly, the message conveyed by xiehouyu represents our conceptualization and experience of the world around us, helping us to make sense of it. For example, Liu Bei jie Jingzhou—you jie wu huan (刘备借荆州— 有借无还 ‘Liu Bei borrowed Jingzhou—borrowed but never returned’; IM: a dead loan, untrustworthy) allows us to use what we know about history to communicate an understanding and interpretation of the contemporary world. It warns us that before we lend things, we should always think twice. Metaphors and xiehouyu play a role in shaping or reshaping our thoughts and actions, revealing common ways of speaking and thinking. For example, baji shifu xue shouyi—zhang dao lao, xue dao lao (八级师傅学手艺—长到老,学到老 ‘an eight-grade master learning craft—people should live and learn’). An eight-grade master in China is supposed to be the best in his trade, yet he never stops improving. This saying carries a profound motto; many such important messages are delivered through xiehouyu.

5. Categorising xiehouyu Xiehouyu is part of sayings (熟语 shuyu) (e.g. Huang and Liao 2002: 317; Zhou 2006: 246), together with proverbs (谚语 yanyu), four-character idioms (成语 chengyu) and threecharacter idioms (惯用语 guanyongyu). Idiom is an expression that is semantically opaque and syntactically frozen (Cruse 2006: 29, 82), for example to pull someone’s leg, its meaning is not a simple adding up of ‘pull’ and ‘leg’ and the structure cannot be changed either. Proverbs are crystallizations of historical, experiential, and popular wisdom (Walls and Walls 402

Chinese xiehouyu

2009: 100). For example, da ren bie da lian, ma ren bie jieduan (打人别打脸,骂人别揭短 ‘avoid hitting someone’s face and cursing someone’s shortcomings’, means ‘avoid attacking people where they are most vulnerable’). Walls and Walls divide four-character phrases into ‘four-character idioms’ and ‘four-character set phrases’ (2009: 52, 55). For simplicity, this chapter uses ‘four-character idioms’ to include both ‘four-character idioms’ and ‘fourcharacter set phrases’. The former are mostly ‘concise summaries of well-known anecdotes or parables from classical Chinese texts’; they note that as educated Chinese ‘derive pleasure from using classical allusions to crystallize the essence of a current situation both in speech and in writing’, using them is admired and appreciated (Walls and Walls 2009: 52). For example, si mian chu ge (四面楚歌 ‘be besieged on all sides, embattled on every front’) originates from the classics of Shiji Xiangyu Benji (《史记•项羽本纪》‘Chronicle of Xiang Yu’ in Records of the Grand Historian). Set phrases do not share this classical background: ben mo dao zhi (本末倒置 ‘confuse cause and effect, put the cart before the horse’) is one such non-classical four-character set phrase. Three-character idioms similarly lack a classical background, being folksier. One such example is zou houmen (走后门 ‘enter through the back door, secure advantages through pull or influence’). As shown in Table 22.1, all four categories are fixed formulaic expressions, the latter two less flexible than the former two. Except for the four-character phrase, they are folksy and tend to be informal. Xiehouyu are not often confused with the latter two, which have distinctive four-character and three-character forms. Four-character idioms can be part of xiehouyu: for example in ni shui xing zhou—bu jin ze tui (逆水行舟—不进则退 ‘sail against the current—either forge ahead or fall behind’; IM: he who does not advance falls backward), the first part is a four-character idiom. In dengcao zhi bu—wang fei xin ji (灯草织布—枉费 心机 ‘weaving cloth with rushes—rack one’s brains in vain’; IM: fruitless effort, waste of time), the second part is a four-character idiom. A less clear-cut boundary exists between xiehouyu and proverbs, because they have similar two-part structures. The difference lies in the relationship between the parts: for a proverb the two parts do not have a pattern of metaphor + intended meaning. For example, ji zou bing, man zou ni (急走冰,慢走泥 ‘walk quickly on ice but slowly on muddy roads, scoot across ice but trudge through mud’), the first and second parts have a coordinated, rather than a clue–answer, relationship. A riddle has a clue and an answer, but it is not a language unit: merely a game where the answer must be figured out by the players. Wen (2011: 4 [preface]) confirms that riddles are a pastime, and not the same as xiehouyu in nature. Table 22.1  Xiehouyu and others Sayings

Style

Form

Origin

Internal structure

xiehouyu

less formal, allegorical, and humorous

two parts, in the form of phrase or sentence

folk

clue–answer

proverb

less formal, insightful

one or two parts, in the form of phrase or sentence

folk

coordinated

four-character idiom

formal, solemn, and elegant

four-character form

classics

N/A

three-character idiom

less formal, user-friendly

three-character form

folk

N/A

Note: The features listed are distinctive but not exhaustive.

403

Grace Zhang

There have been discussions on the legitimacy of xiehouyu in literature. Mao (1954) questioned the literacy value of xiehouyu in literature, but later said that xiehouyu can be used in literary works to create vivid effects. Lao (1958) has reservations about using xiehouyu excessively in writing. The discussions imply that xiehouyu are informal, and possibly do not have the prestige that four-character idioms enjoy. Using four-character idioms is admirable and shows the user is educated and well-learned (Walls and Walls 2009: 52). The former prime minister of China, Wen Jiabao, was famous for quoting classic idioms. While using xiehouyu may not show status as such, it has important functions that other tropes do not necessarily have.

6.  Problematic and inappropriate xiehouyu While the majority of xiehouyu play a positive role, some are absurd. Those xiehouyu created long ago inevitably reflect the now-obsolete and problematic ideologies of their time, no longer compatible with the values upheld by contemporary society. Mao (1954: 94) points out that some xiehouyu consist of feudal ideology and vulgar language, and should be used with care; and there have been calls to clean up unacceptable xiehouyu (e.g. Hao 2012). Some xiehouyu are insensitive towards disadvantaged groups such as the disabled, seniors, widows, homeless people, and overweight people; these are considered inappropriate uses of xiehouyu (Z. Sun and Wang 1982: 4 [preface]). For example, xiazi dian deng—bai fei la (瞎子点灯—白费蜡 ‘a blind man lighting a lantern—a sheer waste of candles’; IM: a total waste) shows disrespect to blind people. Jiaohuazi qi wugeng—qiong mang (叫花子起 五更—穷忙 ‘the beggar rise before dawn—awfully busy’; IM: working poor) implies derision for poor people. Shikelang ban jia—gundan (屎壳郎搬家—滚蛋 ‘a dung beetle shifting house—rolling an egg’; IM: get lost!) has a somewhat abusive tone. The inferential connection between the xiehouyu’s two parts does not always follow common perception and reasoning, and interpretation is limited by factors such as time periods, regional variations, and specific knowledge (Li 2001: 34–5). For example, na zhe yaoshi man jie pao—dang jia bu guan shi (拿着钥匙满街跑—当家不管事 ‘roaming the streets carrying the house keys—the head of a household not taking care of household affairs’; IM: irresponsible, unfit for the post) does not necessarily follow common reasoning, because someone who roams the streets carrying the house keys could well be a very responsible person. Mangshe shang shu—pan sheng (蟒蛇上树—盘升 ‘python climbing up the tree—slowly’; IM: increase in small steps), where pansheng (盘升 ‘moving up slowly and in small steps’) is a stock-market term, so the interpretation of this xiehouyu requires specific knowledge, without which the appropriate inference is far from straightforward. Wen (2011: 16 [preface]) cautions that in terms of xiehouyu use, more is not necessarily better: xiehouyu should be clean and used appropriately. It is important to increase awareness that some problematic xiehouyu should be avoided.

7.  Xiehouyu and translation Why are there so many xiehouyu in Chinese, but not in other languages such as English? D. Chen (2011: 4) proposes that ‘English speakers are good at abstract thinking while the Chinese are apt to think in terms of images. As a result, there are more explanatory words in English and more imagery words in Chinese’. This claim may have some implications for xiehouyu translation, as a free translation might be well received because of the assumed capacity of abstract reasoning of English listeners. 404

Chinese xiehouyu

Xiehouyu is ‘very difficult to translate and it is probably for this reason that not much work has been done in this field’ (C. Sun 1981: vii). There are three major translation methods: literal (direct) translation, free (indirect) translation, and substitution with equivalent English metaphors (e.g. Bao and Bao 2000). Gutt regards metaphors and puns as ‘non-literal uses of languages’ (1992: 48). It is possible that puns are not translatable, so puns in the original text could be simply eliminated in the target language (Liu and Zhang 2006: 66). The cultural element of xiehouyu is perhaps the most complex part of the translation. For the translation of xiehouyu from Chinese to English, Liu and Zhang (2006) propose the direct approach (total interpretive resemblance to the original, normally for people who know Chinese well), indirect approach (adequate resemblance, normally for people who do not know Chinese well), or the substitution of appropriate English metaphors where possible. For example, Zhengyueshiwu tie menshen—wan le ban yue (正月十五贴门神—晚了半月 ‘pasting up pictures of the Door Gods on the fifteenth day of the first lunar month—half a month late’, IM: too late). Direct translation: ‘It’s like pasting up pictures of the Door Gods when the Chinese New Year is already half a month past.’ Indirect translation: ‘It’s too late.’ (Note: the Chinese paste up the pictures of the two Door Gods on the first day of the Chinese New Year. This is done as a way to avert evil and protect the entire household throughout the year.) Substitution of metaphor in English: ‘It’s like shutting the door after the horse had bolted’ (68–9) The indirect translation assumes that the target audience does not have cultural knowledge of the Chinese New Year, so ‘it’s too late’ is a free translation, offered with an explanatory note. The substitution shows that the Chinese metaphor can be replaced with the English metaphor ‘shutting the door after the horse has bolted’, which is vivid and easy to understand. A match-up metaphor in the target language may not always be available, but whenever possible the strategy of substituting metaphors should be utilized (Liu and Zhang 2006: 70). Meaning punning is perhaps easier to translate than sound punning, as extra explanation of sound-play is needed. Because they are packed with linguistic and cultural particulars, the main principle of xiehouyu translation is to offer less opaque, more accessible xiehouyu; advanced learners of Chinese use direct literal translation, otherwise indirect and free translation, with notes that provide adequate information to ensure understanding. Xiehouyu contain both common aspects of folk wisdom and specific traits of Chinese culture. They are a powerful rhetorical device, with a significant role in increasing the effect of language. They are also an indispensible manifestation of experience and philosophy, serving as a fundamental platform for the transformation of thoughts and experience. Appropriate use of xiehouyu makes the Chinese language fascinating and creative.

8. Resources Rohsenow, John S. (1991) A Chinese–English Dictionary of Enigmatic Folk Similes (Xiehouyu), Tucson, Arizona: The University of Arizona Press. Rohsenow’s work is one of the early xiehouyu dictionaries. He provides literal meaning, figurative meaning, and supplementary information in English. The collection has about 4,000 xiehouyu written in simplified Chinese characters, arranged into alphabetical order according to pinyin transliteration, and translated with annotations where needed. 405

Grace Zhang

Sun, Chia-chun (1981) As the Saying Goes, St. Lucia, Queensland: University of Queensland Press. Sun’s work has two parts, the first discusses proverbs etc., and the second specifically deals with xiehouyu. The target readership is learners whose first language is not Chinese; hence elaborate explanations are provided in English throughout. The detailed translation of xiehouyu is a combination of pinyin, word-by-word analysis, literal and free translations, and footnotes. Sun uses full-form Chinese characters. Wang, Chiu-kuei 王秋桂 (ed) (1990) Xiehouyu (《歇后语》(Chinese Riddle-phrases)), Taipei: The Overseas Chinese Library. This work is a picture book in five volumes. It is a selection of basic, most frequently used xiehouyu. The vivid pictures make it easy for learners of Chinese, especially children. It is in full-form Chinese characters, with English translation. Wen, Duanzheng 温端政 (ed) (2004) Zhongguo Xiehouyu Daquan (《中国歇后语大全》(A Comprehensive Dictionary of Chinese Xiehouyu)), Shanghai: Shanghai Dictionaries Press. This dictionary of xiehouyu has about 85,000 entries, probably the highest number of xiehouyu so far. The collection is comprehensive and up to date. It is a monolingual and Chinese-only dictionary, in simplified Chinese characters. Zeng, Dongjing 曾东京 (ed) (2013) Hanying Shuangjie Xiehouyu Cidian (《汉英双解歇后 语词典》(A Chinese–English Explanation Dictionary of Xiehouyu)), Shanghai: Shanghai University Press. This dictionary of xiehouyu consists of about 4,000 entries. It has indexes of Chinese by initial words and strokes, and English by nouns and verbs. All the explanations of xiehouyu are in both Chinese and English. It is in simplified Chinese characters.

Note 1 The author thanks Andy Kirkpatrick, Margaret Johnson, and the reviewer for their helpful feedback.

References Bao, Huinan 包惠南 and Bao Ang 包昂 (2000)《实用文化翻译学》(Translation in Cultural Context), Shanghai: Shanghai Kexue Puji Press. Chen, Dezhang (2011) Contrastive Linguistics between Chinese and English, Beijing: Foreign Language Teaching and Research Press. Chen, Wangdao 陈望道 (1997)《修辞学发凡》(An Introduction to Rhetoric), 2nd ed., Shanghai: Shanghai Education Press. Cheng, Daming 程达明 and Li Yunhan 黎运汉 (1963)〈试谈歇后语〉(‘On Xiehouyu’),《中山大学 学报•社会科学版》(Journal of Zhongshan University—Social Sciences edition) 1: 35–45. Cheung, Hung-nian Samuel 张洪年 (1982)〈粤语中的歇后语现象〉(‘The Phenomenon of Cantonese Xiehouyu’),《新华学报》(Xinhua Journal ) 14(1–2): 51–103. Cruse, Alan (2006) A Glossary of Semantics and Pragmatics. Edinburgh: Edinburgh University Press. Ding, Shizhao 丁时照 (2012)〈一个新歇后语的诞生〉(‘The Creation of a New Xiehouyu’),《青年 记者》(Youth Journalist) 9: 37–8. Gutt, Ernst-August (1992) Relevance Theory – A Guide to Successful Communication in Translation, Dallas, Texas: Summer Institute of Linguistics.

406

Chinese xiehouyu Hao, Tiechuan 郝铁川 (2012)〈保障人权需要一场语言革命—从歇后语对几种人的歧视说起〉(‘The Need of a Language Revolution to Protect Human Rights—Discrimination against Several Groups of People in Xiehouyu’),《学术界》(Academia) 4: 119–25. He, Aijing 何爱晶 (2011)〈歇后语研究新论—心智哲学的观点〉(‘New Theory on Xiehouyu—Views of the Philosophy of Mind’),《现代外语》(Modern Foreign Languages) 34(4): 356–63. Hu, Yushu 胡裕树 (ed) (1996)《现代汉语》(Modern Chinese), 6th ed., Shanghai: Shanghai Education Press. Huang, Borong 黄伯荣 and Liao Xudong 廖序东 (eds) (2002)《现代汉语》(Modern Chinese), 3rd ed., vol. 1, Beijing: Tertiary Education Press. Lakoff, George and Mark Johnson (1980) Metaphors We Live By, Chicago, Illinois: University of Chicago Press. Lao, She 老舍 (1958)〈关于文学的语言问题〉(‘Language Issues in Literature’), in Lao She (ed) 《福星集》(Fuxing Collection), Beijing: Beijing Press, 92–106. Lefevere, André (1992) Translating Literature: Practice and Theory in a Comparative Literature Context, New York: Modern Language Association of America. Li, Xingjian 李行健 (2001)〈关于《歇后语》的规范问题〉(‘On the Standardization of “Xiehouyu” ’), 《汉语学习》(Chinese Language Learning) 4: 34–6. Liu, Chiung-wen and Grace Zhang (2006) ‘Translation of Chinese Xiehouyu (Sayings) and Relevance Theory’, Across Languages and Cultures 7(1): 49–76. Matisoff, James A. (2011) ‘ “Stung by a Bee, You Fear a Fly”: Areal and Universal Aspects of Lahu Proverbial Wisdom’, Bulletin of the School of Oriental and African Studies, 74: 275–304. Mao, Dun 茅盾 (1954)〈关于《歇后语》〉(‘On “Xiehouyu” ’),《人民文学》(People’s Literature) 6: 94–6. Rohsenow, John S. (1991) A Chinese–English Dictionary of Enigmatic Folk Similes (Xiehouyu), Tucson, Arizona: The University of Arizona Press. Smith, Arthur H. (1902) Proverbs and Common Sayings from the Chinese, Shanghai: The American Presbyterian Mission Press. Sperber, Dan and Deirdre Wilson (1986) Relevance: Communication and Cognition, Oxford: Blackwell. Sun, Chia-chun (1981) As the Saying Goes, St. Lucia, Queensland: University of Queensland Press. Sun, Zhiping 孙志平 and Wang Shijun 王士均 (eds) (1982)《歇后语四千条》(Four Thousand Xiehouyu), Shanghai: Shanghai Wenyi Press. Walls, Yvonne Li and Jan W. Walls (2009) Using Chinese: A Guide to Contemporary Usage, Cambridge: Cambridge University Press. Wang, Chiu-kuei 王秋桂 (ed) (1990)《歇后语》(Chinese Riddle-Phrases), Taipei: The Overseas Chinese Library. Wen, Duanzheng 温端政 (1985)《歇后语》(Xiehouyu), Beijing: The Commercial Press. Wen, Duanzheng (ed) (2004)《中国歇后语大全》(A Comprehensive Dictionary of Chinese Xiehouyu), Shanghai: Shanghai Dictionaries Press. Wen, Duanzheng (ed) (2011)《中国歇后语大词典》(Dictionary of Chinese Xiehouyu), Shanghai: Shanghai Dictionaries Press. Wen, Duanzheng and Zhou Jian 周荐 (2000)《二十世纪的汉语俗语研究》(Twentieth Century’s Chinese Folk Sayings Studies), Taiyuan: Shuhai Press. Xiandai Hanyu Cidian (2006)《现代汉语词典》(A Modern Chinese Dictionary), 5th ed., Beijing: The Commercial Press. Yip, Po-Ching (2000) The Chinese Lexicon: A Comprehensive Survey, London and New York: Routledge. Yu, Genyuan 于根元 (2000)《二十世纪的中国语言应用研究》(Twentieth Century’s Applied Chinese Studies), Taiyuan: Shuhai Press. Zeng, Dongjing 曾东京 (ed) (2013)《汉英双解歇后语词典》(The Chinese–English Explanation Dictionary of Xiehouyu), Shanghai: Shanghai University Press. Zhou, Jian 周荐 (2006)《词汇学词典学研究》(Studies on Lexicology and Lexicography), 2nd ed., Beijing: The Commercial Press.

407

23 Classical Chinese Chris Wen-chao Li San Francisco State University, USA

Classical Chinese Classical Chinese refers to the language of canonical texts and formal writings in China prior to the twentieth century. In the early twentieth century, coinciding with the end of imperial rule, language reforms were instituted so that writing was no longer carried out in the classical language, which was understood only by an educated elite, but instead done in an approximation of the modern northern vernacular. The expression ‘Classical Chinese’ has a narrow sense and a broad sense. In its narrow sense, Classical Chinese (古文 gǔwén ‘ancient text’) is the designation for written language from the Zhōu dynasty (1111–256 bce) through to the Hàn dynasty (206 bce – ce 220), encompassing the language of Chinese classics such as the Analects (《论语》Lún Yǔ), the Doctrine of the Mean (《中庸》Zhōng Yōng), the Great Learning (《大学》Dà Xué) and Mencius (《孟子》Mèng Zǐ), and early historical narratives such as the Discourses of the States (《国语》Guó Yǔ), Chronicles of Zuǒ (《左传》Zuǒ Zhuàn) and Strategies of the Warring States (《战国策》Zhànguó Cè). Later imitations of the classical style, used in literature and formal writings from the third century to the early twentieth century, are said to be written in Literary Chinese (文言文 wényánwén ‘literary text’, also referred to as 文理 wénlǐ by biblical scholars). The Classical/Literary divide is in many ways similar to the relationship between Classical Latin and Medieval Latin in Europe, both learned languages, the latter of which emulates the more consistent grammar and lexis of the former. In actual usage, however, the Classical/Literary distinction is not always maintained by Chinese academics, as gǔwén 古文 and wényánwén 文言文 are used interchangeably to refer to formal writing styles prior to the advent of modern vernacular literature (白话文 báihuàwén ‘plain speech text’). It is this broader sense of Classical Chinese, including both early source texts and later imitations, that is adopted in this chapter. In its earliest stages, Classical Chinese evolved out of the written form of Old Chinese (c. 1766 bce – ce 220), traces of whose grammatical patterns can be found in oracle bone and bronze inscriptions of the Shāng (1766–1045 bce) and Western Zhōu (1045–771 bce) dynasties. The inscriptions are for the most part short sentences describing ceremonies and divinations, and are considered to be a more or less faithful record of the spoken language of the day in the Yellow River basin. 408

Classical Chinese

With the breakup of the Zhōu empire in the third century bce however, a new culture of pluralism demanded a more stylized form of writing suited to political oratory. This is reflected in the language of works such as the Confucian Analects and Mencius, which is more concise and structured, richer in rhetorical devices, and shows obvious imitations of earlier classics such as the Book of Songs (《诗经》Shī Jīng) and the Book of Changes (《易经》Yì Jīng). The later Warring States period (402–221 bce) and the Qín (221–206 bce) and Hàn dynasties saw a further move toward allegory and ornamentation, resulting in writing that is stylistically distinct from the vernacular language of the day. It was during this period that Classical Chinese forged an identity as a literary language separate from vernacular speech, and it is the conventions of this period that later authors sought to emulate when writing in the ‘classical style’. For this reason, some sinologists reserve the term ‘Classical Chinese’ (古文 gǔwén) for the writings of the Zhōu, Qín and Hàn dynasties, and refer to the language of later imitations as ‘Literary Chinese’ (文言文 wényánwén). Following the Hàn dynasty, worship of form was taken to an extreme at the expense of substance, giving rise to the belletrist ‘parallel prose’ (骈文 piánwén) of the Southern Dynasties (420–589), in which balance of rhythm, imagery and tonal patterns became of primary concern. This worship of formal elements created a backlash in the Táng dynasty (618–907), in which neoclassicists such as 韩愈 Hán Yù (768–824) and 柳宗元 Liǔ Zōngyuán (773–819) called for a return to substance and the rhetorical styles of the Zhōu, Qīn and Hàn dynasties. From the Táng dynasty onwards, different schools of writing have offered different takes on the classical language, and literary aesthetics have oscillated between form and substance, and between arch conservatism and the adoption of new grammar and lexicon. In the early Republican period, especially in the context of language debates leading up to the 1917–19 Vernacular Language Movement (白话文运动 Báihuàwén Yùndòng) and the 1919 May Fourth Movement (五四运动 Wǔsì Yùndòng), Classical Chinese was often regarded as the logical complement of vernacular Chinese (白话 báihuà), for which definitions vary, ranging from the synchronic to diachronic accounts covering different historical periods. Republican language reformer 胡适 Hú Shì (1891–1962), in his seminal work A History of Vernacular Literature (《白话文学史》Báihuà Wénxué Shǐ, 1928), implied a dichotomy in which vernacular literature is literature written in the spoken language of the day (which may be far removed from the spoken language of today), and classical literature is that which is excluded from this scope by default. By this definition, what is vernacular and what is classical is not a fixed notion, but rather varies with each historical period. Hú’s dichotomy, while similar to the European notion of vernacular vs. mainstream literature, is actually foreign to the Chinese tradition. Hú is often criticized for framing a definition of vernacular language that is too broad, as his definition would necessarily include obscure works of oral literature from remote periods, which speakers of Modern Chinese would have trouble understanding. A more widely accepted definition of vernacular and literary language is that of 吕叔湘 Lǚ Shúxiāng (1904–1998), who states that vernacular literature is written text which corresponds to spoken language from the Táng dynasty (618–907) onwards; all else is relegated to Literary Chinese (Lǚ 1944: 12). Lǚ’s treatment includes in the scope of the vernacular drama and popular writings from the Míng (1368–1644) and Qīng (1644–1911) dynasties while excluding obscure pre-Táng works of oral literature. Lǚ’s definition is sometimes criticized for its arbitrary choice of the Táng dynasty as a divide, but it is worth noting that the intermingling of Sinitic and Altaic-speaking populations in cosmopolitan Táng society accounts for the considerable linguistic gap between Middle Chinese (265–1269) 409

Chris Wen-chao Li

and Pre-Modern Chinese (1269–1795). It is also during this period that basic Chinese word order began to shift from Subject–Verb–Object (SVO) to Subject–Object–Verb (SOV) – a development often taken to be the catalyst for grammatical changes that give the modern vernacular its defining characteristics. It is worth noting also that Classical Chinese, as an abstract form of writing far removed from the Chinese spoken language, was an East Asian lingua franca readily borrowed by neighboring states, and was in wide circulation outside of China proper, often existing as an elite written language learned alongside the local vernacular, comparable to the use of Latin in the nation states of Western Europe. Up until the twentieth century, the educated classes in Japan, Korea and Vietnam possessed a reading knowledge of Classical Chinese (known as kanbun in Japanese, hanmun in Korean and hán văn in Vietnamese – all local pronunciations of the Chinese 汉文 hànwén). Being a semantic rather than phonetic means of representation, Classical Chinese lent well to borrowing across unrelated languages, and did not have a uniform pronunciation, but instead, was incorporated into the pronunciation system of the local language or dialect (Rouzer 2007: xiii–xiv).

Features of Classical Chinese A number of linguistic features serve to distinguish Classical Chinese from its present-day vernacular counterpart. Whereas Modern Chinese acquired metrical constraints that led to the development of a large polysyllabic vocabulary, lexis in the early classical period consisted, by and large, of monosyllabic forms. In the classical language, part of speech assignment is fluid, whereas in the modern language lexical items belong to fixed word classes. Noun classifiers are obligatory in Modern Chinese – not so in its classical predecessor. Classical Chinese does not require a copula, and its pronoun and determiner inventory is significantly different from that of modern-day usage. Many post-verbal prepositional phrases in Classical Chinese have moved to the pre-verbal position in Modern Chinese.

Monosyllabicity of words Classical Chinese, in its early stages, commanded a vocabulary that was predominantly monosyllabic, which stands in contrast to the largely polysyllabic lexis of Modern Chinese. This discrepancy derives chiefly from the fact that Classical Chinese, having developed out of the written mode of Old Chinese (c. 1766 bce – ce 220), represents a language whose phonological structure was vastly more complex than that of any present-day Chinese variety, and is believed to permit consonant clusters in the syllable-onset position and allow a wider range of consonantal endings in the syllable-coda position. These cluster onsets and con­ sonantal endings were merged or dropped from the third century onward due to increased contact with the Altaic languages of northern and northwestern China (Tai and Chan 1999: 232; Janhunen 1996: 165), leading to the loss of Old Chinese superheavy and bimoraic syllables and the subsequent development of disyllabic words to take their place (Feng 1998: 224–6). Hence the words for ‘ear’, ‘mouse’ and ‘son’, for example, are monosyllabic 耳 ěr, 鼠 shǔ and 子 zǐ in Classical Chinese, but are lengthened to disyllabic 耳朵 ěrduo, 老鼠 lǎoshǔ and 儿子 érzi in the modern Beijing dialect. The shift from one to two syllables is achieved via a number of different mechanisms, the most common of which is compounding. Whereas Classical Chinese had the single syllable 父 fù for ‘father’, the same term in colloquial Mandarin becomes 父亲 fùqīn – a noun compound composed of 父 fù ‘father’ and 亲 qīn ‘kinsfolk’ – both elements of which 410

Classical Chinese

have since evolved into bound morphemes that cannot be used independently as words in their own right. Similarly, Classical Chinese 国 guó ‘state’ becomes 国家 guójiā ‘state+estate’ in Modern Chinese, 友 yǒu ‘friend’ becomes 朋友 péngyǒu ‘peer+friend’, and 衣 yī ‘clothing’ must now be expressed as 衣服 yīfu ‘clothing+livery’ in the contemporary idiom. Contact between ethnic Chinese and non-Chinese speaking populations in periods of migration and upheaval served to further add polysyllabic loans to the Chinese language. The Eastern Hàn through the early Táng dynasty (first century–eighth century) saw the spread of Buddhism to China, which brought with it many now firmly ingrained conceptual abstractions, including 世界 shìjiè ‘world’, 剎那 chànà ‘moment’, 缘分 yuánfèn ‘destiny’ and 方便 fāngbiàn ‘convenience’. In the late Míng dynasty, China came into contact with European missionaries and traders, who brought with them (directly from the West and indirectly via Japanese) an influx of words for Western concepts such as 公司 gōngsī ‘company’, 机器 jīqì ‘machinery’, 自由 zìyóu ‘freedom’ and 新闻 xīnwén ‘news’. The process was accelerated after the Opium War of 1839–1842, when the encroachment of Western powers and the import of Western science led to an infusion of polysyllabic transliterated Western terms on a scale never seen before – a process which is continuing to this day, with the introduction of terms such as 咖啡 kāfēi ‘coffee’, 巧克力 qiǎokèlì ‘chocolate’, 麦克风 màikèfēng ‘microphone’, 引擎 yǐnqíng ‘engine’, 吉他 jítā ‘guitar’, 沙发 shāfā ‘sofa’, 马达 mǎdá ‘motor’ and the like. Virtually all foreign loans and transliterations from this period are two syllables or longer, so as to constitute the minimal metrical foot needed to form a prosodic word (Feng 1998: 224–30). We see this at play in the transliteration of country names: whereas polysyllabic country names can be rendered directly, e.g. Italy → 意大利 yìdàlì; Portugal → 葡萄牙 pútáoyá, monosyllabic names must attach the morpheme 国 guó ‘nation; state’ so that there is sufficient length to form a metrical foot, e.g. France → 法国 fǎguó, Deutsch(land) → 德国 déguó.

Word class conversion Classical Chinese differs from the modern idiom also in that words, in their character form, can move freely between one word class and another, exhibiting a fluidity in part-of-speech that is generally not observed in the contemporary language. In Book 12 of the Analects (c. fifth century bce), when Confucius is asked by Duke Jǐng of Qí about government, the master replies with the aphorism: 君君、臣臣、父父、子子。 Jūn jūn, chén chén, fù fù, zǐ zǐ In the master’s response, the first character of each pair represents a noun, i.e. ‘ruler’, ‘subject’, ‘father’, ‘son’, whereas the second identical character represents a verb, i.e. ‘to behave like a ruler’, ‘to behave like a subject’, ‘to behave like a father’, ‘to behave like a son’. The result is a series of clauses in which identical forms are used for subject and predicate, producing a rhetorical effect similar to that of the English: ‘Rulers rule, subjects subject, fathers father, kids kid.’ Conversions in part-of-speech in earlier stages of the Chinese language (and in their present-day remnants) are often accompanied by a change in tone, which linguists attribute to the presence of affixes in Old Chinese. The adjective 好 hǎo ‘appealing’, for example, acquires its verb form 好 hào ‘to be fond of ’ by switching to the departing tone, seen as having developed out of the Old Chinese suffix /*-s/ (Pulleyblank 2000: 29–30). Similarly, 411

Chris Wen-chao Li

in classical writings, the adjective 远 yuǎn ‘far’ alternates with the departing tone verb form yuàn ‘to keep at a distance; to keep away from’; and noun 王 wáng ‘king’ easily converts to the verb wàng ‘to be king’. Suffix /*-s/ induced conversion to the departing tone is also widely used for noun formation, as in 量 liáng ‘to measure’ → liàng ‘measurement; quantity’; 数 shǔ ‘to count’ → shù ‘numeral’; 难 nán ‘difficult’ → nàn ‘difficulty’ (Baxter and Sagart 1998: 54–5). Different Old Chinese prefixes, suffixes and infixes account for some 200 other part-of-speech conversions manifest as tone change in the present-day vernaculars (Downer 1959: 261).

Grammatical changes Classical Chinese retains many of the grammatical features of Old Chinese (c. 1766 bce – ce 220) prior to its transition to Middle Chinese (420–1260) brought about by successive waves of Altaic admixture in the second to sixth centuries, and again in the tenth to thirteenth centuries (Tai 1976: 293–4). The classical language thus exhibits a syntactic structure that is in many ways different from that of Modern Chinese. While Old Chinese has a predominantly Subject–Verb–Object (SVO) word order, Modern Chinese (as characterized by Mandarin) exhibits many features found in languages with a Subject–Object–Verb (SOV) word order (Tai 1973, 1976; S. Huang 1978; Chappell et al. 2007: 188–93). This is evident in the passive constructions of the two languages. Whereas in Old Chinese the passive voice is typically expressed with the passive marker (e.g. 于 yú) and agent positioned after the verb, e.g. 劳力者治于人。(《孟子梁惠王上》) Láolìzhě zhì yú rén (from Mencius, Book 5) ‘Laborers (are) controlled by other people.’ in Modern Standard Chinese the passive marker (e.g. 被 bèi) and agent would have to be moved before the verb in order to be grammatical, e.g. 劳力者被人统治。 Láolìzhě bèi rén tǒngzhì ‘Laborers (are) by other people controlled.’ Mirroring the development is the repositioning of the locative prepositional phrase: whereas Old Chinese typically places locative expressions after the verb, as is typical in SVO languages (cf. English ‘sail on the river’), Modern Standard Chinese moves the locative construction to a pre-verbal position when it is used to predicate the setting of the action. Consider the following passage from the classical work Zhuāng Zǐ (fourth century bce): 庄子与惠子游于濠梁之上。(《庄子•秋水篇》) Zhuāngzǐ yǔ Huìzǐ yóu yú Háoliáng zhī shàng ‘Zhuāngzǐ and Huìzǐ were strolling on a bridge on the River Hao.’ (from Zhuāng Zǐ, ‘Autumn Floods’) Such placement of ‘on a bridge on the River Hao’ after the verb ‘strolling’ would be ungrammatical in the modern language, which would have to rephrase the sentence as follows, with the locative expression preposed: 412

Classical Chinese

庄子与惠子在濠梁之上游走。 Zhuāngzǐ yǔ Huìzǐ zài Háoliáng zhī shàng yóuzǒu ‘Zhuāngzǐ and Huìzǐ (were) on a bridge on the River Hao strolling.’ Another area of divergence is the use of noun classifiers. In Modern Chinese, all nouns behave like English mass nouns, requiring the positioning of a nominal classifier between a number and the noun itself (e.g. 一张纸 yì zhāng zhǐ ‘a piece of paper’, 一根草 yì gēn cǎo ‘a blade of grass’), whereas in Old Chinese nouns were often free to behave as count nouns, which can be directly preceded by a number or a demonstrative (e.g. 一羊 yì yáng ‘one sheep’, 二马 èr mǎ ‘two horses’). Thus whereas Classical Chinese might have 一牛 yìniú ‘one ox’ or 五柳 wǔ liǔ ‘five willow trees’, in Modern Standard Chinese obligatory noun classifiers would have to be assigned to each of the nouns, producing 一头牛 yì tóu niú ‘one head of oxen’ and 五棵柳树 wǔ kē liǔshù ‘five sprouts of willow trees’. The shift from count to mass status for nouns is believed to have been initiated around the second century bce (Wang 1994: 107–12), when noun classifiers began to expand in number and diversify in categorical function, but they appear not to have evolved into an obligatory fixture in ways resembling classifiers in Modern Standard Chinese until at least the first wave of Altaic contact in the second to sixth centuries – considered by many as a turning point in the development of Chinese classifiers (Liú 1965: 27; Wang 1994: 112–13), by which time classifiers began appearing in large numbers in works such as the《世说新语》Shìshuō Xīnyǔ (A New Account of the Tales of the World, fifth century),〈颜氏家训〉Yánshì Jiāxùn (‘Family Traditions of the Yán Clan’, sixth century), and〈洛阳伽蓝记〉Luòyáng Qiélánjì (‘Buddhist Monasteries of Luòyáng’, sixth century) in a manner for the most part identical to contemporary usage. Classical Chinese also differs somewhat from the modern idiom in its inventory of personal pronouns, although when taking into account inflectional variation and historical sound change the differences may not be as pronounced as suggested by their divergent orthographic representations. We encounter the first person pronouns yú (余 or 予), wú (吾) and wǒ (我) in classical writings. In the pre-classical oracle inscriptions of the late Shāng dynasty (1300–1100 bce), yú (余 or 予) is used for the first person singular, whereas wǒ (我) is used for the first person plural (Sagart 1999: 142–43). By the Western Zhōu dynasty, in classical writings such as the Analects and Mencius, yú (余 or 予) falls out of favor, to be replaced by wú 吾 (Middle Chinese /*ŋɔ/) and wǒ 我 (Middle Chinese /*ŋɑ/), the former used more commonly in the subject position or as a possessive attribute, and the latter found predominantly in the object position (Norman 1988: 89) – a pattern of complementary distribution that, paired with their phonological affinities, has prompted scholars such as Karlgren (1920) to posit the existence of a case system in Old Chinese similar to that present in the European languages. In later writings, the case distinction between wú 吾 and wǒ 我 is blurred, although both forms continue to coexist in the literary language. In contemporary Chinese only wǒ 我 remains as the sole first person pronoun. In Classical Chinese, the most common second person pronouns are ěr (爾 – Middle Chinese /*ɲi/) and rǔ (汝 or 女 – Middle Chinese /*ɲio/), both of which occur in preclassical writings of the Western Zhōu period, first exclusively as a second person plural, then as a second person singular as well (Sagart 1999: 142–3). Note the similarities between the two phonological forms, suggestive of an inflectional relationship between the pair, although, unlike with the Classical Chinese first person pronouns, no clear pattern of complementary distribution has been identified to date. The contemporary second person pronoun nǐ 你 is believed to have developed out of a northwestern reflex of the ěr 爾 form (Sagart 413

Chris Wen-chao Li

1999: 143) – note the similarity with the Middle Chinese /*ɲi/ pronunciation, and the fact that the graphic component 尔 is an alternate form of the character 爾. Classical Chinese, strictly speaking, does not have a third person pronoun in the nominative case, unlike Modern Standard Chinese, which has the lexeme tā (他, 她 or 它) at its disposal, most likely the remnant of a pre-classical spoken form (Sagart 1999: 146). We do see in classical writings however the use of qí 其 as a third person possessive, and zhī 之 as a third person object – both still used in stock expressions in more formal contemporary writings, but absent from spontaneous contemporary vernacular speech. It is worth noting also that, since the advent of the classical period, Chinese pronouns no longer distinguished between singular and plural, instead relying on add-on morphemes such as děng 等 ‘ranks’, chái 侪 ‘peers’, bèi 辈 ‘generations’ to indicate collectivity when the occasion called for such. The use of men 们 as a personal collective marker in Modern Standard Chinese is the latest development in this trend, which some argue came about via influence from Mongolian or other Altaic languages in the Yuán (1271–1368) and Míng (1368–1644) dynasties (Haenisch 1931; Iljic 2001: 77–81). Finally, Modern Standard Chinese requires the copula shì 是 for noun predication, a development in the colloquial language of the early Hàn dynasty (second to first century bce) which never quite made its way into the language of formal writings (Pulleyblank 2000: 45). The classical language continued to use the partical yě 也 in an ‘A = B yě’ pattern for identification and classification purposes, while in the spoken language the function was taken over by the new essive shì 是, evolved out of the classical demonstrative pronoun. Thus a sentence like ‘Dǒng Hú is an exemplary historian of antiquity’ would appear in Classical Chinese as 董狐,古之良史也(《左传宣公二年》) Dǒng Hú, gǔ zhī liáng shǐ yě (from the Chronicles of Zuǒ, Book VII) ‘Dǒng Hú, antiquity’s exemplary historian [particle]’ In the modern idiom, however, the particle yě 也 is discarded and the copula shì 是 is inserted between the patient and the classificatory element, giving 董狐是古代的优良史官。 Dǒng Hú shì gǔdài de yōuliáng shǐguān ‘Dǒng Hú is antiquity’s exemplary historian.’

Coexistence with the vernacular Up until the twentieth century, Classical Chinese and vernacular Chinese exhibited pervasive and rigid compartmentization, in which the former was used exclusively by the educated and ruling classes for formal writings, and the latter by the general populace for linguistic settings perceived to be more informal and intimate. This stable divide between an elite, semi-archaic ‘high language’ (H) and a more down-to-earth ‘low language’ (L) is termed ‘diglossia’ by linguists. The Classical–vernacular divide in Chinese, stretching over thousands of years, is seen as a textbook example of diglossia. Linguist Charles Ferguson, who coined the term ‘diglossia’ in 1959, describes Chinese as representing ‘diglossia on the largest scale of any attested instance’ (1959: 337–8). The complementary roles served by the literary and vernacular languages in pre-modern China are in many ways typical of the roles of H and L languages in diglossia, which, in 414

Classical Chinese

the classic sense of the word, implies specialization of function, namely ‘in one set of situations only H is appropriate, and in another only L, with the two sets overlapping only very slightly’ (Ferguson 1959: 328). Such is the case with literary and vernacular Chinese, the former of which was perceived as being a language suited to the expression of sophisticated and elegant thought (Snow 2010: 160), the only vehicle deemed suitable for writing, as it is the language of ‘all works making the least claim to correctness, propriety and chasteness’, owing to which ‘no person would deem his productions fit for the public gaze, and worthy of imitation, who did not write in this style’ (Letter from biblical translators Walter Medhurst, Alexander Stronach, and William Milne to the London Missionary Society (1851), in Zetzsche 1999: 93). Furthermore, in classic diglossia, the H language is typically a written variety which no segment of the population can claim as a mother tongue (Coulmas 1987: 117) – a description which applies fittingly to Classical Chinese, which is often learned through tutors in private academies by those with the means to receive a formal education. For the past two thousand years, the classical language was nobody’s native language, nor has it ever been used by any community for daily conversation (Snow 2010: 160). The H language in classic diglossia often develops out of archaisms frozen by social conventions (Bright 1976: 66), as is the case with Classical Chinese, which traces its roots to canonical writings of the Warring States period, after which time writers continued to emulate the grammar and lexis of this early literary style, forcing the written language to acquire an archaic flavor as the spoken language of the day underwent a very different and by and large independent development (Norman 1988: 83). Perpetuation of this prestige language was helped along by its social status and literary heritage. Furthermore, as access to the more formal occasions which call for the use of the literary language is disproportionately distributed in favor of the educated, literate or otherwise privileged classes most likely to have received formal schooling, the diglossic divide is often characteristic of traditions of restricted literacy in speech communities that are overwhelmingly illiterate (Hudson 2002: 5–6; Walters 1996: 161–2) – as was the case in China, where mastery of Classical Chinese was stock and staple of imperial examinations that open the door to officialdom, often the only path to wealth, status and prestige for those without families ties or background, and was thus viewed by many as literally a road to power. Despite the appearance of class mobility however, the general populace often did not have sufficient schooling to become viable candidates in the examinations, without which the substantial differences between the classical language and the spoken vernaculars effectively served to bar the uneducated masses from civil service. Furthermore, traditional Chinese social hierarchy virtually guaranteed little motivation for the educated elite to promote knowledge of the classical language among the population at large (Snow 2010: 160–1). Ferguson (1959: 332) writes that the type of diglossia typical of the relationship between Classical and vernacular Chinese can persist at least several centuries, and that there is evidence in some cases that it can last well over a thousand years. But it has also been observed that classic diglossia tends to thrive in pre-industrialized civil societies with restricted literacy, and most often disintegrates at the onset of modernization (Neustupny 1974: 40), as it is broken down by the processes of modernization, urbanization, mercantalism and industrialization, which create the need for a literate labor force, and by the demise of small ruling groups, the breakdown of rigid class barriers, and the democratization of education, literacy and knowledge that tend to accompany each other (Hudson 2002: 32). The result is often displacement of the classical language by a new standard more closely related to certain educated strains of the spoken vernacular (Hudson 2002: 30). The process is welldocumented in China, as reflected in the writings of European missionaries, who, early in 415

Chris Wen-chao Li

the eighteenth and nineteenth centuries, initially attempted to translate the Protestant Bible into wénlǐ (文理) – their term for Literary Chinese – seen then as ‘the chaste and correct style of language’ and ‘the classical style in which the Commentaries on the Sacred Books are written’, but later resorted to various compromises between the literary and vernacular languages, and finally, in the late nineteenth and early twentieth centuries, discarded the wénlǐ translations in favor of the now widespread Union Bible version in colloquial Mandarin (Zetzsche 1999). In diglossic speech communities, decline of the classical language is often accompanied by catastrophic political events involving the collapse of classical society itself (Hudson 2002: 34), out of which a newly emerged social structure precipitates the birth of a new literary language molded out of the spoken language then current (Pulgram 1950: 461–2). As China faced encroachment from Japan and the West in the late eighteenth and early nineteenth centuries, there was a growing sentiment that China needed to strengthen itself by promoting mass literacy and education, and that Classical Chinese was not an ideal vehicle for this means, partly because of its association with the weakening empire and traditional values, and partly because it was too unwieldy to teach (Snow 2010: 161). The monopoly of Classical Chinese as the sole vehicle for writing began its steep decline after the Opium War of 1840–1842, as intellectuals, more and more, saw the gap between the spoken and written languages as a hindrance to greater literacy, and called for the removal of the classical language from education and media as part and parcel of the modernization of China. In the New Culture Movement (新文化运动 Xīn Wénhuà Yùndòng) of the late 1910s, promotion of vernacular Chinese gathered momentum through the efforts of noted scholars such as 胡适 Hú Shì (1891–1962), 陈独秀 Chén Dúxiù (1879–1942), 钱玄同 Qián Xuántóng (1887–1939), 刘半农 Liú Bànnóng (1891–1934), and 傅斯年 Fù Sīnián (1896– 1950), culminating in the Vernacular Language Movement (白话文运动 Báihuàwén Yùndòng) of 1917–1919. The success of the movement led to the adoption of vernacular Chinese as the language of textbooks, and influential new works of literature by authors such as 鲁迅 Lǔ Xùn (1881–1936), 老舍 Lǎo Shě (1899–1966), 徐志摩 Xú Zhìmó (1896–1931), and 沈从文 Shěn Cóngwén (1902–1988) began appearing in the vernacular. The abrupt and wholesale replacement of Classical Chinese with vernacular writing in the 1910s coincided largely with the end of imperial rule in China – this happening within less than two decades of the adoption of Western education in China (1898), the abolition of Confucian-style civil service examinations (1905), and the overthrow of the Qīng dynasty (1912). The result of this paradigm shift, officially sanctioned in 1920, was the decision to write in an approximation of contemporary northern vernacular speech, thereby discontinuing the centuries-old practice among literate individuals of writing in the classical style, which had long since ceased to function as a medium of oral communication; in other words, the elevation to prominence of vernacular writing – a writing style always available within society but previously unsanctioned for the purpose of formal and serious writing – was now in full swing (Barnes 1982: 262). While the Vernacular Language Movement of 1917–1919 is viewed today as largely successful, the end result, as is the case in many post-diglossic communities, is not a complete replacement of the literary language with the vernacular, but rather, to a large degree, a merger of the outgoing and incoming norms. It has been noted that when a literary language is replaced or partially merged with the vernacular to produce a new standard, the lexicon of the old classical language, in particular, lives on in the newly minted vernacular-based norm in the form of higher order terminology, especially in areas relating to technology, cultural legacy and abstract reasoning (Kahane and Kahane 1979: 194). Lexicon aside, 416

Classical Chinese

stylistic constraints serve to further distance the new written language from its colloquial counterpart, as languages universally require that the structure of written texts be less casual and more elevated than the natural state of spoken utterances, such that speech communities generally do not sanction ordinary, everyday speech for written use (Hudson 2002: 24; Ferguson 1968: 29–30). Such is the case with Modern Standard Chinese, in which the grammar of the standard written language includes not only the syntax of the vernacular but also remnants of Classical Chinese convention that have made their way into modern standard writing (Zhū 1988: 132). Furthermore, in the contemporary language, there is substantial incorporation of literary elements such as truncated terms, four-character idioms, stock expressions, and classical grammatical constructs (DeFrancis 1984: 244). Unique to Chinese is the fact that phonology plays a role in the choice between literary and colloquial registers, as the modern language is subject to metrical constraints requiring quasi-literary disyllabic forms in a great number of word formation templates (Duanmu 1999; Feng 2006). In other words, Modern Standard Chinese is characterized by ways of amalgamating Classical Chinese with vernacular style that are essentially motivated and licensed by prosody (Feng 2006: 17) – the result of which is a marked distinction between the written and spoken languages, which, while not as glaring as the literary–vernacular divide in pre-modern times, nevertheless, is obligatory and clearly noticeable to the untrained eye. Early Republican philologist Huáng Kǎn 黄侃 (2001: 199) ascribes the divergence to such general tendencies in language as reverence for the past, the need for formality, and the conservative nature of the written medium. In the present age, Classical Chinese, while no longer the medium de rigueur for written communications, lives on in government missives, legal documents, business contracts, academic theses, political speeches and news commentaries – in effect all manner of writing deemed formal. Classical patterns and set expressions are ubiquitous in vernacular prose – more so in Taiwan and Hong Kong than in mainland China – so much so that, with the exception of dialog, contemporary writing that does not incorporate elements of the classical language is rarely considered fit for print. In the Chinese-speaking societies of mainland China, Taiwan and Hong Kong, attitudes toward vernacular writing have shifted since the 1980s, moving away from the May Fourth Movement ideal of complete emulation of oral speech to a more measured approach which sees classical elements as not only desirable but virtually necessary for stylistic effect and rhetorical functions. In all three regions the classical language is taught as part of the Chinese language curriculum from the upper levels of primary school through to universitylevel general education, and is a required subject in entrance examinations for high school and college. The strongest advocacy for reading proficiency in Classical Chinese is to be found in Taiwan, where promotion of traditional culture and values was officially sanctioned in the 1960s as a countermeasure to the Cultural Revolution (1966–1976) on the mainland, where anti-bourgeois sentiment and the purging of ‘old customs, old culture, old habits and old ideas’ effectively discouraged mastery of the literary language. Taiwan followed a decidedly different route, where, after the latest round of reforms in 2010, Classical Chinese texts currently account for some 45% to 65% of the high school Chinese language curriculum, and the four Confucian classics of the Analects, the Doctrine of the Mean, the Great Learning and Mencius are part of the required core. In mainland China, amidst recent calls to revive the study of Classical Chinese, the current percentage of classical texts in the secondary curriculum stands at around 40%; in Hong Kong, close to half of all Chinese language texts studied at the secondary level are pre-modern texts written in the classical language. 417

Chris Wen-chao Li

References Barnes, Dayle (1982) ‘Nationalism and the Mandarin Movement: The First Half-Century’, in Robert L. Cooper (ed) Language Spread: Studies in Diffusion and Social Change, Bloomington: Indiana University Press, 260–90. Baxter, William H. and Laurent Sagart (1998) ‘Word Formation in Old Chinese’, in Jerome L. Packard (ed) New Approaches to Chinese Word Formation: Morphology, Phonology and the Lexicon in Modern and Ancient Chinese, Berlin: Mouton de Gruyter, 35–76. Bright, William (1976) ‘Phonological Rules in Literary and Colloquial Kannada’, in Anwar S. Dil (ed) Variation and Change in Language: Essays by William Bright, Stanford: Stanford University Press, 65–73. Chappell, Hilary, Li Ming and Alain Peyraube (2007) ‘Chinese Linguistics and Typology: The State of the Art’, Linguistic Typology 11(1): 187–211. Coulmas, Florian (1987) ‘What Writing Can Do to Language: Some Preliminary Remarks’, in S. Battestini (ed) Developments in Linguistics and Semiotics, Language Teaching and Learning, Communication Across Cultures (Georgetown University Round Table on Language and Linguistics 1986), Washington D. C.: Georgetown University Press, 107–29. DeFrancis, John (1984) The Chinese Language: Fact and Fantasy, Honolulu: University of Hawaii Press. Downer, Gordon B. (1959) ‘Derivation by Tone-Change in Classical Chinese’, Bulletin of the School of Oriental and African Studies, University of London, 22(1–3): 258–90. Duanmu, San (1999) ‘Stress and the Development of Disyllabic Words in Chinese’, Diachronica 16(1): 1–35. Feng, Shengli (1998) ‘Prosodic Structure and Compound Words in Classical Chinese’, in Jerome L. Packard (ed) New Approaches to Chinese Word Formation: Morphology, Phonology and the Lexicon in Modern and Ancient Chinese, Berlin: Mouton de Gruyter, 197–260. Feng, Shengli (2006) Expressions of Written Chinese, Beijing: Beijing Language and Culture University Press. Ferguson, Charles (1959) ‘Diglossia’, Southwest Journal of Linguistics 10(1): 214–34. Ferguson, Charles (1968) ‘Language Development’, in J. Fishman, C. Ferguson and J. Das Guptas (eds) Language Problems of Developing Nations, New York: Wiley, 27–35. Haenisch, Erich (1931) Untersuchungen über das Yüan-Chʻao Pi-shi, die geheime Geschichte der Mongolen (Abhandlungen der philologisch-historischen Klasse der Siichischen Akademie der Wissenschaften 41/4), Leipzig: Verlag S. Hirzel. Hú, Shì 胡适 (1928) Báihuà Wénxué Shǐ《白话文学史》(A History of Vernacular Literature), Shanghai: Xīnyuè Shūdiàn. Huáng, Kǎn 黄侃 (2001) Huáng Kǎn rìjì《黄侃日记》(Diary of Huáng Kǎn), Nanjing: Jiāngsū Jiàoyù Chūbǎnshè. Huang, Shuan-fan (1978) ‘Historical Change of Prepositions and Emergence of SOV Order’, Journal of Chinese Linguistics 6: 212–42. Hudson, Alan (2002) ‘Outline of a Theory of Diglossia’, International Journal of the Sociology of Language 157: 1–48. Iljic, Robert (2001) ‘The Origin of the Suffix -men 们 in Chinese’ Bulletin of the School of Oriental and African Studies, University of London 64(1): 74–97. Janhunen, Juha (1996) Manchuria: An Ethnic History, Helsinki: Finno-Ugrian Society. Kahane, Henry and Renée Kahane (1979) ‘Decline and Survival of Western Prestige Languages’, Language 55: 183–98. Karlgren, Bernhard (1920) ‘Le proto-chinois, langue flexionelle’, Journal Asiatique 15: 205–33. Liú, Shìrú 刘世儒 (1965) Wèijìn Nánběicháo Liàngcí Yánjiū《魏晋南北朝量词研究》(A Study of Classifiers in the Weijin Nanbeichao Period ), Beijing: Zhonghua Publishing House. Lǚ, Shúxiāng 吕叔湘 (1944) ‘Wenyan yu Baihua’《文言与白话》(Classical and Vernacular Chinese), Guówén Zázhì《国文杂志》3(1): 3–12. Neustupny, Jiri Vaclav (1974) ‘The Modernization of the Japanese System of Communication’, Language in Society 3(1): 33–50. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Pulgram, Ernst (1950) ‘Spoken and Written Latin’, Language 26: 458–66. Pulleyblank, Edwin G. (2000) ‘Morphology in Old Chinese’, Journal of Chinese Linguistics 28(1): 26–51.

418

Classical Chinese Rouzer, Paul (2007) A New Practical Primer of Literary Chinese, Cambridge, MA: Harvard East Asian Monographs. Sagart, Laurent (1999) The Roots of Old Chinese, Amsterdam and Philadelphia: John Benjamins. Snow, Don (2010) ‘Hong Kong and Modern Diglossia’, International Journal of the Sociology of Language 206: 155–79. Tai, James H.-Y. (1973) ‘Chinese as an SOV Language’, in Claudia W. Corum, Thomas Cedric SmithStark, and Ann Weiser (eds) Papers from the Ninth Regional Meeting, Chicago Linguistic Society, 659–71. Tai, James H.-Y. (1976) ‘On the Change from SVO to SOV in Chinese’, in Sanford B. Steever, Carol A. Walker and Salikoko S. Mufwene (eds) Papers from the Parasession on Diachronic Syntax, Chicago Linguistic Society, 291–304. Tai, James H.-Y., and Marjorie K. M. Chan (1999) ‘Some Reflections on the Periodization of the Chinese Language’, in Alain Peyraube and Chaofen Sun (eds) Studies in Chinese Historical Syntax and Morphology: Linguistic Essays in Honor of Mei Tsu-lin, Paris: Ecole des Hautes Etudes en Sciences Sociales, 223–9. Walters, Keith (1996) ‘Diglossia, Linguistic Variation, and Language Change in Arabic’, in M. Eid (ed) Perspectives on Arabic Linguistics VIII: Papers from the Eighth Annual Symposium on Arabic Linguistics, Amsterdam and Philadelphia: John Benjamins, 157–97. Wang, Lianqing (1994) ‘Origin and Development of Classifiers in Chinese’, Doctoral Dissertation, The Ohio State University. Zetzsche, Jost Oliver (1999) The Bible in China: The History of the Union Version or the Culmination of Protestant Missionary Bible Translation in China, Sankt Augustin, Germany: Monumenta Serica Institute. Zhū, Déxī 朱德熙 (1988) ‘Hànyǔ〈汉语〉(Chinese), in Encyclopedia of China – Language Volume 《中国大百科全书: 语言文字》, Shanghai: Encyclopedia of China Publishing House, 128–33.

419

24 Computational linguistics Xiaoheng Zhang Hong Kong Polytechnic University, Hong Kong, China

1. Introduction Computational Linguistics is the scientific study and information processing of language using computers. The purpose is to obtain a better understanding of how languages work and to bring more convenience to language application. Like general linguistics, computational linguistics pays more attention to natural languages (such as Chinese and English, which have naturally evolved through human history) rather than artificial languages (such as logic and programming languages, which were artificially designed). The term ‘Computational Linguistics’ is often employed interchangeably with ‘Natural Language Processing’, though the former usually refers to more theoretical studies while the latter more technical. Rather than introducing computational linguistics in a general sense (Malmkjaer 2010), this chapter will focus on the Chinese language, especially its differences with English and other languages. Chinese computational linguistics is often referred to as ‘Chinese information processing’ 中文信息处理 in China. The following sections will introduce the important areas of Chinese character information processing, word segmentation, natural language understanding and generation, corpus linguistics, and machine translation.

2.  Chinese character input In the English writing system, we use a few dozen characters. In Chinese, we need a much larger character set. There are over 10,000 characters in the Xinhua Dictionary《新华字典》, the most popular pocket dictionary in China (Language Institute of the Chinese Academy of Social Sciences 2011). In the Unicode multilingual character set of about 110,000 characters, over 75,000 (or 68.182%) are Chinese (Unicode Consortium 2014). That means computer processing of Chinese characters is the toughest among other languages. In the present and the following two sections, we will introduce computer input, encoding, and output of Chinese characters, which are the cornerstones of Chinese computational linguistics and information processing. Computer input of Chinese characters is by no means as easy as English. English is written with 26 letters and a handful of other characters, and each character is assigned to a key on the keyboard. Chinese can be input in a similar way. However, that would involve a huge keyboard with thousands of keys. Searching for a character on the keyboard would be a daunting job. 420

Computational linguistics

People did try to ‘shrink’ the Chinese keyboard by putting multiple characters on one key. That turned the original one-step input procedure into two steps for the writer: (i) pressing the key with the character group of your target character, (ii) selecting the target character in the group. The resulting keyboard still remained clumsy, because if you put more characters on one key, the key becomes bigger to make the characters recognizable, and selecting a character from a large group is difficult. Additionally, it is not easy to group the characters evenly in a reasonable and easy-to-learn way. Another drawback of a Chinese keyboard for direct character input is its inconsistency with English input. An alternative way is to encode each Chinese character in English characters, enabling Chinese input on an English keyboard. As a matter of fact, this method has become predominant for Chinese computer input. The software of an encoding input method includes a character-code table (called mabiao or 码表 in Chinese). When an ASCII input code is typed on the English keyboard, the software will search for matching Chinese characters in the table. If there are multiple characters sharing the same code, they will be presented to the user for selection. To make the input method easy to learn, encoding must be based on distinctive features in forms, sounds or meanings of Chinese characters. Because the meaning(s) of a character tend to be more abstract and complicated, encoding is normally based on the sound or form.

2.1.  Sound-based encoding of Chinese character input Sound-based encoding is normally based on an existing Latin character scheme for Chinese phonetics, such as the Pinyin Scheme for Putonghua (National Language Commission of China 1958), and the Jyutping Scheme for the Cantonese dialect (Linguistic Society of Hong Kong 1993). The input code of a Chinese character is its Pinyin letter string followed by an optional number representing the tone. For example, the Putonghua Pinyin input code of 香港 (Hong Kong) is xiang1gang3, and the Cantonese Jyutping code is hoeng1gong2, both of which can be easily input via an English keyboard. In Putonghua Pinyin, there are two letters not appearing on the English keyboard: ê and ü. According to the national standard (National Language Commission of China 2001), ê should be represented by ‘ea’, and ü by ‘v’ in the Pinyin input code. In some Chinese input software ê is also represented as ‘e^’, and ü as ‘u:’ or ‘uu’. The most popular sound-based input methods in Chinese are Microsoft Pinyin, Sogou Pinyin, Google Pinyin, and Jyutping in mainland China and Hong Kong, and New Phonetic (Bopomofo) in Taiwan. There are a number of advantages for sound-based encoding: (i) Easy to learn because most Chinese writers have already got a good command of Putonghua and Pinyin, (ii) Consistent with Chinese language learning, (iii) Allows simplified and traditional Chinese characters to be input in a similar way. The shortcomings of sound-based encoding lie in its high degree of duplicate encoding, with homophone Chinese characters sharing the same code. A Chinese character is normally pronounced with one syllable. Chinese Putonghua only has about 400 different syllables without considering tones, or approximately 1,200 syllables when tones are considered. On the other hand, there are tens of thousands of Chinese characters. That means on average, each syllable has to cover over ten characters. This problem can be largely solved by inputting Chinese word by word instead of character by character, because most words in Modern Chinese consist of more than one character and duplicate encoding is much less frequent at word level. For example, the Pinyin of 香港 (Hong Kong) is unique to the word, while either 421

Xiaoheng Zhang

character 香 or 港 shares its pronunciation with many other characters. Another limitation of sound-based Chinese input is that you must know the pronunciation of a Chinese character before you can input it into the computer. This issue can be solved by form-based encoding.

2.2.  Form-based encoding of Chinese character input A Chinese character can alternatively be input according to its form and structure. Most Chinese characters can be divided into a sequence of components each of which is in turn composed of a sequence of strokes in writing order. For example, Chinese character 福 (good fortune, happiness) can be decomposed as Character: 福 Components: 礻, 一, 口, 田 Strokes: 丶 丨丶, 一, 丨

一, 丨

一丨一.

There are a few hundred basic components (National Language Commission of China 1997), much less than the number of characters. If we represent each component with an English letter and put them in writing order of the character, we get a letter string ready to be used as an input code on the English keyboard. Of course we can also design a rule to select representative letters from the string if it is too long. For example, in the Cangjie input method, character 疆 jiang is encoded as NGMWM corresponding to components 弓土一田一, with some components omitted. Stroke-based coding is simpler than component-based coding. But the codes tend to be longer. There are approximately 30 distinctive strokes of Chinese characters (Unicode Consortium 2014: Table F-1: CJK Strokes). They are usually classified into five categories of heng (一), shu (丨), pie (丿), dian (丶) and zhe (  ) for dictionary consultancy and Chinese input on a mobile phone. For Chinese input with an ASCII keyboard, two strokes can be combined to form 5*5=25 different pairs for mapping to English letters. For example, in input method ZYQ (Zhang 2003), the sequence of stroke pairs ‘一一, 一丨, 一丿, .  .  .  , 丿, 丶, ’ are represented by ‘a, b, c,  .  .  .  , w, x, y’ respectively. Popular form-based encoding methods include Wubi 五笔 in mainland China, Boshiamy (呒虾米) in Taiwan, and Cangjie (仓颉) in Taiwan and Hong Kong. The pros and cons of form-based input methods are complementary to sound-based methods. The major advantage of form-based methods lies in their low degree of duplicate encoding, enabling high speed input of Chinese characters. And the major shortcoming is difficulty of learning. Normally students have to remember over one hundred components and their corresponding English letters. In addition, they have to learn the complicated rules for breaking a character into a sequence of components and making a selection among them.

2.3.  Chinese character recognition and speech recognition Chinese characters can also be input into the computer by optical character recognition (OCR), handwriting recognition, and speech recognition based on technology similar to that of English (White 2008). Compared with English, Chinese OCR and handwriting recognition is more difficult, because there are thousands of different characters instead of 26 letters. Generally speaking, print character recognition is more accurate than handwriting characters because their forms 422

Computational linguistics

are more standardized. There are OCR tools for different fonts, including the popular Song, Kai, and Hei. In comparison with offline handwriting, online handwriting recognition is more efficient, because the computer not only ‘sees’ the written character but also the procedure of writing it. Speech recognition converts a continuous speech signal into a sequence of words. There are two problems: the variation in pronunciation of words by different speakers and the existence of homophones, i.e. words of the same pronunciation such as ‘pair’, ‘pear’, and ‘pare’ in English, and 攻势, 公式, 公示 (gong1shi4) in Chinese. Speech recognition relies on corpus statistical methods and linguistic rules. A helpful feature of Chinese is that each character is pronounced with one syllable. Both Chinese character recognition and speech recognition has reached application level. However neither can guarantee 100% correctness without human proofreading or online character selection.

2.4.  Intelligent Chinese character input The most important feature of intelligent input is application of contextual constraints for candidate characters selection. For example, using Microsoft Pinyin, when we typed input code daxuejiaoshou we got 大学教授 (university professor), when we typed daxuepiaopiao the computer suggested 大雪飘飘 (heavy snow flying). Though the non-toned Pinyin letters of 大学 and 大雪 are both daxue, the computer made a reasonable selection based on the subsequent words. Intelligent Chinese input also makes use of corpus information and linguistic rules. The computer’s selection of ambiguous Chinese characters is not always correct, and further improvement is required.

2.5.  Non-Hanzi Chinese characters input There are Chinese characters which are not complete Han ideographs or Hanzi (汉字, 漢字) in Modern Chinese. Non-Hanzi Chinese characters include punctuation marks (such and 乚), Hanzi radicals (such as 氵, 宀 as 。, 、and《》), Hanzi strokes (such as 丿, and 刂), and Pinyin toned letters (letters with standard overhead marks of tones, such as ā, á, ǎ, à). There are facilities available on MS Windows, MS Office, and the Internet which will enable us to input almost all of these Chinese auxiliary characters, ranging from the input of punctuation marks in general Chinese input methods, to inputting toned Pinyin letters with soft keyboards, to inputting strokes and radicals from the Unicode website and by Unicode-character conversion, as well as the application of special tools on the Web to input Pinyin and other characters. More information on non-Hanzi Chinese character input can be found in a recently published paper (Zhang 2012), which includes a list of 280 non-ASCII non-Hanzi Chinese characters, each annotated with its Unicode and the input code of the author’s design. You can input a character on MS Word by typing its Unicode and pressing keys Alt+X.

3.  Chinese character sets and encodings Inside the computer each character is represented by an internal code. When a character is sent between two machines, it is in information interchange code. Nowadays, information 423

Xiaoheng Zhang

interchange codes, such as ASCII and Unicode, are often directly employed as internal codes. The following sections will introduce the most important encoding standards used in Chinese information technology, including GB, Big5, and Unicode.

3.1.  The GB standard GB stands for Guojia Biaozhun (国家标准, or ‘national standard’) in Putonghua, and is the prefix for reference numbers of official standards issued by the People’s Republic of China. The first GB Chinese character encoding standard is GB2312, which was released in 1980. It includes 6,763 Chinese characters, with 3,755 frequently used ones sorted by Pinyin, and the rest by radicals (部首字序). GB2312 was designed for simplified Chinese characters. Traditional characters which have been simplified are not covered. The code of a character is represented by a two-byte hexadecimal number, for instance, the GB codes of 香港 (Hong Kong) are CFE3 and B8DB respectively. GB2312 is still in use on some computers and the Internet, though newer versions with extended character sets, such as GB13000.1 and GB18030, have been released. The latest version of GB encoding is GB18030, published in 2005. GB18030 supports both simplified and traditional Chinese characters, and is consistent with Unicode’s character set.

3.2.  The Big5 standard Big5 encoding was designed by five big IT companies in Taiwan in the early 1980s, and has been the de facto standard for representing traditional Chinese in computers ever since. Big5 is popularly used in Taiwan, Hong Kong, and Macau. The original Big5 standard included 13,053 Chinese characters, with no simplified characters. Each character is encoded with a two-byte hexadecimal code, for example, 香 (ADBB) 港 (B4E4) 龍 (C073). Chinese characters in the Big5 character set are arranged in radical order. Extended versions of Big5 include Big-5E and Big5-2003, which include some simplified characters and Hong Kong Cantonese characters.

3.3.  The Unicode standard Unicode is the most influential international standard for multilingual character encoding. It is consistent with (or virtually equivalent to) standard ISO/IEC10646. The full version of Unicode represents a character with a four-byte digital code, providing a huge encoding space to cover all characters of all languages in the world. The Basic Multilingual Plane (BMP) is a two-byte kernel version of Unicode with 216=65,536 code points for important characters of many languages. There are 27,522 characters in the CJKV (China, Japan, Korea, and Vietnam) Ideographs Area, including all the simplified and traditional Chinese characters in GB2312 and Big5 traditional. The latest version of Unicode is Unicode 7.0 (Unicode Consortium 2014), with a multilingual character set of over 110,000 characters, among which over 75,000 are Chinese sorted by Kangxi Radicals (康熙部首). Even very rarely used characters are available. The following are some example characters with their Unicode in brackets: 424

Computational linguistics

H (0048) K (004B), 香 (9999) 港 (6E2F), 龍 (9F8D), 龙 (9F99), 龖 (9F96), (2A6A5). 龘 (9F98), All the 5,009 characters of the Hong Kong Supplementary Character Set (HKSCS, http://www.ogcio.gov.hk/en/business/tech_promotion/ccli/hkscs/) are included in Unicode 7.0. HKSCS was developed by the Hong Kong government as a collection of locally specific Chinese characters not available on the computer in the early days, for instance 咗 (already), 嘢 (thing), 脷 (tongue), and 曱甴 (cockroach). As GB, Big5, and Unicode are concurrently used in Chinese encoding, when the computer mistakenly interprets a text with an encoding standard different from its original code, it will be presented with wrong characters, a phenomena called luànmǎ (code confusing), which occasionally happens on the Web or in emails. This problem is often solved by manual selection of encoding or character set (such as the case on Web browsers) or by code conversion beforehand. Unicode is becoming more and more popular. According to a recent report on Google Official Blog, Unicode has covered over 60% of text on the Web (Davis 2012). If we include ASCII, which is compatible with Unicode, nearly 80% of Web documents are in Unicode. It is widely believed that Unicode will ultimately replace all other information interchange encodings, and there will be no more code confusing.

4.  Chinese character output Like English and other languages, Chinese characters are output on printers and screens in different fonts and styles. The most popular Chinese fonts are the Song (宋体), Kai (楷体), Hei (黑体), and Fangsong (仿宋体) families, for example, 中文计算语言学 中文计算语言学 中文计算语言学 中文计算语言学

(Song) (Kai) (Hei or Black) (FangSong)

Fonts appear in different sizes. In addition to the international measurement system of points, Chinese characters are also measured by size numbers (called zihao 字号) invented by an American for Chinese printing in 1859. Table 24.1 is a list of all the font sizes in numbers available on Chinese version MS Word and their equivalent points (Zhang 2006). Table 24.1 is particularly useful for Chinese typesetting on computers not supporting font sizes in numbers. For example, from Table 24.1, we get to know that Chinese size number 3 (三号) is equivalent to 16 points, or 5.62mm high, as shown by the example characters. The image of a Chinese character in a particular font is represented in the computer by a matrix of dots (called dot matrix fonts or bitmapped font) or by outlines (called outline fonts), again like the case in English (White 2008).

5.  Chinese word segmentation and proper noun recognition 5.1.  Segmentation of words It is straightforward to recognize words in English text because they are separated by spaces. However, Chinese words are not separated by any boundary markers, such as 中文计算 425

Xiaoheng Zhang Table 24.1  Chinese font sizes in numbers, points, and mm 字号 (Number)

点数 (pt)

毫米 (mm)

Example

八号 (#8) 七号 (#7) 小六号 (#small 6) 六号 (#6) 小五号 (#small 5) 五号 (#5) 小四号 (#small 4)

 5   5.5   6.5   7.5  9 10.5 12

  1.76   1.93   2.28   2.64   3.16   3.69   4.22

四号 (#4)

14

  4.92

中文

小三号 (#small 3)

15

  5.27

中文

三号 (#3)

16

  5.62

中文

小二号 (#small 2)

18

  6.33

中文

二号 (#2)

22

  7.73

中文

小一号 (#small 1)

24

  8.44

中文

一号 (#1)

26

  9.14

中文

小初 (#small primary)

36

12.65

中文

初号 (#primary)

42

14.76

中文

中文

中文

中文

中文

中文

中文

中文

语言学导论 (Introduction to Chinese computational linguistics). Hence, word segmentation is the first step for text analysis of Chinese. For example, the previous phrase can be segmented into the following sequence of words (with the corresponding English translation written underneath). 中文 计算 语言学 导论 Chinese computational linguistics introduction Chinese word segmentation on a computer is carried out by matching characters in the Chinese text against a lexicon (list of Chinese words) forwardly from the beginning of the sentence or backwardly from the end. There are two kinds of segmentation ambiguities: the intersection-type (交集型歧义字段) and polynomial type (多义型歧义字段) (Xu and Fu 2006). Typically an intersection ambiguity is in the format of ABC, where A, AB, BC and C are all words in the lexicon. 426

Computational linguistics

It is possible to divide the original character string into word AB followed by C, or A followed by BC. For example 美国会 may mean 美/国会 (the US Parliament) or 美国/会 (the US can/will). The most common form of polynomial segmentation ambiguity is AB, where A, B, and AB are all words. That means the character string can be regarded as one single word or be divided into two. For example, 可以 in the following sentences: (1) 你 可以 坐下。 you can sit down You can sit down. 他们 为 样板。 (2) 你 可 以 you can take them as example You can take them as an example. Word segmentation ambiguities can be resolved with contextual information, using linguistic rules and probability of word co-locations derived from Chinese corpora. Usually longer words matching is more reliable. The correctness rate of automatic word segmentation has reached 95% (Xu and Fu 2006). However, there will be no guarantee of 100% correctness in the foreseeable future, because that will involve a complete understanding of the text. An alternative solution is to encourage people to write in a word segmented way, like the case in English (Zhang 1998a). But that does not means computer word segmentation will no longer be needed, because even in English, word segmentation is required for speech analysis.

5.2.  Recognition of proper nouns A proper noun is the name of a person, a place, an institution, etc. and is written in English with the initial letter of each word capitalized, for example, ‘Mr. John Nealon’, ‘America’, and ‘Cambridge University’. However, Chinese proper nouns are usually not marked in any style. Recognition of names of people and places in Chinese text can be supported by a list of names. However, such a list can never be complete, considering the huge number of places and people all over the world, not to mention their dynamic feature of coming, changing, and going. And there are names similar to non-proper nouns. For example, there is a town named 民众 (Minzhong), which is also a common noun meaning ‘people’. Therefore, recognition of names of people and places has to make use of their distinguishing features in internal composition and external context. Corpora with proper nouns annotated can also serve as useful reference. People’s names not in the dictionary can be recognized with a list of surnames and titles, for example 张大方先生, 李经理, where 张 (Zhang) and 李 (Li) are Chinese surnames, and 先生 (Mr.) and 经理 (Manager) are titles. In 张大方说, 张大方 can be successfully recognized as a person’s name by the rule that a Chinese given name normally consists of one or two characters, and the fact that people can speak (说). 427

Xiaoheng Zhang

Names of places also have characteristics useful for computer recognition. For example, in 在广东省中山市民众镇, component words 省 (province), 市 (city), and 镇 (town) are end markers of place names, while 在 (in, at, on) is a preposition frequently appearing in front of a location. The correctness rate of computer recognition has reached around 90% for persons’ names and 95% for place names (Xu and Fu 2006).

6.  Chinese natural language understanding and generation The ultimate goal of computational linguistics is to specify a theory of natural language understanding and generation to such a level of detail that a person could write a computer program that can understand and generate languages.

6.1.  Natural language understanding Computationally, ‘natural language understanding’ refers to the conversion of a sentence into some internal representation which can be used to solve relevant problems (Allen 1995). For example, the sentence 约翰有一本书 (John has a book) can be converted to predicate logic expression ‘own(John, B1) and is-a (B1, BOOK)’, which will allow the computer to answer such relevant questions as ‘Does John own anything?’, ‘What is owned by John?’. Computer analysis for natural language understanding can be separated into three steps: 1. Syntactic analysis, also called parsing, processes a sentence into its syntactic descriptions. For example, the sentence 他买了一个苹果 (He bought an apple) can be parsed into the following representation:

S (NP(N(他)), VP(V(买了), NP(ART(一个), N(苹果)))) he, bought, an, apple

2. Semantic analysis converts the syntactic description of a sentence into a form representing the literal meaning of the sentence independent of the context. A possible interpretation of the example sentence is

PAST EVENT, BUY-OBJ, (AGENT ‘他’ PERSON), (OBJ ‘苹果’ FRUIT/ COMPUTER/MOBLIE), (QUANTITY ‘一个’)),

where 苹果 (apple) may be a fruit, a computer or a mobile phone; an ambiguity not revealed at the syntactic level. 3. Pragmatic analysis maps the semantic representation into a representation of the intended meaning of the sentence. This step makes use of the context and world knowledge, including commonsense.

Suppose the previous sentence was ‘He was thirsty’, then the apple should be fruit, and the computer will output analysis result like:

Bought (他, 苹果) and is-a (苹果, fruit). he, apple, apple

428

Computational linguistics

6.2.  Natural language generation From the point of view of speech act, natural language generation involves plan generation and plan execution. The task of the software is to produce natural language sentences which can achieve given linguistic goals (Allen 1995). Natural language generation can also be divided into three steps: 1. Planning speech actions This is to generate a plan to achieve the goal of speaking. For example, suppose the goal of the computer system is to have a locked door (DOOR1) opened, and the system knows that the key (KEY2) is on the desk called DESK3, and that the user doesn’t know where the key is. A planning procedure to achieve the goal is as follows: Goal ‘is-open(DOOR1)’ is the effect of action OPEN(USER, DOOR1, KEY2), which has two conditions: INTEND(USER, OPEN(USER, DOOR1, KEY2)) and HAS(USER, KEY2). The first condition is the effect of REQUEST(SYS, USER, OPEN(USER, DOOR1, KEY2)), which can be expressed by the computer. The second condition is the effect of action PICK-UP(USER, KEY2) with the precondition KNOW(USER, LOCATION(KEY2)), which is the effect of speech action INFORM(SYS, USER, is-on(KEY2, DESK3)).

Resulting in a plan of two speech actions: REQUEST(SYS, USER, OPEN(USER, DOOR1, KEY2)); INFORM(SYS, USER, is-on(KEY2, DESK3)).

2. Generating semantic meanings For example, speech action REQUEST(SYS, USER, OPEN(OPENER, OBJECT, KEY)) can be achieved with semantic frame OPEN (AGENT = OPENER, THEME = OBJECT, TOOL = KEY). 3. Generating sentences This can be performed using generation rules or augmented transition network (ATN). For our example, the following simplified rules can be used for English and Chinese generation: English: G(ACTION) = G(>AGENT) G(>VERB)G(>THEME){with G(>TOOL)}, Chinese: G(ACTION) = G(>AGENT) {用 G(>TOOL)}G(>VERB)G(>THEME).

For Chinese, the PP representing the tool of an action is located before the action verb.

The previous introduction is based on traditional linguistics. Recently there has been more and more research on natural language processing based on corpora, which will be our next topic.

7.  Chinese corpora and corpus linguistics A corpus is a body of texts stored in a computer. It is normally built to represent a language or a sublanguage. There are mainly three kinds of activities in corpus linguistics: (i) development of corpora, (ii) development of tools for effective use of corpora, (iii) corpus-based linguistic research and language applications. 429

Xiaoheng Zhang

Corpora are built for certain purposes and aim at optimal representativeness of the languages. Texts in a corpus are selected from the natural language the corpus is meant to represent, and normally stored in the computer as files. In the early days the texts were manually typed into the computer from hard copy resources, nowadays they are more often captured from soft copy resources, or generated by OCR and speech recognition. To be more useful, many corpora have been annotated with additional information, such as word segmentation, part-of-speech tagging, syntactic annotation, and semantic annotation. Corpus annotation can be performed by the computer based on statistic models derived from a training corpus manually annotated by human beings (McEnery et al. 2006). As a corpus represents a language with authentic texts, it has many uses for language research and application. For example, we can consult a corpus to test a hypothesis in linguistics inquiry, provide machine translation with bilingual aligned sentences, improve dictionaries with real life word meanings and examples, and support language learning with real life expressions and their frequencies of use, etc.

7.1.  Some important Chinese corpora ● ‘Academia Sinica Balanced Corpus of Modern Chinese’ (Taiwan), simplified as Sinica Corpus, is an influential corpus designed by Professors Keh-jiann Chen and Chu-Ren Huang. Every text in the corpus is segmented and each segmented word is tagged with its part of speech. Texts are collected from different areas to be a representative sample of Modern Chinese language (traditional). The corpus is available at http://www.sinica. edu.tw/SinicaCorpus. ● ‘Linguistic Variations in Chinese Speech Communities’ (LIVAC) synchronous corpus (designed by Professor Benjamin Tsou, Hong Kong) contains texts from representative Chinese newspapers and electronic media of Hong Kong, Taiwan, Beijing, Shanghai, Macau, and Singapore. The collection of materials from the diverse communities is synchronized, and so offers an environment for comparative studies of Chinese. The corpus can be accessed at http://www.livac.org/. ● The ‘National Language Commission Corpora’ (国家语委语料库, mainland China). There are two member corpora. One is a corpus of Modern Chinese language (simplified), which allows free information retrieval from 20 million Chinese characters of texts wordsegmented with part-of-speech tagging. The other one is a corpus of Ancient Chinese language with around 100 million characters of texts from the Zhou dynasty (周朝) to the Qing dynasty (清朝). Both corpora can be accessed at http://www.cncorpus.org/. ● The ‘Center for Chinese Linguistics Corpus’ (the CCL Corpus) at Peking University has three subcorpora: a corpus of Modern Chinese with 581,794,456 characters of texts, a corpus of Ancient Chinese with 201,668,719 characters of texts, and a Chinese–English aligned bilingual corpus especially useful for language translation and comparative studies. It allows disjoint keywords information retrieval and flexible length of contexts on both left and right. The CCL Corpus is available at http://ccl.pku.edu.cn:8080/ccl_corpus/.

7.2.  Corpus information retrieval In corpus linguistics, words and word-forms may not be the same. For example, the English word ‘study’ may appear in different forms of ‘study’, ‘studying’, and ‘studied’. And its Chinese counterpart may appear in traditional Chinese as 學習, or in simplified Chinese as 学习. It is normally safer to use word forms in corpus information retrieval. 430

Computational linguistics

.,

1 2 3 4 5

24414

3.680%

1

29825;

4.534%

23.21%

11165

1.683%

2:

11054

1.680%

-0.1438%

9301

1.402%

3

9190

1.385%

4

7622

1.149%

'..5:,

6

7311

1.102%

7

6186

0.9324%

6

8

5037

0.7592%

9

9

4964

0.7482%

8:

6420

0.976%

30.44%

10

4114:

0.6201%

10;

4525.

0.6879%

10.93%

11 12 13 14 15 16 17 18

4027

0.607%

16

3421；

0.5201% :

3833

0.5777%

19

3781:

0.5699%

12:

3820；

0.5807%

1.899%

3750

0.5652%

23

3027 ；

0.4602%

-18.59%

3688

0.5559%

… 20:

0.5027%

' -9.561%

3642

0.5489%

13

3791；'

0.5763%'

4.985%

3527

0.5316%

21

3244.

0.4931%

-7.234%,

3516

0.5299%

15

3449

0.5243%

" '-1.063%;' '

" "9755V — 1.483% 8359 ". 7107 :

7 ：…

6530 6931!

'

4841

_ ‘ 3311:

’ ’ 3307

-8.261%

1.080%

8：

-9.915%

1.054%

: .

…

13.01%

0.7359%

0.5033%

10466

5:

'-5.956% '

0.9927%

mu

3

5.782%

1.271%

1

2丨 .

：

：

-3.066% "

20.26到

8724:

"'"1.316%

-6.096%'i

7074

1.067%

'-22.94%i

5912

0.8921%

-22.35%

5945

0.8971%

-Ï8.59%|

6654

1.004%

4

7621

1.150%

9:

5855-

0.8835%

7.689%. 51.47%

；

： ；

' ' 18.09% ：

3704

0.5589%

-9.862%:

'3754'

0.5665%

^-6.672% f

‘ 15卜

3734 ：

0.5635%

-2.471%-

11厂

4613:

0.6961%

22:15%:

29 丨

2614

0.3944%

…16|

—

" 1 4 ；.… -12.88%

4.425%

-1.579%： ' —-6.153%!

”

10r

'

W '

4718 ： ' 0.7119%

'

13 !’….

3335

0.5032%

3971

0.5992%

3425 ：

0.5168%

28.08%

：

-8.324% "12.72% .

Ü

； ；

卿

Figure 24.1  Chinese character frequency of Hong Kong, mainland China, and Taiwan

7.2.1.  Frequency lists of word-forms and characters The computer can count the occurrences of each word-form or character in a corpus, and sort them according to their frequency of use. Such a list is very useful to language research, learning, and application. Figure 24.1 is a section of the Chinese Character Frequency of Hong Kong, mainland China, and Taiwan built by the Hong Kong Chinese University. The complete frequency list and much more useful data can be found at http://humanum. arts.cuhk.edu.hk/Lexis/chifreq/.

7.2.2. Concordances A concordance is a collection of the occurrences of a word-form, each in its own textual environment. The program that automatically generates a concordance from a corpus is called a concordance program, which is arguably the most important tool for corpus information retrieval. The most commonly used layout for concordances is KWIC (Key Word in Context), where the word-form under examination appears in the center of each line, and in different color or with extra spaces on both sides. The length of the context is set according to different purposes. Figure 24.2 is a sample of a concordance for keyword ‘中文’ (Chinese) retrieved from the Academia Sinica Balanced Corpus of Modern Chinese. In addition to KWIC, there are line contexts, sentence contexts, paragraph contexts, and even whole-text contexts. The entries of a concordance can be sorted according to their occurrences in the original texts or according to the right/left neighboring words. A concordance provides rich contextual and statistical information on real-life application of the given word. More details on Chinese corpus linguistics can be found in the textbook by Huang and Li (2002). 431

Xiaoheng Zhang

488

I

Figure 24.2  A sample of a concordance for keyword ‘中文’ (Chinese)

8.  Chinese machine translation and machine-assisted translation ‘Machine Translation’ (MT) is translation by a machine, i.e. the conversion of a text or speech in one natural language (source language) into a text or speech of the same meaning in another natural language (target language) by means of a computer.

8.1.  Introduction to Chinese machine translation and machine-assisted translation Chinese machine translation commenced not long after the release of the ‘Weaver Memorandum’ in 1949, which is customarily considered as the starting point of MT research in the world. In 1956, a project on Russian–Chinese machine translation was included in the national plan for scientific research (Feng 2004). The first commercialized MT software in China was Transtar (译星), a rule-based English–Chinese MT system, which was released by the China National Software and Service Co. in 1988. Another important system was 863-IMT/EC (Intelligent MT for English–Chinese translation) developed by the Institute of Computer Technology, Academia Sinica. The software was later commercialized by Huajian Group (华建集团). Translation between Chinese and many other languages is also available on Systran, Microsoft Translator, and Google Translate. In addition to fully automatic translation, machine-assisted translation is also employed for Chinese translation using tools such as Trados and Google Translator Toolkit. Homemade tools of this category include Yaxin CAT by Yaxin Software (雅信) and Huajian IAT by Huajian Group, both supporting development and application of translation memories and terminology databases.

8.2.  Approaches of machine translation An MT system normally consists of a user interface, a source language analyzer, a target language generator, and the supporting dictionaries and knowledge bases. There are two basic approaches for developing MT systems: rule-based and corpus-based. Traditionally, MT is based on linguistic rules and artificial intelligence. Rule-based methods parse a text to create an intermediary symbolic representation, from which the text in the target language is generated. In recent years, a number of MT systems based on sentence-aligned bilingual corpora and statistics calculation have been developed. 432

Computational linguistics Table 24.2  Rule-based English–Chinese machine translation with syntactic analysis a. English syntactic structure:

b. Chinese syntactic structure:

S

S NP

VP

NP PRON

V

He

works

PP P in

PRON NP

He 他

ART

N

the

office

VP PP

P in 在

V NP

works 工作

ART

N

the 这

office 办公室

The following is a simplified example of rule-based English–Chinese machine translation with syntactic analysis. The English source language sentence is He works in the office. The machine translation consists of three steps: Step 1: Find the syntactic structure of the source language sentence according to some grammatical theory, as shown in Table 24.2.a. Step 2: Convert the source language syntactic structure into a structure of the target language, as shown in Table 24.2.b. According to Chinese grammar, the PP subtree has been moved before the verb it modifies. Step 3: Looking up the corresponding words in the bilingual dictionary as shown in Table 24.2.b, and generate a sentence in the target language: 他在这办公室工作。 More often, there are ambiguities which have to be resolved by semantic or even pragmatic analysis. If we change the input into a slightly different sentence such as The thief was caught by the bank. the task becomes much more tricky. If the word ‘bank’ refers to a riverside, prepositional phrase ‘by the bank’ should express the location of the event, and the Chinese translation should be ‘那个小偷在河岸附近被抓到’. If ‘bank’ refers to a financial institution, ‘by the bank’ may still express a location, and the translation is ‘那个小偷在银行附近被抓到’. The financial bank may alternatively play the role of agent, in which case the translation becomes ‘那个小偷被银行抓到’. Selection among these three interpretations needs to be supported by semantic and pragmatic analysis according to context information and background knowledge. 433

Xiaoheng Zhang

8.3.  MT evaluation There are various methods to evaluate the output of machine translation. The traditional way is by human judges to assess a translation’s quality. Even though human evaluation is timeconsuming, it is still the most reliable way. Automated means of evaluation include BLEU, METEOR, and WER. BLEU is the most commonly used metric. It makes use of multiple reference translations prepared by human translators and takes into account matches of both words and word order (Papineni et al. 2002). In his book entitled A Practical Guide for Translators (5th ed.), Samuelsson-Brown (2010: 82) says: ‘Development has been slow since the first serious attempts at MT were made 60 or more years ago. These attempts were limited by contemporary hardware, software and other factors. The facility is becoming more of a viable option, but still needs a skilled translator or language editor to make the result acceptable’. Wholly automatic high-quality MT remains a remote dream, and MT output still needs human post-editing. As for MT between English and Chinese, Chinese–English is poorer than English–Chinese translation because analysis of Chinese is more challenging and less developed.

9.  Chinese–Chinese machine translation Translation not only happens between languages but also within a language. In this section, we will introduce two important types of intra-Chinese machine translation, including Chinese dialect MT and simplified–traditional Chinese conversion.

9.1.  Chinese dialect MT Automatic translation is more achievable here since inter-dialect difference is much less serious than inter-language difference. In the following sections we will discuss inter-dialect MT with emphasis on Cantonese and Putonghua (Zhang 1998b).

9.1.1.  Dialects and Chinese dialects The dialects of a language are that language’s systematic variations, developed when people of a common language are separated geographically and socially. Among this group of dialects, normally one serves as the lingua franca. Inter-dialect differences exist in pronunciation, vocabulary, and syntactic rules. However, they are usually insignificant in comparison with the similarities the dialects have. It has been declared that dialects of one language are mutually intelligible (Fromkin and Rodman 1993: 276). Nevertheless, the seven major Chinese dialects – the Northern dialect (with Putonghua as its standard version), Cantonese, Wu, Min, Hakka, Xiang, and Gan – are for the most part mutually unintelligible, and interdialect translation is required for successful communication, especially between Cantonese, the most influential dialect in South China, and Putonghua, the lingua franca of China (Yuan 1989).

9.1.2.  Linguistic consideration of dialect MT Most differences among the dialects of a language are found in their pronunciations. Words with similar written forms are often pronounced differently in different dialects. For example, the word 香港 (Hong Kong) is pronounced xiang1gang3 in Putonghua, but hoeng1gong21 434

Computational linguistics

in Cantonese. There are also lexical differences although dialects share most of their words. For example, the word ‘umbrella’ is 雨伞 (yu3san3) in Putonghua, and 遮 (ze1) in Cantonese. Differences in syntactic structure are less common but linguistically more complicated and computationally more challenging. For example, the positions of some adverbs may vary from dialect to dialect. To express ‘You go first’, we have (1) Putonghua: 你 先 走。 ni3 xian1 zou3 you first go (2) Cantonese: 先。 你 行 nei5 hang4 sin1 first you go

9.1.3.  Word processing in dialect MT Inter-dialect differences mostly exist in words. In the Cantonese vocabulary, there are about 7,000 to 8,000 dialect words (including idioms and fixed phrases), which are in different forms from their Putonghua counterparts. These dialect words account for about one third of the total Cantonese vocabulary. Because of historical reasons, Hong Kong Cantonese is linguistically more distant from Putonghua than other regions in mainland China. One can easily spot Cantonese dialect articles in Hong Kong newspapers which are totally unintelligible to Putonghua speakers, while Putonghua articles are easily understood by Cantonese speakers. The critical task in Cantonese-to-Putonghua MT is word processing, especially dialect words recognition and translation (Zhang 1999). The most challenging issue is to deal with Cantonese ambiguous words that do not have semantically equivalent counterparts in Putonghua. For example, Putonghua word 桔 ( ju2, orange) has a much larger coverage than the Cantonese 桔 (gwat1). In addition to the Cantonese 桔, Putonghua 桔 also includes the fruits Cantonese refers to as 柑 (gam1) and 橙 (caang2). On the other hand, the Cantonese 行 semantically covers Putonghua 走 (go, walk) and 行 (row).

9.2.  Simplified–traditional Chinese conversion Due to historical reasons, Modern Chinese is written in both traditional characters and simplified characters, which quite frequently renders text conversion between the two scripting systems indispensable. Computer-based simplified–traditional Chinese conversion is available on MS Word, Google Translate, and many language tools on the Internet. Their performance has reached very high precision. However, because of the existence of one-tomany relationships between simplified and traditional Chinese characters (e.g. character 干 in simplified Chinese corresponds to 干, 乾, and 幹 in traditional Chinese.), there is no guarantee of 100% correct conversion. That means human proofreading is needed, especially when high-quality text output is required. Zhang (2011) has developed a tool which goes on to support human proofreading after converting a text between simplified and traditional Chinese. The important features include: 435

Xiaoheng Zhang

● Simplified–traditional Chinese bi-directional conversion with 4 options: (i) simplified to Hong Kong traditional Chinese, (ii) simplified to Taiwan traditional Chinese, (iii) simplified to high-frequency traditional Chinese characters, and (iv) traditional to simplified Chinese. ● Support for human proofreading by: (i) highlighting all characters with one-to-many relationships between simplified and traditional Chinese, (ii) providing relevant dictionary information for reference, (iii) correcting mistakes automatically by a single click. ● Employment of standard and frequently used characters and punctuation marks in the target writing system. The tool is available on the Web at http://myweb.polyu.edu.hk/~ctxzhang/jfj/ and http:// www.mypolyuweb.hk/~ctxzhang/jfj/. Further improvements are in progress. Normally, simplified–traditional Chinese conversion is performed in a character-tocharacter way. For example, simplified 汉字信息处理 (Chinese characters information processing) is converted to traditional 漢字信息處理. MS Word can also translate common terms (though not always correctly), in which case the previous translation becomes 漢字 資訊處理, because in Taiwan and Hong Kong, ‘information’ is more often translated into 資訊 than 信息. A comparative study on simplified–traditional Chinese conversion tools can be found in a recently published paper (Zhang 2014).

Note 1 In this section pronunciation of Putonghua is presented in Pinyin, and Cantonese in Jyutping. Numbers are used to denote tones of syllables. Jyutping is based on Pinyin. That means, across the two Pinyin schemes, words with different Pinyin symbols are normally pronounced differently.

References Allen, J. (1995) Natural Language Understanding, San Francisco: The Benjamin/Cummings Publishing Company. Davis, M. (2012) ‘Unicode Over 60 Percent of the Web’, Google Official Blog (http://googleblog. blogspot.hk/2012/02/unicode-over-60-percent-of-web.html). Feng, Zhiwei (2004)《机器翻译研究》(Studies of Machine Translation), Beijing: China Translation and Publishing Corporation. Fromkin, V. and Rodman R. (1993) An Introduction to Language (5th ed.), Orlando, Florida: Harcourt Brace Jovanovich College Publishers. Huang, C. and Li, J. (2002)《语料库语言学》(Corpus Linguistics), Beijing: Commercial Press. Language Institute of the Chinese Academy of Social Sciences (2011)《新华字典》(Xinhua Chinese Characters Dictionary), Beijing: The Commercial Press. Linguistic Society of Hong Kong (1993) Cantonese Romanization Scheme – Jyutping (粤拼, http:// www.lshk.org/node/31). Malmkjaer, K. (ed) (2010) Linguistics Encyclopedia (3rd ed.), London and New York: Routledge. (Article title: ‘From computational linguistics to natural language engineering’). McEnery, Tony, Richard Xiao, and Y. Tono (2006) Corpus-based Language Studies: An Advanced Resource Book, London and New York: Routledge. National Language Commission of China (1958)《汉语拼音方案》(Pinyin Scheme for the Chinese Language), Beijing: National Language Commission of China. National Language Commission of China (1997) Chinese Character Component Standard of GB13000.1 Character Set for Information Processing, Beijing: National Language Commission of China. National Language Commission of China (2001)《汉语拼音方案的通用键盘表示规范》(Standard for the Scheme of Chinese Phonetic Alphabet Input with Universal Keyboard ), Beijing: National Language Commission of China.

436

Computational linguistics Papineni, K, S. Roukos, T. Ward, and Zhu, W.-J. (2002) ‘BLEU: A Method for Automatic Evaluation of Machine Translation’, in Proceedings of the 40th Annual Meeting of the Association of Computational Linguistics (ACL). Samuelsson-Brown, G. (2010) A Practical Guide for Translators, 5th ed., Bristol: Multilingual Matters. The Unicode Consortium (2014) The Unicode Standard, Version 7.0.0. Mountain View, CA: The Unicode Consortium, (http://www.unicode.org/versions/Unicode7.0.0/). White, R. (2008) How Computers Work, 9th ed., Indianapolis, IN: QUE. Xu, J. and Fu, Y. (2006)《中文信息处理现代汉语词汇研究》(Morphological Studies in Modern Chinese Information Processing), Guangzhou: Guangdong Education Press. Yuan, J. (1989)《汉语方言概要》(Introduction to Chinese Dialects), Beijing: Wenzi Gaige Press. Zhang, X. (1998a)〈也谈汉语书面语的分词问题 – 分词连写十大好处〉(‘Written Chinese Word Segmentation Revisited: Ten Advantages of Word-segmented Writing’), Journal of Chinese Information Processing 3(12): 57–63. Zhang, X. (1998b) ‘Dialect MT: A Case Study between Cantonese and Putonghua’, in Proceedings of COLING-ACL’98 1,460–4. Zhang, X. (1999)〈粤–普机器翻译中的词处理〉(‘Word Processing in Cantonese–Putonghua Machine Translation’), Journal of Chinese Information Processing 3(13): 40–7. Zhang, X. (2003)〈正易全: 一个动态结构笔组汉字编码输入法〉(‘Towards Correctness, Easiness, and Completeness: Building a Chinese Character Coding Input Method Based on Dynamic Structured Stroke Groups’), Journal of Chinese Information Processing 3(17): 59–65. Zhang, X. (2006)〈字形的‘号制’ ‘点制’与‘米制’〉(‘The Number, Point, and Metric Systems of Font Size’), Computer Engineering and Applications 10(42): 175–7, 215. Zhang, X. (2011)〈一个支持人工校对的中文简繁体转换工具〉(‘A Simplified–traditional Chinese Conversion Tool with a Supporting Environment for Human Proofreading’), in Sun, M. and Chen, Q. (eds) Advances of Computational Linguistics in China (2009–2011), Beijing: Tsinghua University Press, 569–75. Zhang, X. (2012) ‘Computer Input of Non-ASCII Non-Hanzi Chinese Characters’, Journal of Modernization of Chinese Language Education 1(2): 18–36 (http://xuebao.eblcu.com/). Zhang, X. (2014). ‘A Comparative Study on Simplified–Traditional Chinese Translation’. In Sun, M., Liu, Y., and Zhao, J. (eds) Chinese Computational Linguistics and Natural Language Processing Based on Naturally Annotated Big Data: Lecture Notes in Artificial Intelligence 8801, pp. 212–22. Switzerland: Springer.

437

25 Corpus-based study of Chinese Tony McEnery and Richard Xiao Lancaster University, UK

This chapter provides a review of the state of the art in the corpus-based study of Chinese. Given that such a review could become unwieldy if every area of linguistics were to be reviewed, we will focus upon language description and applied linguistics, areas where, arguably, corpus linguistics has made its greatest impact to date. The areas considered in this chapter include lexical study, grammatical study, and Chinese interlanguage research.

1.  Corpus-based lexical study Just like English, Chinese as the world’s most spoken language has also benefited greatly from the development of corpus linguistics. The first study of Chinese character frequency in a modern sense dates back as early as the 1920s, with the publication of Li Jinxi’s (1922) ‘A Statistical Analysis of Basic Chinese Vocabulary’. Also in the 1920s, Chen Heqin, a renowned educationalist in China, assembled a paper-based corpus composed of a range of materials such as children’s books, after-class readings for school children, newspapers, women’s magazines, and classical and modern Chinese fiction, amounting to well over half a million Chinese characters. Chen and his nine assistants worked for two to three years to establish a list of the 4,261 most frequently and widely used Chinese characters, which was first published as The Applied Glossary of Modern Chinese in the fifth issue of the fifth volume of New Education in 1922. It was revised later and republished as a book of the same title by the Commercial Press in 1928. Chen’s frequency list of Chinese characters played an instrumental role in promoting literacy in primary education in China by providing empirical evidence for the development of primary teaching materials such as Thousand Character Lessons for Civilians (Tao and Zhu 1923), which was used to teach over half a million civilians basic Chinese characters in more than 20 provinces. Chen’s frequency list of Chinese characters was profoundly influenced by Thorndike’s (1921) The Teacher’s Word Book. However, the contribution of the former to Chinese is arguably more significant than the contribution of the latter to the English language. Chinese as a script language is different from phonetic languages such as English. While such phonetic languages usually have an alphabet of 20–30 characters that make up words which are typically delimited by spaces in writing, there are 6,000–7,000 Chinese characters in general use (with a possible total of over 50,000), many of which are used as monosyllabic words. 438

Corpus-based study of Chinese

With such a large character set, it is of crucial importance to identify, as a means of promoting literacy, both a basic set of the most common Chinese characters and the words which are used most frequently and widely. Chen’s character list is particularly important because it was established in a timely manner, at the initial stage of the development of modern Chinese as used today. The New Culture Movement in 1919 is generally recognised as the demarcation line between classic Chinese and modern Chinese. As such, Chen’s The Applied Glossary of Modern Chinese has not only contributed to primary education and the promotion of literacy in China, it has also helped to shape present-day Chinese. While the data used in Chen was not computerised, his list of basic Chinese characters was nevertheless corpus-based; it is the forerunner of today’s word frequency lists and frequency dictionaries of Chinese derived from computer corpora. For example, since the founding of the People’s Republic of China in 1949, the central government and local authorities have also published a range of lists of Chinese words and characters, including for example, ‘Register of Common Characters’ published by the Ministry of Education in September 1950 which contains 1,017 Chinese characters, ‘List of Common Characters’ published by the Ministry of Education in June 1952 which contains 2,000 Chinese characters, ‘List of Common Characters in Putonghua Common Speech’ published by the Shandong Provincial Department of Education in August 1958 containing 3,000 Chinese characters, ‘Three Thousand Common Words in Putonghua Common Speech’ published by China’s Committee for Chinese Written Language Reform in 1962, ‘A List of Four Thousand Words for Foreign Students’ published by the Beijing Language and Culture University in 1964, and ‘List of Commonly Used Characters’ published by the Beijing Municipal Commission of Education in March 1965 which contains 3,100 Chinese characters. The examples cited above demonstrate that Chinese linguists have a long-standing tradition of studying word and character frequency, because of the size of Chinese character set. This tradition has, unsurprisingly, continued, with the rapid development of corpus linguistics in general and Chinese language processing in particular, into the 1980s and 1990s as well as the new millennium. For example, Liu’s (1973) Frequency Dictionary of Chinese Words gives statistics such as frequency, dispersion index, and usage rate for the 3,059 most frequently used words in Chinese on the basis of a 250,000-word corpus covering five registers (fiction, drama, essays, newspapers and periodicals, technical writing); A Comprehensive Frequency Table of Character Usage in Modern Chinese established by Project Code 748 (1976) lists 4,152 frequently used characters on the basis of a corpus of 21 million characters; Beijing Aeronautical University’s (1985) A Frequency Table of Character Usage in Modern Chinese lists frequently used characters for ten genres and technical domains on the basis of text samples totalling 11.08 million characters; A Frequency Dictionary of Modern Chinese developed by Beijing Language and Culture University (1986) lists 16,593 commonly used words extracted from 1,315,752 word tokens (or 1.82 million characters); the National Language Committee’s (1988) Commonly Used Characters in Modern Chinese lists the most commonly used 2,500 characters and 1,000 commonly used characters on the basis of data collected by Beijing Aeronautic University covering the period 1928–86; the Hong Kong Polytechnic University’s (1991–7) A Chinese Word Bank from Mainland China, Taiwan, and Hong Kong lists 68,011 entries based on a 6 million-character corpus of news texts published during 1990–2 in the three Chinese speech communities. These corpus-based frequency lists and dictionaries are essentially targeted either at native speakers of Mandarin learning their mother tongue (e.g. Chen 1928; National Language Committee 1988), or at language engineers (e.g. the frequency list by Project Code 738) and expert Chinese linguists (e.g. the word bank by Hong Kong Polytechnic University for studying language variation). More 439

Tony McEnery and Richard Xiao Table 25.1  HSK graded lists and words and characters in Chinese HSK level

Words

Characters

Level 1 Level 2 Level 3 Level 4 Levels 1–3 Levels 1–4

1,033 2,019 2,205 3,583 5,257 8,840

800 803 591 671 2,194 2,865

recently, R. Xiao et al.’s (2009) A Frequency Dictionary of Mandarin Chinese, which is based on a sizeable balanced corpus of current spoken and written language from four broad categories – namely spoken, fiction, non-fiction, and news, totalling approximately 50 million word tokens (or 73 million Chinese characters), provides a list of the top 5,000 Chinese words and 2,000 Chinese characters, thus defining a core vocabulary for learners of Chinese as a second or foreign language. With regard to non-native language learning resources for Chinese, one should not fail to mention the Syllabus of Graded Words and Characters for Chinese Proficiency compiled by the Chinese government’s Hanyu Shuiping Kaoshi (the Chinese Proficiency Test, HSK) Committee, which was published in 1992 and revised in 2001. The HSK lexical syllabus lists the words and characters required of learners of Mandarin Chinese as a second or foreign language to pass the Chinese proficiency test HSK, as indicated in Table 25.1. According to the HSK lexical syllabus, learners of Chinese as a foreign language who have learnt about 5,000 words will be able to express their ideas on general issues in Chinese. As can be seen in Table 25.1, this vocabulary is approximate to the total number of words in HSK Levels 1–3. The number of words covered in R. Xiao et al.’s (2009) frequency dictionary is roughly comparable. While it is certainly true that the larger a learner’s vocabulary the better the language learner’s proficiency, it is nonetheless increasingly difficult to learn new words as the learner’s vocabulary grows. This is because, according to Zipf’s law, the frequency of a word is reversely proportional to its rank in the frequency table. As such, there is a 9.27% increase in coverage from top 1,000 to top 2,000 words, whereas the increase in coverage drops to 0.94% from top 8,000 to top 9,000 words (see Table 25.2).

Table 25.2  Coverage of top N words Top N words

Coverage (%)

Increase (%)

1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000

66.24 75.51 80.53 83.81 86.17 87.94 89.33 90.47 91.42

– 9.27 5.02 3.28 3.07 1.77 1.39 1.14 0.94

440

Corpus-based study of Chinese 100 90 80 Coverage

70 60

Increase

50

Drop

40 30 20 10

0 2,

00

0 To p

To p

1,

60 1,

80

0

0 To p

To p

1,

20 1, To p

40

0

0 1, To p

To p

80

00

0

0 60 To p

0 40 To p

To p

20

0

0

Top N characters

Figure 25.1  Coverage of top N characters

In addition to the reference to the HSK syllabus, R. Xiao et al.’s (2009) decision to include 5,000 words is also empirically based on the sharp drop in coverage (from 3.07% to 1.77%) from top 5,000 to top 6,000 words. As can be seen in Figure 25.1, which shows the increase in vocabulary size and the drop in coverage resulting from each additional block of 200 characters, Zipf’s law also applies to characters. Coverage grows very slowly after the top 1,200 characters. The top 2,000 characters cover nearly 98% of our whole corpus, with 4,839 characters accounting for the remaining 2% of coverage. Corpora can be said to have an important role to play in all aspects of lexicography, ranging from selecting headwords for inclusion in a dictionary, identifying word senses, ordering of polysemous and homograph items, to determining a word’s word class and providing illustrative examples of its use. Cui (2011: 85) provides a good summary of the various roles of corpora in compiling New Word Dictionary for Chinese as a Foreign Language (see Table 25.3). According to H. Xiao (2010), a corpus that is annotated with word senses can provide a means of verifying whether the definitions in a dictionary are reasonable and complete. Jiang (2005) observes that major Chinese dictionaries, when evaluated against large-scale corpora, suffer from a number of drawbacks, including, for example, omissions of some commonly used words or word senses, or word senses too narrowly defined, and a lack of illustrative examples. According to E. Li (2002: 95), the process of identifying and arranging word senses in dictionary making can be summarised as three steps: (i) collecting corpus data, (ii) selecting materials from the corpus according to the purpose of the dictionary, and (iii) identifying word senses. This is traditionally a subjective process relying on a combination of slip cards and the lexicographer’s intuitions, a process which is likely to result in omissions. For example, the word 爱护 ‘cherish, treasure; take care of; love and protect’ is defined as 爱惜、保护 ‘cherish; protect’ in A Dictionary of Chinese and as 爱惜并保护 ‘cherish and protect’ in A Dictionary of Modern Chinese. On the other hand, corpora can provide more objective criteria for identification and ordering of word senses in the dictionary on the basis of 441

Tony McEnery and Richard Xiao Table 25.3  A summary of the roles of corpora in lexicography in Chinese Dictionary microstructure

Roles of corpora

Headword

Retrieving new words Identifying headwords Identifying synonyms Determining primary and secondary headwords for cross reference

Pinyin pronunciation

Deciding the pinyin gloss of words with multiple pronunciations

Word class

Determining the word class of the headword Deciding the order of word classes of words with concurrent classes

Translation

Assisting in translation of the headword

Linguistic interpretation

Mining for pragmatic properties of the headword Mining for syntactic properties of the headword

Definition

Assisting in definition of the headword Determining the order of senses of polysemous words

Illustrative example

Providing authentic examples in context

Knowledge window

Providing background knowledge of the headword

exhaustive quantitative and qualitative analysis. For example, a random selection of 1,000 concordance lines of 爱护 aihu from a sizeable corpus of modern Chinese shows that the word is used in three senses: (i) in 651 examples (67.46%), it means 关心、关爱 guanxin, guan’ai ‘care for or show concern for’ (typically seniors to juniors, or elder to young people); (ii) in 221 examples (22.90%), it means 保护,使其不受损害 baohu, shi qi bu shou sunhai ‘protect (from being damaged or hurt)’; (iii) in 93 examples (9.64%), it means 珍惜、不浪 费、不毁坏 zhenxi, bu langfei, bu huihuai ‘cherish or treasure, not waste or damage’. Clearly, the first two are the major senses of the word, and they are interrelated. They should be listed as separate senses in a medium-sized dictionary because of their different collocational behaviour and grammatical properties. On the one hand, word sense (a) typically (95% of the time) collocates with nouns or pronouns denoting humans or animate objects, whereas word sense (b) usually co-occurs with nouns and pronouns referring to inanimate objects (77%). On the other hand, 爱护 aihu with sense (a) can be used either as a verb (71% of the time) or as a noun (27%) or as an adjective or adverb, while sense (b) is normally used as a verb. It is suggested that according to frequency data from the corpus, the word senses of 爱护 aihu should be revised as follows, in the given order: (a) verb or noun: 关心、关爱 (一般对人或生命体,多用于上对下、   长对幼) guanxin, guan’ai ( yiban dui ren huo shengmingti, duo yongyu shang dui xia, zhang dui you) ‘care for or show concern for’ (generally human or animate object; typically seniors to juniors, or elder to young people); (b) verb: 保护,使(物体)不受损害 baohu, shi (wuti) bu shou sunhai ‘protect (usually an inanimate object) from being damaged’; (c) verb: 珍惜、不浪费、不毁坏 zhenxi, bu langfei, bu sunhuai ‘cherish or treasure, not waste or damage’ (see B. Huang 2004). In this arrangement, word sense (c) is close to sense (b); therefore in a small-sized dictionary, it can be merged into sense (b). Language keeps changing over time. Hence, an important area of lexicography is the study of neologisms, which can benefit greatly from the corpus approach, because corpora provide 442

Corpus-based study of Chinese

the necessary sources of data as well as the method for reasonably identifying new words or a new usage of existing words. A Dictionary of New Words (Kang 2003) is the largest Chinese neologism dictionary of its kind, containing 20,000 new words that have gained currency and remained relatively stable between 1978 and around 2000. The dictionary is based on a huge corpus composed of 25 years’ data (1978–2002) from the People’s Daily newspaper, in addition to recent data from 12 newspapers of the People’s Daily newspaper family, nine websites of the people.com.cn family, the Readers magazine (previously known as Readers’ Digest) since its founding, the Southern Weekly newspaper since its founding, as well as online data from newspapers including Guangming Daily, China Education Daily, and Southern Daily. Similarly, Tsou and You’s (2010) A Dictionary of Global Chinese Neologisms includes 1,600 Chinese neologisms that have entered the Chinese language since 2000. These new words are carefully selected from the over 20,000 new words from LIVAC (Linguistic Variations Across Chinese Communities), which is a gigantic, homothematic, and synchronous Chinese corpus designed to monitor pan-Chinese language development and underlying pan-Chinese changing cultural trends. The monitor corpus samples representative local media texts published in Chinese, covering main news, editorials, international news, local news, entertainment, sports, finance, etc. from Chinese-speaking communities in Hong Kong, Taipei, Macau, Singapore, Shanghai, Beijing, Canton, Shenzhen, and Japan. As of 2010, the corpus has included over 16 years’ data, amounting to 400 million Chinese character tokens, with a lexicon of 1.5 million word types (Tsou et al. 2011). In addition to monolingual Chinese dictionaries, corpora, especially parallel corpora, have been used extensively in bilingual lexicography involving Chinese and a foreign language in China. Before corpora made inroads in bilingual dictionary making, the examples in such dictionaries were either invented introspectively by lexicographers (e.g. New English–Chinese Dictionary) or cited from famous literary works (e.g. Chinese–English Dictionary of Idioms) (cf. J. Li and Zhong 2011). The primary aim of a bilingual dictionary is to help the user to establish a link between a lexical unit in the source language and its equivalents in the target language. In a bilingual dictionary, the headword is usually defined in the target language. In most cases, a definition is only partially equivalent to the headword because it is often an abstract generalisation of the typical meanings of the word instead of covering all of its meanings fully. For example, 妥当 tuodang is usually defined as ‘appropriate, proper, suitable, sound’ in a Chinese–English dictionary. However, the word can also mean ‘safe; ready; wise; guaranteed’ in attested language use, which is not covered in the dictionary definition. In this case, bilingual examples cited from parallel corpora can complement missing meanings, for example: 他想把大衣放在最妥当的地方 ta xiang ba dayi fang zai zui tuodang de defang ‘He wanted to put his coat in the safest possible position’ (cf. J. Li and Zhong 2011). One particular issue in making specialised bilingual dictionaries relates to the definition and translation of the domain-specific usage of ordinary words in specific domains. In the business domain, for example, the concept of 表 biao ‘table’ is conventionally expressed as ‘statement’ instead of ‘table’, as in ‘financial statement’ 财务报表 caiwu baobiao, and ‘statement of income and expenses’ 财务收益与费用表 caiwu shouyi yu feiyong biao. Issues such as this can be addressed readily with the help of parallel and comparable corpora of the languages involved. As L. Li (2006) notes, specialised corpora are ideal linguistic and knowledge resources; and corpus-based specialised dictionaries can ensure a systematic coverage of headwords of practical value, accurate definitions, and appropriate illustrative examples. The brief review of the state of the art of corpus-based lexicography in Chinese context suggests that the advantages of using corpora in dictionary making, whether monolingual 443

Tony McEnery and Richard Xiao

Chinese dictionaries or bilingual dictionaries, are self-evident. On the other hand, corpora are a double-edge sword. If used inappropriately, they simply mean labour lost; and worse still, they can lead to falsehood under a scientific and objective disguise. This warning echoes Sinclair’s (2004: 2) observation: ‘A corpus is not a simple object, and it is just as easy to derive nonsensical conclusions from the evidence as insightful ones.’

2.  Corpus-based grammatical study There is presently no corpus-based comprehensive reference grammar of the Chinese language. Research in corpus-based descriptive grammars in Chinese is rather sporadic and fragmentary. What there is has focused on specific linguistic features of interest to individual researchers. For example, C. R. Huang and Ahrens (2003) study the relationship between nouns and nominal classifiers in Mandarin Chinese on the basis of the data from Mandarin Chinese Classifier and Noun-Classifier Collocation Dictionary (C. R. Huang et al. 1995), which is in turn based on data from the Academia Sinica Balanced Corpus of Modern Mandarin Chinese, a grammatically tagged balanced corpus containing a total of 5 million words from Mandarin texts as used in Taiwan. The authors challenge the traditional view that nominal classifiers classify individuals, arguing that ‘classifiers do not simply agree with a noun, but instead coerce a particular meaning from it’ (C. R. Huang and Ahrens 2003: 353). Zhao (2010) studies the grammatical meanings and usage contexts of one function word in Chinese, 来着 laizhe, on the basis of 610 valid instances of the word in a sizeable online corpus of modern Chinese developed by the Centre for Chinese Linguistics of Peking University (hence the PKU Corpus). It is found that the word typically (99.8% of the time) occurs at the end of a sentence or clause. The two main grammatical meanings the word expresses include indicating that something took place in the past (48.4%), and that the speaker seeks to be reminded of something that took place in the past (45.2%). While the first meaning is normally used in declaratives, the second meaning usually occurs in questions. The two types of sentences account for 69.4% and 29.3% respectively of total occurrences of 来着 laizhe in the corpus. It is also of interest to note that sentences that include this function word tend to avoid verb reduplication, verbs denoting termination, as well as perfective aspect markers 了 -le and 过 -guo. Another syntactic condition is that 来着 laizhe is predominantly used in affirmative (99.7%) rather than negative sentences. Q. Zhang (2010) is concerned with a syntactic and pragmatic analysis of a commonly used degree complement structure in Chinese, X 得很 X de hen, on the basis of the PKU corpus. It is found that, semantically, this complement structure primarily expresses the speaker’s psychological feeling, including evaluation of people and attitude towards a particular object or event. Syntactically, the words that can be used in the X slot of the structure include adjectives and verbs. Both qualifying and evaluating adjectives are found to occur in the structure, which can be positive, neutral or negative in evaluative meaning, while the verbs used in the structure are largely psychological verbs. The corpus analysis also suggests the words in the X slot of the structure are mostly disyllabic words, though monosyllabic words are not uncommon. Pragmatically, this complement structure can not only intensify the degree of the quality denoted by the word in the X slot to strengthen the expressive effect, but it can also be used as a parenthesis to express the speaker’s attitude. Siewierska et al. (2010) undertake a corpus-based study of splittable compounds in interaction with morphology, syntax, and pragmatics, with the aim of producing a systematic and realistic account of splittable compounds as attested in 2 million words of authentic spoken and written Chinese data as represented in the Lancaster Corpus of Mandarin Chinese 444

Corpus-based study of Chinese

(LCMC) and the Lancaster Los Angeles Spoken Chinese Corpus (LLSCC). It is found that the typical grammatical pattern of splittable compounds is constitutive of an aspect marker (e.g. 了 -le, 着 -zhe, 过 -guo) or resultative verb complements as post-verbal adjacent elements (54% of all instances), and a quantifier, a classifier, a modifier or a combination of two or more of them which precede the nominal components of splittable compounds. Drawing on morpho-syntactic and phonological criteria, the split uses, together with their combined uses, of splittable compounds with one inserted aspectual morpheme are viewed as words, while the others are regarded as phrases. From a discourse-pragmatic perspective, the split use of splittable compounds is more often found in the spoken genres of Chinese. Insertions of splittable compounds tend to function as mitigation or modification to the verbal heads or final nominal/complement elements. Similarly, H. Wang and Wang (2009) approach splittable compounds from a pedagogical perspective by discussing the implication of their research findings based on the PKU corpus for teaching Chinese as a foreign language. Their investigation of 207 splittable compounds in the Syllabus of Graded Words and Characters for Chinese Proficiency indicates that for most of these splittable compounds, their continuous use is regular whereas their discontinuous use is idiosyncratic in attested language. The discontinuous use of splittable compounds is a colloquial feature which tends to occur in informal spoken genres. The authors suggest that the teaching of splittable compounds should be informed by the core patterns observed in corpus analysis by focusing on the most commonly used splittable compounds and their major forms of insertions in discontinuous use individually on the basis of their frequency bands while also taking account of genre variation and key sentence types so that learners can use splittable compounds appropriately in context. R. Xiao et al. (2006) provide a systematic account of passive constructions in Chinese in contrast with English, covering a range of characteristics of passives in the two languages including various passive forms, long vs. short passives, semantic, pragmatic and syntactic features as well as genre variations. It is found that while passive constructions in both languages express a basic passive meaning, they also show a range of differences in terms of overall frequency of use, syntactic features and functions, semantic properties, and distribution across genres. For example, passive constructions are nearly ten times as frequent in English as in Chinese because Chinese passives are much more restricted in use than their English counterparts. Short passives are predominant in English, whereas Chinese displays a preference for long passives. English passives are more stylistically oriented to make the discourse sound more impersonal, objective, formal, and technical, while Chinese passives are a pragmatic voice that typically expresses a negative semantic prosody. Chinese passives in the predicate position typically interact with aspect; passive constructions with bare verbs in this position are uncommon, though they are frequent in other sentential positions. In contrast, the interaction between passives and aspect is not as apparent in English as in Chinese because all English sentences and finite clauses are formally marked by combined tense– aspect markers. Passives in English occur more frequently in informative than imaginative genres; reports/official documents and academic prose, in particular, show very high proportions of passives. But these two genres have the lowest proportions of passives in Chinese, where mystery/detective fiction and religious writing show exceptionally high proportions of passives. These differences are closely associated with the origins and functions of passive constructions in the two languages. The passive is primarily used as a style marker in English, whereas it is typically an ‘inflictive voice’ in Chinese. R. Xiao and McEnery (2008) explore negation in Chinese on the basis of spoken and written corpora of Mandarin Chinese. They found that in addition to negating functions, 445

Tony McEnery and Richard Xiao

negative adverbs in Chinese also display differences in distribution across genres. The best characterisation of 没 mei is that it negates the realisation of a situation, which distinguishes it from the general purpose negator 不 bu. Regarding the interaction between negation and aspect marking, the actual aspect marked by 了 -le and the experiential aspect marked by 过 -guo are negated by 没 mei, while the durative aspect marked by 着 -zhe and the progressive aspect marked by 在 zai rarely occur in negative sentences. Imperfective aspects of the latter two types usually undergo a viewpoint aspect shift when they are negated. While sentences taking 在 zai can be negated by either 不 bu or 没 mei, co-occurrences of 不 bu and 在 zai are typically in double negation structures or rhetorical questions, which are essentially positive in meaning. Sentences taking 着 -zhe are negated more frequently by 没 mei than 不 bu, and 着 -zhe is often omitted in negative sentences unless it appears in the V-zhe V structure where V-zhe acts as an adverbial. The scope of negation typically extends from the word immediately following the negator to the end of a clause unless the context provides clues that suggest otherwise. Word order is important in determining the focus of negation, which typically falls upon some modifying element that usually follows the negator immediately, or on the end of a clause, unless a contrast present in context suggests otherwise. Transferred negation is uncommon in Chinese. When it occurs, the transferred focus of negation suggests a reduced degree of negation. Double negation is common in Chinese, but the negation of negation often means more than two negators cancelling each other, not only in terms of emphatic force, but in meaning as well. Finally, redundant negation typically occurs in sentences with some element which is inherently negative semantically. R. Xiao and McEnery (2004) provide the first book-length corpus-based comprehensive account of aspect in Mandarin Chinese, encompassing both situation aspect and viewpoint aspect. The corpus-based model of aspect developed in this book represents a significant advance in aspect theory, which explores aspect at both the semantic and grammatical levels. The two levels correspond to the two components of aspect, namely, situation aspect and viewpoint aspect. The former is language independent while the latter is language specific. While the two-level approach to modelling situation aspect taken by them gives a better account of the compositional nature of situation aspect by proposing a set of rules mapping verb classes at the lexical level onto situation types at the sentential level, it also provides a more refined classification of situation aspect, most notably by distinguishing between two types of states (i.e. individual-level state and stage-level stage). Their work is a systematic and structured exploration of the linguistic devices which Chinese employs to express aspectual meanings. In addition to situation aspect, which is inherent in linguistic expressions of situations in human languages, the book identifies, on the basis of corpus data, four perfective and four imperfective viewpoints in Chinese. While some of these viewpoints have already been identified in previous studies of aspect in Chinese, this corpus-based study has corrected many intuition-based misconceptions and associated misleading conclusions readily found in the literature. Some viewpoints, e.g. the completive aspect marked by resultative verb complements (RVCs), were considered for the first time as independent viewpoints based on their behaviours in attested language use. Apart from the corpus studies of specific linguistic features in Chinese that have been reviewed so far, there is hardly any descriptive grammar of Chinese based on or informed by corpora. R. Xiao and McEnery (2010) might be an exception, which provides the first book-length corpus-based contrastive studies of major aspect-related grammatical categories in Chinese and English, i.e. grammatical categories such as aspect markers, completive and durative temporal adverbials, quantifiers, passives, and negation, which all contribute to aspectual meaning by interacting with situation aspect or viewpoint aspect in one way or 446

Corpus-based study of Chinese

another. However, this book is a research monograph more than a reference grammar of Chinese for general or pedagogical use.

3.  Chinese interlanguage research With the rapid development of Teaching Chinese as Foreign Language (TCFL) since the mid-1990s, there has been an increasingly pressing demand for Chinese interlanguage corpora to aid Chinese teaching and learning. As a result, a number of learner corpora of Chinese have been created over the past decade because for teachers of Chinese as a foreign language, learner corpora of Chinese can not only provide more direct and readily available help in teaching and learning, for example, in terms of real-time error analysis, computer-aided teaching, pertinent exercises for individual learners, learning evaluation, but they also play an increasingly important role in syllabus design, materials development, lexicography, and so on. The earliest corpus of learner Chinese is probably the Chinese Interlanguage Corpus created at Beijing Language and Culture University in 1993–5 (Chu et al. 1995). The corpus is composed of 1,371 compositions by 740 students, amounting to 1.04 million Chinese characters. It is encoded with 23 metadata features including, for example, the learner’s gender, age, nationality, and education level, as well as task type, topic, length, and writing time, and annotated with part-of-speech information and learner errors at character, word, and sentence levels. The corpus also comes with an integrated corpus exploration tool. Another corpus of similar nature developed by Beijing Language and Culture University is the HSK Dynamic Composition Corpus, which contains over 4.24 million Chinese characters of learner data in the form of 11,569 compositions written by learners of Chinese as a second or foreign language when they took the HSK Chinese language proficiency tests in 1992–2005. This corpus is encoded with rich metadata and tagged with learner errors at character, word, and sentence levels. The corpus is publicly available online (http://202.112. 195.192:8060/hsk/login.asp). After signing up for a free user account, registered users can search for sentences with a specific string of characters or an error type, or access various statistics as per the encoded metadata. As the compositions included in the corpus are continuous data over time, it is also suitable for longitudinal study. A number of other learner Chinese corpora have been reported in China. Most of them, however, do not appear to be publicly available. Apart from the development of corpora of Chinese interlanguage, a range of corpus-based studies that analyse specific grammatical features of learners’ Chinese interlanguage have been published. For example, Yang’s (2004) study is based on the Chinese Interlanguage Corpus, which compares Japanese learners’ acquisition of Chinese directional complements at different stages and analyses the factors affecting their acquisition. He finds that over-generalisation of object is the main cause of errors while over-generalisation is in turn a result of intralingual or interlingual transfer. Yuan (2005a, 2005b) provides: (i) a detailed analysis of about 100 sentences from a Chinese interlanguage corpus that contain errors with the negator 不 bu and (ii) 100 sentences containing errors with the negator 没有 meiyou, classifying them into various semantic and syntactic types. Explanations of the errors and corrections are also suggested in his papers. M. Wang (2005) investigates foreign learners’ acquisition of the Chinese 比 bi comparative structure, finding that their errors mainly relate to: (i) unnecessary addition; (ii) mismatch of adjectives and nouns; (iii) misuse of the bi structure; and (iv) sentence order inversion. The author argues that the method for correcting errors and improving language ability is to develop learners’ language awareness through intensive practice. 447

Tony McEnery and Richard Xiao

Zheng (2006) investigates the actual use of degree adverbs in the Chinese Interlanguage Corpus and compares her results with the published literature. She finds that the types of errors identified in the literature are all attested in the corpus data, but some error types in the literature which have been proposed on the basis of researchers’ introspection are not common in the corpus; there are also some new error types identified in the corpus which have not been predicted and accounted for in the literature. Hua (2009) analyses the types of errors made with the Chinese preposition 给 gei in Korean students’ interlanguage. Hua explores the reasons for the misuse from the perspective of first language (L1) interference, and proposes remedial teaching strategies to deal with the errors. Shen (2009) undertakes a corpus-based study of international students’ errors with the double-object structure in their Chinese interlanguage, finding that the most common types of errors with the structure relate to the omission of the indirect object or using a preposition to move the indirect object before the verb. Zhou and Hong (2010) investigate the use of figures of speech in the Chinese interlanguage of foreign learners at intermediate and advanced levels. Their results show that foreign learners use fewer types of figures of speech than native Chinese speakers, and they use figures of speech less frequently in their interlanguage; however, the error rate is not low. The errors are due to learners’ insufficient understanding of grammar rules, cultural diversity, and improper creative language use. It is found that the development of figures of speech in learners’ Chinese interlanguage is positively correlated with the level of Chinese language proficiency. B. Zhang (2011) investigates the current situation of and the problems with research on the acquisition of Chinese sentence structures in teaching Chinese as a foreign language. It was found that the central problem, among others, was that the research scope is too narrow and the empirical studies are insufficient. Consequently, we have no clear idea of the real situation of acquisition of Chinese sentence structures by foreigners, and the teaching strategies proposed in the previous studies also lack practical values. In addition to grammatical features, there are also numerous studies of lexical features in Chinese interlanguage. For example, Xing (2003) provides an exhaustive analysis of the 520 erroneous compounds of 17 subcategories in five categories found in the Chinese Interlanguage Corpus. Error analysis shows that foreign learners have strong awareness of the structures of compound words (e.g. words formed with morphemes, structures, and semantic relevance), and they may employ two different ways of acquiring compound words in Chinese, namely, compounding from morphemes and recognising compound words as a whole. It is believed that the first of these methods is dominant. Shi (2008) provides a systematic and comprehensive analysis of the errors with Chinese learners’ use of Chinese-style idioms, 成语 chengyu, on the basis of the Chinese Interlanguage Corpus, explores the reasons for the errors from the perspective of cognitive linguistics, and suggests some strategies for teaching 成语 chengyu to learners at advanced proficiency level. B. Zhang (2008) is concerned with confusable words in Chinese interlanguage. It is argued that confusable words should be inspected separately according to learners’ L1 backgrounds, and cross-linguistic contrast of Chinese with the learner’s L1 could be useful in uncovering the causes of errors they make. In discriminating a group of confusable words, more attention should be paid to the most confusable part, and the meaning characters of the words should be associated with their collocation rules. It can be seen even from this brief review of the development of Chinese learner corpora and various corpus-based studies of grammatical and lexical features of learner Chinese that Chinese interlanguage research has been a focus of TCFL research. On the other hand, the 448

Corpus-based study of Chinese

review above also reveals a number of issues with current Chinese interlanguage research. For example, in comparison with learner English corpora, there are very few existing Chinese interlanguage corpora that are publicly available and can actually be used in teaching and research. In relation to many reference corpora, Chinese interlanguage corpora are typically rather small in size, with annotated corpora composed of only 1 million Chinese characters. Existing Chinese learner corpora also suffer from a lack of balance in terms of learners’ first language backgrounds and the nature of the data included in the corpora. They are seriously biased towards Asian learners such as Korean, Japanese, and Southeast Asian learners, while learners from Europe and the Americas are seriously under-represented. The range of genres in the learner material is limited – such corpora are almost exclusively composed of compositions completed by foreign learners under test conditions. Currently available corpora hardly contain any spoken data. They also suffer from inaccurate and inconsistent annotation and limited public availability. In addition to these issues with corpus resources, corpus-based research of Chinese interlanguage has also been confined to error analysis while the actual usage patterns in learners’ interlanguage, which could be of interest to second-language acquisition research in their own right, have largely been overlooked.

4. Conclusion In this chapter, we have provided a survey of corpus-based study of Chinese by focusing on three major areas of linguistic research that have been deeply influenced by the development of the corpus methodology, namely, lexical study, grammatical study, and Chinese interlanguage research. We hope that this survey of corpus-based study in these areas will contribute to the further development of corpus linguistics in Chinese context.

References Beijing Aeronautical University (1985) Xiandai Hanyu Yong Zi Pindu Biao (A Frequency Table of Character Usage in Modern Chinese), Beijing: Beijing Aeronautical University. Beijing Language and Culture University (1986) Xiandai Hanyu Pinlü Cidian (A Frequency Dictionary of Modern Chinese), Beijing: Beijing Language and Culture University Press. Chen, H. (1922) ‘Yutiwen yingyong zihui’ (‘The Applied Glossary of Modern Chinese’), New Education 5(5). Chen, H. (1928) Yutiwen Yingyong Zihui (The Applied Glossary of Modern Chinese), Beijing: The Commercial Press. Chu, C., Chen, X., Zhang, W., Wei, P., Zhang, W., and Zhu, Q. (1995) Hanyu Zhongjieyu Yuliaoku Xitong Yanzhi Baogao (Research Report of the Corpus of Chinese Interlanguage), Beijing: Beijing Language and Culture University Press. Cui, L. (2011) ‘Yuliaoku zai duiwai Hanyu Xin Ciyu Cidian weiguan jiegou zhong de yunyong’ (‘Application of the Corpus in Microstructure of New Words Dictionary of CFL’), Journal of Chongqing University of Technology (Social Science) 25(10): 84–9. Hong Kong Polytechnic University (1991–7) Zhongguo Dalu, Taiwan, Xianggang Hanyu Ciku (A Chinese Word Bank from Mainland China, Taiwan, and Hong Kong), Hong Kong: Hong Kong Polytechnic University. Hua, X. (2009) ‘Hanguo liuxuesheng xide jieci gei de pianwu fenxi ji jiaoxue duice’ (‘The Error Analysis of Korean Students’ Acquisition of the Preposition Gei and the Corresponding Teaching Strategies’), Journal of the College of Chinese Language and Culture of Jinan University 2009 (1): 24–9. Huang, B. (2004) ‘Yixiang huafen de yiju yu biaozhun’ (‘The Basis and Criteria for Word Sense Division’), Lexicographical Studies 2005 (5): 31–6. Huang, C. R. and K. Ahrens (2003) ‘Individuals, Kind and Events: Classifier Coercion of Nouns’, Language Sciences 25(4): 353–73.

449

Tony McEnery and Richard Xiao Huang, C. R., Chen, K. J. and Lai, C. X. (1995) Mandarin Chinese Classifier and Noun-Classifier Collocation Dictionary, Taipei: Mandarin Daily Press. Jiang, Z. (2005) ‘Dianzi yuliaoku yu yuwen cishu de bianzuan xiuding’ (‘Electronic Corpora and Compilation and Revision of Chinese Dictionaries’), Journal of Sichuan University (Social Science Edition) 140(5): 79–86. Kang, S. (2003) ‘Xin Ciyu Da Cidian de bianzuan’ (‘Compilation of A Dictionary of New Words’), Lexicographical Studies 2003 (2): 12–20. Li, E. (2002) Xiandai Cidianxue Daolun (Introduction to Modern Lexicography), Shanghai: Chinese Dictionary Publishing House. Li, J. (1922) ‘Guoyu jiben yuci de tongji yanjiu’ (‘A Statistical Analysis of Basic Chinese Vocabulary’), Journal of Chinese Association 1(1). Li, J. and Zhong, L. (2011) ‘Jiyu pingxing yuliaoku de jiji-xing Han Ying cidian peili yuanze’ (‘Illustration Principles for Parallel Corpus-based Active Chinese–English Dictionary Compilation), Journal of Jiangsu University (Social Science Edition) 13(2): 44–8. Li, L. (2006) ‘Cong shuangyu shangwu cidian kan zhuanye yuliaoku dui cidian bianzuan de zhongyaoxing’ (‘Specialised Corpus for Dictionary Compilation Seen in a Bilingual Business Dictionary’), Lexicographical Studies 2006 (3): 93–100. Liu, E. S. (1973) Frequency Dictionary of Chinese Words, Berlin: Mouton. National Language Committee (1988) Xiandai Hanyu Changyongzi Biao (Commonly Used Characters in Modern Chinese), Beijing: National Language Committee. Project Code 748 (1976) Xiandai Hanzi Zonghe Shiyong Pindu Biao (A Comprehensive Frequency Table of Character Usage in Modern Chinese), Beijing: Special Research Group of Standard Characters. Shen, X. (2009) ‘Jiyu yuliaoku de waiguo xuesheng shuang binyu ju pianwu fenxi’ (‘A Corpus-based Error Analysis of Foreign Students’ Double-Object Sentences’), Journal of the Chinese Language 2009 (12): 1–4. Shi, L. (2008) ‘Jiyu zhongjieyu yuliaoku de chengyu shiyong pianwu fenxi’ (‘An Interlanguage Corpusbased Error Analysis of Chinese Idioms’), Social Scientist 130(2): 158–61. Siewierska, A., Xu, J., and Richard Xiao (2010) ‘Bang-le yi ge da mang’ (offered a big helping hand): A corpus study of the splittable compounds in spoken and written Chinese’, Language Sciences 32(4): 464–87. Sinclair, J. (ed) (2004) How to Use Corpora in Language Teaching, Amsterdam and Philadelphia: John Benjamins Publishing Company. Tao, X. and Zhu, J. (1923) Pingmin Qian Zi Ke (Thousand Character Lessons for Civilians), Shanghai: Youth Association Book Store. Thorndike, E. (1921) The Teacher’s Word Book, New York: Columbia Teachers College. Tsou, B. K. and You, R. (2010) Quanqiu Huayu Xin Ciyu Cidian (A Dictionary of Global Chinese Neologisms), Beijing: The Commercial Press. Tsou, B. K., Chin, A. C., and Kwong, O. Y. (2011) ‘From Synchronous Corpus to Monitoring Corpus, LIVAC: The Chinese Case’, Paper presented at the Third International Conference on Advances in Databases, Knowledge, and Data Applications (DBKDA 2011), St. Maarten. Wang, H. and Wang, T. (2009) ‘Jianli zai yuliaoku jichu shang de Hanyu benti yanjiu yu duiwai Hanyu jiaoixue: yi xiandai Hanyu liheci yanjiu wei li’ (‘Corpus-based Study of Chinese and Teaching Chinese as a Foreign Language: The Case of Splittable Compounds in Modern Chinese), Chinese Script Culture 88(2): 84–91. Wang, M. (2005) ‘Liuxuesheng bi zi ju xide de kaocha’ (‘A Study of Foreign Learners’ Acquisition of Bi Structure’), Journal of the College of Chinese Language and Culture of Jinan University 2005 (3): 28–35. Xiao, H. (2010) ‘Cidian duoyici yixiang guanxi yu ciyi qufen’ (‘The Sense Relations and Sense Distinction of Polysemes in the Dictionary’), Journal of Yunnan Normal University (Social Science Edition) 42(1): 43–8. Xiao, Richard and Tony McEnery (2004) Aspect in Mandarin Chinese: A Corpus-based Study, Amsterdam and Philadelphia: John Benjamins Publishing Company. Xiao, Richard and Tony McEnery (2008) ‘Negation in Chinese: A Corpus-based Study’, Journal of Chinese Linguistics 36(2): 274–330. Xiao, Richard and Tony McEnery (2010) Corpus-based Contrastive Studies of English and Chinese, London and New York: Routledge.

450

Corpus-based study of Chinese Xiao, Richard, Tony McEnery, and Qian Y. (2006) ‘Passive Constructions in English and Chinese: A Corpus-based Contrastive Study’, Languages in Contrast 6(1): 109–49. Xiao, Richard, P. Rayson, and Tony McEnery (2009) A Frequency Dictionary of Mandarin Chinese: Core Vocabulary for Learners, London and New York: Routledge. Xing, H. (2003) ‘Liuxuesheng pianwu hechengci de tongji fenxi’ (‘A Statistical Analysis of Errors with Compound Words by Foreign Learners of Chinese’), Teaching Chinese in the World 2003 (4): 67–78. Yang, D. (2004) ‘Riyu muyu xuexizhe quxiang buyu xide qingkuang fenxi’ (‘A Study of the Japanesespeaking Students’ Acquisition of Chinese Directional Complements’), Journal of the College of Chinese Language and Culture of Jinan University 2004 (3): 23–35. Yuan, Y. (2005a) ‘Shi xi zhongjieyu zhong gen bu xiangguan de pianwu’ (‘An Analysis of Interlanguage Errors Related to Bu’), Language Teaching and Research 2005 (6): 39–47. Yuan, Y. (2005b) ‘Shi xi zhongjieyu zhong gen meiyou xiangguan de pianwu’ (‘An Analysis of Interlanguage Errors Related to Meiyou’), Chinese Teaching in the World 2005 (2): 56–70. Zhang, B. (2008) ‘Di’er yuyan xuexizhe Hanyu zhongjieyu yi hunxiao ci jiqi yanjiu fangfa’ (‘On Confusable Words in Chinese Interlanguage and Related Research Methods’), Language Teaching and Research 2008 (6): 37–45. Zhang, B. (2011) ‘Waiguoren Hanyu jushi xide yanjiu de fangfalun sikao’ (‘On the Methodology of Chinese Sentence Acquisition by Foreigners’), TCSOL Studies 42(2): 23–9. Zhang, Q. (2010) ‘Chengdu buyu X de hen de jufa ji yuyong fenxi’ (‘A Syntactic and Pragmatic Analysis of the Complement of Degree X de Hen), Modern Chinese 2010 (8): 47–9. Zhao, Z. (2010) ‘Zai tan laizhe: jiyu yuliaoku de kaocha’ (‘A Restudy of Laizhe: A Corpus-based Investigation’), Journal of Linyi Normal University 32(2): 90–4. Zheng, Y. (2006) ‘Zhongjieyu zhong chengdu fuci de shiyong qingkuang fenxi’ (‘An Analysis of Foreign Learners’ Use of Degree Adverbs’), Chinese Language Learning 2006 (6): 66–72. Zhou, X. and Hong, W. (2010) ‘Zhonggaoji liuxuesheng Hanyu zhongjieyu cige shiyong qingkuang kaocha’ (‘A Study on the Use of Figures of Speech in the Chinese Interlanguage of the Intermediate and Advanced Chinese Learners’), Teaching Chinese in the World 2010 (4): 536–5.

451

26 Elastic words in Chinese1 San Duanmu and Yan Dong University of Michigan, USA

1.  What are elastic words? Chinese has many long–short pairs of synonymous word forms. Some examples are shown in Table 26.1. The long form is made of the short one plus another morpheme, which can itself be a word. However, the extra part of the long form contributes little new meaning; therefore, it is shown in parentheses. Since the two forms share the same basic part and have the same meaning, many people consider them to be variants of the same word (Karlgren 1918, 1923; Guo 1938; Chao 1948). We follow Guo (1938) and refer to them as elastic words, or words with elastic length. Theoretically, there is nothing unusual for a word to have two (or more) phonetic forms (Bloomfield 1933). For example, ’m [m] (as in I’m) and am [æm] are different forms of the same English word, so are math and mathematics. What is unusual in Chinese is that the long form is often in the shape of a compound, a point we shall return to. Elastic words in Table 26.1 represent the most common cases. Less common cases are also found, shown in Tables 26.2–26.4. In Table 26.2, the long form is trisyllabic. In Table 26.3, the two parts of the long form have similar meanings and either can be used alone. In Table 26.4, a short form has more than one long form. In this chapter, however, we shall focus on the most common cases, such as those in Table 26.1.

Table 26.1  Examples of elastic words in Chinese Short 见 学 煤 泪 蒜 虎 教 写

Long ‘see’ ‘study’ ‘coal’ ‘tear’ ‘garlic’ ‘tiger’ ‘teach’ ‘write’

(看)见 学(习) 煤(炭) (眼)泪 (大)蒜 (老)虎 教(书) 写(字)

452

‘(look)-see’ ‘study-(practice)’ ‘coal-(charcoal)’ ‘(eye)-tear’ ‘(big)-garlic’ ‘(old)-tiger’ ‘teach-(book)’ ‘write-(word)’

Elastic words in Chinese Table 26.2  Monosyllabic–trisyllabic pair Length

Example

Long Short

加拿大 加

‘Canada’ ‘Canada’

加拿大元 加元

‘Canadian dollar’ ‘Canadian dollar’

Table 26.3  Either part of the long form can be the short form Length

Example

Long Short Short

需要 需 要

‘need-want’ ‘need’ ‘want’

不需要 不需 不要

‘not need-want’ ‘not need’ ‘not want’

Table 26.4  More than one long form Length

Example

Short Long Long Long Long Long Long

泪 泪(水) 泪(花) (眼)泪 泪(珠) 泪(液) (热)泪

‘tear’ ‘tear-(water)’ ‘tear-(flower)’ ‘(eye)-tear’ ‘tear-(pearl)’ ‘tear-(liquid)’ ‘(warm)-tear’

The use of elastic words in Chinese is extremely common, so common that most Chinese speakers, scholars or average people alike, hardly feel anything unusual about it. However, when Chinese expressions are translated into English literally, the oddity becomes apparent. Consider the example in Table 26.5–Table 26.7. In Table 26.5, four of the six Chinese words are repetitive (although the sentence can be used without the repetitive parts, too). In Table 26.6, an object is required in a Chinese answer, even though it is obvious from Table 26.5  An ordinary Chinese sentence and its English translation 我 ‘I

刚(刚) just-(just)

(看)见 (look)-see

他 him

(回)到 (return)-reach

家(里)。 home-(inside)’

Table 26.6  Object required in a Chinese answer Question

你想做什么？ you want do what ‘What do you want to do?’

Odd answer

教/写/走/吃. ‘Teach/Write/Walk/Eat’

Normal answer

教书/写字/走路/吃饭. ‘Teach book/Write word/Walk road/Eat meal’

453

San Duanmu and Yan Dong Table 26.7  Object not required in an English answer Question Normal answer Odd answer

‘What do you want to do?’ ‘Teach/Write/Walk/Eat’ ‘Teach books/Write words/Walk roads/Eat meal’

Table 26.8  Questions to be discussed with regard to elastic words a.  b.  c.  d. 

How are elastic words defined? How many words are elastic in modern Chinese? Why does Chinese have so many elastic words? Impact of elastic words on poetry

the verb. For example, in the normal case, what else do you write if not words (or books)? In contrast, as seen in Table 26.7, corresponding English answers are fine without an object, and preferably so. To the Western eye, therefore, Chinese seems highly repetitive or redundant. Naturally, when Karlgren (1918) started learning Chinese, he immediately noted this oddity and reported it. Since then, many others have discussed the topic (e.g. Jespersen 1930; Guo 1938; Chao 1948; Kennedy 1951, 1955; Lü 1963; C. N. Li and Thompson 1981; Pan 1997; J. Ke 2006; Jin 2011), yet many important questions remain unanswered. In this chapter we address such questions. In particular, we examine the issues in Table 26.8. In section 2 we consider (26.8a). In section 3 we consider (26.8b). In section 4 we consider (26.8c). In section 5 we consider (26.8d). In section 6 we offer concluding remarks.

2.  How are elastic words defined? In previous studies, it is often left to the intuition of the reader whether two forms, such as 店 ‘store’ and 商店 ‘business-store’, are different words or variants of the same word. Many scholars assume that pairs like 店 ‘store’ and 商店 ‘business-store’ are variants of the same word (e.g. Karlgren 1918, 1923; Jespersen 1930; Guo 1938; Chao 1948; Kennedy 1951, 1955; Lü 1963; C. N. Li and Thompson 1981; Pan 1997; J. Ke 2006; Jin 2011). For example, Chao (1948) points out that, while most Chinese morphemes are monosyllabic, most Chinese words are disyllabic. Chao’s statement makes sense only if many disyllabic forms are simple words, and not true compounds. The reason is that a true compound is made of two (or more) words that contribute different meanings. For example, 书包 ‘book bag (backpack)’ and 煤店 ‘coal store’ are true compounds, because ‘book bag’ is not the same as ‘book’ or ‘bag’, and ‘coal store’ is not the same as ‘coal’ or ‘store’. If all disyllabic forms are true compounds in Chinese, we cannot reasonably say that Chinese words are disyllabic, such as 书 ‘book’, 包 ‘bag’, 煤 ‘coal’, and 店 ‘store’. Chao’s statement makes sense only if many disyllabic forms are also simple words, such as 书(本) ‘book-(volume)’, 煤(炭) ‘coal-(charcoal)’, and (商)店 ‘(business)-store’. Because the long form of an elastic word looks like a compound, some scholars treat all disyllabic forms as true compounds. In addition, they propose that, even for related pairs, the long form always has a different meaning from the short one (e.g. N. Li 1990; Wang 2002; Wu 2003; H. Ke 2007). Some examples have been offered to support their view, such as that in Table 26.9. 454

Elastic words in Chinese Table 26.9  Semantic difference between short and long forms (N. Li 1990) Length Short Long

死 死亡

‘die’ ‘die-perish’

Basic meaning

Additional meaning

‘die’ ‘die’

abstract concrete

N. Li (1990) observes that, when used as a verb, the short form ‘die’ has a concrete meaning, whereas the long form ‘die-perish’ has an abstract meaning. Therefore, ‘die-perish’ can be used for an abstract entity, such as ‘society’, but not for a concrete entity, such as a person or an animal. In contrast, ‘die’ has a concrete meaning and can be used for concrete entities, such as a person or an animal. However, examples like that in Table 26.9 are sporadic, and quantitative data are lacking. In addition, it can be shown that most elastic words do not have such meaning distinctions between the long and short forms. Moreover, as is often observed before, and as we shall see below with quantitatively data, while both languages use compounds, elastic words occur extensively in Chinese but not in English. To treat elastic words as regular compounds is to miss an important different between the two and an important difference between Chinese and English. Let us now consider an explicit definition of elastic words, shown in Table 26.10 and Table 26.11, based on Duanmu (2012) and Huang and Duanmu (2013). Strictly speaking, (26.10a) is not necessary, but it represents a property of the majority of elastic words in Chinese. Given (26.10a), (26.10b) means that B contains A. (26.10c) means that the extra part of B adds no new meaning to A and is therefore semantically redundant. (26.10d) is expected if (26.10c) is the case. Of the four conditions in the definition, (26.10a) and (26.10b) are easy to verify, so is (26.10d). An example of (26.10d) is shown in Table 26.11. (26.11a) and (26.11b) show that 工(人) ‘worker-(person)’ and 工 ‘worker’ are inter­ changeable. (26.11b) and (26.11c) show that 技(术) ‘skill-(method)’ and 技 ‘skill’ are interchangeable. Let us now consider (26.10c). It can often be a subtle judgment whether two forms have exactly the same meaning. However, we can rely on an objective criterion, which is mutual annotation in dictionary entries. For example, if A is defined as B in a dictionary and B Table 26.10  Defining elastic words in Chinese An elastic word has two length forms, A and B, where a.  b.  c.  d. 

A is monosyllabic and B disyllabic A and B share the same base (morpheme) A and B have the same meaning A and B are interchangeable in some contexts

Table 26.11  Three interchangeable expressions for ‘skilled worker’ a.  技术 b.  技术 c.  技

工人 工 工

‘skill-method ‘skill-method ‘skill

455

worker-person’ worker’ worker’

San Duanmu and Yan Dong Table 26.12  Mutual dictionary annotation of 煤 ‘coal’ and 煤炭 ‘coal-charcoal’ Entry

Annotation

煤 煤炭

noun,  .  .  .  , also called 煤炭 ‘coal-charcoal’ noun, 煤 ‘coal’.

Table 26.13  Word pairs that do or do not form elastic words Short 煤 虎 乐 包

‘coal’ ‘tiger’ ‘glad’ ‘bag’

Long 煤炭 老虎 高兴 书包

‘coal-(charcoal)’ ‘(old)-tiger’ ‘high-mood’ ‘book-bag’

Same meaning

Same base

Elastic

Yes Yes Yes No

Yes Yes No Yes

Yes Yes No No

defined as A, then we consider A and B to have the same meaning. An example is shown in Table 26.12. The definitions are found in Modern Chinese Dictionary (Xiandai Hanyu Cidian 2005), where  .  .  .  is a description of ‘coal’. Our definition of elastic words is both explicit and implementable. Its application is illustrated in Table 26.13. According to our definition, 煤 and 煤炭 are variants of the elastic word ‘coal’ and 虎 and 老虎 are variants of the elastic word ‘tiger’. In contrast, 乐 ‘glad’ and 高兴 ‘high-mood (glad)’ are different words, so are 包 ‘bag’ and 书包 ‘book-bag’.

3.  How many words are elastic in Modern Chinese? Having defined what elastic words are, let us consider how many there are in Chinese. Previous scholars generally agree that there are many. Some even suggest that ‘nearly all’ Chinese words have elastic length (Pan 1997: 140). However, quantitative data have not been offered until quite recently. Three studies are summarized in Table 26.14. Duanmu (2013, written in 2011) sampled 1/50 of 3,000 most commonly used Chinese characters, totaling 60 characters and 84 words (some characters represent more than one word). It was found that elastic words constitute 79% of the samples. Huang and Duanmu (2013) sampled 10% of all morphemes in Modern Chinese Dictionary (Xiandai Hanyu Cidian 2005, MCD for short), totaling 2,000 word senses. Items that have related meanings but differ in part of speech are counted as different. For example, 学习 ‘study’ can be a noun or a verb, and each is counted as a separate item. Of the 2,000 word senses examined, 61% were found to be elastic (after excluding names). Table 26.14  Quantitative studies on the percentage of elastic words in Chinese Study

Data size

Elastic

Duanmu (2013) Huang and Duanmu (2013) Duanmu and Dong (2014)

60 characters; 84 words 2,000 morphemes in MCD All morphemes in MCD

79% 61% 47%

456

Elastic words in Chinese Table 26.15  POS and length properties of Chinese morphemes POS

All

POS %

1-only

Poly-only

Elastic

Noun Verb Adjective Adverb Others All

9,559 5,904 2,709 429 1,286 19,887

48.1% 29.7% 13.6%  2.2%  6.5% 100%

32.5% 56.1% 53.3% 72.5% 75.0% 46.0%

 8.9%  1.8%  9.4%  0.2% 18.0%  7.2%

58.6% 42.1% 37.2% 27.3%  7.0% 46.8%

Duanmu and Dong (2014) examined all morphemes in MCD, totaling some 10,000 word entries and 20,000 word senses (parts of speech distinguished). It was found that 47% of them are elastic words. The details are shown in Table 26.15. The label ‘1-only’ refers to monosyllabic morphemes without a disyllabic form, such as 水 ‘water’, 人 ‘person’, and 狗 ‘dog’. The label ‘Poly-only’ refers to (mostly disyllabic) morphemes without a monosyllabic form, such as 琥珀 ‘amber’, 葡萄 ‘grape’, and 儒艮 ‘dugong’. The label ‘Elastic’ refers to morphemes that can be monosyllabic or disyllabic, such as 虎–老虎 ‘tiger’–‘old tiger’, 眉–眉毛 ‘brow’–‘eyebrow’, and 煤–煤炭 ‘coal’–‘coal-charcoal’. The result of Duanmu and Dong (2014) is likely to be more accurate, since its data coverage is exhaustive and many times larger than previous ones. The average percentage of elastic words is lower than those reported in previous studies. The result seems counter to the intuition of some scholars that most Chinese words are or can be disyllabic in use (e.g. Chao 1948; Pan 1997). It turns out that most morphemes that do not have a disyllabic form are either verbs, which often have an object to occur with, or a noun that must occur with another morpheme in a disyllabic set phrase or compound. Three examples are shown in Table 26.16. None of the morphemes in Table 26.16 has elastic length, and none of them is used alone either. The word 教 ‘teach’ has a default object ‘book’ if it is to be used alone, regardless of whether a book is used or not. The word 辜 ‘guilt’ is used in a few set expressions only, such as ‘no guilt (innocent)’ and ‘extra guilt’. The word 点 ‘point’ is always used with another word, too, which it is a point of, except in geometry, where it has a special meaning and can be used as an independent noun. If we take such cases into consideration, then 98% of Chinese words can indeed occur in disyllabic forms, either in the long form of an elastic word or in combination with another syllable in a set phrase (Duanmu 2015).

Table 26.16  Example: Monosyllabic word without elastic length Morpheme

Use

教 ‘teach’

教书 ‘teach book (teach)’ 教英语 ‘teach English’ 无辜 ‘no guilt (innocent)’ 余辜 ‘extra guilt’ 起点 ‘start point’ 沸点 ‘boil point’ 冰点 ‘ice point (freezing point)’

辜 ‘guilt’ 点 ‘point’

457

San Duanmu and Yan Dong Table 26.17  Morphologically and semantically related word pairs in English Short

Long

Ben prof math dog brow

Benjamin professor mathematics doggy eyebrow

There is no study on how many English words have elastic length. Various examples can be proposed though, such as those in Table 26.17. The first three pairs are based on truncation, where the extra part of the long form is not an independent morpheme. In contrast, the long form of an elastic word in Chinese is typically made of two independent morphemes. The word doggy is arguably made of two morphemes, but it differs in stylistic meaning from dog; therefore, dog and doggy are hardly interchangeable. However, the pair brow–eyebrow does satisfy the definition of elastic words, similar to 泪 ‘tear’ and 眼泪 ‘eye-tear’ in Chinese. Nevertheless, it is reasonable to say that word pairs like brow–eyebrow are fairly rare in English and constitute far lower than 47% of the vocabulary.

4.  Theories on elastic words Having confirmed the abundance of elastic words in Chinese (and their rarity in English), let us consider why this is the case. Various theories have been proposed. Let us consider five of them, shown in Table 26.18. The ‘speech rate’ theory is proposed by Guo (1938). The idea is that in some positions we need to speak faster and in some we need to speak slower. Monosyllabic forms are good for the former and disyllabic forms are good for the latter, when one would start to repeat words. It is further proposed that speech rate is a property of spoken Chinese and that written Chinese ought to use monosyllabic words most of the time, unless it tries to reflect the spoken style, as is the case in modern writing. Moreover, it is proposed that elastic words have existed in Chinese all along, as can be observed in early poetry The ‘processing need’ theory is proposed by Pan (1997: 177). The idea is that the information conveyed by monosyllabic words is too dense for the listener, who cannot process what is heard fast enough. To help the listener, the speaker slows down by using repetitive words. The ‘disyllabic origin’ theory is proposed by Kennedy (1951, 1955). The idea is that most Chinese words were originally disyllabic in the first place, as some still are, such as 蝴蝶 ‘butterfly’. However, because Chinese dictionaries use characters as entries, instead of using Table 26.18  Theories on why Chinese has so many elastic words a.  Speech rate b.  Processing need c.  Disyllabic origin d. Homophony e. Prosody

458

Elastic words in Chinese

words, each disyllabic word is split into two entries. For example, 蝴蝶 would have an entry under 蝴, where the annotation would be ‘butterfly’, and another entry under 蝶, where the annotation would again be ‘butterfly’. As a result, users of such dictionaries get the mis­ impression that each entry is a word and start using them as such, mistakenly. Gradually, most Chinese words have become monosyllabic. The ‘homophony’ theory is proposed by Karlgren (1918). It is based on the fact that Classic Chinese used to have 3,000–4,000 distinct syllables, whereas modern Standard Chinese only has 1,300. Since most Chinese morphemes are monosyllabic, there is a high degree of homophony. If disyllabic words are introduced, homophony can be radically reduced and ambiguity avoided. It is proposed, therefore, that Chinese created disyllabic words precisely to serve this purpose. In addition, since monosyllabic words are still alive, we end up with the situation of many elastic words. The ‘prosody’ theory is what we would like to argue for. The idea is that some positions are stressed (by word stress or phrasal stress) and others not, and that stressed positions need a disyllabic word, owing to the phonological requirement of Foot Binarity (Prince 1980). To satisfy Foot Binarity, disyllabic forms are created. Little evidence has been shown for the ‘processing need’ theory or the ‘disyllabic origin’ theory. We shall, therefore, not discuss them any further. Instead, we shall focus on the homophony theory and the prosodic theory and compare their predictions. The comparison has implications for the speech rate theory, too, to be discussed later. The homophony theory is intuitively quite plausible. Naturally, it is shared by many other scholars (e.g. Jespersen 1930; Guo 1938; Lü 1963; T’sou 1976; C. N. Li and Thompson 1981; Nettle 1995; J. Ke 2006; Jin 2011). Indeed, it has been called the ‘orthodox’ (Kennedy 1955). However, apart from sporadic examples, little evidence for it has been shown. Let us consider its predictions and compare them with those of the prosody theory, shown in Table 26.19. First, the homophony theory assumes that disyllabic words serve to avoid ambiguity. When there is no ambiguity, there is no need to use disyllabic words. Therefore, there ought to be no ‘minimal word’ effect, i.e. a word spoken alone need not be disyllabic but could be monosyllabic, if the meaning is clear. The homophony theory also predicts a positive correlation between homophone density (the number of homophones per syllable) and the percentage of elastic words. Syllables that have more homophones are more likely to have elastic length. On the other hand, stress ought to have no influence on the percentage of elastic words. In addition, since Modern Chinese has much fewer syllables than Classic Chinese, there ought to be a lot more elastic words in Modern Chinese than in Classic Chinese. Moreover, disyllabic words should occur in spoken Chinese only. In written Chinese, Table 26.19  Predictions of the homophony theory and the prosody theory

Minimal word Homophone Stress Spoken Written Modern Classic

Homophony

Prosody

not relevant relevant not relevant yes no yes no

relevant not relevant relevant yes yes yes yes

459

San Duanmu and Yan Dong

the number of distinct graphs remains similar to that in Classic Chinese and far exceeds the number of distinct syllables. Therefore, there is little need for disyllabic words in written Chinese. The prosody theory makes fairly different predictions. First, there ought to be a ‘minimal word’ effect, because a word spoken alone has stress and should be disyllabic, whether it is ambiguous or not. Second, there is no correlation between homophone density and the percentage of elastic words, since the creation of elastic words is driven by Foot Binarity, not by homophony. Third, stress is relevant for the use of disyllabic words. Fourth, there is no reason written Chinese should not reflect the effect of Foot Binarity; therefore, elastic words ought to be found in both spoken and written Chinese. Finally, the prosodic requirement is likely to be the same in both Classic Chinese and Modern Chinese. Therefore, we expect there to be just as many elastic words in Classic Chinese as in Modern Chinese. In summary, the two theories agree that elastic words are extensively used in spoken, Modern Chinese. However, they differ in whether the minimal word effect, homophony, stress, written Chinese, and Classic Chinese are related to elastic words. Therefore, we examine these factors in detail.

4.1.  The minimal word effect It is well known that Chinese has a ‘minimal word’ requirement, by which monosyllables are not spoken alone. Many examples have been offered by Lü (1963). In addition, this is the case even if there is no ambiguity. An example is shown in Table 26.20. A monosyllabic mountain name is usually accompanied by the word ‘mountain’ in order to become disyllabic, presumably because it would otherwise be ambiguous, owing to many homophones. However, when there is no ambiguity, as in Table 26.20, the answer still has to be disyllabic. In general, all monosyllabic names must be used with another syllable when spoken alone, such as 姓王 ‘named Wang’, 沙市 ‘Sha city’, 法国 ‘France country’, 华山 ‘Hua mountain’, and 鲤鱼 ‘carp fish’. In contrast, disyllabic names can be spoken alone without an extra word, such as 欧阳 ‘Ouyang’ (a personal name), 上海 ‘Shanghai’, 荷兰 ‘Holland’, and 峨嵋 ‘Emei’ (a mountain name).

4.2.  Elastic words and homophony Next we consider the correlation between homophone density and the percentage of elastic words. The homophony theory predicts the correlation to exist. Some small-scale studies have been offered (T’sou 1976; J. Ke 2006; Jin 2011), but questions remain (see review in Duanmu and Dong 2014). In contrast, the prosody theory predicts that there is no such Table 26.20  Minimal word requirement in Chinese when there is no ambiguity Question:

你去华山还是泰山？ You go Hua Mountain or Tai Mountain Will you go to Mt. Hua or Mt. Tai?

Normal answer:

泰山。 Tai Mountain.

Odd answer:

泰。 Tai.

460

Elastic words in Chinese

correlation. To verify the correlation, we used a large set of lexical data. In particular, we annotated all morphemes in Modern Chinese Dictionary, totaling some 20,000 items. For each monosyllabic morpheme, we annotated its homophone density (the number of homophones it has) within its POS category. For each set of syllables that share the same homophone density, we annotated its percentage of elastic words. The results for nouns (excluding family names), which constitute the largest POS category, are shown in Tables 26.21 and 26.22 (Duanmu and Dong 2014). It can be seen that, in most homophone density groups, the percentage of elastic words lies between 60% and 70%. Statistic results in Table 26.22 show that the confidence interval of the correlation crosses the zero point, which means there is no correlation between homophone density and the percentage of elastic words.

4.3.  Elastic words and stress To see the effect of stress, let us consider word length preferences in [N N] compounds and [V O] phrases that are made of two words each. If both words have elastic length, there are four length combinations, 2+2, 2+1, 1+2, and 1+1, where ‘2’ is a disyllabic form and ‘1’ a monosyllabic one. However, it is well known that 1+2 is disfavored in [N N] and 2+1 is disfavored in [V O] (Lü 1963). This is exemplified in Table 26.23 and Table 26.24, where ‘*’ indicates a disfavored length combination. The homophony theory has no explanation for such length preferences. In contrast, the prosody theory can offer a natural account in terms of stress rules and Foot Binarity. First, let us consider the phonological requirements, given in Table 26.25. The stress rules (26.25a) and (26.25b) are similar to those in English, where stress is assigned to the left in [N N] and right in [V O] (Chomsky and Halle 1968). They can be combined to a single rule, too, which has been called Stress-XP by Gussenhoven (1983) and Truckenbrodt (1995) and Nonhead Stress by Duanmu (1990, 2007). Foot Binarity is a wellknown requirement in phonology, originally proposed by Prince (1980). Finally, (26.25d) is a very common process in Chinese, as has been noted by Shih (1986), Chen (2000), and Feng (1995). Let us now consider the analysis of [N N], shown in Table 26.26, where S indicates a syllable and boldface indicates stress. In 2+2, each noun forms a binary foot. In 2+1, the first noun forms a binary foot; the second noun has no stress and need not form a foot. In 1+2, the second noun forms a binary foot. The first noun has stress and must form a foot, but it has only one syllable, hence Foot Binarity is violated. In 1+1, the two nouns can form a foot. Next we consider [V O], shown in Table 26.27. In 2+2, each word forms a binary foot. In 2+1, V forms a binary foot; O has stress and must form a foot, but it has only one syllable, hence Foot Binarity is violated. In 1+2, O forms a binary foot; V has no stress and need not form a foot. In 1+1, the two words can form a foot. The analysis shows that stress is highly relevant for the use of disyllabic words. In contrast, the homophony theory offers no explanation for word length preferences.

4.4.  Elastic words in written Chinese According to Karlgren (1918), homophony is a problem only in spoken Chinese, which can only distinguish some 1,000 distinct syllables. In contrast, the number of distinct graphs in 461

San Duanmu and Yan Dong Table 26.21  Homophone density and the percentage of elastic words among noun morphemes in Chinese. Nouns that are polysyllabic only are excluded. A homophone density of 1 means a word has no homophone (but itself ) Homophones

1-only

Elastic

All

Elastic %

 1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 40 42 44 47 48 49 51 56 62 64 All

75 91 101 101 139 129 147 139 141 87 115 132 133 133 105 65 98 72 117 81 54 96 49 59 37 52 77 42 8 17 22 34 11 10 20 26 24 17 15 14 35 19 16 8 23 30 19 3,035

92 125 199 219 226 267 252 237 264 203 248 240 270 203 225 159 174 126 225 139 93 168 89 85 63 104 166 98 21 43 40 62 22 24 50 46 50 23 27 30 59 29 33 43 33 32 45 5,671

167 216 300 320 365 396 399 376 405 290 363 372 403 336 330 224 272 198 342 220 147 264 138 144 100 156 243 140 29 60 62 96 33 34 70 72 74 40 42 44 94 48 49 51 56 62 64 8,706

55% 58% 66% 68% 62% 67% 63% 63% 65% 70% 68% 65% 67% 60% 68% 71% 64% 64% 66% 63% 63% 64% 64% 59% 63% 67% 68% 70% 72% 72% 65% 65% 67% 71% 71% 64% 68% 58% 64% 68% 63% 60% 67% 84% 59% 52% 70% 65%

462

Elastic words in Chinese Table 26.22  Statistics of the data in Table 14.21, which show no correlation between homophone density and the percentage of elastic words Correlation: Confidence interval (95%): Multiple R-squared: F-statistic: P-value:

0.089 –0.066 0.122 0.008 0.3628 on 1 and 45 degrees of freedom 0.55

Table 26.23  In [N N], 1+2 is disfavored 2+2 2+1 *1+2 1+1

技术 技术 技 技

工人 工 工人 工

skill-skill skill-skill skill skill ‘skill

worker-man worker worker-man worker worker’

大蒜 蒜 大蒜 蒜

plant-plant plant-plant plant plant ‘plant

big-garlic garlic big-garlic garlic garlic’

Table 26.24  In [V O], 2+1 is disfavored 2+2 *2+1 1+2 1+1

种植 种植 种 种

Table 26.25  Phonological requirements (boldface indicates stress) a.  b.  c.  d. 

Compound Stress: [N N] Phrasal Stress: [V O] Foot Binarity: A syllabic foot must have two syllables Two monosyllables can form a foot in Chinese

Table 26.26  Analysis of [N N] 2+2 2+1 *1+2 1+1

(SS)+(SS) (SS)+S (S)+(SS) (S+S)

Second N has no stress and need not form a foot First N has stress and must form a foot Two monosyllables can form a foot

Table 26.27  Analysis of [V O] 2+2 *2+1 1+2 1+1

(SS)+(SS) (SS)+(S) S+(SS) (S+S)

O has stress and must form a foot V has no stress and need not form a foot Two monosyllables can form a foot

463

San Duanmu and Yan Dong Table 26.28  Word length patterns for [N N] in written Chinese (token frequencies) Length

Frequency

2+2 2+1 1+2 1+1

21.8% 21.3%  1.1% 55.8%

Table 26.29  Word length patterns for [V O] in written Chinese (token frequencies) Length

Frequency

2+2 2+1 1+2 1+1

16.2%  1.8% 19.1% 62.8%

written Chinese has remained the same, totaling about 10,000. Therefore, there is no need for using disyllabic words in written Chinese. Now let us consider word length preferences in written Chinese, using the Lancaster Corpus of Mandarin Chinese (McEnery and Xiao 2004), which total 1.5 million graphs (1.3 million Chinese characters). The results in token frequencies of [N N] and [V O] are shown in Table 26.28 and Table 26.29, based on Duanmu (2012). The results in type frequencies are similar and are omitted. The results show that word length preferences in written Chinese are similar to those in spoken Chinese, where 1+2 [N N] is disfavored, so is 2+1 [V O]. The results are predicted by the prosody theory but not by the homophony theory.

4.5.  Elastic words in Classic Chinese Because Classic Chinese has three times as many distinct syllables as modern Standard Chinese, the homophony theory predicts that Classic Chinese has a much lower percentage of elastic words. In contrast, the prosody theory predicts that prosodic requirements should be the same; therefore, Classic Chinese should have just as many elastic words as Modern Chinese does. Dong (2015) examines elastic words in Classic Chinese. The result confirms the prediction of the prosody theory.

4.6. Summary Let us summarize our discussion on whether various properties are related to elastic words. In Table 26.30 we compare how many things each of three theories can correctly account for. The homophony theory can account for the high percentage of elastic words in modern spoken Chinese, but none of the other effects. The speech rate theory can account for two more effects: the role of stress, if we assume that one needs to be slower when speaking a stressed word, and the high percentage of elastic words in Classic Chinese. In contrast, the prosody theory can account for all the seven properties. Therefore, it is by far the best theory on both the creation and the use of elastic words. 464

Elastic words in Chinese Table 26.30  A comparison of how many things each theory can correctly account for

Minimal word Homophone Stress Spoken Chinese Written Chinese Modern Chinese Classic Chinese

Homophony

Speech rate

Prosody

No No No Yes No Yes No

No No Yes? Yes No Yes Yes

Yes Yes Yes Yes Yes Yes Yes

Table 26.31  Elastic words created by truncation Original

Truncated

轮船 ‘wheel boat (ship)’ 荷兰 ‘Holland’

轮 ‘ship’ 荷 ‘Holland’

It is worth pointing out that elastic words do not always come from expanding a monosyllabic word to a disyllabic one. The reverse process can happen, too, where a disyllabic word or compound is truncated to a monosyllabic one, so that the latter acquires the full meaning of the former. Some examples are shown in Table 26.31. A powered ship is originally called a ‘wheel boat’. The compound is then truncated to ‘wheel’, which acquires the meaning of ‘ship’ as well. This process creates the elastic word 轮船–轮 ‘ship-(boat)’–‘ship’. Similarly, the disyllabic word 荷兰 ‘Holland’ is truncated to 荷, which acquires the full meaning of ‘Holland’, creating the elastic word 荷兰–荷 ‘Holland’. The prosody theory can account for both processes of creating elastic words, expansion and truncation, because prosody requires both long and short forms. Specifically, disyllabic forms are needed for stressed positions and monosyllabic forms are needed for certain unstressed positions. For example, the ill-formed 1+2 [N N] can be repaired if the second N is truncated to a monosyllable, so that the result is well-formed 1+1. In contrast, the homophony theory assumes that Chinese is already overwhelmed by homophony; therefore, it is hard to explain why Chinese keeps creating more of it through truncation.

5.  Elastic words and poetry The wide use of elastic words no doubt makes Chinese seem repetitive. Consider two examples in Table 26.32, where the first was given earlier and repetitive words are shown in parentheses. Chinese speakers do not find such sentences repetitive though, because they often ignore the internal structure of disyllabic units, an effect called ‘Foot Shelter’ by Duanmu (2007), and treat each disyllabic unit as a single word. Thus, to a Chinese speaker, the first sentence in Table 26.32 has six words, not ten, and the second sentence has three words, not six. The presence of elastic words has made it very easy to compose rhythmic poems. Specifically, when a position needs one syllable, we can use the short form of an elastic word, and when a position needs two syllables, we can use the long form of an elastic word. An example is shown in Tables 26.33 and 26.34, where S and W represent strong and weak beats. 465

San Duanmu and Yan Dong Table 26.32  Repetitive words (indicated by parentheses) in two Chinese sentences 我 ‘I

刚(刚) just-(just)

保(持) Ensure-(keep)

(看)见 (look)-see 车(厢) train-(box)

他 him

(回)到 (return)-reach

家(里)。 home-(inside)’

(清)洁 (clear)-clean

Table 26.33  A poem with extra positions to fill (three missing syllables per line) SW

SW

SW

保 皮

车 扔

洁 外

Ensure Peel

train throw

clean out

Table 26.34  Elastic words can fill extra positions SW

SW

SW

保(持) (果)皮

车(厢) 扔(出)

(清)洁 (窗)外

Ensure-(keep) (Fruit)-peel

train-(box) throw-(exit)

(clear)-clean (window)-out

The poem has six syllable positions per line (three strong beats and three weak beats), but there are just three words each. If we use monosyllabic words, we are missing three syllables per line. If we use elastic words, all positions are filled. Because it is so easy to make rhythmic poems in Chinese, poetry is used everywhere in China, as noted by Link (2013). One can find it in conversations, in speeches, in advertisements, in social media postings (such as WeChat), in slogans, in public notices, and in government documents. Everyone can make poems and everyone can appreciate them. In contrast, in English-speaking societies, apart from nursery rhymes and artistic performances (e.g. classic plays or rap music), it is rare to see or hear rhythmic language in public or private places, much less in government documents. In addition, even highly educated people may not be able to make or appreciate poetry. For example, Professor Rob Burling (personal communications), who has done research on nursery rhymes, said that he had no intuition for iambic pentameter and could not appreciate its rhythm at all.

6.  Concluding remarks Elastic words are those whose length can vary from monosyllabic to disyllabic, without a change in meaning, as shown in Table 26.1. Elastic words have been observed in Chinese for a long time, at least since Karlgren (1918). However, both theoretical and empirical studies have been lacking. Theoretically, it is not clear whether elastic words can be distinguished 466

Elastic words in Chinese

from regular compounds, nor has it been demonstrate what motivated the creation of elastic words. Empirically, it has not been shown how many elastic words there are in Modern Chinese, or how many there were in Classic Chinese. We have offered a clear definition of elastic words that is simple, objective, and feasible. Based on the definition, we have calculated the percentage of elastic words in modern Standard Chinese. Our result shows that about 50% of Chinese morphemes, mostly monosyllabic words, have elastic length. The percentage is lower than most previous estimates but much higher than that in English (or in any other language, to our knowledge). In addition, we have argued that the percentage of elastic words in Classic Chinese is just as high, contrary to the popular view that elastic words are the property of Modern Chinese only. Finally, we have compared two theories of why elastic words are created in Chinese, the homophony theory and the prosody theory. According to the former, elastic words are created to avoid ambiguity caused by homophones. In contrast, according to the prosody theory, elastic words are created to satisfy Foot Binarity, so that stressed syllables can form a binary foot with another syllable. We have shown that most predictions of the homophony theory are incorrect, whereas all predictions of the prosody theory are borne out.

Note 1 We would like to thank Lijun Huang, Mike Opper, and Xinting Zhang, who participated in the work reported here. For useful comments, we would also like to thank the hosts and audiences at the Chinese University of Hong Kong (2013), Ohio State University (2013), Lund University (2014), and University of Wisconsin Madison (2014), where parts of this work were presented.

References Bloomfield, Leonard (1933) Language, New York: Holt. Chao, Yuen-Ren (1948) Mandarin Primer, An Intensive Course in Spoken Chinese, Cambridge, MA: Harvard University Press. Chen, Matthew Y. (2000) ‘From Tone to Accent’, Cashiers de Linguistique – Asie Orientale 29(2): 179–203. Chomsky, Noam and Morris Halle (1968) The Sound Pattern of English, New York: Harper and Row. Dong, Yan (2015) The prosody and morphology of elastic word in Chinese: annotations and analyses, Doctoral Dissertation, University of Michigan. Duanmu, San (1990) ‘A Formal Study of Syllable, Tone, Stress and Domain in Chinese Languages’, Doctoral Dissertation, MIT, Cambridge, MA, Distributed by MIT Working Papers in Linguistics. Duanmu, San (2007) The Phonology of Standard Chinese, 2nd ed., Oxford: Oxford University Press. Duanmu, San (2012) ‘Word-length Preferences in Chinese: A Corpus Study’, Journal of East Asian Linguistics 21(1): 89–114. Duanmu, San (2013) ‘How Many Chinese Words Have Elastic Length?’ in Peng Gang and Shi Feng (eds) Eastward Flows the Great River: Festschrift in Honor of Prof. William S.-Y. Wang on his 80th Birthday, Hong Kong: City University of Hong Kong Press, 1–14. Duanmu, San 端木三 (2015) 汉语单语素名词的词长弹性研究：词典中和实际使用中的“只单不双” 名词 (‘A Study of Elastic Word Length in Monomorphemic Nouns in Chinese: “Monosyllabiconly” Nouns in the Lexicon and in Actual Use’), in Shengli Feng 冯胜利 (ed)《汉语韵律语法 新探》(New Explorations in Chinese Prosodic Grammar), Shanghai: Shanghai Zhongxi Shuju, 4–21. Duanmu, San and Yan Dong (2014) ‘Homophone Density and Word Length in Chinese’, in Hsiao Yuchau and Wee Lian-Hee (eds) Capturing Phonological Shades within and across Languages: Inspirations from the Theoretical Phonology Conferences, Cambridge Scholars Publishing, 213–42. Feng, Shengli (1995) ‘Prosodic Structure and Prosodically Constrained Syntax in Chinese’, Doctoral Dissertation, University of Pennsylvania, Philadelphia. Guo, Shaoyu 郭绍虞 (1938)〈中国语词之弹性作用〉(‘The Function of Elastic Word Length in Chinese’),《燕京学报》(Yen Ching Hsueh Pao) 24: 1–34.

467

San Duanmu and Yan Dong Gussenhoven, Carlos (1983) ‘Focus, Mode and the Nucleus’, Journal of Linguistics 19(2): 377–417. Huang, Lijun 黄丽君 and San Duanmu 端木三 (2013)〈现代汉语词长弹性的量化研究〉(‘A Quantitative Study of Elastic Word Length in Modern Chinese’),《语言科学》(Linguistic Sciences) 12(1): 8–16. Jespersen, Otto (1930) ‘Monosyllabism in English: Biennial Lecture on English Philology’, Proceedings of the British Academy 14: 341–68. Jin, Wen (2011) ‘A Statistical Argument for the Homophony Avoidance Approach to the Disyllabification of Chinese’, in Zhuo Jing-Schmidt (ed) Proceedings of the 23rd North American Conference on Chinese Linguistics (NACCL-23), vol. 1, University of Oregon, Eugene, 35–50. Karlgren, Bernhard (1918) Ordet och pennan i Mittens Rike, Stockholm: Svenska Andelsförlaget. Karlgren, Bernhard (1923) Sound and Symbol in Chinese, London: Oxford University Press. Ke, Hang (2007) ‘A Study of Monosyllabic and Disyllabic Usage in Modern Chinese’, Doctoral Dissertation, Institute of Linguistics, Chinese Academy of Social Sciences, Beijing. Ke, Jinyun (2006) ‘A Cross-linguistic Quantitative Study of Homophony’, Journal of Quantitative Linguistics 13(1): 129–59. Kennedy, George A. (1951) ‘The Monosyllabic Myth’, Journal of the American Oriental Society 71(3): 161–6. Kennedy, George A. (1955) The Butterfly Case, Wennti Papers No. 8 (March), New Haven: Far Eastern Publications, Yale University, 1–47. Li, Charles N. and Sandra A. Thompson (1981) Mandarin Chinese: A Functional Reference Grammar, Berkeley and Los Angeles: University of California Press. Li, Ningding 李临定 (1990)〈动词分类研究说略〉(‘A Brief Discussion on Verb Classification’),《中国 语文》(Studies of the Chinese Language) 4: 248–57. Link, E. Perry (2013) An Anatomy of Chinese: Rhythm, Metaphor, Politics, Cambridge, MA: Harvard University Press. Lü, Shuxiang 吕叔湘 (1963)〈现代汉语单双音节问题初探〉(‘A Preliminary Study of the Problem of Monosyllabism and Disyllabism in Modern Chinese’),《中国语文》(Studies of the Chinese Language) 1: 11–23. McEnery, Tony and Richard Xiao (2004) ‘The Lancaster Corpus of Mandarin Chinese: A Corpus for Monolingual and Contrastive Language Study’, in M. T. Lino, M. F. Xavier, F. Ferreire, R. Costa, and R. Silva (eds) Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC) 2004, Lisbon, 24–30 May 2004, 1,175–8. Nettle, Daniel (1995) ‘Segmental Inventory Size, Word Length, and Communication Efficiency’, Linguistics 33: 359–67. Pan, Wenguo 潘文国 (1997)《汉英语对比纲要》(An Outline Comparison of Chinese and English), Beijing: Beijing Language and Culture University Press. Prince, Alan (1980) ‘A Metrical Theory for Estonian Quantity’, Linguistic Inquiry 11: 511–62. Shih, Chi-lin (1986) ‘The Prosodic Domain of Tone Sandhi in Chinese’, Doctoral Dissertation, University of California, San Diego. Truckenbrodt, Hubert (1995) ‘Phonological Phrases – Their Relation to Syntax, Focus, and Prominance’, Doctoral Dissertation, MIT, Cambridge, MA. T’sou, Benjamin K. (1976) ‘Homophony and Internal Change in Chinese’, Computational Analysis of Asian and African Languages 3: 67–86. Wang, Canlong 王灿龙 (2002)〈句法组合中单双音节选择的认知解释〉(‘A Cognitive Explanation of the Choice between Monosyllabic and Disyllabic Words in Syntax’), in Zhongguo Yuwen Zazhi She 中国语文杂志社 (ed)《语法研究和探索》(Studies and Explorations on Grammar), 11th ed., Beijing: The Commercial Press, 151–68. Wu, Weishan 吴为善 (2003)〈双音化、语法化和韵律词的再分析〉(‘A Reanalysis of Disyllabification, Grammaticalization, and Prosodic Words’),《汉语学习》(Hanyu Xuexi) 2: 8–14. Xiandai Hanyu Cidian《现代汉语词典》(Modern Chinese Dictionary) (2005) Compiled by the Institute of Linguistics, Chinese Academy of Social Sciences, Beijing: The Commercal Press.

468

27 Hakka dialect Lau Chun Fat Xiamen University, China

1.  Definition of the Hakka dialect It seems simple to say that the Hakka dialect is the dialect spoken by the Hakka people, identified to have come to Southern China through five waves of migration over the last 1,500 years (Luo 1933). Since then Hakka has been recognized as a Modern Chinese dialect group (Yuan et al. 1989; LAC 1989), although there is still ongoing scholarly debate about whether it should be grouped together with the Gan dialects (Wang 1998). However, recent studies suggest that the name ‘Hakka’ arose as a product of the dispute between the Cantonese speaking natives and the Hakka speaking intruders in Western Guangdong around 1800 (Constable 1996), and the Hakka identity has been constructed since then (Constable 1996; Leong 1997). Therefore Hakka as a self-labeling identity (for both the language and the people) is only found in those areas which are better informed (where people have been told explicitly that they are ‘Hakka’ in the past three decades), and it is found to be a still ongoing recent event (Z. Liu 2001). Many people speaking Hakka (or languages intelligible to the Meixian dialect) were still denying that they are Hakka people as late as in the 1980s, and named their dialect otherwise. This means that the legend that Hakka as an ethnic subgroup formed around a millennium ago may be some kind of exaggeration, and this also explains why Hakka as a dialect is not as clear-cut as accepted. As the Hakka dialect itself may have been recently formed, there remains controversy over the criteria for identification of a Hakka dialect group, or whether it should be grouped together with Gan (Norman 1986; Sagart 1998). Luo’s definition of Hakka (1933) based on his understanding that a group of people from the ‘Central Plain’ began their migration as early as the end of the Jin dynasty (ca. ad 300–400). Then they carried on their movement in two waves, and arrived in the southern provinces such as Guangdong and Jiangxi by the Song dynasty (ca. ad 1000). When they reached the south, they found that the land was already occupied and they were named Kehu (客户 ‘guest families’). They have been named Hakka ever since then. However, recent studies have found a lot of flaws in this theory. First, the so-called waves of migration were shared by almost all southern ethnic Chinese, who also claim that their ancestors came from the Central Plain. Interestingly, southern Chinese with the same surname can be traced back to a common ancestor about a millennium ago, no matter what dialect they are speaking. For example, most Southern Min, Chaozhou, Hakka, and even some 469

Lau Chun Fat

Cantonese speakers with the surname Li (李) all admit that their common ancestor was Li Huode (李火德). Second, the regions with the most Kehu did not evolve into Hakkaspeaking areas, such as Duanzhou (端州, today’s Zhaoqing 肇庆) and Guangzhou (both with more than 50% Kehu), and yet the major Hakka-speaking areas contained as little as 6% Kehu, for example in parts of northern Guangdong. Third, many people who speak Hakka did not know that their dialect is named Hakka until the 1980s when China opened its doors. This contradicts the assumption that they were aware they are true ‘Chinese’ from the north and are proud to tell others that they are Hakka people (Z. Liu 2001). In a narrow sense, Hakka as a dialect is a tongue represented by the Meixian dialect (梅县话), which can be understood by a majority of Hakka speakers in Guangdong, Guangxi, Taiwan, Fujian, Hainan, Jiangxi, Hunan, and Sichuan, which I have grouped under the name Jiaying subdialect (JYS, see section 2.2). Some other Hakka varieties in Jiangxi and Fujian are unintelligible to most JYS speakers, but are classified in the Language Atlas of China (LAC 1989). Interestingly, JYS can even be readily understood in most Gan dialect areas, raising the question of a possible combination of Hakka–Gan as a single dialect group. The LAC confines the Hakka dialect mainly to the eastern part of Guangdong, the southern part of Jiangxi, and the western part of Fujian as shown in Figure 27.1. There are also some scattered Hakka settlements in Guangxi, Taiwan, Hainan, Hunan and Sichuan. The total population of Hakka speakers is estimated to be about 40 million, with about half of

Jiangsi Subdialect

JYS

W Fujian Subdialect

Jiaying Central GD Subdialect Subdialect

Figure 27.1  Distribution and classification of Hakka

470

Hakka dialect

them living in Guangdong. (The figure varies a lot according to different sources, e.g. http:// www.encyclopedia.com/topic/Hakka.aspx, as younger generations in Hong Kong, Taiwan, and some parts of Guangdong are not capable speakers of the dialect but are counted as being Hakka.) The majority of Hakka dialects belong to the Jiaying subdialect, or Yue-Tai subgroup in the LAC, bearing the names ‘Kejiahua’ 客家话, ‘Ai-hua’ 艾话, ‘Majie-hua’ 麻介话 or ‘Xinmin-hua’ 新民话 in Guangdong and Guangxi, ‘Keji-hua’ 客籍话, ‘Guangdonghua’ 广东话 or ‘Huaiyuanhua’ 怀远话 in Jiangxi, ‘Guangdonghua’ 广东话 or ‘Tu-Guangdonghua’ 土广东话 in Sichuan, etc. All dialects belonging to this subgroup can readily understand the Meixian speech.

2.  Characteristics and subgrouping of the Hakka dialect group 2.1. Characteristics Despite the shortcomings of its definition, Hakka is now recognized as a Chinese dialect group and has some unique characteristics of its own. However, it also shares many characteristics with the Mandarin, Yue, Min, and Gan dialect groups. Both Hakka and Gan dialects share the feature that the Middle Chinese (MC) voiced onsets are pronounced with aspiration in all tones, whereas in Mandarin and Cantonese, these morpho-syllables are not aspirated in the departing and entering tones (Yuan et al. 1989). That means, if a character has a voiced onset and its MC tone is rising/departing (e.g. 断近柱/ 步旧凈) or entering (e.g. 夺局杂), the character is unaspirated in Mandarin, but always aspirated in Hakka or Gan. Morpho-syllables of the level tone are aspirated in Hakka, Gan, Yue, and Mandarin (e.g. 平唐秦). The following are common features of the Hakka dialect:

2.1.1. Onsets 1. Most Hakka varieties of the JYS in Guangdong, Guangxi, Sichuan, and Taiwan have only one set of sibilants. However, most varieties in Fujian and Jiangxi have two sets. The two sets are usually dental and palatal, or dental and postalveolar series. Some in Jiangxi even have all three: dental, palatal, and postalveolar. 2. The occurrence of the labiodental onset [f ] is more frequent than in other dialects. The Hakka [f ] corresponds roughly to both [f ] and [hu-] of Putonghua, and also some [ku-]. 3. The bilabials [b], [p], [m] are more or less the same as Putonghua and Cantonese. Some characters (about 20–35, according to location) of the MC ‘Fei’ series are pronounced as [p] and [ph], which are considered to be the remnants of Ancient Chinese pronunciations (Norman 1986: 328). 4. The onset [ŋ], on the other hand, is palatalized to [ ɲ] and is regarded as an allophone of [ŋ] since they do not form contrasts. 5. In most JYS, [v] occurs as the only voiced onset in Hakka, which roughly corresponds to [w] of Cantonese. This onset is not a retention of the MC voiced onset, but an innovation as a result of fricativization of [u] when the onset is zero. 6. In some areas, [z] and [Ʒ] may also occur as an allophone of [ j]. 7. In western Guangdong, the lateral fricative [ɬ] may be found in place of [s] in characters originating from the MC Xin Initial (心母), which is believed to be a result of language contact. 471

Lau Chun Fat

2.1.2. Rimes 1. In most varieties of Hakka there is no rounded front vowel [y]. However, [y] does occur in varieties of Sichuan, Jiangxi, Guangxi, western and northern Guangdong, and in scattered locations of southwestern Fujian. That means, no less than one-third of the Hakka areas are speaking Hakka dialect with [y]. Some earlier scholars believed that this was a result of language contact, but the occurrence of [ y] in some mountainous regions of Fujian surrounded by the Minnan dialect (which also has no [y]) shows that the phenomenon is not so easily explained. We need more historical linguistic studies to settle this question. 2. The medial [i] is usually present in most varieties. The medial [u] is absent in many varieties, but still preserved in northeastern Guangdong, including the Meixian dialect. [u] as medial can occur only after [k] and [kh] as in Cantonese. [y] as medial occur only in rare cases, the only known example is the Gutian dialect of Fujian. 3. The codas [p], [t], [k], [m] are well preserved in most varieties in Guangdong, Guangxi, and Taiwan, but absence of [m], [p] is common in northern Guangdong. In Fujian, Jiangxi, and Sichuan, most varieties have only [t] or glottal stop, or have even lost these codas totally. 4. The occurrence of [a] and [i] is more or less the same as in Putonghua, but there are more rimes with [ɛ], [u], and [ɔ] as main vowels. In the northeast of Guangdong, [ə] can act as a main vowel, but its occurrence is limited to follow the sibilant onsets only, roughly the characters pronounced zhen, chen, shen, zheng, cheng, sheng, and (some) zhe, che, she in Putonghua. As the front vowels [i], [ɛ], [ə] cannot have [ŋ], [k] as codas, correspondence of the codas with Middle Chinese is not as regular as in Cantonese. 5. The numbers of rimes vary a lot, from 40 to 75, depending on the presence of codas [p], [t], [k], [m], and the number of medials. The Meixian dialect has the largest number of rimes among all varieties of Hakka.

2.1.3. Tones With respect to tones, varieties of Hakka can differ in the tone value (tone realization) or the tonal grouping of Chinese characters, for example, whether the Yinqu or Yangqu are independent or merged. Most Hakka dialects belonging to the JYS consist of six tones: Yinping, Yangping, Shang, Qu, Yinru, and Yangru. That means, the Shang and Qu tones are not divided into Ying and Yang as in Ping and Ru. For most locations in the middle of Guangdong, western Fujian, and some of southern Jiangxi, however, both the Yinqu and Yangqu are independent tones. In Huizhou and Lufeng, for example, the dialects consist of seven tones, which is the maximum number. In Ningdu of western Jiangxi the entering tones are intact, and there are seven tones altogether. For some in western Fujian such as Changting, the entering tones have disappeared, resulting in five tones. Some localities even have only four tones when the departing tone is not split, e.g. Shexi (L. Liu 1999). Therefore, the number of tones can vary from four to seven, but the usual number is six. The tone values also vary a lot, but for most JYS dialects, the six tones are usually 44, 11, 31, 53, 32, and 53. In Wuhua, Shaoguan (and most dialects around it), and Sichuan, the Yangping is usually 35 instead of 11. The tones follow a totally different pattern in the middle of Guangdong and Lufeng, and also in Jiangxi or western Fujian, making them quite difficult for JYS speakers to understand. 472

Hakka dialect Table 27.1  Monosyllabic Hakka words which can be traced back to Middle or even Ancient Chinese Gloss

Putonghua

Hakka (Meixian)

sugar cane paddy couplet trousers house know happy smart

gan zhe 甘蔗 shui dao 水稻 dui lian 对联 ku zi 裤子 fang zi 房子 zhi dao 知道 gao xing 高兴 ji ling 机灵

tsa53蔗 vo11 禾 tui53 对 fu53裤 vuk32屋 ti44知 hin31兴 tsiaŋ44精

2.1.4.  Lexicon and grammar The grammar and lexicon are to a large extent shared by Yue and Gan, and to a lesser extent by the Southern Min. Major features are summarized as follows:   1. There are more monosyllabic words which can be traced back to Middle or even Ancient Chinese. A list of glosses are listed in Table 27.1. Words like /vuk/ (屋, house), /haŋ/ (行, walk), /zɛu/ (走, run), /sət/ (食, eat), /tsɔk/ (着, put on clothes), /lɔk/ (落, fall), /kuɔŋ/ (光, bright), /sɛ/ (细, small), /vu/ (乌, black) are retained in Hakka as their original meaning in Middle Chinese, and most are also shared by neighbouring dialects like Yue and Min.   2. The reversal of morphemes among some words compared with Putonghua, e.g. 紧要 (MX: kin-jau, important; PTH: 要紧 yaojin), 人客 (MX: ɲin-hak, guest; PTH: 客人 keren), 欢喜 (MX: fɔn-hi, like or love; PTH: 喜欢 xihuan), 闹热 (MX: nau-ɲiat, crowded with people, PTH: 热闹 renao). Some modifiers are suffixes rather than prefixes, e.g. the suffixes for the sex of animals /kuŋ/ (公, male animal) and /ma/ (嫲, female animal). e.g. rooster and hen are gong ji (公鸡) and mu ji (母鸡) in PTH, but /kɛ kuŋ/ and /kɛ ma/ in Meixian. This kind of word order is also commonly shared by Yue, Min, and some Gan dialects.   3. The suffixes /kuŋ/, /ma/ can also be the suffixes for body parts and inanimate objects, e.g. /sɛtma/ (舌嫲) stands for tongue, where the suffix /ma/ is merely a meaningless bound morpheme that equals to zi (子) in Putonghua. Likewise, nose is /phi kuŋ/ (鼻公) and ear /ɲi kuŋ/ (耳公), where /kuŋ/ is a meaningless male suffix. There is also a diminutive suffix /tsai/ (仔) in varieties in and around Hong Kong, especially after names of fruits and common objects, but they do not mean ‘small’, e.g. /tshaŋ tsai/ (orange), /phat tsai/ (guava), /sɔn tsai/ (garlic). In Meixian this suffix /tsai/ is reduced to /ɛ/, which can be assimilated to the coda of the stem syllable, e.g. /tsaŋ ŋɛ/ (orange), /gam mɛ/ (mandarin), /gɛu vɛ/ (dog). This resembles the phenomenon of ‘grammatical gender’ of many European languages such as Latin, German, and Russian, but the difference is that only a very small part of the vocabulary is ‘gendered’ in Hakka, and the numbers of glosses also vary a lot in different localities.   4. There is semantic broadening and narrowing in some Hakka words, e.g. re (热, in PTH refers to liquids, solids, and weather), but in Hakka hot for solids is /nat32/, for liquid is usually /sau44/ and for weather is /ɲiat53/ (热, cognate with PTH re); sha 杀, to kill, is valid for both humans and animals in PTH, but most Hakka varieties use different 473

Lau Chun Fat

  5.

  6.   7.   8.   9. 10.

words for them, e.g. Meixian uses /sat/ for humans and /tshi/ (or /tsHɿ/) for animals. Other varieties also use /lɔt/ or /thɔŋ/ for animals. On the other hand, words differentiated in PTH are not differentiated in Hakka, e.g. fei 肥 (fat) refers only to animals in PTH, and humans should be described as pang, but Hakka uses /phui/ (肥) for both human and animals. /ŋai/, /ɲi/, and /ki/ are used for the first-, second- and third-person personal pronouns respectively. There is a special possessive pronominal form existing in many varieties of Hakka by changing the main vowel to [a] or adding an [a] to the personal pronouns, i.e. /ŋa/, /ɲia/, and /kia/. In some varieties, the first person plural pronoun is divided into inclusive and exclusive forms. e.g. in Dabu county or Taiwan, /ɛn tɛu/ or /ɛn/ is inclusive for the first-person plural, while /ŋai tɛu/ is exclusive. The quantifier may be different from that of Cantonese and Putonghua, e.g. /jit tsak ɲin/ (一只人, a person), /jit thiau kɔ/ (一条歌, a song) in Meixian. The suffix for continuous action, equivalent to zhe (着) in Putonghua, is /tɛn/ or /kin/. The suffix for completed action, equivalent to le (了) in Putonghua, varies from location to location. It is /phεt/ or /ε/ in Meixian, but /thεt/ in Xingning county and Taiwan, and /hɔi/ in Bao’an and Huiyang. The adverb suffix comparable to Putonghua de 的 is /ɛ/ in Meixian, but for other varieties it can be /tshi/, /li/, etc. The adverbial use of /tɔ/, /sau/ (多/少, more/less) is placed after the verb as in Cantonese, not in front as in Putonghua. e.g.

着多件衫 tsɔk tɔ khiɛn sam put on more CL cloth Put on one more piece of clothing.

11. The passive auxiliary is /pun/ (分) (/pin/ or /pi/ in some varieties), and like Cantonese pei 被 it must involve an actor. 12. The negation is /m/ in most Hakka dialects, but the tone may be different in different locations. Meixian and Taiwan have a Yangping tone (11), which sounds exactly the same as in Cantonese. Other varieties around Hong Kong have the Yinping tone (44 or 13). The meaning is the same as in PTH bu (不). e.g. ‘I am not going’ is /ŋai m hi/, comparable to PTH wo bu qu. However, sometimes the word order is changed if the negated phrase concerns ability or possibility, e.g. an adverbial phrase with /tet/ (得): e.g. ‘reluctant (to give)’ is /m sa tɛt/ in Hakka but she bu de (舍不得) in PTH; ‘No wonder’ is /m kuai tɛt/ in Hakka but guai bu de (怪不得) in PTH. The structure /m V tɛt/ is very common in Hakka (similar structure also occurs in Cantonese) but ‘V bu de’ is not so frequently used in PTH. Likewise, ‘I cannot beat him’ is /ta ki m jaŋ/ in Hakka and da bu ying ta in PTH. 13. The comparative forms are quite interesting. There are several ways of expressing them. (a)

我大过佢 ŋai thai kuɔ ki I big pass he I am older than him

(b)

我比佢(过)大 ŋai pi ki (kuɔ) thai I than he (pass) big I am older than him 474

Hakka dialect

(c)

我过大 ŋai kuɔ thai I pass big I am older than him



The first form, (a), is similar to Cantonese, and the second form (b) is similar to Mandarin; but the last one (c) is special to Hakka. 14. Sentences involving double objects are similar to Cantonese for short and simple sentences, i.e. 我分间屋佢 ki ŋai pun kan vuk I give CL house him I give him a house In PTH, the word order would be ‘I give him a house’ instead. 15. Questions can be formed by adding /maŋ/ at the end of the confirmative for past events, which is equivalent to the PTH mei; e.g. in Hakka ‘Did you eat?’ is /ɲi sət fan maŋ/; ‘Did you go to Beijing?’ is /ɲi hi həi pɛt kin maŋ/. In Hong Kong and nearby area, /m tshɛn/ can also be used. For present and future tense, e.g. ‘Is he coming?’, the negation /m/ is added and the verb is repeated, e.g. /ki lɔi m lɔi/? One can also add /mɔ/ at the end, /ki lɔi mɔ/? A similar structure is also found in PTH and Cantonese.

2.2. Classification In the LAC, the Hakka dialect is roughly divided into eight major subgroups according to their geographic distribution: (i) Guangdong and Taiwan (粤台 Yue-Tai), (ii) Central Guangdong (粤中 Yuezhong), (iii) Huizhou (惠州), (iv) Northern Guangdong (粤北 Yuebei), (v) Western Fujian (闽西 Min-Xi), (vi) Southern Jiangxi (宁龙 Ninglong), (vii) Western Jiangxi and Eastern Hunan (于桂 Yugui), (viii) Northern Jiangxi (铜鼓 Tonggu). After careful consideration of mutual intelligibility and self-identity of the speakers, the Hakka dialect group is more appropriately classified into four major subdialect groups:

2.2.1.  The Jiaying subdialect The JYS is the most representative subgroup of the Hakka dialect group, and it is also the most abundant in distribution. It includes the Yue-Tai, Yuebei, and southern part of Min-Xi in the LAC and most of the ‘outlying’ Hakka dialects in Guangxi, Hainan, Sichuan, and overseas. The uniformity of the JYS is the result of a series of migration waves. The first started from the southwestern part of Fujian to northern Guangdong about 500 years ago, the second from the Jiaying prefecture around 300 years ago to southern Guangdong, northern Guangxi, southern and western Jiangxi, and Sichuan, and this is why it is so named. About a century later (around 1800) the third wave arose from and around Meixian (today renamed as Meizhou City) to Taiwan, the Pearl River Delta, southern Guangxi, Hainan, and western Guangdong, A fourth wave about 120 years ago sent these speakers to Malaysia, Indonesia, and around the Caribbean in Central and South America. This explains why the Jiaying subdialect has a wide geographical distribution and yet a good uniformity in its phonology and vocabulary. 475

Lau Chun Fat

Within the JYS subdialect, there are two major categories (patois) according to the behavior of tones: (i) The Meixian patoi merges both Yinqu and Yangqu, so that all characters belonging to MC departing tone are pronounced with the same tone; (ii) the Wuhua patoi merges Yangqu with the Shang tone instead, so that voiced characters of MC departing tones have the Shang tone, not Qu. Besides, the Meixian group has a Yingping tone value of 11, but Wuhua has the value of 35 or 24. Most varieties of JYS belong to the Meixian patoi, but those in northern Guangdong and Sichuan and some dialects in western Guangdong belong to the Wuhua patoi. The JYS in northeastern Guangdong, western Guangdong, the Pearl River Delta near Hong Kong, southern Guangxi, Taiwan, and most JYS overseas keep the codas [m], [p], [t], [k] intact. In southwestern Fujian and northern Guangdong, [m/p] merge with [n/t]. In Sichuan, all the three stop codas [p], [t], [k] merge to a glottal stop. Despite these sound changes, all JYS are easily understood by each other.

2.2.2.  The Bendi subdialect The Bendi subdialect group consists of dialects in the eastern middle part of Guangdong, including the Yuezhong Huizhou subgroups in the LAC, and the Hailu dialect of Taiwan. Their onset and rime characteristics resemble closely that of Meixian, but the tones distinguish them clearly from JYS. The tone values are usually different from that of Meixian, and in most dialects, the characters of MC voiced onsets go to the Shang tone, not the Yinping tone. The vocabulary also differs a lot from JYS. Interestingly, their speakers, especially those of the older generation, reject the notion that they have any connection with Hakka (S. Liu 2007). They stress that they are Bendi (本地, local), and named their tongue Bendihua (本地话, the local speech). The people in Huizhou even express animosity towards JYS speakers, stressing that they are intruders who took away their lands. However, because of the ready intelligibility of their dialect to JYS, most scholars would put them into the category of Hakka, although some suggest that the resemblance is due to long-time language contacts. The Bendi subdialect speakers form an identity because of similarities in their speech. Many of their dialect have the [y] rime and a characteristic tone pattern. They have also some special glosses which are distinct from JYS, of which some are shared by Cantonese. Some scholars just attribute these to language contacts, but we can see that before 1980, these speakers especially those in Longchuan county, have virtually no access to Cantonese. We would rather explain them as cognates rather than loan words and say these dialects actually share the same ancestry as Cantonese and their similarity with JYS is the result of contact over centuries.

2.2.3.  The Jiangxi subdialect This group includes all three of the LAC subgroups of Hakka in Jiangxi (Ninglong, Yugui, and Tonggu). Most Hakka dialect speakers in Jiangxi are more or less living interspersed with Gan dialect speakers. In southern Jiangxi, there are also two major subcategories: the Bendi (本地, locals) patoi and Keji (客籍, immigrants, synonym of Hakka) patoi, who have similar sentiments as in Guangdong. The Bendi speakers claim themselves indigenous to the land, and Keji are usually descendants of immigrants from Guangdong about 350 years ago. The speech of the latter sounds closer to JYS. Their speech differs from the former mainly in the tones, but they understand each other, and these so-called ‘Hakka’ dialects have already 476

Hakka dialect

mixed with the speeches around them, so that a JYS speaker from Guangdong may have communication problems mainly because of vocabulary. The Tonggu variety, however, is almost readily intelligible to JYS despite some differences in vocabulary.

2.2.4.  The western Fujian subdialects In this classification, the southern part of the LAC Minxi subgroup still belongs to JYS because of mutual intelligibility with Meixian, but the rest go to this subgrouping as their speeches are not readily understood and some of them do not have the feeling of Hakka. The Hakka dialects in western Fujian are more complicated than the above three subgroups. Some are more readily understood by speakers in Guangdong, such as those from the border areas of Wuping and Yongding, and in scattered locations in Liancheng. However, many Hakka dialects of western Fujian are totally unintelligible for most JYS speakers, being even more difficult to understand than the dialect of Nanchang, the representative Gan dialect. They also have communication problems among themselves, e.g. someone from Liancheng will have difficulty understanding the dialect of Changting or Ninghua. This also raises the problem of dialect definition and dialect boundary, which is out of the scope of our discussion here.

3.  Description of the Jiaying subdialect (JYS) The Meixian dialect is regarded as an unofficial standard of the JYS, although the Sixian (四县) dialect of Taiwan is gaining more influence because of its official status there. Both of them differ little in their phonology, including the tones, but the Sixian dialect contains more vocabulary of southern Minnan origin. On the other hand, the Bao’an–Huiyang variety, which also includes the speech of Hong Kong Hakka, has some significance not only because speakers of Hakka in Malaysia, Central America and Europe are using this tongue, but also because it has been well documented by the Basler Mission since around 1860. The Bao’an– Huiyang variety has a simplified phonology and contains more glosses of Cantonese and other origins, but the tone features are the same for all three varieties.

3.1.  Phonology of Meixianhua The phonological system of the Meixianhua is summarized as follows (Yuan et al. 1989; Huang 1992).

3.1.1. Onsets 1. On the whole, the onset system of Meixian is very similar to that of Cantonese. The pronunciations of the stop onsets [p, t, k, ph, th, kh] and sibilants [ts], [tsh], [s] are the same as Cantonese. The only differences are the presence of the consonants [v] and nasal [ ɲ] but absence of [w]. [ ɲ] is actually an allophone of [ŋ] before the high front vowels as a result of palatalization. 2. The Meixian [v] is a voiced fricative, the only voiced initial. It corresponds to Cantonese [w] and replaces it in most cases. 3. [h] is glottal and pronounced the same way as in Cantonese (and also English), not velar as in Putonghua. 4. [ j] is used when the rimes beginning with [i] have zero onsets, which is similar to the case of Putonghua and Cantonese. 477

Lau Chun Fat Table 27.2  Onsets (18) p t k ts

ph th kh tsh

m n ŋ ɲ

f

v

h s

j

l

Table 27.3  Rimes (75) a ia ua i ɿ u ε iε uε ɔ uɔ iɔ

ai iai uai

au iau

am iam

iu

im əm

ui εu

εm

ɔi m

an ian uan in ən un iun εn iεn uεn ɔn uɔn iɔn n

aŋ iaŋ uaŋ uŋ iuŋ

ɔŋ uɔŋ iɔŋ ŋ

ap iap ip əp εp

at iat uat it ət ut iut εt iεt uεt ɔt iɔt

ak iak uak uk iuk

ɔk uɔk iɔk

3.1.2. Rimes 1. There are altogether seven vowels, but [ɿ] and [ə] are actually allophones because they are not contrastive and are complementary in distribution. Both of them can only occur together with the onsets [ts], [tsh] and [s]. 2. [a] is a low open vowel. The pronunciation of [a] can be affected by the onsets in front and the codas behind as in Putonghua. 3. [ε] is front and half open, and as a suffix it can take over the coda of the syllable before it as its onset, e.g. /kεu vε/ (dog), /phan nε/ (disk), etc. 4. [i] is almost always a standard high front vowel, but can be lowered a bit when combined with labial onsets. It is changed to [ j] as coda. 5. [ɔ] is a mid-back rounded vowel. It can combine with any of the onsets, as in Cantonese and most varieties of Gan. 6. [u] is a high back rounded vowel and a bit lower than the standard. 7. [m], [n], [ŋ] are independent syllables. [ŋ] can be assimilated by the following onsets to [n] if they are alveolars. 8. The complete set of Middle Chinese codas [m], [n], [ŋ], [p], [t], [k] is kept in Meixian, but as [i], [ε], and [ə] cannot combine with coda [ŋ], characters from Middle Chinese Geng (梗), Zeng (曾) group of rimes are usually pronounced [in]/[it], [εn]/[εt] or [ən]/ [ət], instead of [iŋ]/[ik] or [εŋ]/[εk] as in Cantonese or Southern Min.

3.1.3. Tones Tone sandhi occurs with the Yinping and the Qu tones. When the following syllable is of lower tone than the Yinping (i.e. Yangping, Shang or Yinru), then the tone is changed from 44 to 45. In the case of Qu, the tone is changed from 52 to 54 or 55. 478

Hakka dialect Table 27.4  Tones

Tone value Example

Yinping

Yangping

Shang

Qu

Yinru

Yangru

44 参芬声

11 蚕焚成

31 惨粉省

52 杉份盛

2 插忽析

5 杂佛石

3.2.  Regional varieties of the Jiaying subdialect As mentioned above, the Jiaying subdialect consists of hundreds of varieties, even inside Guangdong alone. Although they are mutually intelligible with each other, they can differ quite a lot in the onsets, rimes, tones, and lexicon (but no syntactic differences). For example, there is only one set of sibilants [ts], [tsh], [s] in MX, but the counties around Meixian such as Dabu, Xingning, and Wuhua have both [ts], [tsh], [s] and [tʃ ], [tʃ h], [ ʃ ] (Yuan et al., 1989). In and around Hong Kong, the Hakka speeches are quite similar to that of Meixian, but still there are some obvious differences. In Taiwan, the Sixian dialect is the dominant speech, which sounds very similar to the Meixian dialect.

3.2.1.  The Bao’an–Huiyang dialects Bao’an–Huiyang dialects are spoken in Huiyang, Dongguan, Boluo, and the rural areas of Hong Kong, Shenzhen, Huizhou, and Heyuan, adding together to around 4 million speakers. They are also the speeches used by Hakka speakers in Malaysia, Central America, South America, and Europe, which have the following characteristics: 1. no [n] as initial (which merged into l), 2. no apical vowels, 3. no medial [u], because of these two changes in the rime, the number of rimes is radically reduced to 53, 4. the tone for the negation /m/ is Yangping 11 instead of Yinping.

3.2.2.  The Sixian dialects Sixian dialects, literally the dialects from four counties (Meixian, Pingyuan, Jiaoling, Dabu), are the dominant speeches of Hakka speakers in Taiwan; mainly in Miaoli, Xinzhu, Taoyuan, Pingdong, and scatter patches in Hualian and Taidong. The four counties named above are believed to be the homelands from which their ancestors came around 200 years ago, although immigration never ended until Japanese occupation in 1895. Their dialects closely resemble that of Meixian, with one set of sibilants, apical vowels, and medial [u]. However, the medial [u] does not occur together with [ɔ], so that the rimes /uɔ/ (过), /uɔn/ (官), /uɔŋ/ (光), /uɔk/ (郭) are not spoken with [u] in Taiwan. The second point of interest is that there is a significant difference in the vocabulary, as Sixian contains a large amount of glosses from the Southern Min dialect, such as /phai se/ for ‘sorry’, /tshin tshai/ for ‘as you like’, which are totally unheard by Hakka speakers in Meixian. Sixian also contains some glosses from the Hailu dialect, which shares a lot of glosses with the Hong Kong dialect, such as /lɔk sui/ for ‘raining’, and /thi theu/ for ‘peanut’. 479

Lau Chun Fat Table 27.5  Some lexical differences between Meixian and Hong Kong Hakka Gloss

Hen egg

Peanut

Give birth (to a baby)

Completion aspect

Meixian Hong Kong

kε44lɔn31鸡卵 kai13tshun44鸡春

fan44tεu53 番豆 thi53tεu53地豆

kiuŋ53 jɔŋ13

ε31 hɔi13

3.2.3.  The Xingning dialect Xingning lies only about 80km west of Meixian, but the phonology is significantly different. It has no [ŋ] initial (which merged into [n]). MX has [ɿ] only, but Dabu and Xingning have both [ɿ] and [ʅ]. In Xingning and most varieties of northern Guangdong (Li and Zhang 1992), the bilabials [m], [p] are not used as codas. Other minor differences are the conflation of [ɛu] with [iu] in the rimes, and the medial [u] is always omitted. There are some lexical differences between subdialects. Take the example of Meixian and Hong Kong Hakka in Table 27.5.

3.3.  Non-Jiaying dialects Non-Jianying dialects refer to the three dialects mentioned under sections 2.2.2–2.2.4.

3.3.1.  The ‘Bendihua’ subdialect in Guangdong The Bendihua is best represented by the Huizhou dialect, but other major dialects include the speeches of locations such as Heyuan, Longchuan, Heping, Zijin, Haifeng, Huidong, etc. Geographically, they form a 50–100km-wide strip from the north to the south along the banks of the Dongjiang River and its tributaries. They are named the Central Guangdong subgroup and Huizhou subgroup in the linguistic map of China (LAS), but their common features show that they are indeed a unified group. As mentioned before, the onsets and rime of Huizhou are similar to that of Meixian, except for the presence of [y], [iεm], [yεn], [əŋ] in place of Meixian [i], [iam], [iεn], [uŋ]. Other minor differences in the rime structure include the absence of [u] as medial, and the presence of more [i] medials. However, the tones of the Bendihua group sound a lot different from Meixian. As shown in Table 27.6, Huizhou has the first two tones similar to that of Meixian, but the following five tones are totally different. For Hailu and Shuiyuan, they are all different from Meixian. The tonal features make Bendihua distinctive. Besides, the Bendihua dialect uses [ŋɔi] for its first-person pronoun and [thε] for ‘see’. As a native speaker of the Huizhou dialect, Liu Shuxin (2007) opposes the notion that his tongue is Hakka, and argues that the Bendihua group should not be members of Hakka, Table 27.6  Tones of the Bendihua group

Huizhou Hailu Shuiyuan, Xinfeng

Yinping

Yangping

Shang

Yinqu

Yangqu

Yinru

Yangru

33 53 22

11 55 45

35 13 13

13 31 31

21 11 (31)

5 5 5

2 2 2

480

Hakka dialect

but a branch of the Yue which he names the Dongjiang group. However, his view gains little attention as the phonological structure of these ‘indigenous’ dialects closely resembles that of JYS, and poses no barrier to oral communication.

3.3.2.  The Jiangxi subdialects There is no clear demarcation between Hakka and Gan dialects in Jiangxi as in many locations the two dialects are mutually intelligible, although scholars suggest that Hakka uses [ŋai] for its first-person pronoun, and the Gan dialects use [ŋɔ]. However, the similarities in the phonological structure of the two dialects always arouse debate of whether Hakka and Gan should unite to one dialect instead of two. Today, the Hakka dialects in Jiangxi are situated in southern Jiangxi, geographically linked to the Hakka dialects of Guangdong and Fujian, plus an ‘island’ in the Tonggu area of northwestern Jiangxi, where Hakka immigrants came in the name of Pengmin around the beginning of the eighteenth century (Leong 1997). They add up to about 30% of the total population of Hakka. Southern Jiangxi actually consists of two groups of people and hence also two groups of speeches, the Bendi and Keji, as in eastern Guangdong, but they are all named under the label ‘Hakka’ by linguists. The Keji speakers came from northeastern Guangdong around the beginning of the Qing dynasty and after centuries of contact, their speech is based on JYS with a strong influence from neighboring dialects, which hinders their communication with Guangdong JYS speakers. The Bendi subgroup, however, after long-term contact with the immigrants of Guangdong, have adjusted their speech to the extent that they are readily intelligible to their Keji neighbours. That means both the Bendi and Keji dialects are converging to a common speech and in the course of formation of a Creole, and hence their speeches are now widely recognized as members of Hakka (L. Liu 1999). Tones of Keji speeches are more or less that of JYS, but the tones of Bendi may vary, and roughly resemble that of the Bendihua counterparts in Guangdong. There are some special characteristics of the region, such as some characters with [l] as initials are changed to [t] in the other groups, occurrence of nasalized rimes are common among the Bendi speeches, and loss of coda in the entering tones are common in both Bendi and Keji. JYS speakers may find the Keji speeches easier to understand, but as they form a continuous spectrum of almost perfect JYS to relatively ‘pure’ Bendi, this area is actually a good example of language contact.

3.3.3.  The western Fujian subdialects The western Fujian subgroup forms the most complicated picture of Hakka. The number of speakers makes up roughly 12%, which is the least among the four major ‘homeland’ provinces, but the number of varieties exceeds the other subgroups. Unfortunately, the speeches of this area are not yet well documented and a clear understanding of the picture is not yet possible, but we know that the speeches of Changting, Ninghua, Liancheng and the Gutian township of Shanghang are all unintelligible to JYS, plus these speeches cannot understand each other either. Because of the limitation of space, I only want to introduce the Gutian dialect as an example here. The following information about the Gutian dialect was collected by the author recently. The onsets of Gutian are not very different from the Meixian dialect except that it has another set of sibilants [tʃ ], [tʃ h], [ʃ ], which may include some of the syllables beginning 481

Lau Chun Fat Table 27.7  Some examples of correspondence between the Gutian and Meixian rime systems Gutian Meixian Gutian Meixian Gutian Meixian Gutian Meixian Gutian Meixian

a ai uau uaŋ iŋ in, im, uŋ ɔ a, ak uəʔ uk

ia ia, iak aŋ un, aŋ u u, au, ak iɔ iau ɔʔ ak, ap

ua ua, uai iaŋ iaŋ iu u, iu εi iεn, εu uɔʔ ak, ap

ai an uaŋ un uŋ uŋ yε i, u uʔ ɔk, uk

uai uan ə u iuŋ iuŋ aʔ at, ap iʔ it, ut, ip

au an, am əŋ ŋ o ui, ut iaʔ iak, iap yʔ iuk, ut

iau iaŋ i i, ui uo ui, ut əʔ əp, uk, εt, ut iεʔ iεt, iap

Table 27.8  The Gutian and Meixian tone systems compared

Gutian Meixian

Yinping

Yanping

Shang

Qu

Yinru

Yangru

44 44

53 11

31 31

11 52

35 32

55 55

with [k], [kh], [h] in Meixian. The rime system is much more different and Table 27.7 gives some examples of correspondence. The correspondence is quite complicated as a result of the simplification of the phonology of Gutian. It consists of the loss of the MC codas [m], [n], [ŋ], [p], [t], [k] and their substitution by [ŋ], [u] and glottal stops. There is considerable raising and centralization of vowels, plus some monophonization of diphthongs. Therefore the Gutian dialect is a totally foreign speech for most JYS speakers, although the tonal system is not very different (Table 27.8). The Yinru tone in Gutian has lost the coda but develop a totally different tone to dis­ tinguish it from the other tones. Only the Yangru tone keeps a glottal stop as coda to show that it developed from a syllable with stop coda. Unlike the JYS group, the western Fujian subdialects are the most complicated group of dialects among the Hakka as individual dialects may have developed independent of each other over more than 600 years. This may be explained by the hilly nature of the area and the contact with the Min dialects as their immediate neighbours.

References Chinese Academy of Social Sciences and Australian Academy of Humanity (1989) Language Atlas of China, Longmans. Constable, N. (1996) Hakka Identity in China and Abroad, Washington, D. C.: University of Washington Press. Huang, Xuezhe (1992) ‘Meixian Kejiahua de Yuyin Tedian’, Fangyan 4: 275–89. Leong, S. T. (1997) Migration and Ethnicity in Chinese History, Hakkas, Pengmin, and their Neighbors, Stanford: Stanford University Press. Li, Rulong and Zhang Shuangqing (1992) Kegan Fangyan Yanjiu, Xiamen: Xiamen University Press. Liu, Lunxin (1999) Kegan Fangyan Bijiao Yanjiu, Beijing: Chinese Academy of Social Sciences. Liu, Shuxin (2007) Dongjiang Shangyou Tuhuaqun Yanjiu – Yueyu Huihexi Tankao, Beijing: Zhongguo Shehui chubanshe. Liu, Zhenfa (2001) Lishi de Wuhui, Wuhui de Lishi, Beijing: Xueshu Yanjiu Publishers.

482

Hakka dialect Luo, Xianglin (1933) Kejia Yanjiu Daolun, Guangdong: Shishan Library. Norman, J. (1986) ‘What is a Kejia dialect?’ Proceedings on the Second International Conference on Sinology, Taipei: Academia Sinica, 12/86. Sagart, L. (1998) ‘On distinguishing Hakka and Non-Hakka Dialects’, Journal of Chinese Linguistics 26(2): 281–301. Wang, Futang (1998) ‘Guanyu Kejiahua he Gan Fangyan de Fenhe Wenti’, Fangyan 1: 14–19. Yuan, Jiahua et al. (1989) Hanyu Fangyan Gaiyao, Beijing: Wenzi Gaige Publishers.

483

28 Hanyu Pinyin Lilly Lee Chen Rice University, USA

1. Introduction The Chinese Phonetic Alphabet Hànyǔ Pīnyīn 汉语拼音, sometimes abbreviated Pinyin, is China’s official system for spelling Mandarin Chinese in Roman letters. It was promulgated in 1958. By convention, practically all Chinese dictionaries compiled since then feature it, often as an appendix under the heading Hànyǔ Pīnyīn Fāng’àn 汉语拼音方案 ‘Chinese phonetic alphabet scheme’. Pinyin is what Western tourists see written on street signs below the Chinese script street names (Figure 28.1). In elementary schools in China, Pinyin is used as a pedagogical tool to teach students to read Chinese script (Table 28.1, based on Xiaoxue Yuwen Jiaocai 1987, vol. 2: 113). All publications such as books, film credits, digital media, etc., coming out of China use the Pinyin system for Romanization of Chinese characters.

Figure 28.1  Beijing street sign1

484

Hanyu Pinyin Table 28.1  First-grade textbook Nà 那

biān 边

chí 池

zi 子

li 里

de 的

xiǎo 小

yú 鱼

dōu 都

yóu 游

dào 到

shuǐ 水

miàn 面

shang 上

lái 来

Pinyin is widely adopted in Sinophone areas outside of China, such as Singapore and other parts of the world where Mandarin Chinese is taught – in local Chinese community classrooms, mainstream school systems, and colleges. It is a common input method for Chinese word-processing. Pinyin is fast replacing other previously popular systems of transcription of Mandarin. As early as 1966 the British Library switched to Pinyin for the cataloging of new Chinese acquisitions. The Library of Congress, however, did not undergo the Pinyin Conversion Project until 2000, which converted Romanized Chinese data from Wade-Giles to Pinyin. In 1977 the United Nations adopted Pinyin as the international standard in Romanizing Chinese geographic names. In 1978 the Chinese State Council announced that the alphabetization of Chinese names of persons and places in Pinyin should be applied to all languages using the Roman alphabet, including English, French, German, and Spanish. In 1982, the International Organization for Standardization also adopted Pinyin as the standard Romanization system for Modern Chinese. These official decisions helped to end the quandary of the last hundred years, namely, the existence of several competing alphabetic systems, the most well-known of which being Wade-Giles and Yale, for transliteration of Chinese names and places. However, locations or personal names that are well known in the spelling of the latter two systems or others are left unchanged, for examples, Yangtse River, Sun Yat-sen. Exceptions also are names of people from Hong Kong, Taiwan, and overseas Chinese communities who prefer their own spellings, for examples, Chao Yuen Ren, Lee Kuan Yew, Lee Teng-hui.2 Taiwan has been a notable exception to the current sweeping influence of Pinyin. Zhùyīnfúhào 注音符号 ‘sound annotating symbols’, a non-Romanization system created in 1913 in the early Republican era, commonly called bōpōmōfō by its users, continues to prevail as a teaching tool for children learning characters. Tōngyòng Pīnyīn 通用拼音 ‘universal sound spelling’ designed not long ago in Taiwan using the Latin alphabet, can also be seen in transcriptions of place names. However, Pinyin is slowly making its way into Taiwan, especially since the Taiwan government in 2008 approved Pinyin.

2.  Romanization and Pinyin One may wonder: Why the fuss about Romanization, and why Pinyin in particular? It is in the context of the tradition of Chinese writing that the terms Romanization and Pinyin take on a new meaning. To Romanize or not to Romanize was a gut-wrenching issue facing scholars at a time when China was at her weakest in history in the waning years of the Imperial Qing (1644–1911). The Chinese had lived contently, pridefully even, with their writing scripts, called Hànzì 汉字 ‘characters’, for well over three millennia. It was as much their birthright as their duty to write Hànzì. It never occurred to users of Hànzì to take a 485

Lilly Lee Chen

second look at its fitness as a system to record their own speech sound. Yet the foreigners who came from the West toward the end of the Ming dynasty (1368–1644) and later found the system an immense obstacle to be overcome. The size of the sound that a character represents is no more and no less than one syllable; there are no parts in a character that can possibly be construed as a sound unit smaller than the syllable, such as phonemes. One cannot look at a character and sound it out, which one can in the case of an English word. The foreigners, with their evangelical mission, had to learn the language to be with the people, to read to be accepted by the elite, and to teach to enable the poor to read the Bible and sing the hymnals in the language of the people they were proselytizing. To meet the challenge they enlisted the help of the writing of their own mother tongues: the Latin alphabet. Such efforts foretold partly the unsatisfactory state people found themselves in, centuries later, where numerous co-existing Romanization systems staked out their own turfs. The German, the French, the English each devised their own systems that would reflect better the correspondence between the letters and their phonetic values in their own languages. Regardless, their efforts of using the alphabet for Chinese speech, in which a syllable sound unit (a Chinese character) could be broken down into smaller sound units (letters that are close to phonemes) did not go unappreciated in the land where, until then, the Latin alphabet was almost unheard of, and writing without using characters was practically inconceivable. It inspired Chinese scholars to design similar systems by using either the Latin alphabet or other sound symbols. Pinyin is one such example. The limited number of sound symbols could easily serve as aids to learning Chinese characters, the difficulty of which and the amount of time it took to acquire proficiency in them was blamed for widespread illiteracy. For the Chinese scholars, the main motivation was to help promote literacy.

3.  Predecessors I: systems created by Westerners Matteo Ricci (1552–1610), an Italian Jesuit missionary who arrived at the door of China in 1582 and soon adopted the language and culture, was the first one to come up with a system to represent Chinese sounds in the Latin alphabet. In his Xīzì Qíjì 西字奇迹 ‘miracle of Western letters’ he provided Romanization transcription side by side with his Classical Chinese texts. The spelling was, remarkably, accompanied with tone marks. That was the year 1605 – the text says so: 万历3 三十三年 ‘33rd year of the Wànlì era (of the Ming dynasty emperor Shenzong)’ (Table 28.2).4 The accompanying Romanization notations are: uań lyě sān xǎ sān niên. Later, based on texts like this, scholar Luo Changpei 罗常培 (1899–1958) was able to sort out the entire system that Ricci designed (Ni 1958: 19). This system was the first Romanization system for the language in Chinese history. Unlike Modern Mandarin which has four tones (excluding the neutral tone), Ricci’s system

Table 28.2  Xīzì Qíjì 萬 曆 三 十 三 年

uań lyě sān xǎ sān niên

486

Hanyu Pinyin

has five tone marks: ̄  ̂   ̀   ́  ̌ . That is: ā â à á ǎ. The fifth one  ̌  is for syllables called rùshēng 入声 ‘entering tone’ in the convention of Chinese historical phonology. These are syllables that end in a stop consonant (-p, -t, or -k) in Middle Chinese (7th–12th centuries) (Huang 1993: 7).5 Modern Mandarin has lost these distinctions, but they are maintained in some Southern dialects.6 In Early Mandarin (approximately 12th–17th centuries, the tail end of which was where Matteo Ricci found himself in in China), syllable endings -p, -t, and -k had already disappeared. If they existed at all, they were at best in the form of a mere glottal stop (T. Dong 1968: 76). This is corroborated by Ricci’s Romanization. Ability to recognize entering tone characters is assumed in any Chinese historical or dialectal phonologist. The entering tone mark that he placed on syllables that do not have a -p, -t, or -k ending at least reveals his knowledge in this regard. In the text quoted in Table 28.2, the characters 历 and 十, the 2nd and 4th syllables in the Romanization, belong to rùshēngzì 入声字 ‘entering tone characters’, as Ricci aptly so marked. To understand the entering tone better, let us take a dialect, Taiwanese, a variety of Southern Min, for example, that still retains this distinction. In colloquial Taiwanese, the two entering tone characters, 历 and 十, are pronounced [laʔ] and [tsap] respectively, the tone mark [  ] indicating that the entering tone has a short high pitch. While the final -p in the character 十 ‘ten’, pronounced [tsap], is still present, the syllable-ending sound [ʔ] in [laʔ] 历 ‘calendar’ is a glottal stop, which, in this particular case, is a weakened remnant of a historical -k, which still shows up in the classical reading of this character, [lek], as in the emperor’s reign name. The classical reading of 十 is [sip]. What Matteo Ricci and his missionary colleagues did was a first step that potentially could set Chinese on track to an alphabetic writing system. His follower, Nicolas Trigault (1577–1628), a French Jesuit, not only further improved Ricci’s system, his Xīrú Ěrmùzī 西儒耳目资 ‘aid to the eyes and ears of Western literati’ written in 1626 is also important to the study of the early Mandarin phonological system (T. Dong 1968: 76). The book was written in Chinese, but sounds were given in Romanization, an approach very different from Chinese phonological treatises by Chinese linguists at the time. Works like this and Ricci’s were mainly to help their missionary followers learn Chinese. The ordinary Chinese were not affected by the works of Ricci and Trigault nearly as much as by the works of the nineteenth-century missionaries. Trigault’s Xīrú Ěrmùzī contains three volumes: (i) a general statement and introduction to phonology, (ii) characters arranged according to yùn 韵 ‘the final/rhyme’, which is further categorized according to the five tones (including the entering tone), each being further differentiated according to the initial consonant, and (iii) characters arranged according to the radicals. His goal may have been to compile a complete sound dictionary of all Chinese characters, which he indeed accomplished; but the way he did it astounded Chinese scholars at the time. By employing only 25 alphabet letters (five for vowel sounds and 20 for consonants) and five tone marks he was able to capture all Mandarin syllables. It was a far cry from the fǎnqiè 反切 spelling system that the Chinese used for more than two millennia, in which one obtains/learns the sound of a particular character from two given characters (one for the syllable initial (i.e. consonant) sound, the second for the rest of the syllable sound). This set of 30 sound symbols was able to express unlimited meanings (words). It was so impressive that scholar Fang Yizhi 方以智 (1611–71) even entertained the idea that Romanization should be the writing system for the Chinese language (Ni 1958: 21), not just a sound notation device to be put next to characters. The idea of a Latinized writing system, however, had to be relegated to the back burner for some 280 years until it was rekindled in the heyday of the May Fourth Movement in 1919. 487

Lilly Lee Chen

Few significant Romanization works came out subsequent to Trigault. Missionaries in the seventeenth and eighteenth centuries were relatively restricted by the imperial court in their evangelical activities – not an encouraging sign for more missionaries to come. The nineteenth century witnessed a more active production of schemes designed for regional dialects. In the 1820s the first British Protestant missionary to China, Robert Morrison (1782–1834), published the first Chinese–English dictionary based on his own Romanization system. In the meantime the most successful Romanization for local sound was Pèh-ōe-jī ‘vernacular writing’ in Southern Min pronunciation. Pèh-ōe-jī 白话字 (PY: báihuàzì) also goes by the name jiàohuì luómǎzì 教会罗马字 ‘church Romanization’. It is a full-fledged orthography, originally created in 1820 by Walter Henry Medhurst but has since gone through several modifications. Its current utility has certainly gone beyond what its name ‘church Romanization’ would imply. One finds it, especially in Taiwan, in announcements, personal correspondence, literary works, community newsletters, etc., besides the complete Pèh-ōe-jī Bible and hymnals. The success of Pèh-ōe-jī as a writing system would have pleased the likes of Fang Yizhi some three centuries ago whose wishful thinking it was to have a writing system based on the Latin alphabet (because of its universality, in his view) with which one could write down whatever one says without having to ponder just which character to use. The design of a system of Romanization for Chinese sound passed from the Catholic Jesuits to the nineteenth-century British protestant missionaries, and finally to persons of secular profession. Thomas F. Wade (1818–95) was a British diplomat and sinologist who produced a syllabary based on the Beijing dialect. The system was slightly modified by his successor, another British sinologist and diplomat, Herbert Giles (1845–1935), and became known as the Wade–Giles Chinese Romanization system (or simply Wade–Giles), which was to dominate, for a century, the entire English-speaking world for transcription of Chinese names and words until it was gradually replaced by Pinyin in the later half of the twentieth century. Knowledge accrued from centuries of past efforts by the missionaries certainly benefited the Chinese who, by the turn of the century, found it necessary to design their own system, this time with a different mission, not of a divine nature, but a secular mission to lift China out of illiteracy and poverty.

4.  Predecessors II: systems designed by the Chinese While the missionaries in nineteenth-century China were busy trying to save the souls of the masses in their afterlife, the Chinese intelligentsia were agonizing over the fate of China when treaty after treaty, following each war with foreigners, brought ever more humiliations. The jolt from the defeat in the Opium War (1842) and, particularly, the Sino-Japanese War (1895), awakened China. Scholars were painfully aware that something must be done if China was to be rescued from herself. Backwardness, brought on by poverty and illiteracy, must be eliminated. Poverty and illiteracy fed on each other. The issue of illiteracy now received its overdue attention, and Hànzì, the pride vehicle of Chinese cultural heirloom across time and space, suddenly found itself in the eye of the storm. Literary giants and social critics denounced the role of characters in holding back China’s progress in the modern era. Contrasting the differences between the West, which was rich and strong, and China, which was poor and weak, Lu Zhuangzhang 卢戆章 (1854–1928), scholar and creator of spelling systems for regional dialects, attributed the underlying cause to the writing system. In no ambivalent tone, he said that Chinese characters were the most difficult in the world. A rich and strong state is built on an easy writing system that facilitates love for knowledge among the masses, men and women, young or old. Easy access to knowledge, especially of natural 488

Hanyu Pinyin

sciences, is the foundation for a strong nation. One can save more than ten years by learning an easy alphabetic spelling system rather than committing to memory a much larger set of difficult characters. If people spent these more than ten years of time to focus on mathematics, natural sciences, chemistry, and varieties of practical fields, there would be no need to worry about the wealth and strength of the nation: ‘省费十馀载之光阴,将此光阴专攻於 算学,格致,化学,以及种种之实学,何患国不富强也哉’ (Lu 1892: 3). Radicalists like Qian Xuantong 钱玄同 (1887–1939) advocated total elimination of characters in favor of a Latinized writing (Qian 1918). Although arguments like his were rebuffed as impractical as Chinese vocabulary contains an inordinate number of homophones, the need to have a system of sound symbols like Roman letters to help learn, if not to replace, Hànzì was indisputable. Proposals for new systems mushroomed. Organizations were established and government-supported committees convened, all aiming at language reform. A successful outcome was the Zhùyīnfúhào 注音符号 scheme, which was determined by Zhōngguó Dúyīntǒngyīhuì 中国读音统一会 ‘Commission for Unifying Reading Pronunciation’ in 1913. The commission was led by the renowned scholar Cai Yuanpei 蔡元培, the members of which were no less luminous, such as Lu Xun 鲁迅, Wu Zhihui 吴稚晖, and Wang Zhao 王照, along with representatives from all major areas of China. The system consists of 39 non-Roman symbols (ㄅㄆㄇㄈ, etc.). It was promulgated in 1918 and adopted by schools as a phonetic notation system, an aid to learning Chinese characters. Almost a century later, the system (with some modification) is still robust in Taiwan. Another system that should be mentioned is gwoyeu romatzyh 国语罗马字 (PY: guóyǔ luómǎzì) ‘the national language Romanization’, originally designed by the eminent linguist Chao Yuen Ren 赵元任 (PY: Zhào Yuánrèn). The Romanization is famously known for the way it represents the Mandarin four tones by varying the spelling of syllables – so complicated for the common people that the system failed to gain popularity in China, although it is relatively well known in the West. Luo Changpei 罗常培 in his book on the evolution of Chinese sound spelling (Luo 1934), covering spelling systems up to the time of his writing, discussed or mentioned no fewer than 50 of them. Included are also systems that use non-alphabetic symbols, such as shorthand, brush strokes, Chinese numbers, Chinese archaic characters, Japanese kana symbols, etc. These non-alphabetic scripts were detailed in two charts (one for the consonant initials, the other for the finals) in comparison with Zhùyīnfúhào for reference (Luo 1934: 81). In a similar chronological scope, but giving more coverage and credit to the immediate pre­ decessors of Pinyin that were not treated in Luo (1934), Ni Haishu stated that between 1911 (the year that saw the overthrow of the Qing dynasty and the establishment of the Republic of China) and 1949 (the beginning of the People’s Republic of China) there existed almost 100 various spelling schemes (Ni 1958: 61). In Gao (1993: 26–9) a list was provided of 28 spelling systems, in Latin or non-Latin scripts, that came out between 1892 and 1910. This list duly reflects that Lu Zhuangzhang, as seen in his Yīmùliǎorán Chūjiē 一目了然初阶 ‘Beginning steps in being able to understand at a glance’, published in 1892, was the first Chinese scholar to propose a spelling system for the Chinese language. In this particular work he made use of Latin alphabet letters, along with some modifications of his own. The spelling system was mainly for Southern Min. The immediate predecessor of Pinyin was Lādīnghuà Xīn Wénzì 拉丁化新文字 ‘Latinized new script’, also known as Sin Wenz or Latinxua Sinwenz, its characteristics being without tone marking. It was originated in Moscow in 1928 for Chinese expatriates living in the far east of the Soviet Union to achieve literacy without the medium of Chinese characters. Despite promotional efforts as exemplified in Lin (1950), it was not widespread, eventually being superseded by Pinyin. 489

Lilly Lee Chen

Besides these spelling systems as predecessors of Pinyin, one must mention the languagerelated movements taking place around and after the May Fourth Movement as prerequisites to the successful establishment of the Pinyin project in the 1950s. These were báihuàwén yùndòng 白话文运动 ‘movement for vernacular writing’ and guóyǔ yùndòng 国语运动 ‘movement for a national language’. They advocated for a vernacular writing style, báihuà 白话 ‘plain speech’, instead of the classical wényán 文言 (thus, promoting yánwén yīzhì 言文一致 ‘write as one speaks’), and Beijing-based language as the national language guóyǔ 国语 (thus, the spelling system designed can be used as a communication tool for all Chinese). All these movements paved the way for Pinyin.

5.  On the way to Hànyuˇ Pı¯nyı¯n Fa¯ng’àn The need to unify China and to solve the problem of poverty and backwardness, an urgent two-pronged task for the newly formed government of the People’s Republic of China, prompted a new surge of dialogs and calls for action regarding language reform (see Zhang Xiruo (1956: 50–61), Wu (1956: 62–6), and other articles in Zhōngguó Wénzì Gǎigé de Dìyībù ‘The first step to script reform in China’ (1956)). China was lagging far behind advanced Western nations in the twentieth century, meaning nations like Britain, Germany, France, and the USA, whose writing systems happened to be the Latin alphabet. Their languages, particularly English, commandeered the fields of science, industry, and technology. Thus, if China was to catch up, it was in the interest of the Chinese to have a writing system that could align itself with the international community. The key wording here is yǔ shìjiè jiēguǐ 与世界接轨 ‘to be in line with the world community’, a term recurrent in the narrative of the time. The phrase avoided using words that would otherwise suggest imitating or borrowing from the much-maligned Western nations. Implicit in the rhetoric was the necessary dethronement of the Zhùyīnfúhào system – a system put in place by the previous government and, above all, whose letters were non-alphabetic. Zhùyīnfúhào by then appeared to be relatively entrenched in schools after a generation of practice, as evidenced by the fact that in Hànyǔ Pīnyīn Fāng’àn the Zhùyīnfúhào symbols (ㄅㄆㄇㄈ, etc.) were included alongside the alphabet letters to serve as an aid to their pronunciation. The Hànyǔ Pīnyīn Fāng’àn project began as early as 1949 when Wu Yuzhang 吴玉章, a party official who was soon to take charge of the Pinyin project, asked for a directive from Mao Zedong for principles that would underlie the phonetic writing to be designed (Jiaoyubu 2008). The principles stipulated that the system must be: scientific, in line with the inter­ national community, and mass-oriented. This clearly pointed in the direction of the Latin alphabet. A preparatory committee for Chinese language script reform, Zhōngguó wénzì gǎigé yánjiū wěiyuánhuì 中国文字改革研究委员会, was established in 1952. Between 1950 and 1955 a total of 655 proposals of spelling systems were submitted by the general public. The Chinese Script Reform Committee was formally convened in 1954 and charged with the following: (i) the character simplification project, (ii) the Chinese spelling scheme (i.e. Hànyǔ Pīnyīn Fāng’àn), and (iii) promotion of Pǔtōnghuà 普通话 (which was to be the standard national language, formerly known as guóyǔ 国语, or ‘Mandarin’ in the West). From 1955 onwards the committee for the Chinese alphabetic scheme project, which was under the Chinese Script Reform Committee, studied, designed, and, in 1957, finalized the Hànyǔ Pīnyīn Fāng’àn. It was approved and adopted by the 1st National People’s Congress in February 1958. Unlike the character writing system, which is morpheme-based (i.e. meaning-based), Hànyǔ Pīnyīn is sound-based, and the designated sound was the Beijing dialect. In that vast land of China, most people at the time did not speak nor understand this Beijing-based 490

Hanyu Pinyin

Pǔtōnghuà. Thus, learning Pinyin was tantamount to learning Pǔtōnghuà ‘common speech’. The two had to take place simultaneously. This was a tremendous task, given that even teachers themselves in many regions at the time could not speak Pǔtōnghuà. Promotion of Pǔtōnghuà and Pinyin in teacher’s colleges, and elementary and secondary schools went hand in hand with publications and textbooks used in schools. Were it not for the strong official determination and policies as well as thorough (and required) collaboration from scholars and educators, the scale of which was unprecedented in any previous administrations, the success of Pinyin would not be what we see today. On the other hand the extent of the promotion of Mandarin was such that decades later concerns began to surface of the integrity of regional tongues (Z. Chen 1999) and advantages of bilingualism/bidialectalism (Zhan 1999).

6.  The Pinyin system Hànyǔ Pīnyīn Fāng’àn consists of five components: (1) alphabet letters (zìmǔbiǎo 字母表), (2) initials (shēngmǔbiǎo 声母表), (3) finals ( yùnmǔbiǎo 韵母表), (4) tone marks (shēngdiào fúhào 声调符号), and (5) the apostrophe (géyīn fúhào 隔音符号).

6.1.  Syllable structure and the notions of the initial and final In any meaningful discussion on the Chinese sound system, the notions of the syllable and its two components, the initial and final, are indispensible. They are crucial in both synchronic description (of Modern Mandarin itself, for example) and diachronic analysis (of historical Chinese or relationships between dialects). Hànyǔ Pīnyīn Fāng’àn is no exception. A few words of explanation are due here. ‘Initial’ refers to the syllable initial consonant. If there is none, the syllable is by tradition said to contain a zero initial. ‘Final’ is what is left of the syllable. The tone ranges over the entire syllable. Table 28.3  Chinese syllable structure Tone Initial

Final Vowel Medial

Final Consonant Main vowel

Categorizing sounds in terms of initial shēng 声 and final yùn 韵 is in keeping with the tradition of Chinese phonology, particularly historical phonology. In the Chinese historical linguistic context, the term for phonology is yīnyùnxué 音韵学 ‘the study of sound and rhyme’ or shēngyùnxué 声韵学 ‘the study of initial and final’. The word yùn 韵 means ‘rhyme’. Why should Chinese sounds be classified in terms of the concept final/rhyme? This is because Chinese phonology got its start in rhyme dictionaries, specifically Lu Fayan’s rhyme dictionary Qièyùn 切韵 (ad 601), which has since been lost but recovered via subsequent rhyme books written based on his system. It was from the way he (and his colleagues) arranged the Chinese characters in terms of the categories of tone and rhyme, along with provision of sound by the fǎnqiè 反切 method, that modern linguists were able, with further data from sounds of modern dialects, to reconstruct Middle Chinese. 491

Lilly Lee Chen

Rhyme books, as the name denotes, were for poetry. Classical poetic schemes were sophisticated and highly regulated (regarding what characters rhymed with what characters, e.g. entering tone characters rhyming only with other entering tone characters). Rhyme/final reigned supreme in the tradition of rhyme-based studies – the forerunners of Modern Chinese phonology. Modern phonologists may not have poetry-making in mind, but they have appropriated the terms initial and final in their works – for good reasons. Statements of Chinese historical development crucially rely on the terms initial and final. For example, historical initial p- became f- in certain phonetic environments. The -p in syllable final never did, and, instead, disappeared completely from Modern Mandarin, along with other final consonants, except n and ng. Historical -m in syllable final coalesced with -n in Mandarin, whereas m- in the initial never did. In dialectal comparison, Middle Chinese initial velars (k, k’, h) before high-front vowels underwent palatalization in Mandarin (now [tɕ], [tɕ’], [ɕ] respectively), whereas Southern Min retains them. Also synchronically, the same letter in a different syllabic position has different phonetic qualities. For example, in Cantonese the stop consonants, p, t, and k, in the initial position are released, whereas in the final they are unreleased. In Mandarin, p, t, and k occur only in the initial position, and r in the initial represents [ʐ ], whereas in the final the sound is [ɻ ]. There is a lopsided distribution of consonants: the initial consonants by far outnumber those in the final position in all Hàn (Chinese) dialects. In Mandarin, only consonants -n and -ng are allowed in the final (except -r, due to the érhua process), whereas ng [ŋ] is not permitted in the initial. These are but some examples that bespeak the validity of the concepts of the initial and the final in describing Chinese sounds. In light of this, the convenient and general way of referring to Chinese syllable structure as (C)V(C) (DeFrancis 1984: 45) is very general indeed, as the two Cs in the two positions do not behave the same, either synchronically or diachronically.

6.2.  Hànyuˇ Pı¯nyı¯n Fa¯ng’àn consists of five components All letters and finals in the charts in Hànyǔ Pīnyīn Fāng’àn were accompanied with their corresponding sound symbols in the Zhùyīnfúhào (ㄅㄆㄇㄈ) spelling system. The five components are shown below. 1. 26 Phonetic alphabet letters Table 28.4  Table of phonetic letters 字母表 Zimubiao aㄚ h ㄏㄚ oㄛ uㄨ

b ㄅㄝ iー p ㄆㄝ v 万ㄝ

c ㄘㄝ j ㄐーㄝ q ㄑーㄡ w ㄨㄚ

d ㄉㄝ k ㄎㄝ r ㄚㄦ x ㄒー

eㄜ l ㄝㄌ s ㄝㄙ y ーㄚ

f ㄝㄈ m ㄝㄇ t ㄊㄝ z ㄗㄝ

g ㄍㄝ n ㄋㄝ

The letter v is for sounds of foreign languages, other Chinese dialects, or national minority languages.

492

Hanyu Pinyin

2. 21 initials Table 28.5  Initial consonants 声母表 Shengmubiao b ㄅ玻

p ㄆ坡

m ㄇ摸

g ㄍ哥

k ㄎ科

h ㄏ喝

zh ㄓ知

ch ㄔ蚩

sh ㄕ诗

f ㄈ佛

r ㄖ曰

d ㄉ得

t ㄊ特

n ㄋ讷

j ㄐ基

q ㄑ欺

x ㄒ希

z ㄗ资

c ㄘ雌

s ㄙ思

l ㄌ勒

3. 35 finals Table 28.6 Finals 韵母表 Yunmubiao aㄚ啊 oㄛ喔 eㄜ鹅 ai ㄞ 哀 ei ㄟ 欸 ao ㄠ 熬 ou ㄡ 欧 an ㄢ 安 en ㄣ 恩 ang ㄤ 昂 eng ㄥ 亨的韵母 ong (ㄨㄥ)轰的韵母

i ー衣 ia ーㄚ 呀 ie ーㄝ 耶 iao ーㄠ 腰 iou ーㄡ 忧 ian ーㄢ 烟 in ーㄣ 因 ia ーㄤ 央 ing ーㄥ 英 iong ㄩㄥ 雍

uㄨ乌 ua ㄨㄚ 蛙 uo ㄨㄛ 窝 uai ㄨㄞ 歪 uei ㄨㄟ 威 uan ㄨㄢ 弯 uen ㄨㄣ 温 uang ㄨㄤ 汪 ueng ㄨㄥ 翁

üㄩ迂 üe ㄩㄝ 约

üan ㄩㄢ 冤 ün ㄩㄣ 晕

This chart for finals is followed by a list of six notes of spelling rules (see Xīnhuá Zìdiǎn (2002: 669)). 4. Tone marks Table 28.7  Tone marks 声调符号 Shengdiao fuhao 阴平 yīnpíng —

阳平 yángpíng /

上声 shǎngshēng \/

去声 qùshēng* \

*  Yīnpíng 阴平, yángpíng 阳平, shǎngshēng 上声 and qùshēng 去声 are traditional ways of referring to tone categories that, in the case of Mandarin, are also known as the 1st, 2nd, 3rd, and 4th tone respectively. In the Qièyùn rhyme dictionary, characters were first grouped in four tone categories: píng 平 ‘level’, shǎng 上 ‘rising’, qù 去 ‘departing’ and rù 入 ‘entering’, then further into rhyme categories. Yīn refers to voiceless initial consonants and yáng for voiced ones. The voicing contrast was the condition for the split of the píng tone into Modern Mandarin 1st and 2nd tones. By the 13th century, píng had already split into yīnpíng and yángpíng, and rù disappeared, having dispersed into other tones (Huang 1993: 20).

493

Lilly Lee Chen

The tone mark should be placed above the main vowel of the syllable. The neutral tone is not marked. For example: Table 28.8  Tone mark placement 妈 mā 阴平 yīnpíng

麻 má 阳平 yángpíng

马 mǎ 上声 shǎngshēng

骂 mà 去声 qùshēng

吗 ma 轻声 qīngshēng

5. géyīn fúhào 隔音符号, the apostrophe to avoid conflation of two adjacent syllables. A syllable followed by another syllable that begins with an a, o, or e should have an apostrophe to separate the two syllables when otherwise confusion may occur, for example: pi’ao 皮袄 ‘leather jacket’. Above, (1)–(5) show the way Hànyǔ Pīnyīn Fāng’àn is conventionally presented in a Chinese dictionary since the system’s debut in the mid 1950s. See, for instance, Xīnhuá Zìdiǎn 新华字典 (2002: 667–9).

6.3.  The alphabet and sounds Now for the sounds represented by the Pinyin alphabet. There are six vowel letters, including ü, and two semi-vowels represented by w and y. Vowels: a [a] e [ɛ, ɤ]  i [i, ɨ]  o [o]  u [u]  ü [ü] w [u] y [i] The simplified vowel chart (Table 28.9) based on the IPA is a convenient reference for indicating the phonetic values of the Pinyin vowel letters (in parentheses). The 21 letters and their phonetic values based on the IPA are given in Table 28.10. The four tones are given in Table 28.11. Tones can change due to specific phonetic environments, largely in anticipation of the immediate following tone. These changes are not shown in Pinyin. Table 28.9  Vowel chart FRONT     CENTRAL    BACK HIGH MID LOW

i, ü/y (i u/ü) ------------ ɨ (i) ------------- u (u) ɛ (e) ----------- ə/ɤ (e) ------------- o (o) ------ a (a) ------

Table 28.10  Consonant chart Labial Dental Velars Alveolar Palatal Retroflex

b [p] d [t] g [k] z [ts] j [tɕ] zh [tʂ]

p [p’] t [t’] k [k’] c [ts’] q [tɕ’] ch [tʂ’]

494

m [m] n [n] h [h] s [s] x [ɕ] sh [ʂ]

f [f] l [l] (-ng [ŋ]) r- [ʐ ] / -r [ɻ ]

Hanyu Pinyin Table 28.11  Four tones Tones Pinyin Pitch & Contour IPA

1st — high level ˥˥

2nd / mid rising ˧˥

3rd \/ mid-falling- rising ˨˩˦

4th \ high-falling ˥˩

Neutral tone unmarked unstressed

Below are rules for pronunciation: 1. e is pronounced [ə] when it is the lone vowel in the syllable (te [t’ə] 特); otherwise [ɛ] (bie [piɛ], bei [pɛi]). 2. i is pronounced [ɨ] immediately after alveolar and retroflex consonant series, otherwise [i] (zi [tsɨ], ci [ts’ɨ], si [sɨ]). 3. o is pronounced [uo] immediately after labial consonants (bo [puo], po [p’uo], mo [muo], fo [fuo]), otherwise [o] (gou [kou]). 4. ui is pronounced [uei], saving one symbol, as there is no contrasting [ui] sound in Pǔtōnghuà. 5. iu is pronounced [iou], saving one symbol, as there is no contrasting [iu] sound in Pǔtōnghuà. 6. Palatal consonants ( j, q, x) are always immediately followed by a high-front vowel sound ([i] or [ü]). Thus, ju, qu, and xu are pronounced [tɕü], [tɕ’ü], [ɕü], instead of [*tɕu], [*tɕ’u], [*ɕu] respectively. Similarly, ji, qi, and xi are pronounced [tɕi], [tɕ’i], [ɕi], instead of [*tɕɨ], [*tɕ’ɨ], [*ɕɨ] respectively. 7. Alveolar and retroflex consonants are never immediately followed by a high-front vowel sound. Thus, zu, cu, and su are pronounced [tsu], [ts’u], [su], rather than [*tsü], [*ts’ü], [*sü] respectively. Similarly, zi, ci, and si are pronounced [tsɨ], [ts’ɨ], [sɨ] rather than [*tsi], [*ts’i, [*si], respectively. 8. Velar consonants are never immediately followed by a high-front vowel sound (*gi, *gü, *ki, *kü, *hi, *hü).   Velars and alveolars/retroflexes are in complementary distribution regarding their following vowels, reflecting part of the historical palatalization processes, in which the historical velars (g, k, h) and alveolars (z, c, s), when followed by high-front vowels, became palatals (Wang 1964:19–20; T. Dong 1968: 209–36). 9. ü is written without the two dots when preceded by the palatals j, q, x (thus ju, qu, xu), saving a symbol, as the latter can only be followed by a high-front vowel sound. In other words, j, q, or x serves as a signal that the following vowel is a high-front vowel [i] or [ü]. Conversely the alveolars z, c, s and retroflexes zh, ch, shi, and r serve to signal that the following vowel is non-high and non-front: zu [tsu], cu [ts’u], su [su]; zi [tsɨ], ci [ts’ɨ], si [sɨ]; zhu [tʂu], chu [tʂ’u], shu [ʂu], ru [ʐu]; zhi [tʂɨ], chi [tʂ’ɨ], shi [ʂɨ], ri [ʐɨ].

6.4.  Rules for writing If a syllable contains a zero initial consonant and has more than one vowel, the syllable beginning i, u, or ü is written as y, w, or yu, respectively. If a syllable contains a zero initial consonant and i, u, or ü is the only vowel in the syllable, add y, w, or y respectively. The ü in these cases is written without the two dots. 495

Lilly Lee Chen

iao → yao i → yi

uan → wan u → wu

üan → yuan ü → yu

The rules for pronunciation contained in (3), (4), and (5) above can be stated as rules for writing: [puo] → bo [uei] → ui

[p’uo] → po [iou] → iu

[muo] → mo   [fuo] → fo

The following is about érhuà 儿化 contraction, which Southern Chinese tend to ignore when speaking Pǔtōnghuà. Besides the variety of finals shown in Table 28.6, Pǔtōnghuà also contains a great number of words that end in -r [ɻ ] (huār ‘flower’, qǔr ‘song’), the pronunciation of which is similar to English ‘r’ in ‘bar’. This comes from adding ér to a noun, mostly. The process is called érhuà. Besides causing sound changes, its function can be morphological (gài ‘to cover’, gàir ‘a cover’) or lexical (tóu ‘head’, tóur ‘chief’) (Hu 1981:137–43). The érhuà sound changes are: (1) if the preceding syllable ends in an a, e, o, or u, nothing happens, (2) if the preceding open syllable contains an i, or ü as the only vowel, the syllable acquires an [ə] sound, (3) if the preceding syllable ends in i, or n, both become zero, and (4) if the syllable ends in ng [ŋ], the nasal consonant disappears, but its preceding vowel is nasalized. These phonetic changes are not reflected in Pinyin. The rule is to simply add r to the preceding syllable, shown below. 1. 2. 3. 4.

-a, -e, -o, -u: huā + ér → huār [huaɻ ] -i, -ü: mǐ + ér → mǐr [miəɻ ] qù + ér → qùr [tɕ’üəɻ ] -i, -n: pái + ér → páir [p’aɻ ] guǎn + ér → guǎnr [kuaɻ ] -ng: dāng + ér → dāngr [tãɻ ]

Placement of tone marks: A tone mark is placed above the main vowel if there is more than one vowel, the order of priority being: a > e > o > i/u (tài, xiā, xiào, biǎo, bié, bèi, ròu, ruò). When i and u are the only vowel letters in the syllable, the last one gets the tone mark (  jiù, huì). Many words in Mandarin contain the neutral tone (i.e. unstressed syllables) (Zhang Xunru 1953), which is not marked in Pinyin. Table 28.12 gives all the combinations of the initials and finals in Mandarin.

7.  Obscure antecedents of Pinyin Pinyin in its function as sound notation has its ancient counterparts in the fǎnqiè ‘spelling’ method and dúruò ‘read like’. Besides these, the ancients also had other ways of using characters for phonetic purposes. They are listed below, in the order of how close the methods are to modern sound spelling, xíngshēng being the least similar. Xíngshēng is as old as the character writing system. Tōngjiǎ and jiǎjiè are as old as the earliest Chinese texts. 1. fǎnqiè 反切 ‘spelling (with two characters)’ 2. dúruò 读若 ‘read like’. Other names for the same function: dúwéi 读为 ‘read as’, shēngtóng 声同 ‘same sound’, 读如 dúrú ‘read like’, 声近 shēngjìn ‘sound close to’, zhíyīn 直音 ‘exact sound’, etc. While the last was for the exact same sound, the rest could be relatively loose. 496

yi

ya

yan

yang

yao

ye

i

ia

ian

iang

iao

ie

er

er

en

en

eng

e

ei

e

ei

ang

ao

ang

ao

bie

biao

bian

bi

beng

ben

bei

bao

bang

ban

bai

ai

an

ai

ba

b

a

an

a

Table 28.12  Pinyin Chart

Finals ( yùnmǔ 韵母)

pie

piao

pian

pi

peng

pen

pei

pao

pang

pan

pai

pa

p

mie

miao

mian

mi

meng

men

mei

me

mao

mang

man

mai

ma

m

feng

fen

fei

fang

fan

fa

f

die

diao

dian

dia

di

deng

den

dei

de

dao

dang

dan

dai

da

d

tie

tiao

tian

ti

teng

te

tao

tang

tan

tai

ta

t

nie

niao

niang

nian

ni

neng

nen

nei

ne

nao

nang

nan

nai

na

n

lie

liao

liang

lian

lia

li

leng

lei

le

lao

lang

lan

lai

la

l

geng

gen

gei

ge

gao

gang

gan

gai

ga

g

keng

ken

ke

kao

kang

kan

kai

ka

k

heng

hen

hei

he

hao

hang

han

hai

ha

h

jie

jiao

jiang

jian

jia

ji

j

qie

qiao

qiang

qian

qia

qi

q

Initials (shēngmǔ 声母)

xie

xiao

xiang

xian

xia

xi

x

Table of the Combinations of the Initials and Finals in Mandarin

zhi

zheng

zhen

zhe

zhao

zhang

zhan

zhai

zha

zh

chi

cheng

chen

che

chao

chang

chan

chai

cha

ch

shi

sheng

shen

shei

she

shao

shang

shan

shai

sha

sh

ri

reng

ren

re

rao

rang

ran

r

zi

zeng

zen

zei

ze

zao

zang

zan

zai

za

z

ci

ceng

cen

ce

cao

cang

can

cai

ca

c

si

seng

sen

se

sao

sang

san

sai

sa

s

pou

pu

yo

yong

you

o

ong

ou

iong

iu

o

ong

ou

yue

wei

wen

uo

yu

yuan

yun

üe

ui

un

uo

ü

üan

ün

duo

dun

dui

duan

tuo

tun

tui

tuan

nü

nuo

nüe

nuan

nu

lü

luo

lun

lüe

luan

lu

gong

guo

gun

gui

guang

gua

gu

gou

guan

wan

wang

uan

uang

tu

long lou

g

guai

du

nong nou

liu

ling

lin

l

wai

fu

tong tou

niu

ning

nin

n

uai

mu

dou

dong

diu

ting

t

wu

fou

fo

ding

d

wa

mou

mo

miu

f

u

po

ming

min

m

kuo

kun

kui

huang

kuan

kuai

kua

ku

kou

kong

k

huo

hun

hui

huang

huan

huai

hua

hu

hou

hong

h

jun

juan

ju

jue

jiu

jiong

jing

lin

j

qun

quan

qu

que

qiu

qiong

qing

qin

q

Initials (shēngmǔ 声母)

xun

xuan

xu

xue

xiu

xiong

xing

xin

x

Table of the Combinations of the Initials and Finals in Mandarin

ua

bu

bo

ping

io

bing

ying

pin

p

ing

bin

yin

in

b

Table 28.12 (cont’d)

Finals ( yùnmǔ 韵母)

zhuo

zhun

zhui

zhuang

zhuan

zhua

zhu

zhou

zhong

zh

chuo

chun

chui

chuang

chuan

chuai

chu

chou

chong

ch

shuo

shun

shui

shuang

shuan

shuai

shua

shu

shou

sh

ruo

run

rui

ruan

ru

rou

rong

r

zuo

zun

zui

zuan

zu

zou

zong

z

cuo

cun

cui

cuan

cu

cou

cong

c

suo

sun

sui

suan

su

sou

song

s

Hanyu Pinyin

3. tōngjiǎ 通假 ‘sound borrow’ and jiǎjiè 假借 ‘phonetic loan’ 4. xíngshēng 形声 ‘semantic–phonetic character’, in which one part serves as radical (clue to meaning), the other, phonetic. An example for (1) can be found in annotations by Zhang Shoujie 张守节 (seventh century) for Shǐjì 史记 Records of the Grand Historian (Sima ca. 91 bc): 好, 火到反 (Shǐjì vol. 1: 177). The phrase X, YZ fǎn (  fǎn as in the word fǎnqiè) means: X/好 has the initial sound of Y/火 and the final of Z/到. Thus, 火 huo + 到 dao → 好 hao. The fǎnqiè spelling has been faulted as failing to provide the precise sound, but the ancients’ very core concept that a Chinese character/morpheme/syllable consisted of two parts, called the initial and the final, is the foundation of Chinese phonology, the legacy of which is preserved today in Hànyǔ Pīnyīn Fāng’àn – the sound description is couched in terms of the initial and final. An example for (2) can be seen in Shuōwén Jiězì 說文解字 (ad 121) by Xu Shen 许慎 (ad 58–147), a comprehensive dictionary that provided etymology, meaning, and sound for each character. For the character 鼾, the author says: 读若汗 ‘read like 汗’. An annotator added his own fǎnqiè below Xu’s text: 鼾, 候幹切. That is: Y/hou 候 + Z/gan 幹 → X/han 鼾. An example for (3) can be seen in Shǐjì 史记. Annotator Xu Guang 徐广 (ad 352–425), referring to the character 荐 in 荐绅, stated that it was a phonetic loan jiǎjiè 假借 (Shǐjì vol. 1: 47), and that the character should be 缙. The meaning should be 缙绅 jìnshēn ‘the gentry’, instead of 荐绅 jiànshēn. Jiǎjiè 假借 is the name of one of the six categories of characters.7 As a verb it means ‘make use of’. Characters for phonetic borrowing (i.e. tōngjiǎ and jiǎjiè) were rampant in ancient texts. 90% of characters in oracle bone inscription texts 甲骨文 jiǎgǔwén, the oldest Chinese texts, dated from 1200 bc, were sound borrowings tōngjiǎ (T. Zhang 1991: 284). Yao (1980: 7–39) has a similar view in his study of texts in both oracle bone and bronze inscriptions. For example, in a text of 32 characters inscribed on 利簋 lìguǐ (a bronze vessel, dated eleventh century bc), describing the battle of King Wu 武王 against Zhou 纣, 22 characters are jiǎjiè. These 22 characters were there to represent the phonetic values of the words they stood for – much to readers’ chagrin. Often readers were led astray by deducing meaning from these characters, instead of reading for their sounds. Tōngjiǎ and jiǎjiè characters kept generations of annotators and commentators busy. Examples for (4) can be seen in the following group of semantic–phonetic characters, the phonetic component here being 者, which alone is also an independent character: 者渚赌睹 堵煮署暑箸锗楮褚猪诸屠都奢赭, etc. (Liang 1959: 171). Although today these characters no longer sound the same due to millennia of sound changes, in them the ancients have left plenty of evidence to their descendents that they did try to make the writing reflect their spoken sound when they created these and other semantic–phonetic characters. Today, about 90% of characters are semantic–phonetic characters; this ratio in fact had already been reached by the twelfth century (Li 1986: 21). To be acknowledged along with these xíngshēngzì creators are the practitioners of jiǎjiè and tōngjiǎ in those bygone days. Also, in character simplification in the twentieth century it is often the phonetic component of a semantic– phonetic character that is preserved: 表(錶), 复(復), 家(傢), 卷(捲), 蒙(矇), 面(麵).

8.  Comparison of spelling systems The relationship between symbol and sound is arbitrary. For an ordinary user of a spelling system, it is a matter of getting accustomed to it. Wade–Giles and Yale, as compared with Pinyin, are more user-friendly to English speakers, for the obvious reason that they are closer to the English spelling system in sound and letter correspondences. The letters j, q, and x in 499

Lilly Lee Chen

Pinyin particularly received bad reviews. However, in terms of the system-internal cohesive relationship of symbols, for capturing similarity and contrast in the language, Pinyin deserves a good rating. One quintessential characteristic of the language is that it does not tolerate any cluster of consonant sounds (cf. English: casts, splits). When a Chinese encounters a consonant cluster, the latter will be either broken up (with inserted vowels) or zapped: Amsterdam (three syllables) → Āmǔsītèdān (five syllables). The designers of Pinyin seemed to have taken heed of this by avoiding it, at least in appearance: one symbol for one sound, except for zh, ch, sh, in which case the extra h is to mark retroflexion. A comparison of z, c, s : zh, ch, sh in Pinyin and their counterparts ts, ts’, s : ch, ch’, sh in Wade–Giles illustrates the point about internal contrast and similarity. In Wade–Giles, the uses of c and t do not show the similarity between the two sounds that they represent, whereas in Pinyin the presence of the letter z shows the phonetic relationship between z and zh. The contrasting phonetic feature between the two, z and zh, is a matter of retroflexion represented by h. The series, z, c, s and zh, ch, sh (as contrasted with other consonants) also share something unique: their following vowels are not high-front vowels ([i] and [ü]); rather, the i and u vowel symbols after them are for dull, non-high-front [ɨ] and [u]. In other words, these two series of symbols serve as signals to the vowel quality of the immediately following vowel. Table 28.13 gives a comparison of how the contrast between the retroflex series and alveolar series is represented in different systems. The table is mostly based on Luo (1934: 35–81), with additions and modifications. Table 28.13  Comparison of retroflexes and alveolars in different systems Spelling system A. alphabetic: – Pinyin – Gwoyeu Romatzyh – M. Ricci – N. Trigault – J. Edkins – T. F. Wade – Couvreur – BEFEO – Gabelantz – C. Arendt – F. Lessing – Zhu Wenxiong – Liu Mengyang – Huang Xubai – Liu Jishan – Qian Xuantong – Chao Yuan Ren – Zhou Bianming – Lin Yutang B. non-alphabetic: – Zhùyīnfúhào – Lu Zhuangzhang – Cai Zhang – Cai Xiyong

Retroflex

Alveolar

zh j ch ch ch ch tch tch c ć  dsch tz g̈ cz j gh j c c

ch ch c’h ‘ch c’h ch’ tch’ tch’ c’ ćˈ tsch chz x cx q ch ch ch ch

sh sh x x sh sh ch ch s š sch sz s̈ cs w sh sh sh sh

z tz c, ç ç ts ts (tzŭ) ts ts ts ts ds ts t dz z dz tz z z

c ts c’, ç’ ‘ç t’s ts’ (tz’ŭ) ts’ ts’ ts’ ts’ ts th c tx c ts ts zh ts

s s s s s s (ssŭ) s s (sseu) s s (sz̜ě) s s s s s s s s s

ㄓ

ㄔ

辶 /

尸 /

ㄕ マ 十 )

ㄗ ) 子 (

ㄘ ( 冫 (

ㄙ フ 幺 )

500

Hanyu Pinyin Table 28.14  The use of the letter h in Wade–Giles and Pinyin Pinyin Wade–Giles Character

chi ch’ih 吃

qi ch’i 七

jue chüeh 觉

er erh 儿

xie hsieh 谢

lüe lüeh 略

nie nieh 捏

ri jih 日

tie t’ieh 贴

yue yüeh 月

Another point of comparison is the use of the letter h. In contrast to the unique function of h in Pinyin, that of representing retroflex when accompanying a consonant, Wade–Giles uses it for multiple functions. The symbol’s high frequency (in Wade–Giles) reduces its information value, i.e. being less meaningful (Chao 1968: 205). For contrast, see Table 28.14, where Wade–Giles shows h, the corresponding Pinyin does not. Since Pinyin adheres to single consonant representation (except h as noted above) and its concomitance of keeping alveolar, retroflex, and velar series apart in the letters used, it is more convenient to make statements about Chinese historical phonology when referring to words in Pinyin. For example, historical velars underwent palatalization when followed by a high-front vowel ([i] or [ü]), resulting in the fact that Modern Mandarin’s palatals and velars are in complementary distribution regarding what syllable finals they can have (although this does not necessarily mean all modern palatals were derived from historical velars). In Wade–Giles ch can be for a palatal (江: WG: chiang, PY: jiang) or retroflex (张: WG: chang, PY: zhang). The co-mingling of letters in Wade–Giles in this case, but keeping distinct in Pinyin makes Pinyin more historical-phonology friendly. In phonological practices, what is good for Chinese historical phonology also works well for Chinese dialectology (see e.g. sound comparisons between Mandarin and other dialects (S. Dong 1964: 129–74)).

9.  The future of Pinyin The future of Pinyin depends on the fate of the character writing system. Paucity of syllables actually used in Mandarin (totaling 404 out of possible 1,900 (Duanmu 2011), discounting tones – people usually write Pinyin without tone marks) presents a hurdle for Pinyin on its way to becoming a full-fledged writing system, that is, if it was ever going to be one, as it once was expected to be in the mid-1950s, but formally abandoned in 1986 at the 2nd National Congress on Language and Script Reform (P. Chen 1999: 196). Even among Han dialects, Mandarin has a relatively poor inventory of syllables.8 A comparison of two-syllable words in Table 28.15 (an overwhelming majority of words in Modern Chinese being disyllabic (M. Chen 1981)) between Mandarin and Southern Min proves the point. Only a sample of Mandarin words that have the random sound qishi is presented. (Southern Min in Pèh-ōe-jī spelling system). Another multitude of homophones, this time qi, appears in Table 28.16, all having different meanings as seen in their different characters. Table 28.15  Mandarin homophonous words qishi and their correspondences in Southern Min Mandarin Southern Min Character Meaning

qīshí chhitchàp 七十 70

qīshì chhitsū 七事 7 events

qǐshì khésī 啓示 reveal

qǐshì khésū 啓事 announcement

501

qìshì khìsè 气势 loftiness

qíshì kîsū 奇事 strange thing

qíshí kîsìt 其实 in fact

qíshí hitsî 其时 that time

qǐshì khísī 岂是 how?

qīshì khisè 欺世 deceit

Lilly Lee Chen Table 28.16  Homophonous characters for qi [tɕ’i] 1st tone: 七沏妻凄棲悽萋栖戚嘁期欺蹊漆, etc. 2nd tone: 齐脐祈頎麒奇畦其骑旗棋淇崎岐歧鳍, etc. 3rd tone: 乞岂起企启绮杞玘屺芑棨稽, etc. 4th tone: 气汽訖迄弃泣契砌器憇跂葺, etc.

The examples in Tables 28.15–28.16 are evidence that the character writing system has too many difficult symbols. They speak for the cause of Pinyin. These same charts, on the other hand, also speak for the opposite cause that Pinyin obscures meanings, hence the need to keep characters around. Regardless of the merits of Pinyin as a possible writing system, the virtue of the weighty information load that characters famously carry cannot be easily dismissed, the price for which the Chinese seem willing to pay (despite the seemingly tortuous number of characters one has to learn)9 and this is why, in the rivalry of the two, characters have always gotten the upper hand (namely, clarity over economy), so far and in the foreseeable future. Any educated Chinese would be hard pressed to read a page of text in Pinyin, let alone an entire book. The battle between the two used to be fought on the ground of illiteracy and poverty. With universal education, the fast growing economy of China, and the new information age, the ground is now shifting to where the global mass media is. In the early stage of personal computing, information was entered via none other than the keyboard – the death knell for the characters, so everybody gloomily predicted. Then, fast enough, Pinyin came to the rescue. It turned out that Pinyin provides an easy input method for characters to prance again, no less majestically, this time on the computer screen – far beyond the parameters of oracle bone, bronze, and paper. Today, one can use one’s finger as text input on a trackpad, or directly on the screen in cases like the iPad. As long as characters flourish, Pinyin graciously continues to be an auxiliary means of representing Mandarin and an invaluable learning aid in Chinese language instruction in China and elsewhere across the world.10

Notes   1 Photo courtesy of Prof. Gao Yuan of Beijing University of Aeronautics and Astronautics.   2 The Library of Congress compiled a list of headings (ca. 2,400 names) which were blocked from conversion to Pinyin. http://www.loc.gov/catdir/pinyin/conversion.html#1   3 Traditional characters: 萬曆, as shown in Table 28.2.   4 Table 28.2 is based on a text image of Xīzì Qíjì available at http://a2.att.hudong.com/99/02/01000 000000000119080227159799.jpg.   5 Middle Chinese is historical Chinese as represented in Qièyùn 切韵 rhyme dictionary (by Lu Fayan 陆法言 in ad 601) and subsequent rhyme dictionaries based on the Qièyùn system.   6 Any writer who uses the word ‘dialect’ in referring to a regional variety of Chinese is duty-bound to define it. The seven major Chinese dialects, Beifanghua 北方话, Wu 吴, Xiang 湘, Gan 赣, Kejia 客家, Min 闽, and Yue 粤, are mutually unintelligible (thus, sometimes referred to as ‘languages’); even within one of them, intelligibility is a matter of degree – some can be mutually unintelligible. The Chinese use the word fāngyán 方言 ‘regional/local speech’, instead. Mair (1991: 7) prefers ‘topolect’ and DeFrancis (1984: 53–67) ‘regionalect’ for dialects mutually unintelligible and ‘dialect’ for a variety within. I will continue to use the term ‘dialect’ loosely for both in this chapter, as the subject of the distinction between the use of language vs. dialect in our context is not an issue.   7 The six categories of characters are: pictographs 象形 xiàngxíng, ideographs 指事 zhǐshì, semantic– phonetic 形声 xíngshēng, compound ideographs 会意 huìyì, derivative cognates 转注 zhuǎnzhù, and phonetic loans 假借 jiǎjiè.

502

Hanyu Pinyin   8 On syllable inventories in Chinese dialects, Chao Yuen Ren has this to say, referring to translation of a same text into several dialects: ‘the richer the inventory of syllables a dialect has, the more concise the translation; and the poorer the inventory, the lengthier the translation’ (Chao 1968: 218). To this we may add that the richer the information load a writing symbol has, the shorter the text. To wit, the Chinese version is always the shortest on an in-flight multilingual passenger instruction sheet. Imagine the opposite of Chinese writing – machine language with an inventory of only two: 0 and 1.   9 To be literate, a Chinese needs to be able to read and write at least 2,000 characters, as listed in Sǎománg Yòngzìbiǎo 扫盲用字表 ‘list of characters to eliminate illiteracy’ (Zhong 1995: 15). According to a survey on the number of characters contained in publications in social and natural sciences, conducted by the National Committee on Chinese Script (Sun 1995:19), the average newspaper or magazine contains 3,000 to 7,000 distinct characters. 10 In the USA alone, there are more than 200,000 students studying Chinese at US schools and colleges annually, according to Asia Society September 2013 Newsletter. Available at: http://asiasociety.org/.

References Chao, Yuen Ren (1968) Language and Symbolic Systems, Cambridge: Cambridge University Press. Chen, Mingyuan (1981) ‘Shùlǐ tǒngjì zài Hànyǔ yánjiū zhōng de yìngyòng’, Zhōngguó Yǔwén 6: 466–74. Chen, Ping (1999) Modern Chinese: History and Sociolinguistics, Cambridge: Cambridge University Press. Chen, Zhangtai (1999) ‘Lùn yǔyán shēnghuó de shuāngyǔzhì’, in Chen Enquan (ed) Shuāngyǔ shuāngfāngyán yǔ xiàndài Zhōngguó, Beijing: Běijīng Yǔyán Wénhuà Dàxué Chūbǎnshè, 47–56. DeFrancis, John (1984) The Chinese Language: Fact and Fantasy, Hawaii: University of Hawaii Press. Dong, Shaowen (1964) Yǔyīn chángshí, Beijing: Wénhuà Jiàoyù Chūbǎnshè. Dong, Tonghe (1968) Hànyǔ yīnyùnxué, Taipei: Xuéshēng Chūbǎnshè. Duanmu, San (2011) ‘Chinese Syllable Structure’, Available from: http://www-personal.umich.edu/ ~duanmu/10ChineseSyllable.pdf. [1 August 2013]. Gao, Tianru (1993) Zhōngguó xiàndài yǔyán jìhuà de lǐlùn hé shíjiàn, Shanghai: Fúdàn Dàxué Chūbǎnshè. Hu, Yushu (1981) Xiàndài Hànyǔ, Shanghai: Shànghǎi Jiàoyù Chūbǎnshè. Huang, Diancheng (1993) Hànyǔ yǔyīnshǐ, Hefei: Ānhuī Jiàoyù Chūbǎnshè. Jiaoyubu yuyan wenzi yingyong guanlisi (2008) ‘Zhìdìng Hànyǔ pīnyīn fāng’àn dàshìjì’, Available from: http://www.china-language.gov.cn/64/2008_3_12/1_64_3400_0_1205300308765.html. [1 August 2013]. Li, Xiaoding (1986) Hànzì de qǐyuán yǔ yǎnbiàn lùncóng, Taipei: Linking Publishing. Lin, Tao (1950) Dingshinghua Sin Wenz (PY: Dìngxínghuà Xīnwénzì), Beijing: Shēnghuó Dúshū Xīnzhī Sānlián Shūdiàn. Liang, Donghan (1959) Hànzì de jiégòu jí qí liúbiàn, Shanghai: Shànghǎi Jiàoyù Chūbǎnshè. Lu, Zhuangzhang (1892) Yīmùliǎorán Chūjiē, reprinted in 1956, Beijing: Wénzì gǎigé Chūbǎnshè. Luo, Changpei (1934) Guóyīn zìmǔ yǎnjìnshǐ, Shanghai: Shāngwù Yìnshūguǎn. Mair, Victor H. (1991) ‘What Is a Chinese “Dialect/Topolect”?’ Available from: http://sino-platonic. org/complete/spp029_chinese_dialect.pdf. [2 August 2013]. Ni, Haishu (1958) Hànyǔ Pīnyīn de gùshì, Shanghai: Shàonián Értóng Chūbǎnshè. Qian, Xuantong (1918) ‘Zhōngguó jīnhòu zhī wénzì wèntí’, in Zhao Jiabi (ed) (1935) Zhōngguó xīnwénxué dàxì, vol. 1 Jiànshè lǐlùn jí, Shanghai: Liángyǒu Túshū Yìnshuá Gōngsī, 141–5. Sima, Qian (ca. 91 bc) Shǐjì, 1982, 2nd ed., Beijing: Zhōnghuá Shūjú. Sun, Jianyi (1995) Xiàndài Hànyǔ zìpín cèdìng jí fēnxī, in Yǔyán wénzì yìngyòng yánjīu lùnwénjí, Beijing: Yǔwén Chūbǎnshè, 17–21. Wang, Li (1964) Hànyǔ Qiǎntán, Beijing: Běijīng Chūbǎnshè. Wu, Yuzhang (1956) ‘Zhōngguó wénzì gǎigé de dàolù’, in Zhōngguó wénzì gǎigé de dìyībù, Beijing: Rénmín Chūbǎnshè, 62–6. Xiaoxue Yuwen Jiaocai Lianhe Bianxiezu (1987) Yǔwén, Beijing: Běijīng Chūbǎnshè, vol. 2. Xīnhuá Zìdiǎn 2002, 9th ed., Beijing: Shāngwù Yìnshūguǎn. Xīzì Qíjì image available from: http://www.discuss.com.hk/viewthread.php?tid=14469086. [1 July 2013]. Xīrú Ěrmùzī image available from: https://archive.org/details/02077325.cn. [2 July 2013]. Yao, Xiaosui (1980) Gǔha` nzì xíngtǐ jiégòu jí qí fāzhǎn jiēduàn, in Shang Chengzuo (ed) Gǔwénzì yánjiū, Beijing: Zhōnghuá Shūjú, 4: 7–39.

503

Lilly Lee Chen Zhan, Bohui (1999) ‘Luèlùn fāngyán, gòngtóngyǔ yǔ shuāngyǔ zhìdù wèntí’, in Chen Enquan (ed) Shuāngyǔ shuāngfāngyán yǔ xiàndài Zhōngguó, Beijing: Běijīng Yǔyán Wénhuà Dàxué Chūbǎnshè, 377–89. Zhang, Tieyi (1991) Qiǎnyì wényánwén tōngjiǎzì shǒucè, Changchun: Jílín Dàxué Chūbǎnshè. Zhang, Xiruo (1956) ‘Dàlì tuīguǎng yǐ Běijīng yǔyīn wéi biāozhǔn de pǔtōnghuà’, in Zhōngguó Wénzì Gǎigé de Dìyībù, Beijing: Rénmín Chūbǎnshè, 50–61. Zhang, Xunru (1953) Běijīnghuà qīngshēng cíhuì, Beijing: Zhōngguó Yǔwén Zázhìshè. Zhōngguó Wénzì Gǎigé de Dìyībù (1956) Beijing: Rénmín Chūbǎnshè. Zhong, Zheming (1995) ‘Yǔyán wénzì yìngyòng yánjiūsuǒ shíniánlái chéngdān hé cānyú yánzhì de guójiā biāozhǔn hé bùwěijí guīdìng jiǎnjiè’, in Yǔyán Wénzì Yìngyòng Yánjīu Lùnwénjí, Beijing: Yǔwén Chūbǎnshè, 1–16.

504

29 Hong Kong Sign Language Gladys Tang The Chinese University of Hong Kong, Hong Kong, China

1. Introduction Linguistic study of Hong Kong Sign Language (HKSL) began in the 1990s with James Woodward (1993). Since then, research on the linguistic properties of HKSL at different linguistic levels has been flourishing. In this chapter, we will first provide an account of the historical development of HKSL. Discussion will center upon the emergence of HKSL in the late 1930s and how it has evolved until today. The second part of the chapter focuses on the descriptions of HKSL at the level of phonology, morphology and syntax.

2.  Origin and history of Hong Kong Sign Language HKSL is the name given to the natural sign language varieties used by deaf people in Hong Kong, who either acquired the language through their deaf parents from birth or through their peers when they were educated in local deaf schools. Similar to many other sign languages around the globe, the emergence and development of HKSL is closely tied to the establishment of deaf schools and the language policies of the government. Before 1935, there were no known historical records on formal deaf education, sign language use, or the existence of a deaf community in Hong Kong. The first deaf school – the Hong Kong School for the Deaf (HKSD) – was established in 1935, and it remained as one of the main deaf schools in Hong Kong until its conversion to a mainstream school in 2004. Despite its strict adherence to oralism, old graduates recalled that deaf children created signs and communicated through signs/gestures extensively among themselves after class or in the dormitory (Sze et al. 2013). These signs/gestures spread quickly from the older students to younger ones within the school, and were the earliest form of a locally developed sign language variety in Hong Kong. The major impetus for sign language development, however, came from the establishment of the Overseas Chinese School for the Deaf and Dumb (hereafter OCSD) in 1948. The school was founded by a deaf signing couple from China who migrated to Hong Kong. A branch school was subsequently set up on the Kowloon Peninsula in 1956. Both schools adopted sign language as the medium of instruction. Since the deaf principal was born in Nanjing but received his education in Shanghai (Overseas Chinese Daily News 1962), the sign language introduced by him and his wife in OCSD was likely to be a mixture of Nanjing and Shanghai signs. 505

Gladys Tang

The use of Nanjing/Shanghai signs spread quickly outside of OCSD. Graduates from HKSD recalled their encounters with deaf students/graduates of OCSD in public places such as sports grounds, commenting that they quickly incorporated the signs from OCSD into their own signing. As quite a number of hearing and deaf signing teachers from OCSD were later employed by other deaf schools or clubs for deaf children run by the government, the OCSD sign language variety continued to spread further. Although OCSD closed down in the 1970s, its deaf graduates continued to use the signs that they learnt from school in the deaf community. Frequently they were regarded as signing models by deaf people from other education backgrounds (Sze et al. 2013). Based on these historical facts, it is safe to conclude that the emergence of HKSL was contributed to the establishment of deaf schools in which sign languages thrived among the deaf students. Also, to a great extent it was influenced by the Nanjing/Shanghai signing variety of Chinese Sign Language. Since then, HKSL took its own course of development, enriched by Hong Kong deaf children and adults who from time to time created new signs to meet their communication needs. The historical relation between HKSL and Shanghai Sign Language was confirmed linguistically in Woodward (1993). He observed that HKSL and Shanghai Sign Language share about 66–68% of the cognates in the basic vocabulary.

3.  Sociolinguistics of Hong Kong Sign Language 3.1.  Signing population There are no official statistics on the number of sign language users in Hong Kong. According to the latest census released in 2008, the number of people with hearing impairment stood at 92,200 (1.3% of total population), among which 8,600 (0.1% of total population) were termed as ‘completely deaf ’ (see Hong Kong Government 2008). Since profoundly deaf children are in general far more likely to be placed in deaf schools than those with a lesser degree of hearing loss, one may assume that they are likely to have had some degree of sign language exposure at deaf schools. When the Hong Kong government began to mainstream deaf children in the mid-1990s, more and more hearing-impaired children studied in ordinary schools regardless of their hearing status. As such, profoundly deaf children under the age of 15 are likely to be mainstreamed and thus non-signing, especially if they are not born to deaf parents. Also, due to misconceptions about sign language, even some deaf children of deaf parents are not exposed to sign language at all. In fact, it is not uncommon to see deaf children being raised by hearing relatives rather than by their own deaf parents. Given this situation, it’s only natural that the potential number of deaf signers has been dropping. Seen in this light, the figure from the government statistics is only a rough estimate of the number of deaf people who could possibly have studied in deaf schools and therefore had some exposure to sign language. The actual figure should be much lower, as the importance of sign language has never been fully recognized in deaf education in Hong Kong. Those privately run deaf schools that once flourished in Hong Kong in the 1950s and 60s all closed down by the 1970s. The remaining deaf schools generally adopted an anti-sign policy. These records boil down to the fact that, within the group of profoundly deaf people, those above the age of 30 are more likely to be users of HKSL, whereas the younger ones are either non-signing or less fluent in signing. That the number of HKSL signers will continue to dwindle has aroused an increasing concern in the deaf community and led to efforts initiated by deaf organizations to promote the use of sign language in the hearing society and to petition for the formal recognition of HKSL for official purposes. 506

Hong Kong Sign Language

3.2.  Local deaf education and emergence of sign bilingualism



As mentioned, owing to the government’s advocacy of mainstream education for the deaf and widespread cochlear implantation, the past two decades have witnessed a substantial drop in the number of deaf students acquiring HKSL. Currently, all except one deaf school have closed down. The remaining school claims themselves to be using total communication to facilitate classroom teaching (i.e. using speech, cued speech, signed Cantonese (see below), HKSL); however, there is still much emphasis on oralism in the education process. While many mainstreamed deaf children grow up without exposure to sign language, the experimental programme ‘Jockey Club Sign Bilingualism and Co-enrollment in Deaf Education Programme’, commenced in 2006, has the function of preserving the development of HKSL in Hong Kong. The programme aims to create a sign bilingual linguistic environment to support the education of both hearing students and a critical mass of deaf students in a mainstream environment, hence indirectly creating a context for bimodal bilingual acquisition for both deaf as well as hearing students (see Plaza-Pust and Morales-López 2008; Marschark et al. 2014). In fact, increasing knowledge of spoken language among deaf signers oftentimes results in their ability to express spoken language through a form of signing which we sometimes refer to as a ‘manually coded spoken language’, which is, in the context of Hong Kong, signing in Cantonese. This artificial system adopts the Cantonese word order but superimposes on it with the lexical signs of HKSL unsystematically. Unlike Signed Exact English (SEE) in the American context, where each individual morpheme is supposedly given an artificial signing form, in signing in Cantonese, many verb particles or sentence-final particles of in (1)): Cantonese are missing (see





(1) ngo gam-yat m-guai-heoi sik-faan lo. (Cantonese) Pro adv neg-prt-v v prt (signing in Cantonese) IX-1 TODAY NOT GO HAVE_A_MEAL . TODAY IX-1 HAVE_A_MEAL NOT. (HKSL) ‘I am not going out to eat today’ Systematic research on how deaf children develop sign or spoken language remains scanty. Recently some case studies have been conducted on how a deaf child, CC, acquired HKSL. The data display a gradual increase in code-blending involving Cantonese and HKSL (Fung 2010; Tang and Fung 2009). As shown by CC’s utterance in (2), his production of HAVE_ AN_IDEA in HKSL is blended with ‘gaau nei’ (‘teach you’) in Cantonese. Interestingly, the tag question ‘hou-m-hou’ (‘good-not-good’) followed by a sentence particle ‘aa3’ offers the clue that the code-blended utterance is based on Cantonese syntax. (2) HKSL: IX-1 HAVE_AN_IDEA 0TEACH2__ Can: ngo gaau nei hou-m-hou aa3 pro_1 teach pro_2 good-not-good sfp ‘I got an idea, (I) teach you, alright?’ (CC 5; 8.24) Code-blending is evidential of sign bilingualism in deaf children, demonstrating that they have the flexibility of processing linguistic information via two independent but interacting modalities. In sum, although oralism dominates local deaf education, there is increasing visibility of the importance of sign bilingualism in deaf children’s language development. 507

Gladys Tang

4.  Linguistics of HKSL 4.1.  Lexical variation At present, HKSL continues to be used by the majority of deaf people graduating from deaf schools. Nonetheless, signing proficiency varies a great deal across signers. Due to a lack of formal sign language teaching in deaf education, deaf children in deaf schools very often resort to self-created signs/gestures to meet their daily communication needs. In addition, the level of residual hearing, exposure to spoken language, and language policy in deaf schools may have an effect on the degree of Chinese influence in the formation of lexical signs. These factors gradually give rise to lexical variation that cuts across different age groups and education backgrounds. The exact degree of variation of individual lexical items appears to be a function of frequency of use: the less frequent a lexical item is used in daily communication, the more variable it will be. Below are four different signs for BEE that may offer a glimpse of lexical variation in HKSL (see Figures 29.1a–29.1d). Figure 29.1a is used by signers aged below 30 from the Hong Kong School for the Deaf, with mouthing of Cantonese word ‘mat fung’; Figure 29.1b is used by signers aged below 30 from several other deaf schools, with mouthing of Cantonese word ‘mat fung’; Figure 29.1c is used by signers aged above 45 who graduated from the OCSD, and Figure 29.1d is used by signers aged above 45 who graduated from HKSD. As shown in the examples of BEE, school background is a key factor that contributes to lexical variation. Another noteworthy observation is that older signers tend to use more classifiers and compound forms to express lexical meaning in a more precise manner. In contrast, younger signers tend to use a general form that is potentially ambiguous among several possible meanings, and rely heavily on Cantonese mouthing in disambiguation, e.g. the manual form of Figure 29.1a may also mean FLY or BUTTERFLY; the manual form of Figure 29.1b may also mean MOSQUITO. The heavy use of mouthing among the younger signers can be attributed to the improvement of hearing aids as well as the predominance of oralism in deaf education since 1970s. In recent years, more and more loans signs are found in HKSL. Some have a Chinese origin, due to an increase in the number of deaf immigrants from different parts of mainland China; while others are brought into the signing community through local signers’ increasing interactions with deaf people from other countries. The sign WATER in HKSL, for instance, is probably a loan sign from American Sign Language (ASL).

Figure 29.1a  BEE (variation 1)

508

Hong Kong Sign Language

Figure 29.1b  BEE (variation 2)

Figure 29.1c  BEE (variation 3)

Figure 29.1d  BEE (variation 4)

509

Gladys Tang

4.1.1.  Other signing systems Fingerspelling and character signs can be viewed as products of language contact with the spoken language in the surrounding hearing communities. Both are observed in HKSL but are highly restrictive in use. Unlike Chinese Sign Language where there is an extensive use of fingerspelling or initialized signs with alphabetic handshapes, probably due to Romanization (i.e. Pinyin) of spoken Mandarin, HKSL makes use of the one-handed fingerspelling system of ASL, with some minor modifications in movement and orientation. However, few signs in HKSL are initialized with alphabetic handshapes or simply fingerspelt. Also, there is no evidence for the use of the two-handed British fingerspelling system ever among Hong Kong deaf signers, despite Hong Kong being a colony of Britain for more than 150 years. One possible reason is that oralism in deaf education by the British government had prevented the introduction of British Sign Language in local deaf education. Since fingerspelling is not widely used in Hong Kong, phonological processes commonly observed in the lexicalization of fingerspelt signs in ASL are not attested in HKSL (cf. Brentari and Padden 2001). Character signs that visually mimic the shape of Chinese written characters are not many. They generally fall into two types: (i) signs whose handshape configuration visually represents the radicals and strokes of a Chinese character; and (ii) signs in which one hand represents the shape of a part of the character and the other hand traces how the remaining part should be written. Chinese-based character signs are also reported in Chinese Sign Language (Zhong Guo Longren Xiehui 1988), Taiwan Sign Language (Ann 1998) and Japanese Sign Language. Some of the character signs in HKSL can also be found in these three sign languages. Character signs commonly used by HKSL signers include: PERSON (see Figure 29.2a), NORTH, LABOUR, FIELD, CHINESE, FAIR (see Figure 29.2b), INTRODUCE (see Figure 29.2c), BIG, WEST, CHEESE (partial representation of the Chinese character), MAINLAND-CHINA (partial representation of the character that stands for the Communist Party).

(a)

(b) (c)

Figure 29.2a PERSON 人 Figure 29.2b  FAIR 公平 Figure 29.2c  INTRODUCE 介绍

510

Hong Kong Sign Language

4.2. Phonology 4.2.1.  Headshape, location and movement as phonological parameters As mentioned previously, the linguistic study of HKSL began in the early 1990s when James Woodward attempted to examine the varieties of HKSL. After this initiative, Tang (2007) documents about 1,800 signs commonly used by deaf signers in Hong Kong. In the absence of research precedence, Tang adopts a common assumption in sign phonology that movement, location and handshape with their associated features can be contrastive. For instance, signs showing contrastive orientation change under the movement parameter like BREAK_OFF_A_ RELATION and INCOMPATIBLE are both two-handed signs with symmetrical 1-handshapes (see Figure 29.3a and 29.3b). However, BREAK_OFF_A_RELATION involves supination of the wrist joint. INCOMPATIBLE requires repeated flexion and extension of the wrist joint.

Figure 29.3a  BREAK_OFF_A_RELATION

Figure 29.3b  INCOMPATIBLE

511

Gladys Tang

Figure 29.4a  PIG

Figure 29.4b STUBBORN

Contrastive locations are also quite common in HKSL. The contrasts may be within and between body regions, as specified in Brentari (1998). In HKSL, RED and IMPOLITE are signed on a different location on the face, and PIG and STUBBORN are across different body regions (see Figure 29.4a and 29.4b). Tang (2007) identifies 60 contrastive handshapes. For instance, BE_DETERMINED and SET_OFF are articulated in neutral space with a downward path movement and flexion of the wrist joint. However, BE_DETERMINED assumes an S-handshape ( ) and SET_OFF a 1-handshape ( ) (see Figures 29.5a and 29.5b). Handshapes in sign language phonology are conceptualized as a ‘daughter’ of the articulator node capturing the number and joint position of the selected and unselected fingers, as well as the thumb. Note that the thumb is considered to be independent of the selected

Figure 29.5a  BE_DETERMINED

Figure 29.5b SET_OFF

512

Hong Kong Sign Language

Figure 29.6  THANK_YOU

fingers node. In HKSL, the thumb can be selected independently, as in the sign THANK_YOU, with repeated flexion at the non-based joint (see Figure 29.6). Handshapes for the dominant hand like ‘B’ (w) , ‘1’ ( ) , ‘5’ ( ) , and ‘S’ ( ) are most productive in HKSL, in line with the claim that these are unmarked handshapes in many sign languages. For two-handed signs that do not share similar handshapes, ‘B’ remains the most common non-dominant handshape. Note that ‘ ’ ( ) is the next most common nondominant handshape, as observed in a large number of lexicalized classifier predicates. This sub-lexical unit on the non-dominant hand seems to originate from the Y-handshape, which is used as a classifier to refer to humans or animate entities. It seems that the process of lexicalization has led to phonological deletion of the pinky finger, to cope with the requirement of less complex handshapes for the non-dominant hand. As a result, this handshape loses some of the iconicity of a legged animate entity, as in Figures 29.7a and 29.7b. ( ), ( ), and ( ) are observed in Other relatively less common handshapes like HKSL. However, they are less productive and are mainly adopted by signs that use them on

Figure 29.7a scold+CL_sem:human_entity

Figure 29.7b SCOLD

513

Gladys Tang

the dominant hand (see Figures 29.8a–29.8d). They are seldom adopted for the non-dominant handshapes, due to their complex phonological configuration.

4.2.2.  Feature Analysis Adopting the Prosodic Model developed by Brentari (1998), who proposes that signs may be analyzed in term of bundles of features organized hierarchically into a feature geometry of phonological units, Tang (2007) attempts to document the features manifested by the signs of HKSL. Features in Brentari’s model are either inherent or prosodic. Inherent features refer to those underlying properties that are specified once per lexeme in the lexicon and are realized simultaneously. These features characterize the properties of handshape, place of articulation and orientation. Prosodic features capture the movement dynamics of the sign and may be expressed in terms of changes in handshape, setting, and orientation. Adopting this framework, one may account for the phonological complexity of the least productive handshapes such as ( ), ( ), and ( ) (see Figures 29.8a–29.8d). For instance,

Figure 29.8a TEN

Figure 29.8b  DIVIDE_EVENLY

Figure 29.8c DECREASE

514

Hong Kong Sign Language

Figure 29.8d  INCREASE

( ) originates from a character sign TEN. ( ) violates the Selective Finger Constraint (Mandel 1981), which stipulates that selected fingers are in one finger position. ( ) involves the index and middle fingers as selected fingers but each with a different finger specification. The middle finger is [extended] but the index finger is [flexed] at the non-base joint. Moreover, these two selected fingers are [non-spread]. For ( ), the status of thumb as selected finger is not clear if defined in terms of the inherent properties of sign articulation. In this case, the radial side of the index finger becomes the place of articulation and the tip of thumb traces alongside the radial side of the index finger, in an outward direction meaning DECREASE and an inward direction meaning INCREASE. As for place of articulation, HKSL makes a heavy use of the three-dimensional, neutral space (i.e. X-vertical, Y-horizontal and the Z-midsagittal planes) for one-handed and twohanded signs (see Figures 29.9a–29.9c). Out of these 1,800-odd signs, about 1,000 of them

Figure 29.9a  FAR (X-plane)

515

Gladys Tang

Figure 29.9b  CLOUD (Y-plane)

Figure 29.9c  SIGN LANGUAGE (Z-plane)

Figure 29.9d  FACE-PRETTY

Figure 29.9e  HEALTHY

516

Hong Kong Sign Language

use one of the three planes as place of articulation. Among them, the Y-plane is the most common and the Z-plane the least, in between is the X-plane. The head and its parts like forehead or mouth constitute the next most common places of articulation. Only about 200 signs use the torso or body parts like the shoulder or waist. In fact, 111 signs use the entire torso and 46 signs use the entire head as place of articulation rather than specific parts of the region (see Figures 29.9d and 29.9e). Another noticeable place of articulation is the non-dominant hand. About 406 signs use the non-dominant hand as place of articulation where the finger and palm are the two most common handparts on the non-dominant hand (see Figure 29.7a above). As said, movement features capture the dynamic properties of the sign. Mak and Tang (2012) attempt to analyze the co-occurrence of movement types and features in HKSL. In their analysis, movement in HKSL may simply involve an epenthetic movement to a hold at the place of articulation (i.e. No-MV), resulting in a lack of lexical movement in the entry. In the database, 101 one-handed and 58 two-handed signs do not require lexical movements (see Figure 29.10). Other lexical movements may be simplex or complex. A simplex movement involves independent changes of a Path movement or a Local movement. Path movement defines the two settings of motion. Local movements involve an orientation change of the wrist position (i.e. pronation or supination of the write joint), or an aperture change (i.e. open or close of selected finger). Another phonological analysis of HKSL focuses on the feature [contact]. Wong (2004) argues that it is a phonological feature predictable by other movement features. Path movement with a direction feature ‘to’ or ‘from’ a place of articulation predicts initial or end [contact] (see Figure 29.11a). [Trace] requires continuous contact at place of articulation (see Figure 29.11b). Finally, a setting change at a place of articulation requires two points of contact with an epenthesis in between (see Figure 29.11c). Contrary to Brentari’s observation that contact is predicted by path features alone, in HKSL, orientation change may also predict ‘grazing’ contact at the place of articulation (see Figure 29.11d).

Figure 29.10  LOVE-SOMETHING-ARDENTLY

517

Figure 29.11a STOP

Figure 29.11b  TRY

Figure 29.11c DOCTOR

Figure 29.11d SHARE

518

Hong Kong Sign Language

Being phonological units, handshape, location and movement may undergo assimilation in sign articulation. Assimilation occurs in compound formation involving change of handshape, especially in terms of the change of selected fingers and their position, justifying the phonological status of handshape features in sign language. For instance, the sign THINK^OLD (‘old-fashioned thinking’) originally involve a 1-handshape, followed by -handshape ( ) which also selects the thumb with the feature [extended] (Figures 29.12a and 29.12b). In signing, the thumb together with this feature will be assimilated to 1-handshape, a process of anticipatory assimilation. Phonological reduction in the formation of compounds is also observed and it is particularly obvious in the movement parameter. GAMBLE in HKSL requires repeated tapping against the back of the thumb of the non-dominant hand by the back of the index finger. When this sign enters into a compound with HORSE, glossed as BETTING_ON_HORSES, the trill movement is deleted and replaced by a single path movement to the back of thumb.

Figure 29.12a  THINK and OLD (citation form)

Figure 29.12b  THINK^OLD (assimilated thumb)

519

Gladys Tang

5.  Morphology of HKSL 5.1. Compounds There are endocentric compounds in HKSL, which consist of a head that determines the categorical status and a modifier that restricts its meaning. Sometimes, the head stems from a superordinate term. For instance, common nouns for institutions are usually formed with HOUSE as the head, such as LEARN^HOUSE (‘SCHOOL’) (Figure 29.13) and LAW^HOUSE (‘COURT’). Common nouns for meat consist of name of the animals and MEAT, e.g. PIG^MEAT (‘PORK’) (Figure 29.14) and BEEF^MEAT (‘BEEF’).

Figure 29.13  LEARN^HOUSE

Figure 29.14  PIG^MEAT

520

Hong Kong Sign Language

Figure 29.15 ROCKET

There are exocentric compounds where no member of the set is the head. Common examples are FATHER^MOTHER (‘PARENTS’), GIVE-BIRTH^MALE (‘SON’), GIVE-BIRTH ^FEMALE (‘DAUGHTER’), RESPONSIBLE^CHOP (‘GUARANTEE’), or LEARN^KNEEL (‘APPRENTICE’). Note that some exocentric compounds in HKSL are influenced by Cantonese. One such example is ‘SANDWICH’, which is made up of THREE^TRIANGLE, where the first sign of the compound is ‘three’, which is a direct translation from the first character of a three-character word ‘saam1 men4 ji6’ (‘sandwich’) in Cantonese. The compounds discussed above are signed sequentially. However, there are simultaneous compounds in HKSL. A typical example is WHAT_MONTH_WHAT_DATE, as in ‘When is your birthday?’. Other examples may be found in lexicalized classifier predicates constituted by two distinct morphemes. Note that their categorical status may change, usually from a verb to a noun and vice versa. To be qualified for a simultaneous compound in the current analysis, we assume that there can be a difference in handshape and movement while there should be one location, a result of lexicalization. For instance, the sign ROCKET stems from a classifier predicate ‘shoot_up+CL_sass:cylindrical_object^fire_emit+CL_sass: 3D_volume’. The dominant hand is represented by a size and shape specifier denoting the rocket while the non-dominant hand denotes the size and shape of the fire emitting at the bottom of the rocket. The upward motion of the dominant hand and the opening of the selected fingers of the non-dominant hand combined simultaneously to refer to a rocket (see Figure 29.15). As a lexical item, this sign violates Battison’s (1978) Symmetry Condition for the phonological formation of two-handed signs, in the sense when both hands move independently, they will have the same handshape, place, orientation and movement. Apparently, in this sign, both handshapes as well as movement with each are different.

5.2.  Derivational morphology 5.2.1.  Noun-verb pairs In ASL, reduplication is adopted to distinguish nouns from verbs. In HKSL, no such systematic pattern of movement alternation is observed for noun–verb pairs. However, it is possible for nouns like KEY, SPOON, NOODLES, BISCUIT and BREAD that are normally signed with 521

Gladys Tang

repeated or trill movement to adopt a single movement when used as a verb. As such, the verb sign can be overlaid with a non-manual manner adverbial, which is not possible otherwise, as in (3): (3) DOORa be_located+CL_sass:doora  IX-1  KEY  TAKEb, __________________________________ ‘carefully’ insert+CL_handle:key^turn+CL_sass:key. ‘The door is here; I get the key, insert it (into the hole) and turn it carefully’ In (3), the noun KEY has a trill movement; but the predicates ‘insert (key)’ and ‘turn (key)’ each require a single movement and both are scoped by the non-manual manner adverb ‘carefully’. However, it is not always easy to distinguish noun-verb pairs in HKSL and syntactic word order will be the clue for determining their grammatical status.

5.3.  Inflectional morphology Although character signs are few in number in HKSL, some of them are used fairly frequently by deaf people so that they undergo grammaticalization and acquire certain phonological and morphological properties of lexical signs in the core lexicon. Negative incorporation is observed with some of these signs. For example, where the sign INTRODUCE has acquired a second meaning FEEL-ANNOYED, it may carry a suffix NOT to mean NOT_FEELANNOYED (Figure 29.16a). The character sign FAIR is a static sign without any lexical movement. The sign UNFAIR is derived morphologically from FAIR by incorporating an up-and-down alternating movement, indicating the abruption of equality implied through the side-by-side placement of the two manual articulators in the sign FAIR (Figure 29.16b). In fact, some of the character signs like INTRODUCE also acquire verb agreement morphology, a phenomenon which we shall now turn to. Similar to other sign languages, verbs of HKSL can be broadly categorized into plain verbs, agreeing verbs, spatial verbs and classifier predicates. In this section, we will focus on agreeing verbs and spatial verbs. Lam (2003) argues that if location marking is controlled, verb agreement in HKSL allows a three-way distinction – first, second and third person. Also, similar to other sign languages, verb agreement is optional, meaning that under certain circumstances, the citation form of an agreeing verb can be used. Typical examples of agreeing verbs are

Figure 29.16a  NOT_FEEL-ANNOYED

522

Hong Kong Sign Language

Figure 29.16b  UNFAIR

GIVE or SEE. When the subject is second person or the object is first person, person agreement has to be marked. Therefore, 3GIVE1, 2GIVE3, and 2GIVE1 show obligatory verb agreement while 0GIVE0 may be used in other contexts. INTRODUCE as a character sign also displays this morphological property, as shown in (4): (4) MOVIE GOOD, IX-2 2INTRODUCE1, 0THANK YOU2! ‘The movie is good; thank you for introducing (it) to me’ Verb agreement in HKSL can also be marked by palm or finger orientation, as the sign HELP or BORROW shows. In (5) and (6), there is no observable path movement to mark the direction from subject and object. In (5), the facing of the palm with the sign HELP indicates a third person subject and first person object. In (6), the direction into which the selected finger (i.e. index) points indicates that the indirect object is second person. (5) MALE IXa  3HELP1 ‘The man over there helps me’ (6) MALE IX-a 3BORROW2  MONEY TWO-HUNDRED ‘The male over there borrows $200 from you’ Another type of inflectional morphology stems from the location of the entity in space. In HKSL, spatial verbs must agree with the location of referents in space, as in (7): (7) MALE CUP TWO PUTa PUTb ‘The man put the cups here and here’ In (7), reduplication (i.e. signing the verb PUT twice) marks two different locations in space. As we shall see, the spatial verb PUT assumes a -handshape ( ) with the palm facing down, regardless of the objects being handled. As a lexical sign, this handshape is assumed to be a meaningless, phonological unit. However, there are certain signs in which the handshapes may become morphemic units. 523

Gladys Tang

6.  Classifier predicates Classifier predicates are common among the sign languages of the world. They are morphosyntactic structures comprising a handshape affix and a movement affix. The classifier handshape encodes the arguments in the predicate. Following the conventional literature, classifier handshapes can be categorized into semantic, handle, size and shape specifiers and bodyparts. The movement affix encodes the predicate root with the associated temporal and aspectual properties of an event. In other words, with a classifier verb, each phonological feature of the sign may be analyzed morphosyntactically as representing a meaningful unit. For instance, the handshape of the spatial verb PUT mentioned above may have in parallel other forms of classifier verbs using different handshapes to encode an argument of ‘put’, the object of which represents different noun categories (e.g. put a ring/cup/towel/piece of paper). Also, the locus in space where the classifier verb is signed denotes the location of the object being manipulated. In the literature, some researchers argue that the handshape classifiers are agreement markers (Glück and Pfau 1998; Zwitserlood 2003). In HKSL, systematic study has been conducted in motion and location predicates (Tang 2003; Tang and Gu 2006), causative and unaccusative predicates (Lau 2002), and simultaneous constructions where two independent predicates are signed each on the dominant and the non-dominant hand (Tang et al. 2006). It has been suggested that iconicity is prevalent in classifier predicates (Tai 2005), and classifier handshapes are a case in point. In HKSL, as well as in many Asian sign languages like Chinese Sign Language, Taiwan Sign Language and Thai Sign Language, the Y-handshape may be used as a classifier to refer to a legged animate entity in a motion or location predicate (see Figures 29.17a and 29.17b). Its iconic nature is confirmed by native signers that in its upright orientation, the thumb refers to the head, the pinky the legs and the handpart the chunk of the body. The same handshape may also refer to a size and shape specifier in other classifier predicates, e.g. fall_down+CL_sass: telephone_receiver. Lau (2002) and Tang and Gu (2006) observe that causation in HKSL requires a composite of two subevents – cause and result – expressed by two classifier predicates that occur in sequence (i.e. (8)). Recently, Lau (2012) argued that the two classifier predicates are morpho­ syntactically serial verb constructions, where the two predicates are sequentially concatenated

Figure 29.17a stand+CL_sem:human

Figure 29.17b lie+CL_sem:human

524

Hong Kong Sign Language

and the second predicate is a complement of the first, reflecting that the first predicate is a main event and the second a subevent.   (8) HOUSEa LIGHTNING  strike+CL_sass:lightning^CL_sass:housea  collapse+CL_sass:housea ‘The lightning strikes the house down’ In fact, classifier predicates in HKSL can either be sequentially or simultaneously concatenated in coordinate constructions (Tang et al. 2006; Tang and Lau 2012), probably due to the facility of having two manual articulators with each representing an independent event. (9) shows a person drinking and driving at the same time. This sign violates all phonological constraints of two-handed signs proposed by Battison (1978) where the two independent classifier handshapes each articulates a different movement, hence two different events. (10) shows a simultaneous event in which a person eats chips and drinks cola simultaneously; however, these coordinated events are expressed sequentially although the eating and drinking events are perceived as simultaneous.   (9) MALE drive+CL_handle:wheel, WINE drive+CL_handle:wheel^drink+  CL_handle:bottle. ‘The man drives and drinks from the bottle’ (10) MALE CHIPS COLA eat+CL_handle:chips, drink+CL_handle:can,  eat+CL_handle:chips, drink+CL_handle:can +++ ‘The man eats chips and drinks cola’

7. Syntax 7.1.  Word order Statistically, Subject–Verb–Object (SVO) is the most frequently attested word order in spontaneous discourse in HKSL, regardless of verb or clause types (Sze 2008). SVO order is particularly preferred if the subject and object are semantically reversible (cf. Fischer 1975) and if the verb involved is a plain verb that provides no morphological cues for distinguishing grammatical relations (Sze 2000, 2003), as shown in (11): (11) PETER(name-sign) LIKE JENNY(name-sign) S V O ‘Peter likes Jenny’ If the subject and object are semantically NOT reversible, either SOV or SVO are observed. However, some plain verbs with a negative meaning tend to occur in a clause-final position. Under these circumstances, a SOV order results because NegP in HKSL is head final and we assume that the negative feature triggers verb raising to this position, as shown in (12) and (13): (12) FATHER PIZZA DISLIKE S O V ‘(My) father doesn’t like pizza’ 525

Gladys Tang

(13) IX-deti BOY SIGN LANGUAGE LACK_SKILL S O V ‘That boy lacks the skills of sign language’ As for word order with agreement verbs, if the subject and object referents are each assigned a locus (marked by subscripts ‘a’ and ‘b’ in the data) in the signing space, and if the agreement verb involved carries inflectional morphology, word order becomes flexible, as such either SOV or OSV will result (see (14) and (15)). For three-place predicates involving signs like GIVE or BORROW, the indirect object may occur sentence-finally (16). (14) IXa BOY IXb GIRL IXa aHITb S O V ‘The boy hit the girl’ (15) IXa GIRL IXb BOY bHITa O S V ‘The boy is hitting the girl’ (16) IXa BOY MONEY aGIVEb IXb GIRL O V IO S ‘The boy gives money to the girl’ For predicates involving classifiers, an SOV word order is usually found if a classifier denoting the object referent is incorporated into the verb root, as in (17a) and (17b). Note that in (17a) role shift is necessary and the classifier predicate is signed at a locus for DOG near the right side of the signing space rather than the left or the centre. In (17b), role shift is optional because the handle classifier denotes a volitional, agentive subject, which in this sentence refers only to BOY. ____rs(right) ____rs(left) __________________________________rs(right) (17a) DOG BOY lick+CL_bodypart:tongue^CL_sem:boy S O CL ‘The dog licks the boy’ DOG wash+CL_handle:rag^CL_sem:dog (17b) BOY S O V ‘The boy washes the dog (with a rag)’ To conclude this section, in HKSL, the basic word order is SVO for plain verbs, but word order variation is observed when the verbs have a negative meaning, have to agree with the person in subject or object, or when the verb root is incorporated into a classifier predicate.

7.2. Interrogatives In HKSL, Tang (2006) observes that wh-adjuncts are consistently clause-final and wh-arguments are either in situ or clause-final, as shown in (18)–(20). That wh-arguments are in situ may be due to Cantonese influence, but most signers prefer the wh-arguments to be in a clause-final position. Wh-interrogatives may be marked non-manually with brow furrow but brow raise 526

Hong Kong Sign Language

is observed occasionally. It may scope over the wh-word only, sometimes the predicate or the entire sentence. However, the wh-interrogative becomes illicit when it is not marked with any wh-non-manuals, as in (21). _______wh-q (18) ELDER_BROTHER BUY BOOK  WHEN? ‘When did elder brother buy the book?’ _____________wh-q (19) ELDER_BROTHER  BUY  WHAT? ‘What did elder brother buy?’ ____________________wh-q (20) BUY  BOOK  WHO? ‘Who bought a book?’ (21) *ELDER_SISTER GET_MARRIED  WHAT_MONTH_WHAT_DATE? ‘What month and date will elder sister get married?’ Polar questions share the same word order of the corresponding declarative sentences. They are obligatorily marked by brow raise which scopes over the entire question or the predicate only, as in (22) and (23): ________________________________________Y/N-Q (22) ELDER_BROTHER  GO  WATCH_MOVIE? ‘Did elder brother go to watch a movie?’ _______Y/N-Q (23) IXa MALE  WORK  FINISH? ‘Has that man finished working?’

7.3. Negation Lee (2006) identifies a few manual negators in HKSL, the most common ones being NO, NOT, NOT_HAVE and NOT_YET. NO and NOT are manually similar but differ in the use of non-manuals. NO is mouthed with ‘m-hai’ (see 24a), but NOT is signed with pressed lips (see 24b). ___mouthing ‘m-hai’ (24a) IX-3  TOMORROW  FLY  NO ‘It is not the case that he is flying tomorrow’ ____neg (24b) IX-3  TOMORROW  FLY  NOT ‘He will not fly tomorrow’ These manual negators are syntactically clause-final, as in (25) and (26). (25) BOSS  BUY  CAR  NOT_HAVE ‘The boss didn’t buy any car’ 527

Gladys Tang

_________neg (26) IX-1  BOOK  READ  NOT-YET ‘I haven’t read the book’ In HKSL, manual negators are obligatory and non-manual headshakes are optional, as shown in (25) (cf. Lee 2006). Unlike other sign languages such as ASL or BSL, it is illicit to mark negation with a headshake only in HKSL. When a headshake occurs, it just scopes over the negator and seldom spreads to other constituents preceding the negator.

7.4. Modals As functional categories are head-final in HKSL, modals, like the negators discussed above, consistently occur in a clause-final position (see (27) and (28)). Note that unlike CAN, NEED in HKSL may also be preverbal, probably due to Cantonese influence (Lam 2009). There are negative modals like WILL_NOT or MAY_NOT in HKSL and syntactically they occupy the head of NegP rather than ModP (see (29) and (30)). (31) shows that NegP occupies a higher position than ModP in HKSL, as shown by the negator NO, which is preceded by the modal. (27) IX-3  WORK_HARD  NEED ‘He needs to work hard’ (28) IX-1  CYCLE  ONE_HANDED  CAN ‘I can cycle one-handed’ (29) IX-3  TOMORROW  COME_BACK  WORK  WILL_NOT ‘She won’t come back to work tomorrow’ __________neg (30) IX-3  CAR^COMPETITION  WIN  MAY_NOT ‘He may not win the car race’ (31) IX-3  CYCLE  ONE_HANDED  CAN  NO ‘It is not the case that he can cycle one-handed’ In this section, we outline some basic word order issues of HKSL and suggest that functional categories are head-final, as observed by the fact that the negators and modals occur in a clause-final position. Just like the modals and negators, wh-words are also clausefinal. However, their syntactic position is not yet confirmed and we leave it to future research.

8. Conclusion In this chapter, we have outlined some basic facts about the historical development of HKSL and the sociolinguistics of its community of signers. We have discussed briefly the oralist situation in deaf education during Hong Kong’s colonial era and some latest developments to preserve the language through reintroducing it in deaf education. As a minority language, HKSL has not been recognized as an official language for the deaf in Hong Kong. However, linguistic study of HKSL has begun to unravel its phonological, morphological and syntactic properties, documented in dictionaries and research papers. Further research on language acquisition by deaf children in HKSL, Cantonese and written Chinese will certainly shed 528

Hong Kong Sign Language

more light on the linguistic capabilities of deaf children in becoming sign-bilingual in a context like Hong Kong, which enables educators to design an appropriate curriculum in deaf education.

Acknowledgements The author would like to acknowledge the contributions of Felix Sze, Kenny Chu and Connie Lo in sign language research in Hong Kong. Their dedication has led to a rapid expansion of sign language literature in the territory.

References Ann, J. (1998) ‘Contact between a Sign Language and a Written Language: Character Signs in Taiwan Sign Language’, in Ceil Lucas (ed) Pinky Extension and Eye Gaze: Language Use in Deaf Communities, Washington, D. C.: Gallaudet University Press, 59–99. Battison, R. (1978) Lexical Borrowing in American Sign Language, Silver Spring, MD: Linstok. Brentari, D. (1998) A Prosodic Model of Sign Language Phonology, Cambridge, MA: MIT Press. Brentari, D. and C. Padden (2001) Native and Foreign Vocabulary in American Sign Language: A Lexicon with Multiple Origins, Mahwah, NJ: Lawrence Erlbaum, 87–119. Fischer, S. (1975) ‘Influences on Word Order Change in American Sign Language’, in Charles Li (ed) Word Order and Word Order Change, Austin, TX: University of Texas Press, 3–25. Fung, C. (2010) ‘Code-blending in Early Hong Kong Sign Language: Some Preliminary Results from a Case Study’, Paper presented at 2010 Annual Research Forum of the Linguistic Society of Hong Kong, The Chinese University of Hong Kong, Hong Kong. Glück, S. and R. Pfau (1998) ‘On Classifying Classification as a Class of Inflection in German Sign Language’, in T. Cambier-Langeveld, A. Lipták and M. Redford (eds) Proceedings of ConSOLE 6, Leiden: Sole, 59–74. Hong Kong Government (2008) Social Data Collected via General Household Survey: Special Topic Report No. 48 – Persons with Disabilities and Chronic Disease, The Census and Statistics Department. Lam, S. W. Z. (2003) ‘Verb Agreement in Hong Kong Sign Language’, MPhil Thesis, The Chinese University of Hong Kong. Lam, S. W. Z. (2009) ‘Early Phrase Structure in Hong Kong Sign Language: A Case Study’, Doctoral Dissertation, The Chinese University of Hong Kong. Lau, S. Y. P. (2002) ‘Causative Alternation in Hong Kong Sign Language’, MPhil Thesis, The Chinese University of Hong Kong. Lau, S. Y. P. (2012) ‘Serial Verb Constructions in Hong Kong Sign Language’, Doctoral Dissertation, The Chinese University of Hong Kong. Lee, J. Y. F. (2006) ‘Negation in Hong Kong Sign Language’, MPhil Thesis, The Chinese University of Hong Kong. Mak, J. and Tang, G. (2012) ‘Movement Types and Feature Organization in Hong Kong Sign Language’, in R. Channon and H. van der Hulst (eds) Formational Units in the Analysis of Signs, Sign Language Typology Series, No. 3, Nijmegen: Ishara Press. Mandel, M. A. (1981) Phonotactics and Morphophonology in American Sign Language, Berkely: University of California Press. Marschark, M., Tang, G., and H. Knoors (2014) Bilingualism and Bilingual Education, Oxford: Oxford University Press. Overseas Chinese Daily News (1962) Hong Kong Year Book (1962), Hong Kong: Overseas Chinese Daily News. Plaza-Pust, C. and E. Morales-López (2008) ‘Sign Bilingualism: Language Development, Interaction, and Maintenance in Sign Language Contact Situations’, in Carolina Plaza-Pust and Esperanza Morales-López (eds) Sign Bilingualism, Amsterdam and Philadelphia: John Benjamins 333–379. Sze, F. (2000) ‘Space and Nominals in Hong Kong Sign Language’, MPhil Thesis, The Chinese University of Hong Kong. Sze, F. (2003) ‘Word Order of Hong Kong Sign Language’, in Anne Baker, Bebbie van den Bogaerde, and Onno Crasborn (eds) Cross-linguistic Perspectives in Sign Language Research: Selected Papers from TISLR 2000, Hamburg: Signum Verlag, 163–191.

529

Gladys Tang Sze, F. (2008) ‘Topic Constructions in Hong Kong Sign Language’, Doctoral Dissertation, University of Bristol. Sze, F., Lo, C., Lo, L., and Chu, K. (2013) ‘Historical Development of Hong Kong Sign Language’, Sign Language Studies 13(2): 155–85. Tai, J. H. Y. (2005) ‘Modality Effects: Iconicity in Taiwan Sign Language’, in Dah-an Ho and Ovid J. L. Tzeng (eds) POLA FOREVER: Festschrift in honor of Professor William S-Y. Wang on his 70th Birthday, Taipei: Academia Sinica, 19–36. Tang, G. (2003) ‘Verbs of Motion and Location in Hong Kong Sign Language: Conflation and Lexicalization’, in Karen Emmorey (ed) Perspectives on Classifier Constructions in Sign Languages, Mahwah, NJ: Lawrence Erlbaum, 143–65. Tang, G. (2006) ‘Questions and Negation in Hong Kong Sign Language’, in Ulrike Zeshan (ed) Sign Language Typology: Interrogatives and Negation, Nijmegen: Ishara Press. Tang, G. (2007) Hong Kong Sign Language: A Trilingual Dictionary with Linguistics Descriptions, Hong Kong: Chinese University of Hong Kong. Tang, G. (2009) ‘Assessment of Chinese Grammatical Competence of Hong Kong Deaf Children’, Paper presented at Workshop on Milestones in the First Language Acquisition of Chinese, Language Acquisition Laboratory, The Chinese University of Hong Kong, 29–30 December. Tang, G. and Fung, C. (2009) ‘Code-blending in Early Hong Kong Sign Language’, Paper presented at AHRC Workshop One Brain, Two Languages: Bridging Neuroscience and Linguistics on Intramodal and Cross-modal Bilingualism, Bangor. Tang, G. and Gu, Y. (2006) ‘Events of Motion and Causation in Hong Kong Sign Language’, Lingua 117(7): 1,216–57. Tang, G. and Lau, P. (2012) ‘Chapter 16: Coordination and Subordination in Sign Languages’, in Roland Pfau, Markus Steinbach, and Bencie Woll (eds) Handbook in Sign Languages, Berlin: Mouton de Gruyter. Tang, G., Sze, F., and Lam, S. (2006) ‘Acquisition of Simultaneous Constructions by Deaf Children of Hong Kong Sign Language’, in Myriam Vermeerbergen, Lorraine Leeson, and Onno Crasborn (eds) Simultaneity in Signed Languages: Form and Function, Amsterdam and Philadelphia: John Benjamins. Woodward, J. (1993) ‘Intuitive Judgments of Hong Kong Signers about the Relationship of Sign Language Varieties in Hong Kong and Shanghai’, CUHK Papers in Linguistics 4: 88–96. Wong, Y. W. (2004) ‘Contact Signs: Variation and Implications in Hong Kong Sign Language’, Master Thesis, The Chinese University of Hong Kong. Zhong Guo Longren Xiehui (Chinese National Association of the Deaf ) (1988) Zhongguo Shouyu, Beijing: Huaxia Publishing Company. Zwitserlood, I. (2003) ‘Classifying Hand Configurations in Nederlandse Gebarentaal’, Doctoral Dissertation, Universiteit Utrecht, University of Wales, United Kingdom.

Websites Hong Kong Association of the Deaf: http://www.hongkongdeaf.org.hk/ Centre for Sign Linguistics and Deaf Studies, The Chinese University of Hong Kong: http://www.cuhk. edu.hk/cslds/v2/ The Asia Pacific Sign Linguistics Research and Training Program: http://www.cuhk.edu.hk/cslds/apsl/ The Hong Kong Jockey Club Sign Bilingualism and Co-enrolment in Deaf Education: http://www. cuhk.edu.hk/cslds/jcslco/

530

30 Lexicography Cheung Kam-Siu The Chinese University of Hong Kong, Hong Kong, China

The Chinese origin of the dictionary is the wordbook (字书, zishu), which explains the structure, pronunciation, and meaning of Chinese characters. Each dictionary normally deals with a specific aspect of Chinese characters and accordingly there are three types of wordbook in general: graphically organized dictionaries, semantically organized dictionaries, and phonetically organized dictionaries. In contrast to European philology, grammatical studies remained absent in ancient China. However, since the Han dynasty, Chinese philologists have paid special attention to lexicography, as the wordbook was an essential aid to reading literature (Norman 1988: 170). Character education occupied a privileged place in Classical Chinese learning. As Ban Gu’s 班固 ‘Monograph on Arts and Writings’ 〈艺文志〉 ( , Yiwen zhi) in The Book of the Han Dynasty 《 ( 汉书》, Hanshu) shows, in the Zhou dynasty, the first lesson for an eight-year-old pupil was the ‘six methods of forming the Chinese characters’, namely pictographic (象形, xiangxing), self-explanatory (象事, xiangshi), ideographic (象意, xiangyi), phonetic (象声, xiangsheng), derivative (转注, zhuanzhu), and loan (假借, jiajie). In the Han dynasty, only those candidates who had learned more than 9,000 characters could work as minor officials. The wordbook is a product of the evolution of Chinese characters. The form and structure of Chinese characters have undergone two major stages of evolution: from ancient writing between the Shang and Qin dynasties, to clerical and regular scripts in the Han dynasty and onwards (Qiu 2009: 40). The replacement of ancient writing by clerical script is arguably the greatest leap forward in the transformation of Chinese characters, which made lexicography necessary for understanding and explaining the structure of characters. In the Eastern Han dynasty, Xu Shen 许慎 produced An Explanatory Dictionary of Chinese Characters 《 ( 说文 解字》, Shuowen jiezi) as an attempt to classify and regularize small seal script, which had become popular since the Qin dynasty. Gu Yewang’s 顾野王 The Jade Chapters 《 ( 玉篇》, Yupian), compiled in the Southern Liang dynasty, is a record of regular script that has replaced small seal script since the Wei and Jin dynasties. The lexicon of Chinese wordbooks has been gradually expanded in response to social changes. Shuowen jiezi consists of 9,353 characters. In contrast, the initial edition of Yupian includes more than 16,900 characters and its expanded edition in the Song dynasty amounts to 22,700 characters. The Classified Chapters 《 ( 类篇》, Leipian) by Sima Guang 司马光 531

Cheung Kam-Siu

has a collection of over 31,300 characters. The Ming dynasty scholar Mei Yingzuo’s 梅膺祚 Comprehensive Dictionary of Chinese Characters 《 ( 字汇》, Zihui) collects 2,000 characters more than Leipian. The general lexicon of The Imperial Dictionary of Kangxi 《 ( 康熙字典》, Kangxi zidian) produced in the Qing dynasty covers over 47,000 characters, with an addition of over 10,000 characters. With the evolution of lexicography in ancient China, the number of radicals decreased and the definition and pronunciation of words were presented in a more scientific way. Shuowen jiezi inaugurated an influential form of lexicography based on the sequence of the radicals. Following the structural associations of Chinese characters, Xu Shen classified the Chinese characters into 540 radical types and thus gave an order to a huge and complex array of Chinese characters. Since Xu’s system of radicals was too complicated, the subsequent dictionary-makers strove to combine and simplify the radicals. Zihui consists of 214 radicals, and The Great Chinese Character Dictionary 《 ( 汉语大字典》, Hanyu da zidian), a modern dictionary with a relatively broad lexicon, includes only 201 radicals. Adopting the alphabetic order and the four-corner system in addition, the modern dictionary is more user-friendly. In reference to the ‘six methods’, Shuowen jiezi (Figure 30.1) explicates the characters and analyzes the structure of strokes. Examples are as follows: ● ●

口,人所以言食也,象形。 ‘口’ is wherewith man speaks and eats. It is pictographic. 泉,水原也,象水流出成川形。 ‘泉’ is the wellspring of a river. It represents the shape of water flowing out to form a river.

Figure 30.1  Shuowen jiezi

532

Lexicography

● ● ● ●

丄,高也,此古文上,指事。 ‘丄’ means high. It is the ancient script of ‘上’. It is self-explanatory. 林,平土有丛木曰林,从二木。 Thickets and trees on even ground called ‘林’. It derived from double tree script ‘木’. 裘,皮衣也,从衣,求声。 ‘裘’ means fur clothing. It derived from clothing script ‘衣’ and ‘求’ as phonetic. 齿,口龂骨也,象口齿之形,止声。 ‘齿’ is the gum bone of the mouth. It depicts the shape of the teeth in the mouth and ‘止’ is phonetic.

The subsequent generations of wordbooks largely followed the method adopted by Shuowen jiezi in explicating words. Nevertheless, since Xu Shen compiled Shuowen jiezi as an aid to commentating on Confucian classics, many word definitions in it were either unspecific or based on Confucian classics. Xu’s explanation is at times inaccurate, such as ‘行’. Xu Shen said, ‘行,人之步趋也。从彳从亍。’ (‘行’ is the way man either walks or runs. It derived from scripts ‘彳’ and ‘亍’.) However, the bronze forms of ‘行’ are ‘ ’ and ‘ ’; it shows a pictograph of a crossroad. ‘册’ is another example, in Shuowen jiezi, the definition of ‘册’ is ‘符命也,诸侯进受于王也。’ (‘册’ is a brevet. It is what a feudal lord receives from the king when he presents himself before the king.) But in fact, the ancient scripts of ‘册’ are ‘ ’ and ‘ ’; it is a simple pictograph depicting a set of bamboo strips woven together as a book. In terms of transcription, Shuowen jiezi indicates a character’s pronunciation with two methods: one is to use the sound radical in the phono-semantic characters to show the pronunciation; the other is to use a more common character bearing the same or similar pronunciation as the homophone. Instead, the subsequent wordbooks like Yupian and Kangxi zidian adopted a method known as fanqie (反切), which annotates the pronunciation with two characters, by combining the vowel of one character and the consonant of the other. For instance, the character dong’s (东) pronunciation is spelled through the combination of ‘d’ in de (德) and ong in hong (红). However, neither homophones nor fanqie is perfect, because homophones fail to define pronunciations accurately, while fanqie is difficult for readers to master. In addition, the historical changes in pronunciations often made it difficult to pronounce modern characters in accordance with the system of fanqie in antiquity. The Ready Guide 《 ( 尔雅》, Erya), the identity of whose author remains at issue, is the earliest Chinese dictionary that explicates the meaning of words. Erya is most probably a compilation of word definitions worked out by scholars of the pre-Han and Han dynasties. Therefore, it was neither produced by a certain individual author nor completed over a short period of time. Erya opened the prelude to dictionary compilation in China and for the first time classified words in terms of their meanings and attributes. The modern version of Erya has three volumes, categorized into 19 chapters. The first three chapters, ‘Interpreting Antique Vocabulary’ 〈释诂〉 ( , Shi gu), ‘Interpreting Verbs and Adjectives’ 〈释言〉 ( , Shi yan), and ‘Interpreting Modifiers and others’ 〈释训〉 ( , Shi xun), deal with daily vocabulary. The rest of the chapters cover specialized terms such as artifacts, astronomy, geography, animals, and plants. Thus, Erya may work as an encyclopedia. Erya defines words mainly by their connotations. It normally gives a general definition of a group of synonyms in more frequent expressions. For example, in the first entry of Erya, all the defined characters ‘初, 哉, 首, 基, 肇, 祖, 元, 胎, 俶, 落, 权舆’ (To start making clothes by cutting the cloth, the beginning of grass or tree’s growth, the human head, the base, to start opening the shutter with the hand, the ancestors, the human head, the embryo, to begin an action, leaves begin to fall 533

Cheung Kam-Siu

from trees, the beginning of the growth of plants) are explicated with the same character ‘始’ (the beginning). This method applies to the first three chapters. In contrast, the other chapters subsume words under certain special topics and define each word separately. For example, the fourth chapter Shi qin 〈释亲〉 ( , ‘Interpreting Relatives’) mainly focuses on explaining the kinship terms. The fifth chapter Shi gong 〈释宫〉 ( , ‘Interpreting the Court’) focuses on explaining the architectural terms. The sixth chapter Shi qi 〈释器〉 ( , ‘Interpreting Utensils’) focuses on explaining the names of various utensils and tools, together with verbs having to do with the use of these items. The seventh chapter Shi yue 〈释乐〉 ( , ‘Interpreting Music’) focuses on explaining the names of musical instruments and musical terms. The eighth chapter Shi tian 〈释天〉 ( , ‘Interpreting the Heavens’) focuses on explaining the astronomical, calendrical, and meteorological terms. The ninth chapter Shi di 〈释地〉 ( , ‘Interpreting Earth’) focuses on explaining the geographical and geological terms. The tenth chapter Shi qiu 〈释丘〉 ( , ‘Interpreting Mounds’) focuses on explaining the terms having to do with hills. The eleventh chapter Shi shan 〈释山〉 ( , ‘Interpreting Mountains’) focuses on explaining the names of mountains. The twelfth chapter Shi shui 〈释水〉 ( , ‘Interpreting Water’) focuses on explaining the names of springs and rivers. The thirteenth chapter Shi cao 〈释草〉 ( , ‘Interpreting Grass’) focuses on explaining names of grasses, herbs, and vegetables. The fourteenth chapter Shi mu 〈释木〉 ( , ‘Interpreting Woods’) focuses on explaining the names of trees and shrubs. The fifteen chapter Shi chong 〈释虫〉 ( , ‘Interpreting Insects’) focuses on explaining the names of different insects and worms. The sixteen chapter Shi yu 〈释鱼〉 ( , ‘Interpreting Fishes’) focuses on explaining the names of different fishes. The seventeenth chapter Shi niao 〈释鸟〉 ( , ‘Interpreting Birds’) focuses on explaining the names of different birds. The eighteenth chapter Shi shou 〈释兽〉 ( , ‘Interpreting Beasts’) focuses on explaining the names of wild and legendary animals. The last chapter Shi chu 〈释畜〉 ( , ‘Interpreting Livestock’) focuses on explaining the names of livestock. As the material and cultural changes enriched the Chinese expressions, the later dictionaries followed Erya’s style and expanded its lexicon. All of them are entitled the ‘Ready Guide’ (雅, ya), and one of the earliest is The Pocket Ready Guide 《 ( 小尔雅》, Xiao Erya), which was produced in the late Han dynasty and became popular in the Wei and Jin Dynasties. However, given Xiao Erya’s limited vocabulary, it failed to improve the Guide as The Broad Ready Guide 《 ( 广雅》, Guangya) did. The latter was compiled by Zhang Yi 张揖, a Wei scholar in the period of the Three Kingdoms. Following the organization of Erya, Guangya consists of ten chapters, and it also adopts Erya’s method of definition. As the eminent Qing scholar Wang Niansun 王念孙 noted, Guangya draws upon a considerably wide range of linguistic materials, including the exegesis of Confucian classics, the annotations of Han-dynasty works, the explication of The Songs of Chu 《 ( 楚辞》) and the odes of Han, The Cangjie Primer 〈仓颉篇〉 ( , Cangjie pian) and Shuowen jiezi. As Guangya incorporates the philological materials ranging from the pre-Qin period to the Han dynasty, it stands as the most important classical dictionary since Erya. Guangya was followed by a succession of dictionaries, including The Augmented Ready Guide 《 ( 埤雅》Piya) by the Song scholar Lu Dian 陆佃, The Extended Ready Guide 《 ( 尔雅翼》Erya yi) by Luo Yuan 罗愿, The Rhythmical Ready Guide 《 ( 骈雅》, Pianya) by the Ming scholar Zhu Mouwei 朱谋玮, The General Ready Guide 《 ( 通雅》, Tongya) by Fang Yizhi 方以智, The Distinctive Ready Guide 《 ( 别雅》, Bieya) by the Qing scholar Wu Yujin 吴玉搢, and The Contrastive Ready Guide 《 ( 比雅》, Biya) by Hong Liangji 洪亮吉. Apart from the ‘Ready Guide’ series, The Dictionary of Dialectal Words 《 ( 方言》, Fangyan), compiled by the Western Han scholar Yang Xiong 扬雄, is another significant work in the 534

Lexicography

history of classical dictionaries. Fangyan is the first Chinese dictionary produced on the basis of dialectal survey, and it covers a collection of local oral speeches and compares different dialects. Fangyan imitated Erya in both its form and content. A telling example is that the classification of the words included in Fangyan resembles Erya’s structure. For instance, Volume Four covers words about clothing, Volume Five contains words regarding utensils, and Volume Eight includes words about animals. Fangyan explicates words primarily in two ways: for one, the dictionary lists a headword and explains the local variations of the same headword in various regions, such as: ● 帬,陈魏之间谓之帔,自关而东或谓之襬。 ‘帬’ (qun, a skirt): In the area of Chen and Wei, it is called ‘帔’ (pei), from the east of the Pass, some call it ‘襬’ (bai). ● 箭,自关而东谓之矢,江淮之间谓之鍭,关西曰箭。 ‘箭’ ( jian, an arrow): From the east of the Pass, it is called ‘矢’ (shi), in the areas of River Chang and River Huai, it is called ‘鍭’ (hou), from the west of the Pass, it is called ‘箭’. The other way is to present a group of synonyms and list the local origins of each word, such as: ● 悼、惄、悴、慭,伤也。自关而东汝颍陈楚之间通语也。汝谓之惄,秦谓之悼,宋谓之悴, 楚颍之间谓之慭。 ‘悼’ (dao), ‘惄’ (ni), ‘悴’ (cui), ‘慭’ (yin) mean ‘伤’ (shang, ‘to hurt’), the standard term to the east of the Pass, in the areas of Ru and Ying, Chen and Chu. In Ru it is called ‘惄’, in Qin it is called ‘悼’, in Song it is called ‘悴’, and in the areas of Chu and Ying it is called ‘慭’. ● 虔、散,杀也。东齐曰散,青徐淮楚之间曰虔。 ‘虔’ (qian) and ‘散’ (san) mean ‘杀’ (sha, ‘to kill’). In the Eastern Qi, it is called ‘散’, in the Qing, Xu, Huai, and Chu areas it is called ‘虔’. The vernacular dictionaries following Fangyan included The Augmented Dictionary of Dialectal Words 《 ( 续方言》Xu fangyan) by the Qing scholar Hang Shijun 杭世骏, Supplements and Reifications to the Augmented Dictionary of Dialectal Words 《 ( 续方言补正》, Xu fangyan buzheng) by Cheng Jisheng 程际盛, and The Dictionary of Textual Researches on the Wu-Xia Dialect 《 ( 吴下方言考》, Wuxia fangyan kao) by Hu Wenying 胡文英. The rhyme book appeared during the Wei and Jin dynasties. The first one was the Wei scholar Li Deng’s 李登 Dictionary of Initial Consonants 《 ( 声类》, Shenglei), followed by the Jin scholar Lü Jing’s 吕静 Collection of Rhymes 《 ( 韵集》, Yunji). Among the ensuing studies of rhymes, The Dictionary of Chinese Rhymes 《 ( 切韵》, Qieyun) by the Sui scholar Lu Fayan 陆法言 remained the most influential one. Later, with supplements and amendments to the Qieyun, the Tang scholar Sun Mian 孙愐 produced The Tang Dictionary of Rhymes 《 ( 唐韵》, Tangyun). However, since both Qieyun and Tangyun were lost, the earliest dictionary of rhymes available is The Extended Rhymes Dictionary 《 ( 广韵》, Guangyun) by the Song scholar Chen Pengnian 陈彭年 et al. and its expanded edition, The Rhyme Dictionary 《 ( 集韵》, Jiyun) by Ding Du 丁度 et al. Guangyun (Figure 30.2) has a total of 26,194 character entries and the characters are arranged in 206 rhyme groups, namely 57 level tone rhymes (平声, pingsheng), 55 falling-rising tone rhymes (上声, shangsheng), 60 falling tone rhymes (去声, qusheng), and 34 entering 535

Cheung Kam-Siu

Figure 30.2  Guangyun

tone rhymes (入声, rusheng). The whole book falls into five volumes: level tone rhymes which comprise the largest number of characters are divided into two volumes and each other tone rhyme is respectively included in one volume. In terms of the formula, every entry contains a brief explanation of the character’s meaning and the pronunciation given by fanqie method, and homophonous characters are arranged in same groups. The rhyme book mainly serves to help readers identify each character’s pronunciation and the rhyme group it belongs to. As the explication of the structure and meaning of characters are also included in many cases, it is appropriate to regard the rhyme book as a dictionary. From the perspective of lexicography, rhyme books are more user-friendly than word books like Shuowen jiezi and Erya as the adoption of the fanqie system enables readers to find characters easily. The Ming and Qing dynasty witnessed the culmination of the development of traditional Chinese dictionaries, and Zihui (Figure 30.3), compiled by the Ming dynasty scholar Mei Yingzuo 梅膺祚, and the Kangxi zidian (Figure 30.4), edited by Qing dynasty scholars like Zhang Yushu 张玉书, are widely recognized as two monumental works. The 14-volume Zihui was completed in 1615, and has a collection of 33,179 characters. The work is renowned for reducing the number of radicals used by Shuowen jiezi to 214 and introducing the principle of arranging characters under a radical according to the number of residual strokes. Another contribution by Mei Yingzuo is that he established at the end of Volume One the table of characters whose radicals are difficult to identify. All of the abovementioned innovations have been inherited by Modern Chinese dictionaries. Zhang Zilie’s 张自烈 The Rectified Dictionary of Chinese Characters 《 ( 正字通》, Zhengzi tong), produced in the late Ming dynasty, is an exemplary work modeling on Zihui in terms 536

Figure 30.3  Zihui

Cheung Kam-Siu

Figure 30.4  Kangxi zidian

of radical system and collation. The progress made by the compiler lies in that the work corrected errors in Zihui and expanded the references in the annotations by incorporating the books of Buddhism, Taoism, and medicine. In the Qing dynasty, under the auspices of Kangxi Emperor, Zhang Yushu and Chen Tingjing 陈廷敬 compiled Kangxi zidian from 1710 to 1716. The dictionary adopted the new Zihui and Zhengzhi tong system of 214 radicals, and the characters under the same radical are arranged by the number of additional strokes. In every entry, the compilers listed pronunciations in traditional fanqie spelling given by Tangyun, Guangyun, Jiyun, and A Collection of Rhymes 《 ( 韵会》, Yunhui), different meanings, and quotations from classical books. If there is any textual research, it will be marked with a note. Kangxi zidian has a vocabulary of 47,035 characters, including those rare ones that have not been covered by precedent wordbooks. It terms of lexicon, its scale is much larger than Zihui and Zhengzi tong, containing about 10,000 more characters. The dictionary is eulogized as the most comprehensive and authoritative traditional Chinese dictionary, and has exerted a strong influence on subsequent Chinese lexicography. 538

Lexicography

Since the late Qing dynasty, as the modes of traditional Chinese dictionaries like Shuowen jiezi and Erya failed to cope with the tremendous transformation of Chinese society, new types of dictionaries have emerged, one of which is The Great Character Dictionary of China 《 ( 中华大字典》, Zhonghua da zidian), another large scale dictionary after Kangxi zidian. The compilation was started in 1909 and completed in 1914. Based on the 214-radical system of Kangxi zidian, each character entry includes the fanqie spelling, different meanings, and classical quotations. Zhonghua da zidian’s explication, unlike Kangxi zidian, is noted for its brevity, and corrects errors in Kangxi zidian by incorporating the researches done by Qing dynasty philologists, such as Wang Niansun’s The Rectified Broad Ready Guide 《 ( 广雅疏证》, Guangya shuzheng), and Duan Yucai’s 段玉裁 The Annotated Explanatory Dictionary of Chinese Characters 《 ( 说文解字注》, Shuowen jiezi zhu). The dictionary has a collection of 48,000 entries for individual characters in total, including characters’ standard forms, big seal script forms, archaic forms, popular forms, and variant forms. The dialect words and translated words are also covered. The Origins of Words 《 ( 辞源》, Ciyuan), edited by the Commercial Press, is a contemporary work of Zhonghua da zidian. Ciyuan is an encyclopedic dictionary, which offers definitions of common words and also incorporates idioms, allusions, jargons, names of persons and places, terms concerning astrology, arts, and social institutions, and recently invented words. Following Ciyuan, there were The Dictionary of Words 《 ( 辞海》, Cihai) published by the China Press in 1937 and The Dictionary of the National Language 《 ( 国语辞典》, Guoyu cidian), which was compiled by the Chinese Dictionary Compiling Department and published by the Commercial Press in four volumes between 1937 and 1945. The latter made a significant change in the format of the dictionary. It presented the characters in accordance with the alphabetic order of the system of Phonetic Symbols (注音符号, zhuyin fuhao) and marked the pronunciation in two modes: zhuyin fuhao and the National Romanized Phonetic Alphabet (国语罗马字, guoyu luomazi). The Modern Chinese Dictionary 《 ( 现代汉语词典》, Xiandai Hanyu cidian), compiled by the Institute of Linguistics, Chinese Academy of Social Sciences and published by the Commercial Press after the founding of the People’s Republic of China, is the first mediumsized dictionary of Modern Chinese language. Following the State Council’s instructions on the promotion of Mandarin in February 1956, the dictionary compilation began and worked to set up a standard of Modern Chinese. Although the project started in 1958, the first formal edition was not published until 1978. The 1978 edition contains about 56,000 entries, three quarters of which are ordinary words and the rest are special terms. The pronunciation of words is modeled on the Pinyin system, the People’s Republic of China’s official phonetic system for transcribing the sounds of Chinese characters into Latin script; the word definition is accurate and concise. Instead of specifying a notional word’s part of speech, the dictionary defines it in reference to other words belonging to the same part of speech, namely, nouns are defined by nouns and verbs by verbs. In terms of functional words including numerals, measure words, adverbs, and conjunctions, the dictionary specifies their relevant parts of speech. The dictionary also provides information on the usage of words, marked by abbreviated labels. For example, the label shu (书) indicates a word that is typically used in writing; the label kou (口) refers to a word that is preferred in speech; the label fang (方) indicates a word in certain dialects; the archaic words are labeled as gu (古). Since 1978, five revised editions of this dictionary have been published, and the 6th edition, including 69,000 entries, was published in June 2012. The more than 3,000 new entries have incorporated cyber-slang, loan words, and dialects, as an immediate response to the most recent social changes in China. 539

Cheung Kam-Siu

Apart from Xiandai hanyu cidian, The Great Chinese Character Dictionary 《 ( 汉语大字典》, Hanyu da zidian) and The Great Chinese Word Dictionary 《 ( 汉语大词典》, Hanyu da cidian) are recognized as two influential dictionaries produced in the past four decades. The compilation of the former began in 1979, and the dictionary was successively published in eight volumes between 1986 and 1990. With a collection of 56,000 Chinese characters, it stands as the most comprehensive dictionary of Chinese characters. The compilation of the latter started in 1975 as well, and its 11 volumes were published successively between 1986 and 1993. Composed of more than 370,000 entries, Hanyu da cidian ranks as the most comprehensive dictionary of Chinese words. The above dictionaries which focused on the structure, pronunciation, and meaning of Chinese characters and were edited by Chinese scholars constitute the dominant mode of Chinese lexicography. Exceptions are the Western missionaries’ practices of compiling dictionaries which started in the late Ming and early Qing dynasties. Although the total number of missionaries working in China from the sixteenth to the nineteenth century never reached 1,000, they produced about 200 bilingual dictionaries and grammar books in order to preach Christianity in the Chinese language (Yang 2007: 135). The monumental works include Dizionario Portoghese–Cinese 《 ( 葡汉词典》) edited by the Italian Jesuits Michele Ruggieri and Matteo Ricci, and A Dictionary of the Chinese language 《 ( 汉英词典》) edited by the Scottish Protestant missionary Robert Morrison. Dizionario Portoghese–Cinese is best-known for adopting Latin letters to represent Chinese pronunciation in alphabetical order for the first time, and A Dictionary of the Chinese language, published from 1815 to 1823 in three parts, is widely acknowledged as the first Chinese–English bilingual dictionary, incorporating not only definitions and usages of Chinese characters, but also Chinese idioms and colloquial words. The format of the latter dictionary was inherited by Samuel Wells Williams’ An English and Chinese Vocabulary in the Court Dialect 《 ( 英华韵府历阶》) and Walter Henry Medhurst’s English and Chinese Dictionary 《 ( 英汉字典》). There are also flaws with these bilingual dictionaries, and the most obvious one is the inaccuracy of transcription: for example, Morrison failed to distinguish aspirated/unaspirated contrasts in Chinese language, and Medhurst mixed them. However, these dictionaries constitute an inseparable part of Chinese lexicography, and stand as important documents of witnessing he Chinese and Western cultural exchanges. Since the twentieth century, the purpose of compiling bilingual dictionaries has shifted from helping promote Christianity to enabling foreigners to learn the Chinese language or do researches. Important dictionaries produced in this period include Robert Henry Mathews’ Chinese–English Dictionary 《 ( 汉英 字典》, 1931), Yuen-Ren Chao 赵元任 and Lien-sheng Yang’s 杨联升 Concise Dictionary of Spoken Chinese 《 ( 国语字典》, 1947), John DeFrancis’s ABC Chinese–English Dictionary 《 ( 汉英词典》, 1996) and ABC Chinese–English Comprehensive Dictionary 《 ( 汉英大词典》, 2003). With the progress of information technology and the expansion of cyberspace, hard-copy dictionaries are giving way to CD-ROM and online versions, which have set the current trend of Chinese lexicography. Some databases are constructed on the basis of paper dictionaries. Preserving all the contents of their paper predecessors, the online versions excel further in their user-friendliness by adopting the search function. Famous examples included the online versions of Shuowen jiezi (http://www.shuowen.org/), Guangyun (http://ytenx.org/kyonh/), Kangxi zidian (http://www.kangxizidian.com/index2.php), and Guoyu cidian (http://dict. revised.moe.edu.tw/). There are also databases focusing on specific topics: ‘Chinese Etymology’ (http://www.chineseetymology.org/) incorporates the structure of Chinese characters; Dictionary 540

Lexicography

of Chinese Character Variants (http://dict.variants.moe.edu.tw/) analyzes the structure; Dictionary of Chinese Idioms (http://dict.idioms.moe.edu.tw) explicates Chinese idioms; and ‘Chinese Character Database: Phonologically Disambiguated According to the Cantonese Dialect’ is established to record proper pronunciations of Cantonese (http://humanum.arts. cuhk.edu.hk/Lexis/lexi-can/). The developers, either academic institutions or amateurs, have provided illuminating insights into the Chinese lexicography by fully taking advantage of the Internet.

Appendix List of book titles from Chinese to English with Chinese Titles 艺文志 汉书 说文解字 玉篇 类篇 字汇 康熙字典 汉语大字典 尔雅 小尔雅 广雅 楚辞 仓颉篇 埤雅 尔雅翼 骈雅 通雅 别雅 比雅 方言 续方言 续方言补正 吴下方言考 声类 韵集 切韵 唐韵 广韵 集韵 正字通 韵会 中华大字典 广雅疏证 说文解字注 辞源 辞海

Monograph on Arts and Writings The Book of the Han Dynasty An Explanatory Dictionary of Chinese Characters The Jade Chapters The Classified Chapters Comprehensive Dictionary of Chinese Characters The Imperial Dictionary of Kangxi The Great Chinese Character Dictionary The Ready Guide The Pocket Ready Guide The Broad Ready Guide The Songs of Chu The Cangjie Primer The Augmented Ready Guide The Extended Ready Guide The Rhythmical Ready Guide The General Ready Guide The Distinctive Ready Guide The Contrastive Ready Guide The Dictionary of Dialectal Words The Augmented Dictionary of Dialectal Words Supplements and Reifications to the Augmented Dictionary of Dialectal Words The Dictionary of Textual Researches on the Wu-Xia Dialect Dictionary of Initial Consonants Collection of Rhymes The Dictionary of Chinese Rhymes The Tang Dictionary of Rhymes The Extended Rhymes Dictionary The Rhyme Dictionary The Rectified Dictionary of Chinese Characters A Collection of Rhymes The Great Character Dictionary of China The Rectified Broad Ready Guide The Annotated Explanatory Dictionary of Chinese Characters The Origins of Words The Dictionary of Words 541

Cheung Kam-Siu

国语辞典 现代汉语词典 汉语大字典 汉语大词典

The The The The

Dictionary of the National Language Modern Chinese Dictionary Great Chinese Character Dictionary Great Chinese Word Dictionary

Bibliography Bai, Zhaolin (2007) ‘Xiandai hanyu cidian yu chuantong xungu’ (‘The Modern Chinese Dictionary and Traditional Exegesis’) Lexicographical Studies 1: 37–42. Ban, Gu (2002) Hanshu (The Book of the Han Dynasty) (12 vols), Beijing: Chung Hwa Book Company. Béjoint, Henri (2000) Modern Lexicography, An Introduction, Oxford: Oxford University Press. Béjoint, Henri (2010) The Lexicography of English, Oxford and New York: Oxford University Press. Bussman, Hadumod (1996) Routledge Dictionary of Language and Linguistics, trans. and ed. by Gregory Trauth and Kerstin Kazzazi, London and New York: Routledge. Cao, Xianzuo and Chen Bingcai (1992) Baqianzhong zhongwen cishu leibian tiyao (Classification and Brief Introduction to 8,000 Chinese Dictionaries), Beijing: Peking University Press. Carr, Michael (1997) ‘Internet Dictionaries and Lexicography’, International Journal of Lexicography 10(3): 209–30. Cen, Qixiang (2008) Yuyanxueshi gaiyao (An Outline History of Linguistics), Beijing: World Publishing Company. Chen, Bingtiao (1985) Cishu gaiyao (An Introduction to Chinese Dictionaries), Fuzhou: Fujian People’s Press. Chen, Bingtiao (1986) Guowai xiandai cishu xuanjie (An Introduction to Some Foreign Modern Dictionaries), Fuzhou: Fujian People’s Press. Chen, Bingtiao (1991) Cishu bianzuanxue gailun (An Introduction to Lexicography Compilation), Shanghai: Fudan University Press. Chen, Bingtiao (2013) ‘Xianqin zhuanshu cidian bian zuan gai shuo’ (‘Compilation of Specialized Book’s Dictionaries in the Pre-Qin Period: An Overview’), Lexicographical Studies 5: 16–26. Chen, Zhen (1984) ‘Wuxia fanyankao de zuozhe hu wenying’ (‘Hu Wenying: The Author of The Dictionary of Textual Research on the Wu-Xia Dialect’), Lexicographical Studies 1: 170–73. Chien, David and Thomas Creamer (1986) ‘A Brief History of Chinese Bilingual Lexicography’, in Reinhard R.K Hartmann (ed) The History of Lexicography, Papers from The Dictionary Research Centre Seminar at Exeter, March 1986, Amsterdam and Philadelphia: John Benjamins. Chien, David (1986) Lexicography in China: Bibliography of Dictionaries and Related Literature, Exeter: University of Exeter. Coblin, W. South (1978) ‘The Initials of Xushen’s Language as Reflected in the Shuowen duruo Glosses’, Journal of Chinese Linguistics 6: 27–75. Dai, Jiaxiang (ed) (1995) Jinwen da zidian (A Dictionary of Bronze Inscriptions Characters), Shanghai: Xuelin Publishing House. Ding, Feng (1984) ‘Zhengzitong zhuozhe shi liao wenying’ (‘Liao Wenying: The Author of The Rectified Dictionary of Chinese Characters’), Lexicographical Studies 1: 70. Duan, Yucai (1997) Shuowen jiezi zhu (An Explanatory Dictionary of Chinese Characters), Shanghai: Shanghai Ancient Book Press. Fang, Houshu (1979) ‘Zhongguo cishu shi hua(shang)’ (‘A Narrative History of Lexicographical in China (Part 1)’), Lexicographical Studies 1: 217–28. Fang, Houshu (1979) ‘Zhongguo cishu shi hua(xia)’ (‘A Narrative History of Lexicographical in China (Part 2)’), Lexicographical Studies 2: 219–29. Feng, Fengfeng (1988) ‘Kangxi zidian bianzuan lilun chutan’ (‘A Primary Study on the Compilation Theory of The Imperial Dictionary of Kangxi’), Lexicographical Studies 2: 72–80. Feng, Fengfeng (1992) ‘Wei jin sui tang zidian bianzuan gaiguan’ (‘An Overview of the Theories of Character Dictionary Compilation in the Wei, Jin, Sui, and Tang Dynasties’), Lexicographical Studies 3: 96–108. Guan, Xihua (1996) Erya yanjiu (A Study on The Ready Guide), Hefei: Anhui University Press. Hartmann, Reinhard R. K. and Gregory James (1998) Dictionary of Lexicography, London and New York: Routledge.

542

Lexicography Hartmann, Reinhard R. K. (2012) ‘A Life of Reference: On Dictionary Research Perspectives, with Particular Reference to Learner’s Dictionaries’, Lexicographical Studies 1: 74–80. He, Jiuying (1995) Zhongguo gudai yuyanxue shi (A History of Ancient Chinese Philology), Guangzhou: Guangdong Education Publishing House. He, Jiuying (1995) Zhongguo xiandai yuyanxue shi (A History of Modern Chinese Philology), Guangzhou: Guangdong Education Publishing House. Hu, Qiguang (1987) Zhongguo xiaoxue shi (A History of Chinese Exegetic Studies), Shanghai: Shanghai People’s Publishing House. Hu, Qiguang and Fang Huanhai (2001) ‘Erya chengshu niandai xinlun’ (‘Further Discussion on the Completion Date of The Ready Guide’), Lexicographical Studies 6: 106–16. Huang, Min (1984) ‘Wuxia fanyankao lueshu’ (‘Brief Introduction of The Dictionary of Textual Research on the Wu-Xia Dialect’), Lexicographical Studies 1: 139–44. Jackson, Howard (2002) Lexicography: An Introduction, London and New York: Routledge. Jiang, Lansheng (2013) ‘Xiandai hanyu cidian di liu ban gaishu’ (‘An Overview of Modern Chinese Dictionary (6th ed.)’), Lexicographical Studies 2: 1–19. Li, Jianguo (2002) Hanyu xunguxue shi (A History of Chinese Exegetic Studies) (Revised ed.), Shanghai: Shanghai Lexicographical Press. Li, Rong (1993) ‘Fangyan cidian shuolue’ (‘A Brief Discussion about Dialect Dictionaries’), Lexicographical Studies 4: 130–34. Lin, Yushan (1992) Zhongguo cishu bianzuan shi lue (A Brief History of Dictionary Compilation in China), Zhengzhou: Zhengzhou Ancient Book Press. Lin, Yushan (ed) (1995) Cishuxue gai lun (An Introduction to Lexicography), Fuzhou: The Straits Publishing House. Liu, Qiti (2006) ‘Hanyu da cidian (wangluoban V.2) de tese’ (‘The Characteristics of the Great Character Dictionary of China (Online version no.2)’), Lexicographical Studies 4: 115–18. Liu, Yeqiu (1983) Zhongguo zidian shi lüe (A Brief History of Chinese Character Dictionaries), Beijing: Chunghwa Book Company. Mugglestone, Lynda (2011) Dictionaries: A Very Short Introduction, Oxford and New York: Oxford University Press. Nielsen, Sandro and Sven Tarp (eds) (2009) Lexicography in the 21st Century, Amsterdam and Philadelphia: John Benjamins. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Qian, Jianfu (1979) ‘Shuowen jiezi gaishu’ (‘A Brief Introduction of an Explanatory Dictionary of Chinese Characters’), Lexicographical Studies 1: 229–47. Qian, Jianfu (1986) Zhongguo gudai zidian cidian gailun (An Introduction to Ancient Dictionaries in China), Beijing: The Commercial Press. Qiu, Xigui (2009) Wenzixue gaiyao (Chinese Writing), Beijing: The Commercial Press. Serruys, Paul Leo Mary (1959) The Chinese Dialects of Han Time According to Fang Yen, Berkeley: University of California Press. Serruys, Paul Leo Mary (1984) ‘On the System of the Pu Shou in the Shuo-wen Chieh-tzu’, Journal of the Institute of History and Philology, Academia Sinica 55(4): 651–754. Thern, K.L. (1966) Postface of the Shuo wen chieh tzu, The First Comprehensive Chinese Dictionary, Madison, WI: University of Wisconsin. Wang, Donghai (2013) ‘Hanyu cishu lilunshi de fenqi yanjiu’ (‘Period-based Study of the Theoretical History of Chinese Dictionaries’) Lexicographical Studies 3: 1–14. Wang, Jianli (2012) ‘Lun erya de tongyici cidian xingzhi’ (‘Synonym Dictionary’s Properties of Erya’) Lexicographical Studies 2: 60–4. Wang, Niansun (2000) Guangya Shuzheng (The Rectified Broad Ready Guide), Nanjing: Jiangsu Ancient Book Publishing House. Wang, Ning (2008) ‘Lun cishu de yuanchuangxing ji qi rending yuanze: jianlun xiandai hanyu cidian de yuanchuangxing he yuanchuangdian’ (‘On the Originality of Dictionaries and the Verification Principles: A Concurrent Discussion on the Modern Chinese Dictionary’), Lexicographical Studies 1: 1–13, 47. Wei, Li (2008) ‘Zhonghua da zidian shuping’ (‘On The Great Character Dictionary of China’) Lexicographical Studies 6: 85–94. Xu, Qingkai (2012) ‘Cishu de wenhua jiazhi’ (‘Cultural Values of Dictionaries’), Lexicographical Studies 3: 1–5.

543

Cheung Kam-Siu Xu, Shen (1998) Shuowen jiezi (An Explanatory Dictionary of Chinese Characters), Hong Kong: Chunghwa Book Company Limited (Hong Kong). Xu, Zhongshu (ed) (1988) Jiakuwen zidian (A Dictionary of Oracle Bones Scripts), Chengdu: Sichuan Lexicographical Press. Yang, Huiling (2007) ‘Yezunxiao de hanzi xiyi yu malison de hanying cidian’ (‘Basilio Brollo de Glemona’s Dictionnaire Chinois, Francais et Latin and Robert Morrison’s A Dictionary of the Chinese Language’), Lexicographical Studies 1: 135–42. Yong, Heming and Peng Jing (2007) Bilingual Lexicography from a Communicative Perspective, Amsterdam and Philadelphia: John Benjamins. Yong, Heming (2002) ‘Luobote molisun qi ren qi dian’ (‘Robert Morrison – A Brief Biography and his Dictionary’), Lexicographical Studies 4: 121–25. Yong, Heming (2004) ‘Guanyu zhongguo cidian shi yanjiu de sikao’ (‘Thoughts on the Studies of Chinese Dictionary History’), Lexicographical Studies 2: 14–21. Yong, Heming (2008) Chinese Lexicography: A History from 1046 bc to ad 1911, Oxford and New York: Oxford University Press. Zhang, Bo (2013) ‘Xiandai hanyu cidian di liu ban shiyi xiuding de leixing ji tezheng’ (‘Revision Types and Characteristics of Modern Chinese Dictionary (6th ed.)’), Lexicographical Studies 2: 25–33. Zhang, Minghua (1994) Zhongguo gudai zidian cidian (Dictionaries in Ancient China), Taipei: The Commercial Press. Zhang, Shaoying (2012) ‘Yuwen cidian zhong yixiang pailie shunxu yanjiu’ (‘Research of Senses Ordering in Language Dictionaries’), Lexicographical Studies 3: 24–7. Zhang, Zicai (2008) ‘Hong liangji de biya’ (‘Hong Liangji and his The Contrastive Ready Guide’), Lexicographical Studies 1: 119–24. Zhao, Cheng (1979) Zhongguo gudai yunshu (The Rhyme Dictionaries in Ancient China), Beijing: Chunghwa Book Company. Zhao, Zhenduo (1979) ‘Guanyu hanyu da zidian de bianxie gongzuo’ (‘About the Compilation of the Great Chinese Character Dictionary’), Lexicographical Studies 1: 97–103. Zhao, Zhenduo (1986) Gudai cishu shi hua (Talks on the History of Ancient Word Books and Dictionaries), Chengdu: Sichuan People’s Publishing House. Zhao, Zhenduo (2006) Cishuxue lunwen ji (Collected Papers of Lexicography), Beijing: The Commercial Press. Zhou, Zumo (1993) Fangyan jiaojian (A Collated and Annotated Dictionary of Dialectal Words), Beijing: Chunghwa Book Company.

544

31 Lexicography in the Contemporary Period Huang Chu-Ren and Li Lan Hong Kong Polytechnic University, Hong Kong, China

Su Xinchun Jiageng College of Xiamen University, China

1.  Introduction: historical background of Chinese lexicography The history of Chinese lexicography can be traced back to nearly 1,900 years ago to Xu Shen’s 许慎 Shuowen Jiezi《说文解字》. Xu (131 ce) established a semantics-driven orthographybased framework for lexicography. He analyzed Chinese characters and found that component parts encoding semantic concepts, called bushou (部首 ‘radical’), can be used to identify and classify related characters. In Shuowen Jiezi, each Chinese character (an orthographic unit and an equivalent of a conventionalized sociological word in Chinese) is given an entry according to the radical it contains (and hence its conceptual classification). The entry contains a rough definition of its meaning, often in relation to the basic meaning of the radical; the character composition according to its components (bujian 部件); and very often also gives hint on its pronunciation. Although Erya《尔雅》is often claimed to be an even earlier collection of Chinese ‘words’ in different categories, it is important to note that Erya is a taxonomic collection of terms without linguistic information. Most crucially, the 540-radical system of Shuowen Jiezi has been adopted by all major Chinese dictionaries for nearly two millennia with adaptation and simplification. Indeed, we may conclude that Chinese lexicography started with and has been dominated by Shuowen Jiezi. Although many Chinese dictionaries in the modern era still retain a reorganized and reduced set of radicals as either its primary structure or secondary index the field of lexicography did undergo drastic changes. The vernacular language movement in the early twentieth century created an environment for Chinese lexicographers to focus more on the commonly spoken language and lexical words which may contain one or more characters. Hence the first change is the emergence of the word-based dictionary cidian 辞典 or cishu 词书 versus the traditional character-based dictionary zidian 字典 (literally ‘character dictionary’). In addition, different systems are invented to give phonological representation of the pronunciation so that dictionaries can be organized according to how words are spoken daily instead of written according to the long literary tradition. The earlier phonetic alphabet 545

Huang Chu-Ren, Li Lan, and Su Xinchun

system of zhuyinfuhao 注音符号 is still used in Taiwan, while Pinyin Romanization adopted by mainland China has become the international standard. Based on different phonological transcription conventions, the second change is the emergence of lexicographic system organized according to alphabetic order. The alphabetical order of English is adopted for Pinyin, while ordering according to articulatory location is conventionalized for the phonetic alphabet system (hence the popular name of bopomofo). These two emergent changes in the early part of the twentieth century brought Chinese lexicography to a shared convention with modern lexicography. It also laid the foundation for more recent developments driven by the computerization of the Chinese writing system and the easy accessibility of digital content. In what follows, we will focus on these recent developments.

2.  Fundamental issues: between character zi 字 and word ci 詞, and from character encoding to word segmentation The identification of a lexical unit is the fundamental issue of lexicography. The commonly held (but also often challenged) assumption that the linguistic word should be the most basic lexical unit (e.g. Hartmann 2003; Bloomfield 1926) does not translate into an executable procedure in Chinese lexicography due to its lack of conventionally marked word boundaries (e.g. Huang and Xue 2012) and confusion caused by the competing concepts of character and word (zi 字 and ci 詞 respectively in Chinese). By adopting the neutral term ‘lexical unit’, the ISO 24613:2008 standard for electronic lexicon incorporated a word-like concept in its formal definition and was successfully implemented for a wide range of languages in the world (Francopoulo 2013) including Asian languages (Francopoulo and Huang 2014). Although this result suggests that it is possible to have a common conceptual lexical unit for different languages the character vs. word competition has been, and still is, one of the most critical issues driving research and development in Chinese lexicography in the contemporary period.

2.1.  Character encoding: representation and variations The dichotomy of Chinese dictionaries dictates the definition of lexical entries: characters are lexical entries in a dictionary of characters, and words are lexical entries in a dictionary of words. Although orthographic convention has clearly defined character boundaries, orthographic variations also pose a challenge to the definition of which forms belong to the same character entry. The encoding of Chinese characters, in fact, was one of the first research issues in computational processing of the Chinese language, which brought the field of Chinese writing system (wenzixue 文字学) to the forefront of recent computational studies (Hsieh 1996). Note that a lexical unit typically represents what language users perceive as a single minimal form-meaning pair which allows some variations in forms. In Chinese orthography the variations go beyond graphic variations of the same glyph in different (historical, regional, or typographic) conventions. For instance, the concept of ‘peak’, sharing the same phonological form of feng in modern Mandarin can be represented by either 峰 or 峯, two variants with their components composed differently (left–right vs. top–down). They should be free variants in almost all contexts and be treated under one single entry with rare exception for proper names. However, this is not possible given the traditional character-form based approach. This inconsistency in dealing with glyphic variants can be further exemplified by the ren ‘blade’. The authoritative Kangxi Zidian《康熙字典》listed four homographs 刃刄 546

Lexicography in the contemporary period

刄 separately from the others and Unicode followed suit by giving it a different code. Close inspection will see that these variants differ only in the position and shape of the dot, which serves to refer to the ‘blade’ by marking its location on a knife 刀 dao. In this case, neither the component parts nor the meaning can be differentiated among these variants. A more complicated example involves three glyphs 冲衝沖 chong ‘to charge (ahead)’ and/or ‘to crash (with water)’. In simplified Chinese, the two water-dot 冲 stands for both concepts and will be one lexical unit. For traditional Chinese, the water-based 冲, as well as the nonwater related 衝 ‘charge, onslaught’ are different entries. However, for Japanese kanji, the three water-dot 沖 forms a single entry, while the same character can also serve as glyph variants to the two-water dot 冲 for both traditional and simplified Chinese. The complexity of identifying characters is compounded by the need to identify and represent them in a computer. The computational solution by the Intelligent Chinese Character Encoding System (Jhuang et al. 2005) can provide a way to better define characters as lexical entries. This system can decompose each character based on philological principles, orthographic conventions, and a string of a finite number of component parts. Such ordered sequence can serve as identifiers for characters. Take the 峰 and 峯 variants, for example; they are actually represented by the same unique identifier of 山夂丰 (as 夆 can be further specified as the result of top-down concatenation of two components 夂 and 丰, which cannot be broken down to further components). In addition, variants of the same characters in different historical or regional conventions can be identified by the same sequence. The sequence itself can be taken as instruction on how to realize these variants by combining the component parts using graphs according to the convention. For example, programmes have been developed to render characters in different modern fonts as well as historical conventions such as oracle bones and small seals. In turn, the same encoding sequence can be used to search for different historical orthographic conventions or regional variants. There are two principle ways to generate variants: by instantiating each component in different homographic forms according to the temporal or regional ‘font’ variations, or by implementing a different combinatory procedure (e.g. left to right or top to bottom) while still following the top-left first, bottom-right last general constraint. In terms of (computational) lexicography, the encoding system enables similar characters to be searched and compared in ways beyond the traditional zidian (character dictionaries) classifications of radicals. For instance, it is now possible to link 勞 lao ‘labor’ to 男 nan ‘man’ as both contain 力 li ‘effort’.

2.2.  Identification of words and words as lexical units Words as lexical entries have long been the cornerstone of modern lexicography. However, it should also be noted that identification of words is often dependent on orthographic conventions, and hence identification of words in a language which lacks conventions to mark word boundaries, like Chinese, can be challenging (Huang and Xue 2012). However, take either Bloomfield’s (1926) definition of ‘minimal free form’ or the lexicographic definition of ‘smallest meaningful unit’ (e.g. Jackson 2013; Francopoulo 2013), the main challenge in Chinese remains the lack of a set of operational criteria to define words. For instance, whether compounds or other multi-word units (such as idiom, chunks or proper names of persons and organizations) should be listed as an entry very often depends not on whether they are a word or not, but on the purpose and design criteria of a dictionary. With word dictionaries replacing character dictionaries as the default and more popular form of Chinese dictionaries, a clear operational definition of words as lexical units remains as a critical research topic in Chinese lexicography. 547

Huang Chu-Ren, Li Lan, and Su Xinchun

As words are basic units of a Chinese dictionary, two issues have received attention in recent lexicographic studies: the syllabicity of the Chinese language and the emergence of Romanized words. First, although the earlier fallacy that Chinese is a monosyllabic language has been debunked, the debate on whether a typical Chinese word should be mono- or disyllabic has continued (e.g. Su 2001). It is important to note that the percentage of monosyllabic words is limited by the number of characters, while there is no such constraint on the number of multisyllabic words. The corpus-based study of Huang et al. (2002) shed light on this complex issue. They showed that mono- and disyllabic words account for more than 90% of all instances of words in Chinese; and while there are more disyllabic words (in terms of word types) monosyllabic words tend to have higher frequency. Based on the 5-million-word part-of-speech (POS)-tagged and balanced Sinica Corpus they found that monosyllabic and disyllabic words each contribute to over 45% token frequency in Chinese. In terms of word types, however, disyllabic words compose over 46% of all word types (and monosyllabic words less than 3%, since there are only 6,000 or so commonly used monosyllabic words). In sum, the distributional strength of these two types of words differs in terms of word types (disyllabic words) and word frequency (monosyllabic words) hence either can be considered as the dominant prototype of Chinese. Second, it is crucial for modern lexicographers to recognize that not all Chinese words are rendered as characters. In fact, by different counts, there are at least 100 words in Chinese that are, typically, or only, written with alphabetic characters or a combination of alphabetic and Chinese characters. Examples are CCTV (China Central TV station 中央电视台), 阿Q (a fatalistic protagonist of Lu Xun’s novel meant to be a prototype Chinese person from the past, now referring to all people with that characteristic), and AA制 (‘to go Dutch’). Most Modern Chinese dictionaries now include alphabetic words although they (except for those starting with Chinese characters) are typically put in a separate section and are not listed together with the character-represented words. The lexicographic treatment of alphabetic words in Chinese remains an open research issue.

2.3.  Selection of lexical entries Once the issue of what is a lexical unit is determined, the selection of lexical entries poses yet another challenge. The selection of lexical entries for character dictionaries (zidian), is different from that in word dictionaries (cidian). However radically different they could be in content or format, they do not differ fundamentally in terms of entry selection. This shows that the inventory and use of Chinese characters, is highly conventionalized, just like English dictionaries. A survey of entries in word dictionaries in Chinese (cidian) however by Huang (1998), showed a very different story. He compared five different Chinese dictionaries or lexica compiled between 1993 and 1997 in China and Taiwan (but converted to common character representation), ranging from less than 40,000 entries to over 156,000 entries. They showed that the mutual coverage between any two dictionaries, defined by the mean of the entry coverage of dictionary A over dictionary B and vice versa, ranges from 49% to 68%. In particular, what is surprising is that bigger dictionaries do not have better coverage over smaller dictionaries. The 156,710 entry Revised Mandarin Chinese Dictionary (RMCD, from Taiwan) covers only 68.58% of the 70,325-entry ABC Chinese Dictionary (ABC, from the USA). And the 56,162 entry Xiandai Hanyu Cidian (XH) covers only 72.48% of the 39,459-entry Standard Segmentation Lexicon (GB), even though they are both from China and the compilation of GB consulted XH. Overall, they found that only 21,655 entries were shared by all five dictionaries and that corpus frequency alone was not a good predictor of 548

Lexicography in the contemporary period

these entries. Such variations, especially the failure of much larger dictionaries to include entries from much smaller dictionaries, underline the issue that Chinese lexicography has yet to develop a set of commonly accepted and inter-operable criteria for lexical entry selection.

3.  Chinese dictionaries and corpora By the sixth century ce two major frameworks of Chinese lexicographic arrangement has been established. The first is the radical-stroke system following Shuowen Jiezi, and the second is the rhyming system following Lu Fayan’s 陆法言 rhyme book Qieyun《切韵》(Lu 601 ce). The traditional fanqie 反切 system segments the monosyllabic pronunciation of a character into two parts: initial (consonant) and rhyme. For instance, the pronunciation of 党 dang is represented as 多朗切, which means that it takes the initial (‘d’ in modern Mandarin) from 多 (‘duo’ in modern Mandarin) to combine with the rhyme ‘ang’ (in modern Mandarin) of 朗 lang, forming dang in modern Mandarin. This system is an innovative way to represent phonological awareness without inventing a new set of symbols. It is also surprisingly robust since the initial-rhyme mapping can largely be preserved in spite of sound changes (as sound changes tend to apply to all members of the same class). What the Qieyun contributed to the fanqie is the explicit classification and naming of the rhyme classes. The Shouwen Jiezibased radical-stroke system is more popular and has persisted down to the present because its use requires only the basic knowledge of the Chinese writing system. The Qieyun system, however, requires expert knowledge of rhyme groups (as philologist or as poet) hence is limited to use by scholars. The commonly accepted phoneme-based representational framework in lexicography was not widely applied to the Chinese language before the vernacular language (baihua 白话) movement in the late 1920s. The clear need to represent the language as spoken arose and several different strategies were introduced to ‘write down’ the way people speak. Two of the systems are still commonly used today in both language teaching and in lexicography, as mentioned earlier in the introductory section. They are the phonetic alphabet system of zhuyinfuhao, which relies on a set of invented symbols to represent Mandarin Chinese phonemes, and the Pinyin Romanization which relies on the Roman alphabet. As mentioned before, they each lead to a different lexicographic system for indexing: bopomofo relies on location of articulation while Pinyin relies on English alphabetic order. It should be noted that there were several popular systems of Romanization for Chinese before Pinyin, especially among missionaries and second-language learners. The most prominent among them are the Wade–Giles system and the Yale system. Earlier bilingual dictionaries adopting these two Romanization systems also typically follow English alphabetical orders for organization of their content. Contemporary Chinese dictionaries are normally organized in four ways: 1. By Pinyin Romanization. Most Modern Chinese dictionaries use the Pinyin system for all the characters which makes it possible for dictionaries to be arranged in alphabetic order. Pinyin arrangement is essential for beginning first- and second-language learners who do not recognize Chinese characters yet. Since the users need to know the pronunciation of a word before using the dictionary, Pinyin-only organization can be challenging when a user is trying to look up the pronunciation for a new character/word. Hence it is not uncommon for such dictionaries to also come with a stroke or radical-based index. 549

Huang Chu-Ren, Li Lan, and Su Xinchun

2. By radical. Instead of the 540 radicals from Shuowen Jiezi, most Modern Chinese dictionaries adopt the 214-radical system based on the Kangxi Zidian published in 1716. A radical of a Chinese character is the part that indicates the meaning as related to the basic concept represented by that radical. The radical component has a conventional location in a character accordingly to its radical and its instantiated variant form, most often at left, right, top, or bottom, etc. of the character. In the case of a simple character, the word itself is a radical. The order of radicals is arranged in the Radical Index, usually at the front of a dictionary, according to the number of strokes constituting them, i.e. 木 (4 strokes) mu, which means ‘wood’ or ‘tree’. All the characters having the same radical are then ordered according to the number of strokes. For a word dictionary, the convention is to organize all words with the same initial character as one major entry then order the words with same initial character according to a secondary organization (usually by stroke number, or pronunciation, with words with identical second character clusters together, and so on and so forth). It is important to note that even though most characters in this group have the radical at the left, such as 杆 (3 strokes in addition to radical) gan ‘bar’; and 柱 (5 strokes) zhu ‘pillar’; there are also characters in this group with the referential mark on the character, such as 本 ben ‘root’; characters with the radical at the bottom, such as 柔 rou ‘flexible’; and on top, such as 查 cha ‘to check’. Similarly, since radical-based dictionaries are difficult to use without knowing how to write them or the number of strokes, a modern radical-based dictionary typically also has a stroke or pronunciation-based secondary index. 3. By stroke number. The stroke method refers to the total number of strokes that make up a single character. A monolingual Chinese dictionary usually has a list of ‘difficult words’ in the front matter, arranged in ascending number of strokes. The characters in this list usually have many strokes and it is difficult to determine their radicals. In order to count the strokes correctly it is essential to learn the correct stroke order of Chinese characters. Stroke counting is rarely used as the main organization method of Chinese dictionaries any more. 4. By Four-Corner method (sijiao haoma 四角號碼). The Four-Corner method was invented by Wang Yun-Wu 王雲五 in the 1920s, and the first dictionary organized by FourCorner method was published in 1928 (Wang 1928). This method is based on the fact that Chinese characters can each be considered as a glyph in a square. The assumption is that ten features from each corner of the square will be enough to represent all frequently used characters. The corners are numbered 0 to 9 in relation to their shapes. For instance, 木 has the number of 4090, 杆 4194, 柱 4091, and 及 1724. The advantage of the system is its economy and independence from either knowing how to write or pronounce the character. The disadvantage is, of course, that this new system needs to be learned and memorized independent of learning the language itself. It was popularly used for coding Chinese telegraphs and in early computation but is rarely used nowadays.

3.1.  Chinese character dictionaries (zidian) The modern paradigm of a Chinese character dictionary was set by Kangxi Zidian, which was commissioned by the Qing emperor Kangxi (1662–1722) and launched in 1716. It contains 47,035 characters arranged in a system with 214 radicals and number of strokes. Up to now the dictionary is still one of the most authoritative and widely used dictionaries of the Chinese language and has been used as the source of nearly all Chinese character dictionaries. 550

Lexicography in the contemporary period

The four-volume Zhonghua Da Zidian 《 ( 中华大字典》Comprehensive Chinese Dictionary of Characters), can be regarded as a decedent of Kangxi Zidian. It was edited by Xu, Y., Lu, F. and Auyang, P. and published by Zhonghua Publishing House in Shanghai in 1915. The dictionary added more than 1,000 new characters and corrected over 4,000 mistakes in the original Kangxi Zidian. The 48,000 entries were organised under the 214 Kangxi radicals. For characters with the same pronunciation, or homophones, different meanings were listed in the same entry. Although the dictionary has been updated separately a couple of times in China and in Taiwan, its acceptance has not surpassed Kangxi Zidian. The reason is probably due to its difficulties of use: index by the number of strokes and the lack of radical index in the body text (Teng and Biggerstaff 1971: 131). The mostly widely used Chinese character dictionary is Xinhua Zidian 《 ( 新华字典》New China Character Dictionary), which was published by the People’s Education Press in Beijing in 1953 with Wei Jiangong as the chief editor. The pocket-sized dictionary was compiled for education purposes: to promote Pinyin, Putonghua, and simplified Chinese characters. The book is the best-selling Chinese dictionary, an authoritative reference for the Chinese language and a compulsory tool for children to learn Chinese in primary education. Its 11th edition was published in 2011 by the Commercial Press in Beijing. Organised in Pinyin with a radical index, the dictionary is regarded as very convenient to use. However, its pocket size has limited the inclusion of characters to only 3,500 characters. Its online version is much bigger, with all the contents a Chinese dictionary may have: meaning, Pinyin, grammar, sense demarcation, usage, and different search methods with Pinyin, radicals, or number of strokes.

3.2.  Chinese word dictionaries (cidian) Ciyuan 《 ( 辞源》Sources of Words), edited by Lu Erkui, was published in 1915. It was a groundbreaking effort in Chinese lexicography, regarded as the first word dictionary in Chinese with emphasis on literary, historical, and classical terms (Hartmann 2003: 16). The book has been updated several times and has had many editions and reprints. The latest edition of Ciyuan contains 12,980 head entries, under which are 84,134 definitions of words and phrases, totaling 11.3 million characters. The four volumes are arranged by radicals with a Pinyin index at the end of the dictionary. Ciyuan has been positioned as a reference work for researchers and students of pre-Modern Chinese. Cihai 《 ( 辞海》Sea of Words) is another major dictionary of words in Chinese. Its main feature is encyclopedic coverage of history, philosophy, law, medicine, and science. It was first compiled by Shu Xincheng in 1938 and published by Zhonghua Book Company. The dictionary has been regularly revised since by Shanghai Lexicographical Publishing House. The American sinologist George A. Kennedy (1953: 131) regards Cihai as the basis for sinological studies, and the principal value of the dictionary lies in its explanations for compound expressions and its citations illustrating the use of words and expressions. The most authoritative dictionary of Modern Chinese is Xiandai Hanyu Cidian 《 ( 现代 汉语词典》The Contemporary Chinese Dictionary). The project started in 1958 at the Institute of Linguistics, Chinese Academy of Social Sciences, led by Lü Shuxiang and Ding Shusheng, two renowned Chinese linguists, and finally published in 1978 by the Commercial Press. The dictionary is a milestone in Chinese lexicography because it is the first Chinese word dictionary arranged in Pinyin, and providing a phonetic standard of the language. The Contemporary Chinese Dictionary is characterized by its clearly articulated criteria for selection of entries and for entry format as well as succinct definitions and illustrative examples. It is 551

Huang Chu-Ren, Li Lan, and Su Xinchun

now into its 6th edition (2012) and its bilingual (Chinese–English) version was published in 2002. The 6th edition contains about 69,000 entries including characters, words, and expressions and idioms. While pocket-sized Xinhua Zidian targets native learners at the primary to secondary levels, Xiandai Hanyu Cidian is a popular tool among students at tertiary levels and the general public. By 2002, 40 million copies had been sold. The compilers have attempted to add some new words in each edition, but admit they have been ‘open and cautious in the choice of new words and senses’ (Lü 2002: 8). Only 1,200 new words and new senses of existing words were added from 1996 to 2000. The above Chinese word dictionaries were mostly compiled with the linguistic intuition and expert judgement of the lexicographers involved. In recent revisions, language corpora were used, but probably just for citation or examples. There seems to be limited effort to take a corpus-driven approach in dictionary making in China.

3.3.  Corpora and dictionaries The field of dictionary making has long been influenced by empirical and corpus-based methods. However, the early text collections ‘did not mean to be representative of the language; rather, dictionary makers stressed the normative function of their work, aiming to describe the “proper” use of words’ (McArthur 1996: 235). Corpus today refers to a much larger collection of authentic data which is machine readable and can be processed by a computer with different queries. Language corpora have been used to construct dictionaries since the release of the Collins-Birmingham University International Database COBUILD (Sinclair 1987). The consensus among lexicographers and computational linguists is that statistical word modeling and corpus support are indispensible to modern dictionary compilation. Corpus linguistics benefits lexicography in three aspects: providing authentic texts, building lexical databases, and helping dictionary compilation. A number of Chinese mega-corpora have been compiled in the last three decades; some were sponsored by government, others were developed at institutional level. Compared to English corpora, constructions of Chinese corpora started late when better computer technology became available and corpus linguistic theories had been well developed. Unlike English corpora, few Chinese corpora have so far been constructed for the explicit goal of lexicography. The corpus by the Center for Chinese Linguistics (CCL) of Peking University is a corpus with more than 500 million characters. The data was collected with balanced genres of spoken language, fictions, popular magazines, newspapers and academic journals. Like many Chinese corpora, the main corpus was not segmented or tagged. The small portion (1 million words) which was tagged and annotated with different grammatical and semantic markers and used as the basis of the book The Grammatical Knowledge-base of Contemporary Chinese, has become a reference dictionary for Chinese language processing in many institutions worldwide. The Sinica Corpus was constructed in Academia Sinica in the 1990s under the direction of Keh-jiann Chen and Chu-ren Huang in Taiwan (Chen et al. 1996). It is the first fully POS-tagged balanced Chinese corpus as well as the first Chinese corpus to be available on the Internet. Like many modern balanced corpora its content distribution largely follows the original design of Brown Corpus but is also influenced by the designs of COBUILD and BNC. It is unique among Modern Chinese corpora to have the full corpus manually checked word by word for both its segmentation and POS-tagging after its initial automatic annotations. The Sinica Corpus is publicly available and freely searchable on the Internet (http://app.sinica. edu.tw/kiwi/mkiwi/). Its latest version, the Sinica 5.0, has more than 10 million words. 552

Lexicography in the contemporary period

Another widely used balance corpus of Chinese is the 1-million-word Lancaster Corpus of Mandarin Chinese (LCMC) (McEnery and Xiao 2004). Although smaller and later than the above-mentioned two corpora, the LCMC adopts the Brown/LOB (Lancaster–Oslo–Bergen) Balance Corpus format with 500 texts of roughly 2,000 words from 15 different genres. This conventional set-up allows users of the LCMC to readily compare it to English using the LOB or Brown corpus. However, its size and format constraints also means that it is often inadequate for modern computational lexicographic studies, which typically requires at least 10 million words (i.e. BNC size) of natural and non-truncated texts. Table 31.1 describes some important Chinese corpora. A lexical database generated from a corpus is the starting point of a corpus-based dictionary. It is normally built up by lexicon matching and statistic modeling. Generating an English wordlist is straightforward: words are separated by spaces so there is one-to-one correspondence between orthographic and morpho-syntactic word tokens. Chinese running texts are written without space, which means that words are not identified in the raw data. The first task in data processing is segmentation: to identify wordbreaks or segmented units which can then be used as processing units for other data (Huang and Xue 2012). Since both segmentation and POS-tagging in Chinese is non-trivial, many widely available Chinese corpora are not tagged. In addition, high-quality manually checked corpora tend to be smaller (usually a few million words, with 10-million-word Sinica Corpus being the largest). Larger tagged corpora, such as the 831-million-word tagged Gigaword Corpus (Huang 2009), are automatically tagged with only a small sample checked. The lack of sizeable Chinese corpora with high-quality tagging may have contributed to the fact that a limited number of corpora were used in Chinese lexicography. However, the few examples of corpus-driven dictionaries in Chinese do provide very promising results for future developments. Guoyu Ribao Liang Cidian 《 ( 国语日报量词典》The Mandarin Daily News Dictionary of Classifiers, Huang et al. 1997) published in Taiwan is probably the first fully corpus-driven Chinese dictionary. The Academia Sinica team selected classifiers as the target for the first attempt to compile a corpus-driven dictionary not only because the classifier is a unique feature of Chinese but also because the uses of classifiers depend crucially on their collocation with nouns (Chang et al. 1996). With the fully tagged Sinica Corpus the selection of lexical entries of classifiers can be automated by selecting the POS and setting a frequency threshold. This also means that all attested usages of classifiers and classifier–noun collocations can be extracted and studied for generalization. The research team identified 537 types of measure words from the Sinica corpus and set up a lexical database of the relevant grammatical information for each classifier, which was then exported through a dictionary interface for the dictionary entries. To fully utilize and explicate the corpus-based information, the dictionary contains two parts: a classifier dictionary and a noun-classifier collocation dictionary. The noun-classifier collocation dictionary is organized by the head of noun because the head of the noun determines the semantic class of the noun and hence predicts the selection of classifiers. For instance, regardless of the length and nature of the modifier X, all compound nouns X 牛 niu ‘cow, bull’ will take the classifier 头 tou. Chinese linguists and lexicographers are well aware of this characteristic of Chinese and dictionaries organized by the last (instead of first) character of words were occasionally compiled and referred to as reverse-order dictionary (nixu cidian 逆序辞典). However, such dictionaries are tedious to compile manually. With the tagged Sinica Corpus the compilation of reverse-order noun-classifier collocation involves the same automatic extraction rules, generating comprehensive data much more than the manually compiled ones. 553

Compiler

The State Language Commission of China

Academia Sinica, Taiwan

Chinese Linguistics Research Center, Peking University

Xiamen University Language research Centre, China

Leeds University, UK

Lancaster University, UK

Lexical Data Consortium, University of Pennsylvania, and Academia Sinica

Title

Modern Chinese Corpus

Balanced Corpus of Modern Chinese

CCL

Modern Chinese corpus

Chinese Internet Corpus

The Lancaster Corpus of Mandarin Chinese

Tagged Chinese Gigaword Corpus

Table 31.1  A list of Chinese corpora

2002–2004

1991–1993

2005

2001–2005

1996–2006

1992–2002

Time

Over 1,200 million characters, fully segmented and tagged (= 831 million words)

1 million words fully tagged (Brown-LOB format)

280 million (automatically segmented)

2,500 million characters

58 million characters

14 million characters fully segmented and tagged (= 10 million words)

100 million characters; 50 million characters segmented and tagged

Size

https://catalog.ldc.upenn.edu/ LDC2009T14

http://www.lancaster.ac.uk/ fass/projects/corpus/LCMC/

http://corpus.leeds.ac.uk/ query-zh.html

http://ncl.xmu.edu.cn

http://ccl.pku.edu.cn:8080/ ccl_corpus/

http://asbc.iis.sinica.edu.tw/OR http://app.sinica.edu.tw/ kiwi/mkiwi/

http://www.cncorpus.org

Website

Lexicography in the contemporary period

The next example of a corpus-driven dictionary is also based on a POS-tagged corpus but it is meant to be used both by a computer for natural language processing as well as by human readers. The Institute of Computational Linguistics of Peking University compiled Xiandai Hanyu Yufa Xinxi Cidian 《 ( 现代汉语语法信息词典》Grammatical Knowledge-base of Contemporary Chinese, Yu et al. 2001), listing over 70,000 words in 18 different categories with their grammatical and statistical information. Based on both linguistic and statistical analysis each entry word is marked with POS as well as its syntactic/semantic context and frequency. Working closely with collaborators in the Chinese Department of Peking University the team set up a very detailed segmentation and POS-tagging system. An electronic version of this dictionary has been used as a resource for many applications in Chinese language technology. Routledge’s A Frequency Dictionary of Mandarin Chinese: Core Vocabulary for Learners (Xiao et al. 2009) is a recent example of a corpus-driven Chinese dictionary published overseas for non-native speakers. The dictionary draws on the Lancaster Corpus of Mandarin Chinese (LCMC), a balanced 73-million-character Chinese corpus composed of spoken, fiction, nonfiction, and news texts in current use. The data was processed with the ICTCLAS, a Chinese Lexical Analysis System developed by the Institute of Computing Technology, Chinese Academy of Sciences, an automatic tool widely used in Chinese language processing in China. Since it is automatically processed the dictionary cannot go beyond the original 80,000 words in the system even with mechanism to guess word meaning based on role tagging (Zhang et al. 2002). From this list Xiao et al. found a similar distribution of mono- and disyllabic words as Huang et al. (2002) discovered: disyllabic words consist of most word types. The usually high token frequency of monosyllabic words at 54% is probably due to the fact that automatic segmentation and tagging typically fails to recognize many out-ofvocabulary words and leave parts of these words as monosyllabic words. Xiao et al. (2009) only extracted 84,883 word types from a 73-million-word corpus. In contrast, Huang et al. (2002) extracted nearly 200,000 word types from the manually checked 5-million-word Sinica Corpus, while Huang (2009) extracted nearly 3 million word types from the 831-millionword Tagged Chinese Gigaword Corpus v.2.0. With learners of Chinese in mind the dictionary by Xiao et al. provides the user with a detailed frequency-based list as well as alphabetical and POS indexes. All entries in the frequency list feature the English equivalent and a sample sentence in Chinese character, Pinyin and English translation. The dictionary also contains 30 thematically organized lists of frequently used words on a variety of topics such as food, weather, travel, and time expressions. The authors cherish the wish ‘to enables students of all levels to maximize their study of Mandarin vocabulary in an efficient and engaging way’. Kilgarriff (2013: 78) summarises a number of aspects of dictionary creation supported by the corpus: ● ● ● ● ● ● ● ● ●

Headword list development; For writing individual entries; Discovering the word senses and other lexical units (fixed phrases, compounds); Identifying the salient features of each of these lexical units; Their syntactic behavior; The collocations they participate in; Any preference they have in particular text-types or domains; Providing examples; Providing translations. 555

Huang Chu-Ren, Li Lan, and Su Xinchun

Data-driven Chinese dictionaries are generally for computers to align with words or mark word boundaries in a dataset; they are not for human use. They normally have part of the features in the above list. To meet users’ needs, more lexicographic information, such as collocations, usage notes, and examples should also be provided.

4.  Chinese dictionaries for foreign learners There are three types of learner dictionaries: monolingual, bilingual, and bilingualised, the last one combining information and translations from the previous two. The primary concern is the needs of target users. Some dictionaries are for native speakers to learn their mother tongue; others are for foreign learners to learn the language. The content and organisation can be very different because the learning needs, language problems, and learning focus of users are different. Chinese dictionaries for foreign learners are bilingual dictionaries mostly with Chinese and English. With the increased economic and political influence of China, the interest in learning Chinese is dramatically increasing worldwide. The need for Chinese dictionaries for foreign learners is obvious. Although Chinese publishers have made efforts to produce dictionaries for foreign learners, research has revealed that few foreign learners are using dictionaries published in China (Xie 2010; Li 2013).

4.1.  Features of learner’s dictionary The English learner’s dictionary, which was started in the 1930s in Japan by native English educators, has become a major branch of lexicography and a big business in the publishing world. After the Big Five learner’s dictionaries (by Oxford, Longman, Cambridge, Collins, and McMillan) were published in the UK, Merriam–Webster in the USA finally published its first advanced learner’s dictionary in 2008. Dictionaries of this type have striking features: they are user-cautious, supported by research findings in linguistics, cognitive science, and behavioural science. The word lists are carefully selected and the definitions of the words are explained by a core vocabulary, such as the 3,000 defining words in the Oxford Advanced Learner’s Dictionary (OALD). Most importantly, they are based not only on mega-corpora of native English language but also EFL writings so that the learning problems of users can also be addressed. The compilers make particular effort to help learners with traditional dictionary elements: grammar, pronunciation, definition, and example, as well as with innovative ideas in the order of senses, illustrations, learning panels, and appendixes. The success of English learner’s dictionaries in the world is due to the fact that the com­ pilation focuses entirely on users’ needs, although the importance of English is an obvious reason. The majority of Chinese–English dictionaries published so far target native Chinese users who may need dictionaries to translate texts from Chinese to English; grammatical and phonological information of Chinese is less relevant to this group of users. The dictionary for foreign learners is, by contrast, for users to render their ideas from English to Chinese. Much like a pedagogical grammar which receives sustenance and support from linguistic advances in language description the pedagogical dictionary is an educational aid, a major learning resource, whose form and function must be determined by its users. The best judgment on learners’ dictionaries is probably what Dr. Johnson noticed as long as 250 years ago: ‘In lexicography, as in other arts, naked science is too delicate for the purposes of life. The value of a work must be estimated by its use: It is not enough that a dictionary delights the critic, unless at the same time it instructs the learner.’ (Johnson 1747: 6). 556

Lexicography in the contemporary period

4.2.  Chinese dictionaries for foreign learners The history of Chinese bilingual lexicography reflects the development of China’s relations with foreign countries (e.g. Chien and Creamer 1986). An interesting historical development can be observed based on the chronogially ordered list of early bilingual dicitonaries given in Table 31.2, compiled based on the survey of Chien and Creamer (1986: 41–3) and our own research. Since the early sixteenth century, the interaction between China and the world can be attributed to the European interest in culture, and religion via missionary contact. However, after the Opium War (1839–42), which forced China to open trade opportunities, the role of England becomes more prominent and the emphasis of dictionaries helping foreigners to communicate with local people is underlined. This is attested by multiple dictionaries devoted to local vernacular languages in Southern China, where most direct contacts happen. Another important characteristic is the awareness of directionality of bilingual dictionary, as both English–Chinese (E–C) and Chinese–English (C–E) dictionaries were compiled and published equally frequently. Table 31.2  The first Chinese–foreign language bilingual dictionaries Time

Compiler

Dictionary

1583 1667 1670 1813 1815 1828 1832 1848 1853 1856

Matteo Ricci Michael Boym Michael Boym M. de Guignes Robert Morrison Robert Morrison Walter Henry Medhurst Walter Henry Medhurst Elihu Doty Samuel Wells Williams

1873

Castairs Douglas

1883 1892 1896

John Macgowan Herbert A. Giles Samuel Wells Williams

Dizionario Portoghese–Chinese Chinese–Latin Dictionary Chinese–French Dictionary Dictionnaire Chinois–Français et Latin Dictionary of the Chinese Language (Latin) Vocabulary of the Canton Dialect Dictionary of the Hok-këen Dialect of the Chinese Language English–Chinese Dictionary Anglo-Chinese Manual of the Amoy Dialect A Tonic Dictionary of the Chinese Language in the Canton Dialect Chinese–English Dictionary of the Vernacular Or Spoken Language of Amoy English and Chinese Dictionary of the Amoy Dialect A Chinese–English Dictionary A Syllabic Dictionary of the Chinese Language

The dictionaries in Table 31.2 aimed to help learners/users of Chinese to communicate with local Chinese people, for example, A Syllabic Dictionary of the Chinese Language was arranged according to the WuFangYuanYin 五方元音 (Gong 1660?), and records pronunciation of the characters as heard in Peking, Canton, Amoy, and Shanghai. The WuFangYuanYin was a Chinese word dictionary popular at that time after first being published in the late seventeenth century. It is one of the first Chinese dictionaries arranged phonologically, by multiplying 20 initial consonants and 12 rhyme groups. It is interesting to notice that even today dictionaries used by foreign learners of Chinese are mostly compiled on foreign soil, probably because the compilers, some of whom are even non-native Chinese speakers, can understand the needs of foreign learners better. Many C–E dictionaries have been published in China in recent years, but they are mostly for native Chinese users to write in English or to translate Chinese texts into English, although some claim to be for foreign learners. The entry is normally structured as below: 557

Huang Chu-Ren, Li Lan, and Su Xinchun

Chinese headword Pinyin English translations of the word Chinese example English translation of the example e.g. 曝光 [puguang] (喻) (将隐蔽的或不光彩的事公布于众) expose; reveal; make sth public; 对于贪污腐败给予公开~ give public exposure to graft and corrupt (A New Contemporary Chinese–English Dictionary) Entries like this are not very helpful to foreign learners. First, there is no grammatical information for the word; second, there is no Pinyin of the example. Usually beginner learners of Chinese concentrate on the spoken language rather than learning to read and write Chinese characters. If a reader cannot read characters the Chinese example cited is completely useless. The organization of a C–E dictionary for foreign learners can also be a problem. ‘Chinese characters are without a doubt cumbersome to index for foreign learners’ (Li 2013: 35). To use a C–E dictionary organized in Pinyin the user has to know how to read a Chinese word. To use a dictionary organized in radicals the user has to know how to write a Chinese character. Both are nearly impossible tasks for learners at elementary level. The complexity of Chinese makes the organization of C–E dictionaries very challenging. The Concise English–Chinese and Chinese–English Dictionary by Martin Manser with translation by Wu et al. (1999) is a good attempt to help beginners of Chinese. The 1st edition was published in 1986 by Oxford University Press in Hong Kong and the Commercial Press in Beijing. The book was nicknamed the ‘little red book’ by many foreign learners and received very good feedback. As the title indicates the dictionary has two parts. The A-to-Z English–Chinese section helps users to find both Pinyin and Chinese characters of an English word. The Chinese–English section is organised in Pinyin, enabling users to understand Chinese words. Each entry has a POS marker, usage example of the Chinese word, its English translation, and Pinyin of the whole example. The book has had four editions, with the latest in 2011. It has been among the best-selling Chinese–English books in the world. The ABC Chinese–English Dictionary was edited by John DeFrancis and published by Hawaii University Press in 1999. The dictionary has been welcomed by users in that the entries are organized in Pinyin rather than in characters. Its innovative organization clears up the misconception that the Chinese language is made up of monosyllabic characters. Most meaning units in Chinese are words with multiple syllables and hence with multiple characters. Because of this Chinese characters are without a doubt cumbersome to index. There will always be a few characters where it is difficult to figure out the exact number of strokes or the exact radical to find the character. With the ABC dictionary, as long as one knows the pronunciation of the word then it is very easy to find its meaning. This dictionary is particularly useful for finding words that one hears spoken but is not sure of the meaning. The ABC dictionary was compiled using lexical data from both China and Taiwan with simplified characters and their traditional forms. Where there are differences in usage between the two places, they are noted with PRC or TW in the entry. Another striking feature of the dictionary is that where homophones occur they are ranked in order of frequency. The dictionary does not have a radical index; therefore it is difficult to look up for a word if its pronunciation is unknown. 558

Lexicography in the contemporary period

4.3.  Chinese–English machine-readable dictionaries As Li (2013: 78) observed most foreign learners of Chinese find Chinese words hard to read, hard to recognize, hard to write, and hard to remember. The lack of popularity of Chinese learner’s dictionaries published in China is probably because of the difficulties in finding a word and lack of information on how to use the word. To cater for different users’ needs more space will be required in paper dictionaries. Machine-readable dictionaries (MRD) or electronic dictionaries are the only way to solve the problem. An MRD is a dictionary stored in an electronic form on a computer that can be linked to a database and queried in different formats via application software. MRDs can be loaded onto different electronic media such as on the Internet, on a PC, on a CD-ROM, on a hand-held gadget, and on a mobile phone. Some major Chinese dictionaries have their online versions now, such as Kangxi Zidian (http://kangxizidian.com/), Cihai (http://tool.gaofen.com/cihai/), and Xinhua Zidian (http://xh.5156edu.com/). Users can search a word by Pinyin or by radicals with number of strokes. There are also hyperlinks to related words, different word classes, synonyms, antonyms, grammar, history, and literature. With literally no limit to space, MRDs can have many multimedia features embedded within them, for example, the sound files of entries and examples, static graphics such as photos and colourful images, dynamic graphics such as animations and video clips, and hyperlinks to other resources, which have all been proved to be able to positively stimulate learning from different cognitive channels (Mayer 2005: 46). Mayer’s cognitive theory of multimedia learning (CTML) is based on three principles: the human information processing system includes dual channels for visual and pictorial and auditory/verbal processing; each channel has limited capacity for processing; and active learning entails carrying out a coordinated set of cognitive processes during learning (ibid.: 31). The most attractive advantage of MRDs are their speed and accuracy in locating information. The wildcard function can also save input time and correct user’s typing errors. Xie (2010) surveyed his students who were learning Chinese as a foreign language and found that few students used paper dictionaries; the majority used online dictionaries because they are fast and free. Also the high speed of dictionary consultation increased the amount of student reading because looking up a word in a paper Chinese dictionary was very time consuming. He believed that teachers cared more about the quality of e-dictionaries: whether the definitions are correct, the orders of senses are clear, and examples are easy to follow, while the students paid more attention to convenience and speed. Other dictionary user studies are consistant with this finding: learners ‘really want their dictionaries to be cheap, complete, portable, comprehensible and easy to use’ (Nesi 1999; Yu and Du 2010). However, there has been concern that with more reliance on e-dictionaries and online translation software the language acquisition process might not be as effective as in the traditional learning mode (Xie 2010: 61). As mobile devices have become part of our life, e-dictionaries have rapidly transferred to new media, in connection with apps by Android, iOS, and Windows. Similar to the rapid growth of language learning apps, mobile dictionaries have changed the design and use of reference works even further. A free app, Youdao Cidian (有道词典) by EasyNet, can translate words automatically in two directions between Chinese and English, Chinese and French, Chinese and Korean, or Chinese and Japanese. When a Chinese word is keyed in, its English equivalent will pop up. Further links can show Pinyin, grammar, and examples. If a user wishes to know more about the word, there is a link to lead them to the standard Chinese dictionaries Xiandai Hanyu Da Cidian (A Comprehensive Modern Chinese Dictionary) and Xin Hanying Da Cidian (The New Comprehensive Chinese–English Dictionary). It can 559

Huang Chu-Ren, Li Lan, and Su Xinchun

further draw bilingual examples from the iCloud and provide audio pronunciation of the sentences both in English and in Chinese. In addition, it links the search word with specialised dictionaries. Looking up an English word can obtain abundant information at different levels of request. The app also has a built-in look-up function. The user can get a pop-up screen with the meaning and/or translation of a word highlighted in an example. Modern technology can realize many lexicographic functions in MRDs which traditional lexicographers could never dream of. ‘Whatever the dictionary of the future will be like, there is still ample room for improvement, and the metalexicographer is in no danger of being unemployed: there is still much that has to be done in order to adapt the dictionary to its users and different uses’ (Béjoint 1994: 2).

5. Summary Chinese dictionaries are the records of the language and learning tools for both native Chinese users and foreign learners of Chinese. The needs of users differ, so the organization, components and examples of dictionaries should also vary. With the advancement of modern technology and corpus linguistics, automatic processing of Chinese has started to benefit dictionary compilation from the most basic issues of word identification and entry selection to the more advanced issues of word meaning definition and induction, identification of grammatical and pragmatic patterns, selecting examples, and modelling the language. Although the corpus approach has contributed greatly to the advancement of English lexicography (e.g. Sinclair 1987; Kilgarriff 2013), the same cannot be said of Chinese dictionary making yet, even though there were a few successful cases (e.g. Huang et al. 1997; Yu et al. 1998). The fact shows that there are significant gaps between computational/corpus linguists and lexicographers. In particular, there has not been significant dialogue between them, and the only examples of computational Chinese lexicographical works so far were done by computational and corpus linguists. In addition, in also seems that dictionary publishers in China have been more conservative and have not been involved in any significant research projects, such as COBUILD or subsequent projects by other publishers of English dictionaries. With an increasing demand from Chinese learners all over the world for high-quality, stateof-the-art references and dictionaries, the challenge is on now for Chinese lexicographers, corpus linguists, and dictionary publishers to work together to identify the user needs and resolve many outstanding issues in Chinese lexicography.

Bibliography Dictionaries Cihai Bianji Weiyuanhui (eds) (2009) Cihai 《 ( 辞海》Sea of Words) (6th ed.), Shanghai: Shanghai Lexicographic Publishing House. Ciyuan Xiuding Zu (1983) Ciyuan 《 ( 辞源》Sources of Words) (2nd ed.), Beijing: The Commercial Press. Fan, Tengfeng (1660?) WuFangYuanYin《五方元音》. Accessed at https://archive.org/details/02077328.cn. Hanyu Da Zidian 《 ( 汉语大字典》Comprehensive Chinese Character Dictionary) (2nd ed.), Sichuan: Sichuan Lexicographic Publishing House. Huang, Chu-Ren, Chen, Keh-jiann, and Lai, Ching-hsiung (eds) (1997) Guoyu Ribao Liang Cidian 《 ( 國語日報量詞典》Mandarin Daily News Dictionary of Classifiers), Taipei: Mandarin Daily Press. Kangxi Zidian 《 ( 康熙字典》Kangxi Dictionary) (1716) Reprinted 1993, Beijing: China International Culture Press.

560

Lexicography in the contemporary period Li, X. J. et al. (eds) (2004) Xiandai Hanyu Guifan Cidian 《 ( 现代汉语规范词典》Modern Standard Chinese Dictionary) Beijing: Foreign Language Teaching and Research Press. Lü, Shuxiang (2002) Xiandai Hanyu Cidian (Han–Yin shuangyu ben) 《 ( 现代汉语词典》(汉英 双语本) The Contemporary Chinese Dictionary (Chinese–English edition)), Beijing: Foreign Language Teaching and Research Press. Lü, Shuxiang (2012) Xiandai Hanyu Cidian 《 ( 现代汉语词典》The Contemporary Chinese Dictionary) (6th ed.), Beijing: The Commercial Press. Manser, Martin (1999) Concise English–Chinese and Chinese–English Dictionary (translated by Wu Jingrong, Mei Ping, Zhu Yuan, and Liangbi Wang), Beijing: The Commercial Press and Oxford University Press. The Chinese Academy of Social Science (eds) (2011) Xinhua Zidian 《 ( 新华字典》The Xinhua Dictionary) (11th ed), Beijing: The Commercial Press. Wang, Yunwu (1928) Wang Yun-wu Da Cidian 《 ( 王雲五大辭典》The Comprehensive Wang Yun-wu Dictionary). Shanghai: Commercial Press. Xiao, Richard, Rayson, Paul and Tony McEnery (2009) A Frequency Dictionary of Mandarin Chinese: Core Vocabulary for Learners, London and New York: Routledge. Xu, Yuanhao and Fucun Ouyang (1978) Zhonghua Da Zidian 《 ( 中华大字典》Comprehensive Chinese Dictionary of Characters), Beijing: Chinese Publishing House. Yao, Naichiang (2003) Xinhua Dictionary with English Translation. Hong Kong: Commercial Press. Yu, Shiwen (1998) Xiandai Hanyu Yufa Xinxi Cidian 《 ( 现代汉语语法信息词典》Grammatical Knowledge-base of Contemporary Chinese), Beijing: Tsinghua University Press. Zhang, Jing (2002) Dangdai Xinbian Hanying Cidian 《 ( 当代新编汉英词典》A New Contemporary Chinese–English Dictionary), Shanghai: The World Book Publishing House.

References Béjoint, Henry (1994) Tradition and Innovation in Modern English Dictionaries, Oxford: Clarendon Press. Bloomfield, Leonard (1926) ‘A Set of Postulates for the Science of Language’, Language 2: 153–64. (Reprinted in Hockett, 1970: 128–38). Chang, Li-Li, Chen Keh-jiann, and Chu-Ren Huang (1996) ‘The Use of Corpus in the Compilation of Dictionaries’, in Proceedings of the Ninth R. O. C. Computational Linguistics Conference, 255–79. Also in ACL Anthology. Chen, Keh-jiann, Huang Chu-Ren, Chang Li-Ping, and Hsu Hui-Li (1996) ‘Sinica Corpus: Design Methodology for Balanced Corpora’ in B. S. Park and J. B. Kim (eds) Proceeding of the 11th Asia Pacific Conference on Language, Information and Computation, Seoul: Kyung Hee University, 167–76. Chien, David and Thomas Creamer (1986) ‘A Brief History of Chinese Bilingual Lexicography’, in Reinhard Hartmann (ed) The History of Lexicography, Amsterdam and Philadelphia: John Benjamins Publishing Company, 36–45. Francopoulo, Gil (2013) (ed) LMF – Lexical Markup Framework, ISTE/Wiley. Francopoulo, Gil and Chu-Ren Huang (2014) ‘Lexical Markup Framework: An ISO Standard for Electronic Lexicons and its Implications for Asian Languages’, Lexicography 1(1): 37–51. Hartmann, Reinhard (2003) Lexicography: Reference Works across Time, Space, and Languages, London: Taylor and Francis. Hsieh, Ching-chun 谢清俊 (1996)〈电子古籍中的缺字问题〉(The Issue of Missing Characters in Electronic Classics), in《第一届中国文字学会学术讨论会》(Proceedings of the First Conference of the Association of the Chinese Characters), 25–30 August, Tianjin. Huang, Chu-Ren (1998) ‘Criteria for Computational Chinese Lexicography: A Study based on a Standard Reference Lexicon for Chinese NLP’, in Proceedings of ROCLING XI, 87–108, also in ACL Anthology. Huang, Chu-Ren (2009) Tagged Chinese Gigaword Version 2.0. Philadelphia: Lexical Data Consortium, University of Pennsylvania, Available at: https://catalog.ldc.upenn.edu/LDC2009T14 Huang, Chu-Ren and Nianwen Xue (2012) ‘Words without Boundaries: Computational Approaches to Chinese Word Segmentation’, Language and Linguistics Compass 6(8): 494–505. Huang, Chu-Ren, Chao-Ran Chen, and Claude C. C. Shen (2002) ‘The Nature of Categorical Ambiguity and Its Implications for Language Processing: A Corpus-based Study of Mandarin Chinese’, in Mineharu Nakayama (ed) Sentence Processing in East Asian Languages, Stanford: CSLI Publications, 53–83.

561

Huang Chu-Ren, Li Lan, and Su Xinchun Jackson, Howard (ed) (2013) The Bloomsbury Companion to Lexicography, London: Bloomsbury. Johnson, Samuel (1747) ‘The Plan of an English Dictionary’, in The Works of Samuel Johnson, LL.D.: With an essay on his life and genius, London: Luke Hensard & Sons (printed in 1820). Jhuang, Der-Ming, Jenq-Haur Wang, Chen-Yu Lai, Ching-Chun Hsieh, Lee-Feng Chien, and Jan-Ming Ho (2005) ‘Resolving the Unencoded Character Problem for Chinese Digital Libraries’, Paper to the Joint Conference on Digital Libraries 2005, Denver, Colorado, 7–11 June 2005. Kennedy, George (1953) ZH Guide: An Introduction to Sinology, New Haven: Yale University Press. Kilgarriff, Adam (2013) ‘Using Corpora as Data Sources for Dictionaries’, in Jackson, H. (ed) The Bloomsbury Companion to Lexicography, London: Bloomsbury, 77–96. Li, Yan (2013) ‘On the Compilation of General Purpose Chinese Dictionaries for Foreign Learners of Chinese’, Cishu Yanjiu (Lexicographical Studies) 5: 34–9. Mayer, Richard (ed) (2005) The Cambridge Handbook of Multimedia Learning, Cambridge: Cambridge University Press. McArthur, Tom (1996) The Oxford Companion to the English Language, Oxford: Oxford University Press. McEnery, A. and Xiao, Z. (2004) ‘The Lancaster Corpus of Mandarin Chinese: A Corpus for Monolingual and Contrastive Language Study’, Paper presented at LREC 2004, May 2004, Lisbon. Nesi, Hilary (1999) ‘A User’s Guide to Electronic Dictionaries for Language Learners’, International Journal of Lexicography 12(1): 55–66. Sinclair, John (1987) Looking Up: An Account of the COBUILD Project in Lexical Computing and the Development of the Collins COBUILD English Language Dictionary, London: Collins COBUILD. Su, Xinchun (ed) (2001) A Statistical Analysis of the Chinese Lexicon, Xiamen: Xiamen University Press. Teng, Suyu and Knight Biggerstaff (1971) An Annotated Bibliography of Selected Chinese Reference Works (3rd ed.), Cambridge, MA: Harvard University Press. Xiao, Z., A. McEnery, P. Baker, and A. Hardie (2004) ‘Developing Asian Language Corpora: Standards and Practice’ in V. Sornlertlamvanich, T. Tokunaga, and C. Huang (eds) Proceedings of the Fourth Workshop on Asian Language Resources, 25 March, Sanya, 1–8. Xie, Tianwei (2010) ‘Using Online Dictionaries in Teaching Chinese’, Journal of Chinese Language Teachers Association 45(3): 53–65. Xue, Shiqi (2003) ‘Chinese Lexicography Past and Present’, in Lexicography: Critical Concepts II, London and New York: Routledge, 158–73. Yu, Pingfang and Jiali Du (2010) A Comparative Study on Learner’s Dictionaries in English and in Chinese, Beijing: China Social Sciences Press. Zhang, Huaping, Liu Qun and Cheng Xueqi (2002) ‘Automatic Recognition of Chinese Unknown Words Based on Role Tagging’, in Proceedings of the 1st SIGHAN Workshop, COLING 2002, Taipei, 71–7.

562

32 Loanwords Miao Ruiqin Shanghai Jiao Tong University, China

1. Introduction Loanwords are called wàiláicí 外来词 (words from foreign languages) or jiècí 借词 (loan words) in Chinese. In the history of loanword studies on Chinese, various other terms have been proposed, for example, yìyǔ 译语 (translated expressions), yìmíng 译名 (translated terms), wàiláiyǔ 外来语 (expressions from foreign languages), jièyǔ 借语 (loan expressions), and jièzì 借字 (loan characters) (e.g. Masini 1993: 129–30; Yeung 2005: 8–38; Shi 2013: 13–18). At present, wàiláicí and jiècí are the most commonly used. The Chinese language is noted for its resistance to lexical borrowing from foreign languages, and especially for its disfavor against loan adaptation through phonemic transliteration (vs. meaning translation) (Cheng 1985; Norman 1988: 16–22; Chan and Kwok 1990: 18–21; Kuiper 1993: 118; Masini 1993: 141–5; Chen 1999: 105). The main reasons for this are the linguistic differences between Chinese and other languages (mainly Indo-European ones), ranging from phonology, morphology to orthography. Other factors include the desire for linguistic ‘purity’ and national or social attitudes towards loan nativization (Cheng 1985; Chan and Kwok 1990: 21–3). Despite its intolerance towards heavy borrowing, Chinese has imported a large number of foreign terms in diverse fields. Loans entering the language in various historical periods constitute an integral part of the Chinese lexicon. They play an important role in bridging communication between Chinese and other peoples or cultures. Some of them have become so deep-rooted and extensively used in daily life that Chinese speakers seldom realize that these words were in fact borrowed from other languages, for example chànà 刹那 ‘a short moment’ from Sanskrit, mǎdá 马达 ‘motor’ from English, and jīngjì 经济 ‘economy’ from Japanese. This chapter aims to give a general introduction to loanword adaptation and use in Chinese. Before detailed discussions are given, the use of some terminologies and format of example words in the text need to be noted here. First, the term ‘loanword’ (or ‘loan’) is used in a broad sense to include all types of borrowings, including phonemic loans, semantic loans, graphic loans, and hybrid loans. (For various views on the definition of ‘loanwords’ in Chinese, see Gao and Liu 1958: 7–10; Masini 1993: 128–34; Shi 2013: 238–41). Second, unless otherwise noted, ‘Chinese’ refers to ‘Mandarin’ (or interchangeably ‘Mandarin Chinese’), which is the standard Chinese variety in mainland China (i.e. Pǔtōnghuà 普通话) and Taiwan (i.e. Guóyǔ 国语). Occasionally it is used as an umbrella term covering all Chinese dialects. 563

Miao Ruiqin

Third, loanword examples and other Chinese words are presented in both Pinyin and simplified characters, with Pinyin in italic type. Lastly, the symbol ‘ [phonemic, semantic] hybrids, phonemic loans with semantic associations > pure phonemic loans > [graphic, semantic] hybrids > [graphic, phonemic] hybrids > alphabetic words

native-like adaptations tend to be used in informal and colloquial situations, whereas more native-like forms are more likely to occur in formal speech and writing. Based on the classification of loanwords discussed in §3, the native-likeness of different types of loans is roughly ranked in Table 32.4, in which the symbol ‘>’ means ‘more nativelike than’. The most native-like loans are semantic loans and Japanese graphic loans, with the least ones being alphabetic words. It can be predicted that among alternative adaptations of a same foreign word, a semantic loan may appear more frequently in formal contexts than other versions (e.g. a phonemic loan). In contrast, less native-like forms such as phonemic loans and alphabetic words may be preferred in informal domains (e.g. an online chat room). Take the two Chinese adaptations of English ‘E-mail’ as examples. Currently there are two loan forms, i.e. diànzǐ yóujiàn 电子邮件 and yīmèiěr 伊妹儿. The first one is a semantic loan originating from a morphemeby-morpheme translation of the English word. It occurs more often in formal speech and writing, e.g. news reports and technological texts. In contrast, the second one is a phonemic loan. It is more likely to be used in casual talk and by young people, and sometimes it conveys a facetious or humorous tone.

5.2.  Regional variations Due to historical and political reasons, sociolinguistic situations of language borrowing vary across the Chinese-speaking communities in mainland China, Taiwan, Hong Kong and Macao, and hence discrepancies exist between loanword adaptations in these places. For instance, ‘toast’ is phonemically transliterated as dosì 多士 in Hong Kong Cantonese, but semantically translated as kǎomiànbāo 烤面包 in mainland Mandarin. More examples are given in Table 32.5. (Macao generally follows the same pattern as Hong Kong, and thus it will not be discussed separately.) Table 32.5  Regional variations English

Mainland China

Taiwan

Hong Kong

hacker▲

黑客 hēike

骇客 hɑ́ikè

show▲

演出 yɑ̌nchū 表演 biɑ̌oyɑ̌n 悉尼 xīní 布什 bùshí 新西兰 xīnxīlɑ́n 里根 lǐgēn 程序 chéngxù

秀 xiù 表演 biɑ̌oyɑ̌n 雪梨 xuělí 布希 bùxī 纽西兰 niǔxīlɑ́n 雷根 léigēn 程式 chéngshì

黑客 hēikè 骇客 hɑ́ikè 骚 sɑ̄o 表演 biɑ̌oyɑ̌n 雪梨 xuělí 布殊 bùshū 纽西兰 niǔxīlɑ́n 列根 liègēn 程式 chéngshì

Sydney▲ Bush▲ New Zealand Reagan▲ program*

Notes: (a) Sources: ▲ Shi (2013: 91–92), and * Lü and Dai (2000: 251–252); (b) For all varieties, the pronunciation given is in Mandarin Pinyin.

575

Miao Ruiqin

Generally speaking, the major differences between these places lie in two areas: (i) Phonemic loans in mainland China (M) are more accurate transliterations of foreign pronunciations than those in Taiwan (TW) and Hong Kong (HK), e.g. in ‘Sydney’ → xīní 悉尼 (M) vs. xuělí 雪梨 (TW, HK); (ii) Foreign person names (especially surnames) tend to be adapted into Taiwan by using a Chinese name too, which is seldom observed in mainland China and Hong Kong, e.g. ‘Reagan’ → léigēn 雷根 (TW) vs. lǐgēn 里根 (M) vs. liègēn 列根 (HK). (For more detailed discussions, see Lü and Dai 2000: 249–54; and Lou 1992).

5.3.  Loanword standardization Variant adaptations, either within the same Chinese community (e.g. mainland China) or across different communities, cause potential problems of miscommunication. Standardization of loanword adaptation and use is drawing increasing attention among Chinese scholars. Various proposals have been advanced regarding how to establish a standard form among alternative adaptations of a foreign word. An influential work is Gao and Liu (1958), who proposed a ‘Three One’ Principle, i.e. to establish one standard written form and one standard pronunciation for the loan form of one and a same foreign word. They further expounded this principle by advancing six rules, including extensive use, faithful transliteration, simplicity, historical conventions, semantic explicitness, and conformity to Chinese word-building grammar (1958: 178–81). Meanwhile, they recognized the necessity of allowing for the co-existence of variant adaptations (e.g. for stylistic distinctions) (1958: 181–3). Although advanced over 50 years ago, Gao and Liu’s proposal is still applicable today. Other selection criteria such as syllable structure, word length and ease of pronunciation have also been suggested (e.g. Shi 2003: 219–20). In mainland China, there are government institutions in charge of the adaptation of foreign scientific and technical terms, e.g. National Commission of Language and Script (国家语言 文字工作委员会) and General Administration of Press and Publication (新闻出版总署). Standard lists of characters are recommended for transliteration of foreign places and surnames (Lou 1992). In addition, standardization of alphabetic words has attracted much discussion in recent years. One major question is how to pronounce them. For example, in the case of alphabetic words borrowed from English, it is not clear whether they should be read according to the pronunciation of the letters in English, or in a Sinicized version of English, or in Chinese Pinyin (Zou 2012: 6–8). To reduce and eliminate potential miscommunication, it is advisable for government and academic institutions in Chinese-speaking regions to cooperate on the standardization of loanwords. For instance, comparative studies can be conducted to outline the relevant norms and policies in different communities, and then general guidelines for loan adaptation and use can be drawn up.

6. Conclusion Language borrowing into Chinese occurs along with and as a result of China’s cultural, political and economic contacts with other countries and regions in the world. Loanwords have not only played an important role in bringing new knowledge, thoughts, concepts and objects into China but also contributed significantly to the development of the Chinese language itself. 576

Loanwords

With the pace of China’s economic and social development increasing, it can be safely predicated that lexical borrowing will continue to present a dynamic scene in the coming decades. In particular, two trends have emerged. One is that the proportion of loanwords in the Chinese vocabulary is rising. According to Wang’s (1993) survey, loanwords account for 3.4% of the total Chinese vocabulary in the post-1978 period, which contrasts with a proportion of less than 0.75% in the previous 30 years (cited in Zuo 2005: 284). The other trend is that the number of direct borrowings, especially alphabetic words from English, may increase rapidly. English functions as a global language and the English proficiency of average Chinese speakers is improving, and thus the coming decades may see even more alphabetic words enter daily Chinese speech and writing. These situations present challenges for scholars and policy-makers working on the standardization of the Chinese language in the twenty-first century.

Acknowledgement The writing of this chapter was supported by the Shanghai Pujiang Program, China (Grant No. ZXDF069).

References Chan, Mimi and Helen Kwok (1990) A Study of Lexical Borrowing from English in Hong Kong Chinese, Hong Kong: University of Hong Kong. Chen, Ping (1999) Modern Chinese: History and Sociolinguistics, Cambridge: Cambridge University Press. Cheng, Robert L. (1985) ‘Group Interest in Treating Words Borrowed into Mandarin and Taiwanese’, in Bjorn H. Jernudd (ed) Chinese Language Contact, Anthropological Linguistics 27(2): 177–89. Gao, Minkai 高名凯 and Liu, Zhengtan 刘正埮 (1958) Xiandai Hanyu Wailaici Yanjiu《现代汉语 外来词研究》(Foreign Loans in Modern Chinese), Beijing: Language and Writing Reform Press. Kuiper, Koos (1993) ‘Dutch Loan-words and Loan-translations in Modern Chinese: An Example of Successful Sinification by Way of Japan’, in Lloyd Haft (ed) Words from the West: Western Texts in Chinese Literary Context, Leiden: Centre of Non-Western Studies, 116–44. Lou, Chengzhao (1992) ‘Transliterating Non-Chinese Proper Nouns into Chinese: A Comparative Study of Usage in Mainland China, Taiwan and Hong Kong’, International Journal of the Sociology of Language 97: 121–33. Lü, Jiping 吕冀平 and Dai, Zhaoming 戴昭铭 (eds) (2000) Dangqian Woguo Yuyan Wenzi de Guifanhua Wenti《当前我国语言文字的规范化问题》(Current Issues on the Standardization of Language and Writing in China), Shanghai: Shanghai Educational Publishing House. Masini, Federico (1993) The Formation of Modern Chinese Lexicon and Its Evolution towards a National Language: The Period from 1840 to 1898, Journal of Chinese Linguistics, Monograph Series. No. 6. Miao, Ruiqin (2005) ‘Loanword Adaptation in Mandarin Chinese: Perceptual, Phonological and Sociolinguistic Factors’, Doctoral Dissertation, Stony Brook University, available from: https:// linguistics.stonybrook.edu/sites/default/files/uploads/Miao_2005.pdf. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Shi, Youwei 史有为 (2004) Wailaici: Yiwenhua de Shizhe《外来词: 异文化的使者》(Loanwords: Envoys of Foreign Cultures), Shanghai: Shanghai Lexicographical Publishing House. Shi, Youwei (2013) Hanyu Wailaici (zengdingben)《汉语外来词 (增订本)》(Loanwords in Chinese (Rev.)), Beijing: The Commercial Press. Silverman, Daniel (1992) ‘Multiple Scansions in Loanword Phonology: Evidence form Cantonese’, Phonology 9: 289–328. Yeung, Hong-ting 杨康婷 (2005) ‘A Study of Loan Words in Chinese Language in Hong Kong’, Master Thesis, The University of Hong Kong, available from: http://hub.hku.hk/handle/10722/32172.

577

Miao Ruiqin Zhou, Zhenhe 周振鹤 and You Rujie 游汝杰 (1986) Fangyan yu Zhongguo Wenhua (xiudingben) and Chinese Culture (Rev.)), Shanghai: Shanghai People’s 《方言与中国文化(修订本)》(Dialects   Publishing House. Zou, Yuhua 邹玉华 (2012) Xiandai Hanyu Zimuci Yanjiu《现代汉语字母词研究》(A Study of Alphabetic Words in Modern Chinese), Beijing: Language and Culture Press. Zuo, Xiulan (2005) ‘Language Planning with Respect to English into China’, Terminology 11(2): 283–292.

578

33 Mandarin Shi Dingxu Hong Kong Polytechnic University, Hong Kong, China

1. Introduction The term ‘Mandarin’ usually stands for two things in language studies. In a broad sense, Mandarin refers to the Northern dialect family, whose speakers live north of the Yangtze River. In a narrow sense, Mandarin refers to the dialect of Beijing and the nearby area. The name Mandarin was borrowed into English in the seventeenth century from the Portuguese stationed in Macao. It originally referred to the military officials of the Ming and Qing dynasty and consequently became the name of the language commonly used by officials. The city of Beijing took shape 2,000 years ago. It was one of the capitals of the Liao dynasty (ad 907–1125) and the capital of the Jin dynasty (ad 1115–1234), which were established by the nomadic people outside China proper. When the Mongols defeated the Jin and conquered China to establish the Yuan dynasty (ad 1260–1368), they set up their political headquarters in this city and named it Dadu, the Big Capital. When the Chinese drove the Mongols out to establish the Ming dynasty around 1368, the new capital was in Nanjing, the South Capital. In 1402 a prince usurped the throne and moved the capital to the site of Dadu to stay away from the establishment. He ordered a new city to be built next to the ruins of Dadu and named it Beijing, the North Capital. The construction of Beijing was completed in 1421 but the city had very few inhabitants. The emperor drafted a large number of people from various places and moved the whole imperial court to Beijing. The population grew quickly, and people being transplanted to Beijing had to communicate with each other in their own dialects, most of which belonged to the Northern dialect family. It is likely that a new dialect was formed on the basis of these sub-dialects, especially the variety of Jianghuai dialect in and around Nanjing, but the new dialect only lasted for 200 years because the Ming dynasty was overthrown by a peasant uprising. The rebels occupied Beijing for a while but were forced out in 1644 by the Manchu army, known as the Banner Army. When the Manchu rulers decided to move their imperial court from Manchuria to inland China, Beijing became the natural choice for the capital, since the Imperial Palace was in one piece and most residents had stayed behind. The Qing court took over the Forbidden City; the Manchu nobles, the Banner Army and their families took the inner city; and the Chinese people were driven to the outer city. 579

Shi Dingxu

Once again, Beijing became a city of mixed races but this did not change the language situation too much, since the Manchu people quickly adopted the Chinese language even in the imperial court. There were also a large number of Chinese speakers within the Qing army, in the Chinese wing of the military force known as the Han Banner Army or as servants in the Manchu Banner Army. The Chinese speakers in the Qing army came from the area around modern-day Shenyang and their dialect inevitably had a strong influence on the Chinese language spoken in Beijing. By the end of the nineteenth century, the linguistic form used by the Banner people had become representative of the Beijing dialect known as Jingpianzi, i.e. the colloquial speech of the capital. To a great extent, that is why the Beijing dialect got the name Mandarin. The Qing dynasty was replaced by the Republic of China in 1911, and the capital was moved several times after that. It was in Beijing from 1912 to 1927, in Nanjing from 1927 to 1937, in Chongqing from 1937 to 1945, and was moved back to Nanjing in 1945. The linguistic situation in Beijing remained almost the same until 1949, when it became the capital of the People’s Republic of China. The new political and economic importance brought huge amounts of people to Beijing and many of the newcomers were adults from the south, whose influence on Mandarin was inevitably strong. By the 1960s, the dialect in Beijing had evolved into two varieties, a new one and an old one. The Old Mandarin is spoken by the original residents of Beijing and retains most characteristics of Jingpianzi. The New Mandarin, used mainly by the second generation of the newcomers, has lost some features of Jingpianzi and come closer to Putonghua, the lingua franca promoted by the Chinese government. What is being discussed here are mainly features of the New Mandarin, while Putonghua will be discussed in the chapter on Standard Chinese.

2.  The sound system Mandarin is a tone language with a relatively small sound inventory and simple syllable structure. Speech sounds in Mandarin can be classified into three categories: consonants, glides and vowels.

2.1. Consonants It is commonly assumed that there are 22 consonants in Mandarin (Table 33.1). The consonants in Table 33.1 are recorded with broad-transcription, which is believed to better reflect the native speakers’ perception. Phonologically there are 19 consonants in Table 33.1  Consonants of Mandarin Plosive

bilabial labiodental dentoalveolar alveolar alveopalatal retroflex velar

unaspirated

aspirated

p

ph

t

th

k

kh

Affricate unaspirated

Fricative

aspirated

voiceless

ts

tsh

f s

tɕ tʂ

tɕh tʂh

580

ɕ ʂ x

nasal

liquid

voiced m n ʐ

ŋ

l

Mandarin

Mandarin, excluding the alveopalatal sounds [tɕ], [tɕh] and [ɕ], which occur only in front of high front vowels [i] and [y], or palatal glides [j] and [ɥ]. They are thus in complementary distribution with the alveolar sounds, the retroflex sounds and the velar sounds. It is reasonable to analyze alveopalatal sounds as allophones of alveolar, retroflex or velar sounds, even though there is no consensus on the actual affiliation. The fricative [ɕ] is alveopalatal in that its place of articulation is close to the back part of the hard palate and the same can be said about the affricate [tɕh]. The unaspirated status of [tɕ] is due to the unaspirated nature of [t]. There is no voiced plosive in Mandarin. The contrast between unaspirated and aspirated plosive is phonemic and the three pairs of voiceless plosives, /p/ and /ph/, /t/ and /th/, and /k/ and /kh/, are all phonemes. Similarly, the voiceless affricates /ts/ and /tsh/ contrast with each other on aspiration and the pair /tʂ/ and /tʂh/ have the same contrast. The pair /tɕ/ and /tɕh/ also contrast with each other on aspiration but they are not phonemes. The retroflexes are quite unique to Mandarin. The sibilant fricative retroflex [ʂ] is an apical palatal sound that is produced by lifting the tip of the tongue and pulling it back against the hard palate. Both sides of the tongue are pushed against the upper teeth to form an opening between the tongue and the palate for the air to go through. The tongue is not curled up as commonly believed, and what is being pushed up against the front palate is an area of the tongue blade close to the tip. The retroflex [ʐ ] is a fricative produced by forcing the air through a narrow opening between the front part of the tongue blade and the hard palate while the tip of the tongue is near the upper alveolar ridge. The dento-alveolar sibilant [s] is a pre-tongue-tip sound and the alveopalatal sibilant [ɕ] is a pre-tongue-blade sound. [s] is produced by pushing the tongue tip against the lower alveolar ridge to form a narrow opening between the area after the tip and the upper alveolar ridge. [ɕ] is produced by lifting the tongue blade up to form an opening between the front part of the blade and the palate. The pronunciation of [tʂ] begins with closing the vocal tract to produce [t] and then releasing the closure to produce [ʂ]. Since [t] is a voiceless unaspirated plosive and [ʂ] is a retroflex, [tʂ] is a voiceless unaspirated retroflex affricate. Similarly, [tʂh] is a voiceless aspirated retroflex affricate, [ts] is a voiceless unaspirated dento-alveolar affricate, [tsh] is a voiceless aspirated dento-alveolar affricate, [tɕ] is a voiceless unaspirated alveopalatal affricate and [tɕh] is avoiceless aspirated alveopalatal affricate. Another fricative in Mandarin is the voiceless velar fricative [x], which sounds a little like the American English voiceless glottal glide [h]. The liquid [l] in Mandarin is quite similar to the [l] in languages like English in terms of place and manner of articulation. The same can be said about the nasals [m], [n] and [ŋ]. The only noticeable exception is that the alveolar nasal [n] has a free variation alveopalatal [ ɲ], when it occurs before high front vowels [i] and [y] or glides [ j] and [ɥ].

2.2. Glides There are three glides in Mandarin as shown in Table 33.2. As semi-vowels, they are all sonorant because the air is unobstructed as it moves through the oral cavity in production. The palatal [j] and the labio-velar [w] are similar to the glides [j] and [w] in English. The pronunciation of [ɥ] is quite similar to that of the high front rounded vowel [y], with the lips protruded out to form a small round opening and the front part of the tongue raised to become the highest point. [ɥ] only occurs before a vowel while [j] and [w] may occur either before or after a vowel. 581

Shi Dingxu Table 33.2  Glides of Mandarin

Glide

Labio-palatal

Palatal

Labio-velar

ɥ

j

w

2.3. Vowels The number of vowels in Mandarin has always been controversial. For phonologists, the number of vowel phonemes ranges from two all the way up to eight or nine; and for phoneticians, the number could go up to over 40. The discussion here is based on the broad-transcription of Mandarin, and the number of vowels is 13 (Table 33.3). The unrounded high front vowel [i] is quite similar to the English [i] in words like sheep, and both are long (tense) vowels. The other two unrounded high front vowels [ɿ] and [ʅ] are quite unique. [ɿ] only occurs in syllables with [ts], [tsh] or [s] as the onset. When the production of the onset is completed, the tongue stays in the position against the lower alveolar ridge, the front part of the tongue blade is close to the upper alveolar ridge, and the vocal cords start to vibrate. The whole process clearly starts with a voiceless sibilant and ends with the vibration of vocal cords without apparent obstruction of the airflow. The two segments form a tone-bearing unit and the tone falls on the second part. It is therefore a syllable and the voiced part is an unrounded high front vowel. The highest point of the tongue is near the tip for [ɿ], while the highest point is the middle of the blade for [i]. [ʅ] functions as the rime of syllables with [tʂ], [tʂh] or [ʂ] as the onset. When the pro­ nunciation of the onset is completed, the tip of the tongue is against the front part of the hard palate with the area behind the tip as the highest point. The vocal cords start to vibrate at that moment without changing the position of the tongue. The two segments form a tone bearing unit and the tone falls on the second segment. [ʅ ] is an unrounded high front vowel. The pronunciation of [ʐ ʅ] is very similar to these syllables, although [ʐ ] itself is voiced. The syllable can be clearly divided into two parts and only the second part bears the tone. There is also a rounded high front vowel [y], which is similar to the French [y] in words like tu ‘you’. Apart from the roundedness of the lips, the highest point of the tongue for [ y] is behind that for [i]. Mandarin has only one high back vowel [u], which is rounded just like the [u] in American English words like cool. Table 33.3  Vowels of Mandarin Front

High Mid-high Middle Mid-low Low

Central

Unrounded

Rounded

iɿʅ e

y ǝ

ɛ æ

a

582

Back Unrounded

Rounded

ɤ

u o

ɑ

Mandarin

The unrounded mid-high front vowel [e] is similar to the British English [e]. It only co-occurs with the glide [j] to form a diphthong [ej]. The unrounded mid-high back vowel [ɤ] can stand alone as a syllable or function as the rime of a syllable. It does not form a diphthong with any glide. The rounded mid-high back vowel [o] may serve as the rime of a syllable that has the glide [w] as the onset. There are several syllables in Mandarin that have [o] as the rime and a labial consonant as the onset, but it is often argued that there is a glide [w] between the labial onset and [o]. The main function for [o] is to form a diphthong with the glide [w]. The mid-central schwa [ǝ] never functions alone as the rime. It is always followed by a nasal, either [n] or [ŋ]. The mid-low front vowel [ɛ] must follow a glide, either [j] or [ɥ] in a syllable. The low front vowel [æ] has a very limited distribution, only occurring between a palatal glide [j] or [ɥ] and the nasal [n] in syllables like [pjæn] and [tɕhɥæn]. The low central vowel [a] functions only as part of the diphthong [aj] or the combination [an]. The low back vowel [ɑ] in Mandarin can stand alone as a syllable or function as the rime of a syllable. It can also occur in the diphthong [ɑw] or the combination [ɑŋ]. There is a very interesting Mandarin sound, the status of which is so controversial that it has been treated as a vowel, a diphthong, a vowel–consonant combination or none of the above. Descriptively, the sound can be transcribed as either [ǝɹ] or [ɚ] to represent the fact that a sound resembling the American English [r] or an r-coloring feature is added to the mid-central vowel [ǝ]. [ɹ] is a retroflex approximant but whether it has the status of a full consonant is controversial, since many phoneticians believe that [ɹ] is just rhoticization of a vowel. [ǝɹ] or [ɚ] always stands alone and does not combine with any sound.

2.4.  Syllable structure Descriptively, the basic syllable structure of Mandarin is (C)(G)V(N/G), where C is a consonant, G is a glide, V is a vowel and N is a nasal. The parentheses indicate that the sound in question is optional. The (N/G) alternation means that a syllable could have either a nasal or a glide, but not both, as the coda. The simplest syllable is a single vowel, and the vowel-only syllables are: [ɑ], [ɤ], [i], [u] and [y]. Either [ɛ] or [o] may also stand alone as an onomatopoeia. Some speakers would add a glide as the onset of such syllables to form GV syllables like [ji], [wu] and [ɥy]. Nasals [m], [ŋ] and [n] may be used alone as onomatopoeias which could be considered syllables. The VG combination is of course a diphthong. The diphthongs [aj], [ɑw] and [ow] can stand alone as syllables, while [ej] might stand alone as an onomatopoeia. Out of the three nasals in Mandarin, only [n] and [ŋ] can appear as a coda and the co-occurrence of V and N is subject to certain phonotactic constraints. The alternation of G and N in the basic syllable structure means that diphthongs cannot occur in front of a nasal. The actual VN combinations found in Mandarin include [in], [ǝn], [an], [iŋ], [ɑŋ] and [ǝŋ]. The vowels in these syllables are nasalized. A VN sequence can stand alone as syllables or function as the rime in other syllables. Two VN combinations need special attention. One is [æn], which only appears in a GVN sequence [ jæn] or [ɥæn]. The other one is [uŋ], which is treated as a variation of [wǝŋ] in 583

Shi Dingxu Table 33.4  Actual syllables with a glide as the onset j w ɥ

jɑ wɑ

wo

jɛ ɥɛ

waj

wej

jɑw

jow

wan

jæn ɥæn

wǝn ɥǝn

jɑŋ wɑŋ

wǝŋ ɥǝŋ

some analyses. For those who make a distinction between [uŋ] and [wǝŋ], it is often stated the two are in complementary distribution, namely, [wǝŋ] only stands alone as a syllable and [uŋ] only functions as the rime of a syllable. The glides can be the onset of a syllable with a single vowel, a diphthong or a VN combination as the rime. Not all possible GV or GVN combinations are actually found. The actually existing ones are listed in Table 33.4. Most consonants in Mandarin can be the onset of a syllable, except for [ŋ]. For people with a Southern dialect background, syllables like [ŋɤ] and [ŋɑŋ] are possible free variations. When the onset is a consonant, the rime can be a single vowel, a diphthong, a VN combination or a GVN sequence. However, not all possible syllables actually exist in Mandarin. For example, if the rime of a syllable starts with a high front vowel [i] or [y], or a glide [ j] or [ɥ], the onset must be an alveopalatal sound [tɕ], [tɕh] or [ɕ]. No other consonant can ever occur in the same environment. Another case is the limited distribution of the high front rounded vowel [y] and the related glide [ɥ]. Apart from alveopalatal [tɕ], [tɕh] and [ɕ], only [l] and [n] can be followed immediately by [y] or [ɥ]. There are also some accidental gaps in Mandarin. Although [m] is a versatile onset that can take almost all kinds of V, GV, GVN and diphthong as the rime, [mja] or [mjɑŋ] does not exist. Because of the phonotactic constraints and the accidental gaps, only a little over 400 syllables are actually attested in Mandarin, out of the about 7,500 syllables mathematically possible.

2.5. Tones Mandarin is a tone language in which pitch difference is used to indicate meaning contrast of syllables. Every syllable in Mandarin bears a tone, if the neutral tone or light tone is also counted as a tone value. The neutral tone is a neutralized form of other tones and it always falls on an unstressed syllable, which is reduced in length and loudness. Very few syllables always bear the neutral tone and in most cases, the neutral tone is the surface realization of another tone, which determines the meaning of the morpheme represented by the syllable. Tones are fixed pitch patterns that can be described by the relative highness of the pitch, such as H(igh), M(id) and L(ow) or a five-point scale. Values on the five-point scale are sometimes represented graphically by lines, which are then translated into diacritics with similar shapes. There are four major tones in Mandarin. The first tone is a high level tone, which can be transcribed as HH, 55 or represented by Figure 33.1. The five points on the right of the perpendicular line indicate the highness of the pitch and the arrow on the left indicates how the pitch changes over time. The arrow for the first tone is a level line and its diacritic on the vowel has the same shape, as in [mī] ‘close eyes a little’. 584

Mandarin High Level Tone

5 High 4 Mid-High 3 Mid 2 Mid-Low 1 Low

Figure 33.1  The first tone

The second tone is MH, 35 or a mid-rising tone. It can be represented by the arrow in Figure 33.2 or a diacritic with the same shape on the vowel as [mí] ‘riddle’. 5 High 4 Mid-High Mid Rising Tone

3 Mid 2 Mid-Low 1 Low

Figure 33.2  The second tone

The third tone can be transcribed as MLH but is actually 214. It is a contour tone that can be represented by the arrow in Figure 33.3 or a concave contour diacritic on the vowel as [mǐ] ‘rice’. 5 High 4 Mid-High 3 Mid Contour Tone

2 Mid-Low 1 Low

Figure 33.3  The third tone

The fourth tone is HL or 51, and is thus a high falling tone. The arrow to represent it falls sharply as in Figure 33.4 and its diacritic has a similar shape as in [mì] ‘secrete’. 5 High High Falling Tone

4 Mid-High 3 Mid 2 Mid-Low 1 Low

Figure 33.4  The fourth tone

585

Shi Dingxu

Certain tones may change their surface value in a given environment. This is commonly known as tone sandhi. A typical case is when two third tones are juxtaposed, usually within a syntactic unit, the first one will surface as a second tone. The word [mějtɕjǒw] ‘tasty wine’ will be pronounced as something similar to [méjtɕjǒw] in actual speech. The third tone sandhi is obligatory but other tone sandhis are becoming optional and gradually disappearing. For speakers of Old Mandarin, [tɕhī] ‘seven’ is pronounced as [tɕhí] when it occurs in front of a fourth tone, but for speakers of New Mandarin, this tone sandhi is optional and is a free variation.

2.6. R-suffixation There are not many inflectional affixes in Mandarin and the diminutive suffix is one of them. When a noun or, in some rare cases, a verb takes this particular suffix, it will acquire some special meaning, such as casual, little and lovable. The noun [ɕìn] usually means a written letter which is rather formal. If a diminutive affix is added to it, the same noun will mean an oral message, which is more informal and less important. The diminutive suffix is a consonant [ɹ] added to the root or an r-coloring feature that rhoticizes the final syllable and brings in complicated changes. The pronunciation alternations vary a lot between speakers of Old and New Mandarin. A typical case is the weakening of high front vowels by New Mandarin speakers. In words like [ɕjɑˇwtɕī] ‘little chicken’, the r-suffixation will change the final syllable to [ɕjɑˇwtɕǝˉɹ] or [ɕjɑˇwtɕɚ ˉ ], with the vowel [i] being weakened to a schwa [ǝ]. Another change is the loss of syllable final nasal with possible alternation of the rime. The word [ɕjɑˇwjɑˊŋ] ‘little lamb’ becomes [ɕjɑˇwjɑ˜ɹ] after the r-suffixation, losing the final nasal [ŋ] but keeping the nasalization of [ɑ] to have [ɑ˜]; while [mùpǎn] ‘wood board’ becomes [mùpɐˇɹ], losing the nasal [n] but retaining the nasalization of [a] as [ã].

2.7.  Written Mandarin Mandarin could be written with Chinese characters but certain features would be sacrificed, since some morphemes are not represented in the existing character set. A convenient way is to use the Pinyin system, which is the alphabet designed for Putonghua, the official lingua franca for speakers of all Chinese dialects. The Pinyin system is a phonemic representation of Putonghua while the phonemic and phonological system of Putonghua is based on that of Mandarin. Pinyin is therefore an ideal means to represent Mandarin morphemes. In the Pinyin system each syllable consists of an initial and a final, represented by two symbols respectively. The initial is the (C) portion of the basic syllable structure (C)(G)V(N/G) and is either a consonant or a zero form. The zero initial has no representation and the consonantal initials are represented by 21 symbols listed in Table 33.5. The symbols b, d and g are assigned to unaspirated voiceless plosives [p], [t] and [k], while p, t and k are assigned to aspirated voiceless plosives [ph], [th] and [kh] respectively. The three retroflex sounds [tʂ], [tʂh] and [ʂ] are represented by zh, ch and sh respectively. The three dento-alveolar sounds [ts], [tsh] and [s] are represented by z, c and s. The three alveopalatal sounds [tɕ], [tɕh] and [ɕ] are represented as j, q and x respectively. The retroflex fricative [ʐ ] is represented by r while the velar fricative [x] is represented by h. 586

Mandarin Table 33.5  Initials of Pinyin b (p)

p (ph)

d (t)

t (t )

g (k)

k (kh)

h

m (m)

f (f ) s (s)

z (ts)

c (tsh)

j (tɕ) zh (tʂ)

q (tɕh) ch (tʂh)

x (ɕ) sh (ʂ) h (x)

n (n)

l (l)

r (ʐ)

Table 33.6  Finals of Pinyin Vowel and glide-vowel

a (ɑ) o (o) e (ɤ) ê (ɛ) i (ɿ), i (ʅ ) er (ɚ)

i (i) ia ( jɑ)

u (u) ua (wɑ) uo (wo)

ie ( jɛ)

Diphthong and Glide-Diphthong

ai (aj) ei (ej) ao (ɑw) ou (ow)

iao ( jɑw) iou ( jow)

Nasalized

an (an) en (ǝn) ang (ɑŋ) eng (ǝŋ) ong (uŋ)

ian ( jæn) in (in) iang ( jɑŋ) ing (iŋ) iong (ɥǝŋ)

ü (y)

üe (ɥɛ)

uai (waj) uei (wej)

uan (wan) uen (wǝn) uang (wɑŋ) ueng (wǝŋ)

üan (ɥæn) ün (ɥǝn)

The finals in the Pinyin system are the (G)V(N/G) portion of the canonical syllable structure. All the finals are listed in Table 33.6. The finals are phonemic representation of vowels and vowel combinations. The letter a stands for [ɑ], [a] and [æ]. The letter e represents [ɤ], [ɛ], [ǝ] and [e]. The letter i is assigned to [i], [ɿ] and [ʅ ]. The letter o signifies [o] and [u] while the letter u also stands for [u]. The difference is that o stands for the [u] in [uŋ] and u stands for the vowel [u]. A special letter ü is chosen to represent [y], since there is no other choice in the alphabet. Glides are represented by letters for vowels. The letter i is the symbol for [ j], either as a pre-vowel glide or as the ending part of a diphthong. The letter ü is assigned to the glide [ɥ], while u is for [w] in most cases except for [ɑw] which is written as ao, since au might be mistaken for an in hand writing. Another possible confusion comes from the nasal [n] at the end of a syllable and the glide at the beginning of the following syllable, since the [n] might be mistaken as the initial of the second syllable. It is therefore stipulated that a glide have a different shape when it occurs at the beginning of a syllable. The glide [ j] will be represented by y, [w] will be represented by w while [ɥ] will become yu, as in yan [ jæn], wan [wæn] and yuan [ɥæn] respectively. For the same reason, a y is added in front of i or ü when it appears at the beginning of a syllable, as in ying [i] and yu [y], while a w is put in front of u if it occurs at the beginning of a syllable, as in wu [u]. 587

Shi Dingxu

The nasal [ŋ] is represented by ng, as in ang [ɑŋ]. There are a few conventions that are designed to decrease the number of letters in some syllables. The final uen thus can be written as un, uei as ui, and iou as iu. The letter ü could be written as u except for cases where ü occurs after the letter l or n. The Pinyin system will be used in the following text for convenience.

3.  Words and morphology Mandarin is predominantly a non-inflectional language since it does not mark agreement, case or gender. Its personal pronouns have number markers but its nouns are mostly not marked for number. Most aspects in Mandarin have their own markers, and there are arguably some markers for tense.

3.1. Morphemes A morpheme is the smallest unit with meaning. Most native Mandarin morphemes are monosyllabic, such as nominal lí ‘pear’ and shǒu ‘hand’, and verbal hē ‘drink’ and tiào ‘jump’. Disyllabic morphemes are few and some of them came from other languages, such as Mongolian hútòng ‘lane’ and Manchu gézhi ‘tickle’. Native disyllabic morphemes include those with two identical syllables like bōbo ‘bum’ and qūqu ‘cricket’, onomatopoeia such as būjī ‘sound of squeezing water out’ and kāchī ‘sound of breaking hard object’, and those with no clear origin or meaning, such as lāngdang ‘wag (a tail)’ and hàsa ‘shake (one’s legs)’. Many morphemes are free in that they can stand alone as words. Most disyllabic morphemes are free and so are some monosyllabic morphemes like zǒu ‘walk’, ná ‘take’, cǎo ‘grass’, lù ‘road’, qiào ‘pretty’, zài ‘at, on, in’ and dǎ ‘from’. A large portion of monosyllabic morphemes are bound in that they cannot stand alone as words. Some bound morphemes are affixes, such as the experiential aspect marker –guo, plural suffix –men, and the prefix for intimate address form lǎo–, while the majority were free morphemes historically and are now building blocks for compound words. They include nominal ones like the yǒu ‘friend’ in hǎoyǒu ‘good friend’, verbal ones like the tú ‘slay, slaughter’ in túdāo ‘slay-knife, butcher’s knife’, adjectival ones like the luó ‘curved’ in luóguō ‘hunchback’, and adverbial ones like the qióng in qióngmó ‘repeatedly ask, pester’.

3.2.  Word Formation Morphological derivation is possible in Mandarin but is very limited. A typical derivational morpheme is the nominal suffix –zi, which is attached to a morpheme to form a disyllabic word without contributing any meaning. Words formed this way include zhuōzi ‘table’, yāzi ‘duck’, shàozi ‘whistle’ and quézi ‘cripple’. A more productive way to construct a compound is to put morphemes together according to syntactic rules, namely, the relationship between these morphemes can be classified as coordination, modification, predication, verb–object and verb–complementary. Coordination is between two morphemes with the same status. If two nominal morphemes are conjoined, the combination is usually a nominal compound, as in the case of péngyǒu ‘friend’ and fùmǔ ‘father-mother, parents’. The same could be said about conjoined verbal morphemes, such as xǐhuān ‘like-like, like’ and kēngmēnguǎipiàn ‘trap-deceive-kidnap-cheat, swindle’; and conjoined adjectival morphemes, as in piàoliang ‘pretty-bright, beautiful’ and guāngliu ‘smooth-slippery, slippery’; and adverbs like mànyou ‘slow-casual, casually’ and qiāoqiaoniēniē ‘quietly-softly, quietly’. 588

Mandarin

Modification is a subordinate relation between two morphemes with different functions, namely, the modified and the modifier. When the modified is a nominal morpheme, the result is usually a compound noun. The modifier could be a nominal, as in chēdēng ‘car-lamp, headlight’; an adjectival morpheme, as in hóngguǒr ‘red-fruit-R, haw’; or a verbal one, as in shuōcír ‘say-word-R, excuse’. If the modified is a verbal morpheme, the compound is usually a verb. The modifier in a verb compound can be an adverb, as in sīshēng ‘illegal-born, born out of wed-lock’; a noun as in dāopī ‘knife-cut, hack with a knife’; or a verb as in tōukàn ‘steal-look, peek, peep’. If the modified morpheme is adjectival, the compound is usually an adjective, such as zéisuān ‘extremely-sour, very sour’ and zèngliàng ‘very-shining, shining’. The two morphemes in a compound could have the function of subject and predicate respectively, as in yǎndú ‘eye-fierce, eyes are sharp, having good insight’ where yǎn ‘eye’ is the subject and dú ‘fierce’ is the predicate. Some predication compounds are constructed with three or more morphemes, as in gǒuchīshǐ ‘dog-eat-shit, (fall) face-down’ and yǎnpízigāo ‘eye-lid-high, eyelids are high, picky’. The verb–object sequence is a common form for both verb phrases and compound words. A V–O sequence is a compound if both the V and the O are monosyllabic and one of them is a bound morpheme, as in the cases of liùjiē ‘stroll-street, stroll’, zhāduīr ‘make-group-R, go together’ and kāntānr ‘look-booth-R, look after the booth, sell things’. Most V–O compounds are intransitive verbs, since they already contain an object, but some of them can take another object. A typical case is the compound huáiyí ‘hold-doubt, doubt, suspect’ which can take another object as in (1). (1) Wǒ hěn huáiyí nǐ dìdi. I very suspect your younger-brother ‘I suspect your brother’ The term ‘complementary’ refers to a particular syntactic constituent in Mandarin. The adjectival morpheme hēi ‘black’ in sentence (2) represents the result of the action tú ‘paint’, and the two form one unit that takes the noun phrase dàménr ‘door’ as the object. Morphemes with functions similar to the hēi ‘black’ in (2) are termed bǔyǔ ‘complementary’. Some such compounds have become fixed expressions, such as mǒhēi ‘put-on-black’, which has a meaning similar to túhēi ‘paint (it) black’ but also has a derived meaning of ‘slander, smear’. dàménr. (2) Tā túhēile he paint-black-Perf. door-R ‘He painted the door black’ The meaning change sometimes goes through a different path. The word zhǎngguì ‘managecounter, manage the counter’ was a V–O compound that should be a verb. However, the most prominent meaning nowadays is ‘store manager’ or even ‘husband’, namely, the person who runs the business.

3.3.  Word Classes The role a word can play is determined by its meaning and its syntactic properties are derived from the meaning. Words are therefore categorized according to their meaning: verbs, nouns, numerals, classifiers, adjectives, adverbs, prepositions, coordinators, sentence-final particles and two less significant categories onomatopoeia and interjections. 589

Shi Dingxu

3.3.1. Verbs Verbs in Mandarin can be classified as intransitive verbs, transitive verbs and ditransitive verbs. There is no morphological marker to distinguish these categories and the classification is based mainly on how many arguments each verb can take in a given clause. In terms of the semantic content and syntactic functions related to meaning, they can be classified as action verbs, state verbs, copulas and modals. There are also verbs that are semantically ‘light’ in that their contribution to the predicate depends on their complements like the jínxìng ‘carry out’ in (3). (3) Jǐngchá duì shìgù jínxìngle diàochá. police about accident carry-out-Perf. investigation ‘The police did an investigation on the accident’ Predicative verbs and adjectives are not overtly marked for person, gender, number or case, but they are marked for most aspects and arguably for certain tenses. A verb will appear in its root form when it heads an imperfective simple predicate, and will bear an aspect marker otherwise. The root form is the only choice if the verb directly modifies a noun, a verb or an adjective. There are certain sentence patterns that are comparable to passive sentences in other languages; and there are certain ways to express counterfactual ideas.

3.3.2.  Nouns and pronouns Nouns stand for entities that could be concrete, abstract or imaginary. Pronouns represent entities indirectly and are treated as a subcategory of nouns. A Mandarin noun could be a single free morpheme like guōguō ‘grasshopper’ or a compound, which consist of two or more morphemes. Case, gender, person and agreement are not overtly marked. Occasionally the plural marker –men is attached to some plural nouns that represent human beings to serve certain pragmatic purposes, such as showing affection. Sentence (4) is thus pragmatically loaded in that it is typically used by parents to refer to their own children. (4) Háizimen dōu zhǎngdàle. children all grown-up-Perf. ‘(Our) children have all grown up’ There are three sets of personal pronouns in Mandarin, one for each person. Genders and cases are not differentiated, but numbers are. There are also some special pronouns to indicate honorability and inclusiveness, as shown in Table 33.7. Table 33.7  Mandarin pronouns Singular

First person Second person Third person

Plural

Ordinary

Honorific

wǒ nǐ tā

nín tān

590

Exclusive wǒmen

Inclusive nǐmen tāmen

zánmen

Mandarin

Plural pronouns are derived forms, constructed by adding a plural suffix –men to the singular ones. Possessive pronouns are not listed here since possession is represented by adding the possessive suffix –de to a personal pronoun, even though it could be omitted under certain circumstances. There is no corresponding plural form for the singular honorific pronouns, nor is there any singular counterpart for the plural inclusive first person pronoun zánmen. In casual conversation zánmen could be shortened to zán, but it is still a first person plural inclusive pronoun. The third person honorific pronoun tān is seldom used nowadays. The second person honorific pronoun is used not only in informal communication but also in formal or official communication. When a speaker uses inclusive zánmen ‘we’, he is bringing the listener(s) into the activity in question; if the speaker uses the exclusive pronoun wǒmen ‘we’, the listener is kept out of the activity. There are two types of reflexive pronouns in Mandarin. The basic form zìjǐ ‘self ’, with its colloquial variations zìgěr, jìgěr and zìjǐgěr, could have an antecedent of any person, number or gender; and could be embedded far away from its antecedent, as in sentence (5). (5) Wǒ zhīdào dǔbó hàile zìjǐ. I know gambling hurt-Perf. self ‘I know that gambling has hurt me’ The derived reflexive is the combination of a personal pronoun and zìjǐ. A derived reflexive must agree with its antecedent in person and number; and it must occur in the same clause with its antecedent and the two must be in certain structural configuration. Interrogative pronouns in Mandarin are shénme ‘what’ and shéi ‘who’. Derived interrogative compounds include shěnme-shíhou ‘what-time, when’, shěnme-dìfang ‘what-place, where’, shěnme-rén ‘what-person, who’ and so on. There are also some ad hoc ones created to make specific inquiries, such as shěnme-zhuōzi ‘what table’ and shěnme-chá ‘what tea’. An interrogative pronoun will have an indefinite reading when it falls in the domain of a negative element or appears in a yes–no question, as in (6) and (7) respectively. shéi. (6) Wǒ bù zhǎo I not look-for who ‘I am not looking for anybody’ (7) Nín chī diǎnr shénme ma? you eat a-little-R what Part. ‘Would your honor eat something?’

3.3.3.  Determinatives, numerals and classifiers A full-fledged Mandarin nominal phrase is in the shape of a determinative–numeral– classifier–noun sequence like that in (8). The determinative establishes the relationship between the nominal phrase and the discourse or the immediate environment; namely, it determines the definite or indefinite status of the nominal phrase. Numerals stand for the result of counting while classifiers are the measurement units or the natural division units with which the counting is done. 591

Shi Dingxu

(8) Wǒ yào nèi-liǎng-zhāng zhuōzi. I want that-two-unit table ‘I want those two tables’ The proximal determinative is zhè ‘this’ and the distal demonstrative is nà ‘that’. The proximal determinative is often pronounced as zhèi when it appears before a numeral or a classifier, while the distal one is sometimes pronounced as nèi in the same environment. They may have nasalized forms when they take a classifier beginning with a nasal consonant, as in zhènme ‘this way’, zènme ‘this way’ and nènme ‘that way’. Such determinatives may sometimes take an r-suffix to form a compound, such as zhèr ‘here’ and nàr ‘there’, which may vary as zhèrhàr ‘here’ and nàrhàr ‘there’. The distributive determinative is měi ‘every’ as in měiyìduǒ huār ‘every flower’, while the existential quantifier determinative is rènhé ‘any’ as in rènhé liǎngge xiāngzi ‘any two boxes’. There is also an interrogative determinative nǎ ‘which’ that is used in a nominal phrase representing a subset to be identified from a known set, as in nǎsānge héshang ‘which three monks’. Numerals are built with digitals and counting units. Digitals are líng ‘zero’, yī ‘one’, èr ‘two’, sān ‘three’, sì ‘four’, wǔ ‘five’, liù ‘six’, qī ‘seven’, bā ‘eight’ and jiǔ ‘nine’. Counting units are shí ‘ten’, bǎi ‘hundred’, qiān ‘thousand’, wàn ‘ten thousand’ and yì ‘hundred million’. The numeral 356,847 is thus read as sānshíwǔwàn-liùqiān-bābǎi-sìshí-qī ‘35-ten-thousand+6thousand+8-hundred+4-ten+7’. As measurement units, classifiers are in essence nouns that represent containers or natural units of things, but many classifiers have been bleached of most original meaning. The most versatile classifier gè is a typical case. It can be used with nouns representing people as in yígè xiǎoniūr ‘one-unit-little-girl-R, a little girl’, with nouns standing for food like shígè jiǎozi ‘ten-unit-dumpling, ten dumplings’ and nouns representing abstract things as in yígè gùshi ‘one-unit-story, a story’. Common nouns may be borrowed as ad hoc classifiers as in liǎngpénr shuǐ ‘two-basin-R-water, two basins of water’ and yìwūzi xiǎoháir ‘one-room-child-R, a roomful of children’. Some ad hoc classifiers have lost most of their semantic content to become permanent classifiers, and some may retain certain residual meaning so that they can only be used with certain nouns. Tiáo is such a classifier that can only be used with nouns representing long and thin items, as in yìtiáo shéngzi ‘one-unit-rope, a rope’ and liǎngtiáo yú ‘two-unit-fish, two fish’. A numeral–classifier sequence occasionally functions as a predicate, as in (9). (9) Sān-kuāng nánguā yígòng sānshíbāge. three-basket pumpkin altogether thirty-eight-classifier ‘There are altogether 38 pumpkins in the three baskets’

3.3.4. Adjectives Adjectives predominantly represent properties of entities, including size, shape, color, age, state, hardness, heaviness, dimension, speed, kindness, cruelty, simplicity, complication and so on. A major function of most adjectives is to head a predicative phrase, which usually consists of an adjective and a degree adverbial, like the tǐng ‘rather’ in (10). Most adjective phrases could function as subject or object. 592

Mandarin

(10) Zhè gǒur tǐng jīling. this dog-R rather smart ‘The dog is pretty smart’ Many adjectives can modify nouns directly and the two form a compound, such as hēiliǎntáng ‘black face’, báicai ‘cabbage’ and zāngdōngxi ‘dirty thing’. Many adjective phrases can combine with the marker de to modify a nominal phrase as in zéi liàng de dēng ‘a lamp that is very bright’. Mandarin adjectives often partially change their meaning and function by undergoing morphological derivation, mainly suffixation and reduplication. When the simple adjective hēi ‘black’ undergoes suffixation, each newly created adjective will have a meaning that slightly differs from the original one, as in hēibuliūqiū ‘muddily black’, hēicāncānr ‘shiny black with some shade of dark red’ and hēigulóngdōng ‘pitch dark’. Reduplication of an adjective could apply to the whole word, such as hónghóng ‘pretty reddish’ from hóng ‘red’ and zèngliàngzèngliàng ‘very shiny’ from zèngliàng ‘shiny’. It may apply to each syllable of a disyllabic adjective, such as wěnwěndingding ‘very reliable’ from wěndìng ‘steady, reliable’, or come with an infix, such as luōleluōsuo ‘very repetitive’ from luōsuo ‘repetitive’. Compounds created via suffixation or reduplication still describe certain properties, but they could not modify nouns directly, except for being part of a de phrase, as in hēicāncānr de liǎntáng ‘a face that is shiny black’ and luōleluōsuo de lǎo tàitai ‘an old lady who is extremely repetitive’.

3.3.5. Adverbs Adverbs describe particular aspects of an action or a state, depicting the degree, scope, frequency, contextual relations of a predicate, or indicate the speaker’s stance about a proposition. An adverb could modify a verb or an adjective in a compound, as in chángbèi ‘constantly prepared’ and juémiào ‘absolutely wonderful’. An adverb phrase could modify a predicate, like the suíyì ‘casually’ in (11) and the lǎo ‘always’ in (12). (11) Lǎoshī suíyì huàle ge yuánquānr. teacher casually draw Perf. Cl. circle-R ‘The teacher drew a circle casually’ (12) Tā lǎo zhǎo wǒ yào qián. he always from I ask money ‘He always asks me for money’ The gǎnqing ‘actually’ in (13) indicates the speaker’s surprise at an event, while the shízài ‘really’ in (14) represents the speaker’s evaluation of the state of an action. Both are sentential adverbials. (13) Gǎnqing nǐ shūle. actually you lose Perf. ‘It turned out that you lost’ (14) Gāo! Shízài shì gāo! great really be great ‘It is great! It is really great!’ 593

Shi Dingxu

3.3.6. Prepositions A preposition establishes the relation between its object and a verb phrase. The oblique object of a preposition is typically a nominal phrase as in (15), but it could also be an adjective phrase, a verb phrase as in (16), or a clause. Most prepositions are monosyllabic but disyllabic ones like the wèile ‘for’ in (16) are not uncommon. The most important function of a preposition phrase is to modify a predicate as the adverbial of time, location, reason, instrument and so on. liáotiānr (15) Tāmen zài jiā they at home chat-R ‘They are chatting at home’ (16)

Bié wèile zhuàn qián bú yào mìng. don’t for make money not want life ‘Don’t make money at the cost of your life’

A preposition phrase cannot modify a nominal phrase directly. In the few cases where a preposition phrase is part of the modifier of a nominal phrase, a marker de is obligatory, as in (17). (17) duì wèntí de kànfǎ towards issue DE opinion ‘the opinion about the issue’

3.3.7.  Coordinators and connective adjuncts Coordinators mark the conjunction of two or more elements but contribute little to the overall meaning of the coordination. Those signifying conjunctive relations include hé ‘and’, gēn ‘and’, yǔ ‘and’, yǐjí ‘and’, bìng ‘and (for verbal elements)’, dàn ‘but’, dànshì ‘but’, kě ‘but’, kěshì ‘but’, ér ‘but’ and so on; while those indicating disjunctive relations are huò ‘or’ and huòzhě ‘or’. Connective adjuncts indicate the logic relation between clauses in compound sentences and certain complex sentences, and are therefore commonly used in formal genres. They have some syntactic functions in the clause, as adverbials or modals. Those marking compound sentences typically work in pairs, such as like jì  .  .  .  yòu  .  .  .  ‘while  .  .  .  while  .  .  .’, yī  .  .  .  jiù  .  .  .  ‘as soon as  .  .  .  then  .  .  .’ and yàome  .  .  .  yàome  .  .  .  ‘either  .  .  .  or  .  .  .’. Some connective adjuncts appear alone, such as tóngshí ‘at the same time’ and fǎndào ‘on the contrary’ for conjunctive relation; and háishì ‘or’, háibùrú ‘would be better’ and nìngkě ‘would rather’ for disjunctive relation. Connective adjuncts representing the logic relation between the main and subordinate clauses include pairs for cause and effect like yīnwéi  .  .  .  suǒyǐ  .  .  .  ‘because  .  .  .  therefore  .  .  .’, for induction like jìrán  .  .  .  jiù  .  .  .  ‘given that  .  .  .  therefore  .  .  .’, for purpose like wèile  .  .  .  jiù  .  .  .   ‘in order to  .  .  .  then  .  .  .’ and for concession like jiùsuàn  .  .  .  yě  .  .  .  ‘even if  .  .  .  still  .  .  .’. Connective adjuncts in conditional sentences include pairs like rúguǒ  .  .  .  jiù  .  .  .  ‘if  .  .  .   then  .  .  .’ and yàobúshì  .  .  .  jiù  .  .  .  ‘if not  .  .  .  then  .  .  .’ for factual conditions; zhēnyàoshì  .  .  .   nàme  .  .  .  ‘if indeed  .  .  .  then  .  .  .’ for counterfactual conditions; zhǐyǒu  .  .  .  cái  .  .  .  ‘only if  .  .  .  then  .  .  .’ for necessary condition; and zhǐyào  .  .  .  jiù  .  .  .  ‘as long as  .  .  .  then  .  .  .’ for sufficient condition. 594

Mandarin

3.3.8.  Sentence-final particles Sentence-final particles are predominantly used in spoken genres and typically appear at the end of a main clause, although few could appear at the end of certain subordinate clauses. Particles of the latter type, such as the le in (19), have rather vague meaning but clear propositional functions. When a le is added to (18), the new sentence (19) retains the basic proposition of (18) that the speaker does not smoke but changes it in such a way that the current proposition is that the non-smoking is a new state after the reference time. (18) Wǒ bù chōuyān. I not smoke ‘I do not smoke’ (19) Wǒ bù chōuyān le. I not smoke Part. ‘I do not smoke anymore’ Most particles only attach to the main clause to represent the speaker’s attitude or intension. The clause in (20) might stand alone as a statement, but when it is combined with the particle bā to create (21), the result is a question that stands for the speaker’s uncertainty about the fact and his intension to get confirmation for his conjecture. (20) Tā shì xīn xiàozhǎng. he be new president ‘He is our new president’ (21) Tā shì xīn xiàozhǎng bā? he be new president Part. ‘Is he our new president?’ Particles like bei represent the speaker’s attitude towards a proposition in the discourse. The scenario for (22) is that the listener worries about the consequence of their son not being able to buy a house and the speaker promises to pay for it. He uses bei to indicate that it is not a big deal. chū qián bei. (22) Zánmen we-inclusive pay money Part. ‘We’ll pay for it!’

4. Phrases Mandarin phrases are constructed with words and phrases according to a hierarchical structure. Phrases are commonly classified according to their head. Verb phrases have many shapes. An intransitive verb phrase consists of just a verb, with or without modifier or morphological derivation, such as pū ‘(milk) go over (the edge of the pot)’, liūda ‘walk (slowly)’ and lāolelāodāo ‘talk garrulously’. A transitive verb phrase has a verb and its object, as in liào tiāozi ‘put down carried things, give up’ and jǐdui háizi ‘force the child (to a bad situation)’. A ditransitive verb phrase has two objects, as in gāi tā sānkuài qián ‘owe him three dollars’ and sòng nǐ yìpī mǎ ‘give you a horse’. 595

Shi Dingxu

A full-fledged nominal phrase is in the shape of determinative–numeral–classifier–noun, as in zhèi sānge qūqur ‘these three crickets’, while a nominal phrase with any one, two or three of these four elements is possible, as long as the order is kept intact. Phrases like nà ‘that’, zhèi xiǎozi ‘this guy’, zhèzhī bǐ ‘this pen’, liǎng běnr shū ‘two copies of the book’, yì huǐr ‘a moment’ and even gè kǎnr ‘a ridge, an obstacle’ are all possible. Adjective phrases could be head-only but often consist of a degree adverbial and a head, as in tǐng dòu ‘rather funny’. Certain adjectives may have a nominal phrase as the object, and they are mostly semi-fixed usages with a non-literal reading, such as duō shìr ‘eventful, trouble-making’ and shǎo fèihuà ‘don’t talk nonsense’. Preposition phrases always appear in the form of preposition–object like cóng jiāli ‘from home’, jiù shǒur ‘within reach, with ease’ and bàng shǎngwǔ ‘near noon’, since preposition stranding is not allowed in Mandarin. The internal structure of phrases could also be described with other syntactic relations, such as coordination and modification. Modification at the phrase level is done by putting the modifier before the modified. Verb phrases, adjective phrases and nominal phrases (including noun phrases, numeral phrases and determinative phrases) could have modifiers. The modifier of a verb phrase could be an adverb phrase as in qióng duìfu ‘make-do by all means’, a preposition phrase as in wǎng zuǐli húlu ‘put liquid food into the mouth’ or a noun phrase as in chéngtiān shǔluo ‘criticize all day long’. An adjective phrase may be modified by an adverb phrase as in tè dòu ‘extremely funny’, a preposition phrase as in zài nèi shàngmiànr hélǐ ‘reasonable in that respect’ or a noun phrase as in zuótiān hěn rè ‘hot yesterday’. The modifier of a nominal phrase is a de phrase, which consists of a phrase final morpheme de and an adjective phrase, a preposition phrase, a verb phrase or a nominal phrase. These phrase modifiers are exemplified by píshi de xiǎo xiǎozi ‘a strong little boy’, zài chǎng de nèi sānwèi ‘the three who were there’, chī de dōngxi ‘things to eat, food’ and mùtou de dìbǎn ‘wooden floor’ respectively. The de could be combined with a clause as well and that is usually considered a relative clause.

5. Clauses The main components of a clause are subject and predicate. Although the subject is typically a nominal phrase, it could be another clause, a verb phrase or an adjective phrase. The predicate is typically a verb phrase, an adjective phrase or, to a limited extent, a nominal phrase. There could be peripheral components in a clause, such as clausal adverbial and connective adjuncts. Although the subject of a clause could be a clause itself, as in (1), as well as an adjective or verb phrase, as in (2), the subject is predominantly a nominal phrase, as in (3), (4), (5) and (6). (1) Zán yíkuàir zǒu bú xíng ma? we-inclusive together-R go not fine Part. ‘Is it not OK if we go together?’ tài wēixiǎn. (2) Zuò fēijī sit airplane too dangerous ‘It is too dangerous to take a plane’ 596

Mandarin

A verbal predicate describes an action or a state, as in (3) and (4) respectively. The main verb in such a clause could be an action verb like the mǎi ‘buy’ in (3), an existential verb like the yǒu ‘exist’ in (4), a copular as in (5) or a modal as in (6).   (3) Wǒ míngnián mǎi fángzi. I next-year buy house ‘I will buy a house next year’   (4) Chuáng dǐxià yǒu hàozi. bed underneath have rat ‘There are rats under the bed’   (5) Wǒ lǎopo shì Guǎngdōng rén. I wife be Canton person ‘My wife is from Canton’ (6) Zài nín gēnqian tā zhǔn néng yǒu chūxi. at your-honor front he sure can have future ‘He will certain have a good future (working) with your honor’ The predicate could also be an adjective phrase, as in (7) and (8). A degree adverbial is characteristically needed for an adjectival predicate to work, unless the clause has a special function such as expressing a strong feeling in the exclamatory statement (8).   (7) Zhè háizi tè tiǎopí. this child very naughty ‘The kid is naughty’   (8) Jiù nǐ néng! just you capable ‘Just you are capable (=Don’t show off!)’ Sometimes in spoken Mandarin is found a clause with a nominal phrase as the predicate, like that in (9) and (10). An interesting feature of these clauses is that a verb is needed to support negation, such as méi yǒu ‘not have’ for (9) and bú shì ‘not be’ for (10).   (9) Nèi zhuōzi sāntiáo tuǐr. that table three Cl. leg-R ‘The table has three legs’ (10) Wǒ Běijīng rénr. I Peking person-R ‘I am from Peking’ The predicate could also appear as a verb and complementary combination, which has several shapes. A common one is for the verb and the complementary to form one unit without any marker, such as the V–C combination shāo-rè ‘burn-hot’ in (11). Although the V–C functions as one verb, it heads two predications semantically. One is tā shāole huǒkàng 597

Shi Dingxu

‘he made a fire under the hot bed’, and the other is huǒkàng rèle ‘the hot bed became warm’, namely, an action and the result of the action. Similarly, the action bān ‘move’ in (12) gets the result that the tables are chū ‘out’. Hence the term ‘resultative construction’. (11) Tā zǎozāor de shāo-rèle huǒkàng. he early-R DE burn-warm-Perf. hot-bed ‘He warmed up the hot bed ahead of the time’ (12) Dàhuǒr bān-chū jǐzhāng zhuōzi. all the people move-exit several-Unit table ‘Everybody worked together to move out several tables’ The complementary could also be in a predication relation with the main verb. The kuài ‘fast’ in (13) describes the status of zǒu ‘walk’. (13) Zán zǒu-kuài diǎnr. we-inclusive walk-fast a little-R ‘Let’s walk a little faster’ The complementary could be a clause if the verb is marked with a de. The resultative complementary hěn lèi ‘very tired’ in (14) is semantically a predicate of wǒ ‘I’, which has the verb pǎo ‘run’ as the syntactic predicate. The sentence means that the speaker has run for some time and become tired. The dōu hóngle ‘all red’ in (15) is a predicate of yǎnjīng ‘eyes’, which is syntactically the object of the main verb kū ‘cry’. (14) Wǒ yǐjīng pǎo-de hěn lèile. I already run-DE very tired Perf. ‘I have run so much that I am tired’ (15) Tā kū-de yǎnjīng dōu hóngle she cry-DE eye all red Perf. ‘She cried so much that her eyes became red’ The bù xíng ‘not OK’ in (16) is a predicate of zāng ‘dirty’. The same can be said about zhēn hěn ‘really relentless’ in (17), which describes the extent of the action of zǎi ‘cheat’. (16) Nàr zāng-de bù xíng. there dirty-DE not OK ‘That place is so dirty that no one can tolerate it’ (17)

Zhèi xiǎozi zǎi rén zǎi-de this rascal cheat people cheat-DE ‘This guy cheats people so relentlessly’

zhēn really

hěn. relentless

The V–C resultative construction has a potential form, in which a potential infix –de– is inserted between V and C to convey the meaning that there is a potential for the result to be realized. The question in (18) is whether it is potentially possible for the listener to wash the clothes clean, depending on factors such as the listener’s personal ability, the dirtiness 598

Mandarin

of the clothes, the time allowed, the equipment available and so on. The negative counterpart of the infix –de– is the infix –bù–, as in (19). (18)

Nǐ xǐ-de-gānjìng zhèixie yīfu ma? you wash-potential-clean these clothes Part. ‘Is the potential there for you to wash these clothes clean?’

(19) Wǒ xǐ-bù-gānjìng. I wash-not-clean. ‘The potential is not there for me (to do it)’

6. Sentences A sentence is an independent unit in the discourse, while a clause depicting an action or an event may not have independent discourse function. The clause in (1) represents a coherent proposition but is likely to be considered incomplete when appearing alone. If the clause takes a final particle like the le in (1ii), it will become a sentence and be acceptable. When the clause takes a final particle ma to assume the status of a question as in (1iii), it is also an acceptable sentence. (1) (i)

Tā chōule yān. he smoke Perf. cigarette ‘He has smoked’

(ii) Tā chōule yān le. he smoke Perf. cigarette Part. ‘He has done the smoking’ (iii) Tā chōule yān ma? he smoke Perf. cigarette Part. ‘Has he smoked?’ Although (1i) is not acceptable in isolation, it will be acceptable if it is used as an answer to (1iii). Another way for a clause to acquire the status of a sentence is to function as an exclamatory statement or a directive. If a clause is a generic statement as in (2) or describes an action taking place regularly as in (3), it is common for it to assume the status of a sentence. (2) Xióngmāo xǐhuān chī zhúzi. panda like eat bamboo ‘The panda likes to eat bamboo’ (3) Lǎoyé lǎo zài yuànrlǐ shài tàiyang. grandpa always at courtyard-R-in bask sun ‘(Maternal) grandpa always sunbathes in the yard’ Sentences could be classified into statements, questions, exclamatory statements or directives according to the illocutionary force they bear. A statement usually needs a sentence-final particle if it describes a one-off action, but it may not need one if it depicts a generic state. 599

Shi Dingxu

Some questions come with a fixed number of answers and the listener is asked to choose one. They could be a polar question like (4i), an alternative question like (4ii) or a V-not-V alternative like question (4iii). (4) (i)

Míngtiān hái huì xiàyǔ ma? tomorrow still will rain Part. ‘Will it rain again tomorrow?’

(ii) Zámen shàngwu zǒu háishì xiàwu zǒu? We-inclusive morning leave or afternoon leave ‘Are we leaving in the morning or in the afternoon?’ (iii) Nǐ hē-bù-hē kāfēi? you drink-not-drink coffee ‘Do you drink coffee?’ Some other questions are open ones that have many answers, such as the variable question in (5). (5) Wǎnfàn chī shěnme ne? dinner eat what Part. ‘What should we have for dinner?’ Exclamatory statements are characteristically marked with a special degree adverb like the hǎo ‘how’ in (6i); a sentence-final particle like the a in (6ii); or both like the zhēn ‘really’ and na in (6iii). (6) (i)

Hǎo piàoliang de qúnzi! how beautiful DE skirt ‘What a beautiful skirt!’

(ii) Zhège xiǎoguǐ a! this-unit little-devil Part. ‘What a cute little kid!’ (iii) Zhēn cǎn na! really miserable Part. ‘How miserable it is!’ Directives are uttered to perform illocutionary acts, such as a command like (7i), a request like (7ii) or a permission like (7iii). (7) (i)

Gǔn chū qù! roll out go ‘Get out!’

(ii) Qǐng bǎochí ānjìng. please keep quiet ‘Keep quiet, please’ (iii) Nǐ kěyǐ chū yuàn le. you can leave hospital Part. ‘You can be discharged from the hospital now’ 600

Mandarin

7.  Aspect system There is no visible tense marker attached to verbs in Mandarin but there are aspect markers attached to verbs to indicate the status of the event or action as well as the relation between action time and reference time. The simple aspect is not marked with anything, but perfective aspects are. A common perfective aspect marker is the perfect le, which indicates that, when evaluated in its totality, the action of the predicate has completed before the reference time and the result remains unchanged at the stage after. Another perfective aspect marker is the experiential –guo, which indicates the completion of an action before the reference time and the discontinuation of the result at the stage after. In both (1) and (2), the action of going somewhere has completed, but the le in (1) indicates that the father is still in Hong Kong at the time while the –guo in (2) indicates that the uncle is no longer in Paris. (1) Wǒ bà qùle Xiānggǎng, bú zài zhèr. I father go Perf. Hong Kong not at here ‘My father has gone to Hong Kong and is not here’ (2) Jiùjiu qùguo Bālí, tā kěnéng zhīdào. uncle go Exp. Paris he might know ‘Uncle has been to Paris and he might know’ The imperfective continuous marker zhe in (3) indicates the speaker’s evaluation of a subinterval of time within a situation, and it says that the state represented by guà ‘hang’ is ongoing at the reference time. A marker with similar properties is the imperfective progressive zài in (4), which appears before the verb to show that the rain is falling at the moment. (3) Qiángshàng guàzhe yífù yóuhuà. wall-top hang Cont. one-unit oil-painting ‘There is an oil painting hanging on the wall’ (4) Wàimianr zài xià dàyǔ. outside-R Prog fall heavy-rain ‘It is raining heavily outside’ There are several sentence-final particles in Mandarin that indicate the relation between the speech time and the action time of an event, or the state of an action with regard to the reference time. The sentence-final particle ne in (5) tells the listener that the action of kàn diànshì ‘watch TV’ is in progress at the reference time, while the particle lázhe in (6) indicates that the action of dǎ pái ‘play cards’ takes place before the speech time, continues for a while and stops before the speech time. kàn diànshì ne. (5) Wǒ bà I father watch TV Part. ‘My dad is watching TV’ (6) Wǒmen dǎ pái láizhe. we play card Part. ‘We were playing cards’ 601

Shi Dingxu

8. Negation The predominant way to negate a proposition is to use a negative morpheme. The commonly used Mandarin negative morphemes are bù ‘not’ that negates predicates with an imperfective simple aspect as in (1) and méiyǒu ‘not’ or its shortened form méi ‘not’ that negates predicates with a perfective experiential aspect or a perfect aspect as in (2). (1) Zhèr bù ānquan. Here not safe ‘It is not safe here’ méi jìn. (2) Qiúr Ball-R not enter ‘The ball did not go in’ Certain negative morphemes only appear in formal genres, like the wèi ‘not’ in (3) and the wù ‘(should) not’ in the directive (4). Another negative morpheme that appears in directives only is the bié ‘do not’ in (5), which typically appears in colloquial pleas. (3) Wèi jīng xǔkě, bù kě not via permission not can ‘Don’t enter without permission’

rùnèi. enter

(4) Qǐng wù xī yān. please no inhale smoke ‘No smoking please’ (5) Bié shuāizháo le. don’t fall Part. ‘Don’t fall’

9. Comparison A common scalar comparison construction is built with the morpheme bǐ ‘than’, as in (1). The comparison is between the two terms with regard to some shared property, such as the hotness in (1). Another construction of scalar comparison is marked with chū ‘over’, which indicates the extent of difference as in (2). (1) Hǎinán bǐ Běijīng rè. Hainan than Beijing hot ‘It is hotter in Hainan than in Beijing’ (2) Gēge yào gāo-chū yí dàjiér. elder-brother really tall-over one big-portion-R ‘The elder brother is a lot taller’ Non-scalar comparison is about identity or likeness and marked with the pair xiàng  .  .  .  yíyàng ‘like  .  .  .  the same, the same as’. In such constructions, the speaker states that the primary term is equal to the secondary term in a certain aspect, ‘talking’ in the case of (3). The gēn wǒ yìbān ‘as  .  .  .  the same  .  .  .’ in (4) has a similar function. 602

Mandarin

(3) Nǐ shuōhuà zěnme xiàng wǒ mā yíyàng? you speak how like I mother the-same ‘How come you sound the same as my mother?’ gāo. (4) Nánpéngyou gēn wǒ yìbānr boy-friend as I the-same-R tall ‘My boyfriend is as tall as I am’

10.  Information-packaging constructions The canonical constituent order of a Mandarin clause is subject–predicate while the leastconstrained order in a verb phrase is verb–object. There are a number of informationpackaging constructions where the constituents appear in a non-canonical order to serve certain purposes. Constructions like (1) are the equivalent of passive sentences in other languages in that the patient phrase occurs as the subject, the agent phrase appears in a phrase marked with the preposition jiào and the verb is marked with the passive marker gěi. Sentence (1) is typically used in spoken genres and its equivalent in written genres is marked with bèi and the agent phrase may not appear, as in (2). (1) Gūgu jiào chē gěi zhuàngle. Aunt JIAO car GEI hit Perf. ‘Aunt was hit by a car’ (2) Huǒzāizhōng yìmíng chéngkè bèi shāosǐle. fire-inside one-unit passenger BEI burn-die-Perf. ‘A passenger was burnt to death in that fire’ Constructions marked with bǎ, which marks the preverbal patient nominal phrase, represent another way to pack the information. As shown in (3), the patient nominal phrase marked with bǎ occurs in a preverbal position and such a nominal phrase is always definite. (3) Wǒ yǐjīng bǎ jiù qiáng chāile. I already BA old wall dismantle-Perf. ‘I have already dismantled the old wall’ The sentence in (4), which is marked with the universal coverage adverb dōu ‘all’, serves a different pragmatic purpose. The pronoun shéi ‘who’ under the scope of dōu ‘all’ is interpreted as a universal quantifier and the sentence shows the speaker’s arrogance that his team is not afraid of anyone. The lián  .  .  .  dōu  .  .  .  ‘even include  .  .  .  all  .  .  .’ pair in (5) turns the in-between phrase into a standard, which is either the highest or the lowest case possible, and produces a reading that the problem is so easy that even the least capable person can solve it. shéi dōu bú pà! (4) Wǒmen Guóān Duì we Guoan Team anyone all not afraid ‘We, Team Guoan, are not afraid of anyone’ 603

Shi Dingxu

  (5) Zhèi tí lián xiǎoxiéshēng dōu huì zuò. such problem even include pupil all can do ‘The problem (is so easy that) even a primary school pupil can solve it’ The topic construction is another way to convey special information. It changes the usual constituent order by moving a nominal phrase to the topic position, which is sentence initial. The topic in (6) is the patient phrase which usually appears in the object position, the topic in (7) is the possessor of the subject phrase, while the topic in (8) is the object of a preposition, whose original position is now filled with a resumptive pronoun to avoid preposition stranding.   (6) Nèi wányìr zán bù xīhan. that thing-R we-inclusive not care ‘That stuff, we don’t think very high of it’   (7) Lǎo Nà yǐqián jiāli hěn yǒu qián. Lao Na previously family very have money ‘As for Lao Na, her family used to be rich’  (8)

Zhèi chúr a, rénjia dōu ná tā this chick Part. others all take he ‘This novice, people all consider him a fool’

dàng as

shǎmàor. fool-R

A topic represents an entity known to both the speaker and the listener in the current discourse, and is therefore always definite. The way a topic construction packs information is similar to the focus construction, in which the focused nominal phrase between the subject and the main predicate. The patient phrase in (9) occurs between the subject and the adverbial while that in (10) appears after the adverbial. Both are the focused element.   (9) Wǒ xiāngzi yǐjīng mǎilái le. I suitcase already buy-come Part. ‘I have bought the suitcase already’ (10)

Nǐ jīnr ge zuóyè hài méi you today-R unit homework yet not ‘You have not done your homework today’

xiě write

ne. Part.

Bibliography Gao, Aijun and Min Fu (2001) Beijinghua Ciyu (Words and Expressions of Peking Mandarin). Beijing: The Peking University Press. Qi, Rushan (2008) Beijing Tuhua (Native Expressions in Peking Mandarin), Shenyang: Liaoning Education Press. Shi, Dingxu (2004) Peking Mandarin, Muenchen: Lincom Europa. Zhou, Yimin (1998) Beijing Kouyu Yufa (Grammar of Spoken Peking Mandarin), Beijing: Yuwen Press. Zhou, Yimin (2002) Xiandai Beijinghua Yanjiu (Study of Contemporary Peking Mandarin), Beijing: The Beijing Normal University Press.

604

34 Mandarin and other Sinitic languages Hilary Chappell Ecole des Hautes Etudes en Sciences Sociales, Paris, France

Li Lan Chinese Academy of Social Sciences, Beijing, China

1. Introduction China is a vast country with a population currently approaching 1.4 billion, yet not all these people speak Mandarin, nor necessarily any other form of Chinese. At least 9% of the population belong to one of the 55 recognized ethnic minorities, each of which by the standards of any European country would represent a small nation in its own right. For example, there are an estimated 16.2 million speakers of the Tai languages, known as Zhuàng 壮 in China, who are principally located in the Guangxi Autonomous Region in South China, 1.9 million speakers of Korean, and 8.4 million speakers of the Uyghur language in Xinjiang (Zhang & Xiong 2012: 15–16). In this description, we will concentrate on describing the Sinitic languages of China spoken by the other 91% of the population. ‘Sinitic’ is the technical term used in linguistics to refer to Chinese languages and dialects which are as different from one another as the languages of Europe are, for example, as Romanian is from Portuguese. The Sinitic languages are historically related to the Tibeto-Burman languages, together forming the large SinoTibetan language family which is spread over most of East Asia in addition to many areas of Southeast Asia.

1.1.  Classification of Sinitic languages On the basis of their sound systems (phonology) and their lexicon (vocabulary), Sinitic languages have been classified into ten main branches, as in the Language Atlas of China (1987, 2nd ed., 2012). This classification essentially relies upon the findings of Chinese historical phonology with respect to a set of crucial sound changes that took place from the period of Middle Chinese (sixth–seventh centuries) onwards and is based on the following phonological tradition: the syllable is divided into an initial consonant (shēngmǔ 声母) and a rhyme ( yùnmǔ 韵母). The rhyme carries the prosodic value of tone and segmentally 605

Hilary Chappell and Li Lan

consists of either a simple vowel, a diphthong (two vowel qualities) or a triphthong (three vowel qualities), which in turn may combine with a final nasal or plosive coda (ending), depending on the dialect or the period of Chinese reconstructed: cf. Mandarin tǎ 塔 ‘pagoda’ [th+ɑ214], miào 庙 ‘temple’ [m+iɑu51], zhōng 钟 ‘bell’ [tʂ+ʊŋ55]. The syllable initial may of course be an unfilled position, that is, a ‘zero’, as in ēn 恩 ‘grace’ [∅+ən55]. In addition, the tones are classified into four categories: the level tone ( píngshēng 平声), the rising tone (shǎngshēng 上声), the falling tone (qùshēng 去声) and the entering tone (rùshēng 入声). Below are some of the main phonological criteria that have been used by various scholars to classify Chinese dialects into their principal branches, which vary from seven to ten groups noting that these classifications are not necessarily based on the principle of shared innovations: the evolution of the Middle Chinese voiced obstruent initials (quán zhuó 全浊): *b-, *d-, *g-, *dz-, etc.1 (ii) distribution of the rhymes across across the four main tone categories in the modern Sinitic languages, as conditioned by the voicing and aspiration of the consonant initials (iii) evolution of the Middle Chinese rùshēng (entering tone) (iv) the loss or retention of plosive and nasal codas /–p, –t, –k, –m, –n, –ŋ/ (v) palatalization of the Middle Chinese velars: /k–, k’–, g–/ before high front vowels

(i)

For further reading, see Ting (1982), Yuan (1989), Norman (1988), Wang (1996), L. Li (2002), Yan (2006: ch. 2), Yue (2003) and Kurpaska (2010: ch. 4) inter alia. The different dialect groups are not mutually intelligible in their spoken form. For example, a Min dialect such as Chaozhou (Teochew), spoken in Southern China is not directly understandable to speakers of one of the Jin dialects in the north, that is, without first having been exposed to it, and vice versa. Many special vocabulary items found in these dialect groups often prove not to have a corresponding cognate (related word) in Standard Chinese, and thus no written form, unless a special character has been devised, as is the case in Hong Kong Cantonese (§2.9) and in Taiwanese Southern Min. It is noteworthy that even in the early twenty-first century, the extent of this striking linguistic diversity is almost unknown to linguists and scholars outside of China. The ten main branches are listed in Table 34.1, which is based on the Language Atlas of China (Wurm et al. 1987, Zhang & Xiong, 2nd edn, 2012), hereafter referred to as ‘Atlas’. Mandarin itself is divided up into eight dialect subgroups and covers a very large expanse of China’s territory from Manchuria in the northeast to Sichuan, Guizhou and Yunnan in the southwest. For the remaining non-Mandarin Chinese languages, the greatest concentration is thus to be found in the east and southeast of China.

1.2.  Principal characteristics of Sinitic languages Chinese languages are tonal: the prosodic property of pitch level and contour can be used to distinguish lexical meaning (see descriptions below). The syllable structure for Sinitic languages is (C1)VT(C2) where C2 can be realised as /–p, –t, –k, –m, –n, –ŋ, –Ɂ/ (depending on the dialect), V is a vowel or diphthong, and T is the tonal value. The main sentence word order is Subject–Verb–Object (SVO), while SOV is a quite common alternative. Modifiers generally precede the item that is modified, for example, adjectives precede their head nouns. Sinitic languages are generally described as belonging to the language type known as ‘isolating’ or ‘analytic’, and principally make use of derivational morphology to form words, as opposed to inflectional morphology, which they lack. While 606

Mandarin and other Sinitic languages Table 34.1  Sinitic Languages of China Language branch

Region of China

Population (millions)

Representative variety

  1. Mandarin (Beifanghua) 北方话   2. Jin 晋

North, northeast, southwest of China Shanxi, Inner Mongolia, parts of Shaanxi Hunan Jiangxi Anhui Zhejiang, Southern Jiangsu Fujian, northeastern Guangdong, Taiwan Southwestern Fujian, northeastern Guangdong, scattered over South China Guangdong and Guangxi Guangxi Southern Hunan and northern Guangdong

799

Beijing

 63

Taiyuan

 36  48  3.3  74  75  42

Changsha Nanchang Jixi Shanghainese Hokkien, Teochew Meixian Hakka

 59  7.8

Cantonese Nanning, Guilin

  3.   4.   5.   6.   7.

Xiang 湘 Gan 赣 Hui 徽 Wu 吴 Min 闽

  8. K  ejia 客家 (Hakka)   9. Yue 粤 10. Pinghua 平话 and Tuhua 土话

TOTAL: 1,206,895,000 (1.2 billion)* * These figures are based on those given by Xiong Zhenghui 熊正辉 and Zhang Zhenxing 张振兴 (2012:3, 8) for the new version of the Language Atlas of China and have been rounded up. Xiong and Zhang explain that they have used the 2004 China Administrative Regions Yearbook《中国行政区划 简册》for the population figures.

most Sinitic languages possess an albeit rather small repertoire of affixes (prefixes, infixes and suffixes) to derive new words, such as nouns from verbs, the forms of nouns are not usually changed for case (nominative, accusative, genitive, ablative etc.) (but see §2.4), nor verbs for tense (past, present, future), voice (active, passive), mood (conditional, subjunctive etc.) or number (singular, dual, plural etc.). Instead, analytic methods are employed to code these meanings: prepositions and particles may be used to indicate the role of a noun in a sentence, such as gěi 给 for the dative ‘to, for’ and de 的 for the genitive in Standard Mandarin: tā de zìxíngchē 3SG-GEN-bicycle 她的自行车 ‘her bicycle’. The category of aspect, rather than tense, is prevalent as a modification of the predicate: aspectual particles may be used to indicate whether or not an action or event is completed, as in (3) below, similar to the perfective in Russian. There is an apparently high degree of correlation between the syllable and the morpheme in Chinese languages, each of which generally corresponds to a written character. Nonetheless, morphemes (the smallest linguistic units that have meaning) are not always equivalent to words, even though the written language might suggest this with its evenly spaced Chinese characters. Just as in European languages, morphemes may be bound elements found at a level below that of the grammatical word such as the root form jiǎn– and the suffix –zi in jiǎnzi 剪子 ‘scissors’. Neither can stand alone as nouns. Fusional or portmanteau morphology is more widespread than believed (e.g. fusion of plural suffixes on pronouns, see Table 34.18 below). Another interesting feature of Sinitic languages, which they share with Southeast Asian languages, is the use of classifiers in the enumeration of nouns. This means that a classifier 607

Hilary Chappell and Li Lan

is an obligatory element when a noun occurs with a numeral for counting purposes. In addition to this, classifiers group together sets of objects that share a salient characteristic and thus fulfil a role of semantic categorization. For example, in Southern Min (Hokkien), nouns denoting sea creatures and reptiles are classified by buè 尾, a classifier that developed from a noun meaning ‘tail’. The word order follows the principle of modifier–modified, as mentioned above: (1) NUM – CLF – NOUN sì buè chuìbè  四尾水马 four CLF seahorse ‘four seahorses’ The set of nouns classified by buè thus includes different types of fish, crocodiles and snakes, all of which can be regarded as having a tail or similar profile. A second major feature of Sinitic languages is their use of in situ interrogatives. This means that the interrogative pronouns for ‘who’, ‘what’, ‘where’ etc. are simply placed in the original position of the item being questioned or asked about. This word order contrasts greatly with Germanic languages such as English, where the entire sentence needs to be re-arranged, due to a requirement for interrogative pronouns to occur in sentence-initial position. (2) Nǐ wèn-guo shéi?  你问过谁?  (Standard Mandarin) 2SG ask-PERF who ‘Who did you ask?’ Adversity passives are a third important feature that are widespread in Sinitic and also found in Southeast Asian languages such as Thai, Khmer, Vietnamese and as far east as Japan: an event is construed negatively as detrimental for the person or thing who undergoes it (Clark 1974, 1989). (3) Cantonese passive with 畀 bei 2 < ‘give’ 我畀隻狗咬咗一親. jek3 gau2 ngaau5-joh2 yat 1-chan1. Ngoh5 bei2 1SG PASS CLF dog bite-PFV one-time ‘I was bitten by a dog’

2.  An overview of Sinitic languages In this section, a short description of each of the ten main Sinitic branches is presented.

2.1. Mandarin dialects 官话 The Mandarin dialects form the largest branch of Sinitic, numerically speaking. Mandarin dialects constitute at least 66.5% of the population of Chinese language speakers in China, according to Xiong and Zhang’s estimation (2008, 2012) (approx. 799 million out of 1.2 billion speakers). According to the latest revised edition of the Language Atlas of China (2012), eight subgroups have been identified for Mandarin. These are listed in Table 34.2. 608

Mandarin and other Sinitic languages Table 34.2  The eight dialect subgroups of Mandarin Mandarin subgroup

Region of China

Population (millions)

Representative variety

1. Northeastern 东北官话 2. Beijing Mandarin 北京官话 3. Jilu 冀鲁官话

Manchuria Beijing Municipal City and surrounds Western Shandong, parts of Hebei Eastern Shandong, southern Liaoning Henan, parts of Shanxi, Shaanxi and Gansu Gansu, Ningxia

 98  27

Harbin, Shenyang Beijing, Chengde

 89

Jinan, Baoding, Tianjin

 35

Yantai, Qingdao, Dalian

186

Xi’an, Luoyang, Xuzhou, Zhengzhou Lanzhou, Yinchuan

Jiangsu

 86

Sichuan, Guizhou, Yunnan, parts of Hubei and Hunan

260

4. Jiaoliao 胶辽官话 5. Zhongyuan or Central Plains 中原官话 6. Lanyin or Northwestern 兰银官话 7. Jiang-Huai or Southern 江淮官话 8. Southwestern 西南官话

 17

Nanjing, Nantong, Yangzhou Wuhan, Chengdu, Guiyang, Kunming, Changde

Total: approx. 799 million

The most prominent member in terms of its international status and as the object of high-level linguistic research is undoubtedly Standard Chinese, or Mandarin, which belongs to the Beijing dialect subgroup. Note also that Mandarin is one of the six working languages authorized by the United Nations Organization for use in both domains of official speeches and documents. As the official language of China, it is called pǔtōnghuà 普通话, which literally means ‘the common language’. Pǔtōnghuà is not however equivalent to Pekingese or Beijing Mandarin, as it is to some extent an artificially created language, subject to influence from the local dialect of its L2 speakers. Consensus for the definition of ‘Standard Chinese’, or pǔtōnghuà, was achieved at the Symposium on the Standardization of Modern Chinese held in China in 1956. It was determined that the official language would be based on the pronunciation of the Beijing dialect, the lexicon of the Northern dialects of Mandarin and the grammar of modern vernacular works (Chappell 1980; P. Chen 1999: 124). This variety of Chinese was adopted as the official language in 1958. Mandarin is claimed to be the language with the largest number of first-language speakers in the world, numbering 848 million, according to Ethnologue (Lewis et al. 2014, 17th ed.), followed at some distance by Spanish (414m), English (335m) and Hindi (260m). Of course, this claim includes but does not take into account the differences between the eight main subgroups of Mandarin dialects, which can be considerable, being a consequence of historical and geographical factors. Mandarin is known as Guóyǔ 国语 ‘the national language’ in Taiwan (Republic of China), while in the Republic of Singapore, Huáyǔ 华语 ‘the language of the Hua Chinese’ is one of four co-official languages, in addition to English, Tamil and Malay. The etymology of the term ‘Mandarin’ appears to involve a borrowing from Portuguese mandarim from the verb mandar ‘to command’ (OED) and refers undoubtedly to Chinese officials and their language, Guānhuà 官话.2 609

Hilary Chappell and Li Lan Table 34.3  Standard Mandarin tones Register

Level tone

Rising tone

Falling tone

Entering tone

Upper

55 鲜 35 咸

xiān ‘fresh’ xián ‘salty’

214 xiǎn 险 ‘danger’

51 xiàn 县 ‘county’

–

Lower

In terms of phonological criteria, the series of Middle Chinese (MC) voiced obstruent initials has been devoiced, developing into aspirated and unaspirated counterparts, while palatalization of MC velar initials preceding high front vowels is the rule, e.g. tɕiɛn51 见 ‘to see’ < MC *kenH. Another shared phonological trait of the Mandarin dialects includes the loss of the reconstructed Middle Chinese consonant finals: –p, –t, –k and –m. In syllablefinal position, only the nasals –n and –ŋ [ng] are found, although a new set of finals should be added for many of these dialects in which the retroflex suffix –r [ɚ] has been fused with the erstwhile coda. Beijing xìn /ɕin/ 信 ‘letter’ → xìr /ɕiɚ/ 信儿 ‘message’. Mandarin dialects generally possess a four-tone system, apart from the Jianghuai or Southern subgroup which has preserved the rùshēng or entering tone, yielding five tones. Table 34.3 displays the different meanings for the syllable xian /ɕiɛn/ for the four lexical tones of Standard Mandarin, using pīnyīn Romanization. The tone values use a pitch scale from the lowest, 1, to the highest, 5, a scale popularized by the renowned linguist Yuen-Ren Chao (1968: 25), while the names of the tone categories reflect their correspondance with Middle Chinese. The Mandarin dialect groups vary considerably in their basic morphology and syntax. Taking one example of the pronominal paradigms, let us begin with Standard Mandarin as a basis for comparison. Standard Mandarin has a polite form for second person singular and also makes the distinction between inclusive and exclusive forms for first person plural (see Table 34.4). The ‘inclusive’ pronoun includes the addressee [1+2] as part of its denotation, whereas the ‘exclusive’ refers to the speaker and some other third party, thereby excluding the addressee [1+3]. Nonetheless, wǒmen may have a neutral use as well. The inclusive forms are widespread in Mandarin dialects and are also found in Southern Min, Jin, Gan and Southern Wu. It has been claimed that the Mandarin inclusive category was borrowed from Altaic languages. In Sinitic languages, where polite pronouns are not available, respectful titles of address, kin terms and honorifics are used in their place. Most Mandarin dialects north of the Yangtze use causative verbs as the source of passive markers, mainly JIÀO 叫 < ‘to make’, but also RÀNG 让 < ‘to let’, in addition to the SUFFER class verb BÈI 被. The Jilu and Jiaoliao groups in Shandong use causative ZHUÓ 着 as does Southwestern Mandarin, in addition to ÁI 挨 < ‘to suffer’ (L. Li 2006; Chappell, Table 34.4  Pronouns of Standard Mandarin Singular

Plural

1. wǒ

我

2. n ǐ nín (polite form) 3. tā

你 您 他

610

wǒmen zánmen (inclusive) nǐmen

我们 咱们 你们

tāmen

他们

Mandarin and other Sinitic languages

2015). The marker of the comparative construction is generally the preposition BǏ 比 < ‘to compare’, marking the standard noun NPB, apart from the Southwest and Shandong which employ a Surpass strategy, with verbs such as GUÒ 过 ‘to surpass’ and QǏ 起 ‘to arise’ (L. Li 2003; Chappell and Peyraube, 2015). (4)

Comparative of inequality in Standard Mandarin: NOUNA – Comparative Marker – NOUNB – VERB Xiǎo Wáng bǐ Xiǎo Lǐ gāo  小王比小李高。 CM NPB tall NPA ‘Xiao Wang is taller than Xiao Li’

The object marking or ‘disposal construction’, as it is commonly known in Chinese linguistics, predominantly makes use of the marker bǎ 把, derived from a verb meaning ‘to hold, grasp’ whereas in the Jianghuai Mandarin, GĚI 给 < ‘to give’ is common (Li and Cao 2013a, 2013b). (5) Object marking ‘disposal construction’ in Standard Mandarin NOUNSUBJECT – bǎ 把 – NOUNOBJECT – VERB   她把我写的字擦掉了。 Tā bǎ wǒ xiě de zì cā-diào le 3SG BA 1SG write LIG character erase-away PFV ‘She erased the words I wrote’ In Mandarin dialect subgroups, in the case of ditransitive verbs which take both a direct (DO) and an indirect object (IO), the IO typically precedes the DO. This is called the ‘double object’ construction.3 (6) Double object construction in Standard Mandarin NOUNSubject – VERB – NOUNIO – NOUNDO  他送我一盒糖。 Tā sòng wǒ yīhé táng 3SG present 1SG one:CLF sweets ‘He gave me a box of sweets (as a present)’

2.2.  Jin dialects 晋语 Jin dialects are spoken by more than 60 million people in Northern China, in most of Shanxi province, central and western Inner Mongolia, and also in parts of Henan, western Hebei, and northern Shaanxi. The Jin dialects share the distinctive feature of a rùshēng or entering tone on syllables with a glottal stop coda. This is not found in the surrounding Mandarin dialects, nor in Standard Mandarin for that matter (Hou 1989; Wurm et al. 1987: Map B-7; Zhang & Xiong 2012: Map B1-13). Nonetheless, apart from the nasals, the other plosive codas have generally been lost, as too the voicing distinction in plosive initials. A widespread feature in the Jin dialects is the presence of syllable splitting, known as fēnyīncí 分音词 or ‘crossover words’. In the Taiyuan dialect 太原话, this morphological process of tmesis typically involves the infixation of the consonant –l– preceded by əɁ/uəɁ to form a new disyllabic word (L. Li 2002). The Taiyuan dialect has over 80 such disyllabic words. 611

Hilary Chappell and Li Lan

(7) 笨: pəŋʾ  →  薄愣 pəɁ˒̠ ləŋʾ   滚: ʿkʊŋ  →  骨拢 kuəɁ˒-ʿlʊŋ ‘stupid’ ‘to roll’ A second derivation process which is similarly both common and striking in the Jin dialects is the use of the prefix kəɁ– on verbs, count nouns and to derive adjectives. For verbs, it denotes that an action is repeated or iterative. The data in example (8) are from the Yangqu dialect 阳曲话 (Sagart 2001). (8) kəɁ-sua ‘brush back and forth’ (< sua 刷 ‘to brush’) kəɁ-pha ‘crawl back and forth’ (< pha 爬 ‘to crawl’) Jin dialects have tone inventories varying between five and seven categories; Taiyuan, the provincial capital of Shanxi, has five (Hou and Wen 1993: 37) (see Table 34.5). With regard to pronominal systems, Jin dialects can be divided into two types according to whether they use –MEN 们, realized as /.mɤ/ in the Taiyuan dialect: ɣə53.mɤ ‘1PL’ or –JIA 家 as their plural suffix, as in the Shanyin dialect 山阴话 (Shaanxi) where 1PL is ua51, a contraction of 我家 ‘1SG+family’. Shanyin possesses two polite forms (data from Qiao 2000: 115) (see Table 34.6). The double object construction follows the order of IO–DO in all the Jin dialects surveyed in Hou and Wen (1993: 300): (9)

Double object construction in the Taiyuan dialect of Jin: IO–DO 给我一本书。 kei53 ɣə53 iəɁ 2 pəŋ53 su11 give 1SG one CL book ‘Give me a book’

Table 34.5  Tone system in the Huojia dialect of Jin 获嘉方言 Tone

Level tone

Upper register

33 腔 31 墙

Lower register

tɕhiaŋ33 ‘accent’ tɕhiaŋ31 ‘wall’

Rising tone

Falling tone

Entering tone

53 tɕhiaŋ53 抢 ‘scrape, shovel’

13 tɕhiaŋ13 呛 ‘choke’

33Ɂ kuɁ 33 骨 ‘bone’

Table 34.6  Pronoun paradigm for the Shanyin dialect of Jin (Shaanxi) Singular

Plural

1. ue51

我

2. ȵi51

你

3. t ha213

他

ua51 tsa53.mɤ (inclusive) ȵiəu51 ȵiʌr213 t həɁ 5 t hʌr 213

612

我们 你们 您儿 Polite form 他们 他儿 Polite form

Mandarin and other Sinitic languages

2.3.  Xiang dialects 湘语 Xiang dialects are spoken by approximately 36 million people in Hunan province. The name is taken from the main river in this region, which served as a major immigration route between north and south in the first millennium. The Xiang dialects are spread over most of Hunan province, except for the northwest and some southern areas where Southwestern Mandarin has encroached upon Xiang territory. Pockets of Xiang dialects are also found in Sichuan and in Guangdong and Guangxi. Typically divided into New and Old types, the more archaic features of Old Xiang are now better preserved in the southern dialects. Notably, some of the Old Xiang dialects have preserved the MC voiced initials. The dialect of the provincial capital Changsha is an example of New Xiang, which has steadily converged towards Southwestern Mandarin (Wu 2005). The Changsha dialect of New Xiang has six tones (Table 34.7), a typical number for a Xiang dialect. The Xiang dialects have lost their plosives /–p, –t, –k/ in final position of the syllable, just as in Mandarin. The data are taken from Wu (2005). The pronoun system for a New Xiang dialect (Table 34.8), such as Changsha, reminds one clearly of Standard Chinese. While there is no inclusive–exclusive distinction, a polite form can be used for both younger or older addressees, li41 lau41 zən13ka33 你老人家 literally: 2SG-old-person. This has also developed a fused form: lan33 郎 (data from Wu (2005: 114)). In the Old Xiang dialects, however, including Longhui 隆回, Xinhua 新化, Chengbu 城步 and Lianyuan 涟源, we find the cognates to the main Southern Sinitic form, for example, tɕi3 佢 in the Chengbu dialect. In the Longhui dialect, tone sandhi (change of tone) is employed to code the plural pronouns: The singular pronouns have the tonal value of 31 while the plural forms all have 35 (Table 34.9). Table 34.7  Changsha Xiang tones Register

Level tone

Rising tone

Falling tone

Entering tone

Upper Lower

33 sɨ 33 ‘poem’ 诗 13 sɨ13 ‘time’ 时

42 sɨ 42 使 ‘send’

55 sɨ55 试 ‘try’ 21 sɨ21 事 ‘matter’

24 sɨ24 识 ‘know’

Table 34.8  Pronouns of Changsha Xiang Singular 1. ŋo31 2. ɳi 31~n31~li31 li 41 lau41 zən13ka33 3. t’a33

Plural 我 你 你老人家 (polite form) 他

Table 34.9  Pronominal paradigm: Longhui Xiang 隆回话 Singular

Plural

1. o31 2. n31 3. tɕi31

o35 n35 tɕi35

613

ŋo31 mən li31 mən

我们 你们

t’a33 mən

他们

Hilary Chappell and Li Lan

The disposal construction uses verbs of giving as its source, overwhelmingly the marker BǍ 把 in 26/33 Xiang dialects surveyed in Wu (2005). Ambiguously, BǍ also turns up as a marker of the passive. Comparatives use either GUÒ or BǏ. The double object construction shows the Northern order of IO preceding DO: (10) Double object construction in Changsha Xiang (Hunan): V–IO–DO 妈妈诶把我两块钱咯。 pa41 ŋo41 lian41khuai41 tɕiɛ̃13 lo ma33ma ei, mother PRT give 1SG two:CLF money PRT ‘Mother, give me two dollars please’ The classifier 只 tsa24 is the most common in the Xiang dialects, being used indifferently for human beings, animals and artifacts: 一只学生 ‘one student’ (11) i24 tsa24 ɕio24 sən33 一只牛 ‘one ox’ i24 tsa24 ȵiəu13 i24 tsa24 tiɛ̃45 sɨ45 tɕi33 一只电视机 ‘one television’

2.4.  Gan dialects 赣语 The focal territory for Gan, whose speakers number 48 million, is Jiangxi province. Gan dialects overflow into eastern Hunan, southeastern Hubei and parts of Anhui, and, like the Xiang dialects, are similarly classified into Northern and Southern types due to different pathways of development. Nanchang, the capital of the province, is regularly chosen as the representative dialect. Northern Gan was formed prior to the Tang dynasty in the period from the third to seventh centuries ad when speakers of various northern Chinese dialects flowed into Jiangxi province along the Gan River valley, merging with the local Jiangxi dialects (You 1992). Southern Gan subsequently developed on the basis of Northern Gan in the central Jiangxi region, under the influence of the Chang’an koine, brought by migrants from this northern capital during the late Tang dynasty (ad 618–907) (You 1992; Sagart 1993). The Gan dialects were originally grouped together with Hakka due to sharing certain major sound changes from the time of Middle Chinese, in particular, the evolution of voiceless aspirated plosives from the voiced obstruents. However, certain lexical differences, combined with exceptions to these sound rules, have led linguists more recently to treat them as separate groups, although the relationship is certainly close. Gan dialects usually possess six or seven tones; Nanchang Gan has seven. The data in Table 34.10 are taken from Sagart (1999) and Xiong (1995) (Nanchang Dialect Dictionary). Table 34.10  Tone system of Nanchang Gan Level tone Upper Lower

42 枪 24 墙

tɕhioŋ42 ‘gun’ tɕhioŋ24 ‘wall’

Rising tone

Falling tone

213 tɕhioŋ213 呛 ‘choke’

45 将 31 像

tɕioŋ45 ‘general (Chinese chess)’* ~ 21 tɕhioŋ21 ‘resemble’

Entering tone 45ʔ tɕhiɔʔ 45 雀 ‘sparrow’ 2ʔ tɕhiɔʔ 2 嚼 ‘talk nonsense’

* We were not able to locate a complete tonal paradigm for the syllable, tɕhioŋ24, and so have taken the liberty of including its unaspirated counterpart, tɕioŋ24, as the closest candidate.

614

Mandarin and other Sinitic languages Table 34.11  Pronouns of Nanchang Gan (Jiangxi province) Singular

Plural

1. ŋo3

我

2. n3  n3len (polite form) 3. tɕie3

尔 你人 佢

ŋo3-ko li ŋo3-mɨn ŋo3-tɨn n3-mɨn

我个哩 我们 我等 尔们

tɕie3-mɨn

佢们

Table 34.12  Singular personal pronouns in the Gan dialect of Qianshan Subject/Possessive

Direct object

1. a33 2. n33 3. kɯ33

a33 li n33 li kɯ33 li

The pronominal paradigm in Nanchang Gan (Table 34.11) shows resemblances to Mandarin for first and second persons, but the third person pronoun is classic Southern Sinitic. For the plural forms, the native suffix for first person is: –ko li 个哩 (ŋo3 koli) or ŋo3-tɨn 等, but a borrowing of the Mandarin suffix is commonly used: mɨn 们 as in ŋo3 mɨn etc. as well as for the second and third persons. Note that neither the Nanchang Dialect Dictionary nor Sagart’s data have 2pl or 3pl forms with –ko li 个哩, which explains the gap in Table 34.11. It has been claimed that certain Gan dialects show an incipient case system in the pronominal paradigms, using distinct forms for the subject and the object. Such is the case for the Qianshan dialect 铅山话, in which the singular personal pronouns take the suffix –li, when functioning as a direct object (C. Chen 1995) (Table 34.12). In many Gan dialects, as for Xiang, the main classifier is tsaɁ 7 只, which can be used for human nouns, unlike Mandarin dialects. With respect to grammar, Gan dialects belong to the heart of an area where TAKE and HOLD verbs have ‘converted’ to GIVE verbs in the context of ditransitive constructions (Zhang 2011). Consequently, in many Gan dialects, verbs of giving are identical to the disposal markers. As in a majority of Xiang dialects, this verb is often BǍ 把; for example, Nanchang pa213 and Yichun pa42. Other common markers in the Gan dialects are 搦 LAɁ7, 拿 NÁ and 摆 BǍI, all meaning ‘give’ in their basic verbal use (Liu 1999: 633). Apart from similarly using GIVE verbs such as tɛ42 得 as a source of passive markers, Gan dialects also constitute the core area for the use of ‘wait’ ten213 等 in this function. Gan comparatives typically make use of a double marking strategy which combines Northern Chinese BǏ 比 with an adverb of degree such as GÈNG 更 ‘more’ (see Liu 1999: 749). Probably due to Mandarin influence on the Nanchang dialect, the double object construction shows the Northern order of the IO preceding the DO. (12) Double object construction in Gan (Nanchang) (双宾结构) V–IO–DO 你人接了佢三只锅 tɕia1 le tɕie3 san1 tsaɁ7 wo1 n3len 2SG:POL lend PFV 3SG three CLF pot ‘You lent him three pots’ 615

Hilary Chappell and Li Lan

By way of contrast, with respect to a majority of Gan dialects, the opposite order is more common with the DO preceding the IO, as in Yichun (Li Xuping, 2015): (13)

我得本书你。 ŋo34 tɛ42-33 pun42 ɕy34 ȵi34 1SG give CLF book 2SG ‘I give a book to you’

2.5.  Hui dialects 徽语 The Hui dialects are spoken by more than 3 million people in the southeastern, mountainous areas of Anhui province, but also in parts of northeastern Jiangxi and western Zhejiang. Hui comprises six main dialect subgroups and these are in the main mutually unintelligible, a consequence of the historically isolated nature of Hui communities (Hirata 1998). Although the Hui dialects were moulded by important migrations from Northern China, some scholars have classified them as Wu, due to a identifiable basis of Old Wu features. As a result, their classification as an independent group remains somewhat controversial. Hui dialects, like Gan and Hakka to the south, mainly show voiceless aspirated initial consonants for all the MC voiced obstruents: p’–, t’– and k’–. The large number of nasalized vowels points to the loss of nasal codas, which resembles Southern Wu (Yan 2006: 229): 冷 (Mandarin lɐŋ214) /nẽ/ ‘cold’. The MC final plosive consonants – /–*p, –*t, –*k/ – have all been merged as the glottal stop /-ʔ/: 脚 (*kĭak) > /tɕioʔ/ ‘foot’. The typical Hui tone system has six categories, with the level píng and the falling qù tones undergoing a register split. We choose the Jixi dialect 绩溪 to exemplify this group (all data are taken from Hirata 1998) (Table 34.13). The pronominal system (Table 34.14) is interesting in that it contains an emphatic form in the singular series which is used for both subjects (nominative) and to code possession (the genitive). The plural suffix iã has its source in the word for ‘people’. In contrast to Mandarin, there are no polite or plural inclusive forms. Table 34.13  Tone system of the Jixi dialect of Hui (Anhui) Register

Level tone

Upper

31 西 44 婆

Lower

sɨ31 ‘west’ phə44 ‘woman’

Rising tone

Falling tone

213 kuᴐ 管 ‘pipe’

35 张 22 他

213

tɕiõ31 ‘CLF’ t ha22 ‘3SG’

Entering tone 32 ieʔ 32 入 ‘enter’

Table 34.14  Jixi pronominal paradigm (Anhui) Singular 1. ᴐ213

我

2. n213

尓

3. ki44

渠

Singular NOM/GEN

Plural

ᴐ213nɤ 我仂 n213 nɤ 尓仂 ki44 nɤ 渠仂

ᴐ213 iã

我人

n213 iã

尓人

ki44 iã

渠人

616

Mandarin and other Sinitic languages

In Jixi, /ko324/ 个 is the most common classifier, followed by /tseʔ32/ 只, which has a much wider application than in Mandarin. Classifiers can also be used as possessive markers: /ɑ55pã55-53 ɕy21/ 我本书 1SG-GEN-book ‘my book’. The disposal marker in Jixi is 帮 /põ21/ ‘help’ and the passive marker is the verb ‘to give’: /xɑ̃35/. The comparative uses the cognate marker to Mandarin, 比 pi55, and, similarly, the double object construction uses the Northern word order. These features are not surprising, given the position of Hui dialects in the central, transitional area between North and South China. (14) Double object construction in Jixi: IO–DO 渠囗我 一合桃 ki44 xã53 ᴐ213 (ieʔ 32) xɤʔ 32 t hɤ44 3SG give 1SG one CL peach ‘S/he gives me a peach’

2.6.  Wu dialects 吴语 As for Gan, the Wu group is often split into Northern and Southern types by linguists. With an estimated 74 million speakers, the focal area for Wu dialects is Zhejiang province on the central eastern coast of China, with the more conservative dialects found in the south. Northern Wu dialects are spoken in the contiguous regions of Jiangsu in the area south of the Yangtze River, as well as in southeastern Anhui province. The Wu dialects are classified into six subgroups (Zhang & Xiong 2012: Map B1-14). Shanghainese has unquestionably the most prominent profile in this dialect group but is not necessarily the most representative. Other important dialects are Suzhou, Ningbo, Wenzhou and Jinhua. In general, the Wu dialects can be distinguished by their retention of the voicing distinction and, thus, display the three Middle Chinese series of aspirated voiceless, unaspirated voiceless and voiced initials. Consequently, they have also maintained the upper and lower register distinctions and generally possess seven or eight tones. Shanghainese is an exception with just five tones. Hence, where Mandarin has only the contrast between a series of aspirated and unaspirated voiceless plosives, such as, respectively, /ph/ and /p/, the Wu dialects generally have an additional voiced series: /ph/, /p/ and /b/, the latter being murmured or whispery in quality in Shanghainese. The data and examples in Table 34.15 are from Zhu (2006: 16–20). The paradigm for personal pronouns in the Shanghai dialect (Table 34.16) contrasts greatly with that of Standard Mandarin. Apart from the first person singular pronoun which is pan-Sinitic in nature, quite distinct forms exist for the rest. There is no polite second person form in Shanghainese, nor is there an inclusive first person form, although inclusive pronouns can be found in other Wu dialects.

Table 34.15  Tone system of Shanghainese Wu Register

Level tone

Rising tone

Falling tone

Entering Tone

Upper

53 t hi53 天 ‘sky’

–

34 ti34 店 ‘shop’ 14 di14 田 ‘field’

Ɂ44 tiɁ 44 滴 ‘drop’ Ɂ24 tiɁ 44 笛 ‘flute’

Lower

617

Hilary Chappell and Li Lan Table 34.16  Pronouns of Shanghainese Singular 1. ŋu13 2. nuŋ13 3. ɦi13

Plural aʔ 5 la53 nA13 ɦi13 la53

我 侬 伊

阿拉 亻那 伊拉

As regards disposal markers, there are four main sources in Wu dialects: (i) TAKE verbs, (ii) GIVE, (iii) HELP verbs and (iv) comitatives (Xu and Tao 1999, their example below). (15) Wenzhou dialect 温州话 disposal construction with dei11 < ‘help’: NPobject(i) – [dei11+pronoun object(i)] – VP 苹果代渠吃交 beŋ31ku35 dei11 gei31 tshɨ313 ɦuᴐo apple OM 3SG eat PRT ‘Eat up the apple!’ [more literally: apple, take it and eat] To form the passive, GIVE verbs such as pəɁ 5 拔 are common in Northern Wu. Shanghainese shows a strong contrast between age groups in its speech registers, with a predictable convergence towards Standard Mandarin in the speech of the younger generation. This is clear in the word order for the double object construction which has the IO preceding the DO, as in Mandarin, for younger speakers (data from Xu and Tang 1988): (16) Double object construction in Shanghainese: 老派:

Older generation V–DO–IO 拨张纸头我 pəʔ 5 tsaŋ53 tsɨ34dɤ13 ŋu13 give CL paper 1SG ‘Give me a sheet of paper’

新派:

Younger generation V–IO–DO 拨我一张纸头 pəʔ 5 ŋu13 iɪʔ 5tsaŋ53 tsɨ34dɤ13 give 1SG one:CL paper ‘Give me a sheet of paper’

In Shanghainese, the classifier with the widest range is tsəʔ 5 只, followed by gəʔ1 个 for human and abstract nouns (Zhu 2006: 71).

2.7.  Min dialects 闽语 The Min group is a very archaic and heterogeneous group of dialects whose subdivisions are, in the main, not mutually intelligible. The diverse varieties of Min are spoken by an estimated 75 million people in China. Its speakers live principally in the southeastern coastal province of Fujian, and on the island of Taiwan. Other important Southern Min areas are located in Chaozhou and Shantou in northern Guangdong and on the Leizhou peninsula, as 618

Mandarin and other Sinitic languages

well as the island province of Hainan, the outcome of southward migrations along the coast from the late Song period onwards. Min dialects can be divided into Inland or Western Min which are less well-studied, and Coastal or Eastern Min, to which the better-known Southern Min and Fuzhou dialects belong, albeit in quite separate subdivisions (Pan et al. 1963; Norman 1988, 1991). Several strata can be detected in the Min lexicon which have led to a contrast between literary and colloquial pronunciations, known as wén-bái yì-dú 文白异读. For example, ‘name’, represented by the character 名, is pronounced mian5 in its colloquial form but beng5 [meŋ] in its literary pronunciation. Another example is sai55 (C) and su55 (L) for ‘teacher, master’ 师. The origins of the literary register are reading pronunciations for reciting classical texts that reflect the late Tang koine of Medieval Chinese (Norman 1991: 338–9). Min dialects stand out from other branches of Sinitic by virtue of the absence of the change from bilabial sounds to labiodentals, that is, the change /p > f/ found in the Northern dialects. Mandarin /fei55/ 飞 ‘to fly’ is pronounced /pue44/ in Taiwanese Southern Min. The Middle Chinese voiced obstruents have evolved into voiceless consonant initials which are in the majority unaspirated (Ting 1982: 171). There is no palatalization of velar initials such as /k-/ before high front vowels; for example, Mandarin /tɕhy51/ : S. Min /khi21/ 去 ‘to go’. The finals /–p, –t, –k, –Ɂ, –m, –n and –ŋ/ are retained above all in the Southern Min dialects, a feature they share with Yue and Hakka and notably one which sets them apart from Northern and Central dialect groups. Southern Min can also be typified by its denasalization of initials that occur in open syllables, producing the allophonic voiced sounds /b–, l– and g–/ from /m–, n– and ŋ–/ (Mandarin /mu51/ : S. Min /bak2/ 目 ‘eye’). Most Min dialects have six to eight tone categories, with seven being common for Coastal Min, including Taiwanese, which is illustrated in Table 34.17. Tone sandhi, indicated in italics in the same table, is very complex – particularly for the Southern Min dialects – and takes effect when a syllable is followed by another fully tonal syllable within the same phrase. Let us take an example of the high level tone (Yīnpíng) in Taiwanese Southern Min in a colloquial word formed with a literary register suffix: i55 + su55 → i33– su55 医师 ‘doctor’. The tone value on the first syllable changes from high 55 to mid 33 pitch when it precedes another syllable in fluent speech. The personal pronouns of Southern Min dialects typically show different sources for second and third person to those found in the Northern Chinese dialects, while the plural forms are formed with a suffix –n, a fusion with the suffix lang 5 侬 ‘people’, as Table 34.18 shows. Passives are constructed with verbs of giving, predominantly khit 8 乞, but with hou7 与 in the Taiwanese and Xiamen dialects (Chappell 2000). A special feature of Min dialects is the use of ka7~kang7 共 as the marker of the disposal construction which comes from a comitative preposition meaning ‘with, for’ (Li and Cao 2013a, 2013b; Chappell 2013). The Table 34.17  Southern Min tone system (Taiwan) Register

Upper Lower

isolated tone

tone sandhi for context tones

Level tone

Rising tone

Falling tone

Entering tone

55 33 su55 书 ‘book’ 25 21/33 su25 词 ‘an expression’

51 55 su51 使 ‘to cause’

21 51 su21 思 ‘be anxious’ 33 21 su33 事 ‘business’

25 sut 2 率 ‘to lead’ 52 sut 5 述 ‘to recite’

619

Hilary Chappell and Li Lan Table 34.18  Pronouns of Southern Min (Taiwan) Singular

Plural

1. gua2

我

2. li2 3. i1

汝 伊

gun2 lan2 (inclusive) lin2 in1

阮 咱 恁 亻因

classifier ê 个 shows a tendency to replace more specific classifiers in Taiwanese, whereas chiah8 只 is much more limited in use to animals, a small number of artefacts and vehicles in Quanzhou and Zhangzhou than it is in Xiang, Hakka and Wu. Finally, Southern Min dialects make use of the Northern word order with ditransitive verbs, as example (17) shows: (17) Double object construction in Southern Min: V–IO–DO 伊着与尔一支竹仔。 tiak4-a2 i1 toh8 hou7 li 2 chit 8 ki7 3SG then give you one CLF bamboo-stick ‘then he gave you a bamboo stick (to push down the rice stalks)’

2.8.  Kejia or Hakka 客家话 The Hakka, whose name means ‘guest people’, account for 42 million of China’s population. Northeastern Guangdong, southern Jiangxi and southwestern Fujian constitute the modern heartland of this dialect group with further pockets of speakers scattered across Jiangxi, Hunan, Guangxi and as far west as Sichuan. In Taiwan, the Hakka constitute 12% of the current population to where migration began in the early Qing dynasty. Sizeable communities also exist in Malaysia (Sabah), in Calcutta and French Guiana. Some scholars claim that Hakka developed on the basis of Southern Gan by the end of the Song dynasty (thirteenth century), and includes a substrate of non-Chinese languages, such as She and Yao (Hmong-Mien) (Sagart 1982: 148; You 1992: 103). In contradistinction to this, Norman (1988: 221–8) argues that Min and Hakka share a period of common development and that any connection with Gan is at best superficial. Dialectologists are nonetheless generally in agreement that Hakka is a full-fledged Southern Sinitic language, even if the ancestry of its speakers is northern (Norman 1988: 222). Its recent development would explain the relative uniformity of the Hakka sound system from one dialect to another (Hashimoto 1992; Sagart 1982). Hakka shares with Gan the feature of MC voiced obstruents developing into aspirated voiceless initials in all tone categories, whereas the absence of palatalization of the velars before high front vowels is shared with Min and Yue dialects. They generally possess five to six tones, distinguishing upper and lower registers for the level and the entering categories (Ting 1982: 173), as shown in Table 34.19 below. Final consonants typically include the same set as found in Cantonese Yue: /-p, -t, -k/ and /-m, -n, -ŋ/, apart from Western Fujian Hakka where these have merged to /Ɂ/ (Yan 2006: 185). The data in Table 34.19 are taken from Sagart (1982) and Chappell and Lamarre (2005) based on the same variety of Huizhou 惠州 Hakka spoken in Sung Tim Tong 崇谦堂 and in Bao’an 宝安 near Hong Kong. 620

Mandarin and other Sinitic languages Table 34.19  Tone system of Sung Him Tong Hakka (New Territories, Hong Kong) Register

Level tone

Rising tone

Falling tone

Entering tone

Upper

23 šu23 书 ‘book’ 11 šu11 薯 ‘tubers’

32 šu32 手 ‘hand’

53 šu53 树 ‘tree’

3 lip3 粒 ‘a grain’ 5 lip5 立 ‘to stand up’

Lower

Table 34.20  Hakka pronominal paradigm (Bao’an, Huizhou subgroup) Singular 1. ŋai11 2. ŋi11 3. ki11

Plural 我 你 佢

Possessive

ŋai11 teu 23 ŋi11 teu 23 ki11 teu 23

我兜 你兜 佢兜

ŋa 23 nia 23 kia 23

The Hakka pronominal system (Table 34.20) is fairly uniform across the dialect group. It possesses the interesting series of portmanteau (fused) possessive pronouns for the singular series: ‘my’, ‘your’ and ‘his/hers’ in addition to the use of the analytic forms composed of the basic pronouns and kai53, the possessive or genitive marker, for example, ŋa23 kai53 ‘my’. As in the Xiang and Gan dialects, the general classifier is /tšak5/ 只: (18)

General classifier 个只人好老 kai53 tšak 5 nyin11 hau32 lau32 that CLF person very old ‘That man is very old’

The grammatical characteristics of Hakka dialects tend to be areally determined: for the disposal construction, Hakka dialects in Fujian tend to pattern like Min dialects and use a comitative marker lau11 㧯, whereas in Guangdong, they tend to pattern like Yue dialects and avoid the use of the object-marking construction at all. In Hunan and Jiangxi Hakka, TAKE~GIVE verbs are used in this function. Hakka is typically Southern Sinitic in employing a GIVE verb as a marker of the passive construction, such as PUN~PIN 分. The comparative construction uses the surpass strategy with GUÒ but may also employ double marking with the addition of an adverb such as JIÀO 较 ‘more’. The DO precedes the IO, thus resembling the word order in Yue and many Gan dialects: (19) Double object construction in Hakka: V–DO–IO 分块面包我 pin 23 khwai4 men4pau23 ngai11 give clf bread 1sg ‘Give me a piece of bread’ 621

Hilary Chappell and Li Lan

2.9.  Yue dialects 粤语 Cantonese is the best-known member of the Yue dialects spoken by approximately 5% of China’s Han population or 60 million. The core area for Yue is Guangdong province (in the south and central areas) and the southeastern part of the Guangxi region in South China. In the Hong Kong Special Administrative Region, Cantonese is spoken by 90% of the population who originate mainly from neighbouring Guangdong province. Sizeable Cantonese-speaking communities are also found in Southeast Asia, Indonesia, North America and Australia. The Yue dialects were formed in a frontier area, first annexed as Chinese territory in the Qin dynasty (221–206 bc), in contact with languages of non-Han peoples, possibly the ancestors of the Tai people, known as Zhuàng 壮 in Chinese. In fact, a very ancient Tai substratum in the Yue is generally recognized. This is seen in the presence of a contrasting series of long and short vowels and a sizeable number of Tai loanwords (F.-K. Li 1977; Yue-Hashimoto 1991; Bauer 1996): [sɐm1] 心 ‘heart’, [saːm1] 三 ‘three’. A distinguishing feature of the Yue dialects, compared with all other Sinitic languages, is precisely the presence of vowel length, just mentioned. Like Mandarin, the MC voiced obstruents have in general been devoiced, developing into voiceless aspirated initial consonants in the píng level tone category and voiceless unaspirated in the other three (nonetheless, the divergent Western Yue dialects in Guangxi behave somewhat differently). Tone systems are surprisingly large with nine or ten tone categories possible, due to the division into upper and lower registers. Cantonese has six main tones, but nine counting the tone types found in the three closed syllables, conditioned by vowel length (Tables 34.21 and 34.22). The data are from Bauer and Benedict (1997). Other salient properties of Yue sound systems are that syllables may end in a large range of final consonants including /–p, –t, –k, –m, –n, –ŋ/, as we saw is equally possible in Min and Hakka, while velar initials have not palatalized before high front vowels: Cantonese /kei33/: Mandarin /tɕi51/ 记 ‘to record’. The Siyi 四邑 dialects possess a voiceless lateral fricative /ɬ/, an uncommon sound in most Sinitic languages, apart from Pinghua (§2.10) and the Puxian dialect of Min. Table 34.21  Tone system of Hong Kong Cantonese Register

Level tone

Rising tone

Falling tone

Entering tone

Upper

55/51

25

33

Lower

21

23

22

5 (checked 3 syllables) 2

Table 34.22  Examples of Cantonese tonal syllables Name

Example

Name

Example

1. High level 2. High rising 3. Mid level

yau55 忧 ‘to worry’ yau25 油 ‘to paint’ yau33 幼 ‘juvenile’

4. Mid-low falling 5. Mid-low rising 6. Mid-low level

yau21 游 ‘to swim’ yau23 有 ‘to have’ yau 22 右 ‘right’

7. High checked 8. Mid checked 9. Low checked

jik 5 益 ‘benefit’ jaak 3 吃 ‘to eat’ jik 2 亦 ‘also’

622

Mandarin and other Sinitic languages Table 34.23  Cantonese pronominal paradigm (Hong Kong) Singular 1. ŋᴐ:5 2. lei5 3. k’ɵi5

Plural ŋᴐ:5 tei6 lei5 tei6 k’ɵi5 tei6

我 你 佢

我哋 你哋 佢哋

Hong Kong Cantonese is one of the few Sinitic languages, apart from the standard written form of Mandarin, with its own flourishing popular literature based on the vernacular (see Bauer 1984). Specially created characters, evident in newspapers, comic books and novels are used for morphemes that have no cognates (related forms) in Mandarin, for example, lexical items such as leng3 靓 ‘pretty’ and tai2 睇 ‘to look’ and mo5 冇 ‘not have’. There are also many loanwords from English: au1lo4 欧罗 ‘euro’, so1daa2 梳打 ‘soda’ and dik 1si 6/2 的士 ‘taxi’ (apart from IPA in Table 34.23, examples use the Sidney Lau system of romanization for Cantonese in this section). The Cantonese pronominal system (Table 34.23) includes the Southern Sinitic third person singular form, k’ɵi 5 佢. The paradigm is analytic in nature, using the suffix –tei 6 哋 to form the plural series. Other dialects, including Taishan 台山话, use tone change to indicate the plural. The passive marker in Yue dialects is based on the GIVE verb bei2 畀 (see example (3) above), and in this dialect group the disposal construction is dispreferred. At best, we find the non-native marker borrowed from the literary register of Northern Chinese: jeung 1 将. Comparatives of inequality make use of a SURPASS verb, gwoh3 過 in preference to Mandarin BǏ 比, which has been similarly borrowed. (20) Hong Kong Cantonese (香港粤语) 佢高過我。 kui5 go1-gwoh3 ngoh5 3SG tall-CM 1SG ‘S/He is taller than me’ Cantonese is often cited for its word order in the double object construction, which contrasts with Mandarin and the Northern dialects in placing the DO before the IO: (21) Double object construction in Cantonese: V–DO–IO 噉然后俾返个琴佢 gam2 yin4hau6 bei2-faan1 goh3 kam4 kui5 so after:that give-return CLdef lute 3sg ‘So after that, he returned the lute to her’ (from Reborn Lady in Red) Example (21) also shows the use of the general classifier in Cantonese, which is goh3 个, as in Mandarin. Here it is used without a preceding numeral or demonstrative to code definiteness of the noun. Classifiers may also be used to mark possession: ngoh5 boon2 sue1 我本书 1SG-CLF-book ‘my book’. This is only found in the some of the Southern Sinitic languages, while it is common across Southeast Asia. 623

Hilary Chappell and Li Lan

2.10.  Pinghua dialects 平话 The Pinghua dialects are spoken by more than 2 million Chinese living mainly in Guangxi whose population is, however, predominantly Zhuang [Tai-Kadai]. They are also found in adjacent areas of Hunan and Yunnan provinces. Pinghua is classified into the Northern and Southern types, Guibei 桂北 and Guinan 桂南, which are not mutually intelligible. Furthermore, the Northern type commonly has a six-tone system while the Southern has eight to ten. Southern Pinghua is spoken in the suburbs of Nanning, alongside Nanning Cantonese, but also in towns scattered throughout Southern Guangxi, while Northern Pinghua is spoken on the outskirts of Guilin, alongside Southwestern Mandarin, as well as in many towns of Northern Guangxi. Both dialect clusters are in close contact with speakers of Western Yue and Zhuang (Tai). The Song dynasty appears to be the crucial formation period for the Pinghua dialects. Historical records relate that large contingents of soldiers, many from Shandong, settled in Guangxi, after being sent to suppress a Zhuang uprising. The status of this dialect group is controversial and less easy to justify than in the case of the Hui dialects. Southern Pinghua dialects show a close relationship with Yue, and some scholars would classify them in this group, whereas Northern Pinghua shows more Mandarin influence and certain resemblances with the neighbouring Xiang dialects and Southern Hunan patois. The Tingzicun dialect spoken in one of the districts of Nanning is used to exemplify Southern Pinghua. In the typical case for the Southern group, the MC obstruents have developed into voiceless unaspirated initials and fall into the lower pitch tones. The finals /–m, –n, –ŋ, –p, –t and –k/ have all been preserved, which resembles Yue. The unusual voiceless lateral fricative turns up in words such as 先 ɬin53 ‘first’. Tingzicun has ten tones (Table 34.24), as the entering rù tone has split four ways, similar to the case in neighbouring Yue dialects (Atlas 1987: Map B-14; Zhang & Xiong 2012; Yan 2006: 233). The pronominal system of Southern Pinghua (Table 34.25) is highly reminiscent of Cantonese, including its plural suffix. In the Shangyao 上尧 dialect of Southern Pinghua, also spoken in the outskirts of Nanning, an inclusive first person form is to be had: 伝(队) wɐn11(tɔi22) while the plural suffix is derived from a word for ‘team, group’ (De Sousa, 2015). Table 34.24  Tone system of the Tingzicun 亭子村 dialect of Southern Pinghua Register

Level tone

Rising tone

Descending tone

Entering tone

Upper

53

33

55

Lower

21

24

22

 5  3 23  2

Table 34.25  Nanning pronominal paradigm Singular 1. ŋo24 2. ni24 3. ky21

Plural ŋo24 ti23 ni24 ti23 ky21 ti23

我 你 佢

624

我哋 你哋 佢哋

Mandarin and other Sinitic languages

The disposal construction is not common in Pinghua dialects due to a preference for the regular SVO word order, as in most Yue dialects. In some cases, however, Mandarin bǎ has been borrowed. The main passive marker in the Shangyao dialect of Nanning Pinghua is /ŋai11/ 挨 < ‘suffer’. Similarly to Wu and Xiang dialects, the general classifier is /tʃət3/ 只: ɐt3 tʃət3 təu11 one-CL-head 一只头 ‘a head’. In the Northern Pinghua dialects (Zheng and Lin 2005: 261–3, 273), the double object construction may use both word orders of IO–DO and DO–IO, as in (22) and (23). (22) Yining dialect of Northern Pinghua: Verb–DO–IO 分个梨子你 fɐn34 ko33 lɐi31-tsə53 nə53 give CL pear 2SG ‘(I’ll) give you a pear’ (23) Verb–IO–DO 分我本书! fɐn34 ŋuə53 puɐn53 ʂəu34! give 1SG CL book ‘Give me a book!’ In contrast, for the Southern Pinghua dialect of Shangyao, only the IO–DO word order is possible.

3. Conclusion We have considered the Sinitic languages from the angle of interdialectal comparison among the ten main groups of Chinese dialects, particularly with respect to features that contrast with Standard Mandarin. Much more research needs to be encouraged in order to appreciate their striking diversity.

Acknowledgements This research has been supported by funding from several sources, including the European Research Council under the European Community’s Seventh Framework Programme (FP7/2007–13)/ERC Advanced Grant Agreement n° 230388 for the ‘SINOTYPE’ project (2009–13) and the French National Research Agency’s (ANR) ‘blue sky’ programme blanc: ‘Typologie des processus synchronique et diachronique en Min-Sud (langue sinitique)’ TYSOMIN n° 11-ISH2-001-01, 2012–14, projet bilatéral France (ANR) et Taiwan (National Research Council) (2013–15).

Notes 1 Obstruents include plosives (stops), affricates and fricatives. Unless otherwise indicated, the Romanization systems in use in this chapter are pīnyīn for Mandarin, the Sidney Lau system for Cantonese, the Church Romanization for Taiwanese Southern Min, and the International Phonetic Alphabet (IPA) elsewhere. 2 Although the Portuguese were clearly not the first Europeans to set foot in China, they were the first Europeans having arrived by a sea-route with the express purpose of setting up trading posts in Canton and then in Macau from the early sixteenth century.

625

Hilary Chappell and Li Lan 3 In fact, this is a type of dative construction where there is no preposition or case marker used to introduce the indirect object or recipient.

References Bauer, Robert S. (1984) ‘The Hong Kong Cantonese speech community’, Cahiers de Linguistique Asie Orientale 13(1): 57–90. Bauer, Robert S. (1996) ‘Identifying the Tai Substratum in Cantonese’, Pan-Asiatic Linguistics. Proceedings of the Fourth International Symposium on Languages and Linguistics, V. Bangkok: Institute of Language and Culture for Rural Development, Mahidol University, 1,806–44. Bauer, Robert S. and Paul Benedict (1997) Modern Cantonese Phonology, Berlin: Mouton de Gruyter. Chao Yuen-Ren (1968) A Grammar of Spoken Chinese, Berkeley: University of California Press. Chappell, Hilary (1980) ‘The Romanization Debate’, Australian Journal of Chinese Affairs 4: 105–18. Chappell, Hilary (2000) ‘Dialect Grammar in Two Early Modern Southern Min Texts: A Comparative Study of Dative Kît 乞, Comitative Câng 共 and Diminutive –Guìa 仔’, Journal of Chinese Linguistics 28(2): 247−302. Chappell, Hilary (2013) ‘Pan-Sinitic Object Markers: Morphology and Syntax’, in Cao Guangshun, H. Chappell, R. Djamouri, and Thekla Wiebusch (eds) Breaking Down the Barriers: Interdisciplinary Studies in Chinese Linguistics and Beyond, 2 vols, Taipei: Academia Sinica, 785–816. Chappell, Hilary (2015) ‘Linguistic Areas in China for Differential Object Marking, Passive and Comparative Constructions’, in Hilary Chappell (ed) Diversity in Sinitic Languages, Oxford: Oxford University Press, 13–52. Chappell, Hilary and Christine Lamarre (2005) A Grammar and Lexicon of Hakka: Historical Materials from the Basel Mission Library, Collection des Cahiers de Linguistique – Asie Orientale 8, Paris: École des Hautes Études en Sciences Sociales. Chappell, Hilary and Alain Peyraube (2015) ‘The Comparative Construction in Sinitic: Synchronic and Diachronic Variation’, in Hilary Chappell (ed) Diversity in Sinitic Languages, Oxford: Oxford University Press, 134–154. Chen, Changyi 陈昌议 (1995) ‘Jiangxi Qianshan Fangyan Rencheng Daici Danshu de “Ge’’〈江西铅 山方言人称代词单数的格〉(‘Case in the Singular Forms of the Personal Pronouns of the Qianshan Dialect in Jiangxi’), Zhongguo Yuwen《中国语文》(Studies of the Chinese Language) 10(1): 45–8. Chen, Ping (1999) Modern Chinese: History and Sociolinguistics, Cambridge: Cambridge University Press. Clark, Marybeth (1974) ‘Submissive Verbs as Adversatives in some Asian Languages’, in Nguyen Dang Liem (ed) South-east Asian Linguistic Studies, vol. 1. Canberra: Pacific Linguistics, Australian National University, 89–110. Clark, Marybeth (1989) ‘Hmong and areal Southeast Asia’, in David Bradley (ed) South-East Asian Syntax (Pacific Linguistics. Series A, no. 77. Papers in South-East Asian Linguistics, no. 11.), Canberra: The Australian National University, Research School of Pacific Studies, 177–230. De Sousa, Hilario (2015) ‘Language Contact in Nanning – From the Point of View of Nanning Pinghua and Nanning Cantonese’, in Hilary Chappell (ed) Diversity in Sinitic Languages, Oxford: Oxford Unversity Press, 157–189. Hashimoto, Mantaro (1992) ‘Hakka in Wellentheorie Perspective’, Journal of Chinese Linguistics 20(1): 1–49. Hirata, Shoji 平田昌司 (1998) Huizhou Fangyan Yanjiu《徽州方言研究》(Research on the Huizhou Dialects), Tokyo: Haowen Publishers. Hou, Jingyi 侯精一 (1989) Jìn Yǔ Yánjiū《晋语研究》(Studies on the Jin dialects), Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa. Hou, Jingyi and Wen Duanzheng 温端正 (1993) Shanxi fangyan diaochao yanjiu baogao《陕西方言 调查研究报告》(A Report on the Survey and Research into Shanxi Dialects), Taiyuan: Shanxi Gaoxiao Lianhe Publishers. Kurpaska, Maria (2010) Chinese Language(s): A Look through the Prism of The Great Dictionary of Modern Chinese Dialects, Berlin and New York: De Gruyter Mouton. Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig (eds) (2014) Ethnologue: Languages of the World, 17th ed., Dallas, TX: SIL International, available at: http://www.ethnologue.com. Li, Fang-Kuei (1977) A Handbook of Comparative Tai, Honolulu: The University of Hawaii Press.

626

Mandarin and other Sinitic languages Li, Lan 李蓝 (2002) ‘Fangyan Bijiao, Quyu Fangyanshi yu Fangyan Qufen – yi Jinyu Fenyinci he Fuzhou Qiejiaoci wei Li’〈方言比较，区域方言史与方言区分 – 以晋语分音词和福州切脚词 为例〉(‘Dialectal Comparison, History of Dialectal Areas, and Dialectal Classification: A Case Study of Fenyinci in Jin Dialects and Qiejiaoci in Min Dialects’), Fangyan《方言》(Dialects) 1: 24–59. Li, Lan (2003) ‘Xiandai Hanyu fangyan chabiju de yuxu leixing’〈现代汉语方言差比句的语序 类型〉(‘Word Order Typology of Comparative Constructions in Modern Chinese Dialects’), Fangyan 《方言》(Dialects)   3: 1–21. Li, Lan (2006) ‘ “Zhuo” Zi Shi De Beidong Ju de Gongshi Fenbu yu Leixing Chayi’〈«着»字式的 被动句的共时分布与类型差异〉(‘Synchronic Distribution and Typological Differences for Passive Constructions Formed with Zhuo’), Zhongguo Fangyan Xuebao《中国方言学报》(Journal of Chinese Dialects), 194–205. Li, Lan and Cao Xilei 曹茜蕾 (= Hilary Chappell) (2013a) ‘Hanyu Fangyan-zhong de Chuzhishi he Bǎ Ziju (Shang)’〈汉语方言中的处置式和把字句(上)〉(‘The Disposal and Bǎ Constructions in Chinese Dialects (Part I)’), Fangyan《方言》(Dialects) 1: 11–30. Li Lan and Cao Xilei (2013b) ‘Hanyu Fangyan-zhong de Chuzhishi he Bǎ Ziju (Xia)’〈汉语方言中 的处置式和把字句(下)〉(‘The Disposal and Bǎ Constructions in Chinese Dialects (Part II)’), Fangyan《方言》(Dialects) 2: 97–110. Li, Xuping (2015) ‘Ditransitive Constructions in Gan Chinese: A Case Study of the Yichun Dialect’, in B. Nolan, G. Rawoens and E. Diedrichsen (eds) Causation, Permission and Transfer: Argument Realisation in GET, TAKE, PUT, GIVE and LET Verbs, Amsterdam: John Benjamins Publishing Company, 177–94. Liu, Lunxin 刘纶鑫 (1999) Ke-Gan Fangyan Bijiao Yanjiu《客赣方言比较研究》(Comparative Research on Kejia and Gan Dialects), Beijing: Zhongguo Shehui Kexue. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Norman, Jerry (1991) ‘The Min Dialects in Historical Perspective’, in William S.-Y. Wang (ed) Languages and Dialects of China, Journal of Chinese Linguistics Monograph Series No. 3, Berkeley: Project on Linguistic Analysis, University of California, 325–60. Pan, Maoding 潘茂鼎, Li Rulong 李如龙, Liang Yuzhang 梁玉章, Zhang Shengyu 张盛裕, and Chen Zhangtai 陈章太 (1963) ‘Fujian Hanyu Fangyan Fenqu Lüeshuo’〈福建汉语方言分区略说〉 (‘An Outline of the Classification of Chinese Dialects in Fujian Province’), Zhongguo Yuwen 《中国语文》(Studies   of the Chinese Language) 6: 475–95. Qiao, Quansheng 乔全生 (2000) Jin Fangyan Yufa Yanjiu《晋方言语法研究》(Research on the Grammar of the Jin Dialects), Beijing: The Commercial Press. Sagart, Laurent (1982) Phonologie du Dialecte Hakka de Sung Him Tong, Paris: Langages Croisés. Sagart, Laurent (1993) Les Dialectes Gan, Paris: Langages Croisés. Sagart, Laurent (1999) Notes on the Nanchang Dialect (Sketch Grammar and Transcriptions), Manuscript. Department of Linguistics, La Trobe University, Melbourne. Sagart, Laurent (2001) ‘Vestiges of Archaic Chinese Derivational Affixes in Modern Chinese Dialects’, in Hilary Chappell (ed) Sinitic Grammar: Synchronic and Diachronic Perspectives, Oxford: Oxford University Press, 123–42. Ting, Pang-hsin 丁邦新 (1982) ‘Hanyu Fangyan Qufen de Tiaojian’〈汉语方言区分的条件〉(‘Features for the Classification of Chinese Dialects’), Tsing Hua Journal of Chinese Studies 14: 257–73. Wang, William S.-Y. (1996) ‘Linguistic Diversity and Language Relationships’, in Audrey Li and James Huang (eds) New Horizons in Chinese Linguistics, Boston: Kluwer Academic, 235–67. Wu, Yunji (2005) A Synchronic and Diachronic Study of the Grammar of the Chinese Xiang Dialects (Trends in Linguistics 162), Berlin: Mouton de Gruyter. Wurm, Stephen, Li Rong, Theo Baumann, Mei W. Lee (1987) Language Atlas of China《中国语言 地图集》, Hong Kong: Longmans. Xiong, Zhenghui 熊正辉 (1995) Nanchang Fangyan Cidian《南昌方言词典》(Nanchang Dialect Dictionary), Nanjing: Jiangsu Jiaoyu. Xiong, Zhenghui 熊正辉 and Zhang Zhenxing 张振兴 (2008) ‘Hanyu Fangyan de Fenqu’〈汉语方言 的分区〉(‘The Distribution of Chinese Dialects’), Fangyan《方言》(Dialects) 2: 97–108. Xu, Baohua 许宝华 and Tang Zhenzhu 唐珍珠 (1988) Shanghai Shiqu Fangyanzhi《上海市区 方言志》(A Record of the Dialect of Shanghai City), Hong Kong: Joint Publishers. Xu, Baohua and Tao, Huan 陶寰 (1999) ‘Wuyu de chu zhishi’〈吴语的处置式〉(‘Disposal Constructions in the Wu Dialects’), in Yunji Wu (ed) Hanyu Fangyan Gongshi yu Lishi Yufa Yantao Lunwenji

627

Hilary Chappell and Li Lan 《汉语方言共时与历时语法研讨论文集》(Anthology   on the Synchronic and Diachronic Grammar of Chinese Dialects), Guangzhou: Jinan Daxue Chubanshe, 135–167. Yan, Margaret Mian 严棉 (2006) Introduction to Chinese Dialectology, Munich: Lincom Europa. You, Rujie 游如杰 (1992) Hanyu Fangyanxue Daolun《汉语方言学导论》(Chinese Dialectology), Shanghai: Shanghai Jiaoyu. Yuan, Jiahua 袁家骅 (1960 [2nd ed., 1989]) Hanyu Fangyan Gaiyao《汉语方言概要》(An Outline of Chinese Dialects), Beijing: Wenzi Gaige. Yue, Anne O. (2003) ‘Chinese Dialects: Grammar’, in Graham Thurgood and Randy LaPolla (eds) The Sino-Tibetan Languages, London and New York: Routledge, 84–125. Yue-Hashimoto, Anne O. (1991) ‘The Yue Dialects’, in William S-Y. Wang (ed) Languages and Dialects of China (Journal of Chinese Linguistics Monograph Series 3), Berkeley: Project on Linguistic Analysis, 294–324. Zhang, Min 张敏 (2011) ‘Hanyu fangyan shuangjiwu dongci jiegou nan-bei chayi de chengyin: Leixingxue yanjiu yinfa de xin wenti’〈汉语方言双及物动词结构南北差异的成因: 类型学研究 引发的新问题〉(‘The Reasons for the North–South Difference for Ditransitives in Chinese Dialects: New Issues Arising in Typological Research’), Bulletin of Chinese Linguistics 4(2): 87–270. Zhang Zhenxing 张振兴 and Xiong Zhenghui 熊正辉 (eds) (2012) Zhongguo Yuyan Dituji Di’erban “中国语言地图集 第二版” (Language Atlas of China, 2nd edition), Beijing: Commercial Press. Zheng, Zuoguang 郑作广 and Lin Yi 林亦 (eds) (2005) Guibei Pinghua yu Tuiguang Putonghua Yanjiu《桂北平话与推广普通话研究》(Research on Northern Pinghua and the Popularization of Standard Mandarin), Nanning: Guangxi Minzu Chubanshe. Zhu, Xiaonong (2006) A Grammar of Shanghai Wu, Munich: Lincom Europa.

628

35 Metaphor in Chinese: Cognition, Culture, and Society Zhuo Jing-Schmidt University of Oregon, USA

1.  Metaphor in language, thought, and culture For over two millennia, metaphor has been an object of study in rhetoric. Generally perceived as the poetics of language, requiring a level of linguistic creativity above and beyond ordinary language, it is considered the frosting on the cake of solid literal language, delectable but not substantial, pleasant but not essential. For scholars who view figurative language as secondary to literal language, metaphor is ‘an ornament that at best breaks up the monotony of ‘plain talk’ and at worst reduces all communication to little more than fancy babble’ (Sapir 1977: 8–9). This view is losing hold in contemporary scholarship of metaphor – especially among linguists and cognitive scientists. In Metaphors We Live By (1980), Lakoff and Johnson proposed the Conceptual Metaphor Theory (henceforth CMT), an experientialist approach to metaphor, arguing that metaphor is essential to our understanding of the world, and our experience in the world. Within the experientialist framework metaphor reflects how the mind conceptualizes experiences. A central tenet of the experientialist view of metaphor is embodiment. That is, metaphor expresses abstract thoughts by referring to basic and ordinary bodily experiences. In fact, Lakoff and Johnson argue that without metaphors of bodily experience some abstract thoughts are impossible. For example, cross-linguistically, emotions are conceptualized experientially as physical sensations such as anger is heat and affection is warmth, or as spatial orientations such as happiness is up and sadness is down (Lakoff and Johnson 1980; Lakoff and Kövecses 1987; Kövecses 1991). Thus, when mapped onto familiar and palpable bodily experiences the psychological complexity and abstractness of emotion becomes conceptually accessible. Seeing metaphor as a mapping across separate semantic domains is not new, but has deep roots both in classical Western philosophy and in ancient Chinese philosophy. Aristotle, in his discussion of rhetoric, saw metaphor as a transfer between levels of classification. In this tradition, scholars recognized that metaphor points to ‘an equivalence between terms from separate semantic domains’ (Sapir 1977: 4). 629

Zhuo Jing-Schmidt

In ancient Chinese philosophy, metaphor also holds a special place in discussions about rhetoric, where it is seen as a central tool of persuasion (Wang 2003). Mencius (fourth century bce) commented about rhetoric, ‘He who speaks of what is far in terms of what is near, is good at speaking’ (言近而指远者,善言也). This comment itself is of metaphorical nature in that it is based on a spatial metaphor, whereby spatial nearness and remoteness take on more abstract conceptual content. Hui-Shi (fourth century bce), a thinker, statesman, and lobbyist of the Warring States, was known as being good at persuasion by means of metaphor. It is reported, in the Eastern Han text on persuasion (刘向《说苑•善说篇》), that he could not speak at all without using metaphor. In response to the King’s request that he speak literally, Hui-Shi said, ‘He who persuades, analogizes what he doesn’t know with what he knows, thereby making it known’ (夫说者,固以其知喻其所不知,而使人知之). Here the two poles of the epistemic continuum make up the two domains of metaphorical transfer, suggesting that metaphor is not merely a linguistic phenomenon, but also an epistemological, thus cognitive, instrument. The distinctive features of contemporary scholarship of metaphor in cognitive semantics comprise three aspects: (i) the insistent emphasis on metaphor as a systematic cognitive principle, (ii) the explicit articulation and theorization thereof, and (iii) the recognition that metaphor is universal in its embodiment but culture-specific in the utilization of conceptual sources. Affirming metaphor as a mirror of the embodied mind, Lakoff and Johnson’s initial publication made a significant contribution to the contemporary scholarship of metaphor as a constitutive principle of our abstract thinking. The cognitive significance of metaphor and its systematic nature found further celebration in subsequent book-size treatments of metaphor. These bear titles which audaciously, if somewhat provocatively, promulgate the cognitive role of metaphor and its embodied nature. Lakoff and Johnson (1999) are assertive about Philosophy in the Flesh, challenging the Cartesian disembodiment of mind and reason. Gibbs (1994) speaks of The Poetics of Mind, emphasizing the conceptual significance of metaphor as an ordinary yet essential tool of thinking. Feldman (2006) traces the bodily origin of metaphorical thinking from a neuropsychological perspective in From Molecule to Metaphor. He argues that the co-activation of experiences from distinct modalities, such as affect and warmth, strengthens ‘the connections between the neural circuits supporting the different modalities.’ This strengthened connectivity constitutes metaphorical mapping. From the perspective of cognitive psychology, Tucker (2007) argues in Mind from Body that subjectivity originates from the brain’s neural structure, lending further credence to the thesis of metaphor as systematic embodiment. The recognition of cultural variation in metaphorical thinking counterbalances the argument for its cognitive universality. While noting the centrality of spatial experience as our most fundamental bodily experience, Lakoff and Johnson (1980: 57) also contend, ‘[E] very experience takes place within a vast background of cultural presuppositions.’ In other words, culture shapes the way we experience the world. Gibbs (1999: 154–5) suggests that our embodied experience is shaped by social and cultural constructions of experience, according to which some aspects of our bodily experience are viewed as more salient than others. Kittay (1987: 15) points out that metaphor as a conceptual tool nevertheless relies on language as an ‘expressive medium.’ In fact, the development of higher intelligence and language itself is driven and sustained by social interaction in culturally defined contexts, both during phylogeny (Deacon 1997; Dunbar et al. 1999) and during ontogeny (Tomasello 1995, 1999, 2003). Thus, differences in cultural cognition necessarily give rise to differences in metaphorical conceptualization in different languages (Barcelona and Soriano 2004; Boers et al. 2004; Charteris-Black 2003; Kimmel 2004; Kövecses 2005). 630

Metaphor in Chinese: Cognition, culture, and society

While the cognitive role and cultural background of metaphor are now widely appreciated, the Lakoffian CMT has not gone unchallenged. In the New York Times Edward Rothstein (1999) commented that by reducing rich metaphors to clear-cut mappings CMT ‘resembles a version of simplistic Freudianism, in which varied experiences are treated as metaphorical transformations of primal relationships.’ In the place of a simple cross-domain mapping, alternative theories propose various mechanisms to explicate the way in which metaphor works. Expanding on their earlier work on blending of mental spaces (Fauconnier 1994; Turner 1996; Fauconnier and Turner 1998), Fauconnier and Turner (2002) emphasize the conceptual complexity of metaphor beyond a single pairwise mapping. They show that conceptual metaphors operate in integration networks, where multiple spaces and mappings occur in a mix, and undergo compression to construct the final conceptual product with emergent structure. Another development on CMT is the explicit differentiation of primary metaphors directly based on universal bodily experiences, from complex metaphors deriving from combinations of primary metaphors and culture-specific experiences (Grady 1997; Lakoff and Johnson 1999; Kövecses 2005). Such a distinction fine-tunes our understanding of metaphorical conceptualization shaped by the two most fundamental experiential sources: body and culture. These recent advances in theories of metaphor bring metaphor research to a new level by revealing the conceptual richness and complexity of metaphor hitherto unexplored. Theories of metaphor have implications for theoretical models of metaphor interpretation. In general, the view of metaphor as secondary to literal language favors the indirect access model (Grice 1975; Searle 1979), wherein the metaphorical meaning is arrived at via the rejection of the literal meaning. The view that metaphor is a natural cognitive mechanism favors the direct access model (Gibbs 1994, 2001; Glucksberg 2003), which generally holds that metaphor in context is processed as effortlessly as literal meaning. This debate has led to a wealth of behavioral and, more recently, neurophysiological studies of metaphor processing. By and large, these studies hinge on two questions: (i) Does metaphor comprehension require the activation of cross-domain mapping? (ii) What is the role of conventionality in metaphor processing? Gentner and colleagues (Gentner and Wolff 1997; Gentner and Bowdle 2001) maintain that the processing of metaphors, even highly conventionalized metaphors, begins with an initial structural alignment that maps the source onto the target. Gentner and Bowdle (2001) and Bowdle and Gentner (2005) further proposed that conventional and novel metaphors are processed differently, although both require an initial stage of alignment. Understood as categorizations conventional metaphors involve lower processing costs than novel metaphors, which are understood as comparisons. Giora’s (1997, 2003) Gradient Salience model argues that the ease of processing is not so much determined by the type of processing as by the salience of the expressions, defined as a function of conventionality, prototypicality, and frequency of use. Both models have found support in neurophysiological studies (e.g. Lai et al. 2009; Ahrens et al. 2007).

2.  The social functions of metaphor The experientialist approach is significant not just in its assertion of the cognitive role and cultural background of metaphor. It distinguishes itself from traditional approaches also in its view of the social functions of metaphor. Specifically, it refutes objectivism, and asserts that metaphor constructs social reality and sanctions action. On this view, metaphor serves real-world purposes, especially in social and political spaces. Western philosophy has a long track record of being concerned with truth-condition. In philosophical discussions on 631

Zhuo Jing-Schmidt

metaphor, the question has been whether metaphor can state truth. Some maintain that the revelation of truth requires literal language. Others insist that ‘metaphors express truths’ and that ‘these truths are no different in kind than those expressed by literal language’ (Stern 2006: 245–6). However, metaphors do not express objective truth, as Sapir points out (1977: 3): [T]ropes belong to that part of language that resists strict logical analysis defined by paraphrase, analytical tautology, truth-value tests, and the like. Their resistance to such analysis has been an irritant to some logicians and a challenge to others. Lakoff and Johnson (1980: 159–222) challenge the very idea of objective truth, arguing that, ‘Truth is always relative to a conceptual system that is defined in large part by metaphor.’ Truth in the experientialist approach depends not so much on the inherent properties of things as on the interactional properties arising from our daily experience of perception and categor­ ization. In this line of inquiry, it is less important to explore the truth-value of metaphor than to explore the purpose of metaphor. As Lakoff and Johnson (1980: 157) argue, metaphor can ‘define reality’ by way of ‘a coherent network of entailments that highlight some features of reality and hide others.’ Based on discourse context and speaker intention, the entailments of a metaphor create reality and sanction action. This functional view of metaphor as a discursive tool is in line with earlier research on metaphor and its roles in politics as a cognitive heuristic, a framing device, as well as a conveyer of emotion and political stance (Bronowski 1972; Edelman 1971, 1977; Miller 1979). It has also foreshadowed the more recent scholarship on the ubiquity of metaphor in discourse construction. Specifically, the instrumental nature of metaphor in political discourse (Mio 1997; Charteris-Black 2003, 2004, 2005; Chou 2014; Peng and Jing-Schmidt 2014, inter alia), medical narratives (Angus and Mio 2011; Heath 2011), economic discourse (Nicaise 2010), literary studies (Bennett 2004; Elford 2014), as well as in the media (Porto and Romano 2013) has been brought into focus. As we will see in the analysis of two monumental metaphors in recent Chinese history, mechanical and familial concepts are mapped onto sociopolitical domains to construct social reality for the articulation of political order, and for accomplishing collective ideological and sociopolitical goals. In the sense that metaphor serves to evoke social responses, to mobilize the public, and to sanction policy and sociopolitical action, its accomplishments go beyond persuasion in the sense of Crocker (1977: 53), who refers to metaphor as ‘a socially effective base for persuasive rhetoric.’

3.  Chinese metaphor Chinese metaphor has been the focus of cognitively oriented scholarship in the English language for over a decade. Numerous studies investigate the figurative sources of common target domains. Basic emotions have attracted much attention, such as anger (Yu 1995, 2008), sadness (Ye 2001), and happiness (Yu 1995, 1998, 2008; Polley 2012). Scholars found both cross-linguistically shared conceptual metaphors such as anger is heat, sadness is down, happiness is warmth/heat, and also metaphors with Chinese-specific conceptual sources such as anger is qi. The time is space metaphor is another focus for cross-linguistic metaphor research. Boroditsky (2000) and Boroditsky et al. (2011) found a verticality bias in Mandarin Chinese speakers’ temporal representation, which contrasted with English speakers’ horizontality bias. Huang and Hsieh (2007) tested the dual perspective model of time is space and 632

Metaphor in Chinese: Cognition, culture, and society

found that although the time-moving and the ego-moving perspectives are both accessible to native speakers of Chinese in Taiwan, the time is a flow metaphor is easier to comprehend than the time is a path metaphor due to first-language influence. This finding, according to Polley (2012), suggests a Chinese bias toward the stationary-ego metaphor, the same bias underlying Chinese metaphors of happiness. Polley (2012) identified the stationary-ego event structure metaphor in Chinese metaphors for happiness. Yu (2003c) describes two metaphors thinking is moving and thinking is seeing, and regards both as constitutive of the metaphor system the mind is body in Chinese. Slingerland (2004) explores conceptualization of the self and the spiritual ideal in Chinese Taoist classics. Jing-Schmidt (2008) discusses both universal and culture-specific sources of metaphors and metonymies of verbal behavior in Chinese, and points out the social function of metaphor in conveying stance. Going from the opposite direction, researchers approach Chinese metaphor by looking at concepts used as source domains and examining the target domains onto which they are projected. Yu is particularly concerned with terms of body parts as metaphorical and metonymical sources in Chinese cultural context. He analyzes the eye as the source for sight and mind (Yu 2004), the gallbladder as the source of courage (Yu 2003a), the heart as a central source of cognition (Yu 2007, 2009), and the hand as the basis of metaphorical and metonymical mappings onto abstract social psychological domains (Yu 2000, 2003b). Yu (2001, 2008) discusses the metaphorical and metonymical conceptualization of dignity and prestige as embodied in experiences with the face, arguing that culture is a filter that selects mappings from the bodily source domain. Yu’s studies of body part terms and their metaphorical mappings into abstract conceptual domains invariably point to the intimate interaction of shared basic bodily experiences and cultural assumptions in shaping conceptualization. The examination of mapping principles between source and target domains is associated with research by Ahrens, who is particularly concerned with operationalizing the sourcetarget domain mappings of a conceptual metaphor by explicating the underlying conceptual link between source and target. Her Conceptual Mapping Model (Ahrens 2002, 2010) deals with linguistic expressions in a source domain in search of patterns of mapping onto realworld target domains in order to make processing predictions. Based on this model, Ahrens and her colleagues (Ahrens et al. 2003; Chung, Ahrens, and Huang 2003; Chung, Huang, and Ahrens 2003) propose frequency-based mapping principles using corpus data within an ontology-based knowledge representation model. Metaphor as knowledge representation across languages is the focus of Xiong and Huang (2015), who explore the bitterness metaphor and the pain metaphor with regard to their respective target domains in Chinese and English. They found that effort is bitterness and intensity is pain are dominant in Chinese, while effort is pain and intensity is bitterness are common in English. While metaphor as knowledge representation with mappings based on bodily and cultural experiences have been intensively studied in cognitive linguistics, the social functions of metaphor in Chinese discourse have received scant attention with the exception of a few recent inquiries (Chou 2014; Link 2013; Jing-Schmidt 2014a, 2014b; Peng and Jing-Schmidt 2014). For example, Jing-Schmidt (2014a) and Peng and Jing-Schmidt (2014) examine antigraft metaphors in Chinese political discourse in the People’s Daily, a daily newspaper widely seen as the voice of the Chinese Communist Party. They found that the choice of antigraft metaphors is systematically different across the two versions of the newspaper. corruption is disease predominates in the Chinese version of the newspaper, whereas corruption is enemy prevails in the English version. From a sociocognitive perspective, they argue that the cultural model associated with the presumed audience determines the choice 633

Zhuo Jing-Schmidt

of metaphor as a cognitive tool for that audience to make sense of political discourse. On the other hand, from the perspective of frame semantics, they argue that the different entailments of the two metaphors have drastically different sociopolitical ramifications, which are intended to influence the respective audiences. The ready-for-combat resolution entailed by corruption is enemy is the intended message for the English-speaking audience, whereas the effort of regaining health by curing disease as a system-internal problem, entailed by corruption is disease, is intended for the Chinese-speaking domestic reader. In what follows I will focus on the functional aspects of two Chinese metaphors with profound sociohistorical repercussions. Section 3.1 explores the socialist metaphor a serviceman in a collective is a rust-free cog in a machine and its sociopolitical functions in historical context. Section 3.2 discusses the career of the metaphor national identity is sibship and how the shift of the target domain reflects China’s shifting sociopolitical priorities in different stages of recent Chinese history. Following the usage-based approach to metaphor research (Deignan 2005, 2006; Hilpert 2006), my analysis in 3.2 is based on corpus data.

3.1. The cog metaphor in socialist China ‘I wish to be a cog forever,’ (我愿永远做一颗螺丝钉) wrote Lei Feng, a young soldier of the People’s Liberation Army, in a diary published shortly after his death in 1963. In the ‘Learn from Comrade Lei Feng’ propaganda campaign that ensued, Lei Feng became the personification of altruism, and devotion to Mao and the people. The metaphor ‘be a forever rust-free cog’ (做一颗永不生锈的螺丝钉) became the crystallization of Lei Feng Spirit. As Lei Feng became a socialist cultural icon, the slogan, widely circulated in the media and in schoolbooks, became a household term (Fraser 1980). A novel metaphor at the time of its creation, this expression gained conventionality among older generations of Chinese. Although it has become the object of cynicism, derision, and satire among contemporary Chinese youth, the metaphor is one of the most enduring imprints in the collective memory of the Chinese. Just as a cog is part of a machine, an individual in China is always a part of the intricate social system. A serviceman in a collective is a rust-free cog in a machine is a complex metaphor that can be deconstructed into two socialist cultural beliefs and two basic metaphors:1 The

value of individuals derives from their membership in

and service to a collective

People

should commit themselves to the service of the

collective

A collective is a machine Misbehavior is rust Understanding a complex system in terms of a machine is widespread wherever human societies have exposure to machines. For example, a person can be described as ‘having a screw loose’ in English or in German (‘hat eine Schraube locker’). These idioms draw on the metaphor the mind is a machine and the common knowledge that a machine is functionally impaired when having a loose screw. It is also common to conceive of a corrosive influence in terms of rust. Both metaphors rely on our experience with the material world. However, the complexity of the Chinese metaphor consists in the blending of multiple conceptual spaces. To understand the reality it constructs, we need to consider: (i) the salient physical 634

Metaphor in Chinese: Cognition, culture, and society

and functional facts about a cog, (ii) the basic metaphors a collective is a machine and misbehavior is rust, and (iii) the culture-specific values attached to the target domain: the self in relation to the collective. In the final conceptual product all these elements are blended, contributing to the metaphorical entailments that define the reality constructed by the metaphor. As Yu (2008: 253) argues, the very selection of physical facts from the source domain is filtered by subjective cultural experiences. Specifically, cogs are physically small, inconspicuous, identical, and devoid of individuality. These facts are of significance in the metaphorical projection onto the social domain of self in relation to the collective. The abstract concepts that emerge from these physical features pertain to the ordinariness, obscurity, and exchangeability of people. Furthermore, although each and every cog is indispensable for the proper operation of a machine, it is useless on its own, or out of its place. These facts provide an apt analogy to the cultural belief that individuals derive their value from their membership in and service to a collective. Still more physical facts are salient through the lens of social and cultural cognition: a cog has to be properly inserted and fastened in order to reliably support the operation of a mechanical system. When rusty, a cog is potentially harmful for the mechanical system and must be replaced to prevent malfunction of the system. Projected onto the social domain, these elements give rise to abstract concepts defining the culturespecific self-society relationship, including the passivity, manipulability, and replaceability of individuals as dictated by the collective system. By deconstructing the complex metaphor a serviceman in a collective is a rust-free cog in a machine into its basic components, we see that the socialist cultural beliefs about self and society motivate the selection of relevant facts from the source concept of a cog. The social projections of these elements, blended with the basic metaphors a collective is a machine and misbehavior is rust, create a new social category – the serviceman – that defines the ideal self in a socialist collective, and positions the self in its ideal slot in relation to socialist collective goals. Given the historical context in which the metaphor originated, the sociopolitical purposes intended with the metaphor are apparent. At a time when class struggle and political upheaval dominate social life, when planned economy chokes competition, creativity, and choice, specifically in the aftermath of the devastating Great Leap Forward, social cohesion is in jeopardy. It is necessary to inspire a sense of purpose and to mobilize mass commitment among ordinary members of society. The creation of the social category of the ideal serviceman by the ‘forever rust-free cog’ metaphor, which embodies Lei Feng spirit, was intended as mass persuasion to inspire and rekindle a sense of meaning and purpose, duty and devotion among the people. As Fraser (1980: 104) observes, when the government creates an idealized ordinary citizen as a model hero, ‘it is not very surprising to discover that ordinary citizens are expected to measure up to the models.’ In other words, social cohesion is expected to come from emulation. Language use follows the zeitgeist of society. So does metaphor. Unsurprisingly, four decades after its initial creation, the cog metaphor has met its demise as a result of dramatic socioeconomic transformations since China’s mobilization for reforms and openness in 1978, which gave rise to a shift of discursive focus on development and the glory of economic success.

3.2.  From sibship to national identity The Chinese word tongbao (同胞) is a compound noun consisting of the morphemes tong ‘same’ and bao ‘placenta,’ and refers to a sibling. This conceptualization of sibling involves 635

Zhuo Jing-Schmidt

the working of metonymy whereby placenta stands for womb, which in turn stands for mother. For most of Chinese history, tongbao is used in its basic sense of sibling, as illustrated by the following examples from the Eastern Han and the Yuan dynasties, respectively:2 (1) 同胞之徒,无所容居,其故何也？ 《答客难》 ‘One’s own sibling has nowhere to stay, what is the reason for that?’ (2) 望官人回嗔作喜,好念着同胞兄弟。 《杀狗记》 ‘May the master turn grudge into pleasure, and consider your own brother’ In a Song dynasty philosophical text expressing universal love, the basic reference to sibling was extended to non-kin and humans in general: (3) 民吾同胞,物吾舆也。 《张载集 西铭篇》 ‘People are my tongbao, natural things are my equals’ In this example, the metaphor of sibship entails the expectation of love and benevolence in one’s interaction with people and nature, and calls for actions of universal love. The propagation of this extended metaphorical use did not happen until after the rise of the Republic, as indicated by increased uses in the colloquial literature: (4) 我等将替四万万同胞除害《民国演义》 ‘We will root out evil for the 400 million tongbao’ (5) 全国同胞,亦各奋公诚,同匡危难《民国演义》 ‘Tongbao of the whole nation, rise in good faith, carry the calamity together’ (6) 凡属同胞百姓,决不惊动《民国野史》 ‘As for all the ordinary tongbao, we do not disturb them’ (7) 我辈湘人,不愿与同胞为难《清史演义》 ‘We the Xiang people, are not willing to cause trouble for our

tongbao’

In these examples, tongbao refers to people of the same nation, thus instantiating the metaphor national identity is sibship. This metaphor involves the basic metaphor a nation is a family with a set of entailments, and the sociocultural belief in filial loyalty: A nation is a family Members of a family share a common heritage Members of a family fulfill expectations related Children of a family are subject to filial duties

to their roles

In the final conceptual product, common heritage, shared emotions, individual roles and duties – especially filial duties – are all projected onto the conceptualization of national identity as the target. These projections create and reinforce a reality of national identity based on one’s familiar experience with the primary social unit of family, and of rules governing the interaction among its members. The construction of such reality through the sibship metaphor has profound sociopolitical repercussions in modern Chinese history. The 636

Metaphor in Chinese: Cognition, culture, and society

use of the sibship metaphor is especially noticeable in public statements addressing the whole nation. In response to the 1937 Marco Polo Bridge Incident marking Japan’s comprehensive invasion of China, the Chinese Communist Party delivered its statement calling for nationwide anti-Japanese resistance and coalition. The statement begins with the appeal:  (8) 全國同胞們！平津危急！華北危急！中華民族危急！ 《中國共產黨號召抗戰宣言》   ‘Tongbao-s of the whole nation! Beiping and Tianjin are endangered! North China is endangered! The Chinese race is endangered!’3 Under threat of a violent foreign invasion, the message of national unity and cohesion is urgent and important. The sibship metaphor is central in calling for national unity and the formation of coalition and a united front by evoking shared emotions and the sense of duties underlying sibship. Similarly, after Japan’s unconditional surrender in August 1945, Chiang Kai-shek made his victory speech using the sibship metaphor, which begins:  (9) 全國軍民同胞們、全世界愛好和平的人士們: 我們的抗戰,今天是勝利了。 蔣介石《抗戰勝利告全國軍民及全世界人士書》 ‘Military and civilian tongbao-s of the whole nation, peace-loving persons of the whole world: our Anti-Japanese Resistance War was won today’ In this speech Chiang used the sibship metaphor nine times, acknowledging the suffering and sacrifice of the people especially those in the occupied regions, strengthening the message of national unity and bond, as in the following excerpt: (10)  我全國同胞自抗戰以來,八年間所受的痛苦與犧牲雖是一年一年的增加,可是抗 戰必勝的信念,亦是一天一天的增強; 尤其是我們淪陷區的同胞們,受盡了無窮 摧殘與奴辱的黑暗,今天是得到了完全解放,而重見青天白日了。這幾天以來, 各地軍民的歡呼與快慰的情緒,其主要意義就是為了被佔領區同胞獲得了解放。 蔣介石《抗戰勝利告全國軍民及全世界人士書》 ‘In the eight years since the beginning of the Resistance War, the pain and sacrifice suffered by our tongbao were greater each year, but our faith in winning the war also became stronger each day; our tongbao in the occupied regions in particular, endured the darkness of endless devastation and humiliation. Completely liberated today, they see the blue sky and the bright sun again. In the last few days, the joyful and cheerful feeling of our military and civilian citizens throughout the country is mainly about the liberation of our tongbao in the occupied regions.’ By making frequent references to ‘sibling’ citizens, Chiang emphasizes the shared wartime experiences of pain and sacrifice, and expresses sympathy, joy, and appreciation. Four years later, Mao Zedong, in his own presidential address declaring the founding of the People’s Republic of China, also addressed the people as siblings: (11) 同胞們！中華人民共和國中央人民政府已於本日成立了！ 《中華人民共和國中央人民政府公告》   ‘Tongbao-s! Today the Central People’s Government of the People’s Republic of China is founded!’ 637

Zhuo Jing-Schmidt

In the next paragraphs we will observe a shift of the target domain under changing historical contexts and shifting political priorities. The shift serves to deliver desired messages to the intended audience, to justify policies, and to motivate actions. Mao’s use of the term in (11) may have been one of the last uses of the sibship metaphor in reference to all Chinese in mainland China’s public discourse. In socialist China the emergence of tongzhi ‘comrade,’ literally ‘same-ideal,’ seemed to have brought the all-encompassing use of tongbao to an end. The switch suggests a shift from national identity to ideological identity as the mainland’s political priority. But the word tongbao is far from disappearing from the mainland Chinese lexicon. To the contrary, it is active in public discourse, except that the target is now different. A collocation analysis based on the XIN sub-corpus of the Chinese Gigaword Corpus revealed that the uses of tongbao are especially salient along geopolitical and ethnic lines. Specifically, the collocates of this word in the lexical modifier slot x of the phrase [x tongbao] exhibit two main tendencies in the construction of desired group identity.

3.2.1.  From national identity to ethnic and cultural identity The phrase Taiwan tongbao outranks all other collocations in XIN, suggesting the strategic motivation and geopolitical significance of sending the message of territorial, ethnic, and cultural unity across the Taiwan Strait. Example (12) below provides an illustration: (12)  系列节目通过台湾同胞在大陆的亲身感受和视点,反映两岸同种、同文、同根、 同源的骨肉亲情,反映两岸统一的必然趋势及改革开放以来祖国大陆的巨大 变化。[Internet-ZH 4783] ‘The TV series shows Taiwan tongbao’s personal experiences with and perspectives of the mainland, and reflects the blood tie between people on both sides of the strait, their common race, common script, common roots, and common origin, and points to the inevitable trend of unification, and to the enormous transformations in the fatherland’s mainland since Reform and Opening.’ The use of tongbao in conjunction with Taiwan sanctions the policy of unification with Taiwan. Similarly, liang-an tongbao ‘tongbao on both sides of the Strait,’ ranking second, portrays the peoples across the Strait as inseparable blood kin despite geographical barriers and political conflicts. The same can be said of phrases such as xianggang tongbao ‘Hong Kong tongbao,’ aomen tongbao ‘Macao tongbao,’ haiwai tongbao ‘overseas tongbao,’ malai tongbao ‘Malay tongbao,’ and gang-ao-tai tongbao ‘Hong Kong, Macao, and Taiwan tongbao,’ whereby the ethnic and cultural tie of greater China and the Chinese diaspora to mainland China is emphasized in the construction of a larger Chinese ethnic identity. While such an identity may be real in the general Chinese unconscious, its explicit verbalization has consequences for mainland China’s Taiwan policy, and policies regarding other historically and politically alienated groups. The underlying message is straightforward: just as siblings are inalienable, the people of Taiwan and mainland China are inseparable.

3.2.2.  Refocusing national identity Among the closest collocates of tongbao in the XIN sub-corpus, names of national minorities form a salient subtype. For examples, zangzu tongbao ‘Tibetan tongbao,’ miaozu tongbao ‘Miao tongbao,’ yaozu tongbao ‘Yao tongbao,’ and yizu tongbao ‘Yi tongbao.’ Here, in the 638

Metaphor in Chinese: Cognition, culture, and society

absence of ethnic and cultural identity, the strengthening of a national identity and territorial integrity is critical. The fact that zangzu tongbao ‘Tibetan tongbao’ ranks higher than reference to any other minorities is not arbitrary. Both historical and contemporary conflicts with Tibet pose challenges to the Chinese government, both in terms of domestic policy and in terms of foreign affairs. More generally the use of the sibship metaphor in reference to groups that are politically significant and socially and ethnically marginal, constructs and reinforces a reality of national unity and territorial integrity based on the entailments of closeness and mutual duties associated with sibship. Subjecting the interethnic relation to the Chinese cultural model of sibship transforms the ethnic relation into familial unity and national identity, and foreordains policies safeguarding unity and identity. The example in (13) illustrates this point. (13)  中央領導同志曾多次到西藏實地考察,幫助藏族同胞制訂發展戰略,解決實際 問題。[XIN-0158] ‘Leaders of the Central Government have visited Tibet multiple times, helping our Tibetan tongbao draft development strategies and solve real problems’ Clearly, refocusing the target domain of the sibship metaphor on national identity has implications for China’s ethnic policy in that it emphasizes ethnic integration based on national unity, and sanctions economic development in the minority regions.

3.2.3.  From identity to sympathy and solidarity In addition to the construction of ethnic and national identity, the use of the sibship metaphor expresses sympathy and solidarity with victims of disasters and other traumatized or dis­ advantaged groups within the framework of national identity. As our collocation data show, in the XIN corpus, there is a group of high attraction collocations that denote the victimhood of the referent of the head noun tongbao, such as yunan tongbao ‘calamity-hit tongbao,’ da-tusha tongbao ‘massacred tongbao,’ zaiqu tongbao ‘disaster-zone tongbao,’ sinan tongbao ‘tongbao who perished in a disaster,’ shouzai tongbao ‘disaster-suffering tongbao,’ and xunnan tongbao ‘tongbao who died a martyr.’ The sibship metaphor in the semantic field of calamity expresses sympathy and solidarity. In the Taiwan-based CAN sub-corpus, modifiers from the semantic field of physical disability were also found in high frequency, including canzhang tongbao ‘disabled tongbao,’ shizhang tongbao ‘vision-impaired tongbao,’ and zhangai tongbao ‘impaired tongbao.’ The use of the sibship metaphor in these contexts expresses care and emotional support. While the above analysis shows changes in the target domain of the sibship metaphor in official discourse, Jing-Schmidt (2014b) points to a recent development in the use of this metaphor in grassroots discourse in the Chinese Internet. Specifically, new collocations representing emerging social categories and group identities have appeared, as in (14): (14) 70后同胞 ‘nineteen-seventies tongbao’ 80后单身同胞 ‘nineteen-eighties single 90后同胞 ‘nineteen-nineties tongbao’ 剩女同胞 ‘left-over women tongbao’ 屌丝同胞 ‘have-nots/losers tongbao’ 宅男同胞 ‘otaku tongbao’ 房奴同胞 ‘mortgage-slave tongbao’

tongbao’

639

Zhuo Jing-Schmidt



愤青同胞 ‘angry youth tongbao’ 蚁族同胞 ‘ants-clan tongbao’

These new uses represent the identities that netizens form with respect to the changing social structure that gives rise to emerging social categories. The sibship metaphor is reclaimed by members of grassroots communities to express their sense of belonging and mutual sympathy with regard to their common plight in China’s socioeconomic transformation. The emphasis on ethnic identity in relation to the non-mainland Chinese populations, the affirmation of national identity in relation to the ethnic minorities, and the expression of sympathy and solidarity, as well as the novel uses among grassroots netizens to consolidate group identity and sense of belonging, can be seen as originating from one general metaphor: group identity is sibship. It is interesting to note in passing that this metaphor is used in the domain of faith in the West, giving rise to religious identity is sibship. In the Chinese context, as Jing-Schmidt (2014b) argues, the varying collocation that specifies the group being referred to enables the creation of sociopolitically important categories, prioritizes messages of group identity, alignment, cohesiveness, and solidarity as the binding foundation of political order and political actions.

4. Conclusion This chapter opened with a brief sketch of the past and present scholarship of metaphor in general and of Chinese metaphor in particular. To fill a gap in research on the social functions of metaphor in Chinese, I have investigated two Chinese metaphors whose uses have profoundly shaped China’s collective unconscious. The metaphor a serviceman in a collective is a cog in a machine comprises the blending of two culture-specific beliefs and two basic metaphors. The final conceptual product creates a new social category – the serviceman. The creation of this category defines the ideal self vis-à-vis the socialist collective and pushes for actions towards intended sociopolitical goals. The metaphor national identity is sibship consists of the conceptual integration of the basic metaphor a nation is a family with a series of entailments, and the culture-specific Chinese belief in filial piety. In usages involving different collocation contexts, this metaphor has undergone shifts of the target domain. The metaphor ethnic and cultural identity is sibship is used to address or refer to non-mainland Chinese populations. The metaphor national identity is sibship is in use for a specified target domain in reference to the ethnic minorities in China. In addition, the focus on identity shifts to sympathy and solidarity when used for injured or disadvantaged groups. The differential specifications are a function of the recognized sociopolitical priorities. Functionally, the contextual uses of the metaphor construct identity and unity, and define and reinforce social alignments in the interest of pursuing desired sociopolitical goals. The study of metaphor in terms of its sociopolitical function and sociohistorical context provides a valuable perspective from which to understand the zeitgeist of a society. The uses of metaphors as keywords of public discourse are symptomatic of social, political, and economic change, reflecting ‘the very age and body of the time.’

Acknowledgements The author appreciates Deborah Grant’s editorial assistance with an earlier version of this chapter, and is grateful to the graduate students in the seminar on Figurative Language at 640

Metaphor in Chinese: Cognition, culture, and society

the Department of East Asian Languages and Literatures, University of Oregon, as well as to the audiences of the 2014 Stockholm Metaphor Festival and AMPRA2014 at UCLA, for the valuable discussions and exchanges. All errors in the writing belong to the author alone.

Notes 1 The notion of ‘basic metaphor’ is chosen here for a non-primary (non-bodily based) metaphor that nevertheless constitutes a concept basic to our experience with the material world. Unlike primary metaphors, basic metaphors can be culture-specific. 2 Pre-modern data were retrieved from the Chinese corpus by Peking University Center for Chinese Linguistics (http://ccl.pku.edu.cn:8080/ccl_corpus/index.jsp?dir=gudai); modern data on Mainland Mandarin and Taiwan Mandarin were retrieved respectively from the Xinhua News Agency subcorpus (XIN) and the Central News Agency sub-corpus (CAN) of the Chinese Gigaword Corpus 5.0 (http://wordsketch.ling.sinica.edu.tw/). Additionally, 蔣介石《抗戰勝利告全國軍民及全世界人士 書》,《毛澤東選集》, and simplified Chinese materials from Internet-ZH (https://the.sketchengine. co.uk/auth/corpora/) were included in the data. Depending on the source, the script varies between the traditional and the simplified version. 3 平津 (Ping-Jin) is a portmanteau word consisting of the second syllables of Beiping (Peking, Beijing) and Tianjin.

References Ahrens, Kathleen (2002) When love is not digested: Underlying reasons for source to target domain pairings in the Contemporary Theory of Metaphor. In Yuchao E. Hsiao (ed) Proceedings of the First Cognitive Linguistics Conference, Cheng Chi University, 273–302. Ahrens, Kathleen (2010) Mapping principles for conceptual metaphor. In Graham Low, Zazie Todd, Alice Deignan and Lynne Cameron (eds) Researching and Applying Metaphors in the Real World, Amsterdam: John Benjamins, 185–208. Ahrens, Kathleen, Siaw Fong Chung, and Chu-Ren Huang (2003) Conceptual metaphors: ontologybased representation and corpora driven mapping principles. Proceedings of the ACL Workshop on the Lexicon and Figurative Language, Japan, 41–53. Ahrens, Kathleen, Liu, Ho-Ling, Lee, Chia-Ying, Gong, Shu-Ping, Fang, Shin-Yi, and Hsu, Yuan-Yu (2007) Functional MRI of conventional and anomalous metaphors in Mandarin Chinese. Brain and Language 100: 163–71. Angus, Lynne and Mio, Jeffrey S. (2011) At the ‘Heart of the matter’: Understanding the importance of emotion-focused metaphors in patient illness narratives. Genre 44(3): 349–61. Barcelona, Antonio and Soriano, Christina (2004) Metaphorical conceptualization in English and Spanish. European Journal of English Studies 8: 295–307. Bennett, Lyn (2004) Boundaries and frontiers: Metaphor and the rhetoric of interdisciplinarity in literary studies. Genre 37(3–4): 505–29. Boers, Frank, Demecheleer, Murielle and Eyckmans, June (2004) Cross-cultural variation as a variable in comprehending and remembering figurative idioms. European Journal of English Studies 8: 375–88. Boroditsky, Lera (2000) Metaphoric Structuring: Understanding time through spatial metaphors. Cognition 75(1): 1–28. Boroditsky, Lera, Fuhrman, Orly, and McCormick, Kelly (2011) Do English and Mandarin speakers think about time differently? Cognition 118(1): 123–9. Bowdle, Brian F. and Gentner, Deidre (2005) The career of metaphor. Psychological Review 112(1): 193–216. Bronowski, Jacob (1972) Science and Human Values. New York: Harper Torchbooks. Charteris-Black, Jonathan (2003) Speaking with forked tongue: A comparative study of metaphor and metonymy in English and Malay phraseology. Metaphor and Symbol 18(4): 289–310. Charteris-Black, Jonathan (2004) Corpus Approaches to Critical Metaphor Analysis. Basingstoke: Palgrave-MacMillan. Charteris-Black, Jonathan (2005) Politicians and Rhetoric: The Persuasive Power of Metaphor. Basingstoke: Palgrave-MacMillan.

641

Zhuo Jing-Schmidt Chou, Clay Y.-C. (2014) Metaphors about the Sunflower Movement in Taiwanese online media. Paper presented at AMPRA14, October 17–19, UCLA. Chung, Siaw Fong, Kathleen Ahrens, and Chu-Ren Huang (2003) economy is a person: A ChineseEnglish corpora- and ontological-based comparison using the Conceptual Mapping Model. In the Proceedings of the ROCLING XV R.O.C. Computational Linguistics Conference XV, 2003, National Tsing-Hwa University, Taiwan, 87–110. Chung, Siaw Fong, Chu-Ren Huang, and Kathleen Ahrens (2003) economy is a transportation_device: Contrastive representation of source domain knowledge in English and Chinese. Proceedings of the Special Session for the International Conference on Natural Language Processing and Knowledge Engeneering: Upper Ontology and Natural Language Processing, 2003, Beijing, China, 790–6. Crocker, J. Christopher (1977) The social functions of rhetorical forms. In J. D. Sapir and J. C. Crocker (eds) The Social Use of Metaphor: Essays on the Anthropology of Rhetoric, Philadelphia: University of Pennsylvania Press, 33–66. Deacon, Terence W. (1997) The Symbolic Species: The Co-evolution of Language and the Brain. New York/London: Norton. Deignan, Alice (2005) Metaphor and Corpus Linguistics. Amsterdam: John Benjamins. Deignan, Alice (2006) The grammar of linguistics metaphors. In Anatol Stefanowitsch and Stefan Gries (eds) Corpus-based Approaches to Metaphor and Metonymy, Berlin: Mouton, 106–22. Dunbar, Robin, Chris Knight, and Camilla Power (eds) (1999) The Evolution of Culture. Edinburgh: Edinburgh University Press. Edelman, Murray (1971) Politics as Symbolic Action: Mass Arousal and Quiescence. Chicago: Markham. Edelman, Murray (1977) Political Language: Words that Succeed and Policies that Fail. New York: Academic. Elford, Christopher (2014) ‘And Never The Twain Shall Meet’: East and West as Conceptual Metaphors in the Study of Chinese Literature. Paper presented at AMPRA14, October 17–19, UCLA. Fauconnier, Gilles (1994) Mental Spaces. New York: Cambridge University Press. [Originally published 1985, Cambridge: MIT Press.] Faucconier, Gilles (1997) Mappings in Thought and Language. New York: Cambridge University Press. Fauconnier, Gilles and Turner, Mark (1998) Conceptual integration networks. Cognitive Science 22(2): 133–87 Fauconnier, Gilles and Turner, Mark (2002) The Way We Think: Conceptual Blendings and the Mind’s Hidden Complexities. New York: Basic Books. Feldman, Jerome A. (2006) From Molecule to Metaphor. A Neural Theory of Language. Cambridge, MA: MIT Press. Fraser, John (1980) The Chinese: Portrait of a People. New York: Summit Books. Gentner, Dedre and Wolff, Phillip (1997) Alignment in the processing of metaphor. Journal of Memory and Language 37: 331–55. Gentner, Dedre and Bowdle, Brian F. (2001) Convention, form, and figurative language processing. Metaphor and Symbol 16(3–4): 223–47. Gibbs, Raymond Jr. (1994) The Poetics of Mind: Figurative Thought, Language, and Understanding. New York: Cambridge University Press. Gibbs, Raymond Jr. (1999) Taking metaphor out of our heads and putting it into the cultural world. In R. Gibbs and G. Steen (eds) Metaphor in Cognitive Linguistics, Amsterdam: John Benjamins, 145–66. Gibbs, Raymond Jr. (2001) Evaluating contemporary models of figurative language understanding. Metaphor and Symbol 16(3–4): 317–33. Giora, Rachel (1997) Understanding figurative and literal language: the graded salience hypothesis. Cognitive Linguistics 8(3): 183–206. Giora, Rachel (2003) On our Mind: Salience, Context, and Figurative Language. New York: Oxford University Press. Glucksberg, Sam (2003) The psycholinguistics of metaphor. Trends in Cognitive Sciences 7(2): 92–6. Grady, Joseph (1997) Foundations of meaning: Primary metaphors and primary scenes. Doctoral dissertation, University of California, Berkeley. Grice, H. P. (1975) Logic and conversation. In Cole, P. and Morgan, J. (eds) Syntax and Semantics 3: Speech Acts, New York: Academic, 41–58. Heath, Shirley B. (2011) Narratives and Metaphors of the Middle. Genre 44(3): 293–300.

642

Metaphor in Chinese: Cognition, culture, and society Hilpert, Martin (2006) Keeping an eye on the data: metonymies and their patterns. In Anatol Stefanowitsch and Stefan Gries (eds) Corpus-based Approaches to Metaphor and Metonymy, Berlin: Mouton, 123–51. Huang, Hsin-mei and Hsieh, Shelley Ching-Yu (2007). Time-moving Metaphors and Ego-moving Metaphors: Which Is Better Comprehended by Taiwanese? The 21st Pacific Asia Conference on Language, Information, and Computation, November 1–3, Seoul National University. Jing-Schmidt, Zhuo (2008) Much mouth much tongue: Chinese metonymies and metaphors of verbal behaviour. Cognitive Linguistics 19(2): 241–82. Jing-Schmidt, Zhuo (2014a) Corruption as Disease or Enemy: Cultural model and metaphor choice in Chinese political discourse. Paper presented at Stockholm 2014 Metaphor Festival, August 27–30, Stockholm. Jing-Schmidt, Zhuo (2014b) Metaphor in sociopolitical discourse: creation and entrenchment. Paper presented at AMPRA14, October 17–19, UCLA. Kimmel, Michael (2004) Metaphor variation in cultural context: Perspectives from anthropology. European Journal of English Studies 8(3): 275–94. Kittay, Eva Feder (1987) Metaphor: Its Cognitive Force and Linguistic Structure. Oxford: Clarendon Press. Kövecses, Zoltán (1991) Happiness: A definitional effort. Metaphor and Symbolic Activity 6(1): 29–46. Kövecses, Zoltán (2005) Metaphor in Culture. Cambridge: Cambridge University Press. Lai, Vicky Tzuyin, Curran, Tim, and Menn, Lise (2009) Comprehending conventional and novel metaphors: An ERP study. Brain Research 1248: 145–55. Lakoff, George and Johnson, Mark (1980) Metaphors We Live By. Chicago: University of Chicago Press. Lakoff, George and Johnson, Mark (1999) Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought. New York: Basic Books. Lakoff, George and Kövecses, Zoltán (1987) The cognitive model of anger inherent in American English. In D. Holland and N. Quinn (eds) Cultural Models in Language and Thought, Cambridge: Cambridge University Press, 195–221. Link, Perry (2013) An Anatomy of Chinese: Rhythm, Metaphor, Politics. Cambridge, MA: Harvard University Press. Miller, Eugene (1979) Metaphor and political knowledge. The American Political Science Review 73(1), 155–70. Mio, Jeffrey S. (1997) Metaphor and politics. Metaphor and Symbol 12: 113–33. Nicaise, Laurent (2010) Metaphor and the context of use: A Multidimensional approach. Metaphor and Symbol 25(2): 63–73. Peng, Xinjia and Jing-Schmidt, Zhuo (2014) Anti-graft metaphors in Chinese political discourse. Paper presented at AMPRA14, October 17–19, UCLA. Polley, Carl A. (2012) Metaphors for happiness in English and Chinese. PhD dissertation, University of Hawaii at Manoa. Porto, M. Dolores and Romano, Manuela (2013) Newspaper metaphors: Reusing metaphors across media genres. Metaphor and Symbol 28(1): 60–73. Sapir, J. David (1977) The anatomy of metaphor. In J. D. Sapir and J. C. Crocker (eds) The Social Use of Metaphor: Essays on the Anthropology of Rhetoric, Philadelphia: University of Pennsylvania Press, 3–32. Searle, John (1979) Metaphor. In A. Ortony (ed) Metaphor and Thought, New York: Cambridge University Press, 92–123. Slingerland, Edward (2004) Conceptions of the self in the Zhuangzi: Conceptual Metaphor analysis and comparative thought. Philosophy East and West 54(3): 322–42. Stern, Josef (2006) Metaphor, literal, literalism. Mind and Language 21(3): 243–79. Tomasello, Michael (1995) Joint attention as social cognition. In C. Moore and P. J. Dunham (eds) Joint Attention: Its Origins and Role in Development, Hillsdale, NJ: Erlbaum, 103–30. Tomasello, Michael (1999) The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press. Tomasello, Michael (2003) Constructing a Language. Cambridge, MA: Harvard University Press. Tucker, Don M. (2007) Mind from Body: Experience from Neural Structure. Oxford: Oxford University Press.

643

Zhuo Jing-Schmidt Turner, Mark (1996) The Literary Mind: The Origins of Thought and Language. New York: Oxford University Press. Wang, Lixia (2003) 隐喻探源 － 兼论我国历史上的隐喻研究 Journal of Hubei Institute of Nationalities 2: 102–6. Xiong, Jiajuan and Chu-Ren Huang (2015) Being assiduous: Do we have bitterness or pain? The synaesthetic and conceptual metaphors of bitterness and pain in Chinese and English. Paper presented at the Chinese Lexical Semantics Workshop (CLSW) 2015, Beijing, China. Ye, Zhengdao (2001) An inquiry into ‘sadness’ in Chinese. In Jean Harkins and Anna Wierzbicka (eds) Emotions in Crosslinguistic Perspective. Berlin/New York: Mouton de Gruyter, 359–404. Yu, Ning (1995). Metaphorical expressions of anger and happiness in English and Chinese. Metaphor and Symbolic Activity 10(2): 59–92. Yu, Ning (1998) The Contemporary Theory of Metaphor: A Perspective from Chinese. Amsterdam: John Benjamins. Yu, Ning (2000) Figurative uses of finger and palm in Chinese and English. Metaphor and Symbol 15(3): 159–75. Yu, Ning (2001) What does our face mean to us? Pragmatics and Cognition 9(1): 1–36. Yu, Ning (2003a) Metaphor, body, and culture: The Chinese understanding of gallbladder and courage. Metaphor and Symbol 18(1): 13–31. Yu, Ning (2003b) The bodily dimension of meaning in Chinese: What do we do and mean with ‘hands’? In E. Casad and G. Palmer (eds) Cognitive Linguistics and non-Indo-European Languages, Berlin: Mouton de Gruyter, 330–54. Yu, Ning (2003c) Chinese metaphors of thinking. Cognitive Linguistics 14(2–3): 141–65. Yu, Ning (2004) The eyes for sight and mind. Journal of Pragmatics 36(4): 663–86. Yu, Ning (2007) Heart and cognition in ancient Chinese philosophy. Journal of Cognition and Culture 7: 27–47. Yu, Ning (2008) Metaphor from body and culture. In R. Gibbs, Jr. (ed) The Cambridge Handbook of Metaphor and Thought, Cambridge: Cambridge University Press, 247–61. Yu, Ning (2009) The Chinese HEART in a cognitive perspective: Culture, body, and language (Applications of Cognitive Linguistics, vol. 12). Berlin: Mouton de Gruyter.

644

36 Modern Chinese: written Chinese1 Feng Shengli The Chinese University of Hong Kong, Hong Kong, China

This chapter, based on Feng (2009) and Sun (2012), shows how written Chinese is different from spoken Chinese in modern times, how the formal register system of written Chinese has developed since the May Fourth Movement, and, finally, what principles the formal register grammar must observe.

1.  Spoken and written Chinese Modern written Chinese (白话文) is a result of the May Fourth Movement (1919). Before then, Chinese intellectuals (which included virtually everyone who was literate) wrote in classical (literary) Chinese.2 Although there were proposals for writing in the vernacular before the May Fourth Movement, with people like Huang Zunxian 黄遵宪 promoting ‘My hand writes [what] my mouth [says]’ (我手写我口) in the late Qing dynasty (1644–1911), the shift from writing in literary Chinese to writing in the vernacular did not actually occur until the Literary Revolution (文学革命), launched by Hu Shi and Chen Duxiu in 1919. One important argument for replacing literary Chinese with the vernacular in writing was, according to Hu Shi, that literary Chinese became a dead language thousands of years ago. However, what is striking is the fact that there are still remnants of literary Chinese within Modern Chinese vernacular writing. For example: (1)  当然,这并不是说,凡是汉族所扩及的地区,原来在这些地方居住的民族都被吸收 为汉族了。事实并不如此,即在目前,汉族聚居的地区里还是有少数民族杂居在 内。── 费孝通《中国少数民族的发展》  ‘Of course, this does not mean that local people have been entirely integrated into the Han nationality in all of the areas where the Han have penetrated. In fact, this is not so because even now there are many minorities which have mixed with the Han race within Han regions.’ ——Fei Xiaotong, Minority Development in China In (1) there are about 44 morphemes (free, as well as bound) and 11 of them are taken from Classical Chinese, e.g. 所扩及 ‘extend to’, 吸收为 ‘take as’, 并不如此 ‘is certainly not like this’, 即在目前 ‘even now’, 聚居 ‘live together’ and 在内 ‘within’. Actually, classical 645

Feng Shengli

expressions like these are not merely remnants, they are actually required to make the written text sound natural. Zhang (2002) promoted that people ‘should incorporate some literary expressions into their own writing’(要把来自文言的东西融会到自己的笔下) (Zhang 2002: 134). Why is this syncretization of literary forms into the vernacular necessary? Feng (2005) argued that it is essentially a result of register requirements, and Sun (2012) also argued that literary Chinese serves a distinctive linguistic function in modern times which has ensured its survival. In other words, modern formal Chinese would have been difficult to develop without utilizing some literary Chinese. Thus, literary and colloquial Chinese cannot truly be divorced in the modern context of language communication as will be explained below. It is well known that if a writing style is too literary it may not easily be understood by ordinary or sometimes even educated people; while if it is too colloquial (informal) it will not be acceptable in formal situations because of its lack of formality. The traditional dilemma of separating colloquial expressions from literary diction in modern written Chinese has arisen from the inseparability of vernacular grammar with literary expressions which makes it possible to create a formal register of writing. Thus, the formal and informal styles of Chinese can be analyzed as in Figure 36.1. Figure 36.1 illustrates several important points which are elaborated on below. First, what is called written Chinese should be defined in terms of the formal style of writing in Modern Chinese. Second, the notions of ‘informal/formal’ and ‘spoken/written’ are not isomorphic, i.e. ‘formal’ does not only imply ‘written’, nor does ‘spoken’ always refer to ‘informal’, and vice versa. Formal Chinese is also an utterable language and is not reserved only for writing, even if ‘written language’ usually refers to formal or elevated registers. The definition given in example (2a) implies that formal Chinese can be both spoken and written.

Classical Chinese

Ancient

Modern Elevated

Written Chinese

Newly Developed Formal Systems

Formal Chinese

Writings in Vernacular 白话 Spoken Chinese

Everyday Speech Slang

Figure 36.1  Diagram of formal and informal Chinese (Feng 2006)

646

Informal Chinese

Modern Chinese: Written Chinese

Third, Classical Chinese and Modern Chinese should be clearly distinguished here: Classical Chinese refers to the language with many linguistic features of the Han and pre-Han periods (i.e. up to the third century) which remained prevalent up until the May Fourth Movement, while Modern Chinese is defined in terms of its auditory comprehensibility to the ordinary people of today. Thus, according to Feng (2009), speech that cannot be understood by means of its sound alone by an ordinary high school graduate will not be considered modern (for a more detailed discussion of this criterion, see Feng 2003a, 2003b; Sun 2012). Fourth, formal expressions in Modern Chinese come from two major sources: Classical Chinese and completely new expressions that developed within the formal system itself after 1911. For example, there are systematically developed formal expressions like jinxing 进行 ‘carry out’: (2) (a) 我们一定要对这个问题进行研究。 Women yiding yao dui zhege wenti jinxing yanjiu. we definitely need towards this issue carry-out study ‘We must study this question (we must carry out study on this question)’ (b) 瞧！您的闸不灵了,得修理一下了。 Qiao! Nin de zha bu ling le, dei xiuli yixia le look you POSS brake not effective Asp must repair a little Asp ‘Look, your (bicycle) brake has some problems, you have to fix it’ (c) 瞧！您的闸不灵了,*得进行修理了. Qiao! Nin de zha bu ling le, *dei jinxin xiuli le. Look you POSS brake not effective Asp must carry-out repair Asp ‘Look, your (bicycle) brake has some problems, you have to fix it’ (d) 闸皮失灵,必须进行修理！ Zha pi shi ling, bixu jinxing brake rubber lose efficiency, must carry-out ‘If the brake is not efficient, one must repair it’

xiuli repair.

As seen above, in oral speech (2b), native speakers do not use the formal expression jinxing at all, as seen in (2c). However, on formal occasions, e.g. regulating transportation, the expression can be paraphrased formally as shown in (2d). Note that even though some formal expressions have been taken from Classical Chinese, they are no longer considered as Classical Chinese because they have been accepted into and processed by the modern system of formal Chinese and are understood as such by native speakers.3 Sun (2012: 74) provides a good example of this: (3) (a) 陈毅及其子 Chenyi ji qi zi Chenyi and his son ‘Chenyi and his son’

(b) 陈毅及其子女 Chenyi ji qi zi nü Chenyi and his son daughter ‘Chenyi and his children’

(c) 陈毅及他的孩子 Chenyi ji ta de haizi Chenyi and his children ‘Chenyi and his children’ 647

Feng Shengli

The above three noun phrases all involve a pronoun, qi 其 ‘it, its, his’ or ta de 他的 ‘his’. Comparatively speaking, the style of (3b) is more modern than that of (3a), and (3b) is also more formal than (3c). These samples represent a typical case of ‘classical grammar used in Modern Chinese’ (Sun 2012: 75). Finally, it is not necessary to master Classical Chinese in order to become a writer of Modern Chinese because the classical forms used today are very limited (only about 250 words and 300 phrasal patterns).4 Thus, it is unnecessary (if not a misconception) to promote the study of Classical Chinese for the purpose of mastering the writing of Modern Chinese. Given the above arguments we are not merely facing traditional questions such as where can we draw the line between modern expressions and classical patterns? and how can we separate them? but also new questions like (i) why does Modern Chinese need expressions from Classical Chinese? (ii) how can classical Chinese be syncretized into modern Chinese? (iii) what types of classical expressions can be syncretized into modern writing? and (iv) how many classical expressions are used in modern Chinese? All of these are important questions raised by this study. Though we will not delve into every detail of these questions here, it is important to clarify why Modern Chinese needs expressions from Classical Chinese. The traditional philologist Huang Kan 黄侃 presciently answered this question long ago: ‘The separability of written language from spoken language  .  .  .  is inevitable’ (文与言判  .  .  .  非苟而已 Huang Kan’s Diary《黄侃日记》1922: 199), because modern written Chinese makes use of Classical Chinese to distance itself from the vernacular. It is well known that writing in the vernacular started with the slogan ‘write what you say’ (写口), but has ended up with a new separation of writing and speaking again today. This may not have been expected by the initiators of the May Fourth Movement, but was accurately predicted by Huang Kan a long time ago. Of course, today’s separation of writing from speaking is not simply a reversion to the past. What is important to note here is the fact that all languages necessitate two relative functions: formal and informal. If ‘write what you say’ results in a purely informal style and if Classical Chinese served as the formal style of the past, then something must have been developed or created to fill the gap when Classical Chinese was abolished by the May Fourth Movement because there was a need for a formal style in social communications. However, why must the creation of a formal register make use of Classical Chinese? As was argued in Feng (2003a, 2003b), a formal style can be established by creating a sense of expressive distance from everyday speech. That is to say, when linguistic expressions are used, the more distant the expressions are kept from everyday speech, the more formal (or elevated) sense they can create in their listeners. The demand for vernacular Chinese to function formally after the May Fourth Movement probably motivated speakers and writers to use some classical forms to satisfy their urgent need for stylistic effect. This is the reason why the separation of writing (formal) from speaking (informal) has been resurrected in modern times.

2.  The prosodic grammar of written Chinese If classical forms must be used to write in a formal style in Modern Chinese then we must ask: in what ways can these classical forms be used? Though there may be a variety of ways to make use of them, one obvious structure should be pointed out in the study of modern

648

Modern Chinese: Written Chinese

formal style; namely, they cannot be used directly in modern formal writing without a prosodic justification. For example: (4) (a) 我校 *我的校 *我们校 *我们的校 wo xiao *wo DE xiao *women xiao *women DE xiao I school I POSS school we school we POSS school ‘my school’ ‘my school’ ‘our school’ ‘our school’ (b) 住嘴 *住嘴巴 zhu zui *zhu zui-ba stop mouth stop mouth ‘shut up’ ‘shut up’ (c) 他困在山里了。 *他(在山里)困了 Ta kun zai shan li le. *Ta (zai shan li) kun le. he stop at mountain inside Asp he (at mountain inside) stop Asp ‘He was stopped in the mountains’ ‘He was stopped (in the mountains)’ It has been shown (Feng 2003a, 2003b; M. Huang 2008) that monosyllabic words in Classical Chinese (Qian’ou ci 嵌偶词) must be used in a disyllabic template in order to be considered grammatical in Modern Chinese.5 In fact, more examples demonstrate that not only must the monosyllabic classical forms be prosodically conditioned in modern formal writings, as we saw above (and in (9), (11) below), moreover newly developed formal expressions in the modern written language (Heou ci 合偶词) are also strictly constrained by prosody (see Y. Wang 2009). For example: (5) (a) VV selects VV (b) VV selects NN 加以 + 批*(判) 坚持 + *(真)理 jiayi + pi*(-pan) jianchi + *(zhen)li give + criticism insist on + truth 予以 + 批*(评) yuyi + pi*(ping) give + criticism

面临 + 危*(险) (临危) mianlin + wei*(xian) (lin-wei) face + danger (face danger)

从事 + 教*(学) 导致 + *(疾)病 congshi + jiao*(-xue) daozhi + *( ji)bing engage in + teaching cause + disease (c) Adv selects VV (d) Adv selects AA 公然 + *(逃)跑 极其 + *(容)易 gongran + *(tao)pao jiqi + *(rong)yi openly + escape extremely + easy 共同 + *(协)商 更加 + *(美)好 gongtong + *(xie)shang gengjia + *(mei)hao jointly + discuss even more + beautiful 并肩 + 战*(斗) 极为 + *(不)满 bing-jian + zhan*(dou) jiwei + *(bu)man shoulder to shoulder + fight extremely + unsatisfied

649

(持理) (chi-li) (insist on truth)

(致病) (zhi-bing) (cause disease)

Feng Shengli

There exist about 400 Heou ci (see Expressions of Written Chinese《汉语书面用语 初编》2006) which must be used in a disyllabic couplet. Given the examples of Qian’ou ci and Heou ci and the fact that formal expressions are prosodically constrained, we are able to conclude that the grammar of formal Chinese is essentially a prosodic grammar, which can be defined as follows. (6) Prosodic grammar If the computational system of a language operates under prosodic conditions, then the grammar of the language is characterized as a prosodic grammar. Regarding the characteristics of the prosodic grammar of written Chinese, several principles and rules have been proposed in previous studies. For example: (7) (a) *[σ] → [σσ]foot Disyllabic Foot Structure: A monosyllable cannot form a foot and thus cannot stand alone. (b) [σσ]PrWd → [σσ]PrWd +[σσ]PrWd Stylistic Coherence Principle: A prosodic word (PrWd), commonly formed by a foot, selects another prosodic word in formal Chinese. (c)  Principle of Prosodic Stylistics: The more formal an expression, the more prosodic words are used to compose it, and vice versa. (7a) gives rise to the constraint that classical monosyllabic words must occur in a disyllabic template in Modern Chinese, while (7b) results in a ‘disyllabic word used in a disyllabic copulate’. (7c) affects the proportion of formal features used in modern writing (see section 4). In addition, Sun (2012) has further specified and developed these principles in terms of parallel tendency 整齐律, stylistic coherence 谐体律, brevity 求简律, and elegancy tendency 趋雅律, which function as four writing rules that determine the usages of classical expressions in modern writings. Given these principles and rules, we would expect there to be an unavoidable interleaving of expressions between literary dictions and colloquial patterns in modern formal Chinese. This turns out to be the case as we will see below.

3.  Interleaving of literary and colloquial dictions Based on the theory and analysis given above, we can view modern formal Chinese as a new language which developed after the May Fourth Movement. It is new in the sense that formal Chinese serves as the register grammar within Modern Chinese. What is this new register grammar composed of? Feng (2005) suggests that modern formal Chinese consists of several components, as illustrated in Figure 36.2. According to Figure 36.2, formal Chinese can be characterized as a formal language composed of a syntactic system based on Modern Chinese (i.e. Mandarin Chinese) with a set of prosodic constraints on its morphosyntactic operations and a lexicon (including grammatical features) consisting of expressions from ordinary speech, Classical Chinese, dialects and foreign languages.6 How can Classical Chinese expressions be used in Modern Chinese? First, in Modern Chinese, many monosyllabic words must be used in a disyllabic template. As seen in (4a), 650

Modern Chinese: Written Chinese MODERN FORMAL CHINESE

Amalgamation Vernacular Syntax

Prosodic Grammar

Process Select

oral words

classical words

dialect words

foreign words

......

Figure 36.2  Modern formal Chinese

we can say wo xiao ‘my school’ or li xiao ‘depart school’, but not *women xiao ‘our school’ or *likai xiao ‘depart school’. There are also disyllabic words that must be used within a tetrasyllabic template. Thus, we may say jinxing gaige ‘carry-out reform’ and weida zuguo ‘great country’, but not *jinxing gai ‘carry-out reform’ or *weida guo ‘great country’. Previous studies have characterized the former as semi-free morphemes or Qian’ou ci (i.e. monosyllabic words that must be used in a disyllabic template), and the latter as Heou ci (i.e. disyllabic words that must be used in a tetrasyllabic template). Although insightful, previous studies suffer from three limitations. First, a theory has yet to be developed to explain why there are such prosodic requirements. Second, it is unclear how large the inventories of Qian’ou ci and Heou ci are. Third, very little consideration has been given to different social–political settings in Hong Kong, Taiwan and Mainland China that drastically influence the use of syllabic-templates. Recent developments on theories of Chinese prosodic grammar have motivated researchers to study the template-dependent phenomena much more thoroughly. First, a new field of prosodic morphology developed in the West has been introduced into Chinese since 1997. The prosodic template analysis has been considered ‘an important contribution to morphological studies’ (Pan et al. 2004: 509). Second, the notion of the prosodic word has been developed as a shape-measuring device to characterize why there are so many disyllabic expressions in Modern Chinese (words, fixed forms, as well as short phrases). Furthermore, linguists have developed a new framework of Stylistic-Register Grammar for characterizing the Qian’ou ci with an Elevated Style and the Heou ci with a Formal Style, thereby making it plausible to measure where and how they appear prosodically, syntactically and stylistically in Hong Kong, Taiwan and Mainland China (Feng 2010a; Shi 2006; Tsou and Mok 2013). Thus, the current chapter will tackle questions theoretically as: (a) How does prosodic template affect not only reduplications but also compounds and phrases? (b) Why are monosyllabic forms used for elevated styles and disyllabic forms for formal? And (c) can prosody be morphologically and stylistically characterized as part of grammar? Empirically, 651

Feng Shengli

we must answer: (a) how many Qian’ou ci and Heou ci are there in total? (b) What are the frequencies and degree of the elevated and formal forms used in different syntactic positions, sociolinguistic occasions and political environments? The intriguing phenomena of monosyllabic and disyllabic words occurring in prosodic templates did not receive serious attention until the appearance of Prosodic Morphology in 1983 and its introduction into Chinese in 1996. A preliminary statistical investigation regarding the new field shows that before 1996, there were only 15 articles relevant to word prosody since 1938, while 87 articles were published in 14 years (from 1996 to 2010) on many aspects of prosodic word formation. Lü (1962, 1963) presents the earliest observations of the relevant phenomena, but they are considered in terms of the ‘semi-free morpheme’ or a [2+2] syllable-constraint, but not a template-dependent prosodic constraint. Although Lü’s pioneer work has profoundly influenced the field of Chinese morphology, there are obvious limitations and problems with his framework. First, in his article ‘On Free and Bound’, Lü listed five circumstances where a ‘bound’ monosyllabic morpheme can be used freely (thus he named such kind of forms ‘semi-free morphemes’): 1) different dialects, such as xie vs. xiezi ‘shoe’; 2) written Chinese, as hua vs. huar ‘flower’; 3) Fachschrifttum usage, such as bi ‘nose’ in iatrology; 4) idioms or fixed expressions, such as bu zhi bu jue [not know not feel] ‘unconscious’, and 5) special structures, such as chu-le ming [out Asp name] ‘become famous’. Apparently, Lü’s ‘bound/free morpheme’ proposal is only a partial understanding, which is insufficient for explaining the full picture. For example, if xiao ‘school’ is conceived as a semi-free morpheme in expressions like wo xiao ‘my school’, gai xiao ‘this school’, dao xiao ‘arrive at school’, li xiao ‘depart from school’, it is difficult to explain why the same expressions become ungrammatical when the words used with xiao are changed to polysyllabic ones, such as in *women xiao ‘our school’, *zhege xiao ‘this school’, *daoda xiao ‘arrive at school’, *likai xiao ‘depart from school’. Given the same syntactic and semantic environments where xiao occurs, the ‘bound/free morpheme’ understanding doesn’t explain why xiao is valid for use in the former group but not in the latter cases. The [2+2] syllable-constraint is proposed in Lü’s (1963) paper, ‘A Preliminary Investigation on Mono-/Disyllabic Issues in Chinese’, where five types of [2+2] phenomena are listed: (i) light verbs like jinxing ‘carry out’ and jiayi ‘infect’ that must take disyllabic gerunds (i.e. 名动词 in Zhu’s [1982] term) as their object; (ii) disyllabic verbs like guanli ‘manage’ that must take disyllabic nouns as their object; (iii) objects must be disyllabic before a disyllabic verb, such as gangtie ‘steel’ in gangtie shengchan ‘steel production’; (iv) disyllabic adjectives must modify disyllabic nouns, such as weida ‘great’ in weida renwu ‘great character’, and (v) a disyllabic word must coordinate with another disyllabic word such as jinxing gaige ‘carry out a reform’. Lü’s observation is very important, but it leaves many open questions to be answered, for example, why the disyllabic constraint is not on the size of the words themselves, but on the environments in which the words occur, and on the accompanying word with which a disyllabic word combines. Obviously, answering the question left by Lü’s work demands the development of a new theory. The issue is first tackled in a study of teaching Chinese as a second language. Feng (2003b: 53–63) characterizes the first-type phenomenon, i.e. the monosyllabic words used with another monosyllabic word in terms of ‘syntactically-free-but-prosodically-bound’ (i.e. Qian’ou ci) in the newly introduced Prosodic Word (PrWd) framework (see Pan et al. 2004 for comments on this work). In this system, xiao is a syntactically free morpheme and hence can be used with functional elements like pronouns, prepositions and determinatives 652

Modern Chinese: Written Chinese

in wo xiao ‘my school’, zai xiao ‘at school’ and gai xiao ‘this school’. However, xiao is not prosodically free (because it is taken from Classical Chinese, see Feng 2003a for detailed analysis), and hence must occur within a disyllabic template (which is defined in terms of a Minimal Foot in prosodic morphology, Feng 2000) in order to be used in Modern Chinese. Thus, *women xiao ‘our school’, *wo de xiao ‘my school’, and *women de xiao ‘our school’ in (4a) are all unacceptable because the outcomes violate the prosodic boundedness constraint. In Feng’s treatment, the ‘semi-free’ morphemes are not simultaneously bound and free as in Lü’s system. The seemingly bound-effects are actually violations of a prosodic constraint of the disyllabic template. In the same article, Feng also proposed that the [2+2] syllableconstraint is indeed a reflection of a PrWd-compound effect in prosodic morphology, which generates many disyllabic words that also require disyllabic words with which to be (according to register grammar as well, see below). This theory provides more powerful explanations to the concerned phenomena (see H. Wang 2008: 310–11 for detailed comments on this work). However, it is still insufficient. It does not consider whether the templative phenomenon is a sporadic effect or an across-the-board constraint. Therefore, to what extent the prosodictemplative phenomena can be considered a grammar is still a puzzle. Following Lü’s ‘bound-free’ tradition and Feng’s disyllabic insight, Dong (2004) defined the monosyllabic forms more strictly as ‘semi-free morphemes in disyllabic environment’. As a result, the qualified ‘semi-free’ forms in her system are only certain types of syntactically conditioned monosyllabic words, such as (i) Adverb+Verb, (ii) Auxiliary+Verb, (iii) Negator+Verb, (iv) Determinative+Noun, and (v) Noun+Localizer. Although Dong’s study has brought much attention to the computer process on Chinese word-hood, many well-tested monosyllabic Qian’ou ci are excluded from Dong’s study. As pointed out by Wang Hongjun (2008: 307), a great number of Qian’ou ci are left out because Dong did not recognize VO forms as phrases, thus nouns occurring in the object position (e.g. chuan yi ‘put-on clothing’, dao xiao ‘arrive at school’) are excluded from her list. As a result, the inventory of all the semi-free monosyllabic words in Modern Chinese is incomplete. In 2006, a handbook, Expressions of Written Chinese was published for Chinese L2 teaching and learning (Feng 2006). In this handbook, 250 monosyllabic words used in disyllabic templates and 400 disyllabic words used in tetrasyllabic templates are listed. This collection of Qian’ou ci and Heou ci, as preliminary as it is, shows that the templative phenomena are not sporadic effects, but rather across-the-board constraints. The problems with this work are that, as pointed out by Wang Yongna (2009: 313), the inventory of monosyllabic words used in a disyllabic environment is not complete. Many Qian’ou ci used in VO phrases are not included in it. Additionally, the classification criteria for identifying monosyllabic words in disyllabic environments are not strictly defined. As a result, much more work needs to be done in this area. The recognition of prosodic templates for monosyllabic and disyllabic word usages was a result of theoretical exploration on the prosodic words in Chinese, as well as an independent investigation on written Chinese. In 1995, influenced by McCarthy and Prince’s prosodic morphology, Feng proposed that a compound word in Chinese must first be a prosodic word even if a prosodic word may not necessarily be a compound. There is another factor that inspired the discovery of syllabic templative words in Chinese. When teaching the language to foreign students, Feng observed that the grammar of written Chinese is structurally different from that of spoken Chinese. For example, the coordinative verb structure of goumai he yuedu baozhi ‘buy and read newspapers’ in formal written Chinese cannot be used in casual spoken Chinese *wo xiang mai he du baozhi ‘I want to buy and read newspapers’. Triggered by facts like these, Feng (2006) discovered more and 653

Feng Shengli

more structural devices for the formal style, as well as monosyllabic words used in disyllabic templates and disyllabic words in tetrasyllabic templates, which has been considered ‘a new area with great research potentials’ 开辟出了一片新的、大有可为的研究领域 (H. Wang 2008: 311). A series of research and publications have appeared since 2003 regarding both prosodic morphology and written Chinese grammar. While the grammars of human language are considered systemized between written and spoken languages, the stylistic differences between formal/informal and elevated/plain are also grammaticalized systematically. For example: (8) Elevated 我国政府出於同样考虑,认为不宜前往。 Wo guo zhengfu chuyu tongyang kaolü, our country government have same consideration renwei bu yi qian wang. decide no should forward go ‘Our government will take the same consideration and decide not to go’

Casual 咱想的跟您一样,就甭去了。 Zan xiang de gen nin yiyang, we think with you same, ‘I agree, let’s not go’

jiu then

beng no-need

qu go

le Asp

Within the theory of prosodic morphology in Chinese and the stylistic-register grammar proposed in recent years (Feng 2010a, 2011), we now understand that the bizarre templative behavior of words is in essence a natural consequence of the interface processes between different components of the grammar. Based on the theory developed and the empirical results accumulated thus far, three fundamental categories of linguistic communication devices for adjusting personal relations when communicating with each other were proposed in the newly developed stylisticregister grammar, namely, formal, informal and elevated styles, as shown in Figures 36.3 and 36.4. It is also argued that the three poetic categories of Feng 风 Airs, Ya 雅 Elegantiae and Song 颂 Eulogia in the Odes of the Zhou dynasty (eleventh century bc to 256 bc) reflect and exactly match the registry categories of formal, informal and elevated. As a result, the poetic

casual 8 Oral 1 5

2

6

4

oral

3

Elevated

7

Formal

Figure 36.3  Tridimensional system of stylistic-register grammar (Feng 2010)

654

Modern Chinese: Written Chinese Elevated (Classic)

Casual

Neutral

Formal

Figure 36.4  Interrelationships among elevated, formal and casual forms 风 Airs

casual 8 Oral 1 5 Eulogia 颂

2

6

4

oral

3

Elevated

7

Formal

雅 Elegantiae

Figure 36.5  Evidence for tridimensional register grammar

categories in the past provide historical evidence for the registry categories of the present (Figure 36.5). As seen above, modern formal and elevated Chinese is composed of a syntactic system based on modern system Mandarin Chinese with a set of prosodic constraints on its morphosyntactic operations and a lexicon consisting of expressions from Classical Chinese, ordinary speech, dialects, and foreign languages. Of course, this does not imply that the expressions from these sources can all be assimilated directly into modern writing without any manipulation and justification. On the contrary, all of the elements taken from spoken, dialect, foreign and classical sources may need to be manipulated according to the prosodic stylistic principles given above in (6). Actually, as represented in Figure 36.2, there are three principles being proposed for organizing the different elements in modern formal Chinese, which are elaborated on below.

4.  Principles of integration 4.1.  Selection principle The selection principle is very simple: a speaker or writer has the freedom to select whatever element they want from the lexicon (including phrasal structures and sentence patterns marked/classified in the system) for the purpose of expressing different stylistic-register grammars of language, from colloquial to formal and to elevated. If register expressions can be characterized in terms of their distance (7a) from everyday speech (or degrees of familiarity), then different sources of expressions can be classified according to their different degrees

655

Feng Shengli

of formality in written Chinese. This is illustrated as in (9) (‘>’ stands for ‘is more formal than’).   (9) Degree of Formalities of Register Sources Classical > foreign > dialect > spoken expressions The more familiar an expression, the closer it is to everyday speech and vice versa. Given the distance-gradation-principle in (9), we are able to rank expressions into degrees of formality according to their stylistic properties, as well as their source. For example: (10) fangfu 仿佛 > sihu 似乎 > haoxiang 好像 ‘seems like’ These three words share the same meaning. The only difference between them is that fangfu is more formal than sihu,7 which is more formal than haoxiang.8 Although there may not always be triplets of degrees of formality like example (10), each individual expression can theoretically be ranked with a proper degree of formality according to the distancegradation-principle. In fact, Expressions of Written Chinese ranks virtually all formal expressions in Modern Chinese into one of the three degrees of formality based on the principles and methods illustrated above. As a result, one can select an expression (Qian’ou ci, Heou ci or a phrase Guju xing 古句型) from this reference book in order to facilitate formal writing and speaking.

4.2.  Process principle Word or phrase-selection from sources or reference books is a preliminary step towards expressing a formal style in Chinese. However, this does not mean that one can freely select whatever they want. Also, it is not the case that every formal element, whether a word or phrase, can be used directly without some necessary grammatical modification. Thus, principles for selecting and using formal expressions are needed. Two such important processing principles were proposed in Feng 2005: the Principle of Auditory Intelligibility (PAI)9 and the Principle of Shaping by Prosody (PSP). The PAI requires any formal expression to be auditorily intelligible, while the PSP requires that auditorily intelligible formal expressions must be shaped within a proper prosodic structure. For example, (11) zhidao 知道 vs. zhi 知 ‘know’ (a) 他知道今天没有课。 Ta zhidao jintian mei you ke. He know today not have class ‘He knows that there is no class today’

(b) *他知今天没有课。 *Ta zhi jintian mei you ke. He know today not have class ‘He knows that there is no class today’ 他不知今天没有课。 Ta bu zhi jintian mei you ke. He not know today not have class ‘He does not know that there is no class today’ 656

Modern Chinese: Written Chinese

(12) jieshi 解释 vs. jie 解 and shi 释; and xian 鲜 (a) 这道题无解。 Zhe-dao ti wu jie This-CL problem no solution ‘This (mathematics) problem has no solution’

(b) ♪这道题无释。 ♪Zhe-dao ti wu shi This-CL problem no solution ‘This (mathematics) problem has no solution’ (c) 多还可以,♪鲜则不行。 duo hai keyi, xian ze more still okay, less however ‘More is fine, less wouldn’t do’

bu not

xing okay

In (11), the monosyllabic counterpart zhi of the common word zhidao is auditorily intelligible to native speakers of high-school level or above, even though (11b) is not grammatical. In (12a) the monosyllabic counterpart of jieshi, jie is also intelligible. However, (12b) and (12c) represent a different scenario: neither shi nor xian are auditorily intelligible by a native high-school-level speaker, although they are free-standing, common words in Classical Chinese. According to the Auditory Intelligibility Principle, both of them must be excluded from Modern Chinese. In other words, a boundary between classical and Modern Chinese can conveniently be drawn according to the Auditory Intelligibility Principle: anything that is not auditorily intelligible is not Modern Chinese. Given this AIP, the frequency ( pindu 频度) of classical forms occurring in modern writing, proposed by Sun (2012), becomes an important criterion for identifying what can be considered as ‘classical elements amalgamated in Modern Chinese’, because frequency is one of the crucial factors for intelligibility. Even though intelligibility is important, prosody must also be taken into account before a formal style can be successfully formed. It has become increasingly clear, based on the study of prosodic syntax over the past 20 years, that something being intelligible does not necessarily mean that it is utterable; thus, the principle of shaping by prosody given in (7). The example given in (11b) shows that the monosyllabic verb zhi ‘know’ cannot stand alone and must be used within what is characterized as a ‘disyllabic template’ in order to be effable. There is mounting evidence that words and phrasal patterns are ineffable without a proper prosodic shape (Feng 2003a, 2003b, 2003c, 2005, 2006). For example:10 (13)

SPOKEN WRITTEN

(a) 一样 (a′) 同 yiyang tong ‘the same’ ‘the same’ (b) 不一样 (b′) 不同 bu yiyang bu tong ‘not (the) same’ ‘not (the) same’ (c) A跟B一样 (c′) A和B同 A gen B yiyang A he B *tong ‘A and B are the same’ ‘A and B are the same’ 657

Feng Shengli

(d) 一样的东西 (d′) yiyang -de dongxi same POSS thing ‘the same thing’

同的东西 *tong -de dongxi same POSS thing ‘the same thing’

(e) *四环以外准鸣笛。 *Si huan yiwai zhun mingdi fourth Ring outside permit honking ‘Outside the Fourth Ring (road) honking is permitted’ (f ) 四环以内不准鸣笛。 Si huan yinei bu zhun mingdi Fourth Ring inside not permit honking ‘Within the Fourth Ring (road) honking is not permitted’ (g) 待上级准假后才能离京。 dai shangji zhun jia hou cai neng li Jing. wait superiors approve vacation after only can leave Beijing ‘Only after his superiors approve his vacation can he leave Beijing’ (h) [口语] [Spoken]

来参观旅游的人(不)应该去。 lai canguan lüyou de ren (bu) yinggai qu Come visit [particle] people (not) should go ‘People who come to visit should (not) go [there]’

(i) [书面] [Written]

观光游客不宜*往 / *宜前往 / 不宜前往。 guanguang youke bu yi *wang / *yi qianwang / bu yi qianwang Visiting tourist not should go / should proceed / not should proceed. ‘Visiting tourists should (not) proceed (further)’

(14)



WRITTEN

SPOKEN

(a) 无法 + 学*(习) (a′) wu fa xue*(xi) no way study(-learn) ‘no way to study’

没办法学 mei banfa xue no way study ‘no way to study’

(b) 禁止 + 说*(话) (b′) jinzhi shuo*(hua) prohibit speak(-word) ‘talking prohibited’

不许说 bu xu shuo not allow speak ‘talking prohibited’

(c) 毫无 + 权*(力) (c′) hao wu quan*(li) little not-have rights(-power) ‘no right at all’

一点权都没有 yidian quan dou meiyou little rights even not have ‘no right at all’

(d) 共同 + 谈*(笑) (d′) gongtong tan*(xiao) together talk(-laugh) ‘talk together’

一块儿谈 yi-kuair tan together talk ‘talk together’

658

Modern Chinese: Written Chinese

Without being shaped by prosody, the examples given in (13c′, d′, e) and (14a–d) are all ungrammatical. Thus, the grammar of written Chinese is essentially a prosodic grammar as was seen in (5).

4.3.  Amalgamation principle I: understandability After selection and processing are complete, we are then faced with the next question: how can we transform formal expressions into spoken expressions? This is what has been called the ‘amalgamating problem’. 采用少数(古代)词语,要怎么样才是融会而不是搀杂。情况千变万化,很难具体 (15)  说明。勉强说,可以用耳朵作个尺子,量一量,凡是听起来生硬,明显觉得不像 日常说话(包括谈论学术问题)的,是搀杂而未融会,反之是已经融会而不是搀杂。 ── 张中行《文言津逮》 ‘When can adopting a few (classical) words into modern writing be considered as amalgamation rather than adulteration? The situation is extremely complex and difficult to pin down. I would have to say that we may use the ear as a measure. Anything that sounds odd to the ear or obviously not like everyday speech (including discussion of academic issues) will be considered adulteration, not amalgamation, and vice versa.’ ——Zhang Zhongxing, Gateway to Literary Chinese (2002) As noted above, it is not uncommon that one ‘should incorporate some literary expressions into their own writing’ (Zhang 2002). However, syncretization does not imply adulteration. Thus, we must clearly distinguish amalgamation (融会) from adulteration (搀杂). The new register of formal Chinese has been formed by a process of amalgamating classical expressions into the modern vernacular. It is not an adulteration, even though there may be isolated examples of it. Thus, the principles of the intelligibility (or frequency in everyday speech) and effability (or degree of amalgamation 融合度) proposed by Feng (2009) and Sun (2012) must be combined together in order to derive the resulting amalgam. (16) is a sample of amalgamation: 吝啬的人,我们说他小气;妒忌的人,我们也说他小气。小气,自然不够伟大; (16)  即使不是十足的小人,至少该说是具体而微的小人。但是,如果小气的人就算 是小人之一种,则小人满天下,而足称为君子者,实在太少了。 ── 取自王力《小气》 ‘Stingy men, we say “they are petty”; jealous men, we also say “they are petty”. Pettiness, naturally is not very great, even if they are not entirely petty men, we should at least say they are trivial men. However, if being a person who is petty is considered as one type of petty man, then petty men are everywhere, and those who can be considered as gentleman, they are truly few.’ ——Wang Li, On Pettiness (2003) The shaded portions are literary expressions from Classical Chinese and here they are amalgamated with spoken expressions. This passage can be read aloud and is completely intelligible and easily understandable. Even though there are many classical expressions in it, the way it is composed, artfully combining classical and spoken Chinese, make it a masterpiece of modern formal (or elevated) Chinese. 659

Feng Shengli

4.4.  Amalgamation principle II: modulate literary dictions with colloquial expressions It is clear by now that modern formal writing must amalgamate literary dictions with colloquial expressions. However, to what degree, and how can one modulate them? Years of research have indicated a tentative conclusion, pointing to a ratio of 2:3 between formal and informal expressions in written Chinese. (17) provides a typical example of this type: (17) 吝啬的人, 我们说他小气; 妒忌的人, 我们也说他小气。 小气, 1 2 3 4 5 自然不够伟大; 即使不是十足的小人, 至少该说是具体而微的小人。 6 7 8 但是如果小气的人就算是小人之一种, 9

则小人满天下, 10

而足称为君子者, 实在太少了。 11 12

(取自王力《小气》)

There are a total of 12 phrases in this passage, among which five contain literary dictions, giving a ratio of 5:7. In other words, we have about 42% formal and 58% informal expressions here. The formality ratio of 40/60 seems to be something like a golden mean for maximizing linguistic style in formal Chinese. Based on measurements of the stylistic ratio, we are able to characterize the process of formal writing in terms of a wave model (Figure 36.6). This Stylistic Wave Model can also be viewed as the alternation between literary dictions and colloquial expressions used in a ratio of 4:6 in formal Chinese. As a result of the theory and the methodology presented here, we are able to break through the barrier of the incalculability of styles in human languages. Of course, there are new questions which have arisen and new areas of research which have opened up which must be left for future studies.

5. Conclusion In this chapter, it is shown that written Chinese is in fact a new language independent of spoken Chinese which has only recently matured. Based on the theory of Prosodic Grammar, the features of formal written Chinese consist of (i) monosyllabic words used in disyllabic templates (嵌偶单音词, M. Huang 2008), (ii) disyllabic words used in disyllabic copulates (合偶双音词, Feng 2006) and (iii) formal patterns (Y. Wang 2009). Although these must

11 6 1 2 3 45

8

Formal

10

9

7

12

Figure 36.6  Stylistic wave (of formal Chinese)

660

Informal

Modern Chinese: Written Chinese

be used following strict principles, such as Stylistic Coherence, Auditory Intelligibility (PAI) and Shaping by Prosody (PSP), they must also be mixed with some colloquial features in order to make the language natural (Sun 2012). As a result, an amalgamation principle which modulates literary dictions with colloquial expressions is proposed here. Furthermore, based on these principles, a method of calculation has been developed for the quantitative analysis of the degree of formality of a specimen of Chinese writing. The study of written grammar in Chinese has significant empirical and theoretical implications. An interesting avenue for future research will be to see how and why new language develops and how the diglossic requirement is forced upon and satisfied by language function. Empirically, this study indicates that the statistical measurement methods which have recently been developed into a new technology11 may perhaps be extended to a wide range of related fields, such as degree of formality measurement, composition testing, readability scaling, style gradation, textbook compilation, L2 learning, literacy acquisition, etc. All of these are important areas for future research.

Notes   1 Part of this chapter was published in Journal of Chinese Linguistics 37(1): 145–62 (2013).   2 Here the term ‘Classical Chinese’ refers to expressions that are commonly found in pre-Modern Chinese writings but are not used in the everyday speech of today.   3 This hypothesis needs further research to explore the mechanism and extent of the assimilation.  4 The classical forms that have crept into Modern Chinese are not necessarily used in colloquial expressions. This is discussed in section 3.   5 See M. Huang (2008) for a detailed discussion of how classical monosyllabic words like xiao 校 ‘school’, zhu 住 ‘shut’ and kun 困 ‘stop’, etc. are used in disyllabic templates.   6 Sun (2012) has pointed out further that it is due to (i) some words only existing in Classical Chinese, (ii) certain styles only being provided by Classical Chinese, such as brevity and elegancy, and (iii) certain grammatical constituency being lacking in Modern Chinese. As a result, classical expressions are unavoidable in modern formal or elevated registers.   7 One could argue that fangfu is more ‘literary’ than sihu, as opposed to more ‘formal’. The question of clearly distinguishing literary from formal is a topic for future research.   8 How to decide whether one expression is more formal than another is a problem that requires more study. Here, I rely on both a native speaker’s intuition and the frequency of occurrences inside relevant syntactic environments within our database.   9 This principle was developed according to Zhang’s (2002) ear-measurement strategy as explained in section 4.3. 10 From Feng 2003b. 11 See Feng, Wang and Huang (2008) for the Automatic Feature Checking Algorithm which has been verified using nearly 4,000 compositions from the HSK (Hanyu Shuiping Kaoshi, or Chinese Proficiency Test), resulting in a precise correspondence between the degree of formality as calculated by the algorithm and the scores on the HSK.

Bibliography Dong, Xiufang 董秀芳 (2004)《汉语的词库与词法》(Chinese Lexicon and Morphology), Beijing: Peking University Press. Downing, Laura (2006) Canonical Forms in Prosodic Morphology, Oxford: Oxford University Press. Duanmu, San (1990) ‘A Formal Study of Syllable, Tone, Stress and Domain in Chinese Languages’, Doctoral Dissertation, MIT. Duanmu, San (2011) ‘Word-length Preferences in Chinese: A Corpus Study’, Paper Presented at IACL19, Tianjin: Nankai University. Fei, Xiaotong 费孝通 (1999)《费孝通文集》(Collected Works of Fei Xiaotong), Beijing: Qunyan Press. Feng, Shengli 冯胜利 (1995) Prosodic Structure and Prosodically Constrained Syntax in Chinese. Doctoral Dissertation, University of Pennsylvania.

661

Feng Shengli Feng, Shengli (1996)〈论汉语的韵律词〉(‘On Prosodic Words in Chinese’),《中国社会科学》(Zhongguo Shehui Kexue) 1: 161–76. Feng, Shengli (2000)《汉语韵律句法学》(Prosodic Syntax of Chinese), Shanghai: Shanghai Educational Press. Feng, Shengli (2003a)〈韵律制约的书面语与听說为主的教学法〉(‘Prosodically Constrained Written Language and the Collective Approach to Teaching Chinese’),《世界汉语教学》(Chinese Teaching in the World ) 1: 87–97. Feng, Shengli (2003b)〈书面语语法与教学的相对独立性〉(‘The Independence of Written Chinese Grammar in Second Language Teaching’),《语言教学与研究》(Language Teaching and Linguistic Studies) 2: 53–63. Feng, Shengli (2003c) ‘Prosodically Constrained Postverbal PPs in Mandarin Chinese’, Linguistics 41(6): 1,085–122. Feng, Shengli (2005)〈论汉语书面语语法的形成与模式〉(‘On the Formation and Modulation of the Grammar of Modern Written Chinese’),《汉语教学学刊》(Journal of Chinese Language Studies) 1: 17–49. Feng, Shengli (2006)《汉语书面用语初编》(Expressions of Written Chinese), Beijing: Beijing Language and Culture University Press. Feng, Shengli (2009) ‘On Modern Written Chinese’, Journal of Chinese Linguistics 37(1): 145–61. Feng, Shengli (2010a)〈论语体的机制及其语法属性〉(‘On mechanisms of Register System and its Grammatical Property’),《中国语文》(Studies of the Chinese Language) 5: 400–12. Feng, Shengli (2010b) ‘A Theoretical Exploration of Prosodic Syntax in Chinese’,《语言学论丛》 (Linguistic Forum) 39: 201–41. Feng, Shengli (2011)〈语体语法及其文学功能〉(‘The Literary Function of Stylistic-register Grammar’), Rhetoric) 4: 1–13. 《当代修辞学》(Contemporary   Feng, Shengli, Wang Jie 王洁, and Huang Mei 黄梅 (2008)〈现代汉语书面语典雅度的自动测量 研究〉(‘Computerization of Formality Testing for Written Chinese’),《语言科学》(Language Science) 2: 113–26. Hu, Mingyang 胡明扬 (1993)〈语体和语法〉(‘Style and Grammar’),《汉语学习》(Chinese Studies) 2: 3–6. Huang, Kan 黄侃 (1922/2001)《黄侃日记》(Huang Kan’s Diary), Nanjing: Jiangsu Educational Press. Huang, Mei 黄梅 (2008)〈汉语嵌偶词的句法分析及其理论意义〉(‘Syntactic Analysis and Theoretical Implications of Classical Monosyllabic Words Used in Disyllabic Templates in Modern Chinese’), Doctoral Dissertation, Beijing Language and Culture University. Link, Perry (2009) ‘The Secret History of Classical Rhythm in Modern Chinese’, in Perry Link (ed) The Scholar’s Mind: Essays in Honor of Federick W. Mote, Hong Kong: The Chinese University Press, 253–81. Lu, Bingfu and San Duanmu (2002) ‘Rhythm and Syntax in Chinese: A Case Study’, Journal of Chinese Language Teachers Association 37(2): 123–35. Lü, Shuxiang 吕叔湘 (1962)〈说‘自由’与‘粘着’〉(‘On Free and Bound’),《中国语文》(Studies of the Chinese Language) 1: 1–6. Lü, Shuxiang (1963)〈现代汉语单双音节初探〉(‘A Preliminary Investigation on Mono-/Disyllabic Issues in Chinese’),《中国语文》(Studies of the Chinese Language) 1: 10–23. McCarthy, John and Alan Prince (1993) ‘Prosodic Morphology I – Constraint Interaction and Satisfaction’, Unpublished Manuscript, University of Massachusetts and Rutgers University. McCarthy, John and Alan Prince (1998) ‘Prosodic Morphology’, in Andrew Spencer and Arnold M. Zwicky (eds) The Handbook of Morphology, Oxford and Malden, MA: Blackwell. Pan Wenguo, Ye Buqing, and Han Yang (2004)《汉语的构词法研究》(Study on Chinese Morphology), Shanghai: East China Normal University Press. Shi, Dingxu 石定栩 (2006)《港式中文两面睇》(Two Aspects of Hong Kong Chinese), Hong Kong: Xingdao Chuban Youxian Gongsi. Sun, Dejin (2012)《现代书面汉语中的文言语法成分研究》(Research on Literary Grammar in Modern Written Chinese), Beijing: The Commercial Press. Tsou, Benjamin 鄒嘉彥 and Yue Hong, Mok 莫宇航 (2013)〈陆港台书面语近百年演变比较—以 语料库为依据 (‘A comparison of written language changes in the past 100 years between Mainland, Taiwan and Hong Kong – A corpus-based approach’), in Feng Shengli 冯胜利 (ed)《汉语书面语 的历史与现状》(Written Chinese: The Present and the Past). Beijing: Peking University Press. 58–75.

662

Modern Chinese: Written Chinese Wang, Hongjun 王洪君 (2008)《汉语非线性音系学》(Non-linear Phonology of Chinese), Beijing: Peking University Press. Wang, Li 王力 (1993)《龙虫并雕斋琐语》(Essays from the Longchong Bingdiao Studio), Beijing: Chinese Social Sciences Press. Wang, Yongna (2009)〈汉语书面正式语体的语法手段〉(‘Study on Grammar Means of Formal Written Chinese’), Doctoral Dissertation, Beijing Language and Culture University. Zhang, Zhongxing 张中行 (1987)《文言与白话》(Literary Chinese and Vernacular Chinese), Harbin: Heilongjiang People’s Press. Zhang, Zhongxing 张中行 (2002)《文言津逮》(Gateway to Literary Chinese), Beijing: Beijing Press. Zhu, Dexi 朱德熙 (1982)《语法讲义》(Lectures on Grammar), Beijing: Beijing Commercial Press. Zhu, Dexi (1987)〈现代汉语语法研究的对象是什么?〉(‘The Subject of Modern Chinese Grammatical Studies’),《中国语文》(Studies of the Chinese Language) 5: 321–9.

663

37 Poetic prosody1 Feng Shengli The Chinese University of Hong Kong, Hong Kong, China

This chapter deals with poetic prosody in Classical Chinese. It is shown that the Archaic Chinese prosody had undergone a change from moraic foot structure (a foot consisted of at least two moras) to syllabic foot structure (a foot consisted of at least two syllables). After the establishment of a disyllabic foot system, the poetic prosody interacting with the morphosyllabic property of the language gives rise to four-, three-, five- and seven-syllable lines within the poetic system of Classical Chinese.

1. Introduction In Chinese, the Shijing《诗经》(Book of Odes) is considered the first collection of poems in history (most quoted in the Eastern Zhou period 770–256 bc). It is well known that while a fixed dipodic prosody (二步律) was used in the Shijing, a caesura-based prosody (顿叹律) was developed in the Lisao《离骚》(attr. Qu Yuan, c. 340–278 bc) during and after the Warring States period (475–221 bc). Although there are five-, six-, seven-, eight- and even nine-syllable lines in the Shijing, its basic rhythmic structure is formed by tetrasyllabic lines (1), and thus the metrical structure of the Shijing is dipodic prosody, which is different from the metrical structure of the Lisao. The Lisao has broken through the dipodic prosody of the Shijing, as shown by (2). (‘X’ represents ‘syllable’, ‘兮=*ha’ is an exclamation particle.) (1)

关雎 关关雎鸠 在河之洲 窈窕淑女 君子好逑

guan guan ju jiu, zai he zhi zhou. yao tiao shu nü, jun zi hao qiu.

Kuan-kuan, the ospreys On the river’s isle Delicate, a good girl: A gentlemen’s fit mate.

朕皇考曰伯庸。 《离骚》Lisao (2) 帝高阳之苗裔兮, X X X X X X兮, XXXXXX Of the god-king Gao Yang, I am the far offspring, my late honored sire bore the name of Boyong. 664

Poetic prosody



纷吾既有此內美兮, 又重之以修能。 《离骚》Lisao X X X X X X X兮, XXXXXX Such bounty I had of beauty within, this was doubled with fair appearance.

灵氛既告余以吉占兮, 厉吉日乎吾将行。《离骚》Lisao X X X X X X X X兮, X X X X X X X Since Holy Fen told me my lot of good fortune, I choose a day of good fortune, and I will set out. 苟余情其信讠夸以练要兮, 长咸页颔亦何伤。 《离骚》Lisao X X X X X X X X X兮, X X X X X X If my nature be truly comely, washed utterly pure, what hurt can I have in long wanness from hunger?2 Clearly, the syntactic as well as the prosodic positions of *ha 兮 can be located between any two syntactic constituents (i.e. NP, PP, VP, AP, etc.) with different numbers of syllables. This implies that 兮 is a phrasal exclamation particle which gives rise to a caesural prosody. By the Han dynasty (206 bc – ad 220) however, flexible caesura prosody was gradually replaced by syllabically fixed patterns: first trisyllabic, then pentasyllabic and finally heptasyllabic. A relevant background of this development is given in (3). (3) (a) Trisyllabic poems can be found only after the Western Han, developing in folk songs (‘颍水清,灌氏宁,颍水浊,灌氏族 When the Ying river is clearing, the Guan family is well-off, when the Ying river is turbid, the Guan family is executed. ) and used in imperial ritual songs, i.e. Jiaosiyue《郊祀乐》; 《史记•灌夫传》   (b) Tetrasyllabic verse is still active, though not as dominant as it was in pre-Qin times (Cui 2006); (c) Pentasyllabic poems did not appear until the Eastern Han and later played a central role in the poetic literature of the Six Dynasties (220–589, Zhong Rong 钟嵘: ‘Pentasyllabic lines are essential in literary composition 五言居文词之要’.); (d) Hexasyllabic poems have never been popular in Classical Chinese literature. It is true that in the Western Han, six-syllable lines were used increasingly in fu (赋) and later in pianwen (骈文). In the Wei–Jin period (220–420), hexasyllabic poems were developed by Kong Rong 孔融, Cao Pi 曹丕 and Cao Zhi 曹植, and in the Song dynasty, there existed some hexasyllabic poems. (e) Hexasyllabic prosody never functioned as a standard poetic formula in Chinese poetic history.3 Heptasyllabic verse appeared as early as pentasyllabic verse (in the Western Han 206 bc – ad 23). However, as Chu (1990: 125) has pointed out, the flourishing of five-syllable poems began earlier than seven-syllable poems even though both started around the same time. In the Han–Wei period, five-syllable poems had already become ‘the essence of literary composition 居文词之要(钟嵘 《诗品序》   ), but seven-syllable poems did not become popular until the Tang dynasty’. All these facts raise many interesting and important questions. What are the crucial factors that determine the choice among all the possibilities for later developments of ancient poetic forms? Second, although trisyllabic prosody developed quite early (during the Western Han), why did it not evolve into an independent poetic form?4 Finally, why did pentasyllabic and heptasyllabic forms develop differently, even though both appeared at roughly the same time 665

Feng Shengli

during the Western Han? In what follows I will see how the poetic prosody influences or even determines the different types of classical poetry.

2.  Poetic prosody in Chinese First, let us look at some fundamental concepts of poetic prosody in general and see what poetic effect is. Following Sperber and Wilson (1995), the notion of poetic effect is defined as: (4) Poetic effect (Sperber and Wilson 1995: 222) Poetic effect is a peculiar effect of an utterance which achieves most of its relevance through a wide array of weak implicatures. The following utterance, however, shows that weak implicatures cannot form a poem in Chinese: (5) 院子里的树叶已经巴掌一样大了,爸爸什么时候回来呢？(Z. Zhu 1998: 198) The tree leaves in the courtyard are already as big as my palm. When will father return? As pointed out by Z. Zhu (1998: 198), there is a strong sense of poetic implicature in the above utterance. However, it cannot be considered a poetic line because it lacks proper poetic prosody.5 Since poems must be formed with poetic prosody, the poetic effect must be defined anew as follows: (6) Poetic effect (Revised, Feng 2006) Poetic effect is a peculiar effect of an utterance which achieves most of its relevance through (i) a wide array of weak implicatures and (ii) a proper prosodic structure. Poetic implicature and poetic prosody together give rise to what is called poetic effect in literature. Now we must see what poetic prosody is accordingly. As demonstrated below, the poetic prosody in Chinese is derived from several prosodic and poetic principles. The first and most important one is the Principle of Poetic Rhythm: (7) Principle of Poetic Rhythm The prosody of poems is organized by rhythm, which in Chinese is realized through both pause and rhyme.6 Given this principle, we are able to derive default structures of poetic prosody for Chinese based on metrical phonology in modern linguistics. First, rhythm is a structure with at least two elements (Liberman 1975; Liberman and Prince 1977). This is because there can be no rhythm in terms of stress, length, pause, etc. with only one element. Thus, we have the first rule of Rhythmic Binarity, first proposed by Liberman (1975). (8) Rule I: Rhythmic Binarity Rhythm must be licensed by at least two rhythmic units. This is because stress or pause or whatever form a prosodic unit takes cannot produce a rhythm without at least two elements. Thus a binary structure is a minimal requirement for rhythm (below ‘R’ stands for the notion of ‘rhythm’; ‘*’ denotes ‘non-existent’ or ‘ungrammatical’; ‘X’ and ‘Y’ stand for any minimal prosodic unit). 666

Poetic prosody

  (9) (a)  (b) *R R   X Y X   What is a rhythmic unit then? Given the theory of prosodic morphology (McCarthy 1993) and prosodic syntax in Chinese (Chen 2000; Duanmu 2000; Feng 2000, 2005), rhythmic units can be defined as: (10) Rule II: Rhythmic Unit A rhythmic unit is formed by a prosodic unit which consists of a Foot (PrWd, Prosodic Word), a Super-Foot (Super-PrWd) or a Compound Foot (Compound PrWd). An elementary prosodic unit is defined in terms of a foot in metrical phonology. A standard foot in Chinese is formed by two syllables while a super-foot is formed by three. Thus, a standard PrWd contains two syllables while a super-PrWd contains three (Feng 1995; Chen 2000). (11)

PrWd

Super-PrWd

f

F

σ σ 下 天 tian xia sky under the world

σ σ σ 头 蛇 两 liang tou she two head snake two-headed snake

Compound-PrWd f

f

σ σ σ σ 箭 雕 一 双 jian shuang diao yi one arrow two bird kill two birds with one stone

The notion of a metrical foot is defined in terms of a syllable or mora in metrical phonology: (12) Foot Binarity Principle A foot must be formed by at least two syllables (as in Chinese) or at least two moras (as in Japanese). Given the above rules and principles, it is then necessary to rewrite the non-poetic utterance above in (5) into poetic lines as follows: (13)

Yuan zhong / shu ye / jin you da 园 中 树 叶 今 又 大 The tree leaves in the courtyard are now getting bigger again,



tian ya / jia fu / he shi gui 天 涯 家 父 何 时 归 When will father return from his long journey?

The necessity of reforming (5) as (13) is due to the fact that Chinese utilizes a syllabic foot structure. Unlike the syllabically irregular utterance in (5), the two lines in (13) are syllabically footed rhythmic structures. However, these two parallel seven-syllable lines have 667

Feng Shengli

no rhyme. Of course, there are unrhymed poems in many languages, but nearly all Chinese poems rhyme. Hence, the perfectly footed parallel lines in (13) must be rewritten to observe the rhyming requirement of Chinese. Rewriting, the result is similar to: (14)

Yuan zhong / lü ye / jin you chang 园 中 绿 叶 今 又 长 The green leaves in the courtyard are now growing bigger again,



he shi / a ba / wei huan xiang 何 事 阿 爸 未 还 乡？ What stops father from returning home?

In addition to the footing rhythm in (13), we must further employ rhyme to create a poem in traditional Chinese, as in (14). I will not discuss the reasons why Chinese poetic lines must rhyme here, but will simply take this for granted and propose a rule to capture the fact that poems in Chinese must rhyme: (15) Rule III: Rhythmic Melody Rhythm must be licensed by rhyme, which is realized between at least two lines, that is:  [LR1 # LR2] where ‘L’ denotes ‘line’, ‘R’ is ‘rhyme (at the end of a line)’, and ‘#’ is ‘break between two lines’. Thus LR1 and LR2 form a minimal melodic unit, i.e. a couplet in poetic terms. Poetry makes use of the musical properties of language, and the intrinsic property of melody is repetition. Thus, there are no one-line poems in classical literature because such poems would lack melodic effect.7 The melodic property of poetry is captured in the rule of Rhythmic Melody in (15). Based on the principles outlined above, some general conditions that govern the process of poetry-making in traditional Chinese have been proposed in the literature (Feng 2011), namely the minimality and optimality conditions: (16) (a) Minimality Condition (of Chinese poetic structure) One syllable cannot form a foot, one foot cannot form a poetic line and one line cannot form a poem (单音不成步、单步不成行、单行不成诗). Thus,        

a a a a

minimal minimal minimal minimal

foot = two syllables line = two feet (or two prosodic units) melodic unit = two lines (a couplet) poem = two melodic units (a stanza, quatrain (绝句))

(b) Optimality Condition (of Chinese poetic structure) According to the system of Chinese poetic structure, minimal is optimal, and thus the optimal poetic structure is: ﴾ { [ (σ x 2 ) x 2 ] x 2 } x 2 ﴿   where ‘( )’ stands for foot, ‘[ ]’ for poetic line, ‘{ }’ for minimal structure of a poem, i.e. a couplet, and ‘﴾ ﴿’ denotes a stanza, a minimal combination of couplets. 668

Poetic prosody

The implication of the Optimality Condition is that the poetic effect in Chinese is both minimally and maximally realized within two lines as a melodic unit with each line consisting of two prosodic units. As a result, two syllables make up one poetic unit, two poetic units make up one poetic line, two poetic lines form a melodic unit (or couplet, 诗联), two couplets form a minimal poem (stanza, quatrain, 绝句) and two minimal poems make a compound or regularized poem (律诗), which is the basic structure for the so-called new style poem (近体诗) in the later development of Chinese poetry.8 Given the theory outlined above, most of the known facts about regulated poetry in the history of Chinese literature can be explained systematically, as will be seen below.

3.  Monosyllable and disyllable prosody The theory of poetic prosody developed above will generate not only the basic rules of Chinese poetry, but also predict some of the historical developments of poetic changes. Here I will outline briefly the development of Classical Chinese poetry in terms of different lengths of syllabic lines, from poems of monosyllabic lines to those of heptasyllabic lines. Let us start with monosyllabic lines. According to the Foot Binarity Principle (12), if a foot is formed minimally by two syllables, then one syllable cannot stand alone (cannot form a standard foot).9 This implies that there will be no disyllabic line poems in Classical Chinese, which has been borne out (see Feng 2008). Of course, an independent phrase formed by a monosyllabic word must use a zero syllable (pause) to license a foot. This is also borne out, as seen in Zuozhuan《左传•僖公 二十八年》: (17) Jin jun qibai sheng, xian__, yin__, yang__, ban__ 晋车七百乘， [革显 __]、[靷 __]、[鞅 __]、[靽 __]。 The Jin army has seven hundred military carriages, belly-harnesses, chest-harnesses, bridles, heel-ropes —— fully equipped As a result, a one-syllable foot (‘xian__, yin__, yang__, ban__’) must occur in a parallel environment, making monosyllabic lines in Chinese extremely context-dependent. Two syllables will form a standard foot and are thus perfectly balanced and function as a basic and formal unit in poetic language. However, they cannot serve a rhythmic purpose for poetry due to the violation of Rule I in (8), and thus cannot form a poetic line.10 If the Syllabic Foot structure is observed in Chinese (after the Spring and Autumn period, 722–481 bc), there will be no disyllabic Chinese poetry, which is most true as predicted by the theory and confirmed by the fact.

669

Feng Shengli

4.  Three- and four-syllable prosody A trisyllabic poem can be analyzed using the following poetic structure: (18)

Rhyme Melody Couplet →

Line

Unit f 六 liu Six

Line Unit f

王 wang kings

毕 bi, over,

[_]

Unit f 四 si Four

海 hai seas

Unit f 一 yi united

[_]

As seen in (18), the rhyming melody is built upon the minimal repetition of two lines, each formed minimally by a repetition of two feet: one is formed in a standard way by two syllables, the other is formed in a non-standard way by a syllable with a pause. Given this structure, we are able to explain why trisyllabic lines were not an orthodox poetic form. First, it is well known that trisyllabic lines are heavily constrained by the prosodic system and tend to cause ungrammatical results.11 Second, there are syntactic constraints on trisyllabic forms as well. For example, the structures subject+predicate and verb+object are very commonly used in literature. (19) Subj+Pred:

云从龙、风从虎。 《易》 Cloud follows dragon, wind follows tiger. 岁将暮、时既昏、寒风集、愁云繁。〈雪赋〉 The year was coming to an end, the day was already dark, Cold winds blew hard, sad clouds thickly gathered.

Verb+Obj:

置 旨酒, 命 宾友; 召 邹生, 延 枚叟。 〈雪赋〉 He set out fine wine, and called over guests and friends. He summoned scholar Zou, invited Elder Mei.12

However, there are hardly any trisyllabic forms which are formed by a combination of a disyllabic adjective and a monosyllabic adjective, i.e. [AA+A] or [A+AA]. For example: (20) 散漫交错、纷氲萧索。 〈雪赋〉 Scattering and spreading, mingling and merging. It swells up, then thins out.13 *散交错、*纷萧索 *散漫交、*纷氲萧

670

Poetic prosody

Also, there were hardly any disyllabic and monosyllabic noun combinations, i.e. [NN＋N] or [N+NN] before the Eastern Han dynasty (25–220) such as: (21) 西郊有上囿禁苑、林麓薮泽。〈西都赋〉 In the western suburbs there are imperial enclosures and the forbidden park. There are woods and forests, meres and marshes,  .  .  .14 *上囿苑、*林麓薮 *囿禁苑、*林薮泽 As a result, the prosodic and syntactic constraints together force trisyllabic combinations to occur favorably in VP, but not in AP and NP structures. Another characteristic of a trisyllabic unit is that it is bigger than a standard prosodic unit which has only two syllables, i.e. [σσ]foot. On the other hand, it is smaller than a dualrhythmic-unit, i.e. the tetrasyllabic template (四字格). In order to make a three-syllable line a dual-rhythmic unit with two feet, the single stray syllable within the trisyllabic unit must take a pause as a placeholder,15 as in (22). f

(22) (a)  σ σ 兔 狡 cunning hare 狗 走 running bound

f (b)  # #

σ 死 dies 烹 boiled

σ 左 left 右 right

#

# # #

σ σ 黄 牵 leashes a dog 苍 擎 holds a falcon

Given this, trisyllabic expressions are strictly limited to a couplet environment only by means of which a placeholder (zero) can be realized: (23) 

f

f

f

f

大 冯 君 0 # 小 冯 君 0， Big Feng Mister Little Feng Mister, 兄弟继踵相因循。《汉书》 Two brothers follow each other side by side. Obviously, trisyllabic forms have some unique prosodic properties.16 First, as pointed out in Feng (2006), the unbalanced prosody [2+1] or [1+2] characterizes informal or even jocose features in casual speech (i.e., a colloquial style in the language). For example, bei heiguo 背黑锅 ‘to take the rap (for someone)’, dai gaomao 戴高帽 ‘to brown-nose’, sa yazi 撒鸭子 ‘take to one’s heels’, liao juezi 尥蹶子 ‘to kick back’, shua dapai 耍大牌 ‘to be a diva’, etc. all function like limerick expressions in Chinese.17 Secondly, trisyllabic lines in literature also exhibit special poetic effects when used in dramatic or action events.18 It would not be an overgeneralization to characterize the literary usage of trisyllabic forms as rhetorical expression. This rhetorical function of trisyllabic expressions is illustrated by the following examples:

671

Feng Shengli

(24) Event Opening (with a dramatic effect):19 Xie Huilian 谢惠连 Snow Fu〈雪赋〉 :   岁将暮, The year approaching its close,   时既昏, The season getting dark,   寒风急, Cold winds gathering,   愁云繁。 Gloomy clouds clustering.20 Du Mu 杜牧 E-Pang Palace Fu〈阿房宮赋〉:   六王毕, The six states are no more,   四海一, The four seas are united,   蜀山兀, The Shu mountains have been shorn of all trees,   阿房出。 And E-pang Palace stands before us.21

Event Processing (with a kinetic effect):22 Sima Xiangru 司马相如 Rhapsody on Sir Vacuous〈子虚赋〉:

  浮文意鸟,扬旌拽。 Drifting on figured heron prows, they hoist banner poles.   张翠帷,建羽盖。 Spread halcyon-plume curtains, raise feathered sunshades.   网瑇瑁,钩紫贝。 They net hawksbill turtles, angle for purple cowries. 扌从金鼓,吹鸣籁。 They strike the metal bell-gongs, blow the singing panpipes.     榜人歌,声流喝。 The chief oarsman sings, his voice fluid then sobbing.   水虫骇,波鸿沸。 The water creatures are startled, waves grandly surge.   湧泉起,奔扬会。 While bubbling fountains spurt, and the dashing swells converge.   蠝石相击,硠硠磕磕。 Giant boulders scrape together, grinding and grating, clashing and clattering.23 Event Closing (with an aphorist effect):24 Yang Xiong 杨雄 Plume Hunt Rhapsody〈羽猎赋〉:   因回轸还衡, Thus he wheels round his carriage,   背阿房, Turns his back on E-Pang,   反未央。 Returns to the Everlasting Palace.25 Given all these three-syllable effects, it has been shown (Feng 2006) that trisyllabic patterns will not characteristically be used in formal or solemn situations such as in a eulogy or monody.26 The prosodic restrictions and the rhetorical functions of trisyllabic forms prevent them from being used as prevalently as other syllabic patterns in poetry. However, they can be used for what Liu Xie described as the function of a prosodic-rhetoric adjustor (应机之权节) in his The Literary Mind and the Carving of Dragons《文心雕龙》. In short, it is their prosody that prevents the trisyllabic patterns from being accepted as erudite poetic forms in Chinese. I will not go into details here on tetrasyllabic prosody and poems because it is clear that the [1+1][1+1] rhythm is a perfectly balanced poetic structure. It is the most popular and powerful poetic pattern in Chinese literature as expected by the theory of poetic prosody27 and demonstrated by the corpus of extant Chinese literature.

5.  Five- (and six-)syllable prosody According to the principles that ‘one foot cannot make a poetic line’ and ‘one line cannot make a poem’, pentasyllabic lines should be favored by poets as much as trisyllabic lines 672

Poetic prosody

because they satisfy the Rhythmic Binarity and Rhythmic Melody requirements as shown in the following structure: (25)

Rhyme Melody →

Couplet

f 青

L

L f

F

# 河 畔 青 Green beyond green, the grass along the river.

草

郁

F

# 园 中 郁 Leaves on leaves, willows in the garden

柳

However, the [2+1] in the [2+[2+1]] line cannot be analyzed as a single unit if the [2+1] trisyllabic element is analyzed as [[1+1]foot + [1+0]foot], where [1+0] forms an independent foot. In fact, trisyllabic lines can usually be analyzed as two feet in the early stages of trisyllabic development. Take the following: (26) 江有汜,之子归,不我以! 不我以,其后也悔。 The Kiang has branches breaking out and reverting; this young lady went to her new home, but she would not take us; she would not take us, but afterwords she had to repent. 江有渚,之子归,不我与! 不我与,其后也处。 The Kiang has islands; this young lady went to her new home, but she would not associate with us; she would not associate with us, but afterwords she had to live with (us). 江有沱,之子归,不我过! 不我过,其后也歌。 The Kiang has the (tributary) T’o; this young lady went to her new home, but she would not pass us on; she would not pass us on, but (now) crooningly she sings.28 ——《诗经•江有汜》 In the above poem, the last two characters of the first line (也悔) show that [[1+1]+[1+0]] parallels [[1+1]+[1+1]] with or without a phonologically realized placeholder (也 is a common word inserted into a line of verse for balance or euphony, a placeholder character, 衬字). According to this, pentasyllabic lines of the form [[2+[2+1]] may be analyzed as [[1+1]+ [1+1]+[1+0]], which is perfectly allowable in a mixed line of a poem, as in: (27) South of the Walls We Fought〈战城南〉 战城南,死郭北, South of the walls we fought, north of the ramparts we fell, 野死不葬鸟可食。 fell in the meadows, left unburied, food for the ravens. 为我谓鸟: Speak to the ravens for us, say: ‘且为客豪, we were brave men, far from home, 野死谅不葬, we fell in the meadows, left unburied, 腐肉安能去子逃？’ how can our carrion flee you? 673

Feng Shengli



Where the waters run deep and clear, the reeds and the rushes are dark: 枭骑战斗死,驽马徘徊鸣。 the horsemen all died in battle, their tired mounts linger and neigh. ‘築室,何以南何北, ‘On the bridge a guardhouse is built – how can we cross south? How can we cross north? 禾黍不获君何食？ If the grain is not taken in harvest, how shall our lord eat? 愿为忠臣安可得？’ We want to be loyal liegemen, but how can this be done?’ ‘思子良臣, ‘We think on you, good liegemen, 良臣诚可思, good liegemen should be in our thoughts, 朝行出攻,暮不夜归。 at dawn you went forth to battle, and at evening did not return’.29 水深激激,蒲苇冥冥:

This mixed-length-line poem is written in an irregular meter and could possibly be chanted as: (28)

水深 蒲苇 枭骑 驽马 築室

/ / / / /

激激, Where the waters run deep and clear, the reeds and the rushes are dark. 冥冥。 战斗 / 死__, The horsemen all died in battle, 徘徊 / 鸣__。 their tired mounts linger and neigh. 何以南 / 何北  .  .  . On the bridge a guardhouse is built – how can we cross south? How can we cross north?

Obviously, this type of metrical structure is incapable of motivating a regular prosody of pentasyllabic lines in later ages. This is because if [σσσ] is analyzed as [(σ σ) (σ__ )], the resulting [(σσ)(σσ)(σ_)] (a three-foot structure) will be a violation of the Minimality and Optimality Conditions in (16). One way to meet the poetic requirements of a maximum of two poetic units per line is to force the trisyllabic form to behave like one foot (a super-foot) in poetry as in the following: (29)

子为王,母为虜。 Zi wei wang, mu wei lu 终日 / 舂薄暮, Zhong ri chong bo mu 常与 / 死为伍。 chang yu si wei wu 相离 / 三千里, Xiang li sanqian li 当谁 / 使告汝。 dang shui shi gao ru

Son is king, his mother is prisoner. All day she pestles till night, With death she lives. Three thousand miles away. Should ask whom to tell you? ——Hanshu《汉书•吕后纪》

However, trisyllabic feet required time to evolve from two feet (or strictly speaking, one and a half feet) into a single foot.30 Given this, it follows that pentasyllabic poetry could not fully develop until trisyllabic lines became a single prosodic unit (a super-foot). This analysis is supported by the following phenomena. First, there were no pentasyllabic poems before the Western Han, nor were there trisyllabic compound word formations. Second, only in the Eastern Han were trisyllabic compounds produced freely and only then did pentasyllabic poems flourish. The following trisyllabic examples gathered from throughout the Eastern Han Lunheng 《论衡》show   the correlation between trisyllabic compounds and pentasyllabic poems (from Han Music Bureau, Long Song 汉乐府《长歌行》): 674

Poetic prosody

(30) (a)  养性书、封禅书、甘泉颂、功曹史、军下卒、马下卒、偃月钩、丧家狗、 茧栗牛、两头蛇、桃象人、魍魉鬼、日月道、日廷图、四坎坛、太阳气、 无妄气、五行气、阴阳气、东南方、西北方、图宅术、五音术、博士宫、 明光宫、建章宫、济阳宫、梁山宫、万岁宫、都尉府、甘泉殿、谷城山、 霍太山、劳盛山、劳成山、度朔山、平原津、上虞江、皖侯国、会稽郡、 永昌郡、钱塘县、昆阳城、泉陵城 (b) 青青 朝露 阳春 万物 常恐 焜黃 百川 何时 少壮 老大

园中葵, Qingqing yuanzhong kui Green, green, garden sunflowers, 待日晞。 zhaolu dai ri xi morning dew waits for brightness. The spring sun spreads its goodness, 布德泽, Yangchun bu de ze 生光辉。 wanwu sheng guanghui Brightly, all things are shining. Afraid of autumn’s coming, 秋节至, Chang kong qiujie zhi 华叶衰。 kunhuang huaye shuai yellow and weak are the leaves. 东到海, Baichuan dong dao hai To the sea, a hundred rivers flow east, 复西归？ heshi fu xigui when will they return to the west? Young man, if you don’t work hard, 不努力, Shao zhuang bu nuli 徒伤悲。 laoda tu shangbei. You will sigh when you are old.31 ——Han Yuefu, Long Song 汉乐府《长歌行》 

The correlation between trisyllabic compounds and pentasyllabic poems indicates that pentasyllabic poetry was formed when trisyllabic forms became a single foot.32 I will not go into a detailed discussion here about hexasyllabic forms, but will simply point out that they are made neither by a basic rhythmic structure (because it would be 3+3, and 3 is not a basic foot structure) nor by an optimal structure (because it would be 2+2+2 which is not minimally composed of two units). As a result, hexasyllabicity is not an ideal poetic pattern according to the Optimality Condition for rhythm as given in (16b). This can be seen from the fact that a hexasyllabic line can only be balanced if it is formed by a structure of either [[1+2]+[1+2]]or [[2+1]+[2+1]]. However, a metrical pattern of [[2]+[2]+[2]] is also possible, though it disrupts the balanced norm. Given these poetically undesirable conditions, the prosodic characteristics and poetic functions of hexasyllabic forms turned out to be only suitable for parallel prose (骈文), as pointed out by Feng (2011) and developed in Lu (2014) and Feng (2015). That is to say, their special rhetorical properties are best employed in prose (as prose prosody), not in poetry (as poetic prosody).

6.  Seven-syllable prosody The most difficult question involving the development of heptasyllabic poems is why they developed so late. There have been attempts to account for their slower development, however most of the explanations are focused on external reasons, not the internal system of the language that gives birth to poetic forms. For example, Yu (1942) cites two main reasons: first, there were few good seven-syllable poems composed in the Western and Eastern Han ‘(两汉七言)佳制太少’, and second, they were not collected by the Han Music Bureau ‘不曾被采入乐府’. The first is more of a result than a reason, while the second is also problematic because even though trisyllabic poems were collected by the Han Music Bureau, their production declined after Han, indicating that even though it certainly had an influence, the development of poetic forms may not have been determined by the Han Music Bureau. Chu (1990: 136) proposed another possibility: ‘it was difficult for poets to master the use of too many characters in a line during the early period of seven-syllable poem formation 675

Feng Shengli

字数增加过多不能很快御驾’ and thus seven-syllable poetry developed very late. However, as seen in the following, it is difficult to attribute the slower development of heptasyllabic poetry to the hypothesis that ‘seven-syllable lines were difficult to handle within a short period of time’. (31) Qu Yuan《离骚》: 帝高阳之苗裔兮 Of the god-king Gao Yang I am the far offspring Xiang Yu《垓下歌》: 气拔山兮力盖世 My energy can lift a mountain, my strength can cover the world Bronze Mirror〈铜华镜〉: 千秋万岁乐未央 Everlasting happiness in Weiyang Palace33 Zhang Heng〈四愁诗〉: 我所思兮在太山 The one I long for is on Tai mountain As we can see, by the time of the Eastern Han, heptasyllabic verse-phrases had already been in use for hundreds of years, thus their length cannot be the actual reason for the lateness of the development of heptasyllabic poetry. Feng (2011) proposed that the reason for the late development is a prosodic effect of the language itself, namely that heptasyllabic lines are inherently in violation of the poetic Optimality Condition, which states that poetic lines are optimally formed with two rhythmic units as in (16). Since heptasyllabic lines contain more than two rhythmic units, they are unsuitable for poetic lines. Contrary to traditional analysis, the so-called seven-syllable poems of early times are not actually considered heptasyllabic in our analysis. Instead these seven syllables must be broken up into two lines, as either: (32) [2] + [2], [1+0] + [2]. or: (33) [2] + [2], [2] + [1+0]. This is because they contain three or even four rhythmic units and hence possess neither a minimal nor a poeticized rhythmic structure. As a result, these heptasyllabic lines cannot form a single poetic line according to the Poetic Optimality Condition in (16b). As we will see below, this is indeed what happened in the development of Chinese poetry. (34)

Old Country Song《旧邦曲》: 旧邦萧条心伤悲, When the old country declined, my heart was full of sadness; 孤魂翩翩当何依, A solitary spirit drifting, and where it belongs; 游士恋故涕如摧, A traveling scholar suffering homesickness and torn with grief; 兵起事大令愿违, War launched and the situation severe destroying hopes; 博求亲戚在者谁? Everywhere searching for relatives but who can be found? 立庙置后魂来归。 Build a temple, pray to the gods, and hope that the spirits will return home.

676

Poetic prosody

There appears to be three couplets of heptasyllabic lines in this poem. However, The History of the Liu Song has shown us something surprising: this six-line poem was actually recorded as having twelve lines. Music Record《宋书•乐志•四》states: ‘The above is Old Country Song which consists of 12 lines; among them, six are formed with three characters, and the other six are formed with four characters 右《旧邦曲》凡十二句,其六句句三字,六句句 四字’. So this apparently seven-syllable poem was in fact viewed differently in ancient times. The following metric groups are more likely how it was parsed at that time: (35) 旧邦萧条, jiu bang xiaotiao When the old country declined,   心伤悲。   xin shangbei   my heart was full of sadness; 孤魂翩翩, guhun bianbian A solitary spirit drifting,   当何依。   dang he yi   and where it belongs; 游士恋故, you-shi lian gu A traveling scholar suffering homesickness,   涕如摧。   ti ru cui   and torn with grief; 兵起事大, bing qi shi da War launched and the situation severe,   令愿违。   ling yuan wei   destroying hopes. 博求亲戚, bo qiu qinqi Everywhere searching for relatives,   在者谁?   zai zhe shui   but who can be found? 立庙置后， li miao zhi hou Build a temple, pray the gods,   魂来归。   hun lai gui34   and hope that the spirits will return home. The historical record reveals clearly that the prosody of these heptasyllabic lines was actually ‘4+3’ broken-lines. This not only supports the theory of Optimality Condition given above in (16), but also the conclusion that the so-called heptasyllabic poems of early times were not actually heptasyllabic, but rather formed by a tetrasyllabic line joined to a trisyllabic line. The theoretical assumptions and empirical analysis given above are further supported by the fact that in the early period of seven-syllable verse, the fourth and seventh syllables usually rhymed,35 as in: (36) Wu jing wu shuang [*doŋ] 五经无双 Within the five classics, there is no equal to (Mr.) Shuchong Xu Xu Shuchong [*srooŋ] 许叔重 ——Hou Hanshu《后汉书》 As noted in Wu (2002), ‘there is a break (comma) between the fourth and fifth characters in this type of line 这样的句子,四字和三字之间都可以加一个逗号’. Also according to Wu, there are 151 examples of seven-syllable lines with this internal rhyming in the collection of ancient verse Old Folk Songs《古谣谚》, 145 of them containing the above 4:3 rhyming structure.36 This lends support to our analysis that these seven-syllable lines are actually broken-line poems,37 which in turn supports our theory that heptasyllabic lines were not easily metricalized (七言尚未‘律化’) at the beginning of their poetic evolution. According to the above analysis, it might appear that heptasyllabic rhapsodies (赋) of the Han dynasty could not develop into a poetic prosody. However, this is not what actually occurred. Heptasyllabic poems became more and more popular during the Sui–Tang dynasties. What was actually happening? Why must seven-syllable lines be broken up into four- plus three-syllable lines at the beginning? How did the original two lines [4]+[3] become joined into a single line in their later development?

677

Feng Shengli

Heptasyllabic strings cannot form a poetic line if they contain more than two prosodic units, since this is a violation of the Minimality Condition. If this is the reason why they are not able to form a single poetic line, then one possibility for making seven-syllable poetry is to force the seven syllables into two, rather than three prosodic units. How can this happen? We know that within a seven-syllable string (2+2+2+1), the last three syllables can newly be analyzed as a single (super) foot after Eastern Han as in five-syllable poems. What about the first four syllables? They must also be re-analyzed as one prosodic unit, otherwise there will still be three prosodic units (2+2+3) in the line. In other words, the Minimal and Optimal Conditions will not be satisfied unless the [2+2] part of the [2+2+3] string becomes a single prosodic unit. This is probably what happened, as will be demonstrated below. In current studies of metrical phonology and prosodic morphology, the [2+2] bi-footing unit can be characterized as a Prosodic Word Compound (see McCarthy and Prince 1993; and Feng 2000). In other words, it is theoretically possible that the [2+2] foot forms a single prosodic unit. It is well known that Chinese four-syllable idioms (四字格) are pronounced in an independent stress pattern: 0213 (M. Yu 1989: 105). And there can be no doubt that the [2+2] metrical pattern is an independent prosodic unit in Modern Chinese. However, theoretical possibility does not guarantee empirical reality. If the four-syllable template is a single prosodic unit in Modern Chinese, the question then becomes: when was it formed in history? We must look for historical evidence of four-syllable strings in Classical Chinese which function as one prosodic unit in essentially the same fashion as they do in Modern Chinese. Although it is not possible to acquire acoustic data for the thousand year-old four-syllable template, the modern theory of prosodic syntax can tell us some of the basic prosodic structures. For example, recent studies on historical syntax have revealed that tetrasyllabic verbs start to appear only after the Eastern Han dynasty (see Shimura 1995: 230): 《法华经•五百弟子受记品》 (37) (a) 但能 护持宣助 佛之政法。 Dan neng huchi-xuanzhu fo zhi zhengfa Only can protect-promote Buddha ‘s doctrine. 《贤愚经5》 (b) 要当推求选择名女,形容色状,殊姿越群。 yao dang tuiqiu-xuanze mingnü  .  .  . Need to inquire-select a famous girl  .  .  ., 《中本起经•下》 (c) 比居一母,闻叹佛尊,驰出求索。 Bi ju yi mu, wen tan fo zun, chichu-qiusuo There is an old woman nearby, she heard the Buddha’s story, then went out and sought. These examples are among the earliest instances, occurring in the Southern and Northern Dynasties (420–589), where tetrasyllabic units began to be used in verbal and verb–object structures, indicating that tetrasyllabic verbs must have been used as a single prosodic unit. Otherwise, such lines would have been prosodically ungrammatical if, for example, huchixuanzhu 护持宣助 ‘to protect-promote’ was being used as two separate prosodic units. Another important piece of evidence for this tetrasyllabic unit comes from Liu Xie’s (刘勰) characterization of tetrasyllabic prosody: ‘the tetrasyllabic unit is dense but not rushed 四字密而不促’. The word ‘dense’ (密) here is key to characterizing four-syllable strings: the elements of this four-character unit are ‘tightly packed together’. This implies that the four-syllable unit had become a single prosodic template. It is the first time in the history of Chinese that the tetrasyllabic string has been recognized as a single unit with the prosodic property of being ‘dense but not rushed’. 678

Poetic prosody

Another important piece of evidence for the tetrasyllabic-unit comes from the development of four-syllable idioms. As calculated by T. Zhang (1999), among the 6,593 four-syllable idioms collected in the Etymological Dictionary of Chinese Idioms《汉语成语考释词典》, 68.07% were produced after the Han and mostly in the Tang–Song and Ming–Qing dynasties (59.33%), appearing especially in poetry and lyrical songs such as ‘qingtian bairi (blue sky and white sun – transparent and good political system)’ coined from the poem ‘qingtian bairi ying loutai 青天白日映楼台’ by the Tang poet Han Yu 韩愈. Though future research is needed, the above statistics indicate that the four-syllable template may have been lexicalized as Prosodic Word Compounding after the Wei–Jin period. When this [2+2] template became a recognized prosodic unit, we then had the following poetic structure:38 (38)

Rhyme Melody →

Couplet

L

Unit 2F f

L Unit F

Unit F

Unit 2F

f

f

f

年 年 月 月 对 君 子， 遥 遥 夜 夜 宿 未 央 Every year, every month Long nights sleeping in facing my lord, Weiyang palace. 吴均《行路难》 The poetic Rhyme Melody is built upon the minimal repetition of two lines, each formed by a repetition of two prosodic units: one is a compound-foot and the other is a super-foot, satisfying the Minimality and Optimality Conditions. All of the above phenomena seem to support the hypothesis that only when the [2+2] structure became a single prosodic unit could seven-syllable poems become rhymed on the seventh syllable at the end of each line, as in Bao Zhao’s 鲍照 Ni Xinglunan《拟行路难》. Thus, based on this analysis, it follows naturally that heptasyllabic verse, as would be expected, developed later than pentasyllabic poetry because when three-syllable lines developed into a trisyllabic unit (superfoot), four-syllable strings had not yet developed into a single unit. And when these four-syllable strings were still two distinct prosodic units, sevensyllable strings could not serve as optimal poetic lines. What we find in the history of Chinese literature is exactly what the theory predicts.39

7.  Final remarks Disyllabic foot formation developed during the Spring and Autumn period as proposed in Feng (1997b) and subsequently studied and examined by Feng (2012), P. Zhao (2014), Shi (2014) and Li (2015). Only later did trisyllabic foot structure develop, requiring the maturity of disyllabic Prosodic Word Formation. It took even longer for Prosodic Word Compounding in the form 679

Feng Shengli

of four-syllable structures to fully develop. Observing the development of the prosodic system from Old Chinese up to Middle Chinese, it would be expected that the development of pentasyllabic lines would have to wait for the maturity of the super-foot formation and that heptasyllabic lines would have to wait for the maturity of PrWd Compounding. The former is seen in the trisyllabic compound formation of the Eastern Han dynasty, while the latter appears in tetrasyllabic verb formation during the Southern and Northern dynasties. According to the theory presented here, these prosodically conditioned poetic effects might become the basis for a theory of prosodic genreology (韵律文体学), an interesting area for future studies.

Notes   1 The major part of this chapter was originally published in Tsing Hua Journal of Chinese Studies (2011) 42(2): 223–58. I would like to thank the editors for their permission to revise it and publish it here.   2 Translations from Owen 1996.   3 The unpopularity of hexsyllabic lines cannot be attributed to anything but prosody.   4 Some have considered it to be too short with too little information in a monotonic prosody 语句 短促,容量太少,音节单调 (Chu 1990: 229). However, the trisyllabic-word poem can be formed either by 2+1 or 1+2 syllabic structures, hence it is more flexible than the 2+2 four-syllable combination. For example (taken from Bao Zhao 鲍照《代春日行》):

入莲池 折桂枝 芳袖动 芬叶披 enter lotus-pond break cassia-twigs fragrant-sleeve move redolent-leaves shedding V+O V+O S+P S+P 1+2 1+2 2+1 2+1

Hence, inflexibility is not the cause of the unpopularity of trisyllabic poems in poetic tradition. Furthermore, trisyllabic lines can also be used to create beautiful poems, even if the lines are short. For example (鲍照《代春日行》):

入莲池,折桂枝。 Enter lotus-pond and break cassia-twigs; 芳袖动,芬叶披。 fragrant-sleeves are moving with redolent-leaves shedding; 两相思,两不知。 the two are lovesick but the two do not each know about it. Shen Deqian 沈德潜 in〈古诗源〉has commented on this poem as ‘full of emotion and heart-moving 盛情骀宕’. Obviously, there were some good trisyllabic poems (see also 李贺 Li He’s《苏小小墓》). This being the case, the unpopularity of the trisyllabic poem cannot be attributed to its literary quality.   5 This utterance can best be considered poetic prose 散文诗, but this is not a poem as we will see below. Zhu also calls it a 话 (utterance), not a 诗 (poem).   6 In Chinese poetic prosody is realized through rhythm, which is realized by means of rhymes and pauses (cf. G. Zhu 1979).   7 Even if there is a monostich, as pointed out by Hollander (2001: 11–12), ‘the one-line poem is almost always really a couplet, an epigram formed by the title and the line itself ’.   8 This system can be rendered in Chinese as 两音=诗步、两步=诗行、两行=诗联、两联=绝句、 两绝=律诗 (Feng 2008).   9 There is mounting evidence for such awkward and ungrammatical monosyllable words used alone in both Modern and Classical Chinese (Feng 2000, 2005). 10 Note that the existence of disyllabic lines in Tan’ge《弹歌》and the Yijing《易经》can be taken as evidence for an archaic moraic foot structure. They are different from the syllabic foot structure we are discussing here. See Feng 1998/2008; P. Zhao 2014; Shi 2014 for a detailed discussion of this topic. 11 For example, *何罪有 what guilt is there for Classical Chinese and *浇灌花 water flowers; *种植 树 plant trees; *阅读报 read newspapers; *购买书 purchase books, etc. for Modern Chinese, are all ungrammatical (Feng 1997a). 12 Translation from Knechtges 1996.

680

Poetic prosody 13 Translation from Knechtges 1996. 14 Translation from Knechtges 1987. 15 In the Wenjing Mifu Lun《文镜秘府论•天卷》(by Bianzhao Jingang 遍照金刚 774–835), it says: ‘In general, there is a pause after the first character and another pause after the last two characters, or a pause after the first two characters and then another pause after the last character – this is for three-character lines 凡上一字为一句,下二字为一句; 或上二字为一句,下一字为一句: 三言。’ (the word ju 句 used here refers to ‘pause’ as denoted by Shaoyu Guo 郭绍虞 1975). Accordingly, the three-syllable strings were chanted either as (22a), or (22b) in ancient times. 16 See Feng 2008 for a detailed characterization of trisyllabic forms. 17 ‘The limerick form is apparently inappropriate for the serious treatment of serious material’. (Laurence 1963: 200). 18 This is sometimes called ‘magic-religious power’. We can also find this poetic function of trisyllabic forms in history. For example, in《赋谱》: ‘Energetic (is the effect of ) three-character phrases 壮, 三字句也’. And in《文镜秘府论•南卷》: ‘To pursue the matter more closely, the achievement of the six- and seven-syllable line is to open up the beginning or to write the ending of a piece. To discuss the reasons for things and to create harmonious tones is the ability of the four- and fivesyllable lines. To embody objects and describe forms, to alternate emotions and ideas, is the essence of the three-syllable line. Although a piece of writing should embody both change and continuity, there is no fixed form to hold on to. Here I have stated the general principles 开发端绪,写送 文势,则六言、七言之功也;泛叙事由,平调声律,四言、五言之能也; 体物写状,抑扬情理,三言 之要也。虽文或变通,不可专据,叙其大抵,实在于兹’. (translation from Bodman 1978) 19 Qian Zhongshu 钱钟书 (1979: 1,295) made the following comment on these three-word opening lines: ‘open with a hard-shocked effect like singing in a rapid tone 发唱惊挺,操调险急’. 20 Translation from Owen 1974. 21 Translation by Wong Sui Kit 2003. 22 Knechtges (2002: 364) has characterized the effect of the three-word line as ‘kinetic effect’. 23 Translation from Knechtges 1987. 24 Qian Zhongshu 钱钟书 (1979: 1,295) commented on the three-word ending effect as ‘snapping shut, in a vigorous and abrupt manner 嘎然而止,简劲顿宕’. 25 Translation from Knechtges 1987. 26 Note that The Three-word Classic《三字经》was not designed for formal occasions, and is more like a jingle (if not doggerel) for children. It is the kind of prosody that children like, i.e. ballad prosody such as ‘小胖子,坐门墩,哭着喊着要媳妇儿 Little fatty sitting by the door, a little missus is what he is, weeping and wailing for’; or ‘狼来了,虎来了,和尚背着个鼓来了 Here comes the wolf, here comes the tiger; Here comes the monk, on his back a sort of trunk’. More importantly, in Hanshu《汉书•礼乐志》: ‘(The emperor) often participated in suburban sacrificial ceremonies and none of the royal court music was elegant  .  .  .  all the suburban temple poems are nothing but lascivious songs of the Zheng Style (天子)常御及郊庙皆非雅声  .  .  .  今汉郊庙诗歌  .  .  .  皆以郑声 施于朝廷’. Thus, even though trisyllabic forms were used in Jiaosi 郊祀 (suburban imperial sacrificial) events, they were still not free from colloquial and folklore influences. 27 There are insightful comments on the poetic effects of tetrasyllabic forms, such as ‘for rhythms of a grand style, one must consider the four-syllable pattern to be the standard, other possibilities are not upright even if they can serve a variety of different styles 雅音之韵,四言为正,其余虽备 曲折之体,而非音之正也,(挚虞〈文章流别论〉); ‘the four-word line is orthodox, refining and smoothing are its essence 四言正体,雅润为本’ (刘勰《文心雕龙》); and ‘in poetry, the four-word line is graceful and gracious, the five-word line is upright and disdainful; the seven-word line is relaxed and flowing; the three-word line is rigid and falling, the six-word line is sweet and pleasant; mixed-length lines are like fragrant flowers. The four-word line is the sound of Daya and thus the pneuma of poetry 诗四言优而婉,五言直而倨,七言纵而畅,三言矫而掉,六言甘而媚,杂言芬葩, 顿跌起伏。四言《大雅》之音也,其诗中之元气乎？’ 陆时雍《诗镜总论》 28 Translation from Karlgren 1974. 29 Translation from Owen 1996. 30 See Feng 2008 for a detailed analysis of why trisyllabic strings must be analyzed as more than one foot before the Western Han. 31 Translation from H. Zhang 2007. 32 Note that pentasyllable and trisyllable lines are found in complementary distribution: when fivesyllable poems flourished, three-syllable poems declined 五言成而三言衰. This observation also

681

Feng Shengli supports the analysis given here, in the sense that the maturity of the three syllable (super) foot encourages pentasyllabic poetry on the one hand, but discourages trisyllabic poems on the other. 33 Taken from Li (2001: 87) 34 There are more examples of this type in Y. Wang (1962:161). 35 In《陔余丛考》Zhao Yi 赵翼 states: ‘Most Han proverbs are made up of seven syllables and in general the fourth character rhymes with the seventh 汉人谚语多七字成句,大率以第四字与 第七字叶韵’. 36 According to Wu (2002), there are 145 instances of seven-syllable lines of this type in Gu Yaoyan 《古谣谚》   , but there are none in the Shiji《史记》. In the Hanshu《汉书》there are only three instances, while in the Hou Hanshu《后汉书》there are 29 examples. These show that sevensyllable broken-lines began in the Western Han and became popular in the Eastern Han. 37 The so-called heptasyllabic poem in《汉书•东方朔传》can therefore be analyzed as:

臣以为龙,又无角; I consider it dragon, but it has no horn; 谓之为蛇,又有足。 I call it a snake, but it has feet; 跻跻促促,善缘壁, Mounting and ascending, it is good at climbing; 是非守宫,即蜥蜴。 It is not a gecko, but a lizard.

38 Note that eight-syllable lines are too unstable for any footing possibilities, giving rise to unfavorable usage in poetic lines in Chinese. Since only odd numbers of syllable lines can avoid multiple footing possibilities, while observing the Minimality Condition, only five- and seven-syllable lines are favorable in Chinese poetry. 39 In light of the development of the heptasyllabic line from tetrasyllabic and trisyllabic lines, one may wonder why do we not expect hexasyllabic lines to evolve from [3]+[3]. The reason is very clear: the parallel [3]+[3] rhythmic structure is easily broken into a three-syllable line with another three-syllable line, resulting in a trisyllabic rather than a hexasyllabic line poem. Furthermore, the trisyllabic rhythm (either [2+1] or [1+2]) is too unbalanced, resulting in a strong limerick-sense (see Feng 2008 for detailed argumentation). These may be why hexasyllabic poems are rarely composed of a [3]+[3] rhythm.

Bibliography Bodman, Richard (1978) ‘Poetics and Prosody in Early Medieval China’, Doctoral Dissertation, Cornell University. Chen, Matthew (2000) Tone Sandhi, Cambridge: Cambridge University Press. Cheng, Xiangqing 程湘清 (1992)〈先秦双音词研究〉(‘Study on the Disyllabic Words in Pre-Qin’), in Cheng Xiangqing (ed)《先秦汉语研究》(Studies on Pre-Qin Chinese), Jinan: Shandong Educational Press, 44–112. Chu, Binjie 褚斌傑 (1990)《中国古代文体概论》(Introduction to Chinese Ancient Stylistics), Beijing: Peking University Press. Cui, Yuxi 崔宇錫 (2006)《魏晋四言诗研究》(A Study of the Four-character Poems of the Wei and Jin Dynasties), Sichuan: Bashu Press. Duanmu, San 端木三 (1999)〈重音理论和汉语的词长选择》(‘Stress Theory and Word Length in Chinese’),《中国语文》(Studies of the Chinese Language) 4: 246–54. Duanmu, San (2000) The Phonology of Standard Chinese, Oxford: Oxford University Press. Fabb, Nigel (2002) Language and Literary Structure: The Linguistic Analysis of Form in Versa and Narrative, New York: Cambridge University Press. Feng, Shengli 冯胜利 (1995) ‘Prosodic Structure and Prosodically Constrained Syntax in Chinese’, Doctoral Dissertation, University of Pennsylvania. Feng, Shengli (1996)〈论汉语的韵律词》(‘On Prosodic Words in Chinese’),《中国社会科学》(Chinese Social Sciences) 1: 161–76. Feng, Shengli (1997a)《汉语的韵律、词法与句法》(Interactions between Prosody, Morphology and Syntax in Chinese), Beijing: Peking University Press. Feng, Shengli (1997b) ‘Prosodic Structure and Compound Word in Classical Chinese’ in Jerome L. Packard (ed) New Approaches to Chinese Word Formation: Morphology, Phonology and the Lexicon in Modern and Ancient Chinese, Berlin: Mouton de Gruyter. 197–260. Feng, Shengli (1998)〈论汉语的自然音步〉(‘On Natural Foot in Modern Chinese’),《中国语文》(Chinese Philology) 1: 40–7.

682

Poetic prosody Feng, Shengli (2000)《汉语韵律句法学》(Prosodic Syntax in Chinese), Shanghai: Shanghai Educational Press. Feng, Shengli (2005)《汉语韵律语法研究》(Prosodic Grammar in Chinese), Beijing: Peking University Press. Feng, Shengli (2006) ‘Prosody and Poetic Evolution in Ancient Chinese’, Paper presented at AAS Annual Meeting. Feng, Shengli (2008)〈论三音节音步的历史来源与秦汉诗歌的同步发展》(‘On the Parallel Developments of Trisyllabic Foot Structure and the Qin–Han Poetry’),《语言学论丛》(Forum of Linguistics) 37: 18–54. Feng, Shengli (2010)〈论韵律文体学的基本原理〉(‘Some Fundamental Principles of Prosodic Stylistics’), 《修辞学习》(Rhetoric Study) 1: 25–36. Feng, Shengli (2011) ‘A Prosodic Explanation for Chinese Poetic Evolution’, Tsing Hua Journal of Chinese Studies 2: 223–57. Feng, Shengli (2012)《上古单音节音步例证 — 兼谈从韵律角度研究古音的新途径》(Evidence for Moraic Foot Structure in Archaic Chinese – A New Way to Look at Old Chinese Phonology),《历史 语言学研究》(Studies of Historical Linguistics) 5: 78–90. Feng, Shengli and Ash Henson (forthcoming) ‘Prosody of Parallel Prose and Spatiotemporal-Free Syntax – A Case Study of Wucheng Fu’. To appear in Journal of Chinese Literature and Culture, 2016. Guo, Shaoyu 郭绍虞 (1938)〈中国语词之弹性作用》(‘The Flexibility of Chinese Words’),《燕京 学报》(Yanjing Xuebao) 24: 1–34. Guo, Shaoyu (1975)《文镜秘府论•前言》(Preface to Wenjing Mifu Lun), Beijing: Renmin Literature Press. Guo, Yuheng 郭预衡 (2003) (ed)《中国古代文学史简编》(A Brief History of Classical Chinese Literature), Shanghai: Shanghai Guji Press. Hanson, Kristin and Paul Kiparsky (1996) ‘A Parametric Theory of Poetic Meter’, Language 72(2): 287–335. Hollander, John (2001) Rhyme’s Reason, New Haven: Yale University Press. Huang, Kan 黃侃 (1934)《文心雕龙札记》(Notes on Wenxin Diaolong), Beijing: Wenhua Xueshe Press. Huang, Kan (2001)《黃侃日记》(Huang Kan’s Diary), Nanjing: Jiangsu Educational Press. Karlgren, Bernhard (1974) The Book of Odes, Stockholm: BMFEA. Knechtges, David (1987) Selections of Refined Literature, vol. 3, Princeton: Princeton University Press. Knechtges, David (1996) Wen Xuan, vol. 2, Princeton: Princeton University Press. Knechtges, David (2002) ‘Narration, Description, and Rhetoric in Yang Shyong’s Yeu-Lieh Fuh: An Essay in the Form and Function in the Hann Fuh’, in Court Culture and Literature in Early China, Ashgate/Variorum, Burlington, VT, 359–77. Laurence, Perrine (1963) Sound and Sense, New York: Harcourt, Brace and World. Li, Guo 李果 (2015) ‘The Typological change of Prosody in Archaic Chinese as Evidenced by Choices between Monosyllabic and Disyllabic Names in Chinese History’, Studies of Classical Chinese 2: 58–66. Li, Lixin 李立信 (2001)《七言诗之起源与发展》(The Origin and Development of Seven-syllable Poem), Taipei: Xinwenfeng Publisher. Liberman, Mark (1975) ‘The Intonational System of English’, Doctoral Dissertation, MIT. Liberman, Mark and Alan Prince (1977) ‘On Stress and Linguistic Rhythm’, Linguistics Inquiry 8: 249–336. Link, Perry (2000) ‘The Rhetoric of Rhythm in Contemporary Chinese’, Manuscript, Princeton University. Lu, Guanzhong 卢冠忠 (2014)〈论六言诗与骈文六言句韵律及句法之异同〉(‘On the Prosody and Grammar of Heptasyllabic Lines in Poetry and Parallel Prose’), Social Sciences Forum 4: 73–86. McCarthy, John and Alan Prince (1993) ‘Prosodic Morphology I – Constraint Interaction and Satisfaction’, Manuscript, University of Massachusetts and Rutgers University. Owen, Stephen (1974) ‘Hsieh Hui-lien’s “Snow Fu”: A Structural Study’, Journal of the American Oriental Society 94(1): 14–23. Owen, Stephen (1996) An Anthology of Chinese Literature, New York: Norton. Owen, Stephen (2003) The Making of Early Chinese Classical Poetry, Cambridge, MA: Harvard University Press.

683

Feng Shengli Pan, Wuyun 潘悟云 (2000)《汉语历史音韵学》(Chinese Historical Phonology), Shanghai: Shanghai Educational Press. Qian, Zhongshu 钱钟书 (1979)《管锥篇》(Guanzhui Pian), Beijing: Zhonghua Book Company. Sagart, Laurent (2004)《上古汉语词根》(The Roots of Old Chinese), translated by Qunhu Gong 龚群 虎译, Shanghai: Shanghai Educational Press. Sapir, Edward (1921) Language – An Introduction to the Study of Speech, New York: Harcourt, Brace and World. Shi, Xiangdong 施向东 (2014)〈关于上古汉语阴声音节的韵尾、韵素和声调问题的探讨〉(‘A Study on Issues of Yin-syllable Coda, Mora and Tone in Archaic Chinese’), Manuscript, Tianjin University. Shimura, Ryōji 志村良治 (1995)《中国中世纪语法史研究》(Studies on the History of Mediaeval Chinese), translated into Chinese by Jiang Lansheng 江兰生 and Bai Weiguo 白维国, Beijing: Zhonghua Book Company. Sperber, Dan and Deridre Wilson (1995) Relevance: Communication and Cognition, Oxford: Blackwell. Wang, Hongjun 王洪君 (2000)〈汉语的韵律词与韵律短语〉(‘Prosodic Words and Prosodic Phrase in Chinese’),《中国语文》(Studies of the Chinese Language) 6: 525–36. Wang, Li 王力 (1985)《汉语史稿》(A History of Chinese Language), Beijing: Zhongguo Shehui Kexue Press. Wang, Yunxi 王运熙 (1962)《乐府诗论丛》(Essays on Poems in the Yuefu), Beijing: Zhonghua Book Company. Wong, Sui Kit (2003) ‘Rhyme-prose on the E-pang Palace by Du Mu (803–852)’, in Rachel May and John Minford (eds) A Birthday Book for Brother Stone, Hong Kong: The Chinese University Press, 337–41. Wu, Qingfeng 吴庆峰 (2002)《音韵训诂研究》(Studies on Phonology and Exegesis), Chengdu: Qilu Press. Yip, Wai-lim (1997) Chinese Poetry – An Anthology of Major Modes and Genres, Durham: Duke University Press. Yu, Guanying 余冠英 (1942)〈七言诗起源新论〉(‘New Arguments on the Origin of Seven-syllable Poetry’), in《汉魏六朝诗论丛》(Works on Hanwei and Liuchao Poems), Beijing: Renmin Wenxue Chubanshe. Yu, Min 俞敏 (1989)《俞敏语言学论文集》(Collections of Yu Min’s Linguistic Work), Heilongjiang: Heilongjiang Renmin Press. Zhang, Haiying (2007) Poetry from the Land of Silk, Chelmsford: Little Bird. Zhang, Tiewen 张铁文 (1999)〈成语的数量及产生年代〉(‘The Totality and Chronology of Chinese Idioms’),《语文建设》(Linguistic Construction) 5: 23–5. Zhao, Pusong 赵璞嵩 (2014)〈从‘吾’、‘我’的互补分布看上古汉语中韵素的对立〉(‘An Examination of Moraic Contrast Based on the Complementary Distribution of *ŋa ‘I’ and *ŋad ‘I’ in Archaic Chinese’), Doctoral Dissertation, The Chinese University of Hong Kong. Zhao, Yi 赵翼 (2003)《陔馀丛考•卷二十三》(Gaiyu Congkao), Shijiazhuang: Hebei People’s Publishing House. Zhengzhang, Shangfang 郑张尚芳 (2003)《上古音系》(Archaic Chinese Phonology), Shanghai: Shanghai Educational Press. Zhu, Guangqian 朱光潜 (1979)《诗论》(On Poetry), Hefei: Anhui Education Press. Zhu, Ziqing 朱自清 (1998)〈诗的语言》(‘Poetic Language’), in《朱自清说诗》(On Poetry by Zhu Ziqing), Shanghai: Shanghai Guji Press.

684

38 Psycholinguistics: Reading Chinese Liang Tao Ohio University, USA

Alice F. Healy University of Colorado, USA

This chapter presents comparative studies on Chinese reading processes, mainly between processing Chinese and English. The two languages are drastically different not only in their respective orthographies, but also in their syntactic patterns and discourse grammars. As Cutler (1998) once proposed: ‘Psycholinguistics is comparatively healthy’. Her argument, with which we agree, is that only by comparison can we see what is universal in human cognitive behavior and what is unique in processing a particular language. Specifically, this chapter presents studies that concern two types of Chinese reading processes: (a) general cognitive processes on words, and (b) unique cognitive processes of Chinese due to its specific orthography and discourse patterns. The theoretical issues reflected in (a) include the unitization hypothesis (e.g. Healy 1976, 1994) and the Stroop effect (e.g. MacLeod 1991; Stroop 1935). Both have been found to hold in typologically diversified languages. For (b), it will be shown that the orthographic features of Chinese invoke special cognitive processes in word identification and in discourse processes (e.g. Tao 1996, 2001, 2010; Tao and Healy 1996, 2005, 2012, 2013; L. Wang 2003). Furthermore, it will be shown that although linguistic differences may promote different cognitive processes in reading, the differences in general are a matter of degree, not absolute differences. This fact reflects how language is processed in general as part of human cognition. Except for the Stroop test, the remainder of the studies presented here all involve cognitive processes during reading of actual texts. We first present a very brief comparison of Chinese and English word formation in Section 1, and illustrate how such differences may promote different cognitive processes. Section 2 presents Chinese reading processes that reflect general human language processes, including a study on the unitization effect in processing Chinese words during reading, and a study of the Stroop effect of color-word interference during reading. Section 3 presents reading processes unique to Chinese, including a study on a word identification test during reading and a study that demonstrates how elliptical expressions in Chinese discourse cause 685

Liang Tao and Alice F. Healy

‘fuzzy syntax’, in Chinese. The elliptical syntactic patterns in Chinese discourse, as the authors propose, impact native Chinese speakers on how they process texts in general. The final study discusses the transfer of such cognitive strategies by native Chinese speakers into processing texts in their second language (English). Section 4 is a general discussion of the studies, where it is argued that the different reading processes in English and Chinese are not absolute. The processes evolve from reading practices with different orthographies, and the differences can be minimized through training.

1.  Visual Differences between Chinese and English Chinese and English have drastically different writing systems. Indo-European languages, including English, all use an alphabetic writing system with a fixed set of alphabetic letters such as the letters of the Roman alphabet to form words. The letters or letter strings (e.g. –t–, –ph–, –th–) in English mostly correspond to the segments of spoken language (e.g. W. S.-Y. Wang 1981), and the letter sequences determine word shape, which varies in length as well. In printed text, there are only spaces between words, not between letters. In contrast, Chinese has developed a logographic writing system in which the smallest meaningful unit is normally a character, which represents a syllable in spoken language. A Chinese character is a morpheme (minimal unit of meaning), and it can be a word or part of a compound word. Each character occupies a fixed square space [e.g. 一 (yi1:1 one) has only one stroke, and 皜 (hao4: white, bright, pure/clear) has 15 strokes. Both occur within the same square space. Currently most of the Chinese characters (over 80%) contain a semantic radical (referred to as ‘radical’ in short) and a phonetic component (Zhou 1978). The semantic radicals present semantic information about the meaning of the entire character. The phonetic components are usually simple, basic characters that convey phonetic information of the character. For example, 妈 (ma1: mother) and 码 (ma3: wharf ) are formed with a female radical 女 or a rock radical 石 plus the phonetic component, the character 马 (ma3: horse). Studies have found that character recognition latencies are affected by the occurrence frequency of the radical and the phonetic component, and the position of their occurrence (in the left or right side of the character) in a compound character (e.g. Feldman and Siok 1997; Taft and Zhu 1997; Taft et al. 1999; Zhang and Peng 1992). In addition, research findings suggest that the phonetic component facilitates character naming (e.g. Flores d’Arcais 1992; Flores d’Arcais et al. 1995). In those studies, when the phonetic component was displayed first before the entire character was presented, response time was significantly faster than when the semantic component was presented first. Chinese children develop character awareness before they learn to read, and reading materials take into consideration the systematic patterns of character formation (e.g. Shu 2003; Qian et al. 2015). In contemporary Chinese only 6% of the words are single characters, and over 72% are two-character compound words (Lexicon of Common Words in Contemporary Chinese 2008 (hereafter: Lexicon)2). Furthermore, characters may have different usage frequencies in forming compound words. In some cases a character may occur only in the initial (e.g. 皈 (gui1) as in 皈依 gui1yi1: to proclaim to be a Buddhist), or only in the final position of a compound, whereas in other cases a character may occur in either the initial or the final position of a compound. (e.g. 魄 (po4) as in 魄力 (po4li4: audacity), and 气魄 (qi4po4: boldness of vision), Lexicon 2008). When naming two-character Chinese compound words, studies found that the occurrence frequency of single characters, the semantic relations of the component characters to the compound, and the position of a character in the compound (initial or final) 686

Psycholinguistics: Reading Chinese

这儿有一个村庄。 There is a village here.

Figure 38.1  Visual differences between a Chinese and an English sentence

all affect response latencies in the task of lexical decision on Chinese compound words. Furthermore, it was found that the whole compound word is first processed, individual characters (morphemes) are processed later (Y. Liu and Peng 1997), suggesting that there is a top-down guidance from the meaning to the morpheme during Chinese word processing. However, previous studies on Chinese word processing (be it a single-character or a compound word) focus on word processing alone without context. Subjects were to process both the component characters and the whole words. For instance, in the task of character/ word naming/identification (e.g. Y. Liu and Peng 1997; Taft and Zhu 1995), a character or a word was presented one at a time; thus subjects could fixate on every character or word. This approach does not offer any insight into how words are processed during reading of continuous texts. In a text, Chinese characters are separated by a small space with no further spaces for word boundaries, although at the present time the majority of Chinese words are formed by two characters (e.g. H.-C. Chen 1992, 1996; Lexicon 2008). Figure 38.1 presents a comparison of the visual differences of Chinese and English, in which both Chinese and English sentences contain five words. One can see that although English presents word boundaries, Chinese only leaves a small space between characters (Tao and Healy 2012). Because of the visual differences between Chinese and English, studies have found large differences in how Chinese and English texts are processed (see, e.g. H.-C. Chen 1992, 1996; Tzeng and Hung 1981). For instance, English reading includes spelling processes, which require overt knowledge of sound-meaning correspondences. Chinese word processes involve character component identification (e.g. Feldman and Siok 1997; Taft and Zhu 1995), which requires overt knowledge of the phonetic components of characters that elucidate the phonetic information. Nonetheless, we propose that as people become fluent readers, word processes in the two languages may become very similar during reading.

2.  Common Cognitive Processes of Words in Chinese and English 2.1.  The unitization hypothesis According to the unitization hypothesis, which describes word processing during reading, common words can be processed as single holistic units, whereas rare words are always processed in smaller units like letters or characters. Evidence for this hypothesis comes from studies of languages having an alphabetic writing system (e.g. Healy 1976, 1994; Greenberg et al. 2004; Minkoff and Raney 2000; Tao et al. 1997). These studies, which used a detection task in which subjects search for a target letter in prose passages, have shown a unitization effect, whereby subjects make more detection errors on common words than on rare words. Most of the models of reading propose that words should be identified only after their components are identified (e.g. McClelland and Rumelhart 1981; Perfetti et al. 2005; Taft and Zhu 1995, 1997). For instance, normally, familiar patterns aid in the identification of their components, as in the case of the ‘word superiority effect’ (e.g. Reicher 1969; Wheeler 1970), which occurs when subjects have to decide which one of two alternative letters occurred in 687

Liang Tao and Alice F. Healy

a brief display containing the letter embedded within a familiar word or within an unfamiliar string consisting of the same letters as the word but in a scrambled arrangement. In this task it is found that letter identification is better for the familiar word than for the unfamiliar letter string. The unitization hypothesis is aimed instead to account for the contrasting finding in the letter detection task, which occurs when subjects have to detect a target letter within a passage of normal text. In this task it is found that letter detection is worse for a familiar word than for an unfamiliar letter string (e.g. Healy and Drewnowski 1983). By the unitization hypothesis (e.g. Drewnowski and Healy 1977; Healy 1976, 1994; Healy and Drewnowski 1983; Greenberg et al. 2004), fluent readers identify high-frequency words in terms of familiar higher-order units (e.g. syllables, words, or phrases) without first identifying their subcomponent letters, whereas they can identify low-frequency words (which do not contain familiar higher-order units) only in terms of their subcomponent letters. This explanation also accounts for a developmental increase in the size of the unitization effect (e.g. Buck-Gengler et al. 1998; Cunningham et al. 1988; Drewnowski 1978, 1981; Gesi Blanchard 1998; Tao et al. 1997). As readers become more fluent in the language, their familiarity with higher-order units increases, so they increase the size of the units in which they read text. Only a few studies tested the unitization hypothesis on languages with a logographic writing system (e.g. H.-C. Chen 1984, 1986, 1987). H.-C. Chen (1984, 1986) examined native Chinese speakers on their ability to detect certain characters and found a greater proportion of detection errors in high-frequency test characters than in low-frequency characters. Seidenberg (1985) found that, based on character and/or word frequency, Chinese characters and English words were processed similarly. He observed that response latency was much faster when naming high-frequency Chinese characters and English words, irrespective of their transparent phonetic information. In other words, the study found that the phonetic information only helped naming Chinese characters or English words of low usage frequency. That is to say, high-frequency Chinese characters and English words must both be processed ‘logographically’ as fused units without any direct phonological information. Seidenberg’s (1985) ‘logographic’ effect was exactly what the unitization hypothesis would predict for reading; however, Seidenberg’s study only examined phonological processes of individual characters/words. It did not consider how the internal components of the test words were processed during normal reading. Most of the previous studies only examined word processes in isolation, and subjects were automatically prompted to process the component characters in addition to the whole words. When the task was to name or identify characters or words (e.g. Liu and Peng 1997; Taft and Zhu 1995), a character or word was presented in isolation; therefore subjects were able to fixate on every character or word. As mentioned above, displays showing a single word or letter string at a time are also employed in the identification task used in studies of the word superiotiy effect. In these studies (e.g. Reicher 1969), subjects are typically asked which letter occurred in the display and must choose from alternatives that are usually shown after the display but are sometimes shown in advance (e.g. H.-C. Chen 1987; Peterzell et al. 1990). In contrast, in studies of the unitization effect using the detection task (e.g. Healy 1994), subjects are typically given text passages including multiple words shown at a time, and a target letter is always provided in advance, with subjects responding successively to each word in the passage that contains the target (e.g. by circling the target). Disadvantages for words relative to letter strings and for high-frequency relative to low-frequency words are found in the detection task but are never found in the identification paradigm. The striking differences between the results in the two situations can be understood in terms of whether the target occurs in text that is fixated (in the fovea) 688

Psycholinguistics: Reading Chinese

or instead occurs outside the area that is fixated (in the parafovea). By the parafovealprocessing hypothesis (Hadley and Healy 1991), familiar words shown in the parafovea can be identified in terms of their overall shape, but the identification of individual letters requires greater acuity, which is only available when the words occur in the fovea. H.-C. Chen’s (1984, 1986) studies did find the unitization effect. However, his findings are limited to processing single-characters in a continuous text. He did not find the unitization effect in whole character detection of Chinese compound words (H.-C. Chen, 1987). Tao and Healy (2002, 2003) explored the unitization effect in two entirely different writing systems: alphabetic (English) and logographic (Chinese). Their findings indicate that high-frequency Chinese words were also processed as holistic units, suggesting that the unitization hypothesis has general application for human reading processes. Specifically, they proposed that for fluent readers, not all words are processed in detail during reading. Reported here is part of the study by Tao and Healy (2003), which explored detection of English letter and Chinese character components to examine two related hypotheses: (a) By the ‘word frequency hypothesis’, a unitization effect (a disadvantage for high-frequency words relative to low-frequency words) will be found in both logographic and alphabet writing systems; (b) by the ‘developmental unitization hypothesis’, the magnitude of the unitization effect will increase as reading fluency increases in both writing systems. It has been well attested that in alphabetic languages, including English, the processing units in reading can be larger than letters and may even be larger than a word (e.g. Drewnowski and Healy 1977). This finding may apply to processing units in other orthographies as well. Chinese words may be a single character or compounds containing several characters, but the overwhelming majority of characters have meanings of their own; therefore the processing units in Chinese may well be larger than a character component in reading continuous texts. This assumption is contrary to findings from most previous studies on single-character processing in Chinese. In the study reported here subjects were native and non-native speakers of Chinese and English. They read passages in the two languages while detecting a specific letter or character component. The study compared detection errors of the target letter or character component in high-frequency words to those contained in low-frequency words. To demonstrate a difference in detection errors between high- and low-frequency words with Chinese material, a special manipulation of word frequency was used that was more similar to that used with English material. Thus in this study both the Chinese material and the English material included a comparison of a single very high-frequency word with words of lower frequency. Due to language change in Chinese, the semantic properties of a radical and the phonetic information may be lost, leaving each just as a component of a compound character. Taft and Zhu (1995) proposed that an orthographic or phonological sub-lexical unit may form a set of units at the conceptual level (e.g. the letter string ceive in words perceive, receive, deceive, etc.). We believe that similar phenomena also happen in Chinese. In this study the target radical in the Chinese material was 白 (white or bright). All the test characters share the radical on their left side, although some of the test characters have lost the concept of ‘white or bright’. The high-frequency single-character test word 的 (de, a grammatical particle ranked #1 in its usage frequency, Lexicon 2008) was one of them. This word originally carried the meaning ‘bright’, for which the semantic radical ‘white’ made sense. That meaning, though, is no longer carried in this word. The right-side phonetic component 勺 (shao2) of this character formerly brought phonetic information to the character; yet over time its pronunciation changed so it now is a mere component of the character. Some of the other test words also belong to this category. In terms of pattern identification, it should be easier 689

Liang Tao and Alice F. Healy

to detect the target radical 白 in the high-frequency single-character test word 的 than in the other two-character test words because the right-side component of 的 only contains three strokes, thus offering less visual interference for detecting the target radical. Alternatively, if the occurrence frequency of the whole word is critical, then the component of the highestfrequency test word 的 would be more difficult to detect while reading continuous texts. Tao and Healy (2003) tested both the unitization effect and reading comprehension. No other study had measured reading comprehension in Chinese while performing the task of character-component detection. But previous studies had found that when performing the English letter detection task subjects did read for meaning (e.g. Proctor and Healy 1985). Measuring reading comprehension is highly important for this study because the unitization effect reflects how words are processed during reading. Native Chinese (Mandarin) speakers from mainland China, native Japanese speakers, and native English speakers participated in this study. The native English speakers had two proficiency levels in Chinese: beginning (first year) and advanced. The native Japanese speakers were included in the first-year Chinese language group because they were in the same classes with the first-year Chinese language group. They were tested near the end of the academic year. One Chinese and one English passage were used. Each passage was followed by a set of four four-alternative multiple-choice comprehension questions. Passage presentation was counterbalanced. The participants were instructed to read the passage at their normal reading speed, and when they came to a letter ‘h’ in the English passage or the radical ‘白’ (either on the left-hand side of a word or as the word ‘白’ itself ) in the Chinese passage, they should circle it with their pen or pencil. Participants were told not to slow down their reading speed because they were not expected to circle all the target items. They were also instructed not to go back to circle the target if they realized that they missed one in a previous word. The instructions also informed participants of a set of multiple-choice questions following each passage that they should answer based on their understanding of the passage. The results of this study conformed to our previous findings (Tao and Healy 2002) on both hypotheses. Detection errors showed that participants who had the highest fluency in a given language showed the largest difference between the high-frequency and low-frequency words in that language. On the Chinese material, the native Chinese speakers made a much greater proportion of detection errors on the high-frequency word than on the low-frequency words. The advanced Chinese language students also made a greater proportion of detection errors on the high-frequency word than on the low-frequency words, but the difference was smaller for the advanced Chinese group than for the native Chinese group. The first-year Chinese language students did not show a unitization effect because they made only a slightly greater proportion of detection errors on the high-frequency word than they did on the low-frequency words. With the English material, all participants made a substantially greater proportion of errors on the high-frequency word than on the low-frequency words, indicating that they all had developed reading units larger than the letter. The results of participants’ accuracy in answering the multiple-choice comprehension questions again confirmed the previous findings in English (e.g. Proctor and Healy 1985) that participants read for meaning while performing the detection task. Despite their language experience and test language, all participants made only a small proportion of comprehension errors in answering the multiple-choice questions. Native English and native Japanese speakers were grouped based on their Chinese proficiency levels to examine the developmental unitization effect for Chinese. Our previous study (Tao and Healy 2002) showed no unitization effect in the native Japanese speakers at 690

Psycholinguistics: Reading Chinese

the same proficiency level as the native Japanese speakers in this study; we therefore had expected similar results from this study. The analysis that separated the data from the native English and Japanese speakers showed, as expected, no significant differences between these groups in detection errors. Together with the study by Tao and Healy (2002), this reported study supports both the word frequency and developmental unitization hypotheses in Chinese as well as in English. In particular, a significant unitization effect (more detection errors on common words than on rare words) was found for a given language only for fluent readers of that language. The size of the unitization effect in the Chinese passage was not necessarily the same for a given reader as the size of the unitization effect in the English passage, implying that the size of the reading units does not transfer across languages. Rather, the size of the reading units employed by a reader in a specific language depends completely on the reader’s fluency in that language. Our findings differ drastically from those of many earlier studies on Chinese character and word processing (e.g. Liu and Peng 1997; Taft and Zhu 1997; Taft et al. 1999, Zhang and Peng 1992; but see H.-C. Chen 1984, 1986, for similar results). The different findings seem to be due to the different methods used to present the test words. Test words were shown individually on a computer screen in the earlier studies. In our experiments, though, test words were embedded within prose passages. Healy et al. (1987) showed previously that the pattern of letter detection errors in an English prose passage depended on the number of words in view at a time, with a significant difference between a condition in which one word was shown at a time and a condition in which the full prose passage was shown, as in the current study. Hadley and Healy (1991) suggested that when a single word is in view at a given time readers must fixate on that word, enabling them to devote extra processing time to the letter level. In contrast, when a full prose passage is in view, a common word might not be fixated on because it can be identified in the parafovea. The unitization effect is attributed to such parafoveal identification, which is not possible when words are shown individually. The current experiment also found that detection of a character component in a full prose passage did not depend on whether the character was first or second in a twocharacter compound word. Word frequency alone could explain the observed unitization effect. From these results, it appears that the unitization hypothesis applies universally to all languages, despite considerable variation in their orthographies. Students learning Chinese progress from reading characters as units (as did our first-year Chinese language students) to reading larger units (as did our advanced Chinese language students). A similar developmental trend was evident in our study using a word-identification task (Tao and Healy 2013, reported in Section 3.1). There are practical implications of these findings for teaching Chinese: Teachers should keep in mind that to improve their students’ reading efficiency they will need to train them to increase the size of their reading units, especially when reading common words.

2.2.  The Stroop effect The Stroop color-word interference task offers insight into human cognitive processes (e.g. MacLeod 1991; Saalbach and Stern 2004; Stroop 1935). We give a detailed summary here of a study by Tao and Healy (2006), who used the Stroop color-word interference task to examine how language background, language fluency, orthography, and phonological information impact unintentional lexical processing in native Chinese and native English 691

Liang Tao and Alice F. Healy

speakers. As discussed earlier, Chinese and English differ drastically in their writing systems. Many previous studies find that the two writing systems evoke different cognitive processes in reading (e.g. H.-C. Chen 1987, 1996). Saalbach and Stern (2004) studied the Stroop effect in native German and Chinese speakers. Their findings support the dual-route model for word reading: English word reading involves both routes: (a) the lexical route which connects a lexical item (e.g. a letter string ‘red’) directly to its sound; and (b) the grapheme-phoneme mapping. Chinese word naming follows the direct lexical-to-sound mapping with no direct involvement of phonology. The present study supported their findings; furthermore, the present study proposes that the dual-route model for word naming is a matter of practice; both routes may work in native and non-native speakers of Chinese and English as they develop their foreign language proficiency in reading. The present study examined the Stroop effect on native and non-native speakers of Chinese and English; it also tested native Chinese speakers with a different linguistic background to support the hypotheses that (a) alphabetic and logographic orthographies promote different cognitive processes, (b) direct phonological information does not aid Chinese word naming, and (c) the dual-route model for word reading is a matter of practice: As people improve their foreign language proficiency, the Stroop effect becomes larger. Seven groups of subjects participated, including: 1. Native Chinese speakers in three different groups: (a) Mandarin speakers with knowledge of both English and Chinese pinyin,3 (b) Cantonese speakers with knowledge of English but not pinyin, and (c) Xiang (Changsha, Hunan) speakers, 65 and older, with no knowledge of pinyin or English; 2. Native Japanese speakers from first-year Chinese language classes; and 3. Native English speakers from three levels of Chinese language classes: (a) first year; (b) second year; and (c) third year. The stimuli included six basic colors with Chinese names that the first-year Chinese language students had learned: black, blue, green, red, yellow, and white. Four types of materials were presented: color patches, incongruent color names in Chinese characters, in pinyin with and without tone markings. Pinyin is only a guide for Chinese word pronunciation and is never used for text reading; therefore the knowledge of pinyin should not impact on Chinese speakers for developing cognitive processing of Chinese characters. Participants were told to use Chinese to name the ink colors of the color patches and Chinese color words, loudly and as quickly as they could. The results indicate that there was a Stroop effect overall (slower responses to incongruent color words than to color patches). Responses to the color patches were significantly faster than those to each of the color word types. There were also overall differences among the subject groups in response times, such that the native Chinese speakers (Mandarin and Cantonese) were faster than the native Japanese students and the first- and second-year native English speakers except the Xiang speakers, who were only faster than the first-year native English speakers. The Xiang speakers, being older, were slower than the Cantonese speakers. The native Japanese speakers were slower than both the second- and third-year native English speakers. The results also demonstrated a very interesting phenomenon concerning participants’ prior knowledge in processing alphabetic versus logographic writing systems. The magnitude of the Stroop effect differed as a combined function of both subject group and stimulus type, such that for both native Chinese and native Japanese speakers the Stroop effect was larger 692

Psycholinguistics: Reading Chinese

with Chinese characters than with the pinyin stimuli, but the opposite pattern was found for native English speakers. Furthermore, the Stroop effect was found with the pinyin stimuli only for subjects who knew English. The findings suggest that logographic and alphabetic writing systems evoke different cognitive processes, and that the cognitive processes develop with language proficiency. Familiarity with the alphabetic system promotes automatic phonological processing during unintentional lexical access. The significant Stroop effect on the pinyin stimuli for the two native Chinese groups (Mandarin and Cantonese) and Japanese speakers provides support for the proposal of automatic phonological processing during unintentional lexical access (e.g. Sumiya and Healy 2004) because pinyin is not the normal orthography used for reading. The finding that native Xiang speakers (who had no knowledge of pinyin or English) showed no interference of pinyin stimuli supports the proposal that familiarity with English or the Roman alphabet promotes automatic phonological processing. The finding that native Chinese and Japanese speakers showed more interference with Chinese characters than with pinyin, whereas native English speakers showed the opposite pattern supports the proposal that the two orthographies promote different cognitive processes. The similarity of these processes is evident from the interference in the Stroop test by both orthographies. The difference between the cognitive processes could be explained by the dual-route model (e.g. MacLeod 1991). For the alphabetic writing system (including pinyin) phonological information is presented directly from the alphabetic spelling; for the logographic system, the phonological information comes directly from the grapheme – because even the phonetic component of a character (e.g. 马 (ma3: horse) as in 妈 (ma1: mother)) requires strict memory of the phonological information of this component character. The comparison of the three groups of native English speakers is consistent with a developmental progression both in color naming speed (as in the study by H.-C. Chen and Ho 1986) and in the automatic processing of Chinese characters.

2.3.  Interim summary The two sets of studies reported here reflect how written words are processed in Chinese. The unitization effect is a universal cognitive process in reading. High-frequency words are processed as fused units in readers of both alphabetic and logographic languages, suggesting a direct mapping of grapheme to sound. The Stroop effect results from color-word interference during unintentional phonological processing. This study indicates that different orthographies invoke different cognitive processes. Next we present studies that reflect unique processes in Chinese reading.

3.  Unique Characteristics of Reading in Chinese 3.1.  The psychological reality of Chinese words Reading is a complex cognitive task that involves hierarchical processing, such as visualword recognition and combining words into larger meaningful units such as phrases, clauses, and sentences. We present in detail a study that focuses on how words are identified during reading in two different writing systems: alphabetic (English) and logographic (Chinese), both by native Chinese and native English speakers. Therefore, the study makes a direct comparison of word identification in Chinese and English texts by native speakers and learners of the two languages, in the participants’ native and non-native languages. 693

Liang Tao and Alice F. Healy

In alphabetic languages words are printed with inter-word spaces to separate them; therefore, visually, words may constitute the basic building blocks in those languages. In logographic languages such as Chinese, the basic individual units in written texts are the characters. A sentence is presented with only small inter-character spaces but no indication of word boundaries (see Figure 38.1), although at the present time over 94% of Chinese words are formed by two or more characters (Lexicon 2008). This special feature of Chinese may require readers to group characters into words during reading. The smallest meaningful unit in a sentence could be the word, which is relatively transparent in English. For instance, Tao and Healy (2012) found it was relatively easy to partition English words in a scrambled-word text with no word boundaries. However, word identification in Chinese was a challenge. Hoosain (1992) argued that two-character Chinese words do not function as single chunks in short-term memory; instead these compounds are like loosely packed units. He reported that in the task of word demarcation in simple Chinese sentences, native Chinese speakers disagreed on word boundaries, and they often demarcated chunks larger than words. This finding is not surprising because word boundaries are not indicated in Chinese, and readers may strive to derive general concepts from a text larger than those contained in a character or a word. Studies confirm the psychological reality of the Chinese word in speakers’ mental lexicon (e.g. Liu and Peng 1997; Shen et al. 2012). For instance, Liu and Peng (1997) used the task of character-to-word priming and found that response time was much faster for twocharacter words when a high-frequency character was used as a prime. However, studies also found that words in written texts cannot be easily identified among Chinese speakers, and sentence parsing often results in disagreement on word boundaries (e.g. Hoosain 1992; P. Liu et al. 2013; Tao and Healy 2012; Tsai et al. 1998; L. Wang 2003), although vocabulary instruction, including compound words, has been provided from kindergarten to high school in China. The majority of current Chinese words are two-character compounds (72%, Lexicon 2008). Studies found that although readers disagree on word boundaries, they tend to partition sentences in two-character units (e.g. Li et al. 2013, 2014; Perfetti and Tan 1996, 1999; Shen et al. 2012). This practice has also been incorporated into computational models for Chinese reading (e.g. K.-J. Chen and Liu 1992; Tsai 2001). The alphabetic and logographic orthographies have special impact on adult Chinese language learners. Native English speakers transfer their knowledge of alphabetic writing into Chinese when first learning the language, sometimes adding a hyphen to break a word (e.g. 白-勺 for the word 的 de: possessive) between lines, and many initially place twocharacter compound words into the same square-block, assuming a word should be presented as a unit as in English. Reported here is a recent study (Tao and Healy 2013), which proposes that the concept of ‘word’ is not a universal category as the basic processing unit in reading. The study tested three hypotheses: (a) Different orthographies promote different cognitive processes in reading; (b) processing units in Chinese expand as readers develop their language proficiency; and (c) the pinyin (Roman alphabet) for Chinese word pronunciation does not affect processing units in Chinese. The third hypothesis challenges some speculation that students from Beijing, mainland China, segment words into smaller units compared to students from Taiwan because Beijing students were taught pinyin, which presented compound words as single units (Tsai et al. 1998). Native Chinese speakers and native English speakers participated in this study. The native English speakers were undergraduate students who were completing their second- to fourth-year 694

Psycholinguistics: Reading Chinese

Chinese language courses. Unlike previous studies, which used single sentences (e.g. Perfitti and Tan 1999; L. Wang 2003), the present study used prose passages to examine word identification. Furthermore, the study also measured text comprehension to show that the seemingly difficult task of word identification does not hinder reading comprehension. Three short passages were used. They were modified supplementary readings from the second-year Chinese textbook. Each passage was presented in three forms: Chinese characters, Chinese pinyin, and English (which was a translation of the corresponding Chinese passage). The character passages were printed as normal Chinese texts with small inter-character spacing; but the pinyin and English passages had no word boundaries, only strings of letters before each punctuation mark. There was a set of four true–false comprehension questions following each passage form. Material presentation was counterbalanced so every participant processed all three passages in three different forms. Participants were instructed to read each passage at their normal reading speed. They were instructed to put a slash ‘/’ between words during reading. They were also informed of the comprehension questions after each passage. Parsing errors were classified into two types: over (units larger than the word) and under (units smaller than the word). The results indicated that word identification was much easier in English than in Chinese texts, evidenced by parsing errors. The results also showed drastic differences in word identification by the two subject groups. Both groups identified English words similarly on the three English passages. However, they differed on word identification in the two forms of Chinese passages. Native Chinese speakers made more parsing errors by selecting units larger than the word, whereas native English speakers made more parsing errors by selecting units smaller than the word. The results reflect the impact of written text format on word identification: Words could be easily identified from strings of letters for the English passages, presumably due to years of reading practice with English texts, in which words are always presented as units. Word identification was difficult for the character passages, and from strings of letters for the Chinese pinyin passages, presumably due to reading practice with Chinese texts, which may not have reinforced the notion of the word. As a result, native Chinese speakers often identified units larger than the actual words in the Chinese passages, whereas native English speakers often broke multi-character compound words into single-character words. This finding is particularly interesting for the Chinese language learners because all Chinese compound words were introduced as units in their regular language instruction, and there was a vocabulary quiz every week just to test their knowledge of words. To these Chinese learners, though, compound words did not seem to be processed as inseparable lexical units during reading. The findings from this study further support the proposal that multi-character Chinese words are processed as loosely packed units rather than single-word units in reading (Hoosain 1992). Concerning comprehension errors, the two subject groups both made a small proportion of comprehension errors on the English passages. Native English speakers, though, made a larger proportion of comprehension errors on the two forms of Chinese passages than did native Chinese speakers. However, even the highest proportion of errors made on the Chinese passages was still better than expected on the basis of chance alone. Therefore, despite parsing errors, all participants were able to comprehend the passages fairly well, thus further supporting the proposal that Chinese words may not be the basic processing units during reading. Furthermore, native English speakers were able to correctly identify some compound words that have high usage frequency (e.g. 朋友: friend; 没有: have not). This observation 695

Liang Tao and Alice F. Healy

provides additional support for the developmental unitization hypothesis: As learners improve their reading proficiency, they increase the size of their processing units (e.g. Healy 1994; Tao and Healy 2002, 2003). This study has found remarkable differences between native speakers of Chinese and English in how they partition words in Chinese and English texts. Both groups of participants parse sentences into small units. The major difference is the size of such units. The smallest such units could be words, which are relatively transparent in English. Words, though, may not necessarily constitute the basic processing units in Chinese. Furthermore, there is a clear developmental progression in the processing of Chinese words as language proficiency improves. This study supports the view that different orthographies evoke different cognitive processes during reading. Despite formal instruction on compound words, the format of Chinese texts may have trained readers to process texts in units larger than the word. Therefore, the text forms of the alphabetic and logographic orthographies promote different cognitive processes during reading. The notion of the word may not be a universal category in text processing. The segmentation processes are cultivated through reading practice; learners of a foreign language develop cognitive processes in addition to learning vocabulary and grammar. In the next section we present studies on the recoverability of elliptical expressions in reading Chinese texts, and reference tracking during discourse processing.

3.2.  Zero anaphora and fuzzy syntax This section discusses discourse processing strategies unique to Chinese, possibly fostered by the discourse grammar of the language. The specific focus is the role of zero anaphora in Chinese discourse. Two studies are discussed in detail here; one examines syntactic recoverability while reading Chinese texts; the other examines transfer of reference-tracking strategies from reading Chinese to reading English discourse (Section 3.3). Anaphora is very important in maintaining cohesion in discourse (e.g. Van Dijk and Kintsch 1983). Zero anaphora is one of the anaphoric devices, an empty grammatical slot in a sentence standing for a previously mentioned nominal referent. In the following expression, if the pronoun she is dropped, the empty slot becomes an instance of zero anaphora: ‘Isabelle Hemingway went to China and (she) taught midwifery there’ The discourse pragmatic constraints for using zero anaphora are equivalent to those for using pronouns (e.g. P. Chen 1986; Tao 1996). These constraints apply cross linguistically and have been described by Fox and Thompson (1990) in terms of ‘information flow’ (297). By this account, the way in which speakers use nominal referents in discourse depends on the information status of the referents (e.g. Chafe 1994; Du Bois 1985, 1987; Du Bois et al. 1993). The anaphoric devices consist of pronouns as well as full noun phrases, and the choice of device depends on whether the referent is new, accessible, or given (Chafe 1994). A referent that is new is inactive in the mind of the listener, so the listener cannot readily access information relevant to the referent. Consequently, the speaker typically chooses to introduce a new referent with a noun phrase that is full and clear (e.g. ‘Mary Jones’). A referent that is accessible is semiactive in the mind of the listener, presumably because it had been mentioned previously but not recently in the discourse, which requires some effort on the part of the listener to identify it. Thus, the speaker typically chooses to use a noun 696

Psycholinguistics: Reading Chinese

phrase but it can be abbreviated (e.g. ‘Mary’). Finally, a referent that is given is active in the mind of the listener because it was recently mentioned, so the speaker typically chooses to employ a pronoun in that case (e.g. ‘she’). In fact, when a referent is active because it has just been mentioned, the coherence of the discourse might be disrupted and the listener might require a longer time to process the discourse when the speaker chooses to use a full noun phrase, which Gordon and Chan (1995: 217) refer to as a ‘repeated name penalty’. The analysis by Chafe (1994) is consistent with that by Givon (1983), who refers to it as topic continuity. By this proposal what is crucial in determining the choice of anaphoric devices is the distance in the discourse from the most recent occurrence of the referent, with the likelihood of a reduced form increased as the distance is decreased. An iconicity principle is suggested by Givon’s notion of distance: As a discourse topic becomes continuous and predictable, it requires less overt expression. Cross-linguistic studies provide much evidence for the theoretical notions of information flow, topic continuity, and the iconicity principle (e.g. Chafe 1994; P. Chen 1986; Givon 1983; Pu 1997; Tsao 1979). These studies, though, were all based on the assumption that language production and processes are both linear in form. There are studies of both written and spoken discourse, though, that demonstrate violations of the iconicity principle and explain these violations with the assumption that language processing is hierarchical rather than linear (Fox 1987; Tao 1996, 1997). Specifically these studies demonstrate that a pronoun in English or an instance of zero anaphora in Chinese may refer to a nominal referent that is several clauses back. In each of these cases of ‘jumping’ back, there are specific cues, such as in the semantics of the verb, which allow the readers or listeners to reactivate and, thus, track these distant referents. There are, however, cases in Chinese, both in formal and informal text, where there is no grammatical marking that indicates the missing information. For example, some instances of zero anaphora that violate topic continuity follow a switch-reference pattern (Tao 1997, 2001). In this pattern, there is a switch in referents, or a change in grammatical subjects from the first to the second clause of a sentence. Consider, for example, the following sentence that illustrates the switch-reference pattern: ‘He bought a small plant, but it bloomed surprisingly well’. This English sentence directly translates a sentence contained in an actual Chinese discourse although the word ‘it’ was not present in Chinese but instead was replaced by zero anaphora. Therefore, readers of Chinese must use special reference-tracking strategies to extract information that is missing (e.g. Tao and Healy 1996, 1998, 2005). Sentences with missing referents become elliptical. Tao (2010) proposes that sometimes it is difficult or even impossible to recover the grammatical patterns from these elliptical expressions. In this situation there is a fuzzy boundary between syntactic structures; yet comprehension is not compromised when processing such expressions. Tao thus proposes that conceptual information seems more essential than sentence grammar for Chinese text processing. Reported here is part of a syntactic recovery test (Tao 2010) in Chinese, which examined the hypothesis of fuzzy syntax. The study first tried to confirm the psychological reality of zero anaphora in Chinese discourse by testing native Chinese speakers on a set of elliptical expressions without context. For instance, one of the test expressions is: 玩儿飞了 Wan2r fei1 le: play-fly-Perfective-Aspect. This example contains a verb phrase in a resultative construction: a main verb (玩儿: to play) and a verb-complement (飞: fly), presenting a causative concept: ‘fly away’ caused by the ‘playing’. It involved two referents in the original narrative, a cat and a moth. In the original narrative, one of its interpretations could be ‘(the moth) flew away caused (by the cat’s playing (with it))’. Without context, all participants reported that these expressions did not make sense. Some tried to make sense 697

Liang Tao and Alice F. Healy

out of them by adding overt referents to these expressions based on the semantic cues from the verbs, plus some general world knowledge (such as a bird with wings). This finding suggests that zero anaphora is possible in Chinese, but only with discourse contexts (e.g. Duranti and Goodwin 1992). Next, two natural passages (one from a transcript of an oral narrative and one from a written text) were used to examine further the role of zero anaphora in Chinese discourse. One of the passages, the Cat Passage (transcription of a narrative), contained the sample elliptical expression with two zero anaphors, a cat and a moth. The cat has been mentioned in the prior four sentences so it is coded by zero anaphora. This example follows the theoretical proposals of information flow (e.g. Chafe 1994) and topic continuity (Givon 1983). But the moth, which is also coded by zero anaphora, has not been mentioned in the immediately preceding clauses. Yet its information should still be accessible given the story-line being a cat playing with a moth. A different group of native Chinese speakers were asked to read the passages and to add missing referents whenever they felt the need to do so. Without any hints provided on the elliptical expressions, almost all participants felt there was no need to add overt referents to the passages. This result indicates that within a proper discourse context, the same elliptical expression not only made perfect sense, the nominal referents were clearly accessible to Chinese readers. Because Chinese discourse allows ample use of zero anaphora, native speakers may have developed special cognitive strategies to track their referents (e.g. Tao and Healy 2005). Therefore, to native Chinese speakers, texts containing multiple zero anaphors may not seem to have any information missing during Chinese reading. The third test is about syntactic recoverability, which examined the hypothesis of fuzzy syntax. Printed on the same page with a passage was a set of statements to interpret the elliptical expressions in the passages. For the sample test expression from the Cat Passage, four statements presented (listed below) complete sentences with different syntactic structures and overt mentions of the two referents. In theory, Statement (a) should be the closest match to the elliptical expression because it has the Cat as the subject/agent, following the prediction of topic continuity. But even this one has to add a grammatical particle BA to allow the overt mention of the patient, the Moth. Participants were instructed to choose any statements as the intended interpretation of the elliptical expression. For the Elliptical expression 玩儿飞了 Wan2r fei1 le: play-fly-Perfective Aspect, it means: a. 猫把蛾子玩儿飞了 (the BA4-object-fronting pattern focusing on the agent’s force on patient); b. 蛾子被猫玩儿飞了 (the Passive Bei4 Construction focusing on the adversity of the patient); c. 猫玩儿蛾子玩儿得蛾子飞了 (the Active-serial-events focusing on the agent’s action); d. 猫玩儿蛾子把蛾子玩儿得飞了 (Similar to ‘a’ with serial events and the adversity of the patient); e. All of the above are acceptable. f. Some of the above (a, b, c, and/or d) are acceptable. Participants were encouraged to offer an explanation for their decision, although very few were able to offer an explanation beyond the comment ‘the choice made sense’. The results revealed that subjects disagreed on the exact syntactic structure of the elliptical expressions. For our sample expression, the responses were highly mixed, but there was a significant preference favoring the ‘All acceptable’ choice. 698

Psycholinguistics: Reading Chinese

The results indicate that, with the same discourse context presented, an elliptical sentence containing two missing referents (i.e. zero anaphora) still could be reinterpreted with multiple grammatical structures. The findings suggest that native Chinese speakers seemed to tolerate the fuzzy boundary of syntax during discourse processing. Language grammaticalizes different syntactic structures for a reason. From the discoursefunctional point of view, pragmatically no two grammatical patterns should present exactly the same meaning. However, the present study shows that for the same elliptical sentence, native Chinese speakers accept different syntactic structures for its interpretation. The findings support the fuzzy or non-discrete nature of syntax in discourse (e.g. Aarts et al. 2004), thus challenging the traditional view of single sentences as fundamental units in linguistic analysis. The study supports the proposal that grammar is not binary and linguistic phenomena are not strictly all-or-none in their behavior (Bolinger 1961), especially in Chinese reading.

3.3.  Transfer of cognitive strategies from Chinese to English discourse English discourse contains many nominal referents that would be absent in Chinese discourse because Chinese makes greater use of zero anaphora. In three sets of studies (Tao and Healy 1996, 1998, 2005), native Chinese and native English speakers were compared in comprehending modified English passages. Chinese speakers showed superior comprehension when English texts contained zero anaphora similar to Chinese discourse (Tao and Healy 1996, 1998, 2005), and in their answers to comprehension questions (Tao and Healy 2005). The findings indicate that native Chinese speakers develop reference tracking strategies that they transfer to comprehending English. Reported here is part of the study from Tao and Healy (2005) that measured comprehension of English passages by native Chinese and native English speakers. The materials included four passages from a standardized English reading test (McCall and Crabbs 1950, 1961). After each passage there were six four-alternative multiple-choice questions to measure comprehension of the passage. Passages were presented in three conditions: full English passages, passages with random deletion of nominal referents, and passages with nominal referents deleted that would correspond to zero anaphora in Chinese discourse. Measured by the error-rate on the answers to the multiple-choice comprehension questions, native Chinese speakers did not do as well as native English speakers with the full passages (Experiment 1). They did about equally poorly as native English speakers on the passages containing random deletion of nominal referents (Experiment 2). However, they did better than native English speakers on passages containing missing referents that corresponded to zero anaphora in Chinese (Experiments 1, 2, and 3). Specifically, Experiment 3 tested native Chinese and native English speakers; all were undergraduate students respectively from the University of Colorado, Boulder, and from the National Chiao Tung University, Taiwan. None of the native Chinese speakers had taken the TOEFL test. However, their performance in a freshman level sociolinguistics course at the University of Colorado was similar to that of native English speaking students who took that course either with them or on another occasion. The results showed that the native Chinese speakers made fewer errors overall than did the native English speakers in answering the multiple-choice questions on these passages, all of which contained zero anaphora that would have occurred in the corresponding Chinese discourse. Furthermore, the results indicate that native English speakers made a greater proportion of errors only on the first passage and that their comprehension errors decreased 699

Liang Tao and Alice F. Healy

on the second passage. As a result, the proportion of errors made on the second passage was the same for the two subject groups. The findings suggest that the experience in processing the first passage was sufficient for native English speakers to be aware of the extra missing referents; thus they altered their reference-tracking strategies when processing the second passage. This result further supports findings from our previous studies (Tao and Healy 1996) that native English speakers were equally capable of recovering missing nominal referents that correspond to zero anaphora in Chinese discourse.

3.4.  Interim summary In summary, the findings reported in Sections 3.2–3.3 indicate that, first of all, Chinese discourse allows ample use of zero anaphora, presumably for the sake of economy in communication. Native Chinese speakers have developed special cognitive strategies in tracking missing referents when reading Chinese texts. Tao (1996) proposed that such strategies most likely involve paying special attention to semantic cues from the predicate verbs plus the local discourse context. As a result, native Chinese speakers were often unaware of any referents missing during reading. They could also tolerate imprecise syntactic forms during reading. Furthermore, native Chinese speakers were able to transfer such reference-tracking strategies into processing English discourse. Because zero anaphora follows the general discourse pattern of information flow, native English speakers, whose discourse does not contain nearly as many zero anaphora, were also able to process passages with nominal referents missing as long as they had minimal sufficient experience and the missing nominals corresponded to zero anaphora in discourse.

4. Conclusion The five studies reported here are all psycholinguistic studies on reading and word processing in Chinese. The findings indicate that although English and Chinese have drastically different orthographies, word processing during reading of the two languages reflects similar cognitive processes as described in the theory of unitization. Words of high usage frequency were processed as fused units, whereas low-frequency words were processed by their subcomponents. The unitization effect challenges the dual-route model of word processes because high-frequency words may be processed ‘logographically’ in alphabetic and logographic orthography. The findings from the Stroop effect in the color-word interference task indicate that language background, language fluency, orthography, and phonological information impact unintentional lexical processing in both native Chinese and native English speakers. Alphabetic and logographic orthographies promote different cognitive processes; therefore, direct phonological information does not aid Chinese word naming, and the dual-route model for word reading is a matter of practice: As people improve their foreign language proficiency, the Stroop effect becomes larger for both native Chinese and native English speakers when processing their non-native language. The word identification task during reading Chinese and English texts yielded findings to support the proposal that orthography in discourse, not words alone, determines how words are processed during reading. The findings also suggest that native Chinese speakers’ reading processes involve conceptual units larger than the word. The results of this study from native English speakers’ processing of Chinese texts further suggest the impact of discourse patterns on word identification – although compound words had been properly instructed, native 700

Psycholinguistics: Reading Chinese

English speakers still had trouble identifying these words while reading Chinese texts. However, the illusive task of word identification did not seem to interfere with reading in Chinese because the proportion of errors in answering the comprehension questions by both participant groups was low. Reading in Chinese involves processing discourse with abundant zero anaphora, which may promote special discourse-processing strategies. The findings from the syntactic recoverability study on native Chinese speakers suggest that such strategies may be the reason that native speakers do not think anything is elliptical when reading Chinese passages (Section 3.2). They are also able to transfer these strategies into processing English passages (Section 3.3). Native English speakers, on the other hand, are also able to handle English passages with nominal referents missing given minimal sufficient experience with such passages (Section 3.3). In summary, Chinese reading may involve special cognitive strategies that have been promoted by the orthography and discourse patterns of Chinese. These strategies may be reflected to various degrees in reading and language processes as part of general human cognition.

Notes 1 Mandarin Chinese has four tones. Each syllable has a designated tone, indicated by the numeral 1–4. In ‘zi4’ the numeral indicates the fourth tone, which is a high falling tone. 2 This is the English title of the book. The literal translation of the Chinese title is: Frequency Chart of Commonly Used Words in Contemporary Chinese. This chart includes a total of 56,008 words. 3 Pinyin means ‘spell-sound’ in Chinese. It borrows the Roman alphabet to provide word pronunciation.

References Aarts, Bas, David Denison, Evelien Keizer, and Popova Gergana (eds) (2004) Fuzzy Grammar, Oxford: Oxford University Press. Bolinger, Dwight (1961) Generality, Gradience, and the All-or-None, Gravenhage: Mouton and Company. Buck-Gengler, Carolyn J., Stephen G. Romero, Alice F. Healy, and Lyle E. Bourne, Jr. (1998) ‘The Effect of Alphabet and Fluency on Unitization Processes in Reading’, in Alice F. Healy and Lyle E. Bourne, Jr. (eds) Foreign Language Learning: Psycholinguistic Studies on Training and Retention, Mahwah, NJ: Lawrence Erlbaum, 273–90. Chafe, Wallace (1994) Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing, Chicago: University of Chicago Press. Chen, Hsuan-Chih (1984) ‘Detecting Radical Component of Chinese Characters in Visual Reading’, Journal of Chinese Psychology 26: 29–34. Chen, Hsuan-Chih (1986) ‘Component Detection in Reading Chinese Characters’, in Henry S. R. Kao and Rumjahn Hoosain (eds) Linguistics, Psychology, and the Chinese Language, Hong Kong: Hong Kong University Press, 1–10. Chen, Hsuan-Chih (1987) ‘Character Detection in Reading Chinese: Effects of Context and Display Format’, Journal of Chinese Psychology 29: 45–50. Chen, Hsuan-Chih (1992) ‘Reading Comprehension in Chinese: Implications from Character Reading Times’, in Hsuan-Chih Chen and Ovid J. L. Tzeng (eds) Language Processing in Chinese, Amsterdam: Elsevier, 175–205. Chen, Hsuan-Chih (1996) ‘Chinese Reading and Comprehension: A Cognitive Psychology Perspective’, in Michael H. Bond (ed) The Oxford Handbook of Chinese Psychology, Oxford: Oxford University Press, 43–62. Chen, Hsuan-Chih and Connie Ho (1986) ‘Development of Stroop Interference in Chinese–English Bilinguals’, Journal of Experimental Psychology: Learning, Memory, and Cognition 12: 397–401. Chen, Keh-Jiann and Liu Shing-Huan (1992) ‘Word Identification for Mandarin Chinese Sentences’, in Proceedings of COL1NG-92, Nantes, 23–28 August, 101–7, Available at: http://acl.ldc.upenn. edu/C/C92/C92-1019.pdf.

701

Liang Tao and Alice F. Healy Chen, Ping (1986) ‘Referent Introducing and Tracking in Chinese Narrative’, Doctoral Dissertation, University of California, Los Angeles (University Microfilms). Cunningham, Thomas F., Alice F. Healy, Nancy Kanengiser, Lynne Chizzick, and Rebecca L. Willitts (1988) ‘Investigating the Boundaries of Reading Units across Ages and Reading Levels’, Journal of Experimental Child Psychology 45: 175–208. Cutler, Anne (1998) ‘Psycholinguistics Is Comparatively Healthy’, Invited talk given at the Institute for the Empirical Study of Language, Ohio University. Drewnowski, Adam (1978) ‘Detection Errors on the Word the: Evidence for the Acquisition of Reading Levels’, Memory and Cognition 6: 403–9. Drewnowski, Adam (1981) ‘Missing -ing in Reading: Developmental Changes in Reading Units’, Journal of Experimental Child Psychology 31: 154–68. Drewnowski, Adam and Alice F. Healy (1977) ‘Detection Errors on the and and: Evidence for Reading Units Larger than the Word’, Memory and Cognition 5: 636–47. Du Bois, John W. (1985) ‘Competing Motivations’, in John Haiman (ed) Iconicity in Syntax, Amsterdam and Philadelphia: John Benjamins, 343–65. Du Bois, John W. (1987) ‘The Discourse Basis of Ergativity’, Language 63: 805–55. Du Bois, John W., Stephan Schuetze-Coburn, Susanna Cumming, and Danae Paolino (1993) ‘Outline of Discourse Transcription’, in Jane A. Edwards and Martin D. Lamper (eds) Talking Data: Transcription and Coding in Discourse Research, Hillsdale, NJ: Lawrence Erlbaum, 45–89. Duranti, Alessandro and Charles Goodwin (eds) (1992) Rethinking Context: Language as an Interactive Phenomenon, Cambridge: Cambridge University Press. Feldman, Laurie B. and Siok Wai Ting (1997) ‘The Role of Component Function in Visual Recognition of Chinese Characters’, Journal of Experimental Psychology: Learning, Memory and Cognition 23: 776–81. Flores d’Arcais, Giovanni B. (1992) ‘Graphemic, Phonological, and Semantic Activation Processes during the Recognition of Chinese Characters’, in Hsuan-Chih Chen and Ovid J. L. Tzeng (eds) Language processing in Chinese, Amsterdam: Elsevier, 37–66. Flores d’Arcais, Giovanni B., Saito Hirofumi and Kawakami Masahiro (1995) ‘Phonological and Semantic Activation in Reading Kanji Characters’, Journal of Experimental Psychology: Learning, Memory and Cognition 21: 34–42. Fox, Barbara (1987) Discourse Structure and Anaphora: Written and Conversational English, Cambridge: Cambridge University Press. Fox, Barbara and Sandra A. Thompson (1990) ‘A Discourse Explanation of the Grammar of Relative Clauses in English Conversation’, Language 66: 297–316. Gesi Blanchard, Antoinette T. (1998) ‘Transfer Effects of First Language Proficiency on Second Language Reading’, in Alice F. Healy and Lyle E. Bourne, Jr. (eds) Foreign Language Learning: Psycholinguistic Studies on Training and Retention, Mahwah, NJ: Lawrence Erlbaum, 291–314. Givon, Thomas (1983) ‘Topic Continuity in Discourse: An Introduction’, in Thomas Givon (ed) Topic Continuity in Discourse: A Quantitative Cross-language Study, Amsterdam and Philadelphia: John Benjamins, 1–41. Gordon, Peter C. and Davina Chan (1995) ‘Pronouns, Passives, and Discourse Coherence’, Journal of Memory and Language 34: 216–31. Greenberg, Seth N., Alice F. Healy, Asher Koriat, and Hamutal Kreiner (2004) ‘The GO Model: A Reconsideration of the Role of Structural Units in Guiding and Organizing Text on Line’, Psychonomic Bulletin and Review 11: 428–33. Hadley, Jeffery A. and Alice F. Healy (1991) ‘When Are Reading Units Larger Than the Letter? A Refinement of the Unitization Reading Model’, Journal of Experimental Psychology: Learning, Memory, and Cognition 17: 1,062–73. Healy, Alice F. (1976) ‘Detection Errors on the Word the: Evidence for Reading Units Larger Than Letters’, Journal of Experimental Psychology: Human Perception and Performance 2: 235–42. Healy, Alice F. (1994) ‘Letter Detection: A Window to Unitization and other Cognitive Processes’, Psychonomic Bulletin and Review 1: 333–44. Healy, Alice F. and Adam Drewnowski (1983) ‘Investigating the Boundaries of Reading Units: Letter Detection in Misspelled Words’, Journal of Experimental Psychology: Human Perception and Performance 9: 413–26. Healy, Alice F., William L. Oliver, and Timothy P. McNamara (1987) ‘Detecting Letters in Continuous Text: Effects of Display Size’, Journal of Experimental Psychology: Human Perception and Performance 13: 279–90.

702

Psycholinguistics: Reading Chinese Hoosain, Rumjahn (1992) ‘Psychological Reality of the Word in Chinese’, in Hsuan-Chih Chen and Ovid J. L. Tzeng (eds) Language Processing in Chinese, Amsterdam: Elsevier, 111–30. Lexicon of Common Words in Contemporary Chinese Research Group (2008)《现代汉语常用词表》 Lexicon of Common Words in Contemporary Chinese Beijing: The Commercial Press. Li, Xingshan, Klinton Bicknell, Liu Pingping, Wei Wei, and Keith Rayner (2014) ‘Reading Is Fundamentally Similar Across Disparate Writing Systems: A Systematic Characterization of How Words and Characters Influence Eye Movements in Chinese Reading’ Journal of Experimental Psychology: General 143: 895–913. Li, Xingshan, Gu Junjian, Liu Pingping, and Keith Rayner (2013) ‘The Advantage of Word-Based Processing in Chinese Reading: Evidence From Eye Movements’, Journal of Experimental Psychology: Learning, Memory, and Cognition 39: 879–89. Liu, Ping-Ping, Li Wei-jun, Lin Nan and Li Xingshan (2013) ‘Do Chinese Readers Follow the National Standard Rules for Word Segmentation during Reading?’ PLoS ONE 8(2):e55440.doi:10.1371/ journal. pone.0055440. Liu, Yi and Peng Danling (1997) ‘Meaning Access of Chinese Compounds and its Time Course’, in Chen Hsuan-Chih (ed) Cognitive Processing of Chinese and Related Asian Language, Hong Kong: The Chinese University Press, 219–32. MacLeod, Colin M. (1991) ‘Half a Century of Research on the Stroop Effect: An Integrative Review’, Psychological Bulletin 109: 163–203. McCall, William A. and Lelah M. Crabbs (1950, 1961) Standard Test Lessons in Reading, New York: Teacher’s College Press. McClelland, James L. and David E. Rumelhart (1981) ‘An Interactive Activation Model of Context Effects in Letter Perception: Part I: An Account of Basic Findings’, Psychological Review 88: 375–407. Minkoff, Scott R. B. and Gary E. Raney (2000) ‘Letter-detection Errors in the Word the: Word Frequency versus Syntactic Structure’, Scientific Studies of Reading 4: 55–76. Perfitti, Charles A., Liu Ying, and Tan Li Hai (2005) ‘The Lexical Constituency Model: Some Implications of Research on Chinese for General Theories of Reading’, Psychological Review 112: 43–59. Perfetti, Charles A. and Tan Li Hai (1996) ‘Are There Structural Principles in Sentence Understanding across Languages?’ Paper presented at the Psychonomic meeting, Chicago, November 1996. Perfetti, Charles A. and Tan Li Hai (1999) ‘The Constituency Model of Chinese Word Identification’, in Jian Wang, Albrecht W. Inhoff, and Chen Hsuan-Chih (eds) Reading Chinese Script: A Cognitive Analysis, Mahwah, NJ, Lawrence Erlbaum, 115–34. Peterzell, David H., Grant P. Sinclair, Alice F. Healy, and Lyle E. Bourne, Jr. (1990) ‘Identification of Letters in the Predesignated Target Paradigm: A Word Superiority Effect for the Common Word the’, American Journal of Psychology 103: 299–315. Proctor, Janet D. and Alice F. Healy (1985) ‘A Secondary-task Analysis of a Word Familiarity Effect’, Journal of Experimental Psychology: Human Perception and Performance 11: 286–303. Pu, Mingming (1997) ‘Cognitive Constraints and Discourse Anaphora in Mandarin Chinese’, in Tseng Chiu-yu (ed) Chinese Languages and Linguistics, IV, Taiwan: Academia Sinica, 453–84. Qian, Yi, Song Yao-Wu, Zhao Jing, and Bi Hong-Yan (2015). ‘The developmental trend of orthographic awareness in Chinese preschoolers’, Reading and Writing 28: 571–86. Reicher, Gerald M. (1969) ‘Perceptual Recognition as a Function of Meaningfulness of Stimulus Material’, Journal of Experimental Psychology 81: 275–80. Saalbach, Henrik and Elsbeth Stern (2004) ‘Differences between Chinese Morphosyllabic and German Alphabetic Readers in the Stroop Interference Effect’, Psychonomic Bulletin and Review 11: 709–15. Seidenberg, Mark S. (1985) ‘The Time Course of Phonological Code Activation in Two Writing Systems’, Cognition 19: 1–30. Shen, Deli, Simon P. Liversedge, Tian Jin, Zang Chuanli, Cui Lei, Bai Xuejun, Yan Guoli, and Keith Rayner (2012) ‘Eye Movements of Second Language Learners When Reading Spaced and Unspaced Chinese Text’, Journal of Experimental Psychology: Applied 18: 192–202. Shu, Hua (2003) ‘Chinese writing system and learning to read’, International Journal of Psychology 38: 274–85. Stroop, John R. (1935) ‘Studies of Interference in Serial Verbal Reactions’, Journal of Experimental Psychology 18: 643–62.

703

Liang Tao and Alice F. Healy Sumiya, Hiromi and Alice F. Healy (2004) ‘Phonology in the Bilingual Stroop Effect’, Memory and Cognition 32: 752–58. Taft, Marcus and Zhu Xiaoping (1995) ‘The Representation of Bound Morphemes in the Lexicon’, in Laurie Feldman (ed) Morphological Aspects of Language Processing, Hillsdale, NJ: Lawrence Erlbaum, 293–316. Taft, Marcus and Zhu Xiaoping (1997) ‘Submorphemic Processing in Reading Chinese’, Journal of Experimental Psychology: Learning, Memory and Cognition 23: 761–75. Taft, Marcus, Zhu Xiaoping, and Peng Danling (1999) ‘Positional Specificity in Chinese Character Recognition’, Journal of Memory and Language 40: 498–519. Tao, Liang (1996) ‘Topic Discontinuity and Zero Anaphora: Cognitive Strategies in Discourse Processing’, in Barbara Fox (ed) Studies in Anaphora, Amsterdam and Philadelphia: John Benjamins, 487–513. Tao, Liang (1997) ‘Topic Choice, Switch Reference and Zero Anaphora: The On-line Construction of Grammar’, in Tseng Chiu-yu (ed) Chinese Languages and Linguistics, IV: Typological Studies of Languages in China, Academia Sinica, 523–54. Tao, Liang (2001) ‘Switch Reference and Zero Anaphora: Emergent Reference in Discourse Processing’, in Alan Cienki, Barbara J. Luka, and Michael B. Smith (eds) Conceptual and Discourse Factors in Linguistic Structure, Stanford, CA: CSLI Publications, 253–69. Tao, Liang (2010) ‘Fuzzy Syntax: Zero Anaphora in Chinese and Sentence Processing’, Poster Presentation at the 51st Annual Meeting of the Psychonomic Society, St Louis, MO, 20 November. Tao, Liang and Alice F. Healy (1996) ‘Cognitive Strategies in Discourse Processing: A Comparison of Chinese and English Speakers’, Journal of Psycholinguistic Research 25: 597–616. Tao, Liang and Alice F. Healy (1998) ‘Anaphora in Language Processing: Transfer of Cognitive Strategies by Native Chinese, Dutch, English, and Japanese Speakers’, in Alice F. Healy and Lyle E. Bourne Jr. (eds) Foreign Language Learning: Psycholinguistic Studies on Training and Retention, Mahwah, NJ: Erlbaum, 193–212. Tao, Liang and Alice F. Healy (2002) ‘The Unitization Effect in Reading Chinese and English Texts’, Journal of Scientific Studies of Reading 6: 167–97. Tao, Liang and Alice F. Healy (2003) ‘The Unitization Effect in the Development of Reading in Chinese and English: Evidence from Letter/Character-component Detection’, in Catherine McBrideChang and Chen Hsuan-Chih (eds) Reading Development in Chinese Children, Westport CT: Greenwood Publishing, 183–96. Tao, Liang and Alice F. Healy (2005) ‘Zero Anaphora: Transfer of Reference Tracking Strategies from Chinese to English’, Journal of Psycholinguistic Research 34: 99–131. Tao, Liang and Alice F. Healy (2006) ‘Stroop Effect in Chinese Characters and Pinyin: Orthography and Language Experience’, Poster presented at the 47th Annual Meeting of the Psychonomic Society, Houston, Texas, 18 November. Tao, Liang and Alice F. Healy (2012) ‘Parsing English and Chinese Strings of Characters into Words by Speakers Varying in Fluency’, Poster presentation at the 53rd Annual Meeting of the Psychonomic Society, Minneapolis, MN, 17 November. Tao, Liang and Alice F. Healy (2013) ‘Word Identification in English and Chinese Texts by Speakers Varying in Fluency’, Poster presentation at the 54th Annual Meeting of the Psychonomic Society, Toronto, Canada, 16 November. Tao, Liang, Healy, Alice F., and Lyle E. Bourne, Jr. (1997) ‘Unitization in Second-language Learning: Evidence from Letter Detection’, American Journal of Psychology 110: 385–95. Tsai, Chih-Hao (2001) ‘Word Identification and Eye Movements in Reading Chinese: A Modeling Approach’, Available from: http://research.chtsai.org/dissertation. Tsai, Chih-Hao, George W. McConkie, and Zheng Xianjun (1998) ‘Lexical Parsing by Chinese Readers’, Poster presented at the Advanced Study Institute on Advances in Theoretical Issues and Cognitive Neuroscience Research of the Chinese Language, University of Hong Kong, Available from: http:// research.chtsai.org/papers/cht_asi.html. Tsao, Feng-fu (1979) A Functional Study of Topic in Chinese: The First Step Towards Discourse Analysis, Taipei: Student Book Co. Tzeng, Ovid J. L. and Daisy Hung L. (1981) ‘Linguistic Determinism: A Written Language Perspective’, in Ovid J. L. Tzeng and Harry Singer (eds) Perception of Print: Reading Research in Experimental Psychology, Hillsdale, NJ: Lawrence Erlbaum, 237–55. Van Dijk, Teun A. and Walter Kintsch (1983) Strategies of Discourse Comprehension, New York: Academic Press.

704

Psycholinguistics: Reading Chinese Wang, Li (2003)《汉语词的社会语言学研究》(A Sociolinguistic Analysis of Chinese Words), Beijing: The Commercial Press. Wang, William S.-Y. (1981) ‘Language Structure and Optimal Orthography’, in Ovid J. L. Tzeng and Harry Singer (eds) Perception of Print: Reading Research in Experimental Psychology, Hillsdale, NJ: Lawrence Erlbaum, 223–36. Wheeler, Daniel D. (1970) ‘Processes in Word Recognition’, Cognitive Psychology 1: 59–85. Zhang, Biyin and Peng Danling (1992) ‘Decomposed Storage in the Chinese Lexicon’, in Chen HsuanChih and Ovid J. L. Tzeng (eds) Language Processing in Chinese, Amsterdam: Elsevier, 131–50. Zhou, Youguang (1978)《现代汉字中声旁的标音功能问题》(‘On the Function of the Phonetic Component in Modern Chinese Characters’),《中国语文》(Studies of the Chinese Language) 146: 172–77.

705

39 Semantic change in Chinese Janet Zhiqun Xing Western Washington University, USA

1. Introduction In the early 1980s, when the study of grammaticalization began its revival, the study of a parallel subject, that of semantic change, also began attracting researchers’ attention (e.g. Sweetser 1990; Heine et al. 1991; Bybee et al. 1994; Traugott and Dasher 2002; Hopper and Traugott 2003; Eckardt 2006). As a result, we have gained a much better understanding of several key elements of semantic change, such as its nature, its cause, and its mechanisms, as well as its role and its relationship to grammaticalization and other (e.g. syntactic) levels of linguistic change. It appears that Western researchers are generally in agreement that the nature and the cause of semantic change are pragmatic and communicative-based. Of the mechanisms of semantic change, only two – metaphoricalization and metonymization – have been developed and popularized. Alas, these studies have focused predominantly on IndoEuropean languages; there is a general lack of comprehensive research on semantic change in genetically unrelated and typologically different languages such as Chinese. As a result, it is unclear, especially to the Western linguistic community, whether the general tendencies of semantic change found in Indo-European languages apply also to Chinese. This chapter intends to take a step toward rectifying this shortfall by exploring the patterns of semantic change in Chinese. It is well known that Chinese is typologically different from Indo-European languages in several ways: writing (i.e. logographic), phonology (e.g. tonal), morphology (e.g. lack of inflections and mostly compounds), and syntax and discourse (e.g. topic prominent and lack of agreement markers). By examining cases of semantic change in grammaticalization and factors that might cause semantic change in Chinese, this chapter aims to provide a comprehensive explanation of the characteristics of semantic change that are unique to Chinese as well as those that are commonly seen cross-linguistically. To understand the nature of semantic change, let us first examine the way Ancient Chinese scholars viewed the relationship among the three elements of language: sound, form and meaning. The earliest treatise on this issue was Shuōwén Jiězì 说文解字 ‘Discussion and Explanation of Characters’, written by the renowned Chinese grammarian Xu Shen in the second century ad. In this work, Xu developed a theory about the creation of Chinese characters, known as liùshū 六书 ‘Six Methods of Creating Characters’. Since then, Xu’s followers have made numerous attempts to interpret his theory, among which Duan Yucai’s Shuōwén Jiězì Zhù 说文解字注 ‘Annotation of Discussion and Explanation of Characters’ 706

Semantic change in Chinese

(1807) has probably been the most widely accepted. The following shows Duan’s (764) explanation of Xu’s theory of the creation order of the three elements of a word/character (meaning, sound, and form), as well as his thoughts on the inter-relationship among these three elements: Scholar Xu’s view is that sound is derived from meaning; meaning attaches to the sound. As far as wise men creating words is concerned, when there is a meaning, there is a sound; when there is a sound, there is a word or form. With regard to learning words, learners have to study the form to learn its sound and meaning. According to Xu, humans had ideas (i.e. meaning) first, and then paired those ideas with different sounds to make different words (i.e. forms). For the acquisition of words, however, Xu indicated that learners (presumably referring here to adult learners) studied the form first to acquire the sound and after they learned the sound, they understood the meaning. Evidently, to Xu, creating a word and learning a word require different processes, as diagrammed below: Creating a word: meaning>sound>form Learning a word: form>sound>meaning It appears that these two different processes proposed approximately 2,000 years ago are parallel to the process by which a word, phrase or sentence is produced by a speaker and interpreted by a listener. That is, when a speaker communicates with a listener, his/her initial purpose/goal of communication is to express a certain idea/concept. When this idea is formalized, the speaker identifies a medium (i.e. sound or form) to express the idea. Listeners/ readers, on the other hand, first perceive either a sound or a (written) form, through which they try to decode the meaning expressed by the sound or form as diagrammed below: Speaker: Meaning → Sound/Form Listener/Reader: Sound/Form → Meaning Assuming that the meaning initially expressed by the speaker is the same as the one interpreted by the listener/reader, we have the following result: [speaker] [sound/form] [listener/reader] Meaning1 → ------------------ → Meaning1 Notice that the first executor of the process is the speaker and the second one is the listener/reader. If, however, the initial meaning expressed by the speaker is not the same as that interpreted by the listener/reader, then we have the following outcome: [speaker] [sound/form] [listener/reader] Meaning1 → ------------------ → Meaning2 I argue that this is how meaning shift from M1 to M2 occurs. This view seems to be shared by some linguists in the West. Kiparsky (2008: 24) suggests that ‘the properties of language change might be explained by the way language is acquired and structured in the mind’, which coincides with the general consensus in the Western linguistic community that the 707

Janet Zhiqun Xing

quintessential reason for semantic change is communication and/or acquisition. What remains unexplained is whether typologically different languages require different conditions or employ different mechanisms when they undergo semantic change. This is the main focus of discussion in Section 2. In Section 3, evidence will be provided to show that semantic change in Chinese is regular but somewhat different from that of Indo-European languages such as English and German. Lastly, Section 4 will summarize the results of this study and draw some conclusions.

2.  Mechanism of semantic change Hopper and Traugott (1993: 86) following Heine et al. (1991) and Lehmann (1985) use the term ‘problem solving’ to describe two commonly discussed mechanisms of semantic change: Semantic change in general, not just grammaticalization, can be interpreted as problem solving. One problem to be solved is that of representing members of one semantic domain in terms of another, and metaphoric strategies serve this purpose. The second problem is the search for ways to regulate communication and negotiate speaker–hearer interaction. We have shown that this is a kind of metonymic change, indexing or pointing to meanings that might otherwise be only covert. Later (see Traugott and Dasher 2002) metaphoric change is referred to as metaphoricalization (henceforth Meta) and metonymic change as metononymization (henceforth Meto). Meta is then further defined as ‘primarily an analogical principle [that] involves conceptualizing one element of a conceptual structure Ca in terms of an element of another conceptual structure Cb’ (Traugott and Dasher 2002: 28), while Meto is defined as a contiguous relation between two conceptual structures/domains. Traugott and Dasher also suggest that Meto entails (inter)subjectification, ‘a change which results in the development of meanings that explicitly reveal recipient design: the designing of utterances for an intended audience at the discourse level’ (ibid: 31). Xing (2013: 228) adds another mechanism to her study of semantic change in grammaticalization in Chinese – semantic reanalysis, which she defines as ‘a process of semantic change whereby the interpretation of a given lexeme is triggered by language users’ dependence on its syntactic position and its relation to surrounding lexemes’. Unlike syntactic reanalysis, semantic reanalysis has not received much attention in cross-linguistic studies of grammaticalization. However, according to Xing, it is an indispensable mechanism in many cases of grammaticalization in Chinese. In what follows, I will demonstrate the roles of the three mechanisms, Meta, Meto and semantic reanalysis, in Chinese and compare them with those that have been reported from studies of other languages. It should be noted that neither Meta nor Meto are themselves processes of grammaticalization, but rather means for conceptualization within grammaticalization (Heine et al. 1991: 48). All the cases of Meta and Meto have a general tendency to shift meaning from a concrete concept to a more general one. In the grammaticalization literature, researchers of both Western languages (i.e. IndoEuropean) and Eastern languages (i.e. Chinese) agree that Meta and Meto play an essential role in semantic change and grammaticalization. Their views differ, however, on the role of Meto and its correlation with (inter)subjectification and most importantly on whether Meta and Meto can, in fact, account for all cases of semantic change in genetically and typologically different languages. Based on Indo-European languages, some Western researchers (Traugott 1989, 1995; Traugott and König 1991; Visconti 2004) claim that subjectification plays an increasingly important role in grammaticalization. Xing (2006) and others, on the other hand, 708

Semantic change in Chinese

provide evidence from studies of verbs in Chinese suggesting that (inter)subjectification, similar to other types of Meto, triggers semantic change throughout the process of grammaticalization at all levels: lexical, sentential and discourse. The examples in (1) from C. Liu (2012) demonstrate such a change at the lexical level. (Fourth Century

bc,

Guoyu)

(Fourth Century (b) 孔某曰: ‘来!吾语女……’ Kǒng-mǒu yuē lái wǔ yǔ rǔ NAME-certain say-that come I tell you ‘A certain Mr. Kong said: “Come here! Let me tell you  .  .  .” ’

bc,

Mozi)

(1) (a) 日,吾来此也,…… rì, wǔ lái cǐ yě other-day, I come here MOD ‘The other day when I came here  .  .  .’

(Third Century bc, Zhuangzi) (c) 虽然,若必有以也,尝以语我来! suí rán ruò bì yǒu yǐ yě cháng yǐ yǔ wǒ lái even-if so you must have reason MOD try by-it tell me come ‘Even so, you must have a reason for this. Would you try to tell me about it, please?’ According to C. Liu (2012), the three instances of lái 来 in (1) represent three of its many semantic functions in the process of its grammaticalization. In (1a), lái 来 expresses the original verbal meaning ‘to come’; in (1b) it is ‘calling others to be in the speaker’s location (hortative function)’, and in (1c) it indicates a ‘call to action’. Liu argues that these varied interpretations are supported by discourse, pragmatic and constructional evidence. It seems to me that this is a good illustration of how subjectification works at the lexical and non-lexical levels – that is the development of the newer meanings in (1b) and (1c) relies on the speaker’s/reader’s subjective interpretation or understanding of the context in which lái 来 appears. In other words, the imperative construction is used as a medium to assist the reader to first derive lái’s 来 lexical hortative meaning in (1b) and then its ‘call to action’ meaning in (1c). At the sentential level, historical texts often show the shift from lexical meaning to sentential meaning to be a result of subjectification as illustrated in (2)–(3). (2) (a) 我被府君密教收督邮。   (Third Century, Sanguo Zhi) wǒ bèi fǔjūn mìjiào shōu dūyóu 1SG receive palace-emperor secret-order detain provincial-officer ‘I received an emperor’s order to detain the provincial officer’ (Third Century, Sanguo Zhi) (b) 玚、祯各被太祖辟 Chàng Zhēn gè bèi tàizǔ bì name name each is-given emperor official-position ‘Chang and Zhen were each given an official position’ (Ninth Century, A Complete Collection of Tang Poems) (3) (a) 莫教人笑汝。 mò jiào rén xiáo rǔ not let people laugh-at you ‘Don’t let people laugh at you’ Or: ‘(You) should not be laughed at by people’ 709

Janet Zhiqun Xing

(b) 春思翻教阿母疑。 (Ninth Century, A Complete Collection of Tang Poems) chūnsī fān jiào āmǔ yí spring-thinking turn-around PASS mother suspect ‘Thoughts of love were nevertheless suspected by [her] mother’ The two bèis 被 in (2) have quite different semantic functions: the one in (2a) is used as a main verb meaning ‘receive’ and the noun phrase preceding it is the agent, whereas the one in (2b) still retains the active verbal meaning while allowing a passive reading. Such a shift from active lexical meaning in (2a) to a passive sentential interpretation in (2b) is not possible in Indo-European languages with morphosyntactic markings such as tense, voice, and agreement. This is clearly a change induced by a combination of the speaker/listener’s subjective perception of the context, namely subjectification, and the flexibility of interpretation of the noun phrase preceding the verb as the agent or the patient, a case of semantic reanalysis which will be discussed further later on. The same can be said of the two jiàos 教 in (3). At the discourse level, or the later stage of grammaticalization, subjectification is also seen to have played a role in semantic change in Chinese. What makes the change at this level different from the changes that occurred at the lexical or sentential levels is that, here, discourse-pragmatic meaning is developed through subjectification, as illustrated in (4). (Twelfth Century, Zhuzi Yulei) (4) (a) 此是连上文…说。 cǐ shì lián shàngwén .  .  .  shuō this is connect/include preceding-discourse .  .  . say ‘This is said to be associated with the preceding discourse’ (b) 众人轰然一笑，连贾珍也撑不住笑了。(Seventeenth Century, Honglou Meng) zhòngrén hōngrán yíxiào, lián Jiǎ Zhēn everybody burst-into laughing, including/even name yě chēngbúzhù xiào le also cannot-stand laughing asp ‘Everyone burst out laughing; even Jia Zhen laughed’ Both liáns 连 in (4) may convey the meaning ‘include/including’, yet the one in (4b) seems ‘plainly’ apparent if one simply interprets the sentence as ‘Everyone burst out laughing including Jia Zhen’. But upon reading the novel Honglou Meng ‘Dream of the Red Chamber’, it becomes clear that Jia Zhen is an important figure in the context where the sentence was used and, to reflect that importance, the emphatic meaning ‘even’ is added to the interpretation. As a result, lián’s discourse emphatic meaning is developed, a shift from (4a) to (4b) that is triggered by subjectification. In addition to Meta and Meto, Xing (2013) adds semantic reanalysis as a distinctive mechanism in grammaticalization. Although first coined by Eckardt (2006), semantic reanalysis, as defined by Xing (2013: 228), is quite different from Eckardt’s take on it. Xing defines it as ‘a process of semantic change whereby the interpretation of a given lexeme is triggered by language users’ dependence on its syntactic position and its relation to surrounding lexemes’. This study supports Xing’s analysis by citing two of her examples in (5) and (6). From these two examples, we see that if a lexeme is placed in the verb position or immediately after a negative marker for verbs, it is most likely to be interpreted (reanalyzed) as having a verbal meaning as shown in (5a) and (6a); if, however, a lexeme is placed immediately 710

Semantic change in Chinese

after a numeral, it is most likely to be interpreted as having a measure word or classifier meaning as shown in (5b–c) and (6b). (First Century, Shiji) (5) (a) 周公旦把大钺,毕公把小钺, Zhōugōngdàn bǎ dà yuè, Bìgōng bǎ xiǎo yuè Zhougongdan hold big weapon Bigong hold small weapon ‘Zhougongdan held a big weapon. Bigong held a small weapon’ (Fourth Century, Baiyujin) (b) 于此水底得一把毛。 yú cǐ shuǐ dǐ dé yī bǎ máo at this water bottom get one handful/MW hair ‘(Someone) got a handful of hair from the bottom of this water’ (Tenth Century, Bei Song Huaben Xuanji) (c) 只见两扇门关着,一把锁锁着 zhǐ jiàn liǎng shàn mén guān zhe, yī bǎ suǒ suǒ zhe only see two door close ASP one CL lock lock ASP CL ‘(They) only saw that the two-paneled door was closed and that there was a lock on it’ (6) (a) 秦将诈称二世使人遣李良书,不封 (First Century, Shiji) Qín jiàng zhà chēng èrshì shǐ rén qiǎn Qin chief cheat claim junior-emperor ask man dispatch Lǐliáng shū, bù fēng Liliang letter not seal ‘The Qin commander pretended to be the Emperor’s son to send a letter to Li Liang. (The letter) was not sealed’ (b) 子渊附书一封 Zǐyuān fù shū yī Ziyuan attach letter one ‘Ziyuan attached a letter’

fēng CL

(Third Century, Luoyang Jialan Ji)

When comparing the three usages of bǎ 把 in (5), we see that the sentence construction of (5a) clearly dictates the meaning of bǎ 把 as ‘to hold’, whereas the construction of (5b–c), with bǎ 把 in between a numeral yi ‘one’ and a noun máo 毛 ‘hair’ in (5b) and suǒ 锁 in (5c), leads the reader/listener to deduce the meaning as some kind of nominal modifier implying ‘a handful of’ and classifier for ‘hands-on’ objects. Similarly, the fēng 封 in (6a) and the fēng 封 in (6b) cannot be interpreted as having the same meaning. Since fēng 封 in (6a) follows a verbal negative marker bù 不, it is most appropriately interpreted as a verb meaning ‘to seal (an envelope)’. The fēng 封 in (6b), on the other hand, follows a numeral yi 一 ‘one’ and is, therefore, most logically interpreted as a classifier for the identification of the noun referent shū 书 ‘letter’. This kind of process of interpretation of the meanings of the bǎs 把 and fēngs 封, reflects the process of semantic reanalysis, which leads to the development of their polysemies. The question arises as to the necessity of semantic reanalysis in grammaticalization and whether the classifier meaning might have been extended via Meto instead of semantic reanalysis. Before addressing this question, a further exploration of the situation in which Meto operates is called for. According to Hopper and Traugott (2003), the key to Meto is that the source meaning and the metonymized meaning enjoy some kind of contiguous or 711

Janet Zhiqun Xing

associative relationship. One problem with this explanation is that it does not clearly specify the type of relationship that constitutes ‘contiguity’, and, as a result, it is very difficult, if not impossible, to know the scope of contiguous relationships they have in mind. For instance, we may say that all of the following pairs of concepts are interpreted as having a contiguous relationship: table/chair, pen/paper, employer/employee, father/son, root/trunk, computer/ table, to close one’s eyes/sleeping, to stand/up, to eat/food/full, to look/eye, to seat/bottom, to give/receive, able/disable, good/bad, often/rare, cellphone/hand, etc. Then, can we say that they are all in Meto-type relationships? Probably not, because even though they can be associated in one way or another, not all of them fall into the same class of instance, which is the key in the traditional understanding of metonymy, as defined by Cruse (2004: 209): Metonymy is responsible for a great proportion of the cases of so-called regular polysemy, where a parallel alternation of meaning applies over a class of instance, such as the TREE–WOOD readings of oak, ash, beech, pine, cherry, etc. With this understanding, the typical metonymical relationships are part for whole, whole for part, possessor for possessed, container for contained, represented entity for representative, place for institution, etc. Now, in examining the relationship between fēng’s action meaning ‘seal/sealing’ and the classifier meaning that modifies ‘letters’, we find that the action ‘sealing’ can apply to different types of objects (earth, tomb, wall, door, letter, etc.); however, the classifier function can only be used to classify enveloped objects like ‘letters’. This functional shift from the physical action ‘to seal/sealing’ to merely classifying enveloped objects does not belong to any of the aforementioned metonymized relationships. What is occurring here, instead, appears to be a process of semantic reanalysis, where fēng is used immediately after the numeral yī ‘one’ and the yī fēng phrase modifies, or rather classifies, the noun shū ‘letter’ located immediately preceding the phrase. Any other interpretation of the fēng phrase or sentence in (6c) (e.g. *Ziyuan sent/attach a letter a sealing) is neither sensible nor plausible. The three bǎ examples in (5) demonstrate a similar pattern with a slight variation in the process of semantic extension. In (5a), both bǎs are used as verbs expressing the meaning ‘to hold something with one’s hand’; in (5b) it cannot be interpreted as having a verbal meaning anymore, and instead is most plausibly interpreted as being a measure word meaning ‘handful’ due to the fact that it is in between a numeral yī and a noun máo ‘hair’ and to the fact that the whole phrase is used as the object ‘a handful of hair’ in the sentence. In (5c), bǎ is also in between a numeral, yī, and a noun, suǒ ‘lock’. Interestingly, this time it is most plausibly interpreted as a classifier modifying ‘a lock’, because other interpretations, either the nominal meaning ‘holding’, the verbal meaning ‘to hold’, or even the measure/adjective meaning ‘handful’, simply do not work with the rest of the elements in the sentence. Thus, we see two stages of semantic shift here: ‘to hold’ > ‘handful’ > classifier for ‘hands-on (objects)’. The first shift may be understood as Meto due to the conceptual and categorical association between ‘hold something with one’s hand’ and ‘handful’. As for the second shift, we argue that it is another case of semantic reanalysis, though it is unclear whether the classifier meaning was derived from the adjectival meaning ‘handful’ or from the verbal meaning ‘to hold’. What we see is that neither bǎ’s physical action meaning ‘to hold’ nor its quantifier meaning ‘handful’ has a direct metonymized relationship with its newly developed classifier meaning for hands-on objects such as ‘locks’. Here we are not saying that there is no association between them at all; rather, we are saying that there is no metonymized relationship between them. Think about it: there are many ways to classify 712

Semantic change in Chinese

‘locks’, such as by using shape or function, both of these typical ways classifiers have developed over time. Why use bǎ to describe the manner in which humans hold a lock? We believe that the classifier function was initiated, like other (semantic) changes, by Chinese language users who knew that the syntactic position of the lexeme in question would assist readers in correctly interpreting its semantic function, even though the newly developed meaning did not have an apparent association with its source meaning. Therefore, we conclude that the classifier meaning was triggered and construed through bǎ’s syntactic position and the plausibility of its meaning within the context, rather than through the conceptual association to bǎ’s earlier meaning. Traugott and Dasher (2002: 28–9) also expand the traditional understanding of Meto to the scope which they consider ‘a conceptual phenomenon’ by citing Stern’s (1968) association of mental state with its object or cause [concern (n.) ‘interest (in some matter)’ > ‘(the) matter that concerns]’. Such an association is closer to, but definitely different from, the relationships demonstrated in the examples in (5) and (6). Evidently, whether concern expresses a mental state or a matter that concerns, there is a clear metonymic association between them, whereas the action of ‘hold (with one’s hand)’ or ‘seal’ has no such a metonymic association with classifying ‘hands-on objects’ or ‘type of sealable object’ respectively. Hence, we say that the process of semantic reanalysis differs from the processes of Meta and Meto in that it is triggered by morphosyntactic reanalysis whereby a source lexical item can be interpreted as having a different syntagmatic function than its earlier function. In comparison, the other two mechanisms arise out of a certain cognitive construal – Meta is triggered by analogy and Meto by the same class of instance. In sum, it has been shown that, in addition to Meta and Meto, Chinese employs a third mechanism, semantic reanalysis, in semantic change and grammaticalization. Furthermore, contrary to what has been reported in studies of Indo-European languages, subjectification can occur at any stage or level of semantic change and grammaticalization in Chinese: the lexical, the sentential or the discourse level and not just at the later stage or discourse level. In the following section, evidence will be provided to show that accompanying those mechanisms, Chinese semantic change also demonstrates an accretive pattern of polysemies, another difference from what has been reported in Indo-European languages.

3.  Regularity of Semantic Change in Chinese According to studies by Western researchers, semantic change develops along the following cline: A>A,B>(B) in which the transition from a conceptual entity A to B undergoes an intermediate stage (A,B) where the preceding (A) and succeeding (B) entities coexist, before the older meaning (A) becomes restricted in register, and therefore recessive, and perhaps disappears completely (Heine et al. 1991: 74; and Traugott and Dasher 2002: 11–12). Such a tendency for semantic evolution seems to apply to most of the cases of grammaticalization in many Indo-European and African languages studied by Western researchers (cf. Heine et al. 1991; Traugott and Heine 1991; Bybee et al. 1994; Heine and Kuteva 2002; Good 2008). In Chinese, however, we see a somewhat different pattern. First let us look at Table 39.1, a summary of 23 lexemes1 and their polysemies developed in the course of their grammaticalization from verbs/nouns to function words. All the polysemies of these 23 lexemes can be found either in Cíyuán 词源 ‘Dictionary of Chinese Etymology’ or in monographs or articles (e.g. Jiang 2002; J. Liu 1989; Ma 1993, 2002; Peyraube 1989a, 1989b, 1989c, 1992, 1994; Peyraube and Li 2012; Z. Shi 1989; Y. Shi and Li 2001; Sun 1996; Wang 1980 [1957–8]; Xing 1994, 2003, 2012; Xu 1992; Zhang 1991; etc.). 713

Janet Zhiqun Xing Table 39.1  Summary of the polysemies of 23 lexemes Lexeme

Original meaning

Polysemies

liǎo 瞭/le 了

‘to complete’

zhuó 着/zhe 着 guò 过/guo 过

‘to adhere’ ‘to pass’

zài 在

‘to exist’

bǎ 把

‘to hold’

jiāng 将 bèi 被

‘to assist’*/  ‘commander’ ‘to cover’*/‘cover’

bĭ 比

‘to compare’

lián 连

‘to unite’

lìng 令 shĭ 使 jiào 叫 ràng 让 huì 会

‘to ‘to ‘to ‘to ‘to

děi/dé 得

‘to obtain’

lái 来 tóng 同

‘to come’ ‘to be the same as’

hé 和

‘to be in harmony’

gēn 跟 jí 及

‘heel’ (N) ‘to catch up’

gòng 共 jiù 就

‘to share/be same’ ‘to approach’

huán 还/hái 还

‘to return’

‘end/finish’; ‘understand’; ‘completely’; aspect marker; sentential particle ‘attach/catch’; ‘wear’; ‘reach’; ‘settle down’; aspect marker ‘pass (time)’ ; ‘miss’; ‘surpass’; ‘over’; ‘celebrate’; ‘mistake’ (noun); aspect marker ‘be there’; ‘rely on’, ‘due to’; ‘at (time)’ (prep.); ‘at (place)’ (prep.); aspect marker ‘take’ or ‘bring’*; ‘take hold of ’; ‘handful of ’ (noun); object marker/preposition ‘carry out’*; ‘support’*; ‘yield’*; ‘take’ ‘bring’*; object marker/preposition ‘surface’ (noun); ‘receive’*; ‘suffer’*; passive marker/ preposition ‘similar’; ‘compete/match’; ‘be next to’; ‘examples’ (noun)*; ‘continuously’ (adv.)* ; comparative marker/prep. ‘connect’ ; ‘consecutive’ (adj.); ‘consecutively’ (adv.); ‘and’, ‘with’ (prep.); ‘including’ (prep); emphatic marker ‘order’ (noun); ‘make’, ‘ask’ ; causative marker ‘make’; ‘use’ ; ‘officer’ (noun)*; causative marker ‘ask’; ‘let’; causative marker; passive marker/preposition ‘give away’; ‘allow’, ‘let’; causative marker ‘meet’; ‘understand’; ‘opportunity/gathering’ (noun); ‘be able’ (aux.); ‘be possible’ (aux.) ‘receive/get’; ‘be satisfied’; ‘can’, ‘may’ (aux.); ‘must’, ‘be obliged’ (aux.) ‘call someone to the come to the speaker’; ‘incite to action’ ‘similar (adj.)’, ‘similarly’ (adv.); ‘together’ (adv.); ‘similarity’ (noun); ‘with/and’ (prep.) ‘adjust’; ‘warm and smooth’; ‘peace/harmony’ (noun); ‘with/and’ (prep.) ‘follow’; ‘with/and’ (prep.) ‘reach’; ‘reach up to’; ‘at’, ‘up to a point (time)’ (prep.); ‘with/and’ (prep.) ‘same’; ‘together’ (adv.); ‘with/and’ (prep.) ‘take/accept’; ‘accomplish’; ‘then’ (adv.); ‘therefore’ (conj.); ‘if’ (conj.); ‘just/only’ (adv.) ‘recall’ ; ‘pay back’; ‘again’ (adv.); ‘still’ (adv.)

order’/‘order’ dispatch’ call’ yield’ gather’

* Polysemies not in use in contemporary Chinese.

Notice that, with the exception of gēn 跟 ‘heel’ and possibly bèi 被 ‘cover’, jiang 将 ‘commander’ and lìng 令 ‘order’, each of the lexemes listed in Table 39.1 developed at least two polysemies in the course of their grammaticalization from verb/noun to function word, i.e. adverb, preposition, conjunction, aspect marker, particle, emphatic marker, auxiliary, or causative marker. What makes their evolutions of meanings intriguing, compared with those in languages genetically unrelated to and typologically different from Chinese, is that almost

714

Semantic change in Chinese

all of those polysemies are still commonly used in contemporary Chinese either as free morphemes or as part of a compound. This seems to suggest that the tendency of semantic change in the grammaticalization of Chinese lexemes is a process of an accretion of meanings over time, namely, A>A,B>A,B,C, in which the multiplication from a conceptual entity A to A,B,C first undergoes an intermediate stage (A,B) where the old meaning (A) and the newer meaning (B) coexist, after which the older meaning (A) or the newer meaning (B) may become lexicalized or continue to develop and extend its meaning to (C) the newest conceptual entity. Intriguingly, all three entities may coexist or the older meaning(s) may become lexicalized. This cline, or rather the tendency of semantic change in Chinese, is clearly different from the one, A>A,B>(B), reported by Western researchers. The key difference between them rests on two precepts: (i) that a polysemy may become lexicalized in the course of grammaticalization; and (ii) that the coexistence of multiple conceptual entities (i.e. polysemies) is possible. Consequently, the cline in Chinese is accretive whereas that suggested by Western researchers forecasts a recessive nature and a possible disappearance of the older meaning completely (Traugott and Dasher 2002: 11).2 As noted in Table 39.1, except in the case of jiāng 将 and bèi 被, the lexemes’ original meanings (i.e. the older meaning) are used side by side with their newer meanings in modern texts. Examples of the usage of their older meanings in Modern Chinese are illustrated in (7). Even though some of the older meanings may only be used in compounds – instances of lexicalization – they clearly serve different functions than their newer meanings. I argue that since compounding is the most productive word-building method in Chinese (more than 80%, approximately, of Modern Chinese words are compounds), it is only natural for some lexemes to be used in compounds when they undergo grammaticalization and lexicalization. In fact, being used in compounds appears to enhance the continuity of the conceptual entity in language use, which I will discuss further later. (7) (a) 了 liǎo ‘to complete’ (b) 着 zháo ‘to catch’ (c) 过 guò ‘to pass’ (d) 在 zài ‘to exist’ (e) 把 bǎ ‘to hold’ (f) 比 bǐ ‘to compare’ (g) 连 lián ‘to connect’ (h) 令 lìng ‘order’ (i) 使 shǐ ‘to dispatch’ (j) 叫 jiào ‘to call’ (k) 让 ràng ‘to yield’ (l) 会 huì ‘to meet’ (m) 得 dé ‘to obtain’ (n) 来 lái ‘to come’ (o) 同 tóng ‘same’ (p) 和 hé ‘to be in harmony’ (q) 跟 gēn ‘to follow’ (r) 及 jí ‘to reach’ (s) 共 gòng ‘to share’ (t) 就 jiù ‘to approach’ (u) 还 huán ‘to return’

as in 了了一桩心事 ‘finish taking care of one thing’ as in 着火了 ‘caught fire’ as in 过日子 ‘pass day/to live’ as in 在不在 ‘is there or is there not’ as in 把着东西不放 ‘hold something tight’ as in 比一比 ‘compare a bit’ as in 连在一起 ‘to connect together’ as in 下令 ‘to give an order’ as in 当佣人使 ‘use as a servant’ as in 你在叫我吗？‘Are you calling me?’ as in 不能让让他 ‘cannot yield to him’ as in 不能聚会 ‘cannot get together’ as in 得到 ‘got something’ as in 他来了 ‘He came’ as in 相同 ‘similar to each other’ as in 这两样不和 ‘These two kinds do not go together’ as in 别跟着我 ‘don’t follow me’ as in 不及格 ‘not reach the cut-off line’ as in 公共场所 ‘public-shared places’ as in 就近上学 ‘go to school nearby’ as in 还了书 ‘returned books’

715

Janet Zhiqun Xing

We see that all the lexemes above strongly follow the accretive cline while bèi 被 and jiàng 将 only cling weakly to it. Interestingly, these two even now retain their original nominal meanings which have also become lexicalized and are used along with their grammaticalized meanings in contemporary Chinese as in bèizi 被子 ‘comforter’ and in jiàngcái 将才 ‘commander’ or shàngjiàng 上将 ‘general commander’. In fact, when examining the usage of the polysemies of the 23 lexemes developed in the course of their grammaticalization, it becomes apparent that many of them have become lexicalized into disyllabic words. It is well documented in Chinese linguistic literature (cf. Peyraube 1988; J. Liu et al. 1995; Feng 1999; Dong 2002) that disyllabic words and serial verb constructions emerged during the Han dynasty (Second Century bc – Second Century ad) and became well established in the Tang Dynasty. Many polysemies of the 23 lexemes studied were, in fact, used in lexicalized compounds after the Six Dynasties (Fourth Century ad), a period that witnessed two major grammatical developments in the Chinese language: a rapid increase of disyllabic words and the emergence of the verb+complement construction (cf. Feng 2005; and Wei 2008). Taking guò 过 as an example, we find that many of guò’s 过 polysemies are used in lexicalized compounds: Char.

pinyin

gloss

English

过活 过去 过火 过节 过门儿 过错 过目 过失 错过 难过 不过 罪过

guòhuó guòqù guòhuǒ guòjié guòménr guòcuò guòmù guòshī cuòguò nánguò búguò zuìguò

pass life pass go over fire celebrate festival pass door over mistake pass eye pass miss miss pass difficult pass not pass crime mistake

‘to live’ ‘to go over’ ‘overdone’ ‘to celebrate a festival’ ‘to marry into another family’ ‘fault’ ‘to look over’ ‘wrong doing’ ‘to miss’ ‘sad’ ‘but’ ‘sin’

Notice that some of these compounds have inherited guò’s original meaning ‘to pass’, whereas others were derived from guò’s later developed polysemies ‘over, mistake, celebrate’, each of which can be used as a free morpheme or in a lexicalized compound in Modern Chinese. It appears that these lexicalized usages not only provide evidence for guò’s polysemous functions but also reinforce those functions in modern communication. In other words, once guò is compounded with another lexeme in a verb+object or verb+complement combination, and once that becomes a frequently used compound, it is unlikely to undergo further immediate change and is likely, instead, to retain its function for a long time before it becomes the target of yet another change. Why does Chinese semantic function evolve in the manner of accretion of meanings over time? Text analysis reveals that this can primarily be attributed to the morphological, syntactic and discourse structure of the Chinese language. At the morphological level, all Chinese characters have one immutable written form beyond which it simply cannot be reduced unless it is systematically simplified through language reforms. Furthermore, two or more characters/words, especially those that have undergone grammaticalization, cannot be combined into one character/word3 as they can in alphabetical languages (e.g. ‘going to > gonna’; ‘you all > y’all’; ‘no one > none’ in English). And, most 716

Semantic change in Chinese

importantly, there are no markings for parts of speech or agreement in Chinese as pertains to nouns or verbs or number, case, gender, tense, voice, etc.4 As a result, the SAME noun form (character) can be used as subject/agent or object/patient and the SAME verb form (character) can be used as a main verb, serial verb, adverb, adjective, conjunction, or preposition. Moreover, a verb can also be used as a noun (cf. Shen 2012 for categorization of verbs in Chinese) and vice versa, e.g. huì 会 and lián 连 as illustrated in (8) and (9). In comparison, nouns and verbs in English (especially Old English) and German behave completely differently, as illustrated by German können ‘can’ in (10) and by English even in example (11). Admittedly, words in both German and English may be converted into different parts of speech, yet the majority of them in both languages are marked either by prefixes or suffixes indicating the part of speech and grammatical function (tense, voice, case, etc.) or by stress (perMIT, verb; PERmit, noun).  (8) huì’s 会 various functions (see Xing 2013 for more detailed examples): (a) NOUN: ‘meeting’ as in 两个会 ‘two meetings’ (b) VERB: ‘to meet’ as in 会一下老朋友 ‘to meet old friends’ (c) SERIAL VERB: ‘meet + verb’ as in 会谈 ‘to meet and talk’ (d) AUXILIARY: ‘might’ as in 会下雨 ‘may rain’ (e) SERIAL VERB/COMPLEMENT: ‘verb+perceive’ as in 学会游泳 ‘to learn to swim’  (9) lián’s 连 various functions (see Xing 2006 for more detailed examples): (a) VERB: ‘to connect’ as in 连起来 ‘to connect together’ (b) VERB/ADJECTIVE: ‘continuing/continuous’ as in 连声叫好 ‘praise it continuously’ (c) ADVERB: ‘continuously’ as in 连忙站起来 ‘stand up in a hurry’ (d) PREPOSITION: ‘including/with’ as in 连人带车 ‘including people and vehicle’ (e) CONJUNCTION: ‘even’ as in 连老师也来了 ‘even the teacher came’ (10)

können’s conjugation: (a) Ich/er kann/konnte kommen. ‘I/he can/could come’ (b) Sie kannst/konnste kommen. ‘You can/could come’ (c) Sie kommen können/könnte. ‘They can/could come’

(11) even’s various functions (the examples are all quoted from OED): (a) adjective: Eeuen/eauen: flat (of a land/ground), smooth, direct, exact, equal Me thinkes the ground is eeuen. (1605, Shakespeare’s Lear) (b) verb: Euenen: to level (ground); to even out Law, whose end is, to euen and right all things (1851, Sidney’s Apol) He hath now euened all his reckonings  .  .  . (1664, Pepys’ Diary) (c) adverb: Euene/evene/even: evenly as the rest goes euen (1601, Shakespeare’ Twel) 717

Janet Zhiqun Xing

The illustrative examples of huì’s and lián’s usages in (8)–(9) clearly show that, regardless of huì’s or lián’s part of speech, they have retained their one immutable form. In comparison, German können in (10) and English even in (11) change their part of speech or form based on tense, aspect, voice or their subject’s number, case, and gender. This leads us to believe that words without part-of-speech markings, such as those in Chinese, have a higher degree of flexibility of interpretation in a sentence than those in languages with part-of-speech markings and it is this that consequently leads to the accretive pattern of polysemies in Chinese. Another angle at which to look at the typological tendency of semantic change in Chinese as opposed to Indo-European languages is to compare the grammatical markers in those languages. Chinese does not have agreement markers or verb conjugations whereas IndoEuropean languages do. In other words, the semantic and pragmatic relationships among syntactic units (i.e. NP + VP + NP + NP) in Chinese are not as evident as they are in languages with form changes that show agreement and conjugation. As a result, an NP preceding a verb (e.g. bèi 被 or jiào 叫) can be semantically reanalyzed and interpreted as either an agent or a patient depending upon the meaning of the main verb (prototypical verbs in a passive construction expressing some kind of adversity meaning) and the overall meaning of the sentence, as shown in (12)–(13). Such flexibility of NP interpretation allows the verb in a sentence to be reanalyzed and coerced into an interpretation that fits the contextual meaning of the whole sentence, a typical instance of semantic reanalysis. (12) (a) Noun preceding bèi 被 as an agent (Twelfth Century, Zhuzi Yulei) 月被其光而明。 yuè bèi qí guāng ér míng moon receive its light then bright ‘The moon becomes bright when receiving its light’ (b) Noun preceding bèi 被 as a patient 阿娘被问来由,不觉心中欢喜。(Seventh Century, Bianwen) ā-nián bèi wèn lái-yóu, bùjué xīnzhōng huānxǐ nanny pass ask come-reason, not-feel heart happy ‘When the nanny was asked the reason, she could not help feeling happy’ (13) (a) Noun preceding jiào 叫 as an agent 他爸爸叫他修理一下那辆自行车。 tā bàba jiào tā xiūlǐ yíxià nà 3sg father ask 3sg fix bit det ‘His father asked him to fix the bicycle’ (b) Noun preceding jiào 叫 as a patient 他叫那条狗咬了一口。 tā jiào nà tiáo gǒu yǎo le 3sg pass det cl dog bite asp ‘He was bitten by that dog’ Or: ‘He got bitten by that dog’

liàng cl

yī one

zìxíngchē bicycle

kǒu bite

In (12a–b), neither bèi nor the noun in the subject position is marked, which leaves room for them to be semantically reanalyzed and reinterpreted: the bèi in (12a) has an active meaning ‘to receive’ and its corresponding subject functions as an agent whereas the bèi in 718

Semantic change in Chinese

(12b) along with another verb has a passive meaning ‘to be asked’ and its corresponding subject functions as a patient. This is true, too, of the two jiàos in (13). Evidently, this kind of interpretation is not possible in languages with clear markings for tense, voice, and agreement with number, gender and case, as in German. At the discourse level, Chinese sentences (including both Classical and Modern Chinese) have consistently been arranged by logical relations, such as sequential order and cause/ reason-result as illustrated by jiù’s functions in (14). This kind of ‘fixed’ discourse structure, a characteristic of isolating and analytic languages, coincides with Kiparsky’s reflection (2008: 24) on the characteristic of inflectional languages when they move in the direction of analytical languages – ‘the loss of inflectional morphology entails fixed order of direct nominal arguments’. I argue that the fixed discourse structure in Chinese plays an important role in the development of jiù’s 就 two functions, namely, its discourse function of connecting two sequential events as in (14a)–(14b) and its function of connecting two logically related events as shown in (14c)–(14d). Certainly, jiù’s lexical meaning ‘to approach’ may also be a contributing factor of its discourse function; however, Chinese discourse structure appears to be instrumental in the development of its polysemous functions. (14) (a) Sequential: …,欲遣就师。(Third Century, Sanguo Zhi) .  .  .  , yù qiǎn jiù shī .  .  . want dispatch approach master ‘(Someone) wants to approach the master’ (b) Sequential: 虎贲舁上殿就坐。(Third Century, Sanguo Zhi) hǔbēn yú shàng diàn jiù zuò commander carry up throne approach/then seat/sit ‘The commander was carried up to the throne to sit down’ (c) Condition–Result: 施薪若一,火就燥也; (First Century, Xunzi) shī xīn ruò yī, huǒ jiù zào yě; add hay like one, fire then dry part; ‘(If you) add one piece of hay, the fire will die;’ (d) Reason–Result/sequential: 资复逊位归第,就拜骠骑将军, (Third Century, Sanguo Zhi) zīfù sùnwèi guī dì, jiù pài biáoqí jiāngjūn leader resign return home, then pay-visit horse-riding commander ‘(If/when) the leader resigns and returns home, then (he) will pay a visit to the commander’ Notice that except for jiù in (14a), all the remaining usages of jiù in (14b–d) can be considered sequential/logical conjunctions or space-fillers. That is, jiù’s meaning coincides with the discourse structures of the Chinese language – sequential and logical. This is evidenced by the fact that even without jiù, language users do not have a problem understanding the given sentence/discourse. In comparison, lexemes from languages with different sentential/ discourse structures are less likely to undergo a similar path of semantic reanalysis and develop a conjunction meaning in the way that jiù has. 719

Janet Zhiqun Xing

It is evidenced that all three levels, morphology, syntax, and discourse, provide space for a different sort of reanalysis/reinterpretation/re-use of Chinese lexemes. As a result, various polysemies have developed and now coexist in modern texts. This aspect is supported by Bisang’s (2008: 586) study of Archaic Chinese showing that a lexical item in a given position is coerced into a particular semantic interpretation associated with that position. Certainly, if Chinese had grammatical/agreement markers on nouns and verbs, the changes in function (that is, nouns functioning as verbs or being interpreted as either agent or patient; verbs being interpreted as adjectives, adverbs prepositions or conjunctions) would likely not be accomplished as easily.

4. Conclusion So far we have discussed the tendencies and mechanisms of semantic change in grammaticalization in Chinese and the differences between Chinese patterns of change and those in Indo-European languages. It has been shown that although researchers generally agree that semantic change is pragmatic in nature and triggered by communicative need, evidence from case studies of semantic change and grammaticalization in Chinese suggests that mechanisms of semantic change in Chinese differs from those of Indo-European languages. We have seen that in addition to Meta and Meto, semantic reanalysis is an indispensable mechanism in semantic change and grammaticalization in Chinese. While based on previous studies of IndoEuropean languages, it appears that semantic reanalysis is not recognized as being important as Meta and Meto. I argued that this disparity is the result of the different typological characteristics of Chinese and Indo-European languages, namely that Chinese nouns and verbs have no markings and conjugations and are, therefore, highly flexible in their interpretation. As a result, it is fairly easy for a Chinese lexeme that undergoes semantic change and grammaticalization to develop various polysemies that are retained. Indo-European languages, on the other hand, with their clear markings indicating parts of speech, their different internal structures (phonological, morphological, sentential and discourse), their different historical developments and reforms (e.g. the Great Vowel Shift and the loss of the case system in English, and a large amount of borrowing from other languages or within their language family) make it difficult for a lexeme to develop polysemies and retain them. For these reasons, the pattern of semantic change can be seen as recessive in Indo-European languages, but as accretive in Chinese.

Notes 1 Some of the lexemes in Table 39.1 have two phonological forms. It should be noted that, generally speaking, the longer lexical form corresponds to its verbal meaning, while the shorter corresponds to its grammatical meaning. 2 The English word ‘have’ seems to behave more like the Chinese lexemes in Table 39.1 that have undergone grammaticalization than like those that follow the evolutionary cline suggested by Western researchers. In modern times, ‘have’ has multiple semantic functions: 1) possessive verb – ‘I have a computer’; 2) perfective aspect – ‘I have bought a computer’; and 3) modal auxiliary – ‘We have to go’. 3 This does not include those words that were created by combining two characters such as bú zhèng 不正 ‘not straight’ for wāi 歪, 女子 nǚ zǐ ‘female child’ for hǎo 好 ‘good’. 4 The only plural marking in Chinese is the suffix mén 们, which attaches to singular pronouns, wǒ 我 ‘I’, nǐ 你 ‘you’, and tā 他/她 ‘he/she’ among others (e.g. péngyǒumen 朋友们 ‘friends’); the only gender marking in Chinese is the third person singular pronouns tā 他/她 ‘he/she’.

720

Semantic change in Chinese

References Bisang, Walter (2008) ‘Precategoriality and Syntax-based Parts of Speech: The Case of Late Archaic Chinese’, Studies in Language 32(3): 568–89. Bybee, Joan, Revere Perkins, and William Pagliuca (1994) The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World, Chicago: University of Chicago Press. Cruse, Alan (2004) Meaning in Language: An Introduction to Semantic and Pragmatics, Oxford: Oxford University Press. Dong, Xiufang (2002) Cíhuìhuà: Hànyǔ Shuāngyīn Cí de Yánshēng hé Fāzhǎn (Lexicalization: The Development of Chinese Disyllabic Words), Chengdu: Sichuan Mingzu Chubanshe. Duan, Yucai (1807/1981) Shuōwén Jiězì Zhù (Annotation of Discussion and Explanation of Characters), Shanghai: Guji Chubanshe. Eckardt, Regine (2006) Meaning Change in Grammaticalization: An Enquiry into Semantic Reanalysis, Oxford: Oxford University Press. Feng, Shengli (1999) Hànyǔ Yùnlǜ Jùfǎ Xué (Prosodically Constrained Syntax in Chinese), Shanghai: Jiaoyu Chubanshe. Feng, Shengli (2005) Hànyǔ Yùnlǜ Jùfǎ Yánjiū (Studies of Prosodically Constrained Syntax in Chinese), Beijing: Peking University Press. Good, Jeff (ed) (2008) Linguistic Universal and Language Change, Oxford and New York: Oxford University Press. Heine, Bernd, Ulrike Claudi, and Friederike Hünnemeyer (1991) Grammaticalization: A Conceptual Framework, Chicago: University of Chicago Press. Heine, Bernd and Tania Kuteva (2002) World Lexicon of Grammaticalization, Cambridge: Cambridge University Press. Hopper, Paul J. and Elizabeth C. Traugott (1993/2003) Grammaticalization, Cambridge: Cambridge University Press. Jiang, Shaoyu (2002) ‘Gěi zì jù, jiào zì jù de bèdòng láiyuán: jiān tán yǔfǎhuà, lèituī gōngnéng kuòzhǎn (‘Origin of the Passive gei and jiao Constructions: Discussion on Grammaticalization, Analogy and Extension’), Yuyanxue Luncong (Linguistic Review), Beijing: Shangwu Yinshuguan, 26. Kiparsky, Paul (2008) ‘Universals Constrain Change; Change Results in Typological Generalizations’, in Jeff Good (ed) (2008) Linguistic Universal and Language Change, Oxford and New York: Oxford University Press, 23–53. Lehmann, Christian (1985) ‘Grammaticalization: Synchronic Variation and Diachronic Change’, Lingua e Stile 20: 303–18. Liu, Cheng-hui (2012) ‘The Grammaticalization of the Directional Verb Lai: A Construction Grammar Approach’, in Janet Xing (ed) The Newest Trend in the Study of Grammaticalization and Lexicalization in Chinese, Berlin: Mouton de Gruyter, 87–114. Liu, Jian (1989) ‘Shìlùn hé zì de fāzhǎn, fùlùn gòng zì hé lián zì (‘The Development of hé ‘and’ and the Development of gòng ‘together’ and lián ‘even’), Zhongguo Yuwen (Studies of the Chinese Language) 6. Liu, Jian, Cao Guangshun, and Wu Fuxiang (1995) ‘Lùn yòufā hànyǔ cíhuì yǔfǎhuà de ruògān yīnsù (‘On Factors Contributing to Lexicalization and Grammaticalization’), Zhongguo Yuwen (Studies of the Chinese Language) 3. Ma, Beijia (1993) ‘Jiècī tóng de chǎnzhēng (‘The Emergence of the Preposition tong ‘same as’), Zhongguo Yuwen (Studies of the Chinese Language) 2. Ma, Beijia (2002) Jìndài Hànyǔ Jiècí (Prepositions in Modern Chinese), Beijing: Zhonghua Shuju. Peyraube, Alain (1988) ‘Syntactic Change in Chinese: On Grammaticalization’, Bulletin of the Institute of History and Philology 59(3): 617–52. Peyraube, Alain (1989a) ‘History of the Comparative Construction in Chinese from the 5th Century bc to the 14th Century ad’, Proceedings on the 2nd International Conference on Sinology, Taipei: Academia Sinica, 589–612. Peyraube, Alain (1989b) ‘History of the Passive Constructions in Chinese until the 10th Century’, Journal of Chinese Linguistics 17(2): 335–71. Peyraube, Alain (1989c) ‘Zǎoqī bǎ zì jù de jǐge wèntí’ (‘Several Questions on the Early bǎ Construction’), Zhongguo Yuwen (Studies of the Chinese Language) 1. Peyraube, Alain (1992) ‘History of Some Coordinative Conjunctions in Chinese’, Paper presented at the First International Conference on Chinese Languages and Linguistics, Singapore, 24–26 June, National University of Singapore.

721

Janet Zhiqun Xing Peyraube, Alain (1994) ‘On the History of Chinese Locative Prepositions’, Chinese Languages and Linguistics 2: 361–87. Peyraube, Alain and Ming Li (2012) ‘Semantic Historical Development of Modal Verbs of Volition in Chinese’, in Janet Xing (ed) Newest Trends on the Study of Grammaticalization and Lexicalization in Chinese, Berlin: Mouton de Gruyter, 149–68. Shen, Jiaxuan (2012) ‘Míngdòngcí de fǎnsī: Wèntí hé duìcè’ (‘Reflections on Deverbal Nouns: Problems and solutions’), World Chinese Teaching 1. Shi, Yuzhi and Na Li (2001) Hànyǔ Yǔfǎhuà de Lìchéng (The Process of Grammaticalization in Chinese), Beijing: Peking University Press. Shi, Ziqiang (1989) ‘The Grammaticalization of the Article le in Mandarin Chinese’, Language Variation and Change 1: 99–114. Sun, Chao-fen (1996) Word Order Change and Grammaticalization in the History of Chinese, Stanford: Stanford University Press. Sweetser, Eve E. (1990) From Etymology to Pragmatics, Cambridge: Cambridge University Press. Traugott, Elizabeth C. (1989) ‘On the Rise of Epistemic Meanings in English: An Example of Sub­ jectification in Semantic Change’, Language 57: 33–65. Traugott, Elizabeth C. (1995) ‘Subjectification in Grammaticalisation’, in D. Stein and S. Wright (eds) Subjectivity and Subjectification, Cambridge: Cambridge University Press, 31–54. Traugott, Elizabeth C. and Richard B. Dasher (2002) Regularity in Semantic Change, Cambridge: Cambridge University Press. Traugott, Elizabeth C. and Bernd Heine (eds) (1991) Approaches to Grammaticalization, vol. I and II, Amsterdam and Philadelphia: John Benjamins. Traugott, Elizabeth C. and Ekkehard König (1991) ‘The Semantic-Pragmatics of Grammaticalization Revisited’, in Elizabeth C. Traugott and Bernd Heine (eds) (1991) Approaches to Grammaticalization, vol. 1, Amsterdam and Philadelphia: John Benjamins, 189–218. Visconti, Jacqueline (2004) ‘Conditionals and Subjectification: Implications for a Theory of Semantic Change’, in Olga Fischer, Muriel Norde, and Harry Perridon (eds), Up and Down the Cline – The Nature of Grammaticalization, Amsterdam and Philadelphia: John Benjamins, 169–192. Wang, Li (1957–8/1980) Hànyǔ Shǐgǎo (Chinese Grammar), Beijing: Zhonghua Shuju. Wei, Zhaohui (2008) Shànggǔ Hànyǔ Liándòngshì Yánjiū (Study of the Serial Verb Construction in Old Chinese), Shanghai: Sanlian Shudian. Xing, Janet Z. (1994) ‘Diachronic Change of Object Markers in Mandarin Chinese’, Language Variation and Change 6(2): 201–22. Xing, Janet Z. (2003) ‘Grammaticalization of Verbs in Mandarin Chinese’, Journal of Chinese Linguistics 31(1): 101–44. Xing, Janet Z. (2006) ‘Mechanism of Semantic Change in Chinese’, Studies in Language 30(3): 461–83. Xing, Janet Z. (ed) (2012) The Newest Trend in the Study of Grammaticalization and Lexicalization in Chinese, Berlin: Mouton de Gruyter. Xing, Janet Z. (2013) ‘Semantic Reanalysis in Grammaticalization in Chinese’, in Zhuo Jing-Schmidt (ed) Increased Empiricism: New Advances in Chinese Linguistics, Amsterdam and Philadelphia: John Benjamins, 223–46. Xu, Dan (1992) ‘Hànyǔ lǐ de zài hé zhe’ (‘The Aspect Markers zài and zhe in Chinese’), Zhongguo Yuwen (Studies of the Chinese Language) 6. Zhang, Wangxi (1991) ‘Bǎ zì jiěgòu de yǔyì jíqí yǔyòng fēnxi’ (‘Semantic and Pragmatic Functions of the bǎ Construction’), Yuyan Jiaoxue yu Yanjiu (Language Teaching and Research) 3.

722

40 Standard Chinese Shi Dingxu Hong Kong Polytechnic University, Hong Kong, China

The concept of Standard Chinese The term Standard Chinese commonly refers to a number of different but related linguistic forms. One of them is Putonghua ‘the common language’, the official Standard Spoken Chinese being promoted by the government of the People’s Republic of China since the 1950s. Another one is the written form accepted by the majority of Chinese people and being taught in schools in all Chinese-speaking communities. There are also a number of related terms. They include Guoyu ‘national language’, which refers to the Standard Spoken Chinese in Taiwan nowadays and in the mainland before 1949; Huayu ‘language of the Chinese people’, which is used in Singapore for Standard Chinese; Mandarin ‘language of the general’, which refers to the dialect of Beijing; as well as Guanhua ‘the language of the officials’, which is typically the dialect of the capital of various empires. Another term is Classical Chinese, which refers to the dominant written form for formal genres before the early twentieth century.

1.  Standard Spoken Chinese For thousands of years before the birth of the Republic of China, the Standard Spoken Chinese was known as Yayan ‘elegant language’ or Guanhua ‘the language of the officials’ and had mainly been the dialect of the capital city, where officials with different dialect backgrounds converged. When they had to communicate with each other with whatever means they had, it was quite natural that the local dialect became the first choice. A Guanhua would lose its status to another dialect when one empire was replaced by another and the capital was moved to a different city. There were some exceptions, and a typical case is the dialect of Hangzhou. When the Northern Song dynasty was threatened by the Jin, it moved the imperial court from Kaifeng to Hangzhou to establish the Southern Song dynasty. The dialect of the old capital went along with the royal family, the officials, the court people and the soldiers. The Kaifeng Guanhua settled down in Hangzhou and merged with the local dialect. The newly formed Hangzhou Guanhua inherited features from the local dialect, which should be a member of the Wu dialect family, as well as features from the incoming Guanhua, which should belong to the Northern dialect. Up till today, Hangzhou dialect still has a lot of features not found 723

Shi Dingxu

in surrounding Wu dialects, such as the diminutive suffix –er attached to nominal elements, and the Hangzhou area is still a dialect island. To a certain extent, even Mandarin was formed in such a way. When the third emperor of the Ming dynasty moved the capital from Nanjing to Beijing, he not only brought the whole imperial court with him, but also drafted a huge number of craftsmen and other professionals to settle down in the new capital. Those from the Nanjing area spoke Lanqing Guanhua ‘imperfect official language’, a proximate of the Nanjing dialect, while other people spoke whatever dialects they had. The dialect of the new capital was thus a mixture. When the Qing dynasty replaced the Ming, they also moved their royal court from the Northeast to Beijing, together with the Northeastern Chinese dialect in addition to the Manchu language. The Beijing Guanhua, namely, Mandarin, thus obtained some Northeastern features. The Qing nobles gradually gave up Manchu to adopt Chinese as their spoken language and thus brought some Manchu elements into Mandarin. The Mandarin as we know it took shape around the mid 1850s.

1.1.  Guoyu ‘national language’ At the beginning of the twentieth century, the idea of having a Standard Spoken Chinese other than the dialect of the capital city emerged and the term Guoyu ‘national language’ was coined to represent the thinking that it should be for all Chinese speakers. A campaign by elite intellectuals in the 1910s led to the establishment of Duyin Tongyi Hui ‘the Committee on the Standardization of Pronunciation’ by the Ministry of Education in 1912. The Committee consisted of experts appointed by the Ministry of Education as well as representatives from each province. After much debate and via a voting process in which each province had one vote, the Committee made a decision on the standard pronunciation of 6,500 Chinese characters. By standardizing the pronunciation of a large number of characters, the Committee had actually defined the sound system of Guoyu and its phonological system. Mainly due to the utopian belief that the national language should incorporate characteristics from as many dialects as possible and should preserve the major features of historic Chinese phonology as presented in Li (1728), the first version of Guoyu was a compromise which resembled each major dialect of Chinese in certain aspects but also differed from each of them in a significant way. A typical case was that the rusheng ‘entering tone’ was included in the Guoyu tonal system, even though the four-tone system of the Northern dialect was adopted as the norm. A syllable bearing an entering tone is typically in the shape of a single vowel or a combination of a consonant and a vowel, and its contour is comparable to a corresponding syllable with a non-entering tone. Such a syllable, however, is shorter than the corresponding one because it takes an added coda, which is an unreleased plosive [ph], [th] or [kh] in Cantonese and Hakka, and is a glottal stop in the Wu dialect. It is generally believed that entering tones were widely spread in Ancient Chinese, but they gradually disappeared after the Yuan dynasty from the major Northern dialects and most of the Xiang and Gan dialects. Another case was that the Wu dialect voiced fricative [v] was included in the sound inventory of Guoyu, but the Wu voiced plosives and affricatives were excluded, while the Northern dialect retroflexes [tʂ], [tʂh], [ʂ] and [ʐ ] were included even though they were not found in most Southern dialects such as Wu. Simply put, the tonal system of the first version of Guoyu was based on the four-tone system of the Northern dialect but the entering tone of Southern dialects was added. The 724

Standard Chinese

result was a five-tone system which no real dialect had. The first Guoyu sound system had some sounds only found in Northern dialect but also contained a Wu dialect sound. Such a Guoyu was so far away from any genuine dialect that no one except a few well-trained linguists could actually speak it. Seeing all these difficulties, the Committee made another decision in the early 1920s to adopt a concrete dialect as the base line for Guoyu and the dialect of the capital city at that time, namely Beijing Mandarin, was chosen. The 1912 version was named lao guoyin ‘old national pronunciation’, while the new one was called xin guoyin ‘new national pronunciation’. The Ministry of Education made an official declaration in 1932, making Beijing Mandarin the base line for Guoyu and ending the debate about old and new national pronunciation. This decision has influenced almost all the decisions made afterwards. The Guoyu being promoted in Taiwan is in essence the second version. The government was a strong supporter of the Guoyu campaign and the Ministry of Education issued several decrees in the 1920s to make Guoyu an integrated part of the education system, especially that of teacher training. It published quite a number of textbooks on Guoyu, several journals to promote Guoyu and a few introductory books on Guoyu. It also set up special schools and short training programs all over China to train Guoyu teachers.

1.2.  Zhuyin Zimu ‘Phonetic Alphabets’ Along with the creation and promotion of Guoyu, several writing schemes were designed to record it and to help people learn it. One of them is Zhuyin Zimu ‘Phonetic Alphabets’, which was adopted by the Committee on the Standardization of Pronunciation in 1913 and announced to the public by the Ministry of Education in 1918. It was based on a characterlike alphabet designed by Taiyan Zhang at the end of the nineteenth century and was improved by adding a number of symbols based on components of Chinese characters. There were 39 symbols in the Zhuyin Zimu system and the symbols were divided into three groups known as shengmu ‘consonant letter’, jiemu ‘glide letter’ and yunmu ‘vowel letter’. This was by and large adapted from the tradition of Chinese phonology known as yinyunxue ‘prosody study’. From the sixth and seventh century on, a number of Chinese scholars started to compile rhyme books to help people use the correct rhyme patterns in poem writing. They got inspiration from the Sanskrit spelling system and designed a special way to record the pronunciation of Chinese characters, each of which was in essence a syllable. Each Chinese syllable was dissected into two parts: an initial and an ending. The initial part was called sheng ‘consonant’ or shengmu ‘consonant letter’, and each shengmu was represented by a Chinese character. The ending part was called yun ‘rhyme, vowel’, which was a vowel, a diphthong, a combination of a glide and a vowel, a combination of a glide and a diphthong, a combination of a vowel and a coda, or a combination of a glide, a vowel and a coda. All possible yun were grouped into yunbu ‘rhyme category’ or yummu ‘rhyme entry’ and each rhyme category was represented by a character. Each syllable was represented by a sheng character and a yun character, and its pronunciation was produced by a method known as fan, qie, fanyu or fanqie, namely, by using the consonant of its sheng character and the vowel or vowel combination of its yun character. The symbols in the Zhuyin Zimu system resembled the shengmu of the rhyme books in many ways. An obvious similarity is that each of the shengmu ‘consonant letters’ was assigned a vowel to make it a syllable so that it sounded like a sheng character and could be easily pronounced by the Chinese people. The 24 shengmu therefore all had the pronunciation of 725

Shi Dingxu

a syllable: ㄅ[po], ㄆ[pho], ㄇ[mo], ㄈ[fo], 万[vo], ㄉ[tɤ], ㄊ[thɤ], ㄋ[nɤ], ㄌ[lɤ], ㄍ[kɤ], ㄎ[khɤ], 兀[ŋɤ], ㄏ[xɤ], ㄐ[tɕi], ㄑ[tɕhi], 广[ ɲi], ㄒ[ɕi], ㄓ[tʂʅ ], ㄔ[tʂhʅ ], ㄕ[ʂʅ ], ㄖ[ʐʅ ], ㄗ[tsɿ], ㄘ[tsh ɿ] and ㄙ[sɿ]. The yunmu of the Zhuyin Zimu, on the other hand, resembled the yun characters of the rhyme books in a similar way. Some of yunmu were simply a vowel or, to use the term of the Committee, a syllable with a zero consonant and a vowel, such as ㄚ[ɑ], ㄛ[o] and ㄜ[ɤ]. Some of them were zero consonant and diphthong combinations such as ㄞ[aj], ㄟ[ej], ㄠ[ɑw] and ㄡ[ow]. Some of them were zero consonant, vowel and nasal combinations like ㄢ[an], ㄣ[ən], ㄤ[ɑŋ] and ㄥ[əŋ]. There are also some special cases. One of them was ㄦ, which was pronounced as [ǝɹ] or [ɚ], namely, a mid-central vowel [ǝ] with an r-coloring. Another case was ㄝ[ jɛ], which actually stood for the vowel [ɛ] but a glide [ j] was added to make it pronounceable since [ɛ] could not stand alone as a syllable in most Chinese dialects. The most interesting case was the three jiemu ‘glide letter’ ㄧ, ㄨ and ㄩ. They stood for the three high vowels [i], [u] and [y], but they also represent the three glides [j], [w] and [ɥ]. When they were invented, these symbols were actually given the status of syllables: ㄧ[ ji], ㄨ[wu] and ㄩ[ɥy], namely, a combination of a glide and a vowel. A tone marker system was also designed for the system, with diacritics as tone markers. Some tone markers imitated the contour of tones, such as a ˊ for the rising tone (second tone), a ˇ for the falling–rising contour tone (third tone) and a ˋ for the falling tone (fourth tone). There was no marking for the high flat tone (first tone), and a dot ˙ was assigned to the neutral tone. The tone marker appeared on the right side of the yunmu if the symbols were arranged vertically from top to bottom but appeared on the top-left side of the yunmu if the symbols were arranged horizontally from right to left or on the top-right side of the yunmu if the symbols were arranged from left to right like the second tone marker in ㄆㄚˊ ‘crawl’. The way these symbols were used also resembled that in the rhyme books. Two symbols were typically involved to mark the pronunciation of a particular character, namely a shengmu and a yunmu as in ㄍㄠˋ [kɑw] ‘tell’. The actual pronunciation of the syllable was produced by using the consonant of the shengmu and the diphthong of the yunmu. In other words, the [ɤ] of ㄍ[kɤ] was not pronounced. It was possible to use one yunmu only, as in ㄞˋ [aj] ‘love’. It was also possible to have a shengmu, jiemu and yunmu combination, as in ㄒㄧㄠˇ [ɕjɑw] ‘small’. Note that the ㄒ in this case was pronounced as [ɕ] instead of [ɕi], while ㄧ was pronounced as [ j] instead of [ji] or [i]. In 1930 the government of the Republic of China changed the term Zhuyin Zimu to Zhuyin Fuhao ‘phonetic symbols’ and the number of symbols was reduced to 37 by stipulating that 万[vo], 兀[ŋɤ] and 广[ ɲi] would only be used for recording non-Guoyu dialects.

1.3.  Putonghua ‘the common language’ The government of the People’s Republic of China also subscribed to the concept of a national language based on one dialect and named such a language Putonghua ‘the common language’. It set up the Committee on the Reform of the Chinese Writing System in 1954, which is actually a committee on language planning, to handle the Putonghua issue and issued a decree in 1956 stating that Putonghua should be promoted as the guojia tongyongyu ‘the language for general communication throughout the country’. Putonghua is in essence a revised version of Guoyu, especially the xin guoyin ‘new national pronunciation’, even though the latter was only loosely defined via standardizing the pronunciation of several thousand characters. 726

Standard Chinese

Putonghua, on the other hand, was given an official definition by the 1956 government decree, which consists of three statements. The first one is that the standard for the Putonghua sound system should be that of Beijing Mandarin. This means the phonetic and phonological system of Putonghua is based on those of Beijing Mandarin, in the sense that the sound inventory of Putonghua is the same as that of Beijing Mandarin, the features of each Putonghua sound are similar to those of its Beijing Mandarin counterpart, and the value of each Putonghua tone is identical to that of its Beijing Mandarin counterpart. The Beijing Mandarin sound system mentioned here refers to that used by the educated elite and does not include variations found in the vogue colloquial speech. The second statement is that the dialectal basis for Putonghua is the Northern dialect. This is actually the definition for the Putonghua lexicon, saying that the basic Putonghua vocabulary mainly contains lexical items shared by various sub-dialects in the Northern dialect family. By choosing the core lexicon of various sub-dialects, most vulgar items, such as curses and obscene words, are naturally excluded from the Putonghua vocabulary. The third statement says that the norms of grammar in Putonghua are those found in the representative works of contemporary vernacular Chinese writers. This is another way of saying that the Putonghua grammar norms are the same as those of Standard Written Chinese, and that grammar patterns of Classical Chinese are excluded and even those of early vernacular writing are also excluded. The Committee on the Reform of the Chinese Writing System was given the mandate of regulating the Chinese writing system, promoting Putonghua, and designing and promoting Pinyin, which was a new alphabetic writing system for Chinese. The Committee started a campaign in the late 1950s to promote Putonghua and stipulated that, as the official language for general communication throughout the country, Putonghua should be the only means of instruction in all schools across the country, the dominating language for broadcasting and the medium for public announcements. The Committee was also responsible for quite a number of publications related to its mandate, including a book on the standardization of Chinese characters that stipulates for each character one variation among all the others, two lists of simplified Chinese characters, a dictionary of Modern Chinese, which is the well-known Xiandai Hanyu Cidian ‘A Dictionary of Contemporary Chinese’, and a book on the standardization of Putonghua pronunciation for Chinese words.

1.4.  Hanyu Pinyin ‘Chinese Phonetic Alphabets’ Before the announcement of the decree which declared Putonghua was ‘the language for general communication throughout the country’ in 1956, the government of the People’s Republic of China had launched a literacy campaign and the promotion of a common language for all people was part of the campaign. Guoyu and Zhuyin Fuhao were the choice for the campaign by default at that time. While the Committee on the Reform of the Chinese Writing System was working on the details of Putonghua, they were also considering a writing system for recording Putonghua. They called for proposals and received about 1,600 designs, of which some were based on Chinese characters, and others were based on Cyrillic alphabets, hieroglyphic symbols, digitals, stenographic symbols as well as Latin alphabets. When the Committee made the final decision, the chosen one was the Latin-based Hanyu Pinyin Fang’an ‘Chinese Phonetic Alphabets System’. 727

Shi Dingxu

The idea of having a spelling system based on Latin alphabets to record Chinese could be traced back to the one designed by Jesuit priest Matteo Ricci (1552–1610) and refined by another Jesuit priest Nicolas Trigault (1577–1629). The spelling system designed by the Jesuits not only borrowed the Latin alphabets but also adopted its syllable structure, namely, using a consonant and a vowel, a diphthong or a vowel–nasal combination to build a syllable. This was different from the fanqie of traditional Chinese phonology in which two Chinese characters, i.e. two syllables, were used to represent another syllable. By the late nineteenth century and early twentieth century, quite a number of alphabetic spelling systems were proposed, mainly by missionaries for translating the Bible into various Chinese dialects. There were also some spelling systems for Guoyu, and the best known were Guoyu Luomazi ‘Romanization for the National Language’ and Beifanghua Lading Xinwenzi ‘New Latin Writing System for the Northern Dialect’. The Committee announced in 1955 a draft of the Hanyu Pinyin system and publicized the final version in 1957, which was then given an official status by a decree from the People’s Congress of the People’s Republic of China. The Pinyin system is designed as a phonemic representation of Putonghua, which is very similar to Guoyu, and therefore resembles Zhuyin Fuhao in many ways. On the other hand, the designers of Pinyin took into consideration the gap between Pinyin symbols and the pronunciation of actual syllables and created more symbols than those in Zhuyin Fuhao. The Pinyin system inherits the term shengmu and yunmu from Zhuyin Fuhao, even though the latter is based on Chinese characters instead of the Latin alphabets. As can be seen from Tables 40.1 and 40.2, Pinyin symbols are categorized as shengmu and yunmu, which are defined in the same way as their counterparts in Zhuyin Fuhao. Each shengmu in Pinyin is a Latin letter and is correspondent to a shengmu of Zhuyin Fuhao. This entails that each Table 40.1  Shengmu ㄅ ㄆ ㄇ ㄈ

b [po] p [pho] m [mo] f [fo]

ㄉ ㄊ ㄋ ㄌ

d [tɤ] t [thɤ] n [nɤ] l [lɤ]

ㄍ g [kɤ] ㄎ k [khɤ] ㄏ h [xɤ]

ㄐ j [tɕi] ㄑ q [tɕhi] ㄒ x [ɕi]

ㄓ zh [tʂʅ] ㄔ ch [tʂh ʅ] ㄕ sh [ʂʅ]

ㄗ z [tsɿ] ㄘ c [tshɿ] ㄙ s [sɿ] ㄖ r [ʐ ɿ]

Table 40.2  Yunmu ㄚ a [ɑ] ㄛ o [o] ㄜ e [ɤ] ㄞ ai [aj] ㄟ ei [ej] ㄠ ao [ɑw] ㄡ ou [ow] ㄢ an [an] ㄣ en [ǝn] ㄤ ang [ɑŋ] ㄥ eng [ǝŋ] (ㄨㄥ) ong [uŋ]

ㄧ i [i] ㄧㄚ ia [ jɑ] ㄧㄝ ie [ jɛ]

ㄧㄠ ㄧㄡ ㄧㄢ ㄧㄣ ㄧㄤ ㄧㄥ ㄩㄥ

iao [ jɑw] iou (iu) [ jow] ian [ jæn] in [in] iang [ jɑŋ] ing [iŋ] iong [ɥǝŋ]

ㄨ u [u] ㄨㄚ ua [wɑ] ㄨㄛ uo [wo] ㄨㄞ uai [waj] ㄨㄟ uei (ui) [wej]

ㄨㄢ ㄨㄣ ㄨㄤ ㄨㄥ

728

uan [wan] uen (un) [wǝn] uang [wɑŋ] ueng [wǝŋ]

ㄩ ü [y] ㄩㄝ üe [ɥɛ]

ㄩㄢ üan [ɥæn] ㄩㄣ üen (ün) [ɥǝn]

Standard Chinese

shengmu is assigned the pronunciation of a syllable, such as [po] for b, [pho] for p, [mo] for m and so on, to help Chinese people to pronounce it. There are far more yunmu symbols in Pinyin than in Zhuyin Fuhao, because in addition to vowels and diphthongs, all combinations of glide and vowel, vowel and nasal, as well as glide, vowel and nasal, are listed as yunmu, as shown in Table 40.2. There are quite a number of reasons to list all the combinations as yunmu, and an important one is that, similar to Zhuyin Fuhao, Pinyin is not a phonetic representation of the Putonghua sound inventory and some symbols actually stand for more than one sound. A typical case is that the symbol a stands for [ɑ] when a appears alone or in certain vowel– nasal combinations such as ang [ɑŋ] and iang [ jɑŋ]. The same symbol stands for [a] in the vowel–nasal combination an [an] and for [æ] in the vowel–nasal combinations ian [ jæn] and üan [ɥæn]. Another case is the symbol e, which stands for [ɤ] when e appears alone as a yunmu; [ɛ] in ie and üe; [ǝ] in en, uen (un), üen (ün), eng, and ueng; and [e] in ei and uei. Such an arrangement is reasonable for linguists but is likely to cause confusion for laymen. A simple way to solve the problem is to list all the possible combinations existing in Chinese, assign a fixed pronunciation to each of them and list them as independent symbols. This sounds redundant to people who are used to putting a consonant symbol and a vowel symbol together to pronounce a syllable, but is extremely helpful for Chinese people who are used to recognizing individual characters and pronouncing them as different syllables. An additional advantage is that, given the idea of pronouncing each symbol as a whole, certain symbols could be altered a little to make them shorter. This is how shortened symbols like un (uen), ün (üen), ui (uei) and iu (iou) are created. Another such alternation is to use ao for [ɑw] instead of using au, which is more common in the spelling of European languages, to avoid possible confusion of au and an in handwriting. Pinyin inherits the tone markers from Zhuyin Fuhao, but these markers are put on the top of the main vowel in yunmu, as in píngguǒ ‘apple’, huángfēng ‘wasp’ and lǘpí ‘donkey skin’. Neutral tones are not marked with anything, as shown by the second syllable in yāzi ‘duck’. If the yunmu is a reduced symbol without the main vowel, the tone marker will fall on the glide, as in guìjūn ‘your honorable troop’. Pinyin has a special symbol ü for the high front rounded vowel [y] and the corresponding glide [ɥ], but the two dots (umlauts) on top of the symbol can be omitted unless it appears after l and n. There are two separate Pinyin symbols for the same Zhuyin Fuhao combination ㄨㄥ. One of them is ong [uŋ] and the other is ueng [wǝŋ]. This is a reflection of the 1950s view that the pronunciation of the word ‘red’ is [xuŋ] and that of ‘old man’ is [wǝŋ], in contrast with to the 1930s belief that one is [xwǝŋ] and the other is [wǝŋ].

2.  Standard Written Chinese Unlike the spoken forms Guoyu and Putonghua, both of which have an official status of the national standard language granted by the government, the written form of Standard Chinese has never been given any official definition by the government, even though there has always been a written Chinese form accepted by the government as the standard. The Ministry of Education of the Republic of China issued in 1932 a Guoyin Changyong Zihui ‘A Collection of Commonly Used Characters in National Pronunciation’ as the standard vocabulary of Guoyu. The Committee on the Reform of the Chinese Writing System also issued in 1958 a Hanyu Pinyin Cihui ‘The Vocabulary of the Chinese Phonetic Alphabets’ which tells 729

Shi Dingxu

people how to pronounce Chinese words correctly in Putonghua. To a large extent, the two vocabularies define the basic lexicon of written Standard Chinese in the twentieth century. Another major difference between the spoken and written forms of Standard Chinese is that the former has undergone quite a number of changes while the later only has two de facto versions up till now. The spoken form of Standard Chinese, commonly known as Guanhua ‘language of the officials’, was the means of communication for government officials, especially those in the capital of an empire, and was typically the dialect of the capital city. During the several thousand years before the Republic of China, it happened many times that one empire was overthrown by another and the capital was moved to a different city. The substitution of Guanhua thus took place every few hundred years or even every few decades. Another factor is that spoken languages, including Chinese and all its dialects, are constantly evolving. The same version of Guanhua would have a shape different from the original one after a few hundred years or even a few decades. This is what has happened to Beijing Mandarin over the past 60 years and to Shanghai dialect in the past 30 years. On the other hand, the old written form of Standard Chinese, which is commonly known as Classical Chinese, had remained virtually the same over 2,000 years. It took shape at the end of Chunqiu Shiqi ‘the Spring and Autumn Period’ (approximately from 770 to 476 bc), when scholars wrote a large number of books to articulate their views on various topics. Quite a number of books survived hundreds of years of turmoil caused by rebellions and wars, and the most important ones on ethics, philosophy and rituals were recommended by scholars in the Song dynasty (ad 960–1279) to be the core readings of Confucianism. They are commonly referred to as the Four Books, namely, Daxue ‘The Great Leaning’, Zhongyong ‘The Doctrine of the Mean’, Lunyu ‘The Analects of Confucius’ and Mengzi ‘Mencius’; and the Five Classics, namely, Shijing ‘The Book of Songs’, Shujing ‘The Book of History’, Yijing ‘The Book of Changes’, Lijing ‘The Book of Rites’ and Chunqiu ‘The Spring and Autumn Annals’. The Four Books and Five Classics became the compulsory reading materials for civil servant examination in the Song dynasty as well as in those after the Song. Such an official status inevitably had an enormous impact on the educated elite because the ability to interpret these classics and to write explanatory proses with Classical Chinese had become the most important skills for moving upwards in society. As a written language, Classical Chinese was based on the spoken form of Chinese in the Spring and Autumn Period. This can be seen clearly from the close resemblance between Lunyu, which is very formal and is attributed to Confucius and his disciples, and Shijing, part of which is believed to be folk songs from people who lived in the same period. A gap between the spoken form and written form of Chinese appeared gradually after that period and became obvious by the end of the Han dynasty (202 bc – ad 220). By the Sui (ad 581–618) dynasty and Tang (618–907) dynasty period, the gap had become so wide that two types of written Chinese became a reality. The so-called Dunhuang Bianwen ‘altered written form found in Dunhuang’, which was discovered in the late nineteenth century, provided clear evidence for the existence of a colloquial written Chinese. Since the Dunhuang Bianwen literature was largely Buddhism stories and talk show scripts, which were written mainly by monks during the Sui and Tang period in order to promote Buddhism among ordinary people, the language being used was naturally based on the spoken Chinese of that time. The differences between Dunhuang Bianwen and Classical Chinese could thus be attributed to the evolution of spoken Chinese from the Spring and Autumn period to the Sui and Tang period. 730

Standard Chinese

The literature in colloquial spoken Chinese kept growing in the Song dynasty and the Yuan dynasty (1271–1368). A large amount of popular literature survived the time and could be seen today. A common form of the colloquial literature is huaben xiaoshuo ‘talking-points fiction’, namely, the scripts for story telling or talk shows; and another form is the scripts for Yuan zaju ‘variety opera of the Yuan period’. Since the targeted audience for the story telling and opera was mainly lower and midle-class city dwellers, these scripts should be a genuine record of the spoken Chinese at that time, and they demonstrated a clear disparity between spoken Chinese and Classical Chinese. An even larger body of popular literature written in a form closely resembling spoken Chinese of the Ming dynasty (1368–1644) and the Qing dynasty (1644–1911) has been preserved, and the best-known representatives are the so-called Four Masterpieces, namely, Sanguo Yanyi ‘The Romance of the Three Kingdoms’, Shuihuzhuan ‘Heroes of the Water Margins’, Xiyouji ‘Pilgrimage to the West’ and Hongloumeng ‘A Dream of Red Mansions’. Several hundred years went by from the time of Sanguo Yanyi and Shuihuzhuan, which were produced at the beginning of the Ming dynasty, to the time of Hongloumeng, which was written in the mid-eighteenth century, and the Chinese used in these fictions displays significant differences. The language in Hongloumeng is quite close to contemporary spoken Chinese and further away from Classical Chinese, which had remained virtually unchanged all these years. By the early twentieth century, the enormous gap between Classical Chinese and spoken Chinese had become so big that it was almost impossible for ordinary people to learn Standard Chinese and become literate. Quite a number of intellectuals who received their education overseas, such as Shi Hu, Duxiu Chen, Bannong Liu, Xun Lu and Xuantong Qian, saw the difficulties and witnessed the consequences. When they were confronted with the setbacks China suffered in the international arena, they blamed the political crisis on Classical Chinese in that it prevented the populace from getting proper education and thus hindered the industrialization and modernization of China. They started a campaign known as the Baihuawen Yundong ‘the Campaign for Vernacular Written Chinese’ to go with the Guoyu Yundong ‘the National Language Movement’, and set up the Duyin Tongyi Hui ‘the Committee on the Standardization of Pronunciation’ as part of the campaign. The idea of using contemporary spoken Chinese, namely, baihua ‘vernacular Chinese’, as the basis for Standard Written Chinese was actually formed in the middle of the nineteenth century by a number of scholars, and the slogan of the Campaign for Vernacular Written Chinese, namely, woshou xie wokou ‘what my hand writes should be what my mouth says’ was also coined by them. With such a slogan as the motto, advocators of the Campaign for Vernacular Written Chinese started with popular literature by writing fictions and proses with baihua, and then turned to shici ‘poetry’, the most privileged part of the Classical Chinese literature, by composing xinshi ‘new poems’ with baihua. The Campaign for Vernacular Written Chinese coincided with the May Fourth Movement, which was political in nature, and quickly gained support from many walks of life. A new written form of Standard Chinese came into being within a very short period of time. By the 1920s, the written vernacular had replaced Classical Chinese in many areas including education and news media, even though Classical Chinese and a semi-Classical version of written Chinese were preserved in some very formal genres such as government decrees and official documents. A written Standard Chinese based on contemporary spoken Chinese apparently had its appeal to the populace but how to educate them with this written form was not an easy job. The literacy campaign in the 1950s provided a stage for baihuawen to show its usefulness, 731

Shi Dingxu

and political considerations by the government of the People’s Republic of China added some momentum to its advance. It quickly became the sole choice not only in educational materials, literature, business documents and mass media, but also in government decrees and in laws and regulations. The compulsory use of baihuawen in education and mass media laid the foundation for its further growth. Even though there has never been any official definition for Standard Chinese or its vocabulary, it is commonly agreed that the vocabulary has a clear core shared by most people and a periphery with limited distribution among people with different backgrounds. The core consists of words and expressions from a number of sources. Most of them come from contemporary spoken Chinese, especially the Northern dialect, to replace Classical Chinese terms from the spoken Chinese of the Spring and Autumn period. For example, instead the Classical Chinese 冠 guan ‘hat’, the core Standard Chinese word is 帽子 maozi ‘hat’ from the Northern dialect, even though 帽 mao ‘hat’ from Southern dialects such as Cantonese is occasionally found. Certain items inherited from Classical Chinese are still found in Standard Chinese, such as 夺魁 duokui ‘win the championship’ and 是可忍,孰不可忍 shi ke ren, shu buke ren ‘if this could be tolerated, what else could not be tolerated’, with various frequencies in actual use. 夺魁 appears in different genres and has become a common Standard Chinese compound, while 是可忍,孰不可忍 is typically used in formal political writings only. Quite a few Japanese words have been imported into Modern Chinese and become part of Standard Chinese. Most of them were coined with Chinese morphemes during the nineteenth century by Japanese scholars who translated into Japanese Western works in science, economy and politics. Those terms were introduced to China in the early twentieth century together with the concepts they represented, such as 哲学 zhexue ‘philosophy’, 革命 geming ‘revolution’, 经济 jingji ‘economy’ and 民主 minzhu ‘democracy’, and have become part of Standard Chinese. Some Japanese words were borrowed into Chinese as loan words in recent years, such as 写真 xiezhen ‘photo’ and 达人 daren ‘smart person, talent’, and typically have limited distribution. Loan words from other languages include 雷达 leida ‘radar’ and 博客 boke ‘blog’ from English, 伏特加 futejia ‘vodka’ from Russian, 蒙太奇 mengtaiqi ‘montage’ from French and 探戈 tan’ge ‘tango’ from Spanish. There are also some claque forms from other languages, such as 热狗 regou ‘hot dog’, 硬盘 yingpan ‘hard disk’ and 宽带 kuandai ‘broadband’ from English. As in other languages, new Standard Chinese words and expressions are being created all the time to represent new concepts and new things. Expressions coined in the 1950s, 60s and 70s are mostly political terms, including 合作社 hezuoshe ‘(farmers’) co-op’, 公私合营 gongsi heying ‘private and public joint ownership’, 大跃进 dayuejin ‘big leap forward’, 阶级斗争 jieji douzheng ‘class struggle’, 文化大革命 wenhua dageming ‘the Great Cultural Revolution’, 再教育 zaijiaoyu ‘re-education’, 独生子女政策 dushengzinü zhengce ‘onechild-per-family policy’ and 四人帮 sirenbang ‘Gang of Four’. New expressions created after the 80s are mostly related to economy, including 改革开放 gaige kaifang ‘reform and opening-up’, 经济特区 jingji tequ ‘special economic zone’, 商品房 shangpinfang ‘houses for private ownership’, 快递 kuaidi ‘express package service’ and 高铁 gaotie ‘high-speed railway’. These expressions would typically phase out when the entities they represented had disappeared. Not many people would know the term 合作社 nowadays except for those who went through that period or those who study the history of the 1950s. The periphery of the Standard Chinese vocabulary exists for various reasons. An obvious one is that speakers of each dialect tend to have their favorite local words or phrases which may not be shared by speakers of other dialects. A typically example is the two verbs 732

Standard Chinese

说 shuo ‘speak’ and 讲 jiang ‘speak’. Although they have almost identical meaning, speakers of Northern dialects tend to use the former while speakers of Southern dialects tend to use the latter. Another example is the word 忽悠 huyou ‘try to fool others’, which was a colloquial form of Northeast China and became popular after a comic star used it in a TV show. The word entered Putonghua and written Standard Chinese quickly but still has a limited distribution because of its vulgar nature. Given the fact that Standard Written Chinese is based on contemporary spoken Chinese, it is quite natural that most grammar rules, phrase structures and sentence patterns are the same in these two forms. Certain phrase and sentence patterns of Standard Chinese are inherited from Classical Chinese, such as semi-Classical Chinese complex nominal phrases like 工作之余 gongzuo zhiyu ‘the leisure time that is available after work’ and 危急之际 weiji zhiji ‘the time when crisis comes’. These phrases are constructed with the Classical Chinese morpheme 之 zhi and Standard Chinese nouns, adjectives or verbs, and actually have the structure, function and meaning of their Standard Chinese counterparts 工作后的 闲暇 gongzuohou de xianxia ‘the leisure time that is available after work’ and 危急的时候 weiji de shihou ‘the time when crisis comes’. Since they carry some Classical Chinese flavor, these phrases are typically used in formal genres. Certain grammatical features of Standard Written Chinese are adopted from other langu­ ages, especially from English via translation. A typical case is the distribution of passive sentences. Although there does exist one or two sentence patterns in each dialect that could be understood as conveying a passive meaning, passive sentences have a quite limited distribution since they are used only to express undesirable situations or unpleasant encounters. Passive sentences in Standard Written Chinese, on the other hand, are not subject to the same restriction and are often used to describe pleasant situations with desirable consequences. It is commonly believed that the difference is due to the influence of translating into Chinese all kinds of works written in foreign languages, especially those in English, French and Russian. Since passive sentences are being used in these languages to describe various events, bad or good, in which the agent of the action is unknown or not important, the most straightforward way to render them into Chinese is to create a sentence pattern corresponding to them directly. The result is the 被 bei passive, which is marked with a morpheme derived directly from Classical Chinese instead of from any dialect and has a much wider distribution.

3.  Variations of Standard Chinese Although there is an official definition for Putonghua and a set of standards based on the definition, a perfect Putonghua speaker remains an ideal to which only a few could come close, as indicated by the very small number of people who reached the highest level in the Putonghua Shuiping Ceshi ‘the National Putonghua Proficiency Test’. Most Putonghua speakers actually speak a variety of Putonghua that has some characteristics of their native or adopted dialect, and that is why there are terms like Shanghai Putonghua or Sichuan Putonghua. Even the two official Standard Spoken Chinese forms, namely, Putonghua in the People’s Republic of China and Guoyu in Taiwan, are not exactly the same. In addition to the minor differences in their sound systems and the distinctive representation of these systems, the main difference now lies in their vocabularies. The basic items in these two vocabularies were similar in early days, as attested by the 1932 Guoyin Changyong Zihui ‘A Collection of Commonly Used Characters in National Pronunciation’ and the 1958 Hanyu Pinyin Cihui 733

Shi Dingxu

‘The Vocabulary of the Chinese Phonetic Alphabets’, but have gradually gone along different pathways to accumulate enough differences, as attested by the 2000 Xinbian Guoyu Ribao Cidian ‘The New Dictionary of Mandarin Daily News’ and the 2012 Xiandai Hanyu Cidian (Diliuban) ‘The Dictionary of Contemporary Chinese (6th ed.)’. The lexical items that appear in one form but not in the other are mainly loan words from dialects, loan words from foreign languages and newly invented terms. The dialect that exports most words into Taiwan Guoyu is Southern Min, while that which gives most words to Putonghua is Northern dialect. Taiwan Guoyu has adopted more foreign words, especially from English and Japanese, than Putonghua has. The evolution of political systems, economic development, and advancement of science and technology constantly create new expressions to describe new things, and the differences between the two sides of the Taiwan Strait will inevitably lead to different results. It is therefore quite understandable that the term for ‘parliament’ is 全国人民代表大会 Quanguo Renmin Daibiao Dahui in Putonghua while it is 立法院 Lifayuan in Taiwan Guoyu. The grammar of the two Standard Spoken Chinese forms is basically the same except in some minor details. The most obvious difference is that some Southern Min morphological processes have been incorporated into Taiwan Guoyu but not into Putonghua. A typical case is the tentative suffix –看看 –kankan, which is borrowed from Southern Min and is used frequently in Taiwan Guoyu to produce compounds to represent actions of a trial or uncertain nature such as 吃看看 chikankan ‘eat to taste’ and 穿看看 chuankankan ‘wear to try’, in addition to the more traditional 吃吃看 chichikan and 穿穿看 chuanchuankan that also appear in Putonghua. Like the written form of other languages, the written form of Standard Chinese is not a precise record of spoken Chinese but has a life of its own. Although the more informal the genres are, the more vernacular features will appear in the written form, the opposite is also true. As the genres become more formal, the written form will have fewer and fewer elements from the spoken vernacular. Expressions and structures of Classical Chinese are usually considered formal features, while long and complicated sentence structures are also perceived as formal. Apart from the varieties due to different genres, there are also some regional variations, and the most prominent cases are the written Chinese in Taiwan and that in Hong Kong. Both of them have some distinctive features that are not found in the written Chinese in mainland China, owing to the economic and political situation in these places. There are many forms of Written Chinese in Hong Kong, including a Written Cantonese that is an almost verbatim recording of spoken Cantonese, a Sanjidi ‘mixture of three’ that is mixture of Classical Chinese, Cantonese and Standard Chinese, and a form known as Hong Kong Written Chinese that is widely used in the serious news pages and the editorial sections of various newspapers as well as government documents. Given its function, Hong Kong Written Chinese is the Standard Written Chinese in Hong Kong, but it is not the same as its counterpart in Taiwan or in the People’s Republic of China. The basic vocabulary and main structures of Hong Kong Written Chinese are similar to the other Standard Chinese, but it also has a lot of unique features coming from English, Cantonese and as innovations. English was the sole official language in Hong Kong for a long time and is still one of the official languages, and it is expected that Hong Kong Written Chinese is under strong English influence in terms of vocabulary, syntax and discourse. In addition to English loan words and calque forms, such as 拉布 labu ‘filibuster’ and 硬碟 yingdie ‘hard drive’ respectively, many expressions in Hong Kong Written Chinese have undergone meaning or function changes due to English influence. The Hong Kong 734

Standard Chinese

Written Chinese word 分享 fenxiang, for example, has retained the Standard Chinese meaning of having good things distributed to all people so that everyone can enjoy them, but it also carries some special meaning in certain Hong Kong Written Chinese expressions, such as 分享痛苦的经验 fenxiang tongku de jingyan ‘share a painful experience’ or 分享文件 fenxiang wenjian ‘share documents’, in which the things being 分享 are not good but are undesirable or at the best neutral. Obviously, the semantic content of Hong Kong Written Chinese 分享 and its distribution is identical to that of English ‘share’, but different from that of Standard Chinese 分享. The syntax of Hong Kong Written Chinese has a lot of Cantonese features, such as using the morpheme 过 instead of 比 for comparative construction. To represent the idea that today’s exchange rate is higher than yesterday’s, a typical Hong Kong Written Chinese sentence would be 今天的汇率高过昨天好多 jintian de huilü gao guo zuotian hao duo rather than the Standard Chinese 今天的汇率比昨天高了很多 jintian de huilü bi zuotian gaole hen duo. There are also some unique discourse features in Hong Kong Written Chinese. One of them is the use of the pronoun 有人 youren ‘someone’ in certain discourse, like the news report quoted below. The main figure in this piece of discourse is a man who was arrested and who admitted that he had robbed ten banks. Under Hong Kong law, a person is not guilty in a court case until he or she has been convicted, and a news report could be considered libel if the person is directly named as a culprit. A very innovative way to get around this constraint is to use the pronoun 有人, which originally means ‘there is a person’. Although the 有人 in this context clearly refers back to the man Luo mentioned earlier, the exact meaning of 有人 in Standard Chinese is that a new person is being introduced to the discourse for the first time. The news reporter could thus argue that it is a different person who is being reported as admitting guilt, even though everyone in Hong Kong understands this and would make the correct inference. 被捕男子姓罗,失业,住北角天后庙道,有人承认由今年六月至九月,先后打劫 十间银行。 ‘The last name of the man arrested was Luo. He was unemployed and lived in Tianhou Miao Road of North Point. Someone admitted that he robbed ten banks from June to September this year.’

Bibliography Guoyu Ribao She (2000) Xinbian Guoyu Ribao Cidian (A New Dictionary by Guoyu Daily), Taipei: Guoyu Ribao Press. Li, Guangdi (1728) Yinyun Chanwei (Details of Prosody), Beijing: Yingwu Palace of the Imperial Court. Norman, Jerry (1988) Chinese, Cambridge: Cambridge University Press. Zhang, Bin (ed) (2010) Xiandai Hanyu Miaoxie Yufa (A Descriptive Grammar of Contemporary Chinese), Beijing: The Commercial Press. Zhongguo Dacidian Bianzuanchu (1932) Guoyin Changyong Zihui (A Collection of Commonly Used Characters in National Pronunciation), Shanghai: The Commercial Press. Zhongguo Shehui Kexueyuan Yuyan Yanjiusuo Cidian Bianjishi (2012) Xiandai Hanyu Cidian (A Dictionary of Contemporary Chinese), 6th ed., Beijing: The Commercial Press. Zhongguo Wenzi Gaige Weiyuanhui Cihui Xiaozu (1958) Hanyu Pinyin Cihui (The Vocabulary of the Chinese Phonetic Alphabets), Beijing: Press of the Committee on the Reform of the Chinese Writing System. Zhongguo Wenzi Gaige Weiyuanhui Cihui Xiaozu (1963) Hanyu Pinyin Cihui Zengdingban (The Extended Version of the Vocabulary of the Chinese Phonetic Alphabets), Beijing: Press of the Committee on the Reform of the Chinese Writing System.

735

41 Syntax Li Yafei University of Wisconsin-Madison, USA

In the past few decades, the syntactic study of the Chinese language has reached unprecedented width and depth both in the amount of phenomena discovered and for theoretical development. The literature is so immense that it becomes impractical to provide a comprehensive coverage of the field while guaranteeing the presentation to be informatively detailed and accurate. For this reason, the current entry is written with two guidelines: (i) The content will focus primarily on the works conducted within the Principles-and-Parameters (PnP) model of syntax; (ii) only a subset of those topics will be covered which have either made contributions to the general PnP theory or shed new light on certain unique phenomena in Chinese. These choices are made solely to keep the entry manageable.

Wh-in-situ J. Huang (1982) was the earliest to integrate Chinese into the government-and-binding (GB) theory, an earlier version of PnP. Since May (1977), the field of syntax had accepted the existence of covert (aka LF) operator-movement to deal with the ambiguities created by quantificational expressions. Meanwhile, Chomsky (1976, 1977) proposed to treat English wh-words as operators undergoing movement (also see Kayne 1979). Following these leads, Huang studied Chinese in-situ wh-expressions. The core discovery came from comparing English with Chinese (with t standing for trace, marking the pre-movement location of wh): (1) (a) Why does she think [CP that Smith resigned t ]? (b) *Why does she wonder [CP who resigned t ]? (2) (a) ta renwei [CP Zhang Qing weishenme cizhi ]? 3s think ZQ why resign (b) ta wen ni [CP shei weishenme cizhi ]? 3s ask you who why resign (1) illustrates the wh-island effect, namely the indirect question prevents why from moving out. In (2a), weishenme ‘why’ inside the object CP turns the entire sentence into a wh-question, producing the so-called ‘wide scope’ reading ‘What is the reason x such 736

Syntax

that she thinks that ZQ resigned for x’ just as why does in (1a) via movement. The same weishenme in (2b), however, does not let the sentence mean ‘What is the reason x such that she asked you who resigned for x?’. The generalization is that why and weishenme fail to make the entire sentence into an interrogative while inside an indirect question, with or without overt movement. For Huang, this fact confirms the previous works on covert movement: Both why and weishenme must move to some Spec of CP, the established destination for wh-movement, and therefore both are restricted by the wh-island. They differ only in doing so overtly in English or covertly in Chinese. The same pattern is found with other islands (see below) which Huang subsumes under his Condition on Extraction Domain (CED). But this unification has a gap: (3) (a) *What does she wonder [CP who broket ]? shenme ]? (b) ta wen ni [CP shei dasui-le 3s ask you who break-asp what While the unacceptable (3a) is again due to the wh-island, (3b) contradicts (2b) because shenme ‘what’ inside the indirect question can actually take the wide scope and mean ‘What is x such that she asked you who broke x?’. Huang’s solution to this discrepancy consists of two parts: (A) the general belief in the GB theory that an adjunct trace must be close enough to its antecedent (aka antecedent-government); (B) the postulation that islands don’t block LF movement. Shenme ‘what’ in (3b) is an object (= an argument). Its covert movement out of the indirect question is insensitive to the wh-island according to (B), making (3b) a legitimate question. In (2b), weishenme ‘why’ is an adjunct. Covertly moving it into the root clause is fine in itself under (B), but creates a trace t separated from the moved weishenme (= antecedent) by the wh-island. This separation fails the closeness requirement in (A), producing the observed sensitivity to islands. Huang’s theory triggered two intertwined questions: Exactly what in-situ wh-expressions are immune to islands and why? L. Xu (1990) is the first comprehensive work to tackle the what-question, noting counter­ examples to Huang’s argument-adjunct distinction. (4) illustrates the adjunct zenme ‘how’ inside a complex-NP island but with the wide-scope reading: (4) [DP[CP ta zenme shao ] de cai ] geng shou huanying? 3s how cook DE dish more get liked ‘What is the way x such that the dish she cooks in x is more liked?’ J. Lin (1992) examined the syntactic differences between weishenme and zenme(-yang ). Tsai (1994) argues for the further need to divide zenme(-yang ) into manner (M) and instrument (I), and weishenme into reason and purpose. For example: (5) (a) [CP ta zenmeyang jianghua ] bijiao heshi? 3s I-how talk more proper ‘What is the means x such that her talking by x is more proper?’ (b) *[CP ta jianghua jiang-de zenmeyang ] bijiao heshi? 3s talk talk-DE M-how more proper ‘What is the manner x such that her talking in x is more proper?’ 737

Li Yafei

Tsai reasons that the instrumental zenmeyang is referential like an argument and thus patterns with the latter in the in-situ context. Concerns over the definition of referentiality are noted in Shi (1994), Hua (2000) and Hu (2002). gongzhong? (6) [DP[CP qiangdao ba renzhi zenme-le ] de xiaoxi ] jinu-le gangster BA hostage how-asp DE news infuriate-asp the.public ‘What is the action x such that the news of the gangsters doing x to the hostage infuriated the public?’ As Shi (1994) points out, zenme is used as a predicate in examples like (6) inside a complexNP island plus a subject island. It has the wide-scope reading over the entire sentence but in what sense is a predicate referential? Interestingly, (6) would be expected from Huang’s original analysis because the predicative zenme is structurally the complement of a lexical INFL and therefore aligns with arguments for movement. Worth pondering, though, is that INFL is now widely taken to be T, representing tense. What does it mean for tense to be lexical in Chinese but functional in English? It is often unavoidable that in active research, one must work with a non-ideal set of facts. As seen from above, the study of wh-in-situ is no exception. The efforts to understand this limited data divide according to whether movement is used, but both approaches aim to lessen the arbitrariness in allowing wh to cross islands in LF only. On the movement side, Nishigauchi (1986, 1990) used Japanese to argue for covert pied-piping: Islands always block movement, but covert movement has the option of moving the entire island containing the in-situ wh to the root Spec of CP. Fiengo et al. (1988) cited scope ambiguities in Chinese to reject simple pied-piping, and suggested a two-step derivation: Pied-pipe the island to some spot and then move the wh out of it, their key assumption being that a moved island is no longer an island, as technically made possible by Chomsky (1986). This analysis is critically reviewed in Shi (1994) and Hu (2002). Multiple authors opted for a (partially) non-movement account of wh-in-situ (e.g. Cheng 1991; Aoun and Li 1993; Shi 1994; Tsai 1994; Hu 2002). While details differ, all such works make use of C. Baker’s (1970) idea that some question-operator in C binds the in-situ wh interpreted as a variable. Since binding, unlike movement, is known to disregard islands, a bound in-situ wh may always take a wide-scope reading. See Cheng (2003) for a comprehensive review of the literature beyond Chinese. The primary question for this approach is how to formulate a coherent binding-based account of all and only those in-situ wh’s that are not restricted by islands. (7) are based on Hu (2002): (7) (a) ta shi wei-shenme cizhi? 3s Foc for-what resign (b) [CP ta wei-shenme cizhi ] geng heli? 3s for-what resign more plausible As noted by L. Xu (1990), either wei or shenme in weishenme may carry the stress. The former option yields the ‘why’ reading and the latter the ‘for what’ reading. In the presence of the focus shi, (7a) is good only when shenme is stressed. Matching this contrast, (7b) has the wide-scope reading for weishenme only when shenme carries the stress. In brief, an insitu wh can take a wide-scope beyond islands if and only if it can be focused. A wh-argument 738

Syntax

and zenme(yang) ‘how’ are compatible with the focus shi, so their ability for wide-scope readings is expected. Because focus is semantically associated with an alterative set, Hu suggests that wh-in-situ is ultimately based on the availability of an alternative set for a given wh-expression, implemented via Reinhart’s (1997) choice function supplemented with local LF movement. Another complication is whether wh-in-situ should have the same analysis for all types of islands. Most literature does not make (relevant) distinctions, but Liu (1986) and L. Xu (1990) pointed out that in Chinese, the wide scope of an in-situ wh inside a wh-island may have to do with it forming an echo question (which merely aims to verify the previous utterance). The examples in (8) are from Hu (2002): (8) (a) ta ganggang mai-le shenme (ne)? 3s just.now buy-asp SHENME Q (b) ta wen ni [CP Zhang Qing ganggang mai-mei-mai shenme ] (*ne)? 3s ask 2s ZQ just.now buy-not-buy SHENME Q Though the ‘question’ word shenme in (8a) is in itself ambiguous, meaning either ‘something’ or ‘what’, the question particle ne forces the latter reading. (8b) has been reported as allowing shenme, interpreted as ‘what’, to take the wide scope and thus ignore the wh-island created by mai-mei-mai ‘buy-not-buy’ (J. Huang’s (1982) A-not-A). If true, the unacceptability of ne in (8b) would be unexpected when compared with (8a). Hu concludes that shenme ‘what’ in (8b) must have only the echo-question reading, and that therefore wh-in-situ inside an indirect question is an entirely separate scenario from wh-in-situ in other types of islands. The nature of wh-movement, as characterized in PnP, leads to studies on Chinese relative clauses (Ning 1993) and topicalization (L. Xu and Langendoen 1985; Shi 1992). Also see Soh (2005) for arguing that all wh-forms be treated uniformly.

The serial verb constructions (SVCs) Broadly defined, an SVC consists of a single clause containing at least two lexical verbs neither of which is or heads a thematic argument of the other (Y. Li 1993a). Descriptively, Chinese has SVCs both in the phrasal form and as compounds: (9) (a) ta zhu xianrenzhang chi. 3s boil cactus eat ‘He boiled cacti and ate them’ (b) ta zhu-shou-le xiangrenzhang. 3s boil-cooked-asp cactus ‘He cooked cacti by boiling them’ Y. Li (1990) was the first to study the compounds in the PnP framework, focusing on explaining their semantic and syntactic relations with their arguments. In (9b), zhu ‘boil’ has two q-roles, Agent and Theme, and shou ‘cooked’ has Theme. In order for the three q-roles to match the two arguments supported by the monoclausal structure, the two Theme roles are merged and collectively given to the object (= Higginbotham’s (1985) q-identification), generating the reading that xianrenzhang ‘cactus’ is both what got boiled and what was cooked. 739

Li Yafei

This analysis also predicts the examples in (10) to exist: (10) (a) ta zhui-lei-le na zhi lang. 3s chase-tired-asp that cl wolf. ‘She chased that wolf and consequently she/the wolf became tired’ (b) ta chang-ku-le (daduoshu tingzhong). 3s sing-cry-asp most audience ‘Her singing made herself/most of the audience cry’ The ambiguity of (10a) results from the q-role of V2 optionally merging with either the Theme role or the Agent role of V1. In (10b), both V1 and V2 are intransitive, providing only two q-roles altogether. With Universal Grammar (UG) capable of syntactically supporting two arguments per clause (presumably via Case), q-identification becomes optional in (10b). With q-identification, the compound assigns both the Agent of V1 and the Theme of V2 to ta ‘3s’, making the latter both the singer and the crier. Without q-identification, ta receives Agent from V1 (due to V1 being the head of the compound) for the singer interpretation and tingzhong ‘audience’ gets Theme from V2 and is understood as the crying party. The analysis is also shown in Li (1993a) to explain why Chinese differs from Japanese in which V2 is the head. From this initial study arose two major questions, the first concerning the conditions under which resultative compounds form and operate. V2 seems to be restricted to stative and unaccusative verbs (Gu 1992; Ma and Lu 1997; D. Xu 2000; Packard 2001), presumably due to the event structure. Sometimes, the subject of the compound is not a proper subject of V1 (Gu 1992): (11) na-ping jiu zui-dao-le wo. That-cl wine be.drunk-fall-asp 1s ‘That bottle of wine made me drunk and fell down’ When zui ‘drunk’ is used alone, it does not take na-ping jiu ‘that bottle of wine’ as the subject. Examples of this kind, as well as the obligatory causer and causee interpretations on their respective arguments, prompted Y. Li (1995) to postulate an algorithm to compute causal roles off of any resultative construction and to let the causal roles override the q-roles of the component morphemes. Using idioms, Shibata et al. 2004 question whether V2 is ever thematically linked to the subject of the compound: (12) ta tong-guan-le mafengwo. 3s poke-accustomed-asp wasp.nest Tong mafengwo ‘poke wasp nest’ either has the literal meaning or refers idiomatically to stirring up a trouble. Both readings are available in (12) even though tong as V1 is separated from its object by V2. Further complicating the matter is the ‘liberal’ thematic readings of certain monosyllabic action verbs such as chi canguan ‘eat restaurant’ (T. Lin 2001). See J. Huang et al. (2009: Ch. 2), for how this phenomenon may interact with resultative compounds, especially when the latter are used like (11). The second major question with the resultative compounds is whether they are formed lexically or syntactically, a Chinese instance of the general debate on the nature of word 740

Syntax

formation since the late 1960s. While Y. Li (1990, 1999, 2005), Gu (1992), and Cheng and Huang (1994) are on the ‘lexicalist’ side, many scholars try to derive such compounds syntactically (Zou 1995; Shi 1998; Sybesma 1999; Shibata et al. 2004; among others). All works in this latter approach rely on a structure schematically shown below, with X moving to V to yield the final compound: (13) .  .  .  [VP  .  .  .  V  .  .  .  [XP  .  .  .  X  .  .  .  ]] where V and X does not necessarily correspond to V1 and V2. As an example, zui-dao ‘drunk-fall’ in (11) may be formed separately and correspond collectively to X, with V being a phonologically empty causative verb. Then (11) means, underlyingly, that that bottle of wine made me drunk and fall (cf. Sybesma 1999). Analyses utilizing (13) are often intuitively straightforward. A potential complication is illustrated with (14), adapted from Y. Li (1995): (14)

(a)



zhexie jiu neng rang Zhang Qing (?yilian san tian) zui-dao. these wine can make ZQ in.a.row three day be.drunk-fall ‘This much wine can make ZQ drunk to the floor (repeatedly for three days)’

(b) zhexie jiu neng zui-dao Zhang Qing (*yilian san tian). these wine can be.drunk-fall ZQ in.a.row three day Intended reading: Same as (14a). The duration phrase (DrP) in (14a) sounds somewhat unnatural, but the same DrP is simply unacceptable in (14b). This contrast is not expected if (14b), presumably sharing the same underlying structure as (11)/(14a), is derived simply by moving the compound up to the covert matrix causative verb. The word-formation debate is unlikely to be resolved soon. A new approach, articulated in Y. Li (2005) and mainly based on Semitic and Amerindian morphologies, combines lexicalist word-formation with his Morphology-syntax Mapping Hypothesis, which requires the semantic relation R between the morphemes in a word to be syntactically represented whenever possible. See Y. Li (forthcoming) for an application of this theory to Chinese resultative compounds. The phrasal SVCs present a different set of questions, the major two concerning the syntactic relation between the verbs and their linear order. The PnP literature on SVCs in other languages (M. Baker 1989; Collins 1997; Carstens 2002) generally assume the first verb V1 to take some projection of the second verb V2 as complement. Y. Li and Ting (2013) argue differently (with t standing for trace): (15) Sranan (Jansen et al. 1978) (a)

San what

Kofi Kofe

(b) (*)San what (c)

Mi 1s

e asp

teki take

Kofi Kofi koti cut

teki take a the

t

koti cut a the

a the

nefi knife

brede bread

brede? bread koti t? cut

nanga with 741

a the

nefi. knife

Li Yafei

(16) Chinese (a)

na-kuair that-cl

(b) *zhe-ba this-cl (c)

mianbao, bread

ta 3s

dao, knife

na take

ta 3s

na take t

zhe-ba this-cl qie cut

dao knife

na-kuair that-cl

qie cut

t.

mianbao. bread

zhe-xie zhongxuesheng, ni jingran rang t zuo daxue de ti! this-pl middle.schooler 2s actually make do college DE problem

That san ‘what’ in (15a) is extractable proves V1 to be the matrix verb. Sranan is a cleanly head-initial language, with both the complement and the adjunct (see (15c)) after the verb. For those Sranan speakers who accept (15b), the VP headed by V2 is taken as V1’s complement, whereas those rejecting (15b) treat this VP as the adjunct island for wh-movement. In contrast, the object of V2 moves freely in (16a) while that of V1 rejects topicalization in (16b). Further­ more, the NP after the causative verb rang ‘make’ in (16c) may also be topicalized. All the facts are expected if in (16a–b), VP1 is the adjunct for the matrix verb V2 and thus blocks movement. Rang ‘make’ is the matrix verb, permitting the following NP to move. In conclusion, even the same type of SVC (instrumental in (15)–(16)) may have different structures, contrary to the mainstream belief. The cross-linguistic data above highlights another trait of SVCs: the syntactic configuration may vary from language to language, but the linear order stays unchanged and, as the functionalist literature notes, iconic to the temporal sequence in which the corresponding events take place – one must get hold of the knife before cutting something with it, thus take precedes cut. Y. Li and Ting (2013) investigated this fact and suggest that iconicity fills up a functional gap of UG in sentence-construction, as was initially articulated in Y. Li (1993a). For a further development of this research, see Y. Li (2014).

The bei passives Earlier studies of the bei passive construction either paralleled it with the English passive, by moving the semantic object to the subject position, or explored generating all constituents in situ plus binding between the subject and the object (gap) of the verb (see J. Huang et al. 2009 for a recount of this literature). Feng (1990/1995) started a different approach, primarily motivated by examples like (17): (17)

(a)



Zhang Qing bei Li Kui henhen ZQ BEI LK severely ‘ZQ received a heavy beating by LK’

da-le beat-asp

(ta) 3s

yi-dun. a-cl

(b) Li Kui bei guanfu pai bing sha-le. LK BEI local.authority dispatch soldier kill-asp ‘LK was killed by the soldiers sent by the local authorities’ (17a) shows the optional use of a resumptive pronoun ta ‘3s’, which is known to replace the trace of movements landing only in non-argument (A’) positions such as the Spec of CP. Consequently, bei must be associated with A’-movement rather than argument (A) movement characteristic of English passivization. The conclusion is further confirmed by (17b), where the 742

Syntax

link between the matrix subject Li Kui and the gap after sha ‘kill’ crosses a clausal boundary. Again, this is known to happen for A’-movement but never to A-movement. (18) illustrates Feng’s structure for (17a): (18) [S1 ZQ bei [S2 Op LK severely beat t/him once ]] Op is a phonologically empty operator, somewhat comparable to a wh-word, that moves from t to the beginning of S2, a typical A’-movement since LK is the subject of S2 and occupies the highest A-position already. Semantically, Op turns a clause ‘back into a predicate’ and is coindexed with ZQ, yielding the reading that ZQ got beaten. Ting (1995, 1998) (also see Cheng et al. 1993) further divided the bei passive into the long form (with LK present in (17a)/(18)) and the short form (without LK ): (19) (a) Zhang Qing bei da-le (*ta) yidun. ZQ BEI beat-asp 3s a.cl

(cf. (17a))

(b) *Li Kui bei pai bing sha-le. LK BEI dispatch soldier kill-asp

(cf. (17b))

(20) (a) Zhang Qing bei Li Kui (suo) qiling. ZQ BEI LK SUO bully ‘ZQ got bullied by LK’ (b) Zhang Qing bei (*suo) qiling. ZQ BEI SUO bully (19) are the short counterparts of (17), and neither resumptive pronouns nor long-distance dependency is acceptable. Furthermore, the archaic suo is optional in the long passive (20a) but banned in the short version (20b). As shown in Chiu (1995) and later in Ting (2003), suo may occur inside a relative clause that contains a gap not associated with the subject or certain adjuncts. Since relative clauses are derived via A’-movement, (20) constitute direct evidence that only the long passive results from A’-movement. In a separate attempt, Feng (ibid.) suggests treating bei and V in the short passive as a compound which somehow loses its ability to provide a Case for the overt resumptive pronoun. But one should be aware that (19a) with ta ‘3s’ does not improve even when adverbs like henhende ‘severely’ are inserted between bei and V, making the compound analysis a more restricted availability (Ting 1995, 1998). That the long passive results from movement rather than base-generation plus binding, is evidenced by its sensitivity to islands (J. Huang et al. 2009; t being where ZQ is thematically interpreted): (21)

t de shihou ] nuqichongtian. *Zhang Qing bei Li Kui [XP zai jiaoxun ZQ BEI LK P teach.a.lesson DE time extremely.angry Intended reading: ‘LK was mad when he was teaching ZQ a lesson’

As for the short passive, there is a general consensus for it to be derived from A-movement. Ting (1995) aligns it directly with the English passive, whereas J. Huang et al. (2009) argue for a VP-internal A-movement (22b), prompted by the fact that the bei-passive, long and short alike, is compatible with a subject-oriented adverb: 743

Li Yafei

(22) (a) Zhang Qing guyi bei (Li Kui) pian-le. ZQ intentionally BEI LK cheat-asp ‘ZQ intentionally got cheated by LK’

(b) [S ZQ adv BEI [VP PRO cheat t ]]

where PRO moves from the t position to the Spec of VP and is bound by ZQ. The logic behind (22b) is that adverbs like guyi only modify verbs with a sentient subject, which means that ZQ receives a subject q-role from bei. Then the q-criterion prevents ZQ from originating as the object of pian ‘cheat’ since otherwise ZQ would end up with more than one q-role. While there is reason for explaining the data above through movement, the non-movement analyses of the bei-passive are not without basis. Consider (23): (23) (a) Li Kui bei guanfu zhua-zou-le yijialaoxiao. LK BEI local.authority arrest-away-asp whole.family ‘LK got all family members arrested by the local authorities’ (b) Zhang Qing bei Li Kui zhaoyin-lai-le henduo guanbing. ZQ BEI LK attract-come-asp many soldier ‘ZQ suffered from LK attracting many soldiers (to where he was)’ In (23a), LK is understood as the possessor of the object yijialaoxiao ‘entire family’, but whether the possessor of an argument can itself move away – aka possessor raising – has remained unclear. (23b) represents the ‘adversative passive’ found in East Asian languages, with the subject ZQ not linked thematically to any part of the post-bei constituents. In Chinese, this type of passive usually requires a context, but the fact that it is possible at all bears on how we understand bei. Overall, the lack of a ( justifiable) gap after bei appears compatible with the non-movement approach. Developing an idea of complex predicates in Huang (1982), J. Huang et al. (2009) proposed to alternatively treat (23a) as follows: (24) [S1 ZQ BEI [S2 Op local authorizes [VP  .  .  .  [V’ t [V’ arrest-away [NP e whole family ]]]]]] Op originates where t is and t binds the phonologically empty possessor e inside the object NP. Semantically, Op is the affected ‘outer object’ of the complex predicate V’ arrest the whole family. Essentially the same analysis is explored by Ting (1995, 1998), inspired by Thompson (1973). Similarly, a structurally even further ‘outermost object’ is postulated to analyze (23b): (25) [S1 ZQ BEI [S2 Op LK [VP t [VP  .  .  .  attract many soldiers ]]]] As such, most variations of the bei-passive now receive a unified analysis, which was also shown to extend to the Southern Min dialect and the short passive. But the analysis is not without concerns. First, an obligatorily empty operator is generally associated only with non-finite clauses (which is why Feng 1990/1995 originally compared the bei-passive with tough-movement in English which bans the Op-containing embedded clause from being finite). In (24), then, S2 would be non-finite and consequently unable to provide a Case for its subject local authorizes. Whether the matrix verb bei can help via exceptional Case-marking (ECM) isn’t clear either, because it is not known elsewhere that ECM can happen across an intervening Op. 744

Syntax

Another factor to consider is how necessary to distinguish between the outer and outermost objects. Both carrying a negative connotation, they may be thought of as the same malfactive argument. Bantu languages license such an argument through an applicative affix. Lacking a similar affix (perhaps due to Chinese being an isolated language), Chinese uses what is available, namely bei, to help bring out the malfactive. See Y. Li (2005) for treating the Bantu applicative affix as the matrix verb and thus structurally comparable to bei. Adding to the complication is the historical fact that bei used to mean ‘suffer’ or ‘receive’. It is logically possible, then, that this old use of bei has survived together with its newer passive version: The passive bei is used when there exists a q-criterion-dictated gap and the old form kicks in otherwise, partially explaining why (23) are generally perceived to be more marked. If tenable, we need no postulation of outer/outermost objects. The subject of bei simply suffers from the situation described by the embedded clause, leaving it to the context to determine the exact manner in which the subject is negatively affected. Such an analysis also extends naturally to the currently popular Internet examples such as bei zisha ‘suffer from suicide’. J. Xu (1999) notes another complication, that the bei-passive and the unaccusative have parallel behavior, especially when inalienable possession is involved: fuqin. (26) (a) Zhang Qing si-le ZQ die-asp father ‘ZQ had his father die on him’ (b) Zhang Qing bei (Li Kui) sha-le fuqin. ZQ BEI LI kill-asp father ‘ZQ suffered from his father being killed (by LK)’ In both, ZQ is the possessor of the postverbal NP object. But how does the postverbal NP satisfy the Case filter? As seen above, at least the short passive is said to pattern with the English passive and so has no object Case. The same holds for unaccusative verbs (Burzio 1986). Pan and Han (2008) futher noted that Belletti’s (1988) partitive Case doesn’t help because the postverbal NP can be definite as well. Pan and Han proposed to treat ZQ as the base-generated topic, with fuqin ‘father’ originating as the object of V, raising to the Spec of TP to get the nominative Case and then moving further to right-adjoin to TP for the purpose of focus. The structure is given in (27), which also explains the contrast in (28): (27) [CP ZQ C [TP[TP ti T [VP  .  .  .  V ti ]] fatheri ]] (28) (a) Zhang Qing, fuqin bei sha-le. ZQ father BEI kill-asp ‘As for ZQ, his father got killed’ (b) *fuqin Zhang Qing bei sha-le. father ZQ BEI kill-asp

(J. Xu 1999)

(28a) is simply (27) minus rightward focus-movement. (28b) is bad because there is no topic position to the left of the initial topic ZQ. Using a base-generated topic plus rightward focus movement of the subject, Pan and Han explained another mysterious fact in Chinese, where the word order appears to contradict the normal q-hierarchy (also see Her 2007): 745

Li Yafei

(29) (a) zhe yi-guo fan chi-le wu-ge ren. this a-pot food eat-asp five-cl person Literally: ‘This pot of food ate five people’

(b) [CP this pot of food C [TP[TP ti T [VP  .  .  .  eat  .  .  .  ]] five peoplei ]]

It is yet unclear whether this analysis may be reconciled with those by Feng and Ting.

Coordination Since the clausal structure was revised into TP/CP (Chomsky 1982), coordination becomes the only syntactic constituent still resisting conversion from exocentricity to endocentricity. Not surprisingly, then, there are continual efforts to decipher the structural nature of this last mystery (see Progovac 1998 for a comprehensive review of the literature). Adapting familiar tests in other languages to Chinese, N. Zhang (2006, 2010) supports an asymmetric arrangement of the conjuncts that lend themselves more easily to the endocentric X’-pattern, e.g.: (30)

(a)

meige xuesheng he tade jiazhang dou yao anshi every student and his parent all should on.time ‘Every student and his parent(s) should show up on time’

canjia. participate

(b) tade jiazhang he meige xuesheng dou yao anshi his parent and every student all should on.time ‘His parent(s) and every student should show up on time’

canjia. participate



The field generally believes that quantifiers like meige ‘every’ must c-command a pronoun for them to coindex to yield the bound variable reading. Since tade ‘his’ can be coreferential with meige xuesheng ‘every student’ only in (30a), one concludes that the first conjunct (Conj1) c-commands the second (Conj2) but not vice versa. Furthermore, Zhang argues for Conj1 and Conj2 to be the Spec and complement, respectively, of the coordinator X (31a) in the standard X’-format, contra alternatives such as Munn’s (1993) which treats X and Conj2 as forming an adjunct to Conj1 (31b). Part of Zhang’s evidence is given in (32)–(33): (31) (a) [XP Conj1 [X’ X Conj2 ]] (b) [Conj1 Conj1 [XP X Conj2 ]] (32)

(a)



Li Kui jueding chumen, yinwei/keshi Zhang Qing yao LK decide go.out because/but ZQ want ‘LK decided to go out, because/but ZQ wanted to play chess’

xiaqi. play.chess

(b) yinwei/*keshi Zhang Qing yao xiaqi, Li Kui jueding because/but ZQ want play.chess LK decide ‘Because/*but ZQ wanted to play chess, LK decided to go out’ (33)

(a)



chumen. go.out

zuowobuning ] de na-ju hua. zhe jiushi [CP Li Kui shuo-le t yushi this is LK say-asp and.then feel.restless DE that-cl words ‘These are the words that LK said and then felt restless’ 746

Syntax

(b) zhe jiushi [CP ta kan-le wode biaoyan yushi xiangchulai t ] de moshu. this is 3s watch-asp my show and.then think.of DE magic ‘This is the magic that she saw my show and then came up with’ The adjunct clause introduced by yinwei ‘because’ can either occur after the ‘kernel’ clause it modifies (32a) or be fronted (32b), as is expected from UG. But the [X Conj2] cluster headed by keshi ‘but’ cannot be moved. This is not predicted by the XP adjunct of (31b), but can be explained with (31a) because, independently, the ‘sub-phrase’ X’ generally resists movement. (33) demonstrates Op-extraction out of either Conj1 or Conj2 in relative clauses, further arguing against treating Conj2 as part of an adjunct (it must be noted, though, that movement out of the Spec is not clearly permitted by UG either, given the well-known subject island). Zhang not only uses such data to subsume coordination under the standard endocentric structure, but also dismisses the long-held view that coordination exhibits unique syntactic behavior. Drawing on the established semantic division between natural and accidental coordination, she suggests that at least on the natural side, coordination does not act like islands, as initially expressed in Ross’s (1967) Coordination Structure Constraint. Whereas Zhang aims to reduce coordination maximally to the better-understood UG system, Y. Li (2009) tries to understand another trait of the construction: that a conjunct may appear not to be a constituent and sometimes yield marginal acceptability: (34)

(a)



Li Kui qu-le sushe liang-ci, bangongshi yi-ci. LK go-asp dorm two-time office one-time ‘LK went to the dorm twice and the office once’

(b) Li Kui cong sushe qu-le bangongshi, ?(cong) bangongshi qu-le canting. LK from dorm go-asp office from office go-asp cafeteria ‘LK went to the office from the dorm and went to the cafeteria *(from) the office’ (c)

Li Kui cong sushe qu-le bangongshi, wo *(cong) bangongshi LK from dorm go-asp office 1s from office qu-le canting. go-asp cafeteria ‘LK went to the office from the dorm, I went to the cafeteria *(from) the office’ The object and a frequency adjunct form a conjunct in (34a) and sound perfect. The conjunct in (34b) is either the entire VP or the VP minus cong ‘from’. The latter option detectably drops in acceptability and requires a strongly parallel intonation. (34c) is the least natural, again with cong missing in the second part of the sentence. Starting with Ross (1967), there have always been efforts to handle such data with two steps: Identify the (smallest) constituent containing everything in a conjunct, and delete the repetitious constituent to yield the superficial outcome. Li argues that this approach is inadequate for the marginal cases. For example, both (34b–c) would involve deleting cong, but they are not equally acceptable. Furthermore, their English translations also indicate that simply deleting from leads to bad sentences, inconsistent with what supposedly happens in (34b). The alternative analysis given by Li is to take preposition-deletion as independently disallowed (cf. *I’m from Madison and she’s, Chicago.) while defining an algorithm to measure the number of phrases that the ‘linking’ must cross between adjacent constituents inside a conjunct. (34a–b) have the relevant structures below, respectively: 747

Li Yafei

(35) (a) .  .  .  [VP [V’ go-to dorm] [ADVP twice ]] (b) .  .  .  [VP [PP from dorm] [V’ go-to office ]]

– for (34a) – for (34b)

The conjunct dorm twice in (35a) has two adjacent constituents NP and AdvP, with VP being the smallest constituent containing both. There is no phrasal node between NP and VP, nor between VP and AdvP. Other than the shared VP, the total number of phrases traversed from NP to AdvP is 0, and the result is good. In (35b), dorm is NP and forms a conjunct with V’. The smallest containing constituent is still VP. Between VP and dorm is PP, and between VP and V’ is nothing. The number of phrases traversed from NP to V’ is 1 and the sentence deteriorates. The rest of the Chinese and English sentences in (34) are all bad due to illegitimately deleting cong/‘from’. See Y. Li (2009) for details. This analysis critically does not equate conjuncts to constituents, thereby forcing a parallel or three-dimensional structure for coordination. How it reconciles with Zhang’s theory is no simple question to answer. Worth noting is that it offers a conceptually simpler account of certain facts that Zhang aims to address, such as interwoven dependency: (36) [NP Which nurse and which hostess ] did Ernest [VP date t and George marry t ]? Standard operations in PnP have a hard time ‘distributing’ each wh-phrase to its corresponding object location. Zhang’s theory appeals to sideward extraction, essentially a formalized admission that standard movement targeting a c-commanding position doesn’t work here. In a three-dimentional structure, each wh-phrase and its matching VP exist on their own plane on which wh-movement can target the c-commanding Spec of CP as in normal situations.

Inside Determiner Phrase (DP) In addition to its quantificational traits (A. Li 1998), Chinese nominal phrases also raise questions on their internal design. Since Abney (1987), NP becomes the complement of D, and numerous works also argue for a NumP mediating between DP and NP (e.g. Valois 1991): (37) [DP  .  .  .  D [NumP numeral Num [NP  .  .  .  N  .  .  .  ]]] But (37) offers no accommodation for classifiers (Cl) in Chinese, which A. Li (1999) addresses through (38) on the basis of the examples in (39) (also see Simpson et al. 2011): (38) [DP  .  .  .  D [Num numeral Num [ClP  .  .  .  Cl [NP  .  .  .  N  .  .  .  ]]]] (39) (a) student-s vs. xuesheng-men student-pl (b) four student-s vs. si-ge xuesheng(*-men) four-cl student-pl An English DP has (37) with the plural –s in Num. N moves up to Num to yield student-s in (39a). This derivation is independent of the numeral four (39b). In Chinese, the plural –men is also in Num, but N-to-Num movement is blocked by the intervening Cl (cf. Rizzi’s (1990) relativized minimality). Hence the numeral and –men cannot cooccur. If ClP is absent (and the numeral goes with it for a separate reason), (38) is in effect identical to (37), allowing N to move directly to Num and produce xuesheng-men. 748

Syntax

This analysis crucially makes Cl reject –men. The examples below seem to suggest, however, that what is incompatible with –men is not CL but a numeral: (40)

(a)



naxie(-ge) xin yuangong(-men) zhengzai those-cl new employee-pl asp ‘Those new employees are having a meeting’

kaihui. have.a.meeting

(b) shiji-ming benke de haizi(-men) neng you shenme gaoshen ten.plus-cl college DE child-pl can have what profound de wenti? DE question ‘How profound are the questions a dozen or so college kids can ask?’ (c)

zhe ji-wei zhuanjia(-men) qing dao tai-shang this a.few-cl expert-pl please come.to stage-on ‘These few experts, please come to be seated on the stage’

jiuzuo. be.seated

Provided the quantity is vague, –men and Cl show no mutual rejection, indicating that the semantics of –men is best described as ‘vague plural’ (Y. Li 2015). Another complication is illustrated with (41): (41) (a) san-ge [ meiyou xue-guo daishu de  ] three-cl have.not study-asp algebra DE ‘three students who haven’t studied algebra’

xuesheng student

(b) [ meiyou xue-guo daishu de ] xuesheng-men have.not study-asp algebra DE student-pl ‘the students who haven’t studied algebra’ (c) xuesheng-men [ meiyou xue-guo daishu de ] student-pl have.not study-asp algebra DE ‘those who haven’t studied algebra among the students’ With Cl as the reference point in (41a), the default position of the relative clause is either inside NP or adjoined to it. Since –men is present in (41b), xuesheng ‘student’ must have moved out of NP to Num under A. Li’s analysis. But such a derivation would wrongly leave the relative clause behind. The predicted word order is indeed possible in (41c), though with a detectably different meaning. Also related is (42): (42) [ xin lai de ] xuesheng he [ yao tuixiu de ] laoshi-men new come DE student and will retire DE teacher-pl ‘the newly arrived students and the retiring treachers’ The default reading of (42) is for –men to have scope over both xuesheng ‘student’ and laoshi ‘teacher’. Since each conjunct is clearly a phrase (shown by the corresponding relative clause), mere head-movement is insufficient to make one –men ‘pluralize’ both nouns. A more likely analysis is for –men to be a clitic onto the entire preceding coordinate structure. The uniformly head-final (43) is a structure capable of producing all the facts so far (but is yet to be tested on other data in A. Li 1999): 749

Li Yafei

(43) [DP  .  .  .  [NumP numeral-cl [NP Relative-clause  .  .  .  N ] Num ] D ] Spec-head agreement guarantees mutual exclusion between a precise-quantity numeral and –men in Num, with the latter cliticizing to the preceding NP. The head-final DP simply models that in Japanese, a strictly head-final language with demonstrative pronouns starting instead of trailing the DP.

And more  .  .  . Significant works in Chinese syntax are definitely not limited to the few topics above. Starting with Wang and Stillings (1984), the long-distance anaphor ziji has triggered a large amount of literature (Cole et al. 1990; J. Huang and Tang 1991; Y. Li 1993b; Pan 1997, 2001; J. Huang and Liu 2001; Hu 2002; among others). The ba morpheme, conditionally introducing an internal argument, is another major issue (A. Li 1990, 2006; Y. Li 1990, 1995; Zou 1995; J. Li 1997; Liu 1997; Sybesma 1999; Bender 2000; to list just a few). A popular approach is to take ba as a matrix V or a light verb with the lexical verb as its complement. Worth keeping in mind but often ignored is that in reality, ba and the NP it helps introduce can generally be fronted: (44) (a) ba zhexie tudou, ni xian xiao-le pi. BA these potato 2s first peel-asp skin ‘You get these potatoes peeled first’ (b) ba Qiaomusiji de shu, ta quandou mai-le. BA Chomsky DE book 3s all sell-asp ‘She sold all the books by Chomsky’ This fact would seem to argue against treating ba as the matrix verb but for paralleling it with a preverbal PP adjunct. Minimally, then, a clarification of the basic data is called for at this stage of the research. While most adjuncts occur before the head in Chinese, frequency and duration phrases are, or can be, postverbal and therefore beg an explanation (Kung 1993; Soh 1998; J. Huang et al. 2009). In addition to the works on the morphology-syntax interface (cf. the section on SVCs), research on the syntax-semantics interface has also been productive, with S. Huang (1996) studying the quantificational dou, J. Lin (1996, 1998, 2003, 2006) on polarity and tense, Cheng and Huang (1996) and Pan and Jiang (1997) on donkey sentences, and P. Lee (2012) on Cantonese aspectual and quantificational particles. Regarding the syntaxphonology interface, Feng (1994, 2000) and H. Zhang (1997) have both investigated how structure and prosody interact in Chinese, with quite different conclusions. The theory of Chinese syntax has also helped prominent interdisciplinary studies, such as the psycholinguistic research on the acquisition of Mandarin Chinese and Cantonese by T. Lee (2003, 2012). As significant as its accomplishments are, Chinese syntax as a field still has room for improvement, one place being data robustness. As an example, recall that the contrast in (5a–b) is used to distinguish between the preverbal instrument zenmeyang ‘how’ and the post-V-de manner variant, with respect to their ability for a wipe-scope reading. But consider (45)–(46): (45) *[CP ta jianghua jiang-de hen jingcai/qingchu/kouruoxuanhe ] bijiao heshi. 3s talk talk-DE very brilliant/clear/eloquent more proper ‘That she talks brilliantly/clearly/eloquently is more appropriate’ 750

Syntax

(46) (a) [CP ta pao-de hen congrong/mantengtengde  ] dui gongsi geng youli. 3s run-DE very leisurely/slow to company more beneficial ‘That he runs leisurely/slowly is more beneficial to the company’ geng youli? (b) [CP ta pao-de zenmeyang ] dui gongsi 3s run-DE how to company more beneficial ‘What is the manner x such that her running in x is more beneficial  .  .  .’ Being the answer to (5b), (45) should be the control sentence. If (45) is bad already, (5b) does not count as evidence that the manner zenmeyang is bad in situ. In fact, when the control sentence (46a) is good, turning the manner adverb into an in-situ wh sounds more acceptable. In brief, whether there is a manner-instrument distinction for wh-in-situ cannot be decided on the basis of (5) alone. Another example is (30a), where the bound-variable reading is taken to mean that Conj1 c-commands Conj2. Now consider (47): tade jiazhang ye (47) meige xuesheng dou yao canjia, every student all should participate his parent also jingkeneng anshi daochang. as.far.as.possible on.time arrive ‘Every student should participate, and his parents should also make an effort to arrive on time’ For the bound-variable readings, my informants find no detectable difference in acceptability between (30a) and (47). But the quantifier phrase meige xuesheng in (47) is the subject of the first clause and thus cannot possibly c-command the bound variable tade ‘his’. Note that (47) doesn’t dismiss c-command for the bound-variable interpretation; it may merely show that c-command in overt syntax is not a necessary condition, at least not in Chinese. Consequently, (30a) loses its force in establishing the asymmetric relation between Conj1 and Conj2 in overt syntax. Another area for future improvement is how the field of Chinese syntax views its own role in general linguistics. There are four ways for a particular language L to participate significantly in a universal theory (UT). Apply UT to L to solve unique problems in L. The works on the bei-passives, the ba-construction and the DP-internal structure are prime cases of this sort. (ii) Use facts from L to assess a hypothesis already postulated in UT. The extensive literature on wh-in-situ falls into this category. (iii) With facts from L, help tidy up part of UT. J. Huang’s CED and, if eventually proven to be valid, N. Zhang’s effort to ‘normalize’ coordination are the representatives thereof. (iv) On the basis of L, motivate entirely new content or affect the design for UT. Y. Li (1993a) argued for the inadequacy of the PnP model to explain SVCs and the need for UG to collaborate with iconicity. With concrete data and detailed (and thus testable) theory, this synthetic approach to language pioneers the investigation of what Chomsky (2007) recognizes as the ‘principles not specific to [the language faculty]’.

(i)

Given the immense amount of research in Chinese syntax, the last type is disproportionally scarce. It will be to the long-term health of the field to ponder the reasons and, hopefully, to eventually develop a more balanced research program. 751

Li Yafei

List of abbreviations 1,2,3 [person feature] adv adverb asp Aspect marker CED Condition on Extraction Domain cl classifier DrP Duration phrase ECP Exceptional Case marking Foc Focus marker GB (the theory of ) Government and Binding INFL inflectional morpheme taken to the head of a clause LF Logical Form Num Number (a functional head inside a nominal phrase) Op Operator (typically in the phonologically null form) [plural] (a number feature when occurring after a person feature pl PnP Principles-and-Parameters Q Question marker [singular] (a number feature when occurring after a person feature) s SVC Serial verb construction UG Universal Grammar wh (referring to) interrogative pronouns that start with wh in English

References Abney, Steven (1987) ‘The English noun phrase in its sentential aspect’, PhD dissertation, MIT. Aoun, Joseph and Audrey Y.-H. Li (1993) ‘Wh-elements in-situ: Syntax or LF?’ Linguistic Inquiry 24: 199–238. Baker, C. Lee (1970) ‘Notes on the description of English questions: The role of an abstract question morpheme’, Foundations of Language 6: 197–219. Baker, Mark (1989) ‘Object sharing and projection in serial verb construction’, Linguistic Inquiry 20: 513–53. Bellitti, Adriana (1988) ‘The case of unaccusatives’, Linguistic Inquiry 19: 1–34. Bender, Emily (2000) ‘The syntax of Mandarin bǎ: Reconsidering the verbal analysis’, Journal of East Asian Linguistics 9: 105–45. Burzio, Luigi (1986) Italian syntax: A government-binding approach, Dordrecht: Reidel. Carstens, Vicki (2002) ‘Antisymmetry and word order in serial verb consructions’, Language 78: 3–50. Cheng, Lisa L.-S. (1991) ‘On the typology of wh-questions’, PhD dissertation, MIT. Cheng, Lisa L.-S. (2003) ‘Wh-in-situ’, Glot International 7, 103–9 and 129–37. Cheng, Lisa L.-S. and James C.-T. Huang (1994) ‘On the argument structure of resultative compounds’, in Matthew Chen and Ovid Tzeng (eds) In honor of William Wang, Taipei: Pyramid Press, 187–221. Cheng, Lisa L.-S. and James C.-T. Huang (1996) ‘Two types of donkey sentences’, Natural Language Semantics 4: 121–63. Cheng, Lisa L.-S., James C.-T. Huang and Audrey Y.-H. Li. (1993) ‘Three ways to get passive’, Ms., University of California, Irvine, and University of Southern California. Chiu, Bonnie (1995) ‘An object clitic projection in Mandarin Chinese’, Journal of East Asian Linguistics 4: 77–117. Chomsky, Noam (1976) ‘Conditions on rules of grammar’, Linguistic Analysis 2: 303–51. Chomsky, Noam (1977) ‘On wh-movement’, in P. Culicover, T. Wasaw, and A. Akmajian (eds) Formal Syntax, New York: Academic Press, 71–132. Chomsky, Noam (1982) Some concepts and consequences of the theory of government and binding, Cambridge, MA: MIT Press.

752

Syntax Chomsky, Noam (1986) Barriers, Cambridge, MA: MIT Press. Chomsky, Noam (2007) ‘Approaching UG from below’, in Uli Sauerland and Hans-Martin Gartner (eds) Interfaces + recursion = language? Chomsky’s minimalism and the view from syntax-semantics, Berlin: Mouton de Gruyter, 1–29. Cole, Peter, Gabriella Hermon and Li-May Sung (1990) ‘Principles and parameters of long distance reflexives’, Linguistic Inquiry 21: 1–22. Collins, Chris (1997) ‘Arguming sharing in serial verb constructions’, Linguistic Inquiry 28: 461–97. Feng, Shengli (1990/1995) ‘Guanyue lilun yu Hanyu de beidong ju’ (‘GB theory and the passive sentences in Chinese’), Zhongguo Yuyanxue Luncong (‘Studies in Chinese Linguistics’) 1: 1–28. Feng, Shengli (1994) ‘Prosodic structure and compound words in Classical Chinese’, in Jerome Packard (ed) New approaches to Chinese word formation, Berlin: Mouton de Gruyter, 197–260. Feng, Shengli (2000) Prosodic Syntax in Chinese, Shanghai: Shanghai Education Press. Fiengo, Robert, James C.-T. Huang, Howard Lasnik, and Tanya Reinhart (1988) ‘The Syntax of wh-in-situ’, Proceedings of the West Coast Conference on Formal Linguistics 7: 81–98. Gu, Yang (1992) ‘The syntax of resultative and causative compounds in Chinese’, PhD dissertation, Cornell University. Her, One-soon (2007) ‘Argument-function mismatches in Mandarin Chinese: A lexical mapping account’, Lingua 117: 221–46. Higginbotham, James (1985) ‘On semantics’, Linguistic Inquiry 16: 547–93. Hu, Jianhua (2002) ‘Prominence and locality in grammar: The syntax and semantics of wh-questions and reflexives’, PhD dissertation, City University of Hong Kong. Hua, Dongfan (2000) ‘On wh-quantification’, PhD dissertation, City University of Hong Kong. Huang, James C.-T. (1982) ‘Logical relations in Chinese and the theory of grammar’, PhD dissertation, MIT. Huang, James, C.-T., Audrey Y.-H. Li, and Yafei Li (2009) The Syntax of Chinese, Cambridge: Cambridge University Press. Huang, James C.-T. and C.-S. Luther Liu (2001) ‘Logophoricity, attitudes and ziji at the interface’, in Peter Cole, G. Hermon, and C.-T. James Huang (eds) Long Distance Reflexives, New York: Academic Press, 141–95. Huang, James C.-T. and Jane C.-C. Tang (1991) ‘The local nature of the long-distance reflexives in Chinese’, in Jan Koster and Eric Reuland (eds) Long-distance Anaphora, Cambridge: Cambridge University Press, 263–82. Huang, Shizhe (1996) ‘Quantification and predication in Mandarin Chinese: A case study of dou’, PhD dissertation, University of Pennsylvania. Jansen, Bert, Hilda Koopman and Pieter Muysken (1978) ‘Serial verbs in the creole languages’, Amsterdam Creole Studies 2: 125–59. Kayne, Richard (1979/1981) ‘Two notes on the NIC’, in A. Belletti, L. Brandi and L. Rizzi (eds) Theory of markedness in generative grammar: Proceedings of the 1979 GLOW Conference, Scuola Normale Superiore, Pisa, 317–46. Kung, Hui-I (1993) ‘The mapping hypothesis and the postverbal structure in Mandarin Chinese’, PhD dissertation, University of Wisconsin-Madison. Lee, Peppina (2012) Cantonese Particles and Affixal Quantification, Dordrecht: Springer. Lee, Thomas H.-T. (2003) ‘Two types of logical structure in child language’, Journal of Cognitive Science 3: 155–82. Lee, Thomas H.-T. (2012) ‘Quantificational structures in three-year-old Chinese-speaking children’, in San Xu (ed) Plurality and Classifiers Across Languages of China, Berlin: Mouton de Gruyter, 243–82. Li, Audrey Y.-H. (1990) Order and Constituency in Mandarin Chinese, Dordrecht: Kluwer. Li, Audrey Y.-H. (1998) ‘Argument determiner and number phrases’, Linguistic Inquiry 29: 693–702. Li, Audrey Y.-H. (1999) ‘Plurality in a classifier language’, Journal of East Asian Linguistics 8: 75–99. Li, Audrey Y.-H. (2006) ‘Chinese ba’, in Martin Everaert and Henk van Riemsdjik (eds) The Blackwell Companion to Syntax, Malden, MA: Blackwell, 347–468. Li, Jen-I Jelina (1997) ‘The ba construction in Mandarin Chinese: A serial verb analysis’, in James Blake and Virginia Motapanyane (eds) Current Issues in Linguistic Theory 140: Clitics, Pronouns and Movement, Amsterdam: John Benjamins: 175–216. Li, Yafei (1990) ‘On V-V compounds in Chinese’, Natural Language and Linguistic Theory 8: 177–207. Li, Yafei (1993a) ‘Structural head and aspectuality’, Language 69: 480–504.

753

Li Yafei Li, Yafei (1993b) ‘What makes long distance reflexives possible?’ Journal of East Asian Linguistics 2: 135–66. Li, Yafei (1995) ‘The thematic hierarchy and causativity’, Natural Language and Linguistic Theory 13: 155–82. Li, Yafei (1999) ‘Cross-componential causativity’, Natural Language and Linguistic Theory 17: 445–97. Li, Yafei (2005) X 0: A Theory of the Morphology-syntax Interface, Cambridge, MA: MIT Press. Li, Yafei (2009) ‘Cong binglie jiegou de jufa tiaojian kan bianyuan yuliao de lilun yiyi’ (‘The syntactic structure of coordination and the significance of marginal data’), Dangdai Yuyanxue 4, 289–98. Li, Yafei (2014) ‘Formal syntax, iconicity and the studies of Chinese’, Zhongguo Yuwen 363(6): 521–30. Li, Yafei (2015) ‘Ye tan Huanyu mingci duanyu de meibu jiegou’ (‘Revisiting the internal structure of Chinese noun phrases’), Zhongguo Yuwen 265(2): 9–104. Li, Yafei (forthcoming) ‘Argument structure’, Encyclopedia of Chinese Language and Linguistics, Leiden: Brill. Li, Yafei and Jen Ting (2013) ‘Is UG like a chunk of Swiss cheese? Evidence for a “third-factor principle” ’, Language and Linguistics 14: 737–54. Lin, Jo-Wang (1992) ‘The syntax of zenmeyang “how” and weishenme “why” in Mandarin Chinese’, Journal of East Asian Linguistics 1: 293–331. Lin, Jo-Wang (1996) ‘Polarity licensing and wh-phrase quantification in Chinese’, PhD dissertation, University of Massachusetts, Amherst. Lin, Jo-Wang (1998) ‘On existential polarity wh-phrase in Chinese’, Journal of East Asian Linguistics 7: 219–55. Lin, Jo-Wang (2003) ‘Temporal reference in Mandarin Chinese’, Journal of East Asian Linguistics 12: 259–311. Lin, Jo-Wang (2006) ‘Time in a language without tense: The case of Chinese’, Journal of Semantics 23: 1–53. Lin, Tzong-Hong (2001) ‘Light verb syntax and the theory of phrase structure’, PhD dissertation, University of California, Irvine. Liu, Feng-hsi (1986) ‘On topic-traces in Chinese’, Proceedings of the West Coast Conference on Formal Linguistics 5: 142–53. Liu, Feng-hsi (1997) ‘An aspectual analysis of ba’, Journal of East Asian Linguistics 6: 51–99. Ma, Zhen and Lu Jianming (1997) ‘Xingrongci zuo jieguo buyu qingkuang de kaocha’ (‘A survey of adjectives as result complements’), in A. O. Yue and Mitsuaki Endo (eds) In Memory of Mantaro J. Hashimoto, Tokyo: Uchiyama Books, 155–72. May, Robert (1977) ‘The grammar of quantification’, PhD dissertation, MIT. Munn, Alan (1993) ‘Topics in the syntax and semantics of coordinate structures’, PhD dissertation, University of Maryland, College Park. Ning, Chunyan (1993) ‘The overt syntax of relativization and topicalization in Chinese’, PhD dissertation, University of California, Irvine. Nishigauchi, Taisuke (1986) ‘Quantification in syntax’, PhD dissertation, University of Massachusetts, Amherst. Nishigauchi, Taisuke (1990) Quantification in the Theory of Grammar, Dordrecht: Kluwer. Packard, Jeremy (2001) ‘The morphology of Chinese: A linguistic and cognitive approach’, Bejing: Foreign Language Press. Pan, Haihua (1997) Constraints on Reflexivization in Chinese, New York: Garland. Pan, Haihua (2001) ‘Why the blocking effect?’ in Peter Cole, G. Hermon, and C.-T. James Huang (eds) Long Distance Reflexives, New York: Academic Press, 279–316. Pan, Haihua and J. Han (2008) ‘Hanyu baoliu bingyu jiegou de shengcheng jizhi’ (‘The syntactic mechanism of retained object constructions in Chinese’), Studies of the Chinese Language 6: 511–22. Pan, Haihua and Yan Jiang (1997) ‘NP interpretation and Chinese donkey sentences’, in Proceedings of the Workshop on Interface Strategies in Chinese, Cornell University. Progovac, Ljiljana (1998) ‘Structure for coordination’, Glot International 3: 3–6, 6–9. Reinhart, Tanya (1997) ‘Quantificational scope: How labor is divided between QR and choice-functions’, Linguistics and Philosophy 20: 335–97. Rizzi, Luigi (1990) Relativized Minimality, Cambridge, MA: MIT Press.

754

Syntax Ross, John (1967) ‘Constraints on variables in syntax’, PhD dissertation, MIT. Shi, Dingxu (1992) ‘The nature of topic-comment constructions and topic chains’, PhD dissertation, University of Southern California. Shi, Dingxu (1994) ‘The nature of Chinese wh-questions’, Natural Language and Linguistic Theory 12: 301–33. Shi, Dingxu (1998) ‘The complex nature of V–C construction’, in Yang Gu (ed) Studies in Chinese Linguistics, Hong Kong: Linguistic Society of Hong Kong, 23–52. Shibata, Natsumi, Yasutada Sudo and Jun Yashima (2004) ‘On (non)-resultative V–V compounds in Chinese’, Nanzan Linguistics 2. Simpson, Andrew, Hooi Ling Soh and Hiroki Nomoto (2011) ‘Bare classifiers and definiteness: A cross-linguistic investigation’, Studies in Language 35: 168–93. Soh, Hooi Ling (1998) ‘Object scrambling in Chinese: A close look at post-durational/frequency phrase positions’, NELS 28: 197–211. Soh, Hooi Ling (2005) ‘Wh-in-situ in Mandarin Chinese’, Linguistic Inquiry 36: 143–55. Sybesma, Rint (1999) The Mandarin VP, Dordrecht: Kluwer. Thompson, Sandra (1973) ‘Transitivity and the ba construction in Mandarin Chinese’, Journal of Chinese Linguistics 1: 208–11. Ting, Jen (1995) ‘A non-uniform analysis of the passive construction in Mandarin Chinese’, PhD dissertation, University of Rochester. Ting, Jen (1998) ‘Deriving the bei-construction in Mandarin Chinese’, Journal of East Asian Linguistics 7: 319–54. Ting, Jen (2003) ‘The nature of the particle suo in Mandarin Chinese’, Journal of East Asian Linguistics 12: 121–39. Tsai, Wei-tien (1994) ‘On economizing A-bar dependencies’, PhD dissertation, MIT. Valois, Daniel (1991) ‘The internal syntax of DP’, PhD dissertation, University of California, Los Angeles. Wang, Jialing and Justine Stillings (1984) ‘Chinese reflexives’, in Chunyan Ning (ed) Proceedings of the First Harbin Conference on Generative Grammar, Harbin: Heilongjiang University Press, 100–9. Xu, Dan (2000) ‘Dongbu jiegou zhong de shangzi yu xiazi’ (‘V1 and V2 in the RVC’). in Yufa Yanjiu he Tansuo (‘Grammar Research and Exploration’). Beijing: The Commercial Press, 112–120. Xu, Jie (1999) ‘Liangzhong baoliu binyu jushi ji xiangguan jufa lilun wenti’ (‘Two retained object constructions and their theoretical implications’), Dangdai Yuyanxue 1: 16–29. Xu, Liejiong (1990) ‘Remarks on LF movement in Chinese questions’, Linguistics 28: 355–82. Xu, Liejiong and Terence Langendoen (1985) ‘Topic structures in Chinese’, Language 61: 1–27. Zhang, Hongming (1997) ‘The c-command approach to morphosyntax’, Chinese Language and Linguistics 3. Zhang, Ning (2006) ‘On the configuration issue of coordination’, Language and Linguistics 7: 175–223. Zhang, Ning (2010) Coordination in Syntax, Cambridge: Cambridge University Press. Zou, Ke (1995) ‘The syntax of the Chinese ba-construction and verb compounds: A morpho-syntactic analysis’, PhD dissertation, University of Southern California.

755

42 Wade–Giles Romanization system Karen Steffen Chung National Taiwan University

The Wade–Giles Romanization system for standard Mandarin Chinese held a distinguished place of honor in Sinology and popular usage from the late nineteenth century until the 1970s, when it began losing ground to Hanyu Pinyin. But that is not to say that the Wade– Giles system was not, and is not still, without its problems, and consequently, its sometimes highly vocal detractors.

Historical absence of a phonetic alphabet, fanqie and tone marking It is surprising that the Chinese did not develop their own phonetic alphabet before the arrival of Western missionaries in China starting in the sixteenth century. The closest they came was the use of the fǎnqiè system, under which two relatively well-known characters, plus the word fǎn 反 or later mostly qiè 切, were given after a lexical item. The reader needed to take the initial of the first and splice it onto the final rhyme and tone of the second, to derive the pronunciation of the item being looked up. A typical entry is dōng déhóng qiè 東 德紅切, i.e. dé plus hóng in the qiè 切 system make dōng. (The second tone had not yet separated from the first at this time, thus the difference in tones.) One big advantage of the system is that the fǎnqiè characters were already familiar to any literate Chinese, so there was no need to learn a new set of symbols. The disadvantage is that there is no way to know with certainty the actual phonetic realizations of the syllables at the time  .  .  . (Chung 2013: 216) And instead of developing an alphabet or syllabary,   .  .  .  the 10th-century monk Shouwen 守溫, who was possibly not an ethnic Han, developed an ‘alphabet’ for phonetic notation of Chinese characters for use in the rhyme books. It is interesting that, in spite of having the Sanskrit Devanāgarī alphabet as a model, he did not develop an alphabet or syllabary, but instead chose 756

Wade–Giles Romanization system

30 existing Chinese characters to represent consonant or vowel initials. They were arranged in an order similar to that of the Sanskrit alphabet, according to, for example, whether a sound was voiced, voiceless, or voiceless aspirated. This set was later expanded to 36. The lack of a set of symbols indicating the values of individual segments is a big drawback of the system, but it does give us valuable categorical information on Middle Chinese. (Chung 2013: 216) Chinese scholars were well aware of the different possible phonetic values of fǎnqiè characters according to geographical dialect and historical period. They apparently did not, however, feel the lack of an alphabetic-type phonetic notation system acutely enough to design and adopt one for Chinese, in spite of its potential usefulness in dictionaries, rhyme books, recording dialects, teaching foreigners Chinese, and other applications. A number of different strategies were adopted over history to indicate the tones of spoken Chinese. Attention to tone in rhyming syllables in early Chinese poetical works like the Book of Songs (Shī Jīng 詩經, ca. tenth–eighth century bc), is proof of early implicit awareness among the Chinese of the tonal categories. Exposure to Sanskrit in the course of translating the Buddhist sutras made the Chinese more explicitly aware of the tones as a key feature of their language. Shěn Yuē 沈約 (ad 441–513), in his Sìshēngpǔ 四聲譜 ‘Tables of the Four Tones,’ unfortunately no longer extant, is attributed with being the first to label and describe the tones, as follows: píng 平 (‘level’), shǎng 上 (‘rising’), qù 去 (‘falling’), and rù 入 (‘entering’), used to refer to syllables ending with a /-p/, /-t/, /-k/ or a glottal consonant stop final. The word labels and table format continued to be the usual way tones were explicitly categorized in Chinese rhyme books, such as the Guǎngyùn 廣韻 ‘Expanded Rhymes’ compiled by Chén Péngnián 陳彭年 (ad 961–1017) (Chung 2013: 215–17). Around the same time, another method, the sìshēng quāndiǎn 四聲圈點 (‘Four Tone Circle Marking’), came into use. In this method, the four tones were indicated by a circle, semicircle, or other mark written in one of the four corners of a Chinese character, starting with the píng being marked in the lower left-hand corner and moving clockwise around the character through the rù, which was marked in the lower right-hand corner (Branner 1997: 251; Simmons: forthcoming).

Early efforts to Latinize Chinese The first efforts at Romanizing Chinese can be traced back to Portugal’s quest for new lands to colonize and the Vatican’s for souls to win for the Church in the fifteenth and sixteenth centuries. In 1455, Pope Nicholas V issued the Romanus Pontifex papal bull in which King Alfonso V of Portugal was given permission, in ‘any of the provinces, islands, harbors, seas, and places whatsoever, acquired or possessed in the name of King Alfonso’, to: .  .  .  found and [cause to be] founded and built any churches, monasteries, or other pious places whatsoever; and also may send over to them any ecclesiastical persons whatsoever, as volunteers, both seculars, and regulars of any of the mendicant orders (with license, however, from their superiors), and that those persons may abide there as long as they shall live, and hear confessions of all who live in the said parts or who come thither, and after the confessions have been heard they may give due absolution in all cases  .  .  . (Romanus Pontifex: 1455) 757

Karen Steffen Chung

Under this Portuguese-directed effort of the Roman Catholic church, European missionaries set out for Goa, India, Japan, and China, among other Asian destinations. A number of them settled in Portuguese-administered Macau, but their effectiveness was limited since they spoke only Portuguese. A request was sent to the Vatican for missionaries who were good at language learning so they could introduce their faith to the local inhabitants in their own tongues. Two linguistically talented Italian Jesuit priests were chosen for this work, Michele Ruggieri (1543–1607), who later went by the Chinese name Luó Míngjiān 羅明堅, and Matteo Ricci (1552–1610), called Lì Mǎdòu 利瑪竇 and styled Xī Tài 西泰 in Chinese. They arrived in Goa in 1578 after an arduous voyage. The next year, Ruggieri, who had already learned Tamil, was sent to Macau to study Mandarin; Ricci followed him in 1582 (Hsia 2010: 41–75). Both priests found Chinese very difficult to learn, citing its lack of inflection, the tones, the written characters, and the many dialects. Ricci described Chinese as ‘nothing like either Greek or German’, and pronounced it ‘the most ambiguous spoken and written language ever to be found’ (Fontana 2011: 35–7). But Ricci, relying on his systematic ‘Memory Palace’ method of memorization (Spence 1984: 1–4), succeeded in becoming literate in Chinese in just a year. Ruggieri and Ricci’s goal was to introduce their religion to the people within China proper, so in 1583, after a previous failed attempt, they moved to Zhàoqìng 肇慶 in Guangdong province. They remained there until their expulsion in 1588, translating Christian writings into Chinese, among other projects (Fontana 2011: 67–78). In order to help other foreigners desiring to learn Chinese, Ruggieri, with the collaboration of Ricci, compiled a 189-page Portuguese–Chinese word list in manuscript form, with the Chinese pronunciations written out phonetically in Latin letters. For centuries it was believed lost but was then rediscovered in 1934 by Pasquale D’Elia, SJ (Dé Lǐxián 德禮賢, 1890–1963) in the Jesuit archives in Rome (Yin 1994: 1–2). This is the earliest known effort at representing spoken Mandarin Chinese in Latin letters. The letters chosen for the Ruggieri–Ricci system were based mainly on Portuguese and Italian pronunciation and orthography. The earliest incarnations of this system were far from rigorous. Aspirated vs. non-aspirated initial stops were not distinguished. This certainly can be attributed to the lack of corresponding equivalents in the Romance languages, in which there is a clear distinction between voiced and voiceless stops, but no secondary marking of the voiceless stops with aspiration, as is the case in English. Nor was there any indication of tone at this point; tone marks, along with aspiration marks, were not added until years later. Furthermore, different symbols were often used to represent the same sound, mostly due to the spelling conventions of Portuguese and Italian, for example, ‘c’ was used to represent both /k/ and /kh/ before /a/, /o/ and /u/, but ‘ch’ was used before /e/ and /i/, while ‘qu’ was used for /kw/ and /khw/. So at this point the system could only be considered an impressionistic transcription. Following a failed attempt to gain permission to reside in Peking in the winter of 1598, Ricci, together with Sebastian Fernandes (Zhōng Míngrén 鍾鳴仁) and another Jesuit priest, Lazzaro Cattaneo (Guō Jūjìng 郭居靜, 1560–1640), had to take a boat to Línqīng 臨清, Shandong. The weather was bad, and it took them a whole month to reach their destination. In order to use their time well, they collaborated on a Chinese–Portuguese Dictionary while in transit. Below is the story of how tone and aspiration markings were added to their Romanization system: And so they spent one month [on the road] before they arrived in Lincin [Linch’ing]. In order to avoid the wasting of time during this journey, those who were older in 758

Wade–Giles Romanization system

the mission with the help of Brother Sebastian who knows very well the language of China, compiled a beautiful dictionary (  ferero un bello vocabulario) that according to rule and order dealt with all things concerning this language so that from now everyone could easily learn the language. Since this language is composed of monosyllabic words or characters, it will be very much necessary to listen and pronounce the accent [i.e. tone] and the aspiration whenever these occur in each of the words; with this kind of pronunciation, they distinguish and understand many characters and words; without this, they appear to be identical; this is what makes this language more difficult to learn. And, to distinguish well the words which are aspirated, they devised five types of different accent marks; in this matter, Fr. Cattaneo helped very much with the knowledge he has of music. He observed and distinguished [tones] very well. For this, they decided to use five marks of accent and one mark of aspiration. And they used these marks in writing the sound of a character along with our [Roman] letters. They wrote everything this way so that all were in uniformity. And, Fr. Matteo [Ricci] ordered that from then on all [the Jesuit missionaries] should observe these rules and did not allow any person to write as he pleased; otherwise, there would be great confusion. In this way, one can communicate with this dictionary. Others that they compile later will be very well understood by everybody; and, in the same way, one could serve another with his own writings and notes with much fruit and usefulness of this science among us. (FR II: 32–33) (cited in Ruggieri and Ricci 2001: 185) Though this volume was unfortunately lost, the Romanization system they settled on survives in Ricci’s 1606 collection of religious essays widely known as Xīzì Qíjī 西字奇跡 ‘The Miracle of Western Writing’ (Yin 1994: 5). The essays were typeset vertically, from right to left, as was standard for Chinese at the time, with Latinized spelling added to the right of each character. Aspiration of initial stops was indicated by a reversed apostrophe, borrowed from classical Greek, in which it was used to indicate the ‘rough breathing’ [h] sound before a vowel, diphthong, or rho (Porter et al. 2010: 5). The tonal system employed matches that of the Nanjing dialect, the prestige speech form of the time. The five tones were marked thus: (1) The yinping 陰平 was marked with a macron over the main vowel, e.g. t’iēn 天 ‘sky’, xīm 聲 ‘sound’; reconstructed value: 33 (mid level); (2) the yangping 陽平 with a circumflex, e.g. gîn 人 ‘person’, yên 言 ‘speech’; value: 21 (low falling); (3) the shǎng 上 with a grave accent: yù 雨 ‘rain’, xùi 水 ‘water’; value 42 (mid falling); (4) the qù 去 with an acute accent: ván 萬 ‘ten thousand’, súi 歲 ‘year’, value: 35 (mid rising); and (5) the rù 入, the entering tone, with a breve: pě 百 ‘a hundred’, nhiě 業 ‘enterprise’, value 45 (high rising); (Ruggieri and Ricci 2001: 53). These additional markings rendered the system a much more complete, accurate and usable one. One interesting quirk of the system: the velar nasal final /ŋ/ is represented as ‘-m’, a choice that is quite confusing to a modern reader. In fact, however, there was by this time apparently no bilabial nasal final in Nanjing Mandarin, so the symbol is unambiguously /ŋ/. This can be fairly easily accounted for – neither Portuguese nor Italian has a phonemic velar nasal, so there was no immediately obvious symbol for the Chinese /ŋ/. ‘m-’ in initial position still represents [m-]. The Flemish Jesuit priest Nicolas Trigault (Jīn Nígé 金尼閣; 1577–1628) further tweaked and refined the Ricci–Ruggieri–Cattaneo system in his 1626 Chinese lexicon for Western learners of Chinese, Xīrú Ěrmù Zī (西儒耳目資 ‘Aid to the Eyes and Ears of Western Literati’), mainly by reducing the number of symbols used. The entries were arranged 759

Karen Steffen Chung

by rhyme, with Romanization provided for each. In addition to simplifying the spellings, it is notable that Trigault carefully maintained the aspiration and tone markings for each character. This same system was also adopted by later scholars and missionaries, including Martino Martini (1614–61), Francisco Varo (1627–87), Prospero Intorcetta (1626–96), and Joseph Henri-Marie de Prémare (1666–1736), (Yang 1989: 221; Coblin 2006: 26; Klöter 2011: 103, cited in Simmons: forthcoming). The tone markings eventually adopted for use with the Zhùyīn zìmǔ 注音字母 phonetic alphabet in the early twentieth century, and later the Yale and Pinyin Romanization systems, represent a continuation of this approach.

The Protestant Missionary Period The nineteenth century was a period of intensive Protestant missionary activity in China, which provided an impetus for the production of several substantial English-language Chinese dictionaries and grammars. Some of the most notable figures in this effort were Joshua Marshman (Mǎ Shìmàn 馬士曼; English; 1768–1837), who very curiously learned his Chinese and published a book on the phonology and grammar of Chinese while living in India; Robert Morrison; Walter Henry Medhurst (Mài Dūsī 麥都思; English; 1796–1857), who published his English and Chinese Dictionary in 1848; and Samuel Wells Williams (Branner 1997: 235–6). All relied on native Chinese sources; each based their dictionaries on an existing Chinese rhyme book or lexicon, adding Romanization, English glosses or definitions, plus their own additional material. James Legge did not compile a dictionary, but was highly prolific in his translations of the Chinese Classics.

Robert Morrison Up through the 1840s, missionary work was both difficult and dangerous, since it was forbidden to preach Christianity in China, and Chinese were also forbidden to teach foreigners the Chinese language (Branner 1997: 235). Punishment for violations was harsh. Knowing he would not be allowed to openly preach in China, Robert Morrison (Mǎ Lǐxùn 馬禮遜; Scottish; 1782–1834) set other more attainable goals for himself: producing a new, colloquial translation of the Bible, a Chinese grammar, and a Chinese dictionary to help others learn the language. He based his three-volume A Dictionary of the Chinese Language, which took him 16 years to complete, on Chén Jìnmó’s 陳藎謨 Chinese rhyme book Wǔchē Yùnfǔ 五車韻府 (Xu n.d.: 3), and referred frequently to the Kāngxī Dictionary 康熙字典. Morrison had many setbacks to deal with, including the theft of the italic type intended for use in printing the dictionary, from the ship bringing it from England. The work was originally published in 1815; slightly revised editions were reissued in 1819, 1865, 1879, and 1907. All in this group of writers, including Morrison, drew heavily from the work of the earlier Catholic missionaries, but as native speakers of English, they Anglicized it con­siderably. Table 42.1, a list of some of the symbols Morrison chose for his own scheme, provides a snapshot of one stage in the process of developing a rigorous, practical system of Romanization for Chinese. Though Morrison’s system was not such yet, he certainly made an effort at consistency and to show all the phonetic distinctions of Chinese with unique spellings. In his preface he writes, ‘Without assuming that the orthography adopted is the best possible, it is affirmed that to enable a person to judge, it is requisite that he first spell all the Chinese words; for to judge of single words only will mislead and subject him to the absurdity of giving the same spelling for different sounds’ (Morrison 1865: vi). Some of the examples 760

Wade–Giles Romanization system Table 42.1  Some of Morrison’s spellings and descriptions of the phonetic values Morrison

IPA

A, as in hard Ă, as in hat AE, broad A coalescing with E, forming a sound like igh in high AOU, broad A and OU coalescing AY, as in may E, final, as in me Ĕ, as in met EU, as the sound of EU in the French word peu [Note: peu is certainly not an accurate example; it is everywhere used here for the rounded high front vowel [y]] EW, as in new ĬH, a sound similar to that given by the Letter I, when in pronouncing the word with, the Reader stops short at the I G, is hard in Gĭh J, as in French Ŏ, as in hot OW, as in how [Note: syllables with this spelling are pronounced [ou] in modern Mandarin, e.g. kh’ow 口 kǒu, ch’ow 酬 chóu; written ‘ɛu’ by Ricci U, nearly like EU, as above [Note: final [u] is spelled as ‘oo’] UH, as in hut ZE, a buzzing sound, which cannot be expressed by the Roman Alphabet [Note: This is used in the zi, ci, si apical-dental series]

[ɑ] final: [ɑʔ], medial: [ə] [ai] [au] [ei] [i] final: [ɛʔ], medial: [ɛ] [y] [iu] [ɪʔ] [g] [ʐ ] [ɔʔ] [ou] medial: [u], [ɔ] final: [əʔ] Used in: [tsɨ], [tsʰɨ], [sɨ]

(Morrison 1815 Ib: xvii).

given to approximate the sounds are less than clear, and often quite misleading, as with the French peu. IPA symbols are added to indicate the probable pronunciations he meant to represent, based on an examination of the entries using each respective spelling. He also uses an umlauted ‘e’ (ë) for an [i] onglide. His choice of ‘how’ (in Scottish English, presumably [hʌu]) as an English approximation to represent the [ou] vowel in 口 kǒu is unexpected. Also, his spelling of 口 kǒu is ‘kh’ow’, which has both an ‘h’ and an apostrophe to indicate aspiration; this practice was not followed with the other initial stops, e.g. 多 duō is spelled ‘to’, with 拖 tuō being spelled ‘t’o’ and not ‘th’o’. With ‘t’ and ‘p’, however, there is the risk that readers would interpret and pronounce them as the English digraphs ‘th’ [θ] and ‘ph’ [f ]; but then why use the ‘h’ with the ‘k’, or at all? This is apparently one example of the system’s lack of rigor. But the author reminds us in the ‘Advertisement’ of the 1815 edition not to ‘find fault with the errors of the Work, merely for the sake of publishing a piece of smart Criticism  .  .  .  The Writer is very far from standing forward with proud pretensions to excellence in his plodding task’. His task was in fact not an easy one. Morrison mentions that ‘the Europeans say that there are five tones, and generally speak of them by “first, second,” &c. according to the order in which they stand above’, but then goes on to give examples of Romanized Chinese syllables with tones indicated by diacritics (Morrison 1815: 20). So designating tones by number was still mainly an informal oral practice, following the native Chinese order that Ricci adopted. 761

Karen Steffen Chung

In part 1, vol. 1 of the 1815 edition, Morrison introduces a system for marking the tones: (1) ‘Ping-shing’ 平聲, is left unmarked ( yín píng and yáng píng are not distinguished here but are described elsewhere); (2) ‘Shang-shing’ 上聲, is denoted by the grave accent (ì); (3) ‘Keu-shing’ 去聲, by the acute accent (í); (4) ‘Juh-shing’ 入聲, by the short accent (ĭ); and the aspirate, by (’h.) (Morrison 1815: xvii). In part 2, vol. 1 of an 1819 edition, under ‘Rules for Using the Dictionary’ (p. xiii), he uses a macron for the first tone. The tones are marked in the 1815 edition, but not in the 1865 one, except for the entering tone. One should perhaps allow that his system was still developing and in flux, but it is quite confusing for the reader. Morrison confirms that the form of Mandarin represented in his dictionary was, as was also generally the case thus far starting from Ricci, ‘rather what the Chinese call the Nanking Dialect, than the Peking’. His descriptions of the differences between the Beijing and Nanjing dialects offer elucidation as to which pronunciations were current in the two dialects during this time, adding that the ‘changes are tolerably regular and uniform, so that it is not difficult in speaking to adopt either the one Pronunciation or the other’ (Morrison 1815: xviii). We learn from this that: 1. The ‘k’ initial in syllables like ‘king’ for 京 jīng and ‘keang’ for 江 jiāng was indeed only a Nanjing pronunciation; in Beijing these would be ‘ching’, and ‘cheang’ or ‘tseang’ in this system, confirming that in Beijing these sounds were already the affricated alveolo-palatal [tɕi], or apical dental [tsi]. 2. Initial [hi] and [he] were in Beijing Mandarin [ɕi] or [si]. 3. ‘chang’ and ‘tsang’, ‘cho’ and ‘tso’, ‘man’ and ‘mwan’, ‘pan’ and ‘pwan’, ‘we’ [wi] and ‘wei’ are often used for each other or ‘confounded’ in some informants. The loss of the labialized onglide of vowels following bilabial initials was subsequently completed with unrounded vowels such as /ɑ/, but it was retained with the rounded /ɔ/, e.g. 波 bō [pʊɔ]. 4. The final glottal stop of the old entering tone syllables had by this time disappeared from Beijing Mandarin, so ‘mŭh’ becomes ‘moo’, ‘pih-king’ becomes ‘pei-ching’. Morrison adds an interesting personal comment on the aesthetics and ease of articulation of these respective sounds: ‘The soft and lengthened sounds are more pleasing to the car; and to a person accustomed to speak English, require less effort than the Short Tones.’ 5. The distribution of initial ‘f’ and ‘p’ had not yet fully stabilized; [f ] derived historically from /p/, and the sound change seems to have been incomplete – and this is true even today; also, ‘nwan’ and ‘lwan’, ‘sh’ and ‘ch’, ‘ts’ and ‘ch’ arc occasionally used for each other. These rules go a long way toward explaining some of the persistent variation observed in Romanization systems of Mandarin, also the origin of ‘Peking’ and ‘Nanking’ type spellings, especially those used in the French-influenced ‘Postal’ Romanization system – French was the language of the international postal system. Aspiration marks were not included in first edition of the dictionary, but were added in the 1865 reprint. The author notes: ‘The reprint was commenced on the principle of the original, without a distinctive representation of the aspirates, but from an early period their importance was recognized, and they were accordingly introduced’ (p. ix), more or less retracing the steps of Ruggieri and Ricci. Morrison says further in his 1815 Grammar: ‘Another variety in the Chinese syllables is marked by an aspirate placed with the other marks  .  .  .  The aspirate the Chinese do not seem to consider a modification of the same 762

Wade–Giles Romanization system

syllable, but a quite different initial sound.’ And regarding the tones: ‘The pronunciation of the Tones can only be learned from a living instructor. They are not absolutely necessary to be understood in speaking Chinese; but are yet essential to good speaking. Hence an early attention to them is advisable’ (Morrison 1815: 21). Westerners seemed to find the aspirated– unaspirated distinction a major stumbling block to learning Chinese well, almost as difficult as the tones. The view starting with Ricci that aspiration is ‘something added’ was perpetuated in the Wade–Giles system, and in fact was probably its biggest fatal flaw.

Samuel Wells Williams and James Legge The problem of designing and popularizing a transparent, consistent system of Romanization was a major concern of Samuel Wells Williams (Wèi Sānwèi 衛三畏; American; 1812–84), a contemporary of Thomas Wade. Williams was born in Utica, New York, and began his career in China in 1833 as a printer for the Canton Mission Press, but later became secretary-interpreter for the US legation to China, in 1856. From 1860 to 1862 he was in the United States but returned in 1862 to the US legation in Peking, where he remained until 1876. During this period, he compiled his A Syllabic Dictionary of the Chinese Language, which was published in 1874. The rhyme book Williams based this work on was the Wǔfāng yuányīn 五方元音. As described in his preface, he consciously chose a ‘general’ Mandarin pronunciation style not tied to any one specific locality. This brought his Romanization system a giant step closer to the Beijing-based Wade–Giles system soon to follow. It still had some pre-modern features, such as ‘h’ for some syllable types (‘heung’ 兄 xiōng, ‘hew’ 休 xiū; probably pronounced with an initial [x-] at the time) and ‘s’ for others (‘sü’ for 須 xū and 徐 xú) that start with a [ɕ] initial in contemporary standard Chinese. ‘Chia’-type spellings superseded the previous ‘kia’-type spellings for syllables like 加 jiā. He marked the tone of each character using the semicircle method. In addition, Williams hoped that his Romanization system could be easily convertible into China’s many local dialects. To this end, his dictionary features pronunciations for the speech of (in his orthography) Canton, Swatow, Amoy, Fuchau, Shanghai and Chifu, in addition to Mandarin. Williams sums up his view of the problem thus: ‘If the difficulties of illustrating and analyzing the sounds in their language are almost insurmountable to Chinese philologists, the results of various attempts of foreigners to do so have not the less proved the inherent difficulties of the attempt; and a comparison of their various systems does not encourage the hope that anything like uniformity will ever be attained’ (Williams 1874: xviii). With this dictionary, Williams certainly tried his best to achieve this, though like with Morrison, his system was in constant flux. James Legge (Lǐ Yǎgè 理雅各; Scottish; 1815–97), another contemporary of Thomas Wade, viewed himself primarily as a missionary, but at the same time also devoted himself whole-heartedly to the translation of the Chinese Classics during his more than 25 years in the Far East, from 1839 to 1867. He continued his translation work full time for 20 more years after assuming the new Chair of Chinese Language and Literature at Oxford in 1876. His goal in this work was to help the rest of the world to ‘really know this great Empire’ and also that ‘our missionary labours among the people should be conducted with sufficient intelligence and so as to secure permanent results’ (Ride 1991: 1). In his first edition of the Chinese Classics, Legge adopted Morrison’s Romanization system, without tone marks, though he did use the circle method directly by the Chinese character to indicate the tone of characters with an alternate reading. For his Sacred Books 763

Karen Steffen Chung

of China series (1879–91), however, he began using Thomas Wade’s system. Legge’s work thus reflects the historical transition from Morrison’s systems to the next step in the development of a more rigorous Romanization system that received wide public acceptance, Wade–Giles.

Thomas Francis Wade We now move into the period of Western – mainly English – learners and interpreters of Chinese language and culture who found themselves in China for secular reasons, government service in particular. Sir Thomas Francis Wade (Wēi Tuǒmǎ 威妥瑪, occasionally also Wěi Dé 偉德; English; 1818–95) had a multicultural upbringing. He was born in London, and educated in Mauritius, in Cape Colony, South Africa, at Harrow School in London, and at Trinity College, Cambridge. He was known for his excellent memory and love of languages. Wade’s father cut his university education short in 1838 to enlist him in the military. Wade served in England, Ireland and Greece, where he learned Greek and Italian, before being sent to Hong Kong in June 1842. Wade plunged into his study of Chinese during the long journey to his new post. His knowledge of Chinese, something quite rare among Westerners at the time, led to his being appointed to various positions as interpreter, including as interpreter of Cantonese to the Supreme Court of Hong Kong. He was later appointed assistant Chinese secretary to superintendent of trade Sir John Davis, then as vice-consul at Shanghai, during which time he concurrently helped establish the foreign maritime customs. He served in further appointments in Hong Kong and China, and assisted in the negotiations which led up to the Treaty of Tientsin of 1858. Wade was knighted in 1875. After over 40 years in the British foreign service in China, he returned to England in 1883. Three years later he donated 4,304 Chinese books, mainly literature, to the Cambridge University Library’s Oriental Collection. He served as president of the Royal Asiatic Society from 1887 to 1890. In 1888, he was elected the first Professor of Chinese at the University of Cambridge, a position he held until his death at 77 (Douglas 1899: 420). Wade produced a number of pedagogical works on the Chinese language for foreigners. Of these, two had a lasting impact. The first was his Peking Syllabary, subtitled: being a collection of the characters representing the dialect of Peking; arranged after a new orthography in syllabic classes, according to the four tones [emphasis added], designed to accompany the Hsin Ching Lu (Xúnjīnlù 尋津錄), or, Book of Experiments, Being the First of a Series of Contributions to the Study of Chinese, published in Hong Kong in 1859. Wade’s intention in this work was to provide a Chinese reader useful to ‘student interpreters in service of the British government’. It consists of a collection of specially written Chinese phrases and sentences, plus Emperor Kangxi’s Sacred Edict, first in English, with notes, and tone and pronunciation exercises; then in the original Chinese, typeset vertically, with the pronunciation of each character given in Romanization, reminiscent of Ricci’s essays. An alphabetically arranged character index is appended at the end. Wade tried in earnest to produce a practical work, useful in learning everyday conversation. But with the 1919 May Fourth vernacular literature movement still a ways into the future, there was little precedent for such. So the content and language of the work are in fact quite stilted and far from colloquial, nor is the subject matter particularly engaging. But it comprised a rare bilingual text potentially useful to anyone of the time wishing to advance their knowledge of Chinese. The influence of the Syllabary, however, turned out to be considerably more far-reaching than that of the texts themselves. 764

Wade–Giles Romanization system

The other highly influential and popularly successful work was Wade’s Yü-yen Tzu-erh Chi: A progressive course designed to assist the student of colloquial Chinese, London, 1867, with a number of subsequent editions. The Chinese title, Yǔyán Zì’ěrjí 語言自邇集 is based on a quote from chapter 15 of The Doctrine of the Mean (Zhōng Yōng 中庸): Pì rú xíng yuǎn bì zì ěr; pì rú dēng gāo bì zì bēi 辟如行遠必自邇,辟如登高必自卑 (J. Legge 1991: 396): ‘To go to a distant place, you must begin by treading the ground nearby; to ascend a hike peak, you must begin from a lower level.’ Starting from the release of the first edition, it was certainly the most often-used text by beginners of Chinese at the time; it was also adapted for use in Japan (Sinclair 2003: 147–74). Useful as it was, especially in the near-absence of viable alternatives, like the Hsin Ching Lu, it contained outdated, naïvely quaint, and impractical material in parts, some of which was edited out in later editions (in the 1903 edition, the sections entitled ‘The Hundred Lessons’ and the notorious ‘Graduate’s Wooing’ were taken out); it was also criticized for presenting too much material too quickly, as noted by the author himself in his Preface to the 2nd edition. It did, however, help uncounted people in the daunting task of learning the Chinese language. These works mark two significant developments in the history of Romanization in China. First was the shift from the Nanjing to the Beijing dialect as the standard for Mandarin Chinese. Wade puts it thus in his introduction to the Hsin Ching Lu: ‘The Dialect of Peking is to China what the Parisian of the salons is to France. It is forty years since Dr. Morrison predicted that it would corrupt the general language of the Empire, and we make bold to say that this prediction has been to a great extent fulfilled. The officials born at a distance from Peking strive generally to catch the Peking accent; it is the fashion to acquire it.’ (Wade 1859: Introduction) Secondly, it set down the Wade system of Romanization in preliminary form, and helped establish it as the clear winner among all the proposed orthographies up to that time. The Romanization of Mandarin had been in flux for centuries, due to the different language backgrounds of the writers involved in the transcription process, and also on account of some of the particular features of Mandarin, such as aspiration and the tones, that tended to confound Westerners. Wade did not find it easy to satisfactorily sort everything out either. ‘The best orthography, doubtless, would be one which conformed exactly to the alphabetic prejudices of the person who had to use it; but the anomalies of English pronunciation make it very difficult indeed to avoid shocking these’, Wade observed (Wade 1859: 82). In Wade’s system, aspirated initial stops and affricates were consistently marked with a reversed apostrophe, similar to Williams, who used the symbol ʿ. The entering tone was now out of the picture, though its loss resulted in some irregular and unstable vowel finals, true even to the present day. And, following Thomas Taylor Meadows (Meadows 1847: 59–66), the four tones were indicated with superscript numbers, whereas Williams used semicircles. Some other notable features of Wade’s system:   1. Wade employs two diacritical marks, in addition to the apostrophe. a. Like Williams, he uses an umlaut over the ‘u’: ‘ü’, certainly borrowed from German, to represent the rounded high front vowel /y/; previously it had been written as ‘yu’, ‘iu’ and ‘iuu’, by Ricci, and as ‘eu’, and sometimes ‘ü’, by Morrison. 765

Karen Steffen Chung



b. He uses a circumflex over ‘e’ for [ə] as in 很 hěn [hən], writing it as ‘hên’; compare to ‘hân’ in Morrison and Williams.   2. He uses ‘o’ for an open-syllable [ɤ] vowel, e.g. 可 kě is k’o, apparently because that is mainly how it was pronounced at the time, although it probably belongs to the same phoneme as ‘ê’.   3. He has an unexplained aspiration mark before some but not all syllables with a /h-/ initial, e.g. ‘’hsia’ for 蝦 xiā, but plain ‘hsiang’ for 鄉 xiāng.   4. Morrison’s ‘hi-’ and ‘si-’ are now ‘hs-’, e.g. ‘hsiung’ 兄 xiōng, ‘hsün’ 訓 xùn; compare to Williams’ ‘hiung’ for 兄 xiōng, ‘siün’ for 訓 xùn.   5. Initial ‘ng-’ is gone, e.g. in ‘ai’ 愛 ài and ‘wo’ 我 wǒ: compare to Williams’ ‘ngai’ and ‘ngo’.   6. He uses ‘hu-’, rather than ‘hw-’ as in Williams.   7. There is no jiāntuán 尖團 distinction, i.e. between alveolo-palatal ‘ch-’/‘ch-’ ’ 經/輕 and dental ‘ts-/ts-’ ’ 精/青; both sets are ‘ching’/‘ch’ing’ jīng/qīng.   8. Both ‘yi’ and ‘i’ are used.   9. A final ‘h’ is used in ‘yeh’ yè 葉; compare to ‘yé’ in Williams, with ‘yeh’ for entering tone syllables. 10. There are alternate forms for some of the syllables, mainly the aftermath of lost entering tones, e.g. ‘yo’ (or ‘yao’) for 約, now pronounced ‘yüeh’/yuē; and for 藥 yào, also listed under ‘yao’; and 若 ruò, also under ‘jo’. 11. He reflects the [ə] offglide in /-un/ finals, giving both e.g. ‘kun’ and ‘kuen’ for 滾 gǔn [kʊən]. 12. On the other hand, no onglide is given in e.g. ‘to’/’t’o’ [tuɔ]/[tʰuɔ] 多/拖 duō/tuō, ‘tso/ts’o’ [tsuɔ]/[tsʰuɔ] 做/錯 zuò/cuò, as in Williams. This is perhaps defensible in the case of ‘po’, ‘p’o’, ‘mo’, and ‘fo’ because lip rounding was considered to be inherent in labial initials, but it is not immediately clear why it was also done for the dental/ alveolar series of initials. Perhaps because it was because omitting it does not cause any ambiguity. ‘u’ is added after the velar initials: ‘kuo’, ‘k’uo’, ‘huo’, where omitting it would cause confusion with ‘ko’, ‘k’o’, ‘ho’. 13. He simplifies triphthongs, as is now done in Pinyin, though inconsistently, e.g. the ‘e’ is included in ‘kuei’ 歸 guī and ‘k’uei’ 虧 kuī, but not in ‘shui’ 水 shuǐ. 14. He uses ‘urh’ instead of ‘êrh’ for 而 ér; this was later modified. Wade’s system was quite close in many ways to Williams’ scheme, making allowances for their different dialect bases of Nanjing vs. Beijing, which dictated the inclusion vs. omission of the entering tone, and ‘k-/k’-’ vs. ‘ch-/ch’-’ initials. Since the two were contemporaries, it is hard to know from which direction the influence primarily flowed, or if it was bidirectional.

Herbert Allen Giles Herbert Allen Giles (Zhái Lǐsī 翟理斯; English; 1845–1935) began his career in the British foreign service in Peking, after having passed the competitive examination for a student interpretership, the usual starting point for junior trainees. Giles served as a British consular official in various parts of China (1867–92). His first post was to Taiwan in 1867, and he also served as British Consul at Tamsui (Danshui) (1885–87). He served as interpreter at Tientsin, Ningpo, Hankow and Canton, then also in various positions in Swatow, Amoy, Pagoda Island, Shanghai, and Ningpo. Giles was apt to express views that did not accord 766

Wade–Giles Romanization system

with official policy or public opinion at home. It probably for this reason he was often transferred to different posts, and did not rise high in the foreign service. This seemed to suit him reasonably well, because by ending up in positions with fewer responsibilities, he won more time to engage in his own ‘real’ work, the translation of major Chinese language works, and his own writing. Giles resigned his post on health grounds in 1893 after 25 years in the service. He had by then made a name for himself in Sinology, and in 1897, despite his lack of formal qualifications, he succeeded Thomas Wade as Chair of Chinese at Cambridge. He was the only Sinologist at Cambridge and had few students, so he was able to devote himself almost full time to reading the Chinese books donated by Wade, of which he became Honorary Keeper, and translating and publishing what he gleaned from his wide reading. He retired in 1932, and passed away in 1935 at the age of 90. Giles was curmudgeonly, irascible, and did not ‘suffer fools gladly’, even judging solely from his own memoirs. He was quick to denigrate others, especially their writings, often not content to merely point out errors, but adding a few choice epithets in a personal attack as well, which resulted in more than a few irreparable ruptures with others. Giles was particularly disparaging of Thomas Wade, most of all for his Tzu-erh Chi, with which Giles began his own study of Chinese. This may be surprising in view of how closely Giles’ and Wade’s names later came to be associated with each other – some have even assumed that Wade–Giles was the hyphenated name of a single individual (Language Hat 2006) – but is perhaps better viewed as simply a relatively extreme example of ‘literati looking down on one another’ wénrén xiāngqīng 文人相輕 in order to shore up their own image and position. At the end of his life, Giles was said to be on speaking terms with only one of his surviving children. His memoirs, however, show him to have been a dedicated family man, and he gave his second wife Elise frequent and profuse thanks for all her careful, painstaking proofreading of his writings (Aylmer 1997: 1–6). Giles was a highly prolific translator into English of significant Chinese literary works, such as the Confucian Analects, Zhuang Zi, The Three Character Classic, The Hsi yüan lu, or, Instructions to coroners, and Pu Songling’s Strange Stories from a Chinese Studio. These, combined with his own writings on the people, language and culture of China, did much to give English readers a firsthand acquaintance with some of the wealth of Chinese culture. Most influential in further establishing the Romanization scheme first set down by Wade was Giles’ 1,415-page A Chinese–English Dictionary, which became a standard reference work soon after its release in 1912. The orthography it employed came to be known as the Wade–Giles system of Romanization, and it was soon adopted by English-language academia, and then by the media and general public. In fact Giles’ Romanization was only very slightly modified from Wade’s – the differences are miniscule. Tones continued to be marked in the Wade–Giles system with numeral superscripts, with the neutral tone either being unmarked, or occasionally given the number ‘0’ or ‘5’. Giles probably had a greater role in popularizing the system, due to his high output as a scholar, translator, and writer, as compiler of his Chinese–English Dictionary in particular, while Wade only wrote a small number of – albeit quite widely used – Chinese language teaching texts. In addition to his more serious and scholarly works, Giles also wrote a ‘Teach Yourself ’ manual, entitled Chinese Without a Teacher (1872), for foreign residents in China needing minimum proficiency in Mandarin Chinese for everyday use. It included no written Chinese characters, skipped over any teaching of the tones, but it does use the apostrophe to dis­ tinguish aspirated from unaspirated initial stops and affricates. Its fanciful, ad hoc spellings 767

Karen Steffen Chung

reflect Giles’ awareness that learning Wade–Giles Romanization was not particularly straightforward or intuitive for the average English speaker: I was naturally a good deal laughed at in a friendly way and exception was given to the absence of Tones. This lack was vigorously defended by a Chinese-speaking captain in the mercantile marine, who quoted Sir Harry Parkes’ dictum, ‘never trouble yourself about the Tones,’ – a most erroneous view, with which I have never been in sympathy. My little book, however, was only a jeu d’esprit, in which Tones would have been wholly out of place. My object was to transliterate Chinese strictly according to the values of the English vowels and consonants, so that anyone could pick up the book and read off a simple sentence with a good chance of being understood. Thus, instead of the necessarily arbitrary system for students, which provides ‘ni kei wo mai’ 你給我買 = you buy it for me, I gave ‘nee kay waw mi,’ which no one who knows the English alphabet would have to learn to pronounce. (Aylmer 1997: 13) The book was quite popular at the time and went into many editions. It is a further illustration of the difficulty of combining native language-based intuitiveness and rigor in the same system. Wade–Giles was the undisputed standard for Romanization of Chinese in English-language writing until the 1970s, when the People’s Republic of China began opening up to the rest of the world. Gradually world news media began replacing Wade–Giles spellings with Pinyin, and academia soon fell into step as well. Wade–Giles is still seen in older publications, in a small number of established Chinese loanwords in English like Shih-tzu and Tai-chi – though probably more are in Romanized Cantonese than Mandarin – and in some Chinese place names and personal names, now mostly restricted to Taiwan.

User Feedback on The Wade–Giles Romanization System If there is one aspect of Wade–Giles Romanization that is apt to lead a litany of complaints about the system, it is certainly the use of the apostrophe (’) to mark the aspirated voiceless initial stops and affricates, together with the use of unmarked ‘p’, ‘t’, and ‘k’ for the unaspirated voiceless ones. English speakers typically express deep puzzlement over why Wade and Giles didn’t simply use plain ‘p’, ‘t’, and ‘k’ for the aspirated voiceless stops, and ‘b’, ‘d’ and ‘g’ for the unaspirated voiceless ones. Phonologically, English stops have a voiced-voiceless opposition: /b/ vs. /p/, /d/ vs. /t/, /g/ vs. /k/. Phonetically, however, aspiration of the voiceless stops is often the only way that the voiceless stops are distinguished from their voiced counterparts when they are in utterance-initial position, e.g. in ‘Do it’, the /d/ is seldom fully voiced. It is because of this allophonic rule in English that the Chinese initial stops /p/ and /pʰ/ and so on, sound pretty much just like initial /b/ and /p/ in English. In fact the Mandarin aspirated stops are more strongly aspirated than the English ones, but the difference is relatively slight. This rule makes it difficult for the average English speaker to get a clear understanding of the differences and relationships between voicing and aspiration at all, since they are so intimately intertwined in English. (For a detailed discussion of the ‘aspiration problem’ in Western descriptions and transcriptions of Mandarin, please see Branner 1997.) Most English-speaking learners of Chinese today would in any case be much happier with a ‘b’ vs. ‘p’-type representation. This was the approach adopted by Sinologist George 768

Wade–Giles Romanization system

Kennedy in 1943 in his Yale system of Romanization, probably the most English-user-friendly of any Romanization system ever devised for Mandarin; for example, Pinyin zi is written as dz in Yale, making it easy for an English speaker to get it about right. Tones are marked with diacritics, which were later adopted in Pinyin as well. One can only think it unfortunate that the Yale system never gained wider currency. There are inherent problems with the use of both diacritics and numerical superscripts. First, they are more difficult to typeset or input. It is slightly amazing that both survived as long as they did, considering the extra effort required to produce texts with the correct tone marks, particularly in a pre-computer era. English speakers are less accustomed to adding lots of diacritics to printed texts, and often omit them. But at least English has some words that may include diacritics, e.g. French loans such as naïve, façade, and déjà vu, so the markings still blend into a printed page fairly well. The same is not true of numerals. People are understandably not keen to have a name like Ch’en2 Chih4-hao2 appear on their passport – with the result that the numbers, and all the tonal information they carried, were simply dropped across the board in most situations where Wade–Giles was used. Many decades after the establishment of the Wade–Giles system, Yale and then Pinyin ended up adopting tone marks, basically a reversion to Ricci and Ruggieri’s approach, and these tend to be omitted as well. And the aspiration marks were also widely omitted. Such a style of Wade– Giles, denuded of two of the most fundamental phonological features of spoken Mandarin, comes up seriously short. There was perhaps a good chance to incorporate the tones in a relatively unobtrusive, aesthetic and rememberable way into a Chinese Romanization system with the adoption of the Gwoyeu Romatzyh system (GR) by the Nationalist government in 1928. It was designed by Y. R. Chao 趙元任 (Zhào Yuánrèn), though the idea of using tonal spelling apparently originated with Lín Yǔtáng 林語堂. It was officially referred to as Zhùyīn Dì’èrshì 注音 第二式 in Chinese. But this effort was a popular failure, for two big reasons: (i) the system was only an auxiliary system that existed in parallel to the Mandarin Phonetic Symbols 注音符號, the system universally adopted for teaching reading in schools, and GR was never widely taught or promoted – people can’t use something they haven’t learned, or don’t even know exists; (ii) The tonal spelling rules were so complex, with different rules applying to different syllable types, that they were deemed too difficult to be practicable. Too ambitious an agenda seems to have been taken on in designing the GR system – rather than merely Latinizing Mandarin, it strove at the same time to teach Mandarin syllable distribution patterns. For example, GR has separate rules for marking the first tone of syllables with an initial sonorant (m-, n-, l-, r-), since these are relatively rare in Mandarin, due to historical reasons. So the GR tonal spellings were later completely dropped in Taiwan, meeting with the same fate as the Wade–Giles numerical superscripts. This, along with the failure of the Yale system to catch on, were probably the biggest missed opportunities in the history of Chinese Romanization efforts. The Pinyin system is now nearly universal in the world, mostly for political reasons. It satisfactorily solved some problems, such as how to best represent for English speakers the aspirated vs. unaspirated voiceless stops; left some issues as they were, e.g. easy-to-ignore diacritics to mark the tones; and it introduced some new problems, e.g. how to pronounce the sounds represented by the ‘leftover letters’ c, q, x, and z (‘v’ is the only letter of the Latin alphabet left unused, so it is employed in some input systems to call up the character + diacritic ‘ü’). Some good choices still ended up causing confusion, such as the now nearly universal pronunciation of the Pinyin ‘j’ as in ‘Beijing’ as [ʒ] instead of [dʒ] as in jingle, which would be a very close approximation of the correct [tɕ]. Yet somehow the media 769

Karen Steffen Chung

usually manage to pronounce the much less transparent ‘x’ more or less correctly as ‘sh’, e.g. as in Xí Jìnpíng 習近平. Each Romanization system has its own historical context, strengths, and logic, and is viable as long as it is comprehensive and consistent, and both the transcriber and reader are adequately trained in it. Each system also has its irremediable weirdnesses and inconveniences, which users must, with a bit of indulgence, simply take in their stride. If the same message that was encoded in the system emerges intact when decoded by someone familiar with the system, without overly much effort, it has done its job.

Appendix WG

Morrison

MPS

Pinyin

WG

Morrison

MPS

Pinyin

a ai an ang ao cha ch’a chai ch’ai chan ch’an chang ch’ang chao ch’ao chê ch’ê chei chên ch’ên chêng ch’êng chi ch’i chia ch’ia chiang ch’iang chiao ch’iao chieh ch’ieh chien ch’ien chih ch’ih chin

a ngae an ang aou cha ch’a chae ch’ae chan ch’an chang ch’ang chaou ch’aou chay ch’ay chay chin ch’in ching ch’ing ke kh’e këa kh’ëa këang kh’ëang keaou kh’eaou keae k’eae këen kh’ëen che ch’e kin

ㄚ ㄞ ㄢ ㄤ ㄠ ㄓㄚ ㄔㄚ ㄓㄞ ㄔㄞ ㄓㄢ ㄔㄢ ㄓㄤ ㄔㄤ ㄓㄠ ㄔㄠ ㄓㄜ ㄔㄜ ㄓㄟ ㄓㄣ ㄔㄣ ㄓㄥ ㄔㄥ ㄐㄧ ㄑㄧ ㄐㄧㄚ ㄑㄧㄚ ㄐㄧㄤ ㄑㄧㄤ ㄐㄧㄠ ㄑㄧㄠ ㄐㄧㄝ ㄑㄧㄝ ㄐㄧㄢ ㄑㄧㄢ ㄓ ㄔ ㄐㄧㄣ

a ai an ang ao zha cha zhai chai zhan chan zhang chang zhao chao zhe che zhei zhen chen zheng cheng ji qi jia qia jiang qiang jiao qiao jie qie jian qian zhi chi jin

ch’in ching ch’ing chiu ch’iu chiung ch’iung cho ch’o chou ch’ou chu ch’u chua ch’ua chuai ch’uai chuan ch’uan chuang ch’uang chui ch’ui chun ch’un chung ch’ung chü ch’ü chüan ch’üan chüeh ch’üeh chün ch’ün ê ei

kh’in king kh’ing kew kh’ew keung kh’eung chŏ ch’ŏ chow ch’ow choo ch’oo chwa ch’wa chŭh ch’ŭh chuen ch’uen chwang ch’wang chuy ch’uy chun ch’un chung ch’ung keu kh’eu keuen kh’euen këŏ kh’ëŏ keun kh’eun ngeh

ㄑㄧㄣ ㄐㄧㄥ ㄑㄧㄥ ㄐㄧㄡ ㄑㄧㄡ ㄐㄩㄥ ㄑㄩㄥ ㄓㄨㄛ ㄔㄨㄛ ㄓㄡ ㄔㄡ ㄓㄨ ㄔㄨ ㄓㄨㄚ ㄔㄨㄚ ㄓㄨㄞ ㄔㄨㄞ ㄓㄨㄢ ㄔㄨㄢ ㄓㄨㄤ ㄔㄨㄤ ㄓㄨㄟ ㄔㄨㄟ ㄓㄨㄣ ㄔㄨㄣ ㄓㄨㄥ ㄔㄨㄥ ㄐㄩ ㄑㄩ ㄐㄩㄢ ㄑㄩㄢ ㄐㄩㄝ ㄑㄩㄝ ㄐㄩㄣ ㄑㄩㄣ ㄜ ㄟ

qin jing qing jiu qiu jiong qiong zhuo chuo zhou chou zhu chu zhua chua zhuai chuai zhuan chuan zhuang chuang zhui chui zhun chun zhong chong ju qu juan quan jue que jun qun e ei

770

Wade–Giles Romanization system

WG

Morrison

MPS

Pinyin

WG

Morrison

MPS

Pinyin

ên êrh, ‘rh fa fan fang fei fên fêng fo fou fu ha hai han hang hao hei hên hêng ho hou hsi hsia hsiang hsiao hsieh hsien hsin hsing hsiu hsiung hsü hsüan hsüeh hsün hu hua huai huan huang hui hun hung huo i, yi jan jang jao jê jên

ngăn urh fă fan fang fei fun fung fŭh fow foo

ㄣ ㄦ ㄈㄚ ㄈㄢ ㄈㄤ ㄈㄟ ㄈㄣ ㄈㄥ ㄈㄛ ㄈㄡ ㄈㄨ ㄏㄚ ㄏㄞ ㄏㄢ ㄏㄤ ㄏㄠ ㄏㄟ ㄏㄣ ㄏㄥ ㄏㄜ ㄏㄡ ㄒㄧ ㄒㄧㄚ ㄒㄧㄤ ㄒㄧㄠ ㄒㄧㄝ ㄒㄧㄢ ㄒㄧㄣ ㄒㄧㄥ ㄒㄧㄡ ㄒㄩㄥ ㄒㄩ ㄒㄩㄢ ㄒㄩㄝ ㄒㄩㄣ ㄏㄨ ㄏㄨㄚ ㄏㄨㄞ ㄏㄨㄢ ㄏㄨㄤ ㄏㄨㄟ ㄏㄨㄣ ㄏㄨㄥ ㄏㄨㄛ ㄧ ㄖㄢ ㄖㄤ ㄖㄠ ㄖㄜ ㄖㄣ

en er fa fan fang fei fen feng fo fou fu ha hai han hang hao hei hen heng he hou xi xia xiang xiao xie xian xin xing xiu xiong xu xuan xue xun hu hua huai huan huang hui hun hong huo yi ran rang rao re ren

jêng jih jo jou ju juan jui jun jung ka k’a kai k’ai kan k’an kang k’ang kao k’ao kei kên k’ên kêng k’êng ko k’o kou k’ou ku k’u kua k’ua kuai k’uai kuan k’uan kuang k’uang kuei k’uei kun k’un kung k’ung kuo k’uo la lai lan lang

jing jĭh jŏ jow joo juen juy jun jung

ㄖㄥ ㄖ ㄖㄨㄛ ㄖㄡ ㄖㄨ ㄖㄨㄢ ㄖㄨㄟ ㄖㄨㄣ ㄖㄨㄥ ㄍㄚ ㄎㄚ ㄍㄞ ㄎㄞ ㄍㄢ ㄎㄢ ㄍㄤ ㄎㄤ ㄍㄠ ㄎㄠ ㄍㄟ ㄍㄣ ㄎㄣ ㄍㄥ ㄎㄥ ㄍㄜ ㄎㄜ ㄍㄡ ㄎㄡ ㄍㄨ ㄎㄨ ㄍㄨㄚ ㄎㄨㄚ ㄍㄨㄞ ㄎㄨㄞ ㄍㄨㄢ ㄎㄨㄢ ㄍㄨㄤ ㄎㄨㄤ ㄍㄨㄟ ㄎㄨㄟ ㄍㄨㄣ ㄎㄨㄣ ㄍㄨㄥ ㄎㄨㄥ ㄍㄨㄛ ㄎㄨㄛ ㄌㄚ ㄌㄞ ㄌㄢ ㄌㄤ

reng ri ruo rou ru ruan rui run rong ga ka gai kai gan kan gang kang gao kao gei gen ken geng keng ge ke gou kou gu ku gua kua guai kuai guan kuan guang kuang gui kui gun kun gong kong guo kuo la lai lan lang

hae han hang haou hăn hăng hih, heh how he hëa hëang hëaou hëĕ hëĕn hin hing hew heung heu, sü, süh heuen hëŏ, süeh heun, süen hoo hwa hwae hwan hwang hwuy hwăn hung ho yĭh jen jang jaou jĕ jin

771

kae kh’ae kan kh’an kang kh’ang kaou kh’aou kăn kh’ăn kăng kh’ăng keh kh’eh kow kh’ow koo kh’oo kwa kh’wa kwae kh’wae kwan kh’wan kwang kh’wang kwei kh’wei kwăn kh’wăn kung kh’ung kwo kh’wo la lae lan lang

Karen Steffen Chung

WG

Morrison

MPS

Pinyin

WG

Morrison

MPS

Pinyin

lao lo, le lei lêng li lia liang liao lieh lien lin ling liu lo lo lou lu luan lun lung lü lüeh lün ma mai man mang mao me mei mên mêng mi miao mieh mien min ming miu mo mou mu na nai nan nang nao ne nei nên

laou leh lei lăng le lëa lëang leaou lëĕ lëen lin ling lew lo lo low lu lwan lun lung leu lëŏ

ㄌㄠ ㄌㄜ ㄌㄟ ㄌㄥ ㄌㄧ ㄌㄧㄚ ㄌㄧㄤ ㄌㄧㄠ ㄌㄧㄝ ㄌㄧㄢ ㄌㄧㄣ ㄌㄧㄥ ㄌㄧㄡ ㄌㄛ ㄌㄨㄛ ㄌㄡ ㄌㄨ ㄌㄨㄢ ㄌㄨㄣ ㄌㄨㄥ ㄌㄩ ㄌㄩㄝ

lao le lei leng li lia liang liao lie lian lin ling liu lo luo lou lu luan lun long lü lüe

năng ne

ma mae man mang maou

ㄇㄚ ㄇㄞ ㄇㄢ ㄇㄤ ㄇㄠ ㄇㄜ ㄇㄟ ㄇㄣ ㄇㄥ ㄇㄧ ㄇㄧㄠ ㄇㄧㄝ ㄇㄧㄢ ㄇㄧㄣ ㄇㄧㄥ ㄇㄧㄡ ㄇㄛ ㄇㄡ ㄇㄨ ㄋㄚ ㄋㄞ ㄋㄢ ㄋㄤ ㄋㄠ ㄋㄜ ㄋㄟ ㄋㄣ

ma mai man mang mao me mei men meng mi miao mie mian min ming miu mo mou mu na nai nan nang nao ne nei nen

nêng ni nia niang niao nieh nien nin ning niu no nou nu nuan nun nung nü nüeh ou pa p’a pai p’ai pan p’an pang p’ang pao p’ao pei p’ei pên p’ên pêng p’êng pi p’i piao p’iao pieh p’ieh pien p’ien pin p’in ping p’ing po p’o p’ou

ㄋㄥ ㄋㄧ ㄋㄧㄚ ㄋㄧㄤ ㄋㄧㄠ ㄋㄧㄝ ㄋㄧㄢ ㄋㄧㄣ ㄋㄧㄥ ㄋㄧㄡ ㄋㄨㄛ ㄋㄡ ㄋㄨ ㄋㄨㄢ ㄋㄨㄣ ㄋㄨㄥ ㄋㄩ ㄋㄩㄝ ㄡ ㄅㄚ ㄆㄚ ㄅㄞ ㄆㄞ ㄅㄢ ㄆㄢ ㄅㄤ ㄆㄤ ㄅㄠ ㄆㄠ ㄅㄟ ㄆㄟ ㄅㄣ ㄆㄣ ㄅㄥ ㄆㄥ ㄅㄧ ㄆㄧ ㄅㄧㄠ ㄆㄧㄠ ㄅㄧㄝ ㄆㄧㄝ ㄅㄧㄢ ㄆㄧㄢ ㄅㄧㄣ ㄆㄧㄣ ㄅㄧㄥ ㄆㄧㄥ ㄅㄛ ㄆㄛ ㄆㄡ

neng ni nia niang niao nie nian nin ning niu nuo nou nu nuan nun nong nü nüe ou ba pa bai pai ban pan bang pang bao pao bei pei ben pen beng peng bi pi biao piao bie pie bian pian bin pin bing ping bo po pou

mei mun măng me meaou mëĕ mëĕn min ming mew mo mŭh moo na nae nan nang naou nŭh nă nun

772

nëang neaou nëĕ nëĕn ning new no now noo nwan nun nung neu nëŏ ngow pa p’a pae p’ae pan p’an pang p’ang paou p’aou pei p’ei pun p’un păng p’ăng pe p’e peaou p’eaou pëĕ p’ëĕ pëĕn p’ëĕn pin p’in ping p’ing po p’o p’ow

Wade–Giles Romanization system

WG

Morrison

MPS

Pinyin

WG

pu p’u sa sai san sang sao sê sên sêng sha shai shan shang shao shê shei shên shêng shih shou shu shua shuai shuan shuang shui shun shuo so sou su suan sui sun sung szŭ, ssŭ ta t’a tai t’ai tan t’an tang t’ang tao t’ao tê t’ê tei

poo p’oo să sae san sang saou she săn

ㄅㄨ ㄆㄨ ㄙㄚ ㄙㄞ ㄙㄢ ㄙㄤ ㄙㄠ ㄙㄜ ㄙㄣ ㄙㄥ ㄕㄚ ㄕㄞ ㄕㄢ ㄕㄤ ㄕㄠ ㄕㄜ ㄕㄟ ㄕㄣ ㄕㄥ ㄕ ㄕㄡ ㄕㄨ ㄕㄨㄚ ㄕㄨㄞ ㄕㄨㄢ ㄕㄨㄤ ㄕㄨㄟ ㄕㄨㄣ ㄕㄨㄛ ㄙㄨㄛ ㄙㄡ ㄙㄨ ㄙㄨㄢ ㄙㄨㄟ ㄙㄨㄣ ㄙㄨㄥ ㄙ ㄉㄚ ㄊㄚ ㄉㄞ ㄊㄞ ㄉㄢ ㄊㄢ ㄉㄤ ㄊㄤ ㄉㄠ ㄊㄠ ㄉㄜ ㄊㄜ ㄉㄟ

bu pu sa sai san sang sao se sen seng sha shai shan shang shao she shei shen sheng shi shou shu shua shuai shuan shuang shui shun shuo suo sou su suan sui sun song si da ta dai tai dan tan dang tang dao tao de te dei

tên têng t’êng ti t’i tiao t’iao tieh t’ieh tien t’ien ting t’ing tiu to t’o tou t’ou tsa ts’a tsai ts’ai tsan ts’an tsang ts’ang tsao ts’ao tsê ts’ê tsei tsên ts’ên tsêng ts’êng tso ts’o tsou ts’ou tsu ts’u tsuan ts’uan tsui ts’ui tsun ts’un tsung ts’ung tu

sha shae shan shang shaou shay shin shing shĭh show shoo shwa shwae shwang shwuy shun shŏ so, sho sow su swan suy sun sung sz’, se ta t’a tai t’ai tan t’an tang t’ang taou t’aou the t’eh

773

Morrison tăng t’ăng te t’e teaou t’eaou tëĕ t’ëĕ tëĕn t’ëĕn ting t’ing tew to t’o tow t’ow tsă ts’ă chae chae tsan ts’an tsang ts’ang tsaou ts’aou tseh ts’eh

tsăng ts’ăng tso ts’o tsow ts’ow tsŭh ts’ŭh tswan ts’wan tsuy ts’uy tsun ts’un tsung ts’ung too

MPS

Pinyin

ㄉㄣ ㄉㄥ ㄊㄥ ㄉㄧ ㄊㄧ ㄉㄧㄠ ㄊㄧㄠ ㄉㄧㄝ ㄊㄧㄝ ㄉㄧㄢ ㄊㄧㄢ ㄉㄧㄥ ㄊㄧㄥ ㄉㄧㄡ ㄉㄨㄛ ㄊㄨㄛ ㄉㄡ ㄊㄡ ㄗㄚ ㄘㄚ ㄗㄞ ㄘㄞ ㄗㄢ ㄘㄢ ㄗㄤ ㄘㄤ ㄗㄠ ㄘㄠ ㄗㄜ ㄘㄜ ㄗㄟ ㄗㄣ ㄘㄣ ㄗㄥ ㄘㄥ ㄗㄨㄛ ㄘㄨㄛ ㄗㄡ ㄘㄡ ㄗㄨ ㄘㄨ ㄗㄨㄢ ㄘㄨㄢ ㄗㄨㄟ ㄘㄨㄟ ㄗㄨㄣ ㄘㄨㄣ ㄗㄨㄥ ㄘㄨㄥ ㄉㄨ

den deng teng di ti diao tiao die tie dian tian ding ting diu duo tuo dou tou za ca zai cai zan can zang cang zao cao ze ce zei zen cen zeng ceng zuo cuo zou cou zu cu zuan cuan zui cui zun cun zong cong du

Karen Steffen Chung

WG

Morrison

MPS

Pinyin

WG

Morrison

MPS

Pinyin

t’u tuan t’uan tui t’ui tun t’un tung t’ung tzŭ tz’ŭ wa wai wan wang wei wên

t’oo twan tw’an tuy t’uy tun t’un tung t’ung tsze ts’ze wa wae wan wang wei wăn

ㄊㄨ ㄉㄨㄢ ㄊㄨㄢ ㄉㄨㄟ ㄊㄨㄟ ㄉㄨㄣ ㄊㄨㄣ ㄉㄨㄥ ㄊㄨㄥ ㄗ ㄘ ㄨㄚ ㄨㄞ ㄨㄢ ㄨㄤ ㄨㄟ ㄨㄣ

tu duan tuan dui tui dun tun dong tong zi ci wa wai wan wang wei wen

wêng wo wu ya yai yang yao yeh yen yin ying yu yung yü yüan yüeh yün

ung wo woo ya yae yang yaou yay yen yin ying yew yung yu yuen yuĕ yun

ㄨㄥ ㄨㄛ ㄨ ㄧㄚ ㄧㄞ ㄧㄤ ㄧㄠ ㄧㄝ ㄧㄢ ㄧㄣ ㄧㄥ ㄧㄡ ㄩㄥ ㄩ ㄩㄢ ㄩㄝ ㄩㄣ

weng wo wu ya yai yang yao ye yan yin ying you yong yu yuan yue yun

Bibliography Aylmer, Charles (1997) ‘The Memoirs of H. A. Giles’. East Asian History 13/14: 1–90. Biographical Dictionary of Chinese Christianity: Samuel Wells Williams, 1812–1884 (web page). Accessed June 28, 2013 at: http://www.bdcconline.net/en/stories/w/williams-samuel-wells.php. Branner, David Prager (1997) ‘Notes on the beginnings of systematic dialect description and comparison in Chinese’. Historiographia Linguistica 24(3): 235–66. Branner, David Prager (1999) ‘A study of Edward Harper Parker, an early Western dialect fieldworker in China’. Journal of the American Oriental Society 119(1): 12–34. Carrington Goodrich, Luther and Fang Chaoying (eds) (1976) ‘Cattaneo, Lazzaro’, in Dictionary of Ming Biography, 1368–1644, vol. 1, New York: Columbia University Press, 32. Casacchia. Giorgio and Mariarosaria Gianninoto (2013) ‘Matteo Ricci e la lingua cinese’. Ms. Accessed May 17, 2013 at: http://www.associazionematteoricci.org/drupal/sites/default/files/documenti/12_G. CASACCHIAeM.GIANNINOTO_Matteo Ricci e la lingua cinese.pdf Chung, Karen Steffen (1989) ‘Language’, in Republic of China Yearbook 1989, Taipei: Kwang Hwa Publishing, Ch. 5. Chung, Karen Steffen (2013) ‘East Asian Linguistics’, in Keith Allan (ed) The Oxford Handbook of the History of Linguistics. New York: Oxford University Press, Ch. 10. Cordier, Henri (1985) Nécrologie: ‘Thomas Francis Wade’. T’oung Pao 6(4): 407–12. Criveller, Gianni (2010) ‘The Missionary Method of Matteo Ricci’. Tripod 30(158): 400th Anniversary of Matteo Ricci 1552–1610. Douglas, Robert Kennaway (1899) ‘Wade, Thomas Francis’, in Leslie Stephen and Stephen Lee (eds) Dictionary of National Biography, 1885–1900, vol. 58, London: Smith, Elder & Co. Fontana, Michela (2011) Matteo Ricci: A Jesuit in the Ming Court. Lanham, MD: Rowman & Littlefield. Giles, Herbert Allen (1879) On some translations and mistranslations in Dr. Williams’ Syllabic dictionary of the Chinese language. Amoy: A. A. Marcal. Giles, Herbert Allen (1901) Chinese Without a Teacher, being a collection of easy and useful sentences in the Mandarin dialect, with a vocabulary. (5th and rev. ed.) Shanghai: Kelly & Walsh. Giles, Herbert Allen (1912) A Chinese–English dictionary. Shanghai: Kelly & Walsh. Hsia, R. Po-chia (2010) A Jesuit in the Forbidden City: Matteo Ricci 1552–1610. Oxford: Oxford University Press. Killingley, Siew-Yue (1998) Learning to Read Pinyin Romanization and its Equivalent in Wade–Giles: A Practical Course for Students of Chinese. Lincom Studies in Asian Linguistics 05. München: Lincom Europa.

774

Wade–Giles Romanization system Language Hat (blog): ‘Wade vs. Giles’, October 5, 2006. Accessed on July 8, 2013 at: http://www. languagehat.com/archives/002507.php. Legge, Helen Edith (1905) ‘James Legge, Missionary and Scholar’. Web page. Accessed on July 4, 2013 at: http://www.electricscotland.com/history/legge/index.htm. Legge, James (1991) The Chinese Classics, with a translation, critical and exegetical notes, prolegomena, and copious indexes; vols 1 and 2: Confucian Analects, The Great Learning, The Doctrine of the Mean; The Works of Mencius. Taipei: SMC Publishing; reprint of last editions of Oxford University Press. Luo, Chang-pei 羅常培 (1940) Yēsūhuìshì zài yīnyùnxuéshàng de gòngxiàn, Bǔ 耶穌會士在音韻學上 的貢獻, 補. (‘The Contributions of the Jesuits in Chinese Phonology, addendum’). Addendum issued ten years after original publication in Beijing Jikan 北京季刊 (‘Beijing University Quarterly’) 7(2), 1930. Marshman, Joshua (1814) Elements of Chinese Grammar. Serampore: Mission Press. Meadows, Thomas Taylor (1847) Desultory Notes on the Government and People of China. London: Wm. H. Allen and Co. Medhurst, Walter Henry (1848) English and Chinese Dictionary. Shanghai: Mission Press. Morrison, Robert (1815) Tōngyòng Hànyán zhī fǎ 通用漢言之法 (‘A Grammar of the Chinese Language’). Serampore: Mission Press. Morrison, Robert (1865) A Dictionary of the Chinese Language in Three Parts, vol. 3, English and Chinese. Macao: P. P. Thoms, The East India Company Press. Moule, A. C. (1935) Obituary notice: ‘Herbert Allen Giles’. The Journal of the Royal Asiatic Society of Great Britain and Ireland 3: 578. Porter, Stanley E., Jeffrey T. Reed, Matthew Brook O’Donnell (2010) Fundamentals of New Testament Greek. Grand Rapids, MI: Wm. B. Eerdmans. Ricci, Matteo (1606) ‘Xīzì qíjī’ 西字奇蹟 (‘Remarkable examples of Western writing’), in Chéng Shì Mò Yuàn 程氏墨苑 (‘The Ink Garden of the Cheng Family’). Beijing. Photocopy. The Ricci 21st Century Roundtable on the History of Christianity in China database 點睛論壇. Page on Xiru ermu zi 西儒耳目資. Ricci Institute, University of San Francisco. http://ricci.rt.usfca.edu/ bibliography/view.aspx?bibliographyID=1868 Ride, Lindsay (1991) ‘Biographical Note’, in James Legge The Chinese Classics; with a translation, critical and exegetical notes, prolegomena and copious indexes, Vol 1, 3rd ed.: Confucian Analects, The Great Learning, The Doctrine of the Mean. Taipei: SMC Publishing, 1–29. The Bull Romanus Pontifex (Nicholas V), January 8, 1455, in English translation. Accessed on May 17, 2013 at: http://www.nativeweb.org/pages/legal/indig-romanus-pontifex.html. Ruggieri, Michele and Matteo Ricci (2001) Dicionário Português–Chinês (Pú Hàn Cídiǎn 葡漢辭典, 1585); intro. by John W. Witek; historical linguistic introduction by Paul Fu-mien Yang. Lisbon: Biblioteca Nacional de Portugal, Macau: Instituto Português do Oriente (IPOR), and San Francisco: Ricci Institute for Chinese–Western Cultural History. Simmons, Richard VanNess (forthcoming) ‘Pre-modern tonal notation for Chinese’ 早期漢語標調的 方法, in Rint Sybesma, Wolfgang Behr, Yueguo Gu, Zev Handel, C.-T. James Huang and James Myers (eds) Encyclopedia of Chinese Language and Linguistics. Leiden and Boston: Brill. Sinclair, Paul (2003) ‘Thomas Wade’s Yü yen tzǔ êrh chi and the Chinese Language Textbooks of Meiji-Era Japan’. Asia Major: A British Journal of Far Eastern Studies. Third Series, Academia Sinica 16 (1): 147–74. Spence, Jonathan D. (1983) The Memory Palace of Matteo Ricci. New York: Penguin. Trigault, Nicolas Xīrú ěrmù zī (1626) 西儒耳目資 (‘Aid to the Eyes and Ears of Western Literati’), vol. 3. Hangzhou: Wang Zheng. Facsimile reprint: Beijing: National Beijing University (1933), and in Pīnyīn wénzì shǐliào cóngshū 拼音文字史料叢書 (1957), Beijing: Wenzi gaige chubanshe. Scanned manuscript. http://archive.org/details/02077327.cn. Wade, Thomas Francis (1859) The Hsin Ching Lu 尋津錄, or, Book of Experiments; Being the First of a Series of Contributions to the Study of Chinese, Hong Kong: n.p. Wade, Thomas Francis and Walter Caine Hillier (1867) Yü-yen Tzu-erh Chi: A progressive course designed to assist the student of colloquial Chinese as spoken in the capital and the metropolitan department. In three volumes. Shanghai: The Statistical Department of the Inspectorate General of Customs. Wade, Thomas Francis (2002) Yǔyán Zì’ěrjí: shíjiǔshìjì zhōngqí de Běijīnghuà 語言自邇集: 19世紀 中期的北京話. Zhang Weidong 張衛東, trans., with introduction and commentary. Beijing: Beijing University.

775

Karen Steffen Chung Wellisch, Hans H. (1978) The conversion of scripts, its nature, history, and utilization. New York: John Wiley & Sons. Williams, Samuel Wells (1874) A Syllabic Dictionary of the Chinese Language; arranged according to the Wu-Fang Yuen Yin, with the pronunciation of the characters as heard in Peking, Canton, Amoy, and Shanghai. Hàn Yīng Yùnfǔ 漢英韻府. Shanghai: American Mission Press. Xu, Wenkan 徐文堪 (n.d.) ‘Robert Morrison and His Researches on Chinese Language’. 馬禮遜及其 漢語研究簡論. Studies on Traditional China 傳統中國研究 site. Accessed on June 22, 2013 at: http://www.historicalchina.net/admin/WebEdit/UploadFile/MorrisonWK.pdf. Xue, Zhixia 薛志霞 (2008) 明末傳教士漢語羅馬字注音方案性質考 (‘Study on the Source Language of the Roman Transcription of Chinese Words by Western Missionaries in the Late Ming Dynasty’) 晉中學院學報 (‘Journal of Jin Zhong University’) 25(4). Accessed on June 13, 2013 at: http:// d.wanfangdata.com.cn/periodical_ jzsfgdzkxxxb200804001.aspx. Yin, Binyong 尹斌庸 (1994) ‘How Was the First Romanized Spelling System for Sinitic Produced?’ (Diyi ge Lading Zimu de Hanyu Pinyin Fang’an Shi Zenyang Chansheng de? 第一個拉丁字母的 漢語拼音方案是怎樣產生的?) Sino-Platonic Papers 50: 1–7. Zhang, Xiping (2006) Following the Steps of Matteo Ricci to China. Ding Deshu and Ye Jinping, trans. Beijing: China Intercontinental Press. Zhao, Jianmin (2010) ‘Matteo Ricci: The Pioneer of ‘Renaissance’ in China’. Purple Kwong, trans. Paper presented at the International Symposium in Commemoration of the 400th Anniversary of Matteo Ricci 1552–1610, Fu Jen Catholic University, Taiwan, April 19–22, 2010.

776

General index

abbreviation 128, 131, 225, 228 –30, 232, 235 – 6, 348, 382, 569 ABC Chinese-English Dictionary 540, 558 Academia Sinica 65, 205, 209, 430 –2, 444, 552 – 4 Academia Sinica Balanced Corpus of Modern Chinese 430 –1, 444, 554 accent 36, 41– 4, 51, 129 –31, 759; acute 759, 762; Beijing 129; conservative 36, 42; gangtai 131, 129; grave 41, 759, 762; Hong Kong 131; macron 41; nonstandard 131; Peking 765; short 762; standard 42, 44; Taiwanese 131, 129 Account of the Old Toper’s Pavilion, An 79 acronym 126, 131, 225, 569 ACTFL Program Standards for the Preparation of Foreign Language Teachers 108 Advanced Placement (AP) 122 adverb: attitude 255 – 6; manner 256, 257, 522, 751; non-manner 256 –7; sentence-level 255, 257; time 255, 257; VP-level 256 affix: applicative 745; derivational 205, 218 –19; diminutive 586; grammatical 218, 220; handshake 524; inflectional 4, 218, 220, 222, 586; movement 524; word-forming 218 –19 affixation 4, 205, 215, 221–2 ambiguity: lexical 208, 235 – 9, 319; polynomial segmentation 427; single-syllable 319; structural 204; syntactic 319 American Council on the Teaching of Foreign Languages (ACTFL) 107 – 8, 144 Analects 69, 78, 164, 303, 329, 408 – 9, 411, 413, 417, 730; 767; see also Confucian Analects; Lunyu Analysis and Explanation of Characters, An 55, 57; see also Shuowen Jiezi; Explanatory Dictionary of Chinese Characters, An

Annotated Common Sayings 65 Annotated Explanatory Dictionary of Chinese Characters, The 538; see also Shuowen jiezi zhu anti-Japanese resistance 637 Anti-Rightist Campaign 338 Applied Glossary of Modern Chinese, The 438 – 9 argument: agent- 357, 360, 364, 366; experiencer 358; external 250 –1; internal 250 –1, 253, 750; linking 207; patient- 357 – 8, 360, 364 – 6; -predicate structure 194; structure 354, 361; wh- 526, 738 Aristotle 206, 329, 629 As the Saying Goes 406 ASCII 43, 421–5 Augmented Dictionary of Dialectal Words, The 535; see also Xu fangyan Australian Federation of Modern Language Teachers Associations 108 Baayen’s P value 205 baguwen 333 –5, 338 Bai Yue 20, 155 Bai, Yiping see Baxter, William baihua 94, 335, 409, 549, 731; see also baihuawen baihuawen 408, 731–2; Yundong 409, 416, 490, 731; see also baihua; baihua vernacular movement; Vernacular Language Movement; Written Vernacular Chinese Movement Banner Army 579 – 80 Basic Law 19 Battle of Fei River 69 Baxter, William 164; see also Bai, Yiping bei passive 733, 742 –3; see also bei-construction

777

General index beifanghua 123, 169, 502, 607, 728; see also guanhua; guoyu; Mandarin; Mandarin Chinese; putonghua Beijing Aeronautic University 439 Beijing International Chinese Character Research Association 101 Beijing Language and Culture University 439, 447 Beijing University 85, 101; see also Peking University Bendihua 480; see also Bendihua subdialect bidialectalism 491 Bieya 534 Big5 424 bihui 379, 387 bilingualism 491, 507 Binding Theory 318, 356, 372 Biography of Huo Guang 69 Biography of Wang Zun 82 Biya 534 Book of Changes, The 328, 335, 409, 730; see also Yijing Book of Han 70, 82, 531; see also Hanshu Book of Odes, The 2, 48, 664; see also Book of Songs; Shijing Book of Songs, The 409, 730, 757; see also Book of Odes, The; Shijing Bopomofo 421, 485, 546, 549 borrowing: graphic 567 – 8, 571, 573; lexical 154, 563, 565 – 6, 577; phonetic 24, 499 Bound root 205, 218 –19, 227 – 8 Branch: Gao-Yang 22; Guangfu 21; Guinan 23; Siyi 21–2; Yuehai 21–3 Buddhism 78 – 9, 120, 167, 303 – 4, 386, 388, 411, 537, 564, 730 Buddhism Idioms 70 calligraphy 52, 112 cangci 396 Cangjie pian 534 Cantonese 18 –31, 35 – 49, 98, 123, 125, 129, 162, 166, 344 – 6, 349, 421, 424, 434 – 6, 469 –72, 474 – 8, 492, 507 – 8, 521, 526, 528, 541, 573, 575, 606 – 8, 620, 622 – 4, 692 –3, 724, 732, 734 –5, 750, 764, 768; see also Cantonese dialect; Yue dialect Cantonese Dictionary: Cantonese-English, English-Cantonese 40 Cantonese Made Easy 37 – 8 Cantonese Primer 38 Cao, Cao 72 Cao, Pi 665 Cao, Zhi 72, 665 captatio benovolentiae 330 Cattaneo, Lazzaro 758 – 9 Center for Chinese Linguistics Corpus (CCL) 430, 552, 554; see also PKU Corpus

Chang E. 72 Chao, Yuen-Ren 38, 164, 173, 485, 489, 540, 610, 769 Chen, Di 163 Chen, Duxiu 416, 645, 731 Chen, Heqin 438 Chen, Kui 332 –3, 336 Chen, Pengnian 535, 757 Chen, Wangdao 336, 397 Chen, Zhenqiu 336 chengyu 64, 235, 295, 312, 344, 402, 448; see also idiom Chiang, Kai-shek 301, 637 Children’s Knowledge Treasury, The 65 China Central Television (CCTV) 66, 344, 548 China Digital Times (CDT) 127 China Idioms Conference 81 China Internet Network Information Center (CNNIC) 126 Chinese as a Foreign Language (CFL) 58, 105 – 8, 440, 445, 447, 559; New Word Dictionary for 441; Teaching (TCFL) 447 – 8; see also Chinese as a Second Language Chinese as a Native Language (CNL) 137, 199, 320, 439, 552, 688 –701 Chinese as a Second Language (CSL) 106, 115, 122, 137 – 8, 140 –3, 147 –50, 440, 447, 652; see also Chinese as a Foreign Language Chinese as an object of learning 109, 111–12 Chinese character 51– 63, 141–2, 176, 206, 315 –16, 439, 545 –7, 550 –1, 558, 686; complex 53; formation method 55, 531; frequency 438 – 9; inherited 53 – 4; input 420 –3; Modern 52 – 9; normalization 52 –3; normalized forms 53; regularization 52, 531; simplification 52 –3, 96, 172, 490, 499; simplified 53 – 4, 96 –7, 102, 142, 171–2, 424, 435 – 6, 551, 569, 727; standard forms 53, 538; standardization 53, 171, 727; systemization 53; traditional 53 – 4, 96, 101–2, 142, 171–2, 424, 435 – 6, 569 Chinese Character Database: Phonologically Disambiguated According to the Cantonese Dialect 541 Chinese Chrestomathy in the Canton Dialect 37 Chinese Communist Party (CCP) 93, 120, 338, 633, 637 Chinese crisis 121–1 Chinese Dictionary in the Cantonese Dialect, A 37 Chinese Interlanguage Corpus 448 Chinese Language Law 98, 124; see also Law of the People’s Republic of China on the Use of Chinese Language and Chinese Characters Chinese language: curriculum 417; education 66, 104 –17; learning 122, 421; pedagogy (CLP) 137, 142; Proficiency Test 58, 128, 447;

778

General index reform 53, 93 – 9, 334, 489 – 90; standardization 53, 123, 131, 171, 574 –7, 609, 724 –35; teaching 104, 137, 767 Chinese Riddle-phrases 406 Chinese Syllabary Pronounced according to the Dialect of Canton 45 Chinese without a Teacher 767 Chinese Word Bank from Mainland China, Taiwan, and Hong Kong, A 439 Chinese Yellow Emperor Association 101 Chinese: Ancient 1–15, 76 –7, 82 –3, 204, 430, 471, 473, 706, 724; Archaic 1, 9, 664, 720; Classical 1–15, 58, 65 – 6, 70 –3, 79, 93 –5, 228, 403, 408 –17, 486, 531, 645 – 61, 664 – 80, 723, 727, 730 – 4; Contemporary 1, 4, 10, 14, 130, 132, 204, 312, 329, 382, 413, 549, 686, 714 –16; formal 646 – 60; Hong Kong Written 734 –5; literary 395, 408 – 9, 416, 645 – 6; Mandarin 35, 51, 57, 90, 102, 195, 344, 444, 446, 484 –5, 563, 572, 650, 655, 750, 756 –70; Medieval 4, 619; Middle 1– 4, 164 – 9, 177, 409 –14, 471–3, 478, 487, 491–2, 605 –19, 680, 757; Modern 1, 52 – 9, 65, 76, 79, 82, 102, 132, 175, 204, 228, 247, 336, 381, 385, 391, 409 –13, 421–3, 430, 435, 439, 456 – 67, 469, 485, 492, 501, 536, 539, 548 –54, 567 –74, 645 – 61, 678, 715 –16, 727, 732; Old 1– 4, 24, 155, 164, 166 –7, 169, 171, 408, 410 –13, 680; Pre-Archaic 10; Pre-Classical 1; Pre-Medieval 1; Pre-Modern 410, 551; simplified 171–2, 424, 435 – 6, 547, 551, 727; spoken 83, 92 –3, 97, 109 –10, 400, 445, 458 – 9, 461, 464, 465, 645, 653, 659 – 60, 723 – 9, 757; Standard 24, 45, 107, 122, 145, 346, 412 –17, 459, 464, 467, 563, 580, 605 – 9, 723 –35, 763; traditional 435 – 6, 547, 668; vernacular 65, 345, 409, 414 –16, 648, 727, 731; written 19, 55, 70, 93 –7, 444, 459 – 65, 528, 645 – 61, 727, 729 –35 Chinese-English Dictionary: of Enigmatic Folk Similes, An 405 Chinese-English Explanation Dictionary of Xiehouyu, A 406; see also Hanying Shuangjie Xiehouyu Cidian Chineseness 122 –5 Chinglish 128 – 9, 131 Chronicles of Zuo 408; see also Zuozhuan Chuci 155 – 6; see also Songs of Chu, The Chung, Karen Steffen 756 Chunqiu 730 cibenwei 139; see also word-based pedagogy Cicero 330 cidian 545, 548, 551; see also word-based dictionary Cihai 83, 343, 538, 551, 559; see also Dictionary of Words, The; Encyclopedic Dictionary of Chinese, The; Sea of Words

Ciyuan 538, 551, 713; see also Origins of Words, The; Sources of Words classifier: 15, 26 –7, 112, 208, 223, 242 –5, 247 – 8, 269 –74, 281, 286 – 8, 354, 356, 368 –70, 410, 413, 444 –5, 508, 513, 521–2, 524 – 6, 553, 589, 591–2, 596, 607 – 8, 614 –5, 617 – 8, 620 –1, 623, 711–3, 748; general 248, 273, 370, 625; sortal 244, 247, 269 –73, 276 –7 clerical change 52 –3 cognitive processes 559, 685, 687, 691– 4, 696, 700 cognitive strategies 686, 698 –701 College Entrance Exams 66, 121, 417 Collins-Birmingham University International Database 552 Commercial Press 67, 438, 538 – 9, 551, 558 Commission for the Management of Language Use 128 Committee for Chinese Language Reform 53, 726 – 9 competition model 318 Complete List of Simplified Characters 53 composition 221 compounding 4, 14, 126, 221, 227, 237 – 9, 410, 448, 715; Prosodic Word (PrWd) 679 – 80 Comprehensive Chinese Dictionary of Characters 551; see also Great Character Dictionary of China, The; Zhonghua Da Zidian Comprehensive Dictionary of Chinese Characters 532; see also Zihui Comprehensive Frequency Table of Character Usage in Modern Chinese, A 439 computer-mediated communication (CMC) 126 Conceptual Mapping Model 633 Conceptual Metaphor Theory (CMT) 629, 631 Concise Dictionary of Chinese Idioms, A 66 Concise English-Chinese and Chinese-English Dictionary 558 concordance 431–2 Condition on Extraction Domain (CED) 737, 751 Confucian Analects 4, 9 –12, 14, 409, 767; see also Analects; Lunyu Confucianism 94, 100, 379, 730; neo- 388 Confucianist 95 Confucius 78 – 9, 122, 152, 295 – 6, 298, 300 –1, 303 – 4, 308, 328, 330, 379, 386, 388, 402, 411, 730; see also Kongzi Confucius Institute 101, 122 conjunction: clausal 266 – 8; phrasal 264 construction: ba- 140, 354, 357, 365, 367 – 8, 751; bei- 130, 354, 357, 366, 698, 742; bi- 28; comparative 29, 611, 621, 735; dative 6, 251; de- 128, 185; discontinuous 30; disposal 15, 120, 611, 614, 618, 619, 621,

779

General index 623, 625; double-object 13, 24, 27 – 8, 251, 611–12, 614, 617 –18, 620 –1, 623, 625; information-packaging 603; passive 8, 13, 358, 412, 445, 621, 718, 742; pivotal 5; resultative 15, 598, 697, 740; serial verb (SVC) 13, 524, 716, 739, 741–2, 751–2 Contemporary Chinese Dictionary 66, 67, 227, 551, 734; see also Dictionary of Modern Chinese; Modern Chinese Dictionary; Xiandai hanyu cidian Contextualism 193 –5 corpus linguistics 420, 429 –31, 438 – 9, 449, 552, 560 couplet 193 – 4, 473, 650, 668 – 9, 670, 671, 673, 677, 679 Crazy English 121 Cultural Revolution 96, 120 –1, 225, 337 – 8, 397, 399, 417, 732; see also Wenhua Dageming curriculum design 104, 143, 145, 148 Daily: China Education 443; Guangming 443; Peopls’s 132, 443, 635; Shanghai 228; Southern 443 Daxue 408, 730; see also Great Learning, The Dazangjing see Tripitaka Defrancis, John 20, 124 –5, 232, 417, 540, 558 Deng, Xiaoping 389; Selected Works of 67 determiner phrase 354, 370, 748 dialect 18 –30, 43, 51, 92, 95 – 9, 107, 119 –20, 123 –31, 163, 169 –70, 173 – 8, 195, 343 – 8, 380, 390 –1, 410, 434 –5, 469 – 82, 487 – 92, 501–3, 535, 538 – 41, 563, 579 – 80, 586, 605 –25; Beijing 19, 97, 223, 390 –1, 410, 490, 580, 609, 765; Cantonese 435, 573; Gan 469, 473, 476 –7, 481, 614 –16; Hakka 22, 469 – 82, 607, 620 –1; Hong Kong 479; Hui 616 –17; Jin 606, 611–12; Mandarin 608 –11; Min 22, 473, 482, 618 –20; modern 166 –7; Northern 434, 579, 609; Pinghua 624 –5; social 131; Southern 24, 26, 123, 487, 584; Southern Min (Minnan) 472, 479, 620; Wu 125, 617 –18; Xiang 613 –14; Yue 21, 473, 622 –3 dialectal: background 274; basis 727; border 346; comparison 23 –30; 492, 625; variation 13, 398 dialect-language dichotomy 18 –20 dialectology 501 Dictionary of Chinese Character Variants 541 Dictionary of Chinese Grammar and Rhetoric (DCR) 334 Dictionary of Chinese Idioms 65, 67 – 9 Dictionary of Initial Consonants 535; see also shenglei Dictionary of Modern Chinese 132, 727; see also Contemporary Chinese Dictionary;

Modern Chinese Dictionary; Xiandai hanyu cidian Dictionary of Standard Cantonese Pronunciation, The 46 Dictionary of Textual Researches on the Wu-Xia Dialect, The 535 Dictionary of the National Language, The 538; see also Guoyu cidian Dictionary of Words, The 538; see also Cihai; Encyclopedic Dictionary of Chinese, The; Sea of Words dictionary: bilingual 443, 539 – 40, 549, 556; Cantonese 36 –7; character-based 545, 550; Chinese 45, 111, 441, 443 – 4, 458, 484, 536 –7, 545 –50, 556 –7, 559 – 60, 760; corpus-driven 553; e- 559; frequency 439; graphically organized 531; learner’s 556; Machine-readable (MRD) 559; phonetically organized 531; radical-based 550; rhyme 1, 166 –7, 491; semantically organized 531; specialised 443, 560; word-based 545, 551–2; xiehouyu 397, 405; see also cidian; zidian Discourses of the States 408 Dizionario Portoghese-Cinese 539, 557 Doctrine of the Mean, The 408, 417, 730, 765; see also Zhongyong Dream of the Red Chamber 67 – 8, 710; see also Honglou Meng Dunhuang Bianwen 730 Duyin Tongyi Hui 489, 724, 731 dynasty: Eastern Han 161, 531, 564, 630, 636, 665, 671, 674 – 82; Eastern Jin 52; Eastern Zhou 161, 166; Han 52, 54, 65, 93, 153, 166 –7, 169, 381, 388, 408 – 9, 414, 531, 533 – 4, 564, 665, 677, 716, 730; imperial 159; Jin 161, 167, 469, 579; Liao 160 –1, 167, 169, 579; Ming 65, 77, 82, 161, 379, 383 – 4, 409, 411, 414, 486, 532, 536, 540, 565, 579, 724, 731; Ming-Qing 536, 679; Northern Song 161, 723; Qin 8, 51, 54, 69, 72, 92, 159, 161, 171, 396, 409, 531, 622; Qing 57, 65, 82, 96, 154, 161, 163, 167, 335, 344, 379, 383, 388 – 9, 409, 416, 430, 481, 485, 489, 532, 536 – 9, 565, 579 – 80, 620, 645, 724, 731; Qin-Han 379, 387; Shang 51–2, 54, 160 –1, 163, 169, 174, 344, 408, 413, 531; Six 665, 716; Song 52, 64 –5, 68, 79, 166, 168 – 9, 308, 332 –3, 379, 387, 391, 396, 469, 531, 565, 620, 624, 636, 665, 730 –1; Song-Yuan 65; Southern Liang 531; Southern and Northern (Nanbei Chao) 161, 167, 564, 680; Southern Song 161, 332, 723; Sui 161, 164, 167, 169, 333, 564, 730; Sui-Tang 379, 388, 564, 677; Tang 18, 52, 65, 68, 71, 94, 98, 120, 161, 167, 176, 308, 332 – 4, 384, 388, 396, 409 –11, 564, 614, 665, 679, 716, 730; Tang-Song 308, 387,

780

General index 679; Wei-Jin 92, 531, 534 –5, 679; Western Han 161, 564, 674 –5, 682; Western Xia (Xixia) 161, 164, 167; Western Zhou 161, 408, 413; Wudai Shiguo 161, 167; Xia 161; Yuan 52, 65, 82, 154, 161, 167 – 9, 334, 414, 565, 579, 636, 724, 731; Zhou 51, 54, 161, 163, 166, 174, 227, 379, 386 –7, 408, 430, 654 educational affordances 109, 113 Emperor: First 159, 171; Kangxi 57, 153, 537, 550, 764; Qin Shi Huang 331, 389; Shenzong 486; Taizong 388; Yellow 92, 98, 101 Empire: Chinese 92, 567; Han 166, 381; Qin 331; Qing 335; Qin-Han 379; Roman 153; Zhou 409 Empress Dowager Ci Xi 311 encoding: character 424, 546; form-based 422; morphological 322; sound-based 421 Encyclopedic Dictionary of Chinese, The 343; see also Cihai; Dictionary of Words, The; Sea of Words Engels 337 English and Cantonese Pocket-Dictionary 37 English and Chinese Vocabulary in the Court Dialect, An 540 English as a Second Language (ESL) 121 erhua 492, 496 Erya 533 –7, 545; see also Ready Guide, The Erya yi 534 Ethnologue 608 – 9 Etymological Comprehensive Dictionary of Chinese Idioms 67 event-related potential (ERP) 80, 318, 320 –1 evolution: Chinese characters 531; Chinese sound spelling 489; idiom 81, 157, 169; language 66, 229; lexicography 532; poetic 677; semantic 713 Explanatory Dictionary of Chinese Characters, An 531; see also Analysis and Explanation of Characters, An; Shuowen Jiezi Expressions of Written Chinese 653, 656 Extended Rhymes Dictionary, The 535; see also Guangyu extra-metrical character 673 Fang, Yizhi 487 – 8, 534 Fangyan 166, 534 –5 fangyan 20, 123 –5, 344 fanqie 167 – 8, 487, 491, 496, 499, 533, 536 –7, 549, 725, 728, 756 –7 Fengsu Tongyi 71 Fernandes, Sebastian 758 figure of speech 70, 193, 395 first language (L1) 105, 109 –10, 113, 115 – 6, 406, 448 – 9 Foot Binarity 459 – 61, 467, 667, 669

Former Ode to the Red Cliff 82 FrameNet 209 framework: orthography-based 545; phonemebased representational 549; Pragmatic 142 –3, 145 – 6, 148; Principles-and-Parameters (PnP) 739; Prosodic Word (PrWd) 652 Frequency Dictionary of Chinese Words 439 Frequency Dictionary of Mandarin Chinese: Core Vocabulary for Learners, A 555 Frequency Dictionary of Modern Chinese, A 439 – 40 Frequency Table of Character Usage in Modern Chinese, A 439 fu 665 Fu, Sinian 416 Fudan University 336 fuzzy syntax 686, 696 – 8 gangtaiyin 129; see also Cantonese- and Taiwanese-accented Mandarin; gangtai accent Gao, Benhan 163; see also Karlgren, Bernhard General Administration of Press and Publication (GAPP) 131, 576 General Outline of the Chinese Vocabulary Levels and Graded Chinese Characters 58 Giant Dictionary of the Chinese Language, 57 Giles, Herbert 172, 488, 557, 766 Golden Bough 380 Gong, Zizhen 335 Gong-sun, Long 329 Government Language School 46 Grammatical Knowledge-base of Contemporary Chinese 552, 555 grass-mud horse 127, 352 Grass-Mud Horse Lexicon 127 Great Character Dictionary of China, The 537; see also Comprehensive Chinese Dictionary of Characters; Zhonghua Da Zidian Great Chinese Character Dictionary, The 532, 539; see also Hanyu da zidian Great Chinese Word Dictionary, The 539; see also Hanyu da cidian Great Firewall of China 127, 351; see also Jindun gongcheng Great Leap Forward 338, 635 Great Learning, The 408, 417; see also Daxue Guan Ju 164 Guangya 534 Guangya shuzheng 538 Guangyun 535 –7, 540, 757; see also Extended Rhymes Dictionary, The guanhua 123, 168, 609, 723, 724, 730; see also beifanghua; guoyu; Mandarin; Mandarin Chinese; putonghua guanyongyu 402 Gui Guzi 329 –30, 336 Guo, Jujing see Cattaneo, Lazzaro

781

General index Guojia Biaozhun (GB) 424 –5 Guoyin Changyong Zihui see Zihui guoyu 95, 97, 125, 490, 563, 609, 723 – 9, 733 – 4; see also guanhua; Mandarin; Mandarin Chinese; putonghua Guoyu cidian 538, 540; see also Dictionary of the National Language, The Guoyu Luoma zi 173, 489, 539, 728; see also Gwoyeu Romatzyh Guoyu Ribao Liang Cidian 553; see also Mandarin Daily News Dictionary of Classifiers Guoyu Yundong 490, 731; see also National Language Movement guwen 331–5, 408 – 9 Gwoyeu Romatzyh (GR) 173, 489, 500, 769; see also Guoyu Luoma zi Han Chinese 18, 20, 90 –1, 96, 98 –100, 379, 388, 565 Han Da Comprehensive Dictionary of Chinese Idioms 67 Han Fei 72 Han Feizi 330, 335 Han Feizi 64, 71, 330 Han, Xin 72 Han, Yu 332 –3, 388, 409, 679 Hanban 106, 108, 122; see also Office of Chinese Language Council International; Office of International Chinese Hanshu 531; see also Book of Han Hanying Shuangjie Xiehouyu Cidian 406; see also Chinese-English Explanation Dictionary of Xiehouyu, A Hanyu 1, 101, 153, 162 –3, 167, 169, 170, 177 Hanyu da cidian 539, 559; see also Great Chinese Word Dictionary, The Hanyu da zidian 532, 540; see also Great Chinese Character Dictionary, The Hanyu Pinyin 169, 171–3, 484, 727 – 8, 756; Cihui 729, 733; Fang’an 484, 490 –2, 494, 499, 727; see also pinyin Hanyu shuiping kaoshi (HSK) 128, 440; see also Chinese Language Proficiency Test Hanzi 170 –1, 423, 485, 488 – 9, 566 hapax legomena 205 Heou ci 649 –53, 656 heritage speaker (HS) 105 Heroes of the Water Margins 731 hidden complexity 354 – 6, 374 History of the Liu Song, The 677 History of Vernacular Literature, A 409 homophony 174, 459 – 61, 464 –5, 467 Hong Kong Institute of Education (HKIEd) 46 Hong Kong Polytechnic University 439 Hong Kong School for the Deaf (HKSD) 505 – 8

Honglou Meng 710, 731 HowNet 208 Hsin Ching Lu 764 –5 HSK Dynamic Composition Corpus 447 Hu, Shi 330, 335 – 6, 409, 416, 645 huaben xiaoshuo 731 Huainanzi 71 Huang, Kan 648 Huang, Tingjian 52 Huang, Zunxian 645 Huayu 167, 609, 723 Hui Shi 329 ICTCLAS 555 identity: bicultural 116; cultural 107, 125, 638 – 9; ethnic 638 – 9; group 91, 638, 640; Hakka 469; ideological 638; linguistic 132; morphemic 215 – 6; national 90 – 4, 97 –102, 635 – 6, 638 – 9; religious 640; self- 475; socio-cultural 132; state 91, 98; youth 127 idiom 64 – 86, 177, 217, 344, 411, 413 –14, 435, 448, 538, 540, 552, 634, 652, 740; fourcharacter 312 –13, 402 –3, 417; four-syllable 678 – 9; homophonic 235; literary 295; three-character 402 – 4; see also chengyu Imperial Dictionary of Kangxi, The 532; see also Kangxi Dictionary; Kangxi zidian informality 98, 342, 343 –7 information: and communication technology (ICT) 113; conceptual 206, 697; contextual 143, 195, 199, 208, 427; Extraction 211; flow 696 –7, 700; grammatical 205, 217 – 8, 354, 553, 558; load 502; morphemic 216, 218; -packaging 603; phonetic 169, 686 – 9; phonological 317, 556, 688, 691–3, 700; procedural 188; processing 420, 559, 567; retrieval 430 –1; semantic 204, 233, 271, 316 –7, 320, 686; structural 68, 143; technology 424, 540 inscriptions: bronze 10, 14, 51, 54, 161, 408, 499, 502, 533; clerical script 52 –3, 54, 531; oracle-bone (OBI) 1, 9–10, 14, 51, 54, 161–2, 408, 499, 502; regular script 52 –3, 54, 531; seal script 51–2, 54, 531, 538 Institute of Computational Linguistics 555 Institute of International Education (IIE) 121–2; Institute of Linguistics 177, 539, 551 Intelligent Chinese Character Encoding System 547 International Confucius Foundation 101 Internet-mediated communication 119 Intorcetta, Prospero 760 Jia, Pingwa 124 jiachanghua 395 Jiaguwen 161–3, 169, 499 Jiang, Zemin 334, 351

782

General index Jiangsu Province Research Association of Chinese Idioms 86 Jin, Nige see Trigault, Nicolas Jindun gongcheng 127; see also Great Firewall of China Jiu Tangshu Zheng Qi Liezhuan 396 Jiyun 535, 537; see also Rhyme Dictionary, The Jockey Club Sign Bilingualism and Coenrollment in Deaf Education Programme 507 jubenwen 141; see also sentence-based pedagogy Jurchen 169 Jyutping 41, 43, 421 Kangxi Dictionary, The 57, 760; see also Imperial Dictionary of Kangxi, The; Kangxi zidian Kangxi zidian 532 –3, 536 – 8, 540, 546, 550 –1, 559; see also Imperial Dictionary of Kangxi, The; Kangxi Dictionary Karlgren, Bernhard 163; see also Gao, Benhan Kejia see Hakka dialect Key Word in Context (KWIC) 431 Khitan 160, 169; see also Qidan Kong, Rong 665 Kongzi 152 –3, 163 – 4, 166; see also Confucius L1 see first language L2 see second language Ladinghua Xin Wenzi 489 Lancaster Corpus of Mandarin Chinese (LCMC) 444 –5, 464, 553 –5 Lancaster Los Angeles Spoken Chinese Corpus (LLSCC) 445 Lancaster University 554 Language Atlas of China (LAC) 20, 470, 475 –7, 605 – 8 language: agglutinative 215; analytic 4, 215, 606, 719; Austroasiatic 20, 155 – 6; Austronesian 15, 20, 154, 155 –7; change 39, 93, 228, 689, 707; contact 20, 132, 155, 177, 471–2, 481, 510; Flagship Program, The 122; Indo-European 141, 195, 242 –5, 249, 252, 254, 260, 264, 269, 271, 286, 563, 568, 572 –3, 686, 706, 708, 710, 713, 718, 720; inflecting 215; Internet 120, 126 –7; isolating 4, 215, 243, 574, 606, 719; as an object of learning 109; Sinitic 17, 287, 605 – 8, 610, 622 –3, 625; socially embedded 130; structure (LS) 137 – 9, 142 – 4, 146 –7; syllabic 139; as a system of systems 138; Tibeto-Burman 9, 167, 605; use (LU) 120, 123, 128, 132, 137 – 8, 142 – 8, 184, 195, 443, 446, 448, 505, 635, 715; youth 126 Lao, She 416

Law of the People’s Republic of China on the Use of Chinese Language and Chinese Characters 97, 123; see also Chinese Language Law learner-centred teaching 114 Leeds University 554 Legend of the Condor Heroes, The 84 Legge, James 760, 763 Lei, Feng 634 –5 Leipian 531–2 Lenin 94 lexical access 315, 317, 693 lexical semantic relations (LSRs) 210 –2 lexical semantics 203, 207 – 9, 212 Lexical-Functional Grammar 356 Li, He 388 Li, Huode 470 Li, Madou see Ricci, Matteo Li, Shangyin 396 Li, Yage see Legge, James Lian, Po 72 Liezi 70 –1, 312 Liji 379, 385, 390 Lijing 730 Lin, Xiangru 72 Lin, Yutang 500, 769 Linguistic Variations in Chinese Speech Communities (LIVAC) 430, 443 Lisao 664 List of Cantonese Pronunciation of Commonly-used Chinese Characters 46 List of Chinese Character Forms for General Printing 53 List of Four Thousand Words for Foreign Students 439 List of Frequently Used Characters in Modern Chinese 53, 58, 439 List of Generally Used Characters in Modern Chinese 53 – 4 List of the First Group of Standardized Forms of Variant Characters 53 Literary Mind and the Carving of Dragons, The 331, 672; see also Wen Xin Diao Long Literary Revolution 645 Little Red Book 337, 558 Liu, Bannong 416, 731 Liu, Bei 397, 402 Liu, Gongquan 52 Liu, Xiaobo 338 Liu, Xie 331–2, 672, 678 Liu, Zongyuan 71, 82, 409 Liushu 706 liyu 343 –5, 347 –52; Internet 347; regional 344; see also slang loan: graphic 567 – 9; hybrid 567, 569; phonetic 55 – 6, 224, 499, 563; phonetic-semantic 224,

783

General index 499, 502; semantic 224, 238, 563, 567 – 8, 570 –1, 575 localizer 279 – 81, 286 –7 logograph 55 – 6, 317, 571, 573, 686 –700, 706 Lu, Ban 80, 82 Lu, Fayan 491, 535, 549 Lu, Xun 347, 382, 416, 489, 548 Lu, Zhuangzhang 488 – 9, 500 Lü, Shuxiang 177 Lunheng 674 Lunyu 14, 408, 730 Luo, Changpei 176, 486, 489 Luo, Mingjian see Riggieri, Michele Luoyang Qielanji 413 Ma, Lixun see Morrison, Robert Ma, Shiman see Marshman, Joshua machine translation (MT) 128, 420, 430, 432 – 4; Chinese 432; dialect 434 –5; evaluation 434; rule-based English-Chinese 433; word-byword 128 machine-assisted translation 432 machine-learning 208 Mai, Dusi see Medhurst, Walter Henry Mandarin 18 – 9, 21–31, 35, 40, 42, 44 – 6, 65, 119, 122 – 9, 168 – 9, 172, 177, 197, 199, 216, 225, 318 –22, 343 –50, 350, 410, 412, 416, 471, 473, 475, 485 – 93, 496 –502, 510, 539, 546, 549, 555, 571–3, 575, 579 – 625, 632, 690, 692 –3, 723 –7, 758, 762 –3, 765, 768 – 9; Beijing 129, 131, 345, 609, 725, 727, 730, 762; Cantonese- and Taiwanese accented 129; Early 168; Guizhou 125; Jianghuai 611; Mainland 575; Nanjing 759; New 580, 586; nonstandard 129; northern 124; Old 1–2, 580, 586; Shaanxi 127; Shanxi 124; Southwestern 610, 613, 624; Standard 119, 122 – 4, 131, 344 –5, 349 –50, 608, 610 –11, 617 –18, 625; see also beifanghua; guanhua; guoyu; Mandarin Chinese; Putonghua Mandarin Chinese Classifier and Noun-Classifier Collocation Dictionary 444 Mandarin Daily News Dictionary of Classifiers 553; see also Guoyu Ribao Liang Cidian Mao era 349 Mao, Zedong 67, 84, 94, 120, 311–2, 397, 399, 490, 637; Selected Works of 67 marker: aspect 24, 26, 128, 199, 217, 222, 248, 250, 261, 365, 444 – 6, 588, 590, 601, 714; causative 367, 714; collective 244, 414; connective 9; inflection 222 –3; morphological 8, 590; negative 710 –11; nominalization 218; object 318, 714; passive 29, 130, 318, 412, 603, 610, 615, 617, 623, 625, 714; prepositional 6; preverbal 10; simile 77; tone 726, 729 Marshman, Joshua 760

Martini, Martino 760 Marx 94, 337, 386 Marxism 100, 386 Marxist 94, 100; -Leninist 96 Master of Chinese as a Foreign Language (MCFL) 106 meaning: counterfactual 195 – 6, 198 – 9; sentence 184; word 184, 207 – 8, 555, 560 Medhurst, Walter Henry 415, 488, 540, 557, 760 Mei, Yingzuo 532, 536 melodic unit 668 – 9 Mencius 14, 64, 69, 71, 310, 408 – 9, 413, 417, 630, 730; metacompounding 228, 230, 232, 237,239 Mi, Fu 52 Mid-Autumn Festival 112 mingbianxue 328 minimal word effect 460 Ministry of Education 94, 439, 724 –5, Chinese 122, 729; of Taiwan 65 – 67 minorities 90, 98 –101, 565 – 6, 605, 638 – 9, 640 Mo, Yan 124 Modern Chinese Dictionary, A 395, 456, 461, 539; see also Contemporary Chinese Dictionary; Dictionary of Modern Chinese; Xiandai hanyu cidian Mohism 328 Monkey King 70 morpheme: bound 204 –5, 234, 239, 411, 473, 574, 589, 652; bound content 218; bound function 218 –20; character- 206; derivational 4; disyllabic 588; free 204 –5, 217, 590, 652, 715 –16; free content 217; free function 218; monosyllabic 457, 461, 588, 652; negative 216, 602; passive 29; polysyllabic 233; semi-free 651–3 morphological: changes 574; derivation 588, 593, 595; process 4, 215, 220, 611, 734; productivity 204 –5; typology 215 –16; units 232; variation 24 morphology 4, 80, 195, 203, 205, 215, 218 –19, 232, 258, 355, 444, 505, 520 –3, 526, 563, 571, 588, 606 –7, 610, 706, 720; Chinese 205, 215, 652; derivational 177, 521, 606; HKSL 520; inflectional 258, 522 –3, 526, 606, 719; prosodic 651– 4, 66 – 8; -syntax interface 750; -syntax Mapping Hypothesis 741 Morpho-Semantic Linkage (MSL) 211 morpho-semantics 203 – 4, 211–2 Morrison, Robert 37, 488, 540, 557, 760 mother tongue 51, 123 – 4, 131, 415, 439, 556 Movement: baihua vernacular 120; Common Speech 397; language-reform 330; May Fourth 409, 417, 487, 490, 645, 647 – 8, 650, 731, 764; National Language 731; New

784

General index Culture 416, 439; pro-Cantonese 125; Vernacular Language 409, 545, 549; Written Vernacular Chinese 65; see also Baihuawen Yundong; Guoyu Yundong; Wusi Yundong; Xin Wenhua Yundong Music Record 677 Nanchang Dialect Dictionary 615 Nankai University 85 nation 90 –1; Chinese 90 –1, 93, 98, 100 –1; ethnic 91; linguistic 96; multi-ethnic 100; official 91; -state 90 –1, 119, 123 – 4, 410; territorial 91 National Commission of Language and Script 576 National Education Development Statistical Bulletin 121 National Language Commission 53, 421, 439 National Language Commission Corpora 430 nationalism 90 –3, 100, 121; cultural 91, 102; ethnic 96, 101–2; political 91, 93, 96, 102; racial 100 nationalist 93 –7, 122; cultural 100 –2; political state 102 Natural Language: generation 429; Processing (NLP) 204, 420, 429, 555; understanding 420, 428 neologism 120, 127, 130, 225, 227 – 9, 231– 9, 442 –3; compound 236 –7; disyllabic 232; -forming strategy 229; monosyllabic 228 New Oriental School 121; see also Xindongfang xuexiao Ni Xinglunan 679 Niannujiao Kunlun 84 Notebook during Home Leisure Period, The 65 noun: abstract 244, 247 – 8, 278, 618; collective 244, 246 –7, 274; count 244 –5, 270, 278, 413, 612; mass 243 –5, 269 –70, 277 – 8, 369, 413 Office of Chinese Language Council International 122; see also Hanban; Office of International Chinese Office of International Chinese 106; see also Hanban; Office of Chinese Language Council International Old Country Song 677 Old Folk Songs 677 ontology 206, 633 Opium War 411, 416, 488, 557, 565 Optimality Condition 668 – 9, 675 –7 Ordinary Language Philosophy 184 Origins of Words, The 538; see also Ciyuan; Sources of Words orthographic: convention 546 –7; information 317; -to-phonological conversion 317; system 109; unit 204, 216, 220, 545, 689; variation 546

Ouyang, Xiu 79, 82, 332 Ouyang, Xun 52 Overseas Chinese School for the Deaf and Dumb (OSCD) 505 – 6, 508 Oxford Advanced Learner’s Dictionary (OALD) 556 Oxford Chinese-English, English-Chinese Dictionary 132 Oxford Handbook of Chinese Linguistics 176, 178 pan-Chinese 443 parallel prose 331, 409, 675; see also pianwen particle: Cantonese 24 –5, 29 –30; Classical Chinese 6 –7, 14, 73; Mandarin 595, 599 – 601, 607; modern Chinese 281 pedagogy 104, 106, 137 –50; character-based 139; CSL 142, 149; sentence-based 141; word-based 139; see also cibenwei, jubenwei, zibenwei Peking Man 157 – 8 Peking Syllabary 764 Peking University 65 – 6, 85, 225, 430, 444, 552, 554, 555 Peking University 65 – 6, 85, 225; see also Beijing University Peking University Center for Chinese Linguistics (PKUCCL) 82, 430, 444, 552, 554 Peng, Dehuai 337 People’s Republic of China (PRC) 19 –20, 23, 90, 94 –102, 112, 161, 336, 338, 345, 350, 424, 439, 489 – 90, 539, 566, 580, 637, 723, 726 – 8, 732 – 4, 768 period: ancient 51–2, 54; Archaic 1, 15; Classical 8, 410, 414; Contemporary 545 – 6; Eastern Han 564; Eastern Zhou 664; Han 1, 5, 14, 120; Han-Wei 665; Holocene 158; May Fourth 120, 397; Medieval 1, 8, 15; Middle 1, 605; modern 51–2, 54; post-1978 577; post-imperial 94; post-socialist 124, 130; Pre-Archaic 14; Pre-Han 533, 647; Pre-Medieval 5; Pre-Qin 71, 382, 534, 665; Protestant missionary 760; reform 124, 349; Republican 409; Song 619; Spring and Autumn 669, 679, 730, 732; Sui-Tang 388, 730; Tang 564, 730; Three Kingdoms 52, 161, 534; Warring States 15, 295, 409, 415, 664; Wei-Jin 665, 679; Western Zhou 413; Yuan 396, 731 phonemic transliteration 563, 567, 569, 571 phonetic compounds 316; see also radical phonogram 166, 171–2; see also xingsheng zi phonological: activation 317 –22; adaptation 571–3; approximation 126; coding 317; inclusiveness 43; information 317, 556, 691–3, 700; processing 317, 688, 693; reduction 519; structure 137, 410, 481, 720;

785

General index system 109, 477 – 9, 586, 486 – 96, 724 – 9; unit 228, 514, 519, 523, 689 phonology 1–5, 138 – 9, 155 – 6, 164, 177, 216, 317, 487, 491–2, 511, 571–3, 605 – 6, 688, 691–3, 724 – 9, 768 – 9; Cantonese 35 –7; Hakka 477 – 82; Hong Kong Sign Language 511–19; Mandarin 487, 580 – 8; metrical 666 –7, 678 phrasal boundary 322 pianwen 331–5, 409, 665; see also parallel prose Pianya 534 pictograph 55 – 6, 141, 531–3 pijieyu 396 Pilgrimage to the West 731 Pingju 85 pingsheng 535, 606; see also level tone pinyin 35, 42, 44, 46, 96, 102, 110, 115, 128, 169, 173 –5, 405 – 6, 421, 423 – 4, 442, 484 – 6, 488 – 91, 494 – 6, 499 –502, 510, 539, 546, 549, 551, 555, 558 – 9, 564, 569 –73, 576, 586 – 8, 610, 692 –5, 727 – 9, 760, 766, 768 – 9; see also Hanyu Pinyin Piya 534 PKU Corpus 444 –5, 554b; see also Center for Chinese Linguistics Corpus poem 84, 114, 162 – 6, 194, 334, 465 – 6, 664 – 6, 668 –70, 672 – 82, 725, 731; minimal 668 – 9; new style 669, 731; regularized 669 poetry 65, 105, 155, 164, 193, 308, 458, 465 – 6, 492, 666, 668 – 9, 672, 674 – 6, 678 – 9, 731 popular sayings 295, 344; see also suyu Portable Dictionary of Chinese, A 67 positron emission tomography (PET) 320 post-verbal element 24, 29 –30, 445 Practical Guide for Translators, A 434 pragmatic point 143 – 4 presupposition 184 – 6, 188 – 90, 192 –3 principle: amalgamation 659 – 61; of Auditory Intelligibiity (PAI) 656 –7, 661; of integration 655; process 656; selection 625; of Shaping by Prosody (PSP) 656 –7, 661 processing: Chinese character 315, 420 –36; Chinese information 420; Chinese language 420 –36, 439, 546, 552, 555; Natural Language (NLP) 204, 420, 429, 555; script 315; sentence 315, 318 –20; word- 41, 43, 435, 485, 687, 700 Program Specific Teacher Training (PSTT) 148 prosodic grammar 648, 650 –1, 659 – 60 prosodic morphology 651– 4, 667 – 8 Prosodic Word Compounding 679 prosodic-rhetoric adjustor 672 prosody: five-syllable 672; monosyllable and disyllable 669; poetic 664, 666, 669, 677; seven-syllable 675; three- and four-syllable 670

proverb 64, 67, 69 –70, 83, 177, 295 – 6, 395, 397, 402 –3, 406 psychological reality 187, 321, 693 – 4, 697 punning: meaning 400, 405; sound 400 –2, 405 putonghua (PTH) 19, 96 – 9, 101, 123 – 6, 129 –31, 155 – 6, 162, 164, 166, 169, 173 –5, 421, 424, 434 – 6, 471–5, 477 – 8, 490 –1, 495 – 6, 551, 563, 580, 586, 609, 723, 726 –30, 733 – 4; see also beifanghua; guanhua; guoyu; Mandarin; Mandarin Chinese Qian, Xuantong 416, 489, 500, 731 Qian’ou ci 649 –53, 656 qiaopihua 395, 401 qi-cheng-zhuan-he 334 –5 Qidan 160; see also Khitan Qieyun 2, 491, 493, 535, 549; see also Dictionary of Chinese Rhymes Qu Yuan 156, 664, 676 qusheng 2, 493 – 4, 535, 606; see also departing tone radical 62 –3, 141, 152, 166, 172, 206, 212, 233, 315 –17, 381, 384, 423, 424, 487, 499, 510, 532 –3, 536 –7, 545, 547, 549 –51, 558 – 9, 573, 686, 689 – 90 Ready Guide, The 533; see also Erya recognition: character 316, 686; Chinese character 315, 422 –3; optical character (OCR) 422 –3, 430; proper noun 425 –7; speech 321, 422 –3, 430 reconstruction 2 –3, 155 – 6, 162 – 4, 166, 169 Record of Eternal Words, The 65 Records of the Grand Historian 70, 72, 403, 499; see also Shiji reduplication 4, 75, 223 – 4, 253 – 4, 257, 444, 521, 523, 593, 651; bi-syllabic 224; single-syllable 223 reform and opening-up policy 120 –1, 566, 635, 732 regionalism 127 Republic 636; of China 161, 345, 489, 580, 609, 723, 726, 729 –30; of Singapore 609 Reserve Officers’ Training Corps (ROTC) 122 Revised Mandarin Chinese Dictionary (RMCD) 548 rhetoric 70, 131, 194, 328 –39, 351, 396, 490, 629 –30, 632; see also xiuci Rhetoric in Ancient China 328 Rhetorical Style 336 Rhyme Dictionary, The 535; see also Jiyun Rhyming Glossary 46 Ricci, Matteo 155, 172, 486 –7, 500, 540, 557, 728, 758, 759, 761–5, 769 Role and Reference Grammar 356 Romance of the Three Kingdoms, The 67, 731

786

General index Romanization 35 – 49, 95 – 6, 101, 110, 140, 172 –3, 484 – 9, 549, 728, 758 – 60, 762 –5, 767 –70; Chao’s 38; Church 488; Government 38 – 40, 42 –3, 46; ILE 41; Lau’s 46; LSHK 41–3; Morrisons’s 37, 763; Pinyin 546, 549, 610, 760; Postal 762; Rao’s 43 – 4; Riccie’s 487; Wade-Giles 488, 756, 767 – 8; Yale 39 – 40, 42, 46, 769 Ruggieri, Michele 540, 758 – 9, 762, 769 Rules of Writing 332 –3 rusheng 487, 536, 606, 610 –11, 724; see also entering tone Sacred Edict 764 San Yan Er Pai 82 Sanguo Yanyi see Romance of the Three Kingdoms, The Sanitarium for Sick Plumtrees, A, 335 Sanjidi 734 Scheme for Simplifying Chinese Characters 53 Sea of Words 83, 551; see also Cihai; Origins of Words, The Search for Origins of Common Sayings, The 65 second language (L2) 21, 105 –7, 113, 115 –7, 121–2, 137, 148, 652, 686; see also Chinese as a Second Language; English as a Second Language Second Ode to the Red Cliff 79 Selective Finger Constraint 515 semantic duality 67 – 8 Semantic Minimalism 193 –5 sentence word order 606 sentence: comparative 28, 447, 735; complex 9, 112, 594; compound 594; concessive 9; conditional 9, 594; copular 7 – 8; declarative 14, 283, 527; exclamatory 7, 14, 599 – 600; -final 29 –30, 217, 269, 281– 8, 507, 589, 595, 599 – 601; imperative 7, 286; -initial 255 – 6, 604, 608; interrogative 7, 14, 283, 285 – 6, 526 –7, 591, 608, 737; negative 8, 10, 446; passive 8, 29, 412, 445, 590, 603, 608, 733, 742; pattern 138, 141, 143, 655, 733; simple 7, 65 Shanggu Hanyu 1, 162; see also Old Chinese Shanghai Academy of Social Sciences 124 Shanghai Jiao Tong University 85 Shanghainese 196, 198, 344, 607, 617 – 8 shangsheng 2, 493 – 4, 535, 608; see also rising tone Shen, Congwen 416 Shen, Yue 757 Shenglei 535; see also Dictionary of Initial Consonants shift: diachronic 229; Great Vowel 720; language 21; meaning 208, 707; paradigm 416; role 526; semantic 126, 130, 712 shihui 379 Shiji 388, 403, 499, 682

Shijing 2, 163 –5, 169, 409, 664, 730, 757; see also Book of Odes, The Shishuo Xinyu 81, 413 shizi see full word Shuihuzhuan see Heroes of the Water Margins Shuowen Jiezi 14, 166, 206, 499, 531– 4, 536 –7, 540, 545, 549 –50, 706; see also Analysis and Explanation of Characters, An; Explanatory Dictionary of Chinese Characters, An Shuowen jiezi zhu 538, 706; see also Annotated Explanatory Dictionary of Chinese Characters, The shuyu 64, 395, 402 Shuyuan Zaji 390 Sign Language 381, 505 – 8, 510, 512 –13, 519, 524; American (ASL) 508, 510, 521, 528; Asian 524; British (BSL) 510, 528; Chinese 510, 524; Hong Kong (HKSL) 505 – 8; Japanese 510; OCSD 505 – 6, 508; Shanghai 506; Taiwan 510, 524; Thai 524 Signed Exact English (SEE) 507 Sima, Guang 531 Sima, Qian 329, 388 Simplified-traditional Chinese conversion 435 Simpson, Andrew 90 sinogram 152, 155, 161– 4, 166 –72, 176 –7 Sino-Japanese 566; War 565 Sinophone 116, 125, 485 Sisheng quandian 757 Sishengpu 757 size number 425 slang 123 – 4, 131, 342 –52, 381–2, 392, 539, 646; generational 349; see also liyu Songs of Chu, The 534; see also Chuci Sources of Words 551; see also Ciyuan; Origins of Words, The Speaking Proficiency Guidelines 144 speech: perception 315, 320; production 315, 321–2 standard foot 66, 228, 667, 669 Standard Segmentation Lexicon 548 Standard Stroke Order for Generally Used Characters in Modern Chinese 53 State: Administration of Radio, Film and Television (SARFT) 126, 131; of Lu 328; Qin 51; of Song 64 Strange Stories from a Chinese Studio 767 Stratagems of the Warring States 69 –71, 79, 396 stroke 51–3, 58 – 63, 96, 110, 171–2, 315, 422 –3, 510, 532, 549 –51 Stroop effect 685, 691–3, 700 Student’s Cantonese-English Dictionary 38 Stylistic-Register Grammar 651, 654 –5 Su, Shi 52, 69, 79, 82 subdialect: Bendi 476, 481; Bendihua 480; Jiangxi 476; Jiaying (JYS) 470 –2, 475 –7, 481–2; western Fujian 477

787

General index Sui Code 388 Sun Yat-sen 22, 91, 94, 485 superior temporal gyrus (STG) 320 Supplements and Reifications to the Augmented Dictionary of Dialectal Words 535; see also Xu fangyan buzheng suyu 64, 295, 344 –5, 395; see also popular sayings Swatow see Chaozhou dialect swearword 343 – 4, 346 – 8 Syllabic Dictionary of the Chinese Language, A 557, 763 Syllabus of Graded Words and Characters for Chinese Proficiency 440 symbol: hieroglyphic 727; phallic 170; phonetic 41, 45, 55, 539, 725 – 9, 756 – 69; pictographic 55, 141; semantic 206; sound 485 –502; stenographic 727; writing 54 Symmetry Condition 521 Tagged Gigaword Corpus 553 –5 Taiwanese 125, 487, 606, 619 –20 Tang Code 388 Tang Dictionary of Rhymes, The 535; see also Tangyun Tang, Yue 336 Tangyun 535, 537; see also Tang Dictionary of Rhymes, The Tao, Liang teacher training 104 –7, 122, 147 – 9, 725 Testing Centre of the Office for the National Chinese Language Proficiency Test Committee 58 tetrasyllabic template 651, 671 Three Admonitions 71 Three Character Classic, The 767 Tianjin University 85 tone: 2 – 4, 23, 29, 36, 38, 41, 43, 45 – 6, 48 – 9, 57, 86, 109 –11, 113, 115, 129, 137 – 9, 147 – 8, 155 – 6, 163 – 4, 166 – 8, 173 –7, 179, 185, 201, 223, 235, 320 –2, 347 – 8, 350 –1, 404, 411–12, 421, 423, 471–2, 474, 476 – 82, 486 – 9, 491– 6, 501–2, 535 – 6, 571–3, 580, 582, 584 – 6, 605 – 6, 610 –14, 616 –17, 619 –24, 692, 724 –7, 729, 756 – 69; Dark Departing 45; Dark Flat 41; Dark Rising 45; departing 2, 4, 38, 412; entering 23, 38, 41, 43, 45, 606, 611, 724, 759; falling 156, 175, 606; level 3, 29, 471, 535, 536, 584 –5, 606, 610, 612 –14, 616 –7, 619, 621, 622, 624; light 139, 584; Light Departing 45; Light Entering 46; Light Flat 41, 45; Light Rising 45; low 156; neutral 129, 496, 584, 729; rising 2, 38, 606; see also pingsheng; qusheng; rusheng; shangsheng; yangping; yinping Tongya 534

Tonic Dictionary of the Chinese Language in the Canton Dialect, A 37 training: macrolevel 148; microlevel 147 – 8 Trigault, Nicolas 172, 487 – 8, 500, 728, 759 – 60 Tripitaka 386 Tsinghua University 65, 85 Unicode 420, 423 –5, 547 Universal Grammar (UG) 354, 356, 740, 742, 747 University of Pennsylvania 65, 554 Varo, Francisco 760 verb: action 248, 250, 429, 590, 597, 740; copular 8, 252; ditransitive 249, 251, 364 –5, 590, 595, 611, 620; intransitive 4, 11–12, 130, 249, 253 –5, 358 – 9, 361, 364, 589 – 90, 595; stative 26, 217 – 8, 223 – 4, 248, 250, 252; transitive 6, 8, 250, 265 – 6, 361, 364, 366 –7, 590, 595 Vocabulary of the Canton Dialect, A 37, 557 Wade, Thomas F. 172, 488, 763 –5, 767 Wang, Anshi 333 Wang, Chong 329 Wang, Li 337 Wang, Niansun 534, 538 Wang, Xianzhi 52 Wang, Xizhi 52 Wei, De see Wade, Thomas F. Wei, Sanwei see Williams, Samuel Wells Wei, Tuoma see Wade, Thomas F. Well-Acclaimed Buzzwords (WABs) 228, 234 – 9 Wen Xin Diao Long 331–2; see also Literary Mind and the Carving of Dragon, The Wen, Jiabao 389, 404 Wenhua Dageming 225, 732; see also Cultural Revolution wenli 408, 416 wenyan 1, 335, 490 wenyanwen 408, 409 Wenze 332 Williams, Samuel Wells 540, 557, 560, 760, 763; word: -based pedagogy 139; comparison 28 – 9; complex 221; compound 204, 227, 229 –31, 234, 237 – 9, 653, 674, 686 –7, 691; derived 221, 227; disyllabic 24, 218, 321–2, 459 – 60, 464, 501, 555, 611, 652 – 4, 660, 716; elastic 452 – 67; empty 11, 13; -external inflection 222; -form 430 –1; -formation 4, 227, 237 – 9, 588, 679, 741; full 11, 14; grammatical 11, 221–2, 322, 607; hybrid 384; identification 560, 685, 693 –5, 700 –1; -internal inflection 222; letter-128, 569; loan 15, 24, 40, 43, 120, 128, 224, 228, 232 –7, 239, 476, 563 – 4, 567 – 8, 570 –7, 622 –3, 732,

788

General index 734, 768; measure 112, 245 – 8, 269 –71, 274 –5, 277 – 8, 711–12; monomorphemic 232 – 4, 236; monosyllabic 4, 24, 438, 458 – 9, 465 – 6, 473, 649 –54, 660, 669, 759; multimorphemic 215; multisyllabic; polysyllabic 236; quadrasyllabic 66; segmentation 420, 425; Sense Disambiguation (WSD) 208; simple 220; three-character 521; type 217, 220, 221, 223, 443, 548, 555, 692 WordNet 207 –10, 212; Chinese (CWN) 210 –11; Euro 210; Princeton 210 writing system: alphabetic 51, 487 – 8, 490, 573, 686 –7, 693; character-based 96, 490, 496, 501–2; Chinese 51, 95 –7, 124, 127, 139, 141, 169, 203 – 6, 315, 317, 436, 487, 546, 549, 566, 570 –1, 726 – 9; English 420, 686; ideographic 204; Latinized 487 – 8, 728; logographic 19, 686, 688; Pinyin 115 Wu, Changling 82 Wuche Yunfu 760 Wufang Yuanyin 557, 560 Wujing 379 Wusi Yundong 409; see also May Fourth Movement Wuxia fangyan kao 535; see also Dictionary of Textual Researches on the Wu-Xia Dialect, The Xi Wangmu 72 Xiamen University 554 Xiandai hanyu cidian 132, 395, 400, 456, 539, 548, 551–2, 727, 734; see also Contemporary Chinese Dictionary; Dictionary of Modern Chinese; Modern Chinese Dictionary Xiandai Hanyu Xinxi Cidian 555; see also Grammatical Knowledge-base of Contemporary Chinese Xiang Yu 72, 676; Benji 403 Xiao Erya 534 xiehouti 396 Xiehouyu 406; see also Chinese Riddle-Phrases xiehouyu 64, 395 – 405 Xiehouyu Daquan 406 Xin Wenhua Yundong 416; see also New Culture Movement Xindongfang xuexiao 121; see also New Oriental School xingsheng zi 171, 499; see also phonogram Xinhua Dictionary 420; see also Xinhua Zidian Xinhua Zidian 384, 494, 551–2, 559; see also Xinhua Dictionary Xiru Ermuzi 487, 759 xiuci 328, 335 – 6 Xiucixue Fafan 336 Xiyouji see Pilgrimage to the West

Xizi Qiji 486, 759 Xu fangyan 535; see also Augmented Dictionary of Dialectal Words, The Xu fangyan buzheng 535; see also Supplements and Reifications to the Augmented Dicitonary of Dialectal Words Xu, Shen 57, 166, 206, 533, 545, 706 Xu, Zhimo 416 Xunjinlu see Hsin Ching Lu Xunzi 153, 166, 329 xuzi see empty word Yan, Lianke 124 Yan, Youyi 396 Yan, Zhenqing 52 Yang, Xiong 70, 166, 534 yangping 82, 472, 474, 478 – 80, 493 – 4, 759 Yanshi Jiaxun 413 yanyu 64, 295, 395, 402 yayan 723 Ye Gong (Lord Ye) 397 Yijing 392, 409, 730; see also Book of Changes, The Yimuliaoran Chujie 489 yinping 472, 474, 476, 48 – 80, 482, 493 – 4, 619, 759 yinzhuyu 396 Yiwen zhi 531 Yiyuan Cihuang 396 Youdao Cidian 559 Yue Ren Ge 155 – 6, 164 Yunhui 537 Yunji 535 Yupian 531, 533 Zazuan 396 Zeno 329 zero anaphora 696 –701 Zhai, Lisi see Giles, Herbert Zhang Wei 124 Zhanguoce Chuce Si 396 Zhao, Gu 387 Zhao, Mengfu 52 Zheng, Qing 396 Zhengzhang, Shangfang 155 – 6, 164 Zhengzi tong 536 Zhong, Mingren see Fernandes, Sebastian Zhong, Yao 52 Zhongguo Duyintongyihui see Duyin Tongyi Hui Zhongguo wenzi gaige yanjiu weiyuanhui 490 Zhonghua Da Zidian 537 – 8, 551; see also Comprehensive Chinese Dictionary of Characters; Great Character Dictionary of China, The Zhongyong 408, 730, 765

789

General index Zhongyuan Yinyun 2, 168 – 9 Zhou, Youguang 334 –5 Zhuangzi 69 –71, 78, 295, 298, 313, 767 Zhuangzu 155, 177 Zhuge, Liang 72, 297 Zhuyin Fuhao 485, 489 – 90, 492, 539, 546, 549, 726 – 9 Zhuyin Zimu 725 – 6, 760 zibenwei 139; see also character-based pedagogy

zidian 545, 547 – 8, 550; see also word-based dictionary zihao 425 Zihui 532, 536 –7; see also Comprehensive Dictionary of Chinese Characters zimuci 128, 569; see also letter-word zishu 531 zongheng school 329 Zuozhuan 408, 669; see also Chronicles of Zuo

790