An Essay on Philosophical Method-Oxford University Press, USA (2005)

SAVE OFFLINE

AN ESSAY ON PHILOSOPHICAL METHOD

This page intentionally left blank

AN ESSAY ON

PHILOSOPHICAL METHOD R. G. COLLINGWOOD Revised Edition with The Metaphysics of F. H. Bradley The Correspondence with Gilbert Ryle Method and Metaphysics EDITED WITH AN INTRODUCTION BY JAMES CONNELLY AND GIUSEPPINA D'ORO

CLARENDON PRESS . OXFORD

OXFORD UNIVERSITY PRESS

Great Clarendon Street, Oxford 0x2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Teresa Smith 2005 Introduction and editorial material © James Connelly and Giuseppina D'Oro 2005 The moral rights of the authors have been asserted Database right Oxford University Press (maker) First published 2005 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by Newgen Imaging Systems (P) Ltd., Chennai, India Printed in Great Britain on acid-free paper by Biddies Ltd, King's Lynn, Norfolk

ISBN 0-19-928087-8

978-0-19-928087-2

1 3 5 7 9 1 08 6 4 2

ACKNOWLEDGEMENTS We would like to thank Teresa Smith for granting permission to publish previously unpublished manuscripts; Michael Beaney for his comments on the contents and the Introduction; the reader for the Oxford University Press for valuable comments on the Introduction; Charlotte Vrijen for suggestions on the transcription of the correspondence with Gilbert Ryle; David Boucher, Peter Johnson, Colin Harris, Peter Momtchiloff, and Rex Martin for their help and support over many years.

Editors' Note: In the essays new to this edition Collingwood's original notes are indicated by a dagger(t) after the footnote number; all footnotes in the Essay on Philosophical Method itself are Collingwood's own and appear as they did in the original edition.

This page intentionally left blank

EDITORS' NOTES TO AN ESSAY ON PHILOSOPHICALMETHOD As he notes in his correspondence with Ryle, in the Essay Collingwood does not name those whom he criticizes. These include G. E. Moore (pp. 47-8) in the discussion of precarious margins; Mill (p. 49) under the head of identified coincidents; Croce in chapter III, §2,8, and Gentile in §2,9. In each case he expounded the main principles of their philosophy and went on to criticize them both as fallacious in §10-13. On pp. 71-2 he criticizes W. D. Ross's calculus of goods without naming him; on p. 78 he names him as conceiving the relation between pleasure, knowledge, and virtue as constituting a scale of forms. On p. 79 he mentions H. W. B. Joseph's doubt whether pleasure is a good at all, but does not name him. The brief discussion of the history of the ontological argument between pp. 124-7 summarizes the more extended discussion in Collingwood's 1919 Lectures on the Ontological Proof of the Existence of God, pp. 22-44.

This page intentionally left blank

CONTENTS ABBREVIATIONS EDITORS' INTRODUCTION SELECT BIBLIOGRAPHY

xi xiii cxvii

AN ESSAY ON PHILOSOPHICAL METHOD

cxxiii

The Metaphysics of F. H. Bradley: An Essay on Appearance and Reality' The Correspondence between R. G. Collingwood and Gilbert Ryle Method and Metaphysics INDEX

227 253 327 357

This page intentionally left blank

ABBREVIATIONS A EM EPM IH IN MB MM NL PA RP SM

An Autobiography An Essay on Metaphysics An Essay on Philosophical Method The Idea of History The Idea of Nature 'The Metaphysics of F. H. Bradley: An Essay on Appearance and Reality' 'Method and Metaphysics' The New Leviathan The Principles of Art Religion and Philosophy Speculum Mentis

This page intentionally left blank

EDITORS' INTRODUCTION (A) The Place of An Essay on Philosophical Method in Collingwood's Work Writing An Essay on Philosophical Method

Whatever critics might say, it is clear that Collingwood himself thought highly of his Essay on Philosophical Method. In proposing it to the Clarendon Press he presented it as: A systematic treatment of the methods and peculiarities generally of philosophical thought: it is in fact a new Traite de Methode, attempting to disentangle all the relations concerning philosophy (on its formal side) and science whether mathematical or empirical.1

In An Autobiography he remarked that he wrote the book during a long illness in 1932, and that 'it is my best book in matter; in style, I may call it my only book, for it is the only one I ever had the time to finish as well as I knew how, instead of leaving it in a more or less rough state' (A 117-18). In fact he wrote the book afterthe illness which afflicted him throughout 1931—32. In November 1931 he requested leave of absence for Hilary Term 1932 on grounds of ill health. In January 1932 he wrote to his friend Guido De Ruggiero that 'I never 1

Letter to the Clarendon Press, 9 March 1933. This passage forms the basis of the 'blurb' on the dustwrapper of the book.

xiv

EDITORS' INTRODUCTION

answered your last letter, because I have been ill, rather seriously ill, for a whole year, and unable even to do such simple things as writing letters. It was not so much a definite disease as a kind of breakdown in health, which doctors said was due to doing too much work; anyhow, whatever the cause, 1 have had to stop all kinds of work entirely for a year past. I am now well enough to leave England, and I hope soon to go abroad and travel until the end of April, when I mean to begin work again.'2 The illness took the form of complications arising out of chicken pox suffered in April-June 1931.3 On his return from travelling in Greece and Italy during the summer of 1932 Collingwood rewrote his lectures on moral philosophy. They included a completely new version of the methodological introduction which had been a progressively more important and elaborate feature of the lectures since 1923. He began writing the Essay at the end of November. In Trinity term 1933 he lectured on 'Philosophy, its Nature and Method'—presumably from a draft of the Essay, which at that time was still known as What Philosophy Is, although Collingwood was not satisfied with the title.4 On 2 May he wrote to the press stating that he had 'practically finished the book that the delegates discussed on Friday—if they accepted it!' 2

Letter to Guido de Ruggiero, 4 January 1932. Collingwood delivered no lectures in either Trinity or Michaelmas terms 1931. 4 Letter to Clarendon Press, 9 March 1933. 3

EDITORS' INTRODUCTION

xv

Apparently, when the book was discussed at the Delegates' meeting, Collingwood had not returned to the room after the discussion and was therefore unaware that it had been 'accepted by acclamation'.5 This letter indicated that the title was now to be An Essay on Philosophical Method. The book went to the printer towards the end of June, proofs were corrected in July and the book was published in October 1933. The letter in March offering the Essay to the Clarendon Press contained an abstract of its contents and was sent to W. D. Ross, the Provost of Oriel, who had been asked to comment on it for the press. As the letter makes clear, Collingwood had high hopes for the book. He wrote that: I have gradually been forced to the view that, at present, the progress of philosophy is held up because people do not face these questions of method, and even the best of them are consistently being tripped up by the consequences of this neglect. The book is in a sense pioneer work—nothing like it exists, and although I can't actually hope that it will create the interest that its subject deserves . . . it is by far the most important thing I have written—it is in fact my first genuine, technical, philosophical work. I have written it in a much chaster and less exuberant style than Speculum Mentis, which was an introduction to a philosophy: here the philosophy itself is beginning to take shape, and the style aims at elegance and economy6

For the most part Collingwood's proposal does not deviate from the structure of the published book, but there are a couple of notable differences. The 5

Letter from R. W. Chapman at the Clarendon Press, 3 May 6 1933. Letter to the Clarendon Press, 9 March 1933.

xvi

EDITORS' INTRODUCTION

first is that originally Chapter I was to be an introduction followed by a separate chapter with historical illustrations. In the proposal Collingwood indicated that he had not yet written these chapters; later he merged the two to form the Introduction as published. The other main point of difference concerns Chapters V, 'The Philosophical Judgement' and VI, 'Philosophy as Categorical Thinking'. Originally there was to have been only one chapter entitled 'The Philosophical Judgement' making the claim that the philosophical judgement is universal, yet different from judgements in mathematics and empirical science in being not hypothetical but categorical; this would then lead to a consideration of the ontological proof. Although Collingwood indicates that this chapter is essentially complete in manuscript, it would appear that he had second thoughts and divided it into two. In the published book the title 'The Philosophical Judgement' is reserved for a discussion of affirmation and denial (topics not mentioned in the proposal) followed by a section on the universality of the philosophical judgement. Discussion of the ontological proof and of the claim that philosophy consists of categorical judgements is postponed until the next chapter, now entitled 'Philosophy as Categorical Thinking'. It is clear that Collingwood had always envisaged the book containing a discussion of the ontological proof; the material added at this late stage was that contained in the published Chapter V.

EDITORS' INTRODUCTION

xvii

So much for the immediate circumstances of writing the Essay, but what of its origins in Collingwood's thought and writing? Although the Essay was written afresh, nonetheless it did not emerge from a vacuum; it had a pre-history; it emerged as the outcome of many years of thought on philosophical method in a variety of contexts. This is apparent in Collingwood's short PS in answer to a query by R. W. Chapman of the Clarendon Press: You ask how I do it: but it does itself. The subject 'took charge' about the end of November, and has forced me to write in every odd moment since, neglecting everything else, and welding together all the many thoughts about these questions that have been occupying me at times for many years. One can't stop when once the thing has begun writing itself out.7

Taking this as our cue, let us examine the origins of the Essay a little further. We have already indicated that it began life as the methodological introduction which Collingwood incorporated into his lectures on moral philosophy from 1923. But what of the themes of the Essay, the particular positions, arguments and doctrines which it develops? The Themes of the Essay

Before exploring the origins of the arguments and themes of the Essay, it is important to contextualize them through a brief account of its leading themes 7

Letter to the Clarendon Press, 9 March 1933.

xviii

EDITORS' INTRODUCTION

of conceptual overlap and the scale of forms. Collingwood's argument can be summarized thus: in philosophy the specific instances of concepts tend to overlap so that two or more concepts may be exemplified in the same instance: 'any distinction in philosophy may be a distinction without a difference, (that is), where two philosophical concepts are distinguished Aristotle's formula may hold good, that the two are the same thing but their being is different' (EPM 50). Aristotle's formula explains the essential insight lying behind the doctrine of the overlap of classes. The scale of forms supplements and develops this by suggesting that there is overlap not only in extension but also in intension. The subject matter of philosophy, owing to this overlap of classes, does not admit of classification into mutually exclusive species of a common genus of the sort to be found in the natural sciences. Philosophical concepts are generic; the species of a philosophical genus differ from each other both in degree and in kind; and in a philosophical scale of forms 'the variable is identical with the generic essence itself (EPM 60). Differences of degree between philosophical concepts cannot be measured because they are different both in degree and in kind (EPM 70-1); further, 'if in philosophical thought every difference of kind is also a difference of degree, the specifications of a philosophical concept are bound to form a scale; and in this scale their common essence is bound to be realized differentially in degree as well as differentially in kind' (EPM 77).

EDITORS' INTRODUCTION

xix

The terms on a scale of forms are related both by opposition and by distinction. Each term is distinct from its neighbours, but also opposed to it. This opposition is not absolute: 'if the variable is identical with the generic essence, the zero end forms no part of the scale; for in it the generic essence is altogether absent. The lower end of the scale, therefore, lies not at zero, but at unity, or the minimum realization of the generic essence' (EPM 81). This might seem to imply that all opposition disappears and that we are left only with distincts, but in a philosophical scale of forms there is a fusion of distinction and opposition and therefore the scale does not consist merely of distincts (EPM 81-6). But opposition within a scale of forms does not imply the real existence of either end of the scale; for example, it does not imply the existence of pure wickedness or pure goodness: 'the lowest member of the scale, the minimum realization of the generic essence, is already, so far as it goes, a realization of this essence, and therefore distinct from other realizations; but, as the limiting case, it is an extreme, and therefore an opposite relatively to the rest of the scale' (EPM 82). Opposition appears at any point in the scale: 'the same relation which subsists between the lowest member of the scale and the next above it reappears between any two adjacent forms. Each is good in itself, but bad in relation to the one above, and hence, wherever we stand on the scale, we are at a minimum point in it; and conversely, however far down we go,

xx

EDITORS' INTRODUCTION

there is always the possibility of going lower without reaching absolute zero' (EPM 84). Infinity also has no place in the scale: 'each term in the scale . . . sums up the whole scale to that point. Wherever we stand in the scale, we stand at a culmination. Infinity as well as zero can thus be struck out of the scale, not because we never reach a real embodiment of the generic concept, but because the specific form at which we stand is the generic concept itself, so far as our thought yet conceives it' (EPM 89). This refines the notion of overlap. The classes of a philosophical concept overlap so that 'the higher term possesses not only that kind of goodness which belongs to it in its own right, but also the kind which originally or in itself belonged to its neighbour' (EPM 86-7). Each term has a double relation to its neighbours: 'in comparison with the one below, it is what that professes to be, in comparison with the one above, it professes to be what it is' (EPM 87). Bringing the argument together, this leads to the idea of determinate negation: 'the higher term is a species of the same genus as the lower, but it differs in degree as a more adequate embodiment of the generic essence, as well as in kind as a specifically different embodiment' (EPM 88). The higher term is therefore not only distinct from the lower, but also opposed to the lower. It 'possesses not only its own specific character, but also that which its rival falsely claimed' (EPM 88). Negation is determinate because it both cancels and reaffirms: 'the higher term thus negates the

EDITORS' INTRODUCTION

xxi

lower, and at the same time reaffirms it: negates it as a false embodiment of the generic essence, and reaffirms its content, that specific form of the essence, as part and parcel of itself (EPM 88). Finally, Collingwood further develops the notion of overlap and the nature of the philosophical scale of forms by explaining that the higher of any two adjacent forms overlaps the lower because it includes the positive content of the lower as a constituent element within itself; but it rejects the negative element in the lower, and this negative element is the denial that the generic essence contains anything more than the lower itself provides. It is this denial that constitutes its falsehood. 'The lower overlaps the higher in a different sense: it does not include the higher as part of itself; it adopts part of the positive content of the higher while rejecting another part' (EPM 90). Taken together, 'the overlap consists in this, that the lower is contained in the higher, the higher transcending the lower and adding to it something new, whereas the lower partially coincides with the higher, but differs from it in rejecting this increment' (EPM 91). These are the leading themes of the Essay in brief; they will be amplified and expanded below. Origins of the Essay

In considering the place of the Essay in Collingwood's work it is instructive to consider the development of his attitude towards dialectic,

xxii

EDITORS' INTRODUCTION

degrees of truth and reality, the ontological proof, and one of the central themes of the Essay, the scale of forms. Collingwood began lecturing on Aristotle's De Anima in 1912. Although he observed Aristotle's use of the scale of forms in De Anima (EPM 102), this manner of presentation and analysis did not become ingredient in his own thinking until later. In his first published work, Religion and Philosophy (published in 1916 and written over a three-year period from 1912) Collingwood did not take a dialectical approach. In fact, his procedure there betrayed certain tendencies which later in the Essay he classified as errors. For example, in the earlier work, where he found there to be overlap between religion, theology, and philosophy, he identified the categories and asserted their identity, thereby committing what in the Essay he dubbed 'the fallacy of identified coincidents' (EPM 48). In writing Speculum Mentis (an explicitly dialectical work) a few years later Collingwood reconsidered their relationship and drew an important distinction between implicit and explicit features of experience. In notes inserted into his copy of Religion and Philosophy Collingwood commented on his rejection of realism: This book was written in (and before) 1914 (begun 1912) and represents the high-water mark of my earliest line of thought—dogmatic belief in New Realism in spite of an insight into its difficulties which I think none of my teachers shared. . . . The whole thing represents a point of view

EDITORS' INTRODUCTION xxiii I should entirely repudiate, and its complete failure with the public gives me great satisfaction.8

And in his notes on 'The Devil', published in the same year, he observed that it: represents the breaking point of my earlier philosophical beliefs. It is still realism, sharpened and hardened: The doctrine of God is not thought out: the general position is one of transcendence, and the coarseness and clumsiness of the work reflects the influences of the environment in which 'Prayer' was written. The flagrant superficiality of it, I think, drove me back upon my real convictions, and led to a year of negative criticism (1916) and the building-up of a new dialectical idealism in 1917.9

This confirms that Collingwood understood his own thought at this point to be taking a new turn. But there was still something missing: he had identified his target but not yet fully worked out and articulated his alternative philosophical approach. In 1917 the 'building up of a new dialectical idealism' took the form of a full-length unpublished book entitled Truth and Contradiction. As only the second chapter survived, commentary on the whole book is impossible.10 However, in this chapter he analysed the strengths and weaknesses of the 8

Collingwood had bound the proofs of Religion and Philosophy together with 'The Devil' and wrote these comments on the end paper in about 1918. The volume is in the 9 possession of Teresa Smith. Ibid. 10 Although Collingwood stated in his Autobiography (99) that he had destroyed the manuscript of Truth and Contradiction, nonetheless Chapter 2 survived.

xxiv

EDITORS' INTRODUCTION

coherence theory of truth and rejected the associated notion of degrees of truth and reality. In this endeavour we can see the truth of his later comment that he was building up a new dialectical idealism: but at this stage had still not yet achieved it; Truth and Contradiction was his first serious attempt, but he clearly was not happy with it. In the early 1920s he wrote a 'Sketch of a Logic of Becoming', 'Notes on Hegel's Logic', a draft of opening chapters of a 'Prolegomena to Logic' and Libellus de Generatione, all devoted to working out a dialectical logic of becoming—but he was still not yet satisfied by any of them. And so Collingwood found himself in possession of a serviceable philosophical method which (although he employed it in his lectures and in Speculum Mentis) lacked proper philosophical justification. It took him another ten years to work out the answer to his own satisfaction through lengthy and detailed considerations of the nature of philosophy and the distinctive character of philosophical concepts in a variety of contexts, most especially in philosophy of history and moral philosophy. Given Collingwood's constant insistence that philosophy has to give an account of its own presuppositions, he was never going to be ultimately satisfied with a method justified solely on pragmatic grounds. However, he was obviously confident that he would clear the matter up eventually, and by 1923 he was sufficiently confident in the idea of a scale of forms (which was, with the directly related logic of the overlap of classes, the central theme of

EDITORS' INTRODUCTION

xxv

the Essay) to use both the phrase and the arrangement. The relation between art, religion, science, history, and philosophy is displayed, then, in Speculum Mentis as constituting a scale of forms of experience in which each term in the scale renders explicit what for the previous term was only implicit. Although the term does not appear there, Speculum Mentis is arranged dialectically as an overlapping scale of forms. The forms of experience are articulated phenomenologically, with each achieving more adequately what its predecessor was striving (yet failing) to achieve. Collingwood worked out the idea of an overlapping scale of forms in conjunction with his lectures on moral philosophy, which he had delivered from 1921 and which he rewrote regularly. The courses of lectures from 1921, 1923 (amended in 1926), 1929, 1932, and 1933 clearly show the evolution of his thought concerning the scale of forms. The earliest lectures post-date the dialectical turn but precede the formulation of scale of forms analysis; by 1923 however Collingwood had made the decisive shift. In Part II of that year's lectures on moral philosophy he for the first time offered an account of philosophical method in which he explicitly introduces the idea of a scale of forms and briefly outlines its main features. He attributed the idea to Aristotle, and cited the analysis in De Anima of the different forms of the soul differing in function and capacity as an instance. These lectures were written in September of 1923,

xxvi

EDITORS' INTRODUCTION

as Speculum Mentis was being prepared for the press. His discussion of the idea in the lectures was prompted by the approach taken in the book and this marked the beginning of his working out of the logic of a scale of forms. In Speculum Mentis, Collingwood had developed the distinction between implicit and explicit features of experience and philosophy as experience raised to selfconsciousness, rendering explicit the principles implicitly informing experience, and showing how each form gives rise to its successor dialectically as a scale of forms. However, although under the influence of Croce and Gentile he had moved towards a dialectical manner of thinking and presentation, Collingwood had not yet developed a fully philosophically adequate account of dialectic. In the conclusion to the section on the scale of forms in the 1923 lectures Collingwood stated that: Our series is to be a series of the forms of action; and action is the opposite of passivity. Hence at the bottom of the scale we ought to find pure passivity and at the top, pure activity. Every stage in the scale ought to be more active than the one below it and more passive than the one above it; and the dialectical process leading from one to the next must be based on the lower stage's incomplete self-sufficiency, its dependence on a principle which it does not itself include or possess. For this means an incomplete freedom and therefore an incomplete activity.11 11

'Action': lectures on moral philosophy, 1923, 42. This version of the conclusion, which was an amended version of the original, probably dates from 1926. The references to De Anima are to be found at p. 41.

EDITORS' INTRODUCTION

xxvii

This indicates that the scale of forms was explicitly identified by Collingwood as a series of terms in dialectical relationship. A key feature of a dialectical relationship is the distinction between implicit and explicit, with later terms making implicit what remains only implicit in earlier terms. In Speculum Mentis he wrote that: I may perhaps be permitted here to refer to a book called Religion and Philosophy which I published in 1916, and in which I tried to give a general account of the nature of the religious consciousness, tested and illustrated by detailed analyses of the central doctrines of Christianity. With much of what that book contains I am still in agreement; but there are certain principles which I then overlooked or denied, in the light of which many of its faults can be corrected. The chief of these principles is the distinction between implicit and explicit. I contended throughout that religion, theology, and philosophy were identical, and this I should now not so much withdraw as qualify by pointing out that the 'empirical' (i.e. real but unexplained) difference between them is that theology makes explicit what in religion as such is always implicit, and so with philosophy and theology. This error led me into a too intellectualistic or abstract attitude towards religion, of which many critics rightly accused me. (SM 108n)

Progress on working out the dialectic of implicit and explicit and developing a satisfactory conception of the scale of forms accelerated from 1927-9 onwards, in his writing on moral philosophy, politics, and the philosophy of history. This required consideration of the nature and distinctive character of philosophical concepts. In his essays on 'The Idea of a Philosophy of Something' (1927) and

xxviii

EDITORS' INTRODUCTION

'Outlines of a Philosophy of History' (1928) he argued that there can be a philosophy of something if (and only if) that something is a universal and necessary form of experience; this view was repeated in his pamphlet on The Philosophy of History (1930). In these essays and in his 1929 lectures on moral philosophy he referred to the concepts distinctive of philosophy as transcendentals:

A philosophical concept is universal in the sense that it arises necessarily whenever anybody thinks about a subject . . . the subject itself must be a philosophical, or universal, concept; and that can only mean a concept applicable to everything that exists. It is a familiar idea in philosophy that there are such concepts; in scholastic terminology they are called transcendentals, and you will find, in Spinoza for instance, that ens, res, and unum are given as examples of transcendentals . . . The view which I am putting forward, then, is that the concepts which compose the body of philosophy are transcendentals. 12

In the 1929 lectures Collingwood wrote that 'philosophy deals with conceptions of a particular kind, namely those that in traditional philosophical language are called transcendentals'.13 In these papers and lectures, up to and including the 1932 lectures, Collingwood was working out the themes of the Essay through an exploration of the differentiaeof philosophical thinking, the distinctive nature of philosophical concepts versus non-philosophical or class concepts, and the 12

'The Idea of A Philosophy of Something', in IH 2nd edn., 351—2. I3Lectures on Moral Philosophy, 1929, 6.

EDITORS' INTRODUCTION

xxix

implications of conceptual overlap for philosophical analysis. However, the detailed logic of the overlap of classes had not yet been fully developed and the methodological explorations as a whole had not yet been translated into the vocabulary of An Essay on Philosophical Method. In the Essay Collingwood finally demonstrated to his own satisfaction how a scale of forms, a dialectical progression, was possible. This demonstration hinged on the distinction between empirical or non-philosophical concepts as class concepts and philosophical concepts which, considered in their relations, displayed identity and difference, fusion of opposition and distinction, and fusion of degree and kind. Thus in a philosophical scale of forms the higher terms make explicit what is merely implicit in the lower. The starting point for Collingwood's reflection on method is Socratic. In philosophy we come 'to know better something which in some sense we knew already' (EPM 11). This principle is reasserted throughout the book, and he also remarks that 'every school of philosophical thought has accepted this principle' (EPM 161). There was thus nothing distinctive (and Collingwood did not claim that there was anything distinctive) about the mere claim 'that philosophy brings us to know in a different way things which we already knew in some way' (EPM 161). Indeed, Susan Stebbing, in the article Collingwood subjects to critical scrutiny in Chapter VII, wrote that 'in using the method of

xxx

EDITORS' INTRODUCTION

metaphysical analysis we are not using a method of discovering reasons; we are using a method of discovering what it is precisely which we already in some sense knew.'14 Collingwood's originality lies, then, in his systematic elucidation of an account of philosophical method in which this claim acts as a clue to the nature of philosophical concepts. Philosophical concepts overlap in their instances and out of the basic principle that philosophy renders explicit what is otherwise implicit, Collingwood is able to demonstrate the logic of the scale of forms. A further merit of Collingwood's methodological approach, according to Michael Beaney, is that it avoids the so-called 'paradox of analysis', a problem which he identified several years before the phrase was coined. The paradox is that either the analysandum is the same in meaning as the analysans or it is different. In the first case the analysis is true but trivial; in the second it is interesting and informative but false. From this it would seem to follow that an analysis cannot be both correct and informative. Collingwood's solution lay in his conception of the scale of forms of progressively more adequate and comprehensive knowledge.15 14

L. S. Stebbing, 'The Method of Analysis in Metaphysics', 93. This is the article discussed in Chapter VII; see also, for example, C. D. Broad's Introduction to his Scientific Thought, London, Kegan Paul, 1923. 15 For extended discussion of Collingwood's work in relation to the emerging analytical school see M. Beaney, 'Collingwood's Critique of Analytic Philosophy'.

EDITORS' INTRODUCTION

xxxi

Collingwood's achievement lay therefore in his attempt to think through clearly the requirements of a scale of forms which did critical justice both to his employment of it and to the philosophical work of his forebears. It was the culmination of an engagement with Hegel, Plato, Bradley, Croce, and Gentile on the nature of a dialectical scale and on the nature of the philosophical concept. To take merely one instance, Collingwood did not mention Croce in the Essay, but his presence can nonetheless be felt because one of his concerns was to take account of Croce's criticism of Hegel's notion of dialectic.16 For Croce, philosophical concepts which are related by opposition exhibit dialectical relations; philosophical concepts related only by distinction, however, cannot enter into dialectical relations. Collingwood's view, however, is that this distinction (by which Croce rids himself of much of Hegel's dialectic) ends up by throwing away the possibility of philosophy itself. In his view, because philosophical concepts are related by both opposition and distinction, and because they are both universal and categorical, they are related dialect ically and hence arrange themselves as a scale of forms. Collingwood's conception of dialectic is thus a modification of both Croce's and Hegel's. In the Essay Collingwood reached his goal with a formulation of method which allowed him to 16

Collingwood identified his unnamed interlocutors in his letter to Ryle of 9 May 1935.

xxxii

EDITORS' INTRODUCTION

accept the idea of a dialectical scale of degrees of knowledge. Through this he was enabled to reappraise the work of philosophers (including Plato, Aristotle, Leibniz, Locke, Kant, and Hegel) whom he explicitly cited as following this method. In particular after completing the Essay he directed his method onto Bradley's philosophy, and was able to admit Bradley's doctrine of degrees of truth and reality as valid—at least when understood as constituting a scale of forms. This he did in the essay on Bradley published below, and the results were also made public in his lecture on the 'Nature of Metaphysical Study'.17 In Collingwood's writings from the early 1920s onwards we can trace both the use of scale of forms analysis and its progressive developing theoretical justification. These were only properly brought into mutual balance in the writing of the Essay which was born out of reflection on the scale of forms as previously employed in his philosophical work. Hence we can largely agree with Louis Mink's claim that the Essay is 'throughout, an ex post facto justification of the dialectical system of Speculum Mentis' .l8 The reservation lies in recognition of the fact that in the Essay Collingwood was not trying to produce a philosophical system or to justify the system developed in Speculum Mentis; 17

The second lecture contains a distillation of the argument developed in Collingwood's essay on Bradley's Appearance and Reality and is published in the revised edition of EM. 18 Mink, Mind, History and Dialectic, 73.

EDITORS' INTRODUCTION

xxxiii

he was seeking to elucidate the logic of philosophical thought through an analysis and exploration of the nature of philosophical concepts. This returns us to the point in November 1932 when Collingwood, following his rewriting of his lectures on moral philosophy in the summer of that year, finally embarked on the work which was the culmination of many years of effort, arising both out of substantive philosophizing and also out of determined but elusive efforts to articulate the logic of philosophical method. But it is also important to recognize that, although the foregoing constitutes a very important and distinctive part of the Essay, there were other important sources as well. After all, Collingwood stated that he was 'welding together' the many thoughts about these questions that had been occupying him for many years. One of these sources was previous work in aesthetics which was distilled into the final chapter concerning 'Philosophy as a Branch of Literature'. Collingwood had developed his views on aesthetics in Speculum Mentis and Outlines of a Philosophy of Art (1925), and published papers on 'The Place of Art in Education' (1926), Aesthetic' (1927), and 'Form and Content in Art' (1929) together with a cluster of substantial reviews in the late 1920s and early 1930s. For our present purposes, however, perhaps the other most important source is the series of lectures on the ontological argument written in late 1919. Collingwood first lectured on the ontological proof

xxxiv

EDITORS' INTRODUCTION

in 1920 and again in 1921 and 1922. The lectures were extensive (36,000 words) and consisted (as was his wont) of a theoretical and an historical part. Chapter VI, 'Philosophy as Categorical Thinking', can be seen as a distillation of these lectures.19 Collingwood also discussed the ontological proof in Religion and Philosophy and Speculum Mentis, and a few years later returned to it in An Essay on Metaphysics. As an indication of Collingwood's approach to the ontological argument, consider these passages from the lectures: [T]aken by itself, in abstraction from any context or rather content of positive theological and philosophical conceptions, the ontological proof [is] purely formal and empty. The trouble with it, taken thus in abstraction, is not (as some critics have said) that it proves the existence of an Absolute but not of God still less of the Christian God. The trouble is, rather, that it does not prove anything positive at all. It is the mere skeleton or framework of proof, without any determination towards proving this rather than that. It shows that reality exists as we conceive it: but it does not in itself determine how we shall conceive reality: that is to say it is a pure form which does not dictate its own content. In its special religious bearing, it is of value in insisting that religion is conception, i.e. is not mere imagination, but claims truth: but it does not tell us whether one particular religion is truer than another. Similarly in its philosophical bearing it does not advance any special theory of the nature of reality. Now this purely formal character of the ontological proof must be reconciled with one very conspicuous fact: namely that all its 19

See, e.g., EPM 124-7. Inspection of pages 1, 22, 28, 35, 41-2, 49 of the Lectures on the Ontological Proof of the Existence of God will confirm this.

EDITORS' INTRODUCTION

xxxv

supporters have emphatically asserted that it could only apply to one conception, not to any and every conception. It is the first thing that strikes one on studying the literature, that Anselm, Descartes and his followers, Hegel, and the rest all insist that there is only one idea which is affected by the proof, an idea which they variously define as id quo majus cogitari nequit, ens realissimum, the one substance, the Absolute Conception, or God. All their critics, on the other hand, from Gaunilo to Kant, get at cross-purposes with them by trying to apply it to other ideas—ideas either of things perfect in their kind, not absolutely perfect, or else of ordinary objects of experience.20

In a related passage which moves towards the argument in the Essay he restates the same point by arguing that: Anselm's ontological argument, taken in abstraction from any specific metaphysical or religious doctrine, is empty of all determinate content: it does indeed prove the existence of a reality of some kind, but it is only religion, or metaphysics in so far as metaphysics means the logical development of religion, that it proves the existence of the God of Christian belief. Except to a mind steeped in Christianity, or at any rate in a theistic system of thought, the ontological argument has no definitely theistic bearing at all: the reality whose existence it proves may be an Absolute but it is not God.21

In relation to the position maintained in the Essay, Collingwood's summary of the significance of the argument is important. He suggests that 'the ontological proof is really no less than the conviction that thinking is worth while; a conviction without which thought would never have arrived at any 20

Lectures on the Ontological Proof, 22.

2I

Ibid. 44.

xxxvi

EDITORS' INTRODUCTION

results at all'. 22 In other words, we have to believe that our thought is valid, that is, that valid reasoning exists. We cannot think unless we presuppose that we are thinking truly about a real object and this real object—whatever else it might be— includes our own thought. Further, thought is criteriological, that is, it judges acts of thinking according to standards which in turn are selfreflexively applied to itself. Philosophy is the study of thinking, and thinking is always implicitly criteriological, that is, governed by criteria by which the success or failure of each piece of thinking is judged. Philosophy is self-referential; it is an instance of its subject matter and therefore in its own performance exemplifies the principles and criteria of the performance it takes as its starting point, and appeals in judgement to the same criteria. As he expressed it in The Principles of Art, 'in order to study the nature of thinking it is necessary to ascertain both what persons who think are actually doing and also whether what they are doing is a success or a failure': 23 the philosopher is therefore obliged to judge his or her own performance as a thinker. Philosophy, then, is criteriological in two respects: it judges the success or failure of the performance it philosophizes about and at the same time judges its own success or failure as a piece of philosophical thinking, using in the latter instance 22 23

Lectures on the Ontological Proof, 21. PA 171n; see also EM 107 and 109.

EDITORS' INTRODUCTION

xxxvii

criteria applying both to itself and to its object. This takes the form of a simple argument: philosophical thought concerns standards; because it is thinking about thinking it is self-reflexive; therefore it necessarily concerns its own standards and has to exemplify and live up to them. This appears to be the point of the ontological argument as Collingwood uses it in the Essay; it was also one of Gilbert Ryle's key points of disagreement. Reception of the Essay

Most of the reviews of An Essay on Philosophical Method were positive, as was the reception by Collingwood's friends. In reviewing it for the Oxford Magazine, T M. Knox described it as 'a philosophical classic'; the review in Mind by F. C. S. Schiller was typically quirky but appreciative; that in Philosophy by L. J. Russell was broadly sympathetic—and the editor of Philosophy, Sydney Hooper, sent Collingwood a letter expressing his appreciation of the book. Charles Hartshorne reviewed it favourably in The International Journal of Ethics and it was also reviewed in many other places.24 Collingwood also received letters from Samuel Alexander, H. W. B. Joseph, and H. H. Joachim, all expressing appreciation, and the book also sparked 24

For details see C. Dreisbach, R. G. Collingwood: A Bibliographical Checklist.

xxxviii EDITORS' INTRODUCTION correspondence with his old mentor J. A. Smith. Despite this acclaim, the essay had its critics and Ryle was not the only one; and it is interesting to note that Ryle was not the only one to raise concerns about the use of the ontological argument in the Essay. Father M. C. D'Arcy also did so, although his comments tended in the opposite direction. Whereas Ryle bemoaned what he saw as logical backsliding with Collingwood claiming too much for the constructive powers of philosophy, D'Arcy was more worried that perhaps it established too little. He noted Collingwood's 'remarkable statement that "with Hegel's rejection of subjective idealism, the Ontological Proof took its place once more among the accepted principles of modern philosophy, and it has never again been seriously criticized", and went on to comment that 'I like this remark especially as some time ago I was told by two Cambridge philosophers that the greatest achievement of their school consisted in the final refutation of this very argument! I wonder, too, whether St Anselm would have been satisfied with the kind of object Mr Collingwood wishes to prove.'25 Oddly enough, Collingwood's modified defence of the ontological argument did not worry most reviewers: it did not figure in any way in the reviews in Mind, Philosophy or The International Journal of Ethics.26 25

M. C. D'Arcy, review of EPM. Collingwood had previously written a favourable review of D'Arcy's book The Nature of Belief. 26 By F. C. S. Schiller, L. J. Russell, and C. Hartshorne respectively.

EDITORS' INTRODUCTION

xxxix

Some reviewers, whether friendly or hostile, took issue with Collingwood's use of language. So, for example, D'Arcy regarded the Essay as an interesting and important book whilst C. J. Ducasse, for one, took the opposite view and considered it to be fundamentally mistaken and misconceived;27 but both agreed that Collingwood's language was imprecise, sometimes vague, and insufficiently analytical. This criticism was later echoed by A. J. Ayer in his Philosophy in the Twentieth Century where he remarked that 'An Essay on Philosophical Method is a contribution to belleslettres rather than philosophy. The style is uniformly elegant, the matter mostly obscure.'28 General reviews aside, the two most substantial, critical and detailed contemporary responses to the Essay were the pieces by Ryle29 in 1935 and Ducasse in 1936. Ryle's article is considered below in the context of the correspondence. Ducasse 27

'Mr. Collingwood on philosophical method'. In 1931 Collingwood had given a favourable but critical review of Ducasse's The Philosophy of Art. It is ironic, in the light of Ducasse's criticisms of his use of language, that Collingwood should praise Ducasse for adhering to his determination to avoid 'the vagueness and logical looseness which have been the bane of philosophy'. 28 Ayer, Philosophy in the Twentieth Century, 193. It is interesting to note that R. W. Chapman at the Clarendon Press remarked in a brief note to W. D. Ross that 'I am no judge of its probable soundness or merit as philosophy, but it looks as if it might satisfy the condition of good literature.' Letter to Ross, 10 March 1933, Clarendon Press archives. 29 G. Ryle, 'Mr. Collingwood and the Ontological Argument'.

xl

EDITORS' INTRODUCTION

wrote a substantial critical essay on Collingwood's views of philosophical method. He correctly identified that Collingwood was attempting to elucidate a method for philosophy grounded in the nature of philosophical concepts. For Collingwood, whereas in scientific concepts the species of a genus are mutually exclusive, in philosophy they overlap, and the species of a philosophical genus constitute a scale of forms combining differences of degree with differences in kind. Ducasse's main line of attack was to address head-on what it means to be a species of a genus. He argued that for Collingwood to establish his claim that philosophical concepts are distinguished from scientific concepts by virtue of the fact that species of the genus overlap in philosophy, he would have to show that the relevant species are coordinate species of the genus: 'for the overlapping of species of a genus that are not coordinate is a ubiquitous fact, in no way distinctive of philosophical concepts'.30 Again, 'coordinateness of overlapping species of a genus is the only thing that would have been distinctive, or that was in any need of demonstration, and without it Mr Collingwood's examples are only a waste of time.'31 He suggested that although Collingwood did not explicitly state that they are coordinate species, he tacitly claimed this to be so.32 However, 30

C. J. Ducasse, 'Mr. Collingwood on philosophical 3I method', 98. Ibid. 99. 32 EPM, 35. Elsewhere, however, Collingwood seems to expressly deny that they are coordinate species, see,

EDITORS' INTRODUCTION

xli

he then accused Collingwood of never even attempting to show that the species that he mentions are coordinate and then tried to impale him on the horns of a logical dilemma by stating firmly 'that they are not coordinate automatically follows from the fact that they overlap. But if Mr Collingwood did not claim them to be coordinate, their overlapping, as already pointed out, could not then be claimed to be something distinctive of philosophical concepts.'33 In other words, Ducasse willingly grants conceptual overlap, but denies its significance unless Collingwood were clearly arguing that overlap were overlap of coordinate species—and this he maintained he had not and could not do. His argument is reminiscent of that employed by Ryle.34 For Collingwood, the central claim is that philosophical concepts escape the rules of classification exhibited by empirical or class concepts. In response, then, he would presumably have refused to accept the horns of the dilemma on which Ducasse sought to impale him. For Ducasse, Collingwood evades the issue by refusing (in his imprecision) to see the dilemma; for Collingwood, e.g., 'Method and Metaphysics', below, note 4. The point, presumably, is that Collingwood denies what Ducasse affirms, that is, that the standard logic of classes is applicable to philosophical concepts. Given this, one can see that the two philosophers were bound to talk straight past each other. 33 Ducasse., op. cit., 99. 34 In a footnote Ducasse explicitly states his agreement with Ryle's criticisms of the categorical nature of philosophical propositions.

xlii EDITORS' INTRODUCTION Ducasse (and those like Ryle who share his view) miss the point by falsely assuming that all concepts are empirical classes. As he expressed it in the Essay: [W]hen a concept has a dual significance, philosophical and non-philosophical, in its non-philosophical phase it qualifies a limited part of reality, whereas in its philosophical it leaks or escapes out of these limits and invades the neighbouring regions, tending at last to colour our thought of reality as a whole. As a non-philosophical concept it observes the rules of classification, its instances forming a class separate from other classes; as a philosophical concept it breaks these rules, and the class of its instances overlaps those of its co-ordinate species. (EPM 35)

Philosophy, then, ought to recognize these distinguishing features of philosophical concepts (or of the philosophical phase of concepts). Collingwood was well aware that certain approaches to philosophy, which he had earlier dubbed 'scientific philosophy'35 and now referred to as analytical philosophy, agreed in maintaining that all concepts (philosophical concepts included) are class concepts where the phrase 'class concept' refers to the concepts typical of empirical science. This reduction is of course exactly what Collingwood was trying to escape: Where the generic concept is non-philosophical, as here, the affirmation of one specific form involves the indiscriminate denial of all the rest, for their structure is that of a group of co-ordinate classes where each excludes each and therefore any one excludes all the rest, none more than another. But 35

See SM, 49-50.

EDITORS' INTRODUCTION

xliii

where the generic concept is philosophical, specified in a scale of forms of which the judgement is intended to affirm the highest (which it always is, because every one necessarily conceives the highest specific form known to him as the true form of the generic concept, and so affirms that), its denial of all the inferior forms is summarized in one denial, namely that of the proximate form; since each summarizes the whole scale up to that point, and the denial of that involves the denial of all that it summarizes. (EPM 107-8)

This is a claim about the distinctiveness of philosophy, philosophical reasoning, and philosophical concepts, expressing the view that they do not obey the laws of formal classificatory logic. It would therefore follow that simply invoking those laws against Collingwood could not have persuaded him, as it is their very status and character that is the point at issue. But Ducasse thought that by the looseness of his language at critical points in his argument Collingwood made matters easy for himself. Collingwood was himself well aware that such a charge might be brought against his manner of philosophizing by members of the analytical school who adopted a technical, scientific view of the nature of language. In Speculum Mentis Collingwood had remarked that 'to suppose that one word, in whatever context it appears, ought to mean one thing and no more, argues not an exceptionally high standard of logical accuracy but an exceptional ignorance as to the nature of language' (SM 11). This remark prompted Stebbing in reviewing the book to make

xliv

EDITORS' INTRODUCTION

the riposte that 'presumably, to expect that such important words as true, identical, real should have a clear and unambiguous meaning, is to be a "verbal pedant" who uses "jargon" that is neither English nor "plain". The critic is thus given to understand at the outset that he must not expect precision of statement whether or not there be clearness of thought.' Collingwood's argument in the final chapter of the Essay that philosophy should be written in literary rather than technical scientific language was therefore no new departure: but Collingwood was no nearer persuading Ducasse of its merits than he had earlier persuaded Stebbing. Indeed the claim provoked a comparable outburst from Ducasse: Many others of Mr Collingwood's contentions testify no less eloquently than those already considered how unfailingly fatal in philosophical investigations is a method which, in no matter what eulogistic terms described, essentially consists of a systematic refusal to be precise whenever precision would require of the reader some effort of attention, and would thus interfere with his literary enjoyment of what he reads.36

Ducasse, it is fair to say, was not convinced by the Essay. Presumably Collingwood was aware of his article, but there is no evidence that he considered its criticisms either in unpublished manuscripts, private correspondence, or in print. Ducasse later 36

L. S. Stebbing, review of Speculum Mentis, 566. Ducasse, op. cit., 104.

EDITORS' INTRODUCTION

xlv

reviewed An Essay on Metaphysics, which he much preferred to the earlier work; indeed, he seemed to prefer it in direct proportion to the extent to which, in his view, it ignored or controverted the principles and precepts of its predecessor.37 The Essay and after

Having looked at some of the reviews and critical responses to the An Essay on Philosophical Method we can now examine Collingwood's own use of his fresh-minted method and its relation to his later thought. After publication Collingwood employed the notion of the scale of forms and its associated concepts extensively. For instance, it is to be found in the manuscripts such as 'Notes Towards a Metaphysic', and in lectures such as 'Method and Metaphysics'. It is employed historically in both The Idea of Nature (largely written in 1934 and revised in 1937) and The Idea of History (mostly written early in 1936). It is present in The Principles of Art and forms the backbone of The New Leviathan. In the Essay Collingwood does not distinguish or systematically explore all of the possible applications of the scale of forms. For example, in his work we find consciousness, forms of experience, concepts, and historical development all variously arranged as scales of forms. But there is 37

Ducasse, review of EM.

xlvi

EDITORS' INTRODUCTION

a question: does scale of forms analysis require modification if applied to these different objects? The Essay is largely silent on this, perhaps because it makes the implicit assumption that there are no essential differences worth remarking. Although it could be argued that use of the method in different contexts requires modification,38 it should be noted that neither in the Essay nor later did Collingwood indicate that he saw any need for fundamental modification to the argument of the Essay. In the Essay itself he has no qualms with employing the scale of forms historically; 39 he might be wrong to be so sanguine, but equally, perhaps we should accept that the work is an essay not a treatise, and that Collingwood limited his concerns accordingly. The Principles of Art (1938) was explicit in its references to the Essay and quietly unobtrusive in its employment of the scale of forms and the concept of overlap. One important passage is discussed below; another is where Collingwood refers to the overlap between art and craft and urges his readers to avoid the fallacy of precarious 38

In his introduction to the revised edition of EM, Rex Martin points out that a historical use of the scale of forms requires modification as its particular forms might not be co-present. In a historical process, for example, it may be that a later phase has no point of temporal coincidence with an earlier phase and hence the two cannot overlap in the way in which they can and do in a purely conceptual scale of forms 39 (EM xxxviii—xlv). e.g., see EPM 190—3.

EDITORS' INTRODUCTION

xlvii

margins (PA 22n). This is an important point and one often missed by hostile critics who frequently take Collingwood to be claiming that what is craft is ipso facto not art and vice versa. In 'What Civilization Means', written in 1939-40, the method is employed in an important and explicit rebuttal of historical relativism. In this essay Collingwood analysed the ideals of civilized conduct and demonstrated that their interrelations are those of forms in a scale of forms. There cannot be, therefore, a single unilinear scale in which civilizations differ only in degree: civilizations differ both in degree and in kind. 40 It is worth noting that, generally, Collingwood did not dwell on the issue of method, preferring instead to let the dialectic emerge from the flow of the argument itself. This also follows his general approach to system building: he was systematic, but did not seek to produce a complete system unified in structure and terminology. There remains a curiosity, however, in the fact that the scale of forms was not obviously employed (and was certainly not mentioned) in An Essay on Metaphysics, where, given its immediate predecessors and successors which used or recommended the method, together with its status as a companion volume to the Essay on Method, one would expect it to have been. We examine this curiosity below. 40

For an analysis of this piece (reprinted in the second edition of NL), in relation to the Essay, see J. Connelly, Metaphysics, Method and Politics, ch. 6.

xlviii

EDITORS' INTRODUCTION

Collingwood's last published book during his lifetime was The New Leviathan. This drew extensively on themes emerging directly out of the lectures on moral philosophy, especially his most recent series 'Goodness, Rightness, Utility', written at the turn of the year 1939-40; it incorporates the analysis to be found in manuscripts such as 'What Civilization Means' (1939-40) and draws on the account of mind and consciousness developed in The Principles of Art. It is explicitly arranged as a scale of forms. However, it does make an apparent (but unsignalled) modification to the doctrine by introducing 'the law of primitive survivals'. This states that 'when A is modified into B there survives in any example of B, side by side with the function B which is the modified form of A, an element of A in its primitive or unmodified state' (NL 9.51).41 Philosophical Method and Cosmology

Before embarking on an account of the previously unpublished manuscripts reprinted below, it is 41

David Boucher, The Social and Political Thought of R. G. Collingwood, 96, argues that the presence of an unmodified residue means there can be no complete overlap of forms on a scale. It could be argued in reply that the law of primitive survivals is found in historical, but not in purely conceptual scales of forms; however, this would be to concede the unity of scale of forms analysis. A different reply would be to argue that the law is implicit in the Essay anyway. For example, Collingwood states that the higher of two adjacent forms 'fails to include the lower in its entirety because there is a

EDITORS' INTRODUCTION

xlix

instructive to look at the cosmological notes that Collingwood made immediately after composing the Essay. According to Knox, after completing the Essay, Collingwood remarked that having 'propounded a theory of philosophical method, he was now proceeding to apply it to a problem which had never been solved, namely, to the philosophy of nature'.42 To this end he began a series of notebooks on cosmology—'Notes Towards a Metaphysic'. There were five notebooks in all, covering some 522 pages (about 130,000 words). Much of this work contributed to his lectures on 'Nature and Mind' (and subsequently to The Idea of Nature). One substantial outcome was the sketch of a cosmology which formed the original conclusion (1934) to the lectures which comprise the bulk of The Idea of Nature.43 In the opening remarks of the notebooks, begun in September 1933, Collingwood negative aspect in the lower which is rejected by the higher: the lower, in addition to asserting its own content, denies that the generic essence contains anything more, and this denial constitutes its falsehood' (EPM 90). It is a moot point. 42

T M. Knox, Prefatory Note to The Idea of Nature, v. The conclusions to the lectures on nature and mind are reprinted in The Principles of History. The 1935 conclusion is similar to the conclusion published in The Idea of Nature in that it simply marks the transition from the idea of nature to the idea of history. The 1934 conclusion, by contrast, is a succinctly stated philosophical cosmology tracing the emergence and evolution of matter, life, mind, and God. For details of these manuscripts, see D. Boucher, 'The Principles of History and the Cosmology Conclusion to The Idea of Nature', and 'The Significance of R. G. Collingwood's Principles of History'. 43

1

EDITORS' INTRODUCTION

made the relationship between his current enterprise and his newly published views on philosophical method clear, in terms which were later echoed in 'Method and Metaphysics': The main principle of my method is the Scale of Forms— in this problem we have a genus (I suppose its name is Reality) divided into a scale of species Matter, Life, Mind. According to the method, these should be kinds of reality and also degrees of reality, the highest the most real: each should be distinct from the next and also opposed to the next—and Matter-Mind are just as truly 'nexts' as MatterLife & Life-Mind, Life here forming the 'twilight-term' between them, and the whole triad being only a rough preliminary triangulation for a much more detailed survey—each should sum up the whole scale to that point and from its own point of view be identical with the genus, so that, from the point of view of a lower term (e.g. Matter) that term is the genus (Reality) and the higher terms (Life, Mind) are simply nothing—the names of errors. Whereas from the point of view of a higher term (e.g. Mind) although that term is the genus (Mind = Reality) the same term includes the lower terms (Matter, Life) and, instead of denying these, asserts them as implicates of itself. It does deny them in one sense, but only in the sense that it denies their denials, viz. their claim to be the only or highest term in the scale. Now what about the nisus? i.e. what is the nature of the force which drives us from a lower term to the next higher, or if you like converts that into this? On this subject I have said nothing in the Essay. It is what Spinoza would call an immanent causality: something in the first term which converts it into the second.44 44

'Notes Towards a Metaphysic', 1.

EDITORS' INTRODUCTION li This statement goes beyond the Essay in its assertion that reality constitutes a scale of forms: this is the unifying theme of Collingwood's theoretical cosmology. But is Collingwood asserting that reality in re constitutes a scale of forms or that reality, qua philosophical concept, necessarily articulates itself as a scale of forms? The Essay itself confined itself to the analysis of concepts and remained agnostic on these wider issues. 'The Metaphysics of F. H. Bradley' Over Christmas 1933 Collingwood wrote an extended essay on Bradley with the title 'The Metaphysics of F. H. Bradley: An Essay on Appearance and Reality'. The core argument of this essay also became a central part of the second of his lectures on 'The Nature of Metaphysical Study', delivered in early 1934. Perhaps the essay was itself delivered as a lecture in its own right; certainly Collingwood's discussion of Bradley made an appreciable impact on Donald MacKinnon, who remarked that he 'had the advantage of hearing three courses of lectures by that remarkable philosopher (and indeed his Inaugural lecture as Professor, not to mention two remarkable lectures on Bradley's Appearance and Reality)' .45 The essay provides an account of Bradley's central metaphysical doctrines, their origin in his 45

D. M. MacKinnon, Explorations in Theology, 29.

lii

EDITORS' INTRODUCTION

rejection of the phenomenalism of Mansel and Mill, and proceeds to make the claim that Bradley was enunciating some of the central doctrines of modern realism and was not the idealist he was typically taken to be. From the point of view of method the essay is notable because of its reinterpretation and reappropriation of the doctrine of the degrees of truth and reality, something only now possible for Collingwood because he had cleared up the issue of philosophical method. Whereas he had previously rejected the doctrine of degrees of truth and reality, he was now able to embrace them, suitably reinterpreted along the lines of his conception of the scale of forms. In Truth and Contradiction (1917) Collingwood submitted the theory of truth as coherence to searching criticism, although he admitted that it was probably the best theory available. His main objection to the associated doctrine of degrees of truth and reality was that truth could not be arranged on a linear scale, a point repeated in Speculum Mentis, where he states that 'the terms of a dialectical series are not related to one another in terms of degree, but by the assertion in each term of something which in the previous term was wrongly denied' (SM 208). This statement anticipates the doctrine developed at length in the Essay. Following the conceptual labours of the Essay, however, he was now free to use the notions of degrees of knowledge, truth or reality unencumbered by lingering doubts concerning

EDITORS' INTRODUCTION

liii

their validity, principally because he had reconceptualized what is meant by the term 'degrees' in philosophy. He made this clear in The Principles of Art, where in discussing the imagination, he remarked that 'Kant . . . approached the problem along a new line. Instead of trying to conceive real sensa and imaginary sensa as two coordinate species of the same genus, the conception which, in spite of the empiricists' attempt to revive it, had been once for all refuted by the Cartesians, he conceived the difference between them as a difference of degree' (PA 187). The point is clarified in a footnote: Here and elsewhere I use this word in the traditional philosophical sense, where differences of degree are understood as involving differences of kind; as in Locke's 'three degrees of knowledge', where each 'degree' is at once a fuller realization of the essence of knowledge than the one below (more certain, less liable to error) and also a fresh kind of knowledge. See Essay on Philosophical Method, pp. 54-5, 69-77. (PA 1871n)

This reading of 'degrees' became available only on completion of the train of thought culminating in the Essay. In his essay on Bradley Collingwood made Bradley's position his own by projecting on to him a reading derived from his new understanding of what 'degrees' in philosophy meant. He boldly points out that: Bradley is here using the conception of overlapping classes. A distinction has been made between appearance and reality; Bradley accepts it, but points out that the things so distinguished overlap to a vast and unverifiable extent, whereas the

liv EDITORS' INTRODUCTION subjectivist or phenomenalist philosophers have assumed the absence of such overlap. (MB 29)

Shortly after he suggests that Bradley, in putting forward the notion of degrees of truth and reality: is passing from the overlap of classes to the scale of forms, without which further conception that of an overlap of classes must remain incomplete and unintelligible. For his degrees are not measurable degrees; they are variations in which the variable is the same as the generic essence. And by adopting this notion he is able to accept as his own the doctrines of the phenomenalism which he is criticising, while yet criticising them and rejecting them as inadequate and to that extent untrue. (MB 31)

Collingwood's essay on Bradley should be seen as a re-engagement with, and a fresh endorsement of, Bradley's philosophy in the light of his own concerns and made possible by his work in philosophical method.46 Correspondence with Gilbert Ryle

The correspondence with Ryle illustrates, in the course of a discussion centred on nature and value of the ontological argument, Collingwood's understanding of philosophical concepts and his associated attempt to resist their assimilation to enumerative empirical classifications not subject to overlap. For Collingwood philosophical concepts 46

For more on Collingwood's understanding of Bradley, see J. Connelly, 'Bradley, Collingwood and the "Other Metaphysics"'.

EDITORS' INTRODUCTION

Iv

are distinct from others and necessarily overlap; this Ryle denied: the correspondence brings out sharply some of the leading points at issue between the two approaches to philosophy typified by each man. The correspondence took place in May and early June 1935 and comprised a spirited and detailed philosophical debate centred on three lengthy letters, initiated by Collingwood's reaction to Ryle's article in Mind concerning his rehabilitation of the ontological argument, a rehabilitation which Ryle considered scandalous. The discussion rapidly went beyond the ontological proof per se and became a symposium on the nature of philosophy, what could and could not be proved by philosophical inquiry, the nature of philosophical concepts, the nature of universals, and the distinction between an intensional and an extensional understanding of concepts. The letters make implicit reference to Russell's theory of classes and Collingwood's contribution constitutes an attack on the naturalism implicit in the new empiricist movement. His letters thereby also serve to clarify his disagreement with the presuppositions of empiricism and the rising analytical school in British philosophy. The exchange took place shortly after Collingwood had been appointed to the Waynflete Chair of Metaphysical Philosophy in May 1935 and illustrated some of his developing thinking on the role and nature of metaphysics in the period between the earlier essay on Method and the later Essay on Metaphysics.

Ivi

EDITORS' INTRODUCTION

To set the scene for the debate, it is necessary to give a flavour of Ryle's article, 'Mr Collingwood and the Ontological Argument', published in Mind in 1935. It did not purport to be a review of the Essay as such (indeed the essay had already been reviewed in the journal by the self-consciously iconoclastic pragmatist F. C. S. Schiller). Ryle's focus was entirely centred on Collingwood's understanding of philosophical propositions as both universal and categorical, and with his exhumation of the ontological argument—something which he supposed logically dead and buried. In the article Ryle classified Collingwood 'for what such labels are worth, as an Idealist'.47 He began by frankly stating that he intended to be merely destructive in that he only wanted to show the mistakes which Collingwood had made in propounding his claim that philosophical propositions are categorical and stand in a peculiarly close relation to what exists. He claimed that if Collingwood were right about this then constructive metaphysics was the proper business of philosophy and Hume and Kant were wrong in so far as they maintained that a priori arguments could not establish particular matters of fact. For Ryle, Collingwood's claim that philosophical propositions are categorical meant that they must refer to something which exists, or contain or rest on propositions which do so; and, he continued, 'this must mean, to use language which is not 47

G. Ryle, 'Mr. Collingwood and the Ontological Argument', in G. Ryle, Collected Essays, vol. 2, 101.

EDITORS' INTRODUCTION

Ivii

Mr Collingwood's, that philosophical propositions are or contain or rest on propositions embodying either at least one logically proper name or else at least one definite description which does in fact describe something. In short, every philosophical proposition is or contains or rests on a genuine singular proposition', and he confessed that he could not make head or tail of this. Indeed, he was upset 'to find that apparently after all some judgements may be universal and so (I suppose) expressible in purely general terms and yet categorical in the sense of referring to something actually existing. I fear that the principle of the overlap of classes will be brought in to give us carte blanche to have it both ways when it suits our convenience!'48 Ryle then observed that Collingwood's first argument for this conclusion was that the Ontological Argument is valid, and is indeed presupposed by all other philosophical arguments. He notes Collingwood's recognition of the standard objection to the argument, where he states that 'Anselm's argument that in conceiving a perfect being we are conceiving a subject possessed of all positive predicates, including that of existence, so that to think of this is already to think of it as existing, is an argument open to objection on the logical ground that existence is not a predicate; but the substance of his thought survives all such objections'(EPM 125). Ryle was deeply puzzled by this bland acceptance of what, to him, was a decisive 48

Ibid. 105.

Iviii

EDITORS' INTRODUCTION

objection, and still more puzzled (or perhaps affronted) by Collingwood blithely sailing past the objection with the high-handed comment that nonetheless Anselm's argument survives all such objections. This, maintains Ryle, is precisely the point where recent logicians would dig their heels in and 'say that the argument is an obvious fallacy unless existence is a "predicate"; and that existence is not a "predicate" '. And he then asks the question of how a particular matter of fact can be deduced from a priori or non-empirical premisses.49 Collingwood had in Ryle's view 'rather cavalierly' dismissed Kant's refutation of the argument as a result of false subjectivism and scepticism and then compounded the felony by making the scandalous assertion that 'with Hegel's rejection of subjective idealism, the Ontological Proof took its place once more among the accepted principles of modern philosophy, and it has never again been seriously criticised' (EPM 126).50 This bland assertion moved Ryle to bewildered indignation: To my mind this dictum almost merits tears. One of the biggest advances in logic that have been made since Aristotle, namely Hume's and Kant's discovery that particular matters of fact cannot be the implicates of general propositions, and so cannot be demonstrated from a priori premisses, is written off as a backsliding into an epistemological or psychological mistake, and all's to do again.51 49

G. Ryle, 'Mr. Collingwood and the Ontological Argument', in G. Ryle, Collected Essays, vol. 2, 106. 50 5I Ibid. 106. Ibid. 106.

EDITORS' INTRODUCTION

lix

Collingwood had thus (for Ryle) dismissed the whole of what might broadly be called Russellian logic; this dismissal he regarded as preposterous. But at this point he went on to address Collingwood's denial that the ontological argument made any specific theological claims and quoted Collingwood's conclusion that: What it does prove is that essence involves existence, not always, but in one special case, the case of God in the metaphysical sense: the Deus sive natura of Spinoza, the Good of Plato, the Being of Aristotle: the object of metaphysical thought. But this means the object of philosophical thought in general; for metaphysics, even if it is regarded as only one among the philosophical sciences, is not unique in its objective reference or in its logical structure; all philosophical thought is of the same kind, and every philosophical science partakes of the nature of metaphysics, which is not a separate philosophical science but a special study of the existential aspect of that same subject-matter whose aspect as truth is studied by logic and its aspect as goodness by ethics. Reflection on the history of the Ontological Proof thus offers us a view of philosophy as a form of thought in which essence and existence, however clearly distinguished, are conceived as inseparable. On this view, unlike mathematics or empirical science, philosophy stands committed to maintaining that its subject-matter is no mere hypothesis, but something actually existing. (EPM 127-8)

To which Ryle responded by suggesting that 'surely Russell's theory of descriptions and his consequential analysis of existential propositions as a species of general proposition has been before the philosophical public long enough for this

1x

EDITORS' INTRODUCTION

ontological fallacy to merit immunity from any more exhumations.'52 At this point Ryle turned his attention to Collingwood's claim concerning the normative selfreferentiality of propositions in logic. Collingwood had argued that logicians 'enunciate principles of logic in propositions which themselves exemplify those principles. So their propositions exist. So the essence of the principles of logic involves the existence of examples of them.' 53 This argument Ryle found 'so extraordinary' that (presumably to ensure that his readers should believe the evidence of their own eyes) he felt obliged to quote the relevant passages at length. The passage concerned Collingwood's claim that logic has thought for its subject matter and does not give a merely descriptive account of it; further, Collingwood maintains that neither is logic purely normative, because a purely normative science would expound a norm or ideal of what its subject matter ought to be, but would commit itself to no assertion that this ideal was anywhere realized ... in logic the subject matter is propositions, and the body of the science consists of propositions about propositions . . . the propositions of which logic consists must conform to the rules which logic lays down, so that logic is actually about itself; not about itself exclusively, but at least incidentally about itself. It follows that logic cannot be in substance merely hypothetical . . . logic, by existing, . . . constitutes an actually existing subject matter to itself. Thus . . . when we say 'all universal propositions distribute their subject', we are not only discussing 52

G. Ryle, 'Mr. Collingwood and the Ontological Argument', 53 in G. Ryle, Collected Essays, vol. 2, 109. Ibid. no.

EDITORS' INTRODUCTION

1xi

universal propositions, we are also enunciating a universal proposition; we are producing an actual instance of the thing under discussion, and cannot discuss it without doing so. Consequently no such discussion can be indifferent to the existence of its own subject matter; in other words, the propositions which constitute the body of logic cannot ever be in substance hypothetical. A logician who lays it down that all universal propositions are merely hypothetical is showing a true insight into the nature of science, but he is undermining the very possibility of logic; for his assertion cannot be true consistently with the fact of his asserting it. Similarly with inference. Logic not only discusses, it also contains reasoning; and if a logician could believe that no valid reasoning anywhere existed, he would merely be disbelieving his own logical theory. For logic has to provide not only a theory of its subject matter, but in the same breath, a theory of itself; it is an essential part of its proper task that it should consider not only how other kinds of thought proceed, and on what principles, but how and on what principles logic proceeds. If it had only to consider other kinds of thought, it could afford to deal with its subject matter in a way either merely normative or merely descriptive; but towards itself it can only stand in an attitude that is both at once. It is obliged to produce, as constituent parts of itself, actual instances of thought which realize its own ideal of what thought should be. Logic, therefore, stands committed to the principle of the Ontological Proof. Its subject matter, namely thought, affords an instance of something which cannot be conceived except as actual, something whose essence involves existence. (EPM 129-31)

Ryle's exasperation is palpable in his expostulation that 'I shall find it hard to condense within reasonable limits my objections to this argument'. 54 His first point was that it had nothing to do with 54

Ibid. III.

1xii

EDITORS' INTRODUCTION

the Ontological Argument, and his second that it did not establish the general conclusion that the propositions of logic are not hypothetical. He then proceeded to gnaw at three 'subsidiary bones' he had to pick with Collingwood. Mr Collingwood is at pains to show that a logician who denies the existence of any instances of logically regular thinking must be wrong because he himself is producing an instance of that which he denies to exist. Now this might, per accidens, be so (though a man might, if he troubled, deny the occurrence of genuine singular propositions without producing one, or argue against the occurrence of syllogisms in Disamis by syllogisms in Baroco). But it has no bearing on the point. For (general) hypothetical propositions do not deny the existence of their subjects, they only do not affirm or imply their existence. So a man who maintained that all the propositions of logic are (general) hypotheticals would not be denying the existence of anything. So his exposure as himself a producer of propositions would no more disconcert him than a lecturer on canine diseases would be disconcerted by hearing the bark of a dog.55

This became one of the major themes of the correspondence. The second bone dealt with Collingwood's concern that if logic were purely normative 'it would resemble the exact sciences: it would in fact either be, or be closely related to mathematics' (EPM 129). Ryle clearly did not object to such a conclusion and suggested that Collingwood had forgotten that 'this is precisely what is desired for logic by many logicians, past 55

G. Ryle, 'Mr. Collingwood and the Ontological Argument', in G. Ryle, Collected Essays, vol. 2, 111-12.

EDITORS' INTRODUCTION

1xiii

56

and present'. The third and final bone was that, in Ryle's view, it was not a fact peculiar to logical propositions that they belong to the subject matter which they discuss: The English grammarian writes grammatically about grammar; the educationist lectures instructively about lecturing instructively; the signalling instructor may signal instructions about signalling to his pupils; Horace writes his Ars Poetica in poetry. Have these anything to do with the Ontological Argument? I suppose Mr Collingwood would reply that it is accidental if the principles of grammar or elocution or poetry are conveyed in vehicles which themselves exemplify those principles, but it is necessary that logicians' propositions should instantiate the principles which they themselves propound. But even this seems to me not to be so. For after all one can talk about singular propositions in general propositions, negative propositions in affirmative ones, relational in attributive and attributive in relational propositions. One can reason about the syllogism in non-syllogistic arguments and vice versa. But let us suppose that sometimes logicians have to formulate logical principles or rules in propositions which are instances of them. Even so, the writer or reader might and usually would attend to what the propositions say without noticing that the propositions themselves were cases in point, just as he may study grammar without noticing that the grammarian is keeping the rules.57

This is where, in his original article, Ryle left his criticism; the sequel is published for the first time below and the controversy follows directly the lines set in Ryle's original critical foray. 56

Ibid. 112.

57

Ibid. 112.

Ixiv

EDITORS' INTRODUCTION

Leaving to one side the period detail of this crossing of swords and gnawing of bones, it is instructive to make one or two general comments on the relation between Ryle and Collingwood. Although each would have regarded the other as firmly placed in a different philosophical camp, nonetheless, from the distance of half a century of hindsight, it is possible to see that the two philosophers perhaps shared more in common than they might have acknowledged at the time. In many ways Ryle's own philosophy and philosophical procedures, together with some of his conclusions, do not appear so far removed from Collingwood's own. Although Ryle did not acknowledge the point, their views on mind are strikingly similar in that both tend to resolve mind into activity; 'mind is what it does' was a slogan derived from Gentile, a philosopher studied by both Collingwood and Ryle.58 For both thinkers a body-mind dualism was anathema; equally, their view of philosophy was that it is not deductive nor inductive but both at once; that its method is in some sense dialectical; that philosophy reflects on lower-level experience; that philosophy is concerned with uncovering presuppositions and that it makes clear what in some sense we already know.59 We might further 58

See G. D'Oro, 'Collingwood and Ryle on the Concept of Mind'. 59 See 'Philosophical Arguments', Collected Essays, vol. 2, 196-7 and 'Taking Sides in Philosophy', Collected Essays, vol. 2, 162—3.

EDITORS' INTRODUCTION

1xv

speculate that some of Ryle's work was stimulated by the correspondence, for example, his 1937 paper on 'Taking Sides in Philosophy', which takes as its starting point some of the issues which arose in the correspondence with Collingwood.60 'Method and Metaphysics'

The correspondence between Collingwood and Ryle took place in May and early June 1935. On 19 June Collingwood delivered a paper on 'Method and Metaphysics' to the Jowett Society in Oxford. In this lecture he showed how the scale of forms method of analysis and the associated conception of philosophical method could be applied to a metaphysical question, the general nature of reality. He endeavoured to show that reality is not a class concept but a philosophical concept to which the rules of classification appropriate to the empirical or the exact sciences do not apply. The possibility of metaphysics was thereby defended on the grounds that 60

For further contemporary discussion, see E. E. Harris, 'Mr. Ryle and the Ontological Argument', in which he came to Collingwood's defence from a broadly Hegelian viewpoint. Ryle responded to Harris in his 'Back to the Ontological Argument'. Having had his reply to Ryle's reply turned down by Mind, Harris returned to the fray in 1972 with his paper on 'Collingwood's Treatment of the Ontological Argument and the Categorical Universal'. For Ryle's general estimation of Collingwood see Philosophical Arguments and Autobiographical', 13-14.

1xvi

EDITORS' INTRODUCTION

(because reality is not a class concept) metaphysics as the search for the general nature of reality is not an empty search for a bare abstraction. This lecture was delivered two weeks after the final letter to Ryle and its composition must have overlapped the occasion of the correspondence. It serves to amplify some of the points developed in that correspondence, in particular concerning the nature of philosophical concepts and classes; it is evident from the text that there are several direct references to some of the issues dealt with in the correspondence, together with several oblique references to some of Ryle's comments. 'Method and Metaphysics' is interesting not only as a systematic application of the doctrines of the essay on Method, but also as an illustration of Collingwood's acceptance of some of the main themes of Bradley's philosophy, in particular the doctrine of degrees of truth and reality. It was remarked earlier that the Essay finally allowed Collingwood to accept this doctrine which he had previously found objectionable. As we have seen, Collingwood interpreted Bradley in the light of his conception of the scale of forms, finding him thus interpreted perfectly acceptable. The papers on Bradley and on 'Method and Metaphysics' are testimony to Collingwood's willingness, finally after years of doubt, to approve and employ the idea of degrees of truth and reality. Bradleian echoes also sound in the reference to the 'unearthly ballet of bloodless categories'.

EDITORS' INTRODUCTION

1xvii

In 'Method and Metaphysics' Collingwood addresses reality as a philosophical concept: I have to show, then, that when I speak of the general nature of reality I am using the words general nature in a different sense from that in which we speak of the general nature of men or triangles. These are what I call class-concepts; and thus what I have to show, in order to defend the possibility of metaphysics against the kind of objection I am considering, is that reality is not a class-concept. For the purpose of the present paper, I need not show that this demonstrandum is true; I need only show that it follows from, or is contained in, the argument of my Essay. It is in fact a particular case of a generalization there stated in the second chapter: viz. that no philosophical concept is a class-concept. I have argued in that chapter that whereas concepts like man and triangle consist of characteristics common to all men and all triangles and are for that reason amenable to the ordinary logical rules of definition, classification and division, philosophical concepts, of which reality is one, do not consist of common characteristics and are not amenable to these rules. [MM 3—4]

Later he explains that: I should expect the concept of reality, like other philosophical concepts, to exhibit the structure of a scale of forms, where the different kinds of real things differ both in the kind and in the degree of reality that they possess; a scale nowhere going down to zero, so that there is no such thing as the completely unreal, but one in which every term is relatively unreal by contrast with higher terms, except the highest of all, and I should not expect that we could know, except in a very dim and imperfect way, what the absolutely highest term is. (MM 8)

This seems to imply that there is no distinction in philosophy between the concept of reality and reality itself. Our understanding of reality is necessarily

1xviii

EDITORS' INTRODUCTION

conceptual and therefore reality, as we understand it, is arranged as a scale of forms. Collingwood's idealism, revealed here, was recessive (although implicit) in the Essay but emerges more strongly in some of the manuscripts written immediately after its publication. An Essay on Philosophical Method and An Essay on Metaphysics

We noted earlier the peculiarity that An Essay on Metaphysics neither refers to the earlier Essay nor (so it would seem) employs its method. This would be merely an interesting quirk were it not that various commentators have argued that in the meantime Collingwood fundamentally changed his mind on the relation between philosophy and history and embraced an historical relativism in which history swallowed philosophy whole. For this reason it is worth outlining the reasons for believing that the appearance of radical change has no corresponding substance in reality. This should not be taken to imply that there was no change or development in Collingwood's thinking or that it is free of tension, occasional obscurity and contradiction. Nonetheless the later position is compatible with the earlier position. Let us consider how this assertion might be justified. First, Collingwood neither referred to such a change, nor showed any indication that he believed

EDITORS' INTRODUCTION

1xix

that there had been such a change, and (further) all of his later references to the Essay are positive. This is especially clear in An Autobiography and The Principles of Art. Nonetheless, it should be conceded that it seems remarkable that An Essay on Metaphysics should contain no reference to the earlier Essay. Paradoxically, this absence is made more rather than less remarkable by Collingwood's express desire that the two Essays should be regarded as companions. Indeed, he issued explicit instructions to the Clarendon Press that the two Essays should be regarded as Volume I and Volume II in his series of Philosophical Essays; and he changed the title from An Introduction to Metaphysics to clarify their status as companion volumes. This deepens the puzzle as to why he made no reference in the later work to the earlier, but it was not the action of a man who had consciously repudiated the earlier work. Of course Collingwood could have been radically mistaken about the import of his own doctrines and failed to recognise or acknowledge the contradictions and differences. But this is surely implausible. The absence of internal cross reference is, then, remarkable, but it does not point to the conclusion that there was a radical conversion. Of course it should be recognised that Collingwood used a different vocabulary in the later work, but this was hardly unusual for him: in fact it is precisely what he did throughout his philosophical career, fashioning his

1xx

EDITORS' INTRODUCTION

thinking anew for each philosophical occasion, problem and audience.61 Secondly, absolute presuppositions (to which the metaphysician is allotted the task of discovering and cataloguing) clearly have a close affinity with what Collingwood earlier identified as philosophical concepts or transcendentals. They are nonempirical concepts concerning the general nature of reality; they are both universal (not aggregative or enumerative) and categorical. They are not generalizations derived from experience but yardsticks brought to bear on experience. On this understanding they are aligned with the themes of the Essay on Method and other writing pre-dating it, for example, Faith and Reason published in 1928. Thirdly, it is worth considering another level of connection between the two essays. Both were regarded by Collingwood as examinations of the presuppositions of philosophy itself, and he regarded such an examination as vital to philosophy as a self-reflective discipline. In the Essay on Philosophical Method this emerges most strongly in his criticism of the failure of the analytical school to address the issue of their own presuppositions— see Chapter VII, §2 below. There he argues that the 61

Collingwood instructed the Clarendon Press on the relationship between the two Essays in a letter of 3 June 1939. For extended justification of the claim that there was no radical discontinuity in Collingwood's philosophy, see J. Connelly, Metaphysics, Method and Politics, chs. 1-3, and G. D'Oro, Collingwood and the Metaphysics of Experience, chs. 4 and 6.

EDITORS' INTRODUCTION 1xxi methods and procedures of the analytic school rest on unacknowledged and unanalysed presuppositions which need to be brought to light. Just as An Essay on Philosophical Method is a self-reflective account of philosophy directed upon itself, an examination of the presuppositions of philosophy and its method, so An Essay on Metaphysics is also philosophy of philosophy. Like the earlier Essay it is consciously self-reflective; in some ways it operates on a deeper level of analysis. In the earlier Essay Collingwood assumed the existence of philosophical concepts and took that as his starting point; in the later Essay he sought to prove the existence of this class of propositions (and show that metaphysics was possible) by demonstrating that absolute presuppositions as logical features of thought can neither be regarded as analytically true nor derived empirically. The Essay on Method is in agreement, but did not take its task to be the demonstration of the existence of philosophical concepts. Metaphysics sought to demonstrate that absolute presuppositions do not operate as ordinary concepts or propositions do; that these lie at the root of all our thinking; that some are permanent and necessary, whilst others are temporary and contingent; and that the predicates 'true' or 'false' do not apply where they are taken to refer to normal procedures of empirical verification. Concepts of this type are discussed in chapter II and later in chapter VIII of Method where Collingwood considered the principles of induction. Some of

1xxii

EDITORS' INTRODUCTION

these principles are purely logical principles presupposed by all induction whatever; in addition there is another set of principles (for example, that 'the future will probably resemble the past') which are 'necessary assumptions' in inductive thinking. These principles are not confirmed by the successful conduct of the arguments based on them because 'unless we assumed them, we could never conduct arguments of this kind at all', and what is increased by the success of our inductive inquiries is not the probability of such principles as that the future will resemble the past, but the probability of such hypotheses as that fermentation is due to micro-organisms (EPM 166-7). This account is strikingly similar to the remark in Metaphysics that 'we do not acquire absolute presuppositions by arguing; on the contrary, unless we have them already arguing is impossible to us. Nor can we change them by arguing; unless they remained constant all our arguments would fall to pieces. We cannot confirm ourselves in them by "proving" them; it is proof that depends on them, not they on proof (EM 173). There is a difference in that in the earlier work these principles are characterized as assumptions (even if 'necessary' assumptions) whereas later Collingwood drew a sharper distinction between 'assumptions' and 'presuppositions'. But this does not detract from the point because the key features of the 'absolute presuppositions' to which he later drew attention are also the key features of the 'assumptions' of the earlier Essay.

EDITORS' INTRODUCTION

1xxiii

It is essentially a verbal rather than a substantial difference. It has already been remarked that Collingwood did not generally concern himself with terminological consistency between his writings, and here is a case in point. However, on this point he did develop a sharper set of distinctions, the intermediary stages of which can be seen in 'The Nature of Metaphysical Study' and 'Function of Metaphysics in Civilization'. 62 The account of absolute presuppositions in the later Essay therefore maps closely onto the account of philosophical concepts in the earlier Essay. In Method the emphasis is on concepts and on the distinction between philosophical and empirical classifications; in Metaphysics, by contrast, the emphasis is on presuppositions and the distinction between propositions and absolute presuppositions. Absolute presuppositions, like philosophical concepts or transcendentals, are empirically unverifiable and yet must necessarily be postulated if certain forms of inquiry are to be possible.63 A final set of observations serves to bring out the interesting continuities of doctrine between the two Essays. One of Collingwood's central points in both Essays is that reliance on presuppositions is inescapable and that this also applies to philosophy 62

Both reprinted in the rev. edn. of EM. For discussion, see Rex Martin's Introduction. 63 See G. D'Oro, Collingwood and the Metaphysics of Experience, 65.

1xxiv

EDITORS' INTRODUCTION

itself: hence his critical comments in Method on the failure of the emerging analytical school to provide an account of their own presuppositions. Further, he maintained, as applied to philosophy they give rise to the peculiarities of self-instantiation and self-reference considered in Chapter VI of Method. The issue of philosophy's own presuppositions is addressed head on in Metaphysics: Another perplexity as to method, or perhaps only the same one over again, arises from the recognition that metaphysics investigates presuppositions. Surely, it is argued, a science that investigates presuppositions must avoid making presuppositions in the course of its own work; for how can you detect a presupposition in your neighbour's eye if you have a whole faggot of them in your own? So the idea got about that metaphysics must be a science with no presuppositions whatever, a science spun out of nothing by the thinker's brain. This is the greatest nonsense. If metaphysics is a science at all it is an attempt to think systematically, that is, by answering questions intelligently disposed in order. The answer to any question presupposes whatever the question presupposes. And because all science begins with a question (for a question is logically prior to its own answer) all science begins with a presupposition. Metaphysics, therefore, either has presuppositions or is no science. The attempt at a metaphysics devoid of presuppositions can only result in a metaphysics that is no science, a tangle of confused thoughts whose confusion is taken for a merit. Not only has metaphysics quite definite presuppositions, but every one knows what some of them are, for as metaphysics is a historical science it shares the presuppositions of all history; and every one, nowadays, has some acquaintance with the principles of historical thought. (EM 63)

Metaphysics itself, therefore, rests on presuppositions and needs to acknowledge them; these

EDITORS' INTRODUCTION

Ixxv

presuppositions concern the nature of metaphysics itself and hence are both subject and object in metaphysical thought. In his Lectures on Metaphysics delivered in 1940, Collingwood addressed this point: An inquiry into the nature of metaphysics is a metaphysical inquiry. The inquiry conducted in these lectures consists of a historical (critical) examination of the presuppositions of metaphysicians, with particular regard to their presuppositions as to the nature of metaphysics itself. The inquiry is an effort to arrive at a definition of metaphysics. Yet this definition when finally stated will be found to be one of the presuppositions of the very inquiry itself, a presupposition which has determined the method of procedure. If this seeming circularity is not a legitimate objection, history must be critical, and it must be critical in a special sense. If the object of the historian's study is conceived as something quite separate and independent of the historian himself, there must be absolute standards available at the beginning of the inquiry. If the standards are to arise from the inquiry itself, there must be some sort of unity between historian and object. The historian is, in a sense, making history his own; the criticism is, in a sense, self-criticism. In metaphysics, the method of thought becomes the object of thought, i.e. object and method of study become unified. There is no external, independent subject.64

This set of remarks relates the themes of the lectures and of An Essay on Metaphysics directly to the central concerns of An Essay on Philosophical Method. It also indicates clearly that Collingwood's 64

Notes on 'Lectures on Metaphysics' taken by Louis Hector. A copy of these notes is held at the Collingwood and British Idealism Centre, Cardiff.

1xxvi

EDITORS' INTRODUCTION

use of the term 'history' should be considered carefully and that it does not suggest that philosophical problems are promptly resolvable into history. On the contrary, the term 'history' here refers to the actuality of self-critical and self-referential thought. The purpose of this part of the introduction has been to place An Essay on Philosophical Method within the context of Collingwood's philosophical life and work. Some of the philosophical themes arising out of the argument of the Essay are now explored. (B) The Role and Character of Philosophical Analysis The Argument of the Essay

The main task of the Essay is to show that philosophy has a distinctive method and subject matter. Collingwood outlines the distinctive method and subject matter of philosophy by comparing it with that of the natural and exact sciences. Philosophical method differs from that employed in the exact sciences such as geometry and mathematics because philosophical analysis does not proceed from Cartesian first principles known through the natural light of reason. Philosophy, therefore, is not a deductive science. The method of philosophy also differs from that employed in the natural sciences because philosophers are not in the business of formulating empirical hypotheses.

EDITORS' INTRODUCTION

Ixxvii

Philosophy is therefore not an inductive science that establishes conclusions by means of observation and empirical generalization. But if philosophical inferences are neither deductive, as in exact science, nor inductive, as in empirical science, what is their specific character? Collingwood's answer is that, at one level, philosophy resembles the exact sciences because it is concerned with relations of entailment that are known a priori or through reflection, rather than empirically. Yet it differs from exact science because in philosophy the direction of argument is not from logical ground to consequent, but from the consequent to its logical ground. This has an important implication for what Collingwood takes philosophical inquiry to be capable of establishing. Philosophy, for Collingwood, does not enable us to know something completely new; rather it enables us to become clearer about what we already, albeit implicitly, know. In the Essay Collingwood is particularly interested in unveiling the fundamental principles that underpin the practical and theoretical sciences. He claims that what lies at the basis of the distinction between practical and theoretical sciences is the employment of different basic concepts, the concept of mind and matter respectively. These are concepts that we always, no matter how unselfconsciously, employ when making claims that seek to explain either human conduct rationally or natural events causally. The concepts of mind and matter, Collingwood claims, are philosophical concepts for they can be neither abstracted from experience, as

Ixxviii

EDITORS' INTRODUCTION

empirical hypotheses, nor can they be known a priori by means of intellectual intuition, as Cartesian first principles. Philosophy thus has a distinctive method: the method of philosophy is neither deductive nor inductive but regressive or transcendental. Philosophy also has a distinctive subject matter that differs from that of the deductive and inductive sciences. The inductive sciences are concerned with particulars and their claims never have genuinely universal status. The exact sciences, on the other hand, are concerned not with particulars but with universals, irrespectively of whether they are empirically instantiated. The geometer, for instance, is concerned with the concept of triangles, rather than with particular triangles, and claims in the exact sciences have genuinely universal status. For Collingwood, the subject matter of philosophy differs from that of the exact and empirical sciences alike for, like the former, it makes genuinely universal claims but, like the latter, is not indifferent to the existence of its object in so far as philosophical concepts are necessarily instantiated in our practical and theoretical judgements. The philosophical concept of mind, for instance, is instantiated in the explanation of actions since we cannot explain an occurrence rationally without using the concept of mind. Collingwood explains why philosophy has a distinctive subject matter in the context of his discussion of the ontological proof. He believes the

EDITORS' INTRODUCTION

1xxix

importance of this proof to be that of establishing that philosophical concepts exist necessarily in the sense that they are necessarily instantiated in judgements of a certain logical form. As coherent thinking necessarily involves the employment of the principle of non-contradiction, causal and rational explanation necessarily involve the employment of philosophical concepts. Collingwood's attempt to explicate the subject matter of philosophy in terms of a sympathetic appraisal of the ontological proof has generated an enormous controversy that has unfortunately diverted attention away from his account of philosophical method. But whatever one might think about Collingwood's unorthodox rehabilitation of the ontological proof, it is quite obvious that, if he had used the ontological proof to show that 'mind' and 'matter' must necessarily 'exist' in the metaphysical sense of 'necessary existence', he would have been blatantly contradicting what may be justly regarded as one of the central themes of the Essay: the doctrine of the overlap of classes. According to this doctrine, philosophical concepts differ from empirical ones because they overlap in their instances. The species of a philosophical genus, for Collingwood, capture purely conceptual distinctions, or distinctions in the respective subject matters of the sciences of nature and mind. For this reason the same empirical objects can be described in different ways, depending on whether they are investigated as an

1xxx

EDITORS' INTRODUCTION

expression of thought or a manifestation of natural laws. The rehabilitation of the ontological proof is therefore compatible with Collingwood's account of the overlap of classes only to the extent that it is employed to draw attention to the ontological difference, i.e. to the difference between beings and their Being. Interpreted in this way, at least, it can easily be reconciled with Collingwood's claim that 'reality ... is not a class concept'(MM 6). Finally, in the Essay Collingwood not only defends the claim that philosophy has a method and subject matter that differs from that of the exact and the empirical sciences alike. He also puts forward the view that there is a distinctively philosophical notion of 'progress'. The idea of what it means to make progress in philosophy can be elucidated through Collingwood's discussion of the scale of forms. For Collingwood, the concepts in a philosophical scale are organized in a dialectical progression, i.e. they are ordered not chronologically but logically in such a way that the transition from lower to higher exemplifies a better approximation to the philosophical genus. We will return to these themes later. For the moment suffice it to say that, taken in tandem, Collingwood's account of the overlap of classes and of the scale of forms clearly reveal the influence that Kant and Hegel respectively have exercised on his thought. But as is usually the case with great thinkers, such influences are complexly mediated and have been

EDITORS' INTRODUCTION

1xxxi

subtly appropriated to give rise to a truly original contribution to philosophy. The Critique of Naturalism

The argument of An Essay on Philosophical Method was elaborated in response to two positions of which Collingwood was equally critical: philosophical naturalism and metaphysical realism. As the Essay is best understood in relation to Collingwood's dual critique of naturalism and realism, we will outline Collingwood's philosophical project as it emerges from this dual critique. Let us begin by considering Collingwood's critical engagement with philosophical naturalism. Collingwood's critique of naturalism finds expression in the claim that concepts or universals cannot be reduced to empirical classes and that to reduce all concepts to empirical classes necessarily entails losing sight of the specific subject matter of philosophy, in effect denying the possibility of philosophy itself. Natural scientists, for Collingwood, classify, whereas philosophers do no such thing: philosophers make distinctions. In order to do justice to the nature of philosophical inquiry one must accept that not all distinctions 'cut nature at the joints', that is, that not all concepts are empirical concepts. The attempt to reduce concepts to empirical classes is implicit in early twentieth-century discussions of the relationship between language and logic.

1xxxii

EDITORS' INTRODUCTION

Bertrand Russell, in The Philosophy of Logical Atomism,65 and Gilbert Ryle, in 'Systematically Misleading Expressions', pointed out that all too often, in ordinary statements with a subjectpredicate form, the position of the grammatical subject is occupied by qualitative or attributive expressions, that is, by words that, from a logical rather than grammatical point of view, should be described as predicates. Statements such as 'Unpunctuality is reprehensible' or 'Virtue is its own reward' are examples of everyday expressions in which the position of the grammatical subject is occupied by what is, logically speaking, a predicate. Statements such as these are misleading because they may convey the impression that they are used to denote or refer to entities such as 'virtue' or 'punctuality'. Philosophers should be wary of expressions that invest attributes with nominative status, that is, of expressions that contain bogus (logical) subjects since it is the mismatch between the grammatical and logical form of propositions that encourages the philosophical belief in the existence of transcendent metaphysical entities such as Beauty or Goodness. Russell and Ryle place their analysis of the relationship between the grammatical and the logical form of the proposition in the service of an antimetaphysical programme whose goal was to eliminate from philosophical discourse unwanted Platonic entities (Beauty, Punctuality, 65

See in particular his theory of descriptions.

EDITORS' INTRODUCTION

1xxxiii

Virtue, etc.) whose existence seemed to be entailed by the placing of universal terms in the position of the grammatical subject. The positive assumptions on which this antimetaphysical programme rests are only implicit in Russell's and Ryle's discussion of the relationship between the logical and grammatical form of propositions, but they emerge very clearly when one considers the account of universals that is required in order to avoid the ontological traps in which the metaphysicians of the past fell. To avoid an unnecessary multiplication of entities, universals should be construed as empirical classes and propositions employing universals as summative or enumerative propositions about matters of fact. Russell, and Ryle in particular, did not suggest that we should actually change our ways of speaking. But they certainly believed that the philosopher should be clear that to understand what is meant by (for example) 'redness' is to understand propositions of the kind 'xi is red; X2 is red; X3 is red': to know a concept, for Russell and Ryle, is to know its extension. In the Essay Collingwood claims that concepts can and should be considered from two different points of view: extension and intension. When considered from the point of view of their extension concepts are said to be general because they apply to more than one particular. When considered from the point of view of their intension concepts are said to be generic because they subsume within themselves a plurality of other concepts. To illustrate

1xxxiv

EDITORS' INTRODUCTION

the distinction one might consider the concept 'animal'. The concept is general as it applies to more than one thing; it is generic because it can be subdivided into further, more specific concepts that define objects more concretely, such as the concepts 'vertebrates' and 'invertebrates'. Collingwood's rejection of the proposed identification of universals with empirical classes reflects his view that a purely extensionalist account of universals necessarily loses sight of the general-generic distinction. In particular Collingwood is convinced that the kinds of concepts with which philosophy is concerned cannot be defined extensionally because the distinctions with which philosophers deal are distinctions to which there do not necessarily correspond any empirical differences. To illustrate this point one might consider Collingwood's discussion of the generic concept of the good and its internal subdivision into the concepts of the pleasant, expedient, and right. For Collingwood, the philosophical distinction between the specific concepts of the pleasant, the expedient, and the right does not coincide with the empirical distinction between actions that are pleasant, expedient, and right, since it is in principle quite possible for any particular act to exemplify all three concepts. The suggestion that statements containing universals in the position of the grammatical subject should be rephrased by stating, 'xi is . . .; X2 is . . .; X3 is . . .', fails to take into account that not all conceptual distinctions map

EDITORS' INTRODUCTION

1xxxv

onto empirical classes and consequently that not all universals can be defined extensionally. It is a characteristic of philosophical concepts that they may coincide or overlap in their instances. Since philosophical concepts coincide or overlap in their instances, they capture distinctions to which there may correspond no empirical differences. Collingwood reiterates this point when considering the concepts that make the distinction between the domain of inquiry of natural science and of history possible: the concepts of mind and of matter. Since the philosophical distinction between the concepts of the pleasant, expedient, and right does not map on to the distinction between the empirical class of actions that are pleasant, expedient, and right, but rather illustrates the distinction between the hedonist, utilitarian, and deontological principles, the distinction between the concept of mind and the concept of matter is not a distinction between the particulars that exemplify either concepts but between the respective areas of competence of the natural and social sciences. From Collingwood's perspective the proposed identification of universals with empirical classes not only denies universals ontological or metaphysical status, thereby disposing of the problem of the unnecessary multiplication of entities which deeply concerned both Russell and Ryle; it also carries the dubious implication that all classifications and distinctions are empirical classifications and distinctions, that there are no 'distinctions of

1xxxvi

EDITORS' INTRODUCTION

reason'. More importantly, since distinctions of reason (conceptual distinctions to which there correspond no empirical differences), are the subject matter of philosophical inquiry, Russell and Ryle's reduction of concepts to empirical classes leaves philosophy with no domain of inquiry of its own. If philosophy is to have a subject matter and method of its own, therefore, there must be distinctions of reason, that is, philosophical concepts to which there correspond no empirical differences. It is Collingwood's task in the Essay to show that there are such distinctively philosophical concepts and distinctions. One might object to Collingwood's distinction between philosophical and empirical classifications by claiming that extensional overlap is not sufficient to distinguish between philosophical and empirical concepts. Quine, for instance, has pointed out that empirical or non-philosophical concepts too could overlap in their instances, as it is the case, say, with the empirical concepts 'creatures with a heart' and 'creatures with kidneys'. 66 All creatures with a heart have kidneys; hence the two concepts are extensionally equivalent although they are not intensionally the same. There is, however, an important difference between extensional overlap in the creatures with heart/creatures with kidneys case and the mind/matter case. When we describe a creature as 66

W. V. Quine, 'Two Dogmas of Empiricism', in From a Logical Point of View, 21.

EDITORS' INTRODUCTION

1xxxvii

having a heart (but not kidneys) we provide an incomplete empirical description. The semantic gap between the two concepts, therefore, can be closed by showing that having kidneys is a necessary ontological condition of having a healthy, functioning heart. The distinction between the concept of mind and the concept of matter, on the other hand, is not a result of a failure to include some empirical attributes in the description of a thing: when we describe a creature as rational or minded we do not provide an incomplete empirical description as if, say, being minded or rational were an empirical characteristic of man alongside the fact that man is also a featherless biped. We cannot mix and match concepts which belong to anthropology, understood as an empirical science of man, and to history, understood as a normative science of mind, in the way in which we can mix and match empirical concepts. When we describe a being as rational we do not provide a partial or incomplete empirical description of a biological species so that by adding the adjective 'rational' to 'human', to produce the expression 'rational human being', we could close the semantic gap between the concept of a rational and the concept of a human being. To predicate humanity and rationality of x is not the same thing as predicating possession of a heart and kidneys: the concept of mind and the concept of matter have a different meaning not because they capture selectively certain empirical characteristics of a thing (in the way in which the concept of square and red do), but

1xxxviii

EDITORS' INTRODUCTION

because they determine what kind of thing one is talking about, actions or events, mind or matter. Philosophical distinctions are not empirical classifications but distinctions of reason that determine a priori the subject matter or regional ontologies of the first-order sciences. The conclusion to be extracted from Collingwood's discussion of the overlap of classes, therefore, is that there are some concepts—philosophical concepts—which are not empirical but a priori, and that this explains why there is a set of problems that cannot simply be resolved or eclipsed by the progress of natural science, problems which are and continue to be the peculiar province of philosophy. The Critique of Realism

Having considered the Essay in relation to the claims of philosophical naturalism we shall now consider it as a reply to, or an extended argument against, realism. Collingwood's main dissatisfaction with realism is that it offers no way of accounting for how forms of knowledge that rest on mutually exclusive conceptions of reality can coexist. The conception of reality with which natural scientists operate, reality as nature or system of natural laws, for instance, conflicts with the conception of reality that is presupposed by moral discourse, folk psychology, and history. On realist assumptions there is no way of reconciling the claims of theoretical reason with those of practical

EDITORS' INTRODUCTION

1xxxix

reason. Realists must assume that only one of the forms of experience is true of things-in-themselves, as two or more contradictory claims cannot be true of one and the same object. For Collingwood, if the claims of natural science are to be reconciled with those of practical reason, the concepts which lie at the basis of their respective domains of inquiry must be construed not as ontological or metaphysical truths but as epistemic postulates or presuppositions that guide our reasoning qua natural scientists and qua agents. 'Mind' and 'matter', in other words, are not to be construed in a Cartesian fashion, as different kinds of substances, but in a Kantian fashion, as presuppositions or guiding principles of theoretical and practical reason. Collingwood's alternative to metaphysical realism takes the form of a weak antirealism. Collingwood's antirealism is weak because it does not deny the existence of mind-independent objects, by identifying them with ideas, as Berkeley did. What Collingwood denies is only that the fundamental concepts through which reality is experienced are metaphysical entities, like Platonic forms or Cartesian substances. Collingwood's antirealism, like Kant's transcendental idealism, was developed primarily in response to the antinomical nature of thought rather than, like Berkeley's immaterialism, as an answer to the problem of scepticism concerning the external world. What drives Collingwood's antirealism is the (Kantian) view that the incompatible assumptions governing

xc

EDITORS' INTRODUCTION

theoretical and practical reason cannot be reconciled on realist assumptions, rather than the (Berkeleian) conviction that common-sense beliefs in the existence of the objects of perception are best vindicated on immaterialist premisses. Although Collingwood's antirealism more closely resembles Kant's epistemic idealism rather than Berkeley's ontological idealism or immaterialism, there is no trace of the Kantian Ding-an-sich in Collingwood's weak antirealism. Correspondingly, there is no suggestion that reality might be known as it is 'in itself, independently of the concepts and categories through which we finite rational beings know it. Collingwood's position on this issue is not so much that reality as it is in itself is not known, but that it simply does not make sense to speak of knowing things in themselves. As all knowledge is epistemically conditioned, the expression 'epistemically unconditioned knowledge' or knowledge of things-in-themselves is an oxymoron. Consequently the kind of sceptical doubts which may arise as a result of holding that things-in-themselves are unknowable for us is misguided in principle. The reasons why the expression 'epistemically unconditioned knowledge' involves an oxymoron may be best understood if we take a step back and remind ourselves of Collingwood's general—generic distinction as applied to concepts. As we saw, concepts are general in their relation to particulars (they apply to more than one thing) and generic in

EDITORS' INTRODUCTION

xci

relation to other concepts (they subsume within themselves a number of more concrete concepts or marks). For Collingwood a concept can be infinite with respect to its extension (it can have an infinite number of particulars falling under it), but it cannot be infinite with respect to its intension (it cannot be subdivided into an infinite number of marks or more concrete concepts). The inability of concepts to be infinite with respect to their intension explains why, for Collingwood, scepticism of an epistemological nature, that is, scepticism arising from the consideration that reality as it is in itself may not resemble reality as conceptually mediated, is misguided. For Collingwood, if we climb up the conceptual ladder and progressively ascend from species to genus (for instance from 'cow' to 'mammal', from 'mammal' to 'vertebrate', from 'vertebrate' to 'animal', from 'animal' to 'living organism', etc.), we move from more concrete concepts to concepts that are less specific or more abstract. If this process were carried through to its logical conclusion, one would ultimately reach a concept that contains within itself all marks, a concept so abstract that it would be unable to capture any aspect of reality. Conceptually unmediated knowledge or knowledge of things-in-themselves, for Collingwood, is the kind of 'knowledge' one would get at the top of the conceptual ladder. Yet, since the concept at the summit of the conceptual ladder is the most abstract concept, it is also unable to single out any subject matter. In An Essay on

xcii

EDITORS' INTRODUCTION

Metaphysics, Collingwood claims that the concept at the summit of the conceptual ladder is the concept of pure Being and that metaphysics, traditionally understood as an investigation into the ultimate structure of reality (as ontology or the study of pure being), is a science without a subject matter. Reality, for Collingwood, can only be known as a system of natural laws, rational process, and object of aesthetic contemplation, not in itself. The kind of scepticism that arises from the consideration that epistemically conditioned knowledge is defective or lacking in some respect, rests on the erroneous assumption that there is such a thing as epistemically unconditioned knowledge of objects, that epistemically unconditioned knowledge would be a kind of knowledge at all. Whereas Collingwood refuses to naturalize the subject matter of philosophical inquiry, his critique of realism shows that he is equally reluctant to identify the domain of inquiry of philosophy with that of traditional metaphysics. Although he often speaks as if he were asserting the metaphysical existence of philosophical concepts (such as mind and matter) by alleging that philosophical propositions have the form 'mind exists' and 'matter exists', his defence of the possibility of philosophical inquiry does not entail subscribing to the existence of a transcendent realm of real entities. 'Mind' and 'matter' exist not as Platonic forms but as categories embedded in the judgements made by historians and natural

EDITORS' INTRODUCTION

xciii

scientists respectively. Such categories determine, a priori, the kind of reality (natural or historical) with which the practitioners of the first-order sciences are concerned. The subject matter of philosophical inquiry, therefore, is a set of a priori concepts or categories and the possibility of philosophical inquiry rests not on an ontological but on an epistemological notion of transcendence: mind and matter are a priori epistemic forms that cannot be empirically justified but are the conditions of the possibility of knowledge or experience. Philosophy therefore has a subject matter and method that differs both from that of natural science and metaphysics: unlike natural science philosophy is concerned not with empirical classes and classifications but with a priori concepts that lie at the basis of distinctions of reason; unlike metaphysics philosophy is concerned with these a priori concepts as norms governing different forms of inquiry, rather than as transcendent entities. The Nature of Philosophical Problems

Collingwood's dual rejection of naturalism and realism entails a conception of philosophy as a form of categorial analysis that seeks to bring to light the fundamental concepts that govern our experience of reality. This conception of philosophy as categorial analysis bears some resemblance to the tradition of linguistic philosophy, but also differs from it in crucial ways. Like linguistic

xciv

EDITORS' INTRODUCTION

philosophers Collingwood defends a conception of philosophy as conceptual analysis, rather than as an inquiry into the nature of things, but unlike ordinary language philosophers (such as J. L. Austin) he is not concerned with mere linguistic classifications. Whereas the way in which we speak about reality provides an important clue to the way in which we think about it, there is more to philosophical analysis than merely cataloguing linguistic usage. Philosophy, as a form of categorial analysis, takes its cue from the linguistic articulation of reality but also goes beyond this: the philosopher must reach deeper and move from the way in which we speak to the way in which we think, from words to concepts. Hence, although Collingwood rejects the identification of the subject matter of philosophical inquiry with that of traditional metaphysics, he is no linguistic philosopher in the narrower sense. The concepts with which philosophical analysis is concerned (for example, mind and matter) mark the boundaries between domains of inquiry and determine the sort of inferences to be employed in the first-order sciences. Philosophical analysis, therefore, teases out concepts that are implicit in different forms of reasoning rather than providing a mere empirical survey of linguistic usage. The kind of categorial analysis with which Collingwood is concerned brings out the deep structure of thought and reveals that the conflict between theoretical and practical reason has its roots in mutually exclusive assumptions about

EDITORS' INTRODUCTION

xcv

the nature of reality. This feature of categorial analysis entails a particular conception of the nature of philosophical problems and of the goal of philosophical reflection. For Collingwood, the fundamental problems of philosophy are neither historically transitory nor are they artificial problems that are generated by inappropriate theoretical reflection. The problem of mind-body dualism, for instance, could not be dismissed either by showing that it is relative to a prescientific worldview prepared to include immaterial entities or souls within its ontology, or by arguing that such a problem arises from a misunderstanding of the nature of language. Philosophical problems are neither historically contingent nor artificially created by the wrong kind of theoretical reflection: they have their roots in incompatible discourses and they surface when radically different conceptions of reality come into contact and conflict with one another. Collingwood's belief that philosophical problems are perennial rather than historically contingent, and that they have their roots in mutually exclusive conceptions of reality rather than being generated by theoretical reflection, has important repercussions for his conception of the task of philosophy. First, since philosophical problems are perennial problems, philosophical analysis is not meant to herald the beginning of a postphilosophical era that will no longer be dogged by questions such as those concerning the relations

xcvi

EDITORS' INTRODUCTION

between mind and body, freedom and determinism. The task of philosophical analysis is to clarify the presuppositions on which our claims to knowledge rest, in any particular area of experience, and to disentangle the conceptual knots that arise from the application of presuppositions and modes of inference beyond their appropriate domain of inquiry. Secondly, because philosophical problems are not generated by inappropriate theoretical reflection, categorial analysis does not entail a 'therapeutic' conception of philosophy according to which the main task of the philosopher is to diagnose the theoretical errors into which philosophers are drawn when they start reflecting upon our way of speaking about the world. To the extent that philosophical analysis can be employed to expose the extension of principles and methods of inference beyond their appropriate domain of inquiry, there may indeed be a therapeutic dimension to Collingwood's conception of philosophical analysis. But nowhere does Collingwood suggest that the problems with which philosophers have struggled from time immemorial are primarily the result of philosophical mistakes.67 Problems such as those of the relationship between mind and matter or freedom and determinism are ultimately 67

This observation may naturally suggest a comparison with, and possibly a contrast to, the philosophy of the later Wittgenstein. These remarks are directed less against the position of Wittgenstein than against certain oversimplistic accounts of his later philosophy.

EDITORS' INTRODUCTION

xcvii

grounded in our ability to distinguish between agents and patients, actions and events, and will remain with us as long as such distinctions are part and parcel of our conceptual framework. The Autonomy of Philosophy

The Essay is one of the most sustained attempts within the philosophical tradition to articulate a conception of philosophy as first science and to defend the thesis of the autonomy of philosophy. Its main theme, the nature and method of philosophical inquiry, strikes at the very heart of much contemporary debate, since it claims a foundational role for philosophy and argues against a conception of the philosopher as the mere under-labourer of science. Collingwood's concern for the autonomy of philosophy appears particularly prescient in the light of subsequent attempts to deny the existence of any sharp boundaries between philosophy and natural science, notably Quine's repudiation of the analytic-synthetic distinction as the last obstacle to the complete naturalization of philosophy. At the time when Collingwood was writing the Essay, the logical positivists' critique of metaphysics was in the air, even though Ayer's Language, Truth and Logic appeared three years after the publication of EPM. But whereas Collingwood was aware of the logical positivists' critique of metaphysics, he could not have anticipated the kind of antiphilosophical climate represented by Quine's attempt to take

xcviii

EDITORS' INTRODUCTION

empiricism to its logical conclusions by exposing the analytic-synthetic distinction as an unempirical dogma of empiricism.68 The kind of consistent empiricism advocated by Quine, unlike that of the classical empiricists and logical positivists, implies that there can be no meaningful statements that cannot be empirically verified. It denies, in one fell swoop, analytic and a priori knowledge alike, thereby ruling out even the possibility that philosophy might, as a last resort, seek refuge amongst the exact sciences. Collingwood's defence of the autonomy of philosophy may be clarified by examining it in relation to three often-cited distinctions: the distinction between the a priori and the a posteriori, the analytic and the synthetic, and the necessary and contingent. The a priori-a posteriori distinction is an epistemic distinction which divides knowledge with respect to its mode of justification. The analytic-synthetic distinction is a semantic distinction that divides knowledge claims into trivially true (or tautologous) and informative. The necessary-contingent distinction is a metaphysical distinction that divides knowledge claims into those of a necessary and contingent nature. Philosophical knowledge is a priori knowledge, knowledge which cannot be justified empirically. Collingwood makes this point in the 1929 introduction to the 'Lectures on Moral Philosophy'69 where he claims that the method of verification 68 69

Quine, ibid. 20-46. Collingwood, Lectures on Moral Philosophy, 1929.

EDITORS' INTRODUCTION

xcix

available to empirical scientists is not available to the philosopher because whereas empirical concepts are 'not only different conceptions but [also] conceptions of different things',70 philosophical concepts allow in principle for full extensional equivalence. Whereas a geologist can justify the concepts of sedimentary and crystalline rocks by pointing to their respective instances, the philosopher of history cannot justify the concept of action in the same way since anything which falls under the concept of 'mind' could also fall under that of 'matter' and be described as a bodily movement. The justification of philosophical concepts must be sought not empirically but a priori. What, however, is meant by this? Kant defined a priori knowledge as knowledge that is absolutely independent of experience and contrasted it with knowledge that is merely independent of 'this' or 'that' experience.71 For Kant we can, for instance, know that if the foundations of a house are undermined, the house will fall without actually witnessing 'this' event, but such knowledge is not absolutely a priori because it is based on an anticipation of experience based on previous observations. By contrast, the hypothetical 'if x then y', where the slot occupied by x and y remains empty, is absolutely a priori since it could not be known empirically. Kant's point, when claiming that a priori knowledge is absolutely independent of experience, is that the logical form of the judgement could not be known empirically, 70

Ibid. 4.

7I

I. Kant, Critique of Pure Reason, B3.

c

EDITORS' INTRODUCTION

whereas the actual content of the claim is dependent on empirical observation. Philosophical knowledge, for Collingwood, is a priori in the sense outlined above because philosophical concepts are embedded in propositions whose logical form must be presupposed in order for certain kinds of claims to be possible. The concept of 'mind', for instance, is embedded in propositions that establish a rational connection between the explanans and the explanandum, and such a rational connection must be presupposed if mentalistic explanations are to be possible. The philosophical concepts of 'mind' and 'matter' are embedded respectively in the logical form of mentalistic and causal explanations, and the latter are (in a vocabulary that Collingwood will develop later) 'absolute presuppositions' of history and natural science. Since the concepts with which philosophy deals are a priori, the justification of philosophical concepts is an internalist matter. In contemporary epistemology the distinction between internalism and externalism is drawn by saying that whereas for internalists the factors which are required for a belief to be justified must be cognitively accessible to a believer, for externalists this condition does not hold: what is of paramount importance, for externalists, is the causal history of a belief. If the debate between internalism and externalism in epistemology is characterized in terms of the emphasis that these positions place respectively on the notion of justification and that of

EDITORS' INTRODUCTION

ci

causation, Collingwood may be more appropriately described as an internalist since he would not regard the causal history of a belief as providing any evidence (of a characteristically philosophical kind) that a belief is justified.72 The justification of philosophical concepts, for Collingwood, is secured by showing how such concepts relate to certain forms of experience and how they underpin and make them possible, not by ascertaining how they are acquired. Collingwood would probably agree with many present-day internalists that if externalism were correct, epistemology would no longer be a branch of philosophy. It would be indistinguishable from cognitive psychology and hence would have to be reclassified as a branch of natural science. For Collingwood, as indeed for present-day internalists, to drop the requirement that the justificatory concepts/factors be cognitively accessible would be tantamount to abandoning the view that philosophy is an a priori activity to be carried out by reflecting on the nature of experience. If we take the taxonomy employed in contemporary debates in epistemology as a point of reference, Collingwood would best be described as a weak rather than as a strong internalist. Strong internalists claim that a believer must be aware of the factors required for a belief to be justified. By 72

Collingwood does not deny a role for justification of an externalist variety, but believes that it cannot be applied to philosophical concepts precisely because these concepts are not of a perceptual kind.

cii

EDITORS' INTRODUCTION

contrast, weak internalists require only that a believer be capable of becoming aware of them. Collingwood's position in the Essay represents a variety of weak internalism because he does not require (say) social and natural scientists to be conscious of what a priori principles regulate their respective forms of inquiry. For Collingwood the justificatory principles/concepts with which philosophical inquiry is concerned, principles such as those of mind and matter, are implicit in the practices of first-order scientists but first-order scientists may not be aware of them. It is the task of philosophical inquiry to render such concepts explicit. It is clear from this discussion that Collingwood would be opposed to the project of naturalizing epistemology, that is, the attempt to transform epistemology from a normative science concerned with the question of the justification of belief, to a branch of cognitive psychology, or a descriptive science concerned with mapping the relation between sensory inputs and epistemic outputs. Although Collingwood would not have encountered the expression 'naturalized epistemology', he would undoubtedly have been acquainted with the fundamental ideas underpinning it.73 While the project of normative epistemology had been clearly laid out by 73

The expression was coined by Quine in his 1969 essay, 'Epistemology Naturalized' in Ontological Relativity and Other Essays. It should be noted that throughout his philosophical career, Collingwood was quite clear in his view concerning the scope and limits of psychology as a science of mind. His primary objection was that psychology treats thought from

EDITORS' INTRODUCTION

ciii

Descartes in the Meditations, the contrast between epistemology conceived as a normative and as a descriptive science had been made crystal clear by Kant who distinguished between the question concerning the provenance of ideas/concepts from that of their validity and claimed his transcendental philosophy to be concerned with the latter, the quaestio quid iuris. Like Descartes and Kant Collingwood conceived of philosophy as a normative activity to be pursued a priori, through a reflection on the nature of experience and in this respect he shared their conception of philosophy as first science. He thought that philosophy should reflect on the fundamental principles that govern first-order disciplines. In particular Collingwood concurred with Kant in holding that philosophy is concerned with vindicating certain a priori concepts and that such a vindication properly consists in assessing the role they play in any attempt to explain how experience is possible. For Collingwood the justification of philosophical concepts must show that such concepts are necessarily implicated in our ways of thinking and speaking about the world. For instance, the philosophical justification of the concept of mind without, as something external to the psychologist, as a natural object. He objects that psychology is descriptive and not criteriological or normative, in other words, that it ignores the selfcritical character of thought. For Collingwood's views on the relation between philosophy and psychology, see especially RP, 39-42; EM, chs. IX-XIII, esp. 107-8 and 115; and A, 92-5. See also J. Connelly and A. Costall, 'R. G. Collingwood and the Idea of a Historical Psychology'.

civ

EDITORS' INTRODUCTION

entails showing that the notion of the mental is inextricably entwined in our attempts to understand human behaviour by appealing to reasons. Let us now turn to consider how Collingwood's conception of philosophical analysis can be located with respect to the analytic-synthetic distinction. Collingwood's account of philosophical analysis does justice to a widely held view about the nature of philosophical inquiry: the view that although philosophical inquiry is a priori in character, it is not a trivial pursuit or a form of mere mental hygiene. This point may be best understood by considering Collingwood's defence of the possibility of philosophical knowledge against the so-called paradox of analysis. In its more general formulation the paradox of analysis states that there cannot be philosophical analyses which are both correct and informative. The paradox is based on the view that if an analysis is correct, the identity between the analysans and analysandum is trivial. Hence all analyses are either incorrect or tautologous. Collingwood clearly thought that there are propositions which are arrived at by means of reflection and which are nonetheless informative. In fact, in order for a proposition to qualify as a philosophical proposition, it must satisfy these two requirements: it must be known a priori and it must be informative. The proposition which defines the domain of inquiry of history, 'all actions are expressions of thought' is a case in point. This proposition is,

EDITORS' INTRODUCTION

cv

for Collingwood, not an inductive generalization since we do not (and cannot) come to know it empirically by repeated observation. Like analytic claims, philosophical propositions are not empirically refutable: the proposition 'all actions are expressions of thought' has more in common with the proposition 'all bachelors are unmarried men' than with the proposition 'all swans are white', since we cannot discard the concept of the mental in the way in which we discard empirical concepts and hypotheses. If we come across something that does not fit the description of an action, we do not discount the concept of action, but conclude rather that that something is not an action. This much philosophical propositions have in common with self-evidently true analytic statements. But whereas philosophical propositions thus resemble analytic statements, they are not analytical in the sense that they are mere definitions whose truth can be known by reflecting on the meaning of the words used. The claim that 'all actions are expressions of thought' is not the result of an arbitrary stipulation; it is arrived at by reflecting on the explanatory practices of historians and by teasing out, from these, the norms which govern them. Philosophical propositions differ from analytic claims because we do not simply accept them as a matter of linguistic convention or use. To endorse the claim that 'all actions are expressions of thought' requires an understanding of what it means to provide

cvi

EDITORS' INTRODUCTION

mentalistic explanations. The relationship between the subject and predicate concepts in a philosophical proposition is therefore a priori but not analytical and philosophical inquiry yields precisely those kind of analyses that are allegedly ruled out by the paradox of analysis. Although philosophical inquiry does not expand our empirical knowledge, it does add to what we know in the sense that it provides a deeper understanding of things that were only implicitly known. In so far as philosophical inquiry enriches our knowledge it is synthetic. But it is not synthetic in the sense that it is about matters of fact or that it has existential import. If we take the a priori-a posteriori distinction to be a distinction between different modes of justification and the analytic-synthetic distinction to be a semantic distinction dividing trivial from non-trivial claims, philosophical propositions may be described as synthetic a priori. They are a priori because they cannot be justified empirically (justification of an appropriately philosophical nature is an internalist matter) and they are synthetic because they provide informative or non-tautologous analyses. Let us now consider the third distinction on the basis of which we can attempt to explain the nature of Collingwood's defence of the autonomy of philosophy, the necessary-contingent distinction. Collingwood claims that the subject matter of philosophy is necessary, not contingent. He makes this claim by saying that philosophical propositions,

EDITORS' INTRODUCTION

cvii

like analytic statements, are universal rather than enumerative, and that, like contingent empirical claims, they are categorical rather than hypothetical. For this reason philosophical propositions are necessary in the way in which universal statements such as 'all bachelors are unmarried men' are, and yet, like contingent propositions expressing generalizations of an empirical nature ('all cats in this room are black'), they are existential statements. Philosophical concepts, in other words, are rather like the concept of God in the ontological proof: they exist necessarily. Collingwood goes as far as reclaiming the centrality of the ontological proof by saying that the idea of necessary existence (the idea expressed by the word 'God' in the ontological proof) is the very subject matter of philosophy, it is that which we are thinking about when we think philosophically. This claim is controversial for two reasons. First, the ontological proof is widely regarded as having failed and, secondly, Collingwood's account of the overlap of classes appears to be making a purely conceptual point, with no ontological or existential import. Collingwood's attempt to reclaim the centrality of the ontological proof and his view that the idea of necessary existence is that which we think about when we think philosophically therefore requires some careful scrutiny. What, in other words, can possibly be meant by Collingwood's claim that philosophical concepts, concepts such as 'mind' or 'matter', exist necessarily?

cviii

EDITORS' INTRODUCTION

The view that the subject matter of philosophy is not contingent is at least as old as one of the earliest philosophical attempts to articulate the relationship between particulars and universals. Plato, for example, believed the object of philosophical knowledge to be transcendent metaphysical forms and particulars to be imperfect copies of them in the realm of becoming. Whereas the forms or universals were said to be self-same and unchanging, the particulars resembling them were said to be subject to change and decay. The idea that philosophy is concerned with a necessary rather than contingent subject matter is as ancient as the beginning of Western philosophy itself. It is this idea that Collingwood appears to be reviving. The way in which Collingwood reappropriates this venerable thought may perhaps be explained as follows. Let us consider once again the claim that 'all actions are expressions of thought' and let us take this to be a philosophical proposition, a proposition that is definitive of the domain of inquiry of history. The proposition is definitive of history's domain of inquiry in the sense that it tells us what historians are concerned with: actions or rational processes (thought or mind). The philosopher of history, as opposed to a historian engaged in firstorder inquiry, is concerned with delineating the subject matter and nature of historical inquiry and as such s/he is concerned not with particular actions but with the concept of action itself. The subject matter of philosophy (the philosophy of

EDITORS' INTRODUCTION

cix

history in this case) is therefore the notion of action as such, not the empirical instantiations or class of objects that fall under it, the particulars that Plato confined to the realm of becoming. The latter are what practising historians study. Since particulars are not the subject matter of philosophical investigation, the distinction between universals and particulars is not a distinction that plays any role in philosophical thought. Whereas the distinction between actions and events or that between mind and matter is a philosophical distinction, the distinction between particulars and universals is a distinction between what falls within (universals) and what falls without (particulars) the domain of inquiry of philosophy. For this reason, the subject matter of philosophy resembles that of a priori sciences such as geometry and mathematics: the philosopher of history is concerned with the concept of action very much as the geometer is concerned with the concept of a triangle rather than with its instances. The concept of 'action' exists for the philosophical historian in the way in which triangles exist for geometers, that is, as ideal forms or standards. So construed, the idea of necessary existence which belongs for Collingwood to the word 'God' in the ontological proof refers not to particulars but to the concepts in which they partake. 'Mind' exists, and necessarily exists for the philosophical historian, because it is what history is about. Philosophical concepts such as 'mind' and 'matter' necessarily exist for the

cx

EDITORS' INTRODUCTION

philosophy of mind and nature because they define their domain of investigation, their respective subject matters or regional ontologies. In the light of this we can conclude that philosophical propositions may be called metaphysical in so far as they are of a necessary, non-empirical nature; they are metaphysical primarily in the negative sense that they are not subject to empirical verification and refutation. As we have seen, when the philosopher of history says that 'all actions are expressions of thought', he does not mean that some actions (observed ones) are expressions of thought. Since the claim is not an inductive generalization, it is immune from empirical disconfirmation. But philosophical propositions are not metaphysical in the sense that they predicate the existence of philosophical concepts in an unqualified metaphysical or ontological sense of the term 'existence'. The philosophical concept of 'mind' necessarily exists in the heavily qualified sense that mind or actions are what history is about. Actions or mind are not said to exist in an unqualified ontological sense of the word, or independently of the regional ontologies of the first-order sciences. Locating Collingwood's conception of philosophical inquiry vis-a-vis the a priori-a posteriori distinction, the analytic-synthetic distinction, and the necessary-contingent distinction should enable us to understand why philosophy has an autonomous domain of inquiry that differs from that of the natural sciences, the exact sciences, and

EDITORS' INTRODUCTION

cxi

traditional metaphysics alike. Since philosophical claims are a priori, not empirical, philosophy is not a natural science; since philosophical claims are not analytic, philosophy cannot be ranked amongst the exact sciences; finally, since the kind of claims made by philosophers are conceptual not ontological (at least not ontological in an unqualified sense of the word) philosophy cannot be identified with metaphysics either. Progress in Philosophical Inquiry

Having outlined the thesis of the autonomy of philosophy in relation to these three different distinctions, we can now turn to the claim that since philosophical propositions are not inductive generalizations, they are not subject to empirical disconfirmation. It is important to focus on this point because one of the fiercest attacks on the notion of the autonomy of philosophical inquiry is related precisely to this claim. Quine,74 for instance, thought that showing that no claim is immune to revision is tantamount to denying the very notion of a priori knowledge and hence the existence of strict boundaries between philosophical and scientific knowledge. In what sense should we take philosophical claims to be unrevisable? In one sense (the empirical sense) philosophical claims certainly are unrevisable. Since philosophical 74

Quine, 'Two Dogmas of Empiricism'.

cxii

EDITORS' INTRODUCTION

propositions such as 'all actions are expressions of thought' are not enumerative propositions about matters of fact, they cannot be empirically refuted. Yet, for Collingwood, the fact that philosophical propositions are immune from empirical disconfirmation does not mean that philosophical claims are infallible and therefore unrevisable. Philosophical claims do not have the status of indubitable Cartesian first principles, principles that we could not possibly get wrong. In other words, immunity from empirical disconfirmation and infallibility are not equivalent: Collingwood simply excludes revision in the light of empirical evidence. The notion of revisability that applies to philosophical claims is closely connected not to the idea of empirical verification but to that of dialectic. In a dialectical argument a claim is discarded not if it is falsified empirically, but if it fails to provide an appropriate answer to the question it has posed or to the problem it was supposed to solve. For Collingwood, success or failure in philosophical matters must be measured with reference to epistemic goals rather than with reference to ends external to knowledge, such as, for instance, the survival of the human species. Whether a philosophical thesis is to be endorsed or rejected depends on whether it succeeds in vindicating what it was supposed to explain. The criterion to be employed in deciding whether a philosophical claim is cogent, therefore, is whether one can, logically speaking, 'get there from here'. The

EDITORS' INTRODUCTION

cxiii

fact that an explanatory hypothesis might prove useful or expedient to some other extra-epistemic goal is a consideration that is irrelevant to the question of its philosophical cogency or validity. This notion of dialectic is inscribed in Collingwood's account of the scale of forms and implies that progress in philosophical inquiry cannot be measured in terms of natural evolution. For this reason the notion of progress employed by so-called 'evolutionary' epistemology is intrinsically ill-suited to capture what it means to make advances in philosophy. Evolutionary epistemology seeks to combine the idea of the growth of knowledge with that of natural selection by invoking Darwin's theory of the evolution of the species as an explanation for the development of human knowledge. According to Darwin, natural species not suited to their environment do not survive because they are in competition with other species that are better adapted to that environment. The evolutionary epistemologist applies the principle of natural selection to the domain of conceptual formations and claims that epistemic selection occurs along the lines of natural biological selection. For the evolutionary epistemologist the questions to be addressed do not concern issues in the theory of justification. Collingwood would probably argue that evolutionary epistemology, like naturalized epistemology, of which it is a specific variation, is not an 'epistemology' in the traditional sense of the word and carries with it the

cxiv

EDITORS' INTRODUCTION

implication that there is no such thing as a distinctive philosophical method and that there are no distinctively philosophical problems to be addressed. To illustrate the notion of progress in philosophical inquiry one might consider how Collingwood uses the notion of a philosophical scale to vindicate the thesis of the autonomy of history by discarding conceptions of the nature of historical inquiry that fail to delineate a subject matter for history (mind/actions) as distinct from that of the natural sciences (matter/events). The progression from 'common sense' to 'scissors-and-paste' and from 'scissors-and-paste' to 'scientific' history in The Idea of History is an example of a philosophical scale in which different claims concerning the nature of historical inquiry are dialectically revised rather than empirically refuted. Collingwood's defence of the autonomy of history does not imply that the concept of mind serves some useful non-epistemic purpose, but that it is presupposed in any attempt to account for how explanations that employ practical arguments rather than invoke statistical considerations are possible. In some sense, it may be argued, Collingwood is concerned with consequences, but the kind of consequences with which he is concerned are consequences of a logical rather than an empirical nature. It is the task of philosophy to organize our thinking and clarify what conclusions can be drawn from what premisses, but such a task does not require any appeal to the empirical

EDITORS' INTRODUCTION

cxv

consequences of thought. Evolutionary epistemology, like some forms of pragmatism, explains the relative superiority of one knowledge claim over another by reference to non-epistemic factors such as the beneficial effects that certain views might have for the survival of the species, rather than by assessing competing claims to knowledge and establishing to what extent they account for the phenomena they set out to explain. If progress in philosophy were judged by the standards adopted in evolutionary epistemology, there would be no place for a science whose role is to assess the validity of knowledge claims by reference to epistemic goals rather than biological imperatives such as that of survival. The kinds of questions of concern to philosophers would not be questions like 'does such and such an account of knowledge succeed in justifying our beliefs in x, y or z?' but rather 'does such and such a belief system promote survival?' Collingwood's defence of the autonomy of philosophical method, as presented here, is closely connected with his conception of philosophy as first science. On this conception of philosophy, the role of philosophical inquiry is to delineate the a priori concepts or categories that make first-order knowledge possible. This interpretation of the Essay raises a number of questions in relation to the development of Collingwood's thought, in particular to An Essay on Metaphysics. Many interpreters of Collingwood have claimed that by the time he came to write An Essay on Metaphysics

cxvi

EDITORS' INTRODUCTION

Collingwood attempted to assimilate philosophy to history, thereby turning his back on the thesis of the autonomy of philosophy. In fact, Collingwood's theoretical philosophy, if discussed at all, is often associated with this historicist turn, and dismissed as an infamous attempt to reduce philosophy to the sociology of knowledge. The story concerning this alleged turn is complex and would require a careful discussion of whether the changes in terminology from Method to Metaphysics are symptomatic of an underlying change of heart and a substantial reconsideration of the relationship between a prioricity and philosophical analysis, or whether the continuities between Method and Metaphysics run much deeper than these undeniable changes in terminology seem to suggest. But such a discussion would divert the focus from Collingwood's first Essay to his second, and it is now time to let this greatly and unjustly neglected philosophical work speak in its own right.

SELECT BIBLIOGRAPHY (A) Writings by Collingwood Books and Pamphlets Religion and Philosophy. London: Macmillan, 1916. Speculum Mentis. Oxford: Clarendon Press, 1924. Outlines of a Philosophy of Art. London: Oxford University Press, 1925 (repr. in Donagan, 1964). Faith and Reason. London: Benn, 1928 (repr. in Rubinoff, 1968). The Philosophy of History. London: published for the Historical Association by G. Bell, 1930. The Principles of Art. Oxford: Clarendon Press, 1938. The Principles of History and Other Writings in Philosophy of History. Edited and with an Introduction by W. H. Dray and W. J. van der Dussen. Oxford: Oxford University Press, 1999. An Autobiography. London: Oxford University Press, 1939. An Essay on Metaphysics. Oxford: Clarendon Press, 1940; rev. edn. with an Introduction and additional material edited by Rex Martin, 1998. The New Leviathan. Oxford: Clarendon Press, 1942; rev. edn. with an Introduction and additional material edited by David Boucher, 1992. The Idea of Nature. Oxford: Clarendon Press, 1945. The Idea of History. Oxford: Clarendon Press, 1946; rev. edn. with an Introduction and additional material edited by W. J. van der Dussen, 1993.

Articles and Reviews 'The Devil' in Concerning Prayer, ed. L. Dougall. London: Macmillan, 1916 (repr. in Rubinoff, 1967).

cxviii SELECT BIBLIOGRAPHY Review of C. J. Ducasse, The Philosophy of An in Philosophy, vi, 1931 (July), 383-6 (repr. in Collingwood Studies, vol. 2, 1995, 204-7). Review of M. C. D'Arcy, The Nature of Belief, The Criterion, vol. 11, 1932, 334-6 (repr. in Collingwood Studies, vol. 5, 1998, 142-5).

Collections of Articles Boucher, D. (ed.), Essays in Political Philosophy. Oxford: Clarendon Press, 1989. Donagan, A. (ed.), Essays in the Philosophy of Art. Bloomington: Indiana University Press, 1964. Rubinoff, L. (ed.), Faith and Reason: Essays in the Philosophy of Religion. Chicago: Quadrangle, 1967.

Unpublished Manuscripts Truth and Contradiction, ch. 2 (1917). Bodleian Library, Dept. 16/1. Lectures on the Ontological Proof of the Existence of God (1919). Bodleian Library, Dept. 2. Lectures on Moral Philosophy (1921). Bodleian Library, Dept. 4. Action': Lectures on Moral Philosophy (1923). Bodleian Library, Dept. 3/1. 'Outlines of a Philosophy of History' (1928) (published in the rev. edn. of The Idea of History). Lectures on Moral Philosophy (1929). Bodleian Library, Dept. 10. Lectures on Moral Philosophy (1932). Bodleian Library, Dept. 7. Lectures on Moral Philosophy, (1933). Bodleian Library, Dept. 8 (extracts published in Boucher (ed.), Essays in Political Philosophy, 1989).

SELECT BIBLIOGRAPHY cxix 'Notes towards a Metaphysic' (1933). Bodleian Library, Dept. 18/3-7 (extracts published in The Principles of History and Other Writings in Philosophy of History). 'The Metaphysics of F. H. Bradley: An Essay on Appearance and Reality' (1933). Bodleian Library, Dept. 29. 'The Nature of Metaphysical Study' (1934). Bodleian Library, Dept. 18/2 (the second lecture is published in the rev. edn. of An Essay on Metaphysics). 'Method and Metaphysics' (1935). Bodleian Library, Dept. 19/3. 'Function of Metaphysics in Civilization', (1938). Bodleian Library, Dept. 19/7 (published in rev. edn. of An Essay on Metaphysics). 'What Civilization Means' (1939-40). Bodleian Library, Dept. 24 (published in rev. edn. of The New Leviathan). 'The Idea of a Philosophy of Something, and, in particular, a Philosophy of History' (1927). Bodleian Library, Dept. 14/2 (published in rev. edn. of The Idea of History). Correspondence1 Letters to the Clarendon Press are held in the Clarendon Press Archives, Oxford (see Johnson, 6-34). Letters to G. de Ruggiero. Bodleian Library, Dept. 27 (see Johnson, 51-9). Correspondence with Gilbert Ryle, 9 May & 6 June 1935. Bodleian Library, MS. Eng. Lett. d.194 (see Johnson, 100-1).

(B) Writing On Collingwood Beaney, M., ' Collingwood's Critique of Analytic Philosophy', Collingwood and British Idealism Studies, vol. 8, 2001, 99—122. 1

Full details of correspondence can be found in P. Johnson, The Correspondence of R. G. Collingwood: an Illustrated Guide.

cxx SELECT BIBLIOGRAPHY Boucher, D., The Social and Political Thought of R. G. Collingwood. Cambridge University Press, 1989. 'The Principles of History and the Cosmology Conclusion to the Idea of Nature, Collingwood Studies, vol. 2, 1995, 140-74. 'The Significance of R. G. Collingwood's Principles of History', Journal of the History of Ideas, vol. 58, no. 2, 1997, 309-30. Connelly, J., and Modood, T (eds.), Philosophy, History and Civilization: Interdisciplinary Perspectives on R. G. Collingwood. Cardiff: University of Wales Press, 1995. Burchnall, R. A., Catalogue of the Papers of Robin George Collingwood (1889-1943). Bodleian Library, Oxford. Connelly, J., 'Bradley, Collingwood and the "Other Metaphysics" ', Bradley Studies, vol. 3, no. 2, 1997, 89-112. 'Natural Science, History and Christianity: the Origins of Collingwood's Later Metaphysics', Collingwood Studies, vol. 4, 1997, 101-32. Metaphysics, Method and Politics: The Political Philosophy of R. G. Collingwood. Exeter: Imprint Academic, 2003. and Costall, A., 'R. G. Collingwood and the Idea of a Historical Psychology', Theory and Psychology, vol. 10, no. 2, 2000, 147-70. D'Arcy, M. C., Review of An Essay on Philosophical Method, The Criterion, vol. 13, 1933-4, 500-03. D'Oro, G., Collingwood and the Metaphysics of Experience. London: Routledge, 2002. 'Collingwood and Ryle on the Concept of Mind', Philosophical Explorations, vol. 6, 2003, 18-30. Donagan, A., The Later Philosophy of R. G. Collingwood. Oxford: Clarendon Press, 1962 (2nd edn., University of Chicago Press, 1985). Dreisbach, C., R. G. Collingwood: a Bibliographic Checklist. Bowling Green, Ohio: The Philosophy Documentation Centre, 1993.

SELECT BIBLIOGRAPHY cxxi Ducasse, C. J. 'Mr. Collingwood on Philosophical Method', in Journal of Philosophy, vol. 33, 1936, 95-106. Harris, E. E., Nature, Mind and Modern Science. London: Allen & Unwin, 1954. 'Collingwood's Treatment of the Ontological Argument and the Categorical Universal', in Krausz (ed.), Critical Essays on the Philosophy of R. G. Collingwood. 'Mr. Ryle and the Ontological Argument', Mind, vol. XLV, Oct. 1936, 474-80. Hartshorne, C., Review of An Essay on Philosophical Method, International Journal of Ethics, vol. 44, 1934, 357-8. Johnson, P., R. G. Collingwood: An Introduction. Bristol: Thoemmes, 1998. The Correspondence of R. G. Collingwood: an Illustrated Guide. R. G. Collingwood Society, 1998. Kant, I., Critique of Pure Reason, translated by N. K. Smith. London: Macmillan 1929. Knox, T M., Review of An Essay on Philosophical Method, Oxford Magazine, 23 November 1933, 257-9. Prefatory note to The Idea of Nature. Oxford: Clarendon Press, 1945. Preface to The Idea of History. Oxford: Clarendon Press, 1946, v-xxiv. Krausz, M. (ed.), Critical Essays on the Philosophy of R. G. Collingwood. Oxford: Clarendon Press, 1972. MacKinnon, D. M., Explorations in Theology, 5. London: S.C.M., 1979. Martin, R., 'Collingwood's Essay on Philosophical Method ', Idealistic Studies, vol. 4, 1974, 224-50. Mink, L., Mind, History and Dialectic. Bloomington: Indiana University Press, 1969. Quine, W. V., From a Logical Point of View, 2nd edn. New York: Harper & Row, 1963. Ontological Relativity and Other Essays. New York: Columbia University Press, 1969.

cxxii SELECT BIBLIOGRAPHY Ronayne, C. F., Review of An Essay on Philosophical Method, American Review, vol. 4, 1935, 627-33. Rubinoff, L., Collingwood and the Reform of Metaphysics. Toronto: University of Toronto Press, 1970. Russell, B., The Philosophy of Logical Atomism, La Salle, Illinois: Open Court edn., 1985 (first published in 1918 by Open Court). Russell, L. J., Review of An Essay on Philosophical Method, Philosophy, vol. 9, 1934, 350-2. Ryle, G., 'Systematically Misleading Expressions', Proceedings of the Aristotelian Society, vol. 32, 1931-2, 139-71 (repr. in Collected Essays, vol. 2). 'Mr. Collingwood and the Ontological Argument', Mind, vol. 44, 1935, 137-51 (repr. in Collected Essays, vol. 2). 'Back to the Ontological Argument', Mind, vol. 46, 1937, 53-57 (repr. in Collected Essays, vol. 2). 'Taking Sides in Philosophy', Philosophy, vol. 12, 1937, 317-32 (repr. in Collected Essays, vol. 2). 'Philosophical Arguments', Oxford: Clarendon Press, 1945 (repr. in Collected Essays, vol. 2). Autobiographical', in O. P. Wood and G. Pitcher (eds.), Ryle: A Collection of Critical Essays. London: Macmillan, 1970. Collected Essays, two volumes. London: Hutchinson, 1971. Schiller, F. C. S., Review of An Essay on Philosophical Method, Mind, vol. 43, 1934, 117-20. Stebbing, L. S., Review of Speculum Mentis, Hibbert Journal, vol. 28, 566-9. 'The Method of Analysis in Metaphysics', Proceedings of the Aristotelian Society, 1932-3, 65-94. Stock, G., 'Collingwood's Essay on Appearance and Reality: some Contemporary Reflections', Collingwood Studies, vol. 4, 1998, 34-50. Vrijen, C., 'Ryle and Collingwood: Their Correspondence and its Philosophical Context', British Journal for the History of Philosophy, forthcoming.

AN ESSAY ON PHILOSOPHICAL METHOD

This page intentionally left blank

CONTENTS I. INTRODUCTION § I. THE PROBLEM 1. Philosophy must raise the question what philosophy is . 2. Three suggested ways of approaching this question . 3. The way to be followed here: an account of philosophical method . . . . . 4. Importance of this subject at the present time . . § 2. THE METHOD 5. Necessity of restricting the subject under discussion . 6. Significance of the comparison between philosophical and scientific thought . . . . §3. HISTORICAL ILLUSTRATIONS 7. Socrates . . 8 . Plato . . . 9. Descartes . . 10. Kant . . .

.

. .

. .

. .

. .

I 2 .3 4 7 9 .10 . 1 1 .16 .19

. . .

II. THE OVERLAP OF CLASSES § I. THE THEORY OF CLASSIFICATION IN FORMAL LOGIC 1. The traditional theory of classification . . 2. Its application in exact (mathematical) science . 3. Its application in empirical science . .

.

.26 29 .30

§ 2. THE IMPOSSIBILITY OF APPLYING IT RIGIDLY TO PHILOSOPHICAL CONCEPTS 4. Specific classes here overlap one another • 31 5. This fact long recognized in certain cases . . 32 6. The case of concepts having a philosophical and a nonphilosophical phase . . . . 33 7. The overlap of classes in logic . . . .36 8. The overlap of classes in ethics . • .41 9. Significance of these facts: the reader invited to admit their genuineness for the sake of argument . . 43

cxxvi

CONTENTS

§ 3. CONSEQUENCES FOR PHILOSOPHICAL METHOD 10. Methods proper to science (exact or empirical) inapplicable in philosophy, owing to the overlap of classes . 46 11. The fallacy of precarious margins . . .47 12. The fallacy of identified coincidents . . .48 13. Their common ground: the fallacy of false disjunction . 49 14. The first rule of philosophical method: «? 58, 102, 123, 127 Augustine, St. 125 Ayer, A. J. xxxix, xcvii Beaney, M. xxx Bentham, J. 79 Berkeley, G. Ixxxix—xc, 202, 232, 240, 249, 302-3 Bodies 341-54 Boethius 124, 125 Boucher, D. xlvin Bradley, F. H. xxxi, xxxn, h—hv, Ixvi, 229-52, 259 Broad, C. D. xxx n. Burnet, J., 14 n. Chapman, R. W. xvn, xxxix n Classes, overlap of xvni—xxi, xxiv—xxv, xxix, xxx, xl—xlv, liii-liv, Ixx-lxxiii, Ixxix-lxxx, Ixxxiv-lxxxv, Ixxxvi-lxxxviii, Ixxxvi, Ixxxviii, xcviii-ci, cm, Ch II passim, 54—5, 63-4, 74, 76, 78, 86-91, 95, 105, 107, i n , 134-5, 148, 161, 188-9, 200, 213, 221, 244,

245

Cook Wilson, J. 245, 256 Croce, B. xxvi, xxxi, 273

D'Arcy, M. C. xxxviii-xxxix De Ruggiero, G. xm Descartes, R. xxxv, cm, 5, 7, 17—20, 126, 155-8, 202, 213, 232, 249, 343 Dialectic, dialectical xxii, xxv-xxvii, xxxi, lii, 10-15 Ducasse, C. J. xxxix—xlv Essence and existence 127, 131, 133, 266-7, 284, 295, 308-9, 314 Einstein, A. 250 Epistemology normative en naturalized en evolutionary cxiii Externalism c-cii Galileo 232, 248-9, 250 Generic—general distinction Ixxxiv, 27-8,46-7, 113-15, 187-8 see also universals vs empirical classes and intension-extension Generic essence 28, 57, 59-62, 65, 76-7, 81-2, 88-9, 91, 113-15 Gentile, G. xxvi, xxxi, Ixiv, 273 Gorgias 102 Hamilton, W. 233-4, 237, 238, 240 Harris, E. E. Ixv Hartshorne, C. xxxvn Hegel, G. W. F. xxxi, xxxv, Ivm, Ixxx, 103, 123, 126, 159, 213, 230, 231, 243 History civ-cv, cviii-cx, cxiv Homer 343 Hooper, S. E. xxxvn

358

INDEX

Hume, D. Ivi, Iviii, 223, 233, 238-9, 258, 259, 299, 300, 316 Idealism 229, 231, 255-6, 296-7 Identified coincidents, fallacy of xxn, 49, 112 Intension-extension xviii, Iv, Ixxxiii, Ixxxvi, xci Internalism c-cii Joachim, H. H. cccvn Johnson, W. E., 148 n, 262 Joseph, H. W. B. xxxvii, 274, 312 Kant, I. xxxv, liii. Ivi, Iviii, Ixxx, Ixxxix—xc, xcix, cm, 5, 7, 19-25, 58, 102-3, 123, 126, 138, 156, 158-60, 202, 223, 258, 278-9, 300, 316, 326 Knowing in philosophy xxix-xxx, Ixxvii, II, 96-8, 105-6, 161-4, 168 Knox, T. M. xxxvn, xhx Leibniz, G. W. 58, 60, 158-9 Locke, J. liii, 58, 60, 232, 238, 239, 240, 243, 249 Logical atomism 280,312 MacKinnon, D. M. li Mansel, H. L. lii, 234-7, 238, 239-40, 242 Martin, R. xlvi n Mathematics 9, 11-15, 17—20, 117-18, 151-4 Matter 1, Ixxvii, Ixxix, Ixxxv, Ixxxvi-lxxxviii, Ixxxix, xcii, xciii, xcix, c, cii, cvii, cix, 33-4, 302-3 Metaphysics 1, Ixv—Ixvm, Ixxiv, 229-52, 258-9, 318, 327-55 Mill, J. S. lii, 39, 79-80, 237, 238, 240, 273

Mind 1, Ixxvii, Ixxviii, Ixxix, Ixxxv, Ixxxvi-lxxxviii, Ixxxix, c, civ-cv, cviii-cx, 34, 299-300, 302-3, 332,341-54 mind—body relation Ixiv, 332, 243-8 Mink, L. O. xxxii Moore, G. E., 142 n, 245, 273,330 Moral philosophy 42—3, 63, 78—9, 82-3, 88, 90, 102-3, I T 3 > I T 4> 131-3, 172 Naturalism: critique of Ixxxi Necessary-contingent distinction xcvm—xcix, cvi, ex Newton, I. 19, 249, 250. Ontological proof xxxiii-xxxvii, xxxviii, Iv, Ivi, Ivii-lix, Ixi, Ixxix, cvii, 123-33, 254 n, 257, 298-9, 316 Ordinary language philosophy xciv Overlap of classes, see classes, overlap of Paul, St. 56, 89 Paradox of analysis xxx, civ Phenomenalism In, 233—6, 238—41, 242, 244 Philosophical concepts xxviii, 32-6, 40, 45-6, 50-3, 54-70, 73-4, 76-7, 92-3, 95-6, 97-101, 107-8, 113, 135, 187-9, T 93> 221-2 as a priori Ixxvn, Ixxxvin, xcm, xcviii-xcix, c, ci-cii, ciii-civ, cvi, cix-cxi, cxv, 258-9, 299, 300, 318, 326 as transcendentals xxvin, Ixx definition of: Ch IV passim differ in degree and kind xvm, liii, 54-8, 69-81

INDEX not class concepts xxviii-xxix, Ixxxi-vi, 330 related by opposition and distinction xix, 63—9, 74—7, 81-6 Philosophy analytical 141-7 as categorical thinking xvi, xxxiv, Ivi-lvii, lix, Ch VI passim, 153,193, 222 autonomy of xcvn critical 138—41 and deduction 151-63 and experience 170-5 and historical writing 208-12 and induction 164—75 and literature xxxin, xhv, Ch X passim and poetry 212-16 and system Ch Yf^passim technical terms in 201-8 Plato xxxi, hx, cvm, cix, 12—16, 25, 56-61, 101-2, 123, 124, 127, 202, 213, 333 n Pleasure 78-81 Pragmatism 229 Precarious margins, fallacy of 48-9, 112 Presuppositions, absolute Ixx—Ixxvi, c Prichard, H. A. 245 Propositions 'any' 286-8,292,301,314-15 'all' 288, 301, 314-15 categorical xvi, xxxi, xhn, Ivi, Ivn, Ixx, Ixxxni, cvn, Ch VI passim, 260-1, 264, 278, 287, 288-9, 304-8, 323 enumerative liv, Ixx, cvii, cxii, 260, 260 n, 264, 288, 292, 306, 314 'every' 287—8, 292 existence 320-3

359

hypothetical 259, 261-3, 264, 268, 270-1, 278, 292-301, 303-8, 310-12 philosophical xxvin, 32—6, 40,45-6, 50-3, 54-70, 73-4, 76-7, 92-3, 95-6, 97-101, 107-8, 113, 135, 187-9,

T

9 3 > 221-2

universal xvi, xxviii, xxxi, Ivi, Ivn, Ix, Ixi, Ixx, Ixxvni, Ixxxi, Ixxxni—Ixxxv, cvn—cvm, cix, 292—4, 342,350 verbal 319, 323-5 Quine, W. V O. Ixxxvi, xcvii—xcvin, en n, cxi Realism 229, 231, 241, 245-8, 256 Collingwood's critique of xxxviii-xciii Reality 241, 328-55 degrees of, see truth and reality, degrees of Ross, W. D. xv, 78 n, 273-4 Russell, B. Iv, lix, Ixxxi, Ixxxiii, Ixxxv-lxxxvi, 146, 245, 267, 302, 312 Russell, L. J. xxxvn Ryle, G. xxxvn, xxxvin, xxxix, xh, xlii, liv-lxv, Ixxxi, Ixxxv-lxxxvi, 253-326 Scale of Forms xvin—xxi, xxn, xxiv—xxvn, xxx, xlv—xlvni, 1, In—hv, Ixvi—Ixvni, Ch III passim, 97, 100-2, 108, 114, 135, 161, 172, 189-93, 195-8, 245, 334, 339-40, 345, 346, 348 and overlap of classes xvin—xxi, 86-91 Schiller, F. C. S. xxxvii, Ivi

Te

INDEX

Science empirical xlii, 9, 118-21, 259, 264 exact Ixxvii, 9, 93-6, 117-18, 151-4, 259, 264 Smith, J. A. xxxvin, 14 n Socrates, 10-11, 17, 92-3, 102 Spencer, H. 237, 238 Spinoza, B. 1, lix, 33, 126, 127, 157-9, 162, 168, 171, 188, 213 Stebbing, L. S. xxix, xxx, xlin—xhv, 145

Taylor, A. E. 14 n Truth and reality xxic, xxxii, lii, Ixvi, degrees of 243, 246-8, 335,338 Universals vs empirical classes xlii, hv, Ixxxi, Ixxxin—Ixxxvi, xcin, 292 Wittgenstein, L. xcvi