Wittgenstein's Method

Wittgenstein’s Method

To Alan and Shelley, and Geoff and Charlotte, and Nick and Tash

Wittgenstein’s Method Neglected Aspects Essays on Wittgenstein by Gordon Baker Edited and introduced by Katherine J. Morris

ß 2004 by Blackwell Publishing Ltd except for editorial material and organization ß 2004 by Katherine J. Morris BLACKWELL PUBLISHING

350 Main Street, Malden, MA 02148–5020, USA 108 Cowley Road, Oxford OX4 1JF, UK 550 Swanston Street, Carlton, Victoria 3053, Australia The right of Katherine J. Morris to be identified as the Author of the Editorial Material in this Work has been asserted in accordance with the UK Copyright, Designs, and Patents Act 1988. 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, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the prior permission of the publisher. First published 2004 by Blackwell Publishing Ltd Library of Congress Cataloging-in-Publication Data Baker, Gordon P. Wittgenstein’s method: neglected aspects: essays on Wittgenstein/by Gordon Baker; edited and introduced by Katherine J. Morris. p. cm. ‘‘Bibliography for Gordon Baker’’: p. Includes bibliographical references (p. ) and index. ISBN 1-4051-1757-5 1. Wittgenstein, Ludwig, 1889–1951. I. Morris, Katherine J. II. Title. B3376.W564B354 2004 192–dc22 2003020376 A catalogue record for this title is available from the British Library. Set in 10/12.5 pt Galliard by Kolam Information Services Pvt. Ltd, Pondicherry, India Printed and bound in the United Kingdom by TJ International, Padstow, Cornwall The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com

Perhaps! – But who is willing to concern himself with such dangerous perhapses! (Nietzsche, Beyond Good and Evil, Part I: On the Prejudices of Philosophers, § 2)

Contents

Acknowledgements

Introduction by Katherine J. Morris

Part I

ix

1

Reading Wittgenstein

19

A

Methodological Concepts

21

1

Philosophical Investigations §122: Neglected Aspects

22

2

Some Remarks on ‘Language’ and ‘Grammar’

52

3

Wittgenstein’s ‘Depth Grammar’

73

4

Wittgenstein on Metaphysical/Everyday Use

92

B

Applications: the ‘Private Language Argument’

108

5

The Reception of the Private Language Argument

109

6

Wittgenstein’s Method and the Private Language Argument

119

7

The Private Language Argument (Extract: Final Section)

130

Part II Wittgenstein and Waismann A The Analogy with Psychoanalysis

141 143

8

‘Our’ Method of Thinking about ‘Thinking’

144

9

A Vision of Philosophy

179

viii

Contents

10

Wittgenstein’s Method and Psychoanalysis B

Aspects and Conceptions

205 223

11

Italics in Wittgenstein

224

12

Wittgenstein: Concepts or Conceptions?

260

13

The Grammar of Aspects and Aspects of Grammar

279

Bibliography of the Works of Gordon Baker

294

General Bibliography

299

Index

305

Acknowledgements

The original idea of putting together a collection of the ‘later Baker’ articles on the later Wittgenstein came from Talbot Taylor; it was made to Gordon himself, and Gordon would have gone ahead with it had illness not intervened. Rupert Read also suggested it to me after Gordon’s death. I am grateful to Nick Bellorini of Blackwell for his unflagging support for this project. I owe a tremendous debt of gratitude to Graham McFee for his detailed and perceptive comments on earlier versions of this introduction. Edward Kanterian and Andrew Lugg also made many improvements in both style and content. Thanks to Nadine Faulkner for valuable comments and moral support. Alan Baker took on the difficult task of putting together a bibliography for Gordon, included at the end of this volume, for which I am very grateful. Tim Horder’s encouragement and practical help were a sine qua non. I would like gratefully to acknowledge the permission granted to reproduce the copyright material in this book: Essay 1: ‘Philosophical Investigations §122: Neglected Aspects’, originally published in R. L. Arrington and H.-J. Glock (eds), Wittgenstein’s ‘Philosophical Investigations’: Text and Context, pp. 35–68. London and New York: Routledge, 1991. Reproduced by permission. Essay 2: ‘Some Remarks on ‘‘Language’’ and ‘‘Grammar’’ ’, originally published in J. Schulte and G. Sundholm (eds), Criss-crossing a Philosophical Landscape: Essays on Wittgensteinian Themes Dedicated to Brian McGuinness, pp. 107–31. Amsterdam: Rodopi, 1992. Reproduced by permission of Rodopi B.V. Amsterdam.

x Acknowledgements

Essay 3: ‘Wittgenstein’s ‘‘Depth Grammar’’ ’, originally published in Language & Communication, 21 (2001), pp. 303–19. ß 2001 by Elsevier Science Ltd. Reproduced by permission. Essay 4: ‘Wittgenstein on Metaphysical/Everyday Use’, originally published in The Philosophical Quarterly, 52: 208 (2002), pp. 289–302. Reproduced by permission of Blackwell Publishing. Essay 5: ‘La Re´ception de l’Argument du Langage Prive´ ’, originally published (in French) in Fernando Gil (ed.), Acta du Colloque Wittgenstein (Colle`ge International de Philosophie, 1988), pp. 29–40. Mauvezin, France: TER, 1990. Reprinted by permission of E´ditions Trans-Europ Repress. The present translation is by Katherine J. Morris. Essay 6: ‘La Me´thode de Wittgenstein et l’Argument d’un Langage Prive´’, originally published (in French) in J. Sebestik and A. Soulez (eds), Wittgenstein et la philosophie d’aujourd’hui, pp. 261–72. Paris: Me´ridiens Klincksieck, 1992: Reproduced by permission of Me´ridiens Klincksieck. The present translation is by Katherine J. Morris. Essay 7: This extract is the final section (§4) of ‘The Private Language Argument’, originally published in Language & Communication, 18 (1998), pp. 325–56. ß 1998 by Elsevier Science Ltd. Reproduced with permission from Elsevier. Essay 8: ‘ ‘‘Notre’’ me´thode de pense´e sur la ‘‘pense´e’’ ’, originally published (in French) in A. Soulez (ed.), Dicte´es de Wittgenstein a` Friedrich Waismann et pour Moritz Schlick, vol. 2 (E´tudes critiques). Paris: PUF, 1997. Reproduced by permission of Presses Universitaires de France. What is printed here for the first time is Gordon Baker’s original English text. Essay 9: ‘A Vision of Philosophy’, originally published in Figuras do Racionalismo, Conferencias ANPOF 1999, pp. 133–77 (1999). Reproduced by permission. Essay 10: ‘Wittgenstein’s Method and Psychoanalysis’, in K. W. M. Fulford et al. (eds), Nature and Narrative, vol. 1 (2003) of the International Perspectives in Psychiatry and Philosophy series from Oxford University Press. Reproduced by permission of Oxford University Press.

Acknowledgements

xi

Essay 11: ‘Italics in Wittgenstein’, originally published in Language & Communication, 19 (1999), pp. 181–211. ß 1999 by Elsevier Science Ltd. Reprinted by permission. Essay 12: ‘Wittgenstein: Concepts or Conceptions?’, originally published in Harvard Review of Philosophy, 9 (Spring 2001), pp. 7–23. Reprinted by permission of the Harvard Review of Philosophy. Essay 13: ‘The Grammar of Aspects and Aspects of Grammar’, hitherto unpublished. What is printed here is a version of a draft by Gordon Baker, worked up into essay form by Katherine J. Morris. Every effort has been made to trace copyright holders and to obtain their permission for the use of copyright material. The publisher apologizes for any errors or omissions in the above list and would be grateful if notified of any corrections that should be incorporated in future reprints or editions of this book. K. J. M.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

Introduction Katherine J. Morris

Gordon Baker had a long-standing fascination with Wittgenstein and was the author or co-author of a formidable number of articles and books about him. Even those writings which were not explicitly or directly about Wittgenstein were clearly influenced by him. Someone might well have asked ‘When is Baker going to start doing his own work?’1 It is to be hoped that this prefatory essay, in conjunction with Baker’s essays themselves, will make it clear that this question contains a suppositio falsi.2 Baker’s interpretation of Wittgenstein underwent a number of changes over the thirty-plus years of his professional involvement. We might speak of the ‘early Baker’, the author of his DPhil thesis, ‘Criteria: A New Foundation for Semantics’ (1974)3 and ‘Defeasibility and Meaning’ (1977). Here the later Wittgenstein was presented as the creator of a novel, sophisticated and defensible theory of meaning, entirely different from his earlier, ‘picture’ theory of meaning. The ‘Middle Baker’, the coauthor, with Peter Hacker, of the first two volumes of the monumental commentary on the Philosophical Investigations (Wittgenstein: Meaning and Understanding and Wittgenstein: Rules, Grammar and Necessity) among other things,4 roundly rejected the idea that the later Wittgenstein would develop a ‘theory of meaning’ or a ‘foundation for semantics’, new or otherwise. The task of true philosophy, rather, was to replace theorizing with platitudinous descriptions of ‘grammar’ and to police the borders between sense and nonsense, issuing tickets to those philosophers, psychologists and linguists who transgressed the bounds of sense. (An alternative analogy: Wittgenstein was seen as practising corrective therapy – like orthodontics! (Essay 4).) The ‘later Baker’ moved in a very different direction, beginning somewhere around 1986, and established with the publication of ‘Philosophical Investigations §122: Neglected Aspects’ in 1991 (Essay 1 of this volume).5 Wittgenstein was no longer a policeman but a psychotherapist;

2 Introduction

his tools were less ‘factual’ descriptions of grammar than pictures and analogies; his aim was not to get others to toe the line of sense as opposed to nonsense but to free them from their intellectual torment by enabling them to see new aspects; and his method was no longer ‘true philosophy’ but his method, one conception of philosophy. The early Baker – somewhat to the embarrassment of his later incarnations – inspired an ‘ology’, ‘Criteriology’. Philosophers of law, I am told, still read ‘Defeasibility and Meaning’ with admiration. And many Wittgenstein scholars still think of Baker as indistinguishable from the enormously influential if highly controversial Baker-and-Hacker (or B&H, as their students denominated them, more or less affectionately). The present collection is devoted to the later Baker – partly because he is the least well known, and this partly because his writing has until now largely been scattered in journals and book chapters, some obscure, some inaccessible to those who cannot read French.6 But this is perhaps not the only reason for the later Baker’s relative lack of renown. Wittgenstein’s method on the later Baker’s reading is also radically different, not just from earlier versions and from other authors’ readings of Wittgenstein (even to some extent the so-called ‘New Wittgensteinians’),7 but from many anglophone philosophers’ conceptions of ‘the nature of philosophy’, and the unfamiliar can be difficult to perceive even when it is in plain view (just as, for different reasons, the familiar can).8 I highlight two general headings and some themes within these headings around which this collection is organized.9 The aim is to provide an ¨ bersicht of the later Baker’s perspective on the later Wittgenstein. U

Reading Wittgenstein One noteworthy characteristic of the later Baker’s way of reading Wittgenstein is its grounding in careful attention to the details of the text.10 Essay 2 mentions four ‘principles’ of Baker’s ‘me´thode de lire’: (i) a principle of charity: ‘we should proceed on the basis that the texts which Wittgenstein constructed himself consist of carefully thought out arrangements of remarks whose precise wording was of paramount importance’; (ii) ‘a kind of minimalism’ as an antidote to our ‘craving for generality’: ‘we should attach to each expression the interpretation which gives it the minimum generality compatible with the context’; (iii) ‘scrupulous attention to Wittgenstein’s overall therapeutic conception of his philosophical investigations’; and (iv) recognizing ‘the variety of remarks which Wittgenstein offered as descriptions of ‘‘the grammar of our language’’ ’.

Introduction

3

One might think that this me´thode de lire would not distinguish his way of reading Wittgenstein from that of any other respectable scholar; careful reading is what respectable scholars do. Yet Baker’s attention to the text relative to this norm might be said to constitute a ‘transition from quantity to quality’. Part of what it involves is simply scholarly knowledge, brought to bear in a way that makes a difference:11 (1) Biographical knowledge: e.g., the knowledge that Wittgenstein was not well acquainted with Descartes’s work but was with Russell’s might be relevant to understanding his philosophical targets (see Essay 5). (2) Knowledge of the ‘archaeology’ of the various remarks: e.g., in an earlier version of the famous passage ‘What we do is to bring words back from their metaphysical to their everyday use’ (PI §116), Wittgenstein struggled to find the right expression of the second term in the contrast; this might bear on our understanding of both terms (see Essay 4). Such archaeology also contributes to Essay 11. That earlier versions of some passages containing italics were written using ‘scare quotes’ might suggest that italics need not always serve merely as emphasis, and this opens up many new possibilities for understanding some celebrated passages.12 (3) Linguistic knowledge: e.g. the point that the German phrase ‘die Sprache’ is ambiguous between ‘language’ and ‘the (i.e. this) language (say, the language-game just sketched)’ (Essay 2). Which reading we give it has a decisive importance for how we interpret Wittgenstein’s point on any given occasion. Baker’s reading of Wittgenstein involves attending to Wittgenstein’s use of words. (If we see the meaning of a word as its use in the language, we might see the meaning of Wittgenstein’s words as their use in his language; cf. Essay 9.) For example, Essay 12 focuses in part on his use of the words ‘Bild’, ‘Vorstellung’, and ‘Auffassung’; Essay 3 on his use of the expressions ‘depth grammar’ and ‘surface grammar’. It is not that he uses these or other terms in a technical or even a particularly idiosyncratic sense; nonetheless he has a particular conception (Auffassung) in mind and in order to understand it we need to look at his ‘grammar’. (In the latter case, Baker connects ‘depth’ with Wittgenstein’s conception of ‘dimension’ [PI p. 200].)13 Careful reading a` la Baker involves respect (or rather, respect on a certain conception of respect: see below). Few philosophers polish their texts as thoroughly as Wittgenstein did those works that he actually published.14 This suggests that, for example, if Wittgenstein phrases something in the form of a question it should not automatically be assumed to be a ‘rhetorical question’, i.e. a disguised assertion. Again, if Wittgenstein says ‘Here is one possibility’ or ‘Here we might say’, this

4 Introduction

need not simply be a self-effacing way of saying ‘Here is the only possibility’ or ‘This is how it really is’. Again, we should attend to the context of the passage in question; and this includes being sensitive to the possibilities that his uses of words vary from context to context and that a ‘claim’ made in one context, for a particular purpose, may be criticized in another, without drawing accusations of inconsistency. And we should attend not just to what Wittgenstein says but to what he does not say (a point prominent in Essay 5). Baker’s approach to reading Wittgenstein involves ‘seeing connections’ (cf. PI §122). For example, Wittgenstein says that ‘our method’ involves ‘describing grammar’; he also stresses the enormous variety of what we call ‘describing’. Should we not connect these? Yet many commentators impose a very narrow model of ‘description’ upon Wittgenstein’s remarks about grammar (for example, Essay 1, n. 12; Essays 6 and 13). Again, we ought to connect our interpretation of this or that remark or sequence of remarks with what Wittgenstein says he is doing. If we find ourselves ascribing a reductio ad absurdum to him whereas he inveighs against proofs in philosophy, we should at least feel some discomfort; likewise if we find ourselves ascribing theses to him, and defending these against objections (see Essay 6). Finally, Baker’s way of reading Wittgenstein involves openness to alternative possibilities – even imagination and sympathy. At one end, this simply means an openness to alternative ways of reading a particular phrase or passage. (A central feature of Essay 1 is Baker’s identification of an ambiguity in the expression ‘perspicuous representation’; of Essay 3, alternative ways of understanding the phrase ‘can be taken in by the ear’ with which Wittgenstein characterizes ‘surface grammar’, etc.) At the other, it means openness to recognizing alternative conceptions of philosophy. For example, following the consequences of the ambiguity in the phrase ‘perspicuous representation’ leads Baker to a radically different conception of Wittgenstein’s philosophical method, a conception according to which Wittgenstein is doing something very different from traditional philosophers (see below).15 It is all too easy to bring our own preconceptions about philosophy to bear on reading Wittgenstein. If we expect assertions, we are apt to ignore the possibility that questions are meant as questions, and adumbrations of what we are apt to think of as ‘mere possibilities’ may tend to get hardened up into assertions. If we expect to find knock-down proofs and refutations, preferably of important philosophical positions like Cartesian dualism, we are likely to see lacunae in Wittgenstein’s arguments. And then we are prone either to criticize him for this defect or to extend a helping hand by supplying the missing details (see, e.g., Essay 6). Though

Introduction

5

we may see our filling-in of Wittgenstein’s ‘lacunae’ as expressing our respect by rescuing a great philosopher who was drowning in his own material, Baker sees this as the opposite of respect: it is the product of the imposition of our own prior conceptions onto Wittgenstein’s texts. All of the essays in this volume bear the marks of Baker’s me´thode de lire. However, the essays gathered together in Part I.A have in common that they focus on certain of Wittgenstein’s characterizations of his own method ¨ bersicht’, ‘language’, ‘grammar’, ‘depth grammar’, ‘metaphysical vs. (‘U everyday use’) which, Baker suggests, have been widely understood in ways that are not compelled by and are arguably at odds with a careful reading of Wittgenstein’s texts. To read them in the way that Baker urges would therefore change our whole way of looking at Wittgenstein’s philosophical method. As Baker stresses repeatedly, Wittgenstein aims to ‘demonstrate a method by means of examples’; the method is to be judged by its results. Several of these are discussed in some detail in the essays collected here. For instance, Essay 2 works through Wittgenstein’s suggestion that samples might be treated as instruments of language (see also Essay 13) as well as his discussion of ‘intentionality’; Essay 8 explores Wittgenstein’s treatment of Augustine’s question ‘How is it possible to measure an interval of time?’ and the logical problem of colour-exclusion; and so on. But Baker equally aims to demonstrate his ‘method’ of reading Wittgenstein by means of detailed examples. All of the essays here demonstrate this; but the essays collected together in the second half of Part I all focus, in rather different ways, on that ‘chapter’ of the Philosophical Investigations popularly termed ‘the Private Language Argument’. Like Wittgenstein’s method, Baker’s method must be judged by its results.

Wittgenstein and Waismann Baker’s work on Wittgenstein was paralleled by his extensive work on Friedrich Waismann.16 Waismann’s complex personal and working relationship with Wittgenstein makes his work, in Baker’s view, an important resource for understanding Wittgenstein, if used with caution. And changes in the way Baker read Wittgenstein paralleled changes in the way he read Waismann.17 It is tempting to say that the later Baker came to see many points of contact between the later Wittgenstein and, in particular, the Waismann of the title article of ‘How I See Philosophy’. It would be more accurate to say that Baker came to perceive certain aspects of ‘How I See Philosophy’ and that these aspects were ones which he also came to see in the later Wittgenstein. The essays in Part II of this

6 Introduction

volume are collected together under two headings: the analogy with psychoanalysis, and the focus on aspect-seeing and conceptions (as opposed to concepts).

A The analogy with psychoanalysis Wittgenstein certainly suggested that he saw analogies between his philosophical method and psychoanalysis. These were spelled out more fully by Waismann and are discussed in some detail in the essays in Part II.A.18 As a consequence of seeing Wittgenstein through the lens of this analogy, Baker’s reading highlights certain families of concepts which clearly play a role in Wittgenstein’s texts but are seldom accorded a prominent role in interpretations of Wittgenstein.19 (a) Anxiety (torment, uneasiness, etc.). On Baker’s reading, such talk is not hyperbolic (pace many commentators). More than that: a ‘philosophical problem’ is not simply the cause or object of such torment or anxiety, the problem is the anxiety. Philosophical problems are philosophers’ problems: ‘What we call ‘‘philosophical problems’’ are particular personal disquiets [besondere individuelle Beunruhigungen]’ (MS 115.35). It is this anxiety which is to be ‘treated’ by Wittgensteinian ‘therapy’. The problem is dissolved only when the person is no longer tormented.20 (b) Person-relativity. A further consequence: Wittgensteinian therapy is radically individually oriented. This is the point of the analogy which Baker often draws between Wittgenstein’s therapy and general practice (as opposed to public health campaigns) (see, e.g., Essay 2 and Essay 10). Of course many philosophers do not feel anxious or tormented by their philosophical activity. They may say they engage in philosophy simply out of intellectual curiosity or because they enjoy solving puzzles. Wittgenstein described such philosophers as suffering from ‘loss of problems’ (Zettel §456; cf. Essay 8 n. 12). Though Wittgenstein might well attempt to get them to feel uneasy by asking the right kinds of questions, they may simply refuse to (and ‘refuse’ really is the right word); the only conclusion to come to is that they will gain nothing from Wittgensteinian therapy. (c) The interlocutor. Wittgenstein is in dialogue, not with an ‘adversary’ or ‘opponent’21 (as many commentators, including B&H, suppose), but with an interlocutor who is genuinely in torment and who thus wants to ‘get out of the flybottle’ – even if he cannot see the way, and even if

Introduction

7

at times he seems stubbornly unwilling to take the way out which is offered to him! (d) Unconscious pictures. The source of such uneasiness is often, in Wittgenstein’s view, unconscious analogies or ‘pictures’ or objects of comparison (e.g. analogies between propositions and ordinary pictures, or the picture of words as labels stuck on things). What does it mean to talk about an ‘unconscious picture’? It may be that the person has never explicitly articulated the picture which governs his thinking (e.g., that he is thinking of measuring time on the model of laying a ruler up against it). Or it may be that what is unconscious is that the pictures or analogies have that status; the person may respond, ‘Yes, of course, how else could it be?’ In such a case, therefore, what he is unconscious of is the possibility of alternative pictures or analogies or objects of comparison (Essay 8). In applying ‘our method’, ‘I must always point to an analogy according to which one had been thinking, but which one did not recognize as an analogy’ (BT 408). An unconscious picture thus becomes analogous to continuous aspect-seeing (Essay 1). The person behaves intellectually as if his picture represented the only possibility. (e) Prejudices (dogmas, superstitions, etc.). Thus an unconscious picture comes to function as a dogma or prejudice. Dogmas are characteristically expressed with ‘must’ (‘There must be something in common’, ‘Every word must be a name’; cf., e.g., Essays 4 and 12.) These (likewise ‘cannot’, ‘might’, ‘possible’, etc.) are, of course, modal expressions.22 In fact modality (together with the related concept of essence or nature) is the central concept of metaphysics as traditionally understood – yet, curiously, commentators do not associate Wittgenstein’s claim to be bringing words back from their ‘metaphysical’ to their ‘everyday’ use with this point (Essay 4). Moreover, Baker thinks, twentieth- and twenty-first-century philosophers (by contrast with the Scholastics, of course, but also Wittgenstein himself) are rather insensitive to modal expressions.23 This may seem an astonishing claim to make in our era of sophisticated modal logic. But (1) it is demonstrable that Wittgenstein commentators often ignore his modal expressions (e.g. Essay 2, nn. 2 and 4; Essay 11; Essay 12). (2) It is equally apparent that his bringing possibilities to light is often disvalued, as revealing ‘mere possibilities’;24 new ways of looking at things described as ‘mere changes in notation’ (Essay 2). And (3) there is a tendency to treat metaphysical claims as refutable by facts about word use or by scientific discoveries; but this is to misunderstand them (Essay 12).25 (f ) Motives. Thus what is required is not to confront the philosopher with the ‘facts of grammar’ and inform him that he is deviating from

8 Introduction

ordinary use; he is often well aware that he is doing so! – so this could accomplish nothing. Rather one must explore the motives which drive, him to want to say such deviant things (Essay 4, p. 103; Essay 5, p. 118; Essay 10). (g) Freedom. Both a philosophical problem and its treatment are, in various ways, ‘to do with the will, not with the intellect’ (cf. CV 17). ‘The essence of philosophy is freedom’, says Waismann (see esp. Essay 9). Tyranny (captivity, thrall, bondage) and liberation. Dogmas and prejudices are (paradoxically, self-imposed) restrictions on the individual’s freedom of intellectual movement (CV 28).26 It is then as if such pictures ‘hold us captive’ (PI §115). (‘our goal is to break the thrall in which certain forms of expression [Sprachformen] hold us’ [PPI §113].) Though a tormented philosopher wants to be rid of his torment, he also wants to see things in the way that he sees them (hence, as in psychoanalysis, he may resist therapeutic measures); therapy will involve ‘a kind of conversion’, bringing it about that he no longer wants to see things this way. The ‘dissolution’ of a philosophical problem is thus a form of liberation. ‘We free ourselves from the thrall of the ideal [vom Bann des Ideals] by acknowledging it as a picture and locating its source’ (PPI §107). ‘ . . . the bondage in which one analogy holds us can be broken by placing another [analogy] alongside which we acknowledge to be equally well justified’ (TS 220, §99). Discarding the dogma which acted as a kind of weight tied to the intellectual foot (CV 28) is liberating. (ii) ‘No bullying’. There is, as Waismann says, ‘no bullying’ in philosophy on his conception. Like psychotherapy, Wittgenstein’s method requires acknowledgement (that one’s thinking has been dominated by this or that picture or analogy), not simply as a precondition for successful therapy but as a criterion of correctness for the diagnosis. But this acknowledgement must itself be free, not coerced. Likewise, giving up this picture, and adopting a new picture, are to be done freely, and the person is always free to refuse to do so – and not simply in the sense that one is ‘free to talk nonsense if one likes’ (see esp. Essay 9). Proof and refutation (‘QED’) have no place in this philosophical method. (i)

B

Aspects and conceptions

The very title of Waismann’s article ‘How I See Philosophy’ brings out at least two important features of Wittgenstein’s method, in Baker’s view.

Introduction

9

(i) The first-person pronoun immediately signals that what is at issue here is not ‘the nature of philosophy’ (unless that is understood aright: see Essay 9) but ‘my method’ of doing philosophy, or ‘our method’ insofar as others look at things in this way (e.g. Essays 8 and 9).27 (ii) Thus this article expresses one particular conception (Auffassung) of philosophy. (See Essays 12 and 13 for a clarification of this important concept.) Conceptions are, to put it crudely, the intellectual counterpart of visual aspects. The word ‘see’ (especially coupled with the word ‘how’) stresses that what we have here is a ‘vision’ or a case of something like aspect-seeing, a way of looking at things (hence the title of ‘A Vision of Philosophy’). It tolerates other conceptions of philosophy. More generally, (iii) What we have is a conception of philosophy according to which both philosophical problems and their dissolutions involve conceptions or ‘aspect-seeing’. To speak of ‘aspect-seeing’ here itself involves an analogy between philosophical activity and visual perception (Essay 13). Essay 11 is of interest here precisely because, while focusing on a widely neglected orthographical feature of Wittgenstein’s texts (namely his use of italics), it extracts a partly parallel conclusion, namely that italics often signal a particular conception of this or that concept. ¨ bersichten or ‘perspicuous representations’ are Essay 1 argues that U themselves Auffassungen. Wittgenstein is not in the business of opposing one dogma by another: e.g. opposing ‘The meaning of a word must be the object for which it stands’ with ‘The meaning of a word must be its use’.28 ¨ bersichten conOn the contrary, Wittgenstein’s most famous ‘dicta’ are U sisting in analogies whose aim is to try to get us to look at things in a certain way. (‘Philosophia’ (1986) develops this point for ‘Arithmetical equations are rules of grammar’, Essay 12 for ‘The meaning of a word is its use in the language’, etc.) Equally, others of his ‘dicta’ that B&H viewed as ‘very unhappy’ serve just the same role: e.g. ‘Thinking is operating with signs’ (BB; see Essay 8). If one takes this as a statement of a fact of grammar, it seems to be open to all sorts of objections. If one takes it as Baker argues it was meant to be taken, namely as a new centre of variation that might, by virtue of its novelty relative to the picture of thinking as an inner process, get us to look at thinking differently, it is not objectionable in the same way. Indeed, it is not just the pre-Philosophical Investigations ‘dicta’ which are open to objection if taken as statements of facts of grammar (cf. ‘Philosophia’ (1986), essay 4, p. 95, etc.).29 The claim that ‘Meaning is use’ is both vague and contestable; the idea that arithmetical equations really are rules of grammar likewise, and so on. ¨ bersicht ‘Gestures are part of the (Essay 2 develops this point for the U language’; cf. Essay 12, etc.)

10

Introduction

By the same token, Wittgenstein is not in the business of demonstrating that this or that picture or analogy is mistaken (e.g. Essay 8, pp. 157 ff.). There could be no such business (Essay 9). Though a particular analogy can ‘surround the workings of language with a haze’, it makes no sense to call a picture or analogy ‘false’ in the sense of ‘mistaken’30 (Essay 13). To call something an analogy is to imply that there are disanalogies as well (Essay 1, p. 44). Entrenched pictures therefore cannot be combated by confrontation with ‘the facts’ – any more than metaphysical claims (‘musts’) can; they can be combated only by other pictures (other possibilities; see, e.g., Essay 1, p. 34). Thus Wittgensteinian therapy might in this respect be seen as ‘a kind of homoeopathy’ (e.g. pp. 34, 284). All of this contrasts markedly with many other views of Wittgenstein’s ¨ bersichten is drawn in explicit contrast to method. This interpretation of U what Baker terms the ‘Bird’s-eye View’ model, according to which perspicuous representations are ‘maps’ of the ‘logical geography’ of a concept, simplified so as to be easily surveyable at a glance. Several essays here contrast Wittgenstein’s method and Ryle’s;31 there is, to use two of Baker’s favourite phrases, a serious risk of ‘genre-misidentification’, or of misconceiving the ‘spirit’ of Wittgenstein’s remarks, if we think of Wittgenstein as an ‘ordinary-language philosopher’.32 Baker recognizes that a probable response to this way of looking at Wittgenstein’s method is disappointment: if this really is what Wittgenstein is doing, does he merit the place he has been accorded among the great philosophers? In the first place, it looks as if Wittgenstein is a relativist of the worst possible kind; have we not done away with the idea that this or that philosophical thesis is true or false? Indeed, secondly, he has done away with proofs and refutations. What is left of philosophy, of rational argument, if we do away with these? And in the third place, does it not make philosophy just too easy? If all Wittgenstein is doing is revealing alternative possibilities, where does the work, the boulder-rolling that is the day-to-day business of the professional philosopher, come in? We can trade possibilities ’til the cows come home! In response to the first complaint, we need to make some distinctions. Baker is happy to describe Wittgenstein’s method as ‘relativistic’ (see, e.g., Essay 1, n.29). But if by ‘relativism’ we mean ‘Anything goes’, then this conception of philosophy clearly does not entail relativism (Essay 8, p. 153). Visual aspects are perfectly objective (the duck and the rabbit are there to be seen), cf. Essay 13; just as the fact that a poem or painting admits of more than one interpretation does not entail that all interpretations are equally ‘valid’.33

Introduction

11

In response to the second complaint, it might be said that we need a wider conception of what we are prepared to count as an argument.34 What is at issue is rational persuasion, again analogous to the means of getting someone to see a poem or painting according to a different interpretation or to perceive an unnoticed aspect in an ambiguous drawing. We might propose novel objects of comparison (PI §131), describe simplified versions of what we want the person to see (some of Wittgenstein’s language-games have this function), construct intermediate cases, offer new centres of variation or correct the one-sidedness of the diet of examples. A question (e.g., ‘How long before you execute an order must you understand it?’) may effect an aspect-switch (see, e.g., Essay 3, p. 76 ff.). Many such techniques are canvassed here (e.g., Essay 1, pp. 30–1). What we need to combat in ourselves is the prejudice according to which any form of persuasion that is not demonstrative is non-rational.35 In response to the third complaint, the idea that ‘possibilities’ are ‘mere possibilities’ ignores an important point about changes in visual aspects: that though in one sense nothing is changed, in another, everything is. What could be more important than changing someone’s way of looking at things? (For example, to drop the dogma of psychophysical parallelism or the idea that every disposition must be realized in a structure would be ‘a tremendous thing to do’ (UW 434; Essay 2, n. 10).) As for the idea that we can ‘trade possibilities ’til the cows come home’, it may take a genius (of Wittgenstein’s standing) to reveal alternative possibilities in a clear and compelling manner.36 Indeed Baker argues that the philosophical task he envisages is far more delicate and ambitious than the task of arriving at correct analyses of concepts or even producing a comprehensive philosophical grammar. It certainly requires more sympathy, imagination and patience. To change someone’s way of looking at things requires a radical change in his whole intellectual life: the replacement of entrenched habits of thinking (Denkgewohnheiten [BT 423]) by new ones. There cannot be an algorithm for getting someone (a particular troubled philosopher who has got himself into his predicament in a particular way) to change his way of looking at things. The methods for accomplishing this are as various as methods for getting someone to perceive a hitherto unnoticed aspect in an ambiguous drawing. And they are as little guaranteed of success (see, e.g., p. 283.). Obviously, those philosophers with a temperamental preference for the ‘clash of steel upon steel’ and an attachment to the imprimatur ‘QED’ will find this conception of philosophy uncongenial. That, Baker will remind us, is no reason for refusing the label ‘philosophy’ to it.

12

Introduction

Further Directions: History of Philosophy37 This conception of philosophy ought to lead us to read past philosophers differently. At the least, it suggests that we need to attempt to uncover their ‘vision’ of things (Essay 10, p. 208), and this requires us (at the least) to explore their use of words (Essay 9, n. 43), to know enough about their background and intellectual context to have some idea what problems they were facing and what they could and could not take for granted. (‘There is no such thing as grasping the significance of a philosophical thesis independently of understanding the architectonic of the thought of its author’ (‘Philosophia’ (1986), p. 45).) It requires us to show them sufficient respect to prefer, at least ceteris paribus, the hypothesis that we have not yet understood their conception of things to the idea that they were guilty of a stupid error or fallacy. We need to read with imagination, sympathy and openness to alternative ways of looking at things. On this conception of philosophy, ‘philosophical understanding is essentially historical’.38 It might also lead us to engage with other readers of those philosophers differently as well. Interpretations of past texts are expressions of aspects or patterns; and it often happens that one particular interpretation will seem so compelling that it becomes difficult for others to look at the text differently, or at least significantly differently. Ryle’s interpretation of Descartes and Dummett’s interpretation of Frege are cases in point for Baker.39 More locally, the standard interpretation of the so-called ‘Private Language Argument’ as an anti-Cartesian reductio ad absurdum grips people in the same way (see esp. Essay 5). What is needed here is to go through the kinds of therapeutic moves that Wittgenstein himself went through and to present an alternative way of looking at the text sufficiently clearly and compellingly to break the grip of the standard picture that exerts an influence on our reading of the text. Indeed, it might even lead to a new conception of the relationship between philosophy and the history of philosophy. There is an inclination to think that so-called philosophical problems are timeless: that though Descartes, for example, may have given a different answer to the question ‘How are mind and body related?’ than many modern philosophers, at least he was clearly addressing the same question as we are today. From this point of view, history of philosophy might be of instrumental use in allowing us to extract whatever nuggets of gold are to be found in past writings and to learn from others’ mistakes. Yet if the ‘vision of philosophy’ presented here is taken seriously, we might conclude that the questions that past philosophers are addressing are not ours, in part because their conceptions of the relevant concepts (e.g. ‘mind’ and ‘body’, but also, e.g., ‘substance’ and

Introduction

13

‘cause’) are different from ours. (Collingwood saw this very clearly.40) Here history of philosophy might play a quite different role vis-a`-vis philosophy itself: if part of what is required for curing someone of his philosophical distress is getting him to see clearly alternative ways of looking at things, then the history of philosophy might serve as an aid to the imagination: if another highly intelligent philosopher not only looked at things very differently but developed that alternative vision in some detail, understanding that philosopher’s vision might serve an important therapeutic function.

Envoi: A Wittgensteinian Reading of Wittgenstein? The essays in the present collection see themselves as giving a Wittgensteinian reading of Wittgenstein. The reading which Baker offers, as already detailed, tries to uncover Wittgenstein’s vision of things through careful attention to what he says (and does not say). This reading stands in stark contrast to other interpretations which have seemed absolutely compelling (e.g. that of B&H), so much so that it can become difficult for others to read Wittgenstein’s text in any other light; Baker’s alternative reading might help to free readers from the grip of that particular picture. And, just as reading past philosophers who look at things differently may serve as an aid to the imagination in envisioning alternative possibilities (say, for conceiving of ‘mind’ and ‘body’), so might reading Wittgenstein (through Baker) serve as an aid to the imagination in envisioning alternative ways of doing philosophy. I stressed earlier the centrality of close attention to the text as characterizing Baker’s way of reading Wittgenstein. But whereas mere careful reading may allow the respectable scholar to see what is there, what Baker’s careful reading reveals is patterns or aspects. Hence he scrupulously distances himself from the idea that a ‘careful reading’ of Wittgenstein’s text demonstrates that this text must be read in the way that Baker reads it. Indeed he would be the first to acknowledge that many passages are prima facie hard to reconcile with his reading. The idea that it is a Wittgensteinian reading of Wittgenstein is itself contestable, as again Baker would readily acknowledge. Baker sees himself as offering – not the definitive, or the only possible, interpretation of Wittgenstein, but – an alternative way of looking at Wittgenstein. So entrenched is the conception of philosophy he is combating that philosophers are apt to find this unsatisfying: as if only a demonstration that this is right and every other interpretation is wrong is worthwhile. I trust that by the end of this volume they will no longer be inclined to react that way.

14

Introduction

Notes 1 2

3

4

5

6

7

8

Indeed, Patrick Nowell-Smith did. NB. Many of Baker’s essays adopt the convention of using italics in a quotation to mark the author’s own emphasis, and using bold to mark added emphasis. I adopt the same convention here. This, as Baker himself notoriously used to say about Michael Dummett’s Frege: Philosophy of Language and Bernard Williams’s Descartes: A Project of Pure Inquiry, goes wrong in the very title, from the later Baker’s point of view. These co-authored works include Frege: Logical Excavations (1984); Language, Sense and Nonsense (1984); Scepticism, Rules and Language (1984); and the multiple exchanges with Dummett in Philosophical Quarterly (‘Dummett’s Purge: Frege without Functions’, vol. 33, 1983; ‘Dummett’s Dig: Looking-Glass Archaeology’, vol. 37, 1987; ‘The Last Ditch’, vol. 39, 1989) and Mind (‘Dummett’s Frege, or Through a Looking-Glass Darkly’, vol. 92, 1983). Middle-period single-authored articles include ‘Following Wittgenstein: Some Signposts for Philosophical Investigations §§143–242’ (1981). As Baker himself would be the first to stress, identifying ‘early’, ‘middle’ and ‘later’ Baker is a matter of pattern-recognition and is as contestable as any other case. There are, moreover, many ‘new’ elements in some of the ‘middle’ work. Baker had certainly begun to break off from B&H by 1986, as the final footnotes in ‘Alternative Mind-Styles’ and the revised version of this article called ‘Philosophia: Eikon kai Eidos’ make clear. NB. These two articles largely overlap; ‘Philosophia’ contains a new section (§6) that certainly rings in the Later Baker key. By contrast, ‘Moderne Sprachtheorien aus philosophischer Sicht’ is simply a summary of Language, Sense and Nonsense and is pure B&H. (In fact an earlier version was jointly presented by Baker and Hacker at a conference.) The later Baker would never write ‘The [linguistic] theorists’ strategy has all the virtues of theft over honest toil, and a philosopher has a duty as a policeman to deprive them of the enjoyment of their ill-gotten gains’ (from the English version of ‘Moderne Sprachtheorien’). Though Wittgenstein, Frege and the Vienna Circle (1988) is largely ‘later Baker’. This in effect consists of two long essays, one on ‘Philosophy of Logic’, one on ‘Varieties of Conventionalism’. Though there are – independently developed – affinities (Essay 4, n.2). Baker reviewed Alice Crary and Rupert Read (eds), The New Wittgenstein (2000); for various reasons this review remains unpublished, but it clarifies some points of disagreement between his reading and theirs. Much of this review focuses on their interpretation of the Tractatus and Frege (and is hence of less relevance to the present collection); his principal complaint about their handling of the later Wittgenstein is their overly programmatic treatment of Wittgenstein’s ‘distinctive and multi-faceted vision of philosophical therapy’. These two themes, as well as the application of ‘our method’ to the history of philosophy (infra), were picked out by Baker himself, as the outline of a projected book on Wittgenstein’s method.

Introduction 9

10

11

12

13

14 15

16

15

Perhaps his own training as a classical scholar was influential here. Baker majored in mathematics as an undergraduate at Harvard; he came to Oxford to do a second BA and read ‘Greats’ despite not having Greek, which he taught himself over the summer before his course began. Baker does not bring this knowledge to bear in any way that ought to worry a sophisticated critic of the ‘intentional fallacy’. Baker’s point is not that we need to know this or this about Wittgenstein’s biography or the archaeology of a particular remark in order to understand it. It is rather that, though (for example) the target of the remarks that are taken to constitute the Private Language Argument is visible in the text, it may be more easily visible if one has read Russell – just as the rabbit aspect of the duck-rabbit drawing may be made more easily visible by surrounding it with unambiguous pictures of rabbits. Also ‘Quotation-marks in Philosophical Investigations Part I’ (2002). I have not included this in the present volume on the grounds that it makes the same methodological point as ‘Italics’, namely the importance of thinking about Wittgenstein’s use of such devices. A number of Wittgenstein scholars who are generally sympathetic to Baker’s approach find ‘Italics’ uncongenial. What Baker saw in it was this: it presents a kind of ‘research programme’ for Wittgenstein scholars (and Baker was sensible of the irony here). It is one that might be developed in two different directions: on the one hand, attention to other orthographic devices (cf. Baker’s ‘Quotation Marks’, 2002); on the other, the detailed application, to particular problematic passages, of the general remarks in ‘Italics’. Baker did the latter in the unpublished draft ‘To Follow a Rule is a Custom’. He argues for an alternative reading of this celebrated passage that takes the italics, not as a kind of orthographic hammering on the table, but as indicating ‘In a certain sense’; Wittgenstein, absolutely characteristically, leaves some intellectual work to the reader to work out in which sense. Jean-Philippe Narboux called attention to Baker’s stress on the notion of ‘dimension’ at the hommage for Baker (Institut de l’Histoire de Science et Philosophie, Universite´ de Paris, March 2003) organized by Antonia Soulez. Descartes’s Meditations is an obvious and shining exception. Those who heard Baker lecture in French might well see ‘Il nous faut repartir a` ze´ro’ (‘We need to start over again from scratch’) as his credo; it occurs in a number of the essays in this volume. See his ‘Nachwort’ to Waismann’s Logik, Sprache, Philosophie (1976) which Baker also co-edited, and the preface to the second edition of the English translation of this book, The Principles of Linguistic Philosophy (1977); ‘Verehrung and Verkehrung: Waismann and Wittgenstein’ (1979); Baker’s ‘Pre´sentation’ of the Dicte´es de Wittgenstein a` Waismann et pour Schlick, vol. I: Texte (1997) and his introduction to the English translation of these dictations, The Voices of Wittgenstein (2003), both of which he also co-edited. See also ‘A Vision of Philosophy’ and ‘Wittgenstein’s Method and Psychoanalysis’, in this volume, and ‘Friedrich Waismann: A Vision of Philosophy’, forthcoming.

16

Introduction

There are pertinent differences between the earlier and the later discussions of the relationship between Wittgenstein and Waismann. For example, in the 1979 piece in the Luckhardt volume (‘early Baker’ by my categorization, though containing much of great interest), Baker refers to ‘a shift in emphasis in the theory of meaning from hypothesis to criteria’ (p. 255) and to ‘development in the philosophy of mathematics’, away from ‘constructivism’ toward ‘ ‘‘full-blooded conventionalism’’ ’ (p. 271). The preface to the second edition of PLP (‘later Baker’) puts scare-quotes around ‘ ‘‘philosophy of language’’ ’ (p. xix), no doubt to distance himself from the idea that Wittgenstein had a ‘theory of meaning’. Again, whereas a major theme of the Luckhardt piece was that Wittgenstein’s willingness to collaborate with Waismann ‘is the single most effective antidote to the obscurantist dogma that Wittgenstein set his face against any systematic presentation of his philosophical insights’ (p. 280), the PLP Preface refers to PLP as an attempt to ‘do something that Wittgenstein constantly resisted, namely to codify, at least loosely, the methods he used in tackling philosophical problems’ and comments that ‘in any more systematic exposition of his ideas the freshness of Wittgenstein’s approach to each new problem will be lost’ (p. xxii). 17 Baker liked Waismann’s description of working with Wittgenstein: ‘He has the great gift of always seeing things as if for the first time. But it shows, I think, how difficult collaborative work with him is, since he is always following up the inspiration of the moment and demolishing what he has previously sketched out’ (quoted in Baker ‘Verehrung and Verkehrung’ (1979), p. 256). Some might describe working with Baker in similar terms! 18 It would be interesting, though well beyond the scope of this introduction, to compare the later Wittgenstein’s method on the later Baker’s reading to Wisdom’s. Dr Eugen Fischer, a former student of Baker’s, has developed in some detail a therapeutic conception of J. L. Austin’s method (in his Habilitazionsschrift: Philosophical Delusion and its Therapy). 19 The reader will note that the index to this volume highlights these concepts as well. 20 Though it does not follow that any method of getting rid of the anxiety (e.g. pills or a knock on the head) counts as a dissolution of the philosophical problem! Like psychoanalysis, Wittgensteinian therapy is a ‘talk-cure’, cf. Essay 8. 21 This models the dialogue of the Investigations on a battle or a courtroom trial. Other commentators model Wittgensteinian dialogue more on a Socratic dialogue than on a courtroom trial. Here the relationship between Wittgenstein and his interlocutor is at least friendly, and the interlocutor is often not entrenched in his position. The appropriateness of the analogy depends in part on which Socratic dialogues one has in mind. Many Socratic dialogues aim for truth; what may seem most alien to other philosophers is that truth has on Baker’s reading no role to play in Wittgenstein’s method of philosophizing; see below.

Introduction 22 23 24

25

26

27

28

29

30

31

17

It is noteworthy that Baker came increasingly to stress that the ‘Augustinian picture of language’ was a picture of the essence of language (e.g. Essay 12). Also ‘Russell and Wittgenstein on Everyday Use’, forthcoming. Yet from the perspective of modal logic, the negation of ‘‘ p’ is equivalent, not to :p, but to :p. I have argued elsewhere that Wittgenstein’s efforts are directed, in effect, towards persuading the person that :p (Morris, ‘The ‘‘Context Principle’’ in the Later Wittgenstein’). An empirical generalization like ‘All swans are white’ (at least insofar as this is understood as an empirical generalization) is refutable by producing a black swan; but a metaphysical claim like ‘There must be something in common’ is not refutable by producing an example of a concept whose instances lack anything in common – what is at issue is precisely whether there could be any such examples. This is not (of course) to deny that :p entails :‘p! But someone persuaded that there must be something in common will deny the description of the apparent counterexample as ‘lacking anything in common’. Relatedly, Wittgenstein will talk of ‘temptations’, ‘inclinations’ and so forth. These locutions appear in Philosophical Investigations; they are completely unmissable in the Blue Book. Baker had a great deal more time for the Blue Book than many commentators (and had begun an article on this prior to his death); it is strange that it managed to get labelled as ‘Preliminary Studies’ for Philosophical Investigations, but this fact may help to account for other commentators’ tendency to downplay the Blue Book. As n. 5 of Baker’s Preface to the second edition of PLP points out, ‘Our Method’ was ‘the earlier title of section 2 of chapter IV (now called ‘‘Language Games’’)’. Contrast again the middle-period ‘Moderne Sprachtheorien’ (1986). In effect it reads Wittgenstein’s ‘Essence is expressed by grammar’ (PI §372) as a licence for saying that modern linguistic theory misdescribes the essence of a rule and corrects that misdescription (e.g., ‘[Rules] are used as standards of conduct, guides to behaviour and norms for its evaluation. There is no such thing as a rule which has no such role, a fortiori no such thing as a rule which could have no such role’ (p. 17 of English typescript)). This general strategy of interpretation is explicitly rejected in Essays 4 and 12. In fact they have been the subjects of a great number of objections from critics; how do those self-styled Wittgensteinians who attempt to defend these ‘dicta’ against such objections reconcile their activity with Wittgenstein’s insistence that if anyone objects to something he says he will withdraw it? Andrew Lugg (Wittgenstein’s Investigations, §§1–133) is rare among commentators in recognizing this important point. Wittgenstein was often tempted to use phrases such as ‘false analogy’ or ‘misleading picture’, but he also regularly indicates misgivings about these expressions by putting wavy lines under ‘false’ or ‘misleading’ in these phrases (e.g. MS 110: 300, BT 409). A further point: the idea that ‘describing grammar’ is describing rules for the use of words in (say) English. Wittgenstein was not on Baker’s conception

18

32

33 34 35 36

37

38

39

40

Introduction policing the borders between sense (¼ ordinary use) and nonsense (¼ nonordinary use) and arresting those who crossed those borders. Consistent with his individual-oriented therapy, what interested Wittgenstein was not whether or not the individual used words as the rest of us do but how that individual wished to use words. ‘Moderne Sprachtheorien’ (1986) accuses linguistic theorists of multiple ‘category-mistakes’, marking it indelibly as middle-period Baker (as already noted) despite its relatively late date. Baker cites Nietzsche’s perspectivism with approval in Essay 13. Here we have an important point of contact between Baker and the so-called ‘New Wittgensteinians’, especially Cora Diamond. Pace H.-J. Glock’s comments on what he likes to call the ‘No-Position Position’ (e.g. in A Wittgenstein Dictionary). Baker stressed the danger of philosophers’ ‘failing to raise enough questions, or failing to put the question marks deep enough down’ (Essay 9). He used to like to say that philosophy was (or ought to be!) the activity which provided ‘a question for every answer’. This section characterizes, inter alia, what we tried to do in our work on Descartes. Whereas the middle-period Baker treated Descartes as the b^ e te noire of philosophy, the later Baker defended Descartes vigorously against the attacks of the soi-disants ‘Wittgensteinians’, including B&H. That he termed this a ‘Wittgensteinian’ reading of Descartes (Baker and Morris, Descartes’ Dualism, p. 206, n. 21) must have puzzled those unfamiliar with the later Baker! I hope that this section will help to shed light on that remark. See also Essay 5. Baker, Wittgenstein, Frege and the Vienna Circle (1988) p. xv. He explicitly contrasts this aspect of this conception with the view of the business of philosophy as delineating logical geography. On Descartes, see our Descartes’ Dualism (1996); on Frege, see Wittgenstein, Frege and the Vienna Circle (1988), ‘ ‘‘Function’’ in Frege’s Begriffsschrift: Dissolving the Problem’ (2001). Russell is another author on whom Baker worked a great deal; he lectured on Russell on arithmetic for a number of years. See his ‘Bertrand Russell’ (1988) and ‘Russell and Wittgenstein on Everyday Use’ (forthcoming). See his Autobiography (1939). Katherine J. Morris Mansfield College Oxford University

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

Part I Reading Wittgenstein

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

A

Methodological Concepts

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

1 Philosophical Investigations §122: Neglected Aspects*

A main source of our failure to understand is that we do not command a clear view of [u ¨ bersehen] the use of our words. – Our grammar is lacking in this sort of perspicuity ¨ bersichtlichkeit]. – A perspicuous representation [eine [U u ¨ bersichtliche Darstellung] produces just that sort of understanding which consists in ‘‘seeing connections’’. Hence the importance of finding and inventing intermediate cases. The concept of a perspicuous representation is of fundamental significance for us. It earmarks the form of account we give [unsere Darstellungsform], the way we look at things. (Is this a ‘‘Weltanschauung’’?) (PI §122)

First Impressions This text is well known and often quoted. It seems to condense into one short remark much of Wittgenstein’s distinctive conception of philosophy. It lodges a complaint about ‘the grammar of our language’ (a lack of perspicuity), and it suggests that he took as a primary goal remedying this defect in our understanding by providing ‘representations’ of grammar that would make things perspicuous. The search for perspicuity (Durch¨ bersichtlichkeit) is a leitmotif of his later philosophy, clearly sichtigkeit or U audible from the opening of Philosophical Remarks to the close of Last Writings. It must be an integral part of the philosopher’s business of * Originally published in R. L. Arrington and H.-J. Glock (eds), Wittgenstein’s ‘Philosophical Investigations’: Text and Context, pp. 35–68. London and New York: Routledge, 1991. Reproduced by permission.

Philosophical Investigations §122 23

describing the actual use of language (PI §124). Indeed, perspicuity might be taken to be the single ultimate end of all of the activities of the philosopher (cf. §127), the goal to which the dissolution of all particular philosophical problems is subordinate (cf. §§132–3). Though clearly important, Wittgenstein’s concept of a perspicuous representation is not itself perspicuous. This is hardly surprising since it is here introduced with minimal clarification and without a single example. Originally he cited the colour-octahedron as a paradigm (PR 51–2; cf. WWK 42 and PR 278). He regarded this three-dimensional model (or the corresponding two-dimensional diagram) as a ‘representation’ of the grammar of colours. From it we can ‘read off ’, for example, the possibility of a spectrum of shades between red and yellow (hence also the intelligibility of the phrases ‘reddish yellow’ and ‘yellowish red’) and the polar opposition of red and green (i.e. the nonsensicality of the phrase ‘greenish red’). The entire grammar of the network of colourwords seems to be condensed into a diagram that can be taken in at a glance (i.e. a diagram that has the attribute of being surveyable (u ¨ bersichtlich or u ¨ bersehbar)). Conversely it seems as if this whole assemblage of rules can be reconstructed by close inspection of the diagram. Here a sizable domain of grammar seems to be laid up or contained in a single sign (cf. PG 55–6). The power to condense something complex into a simple and manageable symbol seems to be the defining characteristic of what Wittgenstein called ‘a perspicuous representation’. Like the perspicuity which is said to be essential to a proof in mathematics (RFM 95, 143–53, 174), the perspicuity of a representation is apparently to be explained by reference to its memorability and to the possibility of copying or reproducing it. Having inspected the diagram of the colouroctahedron once, most people can draw it again for themselves and make use of it to work out further relations among chromatic colours (e.g. that ‘bluish red’ makes sense, but not ‘bluish yellow’). The colour-octahedron is in fact the sole labelled instance of a perspicuous representation (of grammatical rules) in all of Wittgenstein’s published writings. This generates two major quandaries. First, how is the absence of anything closely comparable to the colour-octahedron from the text of the Investigations compatible with the claim ‘The concept of a perspicuous representation is of fundamental importance for us’? The reader might expect explicit applications of this concept to characterize Wittgenstein’s philosophizing in the same way that Ryle’s criticism of the myth of the ghost in the machine is punctuated by the pinpointing of ‘category-mistakes’. Does the lack of diagrammatic representations of conceptual connections from the text prove that the remark on perspicuous representations is a fossil remnant of an extinct conception of

24

Reading Wittgenstein

philosophy? Should we argue that the term ‘perspicuous representation’ covers a family of ‘representations’ of the uses of symbols and then go on to identify numerous unlabelled instances in the text? There is a prima facie tension in the Investigations which any satisfactory interpretation of PI §122 must seek to resolve. The second quandary is the evident underdetermination of the grammar of the phrase ‘a perspicuous representation’ by the general explanation given in §122, even when this is supplemented by the example of the colour-octahedron (as in PR 52). In fact there are two different directions in which these embryonic explanations of ‘perspicuous representation’ could be expanded. The first conception lays stress on the concept of a representation (Darstellung). It is ‘our grammar’ or ‘the use of our words’1 which is said to lack perspicuity. Hence, presumably, it is ‘a perspicuous representation of our grammar’ which is sought to remedy ‘our failure to understand’. This phraseology may be taken to indicate a contrast between ‘our grammar’ and ‘representations of our grammar’. The expectation of finding a contrast is fulfilled in Wittgenstein’s characterization of the colouroctahedron as ‘a perspicuous representation of the grammatical rules [sc. for the use of colour-words]’ (PR 52). Here the purpose of the diagram seems to be producing an order within the motley collection of explanations of what ‘red’, ‘green’, ‘yellow’, etc. mean. Wittgenstein did not avow the intention of substituting the colour-octahedron in place of explanations of the combinatorial possibilities of colour-words, but rather the intention of using the diagram to make perspicuous (to introduce system into) this set of grammatical rules. We could even state the principle for projecting the diagram on to these grammatical rules, i.e. a method of representation (e.g. that the vertices marked ‘red’ and ‘yellow’ are connected by a single edge implies that shades of red can be ordered on the scale ‘more or less yellow’ and shades of yellow on the scale ‘more or less red’2). Consequently there is a straightforward and important sense in which the colour-octahedron is a representation (or even ‘picture’) of grammatical rules. If we follow up this line of thought, we will conclude that what Wittgenstein called ‘a perspicuous representation of grammatical rules’ must have the characteristic that the representation is distinct from what is represented. (Neither identity nor permutation count as what we call ‘a method of representation’.) In consequence, no verbal formulations of grammatical rules (and no assemblage of grammatical rules) can properly be called ‘a perspicuous representation of grammatical rules’, and conversely the colour-octahedron cannot properly be viewed as a mere compendium of the combinatorial rules for colourwords. An advocate of this interpretation might summarize his conception

Philosophical Investigations §122 25

in the claim that what Wittgenstein called a perspicuous representation (Darstellung) of grammatical rules is essentially different from a perspicuous arrangement (Zusammenstellung) of grammatical rules. The second conception of a perspicuous representation is the polar opposite of the first. It seems to be the dominant interpretation offered by commentators who follow Wittgenstein in assigning a fundamental importance to the concept of a perspicuous representation in his practice of philosophizing. This second interpretation might be developed from the description of the colour-octahedron as ‘a rough representation of colour-space, . . . a grammatical representation, not a psychological one’ (PR 51). Apparently he meant to contrast the a priori relations among colours with relations that might be established by empirical investigation (e.g. that a red after-image may be produced by looking at a white wall after staring hard at a green circle (cf. PR 51–2)). This suggests that the role of the colour-octahedron is to describe colour-space by presenting the grammar of colour-words. But an enumeration of combinatorial rules for colour-words (e.g. ‘There is no such thing as a reddish green’, ‘A shade of red may be more or less yellow’) has exactly the same role. Consequently, the diagram and the assemblage of combinatorial rules may each be called ‘a grammatical representation of colour-space’ or ‘a description of the use of colour-words’; the diagram simply differs in virtue of being more comprehensive than any one rule (or any succinct list of rules). The hallmark of what Wittgenstein called ‘a perspicuous representation’ is that it is a presentation or arrangement of grammatical rules which can be taken in as a whole. An exponent of this second interpretation might epitomize his conception by claiming that a perspicuous ‘representation’ is nothing more than a particular kind of arrangement (Zusammenstellung) of descriptions of ‘the use of our words’. This second interpretation might be reinforced by noting that somebody who knows how to make use of the colour-octahedron can ‘read off from it’ all of the combinatorial rules for chromatic colours. At the same time, this same thought seems to highlight the crucial difference between a diagrammatic representation of colour-space and a piecemeal description by the enumeration of many rules. In the first case, this whole domain of grammar is depicted in a single surveyable symbol, whereas in the second the full description will be impossible to take in at a glance. It would perhaps be less misleading to claim that Wittgenstein intended the colouroctahedron to be a formulation of the grammar of our language than to claim that he meant it to represent or describe this grammar. Moreover, since he often called grammar itself ‘a form of representation’, this point perhaps explains why he called a perspicuous representation ‘a form of representation’ (cf. PI §122).3 It must be seen as an element of grammar

26

Reading Wittgenstein

or as a member of the set of ‘descriptions of our language’ (just as an architect’s blueprint is to be regarded as a description of a building). Within this genus, perspicuous representations are differentiated by being surveyable sets of rules which constitute complete explanations4 of how to use ‘the words of our language’. For the moment we shall drop the first interpretation and explore the second one. By following up and elaborating its account of the role of the colour-octahedron, we do arrive at a conception of perspicuous representations which has the merit of giving them a central place in Wittgenstein’s account of the nature of philosophy.5 On this view, the general aim of the philosopher must be to produce surveyable descriptions of the uses of words which have a high degree of comprehensiveness and which can therefore be employed to clarify sizeable domains of grammar and to dissolve many different philosophical problems all at once. What is required is the careful selection and arrangement of grammatical trivialities (‘quiet weighing of linguistic facts’ (Z §447)) to bring out features of the uses of symbols of which we have lost sight in the welter of everyday explanations of what symbols mean. The philosopher reminds us of familiar facts which have slipped out of our field of attention, and he directs our thinking into less convoluted paths by plotting a grammatical nexus embracing a wide range of related concepts (primarily by producing schemata that avoid confusing detail).6 The production of perspicuous representations is the most general characterization of the proper goal of philosophy once we have come to acknowledge that philosophy can only describe the grammar of our language (PI §§109, 124). The philosopher strives to command a clear view of ‘the grammar of our language’. His success is comparable to someone’s looking down on a city from a height, thereby commanding a clear view of the layout of its streets and squares. The philosopher has the additional task of giving a clear description of what he surveys. Here his success is comparable to a cartographer’s producing an aerial map. The analogy embedded in Wittgenstein’s discourse about perspicuous representations is a bird’s-eye view of a city. In describing ‘the grammar of our language’, he is allegedly engaged in ‘logical geography’ a` la Ryle, but he aimed to present his findings in a distinctive style of conceptual map. This second conception of the nature of perspicuous representations (which I will refer to as the Bird’s-eye View Model) has several important corollaries. First, it suggests that Wittgenstein left room for a positive role for philosophy which stands in contrast and supplements its predominantly negative or therapeutic task.7 Though he treated particular problems like illnesses (§255) and applied various particular therapies to make them disappear completely (§133), his ideal of constructing perspicuous repre-

Philosophical Investigations §122 27

sentations seems to indicate a striving towards a more systematic philosophy.8 Did he not envisage descriptions of some domains of grammar which would provide accurate maps and thereby constitute a permanent prophylaxis against whole sets of philosophical problems? For example, he seems to have aimed at a general classification of ‘psychological concepts’ in which the correct treatment of each would throw light on all and thereby secure an understanding of conceptual connections that would dissolve many of the problems of philosophy of mind (cf. Z §§464–5).9 Second, a perspicuous representation of grammar is evidently to be contrasted with the various ‘mythologies of symbolism’ that philosophers are inclined to elaborate (PG 56). This contrast might suggest that the aim is simply a correct description of the uses of words (‘the account books of language’ (PG 85)) which will explode various mistaken philosophical theories, ranging from Platonism and intuitionism in philosophy of mathematics through Cartesian dualism and behaviourism in philosophy of mind to logical atomism and the picture-theory of the proposition in philosophy of logic. Wittgenstein often suggested that these theories can be traced to a more fundamental set of ‘grammatical illusions’ (PI §110), which are linked by ‘Augustine’s conception of language’ (§4, cf. §1). His intention was not to replace one mythology of symbolism by another parallel theory (e.g. to replace Platonism by ‘strict finitism’ or Cartesian dualism by behaviourism); not even to replace ‘Augustine’s picture’ by some alternative generalizations about the nature of language (e.g. ‘anti-realism’ or ‘semantic idealism’). Instead, he aimed at a surveyable description of ‘our grammar’ which neutralizes the appeal of all standard philosophical theories. On his view perspicuous representations are to stand on a different level. They make no pretensions to explain ‘the uses of our words’. Hence they abstain from metaphysical fantasies and stick to describing the down-to-earth, ordinary (hausbacken, gewo¨hnlich) uses of ‘our words’ (TS 213, 412). Their function is to destroy idols or to clear the ground of derelict structures and rubble. Whereas Platonism or Cartesian dualism might be described as ‘ways of seeing things’ (indeed, as poor ways of seeing things which generate philosophical puzzles), a perspicuous representations does not ex officio embody any point of view.10 (In this regard it resembles a report of a visual perception rather than a description of a visual aspect of things.) Third, according to this interpretation, it is a pleonasm to say that the subject-matter of a perspicuous representation is ‘the grammar of our language’ or ‘the use of our words’. Nothing other than descriptions of grammar are even candidates for being called ‘perspicuous representations’. Consequently, there is no possibility of identifying any form of scientific theory, any hypothesis about historical development, or any

28

Reading Wittgenstein

observation about a religious ritual or a work of art as ‘a perspicuous representation’. In particular, it would be a mistake to call Goethe’s scheme of plant morphology (‘All the organs of plants are leaves transformed’) or Darwin’s evolutionary scheme for the classification of species (‘Taxonomy recapitulates phylogeny’?) ‘perspicuous representations’, at least in the context of clarifying the employment of this expression in Wittgenstein’s method of philosophizing.11 From this sketch of the Bird’s-eye View Model, we can milk out an account of the use of ‘perspicuous representation’ in Wittgenstein’s writings: (i)

(ii) (iii)

(iv)

(v)

(vi)

A perspicuous representation is an ordering or arrangement of grammatical rules or of descriptions of ‘the use of our words’. It is tautologous to add ‘of grammar’ to the phrase ‘a perspicuous representation’ (just as it would be to add ‘of a person’ to the phrase ‘a portrait’). What it represents is exactly what its components severally describe, namely the employment of symbols of ‘our language’. The adjective ‘perspicuous’ in the phrase ‘a perspicuous representation’ is used attributively. It ascribes a property to a particular arrangement of grammatical rules, namely that they can be taken in at a glance, remembered with comparative ease, and reproduced with a minimum of errors. The adjective ‘perspicuous’ lacks any comparative form in the phrase ‘a perspicuous representation’. A description of grammar which does not meet the appropriate criteria is simply not a perspicuous representation just as a sequence of mathematical propositions which does not exhibit perspicuity is simply not a proof. One representation of grammar cannot be ‘more (or less) perspicuous’ than another any more than one axiom of geometry can be more (or less) self-evident than another. There is no such thing as a mode of representation or a way of looking at rules of ‘our grammar’. If there are different perspicuous representations of a single domain of grammar, they differ merely in the selection and arrangement of grammatical rules; they would be different orderings ‘in our knowledge of the use of language’. Perspicuous representations are roughly additive. By pasting together a map of Buda and a map of Pest we obtain a map of the conglomeration called ‘Budapest’. Descriptions of grammar are similar in this respect; two of them can always be amalgamated into a more comprehensive whole. For perspicuous representations we must add a caveat since the combination of two individually

Philosophical Investigations §122 29

surveyable descriptions may overstep the threshold of what can be taken in at a glance; perspicuity could then be restored only by some further simplification or schematization. (vii) The criteria of identity for perspicuous representations have an indeterminacy which parallels that for such concepts as explanations of what signs mean or descriptions of character. A fragment of a description of ‘our grammar’ may or may not qualify as such a description on its own; certainly a fragment of a perspicuous representation may drop below the threshold of completeness necessary for a perspicuous representation. Conversely, any description of ‘the use of our words’ (or any arrangement of grammatical rules) may be swallowed up in a more embracing one. There is no clear answer to the question ‘How many perspicuous representations are there in the Investigations?’, just as there is none to ‘How many descriptions of character are there in Emma?’ (viii) There are equally no clear criteria for the success or adequacy of a perspicuous representation. Presumably each component, i.e. each individual grammatical rule, must be correct,12 but in addition the whole set have a certain degree of comprehensiveness or completeness. This crucial vagueness might give ‘perspicuous representation’ many of the features of an essentially contested concept. (ix) Although to somebody at some time ‘the uses of our words’ may not be perspicuous, whether globally or locally, it seems that it must be possible for anybody to command a clear view of the use of any word or family of words (e.g. the use of ‘psychological concepts’). This seems to be a corollary of the insight that there are no discoveries to be made in grammar given that it consists of an array of rules which competent speakers follow in speaking a language. Since nothing is hidden from us, each of us can in principle construct a map of any domain of ‘our language’, however extensive it may be. ‘It is not a contingent feature of language that its grammar is surveyable.’13 The question to be addressed here is whether this description of its grammar squares with the actual use of ‘perspicuous representation’ in Wittgenstein’s language. Close examination of the immediate progenitor of §122 gives some grounds for serious doubts. Careful exploration of the context of this remark leads us back to the first antithetical interpretation of his conception of a perspicuous representation. This in turn opens the way to realizing that his writings are rife with examples of what he called ‘perspicuous representations’. This point is not merely scholastic.

30

Reading Wittgenstein

Acknowledging it calls for a radical redescription of what he called ‘our method’.

Parentage: TS 220, §100 The immediate precursor of PI §122 is a remark in the ‘Early Version’ of Philosophical Investigations (TS 220, §100). It is immediately preceded by two remarks which trace philosophical problems to analogies which have been absorbed into the forms of our language (TS 220, §98b ¼ PI §112). The source of the difficulty is that ‘the form of representation of our language’ (die Darstellungsform unserer Sprache) has assumed ‘a disquieting aspect’ (einen uns beunruhigender Aspekt) (TS 220, §98a). Some examples are listed. The noun ‘time’ may suggest some mysterious medium. ‘(But here there is surely nothing! – But here there is surely not nothing!)’ [‘Aber hier ist doch nichts! – Aber hier ist doch nicht nichts!’] (TS 220, §98c). It seems problematic that we can measure the duration of an event because it is never present in its entirety. And the use of ‘is’ both as copula and identity-sign generates an apparent paradox. ‘The rose is red and also it is surely not red’ [‘Die Rose ist rot, und ist doch wieder nicht rot.’] Wittgenstein then proposed a strategy for dissolving these problems. We then change the aspect by placing side-by-side with one system of expression other systems of expression. – The bondage in which one analogy holds us can be broken by placing another [analogy] alongside which we acknowledge to be equally well justified. (TS 220, §99)14

For example, the ‘problem of identity in difference’ can be made to disappear by adopting a notation in which ‘is’ is replaced in some contexts by ‘¼’ and in others by ‘e’ (which symbolizes set-membership). Exhibiting this possibility side-by-side with our use of the word ‘is’ suffices to break the spell of the form of representation of our language. ‘It was the system of expression which held me in bondage’ [‘Es war das System des Ausdrucks, welches mich in Bann hielt’] (TS 220, §99b). Juxtaposing one notation with another is intended to effect a change of aspect; success would consist in our seeing the use of ‘is’ differently, i.e. in our looking at the use of this word as decomposing into two distinct uses. The ‘problem of identity in difference’ will simply vanish for anybody who adopts this point of view since the apparent contradiction depends on seeing the two occurrences of ‘is’ as having the same use. Since the remark on the importance of perspicuous representations immediately follows this discussion, there is a prima facie case for linking its content to the idea of exposing new aspects of systems of expression in

Philosophical Investigations §122 31

order to break our bondage to analogies absorbed into the forms of our language. How can this suggestion be filled out? The obvious thought is that Wittgenstein moved from illustrating some particular therapies for some particular philosophical problems to giving a more general description of the method that he has just exemplified. He suggested that the paradoxes or antinomies are themselves manifestations of the lack of perspicuity in ‘our grammar’. If every particular philosophical problem is rooted in our failure to command a clear view of the uses of the words which are employed to frame it, then in making the ‘problem of identity in difference’ disappear by juxtaposing with it a new notation, Wittgenstein must have thought himself to be remedying this defect in a particular case by making (one aspect of) the grammar of the word ‘is’ perspicuous to us. We might further conjecture that he thought of this object of comparison itself15 as a perspicuous representation of the use of the word ‘is’, at least if it is successful in making perspicuous an important aspect of the use of ‘is’. (Its success would not entail either that ‘is’ has two different uses exactly matching those of ‘¼’ and ‘e’ in the notation of set theory or even that ‘is’ as a matter of fact has two distinct uses, but merely that it can be seen as having two uses.) The maximum of continuity in the sequence of remarks in TS 220 is secured if the observation about perspicuous representations in PI §100 is taken to be a generalization occasioned by an exhibition of specific instances in §§98–9. This reading rests on two assumptions. The first is that Wittgenstein called ‘a perspicuous representation of our grammar’ anything which has the function of introducing ‘perspicuity’ into some aspects of the use of some of ‘our words’ (i.e. anything which manifestly helps somebody to know his way about by dissolving some philosophical problems which bother him). There is no general restriction on what form a perspicuous representation may take. In particular, it need not be either a diagram (like the colour-octahedron) or an assemblage of grammatical rules for the use of ‘our words’. The set-theoretic system of notation differs from both although it is clearly introduced in order to clarify an aspect of the use of the word ‘is’ whose neglect generates the ‘problem of identity in difference’. In this case (and in others too (TS 220, §6 ¼ PI §5)), it is describing a different language-game (a clear and simple language-game or a calculus which proceeds according to definite rules (cf. PG 63)) and comparing this other system of notation with the complicated use of ‘is’ which enables us to command a clear view of ‘the use of our words’. If the object of comparison itself is to be called ‘a perspicuous representation’, then there need be no conflict between calling something ‘the description of a language-game different from our own’ and calling it ‘a perspicuous representation of the grammar of our language’!

32

Reading Wittgenstein

The second assumption is that the term ‘representation’ can be applied to an object of comparison which is employed to exhibit aspects of ‘the use of our words’. This seems to lead to two related difficulties: the first is to make sense of the idea that the object of comparison is itself a symbol, and the second is to make sense of the idea of a method of projection which relates the symbolic representation to what is represented. The first hurdle can be surmounted by realizing that whether something counts as a symbol or not depends on how it is used. If we make use of a landscape painting to say something about a proposition or of the notation of set theory to say something about ‘The rose is red’, then in this particular context the painting is a symbol for a proposition and the set-theoretic notation a symbol for contrasting aspects of the use of ‘is’. Anything can be used as a symbol, even if it is concrete, and something commonly used as a symbol may fail to be so used, as in using the colour-octahedron or arithmetical calculations as elements in a wallpaper design. The problem about specifying a method of projection itself requires refinement. On the one hand, there are objects employed in comparisons or similes, e.g. a landscape painting is taken as a model for part of the use of ‘proposition’. Here a non-trivial method of projection can easily be supplied, e.g. a correlation of picture-elements with the words comprising a proposition. On the other hand, there are simple or ‘primitive’ language-games which Wittgenstein used to promote a clear view of our complicated patterns of speech. Here the notion of a method of projection seems to lose its grip, at least in so far as the simple language-game is taken to be a fragment of ‘our language’. In what sense can a part of our language-game with numerals or colour-words (e.g. the use of natural numbers in counting or of colour-words without any ‘is’/‘looks’ contrast) be said to represent the whole? Or is it supposed to represent only itself ? This difficulty can be alleviated by considering projections in geometry. Three-dimensional surfaces may be projected on to a plane, and the resulting figures will count as ‘representations’ (in two dimensions) of configurations of three-dimensional surfaces. A simple language-game has a parallel role and hence has equal right to be called ‘a representation’ of the use of ‘our words’. The converse movement of thought, from the simplified object of comparison to the full complexity of ‘our language’, is best conceived as adding a new dimension to the grammar of the symbols of the simple language-game (cf. PI pp. 200–1). Wittgenstein often advised that it would be best not to view the simple game itself as a fragment of the more complicated one; rather we should look on the simple one as a complete language-game which can be projected on to an isomorphic subsystem within the complex one (on the model of correlating the natural numbers with the non-negative whole numbers within the

Philosophical Investigations §122 33

system of the rationals16). On this understanding, there seems no problem about describing a simple language-game as ‘a grammatical model’17 or ‘a perspicuous representation’ of the use of ‘our words’. Integrating TS 220, §100 into its context gives a highly unified account of the method which Wittgenstein tried to demonstrate by examples (TS 220, §116 ¼ PI §133). Although there is not one single method which can be mechanically applied to dissolve every philosophical problem, there is a general strategy exhibited in all the various therapies, and the possibility of mastering it and transferring it to new problems gives substance to the conviction that the correct treatment of each problem casts light on the correct treatment of all (Z §465). The unity of the method turns on the application to grammar and language of the concept of an aspect (and of the related concepts of seeing an aspect and being blind to an aspect). This concept is prominent in the text of TS 220. Philosophical problems arise ‘because the forms of representation of our language have taken on a disquieting aspect’ (TS 220, §98; emphasis added). The tyranny of a system of expression is to be broken and the problems dissolved by our effecting a change of aspect through juxtaposing with our language other systems of expression (TS 220, §99). We are often blind to ‘the philosophically most important aspects of things’ because of their simplicity and familiarity (TS 220, §105; cf. PI §129; emphasis added). And we must free ourselves from the thrall of the ideal by acknowledging it as a picture and finding its source; unless we find the concrete image that gave rise to it (Urbild), we cannot free ourselves from its misleading aspect. We can avoid ineptness or emptiness in our assertions only by presenting the model (Vorbild) as what it is, as an object of comparison (TS 220, §107). Not only is the concept of an aspect highlighted, but also the method of dissolving philosophical problems by effecting changes of aspect is demonstrated. The treatment of the ‘problem of identity in difference’ is offered as an illustrative example (TS 220, §99). Wittgenstein then put this method to work in the text of TS 220, addressing himself to the nature of the proposition (Satz) as envisaged in the Tractatus. He had characterized a proposition as a concatenation of names each of which corresponds to an object (TS 220, §108; cf. TLP 3.22, 4.22); looking back on this thesis, he now observed that ‘A picture held us captive’ (TS 220, §109 ¼ PI §115). He had mistaken the possibility of a comparison (the possibility of rephrasing ‘Socrates was older than Plato’ as ‘Socrates stood in the relation of being older than to Plato’) for a state of affairs of the highest generality (namely, that all words with material meanings are distributed in networks of purely logical relations (TS 220, §110)). We are inclined to predicate of the thing represented what lies in the method

34

Reading Wittgenstein

of representing it; through a kind of optical illusion, we seem to see in the inmost nature of the thing what is etched on to our spectacles (TS 220, §110). The expression of this illusion is the metaphysical use of our words (TS 220, §110). The cure is to encourage surrender of the dogmatic claims ‘Things must/cannot be thus and so’ by exhibiting other intelligible ways of seeing things (other possibilities), i.e. by showing that we can take off the pair of spectacles through which we now see whatever we look at (TS 220, §92 ¼ PI §103). To the extent that philosophical problems take the form of the conflict between ‘But this isn’t how it is!’ and ‘Yet this is how it must be!’ (TS 220, §98b ¼ PI §112), they will obviously be dissolved away once the inclination to say ‘must’ has been neutralized by seeing another possibility. Wittgenstein tried to liberate our thinking from enslavement to particular analogies by bringing to light other analogies which are equally well supported as the ones of which we unconsciously make use (TS 220, §99). A simile belongs to our structure; but we cannot draw any conclusions from it; it does not lead us beyond itself, but it must always remain as a simile. – We can draw no conclusions from it. This is so if we compare a proposition with a picture . . . or the application of propositions, operating with propositions, with the application of a calculus, e.g. the calculus of multiplication. (TS 220, §102; emphasis added)18

Wittgenstein did not either recommend or practise a method of eschewing analogies in the clarification of grammar. On the contrary, he constantly introduced fresh ones. But he employed them deliberately for specific purposes, emphasizing that he was dealing with objects of comparison and hence avoiding falling into metaphysical assertions by drawing conclusions about how things must or cannot be (TS 220, §107; cf. PI §131). Analogical descriptions of grammar stand on the same level as the unexamined analogies which they are intended to displace in dissolving particular philosophical problems. Wittgenstein’s therapy is, as it were, a kind of homoeopathy. Conscious analogies and comparisons are useful tools for curing diseases of the intellect, whereas unconscious ones generate insoluble problems by exercising an imperceptible tyranny over our thinking.19 When we are held captive by a picture or analogy ‘embedded in our language’, we are unable to see something in more than one way. We think, e.g., that a proposition must consist of a set of names scattered in a nexus of purely logical relations (TS 220, §109), and we express this conception in the form of the report of a perception of a (metaphysical)

Philosophical Investigations §122 35

fact, i.e. of an insight into the nature of a proposition. Our position is comparable to that of someone who continuously sees a single aspect in the duck–rabbit diagram (PI p. 194). He is aware of no other possibility, he is unable to see the figure other than as a picture-rabbit, and he reports his seeing this aspect of the figure in the form of a perceptual report (PI pp. 194–5). In both cases there is a kind of blindness to aspects. The remedy in both cases must be to bring hitherto unnoticed aspects of things to a person’s awareness, i.e. to get him to see things differently. The aim is to effect not merely a change of opinion, but a kind of conversion (die Umstellung der Auffassung (TS 220, §116)). Wittgenstein suggested that one method is to place other objects side-by-side with the thing to be reconsidered. We change the aspect of ‘the use of our words’ by juxtaposing with it another system of expression, real or imagined (TS 220, §99). Indeed, ‘our method is not merely to enumerate actual uses of words, but rather deliberately to invent new ones, some of them because of their absurd appearance’ (BB 28). Surrounding our practice with new possibilities (language-games) may have the consequence that we see matters differently; it brings it about that we compare it with this rather than with that, and thereby it changes our way of looking at things (TS 220, §126; cf. PI §144). Indeed, juxtaposing ‘our language’ with a simple language-game may bring about our commanding a clear view of the use of our words! (TS 220, §6 ¼ PI §5) This procedure parallels bringing someone to notice a new aspect of the duck–rabbit diagram by surrounding the figure with other picture-rabbits (cf. LW §165). In both cases there is an inclination to exclaim: ‘Nothing has changed, yet everything looks different!’ Provided that we are prepared to call ‘perspicuous representations of our grammatical rules’, inter alia, whatever objects of comparison serve to make perspicuous to us (to bring us to command a clear view of ) ‘the grammar of our language’, there seems to be a cast-iron case for calling some language-games ‘perspicuous representations of our grammar’. The very same argument would apply to some diagrams (e.g. the diagram of a graduated ruler in the clarification of the mutual exclusion of parallel determinates under a single determinable20 (WWK 64; PR §84)) and to certain things introduced as analogues in the clarification of ‘the use of our words’ (e.g. representational paintings in the philosophical investigation of propositions, or everyday rules of grammar in the clarification of mathematics). The range of perspicuous representations of ‘our grammar’ is extremely catholic. Neither divergence between the object of comparison and ‘the use of our words’ (e.g. its being a ‘clear and simple language-game’) nor lack of pretensions to describe ‘the grammar of our language’ debars something from qualifying as a perspicuous

36

Reading Wittgenstein

representation. On the contrary, simple language-games mesh with one preferred paradigm that Wittgenstein employed to explain what he understood by ‘a perspicuous representation’, namely Goethe’s conception of the primal plant (Urpflanze). Goethe’s enterprise was to describe all of the parts of plants in terms of the degree to which their forms deviated from an archetypal leaf-form; he used the primal plant as a centre of variation for plotting plant morphology.21 Wittgenstein frequently used simple or ‘primitive’ language-games for exactly this purpose (cf. LPP 25, 142). We may come to command a clear view of the complicated and interwoven uses of ‘our words’ (to see ‘our language’ in a particular way) by comparing them with language-games in which symbols have clearly defined uses and noting similarities and differences between ‘our language’ and these prototypes (TS 220, §115; cf. PI §130).22 Provided that we acknowledge the point of calling objects of comparison (when used in this manner) ‘perspicuous representations’, there is no obstacle to seeing that perspicuous representations have as important a role in ‘our method’ as Wittgenstein claimed. At the same time, the sequence of remarks in the ‘Early Version’ of the Investigations points towards a very different interpretation of ‘a perspicuous representation’ from the one encapsulated in the Bird’s-eye View Model. Many points of discrepancy are noteworthy. First, if the paradoxes of PI §98 are instances of problems arising from our failure to command a clear view of ‘the use of our words’, and if comparing ‘our notation’ with a different one exemplifies the strategy of making perspicuous a particular use of one of ‘our words’ (PI §99), then the method of constructing perspicuous representations (or specifying objects of comparison) is inseparable from the task of dissolving particular philosophical problems. The notation which distinguishes ‘¼’ from ‘’ makes perspicuous an aspect of the use of the word ‘is’ inattention to which generates the ‘problem of identity in difference’; using this as an object of comparison makes the problem completely disappear by exhibiting a new possibility and thereby altering our way of looking at the grammar of ‘is’. Here giving a perspicuous representation is essential to a conceptual therapy, not something to be contrasted with particular therapies as an independent project manifesting Wittgenstein’s hankering for a more global, positive, or systematic role for philosophy. Second, comparison with an alternative notation is clearly one method for introducing an order into ‘the use of our words’, but in every case it is also relative to a specific purpose, namely the dissolution of a particular problem. If each philosophical problem arises from a particular simile ‘which is taken up into the forms of our language’ (TS 220, §98), if each problem manifests ‘the bondage in which an analogy holds us fast’ (TS 220, §99), then Wittgenstein’s goal, in each case, must be to ‘break

Philosophical Investigations §122 37

the bondage in which certain forms of language hold us fast’ (TS 220, §113). Consequently, in respect of each philosophical problem he wished ‘to establish an order in our knowledge of the use of language’ (TS 220, §113); ‘an order for a particular purpose, one out of many possible orders’ (TS 220, §113, emphases added; cf. PI §132). Both the claim that the order is purpose-specific and the acknowledgement of the possibility of different orders indicate that his aim was to produce for each problem an order which would make it completely disappear,23 not to establish a single order which would make every problem disappear (or even every problem within some range of problems). It exemplifies a basic quantifier-shift fallacy to draw the conclusion that, on Wittgenstein’s view, ‘it is the task of philosophy to achieve an order, an order which gives complete clarity, . . . an order which makes everything surveyable’.24 Third, provided that giving a perspicuous representation is the method for making a philosophical problem disappear, and provided that an effective therapy may be problem-specific, there is clearly no commitment whatever to the idea that perspicuous representations must be (even roughly) additive. The attempt to amalgamate two perspicuous representations (e.g. the colour-octahedron and the graduating marks on a yardstick which exhibit two aspects of the use of colour-words) may produce nothing intelligible, or it might happen that one exacerbates the problem that the other is designed to eliminate (e.g. taking the durations of intervals on the model of the graduating marks on a yardstick may aggravate the difficulty expressed by the question ‘How is it possible to measure time?’). There is no more reason to suppose that perspicuous representations are additive than to claim that the successive seeing of two different visual aspects in the duck–rabbit diagram can be combined into a single visual experience of seeing both aspects at once.

Mature Reflections Aside from TS 220 there is more textual evidence for a conception of perspicuous representations which differs radically from the Bird’s-eye View Model. I cannot present and discuss all of the relevant material here.25 Instead, I shall suggest how sensitive attention to the texts of both the ‘Early Version’ and the ‘Final Version’ of the Philosophical Investigations leads us in the direction of elaborating the first (and hitherto neglected) strand of his thinking about ‘perspicuous representations’. Perhaps the most direct route to attaining an overview of ‘his method’ is to address the question of what he meant by the remark ‘The concept of a perspicuous representation . . . earmarks our form of representation, the

38

Reading Wittgenstein

way we look at things’ (§122; emphasis added). At the very least, this concept must apply to what he considered to be his own method of philosophizing (whether or not he held that it had some claim on the attention of other philosophers or that it lay in the nature of philosophy). Consequently, it must play a role in his way of looking at and representing what he took to be the subject-matter of his investigations, namely ‘the use of our words’ or ‘our grammar’. Does it make sense to speak of ‘a form of representation of the grammar of our language’? This phrase may seem problematic because he sometimes referred to ‘our grammar’ itself as ‘a form of representation’.26 Yet Wittgenstein saw no objection whatever to this locution. On the contrary, it might be said that he regarded this idea as of fundamental importance for philosophy, and that he focused his attention not so much on accumulating data on ‘the use of our words’ as on confronting more general prejudices or preconceptions which stand in the way of our making effective use of these data. His therapies for particular philosophical problems involve tracing the roots of our thinking back to certain underlying ‘grammatical illusions’ (cf. §110), i.e. unacknowledged or unconscious ways of looking at symbolism (or ‘our grammar’). This ‘second-order’ concern with different forms of representation of grammar (i.e. different forms of representations of forms of representation) is distinctive of what he called ‘our method’, and it clarifies why he held out the hope of demonstrating a method by means of examples (§133). To make sense of the phrases ‘our form of representation’ and ‘our way of looking at things’ in this context presupposes that there are other contrasting forms of representation and other ways of looking at ‘our grammar’. Are these ideas incoherent? Some philosophers may show extreme resistance to accepting them, but Wittgenstein identified and explored many examples of alternative conceptions (or ‘pictures’) of ‘the grammar of our language’.27 One which is prominent in many texts is what he called ‘Augustine’s conception of language’ (§4, cf. §1). According to this schema, a difference in the meanings of two words must always be explained by reference to a difference between the two objects which they severally stand for, and a difference between the uses of two sentences must be accounted for by reference to a difference between the two facts which they severally describe. In adopting ‘this general notion of the meaning of a word’ (§5), we are asking for the expression ‘This word signifies this’ to be made part of the description [of the use of any word]. In other words the description ought to take the form: ‘The word . . . signifies . . . ’. Of course, one can

Philosophical Investigations §122 39 reduce the description of the use [of any word] to the statement that this word signifies this object. (§10)

Nonetheless, ‘this general notion of the meaning of a word surrounds the working of language with a haze which makes clear vision [das klare Sehen] impossible’ (§5). This conception of language is manifestly a norm (or form) of representation of the ‘uses of our words’ to which any use whatever may be fitted (§13). It is not a theory which excludes anything or which could be held to misdescribe (or describe!) anything. Nothing defies being cast into this pattern, though there may be much more difficulty with some words than with others. Augustine’s picture of language can be seen as a dogma that controls the expression of all descriptions of the uses of words and sentences. In this way it exerts an absolute, palpable tyranny over philosophers who cannot complain, however, that they are not free. It is not a wall setting limits to what can be noted and described in ‘our grammar’, but rather a brake which serves the same purpose (cf. CV 28). At the same time it appears to have no possible influence. It seems to be a mere form of representation which alters nothing in what is represented, but simply our ways of representing things. Viewed in this way it seems completely anodyne. ‘When we say: ‘‘Every word in language signifies something’’ we have so far said nothing whatever’ (§13). (We might even say that a philosopher who repudiated Augustine’s picture would merely be objecting to a convention. What he wants is only a new notation, and by a new notation no facts of logical geography are changed! (cf. BB 57).) The pervasive, subtle, and pernicious influence of forms of representation is the very aspect of the description of ‘the grammar of our language’ which Wittgenstein laboured to bring sharply into focus, and his treatment of Augustine’s picture of language is the prime exhibit. On this view of his intentions, it betokens grave misunderstanding of the spirit of the Investigations to treat Augustine’s picture as a theory which is the least common denominator of a wide ranges of theories of meaning and to interpret his critical exploration of it as a wholesale reductio ad absurdum of these theories (Frege’s ‘philosophy of language’, the logical atomism of Russell and the Tractatus, etc.).28 Wittgenstein tried to show that the roots of many more local philosophical views (both superstitions and insights!) lie in general pictures of the nature of symbolism. These explorations ranged over the conception of logical constants in the Tractatus, Platonism in philosophy of mathematics, behaviourism in philosophy of mind, etc. In every case he brought sharply into focus general conceptions about the essential nature of symbolism. All of them present particular forms of representation of ‘our

40

Reading Wittgenstein

grammar’, and a critic might call each and every one of them ‘a mythology of symbolism’ (PG 56), ‘a grammatical fiction’ (PI §307), or ‘a false grammatical attitude’ (PG 85; emphases all added). The influence of these ‘second-order’ forms of representation (or ways of seeing things) on philosophers’ descriptions of ‘our grammar’ is clearly visible not only in generating many of the celebrated ‘isms’ which dominate philosophical debates, but also in shaping the description of the most basic ‘data’ supporting their ‘metaphysical uses’ of the words of ‘our language’. One instance is the debate about the status of ostensive definitions. It is clear enough that we do in fact teach the word ‘red’ by pointing to a ripe tomato and saying ‘That ! is red’; moreover, it is also clear that this procedure is one paradigm of what we call ‘explaining what ‘‘red’’ means’ or ‘explaining what red is’. Yet one philosopher insists that this practice cannot be called ‘defining or explaining what ‘‘red’’ means’, since it really consists in correctly applying the word ‘red’ to a perceptible object. Another philosopher argues, on the contrary, that it is the fundamental case of assigning meaning to a word, since it actually connects a word with what it stands for (unlike the usual ‘verbal’ or ‘dictionary’ definitions). A third claims that it is indeed an explanation, but only in virtue of its endowing the recipient with the capacity to recognize the colour red when he sees it. Wittgenstein himself suggested that the ostensive definition might be viewed as a substitution-rule for symbols (which should be seen as including both a sample and the gesture of pointing). And so on. In all these cases, philosophers’ descriptions of the data of ‘the grammar of our language’ are shaped by the adoption of particular and identifiable forms (or norms) of description. There is just as compelling a case for claiming that descriptions of ‘our grammar’ must conform to particular forms of description (employing, as it were, particular systems of coordinates) as that scientific descriptions of the world must do so (in virtue of being framed in a symbolism which has a particular ‘grammar’). There seems no good reason to think (or even to think that Wittgenstein thought) that the subservience of descriptions to forms of description, and hence the ‘relativism’ of acknowledging the possibility of different forms of description, comes to an end at the frontiers of empirical discourse! From close inspection of his method in conducting his own philosophical investigations, we can see clearly that Wittgenstein was committed to the intelligibility of there being different forms of representation of ‘our grammar’ or different ways of looking at ‘the use of our words’. He thought that philosophical problems could be exacerbated or alleviated, generated or annihilated by shifting from one of these ‘second-order’ forms of representation to another. For example, he thought that the

Philosophical Investigations §122 41

sterile debates of philosophy of mind could be dropped altogether if we could free our thinking from the grammatical illusion that every sentence must describe something. He constantly tried to sketch new possibilities (e.g. PI §244), to make visible hidden aspects of ‘the use of our words’ (§129), and to encourage us to look at things like this, not like that (LPP 168). What makes a remark a perspicuous representation of ‘the use of our words’ is not its intrinsic features, but its function in making ‘our grammar’ perspicuous, i.e. in providing some sort of landmarks, patterns, analogies, pictures, etc. which enable us to find our way about in the motley of ‘our language’ (cf. §123). Such a remark need not have one form; in particular, it need not consist of a mere selection and arrangement of grammatical rules. In fact, perspicuous representations in Wittgenstein’s writings have several radically different forms. Their diversity matches the diversity of procedures for bringing somebody to notice a new aspect in a drawing. This may take the form of tracing certain lines in a particular sequence, surrounding parts of the drawing with other figures, comparing the drawn figure to various three-dimensional models or other perceptible objects, or showing how the drawn figure might result from a sequence of modifications to another drawing. All of these procedures have analogues in making visible unnoticed aspects or patterns in ‘the use of our words’. We may arrange various cases of following instructions for using symbols by reference to samples in order to lead someone to see a connection between the obviously normative role of the standard metre in the language of metric linear measurement and the function of everyday ostensive definitions of the word ‘red’ by reference to ripe tomatoes or pillar boxes (cf. PI §§50–4; BB 85–90). We may imagine a familiar use of symbols to be embedded in very unfamiliar or abnormal contexts (PI §142; BB 9, 28, 49, 61–2). We may compare ‘our grammar’ with various ‘clear and simple’ language-games, noting respects of similarity or difference (PI §§5, 130–1; TS 220, §99). We may demonstrate how ‘our complicated uses of words’ might have grown out of ‘primitive language-games’ by the gradual accretion of new expressions and the adding of new joints to language (BB 77–125; LPE 293, 295–6; LPP 23–5, 96; Z §§418–25). Provided that we acknowledge the diversity in the forms of what he called ‘perspicuous representations (of our grammar)’, we should have not the slightest difficulty in seeing that the concept of a perspicuous representation does indeed have a fundamental importance in his form of representation. Failure to comprehend this aspect of his method reflects a misunderstanding of his use of the phrase ‘a perspicuous representation’. The Bird’s-eye View Model misrepresents almost every aspect of this use;

42

Reading Wittgenstein

indeed, it interposes a veil which makes clear vision impossible. I shall redraw each of its lines to exhibit a different aspect of the grammar of ‘perspicuous representation’ (see pp. 28–9). (i)

A perspicuous representation is a representation of something which makes what is represented perspicuous (or orderly) to someone to whom it is given. It need not be a representation of ‘our grammar’; its subject-matter may be the morphology of plants, the classification of animals, or religious ceremonies or magical rites. It need not consist of ‘rules of grammar’ (arranged in an ordering). It is not a pleonasm to speak of ‘a perspicuous representation of grammar’. (ii) The components even of a perspicuous representation of ‘the grammar of our language’ need not be descriptions of the employment of the symbols of ‘our language’. They may be ‘centres of variation’ for giving descriptions of the use of ‘our words’ (LPP 25); or they may be descriptions of different language-games (even of ones which have an absurd appearance (BB 28)) which may serve as objects of comparison in order to induce us to see things differently (or to change our way of looking at things (PI §144)). (iii) The adjective ‘perspicuous’ in the phrase ‘a perspicuous representation’ is not used attributively. Whether a representation is perspicuous is not an intrinsic feature of it (e.g. whether it can be taken in at a glance or easily reproduced accurately from memory), but rather a characterization of its role or function. It is a representation which makes perspicuous what is represented. In this respect, the adjective ‘perspicuous’ in the phrase ‘a perspicuous representation’ has a use similar to the adjective ‘clear’ in the phrase ‘a clear description’. (iv) Wittgenstein never called one representation ‘more perspicuous’ than another. On the other hand, there seems no reason why one representation of ‘the use of our words’ should not be more comprehensive than another; it might illuminate more aspects of the use of certain words. Certainly there seem to be degrees of ‘knowing one’s way about a city’, and corresponding differences in the knowledge conferred by maps. Wittgenstein seems to have invited some analogy with maps in drawing a comparison with ‘knowing one’s way about the use of our words’. (v) It makes sense to speak of modes of representation or ways of seeing ‘our language’, ‘our grammar’, or language-games (whether actual or imaginary). Indeed, both ‘Augustine’s conception of language’ and Wittgenstein’s proposal to view the meaning of a word as its use in the language are specific forms of representing ‘our grammar’ or particular ways of looking at it. What seems doubtful, on the

Philosophical Investigations §122 43

contrary, is whether it makes any sense to speak of ‘descriptions of our grammar’ which do not exemplify particular modes of representation or ways of seeing things. In particular, it seems doubtful whether we can legitimately contrast Wittgenstein’s investigation of ‘psychological concepts’ (or his treatment of mathematics) with standard philosophical positions (behaviourism, Cartesian dualism, etc., or Platonism, intuitionism, etc.) by claiming that he merely described ‘the uses of our words’ whereas other philosophers looked at the facts of grammar through the distorting spectacles of particular conceptual schemes or metaphysical prejudices. In claiming the importance of perspicuous representations for ‘our form of representation’, Wittgenstein seems to have been explicitly subscribing to a form of relativism which most of his would-be followers reject.29 (vi) Perspicuous representations need not be (even roughly) additive. If individually each one makes visible some one aspect of the use of some particular ones ‘of our words’, and if there are typically different aspects of the uses ‘of our words’, then there is no necessity that there be any single way of making simultaneously visible several different aspects ‘of our grammar’, still less that the combination of two perspicuous representations of the use of the same symbols make perspicuous two aspects of the use of these symbols. The mutual exclusion of different visual aspects of a diagram or drawing may be matched by the complementarity of perspicuous representations of ‘the use of our words’. (vii) There are tolerably clear criteria of identity for perspicuous representations which license the conclusions of the form that Wittgenstein sometimes offered several distinct perspicuous representations of the use of a particular symbol or set of symbols. In this respect, ‘perspicuous representation’ is a count-noun whose use parallels the use of ‘landmark’ or ‘point of reference’. There is no doubt that the dome of St Paul’s, Nelson’s column, and Victoria Tower are distinct landmarks in the cityscape of London. Similarly, there is no doubt that the colour-octahedron and the yardstick marked off with colour coordinates are two distinct perspicuous representations of the use of words for chromatic colours. (viii) The criteria of success in giving a perspicuous representation are strictly relative to particular situations. Adequacy must be judged with reference to the elimination of a particular person’s not knowing his way about in a particular situation. In this respect a perspicuous representation ‘of our grammar’ is fundamentally

44

Reading Wittgenstein

different from the philosophical ideal of a correct conceptual analysis. In particular, it cannot be faulted on the grounds that it does not give a description that fits every use (or all aspects of the use) of the words whose use is represented. On the contrary, it is adequate provided that it makes clear some puzzling aspect of ‘the use of our words’ and that it can serve as a centre of variation or a point of reference for making clear a large class of cases of the use of certain symbols.30 For example, the slogan ‘Thinking is operating with signs’ (BB 6), understood as a perspicuous representation of the use of the verb ‘think’, should not be rejected on the grounds that a person may think without speaking or writing or on the grounds that a parrot may utter words without cogitating or having thoughts.31 On the contrary, its utility is to be judged primarily by whether or not it effects a certain change in the way of looking at ‘our language’, namely whether it dissolves the question ‘What gives life to dead signs?’ (BB 4) by clarifying the point that ‘Thinking’ is an absurd answer. To conclude from Wittgenstein’s failure to repeat this slogan in later writings about other aspects of the concept of thinking that he repudiated this earlier idea is to judge perspicuous representations by inappropriate criteria. (ix) It is clearly not necessary that there be one perspicuous representation which simultaneously dissolves every philosophical problem into which people naturally fall in reflecting about ‘the use of our words’. Moreover, it seems doubtful whether it even makes sense to delineate in advance all the possible aspects of the grammar of any word(s); there seems to be no such thing as a perspicuous representation of all aspects of ‘the use of our words’. In respect of each particular aspect, it is a creative achievement (not a mechanical procedure) to find a means for bringing it to another’s notice, and it is a task of persuasion (not a demonstrative proof ) to bring it about that another sees things differently. If this synopsis of the grammar of ‘perspicuous representation’ captures Wittgenstein’s conception of the nature of a perspicuous representation (or even one facet of the employment of a family-resemblance concept), then there are important consequences for the interpretation and exploitation of his philosophical ‘insights’. He constantly advocated new ways of looking at things: he urged us to examine language under the guise of a calculus, to view the meaning of a word as its use or the sense of a sentence as its employment, to regard an ostensive definition as a substitution-rule for symbols or a sample as part of ‘our language’, and to look at an avowal

Philosophical Investigations §122 45

of pain as a manifestation or expression of a sensation or at an arithmetical equation as a rule of grammar. In all these cases he was manifestly trying to clarify aspects of ‘the use of our words’. If it is correct to link up perspicuous representations of grammar with seeing aspects of the employment of symbols, then each and every one of these well-known leitmotifs from Wittgenstein’s later philosophy appears to be a paradigm of what he understood by ‘a perspicuous representation’. The application of this concept to what he called ‘descriptions of grammar’ bears on what he understood by the remark that he advanced no theses, gave no explanations, and avoided dogmatism in philosophy (PI §§109, 126, 128, 131). If ‘perspicuous representations’ are connected with exhibiting aspects of ‘the use of our words’, then they do not conform to many of the features of fact-stating discourse. On the contrary, they more closely resemble descriptions of possibilities. If his remark that an ostensive definition of ‘red’ can be seen as a substitution-rule for symbols is intended to make visible an aspect of a form of explanation of word-meaning which dissolves the questions ‘How is language connected with the world?’ and ‘Must any language not include indefinable or primitive concepts?’, then there is no need (indeed no point) in debating whether his observation is a correct application of the everyday expression ‘a substitution-rule for symbols’, that the remark encompasses all cases of what we call ‘ostensive definitions’ (e.g. explanations of proper names of the form ‘That ! is Mont Blanc’), or that it makes clear all of the important aspects of ostensive definitions (e.g. the differences between them and verbal or dictionary definitions). Precisely parallel points hold for the observations that arithmetical equations are rules of grammar, that the meaning of a word is its use in the language, that the utterance ‘I am in pain’ is an expression of pain, etc. No fact (even one about ‘our grammar’) is stated, no thesis advanced. There is nothing to attack, hence nothing to defend against criticism. Wittgenstein advocated nothing more (and nothing less!) than different possible ways of looking at things which he offered in particular argumentative contexts for certain specific purposes. One might object that the conception of a perspicuous representation which I have tried to pin on Wittgenstein would rob his writings of all interest and importance. Did his work in philosophy of mind and philosophy of mathematics in the end just amount to advocating mere changes in the form of representation of ‘our grammar’? Did he really suggest that we can dissolve our philosophical problems at will by deciding to see things differently as we can shift at will from seeing the duck–rabbit diagram as a duck to seeing it as a rabbit? These dismissive responses ignore the

46

Reading Wittgenstein

important fact that forms of representation are deeply embedded in our thinking and even in our pattern of activities (our forms of life). To change our form of representation (e.g. to drop the dogma of psychophysical parallelism) may be an enormous thing to do (UW 434). There is every reason to suppose that changes in forms of representation of ‘our grammar’ will be equally momentous and full of far-reaching consequences for our ways of thinking. Not only momentous, but also difficult to achieve (cf. CV 48). It is not a simple matter to effect a total reorientation of one’s whole style of thinking (a conversion to a new point of view). Indeed, this may be impossible in practice for many individuals at many times (cf. CV 61). The point of calling aspect-seeing ‘voluntary’ (and in this respect contrasting it with perception) is not to claim that it can be brought about on a whim, but rather that it makes sense to ask somebody to look at things differently, to say that a person has complied with this request, or equally that he has refused to see an aspect which is perfectly visible to others (LPP 334; RPPI §899). We might say that changing one’s way of seeing things is difficult because it is voluntary, because one has to surrender what one has always wanted to see (cf. CV 16–17). What I have tried to do is to bring to light a new way of looking at Wittgenstein’s own conception of a perspicuous representation. This enterprise consists in making visible hitherto unnoticed aspects of his own philosophical methods. If the general line of my argument is correct, the most that I can accomplish is to expose a possibility for interpreting his later philosophy, a possibility the acceptance of which would amount to a change in our way of looking at his ‘descriptions of the grammar of our language’. It is not possible to prove or demonstrate contra mundum that this interpretation captures the nature of his philosophy or that it alone fits his practice and pronouncements; it is at best possible to persuade a willing reader to explore this possibility for himself. Working in philosophy – like work in architecture in many respects – is really more a working on oneself. On one’s own interpretation [Auffassung]. On one’s way of seeing things [Daran, wie man die Dinge sieht]. (And what one expects of them.) (TS 213, 407 ¼ CV 16)

This observation is important for understanding the point or spirit of the particular perspicuous representations of ‘our grammar’ scattered through the Philosophical Investigations. It is no less important for appreciating the obstacles standing in the way of a contemporary analytic philosopher’s grasping Wittgenstein’s concept of a perspicuous representation.

Philosophical Investigations §122 47

Notes 1

2 3

4

5 6

7 8 9

10

11

I will usually put these expression in ‘scare quotes’ to remind the reader that there are many dimensions of uncertainty about how to interpret them (as well as many related expressions such as ‘a simile that has been absorbed into the forms of our language’ (PI §112) or ‘a picture . . . which lay in our language’ (§115)). These forms of expression in Wittgenstein’s discourse need careful investigation. This principle of projection applies to the simple octahedron (WWK 42), not to the more complex hexkaidekahedron (PR 278). In fact, the text does not state or imply that perspicuous representations are ‘our form of representation’, or even that they exemplify a (single) form of representation. Instead, it claims that the concept of a perspicuous representation characterizes ‘our form of representation’. The notion of completeness is fairly rough-and-ready. The criteria for distinguishing complete from incomplete explanations may be both purposerelative and context-dependent (cf. Baker and Hacker, Understanding and Meaning, pp. 78–81). Cf. Kenny, ‘Wittgenstein on the Nature of Philosophy’, pp. 42–3; Baker and Hacker, Understanding and Meaning, pp. 488–91, 531–45. Though comprehensiveness and schematization are logically distinct features of descriptions, they seem to be connected once the requirement of surveyability is imposed on an assemblage of rules of ‘our grammar’. A perspicuous representation must focus on ‘the central structure of the net of language, not the local refinements’; it would turn its back on ‘the lavish detail of an Austin’s descriptions of features of English grammar’ because of the danger of ‘getting lost in the details’ (Baker and Hacker, Understanding and Meaning, p. 543). See Kenny, ‘Wittgenstein on the Nature of Philosophy’, pp. 38–9, 42–3; and Hacker, Insight and Illusion, p. 151. Baker and Hacker, Understanding and Meaning, pp. 488–91; cf. Hacker, Insight and Illusion, pp. 177–8. Cf. Hacker, Insight and Illusion, p. 177. This reading arguably conflates two different things, the ‘plan for the treatment (or classification of psychological concepts’ (Z §§472, 488) and ‘the genealogical tree [Stammbaum] of psychological concepts’ (Z §464; cf. RPPI §722). According to this second method of representation, certain concepts are regarded as belonging to ‘primitive language-games’, while others are viewed as adding new joints to these language-games (cf. Z §425). No logical (or psychological) priority seems to be asserted (cf. LPP 23–5). It is, as it were, a view from nowhere. In this respect, a perspicuous representation of grammar might be said to be the heir of ‘the correct logical point of view’ (TLP 4.1213), or a part of seeing the world sub specie aeterni (TLP 6.45). (Cf. Hacker, Insight and Illusion, p. 151.) Waismann did exactly this (PLP, 80–1), but one might doubt (in absence of independent evidence) whether he correctly grasped Wittgenstein’s

48

Reading Wittgenstein

thinking about this point. There is a risk of blurring the all-important distinction between grammar and science, i.e. between description and theoryconstruction (Baker and Hacker, Understanding and Meaning, p. 539). 12 This requirement might seem to be completely unproblematic. Yet some degree of simplification, schematization, or exaggeration might be tolerated (as in the qualification ‘in a large class of cases – though not for all’ which accompanies the identification of the meaning of a word with its use in the language (PI §43)). Likewise ‘descriptions of our grammar’ may include similes, analogies and metaphors, and in these cases the true/false dichotomy has no straightforward application. What Wittgenstein called ‘descriptions of grammar’ have a considerable variety. The criteria for correctness of descriptions must be as variable as the language-games of giving descriptions of grammar (or the language-games of giving ‘descriptions of what is seen’ (PI p. 200)), and therefore as indeterminate as the criteria for judging the correctness of the description of a painting (or the ‘life-likeness’ of a sketch of a person or even the accuracy of a sketch of a painting of an animated battlescene). Indeed, the project of describing grammar, as Wittgenstein both envisaged and practised it, is not a pedestrian activity, and the skill of a philosopher is as much a matter of educated judgement as the talent of ¨ higkeit, von a portrait-painter. ‘Die Anlage zur Philosophie beruht auf der Fa einer Tatsache der Grammatik einen starken und nachhaltigen Eindruck zu empfangen’ (TS 220, §104). 13 Ibid., p. 544. It is not clear what licenses this conclusion. Need any enumeration of rules for the use of ‘our words’ have the property of surveyability? Or even the property that some rearrangement must exist which would give it this property? Perhaps the idea is rather that by choosing a suitably large scale and omitting enough fine detail any map can be transformed into one which can be taken in at a glance and reproduced flawlessly from memory. Not only are these theses dubious, but each also lacks any clear grounding. Could any of them be derived from the grammar of the word ‘grammar’? or of the phrase ‘a description of grammar’? These proposals seem to be nonstarters as long as the term ‘surveyable’ (‘u ¨ bersichtlich’) is not taken to be vacuous. ¨ ndern nun den Aspekt, indem wir einem System des Ausdrucks andere 14 ‘Wir a ¨ lt, an die Seite stellen. – So kann der Bann, in dem uns eine Analogie ha gebrochen werden, wenn man ihr eine andere an die Seite stellt, die wir als gleichberechtigt anerkennen’ (TS 220, §99). 15 Alternatively, he might have thought that perspicuously representing this aspect of ‘is’ consists in the activity or procedure of placing the notation of set theory side-by-side with the use of ‘is’ and noticing the similarities and differences. Here the making of the comparison would count as ‘giving a perspicuous representation’, although there would be nothing that counts as ‘a perspicuous representation’ which is given. (Describing ‘our grammar’ in this way would be comparable to drawing a figure in the air, where no drawing is produced.)

Philosophical Investigations §122 49

16 17

18

19

20

Yet another possibility is that he thought of the listing of points of similarities and differences with an object of comparison as a distinctive form of describing ‘our grammar’ which he labelled ‘a perspicuous representation’. Since my aim is to expose a possibility of interpretation, not to exclude alternative ones or even the possibility that ‘perspicuous representation’ is a family resemblance concept, I shall not explore the merits and demerits of these two alternative conceptions of Wittgenstein’s use of objects of comparison. It is clear enough what speaks in favour of treating objects of comparison themselves as perspicuous representations. The initial example of the colour-octahedron introduces a specific diagram which is called ‘a [perspicuous] representation’ (PR 52). Likewise, Wittgenstein later applied this label to the formula ‘Thinking is operating with signs’ (F. Waismann, ¨ nde’, ‘F 15’ in The Voices of Wittgenstein, 2003–see note 25 for ‘Zwei Einwa details). Finally, he cited (or alluded to (GB 69) Goethe’s notion of the primal plant (Urpflanze) as a paradigm of ‘a perspicuous representation’. In all these cases, something specific is isolated from a general description and singled out as a perspicuous representation. For a careful exposition of this point, see F. Waismann, Einfu ¨ hrung in das mathematische Denken, ch. 2. This phrase is employed by Waismann, who emphasized the importance of ‘the method of constructing grammatical models’ for the purpose of obtaining ‘a compendious scheme [ein u ¨ bersichtliches Schema] with which we can compare language’ (Waismann, Einfu ¨ hrung, pp. 72, 74). ¨ ude; aber wir ko¨nnen auch aus ihm ‘Ein Gleichnis geho¨rt zu unserem Geba keine Folgen ziehen; es fu ¨ hrt uns nicht u ¨ ber sich selbst hinaus, sondern muss als Gleichnis stehen bleiben. – Wir ko¨nnen keine Folgerungen daraus ziehen. So wenn wir den Satz mit einem Bild vergleichen . . . oder die Anwendung der ¨ tze, das Operieren mit Sa¨tzen, mit der Anwendung eines Kalku¨ls, z.B. des Sa Multiplizierens’ (TS 220, §102; emphasis added). Here ‘our construction’ (‘unserem Geba¨ude’) clearly refers to Wittgenstein’s own clarification of the grammar of the term ‘proposition’ (‘Satz’); the simile to which he called attention is the one between propositions and pictures. This is one point of resemblance between Wittgenstein’s method of philosophical therapy and psychoanalysis: ‘Der Philosoph trachtet das erlo¨sende Wort zu finden, das ist das Wort, das uns endlich erlaubt, das zu fassen, was bis dahin, ungreifbar, unser Bewusstsein belastet hat’ (TS 220, §106; cf. TS 213, 409). To bring something to the patient’s consciousness (to bring him to acknowledge it) is the principal step in effecting a cure. Cf. Kenny, ‘Wittgenstein on the Nature of Philosophy’, pp. 39–41. Note that there is strong, though indirect, textual evidence for calling this diagram ‘a perspicuous representation’: namely it is offered as an instance of finding ‘das lo¨sende Wort’ (WWK 77), and Wittgenstein used this phrase (and the variant ‘das erlo¨sende Wort’) to tag what he called ‘perspicuous representations’ (Baker and Hacker, Understanding and Meaning, pp. 552–3). This case of ‘clearly articulating the rule we have been applying unawares’ has

50

21 22

23

24 25

26

Reading Wittgenstein certain striking features. First, it is an example of what it clarifies: a rod can only be assigned one length, i.e. it can coincide with only one graduating mark when held up against a yardstick (cf. WWK 77–8). Second, it provides merely an analogy for clarifying the grammar of other determinables, e.g. colour-words or specifications of time-durations: there is no such thing as a yardstick marked off with colours, and the closest analogue (namely a colourchart) differs essentially in lacking a total linear ordering (cf. WWK 76). Third, this perspicuous representation can come into conflict with other clarifications of ‘the grammar of our words’: in particular, the comparison of measuring the duration of events with ascertaining the length of a rod is the root of philosophical perplexity about how it is possible to measure time and how it is possible to ascribe durations to events (BB 26; cf. TS 220, §98). For a slightly fuller clarification, see Waismann, Einfu¨hrung, pp. 80–1. Arguably, it is not sufficient for commanding a clear view of ‘the use of our words’ to have a discursive knowledge of points of similarity and difference with certain prototypes of language-games. In addition, it seems, we have to see things in a certain way. Compare: ‘Wie ich Philosophie betriebe, ist es ihre ganze Aufgabe, den Ausdruck so zu gestalten, dass gewisse Beunruhigungen // Probleme // verschwinden. [(Hertz.)]’ (TS 213, 421). Kenny, ‘Wittgenstein on the Nature of Philosophy’, p. 43, emphasis added. In particular, the immediate ancestor of TS 220, §100, namely TS 213, 417, places the concept of a perspicuous representation in the context of a discussion which emphasizes that philosophical problems arise as much from defects of the will as from those of the intellect (i.e. from how we want to see things); it also connects this concept with Goethe’s poem ‘Die Metamorphose der Pflanzen’ and with Hertz’s strategy of dissolving philosophical questions. The yet more remote ancestor, namely GB 69, applies the same concept to the description of religious rituals and magical rites (i.e. to subject-matter other than ‘the grammar of our language’), and it seems to countenance various distinct forms of ‘perspicuous representations’ (not merely the ordering or rearrangement of descriptions of phenomena, but also comparisons with a centre of variation and fictional accounts of evolutionary development). In addition, there is important evidence from hitherto unknown dictations recorded by Waismann in the early 1930s and partly incorporated into his book The Principles of Linguistic Philosophy (PLP). (Later references to them will include the catalogue number given in the deposit of Waismann’s papers in the Bodleian Library, Oxford.) These are now available in English under the title of The Voices of Wittgenstein. The idea that this would be paradoxical seems to underlie a certain narrowing of the possible interpretations of PI §122 and the elimination of one possible interpretation of §131 (cf. Baker and Hacker, Understanding and Meaning, pp. 546, 556).

Philosophical Investigations §122 51 27

He even referred to ‘false representations [Darstellungen]’ of ‘the every-day language-game [sc. of describing what is seen]’ (PI p. 200b). 28 Pace Baker and Hacker, Understanding and Meaning, p. 47. 29 The final parenthesis of §122 may point towards the same conclusion. ‘Is this a ‘‘Weltanschauung’’?’ seems meant as a genuine question, though often taken to be a rhetorical one. The case for giving a positive answer is that Wittgenstein openly adopted a particular way of carrying on philosophical investigations, namely by looking at ‘the use of our words’ (e.g. focusing on the role of mathematical propositions in relation to empirical countstatements rather than on the grounds of mathematical truth or the sources of mathematical knowledge). By taking up this point of view, he cast his descriptions of the problems of philosophy in a particular form, and his procedure therefore parallels descriptions of history, social practices, or natural history which are said to manifest the influence of particular ‘Weltanschauungen’. On the other hand, there is a strong case for giving a negative answer: anything labelled ‘a Weltanschauung’ must be a way of seeing the world, but the subject-matter of most of Wittgenstein’s own philosophical investigations is taken to be ‘our language’ (i.e. not ‘the world’, but rather ‘our means of representation of the world’). A self-avowed relativism seems to be the only reason for his raising the question whether his own form of representation is a ‘Weltanschauung’. ¨ berzeugung tut uns dieselben Dienste wie die U ¨ berzeugung, 30 ‘Der Brustton der U oder vielmehr bietet er uns eine einfache und u ¨ bersichtliche Darstellung der ¨ berzeugung’’ an, die in einer grossen Zahl von Fa ¨ llen Grammatik des Wortes ‘‘U ¨ berzeugung’’ gerecht wird’ (F. Waismann, dem Gebrauch des Wortes ‘‘U ‘Glaube’ (F 5), p. 3). 31 This important point is not merely conjectural. Wittgenstein offered ‘Das Denken . . . [ist] das blosse Kalkulieren mit der Sprache // ein Operieren mit Worten oder sonstigen Zeichen’ as one paradigm of what he called ‘eine ¨ nde’ (F 15), p. 1; cf. u ¨ bersichtliche Darstellung’ (F. Waismann, ‘Zwei Einwa ‘Glaube’ (F 5), p. 3). He canvassed various obvious objections to this description of the use of ‘denken’ (e.g. the possibility of lying or the phenomenon of trying to find the right words to express a thought). As we can infer from a parallel treatment of identifying belief or conviction with the intonation of uttering a sentence, he concluded that nonetheless he regarded this formula as the best epitome of the grammar of ‘denken’.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

2 Some Remarks on ‘Language’ and ‘Grammar’*

My original intention was to present a systematic, if schematic, account of Wittgenstein’s employment of the words ‘Sprache’ and ‘Grammatik’. The ultimate aim was to effect a kind of genre-identification, to make some tentative steps towards clarifying his conception of language and grammar in order to locate his work on the map of modern philosophy. By making use of unsuitable landmarks, there is an acute risk of misconceiving the whole spirit of his philosophical investigations and consequently of misunderstanding many individual remarks.1 To what extent is it illuminating to label him an ‘analytic philosopher’ or a ‘philosopher of language’? What are the resemblances and differences between his ‘descriptions of the grammar of our language’ and Ryle’s mapping of the ‘logical geography of ordinary language’? How do his investigations compare with Austin’s fine-grained examination of English idiom or with Strawson’s account of the most general features of our conceptual scheme? How does his conception of meaning as use relate to truth-conditional semantics a` la Carnap or Davidson, to Dummett’s ‘antirealism’, to Chomsky’s transformational generative grammar, to theories of speech acts and illocutionary force, etc.? Or we might raise the yet more radical question whether it even makes sense to locate Wittgenstein’s thinking within the logical space generated by these fixed points within analytic philosophy. Any detailed treatment of these questions was clearly beyond the scope of a single paper of manageable proportions, but I found the same to be true even of the more modest synopsis that I had first envisaged. The data are extraordinarily complicated and wide-spread, while unpacking * Originally published in J. Schulte and G. Sundholm (eds), Criss-crossing a Philosophical Landscape: Essays on Wittgensteinian Themes Dedicated to Brian McGuinness, pp. 107–31. Amsterdam: Rodopi, 1992. Reproduced by permission of Rodopi B.V. Amsterdam.

Remarks on ‘Language’ and ‘Grammar’

53

the subtleties of individual remarks often requires extensive argument and textual comparisons. The attempt to produce an u¨bersichtliche Darstellung of this material risks degenerating into a series of controversial but unsupported dogmas, and that eliminates the possibility of achieving anything very important in this way. The heir of the original paper is an unsystematic description of a few aspects of Wittgenstein’s use of the terms ‘Sprache’ and ‘Grammatik’, particularly in the phrases ‘die Sprache’, ‘unsere Sprache’, ‘die Grammatik’, and ‘unsere Grammatik’. Since my aim is principally to raise questions rather than to answer them, I shall make no pretence at giving a complete or exhaustive treatment of these expressions in his writings. Moreover, I shall restrict attention to his speech-patterns (his idiom, his jargon, etc.), leaving out of account such vexed questions as whether he himself did not misuse or stretch the term ‘Grammatik’ (e.g. in his characterizing the arithmetical equation ‘2 þ 3 ¼ 5’ and an ostensive definition of ‘red’ as ‘substitution-rules for symbols’ or ‘rules of grammar’). My modest hope is that some careful scrutiny of some of his own carefully formulated remarks may encourage others to further investigations which might throw some light on a wide range of issues which seem to be imperfectly understood and inappropriately debated.

‘Die Sprache’: the Craving for Generality There is a tendency among translators to treat ‘die Sprache’ and ‘die Grammatik’ as abstract mass-nouns, parallel to ‘der Gedanke’, ‘das Denken’, or ‘die Wirklichkeit’; and there is a corresponding tendency among commentators to take remarks incorporating these phrases as generalizations (e.g. about natural languages or about any conceivable sign-system, whether actual or invented). In many cases this pattern of interpretation clearly distorts what Wittgenstein intended us to understand by particular remarks. One instance is the remark: ‘Es ist das Natu¨rlichste . . . wenn wir die Muster zu den Werkzeugen der Sprache rechnen’ (PI §16). This is regularly paraphrased as a generalization about the samples used in ostensive definitions of words in any language whatever; moreover, it is often cited to support ascribing to Wittgenstein the (dogmatic) thesis that such samples are part of language (often reinforced by the citation of §50). In fact, this remark has as its subject-matter the colour samples shown by A to B in the language-game of §8, and it suggests that they can be regarded as instruments of this language.2 The fact that these samples are imagined to have an institutionalized use may be important or even

54

Reading Wittgenstein

essential to making sense of this way of seeing them, and hence absence of any such role for many of the samples used for explaining colour-words in natural languages would block any generalization of this remark even over all colour-samples,3 let alone over all samples whatever. Another clear instance is the remark: ‘Wenn wir sagen: ‘‘jedes Wort der Sprache bezeichnet etwas’’ so ist damit vorerst gar nichts gesagt’ (§13). In the authorized version this is translated: ‘When we say: ‘‘Every word in language signifies something’’ we have so far said nothing whatever.’ This suggests, and is usually taken to be, a general thesis about the vacuity of the principal component of Augustine’s picture of the essence of human language, namely ‘Die Wo¨rter der Sprache benennen Gegensta¨nde’ (PI §1). But the subsequent parenthesis (in §13) makes it clear that the remark is intended to be one about ‘die Sprache (§8)’, i.e. the languagegame 2 extended by adding numerals, colour samples, and demonstratives used in connection with a gesture of pointing. This remark is no more a generalization about all sign-systems than the question raised in §10 (‘Was bezeichnen nun die Wo¨rter dieser Sprache?’) is a general question about languages. The quoted sentence from §13 should be translated ‘Every word in the language signifies something’, where the definite article has the role of a quasi-demonstrative as in ‘There is some wine in the cellar’; indeed, to avoid misunderstanding, it might even be translated as ‘Every word in this language signifies something’. The tendency to exaggerate the scope of strictly limited generalizations and the inclination to mistake remarks about particular sign-systems for generalizations about all possible languages arise in part from a feature of German grammar, namely the demand for the definite article with abstract nouns. This frequently introduces the possibility of an ambiguity for which there is often no exact counterpart in English. When Wittgenstein made use of the phrase ‘die Sprache’, we must always ask ourselves whether or not it makes sense to ask ‘Welche Sprache?’ (‘Which language?’). In some cases this question seems absurd (e.g. in Augustine’s delineating ‘ein bestimmtes Bild von dem Wesen der menschlichen Sprache’ (PI §1) or in our investigating ‘die Erscheinungen der Sprache’ in primitive forms of communication (§5)). In others we might well doubt whether this question is appropriate or not (e.g. in the remark that Augustine’s picture of language surrounds ‘das Funktionieren der Sprache’ with a fog (§5) or in the question how one can even want ‘mit der Sprache noch zwischen die Schmerza¨usserung und den Schmerz treten’ (§245)). But in some cases the question ‘Which language?’ seems to be appropriate and to demand an answer (e.g. in the remark that we might say that ‘a’, ‘b’, etc. in the language-game (8) signify numbers to remove the mistaken idea that these signs play the part

Remarks on ‘Language’ and ‘Grammar’

55

actually played ‘in der Sprache’, i.e. in this particular language, by the signs ‘block’, ‘slab’, etc. (§10)). Much hangs on how to sort out the cases which fall between the two clear extremes. For example, is the often-quoted remark ‘Was es, scheinbar, geben muß, geho¨rt zur Sprache. Es ist in unserem Spiel ein Paradigma . . . ’ (§50) a generalization which is meant to encompass the ‘simples’ (‘objects’) of the Tractatus, or is it comment specifically about the elements (the coloured squares) in the language-game of §48? (We might well say that our noting that this remark contains an ambiguity is an important observation about our method of representation of Wittgenstein’s philosophy!) Persistent neglect of this particular ambiguity in the phrase ‘die Sprache’ has arguably distorted much discussion of Wittgenstein’s ideas,4 and it has generated some clearly mistaken (and many disputable) translations into English. There is a constant danger of endowing quite specific comments with a spurious appearance of generality. Perhaps we could learn to take at its face-value his claim to have said nothing about the essence of language throughout his critical discussion of Augustine’s picture of language (§65).

‘Die Grammatik’ There are many close parallels in the patterns of use of the expressions ‘language’ and ‘grammar’ (‘die Sprache’ and ‘die Grammatik’). Though they are far from interchangeable, either in Wittgenstein’s writings or in general, they occur in phrases with the same structures, and they display parallel ambiguities. In particular both can be used either as abstract massnouns or as count-nouns. As abstract nouns, both require the use of the definite article in German. And both can be used with the applicative ‘unsere’ (‘our’). In translating and interpreting Wittgenstein’s work, philosophers show a penchant for taking ‘grammar’ as a mass-noun, and correspondingly they treat ‘our grammar’ (‘unsere Grammatik’) as equivalent to ‘the grammar of our language’. The question ‘Which grammar?’ seems to be rejected without serious consideration as being logically inappropriate, at least in any context in which we cannot raise the question ‘Which language?’. We seem prepared to distinguish ‘the grammar of English’ from ‘the grammar of German’, or ‘the grammar of sensation-words’ from ‘the grammar of number-words’, but it seldom occurs to anybody to canvass the possibility that there might be radically different grammars of our language (e.g. of English or of sensation-words) or that ‘pictures absorbed into the forms of our language’ might sometimes refer to

56

Reading Wittgenstein

different ways of describing the use of our words. Wittgenstein’s enterprise is usually taken to be presenting the grammar of our language, as if any deviation from his descriptions of how we speak would amount to a misdescription of our language (except in so far as it were either more or less detailed than his own). This understanding of the term ‘grammar’ seems conspicuous in gloss¨ bersiching Wittgenstein’s remarks on the importance of surveyability (U tlichkeit). He diagnosed one main source of our failing to understand the use of our words as our inability to command a clear view of (u ¨ bersehen) the use of our words; ‘Our grammar is lacking in this kind of perspicuity’ ¨ bersichtlichkeit’). He sought perspicu(‘Unserer Grammatik fehlt es an U ous representations (u ¨ bersichtliche Darstellungen) in order to remedy this defect and to produce a kind of understanding which consists in ‘seeing connections’ (PI §122). Our understanding of his programme is no more secure than our understanding of the problem which he addressed.5 This is turn depends on grasping exactly what he meant here in speaking of ‘our grammar’.6 One possibility (the only one usually considered) is that ‘our grammar’ here refers to what he had just called ‘the use of our words’. To say that our grammar lacks perspicuity is simply to repeat that we do not command a clear view of the use of our words. The criterion for this failure is presumably that we do not know our way about in our own language7 (§123), that we get entangled in our own rules (§125) and sustain intellectual bruises from running up against the limits of language (§119),8 and that we fall into antinomies, saying ‘This isn’t how it is, yet this is how it has to be!’ (§112). Lack of this special kind of understanding is manifested in certain very specific inabilities. It might seem paradoxical, however, to summarize what seems a contingent and variable feature of certain persons’ degree of understanding of the use of their own words in the comment that our grammar (the use of our words) is lacking in the property of perspicuity (as it were globally and absolutely). Does it make sense to find fault with a natural language (English, German, etc.) on the grounds that we sometimes fall into conceptual confusions by blindly following the lead of ‘surface grammar’? Would this not be comparable to claiming that gravity is blameworthy on the grounds that some people sometimes fall and injure themselves? Equally, it might seem odd for Wittgenstein, who professed to leave everything as it is (§124), to avow the intention to bestow on our grammar a property which it now lacks by constructing perspicuous representations which will henceforth make the use of our words perspicuous. In discussing proofs in mathematics, did he not make the antithetical point that symbolic abbreviations do not leave everything as it is precisely because

Remarks on ‘Language’ and ‘Grammar’

57

they may make a complex array of symbols surveyable for the first time and hence transform something that was not previously a proof into something that is one? Indeed, did he not constantly oppose the common dismissive strategy of labelling certain transformations as ‘mere changes of notations’9 or ‘mere alterations in our form of representation’?10 Of course, neither of these objections to the prevailing interpretation is insurmountable. Both can be removed by treating perspicuity as a property of the use of our words (our grammar) which admits of degrees and is relative to a point of view. Our grammar might be declared to lack surveyability because there is a high probability that we will get lost in trying to give descriptions of the use of our words. Moreover, this tendency to error seems to depend on our point of view: though we may not command a clear view of something complicated, merely altering the angle or distance from which we view it (as in looking down on a maze from on high) may make everything easy to grasp without altering anything in what we are looking at. This interpretation not only exploits the metaphor of a bird’s¨ bersicht’), but also has the merit of showing that eye view (one sense of ‘U it must be possible to achieve Wittgenstein’s avowed goal since speaking a language is a normative practice of which any speaker must in principle be able to attain mastery (cf. §199: ‘To understand a language means to be master of a technique.’).11 Though we may well not succeed at the first attempt, we must in principle be able to construct a description of the use of our words which embodies our own practical ability to speak a language, just as it must be possible to produce a map which would encapsulate our ability to find our way about in the streets of a city; success may be difficult to achieve, but it demands no more than the right choice of landmarks and the selection of a suitable scale (or degree of detail).12 This first (and now standard) interpretation fails even to consider the possibility of an ambiguity in the phrase ‘our grammar’. Rather than the use of our words, could Wittgenstein not have meant by it our descriptions of word-use, perhaps (more narrowly13) our schedule of rules for the use of words whose meaning we are attempting to clarify? On this view, the statement ‘Our grammar is lacking in perspicuity’ is not simply a reformulation of the preceding statement ‘A main source of our failure to understand is that we do not command a clear view of the use of our words’. Instead it traces back our failure to command a clear view of our own word-use to a general defect in our descriptions of what we say or in our compendia of rules by which we attempt to delineate the use of certain ones of our words. For example, when we try to describe what we say, we might make use of forms of description which are inappropriate for our purposes or liable to produce or exacerbate misunderstandings in a particular context; e.g. instead of describing various uses of the word ‘time’

58

Reading Wittgenstein

(or of ‘language’), we may try to frame a definitive once-for-all answer to the question ‘What is time?’ (or ‘What is language?’), or instead of describing how to use the connective ‘if ’, we may try to fill out the schema ‘ ‘‘if ’’ signifies . . . ’ (cf. BB 26–7; PI §§16, 89, 92, 293). Because we are prone to over-hasty generalization (BB 7) and also inclined to be dissatisfied with careful enumeration of cases or detailed attention to the vagaries of word-usage (BB 20), when we describe the use of our words, we may lay out rules in which we then become entangled (PI §125). We find ourselves making assertions which would ordinarily strike us as absurd, e.g. that we cannot measure time (BB 26), that ‘this’ and ‘that’ are the only genuine proper names (PI §38), that I can never know whether another person is in pain (§245), or that ‘I am annoyed by your repeated thoughtlessness’ is a description of the speaker’s behaviour, actual or potential (cf. §244). In all these cases the fault lies in the very forms of statement which we employ in attempting to describe what we say (how we use our words). If Wittgenstein meant to give the second diagnosis of the origins of (some) philosophical problems, then the remedy which he envisaged must obviously be our modifying our previous descriptions of the use of our words in particular circumstances and for particular purposes, i.e. our producing a different grammar or even new forms of grammatical descriptions. This seems to be exactly what he did have in mind when he first introduced the notion of a perspicuous representation (PR 52). The colour-octahedron is described as a perspicuous representation of the grammatical rules, namely the rules which describe the possibilities for significant combinations of colour-words (cf. PR 273–80). This diagram is evidently used to remedy in the particular case of colour-words what is the chief trouble with our grammar, i.e. with our (current) presentation of the rules for the use of our words; though each rule is clear enough on its own (e.g. that we can speak of ‘yellowish red’, but not of ‘greenish red’), the totality of the rules seems disorderly, a motley or hotch-potch which displays no overall pattern. Wittgenstein’s earliest search for a perspicuous representation was addressed to this problem: he sought to display an order within a particular set of grammatical rules (our own particular descriptions or register of rules setting out how we speak about colours). Might the later version of the same remark (‘Our grammar is lacking in perspicuity’ (PI §122)) not express the very same complaint? In so far as the Philosophical Investigations is a search for perspicuous representations, the question to be answered is what it is that any of Wittgenstein’s perspicuous representations represents. Or perhaps the question should be whether this question even makes sense. Is the answer not tautologically ‘our grammar’? Our way of regarding the question

Remarks on ‘Language’ and ‘Grammar’

59

would be transformed, however, if we acknowledged that the phrase ‘our grammar’ is itself ambiguous and hence that the question had at least two different answers. ‘Our grammar’ here might mean either ‘the use of our words’ (what we describe in various ways) or ‘our descriptions of worduse’ (especially, but not exclusively, our compilations of what we call ‘explanations of what our words mean’). If the answer construed ‘our grammar’ in the first sense, then what is represented in a perspicuous representation would be something relatively fixed and stable, something which could (save for alterations in how we speak) be described once for all independent of all future experience (cf. PI §92). On the other hand, if the answer construed ‘our grammar’ in the second sense, then these implications would not hold; indeed, there might be good reasons to replace one form of description of the use of our words with another (e.g. to drop the uniform application of the schema ‘ . . . signifies . . . ’) even though what is described has undergone no change at all. We might come to see that this form of description of how we use our words is what surrounds the workings of our language with a fog (PI §5). Correlatively, we might have to re-examine the question whether the demand for ¨ bersichtlichkeit is unequivocally well founded or healthy, whether the U possibility of mastering the practice of speaking a language guarantees that it can be met, and whether the possibility of meeting the demand in some cases might not call for creativity in the form of novel ‘descriptions of the use of our words’ or even new forms of ‘grammatical descriptions’ (e.g. the truth-table notation of the Tractatus). We might consider recasting some of Wittgenstein’s best-known dicta in the role of presenting different ways of seeing the use of our words, i.e. different forms of representation of the grammar of our language. This would require a different attitude towards assessing the merits and demerits of his calling arithmetical equations ‘rules of grammar (or syntax)’ (PR 130, 143; PG 347), of his viewing ostensive definition as substitution-rules for symbols which remain within language rather than connecting words with the world (cf. BB 109), of his criticizing the application of the model of ‘object and designation’ to describing the use of ‘pain’ (PI §293), etc. Perhaps his principal goal was not to establish any facts of grammar, but rather to reveal or bring to the attention of willing readers neglected aspects or unnoticed patterns in what we say.

What We Say: Language and Speech Contemporary philosophers are concerned to make various distinctions. They commonly differentiate between type-sentences and token-sentences,

60

Reading Wittgenstein

between the sentence uttered by a speaker and the statement which is made by the speaker’s utterance, between a form of words and the uses to which it can be put in different circumstances, or between what a sentence means and what a speaker means by uttering it. These distinctions are often thought to be subtle and easily overlooked, but also to be of fundamental importance (e.g. in developing a sound philosophy of logic). All of these distinctions can be presented as ambiguities in the phrases ‘what is said’ or ‘to say the same thing’, and many of them can be marshalled under the two general headings of ‘language’ and ‘speech’ (or among linguists under the headings ‘langue’ and ‘parole’ or ‘semantics’ and ‘pragmatics’). Anybody properly trained in these matters will wish to apply these established distinctions to Wittgenstein’s thinking about symbols, calculi, language-games, etc. In translating or expounding his ideas we will be faced by the question what concept or set of concepts he expressed by each of his various uses of the phrases ‘die Sprache’ (and ‘unsere Sprache’). Sometimes he seems clearly to have meant ‘language’, while at others it seems to be ‘speech’ (or even ‘speaking’ or ‘speaking a language’). Of course, this problem would not be so very urgent were philosophers not inclined to attach such great importance to subtle differences between the various senses of ‘what is said’. But in view of the entrenched propensity to think that all these differences are BIG differences and to try to institutionalize them in semi-technical usage of ‘language’, ‘speech’, ‘sentence’, ‘statement’, ‘proposition’, ‘Gedanke’, etc., the exact understanding of each occurrence of ‘die Sprache’ comes to have a decisive importance for the overall interpretation of Wittgenstein’s philosophical investigations. I will examine two celebrated remarks whose content might be transformed were we to associate the phrase ‘die Sprache’ with what philosophers now call ‘speech’ rather than what they call ‘language’. (The authorized translations invite English-speaking philosophers to neglect alternative interpretations of these remarks. The more neutral, though less literal, translation of ‘die Sprache’ as ‘what is said’ or ‘what we say’ might have the merit of encouraging further thought about what Wittgenstein may have meant.) (1) In various contexts he tackled the ‘prejudice’ of limiting ‘die Sprache’ to the words and sentences which speakers utter. For example, it is this preconception which fosters the idea that there are many things which are ‘indescribable’ and which we ‘cannot put into words’ (e.g. the aroma of coffee or the sound of a clarinet (PI §§610, 78)). He suggested trying to look at things differently: we can consider gestures (e.g. pointing at something), miming an action (e.g. wiping sweat from one’s brow), drawings (e.g. geometrical diagrams accompanying a Euclidean proof ), or

Remarks on ‘Language’ and ‘Grammar’

61

samples (e.g. the colour-samples in language-game (8)) as instruments or parts of ‘die Sprache’ (§§16, 50), or as something belonging to the means of representation rather than what is represented (§50)). What should we make of this proposal? One response might be that Wittgenstein has here misused or abused the term ‘language’. Dictionaries, phrase books, and grammars exhaust the description of what is called ‘the English language’ (and the description of an imaginary language, e.g. the language-game of §8 or §48, would differ solely in being less complicated). These descriptions of a language do not include gestures, actions, drawings, samples, etc. Indeed, it seems odd to reckon most of these concrete symbols among the elements of English: some of them have no institutionalized status among English-speakers (e.g. the particular red wooden brick which I employed yesterday as a sample in teaching one of my children the word ‘red’), and others are not language-specific since they have a use among speakers of many different languages (e.g. the gesture of pointing or a geometrical diagram). In addition it seems inconceivable that we could incorporate gestures, actions, samples, etc. in what we call ‘a dictionary’, i.e. within the covers of a book. A critic might well turn Wittgenstein’s own strategy against him, urging that the expressions ‘language’, ‘parts of language’, ‘instruments of the language’, etc., should be brought back to their everyday uses! Even if he did not fall into literal nonsense by affirming that colour-samples are parts of language, how can it be any more useful to regard samples as if they were parts of language than it would be to look at numbers as if they were perceptible objects? Did he not simply abuse the term ‘language’ without producing any good reasons for extending the boundaries of what we commonly conceive of as language? These obvious objections could be short-circuited if we replaced the term ‘language’ with the phrase ‘what we say’. There seems no doubt that what someone says may be altered by his pointing to one object rather than another, by his producing one geometrical diagram rather than another, or by his indicating one sample rather than another. If we were to reckon among the ‘parts of what we say’ on a particular occasion anything which makes a difference to what is said (especially to a report of what is said in indirect speech), then we should in many cases include such items in this class. A witness in a trial would, e.g., have misrepresented what I said if he reported my utterance ‘You are lying’ as a statement about somebody who was not present when I spoke or if he reported it as a statement about X when I pointed to Y when I said ‘you’. It seems that Frege followed this line of reasoning in allocating to the sense of a token-reflexive utterance all indications of time, place, speaker, audience, etc.;14 such things must generally be known in these cases in order

62

Reading Wittgenstein

to understand what has been said or to decide whether a report in indirect speech of what has been said is correct. From this relatively uncontentious starting-point, we might make some cautious steps towards treating ‘parts of what we say’ as ‘parts of our language’ in a limited range of cases. First, if we were to compare the effect of substituting one gesture for another, one sample for another, etc., with the effect of replacing one word by another, then we might also view gestures, diagrams, etc. as symbols. On the same grounds, we might also consider the mouth that says ‘I have toothache’ as part of the symbol (cf. XVI. 230), the mechanical model of the ether which accompanies an exposition of the theory of electricity as part of the symbolism of the theory (BB 6) or the time and place of the utterance as parts of the expression of the thought.15 Of course, we might baulk at some or all of these suggestions, perhaps on the ground that symbols must be things the production of which is wholly within our control, whereas samples give hostages to fortune, and I cannot choose the mouth with which I say ‘I have toothache’ (BB 68; LPE 311). On the other hand, we might take a further step in the same direction: if certain gestures, actions, samples, etc. had sufficiently established, regular and institutionalized roles among English-speakers, we might see merit in comparing them with the words of the language. For a limited class of gestures, stylized actions, canonical samples, etc., we might even appropriate the label ‘instruments of the language’, or we might regard them as the elements of a symbol-system (perhaps a ‘gesture-language’) into which we can translate some elements of our word-language. With some caution and at a respectful distance we might follow Wittgenstein in allocating some samples, diagrams, gestures, etc. to language at least for some purposes of clarifying what we say. (This piecemeal approach might highlight some of the many unnoticed subtleties of his remarks.)16 (2) He framed various remarks of the form exemplified in the dictum ‘It is in language that an expectation and its fulfilment make contact’ (PI §445; cf. PR 65–9). Variants are the remarks ‘If you see the expression of the expectation, you see what is expected’ (PG 132), ‘Expecting that p will be the case must be the same as expecting that this expectation will be fulfilled’ (PR 65), ‘The answer to the question ‘‘What is the carrying out of the command p?’’ is a grammatical transformation of the command p and nothing more’ (PLP 119; cf. PR 65–6; PI §458) and ‘The proposition ‘‘p’’ determines that p must be the case in order to make it true; and that means: (the proposition p) ¼ (the proposition that the fact that p makes true)’ (PG 161). Wittgenstein’s position might be encapsulated in the slogan: ‘Like everything metaphysical the harmony between

Remarks on ‘Language’ and ‘Grammar’

63

thought and reality is to be found in the grammar of the language’ (PG 162). These remarks are sometimes regarded as holding the key to the dissolution of the ‘problem of the intentionality of mental states or acts’: the possibility of having the thought that Paris lies to the south of London presupposes command of some means of giving expression to this thought, and hence the thought can be seen to make reference to a spatial relation between two cities in virtue of the power of words (e.g. ‘Paris’ and ‘London’) to refer to things in the world. On this view, Wittgenstein turned on its head the familiar idea that it is the mind (especially mental acts of meaning and understanding) which gives life (i.e. sense or significance) to sounds and ink-marks (cf. BB 4). His claim seems to be that the concepts of thinking, meaning something by a word, understanding, etc. all presuppose independent concepts of signs, symbols, language, etc.: the mind is, as it were, a second-order capacity to acquire mastery of a system of symbols17 whose meanings are to be explained in terms of their use (cf. PI §43; BB 4–6). Whether this pattern of reasoning has any claim to being called Wittgenstein’s solution to the ‘problem of the intentionality of the mental’ depends in large measure on how exactly the phrase ‘in der Sprache’ is to be understood. What could he have meant by remarking that an expectation and its fulfilment make contact ‘in der Sprache’? The most immediate suggestion might be that the expression of an expectation and the description of the event which counts as its fulfilment must be identical or at least substantially the same. This would indeed be a form of ‘contact in language’, a case of forms of expression running together for a distance on parallel tracks (cf. PI p. 192). This idea, however, encounters a host of problems. What is to be counted as ‘the expression of an expectation’? There seem to be at least three different kinds of expressions whose relations to the description of the fulfilment need to be clarified; namely ‘The gun will go off ’, ‘I expect the gun to go off ’ or ‘I expect that the gun will go off ’, and ‘He expected that the gun would go off ’. Which of these expressions was the focus of Wittgenstein’s attention? Does the ¨ ußerung) and description (Beschreidistinction between expression (A bung) matter here? Or the distinction between describing my present mental state (PR 68) and predicting something? With respect to matching the description of the event which fulfils the expectation, each form of the expression of an expectation raises its own peculiar problems, e.g. about variation in the moods and tenses of verbs, systematic differences in pronouns and indexicals, and modifications of word-order; moreover, any claim that any of these expressions must exhibit parallel structures has the disquietingly dogmatic ring of an a priori limitation on the

64

Reading Wittgenstein

possible varieties of linguistic expression.18 The requirements for ‘contact’ might be loosened by allowing for standard forms of grammatical transformations, e.g. producing sentence-questions out of declarative statements, paraphrasing statements in indirect speech, or transforming first-person statements into third-person ones. But even this form of ‘contact’ between expressions does not cater for many variations in mood or tense, etc., or for logical relations (e.g. the temporal generality of a rule and the temporal specificity of an act of compliance with it). Building all of the required complexity into a description of what counts as ‘contact in language’ might make Wittgenstein’s remark lose much of its illuminating power.19 It might even reduce its content to the claim that anybody who can be said to understand any one of the expressions of an expectation must be able to explain what would count as its fulfilment. This would be a grammatical remark about the phrase ‘to understand an expression of an expectation’, but precisely because it involves the (mental) concepts of understanding and knowing how to explain words, it would not seem to provide any proper basis for solving the ‘mystery’ of the ‘intentionality of the mental’. Many of these doubts and difficulties would be circumvented if we interpreted ‘in der Sprache’ in a less restrictive way. Perhaps Wittgenstein could be best understood as remarking that an expectation and its fulfilment make contact ‘in what we say’, perhaps with the intention of undermining our desire to persist in posing the question ‘How does an event fulfil an expectation?’. The schema of ‘making contact in what we say’ is flexible enough to be filled out in many different ways in different cases. The case of expectation and its fulfilment may exemplify a particular pattern of identity in what is said, a pattern which is common to many other speech-acts and descriptions of mental states. For example, if I predict at midnight that the clock will stop at noon, and if the clock does subsequently stop at noon, then what I predicted at midnight (namely that the clock would stop at noon) is what I later do (or would) say actually happened at noon (namely that the clock did stop at noon); recognizing the identity of what is said despite variations in the construction of the two indirect statements is part of the grammar of one use of the phrase ‘what is said’.20 Different forms of identity of what is said might be taken to inform the relation of wishing for something to happen and the occurrence which is its fulfilment and the (different) relation of wondering whether something is the case and finding out that it is the case. In yet a different way, a standing order and particular acts of compliance with it might be said to make contact in ‘what is said’, and this might dissolve another perennial form of philosophical puzzlement.21 We often

Remarks on ‘Language’ and ‘Grammar’

65

say that part of what is said in issuing the general directive ‘Wind the clock every Sunday morning’ is that the child whose chore this is should wind the clock this morning if today is indeed Sunday; hence the general directive and the child’s act of winding the clock today are connected by our saying that part of what the child has been told to do is to wind the clock this morning. (This account calls for no special explanation of the phrase ‘part of what is said’; in particular, it does not make any demand that the general directive be regarded as a conjunction of orders specifying that the clock be wound on particular Sundays.) All of these suggestions about the variety of ways for things ‘to make contact in what we say’ are contributions to the grammar of oratio obliqua (indirect statements, indirect commands, etc.). We might also speak of these forms of ‘making contact in what we say’ as ‘making contact in our language’, provided that we saw that what is primarily at issue is the forms we actually use for describing what we say about things (cf. BB 23). The most important point about interpreting the phrase ‘in der Sprache’ is to capture the spirit of Wittgenstein’s remarks about expectation, thought, desire, intention, etc. He did not offer a novel set of solutions to ‘the problem of intentionality’. In particular, he did not subscribe to the thesis that the concepts of words, signs, symbols, language, etc. are independent of and logically anterior to concepts of ‘the mental’ (expectation, thinking, etc.). On the contrary, he emphasized that the concept of a language is interwoven with the concepts of meaning and understanding, of communicating thoughts, of expressing emotions or sensations, etc. He also did not subscribe to the thesis that something must be added to ink-marks and sounds in order to make them into linguistic expressions capable of expressing ideas or of referring to things. On the contrary, characterizing something as a sign or symbol presupposes that it has a distinctive kind of use within a normative practice. In conformity with his usual procedure, he sought to dissolve the puzzles which make up ‘the problem of intentionality’. The idea that it is mental acts (of meaning and understanding) which connect language and the world is not the wrong answer to an important philosophical question,22 but rather an answer to the wrong question (one which he thought that we can be brought to acknowledge is nonsensical). The antithetical idea that it is linguistic expressions which forge links between thought and reality (or which explain how mental states or acts can refer to things in the world) can equally be made to appear to be an answer to a nonsensical question. In this respect, ‘It is in language that an expectation and its fulfilment make contact’ is precisely comparable to the remark ‘The equation ‘‘2 þ 3 ¼ 5’’ is a rule of grammar.’ Both may give the appearance of being explanations of necessary truths in terms of

66

Reading Wittgenstein

linguistic conventions, but in reality both were intended to demolish the very framework in which the metaphysical harmonies between thought, language and the world appear to be problematic. The two cases we have just examined of Wittgenstein’s use of the phrase ‘die Sprache’ exhibit very clearly the fact that the exact interpretation to be put on this phrase in any particular context may be of decisive importance for the interpretation of remarks which are pivotal for understanding his philosophical investigations. Like a railway wagon which must be placed exactly on the rails in order to move along a track, our expositions of many of his ideas will run on the rails of his thinking only if we arrive at an understanding of precisely what he meant by each employment of the phrase ‘die Sprache’. This is an extremely subtle and delicate task of interpretation which needs to be carried out case by case, not something capable of being settled once-for-all in advance of considering the intricate details of his thinking.

Starting Again from Scratch What I have attempted here is to demonstrate by examples a method of reading Wittgenstein’s texts and expounding his ideas. This demonstration can be broken off at any point, once its nature has been made clear. Piling up more examples would achieve little, since the method must be justified by its works, i.e. by its clarifying Wittgenstein’s remarks from case to case. But a useful coda to this investigation might be an indication of some of the distinctive general features of this method (a parallel to his description of the methods of his own philosophical investigations in PI §§89–133). First, I would advocate a strong principle of charity: we should proceed on the basis that the texts which Wittgenstein constructed himself consist of carefully thought out arrangements of remarks whose precise wording was of paramount importance. (Alas, this principle does not apply either to the texts compiled by editors in various more or less systematic ways from his manuscripts or to texts compiled from lecture notes taken down by others.) In these authoritative texts we have no right to assume that a remark which is repeated more or less verbatim has the same significance in its various occurrences, still less to presume that an earlier occurrence can be used to establish how it is to be understood in a later text; such evidence should be treated as presumptive and defeasible. Correspondingly, we have no business arguing that a word or phrase which recurs in various remarks must have the same significance in them all. This is certainly not true of many expressions which are crucial to the interpret-

Remarks on ‘Language’ and ‘Grammar’

67

ation of Wittgenstein’s thinking. It is demonstrable that such terms as ‘Satz’, ‘Sprache’, ‘Objekt’, ‘Ta¨tigkeit’, ‘Vorgang’, and ‘Grammatik’ have different meanings in some different contexts, and that the most accurate translations into English must vary from one context to another or be taken to show corresponding ambiguities. This is not a defect in Wittgenstein’s writing, something which we should assume that he would have eliminated had it been pointed out to him. Nor does it justify either despair about understanding his ideas or our pinning interpretations randomly on to ambiguous expressions in his writings. We should aim at giving sensitive readings of his remarks which might in many cases fully justify giving various different interpretations of certain words or phrases. Indeed, we might well expect that careful probing of the reasons supporting our variable interpretations of these expressions might be a powerful source of illumination of his ideas. Second, I would urge a kind of minimalism: given our strong cravings for generality and the inclination to extract generalizations from his investigations of specific concepts or language-games, we should try to relate every remark to the specific arguments which support it and to the particular purpose which informs the surrounding remarks, and we should attach to each expression the interpretation which gives it the minimum generality compatible with the context.23 Wittgenstein clearly abstained from drawing many of the general conclusions which are taken to be the leading ideas of his philosophy, especially from employing generalizations as premises for making inferences to other conclusions about the grammar of our language, and we might consider respecting his reticence as an essential aspect of his thinking (an integral part of his avoiding dogmatism) rather than dismissing it as a stylistic quirk. Just as he advocated treating a mathematical theorem as a cross-section of the body of the proof which supports it, so we might try taking each of his philosophical epigrams as a cross-section of the body of comment (reasoning, dialogue, analogies, language-games, etc.) which informs it. Each remark would then be absolutely context-relative and purpose-specific. Perhaps it was this aspect of his work which Wittgenstein sought to highlight in the remark: ‘The solution of philosophical problems can be compared with a gift in a fairy tale: in the magic castle it appears enchanted and if you look at it outside in daylight it is nothing but an ordinary bit of iron (or something of the sort)’ (CV 11). Our very attempt to appropriate and carry off his nuggets of gold for our own purposes may guarantee that we will find nothing but scraps of rusty iron in our hands when we have arrived home with our booty. Third, I suggest scrupulous attention to Wittgenstein’s overall therapeutic conception of his philosophical investigations: far from advocating

68

Reading Wittgenstein

any general positive position (whether anti-realism, conventionalism, anthropological idealism, or whatever) and far from undertaking to give any general outline of the logical geography of our language (or even of the narrower domain of ‘mentalistic’ or ‘psychological’ concepts), he always sought to address specific philosophical problems of definite individuals and to bring to light conceptual confusions which these individuals would acknowledge as a form of entanglement in their own rules. He did not make direct assaults on various standard ‘isms’ (whether Cartesian dualism in philosophy of mind or Intuitionism in philosophy of mathematics); nor did he pretend to overwhelm these positions by an accumulated series of indirect attacks. Rather he confronted specific remarks of specific writers (sometimes Augustine, Plato, James, or Ko¨hler, often Frege, Russell, and the author of the Tractatus). He did not see himself in the role of a public health official whose brief was to eradicate smallpox from the face of the earth (e.g. to eliminate Cartesian dualism once for all by means of the Private Language Argument). Rather he operated as a general practitioner who treated the bumps that various individual patients had got by running their heads up against the limits of language (cf. §119). Or, more accurately, he compared his procedure with psychoanalysis: his aim was to bring each patient to acknowledge the origins of her particular conceptual disorders (especially in the workings of analogies or pictures of which she was not conscious), and the patient’s own acknowledgement of the rules in which she is entangled is a precondition of the correctness of the diagnosis (BT 410) as well as of the effectiveness of the cure (BT 410). Wittgenstein’s practice in philosophizing is not less, but rather more, consistently therapeutic than we commonly recognize. His interest is not to remedy ignorance about segments of a complex practice of which we have individually only partial mastery, nor is it to establish objective facts about shared institutions of which we may have temporarily lost sight and need to be reminded. He was far more concerned with making propaganda for different points of view, with exploring neglected possibilities, with trying to effect changes in our ways of seeing things, indeed with bringing about changes of mind or modifying the will (how we want to see things). For achieving these purposes the construction or assessment of abstract proofs and refutations would be manifestly peripheral or even irrelevant. Wittgenstein’s enterprise is essentially person-relative, and it centres on the dynamics of somebody’s thinking, not on the geometry of thoughts. Finally, it is vital to recognize the variety of remarks which Wittgenstein offered as descriptions of ‘the grammar of our language’ and the variety of roles which they play in his investigations. We are strongly inclined to

Remarks on ‘Language’ and ‘Grammar’

69

regard all of them as extracts from the schedule of rules which govern the normative practice of speaking a natural language and hence to treat them each as part of what competent speakers would acknowledge to be correct explanations of the meanings of certain words. This idea manifestly jars with Wittgenstein’s own procedure. Many remarks have very different roles. Some are enumerations of rules for using symbols in simple or even wholly imaginary language-games (§§2, 4, 8, etc.); they may throw light on our practices, in respect of both similarities and differences, and to this extent they may be called ‘grammatical remarks’, but it would not be natural to describe them as parts of the grammar of our language. Other remarks draw comparisons or offer analogies. For example, to note that a colour-sample in language-game (8) can be regarded as an instrument of the language is certainly not to formulate part of an explicit explanation of any colour-word within language-game (8), nor is it even to give part of any standard explanation of how we are to use the word ‘sample’ in our language. Similarly, to call the equation ‘2 þ 3 ¼ 5’ a rule of grammar is no part of an explanation of how correctly to use the phrase ‘a rule of grammar’, but rather an analogy drawn for the specific purpose of calling attention to an important aspect of the role of equations in our language (or form of life). Yet other remarks concern what we say about what we say. For example, we are invited to attend to how we use the expression ‘a name of a sensation’ (LPE 290–1; cf. PI §261), and the clarification of the concepts of thinking, expectation, hoping, intending, etc. turns in part on describing the grammar of reports in indirect speech (various forms of ‘making contact in our language’). The presumption that all the things which Wittgenstein would presumably have called ‘descriptions of grammar’ are homogeneous in function is a major obstacle to understanding many of his conceptual investigations. Still more damaging is the subsidiary idea that these descriptions are additive. A comparison which is illuminating for one purpose may be unhelpful or even obfuscating in another context, and a pair of seemingly inconsistent analogies may facilitate grasping different aspects of a single thing. What Wittgenstein called a perspicuous representation of our grammar would perhaps never result from mere accumulation of what are given and accepted in everyday practice as correct explanations of the meanings of words. In any case, what brings about somebody’s ‘knowing her way about’ in the language is often the judicious choice of a new object of comparison, a creative analogy, or the revelation of a new aspect of the use of our words. In respect of Wittgenstein’s own ‘descriptions of the grammar of our language’, there is some merit in our appropriating the motto: ‘I’ll teach you differences.’

70

Reading Wittgenstein

Notes 1

2

3

4

5

6 7 8

In respect of his philosophy of mathematics, I have tried to demonstrate that the widespread practice of labelling his thinking ‘conventionalism’ does not square with his overall intentions and distorts understanding of many of his best-known remarks (e.g. ‘Arithmetical equations are rules of grammar’ and ‘The sense of a mathematical proposition is given by its proof ’). Cf. Part II of my book Wittgenstein, Frege and the Vienna Circle (1988). Compare this remark: ‘It is natural for us to call gestures, as those employed in (4), or pictures as in (7), elements or instruments of language’ (The Brown Book, p. 84; italics added). Ironically, the neglect of such qualifications, and even of modal auxiliaries such as ‘need not’, ‘may’, etc., is a conspicuous aspect of many expositions and analyses of Wittgenstein’s ideas, as if these niceties were not worthy of attention among philosophers. If Wittgenstein meant to employ the samples used in language-game (8) merely as objects of comparison (rather than as prototypes intended to support some generalization), then the project of classifying samples into species, in particular of distinguishing ‘optional samples’ from ‘canonical samples’, might be held to be irrelevant to clarifying PI §§16 and 51. (Cf. Baker and Hacker, Wittgenstein: Understanding and Meaning, pp. 196–201.) This point is as true of the Tractatus as of the Investigations. In the former, logical syntax is misrepresented as a kind of abstraction from the syntactic categories of natural languages, whereas it is meant to lay down a priori the essential structure of any possible language; the thesis that every proposition must be complex is treated as a potentially falsifiable generalization about English, German, Latin, etc., whereas it is meant to indicate part of the essence of a propositional-sign, etc. In the latter, remarks (53 ff.) about the particular language-game (48) are generally treated as generalizations about samples, names, rules of language-games, etc. and taken to be explicit criticisms of the doctrine of unanalysable names and simple objects in the Tractatus; the suggestion of a possibility for connecting the word ‘pain’ with primitive expressions of pain (§244) is taken to state how sensation-words in English are in fact connected with sensations or even to formulate how any sensationword in any possible language must be taught and learned; etc. Commentators on Wittgenstein seem to be pulled by powerful gravitational forces towards assimilating all of his remarks to factual observations about the logical geography of natural languages. For a detailed interpretation of this passage, see my paper ‘Philosophical Investigations §122: Neglected Aspects’ (1991), pp. 35–68. [Essay 1 of this volume Ed.] There is a related problem of whether the first-person plural is to be interpreted uniformly throughout this remark. Namely English, German, French, etc. Presumably a matter of making nonsensical assertions and framing unintelligible questions, e.g. because we commit ‘category-mistakes’.

Remarks on ‘Language’ and ‘Grammar’ 9

10

11 12

13

71

Contrast the standard conception of formal definitions among logicians: ‘a definition is, strictly speaking, no part of the subject in which it occurs. . . . Theoretically, it is unnecessary ever to give a definition: we might always use the definiens instead, and thus wholly dispense with the definiendum. . . . theoretically, all definitions are superfluous’ (B. Russell and A. N. Whitehead, Principia Mathematica, vol. 1, p. 10). Wittgenstein had a very rich conception of what he called ‘a form of representation’. Under this heading, he included not only the propositions of arithmetic and geometry, but also such metaphysical propositions as ‘Every event must have a cause’, ‘Every mental state must correspond to a state of the brain’, or ‘Every disposition must be realized in a structure’. Following up this line of thought, he considered the claim that the difference in the development of a poppy and a rose must originate from a difference in the microscopic structures of their seeds; he suggested that this amounts to adopting a form of representation. But he emphatically denied that this way of seeing the claim makes it trivial: on the contrary, to abandon this dogma would be ‘a tremendous thing to do – as great as recognizing indeterminacy’ (UW 434). We risk misunderstanding the dimension of depth in our conventions if we describe this as a mere change in our form of representation. (The deep aspect of this matter easily eludes us!) Compare: ‘It is not a contingent feature of language that its grammar is surveyable’ (Baker and Hacker, Wittgenstein, p. 544). This interpretation has an affinity with Dummett’s idea that a philosophical theory of meaning for a natural language is ‘the theoretical representation of a practical ability’. He too thinks that it must be possible in principle to construct a ‘systematic theory of meaning’. Though he thinks that he detects in Wittgenstein’s work an antipathy to this claim, he argues that this cannot have any cogent support. Not only would it be an unwarrantedly ‘defeatist’ doctrine, but also it would run counter to the obvious fact that anybody who has a mastery of a language has a capacity to understand an infinity of sentences on the basis of ‘an implicit grasp of a number of general principles governing the use in sentences of the words of the language’ (Dummett, Truth and Other Enigmas, p. 451). In addition, Wittgenstein did himself give ‘a complete systematic account of the functioning of a miniature language’ in describing each of his illustrative language-games, and this makes it difficult to see any conclusive reason for denying the possibility of giving similar accounts of an entire natural language. Indeed, it seems that we could ‘command a clear view of the working of our language’ only by attaining to ‘a systematic description’ of the use of our words (Dummett, The Logical Basis of Metaphysics, pp. 13 and 163). He sometimes used ‘descriptions’ and ‘rules’ as alternatives when combined with the adjective ‘grammatical’ or the phrase ‘of grammar’. On the other hand, many of his remarks about the use of our words, though presumably part of his project of describing grammar, are not even plausible candidates for ordinary explanations of word-meaning (or rules for the use of words whose grammar is described).

72

Reading Wittgenstein

14

Frege, ‘Thoughts’, in Collected Papers on Mathematics, Logic and Philosophy, p. 358. Ibid., p. 358. For example, important differences in the role of particular samples in particular language-games (e.g. the coloured bricks shown to a child and a paint manufacturer’s colour-chart, or an ordinary metre-stick and the standard metre in Se`vres used to define the unit of metric linear measurement) might be brought out by contrasting samples which belong to grammar with those which are merely instruments of the language and with those which are merely parts of what we say. This phrase is taken from Kenny, The Metaphysics of Mind, p. 20. The idea is developed by him as the central element of Wittgenstein’s philosophy of mind. This worry should perhaps be set aside. Compare: ‘Could we imagine any language at all in which expecting p was described without using ‘‘p’’? Isn’t that just as impossible as a language in which ~p would be expressed without using ‘‘p’’?’ (PR 69). It might be thought to trivialize the claim that the internal relations between the uses of sentences are to be explained by saying that ‘assertion, question, supposition, command, etc. make contact in language’ (Baker and Hacker, Wittgenstein, p. 131). It would certainly not meet the widespread desideratum of exhibiting some element (a ‘neustic’, ‘proposition-radical’, or ‘descriptive content’) which is common to sentences which differ in ‘mood’ or ‘force’. In some cases – but not in all – there is an even simpler grammatical relation, i.e. an exact identity of two indirect statements. For instance, we might say that it is part of the grammar of ‘belief ’ that the belief that the clock has stopped is the belief which is made true by its being the case that the clock has stopped. The ‘problem of imperatival inference’: namely how is it possible for one order to follow from another order (given that logical consequence depends on the relation between the truth-conditions of the premises of an argument to the truth-conditions of the conclusion)? Cf. Hare, The Language of Morals, cc. 2–3; Kenny, Will, Freedom and Power, ch. 5. In particular, it is not Wittgenstein’s intention to answer the question ‘What must be added to mere signs to give them life?’ by citing ‘use’ or ‘the practice or technique of speaking a language’. A sign is no more separable from its use than a lever is from its employment. (We might say that there is no such thing as a name which does not belong to a particular language-game (cf. PI §49).) This strategy, if applied to some of his ‘methodological remarks’, might transform our understanding of his philosophical activity. Consider: ‘What we do is to bring words back from their metaphysical to their everyday use’ (PI §116) or ‘The concept of a perspicuous representation [u ¨ bersichtliche Darstellung] is of fundamental importance for us ’ (§122, italics added). Are these remarks to be classified as part of his description of the nature (essence) of philosophy?

15 16

17

18

19

20

21

22

23

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

3 Wittgenstein’s ‘Depth Grammar’*

PI §664: In the use of words one might distinguish ‘surface grammar’ from ‘depth grammar’. What immediately impresses itself upon us about the use of a word is the way it is used in the construction of sentences [Satzbau], the part of its use – one might say – that can be taken in by the ear. – And now compare the depth grammar, say of the word ‘to mean’, with what its surface grammar would lead us to suspect. No wonder we find it difficult to know our way about [sich auszukennen].1

In part because its terminology is unique in his published writings, this is a frequently cited remark. It indicates the need to grasp depth grammar in order to avoid philosophical confusion, i.e. to avoid the predicament of not knowing one’s way about (PI §123). Hence it signals Wittgenstein’s own intention to clarify the depth grammar of such expressions as ‘to mean’. Presumably this is his general programme in describing ‘the use of our words’ or ‘the grammar of our language’. We might even rephrase §371: Essence is expressed by ‘depth grammar’.

At the same time, PI §664 gives no clear explanation of the terms ‘surface grammar’ and ‘depth grammar’. What is the intended contrast? The ‘scare quotes’ seem to signal his using ‘depth grammar’ in a special sense. Perhaps they are used specifically to forestall the danger of the * Originally published in Language & Communication, 21 (2001), pp. 303–19. ß 2001 by Elsevier Science Ltd. Reproduced by permission.

74

Reading Wittgenstein

reader’s taking this conception of grammar to come into conflict with his concern to ‘understand something that is already in plain view’ (§89) and to make things ‘surveyable [u ¨ bersichtlich] by a rearrangement [Ordnung]’ (§92). How is all this to be clarified? Wittgenstein is usually taken to draw a distinction here between sorting words into kinds according to crude grammatical classification (e.g. verb, noun, adjective) and allocating words into categories according to ‘logical syntax’ or ‘philosophical grammar’ (e.g. colour-words, numerals, names of persons, sensation-names, propositional connectives, quantifiers, n-place predicates signifying sensible properties or relations, numerical properties or relations, etc.).2 On this reading, PI §664 contrasts different criteria for differentiating word-types [Wortarten] and for determining which combinations of words are well formed. Its concern is to distinguish two different ways of considering sentence-construction [Satzbau]. ‘Depth grammar’ is taken to impose more restrictive conditions on legitimate word-combinations than does ‘surface grammar’. For example, depth grammar is what demonstrates the illegitimacy of the question ‘When did you understand what she said? When she uttered the first word? Not until she reached the end of the sentence? Immediately thereupon? If not, how long after she finished speaking?’ (Compare ‘Before you can execute an order, you must understand it.’ – ‘How long before?’) ‘Depth grammar’ is taken to govern the combinatorial possibilities of words in constructing significant sentences,3 i.e. to delineate the logical types of terms and the logical forms of propositions4 It is assumed that conformity with depth grammar guarantees that a sentence is significant,5 and contravention of these rules ensures that a sentence is nonsensical.6 Surface grammar, by contrast, is useless for philosophers since it may let (logical) nonsense slip through the net.7 According to this conception, combining expressions in contravention of the rules of depth grammar produces strings of words which are logically nonsensical. (This defect may not be immediately apparent even to competent speakers of the language.) It is concerned with the logical dimension of sentence-construction, what Carnap called the logical syntax of language and Ryle called logical geography. Infringements of logical grammar are often called ‘category-mistakes’.8 Some cases may yield patent nonsense, even logical jokes. (‘Who passed you on the road?’ – ‘Nobody.’ – ‘That’s impossible: he would have got here first.’) Others may be subtle and difficult to detect. (The nonsensicality of calling thinking an activity, or the seductive error, allegedly informing Cartesian dualism, of misrepresenting achievement-verbs as task-verbs and capacities as private mental performances.9) The whole business of ‘ordinary language philosophy’ may be thought to be the dissolution of philoso-

Wittgenstein’s ‘Depth Grammar’ 75

phers’ confusions by exhibiting category-mistakes in their utterances.10 (Cartesian dualism is diagnosed as a philosopher’s myth arising largely from misrepresenting the contrast mental/physical as a difference within each of the categories of things, stuffs, attributes, states, processes, changes, causes and effects.11) Wittgenstein is widely considered to have practised precisely this method in the Philosophical Investigations.12 He gave ‘descriptions of the grammar of our language’ for the purpose of correcting philosophers’ misconceptions and regulating the use of words in philosophical discourse.13 This is the message universally extracted from his famous declaration: (§116) What we do is to bring words back from their metaphysical to their everyday use.14

What he did is to impose category disciplines, to put words in their proper conceptual harnesses.15 Though deeply entrenched, this interpretation of his method seems quite speculative.16 What support is forthcoming for it from close examination of the text of PI §664? ‘Surface grammar’ is here identified with what immediately impresses itself upon us about the use of a word in the construction of sentences [Satzbau]. (Note that these italics are omitted in the English translation.) This is described as a part of its use. Namely the part of its use – one might say – which can be taken in by the ear. What is this phrase meant to cover? What is the intended contrast?, i.e. what is the other part of its use? And what is the force of the phrase ‘one might say’ in qualifying the phrase ‘can be taken in by the ear’? The concern with sentence-construction and with ‘the ear’ seems to point towards linking ‘surface grammar’ to the notion of Satzklang (PI §134), i.e. to whether a string of words sounds like a (well-formed) sentence or has the ‘ring’ of a proposition. This seems none too precise, yet Satzklang must be something that can be taken in by the ear when the words are spoken, though presumably only by somebody who has mastery of a particular technique (roughly, speaking the language to which the vocabulary belongs17). The need for this expertise might explain the qualification ‘one might say’: this is a different sense of ‘taking something in by the ear’ than hearing the first consonant in the word ‘help’ or the rising intonation contour of the question ‘Is it raining?’ A comparable case of ‘taking something in by the ear’ might be hearing a string of musical notes as a theme, e.g. as the basis for a set of variations or for the construction of a fugue; or hearing a second theme as an inversion of

76

Reading Wittgenstein

the first. (We might rephrase PI §664 to read: . . . the part of its use which can be ‘taken in by the ear’ – using ‘scare-quotes’ in Wittgenstein’s manner to signal that a phrase is to be taken in a special sense.) We should note, however, that the idea of Satzklang might be explained and developed in very different ways. One would be to invoke only basic syntactic categories (noun, verb, adverb, . . . ) and sentence-forms (subject-predicate, relative clause, . . . ). Then, e.g., ‘This statement is false’, ‘Green ideas sleep furiously’ or even ‘’Twas brillig and the slithy toves did gyre and gimble in the wabe’ might be thought to be well formed in this respect.18 But a different proposal would be to link Satzklang to the word-types typical of logical or philosophical grammar (numeral, colourword, sensation-name, . . . ).19 In some cases, grammatically well-formed sentences seem clearly to fail to make intelligible statements or to be logically well formed; e.g. ‘Middle C on this piano sounds putrid’ or ‘The number p is courageous’.20 In other cases, failure to be meaningful may be less obvious, so that we need to develop sensitivity for the absence of Satzklang (a ‘nose’ for nonsense, like a countrywoman’s ‘nose’ for choice wild mushrooms). But even in this case, why shouldn’t we say that the oddity (or nonsensicality) of any logically ill-formed expression is something that can be taken in by the ear with suitable training? Indeed, isn’t a major skill of an analytic philosopher precisely to find ways of making nonsense, as it were, audible?21 ‘Before you can execute an order you must understand it’ may sound impeccable; but perhaps no longer so once one has been confronted with the question ‘How long before?’ The question may effect an aspect-switch, so that the oddity of the phrase has suddenly become something that can be ‘taken in by the ear’.22 (Someone may suddenly exclaim ‘Now that sounds odd’.)23 More generally, one might say that the category-clashes beloved of certain analytic philosophers (Carnap, Ryle) are meant to cultivate an ‘ear’ for what should sound strange. That seems the basis for judging ‘We don’t say that . . . ’ or ‘It is nonsense to say. . . ’. (‘When I said ‘‘He . . . ’’ I meant the Prime Minister, but I ceased meaning this when I began the next sentence.’) Some philosophers now prefer to phrase such remarks in the form: ‘Our intuitions tell us that you can’t say. . . . ’ Apparently, they do not take such judgements to be mediated by reasoning.24 The skill of an expositor of philosophical grammar (or of ‘logical syntax’?) is to take an expression that sounds legitimate and so manipulate its elements as to produce something that sounds ill formed. This is the standard method for making latent nonsense into patent nonsense. Correspondingly, the understanding of a student is perfected by coming to hear nonsense whenever it crops up, however covert its occurrence may be.

Wittgenstein’s ‘Depth Grammar’ 77

The contrast between these two senses of ‘Satzklang’ gives at least two options for identifying what Wittgenstein meant to pick out as the part of the use of a word – one might say – that can be taken in by the ear. In interpreting PI §664, we are faced with a choice between at least two readings. How is this to be made? Note that both options call for sophisticated senses of the phrase ‘can be taken in by the ear’. Hence both merit the qualification ‘one might say’. In both cases, the relevant features of utterances can be discerned only by somebody who has mastered a specific technique – a different technique in each case.25 Hence neither possible reading of this phrase makes a more straightforward use of the metaphor of something’s being on the surface than the alternative reading. These reflections open up a hitherto neglected possibility. We could take the phrase ‘surface grammar’ to signify the rules of logical grammar which impose restrictions on the combinatorial possibilities of words in framing meaningful (significant) sentences, i.e. the rules which govern sentence-construction (Satzbau) from a logical point of view [what could be called ‘logische Satzbau’ (§102)]. This would connect ‘surface grammar’ with a major preoccupation of the philosophers who are the principal explicit targets of critical investigation in PI: namely Frege, Russell and the author of TLP and RLF. (All three of them focused on the logical analysis of propositions, even though all three schemes of analysis were logically incommensurable.) Restoring the missing italics to the English translation of ‘Satzbau’ might help to reinforce this interpretation. Rules for constructing significant sentences could be precisely what Wittgenstein meant to pick out there by the phrase ‘surface grammar’, hence not by the phrase ‘depth grammar’.26 If so, his declared intention to describe depth grammar was presumably to differentiate his later method of describing the grammar of our language from the standard method of constructing a classification of words into logical categories or types (namely, investigating intersubstitutability salva significatione). His primary concern in ‘depth grammar’ must then be different from pointing out category-mistakes27 or clarifying type-restrictions or combinatorial possibilities and impossibilities. But now there seems to be an overwhelming objection. What work would this deeper conception of ‘surface grammar’ leave for a philosopher to do? How is it possible, on this reading, to allocate anything at all to ‘depth grammar’? What could this phrase mean? Evidently, it must signify what cannot be taken in by the ear at all, i.e. the part of the use of a word which has nothing to do with the construction of sentences.28 What kinds of things are these?29 Are there any other possibilities for philosophical clarification of concepts?

78

Reading Wittgenstein

Wittgenstein called attention to several things which fill this bill, and he stressed their importance over and over again. (1) Differences in the ways individual words are integrated into human activity, the different ways of operating with words. A clear paradigm is noting the differences in the way the shopkeeper reacts to the three words in the order ‘Five red apples’, namely looking up ‘red’ on a colour-chart and repeating the sequence of numerals ‘1, 2, 3, 4, 5’ (PI §1); describing the use of these words is here confined to attending to these two different procedures, and nothing whatever is stated about sentence-construction (e.g. about the oddity of ‘red five apples’). Parallel cases are the different uses that we make of proper names of persons, e.g. summoning them, pinning labels on their lapels, or reserving places at a table with place-cards (PI §27; cf. PLP 212–13); the difference between using cards or numerals in a bounded or in an open game (BB 91–5); the difference between the occurrence of the word ‘restaurant’ in a dictionary and the posting of the sign ‘RESTAURANT’ over the entrance to a building (PLP 317–19); attention to various instruments and procedures used for determining lengths or distances (PI §50; PLP 14; LFM 42, 158, 273–4); or differences in whether word-order functions to determine the order of actions in compliance with a compound command (BB 83). The example of the shopkeeper is meant to introduce a conception of meaning antithetical to Augustine’s picture of the essence of language. Clarifying rules for wordcombination is irrelevant in this case, and these differences in use are located, as it were, in a different dimension. Occasionally Wittgenstein suggested paraphrases to highlight ways of operating with words; e.g. instead of saying ‘in ten minutes’, we might say ‘in ten minutes by the clock’. In a similar spirit, we might paraphrase the order ‘five apples’ in PI §1 as ‘five apples to be counted out’.30 A related idea is clarifying conceptual similarities and differences in respect of human reactions. There is an important similarity between pain and pleasure in that both are evident in someone’s eyes, face or posture.31 Likewise, pain has a unique position among sensations in that it has strong and characteristic manifestations in behaviour just as the emotions of fear, rage and joy do (RPP II §63). Finally, one person’s pain may elicit strong, ‘primitive’32 responses of sympathy or anguish in others (RPP II §27–8). These remarks all contribute to describing the grammar of ‘pain’, but they do so by relating pain-ascription to ‘the stream of life’. (2) Differences in the ways complete sentences are employed (PI §421; cf. §23). This dimension of use falls largely outside the scope of principles

Wittgenstein’s ‘Depth Grammar’ 79

for sentence-construction.33 Paradigms include the distinctive use of some first-person present-tense declarative sentences as signals (§181) or expressions (§244) of abilities, sensations, etc.; the distinctive use of arithmetical equations which is clarified by comparing them with rules of grammar (PR 130; PG 319–20; LFM 55, 103); and the difference in role between ostensive definitions and applications of words (PI §§26, 49; PLP 13–14) or the similarity in function between ostensive and verbal definitions (BB 109; PLP 277–9; cf. PR 78–9). ‘Logical grammar’ as developed from the work of Frege and Russell has little to say about this dimension of the use of symbols.34 Wittgenstein’s analogy of symbols with tools or instruments can be applied directly to sentences as well as words or phrases (PI §§11–12). The different applications of sentences are not presented to us clearly – especially not when we are philosophizing in the ways typical of philosophical analysis.35 (Frege was concerned with what number-words designate, but not at all with what role arithmetical equations play in our lives. As if that were unimportant or would somehow ‘flow from’ the nature of numbers.) These differences in use cannot be taken in by the ear. We have to investigate the whole language-game, particularly the complex patterns of human activities into which utterances are interwoven (PI §7d). We need to clarify not principles of sentence-construction, but the circumstances which surround sentenceproduction. This investigation may yield verdicts contrary to standard ones based exclusively on attention to sentence-construction. ‘2 þ 3 ¼ 4’ is built on the same pattern as ‘2 þ 3 ¼ 5’. So it is normal to consider both as arithmetical propositions, the first one being false, the second true. But in putting emphasis on the use of sentences, Wittgenstein offered a different conception. He notes that ‘2 þ 3 ¼ 4’ has no application at all in our practices of counting and measuring; it is not relevant to the use of numerals in framing numerical propositions. It is not part of the grammar of number-words, hence not part of the language-game of arithmetic. Relative to this practice it is nonsensical. He recommended saying: ‘ ‘‘2 þ 3 ¼ 4’’ is not an arithmetical proposition’ rather than ‘ ‘‘2 þ 3 ¼ 4’’ is a false arithmetical proposition’.36 (And he contrasted the negation of ‘2 þ 3 ¼ 4’, which is also nonsensical, with the arithmetical inequation ‘2 þ 3 : ¼ 4’ [PR 247–51].) This is might seem to be merely a terminological dispute, but it is in fact more like a clash of Weltanschauungen. The point to be stressed is that we don’t operate in the same ways with ‘2 þ 3 ¼ 4’ and ‘2 þ 3 ¼ 5’, and that is a fundamental difference quite invisible from the view-point of sentenceconstruction.

80

Reading Wittgenstein

(3) The dependence of the question whether a particular utterance of a well-formed sentence (one having unexceptionable Satzklang) really makes sense (i.e. has a role in a language-game) on the circumstances surrounding its utterance. Logical grammar tends to treat rules for the ways words are used in the construction of sentences as something fixed and antecedent to speech-production; sentences (type-sentences) are regarded as correct or incorrect absolutely.37 (This is a leading principle of TLP: whether a sentence has sense cannot depend on whether another proposition is true (2.0211–2.022, 4.061).38) Wittgenstein’s way of connecting sense with use leads to a radically different conception. One tokensentence may make sense while another token of the same type-sentence makes no sense. (More generally (AWL 20–1): The words ‘nonsense’ and ‘sense’ get their meaning only in particular cases and may vary from case to case.) For example, the fact there is a familiar use of ‘This is how things are’ does not imply that it makes sense to enunciate this sentence in a philosophical treatise, a fortiori to claim that, taken in isolation, it formulates the general form of the proposition (PI §134). Similarly, if somebody had an arm 10 feet long and put it down a hole, we would have no difficulty in making sense of his statement ‘I feel water ten feet underground’; but if a normal human being makes the same statement in similar circumstances, we would make no sense of it (BB 9–10). Treating the question whether or not a sentence is significant as dependent on the circumstances in which it is uttered completely undermines the philosophical ideal of constructing a unique complete classification of expressions into logical types.39 Indeed, that particular philosophical obsession (or prejudice) may itself need to be eliminated in order to dissolve particular philosophical confusions. What makes sense is also treated as strictly relative to particular language-games.40 We can imagine a language-game of counting with the system of numerals ‘1, 2, 3, 4, 5, many’, and this game would have its own distinctive arithmetic, e.g. 2 þ 4 ¼ 3 þ 5 (BT 209; PLP 78 ff.; PG 321). There is a strong inclination to regard such systems of communication as ‘gappy’ or ‘incomplete’, but Wittgenstein urged us to fight against this. We should say not ‘I don’t know how to answer the question what ‘‘ . . . ’’ means, but rather that there is no such thing as an answer in this game’ (cf. PR 256; RFM 214; BB 25). The primitive game does not contain a shadowy anticipation of our more complex procedure of counting; rather these are two distinct language-games which have a very partial isomorphism.41 In a parallel argument, Wittgenstein discussed a language-game in which a child is taught colour-words and then reports to us the colour of traffic lights that we do not see. He argued that, in this game, we have not yet made the distinction between the child’s acting with or without

Wittgenstein’s ‘Depth Grammar’ 81

understanding, or the distinction between reacting automatically and communicating personal experience; furthermore, the fact that in our own practice we distinguish between ‘seeing red’ and ‘saying ‘‘red’’ ’ does not license our applying this distinction to the child in the imagined game; and this game makes no provision for the child’s cheating. The conclusion is drawn that we can talk of seeing red and not saying ‘red’ only in a specific language-game; also that the child knows no more whether he has the experience of seeing red than we do from recording his reactions.42 (4) The nonsensicality (uselessness) of a proposition based on a wrong calculation. A paradigm is the sentence ‘He cut a 3-metre board into 4 one-metre parts’ (PG 319). This might seem trivial, but it is a transparent case of something important which serves to illuminate the meanings of the phrases ‘to make sense’ and ‘to mean something by a proposition’ (PG 319). More generally, a calculation or proof may be taken to show that an apparently well-formed expression is nonsensical. One important example is using number-theory to show that there can be no such thing as a geometrical construction for ‘trisecting an angle with a straight-edge and compass’ (WWK 36–7; PLP 398–400). A non-mathematical example might be demonstrating the nonsensicality of defining a set by the description ‘the set of all sets which are not members of themselves’. Though these cases all concern sentence-construction, the feature at issue is something that in principle cannot be taken in by the ear. Why? Because recognizing nonsense of this type turns on reasoning or calculation, often quite extensive or complicated. To the extent that we follow the venerable practice of distinguishing the intellect from the senses, the essential role of calculation puts these particular identifications of nonsense beyond the range of what can, even ‘as it were’, be taken in by the ear. In addition, many logicians would resist the claim that sentences argued to be nonsensical in this way should be described as logically illformed. Why? Because admitting this form of argument threatens to reduce to absurdity the whole programme of constructing logical categories of expressions. For example, with sufficient mathematical expertise and ingenuity, we should be able to prove that every pair of natural numbers must belong to different logical categories on the model of showing that ‘the construction of trisecting an angle’ is nonsensical, whereas ‘bisecting’ or ‘quadrisecting an angle’ are perfectly intelligible (cf. PLP 96–7). (5) The construction of imaginary or hypothetical language-games as objects of comparison. Characteristic examples are the slab-game (PI §2); the arithmetic linked to the practice of counting ‘1, 2, 3, 4, 5, many’; a

82

Reading Wittgenstein

language-game with expressions of feelings which lacks the possibility of lying (LPE 254); a language-game with colour-words which lacks the distinction between being red and looking red (cf. PLP 285–6); the system of calculation of cost used by the wood-sellers (RFM 93–4); the construction of the ‘private language’ by associating words with sensations (PI §256). What is the point of these investigations? How could they demonstrate anything about the logical grammar of words in our language?43 In many cases moves in the language-game have no syntax at all (e.g. §2), and even in cases where they do, descriptions of the games say little or nothing about sentence-construction (cf. §§8, 17). Rather, the main focus is the relation between utterances and human activities.44 The description of the slab-game relates simple orders to what the builder’s assistant does (§2; cf. §8); in other cases, methods of calculation are related to economic transactions (RFM 94) or judgements of ability to an individual’s past performance (BB 102). These languagegames are constructed as objects of comparison; they prove nothing, but they are intended to reveal aspects of our own practices, as much by way of dissimilarities as resemblances. Contemplating them may have a role comparable to one possible effect of living abroad: a kind of liberation from preconceptions and a live awareness of previously unimagined possibilities. Nothing is proved and no new facts are adduced; the use of our words is left unchanged, but everything may suddenly look different.45 In this distinctive way, imaginary language-games may help to clarify the use of our words even though they tell us nothing whatever about sentenceconstruction in our language.46 The same effect might be achieved in clarifying concepts by finding fresh objects of comparison, say compelling analogies. The model of a rope whose strength comes from the overlapping of many fibres might change our attitude toward the dogma that there must be something common to everything correctly called ‘games’; and that in turn might enable us to relate our concept of a game to the explanations that we know how to give rather than rejecting these explanations as being clearly inadequate (PI §§65–75). Similarly, we might clear up some of our own muddles about arithmetic by exploring the analogy between equations and everyday rules of grammar (e.g. forming the third-person present of most verbs by adding ‘s’). The point is not to prove that we can say things that we did not want to say (e.g. that the concept of an arithmetical equation fits the concept of a rule of grammar), but that we can do things that we did not dream of doing, possibly because they go against the grain and conflict with what we want to see (LFM 55, 103; CV 17 ¼ BT 406). As with imaginary language-games, the intention is nothing more – and nothing less! – than to change ways of seeing things (cf. PI §144).

Wittgenstein’s ‘Depth Grammar’ 83

(6) Concern with pictures which individuals may associate with the uses of particular words (§308, BB 6, LPE 277 ¼ PO 212; cf. RPP I §549). Paradigms are the picture of thinking as an auxiliary ‘inner’ (or ‘mental’) activity accompanying speech (PI §673; BB 43, 148; cf. §153); the picture of mathematicians as explorers who make discoveries about a realm of ‘abstract objects’ (RFM 99, 111, 136); the picture of mathematics as a body of propositions rather than calculations or constructions (LFM 40); the picture MORE/LESS (or addition of something) as used to clarify certain conceptual contrasts, e.g. exhibiting genuine pain and manifesting mere pain-behaviour (PI §304; LSD ¼ PO 297–8, 302, 304); or the picture of all words as names of objects and of all ‘objects’ as independent of each other (PI §293; PO 235). (Many of these pictures are widespread, or even pervasive in our culture). These pictures are freefloating relative to explanations of word-use which serve to differentiate correct from incorrect sentence-construction; they may play no role when we put words to use ‘in the stream of life’. Yet they often dominate our thinking when we try to reflect on and describe our own speech-practices, to make sense for ourselves of what we say. In this way, they may generate philosophical confusion.47 Frequently we make wrong pictures for ourselves of our conceptual world (RPP I §§554, 803, 369, 646, 692). Such pictures have the power to surround things with a haze (PI §5), to hold us captive (PI §112), and to make us lose our bearings (PI §123; RPP I §549). Through them we become prey to all sorts of philosophical troubles (problems and confusions). We may feel driven to say things which are evidently false (PI §§38–9, 246), to make metaphysical assertions such as ‘Only the present is real’ (AWL 27) or ‘You must understand an order before you can execute it’ (PG 45–6), to criticize everyday speech-patterns for misrepresenting what they are correctly used to describe (PI §402), and to reject ordinary explanations of word-use as superficial and inadequate (PI §§66, 92, 108; PG 74–5). We may conclude that we can never know what another person is thinking (PI §246); or we may feel puzzled about how a mathematician can come into cognitive contact with ideal geometrical figures (lines without breadth) or abstract entities (numbers). Exploring pictures may have a pivotal role in clearing up confusions about the use of our words and in making our concepts surveyable, while mere changes of notation may help to shake off certain pictures and sidestep particular obsessions (AWL 78). It is very often the crucial move eliminating disorientation (PI §123) and bringing it about that we can keep our bearings [sich auskennen]. Interlocutors or readers are urged to bring pictures to consciousness, to recognize their influence, to work out their origins or prototypes (e.g. BB 151), and to clarify confusions about how to apply them (PI §§422–4; cf. PO 264).

84

Reading Wittgenstein

It is especially clear that in these matters individuals’ free acknowledgement is of decisive importance (cf. BT 410). It is also clear that pictures associated with words (‘mere pictures’) lie outside the scope of any established practice of ‘philosophical grammar’. None of these six things ‘can be taken in by the ear’ in any sense at all, they are at best marginally related to anything logicians normally consider in analysing the construction of sentences, and all of them are matters to which Wittgenstein openly directed much attention in his later attempts to dissolve philosophical confusions. All of them seem directly relevant to knowing one’s way about among concepts. Why does this account not offer a better-grounded and more fruitful interpretation of the phrase ‘depth grammar’ than the prevalent one? This interpretation has the great advantage that the term ‘depth’ can be given a straightforward explanation. It would here connote not hiddenness, but rather exploration of new dimensions different from any investigations of the principles of sentence-construction (‘surface grammar’). The model for contrasting depth grammar and surface grammar is the distinction between solid and plane geometry. On this reading, Wittgenstein had the clear intention to encourage us to look in new directions for the purpose of dissolving philosophical confusions (especially our own!). For example, he would urge Frege or Russell to turn attention to the role of arithmetical equations in science and everyday living or to reflect carefully on how proofs or calculations differ from experiments. Such matters are no less in plain view than ‘surface grammar’ is, but they are not ‘on the surface’ in that sense of ‘surface’.48 Minimally, this interpretation of ‘depth grammar’ makes good sense of PI §664. The activities of rearranging what is familiar, establishing orderings to dissolve particular confusions (PI §132), making patterns or aspects visible (§129), and gaining acceptance of alternative pictures (LPP 43, 168, 285) are all ways of clearing away fog that prevents clear vision of the workings of our language (PI §5). All of these methods contribute directly to keeping one’s bearings and finding one’s way about among concepts [sich auskennen]. According to §664, the depth grammar of ‘mean’ is to be clarified for precisely this purpose. How could the mere accumulation of information about details of sentence-construction serve this function? How could that method have freed the author of TLP from the picture that held him captive? On a wider front, this interpretation has the attraction of depicting Wittgenstein as presenting a radical or revisionary method, at least relative to the practice of analytic philosophy in the tradition derived from Frege and Russell and TLP. ‘Depth grammar’ would differentiate his later work

Wittgenstein’s ‘Depth Grammar’ 85

sharply from TLP, whose concern was to establish simple and rigid rules for the logical construction of propositions (PI §102; EPB 124).49 To appreciate his concern with ‘depth grammar’ would put emphasis on aspects of his investigations which have no counterpart there (e.g. his exploring the motivations for philosophical confusions (PI §39)). It might ¨ bersichtlichkeit] and give substance to his conception of ‘surveyability’ [U explain the power of ‘surveyable representations’ to dissolve philosophical confusions.50 What difference does all this make to understanding Wittgenstein’s work? This needs to be explored in detail, but the consequences seem likely to be enormous. At the very least, there would be a shift of emphasis. This might transform the spirit of all his discussions. These would be seem to be more open-ended, imaginative and exploratory – less dogmatic, regulatory and adversarial. Any patterns exposed in ‘depth grammar’ would be essentially contestable, both because it is debatable which dimensions of use are relevant to particular confusions and because criteria for identity and difference in use (in each dimension) are themselves open to negotiation. In particular, patient discussion of pictures with individuals would become a major part of clarifying concepts – in ‘depth grammar’. Through this sort of philosophical investigation, increased awareness of neglected aspects of one’s own speech and thought might bring to consciousness prejudices that restrict one’s own intellectual freedom. ‘There must be qualia.’ ‘There must be a what’s-it-like to being a child.’ ‘There must be experiences over and above human behaviour.’ ‘Understanding a sentence must be a form of information-processing.’ ‘The brain must be a computer, the mind must be its software.’ – How is somebody (somebody else, or even oneself ?) to be liberated from today’s ide´es fixes? From these captivating metaphysical uses of words? By (incontestable?) arguments demonstrating deviation from the standard practice of speakers of English? Mightn’t it be more promising to reflect on the materials from which these pictures are derived and on the subtle thought-transitions by which they become hardened into dogmas? How may we best arrive at a different attitude towards the problems that have come to be obsessive? Perhaps if we could surrender the preconception that Wittgenstein was practising ‘conceptual analysis’ as this phrase is now understood, we might find that his work had more to teach us than we ever dreamt of. We might find that effective dissolution of certain problems can be achieved by investigations in very different dimensions. What I have suggested here is that the prevalent interpretation of PI §664 takes most of the depth out of his practice of ‘depth grammar’. We could see all these philosophical activities as auxiliary or subordinate to the only essential task of philosophy, namely to construct a systematic

86

Reading Wittgenstein

logical grammar. Hence we could persist in denial that these intellectual manoeuvres are, strictly speaking, part of his clarification of the depth grammar of our language. Would this be merely a change of emphasis? (It is surely something that would strike some philosophers as worth fighting about.) Giving conceptual analyses and elucidating logical grammar51 are aspects of Wittgenstein’s philosophical investigations, and these activities may contribute, at least indirectly, to the dissolution of some philosophical problems.52 But they do not begin to exhaust his method. I have tried to argue here that his conception of meaning as use has other more radical aspects, and the ‘depth-grammar’ is meant to reveal further dimensions to his activity of ‘describing the grammar of our language’.53

Notes 1 2

3

4 5

6

In quoting or paraphrasing texts, I shall use bold to signal added emphases, whereas italics are reserved for the author’s own emphases. According to Hacker (Wittgenstein: Meaning and Mind, p. 708), ‘Surface grammar consists of obvious syntactic features of the sentence and the words of which it is composed. In this sense, ‘‘I have a pain’’ has the same surface structure as ‘‘I have a pin’’, ‘‘The carpet is darker than the curtains’’ as ‘‘Red is darker than pink’’, ‘‘3 > 2’’ as ‘‘Jack is taller than Jill’’, ‘‘I am talking’’ as ‘‘I am thinking’’, and so forth.’ Hacker (Wittgenstein, p. 708): ‘The depth grammar, by contrast [with surface grammar], is made evident by a description of the overall use of the relevant expression, by a surview of its combinatorial possibilities and impossibilities, of the circumstances of its use, and of its consequences.’ What follows is a description of the contrast in respect of the word ‘to mean’. The description of its depth grammar given here is exclusively concerned with wordcombinations: ‘We can say ‘‘ . . . ’’, but not ‘‘ . . . ’’.’ Matters not mentioned here include the practice of using ‘By ‘‘ . . . ’’ I here mean . . . ’ to draw precise boundaries around a concept for particular purposes (cf. PI §69) or to forestall misunderstandings, the connection of the possibility of meaning something with the capacity to articulate what one means (cf. BB 158–60, 174–7), or the treatment of what one later says as a criterion for what one previously meant – and the limits to this practice (cf. BB 38–9). Ryle, Collected Papers, vol. 2, pp. 176, 183. Compare TLP 5.4733: Frege says that every well-constructed sentence has a sense. Compare also Waismann (in an unpublished letter to Felix Kaufmann in ¨ ssig (d.h. den Forderungen der August 1929): Ein Satz hat Sinn, sobald er rechtma ¨ ss) gebildet ist . . . (A sentence has sense just as long as it is logischen Syntax gema constructed correctly [i.e. according to the requirements of logical syntax]). RLF 162: By syntax in this general sense of the word I mean the rules which tell us in which connections only a word gives sense, thus excluding nonsensical

Wittgenstein’s ‘Depth Grammar’ 87

7 8 9 10 11 12

structures. RLF 171: . . . a perfect notation will have to exclude [nonsensical constructions] by definite rules of syntax. RLF 162: The syntax of ordinary language, as is well known, is not quite adequate for this purpose [excluding nonsensical structures]. Ryle, Dilemmas, p. 9; Ryle, Collected Papers, vol. 2, pp. 170, 174, 179. Ryle, The Concept of Mind, pp. 18, 33, 152–3. Ryle (Collected Papers, vol. 2, pp. 170, 184) suggests that all philosophers’ propositions are category-propositions. Ryle, The Concept of Mind, p. 19. Some commentators praise this strategy (e.g. Norman Malcolm and Anthony Kenny), while others deplore it (e.g. Bertrand Russell). Many see here a point of continuity with TLP (3.323–3.324): In everyday language it very frequently happens that the same word has different modes of signification . . . In this way the most fundamental confusions are easily produced (the whole of philosophy is full of them).

13

14

15 16

17

18 19 20

21

Kenny (The Metaphysics of Mind, p. v) argues for a close parallelism in strategy between Wittgenstein’s investigations of philosophy of psychology and Ryle’s Concept of Mind, suggesting that the chief differences are stylistic. David Stern (Wittgenstein on Mind and Language, p. 24): ‘What is ‘‘given’’ to philosophy. . . is, in the first instance, our ordinary language . . . ’. According to Norman Malcolm (Nothing is Hidden, pp. 136, 148, 168), crucial advances in Wittgenstein’s later work are discoveries about the use of words in everyday language. Often-cited examples in PI are §§38, 251 and 281. Its entrenchment might be explained by the determination to see him as a major player in the general activity of ‘analytic philosophy’: producing arguments and establishing results about the meanings (uses) of words. This point would need to be refined to deal with various kinds of technical terminology, professional lingo, or esoteric jargon. For example in DescartesSpeak, a ‘mode of thinking’ has an ‘existence outside the mind’; this might elicit a bizarreness-reaction from most modern readers. On this understanding, ‘3 is greater than 2’ and ‘London is to the north of Paris’ have the same (surface) grammatical form. In this, case, ‘3 is greater than 2’ and ‘London is to the north of Paris’ would have different (surface) grammatical forms. Some philosophers have made the reverse discrimination in these cases. Bertrand Russell (An Enquiry into Meaning and Truth, p. 185) argued that ‘the sound of a trombone is blue’ is logically well formed (and false), even if it is grammatically ill formed. (This manifests a different conception of what counts as logical. His differentiation of words into logical types is extremely coarse-grained, just as Frege’s was.) For example, J. L. Austin (Philosophical Papers, p. 133 n.1) on the difference between ‘by accident’ and ‘by mistake’, which is made salient by some telling examples of contrasting application.

88

Reading Wittgenstein

22

Friedrich Waismann (HISP 24) noted that a volley of questions may sometimes serve this distinctive purpose. Other declarative sentences may ‘sound strange’, at least to philosophers, for no very good reason:

23

PI §421: It seems paradoxical to us that we should mix together slap-dash in a single report physical states and states of consciousness: ‘He suffered great torments and tossed about restlessly.’ It is quite usual; so why do we find it paradoxical? Because we want to say that the sentence deals with both tangibles and intangibles at once. Evidently the ‘ear’, like the ear, is subject to illusion. Just as Noam Chomsky treats judgements that strings of words are ungrammatical as pre-theoretical. 25 Cf. PI p. 208: the substratum of this experience is the mastery of a technique. 26 Even the final two sentences of PI §664 do not settle this matter: 24

And now compare the depth grammar, say of the word ‘to mean’, with what its surface grammar would lead us to expect. No wonder we find it difficult to know our way about. It is tempting to relate this in a particular way to discussions of the difference between ‘to mean’ and ‘to think of ’. Namely, to suppose that their ‘surface grammar’ is exhausted by the common label ‘verb’, while all differences belong to ‘depth grammar’. (We could reinforce this idea with appropriate citations: e.g. the observations that substantives send us off in search of substances (or objects), and that puzzles about existence arise from the fact that ‘exist’ is a verb like ‘eat’.) But this reading is not compulsory. Consider that ‘to mean’ (‘Meinen’) is naturally categorized as a species of thinking (Denken); especially when anyone adopts the picture that thinking is what gives life to signs (BB 6; cf. PI §81). Those things might be taken to characterize the ‘surface grammar’ of ‘mean’. Making perspicuous the differences in use between ‘mean’ and ‘think’ would then be the business of ‘depth grammar’. 27 Some remarks may have that appearance. But the overt intention is generally to point out potential for confusion, not mistakes (e.g. in calling thinking a mental activity (BB 6) or remembering a mental process (PI §305)). 28 Arguably, one way in which Augustine’s picture of the essence of language surrounds the workings of language with a fog (PI §4) is that it eliminates the possibility of any philosophical clarification of sentences other than treating them as combinations of names (§1). Only principles of sentence-construction could belong to the essence of a sentence. 29 Frege might reply: We recognize the difference in use between two kinds of number-words, the ones which answer the question ‘How many?’ and the

Wittgenstein’s ‘Depth Grammar’ 89 ones which answer the question ‘How much?’. What a logico-philosophical analysis must offer is an explanation of this difference in use: this must be couched in terms of a difference in the kind of objects which these words designate. But this seems a good paradigm of what Wittgenste in rejected as an explanation of meaning ‘in our sense’. It is what is to be cast aside according to this remark (§109): We must do away with all explanation, and description alone must take its place. 30 31 32 33

34

35

It is not claimed that this form of paraphrase generally produces expressions which are synonymous with the original expressions. Bouwsma, Wittgenstein: Conversations, 1949–1951, pp. 63–4. These are not crude or uncivilized, but rather not mediated by judgements. Sometimes these differences are excluded from consideration as being non-logical or psychological (e.g. NL 96; Frege, ‘Conceptual Notation’, §3). Sometimes some distinctions are drawn: declarative sentences are contrasted with sentence-questions and imperatives, and systematic differences are noted in respect of moods of verbs and word-order. But such distinctions are relatively few and pretty crude. Wittgenstein attended to the countless different uses of sentences which have the same sentence-construction (PI §23). Though it might have more to say if it developed the idea that an expectation and its fulfilment, an order and its execution, a question and its answer, make contact in language. This idea could be elaborated as principles for the construction of complex sentences of the form: ‘The request ‘‘Please open the window’’ addressed to A is fulfilled by the action described by the sentence ‘‘A has opened the window’’.’ But clarification of ‘the harmony between language and reality’ need not be restricted to explanations of this form. It could be fully understood by somebody who lacked the reflexive concepts needed to frame such complex descriptions of how we talk – either in oratio directa or in oratio obliqua. The point of view of TLP seems to be captured in this remark (PR 132): The general form of the application [of arithmetic] seems to be represented by the fact that nothing is said about it. (And if that’s a possible representation, then it is also the right one.)

36

Likewise, he suggests that a proof be seen as showing that a formula is an arithmetical proposition, not showing that an arithmetical proposition is true. (This motivates his criticism of taking Go¨del’s theorem to show that there are arithmetical truths which are unprovable.) 37 Whether or not a well-formed (significant) sentence has a use in particular circumstances is a quite different question, one usually relegated to ‘pragmatics’. Wittgenstein suggested that this conception is a source of confusion:

90

Reading Wittgenstein we are so much accustomed to communication through speaking [sprechen], in conversation, that it looks to us as if the whole point of communication lay in this: someone else grasps the sense of my words – which is something mental: he as it were takes it into his own mind. If he does something further with it as well, this is no part of the immediate purpose of speech [Sprache]. (PI §363)

38

39

40

41 42 43

44

Waismann emphasized this point (in a letter to Felix Kaufmann in August 1929): Von einer Tatsache kann aber nur die Wahrheit oder die Falschheit eines ¨ ngen, aber nie der Sinn. (Only the truth or falsity of a sentence can Satzes abha depend on a fact, but never its sense.) PI §132: we want to establish an order in our knowledge of the use of language: an order with a particular end in view; one out of many possible orders; not the order. [And these particular orderings need have nothing to do with word-combination! They may be based on the functions of words in a practice (§17).] He recommended looking at each language-game as a complete primitive language (PI §17). This strategy should govern the interpretation of the ‘private language argument’: the game set up in §256 gives no warrant for calling ‘S’ the name of a sensation (§261) – even though the description of that game seemed at first sight to capture the whole of what is essential to a symbol’s naming a sensation (namely, that it is correlated directly with a sensation). The reasoning turns on respecting the closure of the game described in §256, i.e. relating ‘S’ solely to that particular language-game. This is one the guidelines of Wittgenstein’s philosophy of mathematics. It is elaborated in Waismann’s Introduction to Mathematical Thinking, ch. 16. This is a paraphrase of material from Margaret Macdonald’s notes of lectures from 24 October to 11 November 1935 (unpublished). They might be relevant to this enterprise if they were meant to be viewed as simplifications or idealizations of our more complicated practices, as it were first approximations ignoring friction and air-resistance (PI §130). But Wittgenstein denied having introduced them for that purpose. This form of description may be made compulsory: The examiner puts the men through a set of standardized tests. He lets them lift certain weights, swing their arms, skip, etc. The examiner then gives his verdict in the form ‘so-and-so can throw a spear’ or ‘can throw a boomerang’ or ‘is fit to pursue the enemy’, etc. There are no special expressions in the language of this tribe for the activities performed in the texts. (BB 102) This specification blocks the possibility of redescribing the language-game in terms of inferential connections between different kinds of statement made within the game itself.

Wittgenstein’s ‘Depth Grammar’ 91 45

This might offer the intended (and radical) interpretation of the notorious remark ‘Philosophy. . . leaves everything as it is’ (PI §124). 46 LPP 43, 168, 285: contemplating the fiction of a soulless tribe may transform one’s whole attitude towards the question ‘What is thinking?’ 47 For example, if pain is one thing, pain-behaviour something different, then it may seem that there can be no essential connections between the two. In precisely this way, construing the grammar of ‘pain’ on the model of ‘object’ and ‘designation’ surrounds the working of sensation-words with a fog (cf. PI §§293 and §5). 48 This interpretation seems strongly supported by XII.132: Um diese Begriffe ubersehen zu konnen, musst du sie anders vergleichen, als ihren Oberflachengrammatik es nahelegt. Du musst andere Teile als homologen auffassen, was einem Kiefer gleichsieht, musst Du einem Fuss vergleichen. [In order to be able to command a synoptic view of these concepts, you must compare them in ways other than the surface grammar suggests. You must see different parts as homologous; what looks like a beetle you must compare with a foot.]

49

50 51

52

53

What would be instances of such far-fetched comparisons? Perhaps comparing arithmetical equations with rules of grammar (rules of accidence and syntax), or comparing understanding a word with mastery of a technique (say, skill at solving cryptic crosswords). Philosophical confusions were to be traced to misunderstandings of syntax or of the modes of signification of symbols (TLP 3.323–3.325, 4.1272). Later he might have remarked: ‘I believe you are looking in the wrong dimension.’ Baker, ‘Philosophical Investigations §122: Neglected Aspects’ (1991). For example, PI §246 points out that ‘know’ is used only where expressions of doubt make sense, and §§224–5 argue that the uses of ‘rule’, ‘agreement’, and ‘the same’ are closely interwoven. (These seem to be reminders about aspects of our normal speech-patterns.) But Wittgenstein refrained from pinpointing ‘category-mistakes’ and from classifying words into logical types. (Rather, he suggested that calling thinking a mental process or mental activity involves drawing misleading analogies between the use of ‘think’ and that of certain other expressions – analogies he attempted to displace in favour of other analogies.) Arguably, not by regulating word-use and showing that certain things cannot be said, but by bringing to consciousness someone’s motives for saying something evidently bizarre and helping him to realize that he no longer wants to say this. These dimensions are conspicuously neglected in cognitive science and related work in philosophy of language, at least in so far as attention is directed at representational content, information-processing, etc.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

4 Wittgenstein on Metaphysical/ Everyday Use*

Wittgenstein remarked: ‘What we do is to bring words back from their metaphysical to their everyday use’ (PI §116). On this basis, his ‘later philosophy’ is generally regarded as a version of ‘ordinary language philosophy’. He is taken to criticize philosophers for making (‘metaphysical’) statements which deviate in different ways from the everyday use of some of their component expressions. This paper marshals textual evidence for a different reading of this remark. It shows that he used ‘metaphysical’ in a traditional way, namely to describe philosophical attempts to delineate the essence of things by establishing necessities and impossibilities. On his conception, ‘everyday’ simply means ‘nonmetaphysical’ (in this precise sense). Comparisons of philosophical utterances with non-philosophical uses of words are meant to call attention to this crucial distinction.

Philosophical Investigations §116b: What we do is to bring words back [zuru ¨ ckfu ¨ hren] from their metaphysical to their everyday use.1 This remark is one of the best-known components in Wittgenstein’s exposition of his distinctive conception of philosophy. Even on its own it seems to be decisive textual evidence for assimilating his ‘later philosophy’ to the genre called ‘ordinary language philosophy’ (associated with Oxford in the 1950s and 1960s, pre-eminently with the work of J. L. Austin and Gilbert Ryle). In this paper I will make a case for a very different reading of this remark. To the extent that it carries conviction, it * Originally published in The Philosophical Quarterly, 52: 208 (2002), pp. 289–302. Reproduced by permission of Blackwell Publishing.

Metaphysical/Everyday Use

93

calls for a radical re-evaluation of the methods of concept-clarification exhibited in the whole corpus of Wittgenstein’s writings.2 As a preliminary, three features of this remark need some clarification: (1) What we do . . . : The italics seem to indicate that the procedure here described is taken to be distinctive of Wittgenstein’s own style of philosophical therapy, not of the nature of philosophy. By implication, he takes many other philosophers not to do this.3 (2) Zuru¨ckfu¨hren: This verb might be applied in a variety of cases of ‘bringing back’. One would be using a sheepdog to drive stray sheep back to the flock. Another might be showing somebody the way out of a forest by taking him to the nearest village. A third might be starting a lost rambler off in the right direction and giving him directions about how to continue. Note that somebody lost in the woods might distrust the help or directions given by a stranger and hence need to be persuaded to take the offered guidance or advice, which he is perfectly free to refuse. It is a fundamental question to what extent Wittgenstein’s conception of bringing words back to their everyday use respects the freedom or independence of his interlocutors. Is it a case of beating somebody with the stick of grammar? or of gently leading him somewhere with his own consent?4 (3) The contrast ‘metaphysical’/‘everyday’: This gives rise to a range of important questions. What is ‘a metaphysical use of words’? What is ‘an everyday use’? Did Wittgenstein intend something tolerably definite to be understood by each of these phrases? If so, need this be a dichotomy? Or is one meant to be simply the antithesis of the other? If so, which of the two is to be given a positive explanation? And how is that to be framed? Finally, does a metaphysical use of words always need to be pinned down to the metaphysical use of a single word?5 This paper will focus on this third range of questions.

I The prevalent view of Wittgenstein’s philosophical therapy offers definite answers to these questions without bothering to marshall much textual evidence. ‘Everyday use’ is taken, so to speak, as the centre of gravity. Though lacking any explanation in §116, this phrase is taken to signify the standard speech-patterns of the English-speaking peoples. Describing everyday use is a matter of establishing facts about a public normative practice, and it is presumed to be relatively uncontentious (objective?)

94

Reading Wittgenstein

what these facts are. It is datum that one can say ‘ . . . ’ or that one cannot say ‘ . . . ’. ‘Everyday use’ is treated as synonymous with ‘ordinary speech (language)’. ‘Metaphysical use’ in §116 is then construed simply as the antithesis of ‘everyday use’.6 The phrase is taken to be pejorative as in Hume’s condemning metaphysics as ‘nothing but sophistry and illusion’. Wittgenstein is understood roughly to follow the logical positivists in using ‘metaphysical’ as a general term in condemning philosophical utterances as nonsensical.7 So ‘metaphysical’ in §116 means roughly the same as ‘non-standard’, ‘deviant’, or ‘abnormal’.8 On this view, the starting point is the idea that philosophers frequently misuse expressions,9 thereby speaking nonsense or ‘transgressing the bounds of sense’.10 Wittgenstein’s therapy is directed at such philosophers, especially at proponents of various ‘isms’ (or ‘wasms’) who gratuitously depart from ordinary language, who misguidedly object to it (cf. §402), or who conflate different language-games (PI p. 231; cf. PI p. 224)11. The aim of the therapy is to expose the nonsense central to major philosophical positions. Thus, on a large scale, ‘the private language argument’ is commonly pictured as a kind of reductio ad absurdum of the Cartesian view of the mind. On a small scale, §38 catalogues reasons why it is mistaken to call ‘this’ and ‘that’ the only genuine proper names. Like Ryle, Wittgenstein polices the bounds of sense,12 sharply reprimanding philosophers who commit offences by uttering nonsense. The activity of clarifying concepts or describing grammer is corrective therapy (like orthodontics). As befits a philosopher, this activity is taken to involve a considerable degree of abstraction or generality, primarily by identifying misallocations of kinds of expressions to logical categories.13 In this way, correct grammatical generalizations may have a wide range of useful applications. The second half of part I of the Philosophical Investigations is often seen as consisting of a series of applications of the lesson learned from the ‘private language argument’ to the clarification of various psychological concepts (thinking, imagining, willing, etc.). According to this interpretation, we might say: in the contrast ‘metaphysical’/‘everyday’, the term ‘everyday’ wears the trousers.14 We are assumed to understand what counts as ‘everyday use’ and how to establish what this is from case to case.15 No attention whatever is given to the possibility that prejudices may interfere with one’s perceptions or descriptions of how words are used (cf. §§5, 304, 340, 363). This familiar picture of Wittgenstein’s later philosophy informs the bulk of secondary literature. Its currency explains why his work is likened to ‘ordinary language philosophy’.16 Indeed, it is diagnostic of this pattern of interpretation that there are taken to be very close parallels in method

Metaphysical/Everyday Use

95

between Wittgenstein and Ryle;17 the differences are sometimes dismissed as primarily stylistic.18 In virtue of following this method, Wittgenstein has the ability to prove other philosophers to be wrong by exposing their misuses of words (mistakes in grammar), and hence perceptive readers of his work experience the exhilaration of hearing ‘the clash of steel on steel’. There are many instances where Wittgenstein is alleged to have followed this model of bringing words back to their everyday use. 1 2 3 4

5

6 7

He shows that ‘I’ is not the name of a person (§410) or not a referring expression.19 He demonstrates that ‘I am in pain’ is a signal or expression (‘avowal’), not a description.20 He establishes that arithmetical equations are really rules of grammar, not descriptions of relations among abstract entities.21 He explains why ‘this’ and ‘that’ are not names (B&H 1.226–7), why ‘pain’ or ‘human being’ does not designate an object, and why ‘thinking’ is not an activity or process.22 He shows that samples employed in ostensive definitions are really part of our language, hence that these explanations do not connect language with reality, but remain within language, connecting symbols with other symbols (B&H 1.183–4, 192). He demonstrates that there is no such thing as the general form of a proposition (B&H 1.567). He proves that ‘Only I can know that I am in pain’ is either false or nonsensical (§246).

This is a sample of the grammatical facts that Wittgenstein is taken to have discovered and described. It is noteworthy that all these moves require some explanation, often some qualifications. They are also likely to generate objections; hence they seem to require more or less elaborate defence. This occasions a second-order problem of how to reconcile the need for these counterarguments with Wittgenstein’s claim not to dispute anything (§128; AWL 97; LFM 22). One possibility might be to argue that the term ‘dispute’ has no legitimate application to descriptions of grammar, but only to scientific or empirical truth-claims. But that remark seems, in turn, to be open to challenge. And so on. In this way, his declared ambition to get particular philosophical problems to disappear completely (§133) begins to look more and more unrealistic. I think that there are many reasons to question whether this interpretation squares with Wittgenstein’s conception of philosophy as an activity or his intention in clarifying various concepts to demonstrate a method by means of examples (§133). In this paper, however, I will focus on a

96

Reading Wittgenstein

specific objection: namely that the prevalent interpretation of §116b) rests on a serious misrepresentation of Wittgenstein’s use of the term ‘metaphysical’ in the corpus of texts that constitute his later philosophy. Though occurrences of this term (and its cognates ‘metaphysics’ and ‘metaphysician’) are relatively sparse, there are fairly clear patterns in its use. Pinning down his conception is the basis for proposing a radically different interpretation of §116b, one which derives an interpretation of the phrase ‘everyday use’ from what he understood to be the meaning of ‘metaphysical’.

II I shall begin from a representative sample of remarks incorporating ‘metaphysical’ or its cognates. §58: One cannot say ‘Red exists’, because if there were not red it could not be spoken of at all. – . . . saying something about the nature of red . . . The same idea – that this is a metaphysical statement about red . . . [that red must exist]. . . . [cf. MS 110.149: Die Regel scheint wie ein Spiegelbild des Wesens in der Sprache.]23 PO 208–9: The difficulty is that we feel we have said something about the nature of pain when we say that one person can’t have another person’s pains . . . [not] anything physiological or even psychological, but something metapsychological, metaphysical. [cf. BB 73.] BB 35: the characteristic of a metaphysical question [is] that we express an unclarity about the grammar of words in the form of a scientific question. [namely ‘What is the object of thought?’] [Compare Augustine’s question ‘What is time?’ (§89); ‘What is a proposition?’ (§92); and ‘Is my person . . . a constituent of the fact that I see or not?’ (LPE 282).] BB 46: we are using the words ‘flux’ and ‘vagueness’ wrongly, in a typically metaphysical way, namely without antithesis. BB 49: we shall be liable to confuse our metaphysical proposition ‘I can’t feel his pain’ with the experiential proposition ‘We can’t have (haven’t as a rule) pains in another person’s tooth.’ BB 55: What we did in these discussions was what we always do when we meet the word ‘can’ in a metaphysical proposition. We show that the proposition hides a grammatical rule. [Perhaps the speaker’s own stipulation: or his ‘avowal of adherence to a form of expression’ (Z §442)] . . . Again, when in a metaphysical sense I say ‘I must always, know when I have pain’, this simply makes the word, ‘know’ redundant; and

Metaphysical/Everyday Use instead of ‘I know that I have pain’, I can simply say ‘I have pain’. §246.]

97

[cf.

MS 113.115: Sage ich nun aber, als metaphysischen Satz, ‘ich kann die /eine/ Ursache immer nur vermuten’, so heisst das: ich will im Falle der Ursache immer nur von ‘vermuten’ und nicht von ‘wissen’ sprechen um so Fa¨lle verschiedener Grammatik von einander zu unterscheiden. . . . Was also 24 an unserem . . . Beispiel falsch ~~~~~~~~~ist, ~~ ist das Wort ‘nur’. . . . BB 56–7: an assertion which the metaphysician makes [may express] discontentment with our grammar. . . . [cf. MS 112.240. And it may formulate an objection [§§39, 402] or express dissatisfaction (§105).] BB 66: ‘Only this is really seen’ . . . a metaphysical assertion. There is a grave danger hidden in the word ‘really’.25]

[Compare:

TS 220 §110: Wir ziehen immer wieder die Ausdrucksform nach und glauben, wir haben die Sache gezeichnet. – Durch eine optische Ta¨uschung scheinen wir im Innern der Dinge zu sehen, was auf unserer Brille gezeichnet ist. . . . Der Ausdruck dieser Ta¨uschung aber ist die metaphysische Verwendung unserer Wo¨rter.26 [Namely, that ‘This is how things are’ (‘Es verha¨lt sich so und so’) expresses the essence of the proposition, the general form of the proposition(cf. PI §136).]

What patterns in the use of ‘metaphysical’ emerge from reflection on these citations? There is no single pattern visible in all these cases, but several patterns seem immediately evident: (1) Expressions of necessity and impossibility, on statements that feature ‘must’ and ‘cannot’: non-scientific statements featuring such modal expressions are taken to be overtly metaphysical. (Of course, not all statements incorporating ‘must’ or ‘cannot’ are to be understood as metaphysical (BB 49), and not all statements of essence need be framed in this way (BB 55, 66; §1).) This manifests the traditional conception of metaphysics as the science of the essences or natures of things, a system of necessary truths. The statement ‘Red must exist’ is a metaphysical statement about red (§58), just as the statement ‘I can’t feel his pain’ may be used to propound a metaphysical proposition which is to be contrasted with the experiential proposition that I do not feel pain in another’s tooth (BB 49). Similarly, Augustine’s picture of the essence of language must be understood to contain as one component the metaphysical proposition ‘Every sentence must be composite (a combination of names)’ (§1; cf. PLP 317–18). More generally, modal propositions formulating what is essential (das Wesentliche) are metaphysical propositions. ‘The metaphysical use of our words’ pinpointed in TS 220 §110 seems to be one case

98

Reading Wittgenstein

where an object of comparison is misconceived as a model to which reality must conform (§131). (2) The use of words without antitheses: if every word in a language must be a name, then there is no such thing as a word’s not being a name; or if every (real) pain is mine, then there is no such thing as a pain being yours (cf. BB 55). Metaphysical propositions seem to be assertions of the greatest moment, formulating the very essence of things. At the same time, the impossibility of things’ being otherwise opens the way to the recurrent criticism that metaphysical statements make applications of the key terms redundant and the assertions themselves empty (§131).27 Here is one model of this argument: ‘when in a metaphysical sense I say ‘‘I must always know when I have pain’’, this simply makes the word ‘‘know’’ redundant;28 and instead of ‘‘I know that I have pain’’, I can simply say ‘‘I have pain’’ ’ (BB 55). Similarly, if we say that every word must be a name, then ‘when we say: ‘‘Every word in the language [of §9] signifies something’’ we have so far said nothing whatever; unless we have explained exactly what distinction we wish to make’ (§13). Again, if I say in my grammar ‘Only this chair can be green’, then it makes no sense for me to say in the same language ‘This chair is green’, but only ‘There is green here’.29 (Compare the criticism of ‘Only my present experience is real’ (PR 84–7). Also MS 110.189–90; and §398.) (3) Non-scientific statements having the form of scientific explanations (theories): formulations of essences often take the form of simple formulae giving definitive answers to questions of the form ‘what is . . . ?’.30 It is characteristic of metaphysics to tackle such questions as ‘What is time?’ (PLP 172) or ‘What is a proposition?’ (§92), and to try to remove puzzlement by giving definitions (MS 112.236, 240), e.g. the formula ‘Time is the form of becoming’ (PLP 175) or a statement of the general form of a proposition (§§92, 114). In this form (in contrast to explicitly modal statements), metaphysical propositions are liable to be particularly misleading (BB 55). They may appear to express ‘a kind of scientific truth’ (BB 55), perhaps to give the impression that ‘a stupendous discovery has been made’ (BB 23; cf. §400). This generates tremendous confusion in debates among philosophers (§402) or debates between scientists and philosophers (BB 23). In thinking of this particular form of the metaphysical uses of words, one might say: ‘The essential thing about metaphysics: that the difference between factual and conceptual investigations is not clear to it. A metaphysical question is always in appearance a factual one, although the problem is a conceptual one’ (RPP I §949; cf. BB 35). The same conception of metaphysics is a recurrent theme: ‘Philosophers . . . are irresistibly tempted to ask and answer questions in the way science does. This tendency is the real source of metaphysics . . . ’ (BB 18). (Strong

Metaphysical/Everyday Use

99

attachment to this scientific form of question stands in the way of accepting the proposal that the proper method for removing puzzlement is to investigate how we use words. That seems to change the subject.31) The Tractatus is full of such metaphysical statements, ones which give the impression of being ‘a kind of scientific statement about ‘‘the nature of . . . ’’ ’ (BB 109). Here is a sample: 1.2 2.061 2.161 3.14

The world divides into facts. States of affairs are independent of one another. There must be something identical in a picture and what it depicts. . . . A proposition-sign is a fact.

These are all contributions to the project of revealing what must be true of any possible language in which anything at all can be said. There are some later echoes of these metaphysical statements; e.g. Alles Wesentliche u¨ber den Gedanken ist damit gesagt, dass der Gedanke dass p der Fall ist nicht die Tatsache ist dass p der Fall ist. Dass der Gedanke eine andere tatsache ist. (MS 110.37)32

(4) Statements about word-use giving explanations grounded in the natures of things: for example, ‘It looks to us as if we were saying something about the nature of red in saying that the words ‘‘Red exists’’ do not yield a sense’ (§58). Likewise, in denying that a human body can have pain ‘it is as though we looked into the nature of pain and saw that it lies in its nature that a material object can’t have it’ (BB 73). It seems to be as though it would be not false but nonsense to say ‘I feel his pains’, but as though this were because of the nature of pain, of the person, etc., as though, therefore, this statement were ultimately a statement about the nature of things. So we speak for example of an asymmetry in our mode of expression and we take it as [auffassen als] a mirror image of the essence of the things. (PO 208–9)

In a similar way, noting an asymmetry in the grammar of temporal expressions may be conflated with ‘a kind of scientific statement about ‘‘the nature of the future’’ ’ (BB 109). In all such cases metaphysical statements are offered as explanations (Erkla¨rungen) or justifications (Begru¨ndungen (RPP I §667)) of familiar patterns of word-use. This gives the misleading impression that what is said is ‘in no way dependent on a convention’ (AWL 172; cf. 176), i.e. that the grammar of certain expressions is somehow answerable to independently given essences of

100 Reading Wittgenstein

things. (The conception of grammar as autonomous is expressed in the practice of turning one’s face against such explanations. It does not call for stating an opinion, but adopting a method of investigation (Betrachtungsweise).33) The outcome of this investigation seems clear. Within the corpus of Wittgenstein’s texts, the term ‘metaphysical’ has a definite and quite traditional meaning: it belongs to a semantic field that includes ‘necessary’, ‘essence’, and ‘nature’. More precisely, the four features just listed seem to characterize various instances of what Wittgenstein picks out as ‘metaphysical statements’ or ‘metaphysical uses of words’. (The textual evidence does not support any distinction in meaning between these two phrases.) In this way, he focuses attention on a tolerably well-defined family of propositions having various definite and interrelated characteristics. The most conspicuous is their containing expressions of modality. But whether or not they contain ‘must’ or ‘cannot’, they are to be understood as stating necessities and impossibilities. These findings are crucial for interpreting §116b. ‘Metaphysical uses of words’ are explicitly related to attempts to grasp the ‘essence’ of things in §116a. Consequently, the word ‘metaphysical’ seems here to bear a traditional meaning. Granted that the phrases ‘metaphysical use’ and ‘everyday use’ are mutually exclusive, the term ‘everyday’ must be glossed as ‘non-metaphysical’. Hence, in this pair, the term ‘metaphysical’ wears the trousers; and the term ‘everyday’ need not mean ‘conforming with standard speech-practice’.

III Though unorthodox, this interpretation is strongly supported by textual archaeology. An ancestor of §116b occurs in BT 411–12: What we do is to bring words back from their metaphysical to their// correct ~~~~~~~~~// normal // use in (the) language.

This has a striking feature: namely that Wittgenstein manifests hesitation about what word to contrast with ‘metaphysical’. He offers alternatives, and with the squiggly underlining of ‘correct’, he indicates particular qualms about that word. This remark suggests that he starts from the word ‘metaphysical’ and then tries to find some down-to-earth expression that means roughly ‘non-metaphysical’. The term ‘everyday’ (‘allta¨glich’) in §116 seems to represent yet another attempt to solve this same prob-

Metaphysical/Everyday Use

101

lem.34 By contrast, if ‘metaphysical’ meant ‘not conforming with ordinary usage’, it would be difficult to explain his qualms here about the term ‘correct’. In BT 411–12, the immediately preceding remark casts light on the same early draft of §116b: Whence does this observation get its importance: the one that points out to us that a table can be used in more than one way, that one can think up a table that instructs one as to the use of a table, that one can also conceive of an arrow as pointing from the tip to the tail, that I can use a model as a model in different ways? What we do is to bring words back from their metaphysical to their// correct ~~~~~~~~~// normal // use in (the) language.

Is the second paragraph not meant to be a straightforward description of the tactics illustrated in the first? The series of statements highlighting ‘can’ are evidently to be understood as antitheses to statements framed with ‘cannot’. It is the presence of these modal qualifications that make those (implicit) theses metaphysical uses of words. The denial of these ‘cannots’ is precisely what makes the antitheses everyday or normal uses of the very same words. This pattern of argument is frequently repeated. One may feel torn between claiming that no rule can determine what is to be done and claiming that there must be rules which never let doubts creep in. Both thoughts are alluded to in §85: A rule stands there like a sign-post. – Does a sign-post leave no doubt open about the way I have to go? . . . But where is it said which way I am to follow it; whether in the direction of its finger or (e.g.) in the opposite one? – And if there were not a single sign-post, but a chain of adjacent ones or of chalk marks on the ground – is there only one way of interpreting them? – So I can say, the sign-post does after all leave no room for doubt. Or rather: it sometimes leaves room for doubt and sometimes not. And now this is no longer a philosophical proposition, but an experiential one.

This remark is meant to win acknowledgement of a correlative pair of possibilities. That has the power to transform one’s way of looking at things [Anschauungsweise] (§144). The emphasis on possibilities in BT 411–12 is conspicuous in the immediate continuation of §116b in TS 220 §111: (The person who said that one cannot step twice into the same river said something false; one can step twice into the same river. – And an object

102 Reading Wittgenstein sometimes ceases to exist if I cease to see it, and sometimes not. – And we do know sometimes what colour another sees, and sometimes not.)

Here the expressions ‘sometimes’ and ‘sometimes not’ evidently function as modal terms, understood as contradicting necessities expressed by using ‘never’ and ‘always’. This parenthetical remark is evidently meant to give some examples of the (dis)solution of philosophical difficulties by bringing words back from their metaphysical to their everyday use (allta¨gliche Verwendung), and the crucial move is to refrain from affirming apparent necessities and impossibilities by acknowledging a wider range of possibilities.35 It is noteworthy in most of these passages that no particular constituent (word or phrase) is being used in an obviously deviant manner. It is the entire utterance ‘This table cannot be used in more than one way’ which constitutes the metaphysical use of our words. We would have difficulty in trying to pin the blame on a misuse of ‘table’ or ‘to be used in a certain way’. The same point holds for such philosophical propositions as ‘A rule always lets doubts creep in about its applications’, ‘We can never know what colour another person sees’, ‘Some pictures cannot be applied in more than one way’ (§140), or ‘All games must have something in common’ (§65). In this last respect, §116a might suggest that the content of §116b has been altered by the change of context from BT and TS 220. The suggestion seems to be that it is always particular words such as ‘knowledge’, ‘being’, ‘object’, ‘I’, ‘proposition’, and ‘name’ which need to be herded back from philosophers’ discourse to the language-games which are their original homes. Arguably, however, this is a crude reading of §116a. It leaves out half the protasis: If philosophers use a word . . . and try to grasp the essence of the thing, one must always ask oneself . . .

What it is to try to grasp the essence of the thing is to dig down to a oncefor-all answer to metaphysical questions such as ‘What is knowledge?’ or ‘What is a proposition?’ (§92). Framing these questions gives a particular direction to philosophical enquiry, diverting attention from concrete cases (BB 20) and generating dissatisfaction with what is ordinarily called ‘knowledge’ or ‘propositions’ (§105). It is in precisely this context (not in general) that Wittgenstein recommends looking at how we ourselves ordinarily speak about these things, i.e. our own non-metaphysical use of words. The point is to persuade us to look and see how we ourselves use

Metaphysical/Everyday Use

103

particular words (‘knowledge’ (BB 27) or ‘proposition’ (§108b)). No claim is made that this use is sacrosanct, or that we have no right to depart from it (e.g. by introducing distinctions which are not commonly made for the special purpose of eliminating certain puzzles). Instead, the point is persuade ‘the metaphysician’ to clarify precisely why he is not content to stick to this familiar use in this particular context, i.e. on why he feels driven to say something different (cf. §39). It is worth noting that this gloss on §116a has important implications for interpreting many of Wittgenstein’s most celebrated remarks. Being advised by him to avoid asking the question ‘What is a proposition?’ (§92), we should presumably take him to avoid asking such questions as ‘What is the meaning of a word?’ or ‘What is following a rule?’; and equally to avoid giving answers that formulate the essences of these things. Is it reasonable for us to interpret him as asserting ‘The meaning of a word really is its use in the language’ (§43) or ‘Following a rule must be a social practice (custom)’ (§199)? The philosopher who makes such claims (allegedly on Wittgenstein’s authority!) is apparently using words while trying to grasp the essence of the things. Thereby he himself makes a metaphysical use of words; and his example serves as a paradigm of ‘The dogmatism into which we fall so easily in doing philosophy’ (§131).

IV This paper has argued for a new interpretation of the crucial contrast in the remark ‘What we do is to bring words back from their metaphysical to their everyday use.’ Though this remark is potentially illuminating, it has in fact contributed to a kind of genre-misidentification in respect of Wittgenstein’s later philosophy. It has led commentators to rummage through his writings to compile dossiers of ‘linguistic facts’ (cf. Z §447) in clarifying particular concepts. But we should instead see him as trying to do justice to individuals’ metaphysical uses of words by bringing to light what motivates their utterances,36 and we should examine in detail how he makes use of possibilities, analogies, comparisons and pictures to loosen the grip of underlying prejudices. There is much more depth to explore in this remark. In particular, we need to clarify what is the point of reminding individuals of their own everyday (non-philosophical) use of words and what are the methods appropriate for leading words back from their metaphysical to their everyday use. We also need to show how his writings quite generally exemplify the claim that this is ‘what we do’. Unsurprisingly, the answers to these

104 Reading Wittgenstein

questions will be at least as surprising as Wittgenstein’s understanding of the contrast ‘everyday’/‘metaphysical’ in §116.

Notes 1

2

3 4 5 6 7 8

9 10

In all quotations, italics mark the author’s emphases, whereas bold type marks added emphases. English translations of unpublished remarks are my own. All references to Wittgenstein’s published works use standard abbreviations. My interpretation has some clear affinities with some ideas variously developed by Stanley Cavell, Burton Dreben, Cora Diamond and James Conant, but I cannot discuss any of the similarities or differences here. Compare BB 25: ‘We, in our discussions . . . , constantly compare language with a calculus proceeding according to exact rules.’ This contrast is drawn by Friedrich Waismann (HISP 18, 20–1, 29). PI §116a seems to invite this gloss. I discuss this point later. The antithesis holds only within the class of apparently significant utterances; i.e. it is not to be applied to gibberish or patent nonsense. In their case, however, the principle of verification was used as the criterion of significance. Glock, A Wittgenstein Dictionary, pp. 261–2: ‘it is constitutive of metaphysical theories and questions that their employment of terms is at odds with their explanations and that they use deviant rules along with ordinary ones’. In this style of criticism, little allowance is made for significant changes in word-use or for authors’ own explanations of their terminology. ‘The therapy for the illusions generated by this ‘‘metaphysical use of words’’ is to examine, with unbiased vision, the ordinary use of these words. It will then be evident that we have been misusing them’ (Baker and Hacker, Wittgenstein: Understanding and Meaning, p. 526; henceforth this work will be cited as ‘B&H 1’ with page numbers.) For example, ‘an impartial description of the facts of language and its use’ shows clearly that ‘ ‘‘proposition’’ and ‘‘language’’, as we employ these terms . . . are family resemblance concepts’ (B&H 1.516), and this demonstrates that, to the question ‘do we have a single concept of a proposition? . . . Wittgenstein’s answer is clearly negative’ (B&H 1.567).

11

12

In the lectures of 1935–6, attention is frequently directed to the inclination to assimilate the grammar of mental images to the grammar of ordinary pictures (portraits and landscapes). Words such as ‘language’, ‘Satz’, and ‘number’, ‘if used intelligibly at all, must have their everyday meanings; this must be true of them even if they appeared in a philosophical grammar’ (B&H 1.341). Consequently,

Metaphysical/Everyday Use

13

14

15

16 17

18 19

20 21 22 23 24

105

reviewing the idea of the general propositional form in TLP, Wittgenstein in §108 condemns this ‘as an illusory theory-based requirement grossly at odds with ordinary language’ (B&H 1.516). More generally, confusions among philosophers have been generated or exacerbated because they ‘have not recognized that the concept of a sentence with a sense (‘‘proposition’’ in its traditional use) is a family resemblance concept’ (B&H 1.497). The notion of a category-mistake is conspicuous in Ryle’s discussions. But Wittgenstein too is taken to identify category-mistakes in treating sensations as objects (§§293, 304) or remembering, understanding, thinking, etc. as processes or states (§§305, 308, 321, 332). So unequal are these two words that Wittgenstein’s philosophy can be described simply as ‘bringing words back to their everyday use’ and that §116b can be cited in the form ‘to bring words back . . . to their everyday use’ (Hanfling, Philosophy and Ordinary Language, pp. 38, 40). On this view, Wittgenstein’s remarks describing the uses of words are defective to the extent that they leave room for disagreement. For example, it might be disputed whether the grocer’s response to the slip ‘Five red apples’ constitutes an everyday use of our language (§1); whether the analysis given of the concept of ‘game’ (§66) is actually correct (Rundle, Wittgenstein and Contemporary Philosophy of Language, pp. 45–54, 64–6); or whether the assertion that there are countless kinds of sentences is not a grotesque exaggeration (Dummett, Truth and Other Enigmas, pp. 437–58). This must be a large part of the explanation of why the Philosophical Investigations has been relegated to the sidelines in contemporary analytic philosophy. See Hacker, Wittgenstein: Meaning and Mind (henceforth cited as ‘H 3’) on the concept of thinking (pp. 311 and 324, but cf. 306n), and Kenny, The Metaphysics of Mind, on most topics. Kenny, The Metaphysics of Mind, pp. v–vii: one is more subtle and indirect, the other more outspoken and aggressive. G. E. M. Anscombe, ‘The First Person’, in Mind and Language, ed. S. Guttenplan (Oxford: Clarendon Press, 1975), pp. 45–65; Norman Malcolm, ‘Whether ‘‘I’’ is a Referring Expression’, in Wittgensteinian Themes: Essays, 1978–1989, pp. 16–26; Anthony Kenny, ‘The First Person’, in The Legacy of Wittgenstein, pp. 77–87. Norman Malcolm, ‘Wittgenstein: the Relation of Language to Instinctive Behaviour’, in Wittgensteinian Themes: Essays, 1978–1989, pp. 66–86. Baker and Hacker, Wittgenstein: Rules, Grammar and Necessity, pp. 56–7, 271–2. H 3.318–26, 347–54, 521–4. Translation: ‘The rule appears to be the mirror-image of essence in (the) language.’ Translation: But if I state as a metaphysical proposition ‘I can always only conjecture the /a/ cause’, then that means: in the case of a cause I wish to

106 Reading Wittgenstein speak only of ‘conjecture’ and never of ‘knowledge’ in order to distinguish in this way between cases whose grammar differs from each other . . . So what in our example is false is the word ‘only’. . . . 25 26

This remark occurs in Friedrich Waismann, ‘Methode,’ in The Voices of Wittgenstein (2003). Translation: We trace over and over again our form of expression, fancying that we have sketched the thing itself. – Through an optical illusion we seem to see in the inner nature of the thing what is etched on to our spectacles. . . . The expression of this illusion is the metaphysical use of our words.

27

As the application of a word is widened, its significance is progressively diminished until finally it disappears altogether. Es geht mit dem Wort ‘Satz’ wie mit dem Wort ‘Gegenstand’ und ¨ nkte Spha ¨ re angewandt sind sie zula ¨ ssig andern: Nur auf eine beschra ¨ re ausgedehnt werden, damit und dort sind sie natu ¨ rlich. Soll die Spha der Begriff ein philosophischer wird, so verflu ¨ chtigt sich die Bedeutung der Worte und es sind leere Schatten. (BT 60) [The word ‘‘proposition’’ stands on a par with the word ‘‘object’’ and other such: Only in a restricted sphere are they properly applied and there they are commonplace. If the sphere is extended, then the concept becomes a philosophical one, and thus the meaning of the words evaporates and they are empty shades.]

This may be the point of TLP 6.53: ‘whenever someone else wanted to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his proposition’. Why? ‘If a sign is useless, it is meaningless. That is the point of Occam’s maxim’ (TLP 3.328). (But I cannot defend this interpretation here.) 29 This is a paraphrase of unpublished notes by Margaret Macdonald of the lecture of 26 February 1936. 30 This question ‘almost sounds like a question of physics’; it is a question ‘in the form of a scientific question’ (BB 35). One might wonder to what extent scientific questions have distinctive forms. Some candidates might be ‘In what does . . . consist?’, ‘What more is there to . . . than to . . . ?’, ‘What accounts for . . . ?’. What Wittgenstein has in mind here is presumably not exhausted by the contrast between putting questions in the material mode and in the formal mode.

28

Metaphysical/Everyday Use 31

32

33

34 35

36

107

This seems to be the basis of Russell’s criticism of the ideas of the later work of Wittgenstein. See The Collected Papers of Bertrand Russell (London and New York: Routledge, 1997), vol. 11, pp. 601 and 615. Translation: ‘Everything essential about thought is summed up in saying that the thought that p is the case is not the fact that p is the case. That the thought is another fact.’ Compare CV 63: what is called ‘denying free will’ or ‘denying responsibility’ is not factual belief, but an attitude: it is manifested in one’s not holding other people responsible. The text of BT 60 (quoted in n. 27 above) suggests that we might well add ‘natu¨rlich’ (‘natural’) to the list. A crucial instance of this pattern of reasoning occurs in PI §244. In quoting this remark as formulating a ‘theory of avowals’, the phrase ‘Here is one possibility’ is frequently omitted. He tries to cultivate individuals’ suspicions about their own metaphysical utterances. ‘Aber ich muss doch einen Befehl verstehen, um nach ihm handeln zu ¨ chtig. Wenn das wirklich ein Muss ist – ko¨nnen.’ Hier ist das ‘muss’ verda ich meine – wenn es ein logisches Muss ist, so handelt es sich hier um eine grammatische Anmerkung. (BT 17) [‘But I must surely understand a rule in order to be able to act in accord with it.’ Here the word ‘must’ is suspicious. If that is really a must – I mean, if it is a logical must – then what is at stake here is a grammatical observation.]

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

B

Applications: the ‘Private Language Argument’

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

5 The Reception of the Private Language Argument*

I will try to clarify an important aspect – perhaps even the most important aspect – of the reception of Wittgenstein’s philosophy amongst anglophone philosophers. It has to do with the so-called ‘private language argument’, which is found at the heart of the Philosophical Investigations. Since 1953, when the book was published, this series of remarks (§§243–315) has remained the most renowned, and the best known, of all the texts of Wittgenstein. Analytic anglophone philosophers accord an important role to the private language argument. They teach this text to philosophy students, and it serves as a point of departure for their own research. Perhaps Wittgenstein came up with some original and even interesting thoughts concerning the meaning of words or on the nature of mathematics, but it is the private language argument and it alone which is accorded an eternal importance, it is this which raises its author to the rank of the philosopher-kings. The initial enthusiasm of anglophone interpreters stemmed from the fact that they perceived a striking similarity between Wittgenstein’s text and Ryle’s masterpiece (The Concept of Mind). Ryle battled against the ‘myth of the ghost in the machine’. This myth puts forward the idea that there are two realms, parallel to one another: the one comprises all physical objects (as well as all physical events, properties and relations, etc.), the other all mental objects (thoughts, emotions, sensations, etc.) as well as all mental activities and events, etc. Ryle claimed that dualism had its roots in the work of Descartes. He also affirmed that dualism is really a myth, a completely mistaken doctrine arising out of a series of categorymistakes. Ryle thus tried to refute Descartes’s myth. He denied that each * Originally published (in French) in Fernando Gil (ed.), Acta du Colloque Wittgenstein (Colle`ge International de Philosophie, 1988), pp. 29–40. Mauve´ ditions Trans-Europ zin, France: TER, 1990. Reprinted by permission of E Repress.

110 Reading Wittgenstein

person possessed an interior realm which was his own private property. He battled against the idea that each person possessed an immediate and infallible knowledge of his own mental life about which no one else could know anything (the myth of introspection), and he opposed the claim according to which the only connections between the mind and the body were causal connections (the paramechanical hypothesis). Ryle’s book had just appeared (1949) when the Investigations was published, and everyone was acquainted with it. It was therefore completely natural to believe that Wittgenstein’s remarks concerning the hypothesis of a private language must have the same raisons d’eˆtre. Anglophone philosophers have tended to see the so-called private language argument, since its appearance, as a courtroom trial, or even as a battle. On one side we find a whole crowd of great philosophers: the king amongst them is Descartes, but his army comprises Locke, Berkeley, Hume, Russell, James, Schlick, etc. On the other side Wittgenstein himself stands all by himself, like David against the Philistines. And unassisted, they suppose, he wins the battle. The private language argument constitutes a kind of reductio ad absurdum which succeeds in refuting all forms of dualism, solipsism and idealism. Such is the framework of the reception of Wittgenstein. It leaves a great deal of freedom for arguing about the details of the private language argument, and anglophone interpreters have profited from this for thirty years. There are as a result a whole gamut of interpretations, with virtually no one suspecting the possibility that they have misunderstood the surroundings, or the genre, of Wittgenstein’s philosophical investigations. Here is a schema of diverse interpretations which are on the market: 1

2

3

4 5

Wittgenstein is in the business of examining the idea that the mind constitutes an inner and private realm. He underlines the defects of this idea, the most celebrated statement of which we owe to Descartes. Wittgenstein is involved in an imaginary dialogue concerning the hypothesis of a private language, and this exchange has very important implications for the dualism expressed by Descartes. Indeed, the argument shows that Cartesian dualism is nothing but a cul-de-sac. Wittgenstein is engaged in a dialogue which focuses on a Cartesian question, that of knowing how I can understand a sign in my own private language. The argument only interests us because of its implications for Cartesian dualism. The private language argument has as its goal to clarify the negative implications which serve to refute Cartesian dualism. Wittgenstein aimed to refute Descartes’s dualism as well as to demolish a whole multitude of related doctrines. The private language

Reception of the Private Language Argument

6

111

argument followed the same strategy as that which served Ryle in The Concept of Mind. Wittgenstein’s argument takes the form of an argument with a spokesman for Descartes. The interlocutor plays the role of adversary in this imaginary dialogue, Wittgenstein’s intention being to contradict everything which this interlocutor formulates in Descartes’s name. Wittgenstein finds himself face to face with a cunning opponent who does his best to defend his partis pris.

These different interpretations of Wittgenstein’s intention are evidently associated with different interpretations of the hypothesis of a private language with which Wittgenstein is occupied. (1) There are arguments about the extension of the word ‘sensation’ (Empfindung) or that of the word ‘experience’ (Erlebnis). Some only take it to include sensations (pain, toothache, nausea, etc.), perceptions (visual, auditory, etc.), and emotions and moods (fear, joy, anxiety, etc.). Others add in ideas, images, memory images, etc.; in their opinion, Berkeley and Hume supported the possibility of a private language by virtue of claiming that the meaning of any word whatsoever is an idea of which the reader or hearer is aware. Still others claim that Wittgenstein meant to take into consideration everything which Descartes called ‘thoughts’ (cogitationes); according to them, the extension of the word ‘sensation’ encompasses intentions, desires, thoughts, judgements, etc. Such disagreement has important implications for the structure of the text of the Investigations. Some consider that the remarks which concern thought and images (§§315–421) pertain to the private language argument, whereas others see these as separate. (2) One could equally argue about the meaning of the word ‘associate’. The hypothesis of a private language carries the presupposition that one associates the sign ‘S’ with a particular sensation. Certain interpreters claim that the word ‘associate’ backs off all mental mechanisms and would therefore reject the idea that the prive´phone (the speaker of a private language) could make use of an image in order to link the word ‘S’ to a sensation. But others consider that the hypothesis of association allows for a mental mechanism of some sort, and in their opinion the consideration of images of sensations plays a very important role in Wittgenstein’s argument. Someone who considers this argument as a sort of philosophical therapy must try to establish which are the philosophers who have contracted the disease which Wittgenstein tries to cure. This judgement is of course as uncertain as the meaning of the word ‘associate’. For my part, I will attempt to show (or at least to suggest) that the framework of all the current interpretations of the private language

112 Reading Wittgenstein

argument bears little resemblance to the actual text of the Investigations. I will note nine difficulties which bar the route to this reception of the argument. There are various paths one might take to make these difficulties perceptible; I myself would advise listening attentively to the descriptions which, from time to time, Wittgenstein gave of his own procedure, to observe what he did not say in the course of his reasoning, and to clarify the most salient differences which distinguish his work from that of Ryle. Thus can we hope that a new light will dawn slowly over the whole landscape which Wittgenstein traversed in the course of his investigations. (1) The private language argument differs in several respects from the model of a reductio ad absurdum of Cartesian dualism. Wittgenstein offered no criticisms whatsoever of the idea that there is a kind of inner perception (introspection) which gives us knowledge of our own sensations. Although he rejects the idea according to which each person knows his own present experiences indubitably, Wittgenstein formulates this critique in a completely undogmatic manner (§246). Moreover, he does not criticize the idea that the mind and human behaviour are linked together according to a causal model; he does not reject the ‘paramechanical hypothesis’, which undoubtedly constitutes one of the principal defects of Cartesian dualism. Nor does he advise us to discard the idea that the mind is something inner and hidden. According to him, this image is far from being worthless, but in using it we must be wary of losing sight of the use of psychological terms. (2) In considering the private language argument as a reductio, there is a tendency to separate the negative argument from the positive remarks which have as their goal the clarification of the grammar of psychological terms. It is thought that we can take the negative argument for something definitive, while the concept of criteria, or the so-called ‘doctrine of avowals’, are considered speculative, doubtful, or even false. But such a separation does not conform to the text. The hypothesis of a private language implies that the names of sensations are not linked to natural expressions in the usual manner (§256). It is stated that human behaviour does not permit giving criteria for judging whether the use of the names of sensations is correct or not (cf. §§258, 580). The hypothesis also implies that the prive´phone makes use of the names of sensations to describe his own experiences (§256), namely in order to note in a calendar the reappearance of the sensation associated with the sign ‘S’ (§258). The absurdity of the hypothesis entails that the names of sensations must possess outer criteria. In rejecting it, one must reject the prejudice that every expression has the same function, namely to describe something

Reception of the Private Language Argument

113

(cf. §304). The critique of the hypothesis implies then the possibility that the expression ‘I have toothache’, and even the prive´phone’s expression ‘S’, do not have the role of describing a mental state (cf. §§180, 244; p. 189). (3) The interlocutor does not always play the role of an adversary. He sometimes seems naive and unbiased, though intelligent and fearful of being mistaken. He draws consequences from what Wittgenstein has just said (§§189, 281, 307). He states truths himself without understanding them (§§246, 253, 263, 296). The interlocutor is not a philosopher who comes out with doctrines borrowed from Descartes, much less does he obstinately defend Cartesian dualism against the accusations which Wittgenstein addresses to him. Rather, he resembles the average member of the audience of Wittgenstein’s seminars at Cambridge. He tends to fall into the traps which our language prepares for us, but he wants to escape from them as much as the fly wants to escape the flybottle (§309). (4) The private language argument does not adhere consistently to the Cartesian point of view. It is often claimed that Wittgenstein encourages the reader to imagine himself in the role of the prive´phone. Each one of us must ask himself: ‘Can I distinguish the correct use of ‘‘S’’ from the incorrect use?’ In this respect, it is supposed, the private language argument resembles Descartes’s Meditations. But the text does not really conform to such a model. Wittgenstein certainly talks on occasion of the language whose function is to describe my own experiences and which would be such that no one but me could understand it (§256). He also imagines that I make a genuine use of the sign ‘S’ in a calendar (§270). But at other times Wittgenstein treats the prive´phone as another person, whose behaviour we observe. The words of the prive´phone’s language must be connected to his immediate private experiences (§243). We must ask ourselves: Why do we call ‘S’ the name of a sensation when we see the prive´phone write ‘S’ in his calendar? (§261). The private language is often regarded as a language game which someone else is playing. (5) People often complain about the illegitimate use of the words ‘sensations’ and ‘experiences’ to which Wittgenstein helps himself in formulating the hypothesis of a private language (§§243, 256). The point is that they suppose that he evidently has the intention to consider all inner states, that is to say all the acts and all the processes which take place in the soul or the human mind. His critique, it is thought, loses its force if it does not concern everything which Descartes called ‘thoughts’. But the words ‘sensations’ and ‘experiences’ have less general meanings. Neither one encompasses images or intentions, much less what are called

114 Reading Wittgenstein

‘thoughts’, in everyday language. Wittgenstein misuses these two words in formulating the hypothesis which he attacks, or at least he fails to correct a very grave defect of the hypothesis of a private language, yet he is supposed to be a doctor who is doing his best to cure this disease. (6) It is often claimed that the very idea of an ostensive definition implies the use of a sample to which one can appeal in applying the defined word. It is supposed, moreover, that it is necessary that the sample be sufficiently permanent to guarantee that the meaning of the word does not alter from one moment to the next. And as a sensation, whatever it is, is more or less ephemeral, the private ostensive definition of the word ‘S’ cannot be effected unless the prive´phone takes something more permanent than a sensation to play the role of sample associated with ‘S’. It seems therefore that he must appeal to an image of the sensation which he wants to associate with ‘S’. Most interpreters of Wittgenstein think that images of sensations play an indispensable role in formulating the hypothesis of a private language, as well as in expressing the reductio which Wittgenstein is trying to construct. Nonetheless, the text of the Investigations does not really fit such an exposition. Wittgenstein makes no mention whatsoever of a sample in the course of the argument. Nor does he think that a permanent sample is needed in order to guarantee a permanent meaning to a word which one defines by means of an ostensive definition; there is no canonical sample of the word ‘sepia’ deposited in a display cabinet in the pavilion of Breteuil a` Sevres, yet we understand it as well as the word ‘metre’ (cf. §50). Finally, mental images only play an absolutely minimal role in the argument of the Investigations (§§265, 300–2). We need to distinguish a justification in relation to an image from an imaginary justification of the use of the symbol ‘S’ (cf. §258). The empiricists stress the distinction between sensations and images, and consequently, for them, the hypothesis according to which the prive´phone associates ‘S’ with a particular sensation is entirely different from the hypothesis according to which he associates it with an image of a sensation (unless the relation of association allows a mental mechanism of some sort tying the symbol ‘S’ to the sensation signified). In claiming that images play an important role in Wittgenstein’s argument, we are in danger of allowing modern empiricists (James, Russell, Schlick, Carnap) to escape this argument. (7) In enlarging the extension of the words ‘sensations’ and ‘experiences’ so as to encompass desires, intentions, expectations, hopes, judgements, etc., the hope is to widen the scope of the private language argument. The claim is that Wittgenstein aimed at the totality of mental

Reception of the Private Language Argument

115

acts and processes. If he is in the business of criticizing the idea that the human mind constitutes a secret and mysterious realm, he must not divide up inner life. On pain of rendering his argument incomplete, he must therefore encompass all the Cartesian cogitationes within the hypothesis of a private language. But this interpretation of the text contradicts a fundamental distinction which Wittgenstein outlines on several occasions. Expectations, intentions, thoughts, etc. are essentially articulate. As a result, the expectation that N.N. is going to take tea at 4:30 p.m. could not be a particular sensation (BB 20–1). Again, the thought that it will rain tomorrow is nothing inarticulate or amorphous (cf. PG 41). It does not correspond to the sentence which expresses it as a pain corresponds to the cry which expresses it (§317). But we could never give an ostensive definition of something articulate (BT 189). It follows therefore that an expectation, a hope, a thought, or whatever could not be defined by means of an association with an interior object. According to Wittgenstein, all those mental objects which are articulate are located outside the scope of the private language argument. (8) Ryle thinks that we are inclined to offer bad descriptions of the grammar of a language whose practice we have mastered. The theory of what one is doing is a long way away from everyday know-how; we often go astray by virtue of neglecting differences of logical category between words whose superficial forms are similar. We need some landmarks, which Ryle strives to bring to light. He stresses for example the differences between attention to what one is doing and a supplementary activity which accompanies the first activity; we must deny, according to him, that thinking belongs to the category of activities. In all these respects, the text of the private language argument fails to conform to Ryle’s pattern of reasoning. Wittgenstein does not think that the difficulty of seeing and describing the use of words arises from the difficulty of discovering a theory capable of properly explaining a practice. He believes, on the contrary, that this difficulty arises from the presence of pictures of grammar which we have constructed in order to orient ourselves. If we could get rid of these obstacles, we could easily describe the use of words. The issue is no longer category mistakes. In making use of categorial terms, we risk masking the use of words which we are trying to describe. The word ‘process’, as indefinite as it is, indicates a path which we must follow in clarifying the nature of what we call processes, and it could mislead us in our endeavours to outline the nature of thought or that of understanding (cf. §308). It is not for Wittgenstein a matter of perceiving that one is committing an error in classifying memory among mental processes (§§305–6); but, according to him, we risk going astray in calling thought

116 Reading Wittgenstein

a mental activity; again this has less to do with faults in logical grammar (he himself, indeed, called thinking the activity of operating with signs) than with a misleading image which we are inclined to associate with the adjective ‘mental’ (BB 6). Wittgenstein’s principal activity did not consist in clarifying or correcting category mistakes, and it differs in this regard from the task with which Ryle occupied himself constantly. (9) Wittgenstein announces the intention conscientiously to avoid dogmatism (§131). He claims to offer us new objects of comparison and advises us to consider things from a different point of view than our habitual one. Moreover, he suggests that we should describe the use of words rather than seeking to explain them, and remarks that many philosophical problems arise from an abuse of our everyday words. Doesn’t all that indicate that his goal was to enumerate grammatical facts, in order to resolve philosophical problems? But perhaps such a project might generate some uneasiness. Don’t we risk falling into dogmatism once again? How could Wittgenstein sustain the thesis according to which any kind of ostensive definition requires a sample? How does he know that neither a sensation nor a mental image can play the role of a sample? One might think that in affirming such generalizations he commits a petitio principii. These are hardly trivialities to which everyone agrees the moment they are pronounced (cf. §128). Wittgenstein in fact pursues a more modest strategy than people have suspected. He limits himself to sketching possibilities: for example, he puts forward the hypothesis that one could get someone to understand the word ‘pain’ by teaching him to replace the natural expression of pain by the utterance of this word (§244). Wittgenstein had no intention of establishing how the German infant really learned the meaning of the word ‘Schmerz’; rather, he wants to undermine the prejudice according to which we always use expressions of pain in the same way, namely to describe one’s own experience (§304). In fact Wittgenstein is rejecting a kind of dogmatism, and to accomplish this negative task there is no need to verify any fact of grammar. He tries to show us an aspect of the everyday use of words to which we are insensitive (§129). In summary, if one considers the private language argument as a reductio of Cartesian dualism, it manifests a large number of defects, and rather grave defects. One might even conclude that Ryle has anticipated what is valuable in the Investigations, and that his manner of casting suspicion on ‘the ghost in the machine’ is more systematic and more complete than that which Wittgenstein later framed. If we grant that we are mistaken about the genre of argumentation to which the private language argu-

Reception of the Private Language Argument

117

ment belongs, can we avoid asking ourselves whether this text really merits the esteem which has been accorded to it? Is it really Wittgenstein’s masterpiece? Or rather, doesn’t the reception of this argumentation testify to multiple errors of interpretation? To understand the private language argument properly, we must start over again from scratch. I will end with some suggestions about the way to recommence the analysis of the text. Wittgenstein was not acquainted with the works of Descartes, and, besides, he does not think that the confusions of today’s philosophers arose from the sins committed by the great philosophers of yesterday. He always tries to anticipate the diseases which we are inclined to contract and only occupies himself with the celebrated treatises of the history of philosophy in order to find typical examples of diseases which seem to him natural. Finally, he claims that one must trace an exact outline of the physiognomy of each philosophical error, in order to cure it. There are then powerful reasons to conclude that Wittgenstein is not engaged in a battle against a more or less definite ‘Cartesian’ adversary. It seems to me closer to the truth to say that Wittgenstein submits to scrutiny the ideas which characterize modern empiricists, especially the works of Russell which he knew very well (The Analysis of Mind and ‘The Limits of Empiricism’). Russell proposes a causal theory of meaning which testifies to the influence of the behaviourists (especially Watson). He believes that the foundations of language are found in a vocabulary which serves to describe immediate experiences. According to him, the child learns to produce the word ‘red’ when he is presented with something red. That is to say, the infant understands what ‘red’ means by virtue of associating the word with a sensation with which he is acquainted, and the meaning of ‘red’ is nothing other than the habit of reacting to the reappearance of the sensation by the production of the word ‘red’. Thus we find in Russell the exact model of a private language (§§256, 258). Russell does away with all consideration of the role played by samples, images capable of linking words to sensations, memory images, etc. Could it be that Wittgenstein tailor-made his reasoning to fit this hypothesis, with which, moreover, we know he was concerned? Perhaps the private language argument is the heir of the critique of the causal theory of meaning attempted earlier. His declared intention is to demonstrate a method by means of examples (§133). How does the private language argument conform to this method? It is generally thought that Wittgenstein aims to outline the framework necessary for a description of the grammar of psychological concepts, right down to the last detail. It is said that he was in the business of logical geography, exactly as Ryle was. The intention of the private

118 Reading Wittgenstein

language argument would from this point of view be negative, Wittgenstein trying to demolish houses of cards which occupy the ground on which he wants to build and prevent him from erecting something more solid (§118). The method which he practises is to describe the actual use of words, and thus to replace metaphysical uses by everyday uses, in order to dissolve philosophical ‘problems’ (cf. §116). Thus the private language argument plays an entirely auxiliary role. But one might consider the Investigations from a different point of view. Even though the use of our language is familiar to us and we have it, so to speak, ‘before our eyes’, we have the greatest difficulty in perceiving it and describing it without going astray (§§89, 92). Perhaps what interests Wittgenstein first and foremost is explaining why this task, which seems so simple, is in fact so difficult. He wanted, then, to teach us how to get rid of grammatical illusions which create obstacles for us (§110). He wanted to make us sensitive to the pictures and prejudices in the grammar of our language which impede a clear view. It is, then, a matter of knowhow. Wittgenstein sought to show us a method by which we can fend for ourselves in each case. The investigations which he pursues here, in respect of psychological concepts, are evidently related to those which he pursues elsewhere with respect to mathematical concepts (PI, p. 232). The goal of the private language argument is to unveil the subtle but catastrophic influence that the picture of the mind as a ‘secret realm’ exercises on someone who wants to clarify the use of words like ‘think’, ‘understand’, ‘have toothache’, ‘be afraid’, etc. It is not an issue of logical error (cf. §110), nor of a reductio. And the importance of the private language argument is not proportional to that of the philosophers who put forward the idea which Wittgenstein is in the process of submitting to examination. (We must not assume ‘Aux grandes reme`des, les grands maux’.) The therapy he aims at is much more general than the simple treatment of Cartesian diseases. The reception of Wittgenstein among anglophone philosophers, as enthusiastic as it is, constitutes an obstacle which continues to interfere with our profiting from the richness of the legacy of the Investigations.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

6 Wittgenstein’s Method and the Private Language Argument*

There appears to be a mismatch between the two most famous ‘chapters’ of the Philosophical Investigations. Wittgenstein’s remarks on the nature of philosophy (§§89–132) seemingly do not accord at all with the method which he practises in formulating the private language argument (§§242–315). Although people are seldom sensitive to this difficulty, we must attempt to ‘dissolve’ this problem of interpretation in order to understand the true content of each of the two chapters. On the one hand, there is an inclination to take the private language argument (PLA) as a reductio ad absurdum akin to the transcendental arguments found in Kant. The hypothesis to be refuted asserts that each word which signifies something inner (e.g. ‘pain’, ‘anger’, toothache’, ‘thinking’, ‘imagining’) is to be explained by virtue of being associated with something mental (e.g. with a particular sensation or an incorporeal process). According to this hypothesis, each person gives himself a kind of private ostensive definition by concentrating his attention on a particular experience, and afterwards he makes use of the word which he has just defined in order to describe his own inner life (§§243, 256, 258). The reductio consists of the demonstration that the very idea of a private ostensive definition is meaningless because the person who speaks a private language (the ‘prive´phone’) lacks a genuine pattern for distinguishing the correct use of the word from its incorrect use (§§258, 265, 270). As a consequence, no one can either understand or misunderstand any word of a private language; it is an illusion to think that one can succeed in explaining something to oneself in this manner.

* Originally published (in French) in J. Sebestik and A. Soulez (eds), Wittgenstein et la philosophie d’aujourd’hui, pp. 261–72. Paris: Me´ridiens Klinksieck, 1992. Reproduced by permission of Me´ridiens Klinksieck.

120 Reading Wittgenstein

Since it is held that Cartesian dualism as well as classical empiricism, idealism and solipsism presuppose the possibility of a private language, it is also supposed that the reductio entails the refutation of the fundamental ideas supported by a multitude of great philosophers. It is often taken as the antithesis of Cartesian dualism and is accordingly labelled ‘antiCartesian’. (Moreover, there is a tendency to link the label ‘behaviourism’ to it despite Wittgenstein’s explicit disavowal of this (§305).) In reading the text, it might be a good idea to distinguish the positive ideas here; some commentators do this and then discuss the merits or defects of the idea that all ‘inner processes’ demand outward criteria, or the idea that the utterance ‘I have toothache’ does not have the function of describing something. But it is believed all the same that at the heart of the PLA lies the reductio of the hypothesis of a private language, and that the negative reasoning remains definitive and unassailable. On the other hand, in examining the remarks concerning the method of philosophy, we discover a total prohibition against formulating any thesis whatsoever (§128) or proving anything whatever (§109). In philosophizing, we must refrain from all explanation; we must limit ourselves to describing the actual use of language (§§109, 124). Moreover, we must seek to distance ourselves from dogmatism (§§130–1). In grammar, there is no harm in making a comparison between two things (e.g. between propositions and pictures), but we must not take these useful comparisons for essential generalizations (e.g. that all propositions are logical pictures of states of affairs). There is a perpetual risk of replacing a comparison with a metaphysical thesis (cf. §116). But by the same token, Wittgenstein would apparently forbid his own method, which involves constructing complicated chains of reasoning and formulating general facts of grammar. Unless he distinguishes between negative proofs and positive proofs, or recognizes the right of philosophers to express negative generalizations, it seems that he left no room at all either for any type of reductio or for an argument concerning the possibility of a private language. If he has ruled out deduction, how could he effect a transformation of latent nonsense into patent nonsense? We seem to have reached an impasse here. What do we do now? There are three obvious possibilities: 1

2

We could make the judgement that Wittgenstein asserts a theory which he does not put into practice. He declines to follow his own advice which clarifies how we must undertake all philosophical investigations. We could water down the beverage which he seems to be offering us. On the one hand, he preaches an embargo on all theory and all

Method and the Private Language Argument

3

121

deduction, on the other he wants to teach us that we must limit ourselves to formulating grammatical truths rather than putting forward something speculative or false (‘dogmatism’). The only objects of his condemnation are deviations from the facts of grammar. We could tailor the PLA to the outline which he gives of an obligatory method for philosophizing well. We could then practise a new method of reading the text. We must get rid of the idea that Wittgenstein nourished the ambition of formulating a reductio in the battle with King Descartes and all his barons. We must attribute to him neither the intention of bringing to light mistakes in logical grammar (‘category mistakes’ as Ryle called them), nor that of making good the accusation that some philosophers among the great give themselves bumps on the head by running up against the limits of language. Far from accurately describing the facts of grammar, he is always ready to renounce any description of the use of words which someone disputes. The so-called PLA consists of nothing but a me´lange of aperc¸us. He only sketches possibilities (§244) and imaginary language games (§§258, 271). It is no part of his aim to crusade against dualism, idealism and solipsism, and he effects nothing definitive.

None of these three strategies gives us a real way out of the impasse. Each of them misconstrues an important aspect of Wittgenstein’s thought in the attempt to salvage some treasure. Indeed, it is we who have fabricated the impasse ourselves, and we could suppress it if we wished. As interpreters of Wittgenstein, we have confined ourselves to a range of possibilities which is too narrow and limited. We must start again from scratch. We must reread the text without prejudices. It might be that we can discover new possibilities which dissolve the problem of the mismatch between the PLA and the methodological remarks. Even in interpreting the Investigations, what makes a solution to a philosophical problem a real solution is solely the fact that it makes the problem disappear, pure and simple. To begin, I will try to make visible an aspect which escapes everyone when they read the famous remarks on philosophy (§§89–132). I think that there is a tendency to misunderstand their role and their content. At the least, it might be disputed whether Wittgenstein wanted to establish the essential nature of philosophy, and I want to expose a different possibility of describing what he is doing. (1) We might wonder how Wittgenstein could lay claim to his own principal declarations (that philosophy is independent of science, that it must abstain from formulating theses or theories, that it can do nothing more than describe the actual use of words, etc.). Unless they are

122 Reading Wittgenstein

prejudices or simply advice, they must describe the grammar of the word ‘philosophy’. But this is not what Wittgenstein means by these remarks, and probably this line will get us nowhere. What do we do now? The insistence that he is asserting these remarks depends on the idea that he is in the process of outlining the nature of philosophy and drawing general conclusions from this. But it might be that he is doing something else. Some of these remarks look as though they are describing what he is doing. The member of the audience at his lectures or the reader who takes part in an imaginary dialogue wonders what he is up to. He needs some landmarks in order to follow Wittgenstein, and Wittgenstein tries to indicate some. What we try to do, he says (§89), is to understand something which lies right before our own eyes. Throughout our investigations, we remind ourselves of the kinds of things which we say, and thus our investigation belongs to grammar (§90). Our considerations are not part of science, and we do away with explanations and replace them with descriptions (§109). ‘We lead words back from their metaphysical to their everyday use’ (§106). ‘The concept of a perspicuous representation is of fundamental significance for us. It earmarks the form of account we give, the way we look at things’ (§122, emphasis added). These pronouncements are presented as descriptions of the actual practice of Wittgenstein. That is the role which he assigns to them. As a result, he does not need to lay claim to them (although one can of course dispute whether he gives a correct description of his own practice). Rather than asserting a method and laying claim to it, Wittgenstein makes use of a procedure, and lays claim to the concrete results of his practice. The crucial question is whether he succeeds in making the particular problems disappear; that is the only thing which is worth arguing about. Wittgenstein intends to demonstrate a method by means of examples (§133). And if it is the details of what he has just done which one wants to dispute, the fact is that he tries to lay claim to the results of his investigations, rather than the method in general. If on the other hand one accepts the results in each case, nothing hangs on whether one also accepts the description of what he has just done. In his view, these remarks about method leave nothing to discuss. We misunderstand them completely if we require them to have a general justification. A method can only be justified by its results. (2) If Wittgenstein is describing his own procedure in saying that philosophy renounces all theory and becomes a purely descriptive activity, we might wonder what he means by the expression ‘describing grammar’. We might also like to clarify what he wants to exclude under the label of ‘theory’. In the first place one would tend to try to distinguish grammar from scientific knowledge, and hence people attribute to Wittgenstein the

Method and the Private Language Argument

123

intention to reject all scientific considerations in establishing relations between concepts. We might then think that he seeks to establish the facts of grammar in the manner of a lexicographer and that he wishes to sketch some schemata for guiding anyone who aims to describe in minute detail the logical geography of a network of concepts (e.g. the schema of mental concepts which he outlines in the Notebooks). Indeed, from time to time he does something of this sort. He takes note, for example, of the everyday use of certain words (§§66, 249), and he apparently accepts general prohibitions which depend on differences of logical category (§§308, 281, 293). Moreover, he often distinguishes the task of analysing concepts from that of putting phenomena in order. But if one characterizes the totality of his activity as ‘describing grammar’, this expression must be given a rather extended and varied sense in order to articulate his practice. The remarks in the Investigations, and even those in the PLA, have very different forms and functions, and what he describes varies a great deal from case to case. We must not suppose that every grammatical proposition approaches a tautology such as ‘No bachelors are married’. To be sure he sometimes highlights the usual use of certain words like ‘knowledge’, ‘red’, or ‘thinking’. Often he clarifies the concepts which we use in describing the functioning of certain symbols. He examines, for example, the use of the words ‘name’, ‘proposition’, ‘name of a sensation’ (§261), ‘definition’ (p. 14n.), and ‘explanation’ (§79). He sometimes describes possibilities without claiming to note grammatical facts (§§142, 244, 312; cf. BB 49, 53). He analyses language-games which are completely imaginary (§§2, 5, 258) or which are simpler than those which we ordinarily play (§§3, 5–7, 19; BB 17, 172–3, etc.). (What he describes in these latter cases is something invented, and his wording resembles the description of a character in a novel.) On several occasions, Wittgenstein appeals to something like an object of comparison, and subsequently describes something else by means of remarks on similarities to and differences from the sample (§130). The description of grammar therefore often takes the form of an analogy or a metaphor. For example, he compares ostensive definition with a rule for substituting one symbol for another (BB 109); he considers the standard metre as a tool of language (§50); he suggests regarding the proposition ‘I have toothache’ as the equivalent of natural expressions of pain (§244; cf. §181; LPP 38–40, 159); and he looks at arithmetical equations as rules of grammar (PG 173, 347; PR 143; WWK 156; cf. LFM 55, 250–1). By means of well-chosen analogies, he tries to show aspects of the use of signs to which we are insensitive, although they are obvious because they are in plain view (§§92, 129). We fail to understand descriptions of grammar chez Wittgenstein if we do not recognize their enormous diversity. There

124 Reading Wittgenstein

are a great many which are very far from being tautologies or platitudes, while some of them testify to the subtle intelligence and striking originality of their creator. To call the enterprise ‘purely descriptive’ implies repudiating the tendency to formulate theories or explanations. What does Wittgenstein aim to reject? What he calls ‘theory’ is akin to dogma (cf. §131), and this must be scrupulously expelled from his own philosophical investigations (cf. WWK 182–4). A dogmatist is someone who adheres to metaphysical generalizations. He thinks that every picture must be a logical picture of a state of affairs, that every word must be the name of something, that there must be something in common between every activity which we call a ‘game’, etc. He is completely convinced in advance that such a generalization is true, without ever having taken the trouble to examine any particular case; he persists in thinking that it is an essential truth, no matter how bizarre the consequences (cf. §113); and he refuses to examine the particular case because he is persuaded that he knows in advance what he will find (§§51, 67). Generalizations serve to shunt apparent possibilities to one side. That is the characteristic role which they fill. But, in the kingdom of grammar, Wittgenstein thinks, there is no inertia. He always advises us to pay attention to the use of words and to report what we find: ‘Don’t think, but look!’ (§66). For him, generalizations are therefore expelled; or more exactly, they are made redundant. If one becomes aware of the fact that one cannot assert a grammatical generalization without proceeding by a complete enumeration of particular cases, there is no longer any role for it to play. That means that one is never authorized to appeal to a generalization to deduce a proposition that one has not already supported by means of a direct proof. For Wittgenstein, grammatical generalizations only serve to give a re´sume´ of things which one has supported fom case to case, and Wittgenstein could thus perfectly well do without them. We misunderstand his thinking if we make a generalization into the nub of an argument. Assuming that I have succeeded above in conveying what Wittgenstein means when he announces his intention to describe grammar and avoid constructing theories, it is clear that we are going to have to start over again from scratch in interpreting the PLA. Unless he distorted the truth in describing his own practice, he could not have intended to construct a complicated argument, nor to deduce any general proposition, even a negative one. There is no place here for the traditional activities of philosophers. Wittgenstein is not seeking a foundation which he can use to deduce some new item of knowledge, and, at least in his view, none of the remarks which he formulates about the grammar of a particular word depends on a single general presupposition. The interpreter of his think-

Method and the Private Language Argument

125

ing must always ask ‘What is the particular target of this remark?’ We will undoubtedly find that the targets of his labour differ from case to case a great deal more than we had previously thought. The second task incumbent upon me is complementary to the first. I will try to bring to light aspects of the PLA which no one sees. Once again, it is a matter of presenting a picture of what Wittgenstein is in the process of doing. Equally, it is a matter of showing a possibility of interpreting the text (§§242–319). I make no claim whatsoever to prove that my interpretation of the text is the best, still less to indicate the only interpretation which fits the text. I am aiming at something more modest, and even more provisional. It will suffice for my purposes to reveal a possibility and say ‘Look at this!’, because the justification for any interpretation of the text lies in the results which it produces in the intelligent and reasonable reader. (1) Wittgenstein never uses the label ‘PLA’, and we risk venturing in a direction opposed to his if we help ourselves to this standard label. Indeed, it implies a particular conception of the ensemble of remarks which are found in the midst of the Investigations. It indicates that Wittgenstein had the intention of formulating a reductio ad absurdum of the hypothesis of a private language (that is, of a language each of whose words signifies an immediate sensation or a private experience, and which leaves no room for being understood by anyone other than the prive´phone). There is a tendency to distinguish the purely negative reasoning from the positive clarification of the actual use of names of sensations. In particular, it is often claimed that the reductio is independent of the idea that each ‘inner state’ stands in need of outward criteria, and of the idea according to which the proposition ‘I have toothache’ signals a pain in the same manner as a groan (as opposed to describing a phenomenon of the inner life of he who emits it). It seems to me that the ideas which get hooked up to the label ‘PLA’ are highly dangerous. We must prescribe a preventative medicine to immunize ourselves against deadly diseases. There are at least three reasons which speak against this interpretation of the text: (a) The very idea of a private language is evidently absurd, quite apart from considerations of private ostensive definitions, the inability of mental images to function as samples, the category difference between inner objects and physical objects, etc. A sign is something which can be understood, which one can explain or define, and of which one can make use to teach someone else something. We need to explain what a sign is by reference to the different roles which everyday words play. But it

126 Reading Wittgenstein

is not at all obvious that in explaining the roles of a sign, one can distinguish between people, between ‘you’ and ‘me’. If it is supposed to be impossible for someone to understand any word in my own private language, it must make sense to say that I explain a sign to myself, while being nonsense to say that I explain it to you. Do we need a specific argument to recognize that this hypothesis is absurd? (b) The alleged independence of the negative argument (‘the PLA properly so-called’) and the positive clarification of the grammar of the names of sensations does not begin to fit the text of the Investigations. On the contrary, the positive remarks are implied by the hypothesis of the PLA itself. In supposing that the sign ‘S’ of the prive´phone is to be explained solely through being assocated with a sensation, Wittgenstein is imagining that the use of ‘S’ does not rely on natural expressions of sensations in the usual manner (§256). The hypothesis implies therefore that the sign ‘S’ lacks criteria in human behaviour, and the absurdity of this hypothesis implies that the names of sensations must be linked to the natural expressions of sensations. The hypothesis equally entails that the prive´phone makes use of the sign ‘S’ to describe his own experiences (§256). We must suppose that he writes ‘S’ in a calendar in order to record the reappearance of the same sensation from day to day, and we must thus judge that he writes something true or false each time that he makes use of the sign ‘S’; that is, we must decide if the use which he has made of it is correct or incorrect (§258). Wittgenstein links the prejudice that every sentence is a description with the prejudice that every word is the name of an object (§§1, 304), and the absurdity of the hypothesis of a private language thus entails the possibility of considering the avowal of a sensation as the equivalent of a cry or a groan (§244; p. 189). As a result, the two positive aspects of the grammar of the names of sensations are explicitly attached to the critique of the model of the relation between name and object (§304). (c) The idea that the PLA is a reductio leads to serious difficulties. Undoubtedly it stands in need of supplementary premises in order to effect the reductio. Consequently, the proof leaves hostages to fortune; the force of the proof could be annulled through doubting any one of the hidden premises. Moreover, they seem rather doubtful, and could even constitute exemplary petitio principiis. How could we demonstrate that every ostensive definition must make use of a sample? And why must a solipsist agree that the understanding of any word whatever requires public criteria? The reductio totters unless each of its supplementary premises is unassailable. It seems that it collapses unless we take the text of the Investigations as a collection of grammatical facts, destined to erect complex buildings on the solid ground of what has been established

Method and the Private Language Argument

127

before. If the text were not cumulative, the alleged reductio would be extremely feeble. (2) The text of the PLA takes in general the form of an imaginary dialogue with an indefinite someone. He (‘the interlocutor’) asks questions, makes inferences from what Wittgenstein has just said, responds to questions, etc. In considering the PLA as a reductio, it is tempting to regard the interlocutor as an adversary, or at least as the spokesman for Cartesian dualism. He becomes the cunning philosopher who supports the possibility of a private language. Commentators anticipate what he is about to say, and hasten to put responses in the mouth of Wittgenstein which he certainly doesn’t give. They also consider the remarks of the interlocutor as objections which he hurls against Wittgenstein’s accusations. But the text does not fit the model of a courtroom trial, and the dramatis personae are more numerous than they think. Perhaps the interlocutor resembles the participants in the seminars which Wittgenstein gave at Cambridge. He could well be unbiased, or even want to escape from his own confusions and problems. And as a result, the task which Wittgenstein faces is not that of making someone leave the paradise which Descartes prepared for us, but to show someone the route by which he got there to enable him to recognize where he is and to leave of his own accord. (3) In considering the PLA as a reductio, there is a tendency to perceive important lacunae in the actual text of the Investigations. Many interpreters notice defects in the reasoning which they attribute to Wittgenstein, and they try to improve the text by making appeal to earlier or unedited texts concerning private language. They think, for example, that the critique of the private ostensive definition of ‘S’ concerns the image of the sensation which is associated with ‘S’; they believe that every ostensive definition requires a sample, and that the sensation itself, being ephemeral, cannot function as a permanent sample. Consequently, they neglect the difference between a sensation and an image (of a sensation) in the very act of formulating the hypothesis of a private language. Moreover, they attribute to Wittgenstein the idea that the image of a sensation cannot function as a sample because the prive´phone lacks a practice for applying it. Nonetheless, Wittgenstein did not, throughout the PLA, make any mention whatsoever of a sample, and the images of sensations only play a competely minimal role in the reasoning (§300). Could it not be that it is our own ide´es fixes which produce the illusion of important defects in the text of the PLA? Is it not unreasonable on the part of an interpreter to hope to express the thoughts of the PLA more completely and more neatly than the author of the Investigations himself ?

128 Reading Wittgenstein

In reading the PLA, we help ourselves to spectacles to bring the text into focus. We are inclined to judge that these passages comprise a reductio of Cartesian dualism, so much so that, lo and behold, the text conforms to this prejudice. In employing such a method of reading, we distort the remarks of the PLA by tailoring them to this Procrustean measure. Over and over again we trace the frame of the spectacles through which we look with the idea that we are going to perceive the essence of the PLA. It never occurs to us to get rid of our spectacles. I will end by suggesting a new object of comparison. It will reveal yet another neglected aspect of the PLA, but will also relieve the tension which exists between the structure of this text and the description which Wittgenstein gives of his own methods of philosophizing. One might say that he aims less at the task of inscribing in minute detail the grammar of the names of sensations (or of inner states) than at the clarification of the temptations that we have to go astray while doing so. It seems rather paradoxical that we find it so difficult to notice and to describe the familar use of words of which we have attained practical mastery. Everything that is required is in plain view, and it seems nonetheless that there is some great thing which we do not understand and that there remains a mystery to break through (§§89–90). Wittgenstein explains this phenomenon by reference to the grammatical illusions which beset us in describing the grammar of the names of sensations. He tries in particular to show that the model of the relation between name and object gives rise to paradoxes (cf. §112). It seems as if we are constrained to choose between two equally false propositions, namely that pain and the behaviour which expresses it are two different things, or that the pain is the same thing as the behaviour which expresses it. It seems as if we are likewise constrained to say that the expression of pain accompanied by pain is something more than the expression of the pain without the pain. It seems as if a sensation is not a nothing, but is not a something either (cf. §304). Wittgenstein remarks that he is trying to get us to see a grammatical fiction (§307). These are houses of cards which he sets himself to destroy (§118). He has no intention of establishing facts of grammar (e.g., that ‘pain’ is not the name of an object) or outlining psychological facts (e.g., that a child learns the meaning of the word ‘pain’ through replacing the natural expression of pain by the emission of articulate sounds). Rather, he adumbrates possibilities and makes use of new comparisons in order to ¨ bersichten) of the use of the names of sensations. He give unified visions (U counsels us to look at something problematic from a new point of view and to bring to light aspects which are hidden from us because of their simplicity and familiarity (§129).

Method and the Private Language Argument

129

The interpretation which I am going to propose has some further merits. It makes it absolutely clear why he thinks that he is in the process of demonstrating a method by means of examples (§133). The strategy which consists of dissolving philosophical problems by revealing grammatical illusions can be seen clearly throughout the text of the Investigations. This interpretation shows the fundamental unity which ties the philosophy of mathematics to the philosophy of psychology (p. 232). All Wittgenstein’s philosophical investigations clarify the same grammatical illusions, namely that every word is a name and every sentence a description. The importance of the method is disconnected from the greatness of the philosophers responsible for the doctrines with which he is occupied. He tries to impart a know-how which we can make our own by applying it to any philosophical problem whatsoever; it would be pointless to give us lessons if we were incapable of fending for ourselves afterwards. It seems to me that the interpretation of the PLA should be oriented by relating it to the landmarks which are found in the midst of the Investigations, while always referring back to the beginning of the book (§§1–5). In my view, all standard interpretations take a direction opposed to that of Wittgenstein. As a consequence, the description which he gives of his own procedure does not conform at all to what they think he is doing in the course of the PLA. In order to make contact with his thinking we must run twice as fast . . . and in the opposite direction.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

7 The Private Language Argument (Extract: Final Section)*

[...]

Fresh Bearings – or Neglected Aspects Direct attack is unlikely to prove efficacious in weakening the grip of the PLA-interpretation. Is there any eligible alternative? It seems difficult even to imagine other possibilities. The anti-Cartesian framework of interpretation almost forces itself on one; it seems not to be the best of the bunch, but rather the sole one that even remotely fits the PLA. If Wittgenstein offers a sort of therapy, what could be the disease he treats if not Cartesian dualism? And if Ryle’s attack on this myth does not correspond closely to the PLA, what might be the model of a philosophical argument which would provide a better object of comparison? There seems to be some risk of getting completely lost in a trackless desert. Isn’t it better to follow a well-blazed trail than to try to make one’s way cross-country in poor light? This kind of anxiety seems groundless. I will finish off by offering a first sketch of a very different framework of interpretation which makes sense of Wittgenstein’s remarks on ‘private’ language. My strategy will be to construct a new object of comparison (Vergleichsobjekt; PI §131) in order to highlight a neglected aspect (§129) of the PLA. I refrain quite deliberately from claiming that the aspect that I stress is the only aspect that can be used to throw light on the text. I am content to make apparent to others an aspect to which they have been blind, hoping thereby to make it possible for them to discover for themselves that it is worthwhile persisting with this approach in their attentive reading of the Investigations. * Originally published in Language & Communication, 18 (1998), pp. 325–56. ß 1998 by Elsevier Science Ltd. Reproduced with permission of Elsevier.

The Private Language Argument (Extract)

131

It seems perfectly conceivable that the real value of the PLA has not been appreciated. The problem is that, in pursuing this goal, just as in seeking for something in the Looking-glass World, everyone starts out running hard in the wrong direction. The secret of finding a better interpretation is to look for landmarks within the Investigations, not outside it. We can take our bearings from the opening remarks about Augustine’s picture of language and also from later remarks about ‘our’ philosophical method. We can then see certain unnoticed patterns in Wittgenstein’s great trek back and forth across vast tracts of the grammar of our language. Undoubtedly, he concerns himself with the grammar of words such as ‘pain’, ‘fear’, ‘thinking’, ‘(mental) image’, etc. But what is his purpose? One might suppose that there can be only one: to observe the data of the everyday institutionalized use of words and to describe the grammar of our language carefully and systematically. The ultimate goal is to obtain a yardstick against which to measure any combination of words to settle whether it is intelligible or nonsensical. In this way one can point out to an errant philosopher exactly how his use of words contravenes the rules of grammar, and one might even hope to prevent future philosophers from making similar ‘metaphysical’ use of everyday words. The therapy that Wittgenstein envisages consists entirely in correcting actual mistakes in logical grammar and in ensuring that the future use of words sticks to the rules determining their ‘combinatorial possibilities’. Ordinary language is ‘in order as it is’, but (other!) philosophers persist in committing the offence of departing from the established use of words. The proper role of a philosopher is to catalogue the everyday uses of words and to enforce strict adherence to these rules. This line of thought certainly neglects some other conceptions of therapy. Surely one might have various motives for studying and treating sicknesses. At one extreme, by considering a particular case as the prototype of a disease of a certain general kind, one might seek to isolate its causes in the attempt to treat the whole range of disorders akin to it. Then one would have both theoretical and practical interests in pursuing a specific line of research, but not an interest in curing just this specific disease or in treating this particular patient. At the other extreme, one might bend all one’s efforts to curing an individual person who has fallen victim to a disorder (someone who is an alcoholic or who suffers from severe depression). Here nothing would matter apart from the efficacy of procedures in this particular case. This too seems an intelligible point of view on ‘diseases of the intellect’. On this conception, every philosophical problem is someone’s problem. Hence one needs to isolate the sources, however idiosyncratic, of the

132 Reading Wittgenstein

confusions of this individual patient, perhaps by calling to his attention certain pictures or similes that may have led him astray. Correct diagnosis depends on eliciting a confession from him (BT 410); and the cure must be to help him to reorient his thinking – with his full cooperation. (In these respects there would be an analogy with the procedures of psychoanalysis.) In aiming at therapy, Wittgenstein might be concerned with treatment of a particular patient rather than with a kind of campaign to improve public health. The Investigations might be, as it were, a set of case histories of a general practitioner, not the execution of a campaign for ridding the world of smallpox (‘Cartesianism’). The text of the PLA certainly exhibits this point of view, even though this is seldom noted. Wittgenstein remarks that the prejudice that the phrase ‘to remember’ signifies an inner process stands in the way of our seeing the actual use of the expression; the picture we attach to the phrase ‘inner process’ leads us astray (PI §305). The notion that ‘pain’ signifies a state or process equally misleads us because we have preconceptions about what states and processes are, as well as a definite concept of what it means to know a state or process better (§308). These remarks offer diagnoses which demand the patient’s acknowledgement if they are to be recognized as correct.1 The mention here of pictures and prejudices should remind an attentive reader of the remarks which open the Investigations and set the stage for everything that follows. There Wittgenstein calls attention to a set of very general ideas that affects almost everybody who reflects on the meanings of words. Those ideas, which he pinpoints in a quotation from Augustine’s Confessions, are that every word is a name, that the meaning of a word is an object correlated with it (especially by means of an ostensive definition), and that every sentence is a combination of words which has the function of describing some state of affairs (§1). Wittgenstein notes that this widespread conception, ‘a particular picture of the essence of human language’ (§1), ‘surrounds the working of language with a fog which makes clear vision impossible’ (§5). In his view, there is no error here, but rather something akin to a superstition (§110), a picture that has its roots deep in the soil of our thinking (cf. CV 83). Such a picture is what he calls ‘a grammatical illusion’, and he suggests that many philosophical problems grow out of such illusions (PI §110). The PLA manifests a clear interest in the very grammatical illusion expressed by Augustine’s picture of language. That elaborates the model of name and designation which he criticizes in respect of the meaning of ‘pain’; he argues that the object supposedly correlated with the word plays no role whatever in respect of its use (§293). It is the same illusion which generates the notion that to describe the grammar of ‘pain’ one must

The Private Language Argument (Extract)

133

admit that it stands for something; Wittgenstein observes that we can only escape from the grammar which tries to force itself on us here . . . if we make a radical break with the idea that language always functions in one way (§304)

namely to describe something or to convey information (cf. §363). The interest of philosophical problems retreats to illusions which lie at a deeper level (cf. §§110–11), and Wittgenstein pursues his investigations by putting question marks deeper and deeper down among these preconceptions. In fact, similarity between the PLA and the opening of the Investigations goes much further. In both cases, Wittgenstein follows the same route in bringing to light the sinister influence of grammatical illusions. One of his tactics is to scrutinize the alleged correlation between words and objects. In asserting that every word in a language-game stands for some object, we really assert nothing at all; it remains an empty form of words unless we explain what distinction we wish to make (§13). In particular, we must still explain what function each of our words has in the practice of speaking the language (§51), because the claim of having correlated words with objects says nothing informative about this. We must look at things from close to and investigate how exactly the words of our language are related to the objects named; we must clarify the kind of ‘referring’ that belongs to each different kind of word, i.e. the kind of use each has (cf. §10). This is just the issue that Wittgenstein first raises in examining the private-language hypothesis: ‘How do words refer to sensations?’ He offers this rather surprising possibility: the word ‘pain’ is a learned replacement for the primitive, natural expressions of pain (§244). This answer might seem to be an evasion, but the real evasion would be to claim that all that is required is the mere association of a word with a sensation (cf. §256). That empty formula obscures the fact that no explanation is thereby given (cf. §261). Another shared tactic is the invention of language-games which strictly conform to the picture under critical investigation. Wittgenstein imagines a very simple system of signs for which the description given by Augustine is right (§§2–3). He then goes on to use this slab-game as an object of comparison to highlight both its differences and its similarities with the everyday use of words (§130) and other systems of signs (§§4, 8–9). The PLA follows this strategy in imagining a language-game with the sign ‘S’ which is taken to conform to the principle that the sole link between the sign and the sensation is simple association – nothing more and nothing less (§256).2 In conformity with this blueprint for a language-game, we

134 Reading Wittgenstein

imagine that the Private-Speaker produces the symbol ‘S’ at the moment of having a particular sensation and thereafter he writes ‘S’ from time to time in his diary. On the assumption that this is a complete description of what he has done, we have no reason whatever to call ‘S’ as it occurs in his diary ‘a name of a sensation’, or even to call it ‘a name of something’ (§261). To justify our applying these phrases to his use of ‘S’, we must establish more complex patterns than what is present, by our hypothesis, in the Private-Speaker’s activities. Comparison between genuine names of sensations and ‘S’ in this imaginary language-game makes perspicuous just how far the everyday use of the words ‘pain’, ‘fear’, etc. deviates from the model of the bare association of signs with immediate experiences. Under Wittgenstein’s guidance, when we reflect on what we imagined to be the ideal case, we find that we have nothing at all. The philosophical activity of dispelling grammatical illusions is very different from the one of working out and mapping the logical geography of ‘psychological concepts’. It more often consists in taking note of possibilities to which one is blind than of establishing facts of which one is ignorant. Too limited a range of possibilities (or too narrow a diet of examples) often gives rise to interminable debates among philosophers, so that a dissolution of problems may consist of calling attention to further possibilities (cf. LPP 47). Wittgenstein regularly practises this strategy (e.g. §§41, 66; BB 41, 43). For example, to the question ‘How do words refer to sensations?’, he replies by suggesting the possibility of teaching the word ‘pain’ to a child by conditioning him to replace natural expressions of pain (say, crying) by a verbal expression of pain (§244). Any philosopher who keeps this possibility in mind will thereby escape from the compulsion to address the question ‘What does an avowal of pain describe?’, and Wittgenstein will have liberated him without having to establish any facts about how anglophone children actually learn to use the word ‘pain’3 (and hence without any risk of falling into dogmatism or engaging in ‘armchair psychology’). This is a model of his procedure: the attention he gives to mere possibilities does not manifest his modesty, but rather his method. A second key to unlocking his distinctive intentions is to be found in his frequent criticisms of what he sometimes called ‘metalogical concepts’ (cf. BT 1, 3, 16, etc.; LPP 25–6, 265–6). In his view, philosophers tend to jump into debates without careful enough examination of the concepts they use in making their own statements. They use category-terms (‘event’, ‘process’, ‘state’, ‘fact’, etc.) with little or no explanation, and they employ as tools of the trade such expressions as ‘name’, ‘proposition’, ‘description’, ‘to refer to something’, etc., without any qualms about how these are to be understood. Wittgenstein advises us to pay

The Private Language Argument (Extract)

135

close attention to our own everyday use of these expressions to forestall our making metaphysical use of them (§116). In philosophizing, we are apt to become dissatisfied with what are ordinarily called ‘propositions’, ‘words’, ‘signs’ (§105). We imagine that we need to penetrate to the essence of propositions (§92), to discover what they must have in common (cf. §§65–6). In this case and many others, it is difficult to see that we must stick to the subjects of our everyday thinking (§106). The PLA includes investigations and clarifications of certain metalogical concepts. It tries to clean philosophers’ tools. Wittgenstein tries to reorient someone who asks ‘How does a word refer to a sensation?’ (§244). He notes that the ceremony of attending to a sensation needn’t qualify as defining a symbol (§258). He asks whether we have any reason to call ‘S’ (as used in §256) a sensation-name (§261). He counsels us to pay attention to what we call descriptions and to the differences between describing a state of mind and describing a room (§§290–1). He suggests that we go astray in speaking of processes and states while leaving their nature unclarified (§308). In short, the PLA does not address the question whether the sign ‘S’ is the name of a genuine sensation, but rather the question of whether it is a genuine name of a sensation. Wittgenstein clarifies the grammatical concept of a sensation-name. More generally, it is distinctive of the Investigations to focus on how we speak about how we speak, i.e. the framework of concepts which we use in describing the grammar of our language or the system of coordinates used in mapping the logical geography of specific concepts. One might say that Wittgenstein is less concerned to work out the details of the grammar of names of sensations (or, more generally, of inner states) than to draw attention to the many temptations, urges and prejudices that have the power to mislead us in doing so (§109; RPP II §87). He wants to make us aware of how our picture of an ‘inner state’ interferes with our seeing the use of ‘think’ or ‘understand’ as it is in our own everyday speech (cf. PI §305), or how our craving for generality stands in the way of our being able to inspect or note how we ourselves use ‘number’ or ‘language’ (§§65–7). It may seem quite paradoxical that there could be such difficulties in noticing and describing the perfectly familiar use of our words (cf. LPP 5–7). Everything that we need to note is open to view, and yet it seems that we fail to understand it; so we may end up feeling that we need to penetrate the surface of something mysterious (§§89–90). How can this happen? Some philosophers explain this puzzle by contrasting practical skill (know-how) and the theory which underlies a practice; apparently we know how to use words without knowing how systematically to describe

136 Reading Wittgenstein

our own practice.4 Wittgenstein, however, gives a very different explanation, at least in the case of some particular philosophical problems (cf. §89–90; BB 26). It may be that some important expressions (e.g. ‘think’ or ‘imagine’) are odd-job words. It is difficult to epitomize their complex and rough-edged employment, and one has some real anxiety about losing one’s bearings here. This generates the conviction that everything is mysterious or difficult to penetrate (e.g. that thinking or imagining has a queer nature). Seeking for guidance, one latches on to simplified pictures of the grammar of these problematic words. For example, the picture of thinking as an auxiliary activity which accompanies the activity of speaking aloud (PI p. 218). Such pictures may easily come to dominate our reflections, even if not obviously appropriate,5 and they often serve to oppose the examination of details of word-use (§52). In the same way, one invents pictures of how symbols in general function (pictures of the essence of language): e.g. the idea that every word names an object or that every sentence describes a state of affairs. It is these pictures, these grammatical illusions, which present obstacles to the business of describing our own use of words. We are determined to look at grammar from a certain point of view, and we persist in doing so even in the face of very severe difficulties. Consequently, the therapy for many philosophical problems must be addressed more to the will than to the intellect (CV 17). A picture may force us to confront certain questions. We feel driven or compelled to ask them; we act as if blind to any possibility of circumventing them. For example, if every word is the name of something, then any difference of word-meaning must be explained as a difference in the objects named. Mustn’t we then admit that pain and pain-behaviour are two different things? (They are surely not the same thing!) Mustn’t we further admit that pain-behaviour accompanied by pain is something more than pain-behaviour without pain (§304)? In the background may lie the natural idea that any two distinct objects must be independent of one another. If so, we have already embraced the conclusion that pain is something logically independent of human behaviour (LPE 290). Although we are free to deny the principle that any two objects must be independent, we tend to accept it unreflectively (or unconsciously), and we then find ourselves sliding down a slope into a form of dualism. It isn’t the picture by itself that brings about this philosophical Fall; rather the unreflective application of a picture which is itself innocent enough (§§422–6). Like a psychotherapist, Wittgenstein tries to neutralize the mischiefworking powers of such pictures and analogies by bringing them to consciousness (PLP 179 ¼ DS 28; cf. PPI §106). One procedure is to

The Private Language Argument (Extract)

137

replace one picture by others, to fabricate other objects of comparison (§131), to look at something problematic from a different angle, and to make apparent aspects of things that are hidden from us because of their simplicity and familiarity (§129). In this way he seeks to dissolve some of the questions that seem most obsessive in philosophy of mind, to cure us of some of our mental cramps. He exhibits neglected possibilities to bring about intellectual liberation. A man will be imprisoned in a room with a door that is unlocked and opens inwards; as long as it does not occur to him to pull rather than push it. (CV 42)

The PLA has as its purpose to bring to light the subtle but pernicious influence of our grammatical illusions on our descriptions of the grammar of our own discourse about sensations and experiences. (For this reason, ‘working in philosophy. . . is working on oneself ’ (CV 16).) From this point of view, one can clearly see the method of the Investigations. Wittgenstein does indeed abstain from any dogmatism. In looking at a particular problem, he advises us to look at things like this, not like that (CV 61); e.g. to consider the avowal of a pain as an expression of a sensation (§290; RPP I §313; LSD 11; LPE 301–2), to view an arithmetical equation as a grammatical rule governing the use of symbols in empirical propositions (PR 130; LFM 55, 251; cf. WWK 156), to regard thinking as the activity of operating with signs (BB 6, 15–17; cf. LPP 142), or to look at a sentence as an instrument and at its sense as its employment (PI §421). In each case, he puts before us a picture in order to bring about a change in our manner of seeing something, and instead of making an assertion, he does nothing more than say: ‘Look at this’ (cf. §144). He might also add the question: ‘Has the problem that was troubling you now disappeared?’ (cf. LPP 168). Quite evidently he exhibits a method by means of examples (PI §133). He does this everywhere, as much in philosophy of psychology as in philosophy of mathematics. His method is not systematically to describe in detail the grammar of symbols; this work can be left to the reader to do for himself (CV 39, 77). His method is surely meant to be transferable. But it is difficult to imagine how specific observations about the use of such words as ‘pain’ or ‘joy’ could be applied to clarify the very different use of the words ‘nought’ or ‘set’. The focus of this method is on grammatical illusions. Wittgenstein teaches us how to single out the prejudices which stand in the way of describing grammar (PI §340) and also how to overcome them in ourselves. This is a skill that can be practised in different circumstances.

138 Reading Wittgenstein

He teaches some know-how, not a body of knowledge (facts, theory, doctrine). He abstains from claiming that his method will dissolve every philosophical problem or that it is the sole correct method for philosophizing. On the contrary, he sees himself the heir of the activity which has long been called ‘philosophy’ (BB 28), and he vindicates his method by showing its efficacy in the business of arriving at a complete dissolution of some particular problems bothering some particular persons (generated by their quite specific pictures, analogies, arguments and concerns). The importance of his method (and hence of his therapy) is measured by the possibility of our putting it to work to dissolve other problems. By seeing the PLA as putting into practice his method of dispelling grammatical illusions, we dissolve any appearance of tension between this celebrated argument and his avowed intention of demonstrating a method by means of examples. This point of view sets us free from the notion that the worth of the PLA depends on the aggregate importance of the philosophers whom Wittgenstein confronts therein. It would be pointless to rake around in the texts of Descartes, Locke, etc., to find misunderstandings worthy of serious attention. Why not rest content with his treating concrete problems apparent among some of his contemporaries (Russell, Schlick, Carnap) or even among participants in his discussions at Cambridge? To the extent that this alternative picture carries conviction, the interpretation of the PLA as a reductio ad absurdum of Cartesian dualism will take on the appearance of a grotesque genre-misidentification. Wittgenstein tries there to demonstrate a sophisticated method and to impart a skill by example, not to lodge objections to certain Great Philosophers and score points off them.6 The spirit of his work is expressed in his attitude towards transfinite arithmetic. In that case, he disavows the desire to expel anybody from ‘Cantor’s paradise’, but he urges others to look around them and see where they really are located with the expectation that they may decide then to leave of their own accord (LFM 103). Perhaps the PLA is no less subversive in intention. It doesn’t try to drive anybody willy-nilly from the Cartesian paradise of the twin Inner/Outer worlds, but it encourages us to leave of our own accord once we have become conscious of the route by which we have come to be there. The focus of his attention is also misunderstood in the desire to assimilate Wittgenstein to the mainstream of analytic philosophy. His interest is not to describe in detail what is in full view (the grammar of such words as ‘pain’, ‘think’, and ‘imagine’), but rather to create the conditions under which we can do this for ourselves by teaching us how to identify and

The Private Language Argument (Extract)

139

eliminate some of the prejudices that stand in the way of our noticing important aspects of what is perfectly familiar (‘facing concepts without prejudice . . . is the principal difficulty of philosophy’ (RPP II §87)). Here we have to learn to understand the subtle and indirect workings of his mind in despite of an urge to misunderstand them.7 If we were to acknowledge Wittgenstein’s radical intentions, we might finally be in a position to participate in the PLA and to enter fully into the inheritance that he has left us in the Investigations.8

Notes 1

Compare this invitation to acknowledgement: Let us now return to Augustine’s question [‘How is it possible to measure time?’]. . . . I believe that it is a problem only for those who have been misled by an analogy, who imagine the measurement of time to be more like the measurement of space than it is. (PLP 42, italics added to my translation)

2

3 4 5

6

This point is radically misunderstood by commentators who claim that §256 makes the supposition that ‘S’ is correlated with a ‘pseudo-sensation’, not a genuine sensation (Kenny, Wittgenstein, p. 190). They suppose that ‘S’ would be tied up with my natural expressions of sensation simply by being associated with a sensation, presumably in virtue of the grammatical principle that every sensation must have outward criteria. This seems a generalization of the thesis: ‘that pain is a sensation . . . is a fact of nature which dictates the logic of ‘‘pain’’ ’ (P. F. Strawson, ‘Critical Notice of Wittgenstein’s Philosophical Investigations’, Mind, 63 (1954), p. 88). This is a double non-sequitur. Not all behavioural criteria need qualify as natural expressions; indeed, pain is distinguished from many other sensations by reason of having natural expressions (RPP II §63). Moreover, even if the sensation associated with ‘S’ had natural expressions, these would not thereby become criteria for understanding ‘S’ or using ‘S’ correctly; that role must be conferred on them by an explicit stipulation that the private-speaker is free to make or to withhold. Unlike Malcolm’s reconstruction of his reasoning (‘Wittgenstein’s Philosophical Investigations’, pp. 67–8) and Kenny’s (Wittgenstein, p. 184). Ryle, The Concept of Mind, pp. 7–8. For example, RFM 146: . . . we are inclined to take the measurement of length with a footrule as a model even for the measurement of the distance between two stars. For example, to refute the representational idealism of classical empiricism (Kenny, Wittgenstein, pp. 179–80; and Hacker, Wittgenstein: Meaning and Mind, p. 141).

140 Reading Wittgenstein 7

It is in this respect that there is the clearest analogy between his method of philosophizing and Freudian psychoanalysis: It is a long superseded idea . . . that the patient suffers from a sort of ignorance, and that if one remove this ignorance by giving him information . . . he is bound to recover. The pathological factor is not his ignorance itself, but the root of this ignorance in his inner resistances; it was they that first called this ignorance into being, and they still maintain it now. The task of the treatment lies in combating these resistances. (Freud, The Complete Psychological Works, vol. 11, p. 225)

8

This is a somewhat revised version of an article originally written and published in French. This appeared in Wittgenstein Analyse´, ed. J.-P. Leyvraz and K. Mulligan (Nıˆmes: Jacquelin Chambon, 1993). It is reproduced here with the permission of the publishers.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

Part II Wittgenstein and Waismann

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

A The Analogy with Psychoanalysis

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

8 ‘Our’ Method of Thinking about ‘Thinking’*

In the early 1930s, Wittgenstein took a distinctive approach to clarifying the grammar of a range of terms that have traditionally been taken to stand for various ‘mental processes’. These include such terms as ‘hope’, ‘wish’, ‘desire’, ‘fear’, ‘believe’, ‘remember’, ‘understand’, all of which are often classified as signifying particular species within the genus called ‘thinking’. We are likely, he thought, to fall into serious confusion in reflecting on these concepts, and we are liable to conclude that they are mysterious, paradoxical and impenetrable. He later diagnosed the source of our difficulties in the remark that the label ‘mental process’, though seemingly non-committal and hence perfectly anodyne, can be seen as the decisive move in the conjuring trick (PI §309).1 In this earlier period, he expressed a closely related idea. He urged us to drop the formula ‘Thinking is a mental process or activity’ in favour of the slogan ‘Thinking is (the activity of) operating (or calculating) with signs’. Negatively, this amounts to advice to discard the label ‘a mental process’. The correlative positive recommendation, at first sight a form of behaviourism, is a recurrent theme in the Blue Book (a kind of refrain), and it is equally prominent in his dictations to Waismann in the early 1930s. Like nearly everything characteristic of Wittgenstein’s ‘transitional period’, this remark has received little explicit attention from commentators. If mentioned at all, it is brushed aside as an immature idea, one superseded in his fully developed philosophy of psychology by his analysis of every form of ‘inner process’ in terms of behavioural criteria. I will suggest that this lack of respect for this early Leitmotif manifests a fundamental misunderstanding of all phases of his ‘later philosophy’, and I will * Originally published (in French) in A. Soulez (ed.), Dicte´es de Wittgenstein a` Friedrich Waismann et pour Moritz Schlick, vol. 2 (E´tudes critiques). Paris: PUF, 1997. Reproduced by permission of Presses Universitaires de France.

Thinking about ‘Thinking’

145

offer an interpretation of the formula ‘Thinking is operating with signs’ which aims to make good sense of it and of its centrality in these early texts. Even though he later dropped this particular formulation from his description of the grammar of ‘think’, much of what he understood by it remained permanently central to his thinking. The key to understanding this slogan is to realize that it is not meant to convey information (e.g. to correct particular mistakes of detail) about the logical grammar of ‘thinking’, but rather to persuade somebody puzzled by the nature of thinking to acknowledge certain grammatical prejudices in himself that generate his conceptual confusion. Wittgenstein’s enterprise has closer affinities with sessions of psychotherapy than with didactic presentations of the logical geography of ordinary language or with descriptions of the actual shared practice of speaking ‘the Queen’s English’.

1

A Psychoanalytic Model of Philosophy

In scattered remarks from the early 1930s, Wittgenstein made some explicit comparisons between his methods of philosophical investigation and Freudian psychotherapy. Two are especially striking: DS 28: Unsere Methode a¨hnelt in gewissem Sinn der Psychoanalyse. In ihrer Ausdrucksweise ko¨nnte man sagen, dass im Unbewusstsein wirkende Gleichnis wird unscha¨dlich, wenn es ausgesprochen wird. Und dieser Vergleich mit der Analyse la¨sst sich noch weithin fortsetzen. (Und diese Analogie ist gewiss kein Zufall.) BT 410: . . . we can only convict somebody else of a mistake if he acknowledges that this really is the expression of his thinking. For only if he acknowledges it as such, is it the correct expression. (Psychoanalysis.) What the other person acknowledges is the analogy I am proposing to him as the source of his thought.2

The analogy with psychoanalysis is not developed very far or at all systematically in these or other texts, and this makes it impossible to establish exactly what Wittgenstein had in view in drawing it. But it seems to have struck Waismann as holding the key to unlocking Wittgenstein’s distinctive method of conducting philosophical investigations. The comparison has a prominent role in Waismann’s exposition of ‘our method’ (G 26), i.e. the method learned from Wittgenstein, demonstrated in the writings of Waismann and Schlick, and advocated as the model for the style of thinking of the Vienna Circle.

146 Wittgenstein and Waismann

Though I want to bring out these particular aspects of ‘our method’, I will proceed indirectly by first sketching some of the most salient features of the psychoanalytic model of philosophy which Waismann elaborated much later. This may well be more extensive and more definite than anything that Wittgenstein himself had in mind. But this may make it all the more valuable for highlighting some commonly neglected aspects of Wittgenstein’s later philosophy. Of course, this model is to be understood as an object of comparison, not as a first approximation to describing Wittgenstein’s account of the nature of philosophy. In this role, it may serve to exhibit unfamiliar aspects of his actual practice of treating problems as well as unnoticed implications of his methodological remarks, not only in his dictations to Waismann, but also throughout his later philosophical investigations. In the paper ‘How I See Philosophy’ (published in 1956), Waismann offered a fully developed conception of philosophical therapy expressly modelled on certain features of Freudian psychotherapy. He was explicit about what he understood by the maladies addressed by psychotherapy (compulsions, obsessions, neuroses) and about what methods are appropriate for their treatment. (His text suggests that he had in view primarily Freud’s early writings.) He developed what he took to be the points of analogy in a set of articulated theses.3 I focus on four propositions that characterize his conception of philosophy. (1) Many philosophical questions are unlike most everyday questions. ‘What is a proposition?’, ‘Are numbers objects?’, ‘Is it possible for two people to have the same thought?’ – someone who asks these questions is clearly not requesting information; answering them may well be trivial, and he surely knows the obvious answers already. (‘How is it possible to measure time?’ – ‘With a clock.’ – The direct answer makes fun of the question.) Instead, such questions are regarded as tokens of the questioner’s intellectual disquiet (Unruhe), sometimes even terror or anxiety (Angst). They arise from intellectual obsessions, compulsions or ‘neuroses’. According to Waismann’s diagnosis, for example, Frege was fixated on the idea that mathematical objects must exist in order to make mathematical statements true; hence he was driven to address the question ‘What is a number?’, and he could not let it drop. Such questions need to be dissolved rather than solved or answered. The difficulty in dealing with them is to make sense of someone’s posing them; for this purpose, we need to trace their origins in his ‘mind-set’ (preconceptions, prejudices). Dealing with these questions requires sympathetic and sensitive attention to the intellectual character of the person who raises them, a kind of therapy, and the philosopher is cast primarily in the role of a

Thinking about ‘Thinking’

147

therapist. Frege, for example, needed a kind of treatment, not a lecture on the rational demerits of Platonism in arithmetic and logic. (2) The puzzlement expressed by these philosophical questions arises from the questioner’s way of seeing things, from kinds of bias or prejudice which affect his perceptions of the uses of words. Hence it can be made to disappear only by changing his way of seeing things, especially by carefully describing the use of the words that he used to frame his questions. (Waismann compared this explicitly with revealing different aspects in diagrams or puzzle-pictures, e.g. with experiencing the drawing of a cube in different ways.4 The point of this comparison is to make clear that describing the grammar of our language is not a matter of making discoveries or supplying new information to the ‘patient’, but of bringing him to apprehend or acknowledge previously unnoticed patterns in what was already ‘visible’ and fully familiar to him.) This kind of ‘transformation of the intellectual scene’ may be difficult to effect in someone else; indeed, it may be virtually impossible for even the most highly intelligent philosophers to come to see things differently. Could we even imagine, for example, that Frege might be persuaded to drop his entrenched preconception that the possibility of objective knowledge in mathematics is completely dependent on the thesis that numberwords must stand for things (objects)? This seems to be the axis of revolution of all his thinking, something unshakeable and unassailable in him, as it were, part of his intellectual identity or his Weltanschauung. In any particular case, the philosophical therapist may labour in vain; with regard to prejudices (or demands) there is no friction! (PI §107). On the other hand, it always makes sense to engage in this task of persuasion or conversion. You may urge somebody to try to look at things differently, and he might succeed in complying with your request. (In this sense, how he sees things is voluntary.) Success might come in different orders of magnitude. You may effect a change in the attitude of a single person, or you may succeed in transforming the intellectual scene for many others, for a whole generation, or even permanently. Waismann suggested, for instance, that Hilbert had effected a decisive and widespread change in the conception of the nature of the axioms of Euclidean geometry. But the enterprise is essentially the same, whether conducted wholesale or retail. There are no ways of seeing things that are not someone’s ways of seeing things, and no changes in ways of seeing things that are not changes in the intellectual life of a person. (3) Philosophical discussion is aimed at clarifying grammar. This is apt to suggest (to analytic philosophers) an enterprise of delineating the

148 Wittgenstein and Waismann

contours and internal structure of a supra-personal institutionalized normative system, say on the model of writing a textbook of the English law of contract. The individual speaker might be conceived as participating in a complex practice of which he may have an imperfect understanding, and a philosopher might help to bring to his attention features previously unnoticed or to remind him of things temporarily lost from sight, just as a barrister may try to direct the attention of a judge to certain features of his case. Waismann understood matters completely differently. In his view, the project of therapy is essentially one of cultivating self-awareness. The therapist tries to make the patient conscious of his own rules, of his own practice; especially of his own prejudices and of analogies or pictures that have ‘unconsciously’ guided his own thinking. The language whose grammar needs clarification is his language. He is to look at his own understanding of the words he uses, especially at his own preferred explanations of what they mean. He is, as it were, entangled in his own rules (PI §125). Whether these are shared with others or clearly deviant in comparison with ‘ordinary language’ is of no interest whatever.5 For this reason, we don’t force our interlocutor. We leave him free to choose, accept or reject any way of using his words. He may depart from ordinary usage. . . . He may even use an expression one time in this, another time in that, way. The only thing we insist upon is that he should be aware of what he is doing. (HISP 12)

In dealing with philosophical problems, the aim of the therapy is always the same. In my own case, it is to describe the grammar of my language; in another’s case, to clarify for him the grammar of his language. As in psychotherapy, the goal is improved self-knowledge; in one case, it is intended to give the patient a better understanding of his own behaviour, in the other to give him a better understanding of his own ways of thinking and speaking. (In both cases, the means for achieving this goal may vary from person to person. In particular, there is no presumption that ‘our language’ is uniform among English-speakers entangled in philosophical confusions, or even that the form of describing the grammar of our words must be invariant from person to person or from case to case.) The content of this understanding is dependent on personal acknowledgement or consent.6 The criterion for my having followed a grammatical rule that I have not previously formulated is that I acknowledge it now as the rule by which I have previously proceeded.7 (It is not simply a hypothesis which squares with my behaviour. As it were, I make the rule

Thinking about ‘Thinking’

149

mine by taking responsibility for having been guided by it. There is no such thing as a grammatical rule of my language which I am in no position to acknowledge or which holds in the teeth of my sincerely refusing to acknowledge it. In this sense, there is ‘no action at a distance’ within grammar.) In respect of acknowledging his own rules, the patient is completely authoritative and completely free. (But there is a danger of his acting in bad faith! He may exhibit self-deception in the form of suggestibility or resistance.) Waismann put great emphasis on respecting the patient’s freedom, contrasting this with producing compelling arguments. No philosophic argument ends with a QED. However forceful, it never forces. There is no bullying in philosophy, neither with the stick of logic nor with the stick of language. (HISP 29)

One of his recurrent themes is that in philosophy there are no proofs; there are no theorems; and there are no questions which can be decided, Yes or No. (HISP 1)

The philosopher must try to get his patient to make a decision, to accept a new way of seeing things with his spontaneous consent. Hence, ‘the essence of philosophy lies in its freedom’ (HISP 21). There is nothing in any philosophical remark that anybody must accept.8 (Nothing is comparable to ascertaining the legal principles that are objectively embodied in the common law of contract.) (4) The activity of a good philosopher manifests a high degree of imagination and creativity. He needs imagination in the form of sensitivity and sympathy to the trains of thinking of his patient; there are no recipes to follow mechanically in tracing the roots of confusion and untying knots of misunderstanding. This seems obvious in so far as the goal is to bring the patient to acknowledge analogies or pictures that have unconsciously shaped his own thinking. Imagination in a different form is needed to construct a persuasive analogy, to invent an object of comparison to make visible a new aspect of something, to dream up intermediate cases or new notations, or to make visible neglected possibilities. It is a task of the greatest difficulty and delicacy to get another to see something differently from the way habitual to him and deeply entrenched in his thinking.9 It requires intellectual creativity, especially skill in expressing ideas picturesquely. The philosopher, one might say, is more an inventor than a discoverer. His work must be anything but pedestrian if it is to be effective.

150 Wittgenstein and Waismann

Waismann summed up this point in the claim that the hallmark of a great philosopher is ‘vision’. What is characteristic of philosophy is the piercing of the dead crust of tradition and convention, the breaking of those fetters which bind us to inherited preconceptions, so as to attain a new and broader way of looking at things. . . . What is decisive is a new way of seeing and . . . the will to transform the whole intellectual scene. (HISP 32)

This quality is what he admired in the work of Wittgenstein, Frege, Hilbert, Schlick and other giants in the history of philosophy. ‘How I See Philosophy’ gives expression to a picture or model of philosophy which Waismann illustrated with some episodes mostly drawn from philosophy of mathematics (e.g. the confrontation between Frege and Hilbert about the nature of Euclidean geometry). What he sketched out is deliberately not offered as an account of the nature of philosophy, but rather a conception of philosophizing. That much is clear from the title as well as the content of the article. He set out to explain ‘How I see philosophy’, and the radically therapeutic method that he described and illustrated is certainly not represented as being the only possible (coherent) method of philosophizing. It seems that he wanted to leave each of us perfectly free to accept or reject these methods; there is no imprimatur or ‘QED’, rather we must reach a decision to put them into practice ourselves – or to refuse to do so.

2 Some Aspects of the Psychoanalytic Model Before I make use of Waismann’s model as an object of comparison, I need to convey a more definite idea of what it involves. I will split this task into two parts. First, I will try to put into sharper focus three aspects of the therapeutic method that he held to be of decisive importance for understanding and evaluating its ‘results’. Second, I want to clarify in what respects he saw his method to be analogous to psychoanalysis in order to gain a fuller understanding of its distinctive character. The first of these tasks is to call attention to three aspects of ‘his language’ (his speech, diction or idiom), i.e. the form of discourse in which Waismann tried to discuss particular philosophical problems. All three were mentioned in the previous summary of ‘How I See Philosophy’, but they easily pass unnoticed.

Thinking about ‘Thinking’

151

First, he associated with philosophical problems a wide range of terms signifying troubled or unhealthy states of mind: ‘torment’, ‘Angst’, ‘disquiet’, ‘craving’, ‘prejudice’, ‘superstition’, ‘illusion’, etc. In this context of describing a form of therapy, these expressions are meant to be construed literally, not metaphorically. His concern in addressing ‘a philosophical problem’ is not to solve an enigma, but to bring relief to a person who is manifestly sick and unhappy.10 Second, in linking philosophical problems with compulsions, obsessions and prejudices, he tied them to formulations that make prominent use of the modal terms ‘must’ and ‘cannot’. What appear to be necessary truths generate intolerable conflicts within the patient: ‘This is not how it is – but this is how it must be!’ (cf. PI §112). Consequently, therapy primarily takes the form of persuading the patient to acknowledge that things need not be as he had said they must be, or that things may be as he had said they could not be. Here too modal expressions play a cardinal role. (In this context, whether things are, as a matter of fact, thus-and-so or otherwise is a matter of comparative indifference. Here an unrealized possibility is equipollent with an established fact.) Waismann made extensive use of other phrases that might, in an extended sense, be called ‘modal qualifications’: ‘one might say’, ‘we could say’, ‘it would be better to say’, ‘we say’, etc. In most cases, these expressions are crucial to grasping the purpose and the content of particular remarks. (They cannot be omitted in accurate paraphrases of his remarks.11) Third, Waismann used first-person plural pronouns in a very promiscuous manner: sometimes ‘we’ is interchangeable with ‘I’, sometimes it signifies the author and his coterie (in opposition to ‘others’, ‘the rest’, or ‘they’ (French ‘on’)), sometimes it refers to the patient (as opposed to the therapist), and sometimes it means a more or less indefinite range of persons (say contemporary philosophers, English-speakers, or even human beings in general). The problems of sorting out the different uses of the personal pronouns can be difficult and delicate, yet doing so has decisive importance for construing the content of many individual remarks. (To impose a uniform reading would produce much distortion of his thinking, sometimes plain nonsense.) All of these three features are common to most of the writing that demonstrates ‘our method’, the texts of Wittgenstein as well as those of Waismann. They are seldom treated with seriousness or respect. The inclinations, or temptations, to make little or nothing of them seem to be linked to each other; these bad habits of interpretation constitute, as it were, a syndrome of insensitivity or intellectual blindness. A different approach is essential to making sense of any text that is meant to be read

152 Wittgenstein and Waismann

as the record of a discussion (real or imaginary) between a philosophertherapist and a patient undergoing ‘philosophical analysis’. In particular, this conception excludes taking the interlocutor as an opponent or adversary; it respects his ‘torment’ and ‘disquiet’, it demands scrupulous attention to the modal qualifiers that are indicative of obsessions and compulsions, and it requires maintaining the stance of suspending judgement on whether ‘we’ are the body of speakers of correct everyday English. Further light may be shed on Waismann’s method by exploring in more detail some specific analogies with psychoanalysis. I turn now to this second task. Parallels with Freudian methodology might be argued for in different ways; the various elements of psychoanalytic theory are given different expression even within Freud’s writings, and almost everything is much contested by others. What I will offer is a minimalist account which focuses on five incontrovertible respects of analogy. (1) Psychotherapy is strictly patient-oriented: its primary concern is a form of healing, the treatment of sick individual persons. Waismann emphasized this aspect of philosophy. A philosopher is to engage in discussion or dialogue with a particular person who is genuinely suffering from tormenting intellectual perplexities or confusions, and by altering how this individual sees things, he aims to effect an improvement in how that person copes with his own intellectual environment. Treatment is targeted on a person, someone who is, by his own lights, sick and unhappy.12 This conception is exclusively therapeutic. In the absence of a particular person with a particular complaint, there is literally nothing constructive for a philosopher to do.13 (Waismann leaves no logical space for investigating the geometry – or logical geography – of concepts.) A philosophical investigation must take the form of a dialogue, whether the interlocutor be real (in a class in Cambridge) or imaginary (in the text of the private language argument). (On soigne toujours quelqu’un de malade, jamais une maladie en soi. Peut-eˆtre on s’occupe de quelqu’un de fabrique´, mais jamais d’un malade imaginaire.) (2) In both cases therapy is directed to a particular kind of disease: kinds of internal conflict. The compulsions, obsessions and neuroses treated by psychoanalysts are conceived as conflicts of conative and affective states, of desires and emotions within the patient.14 The philosopher deals with the counterparts of compulsions, obsessions and neuroses in the way his ‘patients’ deal with their own concepts and ways of thinking. He tries to show them how to trace their disquieting confusions and unanswerable questions to conflicts in ways of seeing things – especially in ways of their

Thinking about ‘Thinking’

153

wanting to see things;15 to their own prejudices or dogmas clashing with each other and generating fogs which make clear vision impossible (PI §3). The focus is always on conflicts between different things that the patient feels he must say or that he deeply wants to say.16 (They are not external conflicts, in particular not discrepancies between what he says and what others say. The contrast ‘je dis’ / ‘on dit’ is irrelevant to his therapy.) The patient’s unhappiness is grounded in his experiencing irresolvable conflicts within himself. (External conflicts are irrelevant unless they are felt! They are trivial to resolve unless they arise from serious internal commitments.) (3) In both cases the method of therapy is sharply circumscribed: some form of rational discussion face-to-face with the patient.17 The only admissible treatment is a talk-cure. On the one hand, this restriction excludes any form of mechanical treatment, e.g. by administering sedatives; also any form of manipulation of the patient’s human environment about which he is kept in the dark, e.g. by arranging others’ conduct so as to avoid their bringing up certain topics of conversation in his presence. On the other hand, it excludes non-rational verbal influencing of the patient, e.g. by exploiting subliminal cues or by playing on his fears, cravings or anxieties. These aspects of the method of psychotherapy differentiate this approach to treating mental disorders from others (e.g. from psychiatry and behaviour-therapy). Waismann emphasized the crucial importance of just these features for his psychoanalytic conception of philosophy. The essence of this therapy is discussion; not constructing proofs or refutations, but ‘building up a case’. This may involve subjecting the patient to a volley of questions, running through a series of examples, or offering arresting analogies. Doing these things may surely be called ‘reasoning’ or ‘producing arguments’. The fact that such arguments are not logically compelling does nothing to detract from their rationality (HISP 31),18 Reasoning face-to-face with the patient is the only form of therapy allowed to the philosopher for bringing about a change of attitude or way of seeing things. (Hence the objection ‘Anything goes!’ would be a grotesque distortion of Waismann’s conception.) (4) In both cases the cure of the patient’s disorder consists in his achieving an understanding of himself. This has two components: bringing to consciousness things of which he was partly or wholly unconscious, and tracing things to their origins or roots. Disorders are held to disappear once full self-knowledge is attained. (This Freudian principle19 is sometimes condemned as a bit of unwarranted optimism – or intellectualism.)

154 Wittgenstein and Waismann

The psychotherapist encourages the patient to become aware of motives or feelings that he has repressed, to admit to having them and to explore their origins. According to Freud, understanding them is essentially genetic:20 primarily the patient must learn to relate them to his own sexuality, especially in its infantile stages of development. In Waismann’s view, the philosopher pursues a similar strategy, encouraging his patient to take notice of prejudices, dogmas, analogies and pictures that have unconsciously shaped his own thinking. Full selfunderstanding is attained to the extent that these are traced to their deeper roots in certain very primitive and pervasive pictures, e.g. the picture of possibility as a shadowy actuality. The patient’s becoming conscious of his own prejudices, analogies, etc., will deprive them of their power to work mischief; he will be freed from the spell that they cast on his own thinking. In expounding his conception of philosophy, Waismann emphasized both bringing to consciousness what was unconscious and tracing the origins of conceptual confusions.21 (5) The patient’s own acknowledgement is indispensable to establishing a diagnosis as correct and to effecting a cure. This is a characteristic feature of psychotherapy: only what is ultimately acknowledged by the patient is true,22 and nothing more or less than his acceptance of his own feelings or motives constitutes the removal of his disorder.23 Philosophical therapy is held to be similar in both respects. Waismann stressed the crucial role of the patient’s acknowledgement or spontaneous consent: this holds as much for his previous way of seeing things as for any new way of seeing things. Aspect-seeing, unlike simple visual perception, is voluntary.24 The rules he follows, which comprise the grammar of his language, are none other than what he acknowledges. This is precisely the point highlighted by Waismann’s emphasis on freedom as constituting the essence of philosophy. Acceptance is equally important. A patient succeeds in dissolving his own question only if he arrives at seeing something differently. This is more than the bare recognition of a possibility; it is, as it were, a transformation of what he experiences. (Just as your visual experience of a puzzle-picture is transformed when you are struck by a new aspect of it.) He must willingly drop his original question, or willingly replace it by another one. Here the notion of acknowledgement (Anerkennen) is as central to Waismann’s conception of philosophy as to Freudian psychotherapy.25 In all these five respects, ‘How I see Philosophy’ likened a particular style of philosophical investigation to the general method of psychotherapy. In

Thinking about ‘Thinking’

155

so far as Waismann put his preaching into practice, his own philosophizing should be understood to be distinctively different in content and presentation from the normal didactic information-conveying model of analytic philosophers.26 And in so far as his conception captured an important aspect of the thinking of his mentor, the same conclusion should hold in respect of Wittgenstein’s writings. In both cases, the writer is trying to engage his reader in a dialogue whose principal goal is to effect a change in ways of seeing things. For this purpose, generating ‘concern and unrest’ in the interlocutor is a crucial part of the tactics.27

3 Freudian Aspects of ‘Our Method’ Having developed Waismann’s psychoanalytic model of philosophical therapy, I want now to exploit it to reveal a pattern in the LSP-dictations. (I suspect that Waismann derived it principally from this source, but I doubt whether this historical claim can be clearly substantiated.) This pattern is the one summed up in the remark, ‘Our method is similar to psychoanalysis’ (F 93). This point has considerable philosophical interest. First, it demonstrates that there was a definite phase of Wittgenstein’s thinking in which close comparison with Freud’s methods informed his own conception of philosophical investigation. This phase extended over several years, at least from the composition of BT to the writing of PPI. (That fact makes some sense of his having described himself to Rhees as ‘a disciple of Freud’.) Second, this general orientation of ‘our method’ is important for understanding various of the conceptual investigations carried out in the LSP-dictations (as well as those conducted in BT and PPI). It clarifies the spirit that informs these many fragmentary sketches of the grammar of our language, and this may transform how these ‘grammatical descriptions’ are to be understood. (I will illustrate this point in section 5.) The LSP-dictations demonstrate ‘our method’ by means of examples: it is put to use to treat a range of philosophical problems. Here we have an opportunity to observe Wittgenstein’s distinctive modus operandi in detail, and we can work out how it should be described by examining his therapeutic procedures from case to case. The worries addressed are very diverse: the nature of negation, whether ‘if . . . then . . . ’ can be defined by a truth-table, the limits of ‘experience’ (‘Erfahrung’), the nature of an assumption (Annahme), etc. Prima facie, the goal in each case is to dissolve a particular philosophical problem, to enable somebody to adopt a point of view from which it completely disappears.

156 Wittgenstein and Waismann

In the course of these investigations, Wittgenstein made some general comments about what he was doing. These seem to be intended to help his interlocutor or reader to keep his own bearings. I will present a selection and arrangement of some of these remarks. ‘Our method’ embodies a particular conception of what it means to describe the grammar of our language.28 In each case of clarifying the use of an expression, we characterize a particular aspect of language by investigating the use of signs from the standpoint of a procedure governed by a code of rules (F 21). To follow this method, whenever we find ourselves in philosophical unclarity, we set up for ourselves a schedule of rules (F 21). We answer the question ‘What does the word ‘‘ . . . ’’ mean?’ by describing the use of the word in the form of a rule (Abmachung) or calculus (F 33). We make use of such language-games as objects of comparison; in suggesting that they may be useful in capturing particular aspects of the use of words, we refrain from making the claim that they depict reality (what words really mean) (F 21).29 ‘Our method’ presupposes that there are different general conceptions of the use of words to be contrasted with ours.30 Ours is definitely not the only conception of word-meaning.31 One is the causal theory of meaning (e.g. Russell’s late stimulus–response analysis); another is Augustine’s picture of language, which holds that every word must be the name of some object (PPI 109). These are different grammatical conceptions (Auffassungen) of signs. (It is a philosophical question to explore the merits and deficiencies of each of them (cf. PG 71).) Even within our general conception of language, there are different conceptions (Auffassungen) of the uses of particular words. We can describe our own practices in different ways, from different points of view. We can set up different sets of rules and use them as objects of comparison. What we call ‘a conception (Auffassung) in grammar’, or ‘a conception (Auffassung)’ in the grammatical sense, is always expressed in grammatical rules (F 76). Conversely, any codification of rules for the use of an expression is treated as a conception (Auffassung) of its grammar (cf. F 21). It is important to recognize, for example, that there are different conceptions of what is meant by ‘Satz’ (F 54; PLP ch. 14), ‘tautology’ (F 20, 75; cf. F 7), or ‘rule’ (F 36). Likewise, there are different conceptions of how the rule of double-negation elimination is related to an ostensive explanation of ‘not’ or even to a truth-table explanation; we see some real value in comparing acknowledgement of this rule with making somebody aware of a new aspect of some geometrical solid that has been placed in his hand (F 58). In such cases, it is evidently useful to call attention to different grammatical conceptions of

Thinking about ‘Thinking’

157

particular signs by juxtaposing different sets of rules with each other rather than trying to eliminate all but one from consideration. ‘Our method’ is concerned at two levels with aspects or conceptions. It exploits a particular conception of the meanings of linguistic expressions (‘The meaning of a word is its role in the calculus’), and it describes particular aspects of the uses of particular expressions. In both respects this procedure is essentially anti-dogmatic. Aspects and conceptions (Auffassungen) are essentially multiple. To call an explanation of how to use a word a conception of its meaning requires acknowledgement of available alternatives. This point is fundamental. It is part of the grammar of ‘aspect’ and ‘conception’ (at least in Wittgenstein’s language!). The goal of ‘our method’ is to dissolve particular philosophical problems. The problems addressed are ones that arise from our making ‘metaphysical uses of our words’. The remedy is to get the sufferer to renounce his grammatical prejudices by bringing to his consciousness neglected aspects of how he has always used his words (to remind him of his own practice). Though these statements are familiar and important, they may be understood in various ways. I shall suggest a particular interpretation of them. The problems addressed by ‘our method’ arise from grammatical dogmas (or illusions), and these can be traced to the influence of unacknowledged analogies. In treating a philosophical problem, ‘I must always point to an analogy according to which one had been thinking, but which one did not recognize as an analogy’ (BT 408–9). The patient holds that every proposition must be a complete picture of a state of affairs, that every proposition must be composite, that a state of affairs cannot be named, etc. His thinking is grounded in certain analogies, e.g. the analogy between a portrait and a description of the sitter, the analogy between a state of affairs and a spatial complex. The ‘musts’ and ‘cannots’ that are characteristic of grammatical dogmas pick out what Wittgenstein called ‘the metaphysical use of our words’, and they illustrate what Wittgenstein meant by claiming that pictures hold us captive or that similes (or analogies) are embedded in our language. There is nothing wrong with these analogies in themselves; they are not to be repudiated or discarded. Rather, the ‘patient’ needs to learn to recognize these analogies as analogies, hence to resist the temptation to suppose that they reveal the essences of things. When properly used, they exhibit important aspects of things; for example, there are similarities between propositions and pictures which are worth following up. But fixation on one analogy tends to produce blindness to other important aspects of things; for instance, to the relation of exclusion between the propositions ‘X is red’ and ‘X is

158 Wittgenstein and Waismann

green’ which is highlighted by the analogy of a proposition with a yardstick. The major purpose of ‘our method’ is to break the tyranny of certain pictures or analogies that are entrenched in the thinking of someone tormented by a philosophical problem. That explains why the elimination of the problem turns on eliciting from the patient the acknowledgement that a particular analogy has indeed dominated his thinking (without his having acknowledged it as an analogy) (BT 410; PPI §106). This is effected by bringing him to acknowledge that other analogies are equally justified (PPI §99). That amounts to a confession of his having been misled by the original analogy of which he made use unconsciously, i.e. without consciousness of its being an analogy. (In this sense, it could be called a false or misleading analogy.) It is unacknowledged analogies which are dangerous (DS 28),32 while making them explicit and conscious removes their destructive power (F 57).33 The dissolution of the problem turns on surrendering some prejudice (Vorurteil) (PPI 95) or in effecting a conversion to a new conception of things (die Umstellung der Auffassung) (PPI §116; cf. Z §447). This is the background that makes sense of Wittgenstein’s tactics in conducting his various philosophical investigations in the LSP-dictations. It explains why carefully constructed comparisons have the power to free us from misleading analogies (F 81) and why artificial language-games may be useful for eliminating grammatical prejudices (F 89). If we think of understanding a word on the model of being acquainted with an object which it designates, we should explore the comparison between understanding a word and knowing how to play chess (F 81).34 We might regard a sample length (e.g. the standard metre) and the gesture of pointing as symbols (or parts of language) and consider an ostensive definition as a substitution-rule licensing the replacement of one symbol by others. Then, from this standpoint, an ostensive definition remains within language (F 84), and language speaks for itself (PG 40, 63).35 Accepting new analogies has the power to set aside problem-generating disquiet (F 90). In particular, a model (or u¨bersichtliche Darstellung) may make visible a pattern or system in our word-use where none was apparent (F 35), and in this way it may remove grammatical anxiety (Angst) (F 2). Wittgenstein made extensive use of similes, analogies, comparisons and pictures to expose grammatical illusions, to neutralize the strong attractions of making metaphysical uses of our words.36 This treatment of a problem depends on the patient’s acceptance of what the therapist proposes (F 35; BT 410) and on the patient’s continuing to acknowledge these previously unconscious analogies as pictures (F 66; PPI §107).37 Clarifying the goal of ‘our method’ has the important corollary that the procedure is essentially person-specific. To dissolve someone’s

Thinking about ‘Thinking’

159

philosophical problem turns on evoking an acknowledgement from this particular person. ‘Our method’ focuses attention on aspects of language to which the ‘patient’ is blind, on analogies which he has used unconsciously, on anxieties which he feels, etc. Moreover, it concentrates exclusively on internal conflicts, i.e. with the rules I myself establish, acknowledge and follow in my own language, independently of whether these are in fact shared with others. For example, in pointing to an apple and giving the explanation ‘This colour is red’, I introduce a sample of ‘red’ which I can now use for framing translation-rules for my language (F 33, 36). (It is important that others could participate in this practice, but not that they do.) The therapist reminds me of how I have used certain words (PPI §104) and elicits my acknowledgement of his suggestions; but his description of the grammar of my language is correct only if I acknowledge it (BT 410). The essentially person-specific character of ‘our method’ is a reflection of the fundamental role given to acknowledgement. It is diagnostic of Wittgenstein’s conception of grammar that acknowledgement (anerkennen, bekennen) is the sole criterion for determining the grammar of my language;38 Sich-zu-einer-Regel-bekennen determines the rule by which I have proceeded, at least in cases where I have not previously made it explicit (F 34, 57). There is no independent ‘fact of the matter’ to be ascertained, just as the distinction between what is essential and what inessential is not something to establish (konstatieren), but something for each individual to decide (bestimmen) (F 39). ‘Man kann die Grammatik einer Sprache nur im Einversta¨ndnis mit dem Sprechenden feststellen . . . ’ (F 36); this needs to be settled with each individual speaker. (Compare the case of my searching for the right words to formulate a thought: you make a suggestion, and my accepting it as correct is the criterion for that’s having been precisely what I did think (F 88).) It is important that acknowledgement be contrasted with discovery. Two respects are crucial. First, it is voluntary or free. I cannot be forced to acknowledge a rule against my will (or to modify a calculus) as I can be compelled to revise a catalogue of trees by the discovery of a new species (F 2). Second, what is acknowledged cannot be something new or surprising; it has the aura of something humdrum and familiar. What I acknowledge, e.g. the double-negation rule, seems like something I have always known and hence could have ascertained by myself (F 57–8). For this reason, one might say that philosophical reflection is always a case of recollection (Ru¨ckerinnern) (PPI §104). If we bear in mind the contrast between acknowledgement and discovery, it is literally true that in practising ‘our method’ nothing is discovered! (F 16). Everything is familiar and open to view; nothing is changed or everything remains the

160 Wittgenstein and Waismann

same, in one sense (PI §124), just as it does when a new aspect of a picture dawns on us! (PI p. xi). At the same time, an acknowledgement may transform the impression that everything makes on us, and in this sense everything may be changed completely.39 Philosophical discussion doesn’t add to our stock of information; but it transforms the impression (Eindruck) (PPI §104) made on us by what is familiar. This is the whole point of each individual remark. Isolated from a specific dialogue, it will lack any purpose or point – perhaps any significance. This may be what Wittgenstein tried to convey in this simile: ‘The solution of philosophical problems can be compared with a gift in a fairy tale: in the magic castle it appears enchanted and if you look at it outside in daylight it is nothing but an ordinary bit of iron . . . ’ (CV 11).40 Two features of the LSP-dictations reinforce this impression of affinity between Wittgenstein’s therapeutic philosophy and psychoanalysis. The first is the presence of material bearing the hallmark ‘Freud’, most obviously the use of his technical jargon. Here is a sample: 1 2 3 4 5

Wittgenstein drew attention to the power of unconscious analogies and pictures to work harm (F 24, 57, 71, 93; DS 28). He noted the possibility of resistance and the need to overcome it (F 87, 9). ‘Widersta¨nde des Willens sind zu u¨berwinden’ (BT 406). He stressed the pivotal importance of acknowledgement (F 16, 17, 34, 35, (39), 57, (63), 88). He mentioned both censorship and sublimation in respect of pictures (F 24). Also the repression of doubts (PG 382). He referred to various forms of pathological states (obsession, compulsion and mental cramp) (F 46, 47, 48, 54, 81, 89, 92). For instance, seeking kramphaft (compulsively) for what is common to all propositions (Sa¨tze) (PPI §109); or revulsion from certain ideas as being infantile (PG 382).

In addition to using this terminology, Wittgenstein stressed two general themes that are prominent in Freud’s thought. First, he practised a genetic method by tracing problems to their sources in pictures, analogies and similes which are embedded in our language and our thinking;41 he tried to show that many philosophical prejudices have deep roots, and he suggested that they can be eliminated only by clarifying their sources42 (F 4, 9, 13). Of course, the nature of the origins of the problems differs in the two cases: in psychoanalysis problems are traced to experiences or events, in philosophy to analogies or pictures. Second, Wittgenstein aimed at removing puzzlement by promoting self-awareness (rather than the discovery of something new); he sought to secure confessions or

Thinking about ‘Thinking’

161

acknowledgements of prejudice, bias, compulsion or superstition and to make visible patterns in or aspects of what is already familiar to his interlocutor (F 2, 35, 57–8, 75, 76).43 Of course, the items exemplifying the patterns are different in the two cases: in psychoanalysis we are concerned with patterns of behaviour (e.g. manifestations of an Oedipus complex), in philosophy with patterns in the uses of our words (e.g. connections of mental states with their expressions). The second confirmation of Freud’s influence is the content of the document entitled ‘Method’ (G 26). The procedure there advocated is essentially therapeutic, hence nonadversarial. Any participation in perennial philosophical debates is to be scrupulously avoided. The point is to abstain from entering into these controversies (e.g. between realism and idealism or between formalism and Platonism), but at the same time to transform the whole way of seeing them by revealing new aspects in the use of ‘our’ words. Wir ko¨nnten aber auch einen ganz anderen Standpunkt einnehmen. Wir ko¨nnten sagen: Lassen wir die Frage ganz beiseite, ob die Arithmetik ein Spiel ist oder nicht! Eines ist klar: Eine gewisse Verwandtschaft muss hier bestehen, denn sonst wa¨re wohl niemand auf diese Idee verfallen. Untersuchen wir daher einmal das Spiel! Setzen wir die Untersuchung des Spieles neben die Untersuchung der Arithmetik und lassen wir das eine ein Licht auf das andere werfen! Seien wir ganz gerecht, behaupten wir nichts, sondern lassen wir die Dinge fu¨r sich selber reden! . . . Wie der Jurist bestimmte Streitfa¨lle als Paradigmen behandelt, gleichsam als ideale Fa¨lle, so konstruieren auch wir ideale Fa¨lle, grammatische Bilder, um im Falle eines philosophischen Streitfalles Aspekte zu gewinnen, den Konflikt zu entscheiden. Wir wollen die Sprache nur unter dem Gesichtspunkt des Vorgehens nach einem Regelverzeichnis betrachten, unter einem solchen Aspekt. (G 26)

Nothing is to be proved (F 90). Not even anything about the grammar of our language! (If acknowledgement is the criterion for my having always followed a rule, there is literally nothing to be proved. Unless I acknowledge this as the rule I have been following, it is not the rule I have followed (BT 410), and no argument has the power to compel any such acknowledgement.) Wir brauchen daher keinerlei Behauptungen zu machen und etwa zu sagen: So muss das Wort gebraucht werden: Das ist doch eigentlich der Sinn. In dem Wo¨rtchen ‘eigentlich’ steckt eine grosse Gefahr, man ist dann immer versucht, die Wirklichkeit nach einem Schema wenn nicht a¨ndern zu wollen, so doch auf dieses Schema aufzuspannen. Wir brauchen nichts

162 Wittgenstein and Waismann dergleichen, wir stellen das Schema neben die Wirklichkeit und lassen es soviel Licht darauf werfen, als es wirft. (G 26)

This procedure of employing language-games as objects of comparison is characterized as a non-dogmatic method of philosophizing. It clearly leaves the interlocutor completely free to make what he will of the comparison, and it gives no licence whatever for making any metaphysical uses of our words. Das ist der Gesichtspunkt, unter dem wir die Sprache betrachten wollen. Wir wollen nicht dogmatisieren, sondern wir lassen die Sprache, wie sie ist, und stellen ihr ein grammatisches Bild an die Seite, dessen Eigenschaften wir vo¨llig in unserer Gewalt haben. Wir konstruieren gleichsam einen idealen Fall, aber ohne die Pra¨tension, dass er mit etwas u¨bereinstimmt; sondern wir konstruieren ihn nur, um ein u¨bersichtlich Schema zu gewinnen, mit dem wir die Sprache vergleichen; gleichsam als einen Aspekt, der noch nichts behauptet, also auch nicht falsch ist. Man ko¨nnte also sagen: Wir stellen Regel-Verzeichnisse auf, die mit der wirklichen Sprache gleichsam stu¨ckweise parallel laufen und die dazu dienen, Schwierigkeiten zu beseitigen, die durch Aufstellen von Regeln zu beseitigen sind. (G 26)

Wittgenstein warned of a real and present danger in putting ‘our method’ into practice. ‘Our philosophy’ can very easily fall into dogmatism (PPI §107) if we lose sight of the analogical status of codifications of grammatical rules or language-games. Each description of grammar incorporates a simile (PPI §102). The temptation is to make deductions from these descriptions, or even to take descriptions of language-games to be ideals to which reality must conform (PPI §107; cf. PI §131). We can avoid this danger only by dint of keeping alive the awareness that language-games are objects of comparison whose sole role is to make a confused person conscious of neglected aspects of his own practice. In what respect is this method comparable to psychoanalysis? Wittgenstein’s answer seems clear enough. We make use of carefully selected comparisons in order to bring to the interlocutor’s consciousness certain unacknowledged or unconscious analogies that appear to generate his perplexity, confusion and sense of uneasiness (Unwohlseins) (BT 408, 410; cf. PPI §98). We exhibit the force (Macht) which is apt to fixate us on the questions ‘What is the essence of time? What is the essence of a rule’ etc. (F 94). The juxtaposition of cases has the potential to break the tyranny of pictures which have held us captive (cf. PI §115); accepting other possibilities amounts to refusing to countenance our own previous

Thinking about ‘Thinking’

163

metaphysical uses of our words. (It is clearly self-contradictory to say ‘It must be thus, but it may be otherwise.’) The non-adversarial character of this method does not amount to a temperamental preference for philosophical ‘quietism’. Rather, it is directly motivated by Wittgenstein’s conception of philosophical problems and his therapeutic aspirations. There is literally nothing to ascertain, discover or demonstrate in the grammar of our language. Hence philosophical discussion, according to his conception, must abjure any pretension to being demonstration more geometrico; instead it takes on the aspect of persuasion or propaganda for a point of view (Standpunkt), an attempt at conversion (Umstellung). It is addressed more to the will than the intellect. ‘Resistances of the will are to be overcome’ (BT 406). What makes a subject difficult to understand may be how we want to see it (BT 406 ¼ CV 17).44

4

Some Psychoanalytic Aspects of Wittgenstein’s Dissolutions of Philosophical Problems

A therapeutic conception of philosophical discussion bears on what is to be understood by the phrases ‘describing grammar’ and ‘dissolving problems’. These philosophical activities take on a novel physiognomy in the context of ‘our method’. Analytic philosophers are tempted to think of describing grammar as an activity which is essentially impersonal and context-free, something concerned with the more or less determinate geometry of an institutionalized natural language. The paradigm of dissolving a philosophical problem is then a demonstration that the question cannot be framed without transgressing the bounds of sense or running up against the limits of language.45 On the psychoanalytic model, however, the treatment of problems is essentially patient-specific. It is concerned more with his practice than a public institution, more with his attitudes towards speaking a shared language and less with the facts of this practice,46 more with his motives for raising a question than with the abstract possibility of doing so. Consequently, his question will disappear completely from his life only when he no longer wants to pose it.47 I want to illustrate this unfamiliar practice with two familiar examples of therapy taken from Wittgenstein’s work. Waismann’s psychoanalytic model of philosophy can serve to highlight a neglected aspect of Wittgenstein’s actual treatment of philosophical problems. Once we become aware of this pattern in his reasoning, it will be seen to be of decisive importance for capturing the spirit of many of his descriptions of the grammar of our language. Here ‘we fail to be struck by what, once seen, is most striking

164 Wittgenstein and Waismann

and most powerful’ (PI §129). Wittgenstein’s individual remarks are fully intelligible only when understood as components of a dialogue, an interaction with a person to be understood as endowed with preferences, prejudices, pictures, cravings, anxieties, etc. Hence what is to be counted as descriptions of the grammar of his language is a matter for specific investigation, negotiation and acknowledgement. This cannot in principle be ascertained independently of a discussion of each specific issue with him. Wittgenstein’s treatment of Augustine’s problem, namely ‘How is it possible to measure an interval of time?’, is undoubtedly the most celebrated succinct paradigm of his philosophical therapy.48 Its aim is evidently to make the problem disappear completely. His treatment of this case is meant to clarify exactly how conceptual confusion can be traced back to the influence of an unacknowledged analogy. Augustine argued himself into a difficulty by noting that the past has ceased to exist and the future has yet to come into existence; this seemed to him to carry the implication that it is impossible to compare any two successive intervals of time. Dissolving his perplexity would involve getting him to acknowledge that his own reasoning turns on a particular comparison between measuring time and measuring distance. His problem evidently arose from his exploiting the idea that measuring must always consist in juxtaposing two things on the model of measuring the length of a log with a metre-stick. From inspection of his reasoning, we seem to see as clearly as we can that he made use of this analogy. But, if he did so, he employed it without having formulated it explicitly and without having considered its justification. Untying the knots of his misunderstanding is a delicate task, and any assistance that we could give would depend on his willing cooperation. If we were to get him to reflect on the procedure of measuring the length of a field by pacing it off (or by making a calculation from the readings of a theodolite), he might then see that measuring time-intervals is not peculiarly problematic; this case might then lose its disturbing aura of uniqueness. On the other hand, if Augustine himself refused to acknowledge having exploited the analogy of juxtaposition, nothing whatever would have been achieved by the (imaginary) discussion with him. Our diagnosis of his problem would be unsustainable.49 Likewise, nothing would be achieved if he were to refuse to admit, on reflection, that the analogy is unjustifiable (falsch),50 and he could rationally refuse to do this. He could instead formulate and henceforth adhere to the convention that measurement must consist of juxtaposing a measuring instrument with what is to be measured,51 and then he could treat other cases of so-called ‘nonjuxtapositional measurement’ as generating further conceptual perplexity

Thinking about ‘Thinking’

165

parallel to the initial problem. To his original puzzle might now be added: ‘How is it possible to measure the length of a field by pacing it off ?’ Whether or not the treatment dissolved Augustine’s problem would depend entirely on how he reacted to it. (And the same point holds of anybody who talks himself into the same problem by rehearsing Augustine’s argument.) Whether an analogy is a decisive influence on someone’s train of thinking cannot be settled independently of taking his acknowledgement or disavowal into account. This makes this instance of philosophical therapy essentially person-relative. Moreover, dissolving the patient’s problem must respect his freedom; it makes no sense to foist the diagnosis on him against his will.52 The analogy allegedly ‘laid up in Augustine’s language’ was really operative in his thinking only if he could be brought to acknowledge it. Moreover, its deleterious influence would be neutralized only if he further acknowledged that it is no better justified as a model of measurement than the procedure of pacing off a field. This dissolution of Augustine’s problem proceeds quite independently of laying out the logical geography (or geometry) of fifth-century North African Latin. Wittgenstein’s dissolution of the logical problem of colour-exclusion has a similar physiognomy which is particularly clear in one of the LSPdictations (F 2). It is here taken for granted that a thing cannot be red and green at the same time; this fact about the grammar of colour-words is the point of departure for the discussion. There is no attempt to explain why this fact obtains; in particular, it is not grounded in any analysis of colourconcepts into simpler elements. The express purpose of the discussion is to allay a philosophical worry or disquiet about the mutual exclusion of red and green. Wittgenstein imagined that ‘we’ (himself formerly and his interlocutors with a logical bent) are inclined to think that the possibility of having a surveyable and definitely complete system of logic depends on the possibility of so analysing colour-concepts that all colour-exclusion is revealed as tautologous. Indeed, a parallel analysis must hold for a wide range of sets of determinates ranged under single determinables, e.g. for lengths of rods, ages of persons, days of the week. Otherwise, it seems, language would be so complex that it could conform to no single general system of logic; different kinds of propositions would have their own local geometries.53 Moreover, logic would be at the mercy of ‘experience’ because novel ‘phenomena’ might turn up at any moment which would exhibit novel, hitherto unsuspected logical relations.54 Wittgenstein’s express purpose here is to address these worries. He tried to alter ‘our’ attitude towards the mutual exclusion of red and green by exploiting an analogy with measuring a rod against a metre-stick. The analogy is meant to persuade us that there is no threat in treating

166 Wittgenstein and Waismann

colour-exclusion as something primitive, i.e. as something not in need of any explanation. This conception should not give rise to Angst about anarchy or disorder (Regellosigkeit) in the grammar of our language. It does not ‘hand the reins over to experience’ or ‘leave us at the mercy of future discoveries’, at least not once we appreciate that whether or not something counts as a rule of our grammar depends wholly on our own acknowledgement, which we are completely at liberty to withhold.55 Unless ‘our’ logical Angst about colour-exclusion is removed by this discussion, nothing whatever is achieved by it. We are free to decide that Wittgenstein’s analogy with the metre-stick accomplishes nothing. In particular, we could react by treating the mutual exclusion of different lengths as itself something just as urgently in need of explanation by appeal to ‘analysis’ as the mutual exclusion of colours. (That had been the attitude of the author of the Tractatus!) In this case, the whole discussion based on the analogy would lose its point altogether. Many other discussions are directed to removing someone’s prejudices or bringing about a transformation of his attitude towards how he describes the use of his own symbols. This seems to me to be the overarching goal of his calling samples ‘parts or instruments of our language’, of his criticizing Augustine’s picture of the essence of language, and even of the private language argument (with its distinctive slogans ‘ ‘‘Inner states’’ stand in need of outward criteria’ and ‘Avowals such as ‘‘I have toothache’’ are expressions of sensations, learned replacements for primitive pain-behaviour’).56 In all these cases, the ‘conclusions’ of his investigations are altogether misunderstood unless they are taken in the right spirit.

5 A Grammatical Prejudice about Thinking I want to conclude by demonstrating this general point in one instance, namely in Wittgenstein’s advocacy and defence of the slogan ‘Thinking is operating with signs’ in the LSP-dictations. It is important to realize that this is not an isolated aperc¸u. On the contrary, it is a main theme, variations on which run through much of Wittgenstein’s work in the early 1930s. It is closely linked with some other slogans about thought and thinking. One is that thought is not something hidden behind expressions of thought, but rather that a thought is in its expression (BT 81).57 (Correspondingly, he argued, thinking is not something occult going on behind expressing thoughts; rather, the experience of thinking may be the experience of saying something (BB 43).) A second mnemonic slogan is that, ‘for our purposes’, a thought can always be replaced by its expression (and equally, a mental image by a

Thinking about ‘Thinking’

167

drawing (BB 4), memory of an explanation by a slip of paper on which it is written (PG 49), etc.).58 It is equally important to realize that the slogan ‘Thinking is operating with signs’ is intimately associated with a wide range of remarks about many other ‘mental processes’. Indeed, this formula is a kind of schema. We can regard various other grammatical remarks as substitution instances of it. For example, ‘belief is calculating with signs, framing a linguistic expression, albeit in quite particular circumstances and with quite particular indications of its origins’ (F 5); ‘Expecting is operating in a particular way with words (or images)’ (F 46); and ‘ ‘‘Understanding’’ is . . . the name . . . of more or less interrelated processes against a background, or in the context, of . . . the actual use of a learnt language or languages’ (PG 74 ¼ WiMS 45). (These are blank cheques on explanations of these terms, since they must be filled out by specifying in detail what these ‘quite particular circumstances’ and ‘particular ways’ are.) There are various instances of a fuller exploitation of this schema: ‘Having a strong convic¨ berzeugung) is expressing thoughts with a characteristic emphasis tion (U or intonation (Sta¨rke oder Tonfall der Rede)’ (DS 21; F 5); ‘Making an assumption (Annahme) is expressing a thought in preparation for drawing inferences from it’ (F 72); and ‘Making an inference is transforming one or more propositions into some other proposition (according to some rule)’ (RFM I. §§9, 19). These seem to stand or fall together as ‘descriptions of the grammar of our language’. To make good sense of any of these ideas seems to be to make good sense of them all. Consequently, a lot seems to be at stake in arriving at a good understanding of the underlying idea. What we require to make sense of his slogan ‘Thinking is operating with signs’ is our placing this remark in the context of a philosophical investigation which is conceived on the patient-oriented psychoanalytic model that Waismann took to be ‘our method’. Critics mistake its whole point and purpose in criticizing it as a misguided description of the logical grammar of ‘think’ along behaviouristic lines.59 They rightly point out that it is neither necessary nor sufficient for somebody’s thinking that he produce or calculate with signs. Wittgenstein was perfectly aware of this feature of ‘our use’ of the word ‘think’ even in the early 1930s. In his view, however, it is irrelevant to the question whether ‘operating (or calculating) with signs’ gives an u¨bersichtliche Darstellung of such ‘mental processes’ as thinking, believing, inferring, expecting, etc. How can we grasp his point of view? He employed his slogan to exhibit an unfamiliar form of describing the use of all these expressions,60 i.e. to accomplish something antecedent to giving any detailed ‘analysis’ or ‘description’ of the use of any of them. He

168 Wittgenstein and Waismann

wanted to liberalize our conception of the grammar of ‘think’: he hoped to gain acceptance here for a range of humdrum observations about our applications of ‘think’ which we are strongly inclined to put aside as irrelevant without giving them any serious consideration. In this respect, he made propaganda for a particular conception (Auffassung) of the grammar of terms commonly taken to signify ‘mental processes’. He presented a case in favour of adopting a particular set of criteria for what is to be counted as describing the grammar of these terms. This is the point of his saying: ‘For us thinking is operating with words or other signs’ (F 15). He urged all of us to start to investigate our own use of ‘think’ by describing what uses of symbols by another person we would take to justify saying of that person ‘He has the thought that . . . ’. The problem he faced is not simple blindness, but ingrained prejudice against this procedure. Before embarking on any detailed clarification of the meaning of ‘thinking’, he felt he had to overcome the resistance that his interlocutors (or readers) would feel to proceeding down the road he intended them to follow. He had to engage with their refusal to do what he thought would allow them to escape from their own internal conflicts. One dictation (F 87) is dedicated to making clear this point of view. There he urged as ‘our Auffassung’ the identification of thought with the expression of thought, contrasting this with ‘the traditional Auffassung’ (‘ours’ too, in a different sense!) according to which thought is the inner accompaniment of speech. For our purposes, every so-called inner process can be replaced by an external one, e.g. a sequence of mental images by a series of painted pictures. ‘Our whole Auffassung might be summed up in the words: for us there is nothing essentially private about thought.’ The traditional Auffassung is not justified by the observation that I can utter a sentence without having any thought or that I can, in some cases, think without constructing a sentence. If we drop the search for a single common process lying behind every instance of expressing a thought, we will lose a certain preconception (Voreingenommenheit), the fetters will fall away from our inquiry, and we will no longer have the impulse to struggle against the idea that what we call thinking is simply calculating with language (Sprache). Summing up the reasoning of this one dictation (F 87), we might say: ‘thinking is operating with signs’ is the prolegomenon to a philosophical investigation, not its upshot. This point of view is made clear in various different ways in other texts. Two dictations are devoted to vindicating this slogan against obvious objections. One (F 15) canvasses two objections: the possibility of lying, and the experience of seeking for the right words to express a thought. (To the first, Wittgenstein replied that lying is similar to playing two

Thinking about ‘Thinking’

169

different games simultaneously, one with spoken, the other with unspoken words. To the second, he replied that for us the anticipation of a thought (die Ahnung des Gedankens) is not the thought itself; the two are as different as having a hunch about how to prove a theorem and actually carrying out the proof.) The other dictation (DS: F 5–6) defends the related slogan ‘Belief is calculating with signs’. The substitution of the expression of a belief for the belief itself, the tone of voice of uttering this expression for the intensity of the belief, provides us with a synopsis (konzise Auszug) or u¨bersichtliche Darstellung of the grammar of the word ‘belief ’. (To the objection that the expression of conviction can be dissimulated, but not conviction itself, Wittgenstein noted that someone’s tone of voice is not always within his control, and that, though conviction may be expressed in many ways in behaviour, it is above all expressed in tone of voice; ‘and for this reason we said that tone of voice provides us at least with a synopsis of the grammar of the word ‘‘belief ’’.’) Evidently, he so understood his slogan that it is not vulnerable to objections of this familiar kind, and that requires distinguishing this u¨bersichtliche Darstellung from the standard form of ‘analysis’ of the word ‘think’ (or ‘believe’). These dictations show that Wittgenstein was fully aware of the inclination among philosophers to dispute the correctness of ‘identifying’ thinking with operating with signs. He made clear why he did not give way to these objections: having asserted nothing at all (even about how ‘think’ is used), he had nothing to surrender or withdraw. As an Auffassung of thinking, his slogan (like his strategy) is logically immune to refutation. At the same time, other dictations clearly imply that ‘operating with signs’ is not meant to be an analysis of the particular ‘inner process’ called ‘thinking’. On the contrary, this phrase is used to describe many different ‘mental processes’, e.g. making an assumption, wishing, expecting, inferring, etc. We would be acting perversely if we credited Wittgenstein with the intention of seeking to obscure the differences among these distinct ‘mental processes’. Evidently he sought to explore them by exploiting a general pattern of word-explanation, namely by investigating the uses of symbols which we regard as licensing applications of these words to others. We need to investigate for ourselves the quite particular circumstances of linguistic interaction that we would call ‘hoping that Keynes will come to tea’ or ‘fearing that King’s College is on fire’. Wittgenstein had the hunch or conviction that we could, by carrying out this task, dissolve many of our own philosophical puzzles about the mind. The obstacle that he addressed in offering us therapy is not our ignorance, but our prejudice, i.e. our opposition to his programme of clarifying concepts.61 In respect of our own linguistic practice, just as in our own

170 Wittgenstein and Waismann

day-to-day behaviour, everything is familiar and open to view; nothing is hidden, and no surprising facts await future discovery. At the same time, we may fail to see patterns or to make connections, even though, once seen, these aspects of our own practice may be striking and important. As long as we refuse to look, we will remain blind. We need to be persuaded to do something that we do not want to do. So, Wittgenstein urged us to describe what is apparent and familiar to us about our own use of ‘think’, rather than turning our back on this task because we have the prejudice that it is irrelevant.62 But he foresaw resistance to be overcome, as well as difficulties arising from the interference of entrenched habits of thinking. It is for this purpose that Wittgenstein recommended the strategy encapsulated in the slogan ‘Thinking is operating with signs’. This is a recommendation to break with the conception of thinking as a single mental process that lies behind and accompanies the expression of thought. It is dangerous to consider thinking as ‘a single process’; better to regard it as a whole family of processes of doing things with words! Wittgenstein laboured to remove the bias which prevents us from recognizing that ‘the experience of thinking may be just the experience of saying’ (BB 43). Thinking is much closer to expressing thoughts than we are inclined to think (LPE 232).63 We labour under the false analogy of ‘accompaniment’.64 (Much later he urged the similar strategy of treating as the primitive instance of having a thought the case of someone’s articulating his ideas in a conversation or discussion. This is not the paradigm of thinking, but rather a centre of variation for describing a field of varying examples (cf. LPP 25, 142).) Other dictations contain parallel remarks about a range of terms commonly taken to signify ‘mental processes’. In various degrees of detail, Wittgenstein discussed the strategy for engaging in describing the use of such terms as ‘supposition’, ‘conviction’, ‘dissimulating’, ‘believe’, ‘hope’, ‘fear’, ‘wish’, ‘desire’ and ‘understand’. Much of this material was set aside in the process of constructing the text of Logik, Sprache, Philosophie. Waismann apparently took it to be peripheral to the project of expounding Wittgenstein’s conception of logic and language. Nonetheless, many of the dictations grouped in the chapter ‘Psychological Concepts’ contain variations on the theme ‘Thinking is operating with signs’. The LSP-dictations embody a sustained campaign to effect a reorientation of thinking on the past of the numerous individuals whose reflections are held in bondage by ‘the traditional Auffassung’ that all species of thinking are particular, distinctive mental processes.65 This campaign is a paradigm of Wittgenstein’s method of philosophical therapy, and in its inspiration it is continuous with the later ‘private language argument’.

Thinking about ‘Thinking’

171

The slogan ‘Thinking is operating with signs’ expresses his distinctive conception of what it is to describe the grammar of our language of psychological concepts. The unanimously unfavourable reception given to this key remark is symptomatic of a total failure to understand its spirit – and perhaps the spirit of the whole of his ‘later philosophy’. He engaged in a campaign of trying to shock, badger and cajole others into bringing to consciousness their own pictures or prejudices, especially ones that produce dogmatic or metaphysical uses of metalogical concepts. He tried to bring home to those tormented by philosophical problems their own responsibility for their confusion as well as their freedom to dissolve their own internal conflicts. His therapy was directed at the whole person, and its goal was a contribution to the welfare of individuals. His philosophical investigations have been given a different mien, especially by his admirers. They are treated as describing the logical geography of natural languages and as targeted on exposing the transgressions of the bounds of sense by major philosophers, both the quick and the dead. His readers may find this ‘traditional’ interpretation of his method far less threatening – but it is surely far less challenging and exciting than the activity of philosophizing that he advocated and practised over many years.

Notes 1

Compare F 46: we seem to confront an intractable problem in explaining what expectation is. ‘Ich meine, der Fehler liegt darin, dass man hier von einem ‘‘Vorgang’’ spricht.’ This might be the same diagnosis of the problem. But equally it might not be. That depends where the emphasis is put. PI §309 puts the stress on the term ‘Vorgang’. But F 46 puts the stress on ‘ein (Vorgang)’, as the continuation makes clear. (Compare WiMS 45: ‘ ‘‘understanding’’ is not the name of a single process accompanying reading or hearing’ (the stress is dropped in PG 74).) 2 Bold indicates added emphases, italics the author’s own emphasis. 3 I cite some random passages from Freud that seem to exhibit striking analogies with ideas found in Waismann and Wittgenstein. No analysis of Freud’s thinking is attempted here, and I make no historical claim about the influence of any of Freud’s texts on either philosopher. My quotations are meant to serve as further objects of comparison. 4 In this way, he made explicit a connection between Wittgenstein’s investigation of aspect-seeing and his method of describing grammar. This might bring out an unsuspected dimension of depth in this remark: ‘The aspects of things that are most important for us are hidden because of their simplicity and familiarity’ (PI §129).

172 Wittgenstein and Waismann 5

This is a theme in Wittgenstein’s work – one commonly overlooked when his philosophical investigations are characterized as descriptions of the everyday use (grammar) of ‘our language’: ‘What is it that is repulsive in the idea that we study the use of a word, point to mistakes in the description of this use and so on? First and foremost one asks oneself: How could that be so important to us? It depends on whether what one calls a ‘wrong description’ is a description that does not accord with established usage – or one which does not accord with the practice of the person giving the description. Only in the second case does a philosophical conflict arise’ (RPP I §548). ‘Thinking is . . . an accompaniment of talking. . . . We want to draw attention to the nature of the mistake involved in this. . . . there is a tendency to say that the trouble was verbal. But how can it be verbal? ‘‘Thinking’’ and ‘‘pain’’ can be confused. But in that case philosophical conflict does not arise. There is philosophical conflict only when the person has some right idea of the use. But what is the right idea of the use? Really the person has a use and practises it. The problem here has nothing to do with conflict between different persons; it has to do with conflict within the person himself. Though the person has the right use he has a wrong idea of the use’ (LPP 125). 6 Indeed, we might say that it is wholly dependent on acknowledgement provided we take the issue to be a choice between different formulations of rules of grammar which fit the patient’s speech-patterns equally well. (It would be quite wrong to think that there is nothing to choose in such a case. For example, only Frege’s acknowledgement would settle whether he considered ‘2 þ 3 ¼ 4’ to be a false mathematical statement or not to be a mathematical statement at all, but very different conceptions of mathematics seem to grow out of the choice of one or other of these options.) There seems to be a parallel distinction in the role of acknowledgement in psychoanalysis. The patient has no incontrovertible authority to settle whether particular events occurred in his childhood, but his own acknowledgement is the sole criterion for how he felt about the episodes in his life. (Only within a limited realm is acknowledgement sovereign.) 7 This may be a matter of my now recollecting having previously formulated this rule, but this is not necessary. (It is no more essential to having followed this rule than having said something to myself at the time of uttering ‘Napoleon’ is necessary for my having then meant by this name ‘the victor of Austerlitz’.) Compare: ‘Quite often we do not succeed in bringing the patient to recollect what has been repressed. Instead of that, if the analysis is carried out correctly, we produce in him an assured conviction of the truth of the construction’ (Freud, ‘Constructions in Analysis’, Complete Works, ed. J. Strachey, vol. 23 (London: Hogarth Press, 1957), pp. 265–6). 8 At least, at the price of demonstrating intellectual incapacity (in respect of assessing evidence and reasoning). Cora Diamond (The Realistic Spirit, ch. 11) rightly calls attention to the idea that deafness to forms of argument

Thinking about ‘Thinking’

9 10

11

12

13

14

15 16

17 18 19

20

173

may exhibit other kinds of shortcoming or vice (e.g. lack of imagination or insensibility to morally important feelings). Even without considering the possibility of resistance. That adds further complexity. ‘The primary motive force in the therapy is the patient’s suffering and the wish to be cured that arises from it’ (Freud, ‘On Beginning the Treatment’, Complete Works, vol. 12, p. 143). Such omission is very likely to occur in notes that an auditor makes of a lecture. This must be a major source of uncertainty in any attempt to reconstruct precisely what Wittgenstein thought from notes made by others of his classes at Cambridge. Consequently, a therapeutic philosopher has no business with anybody who is not suffering intellectual disquiet, torment, despair, distress, etc. Perhaps this was the point of Wittgenstein’s observation that Russell (in the 1930s) was a philosopher who was suffering from lack of problems. And also the explanation of his contempt for the bulk of British academic philosophers? Wittgenstein’s notion of therapy is generally taken to be less radical than this. He is thought to address philosophical positions (say, Cartesian dualism) and to demonstrate that they are indefensible or untenable (as it were, in the abstract). This is comparable not to the treatment of individuals by a general practitioner, but to a sustained effort to improve public health (say, by eradicating smallpox from the world). Difficulties in social interactions are treated as ‘symptoms’ of underlying internal disorders, while therapy directed at them is viewed as a palliative, not a cure. The criterion for the presence of these conflicting desires must be the patient’s own sincere confession of them. Conflicts of the will! ‘the . . . pathological [krankhaft] character of solipsism shows itself if we try to draw the consequence that only I, N.N., really see, since we immediately shrink back from this consequence. We immediately see that we didn’t want to say that at all’ (LPE 274). The term ‘rational’ must be so understood that it does not exclude the phenomenon of ‘transference’. But I cannot discuss this complication here. What we call reasoning takes many different forms; e.g. making a case for a stylistic classification of a group of paintings or advocating social reforms. ‘The therapeutic prospects lie in the possibility of getting rid of this ‘‘repression’’, so as to allow part of the unconscious psychical material to become conscious and thus to deprive it of its pathogenic power’ (Freud, ‘On PsychoAnalysis’, Complete Works, vol. 12, p. 208). ‘I have been engaged for many years . . . in unravelling certain psychopathological structures. . . . If a pathological idea . . . can be traced back to the elements in the patient’s mental life from which it originated, it simultaneously crumbles away and the patient is freed from it’ (Freud, ‘The Method of Interpreting Dreams’, Complete Works, vol. 4, p. 100).

174 Wittgenstein and Waismann 21

22

23

24

25

26

27

Of course, there are disanalogies with psychoanalysis. Waismann did not make any explanatory use of the concept of ‘the Unconscious’, and he did not relate the patient’s unhappiness to incidents or events in his earlier intellectual life. ‘I set great store on never taking a step without remaining in agreement with your judgement; I discussed a great deal with you and gave way to your objections – in fact I recognized you and your ‘‘common sense’’ as a deciding factor’ (Freud, ‘Psycho-Analysis and Psychiatry’, Complete Works, vol. 16, p. 243). ‘It was my view. . . (though I have since recognized it as a wrong one) that my task was fulfilled when I had informed a patient of the hidden meaning of his symptoms: I considered that I was not responsible for whether he accepted the solution or not – though this was what success depended on’ (Freud, ‘The Method of Interpreting Dreams’, Complete Works, vol. 4, p. 108). Though frequently neglected and commonly misunderstood, this is an important feature of Wittgenstein’s clarification of the concept of aspect-seeing (cf. PI p. 213; LPP 113, 334). It is closely related to the idea that the flashing of an aspect on us seems ‘half visual experience, half thought’ (PI p. 197) and to the notion that seeing a triangle as lying on its side demands imagination (PI p. 207). A further link between acknowledgement and aspect-seeing may be built into Waismann’s conception of psychoanalysis. It is plausible to think of the therapist not as giving his patient new information about his own behaviour, but as revealing previously unnoticed patterns in behaviour of which he was fully aware. For example, the patient might now see that his behaviour clearly manifests an inferiority complex. This seems a straightforward case of an aspect-switch. (But Waismann did not draw attention to this point.) Certain objections would be deflected. In particular, this dialectical method leaves no scope for worrying about how to verify descriptions of the grammar of our language. There is nothing to discuss about what ‘we say’ (or what ‘we don’t say’) apart from what the participants in a particular discussion openly acknowledge to be their own practice. (Contrast Quine’s worries about how to ascertain whether two expressions in our language are strictly speaking synonymous. This seems a most natural concern for any philosopher influenced by and sympathetic to the spirit of logical empiricism.) In this respect, Drury related Wittgenstein’s writings to a quotation from Kierkegaard: ‘The very maximum of what one human being can do for another in relation to that which each man has to do solely for himself is to inspire him with concern and unrest.’ He then added: ‘Conversations with Wittgenstein were . . . disturbing, and if his writings do not produce the same unrest they have been misunderstood’ (Drury, in ‘Some Notes on Conversations with Wittgenstein’, pp. 33, 36). Compare: ‘You must not take this . . . to mean that I propose to give you dogmatic lectures and to insist on your unqualified belief. Such a misunderstanding would do me a grave injustice. I do not wish to arouse conviction;

Thinking about ‘Thinking’

28

29

30

31

32

33

34

35

36 37 38 39

175

I wish to stimulate thought and to upset prejudices’ (Freud, ‘Psycho-Analysis and Psychiatry’, Complete Works, vol. 16, p. 243). ‘We consider language from one point of view only. . . . Meaning, in our sense, is embodied in the explanation of meaning. . . . we said that by ‘‘meaning’’ we meant what an explanation of meaning explains’ (PG 60, 68). One might say: there is no reality here, i.e. nothing independent of individual acknowledgement. But even if there were, it will play no role in our considerations. We are not interested in an ostensive definition as a cause of someone’s use of a word; we are interested in it only from the point of view of its being a rule (F 14, 36). Compare: ‘[My] technique is the only one suited to my individuality; I do not venture to deny that a physician quite differently constituted might find himself driven to adopt a different attitude to his patients and to the task before him’ (Freud, ‘Recommendations to Physicians Practicing PsychoAnalysis’, Complete Works, vol. 12, p. 111). Another danger lies in sublimating pictures or analogies, especially in treating our actual linguistic practice as an approximation to an ‘ideal’ (F 24, 93; BT 435; cf. PPI §107). (In a discussion, Wittgenstein once remarked: ‘Keep it simple – then we can make some progress.’) The concept of sublimation is another point of contact with Freud’s thinking, but I cannot investigate it here. ‘Every time we come upon a symptom we can infer that there are certain definite unconscious processes in the patient which contain the sense of the symptom . . . ; as soon as the unconscious processes concerned have become conscious, the symptom must disappear’ (Freud, ‘Fixation to Traumas’, Lecture 18 of Introductory Lectures of Psycho-Analysis, Complete Works, vol. 16, p. 279). This diminishes the temptation towards solipsism: ‘// The conception of solipsism // does not extend over games. The other can play chess just as well as I can’ (LPE 258). Similarly, acknowledging the possibility that ‘I have a pain’ is a learned replacement for a natural expression of pain (PI §244) explodes the dogma that this sentence must be a description. We might call it a form of conceptual homoeopathy! Rather than using them as models to which reality must conform! (PI §130–1). At least in respect of choice between rules which square equally well with my own practice. (There are complications which I cannot discuss here.) The grammar of the term ‘anerkennen’ (‘acknowledge’) is subtle and complicated in Freud’s writings, and its use diverges in some respects from everyday employment of the English verb ‘acknowledge’. Though some of Wittgenstein’s uses of ‘anerkennen’ are meant to echo Freud’s (e.g. BT 410), there is room for uncertainty as to how close the correspondence is. In both cases, there are problems about specifying exactly what is acknowledged

176 Wittgenstein and Waismann

40

41 42

43

44

45

46 47

48

49

(the occurrence of an event in childhood? a reaction to an event at the time of its occurrence? taking the occurrence of this event to make sense of a present feeling? etc.), and there are further difficulties about how to distinguish acknowledgement from rationalization and fantasy. I cannot do justice to these issues here, important as they are. ‘The psychoneuroses are substitutive satisfactions of some instinct the presence of which one is obliged to deny to oneself and others. Their capacity to exist depends on this distortion and lack of recognition. When the riddle they present is solved and the solution is accepted by the patients, these diseases cease to be able to exist. There is hardly anything like this in medicine, though in fairy stories you hear of evil spirits whose power is broken as soon as you can tell them their name – the name which they have kept secret’ (Freud, ‘The Future Prospects of Psycho-Analytic Theory’, Complete Works, vol. 11, p. 148). It is not obvious how to interpret these phrases, but I cannot explore this problem here. ‘So befreien wir auch vom Bann des Ideals, indem wir es als Bild anerkennen und seinen Ursprung angeben. – Wie bist Du zu diesem Ideal gekommen; aus welchen Material hast Du es geformt? Welchen konkrete Vorstellung war sein eigentlicher Urbild? Dies mu¨ssen wir uns fragen, sonst ko¨nnen wir seinen irrefu¨hrenden Aspekt nicht los werden’ (PPI §107). ‘Wir a¨ndern nun den Aspekt, indem wir einem System des Ausdrucks andere an die Seite stellen. . . . Die fu¨r uns philosophisch wichtigsten Aspekte der Dinge sind durch ihre Einfachkeit und Allta¨glichkeit verborgen’ (PPI §§99, 105). Here is a striking connection with Freud’s early conception of psychoanalysis: ‘The pathogenic idea which has ostensibly been forgotten is always lying ‘‘close at hand’’ and can be reached by associations that are easily accessible. It is merely a question of getting some obstacle out of the way. This obstacle seems . . . to be the subject’s will’ (Freud, ‘Psychotherapy of Hysteria’, Complete Works, vol. 2, p. 271). One common pattern is to argue that the question rests on a ‘category mistake’, as explained by Ryle. Ryle himself went so far as to suggest that all philosophical problems arise from category mistakes, hence that all of them can be dissolved by establishing ‘the logical geography’ of our language. Though, of course, these are not ‘brute facts’, but rather ones constitutive of (part of) a human practice, e.g. like facts about the powers of chess pieces. This seems to be what Wittgenstein had in mind as bringing to light ‘eine falsche Fragestellung’. It is illustrated by Hertz’s dissolution of the physicist’s questions ‘What is force?’ and ‘What is electricity?’ Exploiting the text of the Brown Book and his shorthand notes, Waismann took it as a model of a philosophical problem and its solution (Lo¨sung) in LSP, ch. 2. ‘How must we look at [Augustine’s question] to see it as a problem? I believe it is only a problem for those who have been misled by an analogy. . . ’ (PLP 42).

Thinking about ‘Thinking’ 50

51

52 53

54

55

56

57

58

59

60 61

177

That is, if he now refused to surrender it. (‘True’ and ‘false’ have a different grammar in application to pictures or analogies from their use in application to empirical statements.) Would this be a clear distortion of ‘the concept of measuring’? No more obviously so than Wittgenstein’s refining or redefining the term ‘Satz’ in order to exclude from its range of application arithmetical equations, rules of chess, or the formulation of Boyle’s law. In sharp contrast to the case of an analytic philosopher’s convicting someone else of committing a category mistake or a fallacy in reasoning. Russell objected to Wittgenstein’s proposal (in PR) to distinguish various different Satzsysteme, each with its own logical grammar of generality; in his view, this would make logic impossible! Logic must, in this sense, be ‘independent of all future experience’ (cf. PI §92). The Tractatus had secured this desideratum by distinguishing sharply between logical syntax (logic) and the applications of logic; what can be decided by logic can be decided a priori, without more ado, but the details of the analysis of ‘the phenomena’ cannot be foreseen (cf. RLF). These remarks are not well understood. In this respect, acknowledging a rule is completely different from making an observation in a laboratory; calling both ‘discoveries’ is liable to generate serious confusion. The point is surely to get ‘us’ to break away from the model of name and designation, from the preconception that the sole function of discourse is to describe things, from the grammatical fiction that pain must be something. Compare F 86: the use of the word ‘sense’ [‘Sinn’] is misleading when it is associated with the conception that the sense of a sentence must be something which is distinct from the words and which stands behind the sentence. ‘If we wish to grasp the relationship of the intensity of a belief to its content, ¨ usserthen we can for our purposes replace this intensity by its expression (A ung) . . . ’ (F 5). This characterization of his formula serves to explain why it must have become obsolete in his later discussions of ‘thinking’: namely because he later repudiated behaviourism! A form of representation (Darstellungsform) of our grammar; as it were, a second-order form of representation. Compare F 22, 33. ‘It is a long-superseded idea, and one derived from superficial appearances, that the patient suffers from a sort of ignorance, and that if one removes this ignorance by giving him information (about the causal connections of his illness with his life, about his experiences in childhood, and so on) he is bound to recover. The pathological factor is not his ignorance in itself, but the root of this ignorance in his inner resistances; it was they that first called this ignorance into being, and they still maintain it now. The task of the treatment lies in combating these resistances’ (Freud, ‘ ‘‘Wild’’ PsychoAnalysis’, Complete Works, vol. 11, p. 225).

178 Wittgenstein and Waismann 62

63 64

65

‘We tell [the patient] that the success of the psycho-analysis depends on his noticing and reporting whatever comes into his head and not being misled, for instance, into suppressing an idea because it strikes him as important or irrelevant or because it seems to him meaningless. He must adopt a completely impartial attitude to what occurs in him, since it is precisely his critical attitude which is responsible for his being unable, in the ordinary course of things, to achieve the desired unravelling of his dream or obsessional idea or whatever it may be’ (Freud, ‘The Method of Interpreting Dreams’, Complete Works, vol. 4, p. 101). This point ultimately rests on acknowledgement. Prior to philosophical therapy, we did not acknowledge this proximity, whereas we now do. Similarly, a false analogy is one that I now disavow, but acknowledge as one that I previously accepted. (It is not something faulted by some objective, external standard of correctness.) ‘This work of overcoming resistances is the essential feature of analytic treatment; the patient has to accomplish it and the doctor makes this possible for him’ (Freud, ‘Analytic Therapy’, Complete Works, vol. 16, p. 451).

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

9 A Vision of Philosophy*

I want to consider the main themes of ‘How I See Philosophy’ (1956). Here Waismann presented a strikingly unusual conception of philosophy in a forceful way. It merits attention in its own right. In addition, his discussion is of special historical interest because it is very closely based on material that Wittgenstein dictated to Waismann during the period 1931–51 when these two were collaborating on the production of Logik, Sprache, Philosophie, intended to be the first volume in the series published by the Vienna Circle under the title Die Wissenschaftliche Weltauffassung. The vision of philosophy that Waismann elaborated in 1956 seems to have been Wittgenstein’s own at least in the early 1930s. The article is often summarized as an extended comparison between conceptual analysis and Freudian psychoanalysis; hence it is described as the elaboration and exploration of an analogy. This summary puts everything in the wrong light. Waismann did not offer an analogy to clarify aspects of philosophical analysis as practised by Wittgenstein, Russell, Carnap, Schlick, logical empiricists, and in a modified form by Austin, Ryle, Quine, etc. Psychoanalysis is in fact meant as a model for developing a distinctive form of intellectual therapy (‘our method’2). In many respects, this therapeutic method is radically different from established procedures of conceptual analysis in analytic philosophy. Waismann outlined not an analogy, but a revolutionary programme. It is a description of a very distinctive method which appears to have dominated Wittgenstein’s work at this period.3 Is the method here described even intelligible? A quick reading of ‘How I See Philosophy’ suggests that a paradox is generated by two of its main themes. On the one hand, freedom, spontaneous decision and vision are * Originally published in Figuras do Racionalismo, Conferencias ANPOF 1999 (1999), pp. 137–77. Reproduced by permission.

180 Wittgenstein and Waismann

the essence of philosophy. On the other hand, ‘our method’ consists in describing and clarifying the grammar of our language. There seems no possible connection between these two ideas. To anybody familiar with prevalent practice of conceptual analysis, the first of these themes appears to be a wry joke. ‘Freedom? Why of course! You are perfectly free to talk nonsense if you wish.’ I will suggest that this is not Waismann’s conception. My main goal in clarifying his article is to make the apparent paradox disappear completely. I hope thereby to make good sense of his programme of enlarging freedom by clarifying grammar. The body of the paper consists of a preamble, a coda, and six interconnected themes clarifying the relations of ‘our method’ to psychoanalysis.4

Preamble As his title indicates, Waismann offered ‘my conception’ of philosophy. (Or more accurately, ‘our conception’: i.e. the one shared with Wittgenstein and Schlick in the early 1930s.) The title implies an open acknowledgement of pluralism, an abstention from any form of imperialism. Consequently, at least to a first approximation, this is not a delineation of the nature of philosophy (at least as that is usually understood). The conception is not prescriptive or exclusionary; it does not purport to state how philosophy must be conducted. And it is certainly not presented as a description of the grammar of the term ‘philosophy’ (if there is any such thing). Tolerance of alternatives is combined with a clear recommendation of ‘our method’.5 Waismann envisaged a possible method for dissolving some6 of the so-called ‘problems of philosophy’ (a loose family of cases). ‘Our method’ is to be distinguished from others,7 and it has certainly not been a practice prevalent in the history of philosophy. We might, with great caution, agree that ‘How I See Philosophy’ does describe the essence of philosophy – but only if the notion of essence is correctly understood. According to Wittgenstein’s conception, essence is created by us; by stipulating how we wish to use our own words. (Our stipulations are answerable to nothing; there can be no justification whatsoever for our giving a word a particular grammar. Here ‘we do not want to give reasons; . . . there is nothing to be discovered in grammar’ (HISP 12).) Waismann exploited this conception of essence when he proclaimed that ‘the essence of philosophy lies in its freedom’ (HISP 21) and when he defined philosophy as ‘the elimination of a peculiar mental unrest’ (GA xxii). The vital point here is that the essence of

A Vision of Philosophy

181

philosophy (understood according to ‘our conception’ of essence) is something we are free to accept or reject. (We can reject it without thereby convicting ourselves of irrationality – or condemning ourselves to talking nonsense.) The whole of ‘How I See Philosophy’ can be seen as an elaboration of Wittgenstein’s remark: ‘Our method is similar in certain respects to psychoanalysis’ (DS 28; F 93).8

The Six Main Themes (1) ‘Our method’ is radically therapeutic.9 It is a method for treating thinkers and their troubles, not abstract problems, confusions or nonsense. (‘The very word ‘‘problem’’, one might say, is misapplied when used for our philosophical troubles’ (BB 46).) ‘Our method’ gives individual treatment to individuals’ particular ‘problems’. It is radically individualistic because it demands the active participation of the ‘patient’ in a discussion. He must explain what he wants to say, how he wants to define expressions; he must acknowledge the pictures that influence him; he is invited to adopt novel ways of ordering things; and so on. Consequently, this procedure is essentially dialectical. It constantly demands that the interlocutor make decisions, often entirely unprecedented ones. There is a tendency to neglect or discount the fact that ‘our method’ focuses on individuals, on the welfare of troubled persons. Therapy must be individualized. Take the remark that Augustine’s picture surrounds the working of language with a fog (PI §5). Commentators often say that this picture generates confusions or diseases of the intellect (especially Platonism in mathematics and Cartesianism in philosophy of mind); and they take Wittgenstein to argue that we must not see words as names, but that we must (or should) see them as instruments. Here generalizations get the upper hand. Wouldn’t it be better (more in the spirit of ‘our method’) to say that Augustine’s picture very often creates disquiet, and that if it does so for me then it would be better for me to regard words as tools? It is difficult to keep Waismann’s conception sharply in focus: especially to resist the temptation to see ‘our method’ as a strategy for dealing with abstract problems which are to be investigated independently of any individual thinkers. Wittgenstein’s therapeutic discussions are apt to be assimilated to other forms of philosophical activity. As if he were engaged in the task of ascertaining the categories of terms10 or in a quasi-scientific investigation into what words really mean (BB 28); perhaps for the purpose of rectifying category mistakes (Ryle) and solving geometrical

182 Wittgenstein and Waismann

problems laid up in the structure of a shared natural language (comparable to finding solutions for chess problems). ‘Our method’ is radically different precisely because its purpose is to help individuals who suffer from philosophical problems. It is also difficult to resist the tendency to construe stretches of dialogue as confrontations of views (various ‘isms’). We are apt to identify a single interlocutor in a sequence of remarks (e.g. in the ‘private language argument’) and to see him as the spokesman for a coherent point of view (e.g. ‘the Cartesian conception of the mind’). Giving in to this temptation distorts the whole intellectual scene. It transforms moves in an interactive dialogue into stages in an adversarial debate.11 (And the reader is then cast into the role of a referee rather than another active participant in the discussion.) ‘Our method’ differs from much philosophical argument in calling for scrupulous respect for individuals, for their subtly varying points of view and their autonomy. Hence it requires us to refrain from producing stereotypes of ‘the opponent’ and from trying to paste together the juxtaposed remarks into Wittgenstein’s positive doctrine. ‘Our method’ resembles psychoanalysis, then, in being individual treatment taking the form of rational discussion with the ‘patient’ of his problems with a view to changing his way of seeing things.12 To whom is it to be applied? To anybody bothered by certain kinds of conceptual blindnesses, dissonances or confusions. This might be a novice struggling with the question, ‘What is the nature of thinking?’ Or it might be Russell when he was tormented by the question, ‘How can a meaningful singular term (a definite description) fail to denote something?’ Or it might be a pair of philosophers rigidly adhering to antithetical theses; e.g. Frege and Hilbert on the question whether the axioms of Euclidean geometry are definitions of the primitive terms. Or it may be oneself. Indeed, as in psychoanalysis, the most important and most difficult form of analysis is self-analysis. ‘Working in philosophy. . . is really more a working on oneself. On one’s own interpretation. On one’s way of seeing things. (And what one demands of them)’ (BT 407; CV 16). (2) The focus is on individuals’ states of mind: unrest (PLP 7, 8; F 93), torment, disquiet (PI §111; LPM 33), discomfort (BB 26; PLP 4), drives (PI §109), obsessions (HISP 18), craving (BB 17; LPM 58), revulsion (BB 15, 57), Angst (BB 27; F 94), irritation (PLP 7; F 62), profound uneasiness of mind (HISP 3), profound mental discomfort (HISP 6), obsessional doubt (HISP 8), shock (PLP 7), troubles (BB 46), compulsions to say things (BB 47), irresistible temptations (BB 18), alarm (HISP 4), etc. That is to say, the focus is on a distinctive range of intellectual emotions or disturbances.13 Hence, too, on individuals; on

A Vision of Philosophy

183

those particular persons who suffer from these feelings. There is, as it were, no disquiet, unrest or torment which is not someone’s troubled state of mind. The feelings targeted by ‘our method’ are differentiated from other feelings by their objects (e.g. conceptual clashes (PI §112), paradoxes, contradictions, entanglements in our own rules (BB 46; PI §125)). There are particular kinds of feelings associated with particular kinds of objects. Different instances of these feelings are individuated by their particular objects; e.g. the frustration of feeling unable to describe an immediate experience in full detail, as if language is too coarse and crude (F 62; cf. PI §106); or the dissatisfaction that basic concepts like number or causality defy all attempts at definition.14 Successful treatment of individuals’ troubles requires, in each case, either eliminating these objects or altering someone’s attitude towards them. Otherwise the treatment is of something else! i.e. of a different discomfort. The goal is to get these particular (quite specific) disturbances (disquiets) to disappear completely; hence, the elimination of a peculiar mental unrest (GA xxii). (This is quite unlike the treatment of somatic disturbances like headaches! These have no objects, only causes. Hence they can be eliminated – wholesale – by countercauses, e.g. drugs; or suspended, e.g. by analgesics.) Idioms associated with psychological disturbance are conspicuous in HISP and in Wittgenstein’s writings. They are often reported or paraphrased. But this terminology is not treated seriously and literally; rather it is discounted as une fac¸on de parler, a stylistic flourish, or hyperbole.15 This is a fundamental misreading of these texts. Waismann meant to focus attention on forms of intellectual discomfort (internal conflict), and he wanted to offer therapy for precisely these ailments of troubled individuals. The aim is to eliminate internal conflicts (tensions). To change what I want to say, with my consent; not to establish limits to what I can say. (Hence to operate through my will, not against it.) The tension must be acknowledged by the ‘patient’. Hence the diagnosis of an internal conflict (problem) is correct only if the interlocutor acknowledges it to be so (BT 409). Consequently, the notion of ‘dissolving problems’ here takes on a distinctive character: clarity about what our words mean is meant to make misguided cravings, drives, urges or prejudices [falsche Bedu¨rfnisse] disappear. For example, it aims to undercut the drive to look for ‘ideal languages’ (F 62). Or the urge to regularize what seems to be the disorderly use of our words (F 94). (Our disquiet (irritation) about the capacity of ordinary language to describe immediate experience is born of a

184 Wittgenstein and Waismann

false analogy which informs misunderstanding of the different uses of ‘exact’ (F 62).) Suffering is the presupposition of ‘our method’: in particular, it is the motive for the patient’s taking the way out of the flybottle when it is made clear to him.16 He feels relief (HISP 20). (‘Our method’ is inapplicable to anybody unaffected by intellectual torment, or who suffers (as it were) from a lack of problems!17) It is distinctive of ‘our method’ to focus on the states of mind of troubled individuals. Where then is it applicable? Some paradigmatic ‘problems’ would be Wittgenstein’s own feeling that ‘Logic’s hell!’ [CV 30]; Russell’s distress before he succeeded in working out the theory of descriptions; or Frege’s twenty years of depression after the collapse of his explanation of how to define numbers as objects. There might be very limited scope for the application of this method to the ruminations and debates of modern academic philosophers! (3) The source of philosophical disquiet, torment, compulsion, etc. is identified as unacknowledged or unconscious analogies, pictures, models, ways of seeing things, dogmas (prejudices), ‘misleading analogies in the use of our language’ (BT 408), or ‘similes absorbed into the forms of our language’ (PI §112). (What is meant by these phrases will emerge only from examination of detailed treatment of particular problems.) Here is a first (and vital) point of clarification: these pictures, analogies, etc. characterize the thought of some individuals (but not all! cf. PLP 211),18 and their influence needs to be acknowledged by the sufferer in every individual case! (PLP 36). For this reason, the diagnosis of someone’s philosophical problem is correct only if the interlocutor acknowledges it to be so (BT 409). The inspiration of ‘our method’ is the idea that unacknowledged pictures generate mental cramps, obsessions, prejudices (and misuses of our words – as we come to see by our own lights); and those in turn prevent our finding a way out of our problems.19 (We may behave like flies who are ‘trapped’ against a sunlit window.) Here Augustine’s problem (How is it possible to measure time?) is a recurrent paradigm for both Waismann and Wittgenstein. ‘When we ask ‘‘How is it possible to measure time?’’ we may well have a picture in mind, say of a band continuously passing by us’ (PLP 41). This seems to make the task self-frustrating, hence impossible. To dissolve this perplexity, we need to detach ourselves from the notion that measurement must consist in juxtaposing two things; we need to remind ourselves of the variety of things that we ourselves count as measuring (e.g. pacing off a field). This is a clear instance of how a picture may have a peculiar power to blind us.

A Vision of Philosophy

185

As long as pictures or analogies are unconscious, we cannot be critical of, or even on our guard against, their influence. They exert tyranny over our thinking; they hold us in thrall; they produce mental cramps; in short, they restrict our freedom of intellectual movement – even though we are unaware of this (CV 28).20 As long as they are unconscious, pictures can never be put to the test; hence, in effect, they function as norms of representation just because they are not open to challenge.21 They seem to fix how things must be; or how they cannot be. Hence they exclude possibilities. In this respect, they are comparable to prejudices (Vorurteile), superstitions (cf. PI §110), or dogmas (PLP 79). Just as in the case of ‘torment’, ‘unrest’, etc., it is crucial to realize that such terms as ‘tyranny’, ‘thrall’, ‘bondage’, etc., are not hyperbolic; they are meant to be literal descriptions of serious loss of freedom – which is real enough, even though we ourselves may be quite unconscious of it. One clear link with psychoanalysis is the idea that this power of pictures or analogies can be broken only by bringing them to consciousness and acknowledging them (DS 28;22 PLP 179; PPI §106). Correcting ‘mistakes’ that they generate would be as pointless as suppressing the symptoms of neuroses. In dealing with our troubled states of mind, we have to penetrate our own motives, to come to understand exactly how we fell into perplexity; especially to retrace the sources and steps of our thinking,23 to trace our confusion back to pictures or analogies. We need to bring these roots of our thinking to consciousness. Hence, e.g., to free ourselves from the picture of possibility as a shadowy reality we need to articulate it clearly (F 57), whereupon it will lose its charm. In this investigation, personal acknowledgement is in principle indispensable (BT 410). Everything depends on it (PLP 36). (Present acknowledgement functions as a criterion for what I have always thought.24) But the task is enormously difficult. We need to work against our own resistances (BT 406)! And our urges to misunderstand (PI §108)! We do not want to acknowledge many of the sources of our own thinking. This is uncomfortable! Particularly because the underlying pictures may well seem very crude or infantile. (We may wish to exercise censorship on them.) Moreover, they are deeply ingrained in our thinking [Denkgewohnheiten], so that we are constantly at risk of lapsing back into them. ‘Our method’ tries to bring to an individual’s consciousness the influence of pictures working unconsciously within him. It strives to combat pictures that generate perplexities or confusions. But this cannot be a matter of refutation. Pictures or analogies are, like shadows, unassailable. It makes no sense to refute them! They lie out of reach of demonstration or refutation. Hence ‘in philosophy there are no proofs; there are no

186 Wittgenstein and Waismann

theorems; and there are no questions which can be decided, Yes or No’ (HISP 1).25 The intellectual discomforts targeted by ‘our method’ are said to stem from similes absorbed ‘in our language’ or ‘in our thought’ or ‘in forms of expression’ (BT 409). (Hence there is a need to ‘escape the domination of linguistic forms’ (HISP 18).) What is to be understood by these expressions? We must investigate this matter by looking at the wide range of examples that Wittgenstein gave. One instance is the idea that the mind is what gives life to dead signs (BB 4). Another is the notion that the head is the real location of thinking (BB 16), yet another that a machine (clockwork) cannot have a ‘private experience’ (or state of consciousness) (BB 16). Other examples: the comparison of visual impressions and memory images with painted pictures (F 62); the idea that denying that numbers are signs makes it necessary to answer the question ‘Then what are they?’ (F 1, 94; BB 27); the picture of possibility as a shadowy actuality or of actuality as moving along the rails of possibility26 (F 57; cf. Z §70); the picture that what we see of a sign is merely the exterior of some inner thing in which the real operations of meaning go on (RFM 108); the model of thinking as a play of images (Z §94); or the picture which tricked us into thinking that the sense of ‘‘p is true’’ differs from the sense of ‘‘p’’ (F 9; cf. PLP 32). There is nothing very uniform here. Both the contents and the roots of these ‘pictures’ seem multifarious. We should try to resist saying that they must have something in common;27 instead we should look at the differences! There is a category of unconscious ‘pictures’ which seems particularly pervasive and pernicious. This is a motley of pictures of ‘how language works’; pictures of how to speak about how we speak. In this connection, Wittgenstein frequently drew attention to aspects of what he called ‘Augustine’s picture of language’ (PI §1). According to that picture, all words are names of objects. This manifests itself in the tendency to construe the grammar of the expression of sensation on the model of ‘object and designation’ (PI §293). Another aspect of this picture is the idea that all sentences are descriptions. This amounts to adhering to ‘the idea that language always functions in one way, always serves the same purpose: to convey thoughts – which may be about houses, pains, good and evil, or anything you please’ (PI §304). Other examples of this category of pictures are the dogma that all items falling under a general term must have something in common (PI §65) and the thesis that every proposition must be complex [zusammengesetzt] (cf. PLP 317–19). The phrases ‘forms of expression’, ‘the forms of our language’,28 etc. need careful elucidation based on detailed examination of the treatment of particular philosophical problems. What Waismann (and Wittgenstein)

A Vision of Philosophy

187

have in mind often seems closer to ways of thinking and speaking (which are variable both between individuals and over time) than to the syntactic structures of natural languages (which are relatively fixed and invariant). On this view, ‘Someone is misled by a form of expression in cases where we think ‘‘he wouldn’t talk as he does if he were aware of . . . ’’ ’ (BB 28) – and that has to be confirmed by his acknowledgement! Accordingly, the misleadingness of a form of expression is considered to be person-specific, and it may well depend on the rest of what he thinks and says. The task of exposing unconscious pictures and of eliciting acknowledgement of their influence is infinitely delicate and subtle. ‘Our method’ calls for imagination, empathy – and skill comparable to producing a ‘speaking likeness’ in a portrait (cf. BT 410). It is far removed from non-interactive collecting of observational data about ‘the grammar of our (public) language’. What the ‘patient’ needs (as in psychoanalysis) is not further information, but rather insight into the nature and sources of his blindness to patterns which are to be found in what is already familiar to him. (4) Analogies, models, similes, pictures play a double role in ‘our method’. They are pivotal both in the genesis and in the cure of diseases of the understanding. In this respect, ‘our method’ could be called a form of homoeopathy: a way of ‘treating’ pictures with other pictures. ‘As analogies give rise to philosophical problems, so analogies are useful in conjuring away philosophical problems’ (PLP 60).29 ‘We juxtapose an analogy against our language as the dissolution of a philosophical problem’ (F 90). In this sense, it is fresh pictures, models, analogies, or language-games which are crucial in clarifying the grammar of our language! Especially in revealing ‘the aspects of things that are most important for us’ (PI §129)! Perhaps for removing the disturbing aura of uniqueness that may surround a particular instance (F 90). (All of this, if taken seriously, is apt to sound paradoxical! It seems inconsistent with the idea that dissolving philosophical problems calls for ‘the quiet weighing of linguistic facts’ (Z §447).) The programme is clearly formulated. Treatment according to ‘our method’ involves making a case for (or getting the ‘patient’ to make use of) a new and different form of representation. ‘The decisive step consists in following a certain analogy with other cases’ (HISP 25). We put a particular picture before him; ‘and his acceptance of the picture consists in his now being inclined to regard a given case differently. . . . I have changed his way of looking at things’ (PI §144).30 ‘Our method’ uses pictures to ‘change the asker’s attitude’ (HISP 18), to get someone ‘to see things in a new way’ (HISP 21). It aims at a ‘conversion of viewpoint’

188 Wittgenstein and Waismann

(Umstellung der Auffassung (PPI §116)). ‘We change the aspect of things by juxtaposing with one system of expression other ones. – Thus can the thrall in which an analogy holds us be broken, if there is juxtaposed with it another analogy which we acknowledge to be equally justifiable’ (PPI §99). One crucial instance of executing this programme is Wittgenstein’s telling a story (we teach a child a verbal replacement for primitive painbehaviour) which reveals the possibility of treating the utterance ‘I am in pain’ as an expression or avowal of a sensation, not a description (PI §244). This remark is commonly misrepresented as stating some proven fact of grammar (Wittgenstein’s ‘doctrine of avowals’); and as stating this with the authority of someone who is destroying a common prejudice (cf. BB 25). But what is described is explicitly labelled ‘one possibility’! The point of the remark is to bring to consciousness the prejudice that ‘I am in pain’ must be the description of something (my experience, my behaviour, a state of my brain or body. . . ). It offers an alternative to one part of Augustine’s picture of language. This is a case where a substitute picture has won widespread acceptance. The pivotal remark ‘Equations are rules of grammar’ has a similar role. It is not meant to be a doctrine (something we are meant to say); but rather a picture whose application is to be explored (‘I want you to do something you don’t want to do, namely to look at equations from this point of view’ (cf. LFM 103)). Wittgenstein tried to win over his readers (or his auditors) to accepting a possibility not previously considered. Rules of grammar (as we now commonly understand them) are to be used as an object of comparison in investigating and describing the use of arithmetical equations. This picture is meant to be a substitute for the pervasive picture that arithmetic is ‘the natural history (mineralogy) of numbers’ (RFM 229). This is a case where an alternative picture has met very strong resistance. In such cases the dissolution of a problem (the elimination of someone’s disquiet) involves persuasion, not demonstration. It turns on winning the interlocutor’s assent to the suggestion of replacing one picture with another; on gaining his acceptance of something that might be dismissively called ‘a mere possibility’ or ‘a mere analogy’. There is no way, through rational discussion, to foist a picture on somebody against his will. It cannot be established ‘by force of argument’. Its acceptance is always voluntary. This cannot be coerced or constrained (HISP 20); I may always persist in refusing to see an aspect – even if it is ‘drawn to my attention’. Hence, in so far as a philosophical dispute turns on adherence to divergent pictures, there is a parting of the ways before it comes to argument.

A Vision of Philosophy

189

The actual practice of ‘our method’ involves inventing similes, pictures, models, language-games, or objects of comparison. It displays possibilities and adds the exhortation ‘Look at this!’ (cf. PI §144). It offers the advice to look at things not like this, but like that! (CV 61). It uses pictures to get someone ‘to see things in a new way’ (HISP 21). This is the method that is demonstrated over and over again in Wittgenstein’s work: in the ‘private language argument’, the ‘doctrine of avowals’, the identification of the meaning of a word with its use in the language, the criticism of Augustine’s picture of the nature of language, etc. Many exponents (and most objectors) fail to appreciate that the decisive moves in all these discussions are exhibitions of new objects of comparison, new similes, new pictures, new forms of representation (all of which, of course, we are free to reject!). Let us remind ourselves of just a few important examples: 1 2 3

the analogy between words and tools (PI §§11–12, 421); (F 66) the comparison of a proposition with a ruler (to clarify the mutual exclusion of red and green); (PG 40) the picture of language as being self-contained, or of speaking for itself. This involves redrawing the boundary between grammar and the application of signs (PLP 13–14) in order to emphasize the difference between ostensive definitions and empirical statements as well as the similarity between verbal and ostensive explanations. In describing certain well-defined language-games, we could say that a colour sample is not something that is represented, but a means of representation or part of the language (PI §50); and then the ostensive definition of ‘sepia’ may be seen not as connecting language with reality, but as a substitution-rule for symbols, some of which are concrete (cf. BB 109).

Each of these pictures introduces not the order, but rather an order into our knowledge of the use of language – an order with a particular end in view (PI §132). The decisive move is to sell the interlocutor some problem-dissolving alternative picture or analogy that he will acknowledge to be no less well ‘justified’ than his original unconscious one. Surrounding things with new possibilities has the power dramatically to change how things look to us. Waismann was honest enough to draw attention to the difficulties and the dangers of this method. The difficulty is the one of making visible what is right in front of our eyes, on the surface, and already in full view (PI §89, 92.) A kind of blindness needs to be overcome. This requires more than simply exhibiting something; it calls for removing the internal

190 Wittgenstein and Waismann

obstacles to seeing what is already evident. We must learn to look into and to recognize ‘the workings of our language . . . despite an urge to misunderstand them’ (PI §109). We must learn to control our own craving for generality (BB 18), our incessant demands for further explanations and the difficulty of stopping (Z §314). Accepting a new picture may require our surmounting powerful resistances. The danger inherent in ‘our method’ lies in the immense power of pictures. If one picture blinds us to some aspects of things, another antithetical picture may well blind us to others.31 This poses a threat: ‘The liberator of yesterday may turn into the tyrant of tomorrow’ (HISP 34). An illuminating object of comparison is likely to generate fresh problems if not accompanied by an explicit reality-disclaimer (PI §§131–2). It is apt to become a fresh dogma. This happens in cases where, impressed by the possibility of a comparison, ‘we predicate of the thing what lies in the method of representing it’ (PI §104). In comparison with standard philosophical analysis, ‘our method’ is truly extraordinary: juxtaposing pictures with pictures, seeking to reveal no more (and no less!) than new aspects or patterns or orders. (5) On this conception, freedom is the essence of philosophy (HISP 21). In fact, it is doubly so. Freedom is characteristic of the conduct of ‘our method’ of philosophizing. There is to be no bullying with the stick of logic or the stick of grammar (HISP 18, 29). In the course of treatment, the ‘patient’ constantly exercises his freedom. He faces choices, about what he now wants to say (decisions (HISP 21)), or about how he wants to explain expressions (PLP 34–532); he is asked freely to acknowledge analogies that have influenced him (acknowledgement must, by definition, be full and free!); and he is invited to consent to replacing one formulation of a problem with another (‘to pass from one way of putting a question to another . . . with our spontaneous consent’ (HISP 21; cf. PLP 399–400)). His freedom is always respected. ‘We leave him free to choose, accept or reject any way of using his words. He may depart from ordinary usage. . . . He may even use an expression one time in this, another time in that, way’ (HISP 12; cf. PLP 35).33 His free decisions are the means for treating his intellectual discomfort. He must be led to see some new aspect ‘with his spontaneous consent’ (HISP 20). Aspect-seeing is voluntary (PI p. 213; LPP 113); it depends more on the will than the intellect (BT 406 ¼ CV 17; cf. HISP 21–2). (We might even say that, on this conception, grammar, like ethics, is ‘a matter of the will, not of the understanding’. So that philosophical remarks ‘have nothing to do with knowledge and error nor with ‘‘true’’ and ‘‘false’’ ’ (EW 45).)

A Vision of Philosophy

191

At the same time, freedom of thought is the goal of philosophical therapy; to promote vision (HISP 32), especially by breaking the tyranny, thrall, bewitchment of our intelligence by means of established ways of thinking and speaking (HISP 18) and opening our eyes or bringing us to see things in a new way – from a wider standpoint (HISP 21). This purpose is apparent in helping others, but most importantly in helping ourselves! We need to free ourselves from too narrow a diet of examples (PI §593), from being nailed down to too limited a range of possibilities (LPP 47); hence we need to alter how we want to see things (CV 17) and what we want to say (BB 28–9). ‘Our method’ aims at effecting liberation from dogmas and prejudices: from the ‘metaphysical use of words’ (PI §116), i.e. from various statements incorporating the words ‘must’ and ‘cannot’ which function as constraints on our thinking by narrowing investigation and restricting possibilities from view. In some cases, successful clarification of concepts may yield conceptual revolutions.34 Waismann took pains to stress that the therapy offered by ‘our method’ is not wholly negative (HISP 32). On the contrary, it is inseparable from a positive achievement. It combats prejudice through opening up new possibilities, new lines of thinking to explore. Hence it necessarily enlarges freedom! One instance of following this strategy is the suggestion con¨ usserungen as expressions or manifestations of pain, fear, anger, sidering A etc. rather than as descriptions of ‘states of consciousness’ (PI §244). Another is the idea of comparing ‘trisection of an angle’ with ‘cardinal number whose square is 10’.35 A kind of creativity or imagination is essential to practising ‘our method’. Blindnesses or prejudices are cured only by acknowledging previously unrecognized possibilities;36 in decisions (stipulations) to extend (one’s own!) logical space (e.g. BB 6–8; PI §244; PLP 81; LPP 47). In this way we are called upon to increase our freedom through exercising our freedom. We are encouraged to construct ‘a background of free possibilities’ (GA xxv). But imagining and acknowledging possibilities are voluntary, free, even creative acts. We do not have to recognize new possibilities; e.g. to acknowledge that a chain of reasons may come to an end (BB 15) or that fetching something ‘from memory’ may not involve any mental image at all (BB 3–4). But we enlarge our freedom of thought through freely acknowledging these possibilities. ‘The essence of philosophy is freedom.’ This slogan defines Waismann’s conception of philosophy. It is crucial to appreciate that the freedom envisaged here is reflexive. The practice of ‘our method’ is based on free acceptance of a particular conception of what is required for the clarification of concepts. It turns on free acknowledgement of Wittgenstein’s ‘revelation’ that the nature or essence of thinking, meaning, understanding, etc.

192 Wittgenstein and Waismann

shows itself in the patterns of use of the words ‘thinking’, ‘meaning’, ‘understanding’ (HISP 37–8). This conception of essence is an exemplary instance of ‘vision’ in philosophy (HISP 37). We are completely free to accept or reject this vision. If we acknowledge it, we must do so spontaneously. But we are therefore also at liberty to refuse to do so. In that case, we make no mistake, but we do place ourselves outside the group of individuals whose welfare can be increased by the application of ‘our method’. (This might be compared with a refusal to abstain from selfcensorship which is the precondition for the possibility of treating an individual’s disturbances by psychoanalytic techniques.) Waismann’s focus on freedom is striking. It is linked to emphasis on other concepts: individuals’ troubled states of mind, acknowledgment, aspect-seeing, logical imagination, and the role of pictures and of decisions in the clarification of meaning. (6) Philosophy is conceived as the clarification of concepts, of ‘our language’, or ‘grammar’. (That explains why Waismann’s practice of ‘our method’ is presented in a book whose English title is ‘‘The Principles of Linguistic Philosophy’’.) But this activity takes on a very distinctive character. As part of ‘our method’, it is in principle person-relative and ‘problem’-oriented; and it primarily concerns pictures, linking problems with unconscious pictures and the dissolution of problems with the tactics of replacing one picture by another one. Hence, too, ‘our method’ must always work through the ‘patient’s’ free decisions, not against his will. (In all these respects, it diverges radically from the standard procedures of analytic philosophers. As it were, ‘our method’ turns its back on the technique of developing the ‘logical geography’ of our language and describing ‘the bounds of sense’.) ‘Our method’ is not passive (observation), fact-ascertaining, objective (contra mundum), still less adversarial (overcoming objections by demonstrations grounded in the accumulation of linguistic data). Rather its practice consists in active dialectical discussions with someone tormented by puzzlement – even with oneself! Meaning is (constantly) negotiated (and renegotiated!) with one’s interlocutor.37 Decisions are made (LPM 149–50): decisions about how we are to use our own words (PLP 33). And we are called upon – or challenged – to be participants in a discussion or dialogue. Clarification of meaning is inventive and imaginative. It involves exposing unconscious pictures and in winning acceptance of new ones, or displaying unnoticed patterns (perhaps by employing new concepts like ‘language-game’ or ‘family resemblance’), or stipulating sharp meanings for certain expressions (e.g. ‘a process involving a rule’ (BB 13)). This

A Vision of Philosophy

193

activity is essentially open-ended and cooperative, even empathetic. It is strictly person-specific and purpose-relative. ‘Description [of grammar] depends on how we approach it’ (LPP 122). This method of clarifying meaning (concepts) has two striking features (seldom noticed): (1) Explanations of how words are used (descriptions of grammar) are flexible and vary from context to context; different orders or patterns may be brought to light for different purposes (PI §132) or in respecting different ways of thinking in different individuals. Hence different explanations (or pictures) of the use of words may be complementary, not discordant. (2) The parameters for framing descriptions of grammar are themselves subject to debate, and they may be chosen (stipulated) differently in different discussions. Nothing here is imposed on us: there are no ‘metalogical’ concepts. Any discrimination of expressions into logical categories is loose, flexible, and purpose-relative. ‘Ordinary language simply has not got the ‘‘hardness’’, the logical hardness, to cut axioms in it. . . . You might just as well carve cameos on a cheese souffle’ (HISP 23). The first point suggests that it is to be expected that explanations are local and purpose-specific, hence variable, not uniform and invariant. This need not raise worries about inconsistency. Nor need we explain variations by claiming that Wittgenstein changed his mind, or that later explanations supersede earlier ones. For example, failure to repeat the slogan ‘Thinking is operating with signs’ (BB 6, 15) in PI does not amount to an open admission that this explanation was incorrect. (This was not intended to give a complete description of the use of the word ‘think’. On the ¨ bersicht or ‘our conception’ (F 87, cf. contrary, it was offered as an U F 5–6), a picture of some central uses of ‘think’, or a centre of variation for describing the very complex and relatively unsurveyable use of ‘think’ (cf. LPP 25, 142).) In fact, the variability and contextualization of explanations is everywhere evident in Wittgenstein’s work. He defined, then made use of the phrase ‘a process involving a rule’ (BB 13). His discussion is surely misunderstood when this is treated as a context-independent analysis of the concept ‘following a rule’ and then criticized as embodying too restrictive an account of that concept.38 He gave a particular definition of the term ‘read’ (BB 119), which then dominates a lengthy discussion of ‘being guided’. ‘Think’ sometimes subsumes ‘meaning’ (meinen) and ‘understand’ (e.g. BB 6; DS 15–16); at other time it is used in contrast with these words (e.g. PI §187). Some of his explanations do have the form: ‘By. . . I here mean . . . ’ (e.g. Z §231). We might read many of his explanations as if they had the same form. This would forestall a variety of

194 Wittgenstein and Waismann

worries and objections. For instance, we might relate his explanation of the term ‘language-game’ at BB 17 solely to the rest of that paragraph, thereby avoiding the difficulty that is generated by treating it as a completely general explanation of all his uses of this term. Treating explanations as localized is important if we are to respect the principle that one’s explanations of what another means are correct only if endorsed or acknowledged by the ‘patient’ himself (BT 408; cf. BB 28; PLP 32, 34). In this practice of local explanations there is a form of neo-scholasticism: Wittgenstein made (created) distinctions to dissolve internal conflicts rather than discovering ready-made ambiguities meeting the criteria imposed by lexicographers for differentiating senses of words. He exercised his freedom to legislate (stipulate) distinctions, and he left us the freedom to accept or reject them. He did not claim to establish the grammar of expressions independently of us (his interlocutors); rather we must do so by freely acknowledging the rules that we are following (cf. GA xxv). Phrases like ‘in our sense’, ‘in a certain sense’, etc. are reminders of previous decisions, and they are ubiquitous and vitally important. For example, ‘Meaning, in our sense, is contained in explanations of meaning’ (PG 60). (Compare: ‘Explanation as the cause of the use of a word does not interest us . . . only an explanation which is part of a calculus’ (F 36).) The second implication of this conception of clarifying meaning (concepts) is that the parameters of describing the meaning (or use) of expressions are themselves open to negotiation and renegotiation. This freedom is constantly exploited: the freedom to choose different forms of representation of our language (of our form of representation).39 Wittgenstein offered a particular conception of meaning (e.g. PI §§43, 421); ‘our concept’ of meaning, as opposed to others’ (PG 60).40 He gave different definitions of ‘Satz’ (‘proposition’) for different purposes (RFM 123; BB 109; PLP 280–303). (Our conception takes a proposition to be the point of application of the calculus of logic, so it here disregards differences between hypotheses, observation-statements, equations, etc. (F 54); we here take ‘proposition’ and language to include drawings and motion pictures, and we thus stretch ‘proposition’ or ‘rule’ well beyond the use normally considered by logicians (F 62).) He suggested that gestures and samples be construed as parts of language, not bits of reality external to language (PI §50; PLP 13–14). He directed attention to how we wish to explain the phrase ‘a name of a sensation’ (PI §261). He defined the term ‘grammar’ in terms of a contrast with the ‘application’ of signs: ‘By grammar we mean everything about language which can be fixed before language is applied’ (PLP 13–14; cf. F 39). He noted that the explanation of a sign, in our sense, is a rule for the use of the sign (F 36). And so on.

A Vision of Philosophy

195

Varying stipulations are suggested in respect of ‘signs’, ‘symbols’, ‘tables’, ‘rules’, ‘name’, and ‘description’ (cf. F 36; PLP 130–1). In this way, the concepts used for the purpose of describing the grammar of our language are themselves objects of philosophical investigation, and their content is the subject of negotiation. Waismann drew attention to precisely this dimension of freedom in noting that Hilbert changed what counts as an explanation (‘definition’) of the primitive terms of geometry. This is the source of the dispute between Frege and Hilbert about the nature of geometry (HISP 17). Against the orthodox conception that explanations must determine the application (Anwendung) of signs, Hilbert viewed the axioms as definitions even though nothing whatever is asserted about the applications of any of the primitive terms. This point of view is not something that can be proved to be correct – in opposition to Frege’s point of view; if it is objected that Hilbert left the primitives completely undefined (as Frege did object), this assertion cannot be demonstrated to be wrong. (Instead, ‘a man like Frege . . . has first to be turned round to see the matter differently. What is required is a change in the entire way of thinking’ (HISP 17–18). This might be a clear case of a mathematician’s having a prejudice (LFM 133).) We may need to reconsider what is to count as a definition and arrive at a new conception of what to call explanations of meaning. Appreciating the freedom exercised in the choice of the forms of representation of the grammar of our language is crucial to understanding fundamental conceptual shifts in the history of our thinking. In effect, Einstein argued that ‘simultaneous’ was not yet defined (PLP 10, 12), or equivalently he argued for (stipulated?) a novel conception of what is required to clarify the meaning of ‘simultaneous’.41 Frege shifted the concept of a concept in arguing that existence-claims and countstatements make assertions about concepts; and he redefined the concept of analytic truth in proving that arithmetical equations are analytic. Cantor introduced a novel concept of proof in his diagonal procedure; this amounts to a new criterion (meaning) of sameness and difference in cardinality.42 Wittgenstein himself constantly introduced new and variable demands on what is to count as describing grammar (cf. LPP 122); e.g. urging that it is part of the grammar of ‘toothache’ to explain what we call ‘getting to know that another has toothache’ (BB 24); encouraging us to examine the sense of mathematical statements by reference to how we put to them to work in science or everyday life (the diametrical opposite of Hilbert’s treatment of geometry!); characterizing the sense of ‘I feel pain’ by viewing it as a verbal replacement of crying (PI §244); or suggesting that explaining the expression ‘the same length’ must be part of the explanation of ‘length’ (cf. LPM 147).

196 Wittgenstein and Waismann

Freedom to choose how to speak about how we speak belongs to the essence of this conception of clarifying our concepts. We have constant need of freedom from being tied down to too limited a morphology of the uses of language (LPP 47; cf. Z §40); freedom to look at things differently; freedom to reject (most?) descriptions of the grammar of our language (e.g. PI §§50, 244, 371, 421); freedom to distinguish concepts (e.g. of classes having the same number of elements (AWL 168)); freedom to depart from entrenched and largely unconscious paradigms of how to describe the use of our words (BT 434; LPP 122). Failure to appreciate this point of view, to acknowledge this form of freedom, is a fundamental blindness in most soi-disant Wittgensteinians. Here at last we can see how to dissolve the paradox which ‘How I See Philosophy’ seemed initially to raise. On this conception (‘our conception’) of the clarification of meaning, ‘grammar’ is invented and voluntary; it is freely negotiated with one’s interlocutor or audience, and it owes its authority to free acknowledgement. ‘The grammar of our language’ is subject to each individual’s decisions about how to use his own words.43 (‘A word has the meaning that someone has given it’ (BB 28)). We need to see how it is used in his language (GA xxiv).44 Correspondingly, we need to appreciate that only recognition, acknowledgement, or decision determines what something means for me (LPM 149). Consequently, grammar is, as it were, the realm of freedom; only in stipulating meanings do we have complete freedom. We are perfectly free to adopt – or to discard – forms of representation. (For this reason, Waismann rejected the notion ‘that philosophy is an exercise of the intellect and that philosophical questions can be settled by argument, and conclusively if one only knew how to set about it’ (HISP 21–2).) On this conception, clarification of the grammar of our own language is liberating; it enhances our individual freedom of thought and promotes vision (it offers us ‘new and broader ways of looking at things’ or ‘a new way of seeing’ (HISP 32)). It opens the way to exploring neglected possibilities or aspects. For example, Einstein’s clarification of simultaneity;45 or Darwin’s ‘dynamic’ concept of biological species. It may be difficult to keep our bearings in respect of this radical conception of the activity of clarifying concepts. But, however unfamiliar it may be, it is undoubtedly central to Wittgenstein’s vision of philosophy in the early 1930s. In fact, it seems to be the immediate corollary of two fundamental ideas. The first is the slogan that essence is expressed by grammar (PI §371). The second is the observation that the distinction between what is essential and what accidental is always one of stipulation or decision, not of discovery (F 39, 83). Hence, it seems, grammar is freely stipulated; it is, as it were, autonomous or arbitrary, answerable to no

A Vision of Philosophy

197

external constraints in reality (DS 3, 16). But by whom is it stipulated? Whose decisions give words their use? Are word-meanings literally created by decisions or stipulations at all? ‘How I See Philosophy’ follows Wittgenstein’s earlier lead in giving a surprising answer to these questions. Grammar is the responsibility of the participants in any particular discussion. It is open to these individuals to negotiate and renegotiate the meanings of their words; they are free to stipulate how their words are to be understood. The sole constraint is that their explanations determine for them (and hence for us too) how the words are to be used or what is to be done within the practice that they set up for themselves. Within those loose limits, there is complete freedom.46 Hence, in this sense, we are all individually free in clarifying the grammar of our words. In respect of grammar, we are more like creators than discoverers.47 ‘Our method’ is grounded in a conception (or concept?) of the clarification of meaning which emphasizes individual freedom in the activity of concept-formation. It goes without saying that we are left perfectly free to reject this conception. Here too the consequence would be that we place ourselves outside the range of those who might benefit from the application of this method of clarifying the grammar of our language. I will finish this section by drawing things together in an U¨bersicht of ‘our method’ for the clarification of grammar. This will consist of two descriptions of how to dissolve philosophical ‘problems’: (a) Having spent much of a term discussing the question of what thinking is, Wittgenstein gave a summary of his procedure by drawing attention to how I reacted to the question with which we started this term: ‘What is thinking?’ In a way I tried to change your point of view: look at it this way. We are inclined to compare some phenomena with something: I ask you to compare them with something else. The question vanished when we classified phenomena not with something happening. We change the concept we have. (LPP 168)

‘Our method’, as it is applied here, is not to impose some categorydiscipline (Ryle); e.g. to pinpoint a mistake in seeing thinking as an activity. Nothing is claimed to be discovered, but an alternative viewpoint is recommended to individuals who are stuck in a rut. ‘The question vanished.’ From where? How? It is not eliminated from the English language by being exposed as an illegitimate (non-existent) move in the language-game. Rather, it is removed from the troubled minds of certain individuals who previously felt driven to ask it. Their

198 Wittgenstein and Waismann

attitude has changed. They no longer want to ask it or to try to answer it. This conception of the dissolution of a question seems to be what Wittgenstein connected with a quotation from Hertz that he originally chose as the motto for PI: ‘our minds, no longer vexed, will cease to ask [the question as to the nature of force]’.48 (‘This question is . . . an expression of mental discomfort’ (BB 26).) The very same conception is conspicuous in his various treatments of Augustine’s question, ‘How is it possible to measure time?’ (b) Waismann addressed the general question: How does clarification of meaning dissolve questions? He replied that it frees us from them by making the meaning of our words and the way they are combined in language so clear that we no longer feel driven to ask the questions. (PLP 4–5)

What is here eliminated is not the question itself (by unmasking it as a pseudo-question (cf. PLP 24)), but rather an individual’s urge to ask it (cf. F 94). This is surely an important distinction! In particular, the criteria of success are wholly different. In the practice of ‘our method’, the discussion brings [someone] gradually to see things in a different light. . . . [He] comes to see that something is wrong with the way he put his question, that the attainment of his object is no longer satisfying. . . . he gives up because he ‘sees’ the question differently. (HISP 20)

This may well be the correct model for what Wittgenstein envisaged as the attainment of complete clarity. Is it not a matter of the will, of attitude, or of modifying what one demands of things, which is needed in order ‘that the philosophical problems should completely disappear’ (PI §133)?

Coda Waismann’s stress on freedom is correlative to a particular conception of personal responsibility for confusions. On this view, philosophical misunderstandings are motivated misconceptions. They are wilful, and we are ourselves responsible for them. ‘If we are taken in by [analogies], it is our fault’ (HISP 19). We cannot escape blame for our own confusions. Hence ‘our method’ embodies an uncompromising – and uncomfortable – moral point of view. On this view, tracing our own confusions to their roots in grammar is not to be taken as exculpation. We are not the hapless victims of

A Vision of Philosophy

199

circumstance; e.g. of syntactic features of English, or even of the grammatical categories of Indo-European languages.50 (As if we couldn’t help but have problems about existence.) Nor are we simply careless and inattentive in ‘theory-building’; liable to make overhasty generalizations on minimal evidence. (So that we must immediately defer to the authority of a more thorough investigation when it is laid out in front of us.) Nor is there a ‘natural’ explanation of our predicament; e.g. the discrepancy between practical know-how and the ability to describe (or explain) what we have mastery of (on the model of the divergence between knowing how to get around a city and being able to draw a useful map (Ryle)). All these familiar responses are feeble excuses. They are attempts to evade responsibility (mauvaise foi!). Waismann’s stress on freedom, and his coupling freedom with responsibility, points towards ‘a moral dimension’ of the activity of clarifying concepts which is conspicuous in Wittgenstein’s work.51 He concerned himself with motivated misconceptions, with deep problems that betray ‘defects of character’. He sought to make others morally uncomfortable with their own muddled thinking. (Wittgenstein’s intention was nothing less than ‘to drive you to self-examination and improvement’.52) In this respect, he exhibited his conviction that ‘the very maximum of what one human being can do for another . . . is to inspire him with concern and unrest’.53 And he recognized that taking responsibility for one’s own confusions calls for enormous courage (CV 52).54 On this view, the sources of our confusions lie in how we want to see things, in what we demand of them (PI §107; BT 407), in how we conceive that they must or cannot be (PI §65; BB 16). One cannot guess how a word functions. One has to look at its use and learn from that. But the difficulty is to remove the prejudice55 which stands in the way of doing that. (PI §340)56

Perhaps too unconscious (often inherited and shared) analogies or pictures; ‘inherited preconceptions’ (HISP 32). For example, the picture of possibility as a shadowy reality (F 57). In this sense, removal of the motivations for misunderstanding requires reformation of personality (intellectual character, or temperament (CV 20)). Like psychoanalysis, the focus of ‘our method’ is revealing to us the motives for our own confusions and exhibiting ways of overcoming them: It is a long superseded idea . . . that the patient suffers from a sort of ignorance, and that if one removes this ignorance by giving him information . . . he is bound to recover. The pathological factor is not his

200 Wittgenstein and Waismann ignorance itself, but the root of this ignorance in his inner resistances; it was they that first called this ignorance into being, and they still maintain it now. The task of the treatment lies in combating these resistances.57

The corresponding philosophical task lies equally out of reach of providing information (details about the workings of a public normative practice). This resemblance in their aims implies a further one: both ‘our method’ and psychoanalysis shift the responsibility for his disease (disorder) on to ‘the patient’. We ourselves are, as it were, responsible for our ‘hang-ups’. And in both cases attaining self-knowledge is conceived as the means for enlarging human freedom!58 The ultimate goal of ‘our method’ is to show how to bring to consciousness our own individual intellectual biases, prejudices, drives, compulsions – how to free our own thinking from their tyranny, how to conduct warfare against our own confusions. This is an intrinsically endless (dynamic) enterprise which enhances our own freedom of thought. ‘How I See Philosophy’ offers a radical, distinctive, and tightly integrated vision of philosophy. It puts emphasis on intellectual discomfort and torment (compulsions or obsessions), on individuals rather than ‘geometrical’ problems, on making unconscious pictures or analogies conscious, on negotiating meanings, on voluntarily coming to see things differently, on assuming responsibility for confusion, and on enlarging human freedom. These elements are woven together into a system of thought and practice: a method demonstrated to good effect by both Waismann and Wittgenstein. This vision has got lost. The moral dimension of ‘our method’ has no place whatever in the sophisticated technology of modern analytic philosophy. There is widespread blindness to the possibility that philosophy might be a positive force in the promotion of freedom and imagination in human thinking. A renewal of Waismann’s vision of philosophy would surely transform the whole post-Wittgensteinian intellectual scene – both the self-image of soi-disant analytic philosophers and their eˆtre pour autrui!

Notes As well as texts that he was given, certainly including parts of PR, the later texts in Part I and most of Part II of PG, a German version of the Brown Book, and a copy of the Blue Book; and probably some parts of BT. 2 Waismann highlighted this phrase by choosing it as the heading for one section of a chapter in LSP. 1

A Vision of Philosophy 3

4 5

6

7 8 9

10 11

12 13

14

201

Arguably it continued to dominate Wittgenstein’s later work. (Cf. Bouwsma’s comment on PPI in Conversations, 32.) I shall not try to document this claim here, though I will draw attention to some relevant passages. Editor’s note: when Baker delivered a version of this paper in Paris, it included a pun: ‘six the`mes’ forming a ‘syste`me’. Compare PLP 13–14 on the recommendation to extend the term ‘grammar’ to include everything about symbols that is fixed before language is applied. A method for dissolving ‘all the problems of philosophy’? These are countless (Drury, The Danger of Words, p. 111). The pretension is modest: ‘[Our method] eliminates those questions which can be eliminated by such a treatment’ (HISP 12; cf. WWK 183). The highest possible achievement is that this method can be demonstrated to be completely effective in some cases, i.e. that somebody’s problem (trouble (BB 46)) may disappear completely. Note: ‘What we do is to bring words back from their metaphysical to their everyday use’ (PI §116). ‘Unsere Methode a¨hnelt in gewissem Sinn der Psychoanalyse’ (dictated in December 1932). Wittgenstein’s conception of philosophy is commonly recognized to be therapeutic (cf. PI §133). But this idea is seldom connected with a recognition that there are many different conceptions of therapy. Different models of administering therapy; especially of the ‘objects’ of treatment (both its subjects and its goals, e.g. Z §382). There is, for example, a tendency to see the ‘private language argument’ as a reductio ad absurdum of Cartesian dualism; i.e. as a strategy for delivering checkmate to the ‘inner/outer’ picture of the mind. This kind of therapy seems comparable to a campaign aimed at the global eradication of a disease (say, smallpox). But this is not the conception of therapy which informs HISP! Bouwsma, Philosophical Essays, p. 26. Consequently, it misses the point that our method focuses primarily on conflicts between competing pictures, and it fails to take note of the extensive negotiation of meaning within many discussions of problems. (See sections (4) and (6) below.) A model might be psychoanalytic treatment of somebody with an inferiority complex. Our method might also address some opposite feelings; e.g. enchantment, fascination, charm, awe, wonder, bewitchment. For example, fascination with the infinite or the unconscious because of its paradoxicality or apparent opacity to reason; or the perverse attraction of Freud’s tracing everything to sexuality because of the repulsiveness of this theory (LA). Clarification of the grammar of ‘infinite’ is meant to dissipate this (pathological!) charm. Cf. BB 95: ‘the idea that its meaning is ‘transcendent’. HISP does not discuss these ‘positive’ feelings (or obsessions?). Boltzmann, Theoretical Physics, p. 137.

202 Wittgenstein and Waismann 15

The same treatment is meted out to Descartes’s comments on the difficulty of breaking habits (at the end of Meditation I). This leads to a complete misconception of the role of the malin ge´nie. 16 Compare: ‘The primary motive force in the therapy is the patient’s suffering and the wish to be cured that arises from it’ (Freud, Complete Works, vol. 12, p. 143). 17 Wittgenstein made this remark about Russell in the 1930s. (It may explain his scorn for academic philosophy and his distaste for philosophical conferences.) 18 Especially because they change dramatically over time – and between intellectual cultures. 19 ‘The enigma of the continuum arises because language misleads us into applying to it a picture that doesn’t fit it’ (PG 471). 20 ‘It is not . . . that one feels any compulsion. The compulsion may be expressed in the words: ‘‘It must be so . . . ’’ ’ (Bouwsma, Philosophical Essays, p. 31). 21 We do not judge the pictures, we judge by means of the pictures. We do not investigate them, we use them to investigate something else (RFM 230). 22

23

24

25

26

27 28

‘In the jargon [of psychoanalysis] it could be said that a simile working unconsciously becomes harmless when it is made explicit.’ Bouwsma stresses this point: ‘The object is to hit upon the analogy that [another] may assent to as what has compelled [him] to say this. The idea is that with this assent the force of the compulsion is broken’ (Philosophical Essays, p. 31). ‘We must become conscious whence we have taken a concept’ (‘Rot und Gru¨n an demselben Ort’, dictation to Schlick). We must ‘work our way back to its roots’ (GA xxii). This echoes psychoanalysis: ‘therapeutic work . . . consists in a large measure in tracing [things] back to the past’ (Freud, Complete Works, vol. 12, p. 152). Just as psychoanalysis involves a new concept of memory (cf. Freud, Complete Works, vol. 12, p. 149). Waismann stated this point forcefully in what might seem to be a non sequitur: ‘Confusion was removed by calling to mind the use of language . . . : it therefore was a confusion about the use of language . . . ’ (HISP 10). Compare: ‘[PI] contains no arguments at all. There are no proofs. It rectifies, nothing. There is nothing to rectify. There are no refutations’ (Bouwsma, Philosophical Essays, p. 24). Compare the scholastic doctrine that God had to create substantial forms before He could create the world! (To prepare niches, only some of which were filled.) Contrast: ‘Everything boils down to surface analogies in language’ (Bouwsma, Philosophical Essays, p. 27). Taken as an example of ‘mythology in the forms of our language’ is the occurrence of the words ‘ghost’, ‘shade’, ‘soul’ and ‘spirit’ in the vocabulary of our language (BT 433–4).

A Vision of Philosophy 29 30 31 32

33

34

35 36 37 38 39

40

41 42 43

203

‘A preference for certain similes . . . underlies far more [philosophical] disagreements than you might think’ (CV 20). Hence, in one clear sense, ‘Philosophy. . . neither explains nor deduces anything’ (PI §126). This might explain why the task of clarifying the grammar of our language is essentially endless (cf. PI §92). ‘We don’t want to say that he cannot explain this to us. He will probably not wish to give this word-combination any sense; just as he could also explain the word ‘‘regular monogon’’, he was free to do so, but he probably didn’t wish to explain it and therefore came to an end with a regular biangle’ (‘Rot und Gru¨n an demselben Ort’). ‘We do not set up the rules of logical grammar, but we ask the other person: ‘‘How do you use this word? Do you follow this rule or that one?’’ ’ (GA xxiv). It is noteworthy that Waismann’s paradigms of clarifying concepts are almost all cases of conceptual revolutions in science and mathematics! For example, Einstein’s redefinition of simultaneity and the definition of the sum of an infinite series (PLP 10–12). ‘Rot und Gru¨n an demselben Ort’. It may seem as if ‘A demon has cast a spell around this position [possibility] and excluded it from our space’ (RFM 56). ‘Only if he acknowledges a particular suggestion is he dealing with a grammatical rule of his language’ (GA xxiv). E.g. Baker and Hacker, Wittgenstein: Rules, Grammar and Necessity, pp. 156–8. In this important respect, our method openly rejects one of the procedural norms of modern science: it does not demand uniformity (consistency) in the use of the terminology employed in describing the grammar of our language. The conception that the meaning of combinations of words ‘depends’ on the meanings of the combined words is, in effect, a stipulation about the use of ‘meaning’ (cf. RFM 107) – one open to replacement by a different conception (e.g. LFM 192). Regarded by the Vienna Circle as a form of verificationism. Cf. ‘Einstein: how a magnitude is measured is what it is’ (PG 459). RFM 135: an alteration in a system of rules; not as it were the discovery of a fact of nature. Failure to appreciate this point blocks any understanding of Wittgenstein’s philosophy of mathematics. Especially the rhetoric of ‘decision’, ‘stipulation’, ‘choice’, ‘ratification’, ‘acknowledgement’, ‘putting in the archives’. (The focus is on the use that individuals make of mathematical symbols; and this is surely not uniform!) This rhetoric is not to be dismissed as metaphorical or hyperbolical! And it is not concerned with the distinction between homework and creative mathematics! (The ‘community view’ of rule-following further darkens understanding of his conception of mathematics.)

204 Wittgenstein and Waismann 44 45 46

47 48 49

In this form, the clarification of concepts should have a major role in the intelligent investigation of the history of philosophy. Fundamental paradigm of clarifying our concepts (PLP) – and manifestly creative. Of course there are broad constraints that constitute the framework within which ‘freedom of speech’ is exercised. But these are not the primary concern of our method. The origins of the intellectual torments of intelligent philosophers are far more subtle than gross violations of the standards of intelligible discourse. And the dissolutions of these problems turn on things within the control of troubled individual thinkers: on how they see things, on what they want to say, on how they choose to explain crucial words. The objects of treatment by our method are always internal conflicts. ‘It was not mistakes, but an urge, a bewitchment, a fascination, a deep disquietude, a captivity, a disorientation, illusions, confusions – these, the troubles of the mixed-up intelligence, that Wittgenstein sought to relieve’ (Bouwsma, Philosophical Essays, p. 28). As Wittgenstein notoriously remarked in respect of mathematics (RFM I. §167). Hertz, Die Prinzipien der Mechanik, p. 9. This would be treatment of a particular intellectual difficulty noted by Boltzmann: What is particularly striking in all this is that the need to ask the question and the tormenting feeling of not finding an answer does not cease once we have recognized that the framing of the question is in itself misguided. (Boltzmann, Theoretical Physics, p. 196)

50 51 52 53 54 55 56 57 58

Rhees suggests that this is Wittgenstein’s ‘mature’ explanation of the source of philosophical confusion (BB Pref. xi). Cf. Drury, The Danger of Words, p. 81. Bouwsma, Philosophical Essays, p. 29. Kierkegaard, quoted in Drury, The Danger of Words, p. 88. Cf. Drury, The Danger of Words, p. 82. ‘We look at the use of these words with some prejudice’ (LPE 233). The importance of combating prejudices in philosophy is a conviction shared with Descartes and Merleau-Ponty. Freud, ‘ ‘‘Wild’’ Psychoanalysis’ (Complete Works, vol. 11, p. 225). The purpose of PI is to help us ‘perfect our skill in the warfare against our own confusions’ (Bouwsma, Philosophical Essays, p. 25).

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

10 Wittgenstein’s Method and Psychoanalysis*

It is known that Wittgenstein conceived of philosophy primarily as a kind of therapy. He offered therapies for particular philosophical problems (PI §§133, 255). Evidently his procedure is rational discussion; he offers a ‘talk-cure’. But what is treated? What are the illnesses? What precisely is the nature of the treatment? What are its goals? There seems to be room for diversity in giving answers to these questions. Even in the case of medicine, we are faced with a choice between different models of therapy. At a minimum there are contrasts between preventative and curative medicine, between the practice of a general practitioner and the conduct of public health campaigns. The targets of these activities are correspondingly different: obesity or lack of aerobic exercise, cancer or smallpox, burns and broken limbs. What is the intended analogy? Which one makes best sense of an activity? This seems an important question to consider, even if any answer is conjectural. The concept of a disorder has a range of applications to non-medical conditions and, correlatively, the concept of therapy extends far beyond the practice of medicine. There are behavioural, affective, sensory, social and intellectual disorders of individuals, and there are all kinds of suggested therapies: drugs, behavioural therapy, ‘cognitive therapy’, etc. Do any of these practices provide an analogy that illuminates Wittgenstein’s philosophical investigations? Some fix on his particular conception of therapy can be derived from his suggestions that there are some similarities between his philosophical investigations and psychoanalysis. This is a muted, but recurrent theme. * Originally published in K. W. M. Fulford et al. (eds), Nature and Narrative, vol. 1 (2003) of the International Perspectives in Psychiatry and Philosophy series from Oxford University Press. Reproduced by permission of Oxford University Press.

206 Wittgenstein and Waismann

He also made definite, though sparing, use of some of Freud’s terminology in the course of his investigations: ‘the unconscious’, ‘acknowledgement’, ‘sublimation’, ‘censorship’ and ‘resistance’. It is, of course, debatable just what similarities he saw and what dissimilarities, and also how much importance is to be attached to comparisons of his philosophical activities with psychoanalysis. ‘Not much’ seems to be the prevalent view,1 but it can scarcely be denied that there is evidence for a contrary assessment. Some of this is internal:2 We can only convict another person of a mistake . . . if he (really) acknowledges this expression as the correct expression of his feeling. For only if he acknowledges it as such, is it the correct expression. (Psychoanalysis.) (BT 410)3 Difficulty of philosophy, not the intellectual difficulty of the sciences, but the difficulty of a change of attitude. Resistances of the will to be overcome. (BT 406)

Other evidence is external: W. had himself talked about philosophy as in certain ways like psychoanalysis . . . When he became a professor at Cambridge he submitted a typescript to the committee. . . . Of 140 pages, 72 were devoted to the idea that philosophy is like psychoanalysis.4

Full clarification of this comparison and resolving the dispute about its importance would be an enormous undertaking even if one could avoid disputes about alternative interpretations of Freud. At a minimum, it would require investigation of the extent of Wittgenstein’s knowledge of Freud, especially of how he understood the methods of psychoanalysis and what he thought to be of interest and value in this practice. It would require thorough investigation of his conception of philosophical problems. Ideally, it would involve a comprehensive survey and detailed analysis of the problems he addressed and the therapeutic methods that he exhibited in his writings, dictations and lectures. And it might even require considering why many analytic philosophers exhibit strong resistance to acknowledging analogies between psychoanalysis and his diverse philosophical investigations. My goal in this paper is more limited: to study in some detail the conceptual investigations which are the immediate context of one remark making an explicit comparison with psychoanalysis. This material occurs in the ‘Diktat fu ¨ r Schlick’:5 a text dictated to Friedrich Waismann, most probably in December 1932. Since the whole text has remained hitherto

Wittgenstein’s Method and Psychoanalysis 207

unpublished and little studied, careful investigation of the remarks in ‘Diktat fu ¨ r Schlick’ (DS 28–30) offers a possibility of throwing some fresh light on Wittgenstein’s therapeutic methods.6 Let’s begin with the methodological remark itself: Our method resembles psychoanalysis in a certain sense. To use its way of putting things, we could say that a simile operating in the unconscious can be made harmless by being articulated. And the comparison with psychoanalysis can be developed even further. (And this analogy is certainly no accident.)7

Six things are immediately noteworthy. First, it is ‘our method’ (not philosophy in general) which is said to have some resemblance with psychoanalysis. Second, the resemblance is claimed to hold only in a certain sense (not in all respects).8 Third, the ostensible topic is the explicit articulation of a simile which is working unconsciously in somebody’s thinking (cf. BT 410). Fourth, it is suggested that a simile may be damaging as long as it is unconscious, whereas it can be rendered harmless by articulating it and acknowledging it as a simile. (This is a specific application of the general therapeutic technique that is characteristic of psychoanalysis.) Fifth, the analogy of ‘our method’ with psychoanalysis holds too in some other (unspecified) respects. Finally, the analogy itself is not accidental, but rather essential to ‘our method’; hence grasping certain respects in which Wittgenstein’s practice resembles psychotherapy is necessary for understanding what he is trying to do. This remark is interpolated into an otherwise uninterrupted discussion. Its placement suggests that the purpose of the remark is to clarify the strategy being pursued in the surrounding text. So the immediate question is how exactly that discussion makes clear that Wittgenstein’s method does indeed resemble psychoanalysis in a certain sense. The immediate topic of discussion is broached a few sentences earlier: If we want to deal with a proposition such as ‘The nothing noths’ or the question ‘Which is prior, nothing or negation?’, then in order to do it justice we ask ourselves ‘What did the author have in mind with this proposition?’ [or] ‘Where did he get this proposition from?’9

The topic is the treatment of a pair of paradigms of philosophical utterances which are taken from Heidegger.10 The first is notorious. And the second one frames a quintessentially philosophical question by making use of the concept of priority. (Other such questions are ‘How is it possible that . . . ?’, ‘What are the real properties of . . . ?’, ‘Of what things can one be absolutely certain?’, etc.)

208 Wittgenstein and Waismann

Given the widespread impression that Wittgenstein’s general business is to frogmarch words back from their metaphysical to their everyday use (PI §116; BT 412), one might expect him to lodge strenuous objections to this pair of propositions. Shouldn’t he condemn Heidegger for uttering nonsense?11 Objections spring immediately to mind. The expression ‘das Nichts ’ (‘the nothing’) is ill-formed: like ‘everything’, ‘nothing’ cannot be combined with the definite article. There is no such verb as ‘nichten’ (‘to noth’). What could it possibly mean to compare ‘nothing’ with ‘something’ in respect of priority? Can ‘negation’ be called some thing? Could a logical operation (negation) be prior to anything other than another logical operation? And so forth. Less pedantically, one might simply make fun of what Heidegger said. ‘When it comes to noth-ing, is it only the nothing that can noth? Or could something noth? Unless something can noth, can there be any such thing as noth-ing? So if the nothing really noths, must the nothing be something? Otherwise there could be no noth-ing. So how then could it be true of the nothing that it noths?’ But these are not Wittgenstein’s procedures. Instead, in order to do justice to such a proposition (um ihm gerecht zu werden),12 we must ask what the author had in mind in framing them, what he might have meant by them, or from what source he took them. These philosophical utterances are patently nonsensical; to the extent that the author is aware of this, it would be pointless to produce an argument in support of that claim. (Heidegger could scarcely have failed to notice this ‘defect’.13) What we need to clarify is what motivates a particular individual to say obviously puzzling things.14 We might do this by getting him to remind himself of how he himself uses words in everyday life – for the specific purpose of encouraging him to reflect on precisely how his philosophical utterances deviate from that pattern. We try to get him to direct his attention to his own motivation. He needs to work out why he feels driven to say what he does. What is pathological in his thinking is not the deviance of his philosophical utterances from everyday speech-patterns, but the unconscious motives which give rise to his behaviour. This strategy is pursued in a distinctive form of investigation – seeking for an unconscious analogy or picture, an unconscious conception or a way of seeing things. Wittgenstein’s diagnosis of a philosophical problem (disquiet) is that a philosopher thinks that to convey an important insight he is compelled to say something which seems, even to himself, empty, self-contradictory or meaningless. He experiences an internal tension or conflict. What needs clarification are the motives which occasion such a conflict.

Wittgenstein’s Method and Psychoanalysis 209 When . . . we disapprove of the expressions of ordinary language (which are after all performing their office), we have got a picture in our heads which conflicts with the picture of our ordinary way of speaking. Whereas we are tempted to say that our way of speaking does not describe the facts as they really are. . . . As if the form of expression were saying something false when the proposition faute de mieux asserted something true. (PI §402)

The suggestion of an unconscious picture as the root of someone’s intellectual trouble clearly requires to be confirmed by this individual’s own acknowledgement. To put this strategy into practice is, in a certain sense, to treat a person seriously and with respect, i.e. to try to make sense of his saying something despite its apparently making no sense at all. Aren’t there parallels with Freud? The psychoanalyst’s task is to uncover the unconscious motives for a patient’s manifestly absurd ideas or behaviour, and to get the patient to acknowledge his motives. This is a difficult and delicate task. In particular, it may require helping the patient to overcome his own resistance. This pattern of investigation dominates the subsequent stretch of ‘Diktat fu ¨ r Schlick’. This explores why (for what reasons) Heidegger expresses himself in an extraordinary way, why he feels driven to run up against the limits of language (PI §119). It looks for a picture that unconsciously shapes his thinking and it seeks to clarify the sources of this picture or the materials from which it has been derived. (Hence this investigation into motives has some general affinity with Freud’s procedure of tracing the origins of neuroses in an individual’s childhood experience.) Wittgenstein first suggests the diagnosis that someone who is tempted to speak as Heidegger does is operating unconsciously with a particular simile. In this case, being or existence is pictured as an island in a sea of nothingness, and this sea is imagined as restlessly eating away at the solid land and dissolving it away. In endlessly throwing up waves and causing erosion, the sea manifests an activity, and that is what Heidegger meant to signify by the verb ‘nichten’ (‘to noth’). (This might be a paradigm of ‘a simile that has been absorbed into the forms of our language’ (PI §112).15) What is the status of this suggestion? How could it be demonstrated that Heidegger must have been influenced by just this picture? (Man kann es garnicht zeigen.) Presumably the individual to whom this suggestion is made must acknowledge the simile as something that did influence his thinking. Only in this case is the diagnosis correct (BT 410; AWL 27, 40). Only is his problem brought to light. Only then have we articulated something that was working unconsciously in his thinking.

210 Wittgenstein and Waismann

His acknowledgement need not be grounded in any memories of actual episodes of reasoning things out; indeed, if the simile was working unconsciously (im Unbewussten), this could not be his ground. Here there seems to be a problem about validation of a statement that has the form of a memory-claim about an individual’s past unconscious motives. (There is a parallel conceptual problem in psychoanalysis.) Does this make Wittgenstein’s diagnosis pointless? Is the suggestion empty? Arguably what underpins my acknowledgement that this simile has been working unconsciously in my thinking is the recognition that reflecting on it now has the power to eradicate the drive or temptation for my making these particular philosophical pronouncements. I am now inclined to say both ‘That’s exactly the way I meant it’ (‘ja, genau so habe ich es gemeint’) and ‘That is completely absurd’ (cf. BT 410). (Here, the manifest absurdity of thinking of non-existence as an activity is indicated by the ironic remark: ‘In this sense of ‘‘activity’’ even resting (das Ruhen) would signify an activity.’) Philosophical theses and questions may be the manifestations of confusion (Verwirrung). If my particular confusion is now dissolved by reflecting on the simile that Wittgenstein proposes, then this discussion has achieved the sole desideratum (das, was wir tun wollten) of a philosophical clarification of concepts, namely eliminating a particular philosophical problem, i.e. a particular individual’s disquiet, unrest or torment, an individual’s tormenting question (qua¨lende Frage). We could imagine Heidegger’s simply dropping the assertion ‘Das Nichts nichtet ’ because it has now lost all its interest or importance for him. This is a very different criterion of success from proving that a question cannot be framed or that a statement cannot be made because it transgresses the bounds of sense.16 In particular, success is strictly personrelative. The problem is somebody’s mental disturbance. Hence philosophical problems, in contrast to all others, can be completely dissolved (BT 421; cf. 431): the individual’s specific disturbance may vanish completely. ‘The problems are dissolved in the literal sense (im eigentlichen Sinne) – like a lump of sugar in water’ (BT 421). In this sense, conceptual analysis is essentially terminable, but other people may develop or continue to have the very same problems – or new ones. For these reasons, the general task of philosophical therapy is, in another sense, essentially endless (cf. Z §447). What seems crucial here is winning someone’s acknowledgement of other possibilities. Pluralism is built into the very idea of a simile or picture: explicitly to regard something as a picture is to recognize that there are alternative conceptions of what is depicted. (This is arguably why articulating an unconscious simile gets rid of its damaging power and

Wittgenstein’s Method and Psychoanalysis 211

renders it harmless. It can no longer exercise tyranny if one acknowledges it as a simile.) If Heidegger were to acknowledge the proposed simile as the root of his thinking, and if he were really uneasy (ungemu ¨ tlich) about saying ‘Das Nichts nichtet ’, then he has available a way out of his difficulty, namely dropping this picture. (This is perhaps easiest to achieve if he sees how to replace the first picture by another picture no less well ‘justified’ by what he says and thinks.) He would then lose the drive to make this philosophical statement. It gave expression to a disquiet that has now vanished away. In what sense does this amount to treatment of a philosophical problem? Understanding why Wittgenstein thinks that a discussion of a picture with an individual is relevant to philosophical investigation depends on clarifying his distinctive conception of philosophical problems. This is exactly what he now proceeds to sketch. ‘It might strike us as strange by what as it were trivial means [durch welche gleichsam trivialen Mittel] we can be freed from deep philosophical disquiets [Beunruhiggungen].’ This does indeed seem remarkable. One might exclaim: ‘How can discarding a mere picture eliminate a conceptual difficulty?’ One might think that more impressive resources must be necessary, something comparable to inventing the theory of types or Russell’s analysis of definite descriptions. It seems that deep problems must require non-trivial solutions. (This idea is expressed in the French proverb: Aux grands maux les grands reme`des.) But ‘against this stands our conception that there is no such thing as a big problem in the intellectual sense’ [im Sinne der Wissenschaft] (BT 407). Wittgenstein tackles this wonder as a sign of confusion. It must be equally astonishing that disquiets may also be eliminated by mere changes in notation.17 In both cases the confusion is to suppose that these means of dissolving difficulties are really trivial. They are only ‘as it were trivial’! We are strongly inclined to speak of ‘mere pictures’ and ‘mere changes in notation’. But these dismissive turns of phrase are misconceived – in respect of many philosophical problems. He discusses two cases of dissolving worries by changes of notation. The first is ‘the problem of identity in diversity’ (TS 220 §99). We say ‘This rose is red.’ Comparing this to the statement ‘Two plus two is four’, we may suppose that the first statement claims that this rose is identical with the colour red. Or we might arrive at the more sophisticated idea that its really being true that the rose is red requires the partial identity of the rose and the colour red. (Even that has obviously paradoxical implications.) Here the problem can be dissolved by introducing two distinct logical symbols, ‘2’ and ‘¼’, to distinguish two senses of ‘is’. The first is used to translate ‘This rose is red’: this statement has the form ‘a 2 A’ and

212 Wittgenstein and Waismann

says that this rose is a member of the set of red things. The second is used to translate ‘Two plus two is four’: this can be written in arithmetical notation as ‘2 þ 2 ¼ 4’. This change of notation frees us from the troubling question: ‘To what extent is this rose the very same thing as the colour red?’. Or, more precisely, our puzzlement is removed provided that we acknowledge that ‘2’ expresses what we meant by ‘is’ in ‘This rose is red’ (PLP 35–6).18 ‘It is characteristic of [a certain kind of philosophical problem] . . . that the questioner is freed from the problem through a certain change in his mode of expression.’19 The second case is the ‘Law of Identity’. The formula ‘A ¼ A’ seems to be a fundamental proposition of logic; it has often been taken to be a logical axiom. But what does it mean? This question seems deep and in a certain sense mysterious. At the same time, ‘A ¼ A’ seems to be utterly stupid and useless (cf. LFM 282–3). Compare this formula with ‘A ¼ B’; that may be construed as a licence to substitute ‘A’ for ‘B’ in any logical formula. But what point could there be in setting up a rule to license substituting ‘A’ for itself ? It is remarkable that we can free ourselves from this disturbing question by introducing a notation which excludes the possibility of the formula ‘A ¼ A’ (TLP 5.53, 5.533). Strange as it may sound, what seemed to us deep, fundamental, a priori about this formula we recognize again in its exclusion from our language by adoption of a new system of signs.20 ‘The deep problem lay, so to speak, simply in the fact that we felt uncomfortable [ungemu ¨ tlich] in the old notation. . . . ’ On this conception, there is nothing in the least astonishing in the observation that a change in notation may eliminate someone’s deep discomforts. What these considerations bring out is Wittgenstein’s distinctive conception of philosophical problems. These are not regarded as abstract puzzles that stand in need of solutions, like chess problems or riddles. They are not conceptual problems which have the accidental property of perplexing and troubling some thinkers. Equally, they are not discomforts which are caused or brought about by conceptual problems.21 Rather they are individuals’ troubled states of mind which have as their intentional objects particular conceptual confusions, tensions, paradoxes or puzzles. A philosophical problem is an individual’s internal conflict: ‘But this isn’t how it is!’ – we say. ‘Yet this is how it has to be!’ (PI §112).22 There are many examples. Frege had the conviction that numbers must be objects if there is any possibility of knowledge in arithmetic, and yet he could see no coherent way to frame logical definitions of numbers as objects. Someone may be disturbed by the impression of disorder in his own concepts (BT 421: ein Bewusstsein der Unordnung in unsern Begriffen) or by an apparent absence of systematicity in rules of inference.23 A person may be

Wittgenstein’s Method and Psychoanalysis 213

seriously irritated at his apparent inability to describe fully his own immediate experience; e.g. his inability to say how many stars or raindrops he now sees.24 He may feel frustrated or exasperated at becoming entangled in his own explanations of word-meanings or at feeling left in the lurch by them (e.g. in trying to explain what ‘A ¼ A’ means). And so on. This point is subtle but crucial. On Wittgenstein’s conception, philosophical problems are deep disquiets (PI §111), feelings of discomfort, torments, irritations, conflicts, etc. He targets ‘philosophical problems, i.e. the particular disquiets of individuals which we call ‘‘philosophical problems’’ ’ (von Wright 1982, Band XI. 35). ‘Our method’ is aimed at getting philosophical problems to disappear completely – in this sense of ‘problem’. ‘As I conduct philosophy, its entire task consists in shaping expressions so that certain disquiets [Beunruhigungen] . . . disappear’ (BT 421). Certain questions give rise to a peculiar mental disquiet, and activity that is philosophy might be characterized as ‘the allaying of this disquiet’.25 This conception of philosophical problems is clear in ‘Diktat fu ¨r Schlick’ even though it is only partially explained there. I have exploited roughly contemporaneous texts to bring it into sharper focus. It is a corollary of this conception of a philosophical problem that philosophical investigation (or the clarification of concepts) must take the form of a discussion with an individual (actual or imaginary). Every problem is someone’s problem, and another’s problem is another problem (cf. FA §27). Hence therapy for confusions is essentially person-relative. It is equally clear that philosophical discussions cannot be adversarial. Suggestions are made, but the interlocutor must acknowledge a picture if it is to be root of his problem (BT 410). Likewise, he must accept and make use of a notation that gets rid of his worry. Everything depends on his free acknowledgement (PLP 36; cf. 148; HISP 18–21). This might leave one wondering how anything is achieved if nothing is proven – or at least why the achievement belongs to philosophy. It is an essential feature of Wittgenstein’s discussions that the interlocutor is free to reject any suggestions made. Someone who makes a philosophical utterance is offered a picture which may inform his thinking, or he is made aware of an alternative way of expressing things. He may be advised to put aside the question ‘What is meaning?’ and to focus instead on the question ‘What’s an explanation of meaning?’ (BB 1). Why not reject these suggestions? They might seem like evasions. He could object that they are not analyses of concepts. The answer indicated here is that the proposals are felt to relieve the problem-troubled individual of his disquiet or torment. His suffering is the motive-force for his accepting a possible escape route when its availability is clear to him.26 (Here is

214 Wittgenstein and Waismann

another respect of analogy with psychoanalysis.) He feels: ‘Thanks be to God! I now see how to drop that burden’ (cf. BT 416). He realizes the possibility of adopting a way of thinking or a notation that no longer leaves him feeling deeply ill at ease [ungemu ¨ tlich]. If I see the possibility of discarding the simile that drives me into saying ‘Das Nichts nichtet ’, or if I see the possibility of replacing that simile by another that doesn’t depict a kind of struggle between being and non-being, then I have found a good reason for laying aside this metaphysical proposition as something without further interest to me. What earlier seemed to me to be gold when I was, as it were, in the magic castle, now in the light of day strikes me as nothing but a piece of rusty metal (cf. CV 11).27 One might describe this enterprise as dissolving ‘the metaphysical aura’ around certain expressions (PLP 81).28 ‘One can’t step twice into the same river.’ ‘Che sera sera.’ ‘What is done cannot be undone.’ ‘Everything is what it is and not another thing.’ ‘Another’s idea is another idea.’ ‘One can never be certain what another thinks or feels.’ ‘There is only one number O.’ The temptation to say such things can be eliminated only by clarifying the motives that lead highly intelligent people into making these useless statements. Suggesting to another person a simile or picture that may have unconsciously guided his thinking is a rather hit-or-miss method of diagnosis. It may be highly effective in some cases; it seems to be a compelling diagnosis of the source of Augustine’s problem of how to measure time (PLP 42). But it calls for imagination and empathy from the philosophertherapist and on for sensitivity and lack of resistance from the problemtormented ‘patient’. (It may fall far short of our ideal of a diagnostic method.29) What else can be done? Wittgenstein now suggests applying his own distinctive method of investigation (unsere Betrachtungsweise). This is to clarify how the individual who formulates a philosophical proposition such as ‘Das Nichts nichtet’ means to make use of it. We should ask him: ‘What inferences can you draw from it? What evidence do you take to support it? What kind of a proposition is it meant to be? Does it have the status of a scientific explanation? Is it a foundation-stone on which other propositions rest? . . . ’. Wittgenstein declares that he will tolerate any answers whatever to these questions; he will take exception to no honest answer, including the negative answers that the proposition has no implications, no grounds, no explanatory status, etc. (Ich erkla¨re mich mit allem einverstanden . . . Ich habe nichts dagegen . . . ) But he insists on knowing precisely what role the proposition has or whether it is an idle wheel in the other’s system of thinking (ob er leer la¨uft). The questions asked are genuine questions calling for sincere answers, and the interrogation

Wittgenstein’s Method and Psychoanalysis 215

is open-ended and exploratory. (Here is yet another parallel with psychoanalysis; the analyst may need to press the patient to answer questions and to cut off evasions, but he should not pass any judgements on the answers given.) The next paragraph envisages the case in which the philosophical proposition turns out to have no real function. It proves to be something similar to a polite formula (Ho¨flichkeitsformel), say an old-fashioned locution for closing business letters (‘I remain, Sir, your most humble and obedient servant’). Propositions of this kind may be found in logical, mathematical and scientific treatises (e.g. the Law of Identity). Here Wittgenstein mentions citations of the law of causality in books on mechanics; this may be included in the preface as the foundation of physics, and yet no further mention is made of it in the body of the work and no consequences are drawn from it. Frege commented on similar propositions included as ‘definitions’ in mathematics texts, and he described them as ‘stucco ornaments’ in contrast to load-bearing members in an intellectual aedifice (PW 212). The final paragraph of this discussion in ‘Diktat fu ¨ r Schlick’ further investigates the motivation driving authors to incorporate such empty propositions into formal, rigorous expositions of their ideas. Thinkers are prey to the illusion (aufgeta¨uscht) that these propositions say things of fundamental importance. They imagine that these principles constitute the foundations (Grundsteine) of their intellectual constructions. The Law of Identity has often been taken to be a basic law of logic.30 Similarly the Principle of Induction or the Uniformity of Nature is often claimed to be the presupposition of all natural science. How can such a radical misconception come about? Wittgenstein offers two very different explanations. The first characterizes philosophical pseudo-propositions as the products of a kind of displacement behaviour. (This explanation builds on an idea which Wittgenstein took from Heinrich Hertz.31) Here the matter is explained by a simile. Suppose that a person usually never eats his fill, i.e. that he rarely fully satisfies his appetite for food. He may then be inclined to take any sensation in his gut to be a form of hunger and to respond to it by eating something, even when the sensation arises in an exceptional case from indigestion occasioned by his having overeaten. In this case his responding by eating serves no purpose – or may even be dysfunctional. We are, Wittgenstein suggests, educated to respond to intellectual discomfort by searching for more fundamental principles. We respond to any felt difficulty by asking for or providing an explanation.32 This habitual response may then occur even when the difficulty is one that cannot be met in this way. That might happen if what bothers us is the apparent

216 Wittgenstein and Waismann

absence of any pattern or system (Regellosigkeit) in what we call ‘rules of inference’33 or ‘propositions’ or ‘languages’. In such cases we ask: ‘What is a rule of inference?’ or ‘What is the general form of the proposition?’34 What we are then likely to produce or accept is something which has the form of an explanation even though it has no real function at all. This way of thinking may culminate in the desire to begin philosophy with something that is the foundation for all the sciences. ‘Here we fall into a similar confusion as the one that could arise from designating as the foundation of a house, at one time the lowest course of bricks, at another time solidity itself.’ According to this explanation, somebody’s producing philosophical pseudo-propositions is something detrimental to his intellectual welfare, i.e. a kind of disease. It prevents his effectively dealing with the sources of his intellectual discomfort. The immediate discomfort in a particular case may be eliminated by reshaping his forms of expression (BT 421). In this way certain vexing questions may disappear completely from his intellectual agenda – because he ceases to want to ask them.35 Therapy for this kind of intellectual dysfunctionality would require teaching another how to differentiate among his various difficulties and how to satisfy his real needs from case to case. The second explanation recognizes the possibility of there being some positive value in philosophers’ nonsense. (This explanation is said to build on architectural ideas expressed by Adolf Loos.36) Stucco ornaments may serve an aesthetic purpose even if not a structural one. They may be requirements of style. Tautologies have this role in structuring arguments; the law of excluded middle may be prefixed to a complicated argument having the form of a constructive dilemma in order to make this pattern conspicuous. In the same way, propositions without any explanatory content may be used to give a particular shape to a body of thought – a shape which is for some reason especially satisfying. There is appeal, for example, in the idea that the whole world is supported by a tortoise, that it is held up by Atlas, or that it was created in six days by God; in form, if not in fact, this closes off the disquieting spectre of an infinite regress. A mere ornament (say a cornice on a house or a wardrobe) has the power to emphasize the limits of what is bounded – and conversely a different pattern of decoration may serve to disguise these limits from view. Either role may yield a deep kind of satisfaction (or dissatisfaction). In this way even a philosophical pseudo-proposition may have an important positive role in somebody’s system of thought. Arguably this was precisely the role of the non-sensical metaphysical propositions of the Tractatus – not to give any kind of justification, but simply to make perspicuous and satisfying the activity of analysis there advocated as the proper activity of philosophizing.

Wittgenstein’s Method and Psychoanalysis 217

What is the upshot of this whole discussion? In particular, what have we learned from his practice about Wittgenstein’s methods for dissolving philosophical problems?37 Three things stand out. First, there is the suggestion that some philosophical utterances can be traced back to similes or pictures that unconsciously shape an author’s thinking. (This might destroy the appearance that these propositions express ‘something deep in the nature of things . . . foundation stones that human thinking can overstep but never displace’ (FA xvi).) Second, there is the suggestion that many philosophical propositions are idle wheels which engage with nothing in the mechanism of thought; they do no real work. (This is a mechanical metaphor which crops up frequently in Wittgenstein’s writings.) They say little or nothing (e.g. PI §13). Yet, despite their vacuity, they impose demands and generate dissatisfaction and confusion. We may feel that we must call the utterance ‘I have toothache’ a description and then agonize over the question of what it describes. In this way, philosophical problems arise when language goes on holiday (freiert), not when we are guided by practical purposes in framing propositions (BT 427). Third, Wittgenstein suggests that philosophical propositions are a kind of displacement behaviour; they attempt to deal with all intellectual disquiets by constructing intellectual explanations (on the model of mathematics or physics), whereas what is needed are alternative pictures to make our own linguistic practices surveyable (u ¨ bersichtlich) – on the model of depicting intensity of conviction by speech-intonation (DS 21). He urges the importance of acknowledging other possibilities for the purpose of breaking the grip of specific prejudices. All three of these activities are evidently meant to be ‘clarifications of concepts’ or ‘descriptions of the grammar of our language’. But this activity takes a very distinctive and unexpected form. Wittgenstein engages in sympathetic, open-ended discussion which demands the interlocutor’s active participation. There is no attempt to assemble a dossier of grammatical facts, and no attempt to frame adversarial, coercive arguments. Discussion is strictly person-relative, and there is scrupulous respect for the interlocutor’s freedom. The goal is to change his ways of thinking – with his consent (cf. HISP 20–1). In these respects, the practice of ‘our method’ is very far removed from the paradigms of ‘conceptual analysis’ to be found in Carnap38 and Ryle.39 Let us now go back to the beginning. What are the resemblances of ‘our method’ with psychoanalysis? There is one quite general one: the purpose of providing therapy for individuals’ troubled states of mind (Beunruhigungen). But there are also resemblances in certain specific respects. Three ¨ r Schlick’. Arguably all three of them are to be are evident in ‘Diktat fu

218 Wittgenstein and Waismann

regarded as essential to the practice of ‘our method’; i.e. they are constitutive of the identity of this distinctive practice of philosophizing. First, the primary concern is helping individuals to bring to consciousness their own motivations and desires. Philosophical utterances which are often patently absurd are taken to manifest unconscious cravings, drives, prejudices and pictures. It is these things which are the targets of Wittgenstein’s therapeutic activities, not the utterances generated by them. An individual needs to acknowledge a picture that holds him captive (PI §115) or prevents his seeing clearly his own use of words (PI §§5, 110). He needs to become aware of his own cravings (BB 17–18), his urges to misunderstand the workings of language (PI §109),40 or his strong desires to see things in particular ways (BT 406–7). Wittgenstein’s purpose is to win acknowledgement from suffering individuals of what unconsciously influences them. Proof and refutation are just as alien to his investigations as they are to Freud’s practice of psychoanalysis. Second, Wittgenstein’s writings resemble Freud’s in respect of teaching a method by means of detailed case-studies. ‘Demonstrating a method by means of examples, and the series of examples can be broken off ’ (PI §133). The only product of the teaching is imparting a skill. (Results are of limited interest. Heidegger’s confusions are unlikely ever to be repeated exactly.) Case-studies are of interest in respect of the methods exhibited in the treatment of absolutely specific individual difficulties: what is demonstrated are procedures for untying the particular knots that someone has tied in his own thinking (PR 52; BT 422) or for putting a derailed wagon back precisely on the rails so that it can roll (BT 410). The importance of the methods is independent of the generality of the difficulties. It is in precisely this respect that ‘our conception [is] that there is no such thing as big, essential problems in the sense in which such problems arise in intellectual disciplines’ (BT 407). Third, ‘our method’ is intended to be therapeutic in a strong sense: the overarching concern of therapy is with enhancing human welfare. Intellectual torment signals intellectual disease (confusion, prejudice). Like pain or neurotic discomfort, it is a form of ill-fare, hence it calls for treatment (cf. PI §599). The programme is to make the unconscious conscious in order to break its thrall and render it harmless (unscha¨dlich); it is to break the grip of prejudices or superstitions. The target is not opinions (or mistakes), but ways of thinking. The aim of Wittgenstein’s therapy is to increase freedom – just as psychoanalysis aims to set patients free from neurotic behaviour-cycles. We cannot engage in critical investigations of aspects of our own thinking of which we are unaware (cf. PI §129). We may be held captive by pictures (§115). We may get stuck in pointless repetition: ‘But this is how it is ——’, I say to myself over and

Wittgenstein’s Method and Psychoanalysis 219

over again (PI §113). The goal is to eliminate neurotic confusion, as it were; but not ordinary confusion resulting from laziness, inattention, ambiguity, vagueness, fallacious inferences, etc.41 Accepting Wittgenstein’s methods of therapy as a form of philosophical investigation presupposes reconceptualizing the boundary between logic and psychology.42 On his view, dealing with compulsions, obsessions, prejudices, torments, . . . is the proper business of philosophy. So too is facilitating or effecting changes in ways of thinking about things. Discussion is less a matter of constructing rigorous arguments from secure premises than of making propaganda for alternative points of view. It is subversive in aiming to shake up entrenched habits of thought (Denkgewohnheiten) (BT 423) – including prejudices about the nature of philosophical argument and philosophical problems. Hence the method that Wittgenstein sought to demonstrate clearly in his practice is likely to encounter yet further severe resistance among analytic philosophers!43 What Wittgenstein offers us is a distinctive conception of the method and purpose of concept-clarification.44 His discussion of Heidegger in ‘Diktat fu ¨ r Schlick’ shows precisely how ‘our method is, in a certain sense, similar to psychoanalysis’. We need to do justice to this statement by clarifying what he had in mind.45 Arguably this particular conception of therapy – the essence of ‘our method’ – runs through and unifies all of his later philosophy. But making a thorough case for that ambitious claim would need book-length treatment. Notes 1 2

3

4

5 6 7

Hacker (Wittgenstein: Meaning and Mind, pp. 90–2) lists a number of respects in which the analogy is illuminating. In all quotations, italics will be used to mark the author’s own emphasis, and bold will mark emphases that I have added. Translations are the author’s own; occasionally they modify published ones. The abbreviation ‘BT’ in the text refers to the ‘Big Typescript’, TS 213 in von Wright, Wittgenstein (1982). References are to section numbers when prefixed with the sign ‘‘§’’, otherwise to page numbers. This is part of a reported conversation in August 1949 (Bouwsma, Conversations, p. 36). The typescript referred to here is the pre-war version of Philosophical Investigations (TS 220). This is TS 302. The original German text together with a translation into English appears in Waismann, Voices of Wittgenstein. ¨ hnelt in gewissem Sinn der Psychoanalyse. In Original text: ‘Unsere Methode a ihrer Ausdrucksweise ko¨nnte man sagen, dass im Unbewusstsein wirkende

220 Wittgenstein and Waismann

8 9

10 11 12

13 14

15

16

17

¨ dlich, wenn es ausgesprochen wird. Und dieser Vergleich Gleichnis wird unscha ¨ sst sich noch weithin fortsetzen. (Und diese Analogie ist gewiss mit der Analyse la kein Zufall.)’ What are the respects in which the analogy does not hold? This important question is not investigated here. Original text: ‘Wenn wir einen Satz wie den ‘‘das Nichts nichtet’’ oder die Frage ‘‘was ist fru ¨ her, das Nichts oder die Verneinung?’’ behandeln wollen, so fragen wir uns, um ihm gerecht zu werden: was hat dem Autor bei diesem Satz vorgeschwebt? Woher hat er diesen Satz genommen?’ Both occur in his inaugural lecture ‘What is Metaphysics?’ given at Freiburg in 1929. ¨ berThis is precisely what Rudolf Carnap did in some detail in the paper ‘U windung der Metaphysik durch logische Analyse der Sprache’ (1932). This purpose is prominent in the ‘Big Typescript’: ‘Our sole task is to be just [ gerecht zu sein] (BT 420); ‘The goal . . . justice [Gerechtigkeit] (BT 414). Later he indicated the need to avoid injustice [Ungerechtigkeit] in his assertions (PI §131). In fact, as Carnap noted, Heidegger turns this ‘logical’ absurdity into a condemnation of the sovereignty of ‘logic’. This strategy is pursued in Wittgenstein (PI §§38–9). It is noted that ‘strange to say, the word ‘‘this’’ has been called the only genuine name’. (This seems a clear allusion to Russell.) Reasons are listed why the use of ‘this’ does not fall within the gamut of word-uses that are characterized by the term ‘name’ (§38c). ‘But why does it occur to one to want to make precisely this word into a name, when it evidently is not a name? – This is just the reason. For one is tempted to make an objection against what is ordinarily called a name (§39). A suggestion is then made: ‘a name really ought to signify a simple’. But this will be a correct diagnosis of the motivation only if it is acknowledged by the imagined interlocutor. The emphasis of this whole passage is not on showing that the philosophical statement is mistaken, but on clarifying an individual’s motivation for making such a patently strange claim. Analogues might be the picture of the sequence of events as logs floating downstream in a river and passing under a bridge (BB 107–8); the simile of immediate experience as the image cast on a screen by a film strip passing through a projector (BT 428); or the conception of measurement as juxtaposition of two things (BB 26; PLP 42). Carnap’s explicit purpose is to show how Heidegger’s questions and assertions violate logical syntax; in a logically adequate language, they could not be expressed. If his argument is cogent, its conclusion is impersonal, a theorem about the ‘geometry’ of language. It is sometimes asserted that he holds the opposite view, as if, in describing our conceptual landscape, he endorsed the objection: ‘What you want is only a new notation, and by a new notation no facts of geography are changed’ (BB 57).

Wittgenstein’s Method and Psychoanalysis 221 18

19 20

21

22

23 24 25 26 27 28 29 30

31 32

33 34

Cf. PLP 148: The rule that a speaker is now following might be one to which he assents when we suggest it to him. In this case it is a revocable decision which validates our taking this to be the rule that he is following (cf. BB 82). F 34: Die Zeichenregel ist dann die Regel, zu der Sprechende sich bekennt. Dieses Sich-zu-einer-Regel-bekennen ist auch das Ende einer philosophischen Untersuchung. Wenn man z. B. die Skrupeln u ¨ ber das Wort ‘ist’ dadurch ¨ umt hat, dass man den Menschen 2 oder 3 verschiedene Zeichen statt weggera des einen zur Verfu ¨ gung stellt, so hinge nun alles davon ab, dass er sich zu dieser Regel bekennt, 2 sei durch ¼ nicht zu ersetzen. von Wright 1982, Band XI (MS 115) 36. This thought might be generalized. If essence is expressed by grammar (PI §372), and if every rule of grammar can be made to disappear by adopting a suitable notation (WWK 239–41), then what makes any proposition the expression of something fundamental and a priori is the possibility of eliminating it completely from language. For this reason, they are wholly unlike headaches. (That comparison is frequently drawn in order to make fun of the idea that Wittgenstein’s therapy is directed at troubled individuals. It seems to inform Ryle’s often quoted remark that Wittgenstein did not have ‘a soothing bedside manner’.) It is common to confuse the objects of emotions with their causes, and thereby to take the relation of an emotion to its object to be external or hypothetical. Wittgenstein emphasized the importance of keeping this distinction clearly in focus. Waismann, Voices of Wittgenstein, §16: ‘die Regellosigkeit’. ¨ nomenale Sprache’. Waismann, Voices of Wittgenstein, §62: ‘Pha ‘Philosophie ist die Stillung dieser Unruhe’ (GA). This point is neglected in Ryle’s ironic question, ‘Why should the fly take the way out of the flybottle when it is pointed out to him?’ This is an application of this simile opposite to the self-effacing one that Wittgenstein made, but it seems no less apposite. One might add that it is in this sense that what we do is to bring words back from their metaphysical to their everyday use (PI §116). That ideal may itself call for therapeutic treatment (cf. PI §§100–8). Leibniz called identities (e.g. ‘Whatever is, is’; ‘Each thing is what it is’; ‘A is A’, etc.) the primary truths of reason, and he showed how to demonstrate the inferences of logic by means of identity (New Essays, pp. 361–3, 366). The 1938 version of PI had a motto from Hertz’s Die Prinzipien der Mechanik (1910) which gave powerful expression to this idea. TS 219 §8: what is disastrous in the scientific way of thinking (which today rules the whole world) is that it wants to respond to every discomfort by giving an explanation. This problem is discussed in the dictation ‘Rechtfertigung der Grammatik’ (Waismann, Voices of Wittgenstein). Contemporary philosophers ask obsessively ‘What is consciousness?’

222 Wittgenstein and Waismann 35

36 37

38 39 40 41 42

43 44

45

This is precisely the message of the motto from Hertz. Even a physicist may ask ‘What is force?’ or ‘What is electricity?’; or he may state that the nature of force is a mystery. In doing so, he seems to be expressing some kind of dissatisfaction or bewilderment in respect of the complexity of the scientific knowledge that he has. ‘When these painful contradictions are removed, the question as to the nature [of force] will not have been answered; but our minds, no longer vexed, will cease to ask illegitimate questions’ (Hertz). In this sense, one might say, ‘offenbar irrt die Frage in Bezug auf die Antwort, welche sie erwartet ’ [evidently the question is mistaken in respect to the answer which it invites] (BB 169). To that extent the question is illegitimate [unberechtigte]. Loos occurs in listings of the thinkers who have most influenced his own thinking (CV 19). His methods might have fruitful application to resolving other kinds of problems such as entrenched disagreements about matters moral, aesthetic or political. Cora Diamond has made some intriguing suggestions along these lines, but such wider possibilities are not explored here. This contrast is particularly striking in respect of their treatments of Heidegger’s metaphysical uses of words. The differences are often thought to be mostly stylistic, the similarities profound. Cf. Kenny, The Metaphysics of Mind. These might be what is flagged by the phrase ‘grammatical illusions’ (PI §110); and the results may be what are called ‘grammatical fictions’ (§307). So too, Freud treats neurotic distress, not everyday unhappiness. It might be better to recognize the variety of boundaries drawn by philosophers between logic and psychology. For example, Frege denied that modality is relevant to discriminating logical forms of judgement or inference, whereas TLP is built on the principle that every significant proposition must be contingent, so that any ‘necessary truth’ must be logically unlike any contingent proposition, and any inference in logic or in mathematics must be completely different from inferences relating significant propositions (TLP 6.1263). It has some real affinities with the methods exhibited in the work of Nietzsche and Merleau-Ponty. Waismann expounded a somewhat similar conception in his article ‘How I See Philosophy’. Much of the description of method (but not the particular examples considered for treatment) is closely based on material that Wittgenstein dictated to him in the early 1930s. However, Waismann says nothing about this pedigree, nor does he claim that the method there advocated is (or was) Wittgenstein’s. It would be a moral defect in us to make fun of this statement along the lines that Carnap makes fun of Heidegger.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

B Aspects and Conceptions

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

11 Italics in Wittgenstein*

Even the most casual and cursory look at the text of the Philosophical Investigations, Part I, makes clear that Wittgenstein made very copious use of italics in this most polished of his later texts. But this practice is equally conspicuous in all his writings, both typescripts and manuscripts. It is even characteristic of texts that he dictated. This raises a problem: How do italics contribute to the intended interpretation of a remark? What do they add? What would be the effect of deleting them? The painstaking care he took in framing his thoughts, his obsessive redrafting of remarks in subtly modified forms, suggest that his italics are not gratuitous ornaments, not at least in any relatively polished text. He must have thought that his use of italics always makes a difference, and he would presumably have been able to explain in each case what difference is made. Can we now make significant headway in clarifying this aspect of his writings? Or must we resign ourselves to being unable to decipher this typographical device? It is noteworthy that italics occur in many remarks that have become the focus of discussion and controversy. It is also noteworthy that commentators pay little attention to this point; at best they observe that the whole remark is emphatic.1 The immediate aim of this paper is to bring to light some general patterns in Wittgenstein’s use of italics; to clarify this single aspect of the grammar of Wittgenstein-Speak. We will discover that italics have a number of fairly definite roles. Once we have identified these patterns, we may learn to spot them in cases where they are not readily apparent to most modern readers. The more remote aim is to suggest some new and better-supported interpretations of a few of his most celebrated * Originally published in Language & Communication, 19 (1999), pp. 181–211. ß 1999 by Elsevier Science Ltd. Reprinted by permission.

Italics in Wittgenstein 225

remarks. This might be seen as some opening moves in a new research programme.

The Problem Of course all readers of his work are familiar with the fact that italics (or ‘emphasis’) occurs frequently in his remarks. This usually takes the form of underlining in his manuscripts and typescripts, but there is occasional use of other equivalent devices: capitalizing an entire word (PI §§253, 263, 398; RFM 183, 330), multiple underlining, or the German stenographic practice of spacing out the letters of a word (‘Sperrdruck’).2 Italics in his writings might be treated as the written counterpart of a well-known feature of his speech: he had the habit of putting very considerable stress on particular words or phrases.3 This mannerism proved infectious with some of his ‘disciples’. It has given much leverage for caricature and mockery (of him and of them); as if a random exaggerated stress on such words as ‘‘this’’ or ‘‘that’’ might suffice by itself to create an aura of profundity. What may not have struck most readers, however, is the frequency with which italics crop up in remarks that seem especially important or pivotal. Here is a small sample: §585: When someone says ‘‘I hope he’ll come’’ – is this a report [Bericht] ¨ usserung] of his hope? about his state of mind, or a manifestation [A §199: To obey a rule, to make a report, to give an order, to play a game of chess, are customs (uses, institutions). §116: What we do is to bring words back from their metaphysical to their everyday use. §109: We must do away with all explanation, and description alone must take its place. §66: Don’t say: ‘‘There must be something common, or they would not be called ‘games’ ’’. . . . §307:

If I do speak of a fiction, then it is of a grammatical fiction.

§371:

Essence is expressed by grammar.

§69: How should we explain to someone what a game is? I imagine that we should describe games to him, and we might add: ‘‘This and similar things are called ‘games’.’’ ¨ bersichtlichBT 417: Above all, our grammar is lacking in perspicuity [U keit].

226 Wittgenstein and Waismann §122: A main source of our failure to understand is that we do not command a clear view [u ¨ bersehen] the use of our words. PI p. 226: life.

What has to be accepted, the given, is one could say – forms of

Z §651: Shrugging the shoulders, shaking the head, nods and so on we call signs first and foremost because they are embedded in the use of our word-language [Wortsprache]. PI §404: . . . there is a great variety of criteria for personal ‘identity’ ¨ t’ der Person]. [Kriterien der ‘Identita Z §477: What is common to sense-experiences? – The answer that they give us knowledge of the external world is partly wrong and partly right. It is right in so far as it is supposed to point to a logical criterion.

Commentators seem almost uniformly to take the italics in all of these remarks simply to emphasize or underline something.4 On this view, their role is uniform and comparable to prefacing the remark with an expression like ‘‘Now hear this!’’ or ‘‘Oyez! Oyez!’’; or perhaps to supplementing the italicized expression with some adverb such as ‘‘definitely’’, ‘‘certainly’’, ‘‘unequivocally’’, or ‘‘literally’’ (‘‘au pied de la lettre’’). The effect of italics is taken to be his insisting on the relevance of all the connotations or implications of the italicized expression. The possibility that italics serve other functions is not even considered, let alone rejected for good reasons (which would surely have to vary from case to case). As a thought-experiment, just imagine that Wittgenstein had adopted some more Derridean device instead of italics, say writing the expression with a line through it (cancellation) or giving the general instruction that anything underlined be printed upside down.5 How much would then be left intact of what we now consider the Leitmotiven of ‘‘Wittgenstein’s later philosophy’’? At the very least, might italics not affect the spirit in which his remarks are to be taken? The currency of the term ‘‘emphasis’’ suggests widespread commitment to the idea that italics in writing simply deputize for stress or emphasis in speech. This idea should undoubtedly not go unchallenged.6 But even its acceptance indicates the need for careful study of italics. Stress in speech has many different forms and many different functions. (Think of Churchill’s line: ‘‘Some chicken! Some neck!’’) Hence the putative correspondence of italics with stress would at best be a paper draft on interpreting a remark, not a complete explanation of its intended significance.

Italics in Wittgenstein 227

Preliminary Patterns In some cases, italics in Wittgenstein’s texts seem to have some of the more commonplace roles of spoken stress (even if these are difficult to survey or to catalogue).7 Here is a sample of the variety of uses emphases may have: Stress on a point that is apt to be neglected, or even expected to give rise to objections: for example, §28:

. . . an ostensive definition can be variously interpreted in every case.

§23: . . . how many different kinds of sentences are there? . . . – There are countless kinds: countless different kinds of use of what we call ‘‘symbols’’, ‘‘words’’, ‘‘sentences’’. §16: What about the colour samples that A shows to B: are they part of the language?

Italics may indicate the focus of a remark. Perhaps a particular concept: for instance, §244:

How do words refer to sensations?

§291:

What we call ‘‘descriptions’’ are instruments for particular uses.

Or an aspect of something: Z §527: . . . the picture of the process of calculating as, so to speak, submerged and going on under the surface. LPE 302: I am saying that there is nothing behind the moaning [when another expresses pain]. BT 81: The sense [of a sentence] does not lie behind it (as does the mental process of imagining something, etc.).

Perhaps a contrast between two concepts: for example, §585: When someone says ‘‘I hope he’ll come’’ – is this a report [Bericht] about his state of mind, or a manifestation [A¨usserung] of his hope? §303: ‘‘I can only believe that someone else is in pain, but I know it if I am.’’ PG 66:

I am only describing language, not explaining anything.

228 Wittgenstein and Waismann

Sometimes a contrast between singular and plural: e.g. §§36, 69, 132, 133. Italics may clarify a back-reference: for instance, §129:

Unless that fact has at some time struck him.

They may focus attention on the positive connotations or associations of a word or phrase. §87: It may easily look as if every doubt merely revealed an existing gap in the foundations. . . . 8

Equally, italics may focus attention on the negative implications of a word or on the absence of a contrary: for example, §43: For a large class of cases – though not for all – in which we employ the word ‘‘meaning’’. . . . §587: Does it make sense to ask ‘‘How do you know that you believe?’’ – and is the answer: ‘‘I know it by introspection’’? In some cases it will be possible to say some such thing, in most not.

Italics may be used to highlight a pattern within a complex remark. For example, drawing attention to detailed points of contrast between closely related sentences: §20: I think we shall be inclined to say: we mean the sentence [‘‘Bring me a slab’’] as four words when we use it in contrast with other sentences such as ‘‘Hand me a slab’’, ‘‘Bring him a slab’’, ‘‘Bring two slabs’’, etc.; that is, in contrast with other sentences containing the separate words of our command in other combinations.

No doubt, in a large class of cases, italics can be said to indicate emphasis. Hence they may have any of the various different functions that emphasis may have in speech. (Their precise function must be worked out from case to case.) But some occurrences seem to exemplify different and more idiosyncratic patterns. Here are some samples: (1) One (with many historical precedents9) is the italicization of everyday expressions of which an author gives his own specific explanation. This device is often employed at the place where the definition is given.

Italics in Wittgenstein 229 BB 13: We shall say that the rule is involved in the understanding, obeying, etc. if . . . the symbol of the rule forms part of the calculation. PG 54: The comparison suggests itself that the word ‘‘is’’ in different cases has different meaning-bodies [Bedeutungsko¨rper] behind it; that it is each time a square surface, but in one case it is the end surface of a prism and in the other the end surface of a pyramid. Imagine the following case . . . [An explanation follows.] F 39: By grammatical rule I understand every rule that relates to the application of language. LSP 43–4: By grammar we understand everything about language which can be put in place before the language is applied at all. The antonym of this will thus be the application of language. . . . If the word ‘‘grammar’’ is taken in this extended sense, . . . [Unter Grammatik verstehen wir alles das an der Sprache, was man festlegen kann, noch bevor man die Sprache anwendet. Der Gegensatz hierzu wird dann die Anwendung der Sprache sein. . . . Nimmt man das Wort ‘‘Grammatik’’ in diesem weiten Sinn, . . . ]10

Italics may subsequently be used to secure back-reference to such an explanation; namely, as equivalents of ‘‘in this sense’’, ‘‘in a certain sense’’, or ‘‘in our sense’’. One clear example of this extended pattern is the lengthy discussion of reading (§§156–71; BB 119–25). This opens with a special explanation: §156: First I need to remark that I am not counting the understanding of what is read as part of ‘reading’ for the purposes of this investigation: reading is here the activity of rendering out loud what is written or printed; and also of writing from dictation, writing out something printed, playing from a score, and so on. (cf. BB 119)

The following eight sections sporadically italicize the word ‘‘read’’ (once in each of §§156–62), sometimes (even if not always11) to remind us to tie down the subsequent remarks to this explanation. The whole discussion makes sense only when pinned down to this particular sense of ‘‘read’’. For example, §159: But when we think the matter over we are tempted to say: the one real criterion for anybody’s reading is the conscious act of reading, the act of reading the sounds off from the letters.

This point holds true even after the italics peter out. For example, §166 opens:

230 Wittgenstein and Waismann I said that when one reads the spoken words come ‘in a special way’.

And §169 opens: But when we read, don’t we feel the word-shapes somehow causing our utterance?

(We might say that the scope of the italics extends over these subsequent occurrences of ‘‘read’’.) Sometimes this use of italics occurs where a special definition is implied rather than expressly formulated. RFM 123: Here one needs to remember that the propositions of logic are so constructed as to have no application as information in practice. So it could very well be said that they were not propositions at all. . . .

This restrictive use of proposition12 is paralleled by other different ones, each related to a context-specific explanation which is either given or understood (§§135–6; cf. RFM 117, 162–4; cf. PLP 280–303 (esp. 290); WWK 106–7). This function of italics occurs in various cases of introducing or using Wittgenstein-speak (his ‘jargon’). This is important for understanding these remarks. For example, BB 24–5: To the question ‘‘How do you know that so-and-so is the case?’’, we sometimes answer by giving ‘criteria’ and sometimes by giving ‘symptoms’. [A case of giving a definition.] LPE 252: When we talk of the private experience which the others don’t know, we don’t originally mean to talk of a shapeless nothing but of a variable with a certain definite value. [A back-reference to a special glossing of ‘‘private’’.]

Italics are certainly not necessary to this practice, especially where the newly explained expression is immediately applied in this special sense, with or without the rider ‘‘in this sense’’ or ‘‘in our sense’’. For example, DS 3: One could put the point roughly like this: we do not step beyond the bounds of language in giving any explanation of what a word means, not even in an ostensive definition, and language is in this sense autonomous. [Man ko¨nnte es beila¨ufig so ausdru¨cken, dass wir durch keinerlei Erkla¨rung der Wortbedeutungen, auch nicht durch die hinweisende, aus der Sprache hinaustreten, und dass die Sprache in diesem Sinn autonom ist.]

Italics in Wittgenstein 231 DS 16: We say that grammar determines which word-combinations have sense and which do not; but on the other hand too, grammar is not answerable to any reality, [so] it is in a certain sense arbitrary. [Wir sagen, die Grammatik bestimme, welche Wortzusammenstellungen Sinn haben und welche nicht; andererseits aber auch, die Grammatik sei keiner Wirklichkeit verantwortlich, sie sei in gewissem Sinn willku¨rlich.]

(Omission of italics in such cases may be a particular characteristic of texts transcribed from dictation.) This use of italics is the counterpart of a familiar scholastic practice. First the author gives an ad hoc explanation: for example, By the word ‘‘thought’’ [‘‘cogitatio’’] I understand . . . And that is why not only understanding, willing and imagining, but also sense-perceiving [sentire] are here the same thing as thinking.13

Then explicit back-reference to this explanation is signalled by some phrase such as ‘‘in this sense’’, ‘‘properly so-called’’, ‘‘strictly speaking’’, etc.: for instance, The word ‘‘sense-perceive’’ [‘‘sentire’’] in this restricted sense [praecise sic sumptum] is nothing other than a mode of thinking.14

Ignoring these purpose-specific explanations or back-references to them is apt to yield flagrant misinterpretations of remarks whose intended meaning is fully clarified. (For instance, the unqualified assertion that Descartes in Meditationes II and Principia I. 9 extends the term ‘‘cogitare’’ to include ‘‘sentire’’ (sense-perception).) The scholastic tradition insisted on the importance of making such local distinctions and of attending to them in interpreting texts, i.e. of scrupulously avoiding the fallacy of equivocation. But where scholastics signalled their ‘disambiguations’ by such phrases as ‘praecise’ or ‘praecise sic sumptum’, Wittgenstein often used the more economical device of italicizing the relevant expression.15 His practice of constantly redrafting and reorganizing remarks might be expected to yield less clear (or defective) instances of this use of italics. In particular, it might happen that his intention is to attach a special sense to an italicized expression, even though the present version of the text where the remark occurs fails to give much if anything of a prior explanation.16 One arguable case in PI is the use of italics apparently to invoke his own conception of grammar: §307:

If I do speak of a fiction, then it is of a grammatical fiction.

232 Wittgenstein and Waismann

Would this remark make sense to any reader unfamiliar with his rather unorthodox use of ‘‘grammar’’? The italics might indicate the need to interpret ‘‘grammatical’’ by reference to his various explanations of the term, though, as it happens, these are given only in other texts (e.g. F 39). A parallel case from PI, perhaps even clearer, is the use of italics in ¨ bersicht. The expression u references to his particular conception of U ¨ ber¨ bersicht and U ¨ bersehen (§122), for instance, picks up a conception of U sichtlichkeit that is developed and illustrated in various scattered texts (esp. GB). Hence italics flag the first introduction of this conception in ¨ bersichtlichkeit. (Once italics have been introeach text; e.g. BT 417: U duced in this role, subsequent uses of a word or its cognates need not be italicized in the same remark since it is natural for the reader to relate them to the initial italicized occurrence. For example, §122: u ¨ bersehen governs ¨ ‘‘Ubersichtlichkeit’’ and ‘‘u¨bersichtliche Darstellung’’; as it were, the scope of the italics includes all three expressions.17 That would explain why the italics present in BT 417 (Above all, our grammar is lacking in ¨ bersichtlichkeit]) were omitted in its redrafting in §122.) perspicuity [U ¨ usserung in PI. For instance: Another clear example is the expression A §585: When someone says ‘‘I hope he’ll come’’ – is this a report about his ¨ usserung] of his hope? state of mind, or a manifestation [A Z §53: The statement ‘‘I am expecting a bang at any moment’’ is an ¨ usserung] of expectation. expression [A

This everyday expression Wittgenstein hijacked to express his own distinctive conception.18 Here, presumably, the explanation to be supplied turns on regarding a first-person expression of hope or expectation as a manifestation of an ‘inner state’, hence on the possibility of regarding these utterances as learned replacements for primitive or natural reactions ¨ usserung is (cf. §244). According to this explanation (absent from PI), A to be contrasted with statement [Behauptung] (Z §549) and report [Bericht] (§585). (A different but related conception is attached to PG 66.) (2) Another quite different use of italics is to indicate the need for the reader to supply a special explanation that is neither peculiar to the author nor explicit in the author’s text. This might be conceived as a distant cousin of the previous use. In that case, italics mean roughly the same as the phrase ‘‘in my sense’’ (or ‘‘in our sense’’); and the onus is ultimately on the author to explain this. In this case, by contrast, italics mean ‘‘in a certain sense’’ or ‘‘in some sense’’ or even ‘‘one might say’’. The reader is then landed with the tasks of identifying which sense is intended and of supplying the appropriate explanation. The onus is on him to work out

Italics in Wittgenstein 233

somehow what is relevant to interpreting the remark. Here, obviously, it is more debatable what the intended interpretation is, and adversarial disputes may defy resolution by discussion. In some cases italics may indicate that an explanation is to be supplied from another source (a particular author or even a specific set of authors). Consider: §256: And now I simply associate names with sensations and use these names in descriptions. –

This remark seems to invoke a conception of bare association that is common in a certain tradition of psychology.19 More commonly, the required explanation must be generated by the reader from his own resources. As in these two cases: §219:

I follow the rule blindly.

PI p. 226: life.

What has to be accepted, the given, is – one could say – forms of

The interpretation of such remarks calls for sensitivity, imagination, and judgement. The most important guide is clues given by the author. In the first example, blindly seems tied to the adjacent comment: When I follow the rule, I do not choose.

Also to the neighbouring comment: §217: If I have exhausted the justifications . . . I am inclined to say: ‘‘This is simply what I do.’’

In an analogous case, Wittgenstein himself gave a gloss on his italics. This might provide a model to follow here. RFM 422: One follows the rule mechanically. . . . ‘‘Mechanical’’ – that means: without thinking. But entirely without thinking? Without reflecting.

Taken together, these remarks seem to make clear in what sense Wittgenstein here used the term ‘‘blindly’’. It does not imply that the agent is not attentive to what he is doing, that he is acting like someone whose mind is ‘elsewhere’, as it were like an automaton or zombie (cf. §156). Rather, it implies that he does not question what he does in response to the rule, hence too that he could offer no reasons (justification) for

234 Wittgenstein and Waismann

following it just as he does. He follows the rule blindly in a certain sense: namely, in the sense that Eurydice followed Orpheus blindly in making her way with him back to Earth from Hades. Unhesitatingly, willingly, in perfect trust: these seem to be the relevant connotations of ‘‘blindly’’. In the other case, the italicized expression Lebensformen (forms of life) might be intended to focus sharply on some quite particular conception (Which one?), and, if so, this might make it irrelevant or pointless for interpreting this controversial remark to investigate and clarify the use of the word ‘‘Lebensform’’ in nineteenth-century German philosophy or science.20 (Wittgenstein’s term may be the heir of this practice, but it need not inherit all the properties of its predecessors.) This method of interpreting the italicized expression involves discrimination and selection. It stands in sharp contrast with the policy of taking any remark incorporating italics to have as its intended content whatever might be obtained by substituting for the italicized expression any legitimate dictionary equivalent. More generally, in this use, italics function as filters, not as licenses for treating all of the connotations of the italicized expression as yielding acceptable interpretations of a remark. (In that respect, this use of italics differs sharply from the picture of ‘emphasis’ which informs most commentaries on Wittgenstein’s writings.) There might well be occurrences of italics which are ambiguous between this use and the previous one. An important instance might be his use of ‘‘grammar’’. For example, §307:

If I do speak of a fiction, then it is of a grammatical fiction.

One interpreter might argue that this is meant to invoke a special explanation of the term ‘‘grammar’’ which is missing from this text.21 Another might call attention to Wittgenstein’s contention that he did not use ‘‘grammar’’ in any special sense22 and then argue that grammar restricts attention to some particular aspect of what we call ‘‘grammar’’, say that rules of agreement and inflection are not answerable to the nature of what words signify. It is not clear that such a conflict can be resolved, or even that there would be any real point in trying to do this. (3) Both previous patterns of using italics have parallels in Wittgenstein’s use of single quotes or ‘scare-quotes’.23 This suggests a more general scheme of interpretation: in many cases the use of italics is equivalent in Wittgenstein-Speak to the use of single quotes. Italics might then be employed for whatever purposes single quotes might be introduced. (This invites a separate investigation of his use of single quotes to narrow down the range of admissible interpretations on both devices.)24

Italics in Wittgenstein 235

In the case of special explanations of everyday expressions, italics and single quotes are used interchangeably to secure back-reference. There are clear examples of this parallelism. The term ‘reading’ in }7 of §156 refers back to the explanation given in }1, just as reading in }5 does. Surely ‘private’ in §256 (‘private language’ in §275 and ‘privately’ in §202) is to be referred back to the explanation given in §243 of the expression ‘‘private language’’; hence it functions just as private does in ¨ usserung’ (Z §19) LPE 252 (and perhaps too in §311). Similarly, ‘A ¨ has the same significance as Ausserung (Z §58); and the phrase ‘‘eine ‘u¨bersichtliche’ Darstellung’’ (cf. GB 241) would be equivalent with ‘‘eine u ¨ bersichtliche Darstellung’’ (cf. §122).25 As in the case of italics, we might replace this use of single quotes with the phrase ‘‘in our sense’’ or ‘‘in this sense’’. For example, we might redraft §256 in this form: In that case my language is not, in our sense, a private one.

It is noteworthy too that italics and single quotes are used together in passages where Wittgenstein gave an idiosyncratic explanation or definition of some everyday expression. (Notably ‘criterion’ and ‘symptom’ occur in BB 24–5.) The second use of italics is also paralleled by one use of single quotes. In certain cases, both seem meant to single out some particular use or implication of an expression, but without any explicit or detailed explanation of the intended conception.26 It is of the utmost importance to work out exactly what is to be understood in such cases. One family of parallels ramifies around §219: I follow the rule blindly. This remark is tied to a metaphor: lines or rails that stretch through the whole of space to infinity (§§218–19). Following a rule is likened to a wagon’s running along tracks. This mythological description of the use of a rule (§221) generates a large family of remarks. RFM 371: But what I want is that one should be able to go on inferring mechanically according to the rule without reaching any contradictory results. RFM 217: I was playing a game and in doing so I followed certain rules. . . . Now I wanted to play this game in such a way as to follow the rules ‘mechanically’ and I ‘formalized’ the same. But in doing this I reached positions where the game lost all point; I therefore wanted to avoid these positions ‘mechanically’. RFM 214: diction.

The misuse of the idea of mechanical insurance against contra-

236 Wittgenstein and Waismann

In these cases it seems transparent that ‘mechanical’ and mechanical are to be understood in the same way (even if the precise interpretation of these passages must be left unsettled here). The discussion of securing consistency within a calculus switches back and forth between these idioms. Similar interchangeability of italics and single quotes typifies a wider range of remarks concerned with contrasting doing things ‘mechanically’ and ‘thinkingly’. BB 33: Now we might say that whenever we give someone an order by showing him an arrow, and don’t do it ‘mechanically’ (without thinking), we mean the arrow in one way or another. §156: Now what takes place when, say, [an Englishman with the usual education] reads a newspaper [aloud]. . . . He may attend to what he reads, or again – as we might put it – function as a mere reading-machine: I mean, read aloud and correctly without attending to what he is reading. Z §103: . . . an important distinction between creatures that can learn to do work, even complicated work, in a ‘mechanical’ way, and those that make trials and comparisons as they work. Z §108: . . . some of the creatures [anthropoid brutes] work ‘thinkingly’, others quite mechanically. RFM 422: One follows the rule mechanically. . . . ‘‘Mechanical’’ – that means: without thinking. But entirely without thinking? Without reflecting.

The equipollence of italics and single quotes in these passages is striking. All these remarks call for effort to work out precisely what contrasts are being drawn. A general investigation of the grammar of ‘‘mechanical’’ or ‘‘thoughtless’’ may well be of minimal use in deciding in what sense an activity is described as ‘mechanical’ or mechanical in any one of these remarks. There are other important instances of this parallel pattern of use of italics and single quotes, or equally of parallels between these reflexive devices and such phrases as ‘‘in this sense’’. Examples include ‘‘arbitrary’’ (‘‘willku¨rlich’’), ‘‘autonomous’’, ‘‘accompaniment’’, ‘‘interpretation’’, and ‘‘primitive’’. These all deserve careful investigation since all of them are crucial to some of Wittgenstein’s most distinctive ideas (e.g. ‘‘the autonomy of grammar’’26). Tracing remarks through repeated redraftings turns up some evidence for Wittgenstein’s sometimes replacing italics by single quotes or vice ¨ bersicht’ and U ¨ bersicht). Detailed investigaversa (e.g. interchanging ‘U tion might show this practice to be common.

Italics in Wittgenstein 237

In some other cases, there is evidence for some degree of interchangeability of italics and single quotes, even though these cannot be classified clearly under either of the two previous patterns. Here is one isolated case: §178: the guiding character of the movement . . . the movement one of ‘guiding’ . . . [cf. §170].

A more systematic interchangeability is evident in the case of certain verbs connected with linguistic communication: namely, ‘‘to mean (something)’’ [‘‘meinen’’], ‘‘to understand’’ [‘‘verstehen’’], ‘‘to think’’ [‘‘denken’’], etc. §609:

Interpreting ‘understanding’ as an atmosphere; as a mental act.

§81: . . . what can lead us . . . to think that if anyone utters a sentence and means or understands it he is operating a calculus according to definite rules. §186: ‘‘The right step is the one that accords with the order – as it was meant.’’ – . . . But that is just what is in question: what, at any stage we are to call ‘‘being in accord’’ with that sentence (and with the meaning you then put into the sentence – whatever that may have consisted in). §665: Imagine someone pointing to his cheek with an expression of pain and saying ‘‘abracadabra!’’. – We ask ‘‘What do you mean?’’. And he answers ‘‘I meant toothache’’. – You at once think to yourself: How can one ‘mean toothache’ by that word? Or what did it mean to mean pain by that word? §33: And what does ‘pointing to the shape’, ‘pointing to the colour’ consist in? Point to a piece of paper. – And now point to its shape – now to its colour – now to its number (that sounds queer). – How did you do it? – You will say that you ‘meant’ a different thing each time you pointed. – How is that done? – You will say you concentrated (or directed) your attention now on the colour, now on the shape.

This is an extensive and important class of cases. It seems likely that italics and single quotes are both regularly employed to mark a ‘fishy’ conception of these things as mental activities or processes, a picture which is liable to mislead us (§154; BB 15–16). These expressions all require careful investigation in detail. For various reasons, the significance of italics and single quotes seems to be roughly the same in many remarks.28 For readers of today’s philosophical journals and monographs, this is surely an unexpected use of italics. It could scarcely be categorized as a form of emphasis or stress. Yet an informed and sensitive reader of Wittgenstein should be constantly alive to the possibility of cashing out italics into single quotes. This is not just a

238 Wittgenstein and Waismann

hare-brained proposal, but a recognition of the author’s intention as made manifest in his texts. (Bearing in mind the possibility of this equivalence might help to reduce dogmatism in interpretations, even to persuade us that we have no idea what is meant by italicized expressions in advance of detailed investigation of ideas that surround each particular remark.) At the same time, this parallelism with single quotes makes it important here to ward off a possible misunderstanding. Italics, even when used together with or interchangeably with single quotes, are more closely akin to using the italicized expression than to mentioning it. This point seems evident.29 On the other hand, they do regularly license back-references of a meta-linguistic kind.30 For example, §101: The idea now absorbs us, that the ideal ‘must’ be found in reality. Meanwhile we do not as yet see how it occurs there, nor do we understand the nature of this ‘‘must’’. §141: Can’t an application come before my mind? – It can: only we need to get clearer about the use of this expression. [cf. §154; RFM 371]

(These cases will remind a well-informed reader of Quine’s two-level analysis of the sentence: ‘‘Georgione was so-called because of his size.’’) There is a danger of misinterpreting Wittgenstein if we do not keep our bearings on this complexity. I have tried to make visible three general patterns, partially overlapping, in Wittgenstein’s use of italics. How far each of them extends calls for investigation, but this can’t be undertaken before we know what to look for. Available secondary literature manifests total aspect-blindness in these matters.

Some Specific Patterns In addition to these general patterns, we can also discern at least three more quite particular patterns in Wittgenstein’s use of italics. All three have some philosophical significance. The first is his using italics for the purpose of disambiguating the first-person plural pronoun ‘we’ (and its variants ‘us’, ‘our’, ‘ours’). It contrasts ‘‘we’’ with ‘‘they’’, ‘‘the rest’’, ‘‘others’’, etc.31 Italics here make a crucial contribution to interpreting many remarks. Indeed, uncertainties about how to interpret ‘‘we’’ from case to case is a notorious source of difficulty and controversy. (With correlative difficulty uncertainty about interpreting second-person pronouns, esp. in interior dialogues. Who is the we who say. . . ? And who is the you who say. . . ?)

Italics in Wittgenstein 239

In some cases it seems clear that I (LW, the author) am the centre of the ‘we-circle’. Discussion is conducted with ‘the royal ‘‘we’’ ’. Here the pronoun ‘‘we’’ is to be taken as referring to Wittgenstein himself and some rather indefinite group of like-minded people who are imagined to surround him (sometimes an idealization of the set of his interlocutors in his seminars/lectures at Cambridge, sometimes a set of sympathetic philosophers including Schlick and Waismann32). Italicizing ‘‘we’’ generally seems to have the purpose of marking off this particular use of ‘‘we’’. For example, §116: What we33 do is to bring words back from their metaphysical to their everyday use.34 BB 25: We, in our discussions . . . , constantly compare language with a calculus proceeding according to exact rules.35 BB 4: But if we had to name anything which is the life of the sign, we should have to say that it was its use.36 PG 87 (¼ BT 58): What interests us in the sign, the meaning which matters for us, is what is embodied in the grammar of the sign.37

This use of italics is particularly conspicuous in Wittgenstein’s dictations to Waismann in the early 1930s:38 F 14: The explanation of what a word means can be seen as the cause of its subsequent use, but in this sense we are not interested in the explanations of words. We are interested in them only as parts of a set of rules and operations (a calculus), i.e. only from the standpoint [Gesichtspunkt] of rules governing the use of signs [Zeichenregeln]. F 15: We say: when a thought swims before one’s mind [einem vorschwebt], this is similar to the case when the solution of a mathematical problem swims before one’s mind: the task is not thereby \completed\ /solved/. . . . For us it all comes down to this, that the presentiment of a thought is not a thought. F 26: . . . ‘The word must be used thus’ or ‘That is really its meaning’. In the word ‘really’ lurks a great danger, for one is then always tempted, if not to alter things, at least to stretch what is actual [the reality of language] to fit this pattern. We have no need to do anything of this kind. We place the pattern beside what is actual [language] and let it throw as much light on that as it can. PLP 278:39 We put forward the following terminology; we suggest that in the case in which the word ‘red’ is explained by means of a colour spot we shall say that one sign is explained by another, that a word is explained by a

240 Wittgenstein and Waismann sign of another sort. . . . The colour spot will be used by us in future as a standard of comparison, a paradigm, and is, in this sense, part of language.40

Once introduced, this device might be taken to disambiguate subsequent comments where parallel italics are absent (as in the last two citations). For instance, we in BB 25 dominates this remark: BB 25–6: Why then do we in philosophizing constantly compare our use of words with one following exact rules? The answer is that the puzzles which we try to remove always spring from just this attitude towards language (cf. PG 63).

One could say that the scope of the italics, though somewhat indefinite, regularly extends over some subsequent uses of ‘‘we’’; as it were, those uttered in the same logical breath. Italics are surely not necessary for using ‘‘we’’ in this sense. Indeed, this reading seems required in many cases where ‘‘we’’ is not italicized. For example, §§132–3: Our clear and simple language–games are . . . set up as objects of comparison. . . . §115:

A picture held us captive.41

BT 410: Indeed we can only convict somebody else of a mistake if he acknowledges. . . . What the other acknowledges is the analogy I am proposing to him as the source of his thought. BT 416:

. . . again and again in our investigations when we show that. . . .

BB 27: Philosophy, as we use the word, is a fight against the fascination which forms of expression exert upon us. BB 28: . . . our method is not merely to enumerate actual usages of words, but rather deliberately to invent new ones. . . . PG 312: Grammar is for us a pure calculus (not the application of a calculus to reality).

Nor are italics sufficient to fix this sense of ‘‘we’’. For instance, Z §529, where we are contrasted with hypothetical creatures or automata. Or §20 contrasts our language (English or German) with unspecified other languages (real or imaginary) which have radically different sentence-forms. Or PG 321: The reason why there are infinitely many cardinal numbers is that we construct this infinite system and call it the system of cardinal numbers.

Italics in Wittgenstein 241

The moral to be drawn is rather indefinite but not useless. Wittgenstein made extensive and varying use of the first-person plural pronoun. Working out its reference is of critical importance for the interpretation of many remarks, and italicization is a clue that we cannot afford to ignore. Moreover, in the absence of italics, crucial ambiguities may arise, and some, perhaps, may have to be left unresolved. For example, we might wonder whether §122 might be clearer if drafted in this form: The concept of a perspicuous representation is of fundamental importance for us. It earmarks the form of account we give, the way we look at things.

Or would this revision give the remark a content different from its intended one? In respect of BT 418, we might canvas a similar redrafting: A simile belongs to our structure. . . . We can draw no conclusions from it. As when we compare a sentence to a picture . . . or when I [sic!] compare the application of language with, for instance, the calculus of multiplication.

Equally, one might consider recasting remarks apparently framed with ‘the royal ‘‘we’’ ’ in versions making use of the first-person singular pronoun. This is a pattern that Wittgenstein himself sometimes exploited. For example, PG 66:

I am only describing language, not explaining anything.

BT 421: As I do philosophy, its entire task consists in expressing myself in such a way that certain troubles //problems // disappear ( (Hertz.)). ~~~~~~~~~~~~

It seems clear enough that these remarks are fully equivalent to others couched in We-speak. For example, §109: We must do away with all explanation, and description alone must take its place. §133: . . . the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear.

Careful attention to the details of Wittgenstein’s use of first-person pronouns, singular and plural, might help to break the tyranny of the idea that he sought to clarify the nature of philosophy, at least if this is taken to be a description of the only possible method of philosophizing.42 Commentators seem here to exhibit the influence of their own craving for

242 Wittgenstein and Waismann

generality. At the least, there is another pattern to be seen in his methodological comments. Look again at this influential remark: §116: What we do is to bring words back from their metaphysical to their everyday use.

And ask yourself what precise significance you attach to his use of italics here. A second frequent and distinctive pattern is the italicization of demonstratives. There are many reasons for this. But one recurrent one is to indicate a particular conception of demonstrative reference, or equivalently a particular conception of the limits of language. The symbol ‘‘this’’ is regarded as composite, as consisting of the utterance of the pronoun ‘‘this’’ together with the gesture of pointing (BB 109; cf. PG 88); it is, as it were, partly concrete, and speaking here includes bodily action in addition to articulating sounds (cf. §669: Here pointing is a part of the languagegame). The italicized this might be taken to call attention to this point; i.e. to an aspect of the fact that the speaking of language is part of an activity (§23). This interpretation might be strengthened by noting that Wittgenstein often used direct quotation with an italicized demonstrative (e.g. ‘‘This is red’’) to record a speech episode where the utterance includes (or requires) a demonstrative gesture (§28; cf. Z §713). In some cases, this usage is further augmented by an arrow inserted after the demonstrative pronoun; the arrow apparently deputizes for the gesture of pointing at something; e.g. in the report ‘‘This ! is red’’ (e.g. LPE 259, 262, 269, 270, 274; F 39). Italics and arrows with demonstratives are clear idiosyncrasies in Wittgenstein’s use of oratio recta, but both seem to be motivated by his philosophical concern to make sure that we include gestures in our description of the use of demonstrative pronouns.43 An important variant on these cases of stressing demonstratives: ‘‘going on like this’’ – where the speaker does not point or gesture to an action or a pattern, but rather produces it himself.44 He may recite the opening of the alphabet, execute a mordant on a harpsichord, exhibit the correct pronunciation of a word or the intended way of declaiming a Shakespearean soliloquy, demonstrate how to hit a backhand at tennis or how to bow to the Vice-Chancellor in a degree ceremony, or show the lie of the land with a gesture (PI p. 198). Somebody understands such an explanation only if he grasps the pattern exemplified in the performance (a way of acting, a method of reproducing or extending a geometrical ornament, a technique of forming numerals) (PI p. 193, 196; RFM 310–11, 317, 320, 328, 331, 332). In this case the italicized demonstrative seems meant to

Italics in Wittgenstein 243

call attention to the way that certain definite behaviour of the speaker is interwoven with a particular use of words; perhaps also to the way that knowing how to go on is interwoven with understanding these words. Italicizing demonstratives seems to be a minuscule but important contribution to the conduct of a major philosophical campaign. Namely to persuade us (his readers) of the need to reconsider our (unconscious? prereflective?) picture or conception of the boundary between language and extra-linguistic things or the boundary between speaking and the varied activities of human life. (We are inclined to take these boundaries to be sharply drawn and permanently fixed rather than vague, fluctuating, and purpose-relative (cf. §§38, 50).) To treat the use of italics with demonstratives as a stylistic quirk or mannerism is to neglect the dimension of depth in this typographical device. A third prominent pattern is the italicization of modal expressions: especially ‘‘must’’, ‘‘may’’, and ‘‘cannot’’. (Also the italicization of expressions used to draw attention to the absence of modal auxiliaries: e.g. ‘‘is not’’, ‘‘does not’’, and ‘‘would not’’.) Modal distinctions are crucial to the intended meaning of many remarks (e.g. §§66, 81, 87, 90, 101, 131), and Wittgenstein employed italics for the purpose of ensuring that readers not overlook these qualifications. (One might say that this use of emphasis is to call attention to his practice of not using modal auxiliaries merely to express different shades of emphasis or assertiveness.45) §66: Don’t say: ‘‘There must be something common, or they would not be called ‘games’ ’’. . . . BB 41: I have been trying . . . to remove the temptation to think that there ‘must be’ what is called a mental process of thinking, hoping, wishing, believing, etc. independent of the process of expressing a thought, a hope, a wish, etc. BB 43: Our investigation tried to remove this bias, which forces us to think that the facts must conform to certain pictures embedded in our language.

Other remarks use italics to stress other modal qualifications, or even their absence. For example, §28:

He may suppose this; but perhaps he does not.

§41: . . . suppose that the tool with the name ‘‘N’’ is broken [in language (8)]. . . . Here one might say: ‘‘N’’ has become meaningless. . . . BB 20: If for instance I expect B to come to tea, what happens may be this: . . .

244 Wittgenstein and Waismann BB 49: . . . the proposition ‘‘An iron nail can’t scratch glass’’ [could be rewritten] in the form ‘‘experience teaches us that an iron nail doesn’t scratch glass’’, thus doing away with the ‘‘can’t’’.

Most of Wittgenstein’s texts contain many instances of this pattern of italicization. Here, arguably, the force of italics is primarily stress. Modal qualifications and modal distinctions have enormous philosophical consequences,46 and many philosophers have a strong urge to disregard them, to minimize their significance, or to misunderstand them.47 For instance, interpreters of Wittgenstein neglect the contrast between saying that every proposition is composite and claiming that any proposition must be composite;48 or the contrast between saying no word in English has a meaning outside a sentence and claiming that any word can have a meaning only in a sentence.49 Wittgenstein followed the opposite policy of attending scrupulously to modal distinctions as an integral part of his campaign to distinguish sharply between grammatical and empirical judgements (RFM 165, 363). He suggested that statements with ‘‘must’’ and ‘‘cannot’’ constitute dogmas or prejudices (BB 130, 137). Their presence is diagnostic of what he called the metaphysical use of words (§116; BB 35). In excluding possibilities, such statements have the function of norms of representation or rules of grammar (whether or not this is intended) (BB 55, 61, 139). Hence, framing a remark with emphasis on ‘‘need not’’ or ‘‘may’’ is the main way of implementing the strategy of bringing words back from their metaphysical to their everyday use (§116; e.g. §66). Conversely, when he himself put emphasis on ‘‘cannot’’ or ‘‘must’’, he meant this remark to be understood as a determination to take a sentence to formulate a rule of grammar, a principle that regulates the use of the other constituent words (e.g. BB 41). In Wittgenstein’s writings, the copious use of italics with modal expressions has clear philosophical motivations. Indeed, in his texts, modal qualifications, with or without italics, are always crucial to the content of a remark.50 (Interpreters seem to operate on the erroneous principle that if they focus on the sense of a remark the cans and musts can (or must?) be left to take care of themselves.) All three of these specific patterns in the use of italics could be called variants of emphasis.51 But in each case emphasis seems to serve a definite purpose, and the purpose is different from case to case. (One might then deny that italics are used ‘merely for emphasis’.) The message carried by this (namely, the importance of seeing gestures as part of language) is different from the message carried by must (namely, that a sentence formulates a rule of grammar); and both differ from the significance of

Italics in Wittgenstein 245

we (namely, contrasting our ways of seeing and describing things with those of an amorphous set of ‘others’). All of this seems tolerably clear. Might it also seem puzzling? How can mere italics carry so much baggage? How can the expression this (or we or must) say all that? This puzzlement should surely be dissolved. It is the extended patterns of Wittgenstein’s different uses of italics that fix how they are to be interpreted in these different cases. His intentions are clearly embodied in his texts. Is it any more puzzling that he could give this single reflexive device all this rich and variable significance than that he could appropriate the ¨ bersicht’’ to convey his own complex and subtle conception of a term ‘‘U form of analogical insight? Italics are no more – and no less – wonderful than other linguistic symbols!

Methods of Investigation At least four different general patterns have emerged in the use of italics in Wittgenstein-Speak. All seem capable of considerable extrapolation. In addition to the various forms of emphasis, we have noted the equivalence of italics with quotation-marks (especially with single quotes); the use of italics to express ‘in our sense’; and their use to express ‘in some sense’ or ‘in a certain sense’. How far these patterns do extend in Wittgenstein’s writings seems worthy of investigation. Even the relatively unfamiliar ones might turn out to be common in his work if we cultivated the habit of looking for them. The recognition of the existence of these patterns seems to open up the possibility of a range of new interpretations of many apparently ‘mined-out’ remarks. As a long-term project, one might make use of these patterns in a systematic re-examination of his texts. By eliciting some degree of order out of an apparent motley, this investigation might well turn up further patterns – more subtle but no less clear than the preliminary ones. One could then take these back to the texts, thereby revealing yet more refined patterns. And so on. The aim would be to set up a virtuous circle! (A hermeneutic circle?) A progressively more sensitive understanding both of his thinking and of his ways of expressing his thoughts. Much of the available interpretation of his work might finish up looking impossibly crude. Provisionally, we might adopt this somewhat laboured method of procedure (at least until we have acquired new habits of thinking). For each remark containing italics we will generate four ‘readings’. Here to serve as a model are four ways of rephrasing the remark: Following a rule is a custom (cf. §199).

246 Wittgenstein and Waismann

1

2 3 4

Following a rule is really, unambiguously, or definitely a custom. (Or: Following a rule is a paradigm of a custom. Or even: Following a rule is what is called ‘‘a custom’’.) Following a rule is, in our sense, a custom. Following a rule is, in a certain sense, a custom. Following a rule is a ‘custom’. (Or even: Following a rule is, as it were, or so to speak, a custom.)

Applying this method, we generate four readings for any italicized expression in any remark. The task of interpretation must minimally include canvassing reasons for or against each of these paraphrases. No doubt this method would prove tedious, but it might be worth the effort if it generated some really eligible new readings of central texts.

Some Applications Let us examine in a bit more detail four remarks whose interpretation seems to turn on working out the precise significance of italics. (1) §371:

Essence is expressed by grammar.

Compare the same italicized expression in §§92, 547. We might paraphrase each of them with single quotes: e.g. (§371) ‘Essence’ is expressed by grammar. All three seem to indicate particular conceptions of essence, evidently not always the same one! What seems to be at issue are different conceptions [Auffassungen] of essence: meanings that we give to the word ‘‘essence’’ (Z §467), not a fact to be discovered about the concept of essence (not some linguistic fact [sprachliche Tatsache (Z §447)] about the use of ‘‘essence’’52). A fortiori, Wittgenstein’s intention is not to establish a principle from which we can deduce such grammatical principles as that ‘the use of a word does not follow from the object it names’.53 Here (in §371) the italics seem to mark Wittgenstein’s own present conception of essence (unlike §92, where essence is contrasted with this conception and picks out his previous conception which informed TLP 54). So we might arrive at this preferred paraphrase: Essence, in our sense, is expressed by grammar. (Or particular applications of this dictum: e.g. what we call ‘‘the essence of a natural number’’ is the grammar of the words ‘‘zero’’, ‘‘one’’, ‘‘two’’, . . . ) If so, omitting these italics would yield a different remark,55 certainly contestable, arguably both dogmatic and false.56 Most interpreters take §371 to formulate some species of conven-

Italics in Wittgenstein 247

tionalism, taking Wittgenstein to endorse the cited remark (§372): ‘‘The only correlate in language to an intrinsic necessity is an arbitrary rule.’’57 By contrast, the remark in §92 requires a different paraphrase from §371: . . . questions as to what is, in a certain sense, the essence of language.

Or even: . . . questions as to the so-called essence of language.

(That conception of essence is explicitly contrasted with Wittgenstein’s own: For if we too in these investigations. . . . For they see in the essence . . . (§92: both emphases added).) Might the gloss here put on the italics in §371 be rejected on the grounds that any alternative conception of essence is incoherent? Is that view clearly expressed anywhere by Wittgenstein? (Perhaps the Aristotelian conception of essence is a paradigm of something that is more a superstition than a mistake (cf. §110).) Did he articulate the view that what are commonly conceived (by whom?) to be the natures or essences of things ‘‘are merely the shadows cast by grammar’’?58 One might say: the topic of §371 is essence, not essence. In other words, italics here belong to the content of this remark.59 (2)

§199: To obey a rule, to make a report, to give an order, to play a game of chess, are customs (uses, institutions).

Do these italics serve to make an emphatic assertion of a point likely to meet with resistance? Some (or even most?) philosophers are allegedly loath to concede that these things must all really be social practices shared within a community; so Wittgenstein took the step of insisting that they are customs, uses, institutions. This idea is widely thought to be the crucial move in his analysis of following rules, especially in his resolution of the paradox of §201.60 What about the alternatives of taking this remark to be equivalent to: . . . are ‘customs’? or to: . . . are, as it were, customs? or to: . . . are, in a certain sense, customs? or to: . . . are, in our sense, customs? Prevalent interpretations of §199 do not consider these alternatives, let alone offer reasons for judging them to be unacceptable. In spite of the enormous body of secondary literature on rule-following in Wittgenstein, nobody examines the import of these italics.61 All available reasoning runs

248 Wittgenstein and Waismann

along these lines: emphasis on ‘‘custom’’ has the implication that the listed activities are all paradigms of what are called ‘‘customs’’; therefore all of the connotations of ‘‘custom’’, a fortiori the property of being shared among the members of a community, must apply to all these activities, a fortiori to following a rule. This reasoning has the obvious merit of building on a natural and appropriate explanation of the term ‘‘custom’’. Indeed, customs might be defined as social habits62 – to emphasize that they are shared among members of certain societies, that they vary from society to society and that they are a product of nurture, not nature. This explanation would license our saying: ‘‘In England, orderly queuing is a custom’’; also ‘‘Among the fellows of St Jude’s, it is the custom to dine at least once a week on high table.’’ The existence of these practices differentiates some groups from others. By contrast, applying the same explanation of ‘‘custom’’ to all Wittgenstein’s examples in §199 runs into difficulties. It would surely not license our saying: ‘‘Giving an order is a custom’’! Why not? Because there is nothing here of the variability from society to society which is both characteristic of customs and integral to our calling them social habits.63 Could we conceive of a society that lacked all forms of interaction which fall under the general headings of giving and following orders? of making reports? of following rules? Or is it conceivable that any society have rules but lack the habit of following them? (Is it a habit that humans walk with their feet on the ground?) On reflection, don’t all these suggestions strike you as absurd? These reflections might suggest that italics here function as a filter on the connotations of ‘‘custom’’. They direct us to consider what is here to be understood by ‘‘custom’’, rather than to import all of the connotations of ‘‘custom’’ (say, being social) and to apply them in interpreting this passage.64 Most non-committally, one could operate with the paraphrase: . . . following a rule is, in a certain sense, a custom. This would force us to confront the question: In what sense (or in what respects) are these things similar to paradigmatic customs? Which implications of ‘‘custom’’ are meant to be highlighted here, and which ignored? Working out a well-documented answer to this question would be a considerable undertaking. But here is one direction that might be explored with profit:65 A custom is a normative practice that lacks any authoritative codification; indeed, it may nowhere be formulated ‘in so many words’.66 Wittgenstein often pointed out that any explicit codification of rules and any practice of appealing to rules to justify particular actions presupposes prior mastery of practices that are inculcated by imitation, drill and practice (Z §419) and thereafter policed without

Italics in Wittgenstein 249

recourse to a rule-book. It might have been for the purpose of drawing attention to this point, as it were the primacy of practices over ruleformulations, that he remarked that following a rule is, in a certain sense, a custom.67 This might be one skirmish in his general campaign to bring to others’ notice the deep insight embodied in Goethe’s maxim: Am Anfang war die Tat. (3)

§665: Imagine someone pointing to his cheek with an expression of pain and saying ‘‘abracadabra!’’. – We ask ‘‘What do you mean?’’. And he answers ‘‘I meant toothache’’. – You at once think to yourself: How can one ‘mean toothache’ by that word? Or what did it mean to mean pain by that word?

Here the topic is: How can one ‘mean [meinen] toothache’ by the word ‘‘abracadabra’’? The italics seem to indicate a particular conception of meaning as an ‘inner process’ which is only contingently related to any outward expression. This gives rise to serious perplexity (cf. §112). On the one hand, it seems that such a mental act must be what gives life to signs (cf. BB 4). On the other hand, it seems opaque how what goes on in me when I utter the word can have the consequence of endowing the word with meaning (cf. PI p. 218). The problem is to explore and clarify what it means to ‘mean toothache’. What is under investigation here is a particular picture or conception of meaning something by a word. Wittgenstein contrasted this (fishy) picture with an alternative picture of the use of ‘‘meaning (something)’’ – one which relates the application of this verb, on the one hand, to mastery of a technique or practice, and, on the other hand, to the circumstances surrounding such humdrum but variable activities as pointing at someone, tracing the contour of something (§35), sitting down at a chessboard, etc. (especially what comes before and after saying things (BB 43)). We are apt to get ourselves thoroughly confused unless we recognize that the expressions ‘mean’ or mean and ‘‘mean’’ have radically different explanations. Both single quotes and italics (§§19, 29, 81, 156, 161, 186–8, 190, 209, 276, 358, 590, 634, 665; BB 65, 80, 142; Z §§27, 397) are regularly used to mark similar conceptions (or misconceptions) of understanding, thinking, etc. These devices seem generally to signal that these concepts are here regarded as instances of ‘mental activities’ or ‘mental processes’, where the term ‘‘mental’’ is understood to indicate that what they describe is altogether extraordinary and mysterious (cf. BB 6). Quotes and italics regularly single out these pseudo-concepts. It is Wittgenstein’s task to bring us to acknowledge that these complex expressions have

250 Wittgenstein and Waismann

no meaning because their alleged (so-called?) ‘explanations’ (‘private’ ostensive definitions) do not determine how they are to be used (§258; cf. PG 60). Here then is a recurrent pattern of use of reflexive devices with ‘psychological verbs’. Failure to spot its instances or to appreciate its significance would be damaging to the understanding of the whole of Wittgenstein’s philosophy of psychology. One might say that what the ‘Big Typescript’ tries to get rid of is ‘meaning’ and ‘understanding’, not meaning and understanding (BT 1ff.). Less obviously, but no less crucially, the topic of §202 is not following a rule (as in §199), but ‘following the rule’. In a large number of remarks about meaning or ‘meaning’, understanding or ‘understanding’, etc. the italics (and single quotes) modify the content. (4) BB 125:

Our method is purely descriptive.

Previous considerations suggest that italics and single quotes might qualify the phrase ‘‘purely descriptive’’ in the same way. Hence we can now spot a variant of BB 125 in this remark: BB 18: Philosophy really is ‘purely descriptive’.

At the very least, it cannot be taken for granted that these two remarks make different points (despite the apparent contrast between ‘‘philosophy’’ and ‘‘our method’’). Both invite the question: In what sense ‘‘purely descriptive’’? Surely neither version makes a statement of fact about the whole of what has been carried on under the label ‘‘philosophy’’. Is the intention to lay down the limits of what makes sense for philosophers to do? (To make a dominative use of ‘‘is’’, as in the constitutional provision ‘‘The House of Representatives is elected every two years’’?) The phrase ‘‘our method’’ definitely suggests something more modest. We might see both remarks as a declaration of Wittgenstein’s own intention about how to proceed: the intention actually embodied in his work. This would carry no pretensions to intellectual imperialism. ‘‘Our method is purely descriptive’’ is itself a simple description of our method: not an attempt to lay down the law about ‘‘the method, the only method, of philosophy. . . ’’.68 What then is the property of being purely descriptive? Here is one possible paraphrase: Our method is, in our sense, purely descriptive. This paraphrase would call for parallel paraphrases of some linked remarks. In particular, since description is contrasted with explanation (BB 18, 125), our method surely requires that we do away with all

Italics in Wittgenstein 251

explanation (§109).69 That is to say, we must avoid giving any explanations, in our sense. (BB 18: I want to say here that it can never be our job . . . to explain anything.) So what is ‘‘description, in our sense’’? And correlatively, what is ‘‘explanation, in our sense’’? These seem to be the questions which need to be answered in order to arrive at the intended interpretation of this remark (and its close relatives). This is a vast topic. Here I can do nothing more than give some directions for thinking about it. What needs to be clarified is what Wittgenstein understood by the project of describing the grammar of our language,70 and there are here at least three dimensions of uncertainty. (a) The explanations to be avoided are not confined to scientific theories (even though scientific discoveries have no authority to effect modifications to grammar). The explanations put on the Index include anything which generates ‘‘musts’’ and ‘‘cannots’’ about how words are used. And hence anything which opposes looking and seeing how language functions (§§5, 66). This will include a large (and variable!) range of pictures, prejudices, dogmas, and philosophical theories of language which have certainly not been advocated as empirical or scientific generalizations. For example, Augustine’s picture of language (§1); Frege’s claims that every judgement can be analysed as the value of a function for an argument and that every concept must be defined for every object (cf. §71); or the theses of the Tractatus that every propositional sign must be a fact or that every proposition must be a truth-function of elementary propositions. It is this kind of claim that is to be excluded by rigorously adhering to the principle that there are no deductions within grammar (§126). Similarly, it is just these metaphysical statements that are to be undermined by bringing such terms as ‘‘sentence’’ and ‘‘language’’ back from their ‘metalogical’ to their everyday use (§108; cf. §116). ‘‘Explanations, in our sense’’, one might say, covers a multitude of sins. (b) Descriptions of the grammar of our language are not concerned with facticity, with establishing what is the actual public practice of the English-speaking peoples. Wittgenstein’s descriptions include possibilities (§244), imaginary language-games (§§1, 2, . . . ),71 objects of comparison (§§130–1), similes and pictures of word-use (§§422–5), and aspects of our use of words (§123).72 Acceptance of his descriptions is always voluntary; it is a form of free acknowledgement (BT 408–9). The aim of any philosophical investigation is to bring to my consciousness an internal conflict (CV 16); a clash between different things that I feel driven to say. (Nothing can be established against my will – as objective facts about the grammar of my language! Because rules cannot operate at a distance! I can follow only those rules which I acknowledge.) Consequently,

252 Wittgenstein and Waismann

‘‘descriptions, in our sense’’ has an application far more open-ended and person-relative than one might expect. (c) Descriptions of the grammar of our language are confined to everyday, non-technical terminology for speaking about how we speak. (Our everyday metadiscourse is itself a major topic of philosophical investigation, not something that stands outside the paths of Wittgenstein’s investigations.) What is excluded here is ‘scientific’ discoveries about word-use; e.g. a grammarian’s definition of the category of ‘static verbs’, or the philosopher’s claim that the word ‘‘that’’ which introduces an indirect statement is really a demonstrative pronoun. According to this conception, there is no meta-language – just as there is no meta-mathematics (PG 296). (In this sense, language must speak for itself (cf. PG 40, 63).)73 Our method is purely descriptive, not purely descriptive. Here italics are to be understood as gesturing at Wittgenstein’s own complex conception of description and explanation. This device should arguably here be cashed out into the phrase ‘‘in our sense’’. This makes a BIG difference to the content of this crucial remark – and to grasping the spirit of all his philosophical investigations. In all four cases, these remarks have been generally interpreted without any careful attention to the occurrence of italics. If noticed at all, emphasis has been taken simply to demonstrate that Wittgenstein took these remarks to be of cardinal importance within the framework of his thinking. This manifests serious misunderstanding in each case. Indeed, available interpretations pin on him various dogmas that he would surely have disowned (e.g. that following a rule must be a public practice shared within a community, i.e. that this is part of the grammar of the term ‘‘rule’’). A better understanding of his varied uses of italics might have helped to forestall these serious misinterpretations. The same point holds for dozens of other important remarks in the corpus of his work. Here is a small sample: §115:

A picture held us captive.

§241: It is what human beings say that is true and false; and they agree in the language they use. That is not agreement in opinions but in form of life. §307: Z §423:

If I do speak of a fiction, then it is of a grammatical fiction. The red visual impression is a new concept.

PI p. 204: A concept forces itself on one. PI p. 208: ‘‘Now he’s seeing it like this’’, ‘‘now like that’’ would be said only of someone capable of making certain applications of the figure quite freely.

Italics in Wittgenstein 253 The substratum of this experience is the mastery of a technique. But how queer for this to be the logical condition of someone’s having such-and-such an experience! PI p. 226: life.

What has to be accepted, the given, is – one could say – forms of

Envoi This paper is unusual in two respects: First, it is intended to make visible an aspect of Wittgenstein’s method, namely, various significant patterns in his use of italics. For this purpose, it offers what I hope are persuasive interpretations of a few pivotal remarks. None of this is, in one sense, definitive or conclusive. Somebody who is determined to uphold a different position (in support of customs’ being shared practices or philosophy’s being essentially descriptive) cannot be dislodged by force of argument. Here, as elsewhere, aspect-seeing is voluntary. Possibilities may be exhibited, but nothing proved. Some philosophers will think that this concession makes my enterprise pointless. What is certainly true is that they will gain nothing from it. Second, I do sincerely think that this paper will achieve little of real value unless it stimulates others to take up the challenge to think out for themselves how to interpret remarks in Wittgenstein’s texts where italics play a significant role. I have tried to demonstrate a method of interpretation in a few cases. There are countless cases where important work remains to be done. Notes 1

For example, Malcolm multiplies italics in his paraphrases as part of arguing that Wittgenstein saw the concept of following a rule to be ‘essentially social’: What is harder to grasp is Wittgenstein’s conception that following a rule . . . is a practice, a custom, an institution. It is a form of life, a feature of the natural history of human beings. (Wittgensteinian Themes, p. 165)

2

As in the ‘Big Typescript’. This German typographical device is not familiar to literate anglophones; hence it may be easily overlooked by readers when the English translation of a German text simply copies this form of emphasis (BT §§86–93 in PO). 3 In the case of dictations, this might be the sole source of italics. Yet he seems to have intervened in inserting italics and quotation marks in the Blue Book, and he may have dictated such punctuation to Waismann.

254 Wittgenstein and Waismann 4

5

6

7

8 9 10

11 12

13 14 15 16

17

So, for instance, §371 is italicized in toto to stress the point that it ‘crystallizes a leitmotif of Wittgenstein’s later philosophy’ (Hacker, Wittgenstein: Meaning and Mind, p. 437). Or, conversely, this remark is picked out as a good epigram, but paraphrased as ‘Essence is determined by grammar’ (Hallett, A Companion, p. 424). Such comments indicate that italics are not taken to contribute to the content of remarks in PI. He suggested this policy in respect of the word ‘‘shock’’ when he wrote the introductory discussion to the Medical Research Council report by Grant and Reeve on the physiology of shock. (His idea was not carried out in the publication.) (See Monk, Ludwig Wittgenstein: The Duty of Genius, p. 447.) In English printed texts of the seventeenth and eighteenth centuries italics serve many distinct purposes, and the pattern of use gradually shifts. For example, they are often used as catchwords in the margins of texts and as headings in indices or in the text; they are regularly used for quotations from authorities and for expressions in foreign languages, sometimes for the citation of words or phrases, for definitions of expressions, or even for every proper name of a historical person. One common in nineteenth-century texts but not evident in Wittgenstein is the practice of italicizing the whole of an author’s definitions of expressions (e.g. Frege’s (Begriffsschrift) definition of ‘‘function’’ (§9)). This may have been intended to facilitate the reader’s locating these definitions – an especially important matter when the practice was to publish books without indices. Here italics seem meant to highlight the fact-claiming force of ‘‘reveal’’; to draw a contrast with ‘‘create’’ or ‘‘invent’’. For example in Locke, Watts, Hume and Schopenhauer. Whether or not the content of this explanation of the correlative terms ‘‘grammar’’ and ‘‘application’’ is Wittgenstein’s, its form conforms exactly to his practice. For example §161. Similarly, neither reading (BB 121) nor read (BB 122) is meant to be a back-reference to the special explanation of ‘‘read’’ (BB 119). To avoid both confusion and pedantry, I have followed this practice: no quotes are used in mentioning expressions which are themselves italicized or enclosed in quotation-marks, single or double. Otherwise, double quotes are used for mentioning expressions. Descartes, Principia Philosophiae, book I, §9 (Adam and Tannery, eds, Oeuvres de Descartes, vol. 8). Descartes, Meditationes II (Adam and Tannery, eds, Oeuvres de Descartes, vol. 7, p. 29). He also used single quotes (‘scare-quotes’) for this purpose. His use of scare-quotes produces similar anomalies. For example, the expression ‘privately’ in §202 must be understood by reference to the explanation of a ‘‘private language’’ which occurs only subsequently in §243. (By relocation, it has lost its original back-reference in MS 180a.) In some other cases, the scope of italics may be meant to extend over other related expressions or over alternative phrasings of a remark. For example, the

Italics in Wittgenstein 255 italics in §199 are probably to be understood to extend over the brackets so that the remark could be redrafted: To follow a rule . . . are customs (uses, institutions). It is noteworthy that the term ‘‘Gebrauch’’ is italicized (underlined) in one MS ancestor of this remark: MS 180a: 23r: Schachspielen, Regeln folgen ein Gebrauch. 18

Perhaps too for the purpose of disambiguating the term ‘‘expression’’ [‘‘Ausdruck’’] (cf. §317). 19 This conception is fundamental to the causal theory of meaning presented in Russell’s Inquiry into Meaning and Truth. 20 An analogous case is the term u ¨ bersehen in §122. It seems quite irrelevant for grasping Wittgenstein’s conception to investigate the long tradition ¨ bersichtlichkeit with the axiomatic presentation of a body connecting U of knowledge (evident in Frege, FA §5 (cf. also Frege, Begriffsschrift, §13) and also in Boltzmann (Theoretical Physics and Philosophical Problems, pp. 4, 16, 96). 21 On this model: PLP 13–14: . . . the grammar of the language . . . [is] everything about language which can be fixed before language is applied. . . . Grammar is, as it were, the installation and adjustment of a system of signs, in preparation for their use. . . . If we take the word ‘grammar’ in this wide sense, . . . [it] includes all the enormous number of conventions which . . . are presupposed in the understanding of everyday language. 22

23

24 25

26

Moore, G. E., ‘Wittgenstein’s Lectures in 1930–33’, Mind, 53 and 54 (1959), pp. 276 ff. (Reprinted in Philosophical Papers (London: Allen and Unwin). References are to this edition.) In most of his writings, both manuscript and typescript, single quotes are used differently from double quotes. This important contrast is too complicated to clarify here. Editor’s note: see Baker, ‘Quotation Marks in Philosophical Investigations Part I’ (2002). For different reasons, neither phrase occurs in the quoted form. Rhees captures the force of the double quotes in GB by replacing them by single quotes in his translation, while in §122 the scope of the italicization of u ¨ bersehen extends over the quoted phrase. Single quotes alone, but not italics, are used to make allusions to other thinkers or texts. Some are clear: e.g. ‘organization’ (PI p. 196) refers to Ko¨hler, ‘speaking with tongues’ (§528) to the Bible. Others are problematic: e.g. ‘Weltanschauung’ (§122) might allude to Spengler (cf. GB 241).

256 Wittgenstein and Waismann 27

Compare: PG 63:

The calculus is as it were autonomous.

DS 3: One could put the point roughly like this: we do not step beyond the bounds of language in giving any explanation of what a word means, not even in an ostensive definition, and language is in this sense autonomous. PLP 40: Grammar is autonomous and not dictated by reality. . . . This is how we should regard grammar. . . . 28

29

One case is familiar to us in English: the convention of using quotation marks in referring to the titles of chapters in a book or articles in a journal, but italics in referring to the titles of books or journals. It is hard to find clear counterexamples in Wittgenstein, but here is one: RFM 309–10: ‘It must be so’ . . . This must shows that he has adopted a concept. This must signifies that he has gone in a circle.

30

31 32 33 34

Though now unusual, this use of italics was standard in many earlier philosophical texts, e.g. Locke and Watts; also in nineteenth-century texts, e.g. Hegel, Schopenhauer and Frege (who used the locution: the concept ‘‘horse’’ (FA §46) interchangeably with the idiom: the concept horse (Frege, Translations, p. 196)). It is also common in contemporary linguistics for the citation of type-expressions (in contrast with spoken or written tokens). The same two points hold of single quotes, used with or without italics. For example §202 (And hence also ‘following the rule’ is a practice) is not a remark about the phrase ‘‘following the rule’’, but about an imaginary language-game (previously described in MS 180a). At the same time, in BB 133 the expression ‘seeing what’s common’ leads to a back-reference in the form: the phrase ‘‘seeing what’s common’’; similarly BB 135 discusses whether a mental strain and a bodily strain are ‘strains’ in the same sense of the word. Also with the indefinite third-person pronoun ‘‘one’’ (‘‘man’’), as in ‘‘one might say. . . ’’ (‘‘Man mo¨chte sagen . . . ’’). As in Waismann’s elaboration of ‘our method’ (‘unsere Methode’) (PLP, LSP). This emphasis is absent in the earlier draft in BT 412. If this comment is descriptive, as it appears to be, then it must be restricted to us, the practitioners of Wittgenstein’s method. If it were taken to apply to all of us philosophers, the host of past and present participants in the intellectual activity named ‘‘philosophy’’, it would be a ludicrous generalization.

Italics in Wittgenstein 257 35

36 37 38 39 40 41 42

43 44

45 46 47

48

49

50

Rhees seems to misunderstand this particular use of we. He takes Wittgenstein to be explaining the source of philosophical problems, hence to be addressing the question why we (philosophers) do constantly compare language with a calculus. The answer Wittgenstein offered is simply that philosophical puzzles ‘always spring from just this attitude towards language’. And he objects: ‘And you might wonder whether this is an answer’ (BB x–xi). It seems a perfectly good answer to a different question, namely one about Wittgenstein’s adopting a particular strategy in his philosophizing; particularly because this strategy is built on a counterfactual hypothesis! The first emphasis is missing from the printed text, though present in Wittgenstein’s own copy. The title of BT §14 emphasizes both occurrences of the pronoun. References are to the catalogue numbers of the dictations. Largely derived from F 36. This conception [Auffassung] is immediately contrasted with ‘the usual view’ (PLP 278). Surely an autobiographical remark, alluding to the Tractatus. This implication may manifest a serious misunderstanding of Wittgenstein’s conception of essence. (For clarification of this point, see ‘A Vision of Philosophy’ [Essay 9 of this volume: Ed.].) This feature is important in distinguishing the uses of demonstratives from standard uses of proper names (cf. §38). MS 109. 286: [Instances of the justification of a course of action as following an order] Or also somebody has told me, if it means ‘Do p and q’, then I ought to do this (whereupon I exhibit the activity). Contrast Frege, Begriffsschrift §3 (strength of evidence) and Russell, An Inquiry into Meaning and Truth (level of confidence). Neglect of them is one reason that Wittgenstein gave for thinking that set theory fosters conceptual confusion (PI p. 232; cf. WWK 213–17). Disappearance of these ‘subtle nuances’ must be the most serious general defect in the various notes made by persons who attended his lectures. Quite naturally they concentrated on what they picked out as ‘the message’. It is this second claim that belongs to the ‘theory of common structure’ as expounded first in the Tractatus and later by Schlick in ‘Form and Content’ ¨ tze). (in his Gesammette Aufsa This gives rise to the odd practice of constructing counterexamples to refute things put forward as necessary truths! (cf. Morris, ‘The ‘‘Context Principle’’ in the Later Wittgenstein’). Compare Freud’s lament: qualifications and exact particularization are of little use with the general public; there is very little room in the memory of the multitude; it only retains the bare gist of any thesis and fabricates an extreme version which is easy to remember. (Freud, The Complete Works, vol. 7, p. 267)

258 Wittgenstein and Waismann 51 Single quotes never take over the role of italics in these cases. 52 It might be wondered to what extent there is any everyday use of ‘‘essence’’ in English (contrast French!). 53 Hacker, Wittgenstein: Meaning and Mind, p. 102. 54 The concern to delineate the essence of any possible language sharply differentiates the doctrines of TLP from any description of the working of any particular languages. All of the remarks about symbolism in TLP are to be understood as formulating necessary truths, i.e. to belong to the metaphysics of sign-systems. For example, the claim that all propositions are composite (zusammengesetzt). Hence they are invulnerable to refutation by observations about the use of words in natural languages. (Among commentators on TLP, Waismann is unique in his sensitivity to this point. Cf. PLP 317–20.) 55 So too would extending italics over the whole remark in paraphrasing it (Hacker, Wittgenstein: Meaning and Mind, p. 437). The first would, in effect, eliminate the rider ‘in our sense’, while the second would alter its scope. 56 In the Aristotelian tradition (e.g. Aquinas and Suarez), the essence (substantial form) of a horse explains a fundamental feature of biological reproduction, namely that the mating of a pair of horses generates another horse. Essence in that sense is surely not expressed by ‘grammar’. (It has been displaced by mechanisms of reproduction which are investigated by molecular biology.) 57 In effect, ignoring the preamble: e.g. Baker and Hacker, Wittgenstein: Rules, Grammar and Necessity, pp. 329–47. 58 Hacker, Wittgenstein: Meaning and Mind, p. 438. (And what justifies the word ‘merely’ in this paraphrase?) 59 This is Waismann’s understanding of the matter. He cited as an example of vision in philosophy Wittgenstein’s sudden insight into the nature of such things as hoping, fearing, intending, meaning, understanding, etc. On this novel conception (itself an instance of vision in philosophy), ‘the nature of understanding reveals itself in grammar, not in experiment’ (HISP 37). This is presented as one conception of essence among others, and hence too as something that we are free to reject. 60 These points are forcefully presented by Malcolm, Wittgensteinian Themes, pp. 146–7, 156, 162, 170. 61 Similarly, but even more surprisingly, nobody even remarks on the fact that the phrase ‘‘following the rule’’ (‘‘der Regel folgen’’) in §202 is enclosed in single quotes, just as ‘‘privately’’ is. Hence, without any justification whatever, commentators give asymmetric interpretations of this single reflexive device within the compass of one short remark; and they ignore a striking difference between the drafting of §199 and that of §202. 62 von Wright, Norm and Action, p. 9. 63 von Wright, Norm and Action, pp. 9–10.

Italics in Wittgenstein 259 64 65 66 67 68 69 70

71 72

73

A parallel use of italics occurs in an earlier version of this remark: (MS 180a. 23r): Schachspielen, Regeln folgen ein Gebrauch. Another interpretation is developed at length in Baker and Hacker, Wittgenstein: Rules, Grammar and Necessity, pp. 154–81. von Wright, Norm and Action, p. 10. An earlier draft (in MS 180a) reads: . . . presupposes a custom. Baker and Hacker, Understanding and Meaning, p. 266. Cf. HISP 12: we only describe; we do not ‘explain’. [Here single quotes seem to be equivalent to the italics in §109.] Some of the ambiguities of ‘the grammar of our language’ are discussed in Baker, ‘Some Remarks on ‘‘Language’’ and ‘‘Grammar’’ ’ (1992), pp. 107–31 [Essay 2 of this volume: Ed.]; and different conceptions of the clarification of grammar are developed in Baker, ‘A Vision of Philosophy’ (1999) [Essay 9 of this volume: Ed.]. CV 74: Nothing is more important for teaching us to understand the concepts we have than constructing fictitious ones. Perhaps the pattern of using arithmetical equations for the purpose of transforming one empirical statement into another; hence the possibility of seeing equations under the guise of rules of grammar (a conception (Auffassung) of arithmetic). This is not what Wittgenstein himself meant by this slogan!

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

12 Wittgenstein: Concepts or Conceptions?*

I want to make a distinction between concepts (Begriffe) and conceptions (Auffassungen), and a related distinction between descriptions (Beschreibungen) and pictures (Bilder). I will then try to put this material to work in clarifying an important dimension in Wittgenstein’s Philosophical Investigations: the difference between conceptual analyses and overviews ¨ bersichten or U ¨ berblicken) (compare F* 16). (U These distinctions have been hidden from view by the widespread preconception that Wittgenstein is always in the business of giving clarifications of the meaning of words which are allegedly grounded in detailed descriptions of their everyday use in ordinary discourse and that his philosophical investigations have a substantial degree of overlap with Carnap’s programme of constructing logical syntax and with Ryle’s mapping of logical geography. The distinctions I discuss are both subtle and contestable, but I will try to make a case for their being of decisive importance. Initially I follow the practice indicated in the index to the Philosophical Investigations, which distinguishes a sense of ‘picture’ which is glossed as ‘conception, model’. Wittgenstein seems frequently to take conceptions to be crystallized in what he call ‘pictures’, and he often uses ‘picture’ to characterize ideas which seem non-pictorial or even unpicturable.1 My modus operandi is more a matter of raising questions than providing answers. But failing to raise enough questions or to put the question marks deep enough down is a main weakness of much philosophy. It generates the danger of falsche Fragestellung. * Originally published in Harvard Review of Philosophy, 9 (Spring 2001), pp. 7–23. Reprinted by permission of the Harvard Review of Philosophy.

Concepts or Conceptions?

261

Preliminary: PI §1 The Philosophical Investigations opens with a quote from Augustine’s Confessions. Wittgenstein then comments: ‘These words, it seems to me, give us a particular picture [Bild ] of the essence [Wesen] of human language. Namely: the words in language name objects – sentences are combinations of such names’ (§1).2 (Elsewhere he remarks that this picture treats naming and describing as the two essential activities in speaking a language; also that it assigns a fundamental role to ostensive definition.) What is meant by calling this a ‘picture’ of the ‘essence’ of language? Essence may be readily understood. What belongs to essence is necessary. What is essential is what cannot be otherwise than it is. So according to Augustine’s picture, every word must be a name, every sentence must be a combination of names. Or, every (meaningful) word must be correlated with an object; and every sentence must be composite. These uses of ‘must’ produce statements that cannot be read as empirical generalizations. In this sense, Augustine’s picture is not comparable to any ‘theory of meaning’. Formulations of essence are the paradigms of the ‘metaphysical uses’ of words (§116). But what does it mean to speak of a ‘picture’ of the essence of language? What is ‘picture’ supposed to contrast with? One relevant point might be that the picture ascribed to Augustine is not explicit in the citation. It is imputed to him on the basis of his description of how he learned to speak. In this case, if Augustine had set out to answer the question ‘What is the essence of language?’, could we no longer say that he gave us a picture? Another point might be that Augustine’s account is impressionistic. Almost certainly it does not report memories of actual speech-learning episodes in his childhood, but rather his idea of how he must have learned to speak. The account is also highly schematic, devoid of significant detail. Does it follow that what is inferred from this account must be equally an impression of the essence of language? Yet another point might be that ‘picture’ here suggests that Augustine’s account is inaccurate or incorrect or incomplete; as it were a mere picture. This dyslogistic use of ‘picture’ is familiar in other contexts. We speak of somebody’s picture of a sequence of momentous events in which he was a participant, contrast one participant’s picture with another’s, and perhaps also contrast the pictures of all the participants with a description of what really happened. Is Wittgenstein preparing to give a correct description of the essence of language (not a picture at all!) to be

262 Wittgenstein and Waismann

contrasted with all other pictures of this essence? Does ‘picture’ here imply that he judges Augustine’s picture to be incorrect? And if it is not incorrect, how can it ‘surround the working of language with a haze which makes clear vision impossible’ (§5)? On the other hand, Wittgenstein often seems to suggest that what he calls ‘pictures’ are not subject to assessment as being correct or incorrect. They might be superstitions, but not mistakes (§110). ‘The picture is there; I don’t dispute its correctness’ (§424). This is because there is nothing to dispute in this respect about a picture. He continues: ‘What is in question is unclarity about how it is to be applied.’ An example would be misapplying the picture of inner–outer by taking something to be hidden behind what is inner in the picture (LPE 281)! Pictures are to be contrasted with applications of pictures (§§374, 422–7). But if a picture cannot be correct or incorrect, then in a sense it cannot give any information, so it must, in a sense, be vacuous. And yet it can be misleading (VI. 248), damaging, constraining. How is this possible? For example, a description might be misleading if it incorporated false information (how?), or implied (or even naturally suggested (to us)) ideas that are false (e.g. ‘objects’ suggests logical independence to us). But what if no information is conveyed? How can a picture be challenged as being mistaken or wrong?3 How can it ‘take us in’ (PI p. 184)? These questions are all-important. Uncertainty about how to construe ‘picture’ affects the overall interpretation of §§1–89, arguably too the interpretation of the private language argument (§§243–321). Is Wittgenstein’s intention to demolish Augustine’s picture? To demonstrate that it is a misdescription of the essence of language? (Does he regard this picture as a botched attempt to describe the everyday use of such terms as ‘language’, ‘meaning’, ‘word’, ‘sentence’, and ‘object’? Hence as a misrepresentation of these concepts?) Is his aim to replace Augustine’s picture with a correct description of the essence of language, that is, with an accurate description of the grammar of ‘language’ (compare §372) – and perhaps also of ‘name’, ‘sentence’, ‘object’, ‘description’, etc.? Or does he propose, and try to win our acceptance of, a different picture of the essence of language? Could it be a picture which he means to present in the slogan ‘The meaning of a word is its use in language – a sentence is an instrument in a language-game’? What would be the implications of our calling this a ‘picture’? Would it thereby lose all philosophical interest? Or any claim to our attention? Would it be a mere picture?

Concepts or Conceptions?

263

Provisional Typology of Remarks in the Philosophical Investigations At least three distinct kinds of remarks stand out in the Philosophical Investigations: (1) Formulations of rules of grammar; or descriptions of the grammar of certain words (whether these be our words in everyday use or symbols of an imaginary language-game). For example: §2: The language of the builder and his assistant. §246: ‘I cannot be said to learn of [my sensations]. I have them. The truth is: it makes sense to say about other people that they doubt whether I am in pain; but not to say it about myself.’ (Incidentally, this perhaps describes the grammar of our current language, but not of Descartes’s!) §199: ‘It is not possible that there should have been only one occasion on which someone obeyed a rule.’ §248: ‘One plays patience by oneself ’. PI p. 222: ‘It is correct to say ‘‘I know what you are thinking’’, and wrong to say ‘‘I know what I am thinking’’. (A whole cloud of philosophy condensed into a drop of grammar.)’ (2) Extremely general facts of nature (see p. 56n), which are important for explaining the importance of certain concepts (see p. 230). Also propositions constituting our Weltbild (C §§93–4, 167). For example: §142: The practice of selling things by weight would ‘lose its point if it frequently happened for [lumps of cheese] suddenly to grow or shrink for no obvious reason.’ When I point to something, another generally looks in the direction in which I have pointed, not at my finger (as a cat does). The world has existed for a long time (C §§85, 182–8, 233, 311). Cats and motor-cars do not grow on trees (C §§279, 282). Mathematicians do not worry about the stability or permanence of paper-and-ink calculations. PI p. 227: ‘There is in general complete agreement in the judgements of colours made by those who have not been diagnosed as abnormal. . . . There is in general no such agreement over the question whether an expression of feeling is genuine or not.’ (3)

Pictures: Bild, Vorstellung, Auffassung. Some clear examples:

264 Wittgenstein and Waismann

The picture of possibility as a shadowy reality, for example §194: ‘The possibility of the movement is . . . supposed to be like the shadow of the movement itself.’ (Compare §448.) §427: ‘ ‘‘While I was speaking to him I did not know what was going on in his head [hinter seiner Stirn; lit. behind his forehead].’’ In saying this, one is . . . thinking of thought-processes. The picture should be taken seriously. . . . We have this vivid picture.’ §352: ‘ . . . our thinking plays us a queer trick. We want . . . to quote the law of excluded middle and to say: ‘‘Either such an image is in his mind, or it is not; there is no third possibility!’’ . . . When it is said ‘‘Either he has this experience, or not’’ – what primarily occurs to us is a picture which by itself seems to make the sense of the expressions unmistakeable: ‘‘Now you know what is in question’’ – we would like to say. And that is precisely what it does not tell him.’ §59: ‘ ‘‘A name signifies only what is an element of reality. What cannot be destroyed; what remains the same in all changes.’’ – This was the expression of a quite particular image: of a particular picture which we want to use.’ The picture of the mind as a private inner world (LPE 279–81).4 I would like to offer some very preliminary comments on this third kind of remark, that is, about a picture as it contrasts with the first two kinds of remarks: A picture gives no information; hence no incorrect information. ‘Not facts; but as it were illustrated turns of speech’ (§295). Perhaps it is more like a way of seeing things; a conception (Auffassung), or a norm of representation. Its adoption may force us to give descriptions having a particular pattern, for example to describe all differences in word-use as differences in the objects signified by the words, or (as with Frege) to describe the inferential powers of judgements in terms of analysis into functions and arguments. A picture may force itself upon us or persist unshakably as part of our thinking. ‘The picture forces itself on us . . . ’ (§103). It is very important that pictures do force themselves on us (RFM 42), or ‘intrude on us’ (CV 50). They may captivate us (§112). They may get a grip on our thinking, holding us in a cramped position or keeping us in thrall. In this way, some pictures resemble prejudices or superstitions. And they may seriously restrict intellectual freedom. They produce mental cramps. A picture seems non-literal, metaphorical (LPE 280), allegorical, figurative, and so forth (for example, of possibility as a shadow; or the outside/inside distinction like a drawer and its contents). Hence, apparently beyond criticism as ‘nonsense’.

Concepts or Conceptions?

265

Changes in grammar alter sense (concepts), but changes in pictures do not. For example, different pictures may be attached to our concept of possibility. It seems that we may replace one with another without altering the meaning of ‘possible’.5 Weltbild and pictures need not be evident or conscious; often they must be discovered by sensitive reflection on our own practice (hence, in a sense, by bringing to consciousness what is unconscious). We may even resist acknowledging them; because they are revolting or too crude. (Whereas grammar can be regarded as being fixed by the explanations we give [compare §75] – though perhaps not without some prompting?) The power of pictures (or prejudices) to distort thinking may largely depend on their being unconscious.6 Two clear points about pictures emerge: 1

2

Pictures are often to be contrasted with rules of grammar or everyday explanations of the meanings of words. (Nobody would explain ‘possible’ by calling possibility a shadowy reality. Or formulate Augustine’s picture to explain how to make use of the expression ‘a language’. Or propose ‘inner’ as a synonym for ‘mental’ or ‘conscious’.) Pictures are of the utmost importance for Wittgenstein’s philosophy. ‘Grammatical problems are connected with the oldest Denkgewohnheiten, i.e. with the oldest pictures embedded in our language’ (V. 224; BT 422–3). ‘Our investigation tried to remove this bias which forces us to think that the facts must conform to certain pictures embedded in our language’ (BB 43).7

The Main Issue The BIG question is: How does Augustine’s picture fit into this tripartite framework? Is it to be treated as a picture of meaning? Or as a misdescription of the grammar of ‘meaning’, ‘word’, ‘sentence’, and so forth? How does Wittgenstein see this matter? (Which view makes best sense of his strategy of investigation? Which best epitomizes the content of §§1–143?) A subsidiary goal here is to attain further clarification of the distinctions among these three kinds of remark and to show what is at stake in confusing them with one another. The use of resolving these matters will be to mark out the danger of crossing two different language-games in discussing Wittgenstein, that is, the danger of genre-misidentification. In particular, to caution against attacking and defending pictures as if they were meant as (literal?) descriptions of grammar. (For example: arithmetical equations are rules of

266 Wittgenstein and Waismann

grammar; the mathematician is an inventor, not a discoverer; in a proof we win through to a decision.) Much secondary literature may be deeply flawed in this respect. A similar danger is taking Wittgenstein to offer grammatical refutations of pictures. For example: ‘Every Satz must be composite’ (PLP 316–17), or ‘The real elements must be indestructible’ (§39), or ‘A propositionsign must be a fact’ (TLP 3.14). Does Wittgenstein intend to canvass exceptions to such statements? To propose counterexamples?

The Grammar of ‘Picture’ This section attempts a first sketch of the use of ‘picture’ (in, as it were, ‘Wittgenstein-Speak’), that is, an exploration of one of Wittgenstein’s uses of the term. (I ignore here his discussions of painted pictures, portraits, and so forth.) These remarks seem to hold equally of his use of ‘conceptions’ (Auffassungen) and ‘ways of seeing things’. (1) ‘Picture’ is quite naturally associated with facticity-disclaimers. Hence a picture might be called a ‘conception’ (‘Auffassung’), ‘way of seeing things’, ‘way of looking at or regarding things’, or ‘aspect’. And it might be marked with qualifiers, for example, ‘in a certain sense’, ‘one might say’, ‘I want to say.’ (Perhaps too with ‘scare-quotes’ and italics!) Wittgenstein emphasizes this point; he refrains from disputing the correctness of pictures (§424). ‘The mind seems able to give a word meaning’ (p. 184). This is not something that seems to be so; it is a picture. Thus we see that a picture is not suggested to us by experience (§59). (Though it may be drawn from familiar materials.) It is not gotten by induction, for example, the picture theory of the proposition (compare Z §444). This point about non-facticity seems independent of the particular content of a picture, especially of whether it is a picture of language or of the world. Corollary: a picture cannot be contradicted by observations or discoveries. It lies outside the range of refutation by facts. (It may be as easily dislodged by fiction as by fact. For example, §244: ‘Here is one possibility: . . . ’.) (2) Pictures are contrasted with explanations of how to use words. Pictures are no substitute for detailed formulations of how words are to be used. Sometimes they may seem to make the sense of an expression

Concepts or Conceptions?

267

unmistakable – but this is precisely what they do not do (§352). In this sense, they are not part of grammar. Corollary: they cannot be misdescriptions of the grammar of our language. (Not mistakes, but, at least sometimes, superstitions (§110).) A good example here is the picture of thinking as either a hidden mental process or as operating with signs (BB 6). This divergence of pictures seems not to engage with the everyday practice of using the word ‘think’. (Pictures are, in this respect, free-wheeling (Leerlaufende Ra¨der).) (3) It is a further (though related) implication of ‘picture’ that acceptance or rejection of pictures is always optional, an exercise of intellectual freedom. Impressed by other possibilities, I may decide to give up a picture, even one deeply entrenched in my thinking (a prejudice or preconception). Wittgenstein, like Descartes, urges us to try to do this. Conversely, I may refuse to accept any picture without displaying any form of stupidity or intellectual ineptitude. By contrast, shutting my eyes in the face of unwelcome facts (even facts of grammar?) is a form of irrationality. There are no conclusive proofs or refutations in respect of pictures. But there may be reasoning or argument – in a way not conforming with the ‘ideal’ philosophical argument. (4) A picture is, in a sense, empty; it tells us nothing and conveys no information. (Even a picture of grammar is empty.) For example, when we say (in accord with Augustine’s picture): ‘ ‘‘Every word in language (8) signifies something’’ we have so far said nothing whatever’ (§13). But it does not follow that every picture is harmless. For example, Augustine’s picture surrounds the working of language with a fog (§5). More generally, Wittgenstein traces the origins of many philosophical problems to the influence of pictures. The source of philosophical problems (disquiet, torment, compulsion, and so forth) is repeatedly identified as unacknowledged or unconscious analogies (BT 410), pictures (Bilder, Vorstellungen) (V.140; VI.40), Iden (V.193; PG 107), models, ways of seeing things (PG 57), comparisons (V.174), dogmas (prejudices), ‘‘misleading analogies in the use of our language’’ (BT 408), or ‘‘similes absorbed into the forms of our language’’ (§112). In dissolving problems, ‘I must always point to an analogy according to which one had been thinking, but which one did not recognize as an analogy’ (BT 408). How is this to be explained? A picture picks out certain things as selfexplanatory, others as problematic; it steers attention towards certain aspects of things and away from others; and it guides the direction that problem-solving takes and helps to set the standard of adequacy for a solution. These matters are of the greatest importance. It seems to set up,

268 Wittgenstein and Waismann

as it were, a form of representation for the description of things. Even for the description of the grammar of our language (compare §131). Nothing more? Rather, nothing less! Choice of a form of representation is answerable to no facts, but it is of decisive importance. It determines a whole intellectual orientation. A picture may be both empty and pernicious – this seeming paradox is fundamental to Wittgenstein’s conception of pictures, that is, to his metaphorical use of ‘picture’ in respect of conceptions or models. (In fact it is not a paradox at all: its emptiness makes the picture irrefutable and beyond the reach of attack and channels intellectual activity into the perverse activity of explaining away apparent counterexamples (CV 28).) ‘A picture held us captive’ (§115), namely, the picture of a proposition as a logical picture of a state of affairs (necessarily isomorphic with what it depicts). The italics call attention to the paradox. How is it possible for a mere picture to hold anybody captive, in thrall, in bondage? How can a picture exercise tyranny over our thinking? Especially if pictures are essentially voluntary or optional? (5) Pictures are locally (transiently or temporarily) exclusionary. ‘I cannot see something in two ways at the same time.’ Such as seeing mathematics as what measures, not as what is measured. Hence they generate (or may even constitute) aspect-blindness. (6) But pictures are globally complementary, not exclusionary. One way of seeing things (Auffassung) does not exclude the possibility of others.8 It is rational to make use of different pictures of a single phenomenon for the purpose of bringing out different patterns or aspects of what is investigated for the purpose of treating different conceptual confusions (compare §132). Ceteris paribus, pluralism is a virtue in making use of pictures. (Unlike constructing theories in science.) I can investigate language as a mechanism of human interaction (in a stimulus-response model), or as a calculus of rules, or as a game with signs (a languagegame). Corollary: we must surrender the desire to establish ‘the only possibility’, ‘the nature of . . . ’, and so forth and the desire to confute philosophical adversaries. (Perhaps a very difficult renunciation?) (7) Consequently, remedying the defects of one picture (or eliminating its tyranny) is a matter of gaining acceptance for other pictures. (This is a kind of homeopathy: pictures are to be treated with pictures.) Another apparent paradox: only pictures have the power to transform the aspects of things. Hence, according to Wittgenstein, what Darwin, Freud and

Concepts or Conceptions?

269

Einstein discovered are primarily new and fruitful ways of looking at ¨ bersichten. Accepting a picture is changing ways things; paradigms of U of seeing things (§144). (8) Hence dissolving picture-generated philosophical problems depends on each thinker’s consent. Accepting a picture is voluntary (like aspectseeing or imagination) (p. 213), not in the sense that it is easy or can be effected at whim, but in the sense that refusal is possible. Hence a philosopher is up against difficulties of the will, not the intellect (CV 17 ¼ BT 406–7). Discarding one picture and adopting another is an exercise of freedom. (Hence, (HISP 21) philosophy is the realm of freedom, that is, of rational choice.) It involves willingness to explore comparisons (LFM 55), or even conversion to a new way of seeing things (Umstellung der Auffassung) (§144). To a considerable extent, Wittgenstein’s philosophy involves negotiations with others (his readers and interlocutors, real or imaginary) about pictures, Auffassungen, conceptions. This is liable to go deeply against the grain of soi-disant analytic philosophers. They relish the clash of steel on steel; the adversarial model of argument, the possibility of proving something, especially the possibility of proving other philosophers to be wrong. They have a definite ideal of philosophical argument, as case-building. Wittgenstein’s admirers want to see him as participating in this activity: they want to extract results from his texts (for example, proof of the incoherence of speaking a private language), and they wish to find in his work cogent justification for their own activity. In their view, swapping pictures or possible ways of seeing things would be useless and repugnant, perhaps to be condemned as a form of ‘relativism’.

A Picture: ‘Meaning is Use’ Given Wittgenstein’s care in drafting remarks, and given this minimal background to his use of ‘picture’ (his conception of a picture?), we should expect him to juxtapose against Augustine’s picture of the essence of language another picture, not a compendium of grammatical rules for using ‘meaning’ or ‘language’ or an alternative ‘theory’. We might set out to look for signs that his slogan ‘The meaning of a word is its use in the language’ is meant to be taken as a picture in his sense, that is, as a counter-picture to Augustine’s picture. This expectation seems to be abundantly borne out. Evidence is everywhere in plain view in his texts. Wittgenstein’s own discussion of meaning,

270 Wittgenstein and Waismann

explanation, language, and so forth is full of the qualifiers characteristic of pictures: ‘in our (my) sense’, ‘for us’, ‘for my purposes’, and so on. These seem to mark off his remarks as presenting an alternative conception rather than facts of the grammar of our language (standard English). Here is a sample of such remarks from the early 1930s: ‘For us, meaning is the correlate of explanations of meaning.’ ‘Meaning, in our sense, is embodied in explanations of meaning.’ ‘For our purposes, mental pictures can always be replaced by drawings or diagrams.’ ‘We are always comparing language with a game played according to strict rules’ (PG 60, 63, 68, 69; BB 4, 25). ‘[I]n meiner Darstellung’ (VI.102). ‘I treat ‘‘etwas meinen’’ as synonymous with ‘‘einer Regel folgen’’ ’ (V.281). This form of discourse continues in the Philosophical Investigations: ‘it causes least confusion to reckon the samples among the instruments of the language’ (§16). ‘We can put it like this: This sample is an instrument of the language used in the ascription of colour’ (§50). ‘What we do is to bring words back from their metaphysical to their everyday use’ (§116). Are these not clear indications of Wittgenstein’s offering what he considered to be (no more – and no less than) a particular conception (Auffassung) of language? (And what is the alternative? Is it plausible to think that careful description of actual use of the phrase ‘instruments of a language’ would clearly endorse the claim (truism?) that samples (and gestures) fall under this concept? If I refuse to go along with his recommendation (§16), do I exhibit mis understanding of the phrase ‘instruments of a language’?) Of course there are many remarks in the Philosophical Investigations that lack locutions that explicitly signal the intention to offer a picture. They may give the appearance of being straightforward descriptions of the grammar of our language. See, for example, p. 224: ‘we do not learn [what ‘‘judging a motive’’ means] by being told what ‘‘motive’’ is and what ‘‘judging’’ is.’ That is, it cannot be explained in this way, as ‘measuring length’ can. More importantly: ‘For a large class of cases – though not for all – in which we employ the word ‘‘meaning’’ it can be explained thus: the meaning of a word is its use in the language’ (§43). This sounds as if it were meant to be definitive of the concept of meaning, not a picture of the essence of word-meaning. It seems to say something authoritative and final (even worryingly dogmatic?). It has the register of fact-stating discourse. So too does the remark: ‘The use of a word in the language is its meaning’ (PG 60). Is there really inconsistency or tension here? This implication might easily be resisted. Provided Wittgenstein thought that he had made it perfectly clear that his general intention was to offer a counter-picture to Augustine’s picture, he might see no point in drafting every individual remark to make this explicit. The whole

Concepts or Conceptions?

271

discussion of §§1–89 might be deemed to be subordinate to this qualification. Just as Montaigne said of Aristotle that his ideas are Pyrrhonism cloaked in assertion, so we might say of Wittgenstein that his criticism of Augustine’s picture consists of a picture (Auffassung) cloaked in assertion. The issue is the spirit in which his remarks are to be read. Augustine’s picture is a widely accepted form of representation of the grammar of our language, of the meaning of words. (It is visible in Frege’s search for the definitions of the natural numbers; or in Quine’s concern with ‘ontological commitments’.) Wittgenstein labours to establish a different form of representation in the thinking of his readers. According to his alternative picture, the meaning of a word is its use in practice (compare the grocer in §1), its meaning is the correlate of everyday explanations of meaning (§§69, 75), meaning is in the expression (not behind it) (DS 4–5; F* 18), speaking and thinking are operating with signs, and it is use which gives life to ‘dead’ signs (BB 4). Just as Augustine’s picture is not criticized by reference to ‘linguistic facts’, so this picture cannot be justified by the claim that it describes them correctly. Wittgenstein does not claim that ‘Meaning is what is explained by an explanation of meaning’ is a true account of the concept of meaning (and the concept of explanation), but rather that this principle characterizes ‘meaning in our sense’ (or his conception of meaning). Similarly, he does not claim that ‘think’ and ‘operate with signs’ are synonymous (or have identical uses), but rather that this rough equivalence should be the centre of variation for describing the complex ¨ bersicht of the use of the term ‘think’ (BB 6). The fact that he offers this U grammar of ‘think’ while acknowledging discrepancies between the uses of ‘think’ and ‘operate with signs’ is a clear indication of his conception that ¨ bersicht of ‘think’ is misunderstood pictures are not fact-stating. (This U when it is taken to formulate a rule of grammar.) Both attacks on his philosophy of language and defences of his conception neglect the crucial point that he is offering an interpretative picture. (This is what these disputes look like. The one party attack Wittgenstein’s form of expression as if they were attacking a statement; the others defend it, as if they were stating facts recognized by every reasonable human being (compare §402).) The same point holds in his philosophy of mathematics, for example, his calling arithmetical equations rules of grammar.

Augustine’s Picture of Language Analysis of Wittgenstein’s discussion of Augustine’s picture pays too little attention to his treating it as a picture of the essence of language. His

272 Wittgenstein and Waismann

methods of investigation are thereby misrepresented. Both ‘picture’ and ‘essence’ are crucial for understanding the contours of his critical investigation.

Essence Augustine’s picture consists entirely of necessary truths; it is concerned with the essence of language. Hence it tells us nothing about the actual uses of any words. In particular, it does not state that all words in English or German have the same use, that there are no categorial (or combinatorial) differences between them, that all English or German sentences contain more than one word, or that all sentences are used in a single way. It cannot be demonstrated to be incorrect by producing counterexamples. That strategy fails to acknowledge the radical difference between a necessary truth and an empirical generalization. (Claims about essence are demands (Erforderungen), not discoveries (Ergebnisse).) Canvassing counterexamples also pays no attention to the distinction between appearance and reality: as in the Tractatus Logico-Philosophicus, apparent nonconformity may mask real (hidden) conformity. Wittgenstein’s own discussion pays scrupulous attention to this difference. Here is one clear example (paraphrase of PLP 317–18): An objection might be made to the thesis that any proposition must be composite (complex). Consider the sign ‘Restaurant’ posted over the door into a building. This simple, one-word (non-composite) sign is apparently sufficient to state that this building is a restaurant. QED, – But, it is replied, what signifies that this building is a restaurant is not the sign alone, but the fact that the sign is positioned over the door. This fact is itself composite, one of its elements being the spatial position occupied by the word ‘Restaurant’. So, if we recognize that the propositional sign must be the whole of what states that this is a restaurant, it is not a single word and it is manifestly composite. (This line of reasoning, derived from a dictation to Waismann (F* 38), could serve as a model of how cogently to rebut any putative counterexample to any of the essential truths of the Tractatus.) How then can one criticize this component thesis in Augustine’s picture? Wittgenstein’s ‘argument’ (derived from a conversation with Sraffa) is to point out that we can use a gesture to make a statement (e.g. the gesture of slitting the throat). He then asks a question: ‘Is it clear what we should call the elements out of which it is composed?’; or (better) ‘What do you want to call the elements of a gesture?’ (compare PLP 318). This movement of thought may be persuasive, but it certainly does not consist in confronting a thesis with a recalcitrant fact. In fact, as an argument, it

Concepts or Conceptions?

273

seems open to an obvious objection: what right has Wittgenstein to take a gesture as an instance of a sentence of a language? (An important aspect of his discussion – even in the Tractatus – is to encourage us to reconsider, even to redraw, the boundary between language (symbols) and the world (what is symbolized) – at least for the purpose of certain discussions. This seems a paradigm of exercising freedom (rational choice) in stipulating concepts.) Wittgenstein seems aware of, and critical of, a tendency among philosophers to disregard modal qualifications (‘can’, ‘cannot’, ‘may’, ‘need not’, etc.) and thereby to assimilate essential truths to empirical generalizations. ‘Science has shown that . . . ’ is a constant refrain, a recurrent form of criticism in modern philosophy. For example, Descartes held that there was no such thing as a thought of which the thinker is not conscius, to which is made the objection: but Freud has shown that there are unconscious states of mind. (Here is an ignoratio facti with regard to ‘thought’ and ‘conscius’). Wittgenstein suggests seeing ‘unconscious toothache’ (or ‘unconscious desire’) as a new convention, rather than seeing this as a ‘stupendous discovery. . . which in a sense bewilders our imagination’ (BB 23). He suggests regarding many revolutions, even in science, as stipulations of new concepts rather than discovery of new facts. For example, admitting action at a distance in mechanics (Newton vs. Descartes) changes what it makes sense to say, or to offer as an explanation.9 Wittgenstein himself constantly emphasizes the radical distinction between necessary and contingent propositions (especially in mathematics): he is very careful in his use of modal qualifications, and he respects them in the thinking and writing of others. In his view, it is a fundamental confusion to propose counterexamples to descriptions of essence. Here, we might say, Wittgenstein emphasizes the distinction between descriptions of grammar and very general facts of nature: the importance of keeping these things distinct. He treats Augustine’s picture as consisting of propositions about essence, not as a schematic, quasi-scientific theory (or proto-theory) of meaning.

Picture Impressed by this last point, one might adopt a new strategy for dealing with Augustine’s picture, one that discounts the importance of the term ‘picture’. (This strategy is widespread in interpreting Wittgenstein.) Since essence is expressed by grammar (§372), Augustine’s ‘account’ of the essence of language must be taken to consist of descriptions of the use of the terms ‘name’, ‘object’, ‘sentence’, ‘combination’, ‘word’, etc.

274 Wittgenstein and Waismann

The essence of language is then ‘given’, namely by the grammar of ‘meaning’, ‘name’, ‘word’, ‘sentence’, ‘object’, ‘combination’, and so forth. It is then tempting to suppose that the use of these terms is relatively fixed or omnitemporal. They belong to the basic vocabulary of everyday discourse – a vocabulary whose application seems more or less invariant. (Compare BT 424: there is no progress in philosophy because language always throws up the same problems.) Combining these two ideas gives scope for trenchant criticism of Augustine’s picture. That is, it gives an incorrect description of the (omnitemporal) grammar of ‘meaning’, ‘name’, and so on. So it stands in need of replacement by the correct account. This provides another pattern of interpreting Wittgenstein’s discussion of Augustine’s picture. Its theses are mistaken. Here is one remark that seems to fit this pattern: having noted that some philosophers call the word ‘this’ the only genuine name, he observes that the kind of use the word ‘this’ has is not among the many different kinds of use of a word that are labelled by the word ‘name’ (§38). Hence to call ‘this’ and ‘that’ names is to misuse the term ‘name’. It is, apparently, a ‘metaphysical use’ (because it is deviant from ordinary practice), and it is to be corrected by bringing the word ‘name’ back to its everyday use (§116). Similar remarks may be made about other features of Augustine’s picture. For example, that it is incorrect to classify ‘Now I understand’ or ‘I am in pain’ as ‘descriptions’; the first is a signal (§180), while the second is an ¨ usserung) of pain, a learned replacement for expression or avowal (A pain-behaviour (§244). Similarly, he seems to claim that it is incorrect to describe ostensive definitions as connecting language with reality; they in fact connect symbols with samples, which are themselves parts of language (§51). In this way, each of the component theses of Augustine’s picture is demonstrated to be mistaken. Correspondingly, Wittgenstein’s positive conception of meaning (identifying meaning with use) is taken to be the correct description of the concept of meaning (or the grammar of ‘meaning’). For example, the slogan ‘Meaning is use’ is itself a description of grammar; it is justified by the observation that the use of ‘the meaning of the word ‘‘ . . . ’’ ’ and the use of ‘the use of the word ‘‘ . . . ’’ ’ run along parallel tracks. He is correct to state that ostensive definitions are substitution-rules for symbols (including gestures and samples) (BB 109), that it remains within language.10 However, neither the negative nor the positive sides of this account seems to correspond closely to what Wittgenstein actually says. The negative case is not simply a clarification of the grammar of these metalinguistic concepts. The discussion does not terminate with identifying ‘mistakes’. For example:

Concepts or Conceptions?

275

§24a: ‘If you do not keep the multiplicity of language-games in view you will perhaps be inclined to ask questions like: ‘‘What is a question?’’ ’ Wittgenstein proceeds to list some answers. The point is not that these are incorrect answers (based on the idea that answers to questions are all descriptions) but that the question itself is pointless, to be discarded as without interest (dissolved). The inclination is misguided, not the particular answers given (as opposed to better ones, as in PLP ch. 20). §27a: Wittgenstein lists some one-word sentences. Why? Not to establish anything. Instead he ends with a question: ‘Are you still inclined to call these words ‘‘names of objects’’?’ (This is directed at the will, not the intellect; compare §261.) §38: Noting that ‘this’ is not what is called a name is not the end of Wittgenstein’s discussion, but the beginning. He seeks to understand why it occurs to one to want to make precisely this word into a name when it evidently is not a name (§39).11 The implication is that this particular application of Augustine’s picture appeals to someone precisely because it involves a deviant use of the term ‘name’. He is already aware of this deviant usage – so pointing it out to him achieves nothing! What needs to be counteracted is the urge to describe the use of ‘this’ in this peculiar manner. Wittgenstein treats Augustine’s picture as if it lay beyond the reach of argument based on cataloguing facts about the grammar of metalinguistic concepts. He discusses this conception of the essence of language as something powerful, deep-rooted and worthy of respect. And he treats it as being, in a certain sense, unassailable. The counterpart, positive, idea that Wittgenstein’s own discussion of the essence of language is intended to be a correct description of the grammar of ‘meaning’, ‘name’, etc. is equally misconceived. (It too begins from the thesis that ‘Essence is expressed by grammar’ is a description of the grammar of ‘essence’, not an insight or way of seeing things (HISP 33).) We have already noted the abundance of textual signals that what he offers is a picture or conception of language (of the meaning of words and the sense of sentences). Such as in §421: ‘Look at the sentence as an instrument, and at its sense as its employment.’ (This is an optional point of view; not an ineluctable fact about the grammar of the phrase ‘the sense of the sentence ‘‘ . . . ’’ ’.) Here, we might say: Wittgenstein emphasizes the distinction between descriptions of grammar and pictures of grammar, the importance of keeping these distinct. A picture is not subject to a reductio ad absurdum – in even the loosest sense. The point of having a clear conception of what Wittgenstein means by a picture is two-fold:

276 Wittgenstein and Waismann

First, Augustine’s picture is primarily a form of representation, a way of seeing things, an intellectual orientation. To displace or replace it is a tremendous undertaking. Wittgenstein aims at nothing less than transforming an entrenched way of thinking, habits of thought (Denkgewohnheiten) which are evidently still dominant among analytic philosophers. This is much more ambitious and radical than correcting a misdescription of the uses of words. Part of what he aims to achieve is to win acceptance of the principle that ‘Essence is expressed by grammar’, that is, acceptance of a particular conception of essence.12 Second, appreciating that Augustine’s picture is a picture is vital for acknowledging that his own ‘theory of meaning’ is also meant to be a picture (conception, way of seeing things). Hence its acceptance (or rejection) is wholly voluntary. This has the implication that most of the discussions that try to refute or rebut Wittgenstein’s ‘theory’, as well as most of those that try to defend it against attack, are misconceived. There is literally nothing to attack – as being incorrect. And nothing to defend – as being an accurate description of the grammar of our language. To engage in these controversies is already to take Wittgenstein’s philosophical investigations in the wrong spirit. Wittgenstein’s positive aim is not to get you to say something you don’t want to say, but to do something you don’t want to do (compare LFM 55), namely to investigate meaning from a very different point of view. To focus on use – as in §1; not on Satzbau (sentence-construction) – as in ‘logical geography’. Soi-disant Wittgensteinians tend to miss this aspect of his philosophy.

The Content of the Picture of Meaning as Use Here is not the place to elaborate this picture, even in its main lines. But what stands out in Wittgenstein’s initial descriptions of word-use is important to note. 1

2

Wittgenstein shows little or no interest in sentence-construction (Satzbau) – even in logische Satzbau (§101). This is not discussed in the grocer’s game (§1) or the slab-game (§2). He directs attention to how we operate with words (§1), how they are integrated into human activities (§6), and how differently sentences function as instruments (§24).

His interests are radically different from the concerns of logicians to clarify sentence structure to carry out the analysis of inferences.

Concepts or Conceptions?

277

We might say that his investigations of language take place largely in a different dimension. (Could that be his point in contrasting ‘surface’ with ‘depth grammar’ (§664)?) In any case, his descriptions of grammar are products of a very distinctive way of regarding the uses of symbols. (This seems to be neglected by interpreters who see his main activity as primarily one of bringing words back to their everyday use.) Just as Augustine’s picture leaves indefinite flexibility in distinguishing kinds of objects (hence kinds of word-use), so too this picture leaves indefinite flexibility in distinguishing kinds of use (hence kinds of wordmeaning). It seems a virtue – not a defect – that ‘use’ is not precisely pinned down. What counts as use is open to negotiation, from case to case. (This is another dimension of freedom in Wittgenstein’s philosophical methods.)

Coda I want to finish by pointing out that what I have done here is at best to indicate some topics that seem worth further exploration and discussion. Here is a brief list of things that need still to be done. .

. .

. . .

Further investigation of the ‘logic of pictures’, especially of their being heterogeneous, strictly purpose-relative, and non-additive. (This ¨ berwould be crucial for understanding §132 and the conception of U sicht in §129.) Close examination of Wittgenstein’s various treatments of different pictures in his writings and lectures; clarifying in detail his methods of therapy. The implication of picture-investigation for his method. His concern with revealing and combating philosophical prejudices and superstitions (which links him to some extent with Descartes, Nietzsche and Merleau-Ponty). The exact place and role of his own pictures in his various ‘descriptions of grammar’. Clarifying his conception of philosophy. (Not: the nature of philosophy.) The possibility of seeing the history of philosophy as a dialectic of conceptions or pictures and of recognizing the fluidity of conceptions (and even decisive shifts in concepts). Finally, exploring motives for resistance to admitting pictureinvestigation to be a major component of his philosophical investigations.

278 Wittgenstein and Waismann

Notes 1

2

3

4 5

6 7 8

9

10 11

12

A more thorough study of the role of conceptions in Wittgenstein’s thinking would have to investigate his use of a wider range of expressions, e.g. ‘way of seeing’ (Auffassungsweise), ‘regarding . . . as’ (ansehen als), ‘mode of investigation’ (Betrachtungsweise), ‘form of representation’ (Darstellungsform), ‘way of thinking’ (Denkweise), and ‘figurative proposition’ (bildhaftiger Satz). For the sake of concision, subsequent remarks from the Philosophical Investigations are referenced by their section number only. Other sources are referenced by an abbreviated title and page or section number. All translations are mine, following, with some exceptions, G. E. M. Anscombe’s translations. See falsche Bild (VI. 10; PI p. 184 (d.h. unzutreffendes); irrefu ¨ hrende Bild (VI. 247); falsche Vorstellung (V. 140; VI. 38–9); falsche Auffassung (VI. 3); falsche Betrachtungsweise (VI. 2). See also ‘false analogies’ (falsche Vergleichen) (VI. 19, 23; 86–7, 229–30, 236, 241, 300) and ‘misleading comparisons’ (falsche Gleichnissen) (V. 174). That is to say, the arena of ‘inner processes’ (§305), ‘inner pictures’ (p. 196), etc. ‘The realm of consciousness’ (LPE 320). Though Wittgenstein sometimes uses ‘concept’ more or less equivalently with ‘conception’ (e.g. §308; p. 196), and, in that sense, the concept of possibility would be transformed. Here is one resemblance of Wittgenstein’s philosophical method with psychoanalysis (BT 408–10). What is meant by the phrase ‘embedded in our language’? or ‘absorbed into the forms of our language’ (§112)? This mirrors a familiar point about visual aspect-seeing, and it is a crucial component of Frege’s explanation of articulating judgements into function and argument (Frege, Begriffsschrift, §§9–10). On this view, there is a lot of important change in many concepts in each of the past four centuries! Though of course, criteria of concept-identity are open to negotiation. This line of reasoning is prosecuted with some vigour in Baker and Hacker, Wittgenstein: Understanding and Meaning. Cf. Why do we have this ‘urge to misunderstand’ things? (§109), And his answer to the question in §39 is ‘That is just the reason’! (‘For one is tempted to make an objection against what is ordinarily called a name.’) The grammar of ‘essence’? Is there any such thing? Outside philosophy?

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

13 The Grammar of Aspects and Aspects of Grammar*

Wittgenstein discusses aspects primarily in connection with visual perception, and he tries to point out differences between the concept of seeing and the concept of aspect-seeing (or seeing-as). I want to clarify a related distinction between the concept of describing concepts and the concept of proposing conceptions (the distinction between Begriff and Auffassung). Concentration on the latter can illuminate aspects of Wittgenstein’s philosophical method, as I will try to bring out, first through working through a particular example in some detail, then by making some more general observations about describing grammar and achieving philosophical understanding. Wittgenstein’s discussion of visual aspects takes its rise from an apparent paradox: when an aspect dawns on me, nothing has changed in what I see, yet everything looks different. ‘I contemplate a face and suddenly I notice its likeness to another. I see that it has not changed; and yet I see it differently’ (PI p. 193). ‘The expression of a change of aspect is the expression of a new perception and at the same time of the perception’s being unchanged’ (PI p. 196). He tries to dissolve this puzzle. In one sense, what I see (e.g. a diagram or drawing) is the same as before; in another sense, what I see (e.g. a rabbit as opposed to a duck) is entirely different. (Each of these senses of ‘what I see’ is correlated with a different sense of blindness.) I will suggest that a closely analogous puzzle arises in respect of his discussions of the grammar of our language. His intention is to leave language just as it was. Philosophy, according to his conception, has no authority to interfere with the actual use of language (PI §124). At the same time, he attempts to reveal aspects of things that are hidden because * Hitherto unpublished. What is printed here is a version of a draft by Gordon Baker, worked up into essay form by Katherine J. Morris.

280 Wittgenstein and Waismann

of their simplicity and familiarity (PI §129). By arranging what we have always known (PI §109), by reordering what lies open to view (PI §92), we may come to understand for the first time what is already in plain view (PI §89). Here too nothing is changed, yet everything appears differently. To clarify his distinctive conception of philosophical investigation requires making this apparent paradox completely disappear. The key to making good sense of these remarks about his method of philosophizing is to recognize the pivotal importance of conceptions (Auffassungen) or ways of seeing things in the activity that he calls ‘describing grammar’. He suggests that philosophical problems are an impression of disorder in our concepts (BT 421) which is manifested in our not being able to find our way about (PI §123). As a remedy, he seeks das erlosende Wort in the form of exhibiting an order (eine Ordnung) which, as if by magic, transforms what seemed chaotic into something ¨ bersichtlich). He tries to make his interlocutor aware of a intelligible (U particular ‘physiognomy’ in the use of our words. Success in this activity is not a case of imparting some information or conveying an opinion, and what is now grasped is not a discovery since it was always in plain view, though previously unnoticed (PI §126). In accepting a new Auffassung, one sees that everything remains the same, and yet everything undergoes a metamorphosis. Only conceptions have the power to work this form of magic within the framework of a philosophical investigation.

Some Aspects of the Grammar of ‘‘Aspect’’ Wittgenstein offers some clarification of the logic (grammar) of purely visual aspect-seeing. (a) Global pluralism: When we see something to be red (or square), this excludes the possibility of our seeing it to be green (or triangular). But when we see a picture as a duck, this does not exclude the possibility of our also seeing it as a rabbit (on another occasion). Aspects do not exclude one another; to acknowledge one way of seeing something does not render illegitimate a different way of seeing it. In this sense aspects are complementary. One might even say that they are essentially plural; to speak of one way of seeing something presupposes that there are others.1 Thus no aspect has exclusionary claim-rights (to be the only possibility). (b) Local incompatibility: Aspects are transiently exclusionary; it is impossible to see a picture simultaneously as a duck and as a rabbit. Visual aspects are essentially non-additive: i.e. there is no such thing as combining two ways of seeing something to produce a single more comprehensive way of seeing it. Consequently, seeing something in one way interferes

Grammar of Aspects and Aspects of Grammar

281

with seeing it differently. And an entrenched way of seeing something may make it extremely difficult or practically impossible to see an unfamiliar aspect at all. (c) Voluntariness: Aspect-seeing (in contrast to seeing) is, in a certain sense, voluntary. This does not mean that one can look at something differently ‘at will’, but that, for example, it makes sense to ask someone to try to see something in this way as opposed to that. Hence, as in exercises of imagination, no information is acquired in the dawning of an aspect; nothing is discovered. Seeing an aspect is not learning something; aspects are, in this sense, cognitively empty. Hence in a sense they are not subject to dispute: to rational support or to disconfirmation by facts. (At the same time aspect-seeing may be facilitated by rational discussion: see infra.) Corollary: getting someone to see this or that aspect may be purposerelative. For example, I may try to get you to see a particular twodimensional diagram as a convex step in order to help you to follow one geometrical demonstration, as a concave step to aid you in following another (cf. PI p. 203). (Hence the pluralism of aspects may be deliberately exploited – like complementarity in atomic physics.) (d) Dependence on concept-mastery: Much aspect-seeing presupposes the mastery of concepts; it is, as it were, half thought, half experience.2 In many kinds of cases, we cannot see something as an F unless we have the concept of being an F (e.g. seeing a triangle as having fallen over on its side). ‘The substratum of this experience is the mastery of a technique’ (PI p. 208).3 (e) ‘Homoeopathy’: An aspect can be displaced only by another aspect. Seeing as belongs, as it were, to a different dimension than seeing. It is the different aspects of the duck–rabbit that exclude each other; by contrast, no feature of the drawing (nothing to be seen in it) can logically exclude my seeing it as a duck. (If it could, then there would be no duck-aspect of the drawing.) Moreover, nothing but seeing it as something different can interrupt my seeing it as a duck. (f) The possibility of ‘blindness’: An aspect may be invisible to someone even though what has this aspect is open to view. It may be seen by one person but invisible to another. Hence we can speak of someone’s being blind to a particular aspect of a particular thing. (Unlike ordinary blindness, this deficiency may be absolutely specific. It is more like prosopagnosia (‘blindness’ to faces) than blindness sans phrase.) Or we might speak of ‘aspect-blindness’ in respect of a more wholesale inability to see certain kinds of aspects of certain kinds of things. (Cf. ‘meaning-blindness.’) Only someone who sees a particular aspect can ascertain that another is blind to this aspect of what is in plain view, and nobody can establish by himself

282 Wittgenstein and Waismann

that he is blind to an aspect (or aspect-blind). I cannot say of myself that I see the duck–rabbit as a duck unless I can also see it as a rabbit (or something else); otherwise I simply see a duck (or a drawing of a duck), though another may say of me ‘He sees it as a duck’. (g) A paradox of modality: I cannot demonstrate to somebody that there is a possibility of seeing a particular aspect of something (a possibility of seeing this as that) without getting him actually to see this aspect. (Cf. proving the possibility of proving an arithmetical or geometrical proposition.) (h) Quasi-objectivity: Aspect-seeing calls for refinement of the notion of what is objective.4 On the one hand, we are tempted to say that different aspects are there to be seen in what is open to view. (The duck–rabbit can be seen as a duck; that aspect is available, just as 4 can be regarded as the square of 2 but not as the square of 3.) On the other hand, a visual aspect does not confront us like Mont Blanc; it cannot be imposed on us against our will (at least not by reasoning grounded in observation or experiment, though perhaps by indoctrination or brainwashing), and failing to see it is not a defect of vision. Aspects are not imaginary or mere creatures of the mind; but, unlike perceptual experiences, they are subject to the will. For this reason, aspect-seeing might be said to be half perception, half imagination. The term ‘subjective’ might capture one aspect of the concept of aspect-seeing, the term ‘objective’ another; but each used alone seems seriously misleading.5 (i) Open-endedness: The set of aspects of a given thing is essentially open-ended; hence the possibilities cannot be exhaustively enumerated (there is no ‘closure’ here). This point is related to their being voluntary and, in part, exercises of the imagination. (j) Rational persuasion without demonstration: Solving a picture puzzle, or engaging in ‘artistic appreciation’, may turn on getting another person to see an aspect to which he is now blind, e.g. the biplanar structure of Monet’s Nympheas (the role of the flowers and leaves of the water-lilies in creating the contrast between the flat surface of the water and the depth of the reflections). This kind of discussion is distinctive; it is a form of rational persuasion without the possibility of proof. Such non-demonstrative rational persuasion can employ various methods: e.g., surrounding the ‘target’ with drawings or models of what it is to be seen as (e.g. surrounding the duck–rabbit with pictures of rabbits, some in appropriate ‘attitudes’); drawing attention to certain features of the ‘target’, and de-emphasizing others (e.g. inking in certain lines in a pencil drawing, or partially erasing others; the point is not to establish visual facts about the drawing that exclude seeing it as a duck); rearranging, gradually deforming, or distorting various elements in the

Grammar of Aspects and Aspects of Grammar

283

‘target’ (e.g. by producing a series of drawings intermediate between the ‘target’ and what it is to be seen as (famously, a portrait of Louis Philippe and a picture of a pear; more recently, a photograph of a certain American president and one of a chimpanzee)); and so on. All of these can perhaps be subsumed under the general heading: suggesting objects of comparison. None of them guarantees success. (None carries ‘the force of argument’.) The interlocutor may remain entrenched in his way of seeing the thing (unaware, of course, that he is seeing the thing in a way – as opposed to simply seeing it). He may not try to see things differently; perhaps he will refuse to make the effort.

Some Aspects of the Grammar of ‘‘Auffassung’’ Waismann attempts in ‘How I See Philosophy’ to clarify Auffassungen in philosophy by means of a comparison with visual aspect-seeing. These are variously called pictures, analogies, similes, Betrachtungsweise, Auffassungsweise, etc. Paradigms of such philosophical Auffassungen might include: 1a Thinking is an inner accompaniment to speaking. 1b Thinking is operating with signs. 2a The meaning of a word is the object for which it stands. 2b The meaning of a word is its use in the language. 3a Instances of concepts are unified by Merkmal-definitions. 3b Instances of concepts may be unified by a kind of ‘family resemblance’. 4a Grammar is something superficial. 4b Essence is expressed by grammar. 5a Mathematical equations describe incorporeal objects. 5b Mathematical equations are rules of grammar.

Analogies between aspects and conceptions All of the points made above about the concept of visual aspect-seeing seem to hold equally for the wider use of ‘aspect’ or ‘way of seeing things’ in application to conceptions in philosophy: (a) Global pluralism: Here too, no conception has exclusionary claimrights; no-one can claim his conception to be the only possibility. We can look at meaning in terms of the Augustinian picture; this does not

284 Wittgenstein and Waismann

exclude the possibility of our also looking at meaning as use (on another occasion) – or vice versa. To acknowledge one conception of meaning does not render illegitimate a different conception of it. Indeed, just as to speak of one visual aspect presupposes that there are others, so too, to speak of one conception presupposes that there are others. Conceptions, like aspects, are essentially plural. (b) Local incompatibility: Like visual aspects, conceptions are transiently exclusionary; it is impossible to see thinking simultaneously as an inner accompaniment to speaking and as operating with signs (cf. BB 7 ff.). Conceptions too are essentially non-additive: attempting to combine them produces, not a more comprehensive way of looking at a concept, but muddle. That conceptions interfere with one another has enormously important consequences in philosophy. While it is imaginable that someone could get so ‘hooked’ on one visual aspect that he was unable in practice to see any other, the consequences would, it seems, be minimal. But when a philosophical Auffassung becomes entrenched, it affects the whole conduct of one’s philosophical investigation. Entrenched conceptions can act as brakes to intellectual movement, or restrictions on freedom of enquiry (cf. CV 28). (c) Voluntariness: Like aspect-seeing, ways of conceiving are in a certain sense voluntary (and hence proposing a new Auffassung may be purpose-relative). Thus it makes sense for Wittgenstein to counsel his readers: ‘Try not to think of understanding as a ‘‘mental process’’ at all. – For that is the expression which confuses you’ (PI §154); or ‘Look at the sentence as an instrument, and at its sense as its employment’ (PI §421; cf. §360); or ‘Don’t say: ‘‘There must be something in common, or they would not be called ‘games’ ’’ ’ (PI §66). Hence too, to conceive of things differently is not to discover something (PI §§124–6; cf. BB 23 on the unconscious). Or better: though it might be called a discovery, it is not the kind of discovery we thought it was: ‘The ‘‘visual room’’ seemed like a discovery, but what its discoverer really found was a new way of speaking, a new comparison’ (PI §400). (d) Dependence on concept-mastery: Conceptions are, as it were, ways of looking at concepts. They are patterns discernible in, or ways of organizing, the use of this or that word. (But: ‘an order, not the order’.) Hence, even more clearly than with visual aspect-seeing, proposing and acknowledging conceptions presupposes the mastery of concepts. (e) ‘Homoeopathy’: An Auffassung can be displaced only by another Auffassung. No feature of our actual use of a word (say, ‘meaning’ or ‘thinking’) can logically exclude my ‘seeing’, say, meaning as naming or thinking as an inner process; and nothing but seeing the concept under

Grammar of Aspects and Aspects of Grammar

285

another conception can interrupt my seeing it in this way. For example, someone might be inclined to say that every sentence must be composed of parts; one might be tempted to see this conception as refutable by such ‘counterexamples’ as the word ‘Restaurant’ on a sign above the entrance to a building. But the person gripped by this conception may, justifiably, respond that ‘it is not the word alone but the word together with the whole situation in which it is integrated that makes up a sentence’ (PLP 317). Corollary: An Auffassung cannot be refuted by pointing to any fact. Nonetheless, a way of seeing things may be damaging or pernicious in a way that is harder to imagine with visual aspects. How? Augustine’s picture does not assimilate all uses of words. Rather, it assimilates the descriptions of all uses of words to a single pattern: we must in every case explain what a word designates. (This is a very different thing.) Far from denying that words have different roles in Satzbau (§664), it recognizes differences in use in terms of differences among the objects named by words.6 Nonetheless, it may encourage or invite neglect of differences in use (§1), and it is in this way that it may ‘surround the working of language with a haze’ (cf. PI §5). (f) The possibility of ‘blindness’: Like visual aspects, Auffassungen may be invisible to one individual, visible to another. Perhaps more obviously with conceptions than with visual aspects, an Auffassung may be invisible to one generation or culture, visible and even salient for another. Thus there can be conception-blindness as well as aspect-blindness. One may be unable to notice something, even though it is always before one’s eyes and in plain view. Hence ‘we fail to be struck by what, once seen, is most striking and most powerful’ (PI §129). Or conversely, we may now fail to be struck by what used to seem most striking and most powerful; an Auffassung may become completely invisible, perhaps through a pervasive change in the climate of thinking. This ‘paradox’ haunts the history of philosophy; for instance, modern ethicists have largely lost sight of the Aristotelian and scholastic conception of volition as rational choice; modern philosophers of mind have lost sight of the idea of seeing freedom and moral values as modes of thinking, of the role of the faculties in calling things into doubt, and of a conception of pain that makes pain a condition of the body. A consequence of this disappearance is that we fail to see these ideas in, say, Descartes’s writings, although they are in plain view. Such blindness is compatible with the greatest intelligence and sophistication in reasoning. Making such a conception visible requires waging war against the habits of a lifetime, and probably against the spirit of the culture in which one lives.7 We need, for example, to learn to mount stiff resistance to ‘our preoccupation with the method of science’ (BB 18). We need to have the courage to stand against the force of the main current of Western

286 Wittgenstein and Waismann

civilization and its fascination with the ideal of progress (PR Preface; PI motto). The goal is nothing less than effecting a ‘radical conversion’, a transformation in ways of thinking.8 (g) A paradox of modality: As with aspects, I cannot demonstrate to somebody that there is a possibility of seeing a concept in a particular way without getting him actually to see it in this way. (h) Quasi-objectivity: Conceptions exhibit the same kind of ‘subjective objectivity’ as visual aspects. Though, like aspects, the adoption of a way of conceiving is in a certain sense voluntary (and in that sense ‘subjective’), it is possible, with an effort of imagination, to work within an alien Auffassung, e.g. scholastic metaphysics, to make sense of their manoeuvres and strategies, of their problems and solutions. (i) Open-endedness: The possible conceptions of this or that concept cannot be exhaustively enumerated – any more than all the possible objects of comparison can be exhaustively enumerated. (j) Rational persuasion without demonstration: The methods for getting another to look at things differently have a general similarity with the ones used for making visual aspects visible to another person: e.g., emphasis, rearrangement, constructing intermediate cases, offering new objects of comparison. I return to this point below.

Disanalogies between aspects and conceptions The similarities are important, but there are also some vital differences. In old-fashioned Faculty-speak, (visual) aspects can be said to stand to the senses as Auffassungen stand to the intellect (or pure reason). This has certain implications: (a) Any Auffassung can be articulated (e.g. PI §81: the picture according to which anybody who says something and means or understands it must be operating a calculus according to definite rules). (b) A philosophical Auffassung or picture may be unconscious yet operative. Someone can be aware of a visual aspect without being aware that it is an aspect (i.e. that there are other ways of seeing the thing in question); but arguably an Auffassung can be unconscious in a further sense: the person may never have made explicit to himself the picture that is exerting an influence on his thinking. Hence Auffassungen may need to be brought to the subject’s consciousness – perhaps despite resistance. This requires articulating his conception and winning his acknowledgement that this picture is indeed the one that has been operative. (c) Pictures or ways of seeing things have a peculiar power to hold us in thrall or captivity (cf. PI §115). They have this power especially if they

Grammar of Aspects and Aspects of Grammar

287

are unconscious. They exert their influence antecedent to argument or investigation, and they are, like shadows, unshakeable or unassailable (cf. RFM 230). They are patterns to which it seems that reality must conform (PI §131). Hence Wittgenstein compares them to superstitions (e.g. PI §110; CV 83). It never occurs to us that a mere picture could exert any tyranny over our thinking: that it could limit our freedom of thought. But pictures may become entrenched in our ways of thinking (Denkgewohnheiten), so that replacing one picture by another may require great determination and effort. Uprooting prejudices – overthrowing habits of a lifetime – is (as Descartes reminded us) an extremely difficult task. (d) Enslavement to a particular Auffassung, and the consequent blindness to others, may be motivated. We may want to see things in a certain way (CV 28); we may have a drive to impose a particular picture on phenomena (§109). (The workings of the intellect have more dimensions of depth than the workings of the senses.) Hence methods for removing such blindness may need to be far more intricate and indirect. It is a very subtle and wide-ranging task to eliminate serious prejudices: nothing less than implanting a whole new way of thinking (HISP). Cf.: ‘When he was old Charlemagne tried to learn to write, but without success: and similarly, someone may fail when he tries to acquire a manner of thinking. He never becomes fluent in it’ (CV 75). What is required is a kind of conversion (Umstellung der Auffassung) or re-education (cf. Freud). Conceptions may be damaging because they make certain lines of thinking impossibly arduous (or unnatural) to pursue. (They exert a palpable tyranny without any apparent restriction of freedom!) (e) The choice of conception may be the topic of rational argument. One may offer reasons for trying to see something in a particular way. If a particular worry or confusion can be traced to one Auffassung, then the prospect of dissolving this problem might be a reason for exploring another Auffassung. But this form of rational argument is nondemonstrative (like giving moral advice or assessing aesthetic values of particular works of art). Nothing is proved; rather, a discussion is undertaken in order to make one option appear more eligible. (This seems a paradigm of rational persuasion, comparable to trying to sell somebody a product on its merits; or to persuade him to pursue a particular course of study in university.)

An Illustration: Samples as Part of Language I want to suggest that Auffassungen, as just characterized, can be seen as playing a vital role in Wittgenstein’s philosophical method. I begin by

288 Wittgenstein and Waismann

working through a particular example. Consider the conception according to which ostensive definition forges a connection between language and reality, hence leads us out of language (PLP 277): ‘on the one hand, we have the sign, on the other, reality’ (PLP 278). Waismann offers an alternative Auffassung: We put forward the following terminology; we suggest that in the case in which the word ‘red’ is explained by means of a colour spot we shall say that one sign is explained by means of another, that a word-sign is explained by a sign of another sort. . . . The colour spot will be used by us in future as a standard of comparison, a paradigm, and is, in this sense, part of language. (PLP 278)9

This is overtly the proposal of a different point of view; a stipulation about how we (Wittgenstein, Waismann and like-minded philosophers) will use the term ‘sign’. Compare: It is natural for us to call gestures [in (4)] or pictures [in (7)] elements or instruments of language. . . . The pictures in (7) and other instruments of language which have a similar function I shall call patterns. . . . We may say that words and patterns have different kinds of functions. (BB 84)10

But any reader is at liberty to refuse to go along with this suggestion, i.e. to persist with the ‘usual view’.11 He might even justify this persistence by pointing to important differences between words and samples; e.g. we cannot copy from a word as we can from a sample (PLP 278), and we cannot compare an object with a word as we can compare it with a sample (BB 84). Thus the new conception of signs is optional or voluntary. It would not manifest clear misunderstanding of ‘sign’ to refuse to accept this proposal. Hence, ‘By saying that samples are a part of language we do not wish to make a dogmatic assertion, we wish merely to draw attention to the similarity between an ostensive definition and an explanation of the form ‘‘a ¼ bDf ’’. We wish to regard both explanations in the same way’ (PLP 278). This alternative Auffassung is not an opinion. Rather, it explicitly treats word-signs as objects of comparison for samples, and it offers an Auffassung of signs, namely that any object can be called a sign if it is used as such (PLP 278). (Consider the use of concrete signs, especially in codes, e.g. ‘One if by land, two if by sea’; cf. TLP 3.1431.) It is a functional conception of signs. (There are other conceptions: for example, that things other than signs can be used to convey information (or to express

Grammar of Aspects and Aspects of Grammar

289

things), e.g. facial expressions, winks and nods, or the gesture of pointing.) Note that the concept of a sign admits of being looked at in various ways: ‘If we abide by the use of language we can say only that the concept [‘sign’] is not exactly delimited. It is best then to give a few typical examples and leave it open to the person to whom we are talking to decide how far he allows the concept to extend’ (PLP 109). This concept, arguably like all philosophically interesting concepts, is essentially openended, even ‘essentially contestable’. Though it doesn’t at all follow that ‘anything goes’, it does follow that facts about our use of the word ‘sign’ are not going to decide between different stipulations about ‘how far [to allow] the concept to extend’. Various conceptions of a sign might go hand in hand with various conceptions of ‘language’ and ‘grammar’. The word ‘language’ will be used here as a term applying to everything that serves the end of expression and communication. It is of set purpose that we extend this concept to cover not only word-languages but also gesture language, picture language, the ‘language’ of maps, the formal ‘languages’ of mathematics and logic, besides signal systems and much else. It follows from the convention that all communication takes place by means of language. (PLP 93)

It follows too that the samples used in PI §8 are elements of language (8). Again, whether we count ostensive definition as part of grammar depends on how we wish to explain ‘grammar’ and ‘meaning’ and the relation between the two notions. There are clearly different possibilities. If I say ‘The meaning of a word is determined by its grammar’ [one Auffassung], then an explanation of the meaning of a word must be the explaining of everything which distinguishes the meaning of this word from the meaning of other words and all this must be considered as part of grammar. In this case the ostensive definition of the words ‘red’ and ‘green’ will be part of their grammar. (PLP 279)

With both questions (are samples part of language? is ostensive definition part of grammar?) we might say, ‘Well, it is as you please’ (PI §16). The task of PLP’s discussion is to persuade us that there are many things that we don’t want to say (e.g. that we don’t want to ‘lay down that the meaning of the word ‘‘red’’ is nothing but the place that this word occupies in the grammar [combinatorics or syntagmatics] of the language of words’ (PLP 278–9)).

290 Wittgenstein and Waismann

This whole discussion is directed at undermining the prejudice that ‘explanation by ostensive definition is something utterly [essentially] different from other sorts of explanation’ (PLP 277). This is achieved in a distinctive way: not by establishing any linguistic facts, not by arguing that it would be incorrect (or in conflict with ordinary usage) to deny that samples are signs or that grammar includes ostensive definition. Rather, it is accomplished by drawing attention to similarities and differences to motivate acceptance of stipulations about how to use ‘sign’, ‘language’, ‘grammar’ in cases where the application of these terms is not determined in advance and hence (essentially) contestable.12 Accepting another possibility (an alternative Auffassung) as a possibility is equivalent to dropping the prejudice that ostensive definition is essentially different from every other kind of explanation.

¨ bersicht of Wittgenstein’s Method of Philosophical U Investigation First, a general question: What does Wittgenstein mean by ‘describing the grammar of our language’? This might cover a wide variety of things – just as in the case of describing a painting, a room, a character in a novel, etc. A catalogue of rules? of facts about standard English? For what purpose? Regulation of Philosophy-Speak? What is ‘l’esprit des re`gles’? Wittgenstein highlights pictures, analogies, Auffassungen. Revealing new aspects of what is perfectly familiar, developing new Auffassungen, is the general method of Wittgenstein’s philosophical investigations. Wittgenstein traces problems (confusions, worries) back to unconscious pictures or analogies, and his method for dissolving these problems is to persuade another to acknowledge such prejudices and to replace them by a different Auffassung. This is a matter of establishing an order in our knowledge of the use of language; not of establishing any linguistic facts about a public practice. (Here is a clear contrast between two conceptions of ‘describing the grammar of our language’.) The pursuit of such an ‘order’ or ‘arrangement’ is the pursuit of ¨ bersichtlichkeit. Hence ‘The concept of u U ¨ bersichtliche Darstellung is of fundamental importance for us. It earmarks the form of account we give, the way we look at things’ (PI §122). (It is this conception which characterizes unsere Darstellungsform, i.e. what we (Wittgenstein et al.) call an adequate description of grammar. The desideratum is not compactness, but pattern-presentation. Cf. Goethe’s Urpflanze.) Philosophical problems are traced to prejudices, and these are addressed by proposing other ways of seeing things.

Grammar of Aspects and Aspects of Grammar

291

This may yield new understanding of what has always been in full view (PI §89). ‘A perspicuous representation [u ¨ bersichtliche Darstellung] produces just that [form of] understanding which consists in ‘‘seeing connections’’ ’ (PI §122). ‘Ich kenne mich nicht aus’ (PI §123) is the lack of this form of understanding (not ignorantia)! And it is remedied by establishing an order (Ordnung) in our knowledge of the use of our language (PI §132); that is what eine u ¨ bersichtliche Darstellung is (PI §122). (These are cognitive achievements of the greatest potential worth. They demand creativity, originality, imagination – of a very high order.) But valuing – or even grasping – Wittgenstein’s descriptions of grammar may depend on a total reorientation about what counts as real understanding in philosophy – or its absence. The understanding yielded by pattern-recognition is strictly localized. An aspect or conception always pertains to something in particular. What is an illuminating aspect of this thing may not belong to that thing at all, or even if it does, it may yield no understanding. Nothing need be transferable from one thing to another, and the importance of this aspect or conception for understanding this thing is not diminished by lack of transferability. Conversely, apprehending a new aspect or conception of something is an insight or vision – even a sudden illumination. (It is an exercise of imagination and ingenuity, sometimes genius.) On some occasions, when we look at things in a certain way, unexpectedly they seem to change as though by magic (HISP 4; cf. CV 11). The principal business in ‘describing grammar’ on this conception is making neglected aspects or conceptions visible to others who are blind to what is in full view. All are presented as pictures, Auffassungen, Betrachtungsweise, objects of comparison, models, etc. – not as discoveries, not as truths. Adopting any of them is voluntary. Refusing to accept any one of them would not manifest misunderstanding of any words; e.g. of ‘essence’ or ‘samples’. Nor would it count as ignorantia. The form of understanding sought by Wittgenstein (cf. PI §89) does not appear on most maps of knowledge in the theory of knowledge. Pattern-recognition seems not to fit at all into the conception of knowledge which has dominated epistemology. Aspects, Auffassungen, ways of seeing things, or prejudices are difficult to fit into the prevalent philosophical ideas of rationality and objectivity: the attempt to establish things by force of argument; the possibility of proving something, once for all; and the possibility of demonstrating that some claims are definitely wrong. ‘QED’: and whether another accepts or rejects the conclusion is a merely psychological problem. ‘Seeing connections’ has the power to transform a way of seeing things (be it ostensive definition or arithmetic or pain) which makes them seem

292 Wittgenstein and Waismann

singular and hence disturbing. These ‘connections’ are as much invented as discovered; they are comparable to seeing analogies.13 But aspects or prejudices seem to lie outside the reach of demonstration or refutation, hence outside the usual conception of the proper business of philosophy. If the discussion here carries conviction, it ought to be possible to persuade philosophers that the standard way of conceiving philosophical understanding is itself one Auffassung to which there are alternatives; one such alternative is what has been sketched here.

Notes 1 It might also imply the possibility of seeing the thing sans aspect. An aspect may dawn out of chaos (Regellosigkeit). 2 Wittgenstein suggests (PI p. 207) that the aspects of the double cross ‘might be reported simply by pointing alternately to an isolated white and an isolated black cross’, and that ‘One could quite well imagine this as a primitive reaction in a child even before it could talk’. This may suggest that some aspect-seeing does not require the mastery of a concept. 3 Wittgenstein notes (PI p. 208) that this may seem paradoxical, even ‘crazy’ (p. 209). It may be tempting to say that the same point applies to many visual ‘experiences’. For example, one might argue, to be able to see a Rembrandt, on one understanding of this phrase, requires familiarity with other works of Rembrandt and the ability to recognize features of his style. Yet this might simply show that the concept of seeing has yet further complexities. There are analogies between this case and that of aspect-seeing: to ‘see a Rembrandt’ in the sense here picked out is to see resemblances between this painting and other paintings; one might even express this experience in the form ‘Now I see . . . ’ (e.g., ‘Now I see what my art teacher was trying to get me to see’) or describe someone as ‘seeing the painting as a Rembrandt’. 4 There might be parallel things to be said here about Nietzschean ‘perspectives’. 5 Making sense of this feature of aspect-seeing is a modern conceptual problem reminiscent of the scholastics’ difficulties in developing a satisfactory picture of entia rationis. The temptation is to dismiss such things as subjective or psychological (e.g. FA 26), especially when they are described as having ‘no existence outside the intellect’ (in standard scholastic jargon). We may be tempted to speak pejoratively of ‘mere aspects’, and to dismiss them from philosophy (a cognitive enquiry), as having no place in reasoning or argument. 6 Hence Frege distinguished the use of natural numbers from the use of rationals: namely as answers to the different questions ‘How many?’ and ‘How much?’. This distinction is basic to his whole analysis of numberwords. But he didn’t treat this difference as an explanation of what numbers are; rather it is something that calls for a deeper explanation in terms of what objects different kinds of number-words designate.

Grammar of Aspects and Aspects of Grammar 7

8

9

10

11

12

13

293

This last point may be the only place where Wittgenstein’s and Waismann’s conceptions of ‘our method’ sharply diverge. Waismann did not express this form of pessimism in his philosophical writings, and he applied ‘our method’ largely to the clarification of concepts in mathematics and physics. ‘I am by no means sure that I should prefer a continuation of my work by others to a change in the way people live which would make all these questions superfluous’ (CV 61). From this point of view, the colour spot is not something described, but part of a means of representation (PLP 278). (Cf. BB 175: ‘the error . . . of believing that an ostensive definition says something about the object to which it directs our attention’.) BB 109: ‘What can be compared with a name is not the word ‘‘this’’ but, if you like, the symbol consisting of this word, the gesture [of pointing], and the sample’. This can be seen as a substitution-rule for symbols. And the same treatment could be given to an ostensive definition of ‘red’. This is not a (dogmatic) assertion, but a comparison to serve the purpose of getting rid of a particular puzzlement (cf. PLP 278). Compare: ‘In introducing the distinction ‘word/pattern’, the idea was not to set up a final logical duality. We have only singled out two characteristic kinds of instruments from the variety of instruments of our language’ (BB 84). But we can easily imagine that somebody would refuse to classify these things together (cf. BB 83). It seems natural to think that we would explain these terms by giving examples, and that the rider ‘and similar things’ leaves latitude (Spielraum!) for developing different conceptions within the scope of these explanations. These concepts, like the concept of number (PLP 278), are essentially openended. Contrast this conception of argument with the idea that arguments by analogy are defective – because not coercive.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

Bibliography of the Works of Gordon Baker

Books Sole author 1988

Wittgenstein, Frege and the Vienna Circle. Oxford: Blackwell.

Co-author with P. M. S. Hacker 1980 1983 1983 1984 1984 1984 1985

Wittgenstein: Understanding and Meaning: An Analytical Commentary on the ‘Philosophical Investigations’, vol. 1. Oxford: Blackwell. An Analytical Commentary on Wittgenstein’s ‘Philosophical Investigations’. Chicago, IL: University of Chicago Press. Wittgenstein: Meaning and Understanding: Essays on the ‘Philosophical Investigations’. Chicago: University of Chicago Press. Frege: Logical Excavations. Oxford: Blackwell. Language, Sense and Nonsense. Oxford: Blackwell. Scepticism, Rules and Language. Oxford: Blackwell. Wittgenstein: Rules, Grammar and Necessity: An Analytical Commentary on the ‘Philosophical Investigations’, vol. 2. Oxford: Blackwell, 1985 (hardback); 1988 (paperback); 1992 (reprinted, with corrections).

Co-author with K. J. Morris 1996

Descartes’ Dualism. London: Routledge (paperback version, 2002). Descartes’ Meditations: First Philosophy, forthcoming.

Works of Gordon Baker 295

Editor 1976 F. Waismann: Logik, Sprache, Philosophie, Stuttgart: Reclam. (Editor (with B. F. McGuinness), and author of Nachwort (Afterword)). 1997 Dicte´es de Wittgenstein a` Waismann et pour Schlick, vol. 1 (texte). Paris: Presses Universitaires de France (translator and editor (with B. F. McGuinness, under the direction of Antonia Soulez), and author of Pre´sentation (Preface)). 2003 The Voices of Wittgenstein. London: Routledge (general editor and one of four translators).

Other Publications 1966 ‘Rules, Definitions, and the Naturalistic Fallacy’ (with P. M. S. Hacker), American Philosophical Quarterly, 3: 299–305. 1974 ‘Criteria: a New Foundation for Semantics’, Ratio, 16: 156–89. 1974 ‘Kriterien: Eine neue Grundlegung der Semantik’ (in German), Ratio, 16: 142–74 (German translation of above article). 1976 ‘Critical Notice: Philosophical Grammar, by Ludwig Wittgenstein’ (with P. M. S. Hacker), Mind, 85: 269–94. 1977 ‘Defeasibility and Meaning’, in P. M. S. Hacker and J. Raz (eds), Law, Morality, and Society: Essays in Honour of H. L. A. Hart, Oxford: Clarendon Press, pp. 26–57. 1979 ‘Verehrung und Verkehrung: Waismann and Wittgenstein’, in C. G. Luckhardt (ed.), Wittgenstein: Sources and Perspectives, Ithaca, NY: Cornell University Press, 243–85. 1980 ‘Wittgenstein aujourd’hui’ (with P. M. S. Hacker) (in French), Critique, 690–704. 1981 ‘Following Wittgenstein: Some Signposts for Philosophical Investigations §§143–242’, in S. H. Holtzman and C. M. Leich (eds), Wittgenstein: To Follow a Rule, London: Routledge and Kegan Paul, pp. 31–71. 1982 ‘The Grammar of Psychology: Wittgenstein’s Bemerkungen u¨ber die Philosophie der Psychologie’ (with P. M. S. Hacker), Language and Communication, 2 (3): 227–44. 1982 ‘Wittgenstein’ (with P. M. S. Hacker) (in Hebrew), in A. Kasher and S. Lappin (eds), Modern Trends in Philosophy I, 205–30. 1983 ‘Dummett’s Purge: Frege without Functions’ (with P. M. S. Hacker), Philosophical Quarterly, 33: 115–32. 1983 ‘Dummett’s Frege or Through a Looking-Glass Darkly’ (with P. M. S. Hacker), Mind, 92: 239–46. 1983 ‘The Concept of a Truth-Condition’ (with P. M. S. Hacker), Conceptus, 17 (40/41): 11–18.

296 Works of Gordon Baker 1984 1985 1985 1985

1985

1985

1986 1986 1986 1986 1987 1988 1989 1989 1990 1990 1991

1992

‘Critical Study: On Misunderstanding Wittgenstein: Kripke’s Private Language Argument’ (with P. M. S. Hacker), Synthese, 58: 407–50. ‘Wittgenstein and the Vienna Circle: the Exaltation and Deposition of Ostensive Definition’ (with P. M. S. Hacker), Teoria, 5: 5–33. ‘Frege’ (with P. M. S. Hacker) (in Hebrew), in A. Kasher and S. Lappin (eds), Modern Trends in Philosophy II, 302–34. ‘Wittgenstein’s Demythologization of Recognition: an Indictment of Logical Empiricism’, in H.-J. Dahms (ed.), Philosophie, Wissenschaft, Aufkla¨rung: Beitra¨ge zur Geschichte und Wirkung des Wiener Kreises, Berlin and New York: Walter de Gruyter, pp. 81–100. ‘Echte en zogenaamde regels I: taaltheorien’ (with P. M. S. Hacker) (in Dutch) [original English title: ‘Rules and Pseudorules I: Theories of Language’], Wijsgerig Perspectief op Maatschappij en Wetenschap, 26 (6): 186–92. ‘Echte en zogenaamde regels II: taal en zogenaamde regels’ (with P. M. S. Hacker) (in Dutch), Wijsgerig Perspectief op Maatschappij en Wetenschap, 26 (6): 192–97. ‘Reply to Mr Mounce’s ‘‘Following a Rule’’ ’ (with P. M. S. Hacker), Philosophical Investigations, 9: 199–204. ‘Alternative Mind-styles’, in R. E. Grandy and R. Warner (eds), Philosophical Grounds of Rationality, Oxford: Clarendon Press, pp. 277–314. ‘Philosophia: Eikon Kai Eidos’ (in English), in S. G. Shanker (ed.), Philosophy in Britain Today, Albany, NY: SUNY Press, pp. 1–57. ‘Moderne Sprachtheorien aus philosophischer Sicht’ (in German), Sprache der Gegenwart, 81: 77–98. ‘Dummett’s Dig: Looking-glass Archaeology’ (with P. M. S. Hacker), Philosophical Quarterly, 37: 86–99. ‘Bertrand Russell’, in R. Harris (ed.), Linguistic Thought in England (1914–1945), London: Duckworth, pp. 27–59. ‘The Last Ditch’ (with P. M. S. Hacker), Philosophical Quarterly, 39: 471–477. ‘Frege’s Anti-Psychologism’ (with P. M. S. Hacker), in M. A. Notturno (ed.), Perspectives on Psychologism, Leiden: Brill, pp. 75–127. ‘Malcolm on Language and Rules’ (with P. M. S. Hacker), Philosophy, 65 (252): 167–79. ‘La Re´ception de l’Argument d’un Langage Prive´ ’ (in French), in T. E. R. (ed.), Acta du Colloque Wittgenstein 1988, pp. 29–40 ‘Philosophical Investigations §122: Neglected Aspects’, in R. L. Arrington and H.-J. Glock (eds), Wittgenstein’s ‘Philosophical Investigations’: Text and Context, London: Routledge, pp. 35–68. ‘Some Remarks on ‘‘Language’’ and ‘‘Grammar’’ ’, in J. Schulte and G. Sundholm (eds), Criss-crossing a Philosophical Landscape: Essays on Wittgensteinian Themes Dedicated to Brian McGuinness, Amsterdam: Rodopi, pp. 107–31.

Works of Gordon Baker 297 1992 ‘La Me´thode de Wittgenstein et l’Argument d’un Langage Prive´’ (in French), in J. Sebestik and A. Soulez (eds), Wittgenstein et la Philosophie aujourd’hui, Paris: Me´ridiens Klincksieck, pp. 261–72. 1993 ‘Re´cherches Philosophiques: l’Argument d’un Langage Prive´’ (in French), in J.-P. Leyvraz and K. Mulligan (eds), Wittgenstein Analyse´, Nimes: Jacqueline Chambon, pp. 238–95. 1993 ‘Prophetic Glimmerings: the New Pythagorean’ (with P. M. S. Hacker), in H. D. Sluga (ed.), General Assessments and Historical Accounts of Frege’s Philosophy, New York: Garland, pp. 315–42. 1993 ‘Descartes Unlocked’ (with K. J. Morris), British Journal for the History of Philosophy, 1 (1): 5–28. 1994 Critical Notice: ‘John Cottingham’s A Descartes Dictionary’ (with K. J. Morris), International Journal of Philosophical Studies, 2 (1): 116–22. 1997 Historical and analytical Preface to F. Waismann, The Principles of Linguistic Philosophy, Basingstoke: Macmillan, pp. xi–xxiii. 1997 ‘Reply to Nadler’ (with K. J. Morris), Philosophical Books, 38: 164–69. 1997 ‘ ‘‘Notre’’ me´thode de penser sur la ‘‘pense´e’’ ’, in Dicte´es de Wittgenstein a` Waismann et pour Schlick, vol. 2 (Etudes critiques), Paris: Presses Universitaires de France, 283–319. 1998 ‘The Private Language Argument’, Language and Communication, 18: 325–356. 1999 ‘Italics in Wittgenstein’, Language and Communication, 19: 181–211. 1999 ‘A Vision of Philosophy’, in Figuras do Racionalismo, Conferencias ANPOF 1999: 133–77. 2000 ‘The Senses as Witnesses’, in S. Gaukroger, J. Schuster and J. Sutton (eds), Descartes’ Natural Philosophy, London and New York: Routledge, pp. 604–29. 2000 Translation into Japanese of Language, Sense and Nonsense (co-author with P. M. S. Hacker). 2001 ‘Wittgenstein: Concepts or Conceptions?’, Harvard Review of Philosophy, 9 (Spring): 7–23. 2001 ‘ ‘‘Function’’ in Frege’s Begriffssschrift: Dissolving the Problem’, British Journal for the History of Philosophy, 9 (3): 525–44. 2001 ‘Wittgenstein’s ‘‘Depth Grammar’’ ’, Language and Communication, 21: 303–19. 2002 ‘Quotation-marks in Philosophical Investigations Part I’, Language and Communication, 22: 37–68. 2002 ‘Wittgenstein on Metaphysical/Everyday Use’, Philosophical Quarterly, 52 (208): 289–302. 2002 ‘Russell and Wittgenstein on Everyday Use’, forthcoming in Essays on Russell and Wittgenstein, ed. R. Harre´ and J. Shosky, Prague: Filosofia. 2003 ‘Wittgenstein’s Method and Psychoanalysis’, Nature and Narrative, ed. K. W. M. Fulford et al., vol. 1 of the International Perspectives in Psychiatry and Philosophy series from Oxford University Press, pp. 57–73.

298 Works of Gordon Baker 2003

‘Friedrich Waismann: A Vision of Philosophy’, Philosophy, 78: 163–79. ‘Wittgenstein and Russell’ (forthcoming). ‘Meditation I and the Logic of Testimony’ (with K. J. Morris), forthcoming in British Journal for the History of Philosophy.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

General Bibliography

Abbreviations Works by Ludwig Wittgenstein PUBLISHED WORKS The following abbreviations are used to refer to Wittgenstein’s published works, listed in chronological order (where possible; some works straddle many years). The list includes derivative primary sources and lecture notes taken by others. NB TLP RLF WWK

PR M LWL PG

Notebooks 1914–16, ed. G. H. von Wright and G. E. M. Anscombe, trans. G. E. M. Anscombe. Oxford: Blackwell, 1961. Tractatus Logico-Philosophicus, trans. D. F. Pears and B. F. McGuinness. London: Routledge and Kegan Paul, 1961. ‘Some Remarks on Logical Form’, Proceedings of the Aristotelian Society, suppl. vol. ix (1929), pp. 162–71. Ludwig Wittgenstein und der Wiener Kreis, shorthand notes recorded by F. Waismann, ed. B. F. McGuinness. Oxford: Blackwell, 1967. The English translation, Wittgenstein and the Vienna Circle (Oxford: Blackwell, 1979), matches the pagination of the original edition. Philosophical Remarks, ed. R. Rhees, trans. R. Hargreaves and R. White. Oxford: Blackwell, 1975. ‘Wittgenstein’s Lectures in 1930–33’, in G. E. Moore, Philosophical Papers. London: Allen and Unwin, 1959. Wittgenstein’s Lectures, Cambridge, 1930–32, from the notes of John King and Desmond Lee, ed. Desmond Lee. Oxford: Blackwell, 1980. Philosophical Grammar, ed. R. Rhees, trans. A. J. P. Kenny. Oxford: Blackwell, 1974.

300 General Bibliography GB

AWL

BB LPE LSD

RFM

LA LFM PI Z EPB UW NL

RPP I

RPP II

LPP

LW

C

‘Remarks on Frazer’s ‘‘Golden Bough’’ ’, trans. J. Beversluis, repr. in C. G. Luckhardt (ed.), Wittgenstein: Sources and Perspectives. Ithaca, NY: Cornell University Press, 1979, pp. 61–81. Wittgenstein’s Lectures, Cambridge, 1932–35, from the notes of Alice Ambrose and Margaret MacDonald, ed. Alice Ambrose. Oxford: Blackwell, 1979. The Blue and Brown Books. Oxford: Blackwell, 1958. ‘Wittgenstein’s Notes for Lectures on Private Experience and ‘‘Sense Data’’ ’, ed. R. Rhees, Philosophical Review, 77 (1968), pp. 275–320. ‘The Languages of Sense Data and Private Experience’ (Notes taken by R. Rhees of Wittgenstein’s lectures, 1936), Philosophical Investigations, 7 (1984), pp. 1–45, 101–40. Remarks on the Foundations of Mathematics, ed. G. H. von Wright, R. Rhees, and G. E. M. Anscombe, revised edition. Oxford: Blackwell, 1978. Lectures and Conversations on Aesthetics, Psychology and Religious Beliefs, ed. C. Barrett. Oxford: Blackwell, 1970. Wittgenstein’s Lectures on the Foundations of Mathematics, Cambridge 1939, ed. C. Diamond. Sussex: Harvester Press, 1976. Philosophical Investigations, ed. G. E. M. Anscombe and R. Rhees, trans. G. E. M. Anscombe. Oxford: Blackwell, 1958. Zettel, ed. G. E. M. Anscombe and G. H. von Wright, trans. G. E. M. Anscombe. Oxford: Blackwell, 1967. Eine Philosophische Betrachtung, ed. R. Rhees, in Ludwig Wittgenstein, Schriften 5. Frankfurt: Suhrkamp, 1970. ‘Ursache und Wirking: Intuitives Erfassen’, ed. R. Rhees, trans. P. Winch, Philosophia, vol. 6 (1976), pp. 391–445. ‘Notes on Logic’, in Notebooks 1914–16, 2nd edn, ed. G. H. von Wright and G. E. M. Anscombe, trans. G. E. M. Anscombe. Oxford: Basil Blackwell, 1979. Remarks on the Philosophy of Psychology, vol. 1, ed. G. E. M. Anscombe and G. H. von Wright, trans. G. E. M. Anscombe. Oxford: Blackwell, 1980. Remarks on the Philosophy of Psychology, vol. 2, ed. G. H. von Wright and H. Nyman, trans. C. G. Luckhardt and M. A. E. Aue. Oxford: Blackwell, 1980. Wittgenstein’s Lectures on Philosophy of Psychology, 1946–7, notes by P. T. Geach, K. J. Shah and A. C. Jackson, ed. P. T. Geach. Hemel Hempstead: Harvester Wheatsheaf, 1988. Last Writings on the Philosophy of Psychology, vol. 1, ed. G. H. von Wright and H. Nyman, trans. C. G. Luckhardt and M. A. E. Aue. Oxford: Blackwell, 1982. On Certainty, ed. G. E. M. Anscombe and G. H. von Wright, trans. D. Paul and G. E. M. Anscombe. Oxford: Blackwell, 1969.

General Bibliography CV R PO WiMS

301

Culture and Value, ed. G. H. von Wright in collaboration with H. Nyman, trans. P. Winch. Oxford: Basil Blackwell, 1980. Ludwig Wittgenstein: Letters to Russell, Keynes and Moore, ed. G. H. von Wright. Oxford: Blackwell, 1974. Philosophical Occasions 1912–51, ed. J. Klagge and A. Nordmann. Indianapolis: Hackett, 1993. ‘Wittgenstein – Moritz Schlick’, a transcript dictated to Schlick after the second version of BT.

Reference style: all references to Philosophical Investigations, Part I are to sections (e.g. PI §1), except those to notes below the line on various pages. References to Part II are to pages (e.g. PI p. 202). References to other printed works are either to numbered remarks (TLP) or to sections signified ‘§’ (Z, RPP, LW): in all other cases references are to pages (e.g. LFM 21 ¼ LFM page 21), or to numbered letters (R). NACHLASS All references to unpublished material cited in the von Wright catalogue (G. H. von Wright, Wittgenstein (Oxford: Blackwell, 1982), pp. 35 ff.) are by MS or TS number followed by page number. Wherever possible, the pagination entered in the original document has been used. The Cornell xeroxes in Bodleian are defective; sometimes a dozen or more papers have been omitted. Consequently, where access to the originals or to complete xeroxes has not been possible, some errors of page reference will unavoidably have occurred. For memorability, the following special abbreviations are used. Manuscripts: Roman numerals refer to the eighteen large manuscript volumes (¼ MSS. 105–22) written between 2 February 1929 and 1944. For example, VI. 241 is to volume VI, page 241. Typescripts: BT The ‘Big Typescript’ (TS 213): a rearrangement, with modifications, written additions and deletions, of TS 211, 1933, vi pp. table of contents, 768 pp. All references are to page numbers. Where the page number is followed by ‘v’, this indicates a handwritten addition on the reverse side of the TS page. PPI ‘Proto-Philosophical Investigations’ (this is not Wittgenstein’s title) (TS 220): a typescript of the first half of the prewar version of the Philosophical Investigations (up to §189 of the final version, but with many differences); 1937 or 1938, 137 pp. The shortened title form ‘ProtoInvestigations’ is used freely. All references are to sections (§). PPI (I) The so-called Intermediate Version, reconstructed by von Wright; it consists of 300 numbered remarks; 1945, 195 pp. All references are to sections (§).

302 General Bibliography

Works by Friedrich Waismann PLP LPM GA HISP LSP F* or F and G

DS EW

The Principles of Linguistic Philosophy. London: Macmillan, 1965; 2nd edn, 1996. Lectures on the Philosophy of Mathematics, ed. Grasshoff. Amsterdam: Rodopi, 1982. ¨ tze. Vienna: Gerold, Preface to M. Schlick, Gesammelte Aufsa 1936. ‘How I See Philosophy’, in How I See Philosophy, ed. Harre´. London: Macmillan, 1968. Logik, Sprache, Philosophie, ed. G. P. Baker and B. F. McGuinness, Stuttgart: Reclam, 1976. (with catalogue number of text): dictations to Waismann from Wittgenstein French translation: Wittgenstein: Dicte´es a` Waismann et pour Schlick. Paris: Presses Universitaires de France, 1997. English translation (with German text): Voices of Wittgenstein: Preliminaries to the Vienna Circle Project, ed. Gordon Baker. London and New York: Routledge, 2003. ‘Diktat fu ¨ r Schlick’ (included in above). Ethics and the Will, ed. McGuinness and Schulte. Dordrecht: Kluwer, 1994.

Works by Gottlob Frege FA

The Foundations of Arithmetic, trans. J. L. Austin, 2nd edn. Oxford: Basil Blackwell, 1959. GA Grundgesetze der Arithmetik, begriffsschriftlich abgeleitet, Band I. Jena: Hermann Pohle, 1893. PW Posthumous Writings, ed. H. Hermes, F. Kambartel and F. Kaulbach, trans. P. Long and R. White. Oxford: Basil Blackwell, 1979.

Works by Bertrand Russell AM The Analysis of Mind. London: George Allen and Unwin, 1921. LK Logic and Knowledge: Essays 1901–1950, ed. R. C. Marsh. London: Allen and Unwin, 1956.

Other Works Adam, C. and P. Tannery see under Descartes. Ambrose, A. see under Wittgenstein: published works.

General Bibliography

303

Arrington, R. L. and H. -J. Glock, eds, Wittgenstein’s ‘Philosophical Investigations’: Text and Context. London and New York: Routledge, 1991. Austin, J. L., Philosophical Papers. Oxford: Oxford University Press, 1976. Baker, G. P. and P. M. S. Hacker, Understanding and Meaning. Oxford: Basil Blackwell, 1980. —— , Wittgenstein: Rules, Grammar and Necessity, Oxford: Basil Blackwell, 1985. Boltzmann, L., Theoretical Physics and Philosophical Problems, ed. B. F. McGuinness. Dordrecht and Boston: Reidel, 1974. Bouwsma, O. K., Philosophical Essays. Lincoln: University of Nebraska Press, 1965. —— , Wittgenstein: Conversations 1949–1951, ed. Craft and Hustwit. Indianapolis: Hackett, 1984. ¨ berwindung der Metaphysik durch logische Analyse der Sprache’, Carnap, R., ‘U Erkenntnis, 2 (1932). Collingwood, R. G., An Autobiography. Oxford: Oxford University Press, 1939. Crary, A. and R. Read, eds, The New Wittgenstein. London and New York: Routledge, 2000. Descartes, R., Oeuvres de Descartes, 12 vols, ed. C. Adam and P. Tannery. Paris: Le´opold Cerf, 1897–1910. Diamond, C., The Realistic Spirit. Cambridge, MA: MIT Press, 1991. —— (1976) see under Wittgenstein: published works. Drury, M. O’C., The Danger of Words and Writings on Wittgenstein, ed. Berman, Fitzgerald and Hayes. Bristol: Thoemmes, 1996. —— ‘Some Notes on Conversations with Wittgenstein’, Acta Philosophica Fennica, 28(1976): 22–40. Dummett, M. A. E., Frege: Philosophy of Language. London: Duckworth, 1973. —— , Truth and Other Enigmas. London: Duckworth, 1978. —— , The Logical Basis of Metaphysics. London: Duckworth, 1991. Frege, G., Collected Papers on Mathematics, Logic and Philosophy, ed. Brian McGuinness, trans. Max Black et al. Oxford: Basil Blackwell, 1984. —— , Begriffsschrift: eine der arithmetischen nachgebildete Formelsprache des reinen Denkens. Halle: L. Nebert, 1879. —— , Translations from the Philosophical Writings of Gottlob Frege, trans. P. T. Geach and M. Black. Oxford: Basil Blackwell, 1960. —— , ‘Conceptual Notation’, in Conceptual Notation and Related Articles, trans. and ed. with biography and introduction, T. W. Bynum. Oxford: Clarendon Press, 1972. Freud, S., The Complete Psychological Works of Sigmund Freud, ed. J. Strachey. London: Hogarth, 1953–74. Glock, H.-J., A Wittgenstein Dictionary. Oxford: Blackwell Publishing, 1996. Guttenplan, S., ed., Mind and Language. Oxford: Clarendon Press, 1975. Hacker, P. M. S., Rules, Grammar and Necessity. Oxford: Basil Blackwell, 1985. —— , Insight and Illusion, rev. edn. Oxford: Clarendon Press, 1986. —— , Wittgenstein: Meaning and Mind, Oxford: Basil Blackwell, 1990. Hallett, G., A Companion to Wittgenstein’s Philosophical Investigations, Ithaca, NY: Cornell University Press, 1977.

304 General Bibliography Hanfling, O., Philosophy and Ordinary Language. London: Routledge, 2000. Hare, R. M., The Language of Morals. Oxford: Clarendon Press, 1952. Hertz, H., Die Prinzipien der Mechanik. Leipzig: Barth, 1910. Kenny, A., Wittgenstein, Harmondsworth: Penguin, 1973. —— . Will, Freedom and Power. Oxford: Basil Blackwell, 1975. —— . The Legacy of Wittgenstein. Oxford: Basil Blackwell, 1985. —— . The Metaphysics of Mind. Oxford: Clarendon Press, 1989. Leibniz, G. W., New Essays on Human Understanding, trans. and ed. P. Remnant and J. Bennett. Cambridge: Cambridge University Press, 1996. Lugg, A., Wittgenstein’s Investigations 1–133. London: Routledge, 2000. Malcolm, N., ‘Wittgenstein’s Philosophical Investigations’, in Philosophical Review, 63, 1954. —— . Nothing is Hidden. Oxford: Basil Blackwell, 1986. —— . Wittgensteinian Themes: Essays 1978–1989. Ithaca, NY: Cornell University Press, 1995. McGuinness, B. F. see under Wittgenstein: Published Works. Monk, R., Ludwig Wittgenstein: The Duty of Genius. London: Jonathan Cape, 1990. Morris, K. J., ‘The ‘‘Context Principle’’ in the Later Wittgenstein’, Philosophical Quarterly, 44 (1), 1994. Pitcher, G., ed., Wittgenstein: The Philosophical Investigations, London: Macmillan, 1966. Rundle, B., Wittgenstein and Contemporary Philosophy of Language. Oxford: Basil Blackwell, 1990. Russell, B. ‘The Limits of Empiricism’, Proceedings of the Aristotelian Society, 36 (1935–6). —— An Enquiry into Meaning and Truth. Norwich: Fletcher and Son, 1973. —— , The Collected Papers of Bertrand Russell. London: Allen and Unwin, 1986. —— , Collected Papers. London: Routledge, 1997. —— and Whitehead, A. N., Principia Mathematica. Cambridge: Cambridge University Press, 1950. Russell, B. see Bibliography: Abbreviations. Ryle, G., The Concept of Mind. London: Hutchinson, 1949. —— . Dilemmas. Cambridge: Cambridge University Press, 1954. —— . Collected Papers, vol. 2. New York: Barnes and Noble, 1971. ¨ tze. Vienna: Gerholt, 1936. Schlick, M., Gesammelte Aufsa Stern, D. G., Wittgenstein on Mind and Language. Oxford: Oxford University Press, 1995. Strawson, P. F., ‘Critical Notice of Wittgenstein’s Philosophical Investigations, Mind, 63, 1954. von Wright, G. H., Wittgenstein. See under Wittgenstein: Nachlass. —— . Norm and Action. London: Routledge and Kegan Paul, 1963. Waismann, F., Einfu ¨ hrung in das Mathematische Denken. Die Begriffsbildurg in der Modernen Mathematik, Vienna: Gerold, 1936. English translation: Introduction to Mathematical Thinking. New York: Harper and Brothers, 1959. Williams, B., Descartes: A Project of Pure Inquiry. Harmondsworth: Penguin, 1978.

Wittgenstein’s Method: Neglected Aspects Edited by Katherine J. Morris Copyright © 2004 by Blackwell Publishing Ltd

Index

A list of philosophical ‘problems’ whose ‘treatments’ are discussed in some detail will be found at the end of this index. acknowledgement of motives, 33, 49n.19, 68, 84, 132, 139n.1, 145, 148–9, 154, 156, 157–60, 161, 164–5, 166, 172n.6, 173n.15, 175n.29, 175n.39, 178n.64, 181, 183, 184–5, 187, 190, 206, 207, 209–10, 211, 213, 218, 220n.14, 240, 265, 286, 290; see also motive; psychoanalysis; unconscious of our own rules, 174n.26, 177n.55, 194, 196, 203n.37, 203n.43, 212, 249; see also stipulation of possibilities, 101–2, 147, 151, 174n.25, 175n.35, 178n.63, 188, 189, 191–2, 217, 251, 280, 284; see also possibility adversary/adversarial, 85, 111, 113, 127, 152, 161, 163, 182, 213, 217, 233, 268, 269; see also argument analogy (comparison, etc.) 48n.12, 177n.50, 264; see also aspect; object of comparison; picture between Wittgenstein’s method and psychoanalysis see psychoanalysis as source of philosophical problems, 30, 31, 34, 47n.1, 91n.51, 120,

132, 136, 138, 139n.1, 149, 154, 157–8, 160, 162, 164–5, 170, 175n.32, 176n.49, 178n.64, 184–6, 198, 199, 200, 202n.22, 202n.27, 203n.29, 207, 208, 209–11, 214, 217, 220n.15, 267, 278n.3, 283, 290 as therapy, 30, 34, 36, 41, 48n.15, 49n.17, 49n.20, 50n.25, 67, 69, 79, 82, 91n.48, 103, 123, 149, 153, 158–9, 165–6, 187–90, 214, 215, 251, 270, 283, 292, 293n.10 Anscombe, G.E.M., 105n.19 anxiety (disquiet), 136, 146, 151, 152, 153, 158, 159, 162, 164, 165–6, 173n.10, 173n.12, 174n.27, 180, 182–4, 187, 188, 197–8, 200, 202n.16, 204n.46, 204n.49, 208, 210, 211, 212–13, 214, 216, 217, 218–19, 221n.21, 267; see also craving; psychoanalysis; therapy argument (on one conception: demonstration, refutation), 39, 44, 46, 68, 82, 87n.16, 94, 95, 149–50, 153, 161, 163, 169, 172n.8, 173n.13, 185–6, 188, 192, 196, 202n.25, 203n.30, 208,

306 Index argument (cont’d) 213, 217, 218, 219, 246, 253, 257n.49, 258n.54, 262, 266, 267, 269, 272–5, 282–3, 285–7, 290–2, 292n.5, 293n.13; see also adversary; ‘our method’; persuasion picture of Private Language Argument as a reductio: Essay 5 passim; Essay 6 passim, 138, 139n.6, 201n.9 aspect (way of seeing things/of looking at things); see also analogy; blindness; conception; pattern the grammar of ‘aspect’, 154, 156, 157, 174n.24, 266, 269, Chapter 13 passim perspicuous representations as highlighting aspects, 30–2, 40–6, 48n.15, 85, 251; see also ¨ bersicht U revelation of neglected aspect as therapy, 33–5, 40–1, 69, 76, 82, 84, 85, 101, 116, 123, 128, 137, 139, 147, 149, 157, 159–60, 161–2, 170, 171n.4, 174n.25, 182, 184, 187–8, 190, 192, 196, 208, 227, 242, 245, 253, 268; (see also conversion; ‘our method’; therapy) Augustinian picture (Augustinian conception of the essence of language), 27, 38–9, 42, 54, 78, 88n.28, 97, 126, 131, 132, 133, 156, 166, 181, 186, 188, 189, 251, 261–2, 265, 267, 269, 271, 272–7, 283, 285; see also conception; essence; picture Austin, J.L., 87n.21, 92, 179 Baker, G.P., 70n.1, 70n.5, 255n.24, 259n.70 Baker, G.P. and Hacker, P.M.S., 47n.4, 47n.5, 47n.6, 47n.8, 48n.11, 49n.20, 50n.26, 51n.28, 70n.3,

71n.11, 72n.19, 95, 104n.10, 104n.12, 105n.21, 203n.38, 251, 258n.57, 259n.68, 278n.10 blindness (to aspects, conceptions, possibilities), 33, 34–5, 130, 134, 136, 151, 157, 159, 168, 170, 182, 185, 187, 189, 190, 191, 196, 200, 238, 268, 279, 281, 282, 285, 287, 291 bondage see tyranny Bouwsma, O.K., 89n.31, 201n.3, 201n.10, 202n.20, 202n.22, 202n.25, 202n.27, 204n.46, 204n.52, 204n.58, 219n.4 captivity see tyranny Carnap, R., 52, 74, 76, 114, 138, 179, 217, 220n.11, 220n.13, 220n.16, 222n.45, 260 categories, category–mistake see Ryle Cavell, S., 104n.2 censorship see psychoanalysis Chomsky, N., 52, 88n.24 combinatorial possibilities (or rules) see Ryle comparison see analogy compulsion see craving Conant, J., 104n.2 conception (Auffassung; way of looking at things, way of thinking, etc.); see also aspect change of conception or new conception as therapy, 46, 60, 61, 80, 86, 100, 147, 152–3, 154, 157–9, 166, 168–9, 170–1, 177n.57, 191, 192, 194, 203n.40, 208, 210–13, 216–19, 243, 245, 249, 255n.19, 257n.40, 259n.70, 259n.72; see also conversion; ‘our method’; therapy the grammar of ‘conception’, 150, 156, 193, 195, Essay 12 passim, Essay 13 passim conflict, 34, 151, 152, 159, 168, 171, 172n.5, 173n.15, 173n.16, 183,

Index 193, 204n.46, 208, 212, 213, 251; see also psychoanalysis consciousness see unconscious convention see stipulation conversion, 35, 46, 147, 158, 163, 187, 269, 286, 287 craving (drive, obsession, temptation, etc.), 67, 80, 83, 85, 98, 103, 134, 135, 136, 137, 146, 151–2, 153, 157, 160, 162, 164, 175n.34, 182–3, 184–5, 190, 198, 200, 201n.13, 202n.20, 202n.22, 204n.46, 208–9, 211, 214, 215, 218–19, 220n.14, 244, 251, 267, 275, 278n.11, 287, 292n.5; see also anxiety; motive; prejudice; psychoanalysis decision see stipulation deduction see argument demonstration see argument dialogue (discussion, negotiation), 67, 85, 110, 111, 122, 127, 152–3, 155, 160, 163–4, 166, 170, 174n.26, 181, 182, 188, 192, 194, 196, 197, 198, 200, 201n.11, 205, 211, 213, 217, 219, 233, 238–9, 269, 277, 278n.9, 282, 287 Diamond, C., 104n.2, 172n.8, 212, 222n.37 dimension, 32, 78, 79, 84, 85, 86, 91n.49, 277, 281, 287 discussion see dialogue disorder (i.e. disease) see therapy disorder (i.e. lack of order) see pattern disquiet see anxiety disturbance see anxiety dogma (ide´e fixe); see also grammatical illusion; metaphysics; prejudice; superstition examples, 39, 46, 71n.10, 82, 85, 127, 175n.35, 186 as targets of Wittgensteinian therapy, 34, 85, 124, 153, 157, 171,

307

184–7, 190, 191, 244, 251, 267 Wittgenstein’s ‘non-dogmatic procedure’, 45, 53, 63, 67, 103, 112, 116, 120, 121, 134, 137, 157, 162, 174n.27, 238, 246, 252, 270, 288, 293n.10 Dreben, B., 104n.2 drive see craving Drury, M.O’C., 174n.27, 201n.6, 204n.51, 204n.53, 204n.54 Dummett, M.A.E., 71n.12, 105n.15 enslavement see tyranny essence (nature) 39, 70n.4, 90n.40, 91n.47, 97–8, 99–100, 102–3, 105n.23, 106n.26, 107n.32, 120, 135, 162, 182, 218, 253n.1, 258n.52, 258n.56, 261–2, 268, 270, 271–3, 278n.12, 290; see also Augustine’s picture; metaphysics; modal qualification ‘the nature of philosophy’ see philosophy ‘our conception’ of essence, 73, 181, 191–2, 221n.20, 246–7, 254n.4, 257n.42, 258n.59, 258n.59, 275–6 essential contestability, 29, 85, 289, 290 freedom (liberation, liberty), 33, 39, 41, 82, 85, 107n.33, 134, 136, 137, 138, 139n.2, 173n.20, 251, 258n.59, 264, 267, 269, 273, 284, 287; see also tyranny; voluntariness; will as ‘essence’ of philosophy on Wittgenstein’s conception, 148, 149, 150, 154, 158, 159, 162, 165, 166, 171, 179–80, 181, 185, 188, 190–2, 194–6, 197, 198, 200, 203n.32, 204n.46, 211, 212, 213, 217, 218; 277; see also ‘our method’

308 Index Frege, G., 39, 61, 68, 72n.14, 72n.15, 77, 79, 84, 86n.5, 87n.20, 88n.29, 89n.33, 146, 150, 172n.6, 182, 184, 195, 212, 213, 215, 217, 222n.42, 251, 254n.7, 255n.20, 256n.29, 257n.45, 264, 271, 278n.8, 292n.6 Freud, S., 140n.7, 152, 154, 155, 160, 161, 173n.19, 173n.20, 174n.22, 174n.23, 174n.27, 175n.31, 175n.32, 175n.33, 175n.39, 176n.40, 176n.44, 177n.61, 178n.62, 178n.65, 201n.13, 202n.16, 202n.23, 202n.24, 204n.57, 206, 209, 218, 222n.41, 257n.50, 268, 273, 287; see also psychoanalysis Glock, H.-J., 104n.8 Goethe, J., 28, 36, 49n.15, 50n.25, 249, 290 grammar; see also language ‘die Grammatik’, 55–9, 232, 234, 255n.21 ‘unsere Grammatik’, 55–9 descriptions of grammar, 42, 48n.12, 58, 59, 69, 71n.13, 115, 122–4, 128, 131, 137, 147–8, 155, 156–7, 159, 162, 163–4, 167–8, 171, 171n.4, 174n.26, 180, 193–6, 203n.39, 217, 251–2, 262, 263, 265, 266–7, 268, 270, 273, 274, 275, 276, 277, 280, 290–1 grammatical illusion (grammatical fiction), 27, 34, 38, 40, 41, 88n.23, 104n.10, 106n.26, 118, 128, 129, 132, 133, 134, 136, 137, 138, 157, 158, 177n.56, 222n.40, 225, 232, 234, 252 representations (pictures) of grammar, Essay 1 passim, 58, 115, 117, 121, 122, 123, 124, 128,

136, 156, 168, 169, 172n.6, 186–7, 195, 244, 267, 271, 275 surface vs. depth grammar, 56, Essay 3 passim, 277 Hacker, P.M.S., 47n.7, 47n.8, 47n.9, 47n.10, 86n.2, 86n.3, 105n.17, 139n.6, 219n.1, 246, 254n.4, 258n.53, 258n.58; see also Baker and Hacker Hallett, G., 254n.4 Hanfling, O., 105n.14 Hare, R.M., 72n.21 Heidegger, M., 207–10, 218, 219, 220n.13, 220n.16, 222n.38, 222n.43 Hertz, H., 50n.23, 50n.25, 176n.47, 198, 204n.48, 215, 221n.31, 222n.35, 241 Hilbert, D., 147, 150, 182, 195 homoeopathy, 34, 175n.36, 187, 268, 281, 284 ide´e fixe see dogma irritation see anxiety James, W., 68, 110, 114 Kenny, A., 47n.5, 47n.7, 49n.19, 50n.24, 72n.17, 72n.21, 87n.12, 87n.13, 105n.17, 105n.18, 105n.19, 139n.2, 139n.3, 222n.39 Ko¨hler, W., 68, 255n.26 language (speech); see also grammar ‘die Sprache’, 53–5, 59–61 ‘in der Sprache’, 63–6 ‘unsere Sprache’, 60 language-game, 31, 32, 33, 35, 36, 41, 42, 47n.9, 48n.12, 50n.22, 51n.27, 54, 61, 67, 69, 70n.3, 71n.12, 72n.16, 72n.22, 79, 80–2, 90n.40, 90n.44, 94, 95, 102, 121, 123, 133, 134, 156,

Index 158, 162, 187, 189, 192, 194, 197, 240, 242, 251, 256n.30, 262, 263, 265, 268, 275 liberation, liberty see freedom logical geography see Ryle Malcolm, N., 87n.12, 87n.14, 105n.19, 105n.20, 253n.1, 258n.60 ‘mere’, 39, 57, 71n.10, 79, 83, 84, 86, 136, 172n.5, 186, 188, 211, 216, 220n.17, 244, 245, 247, 258n.58, 262, 268, 286, 292n.5 Merleau-Ponty, M., 204n.56, 222n.43, 277 metalogical concepts, 134–5, 171, 193, 251 metaphor see analogy metaphysics (metaphysical use of words), 27, 34, 40, 43, 62, 66, 71n.10, 72n.23, 75, 83, 85, Essay 4 passim, 118, 120, 122, 124, 131, 135, 157, 158, 162–3, 171, 191, 201n.7, 214, 221n.28, 222n.38, 244, 251, 258n.54, 261, 270; see also dogma; essence; modal qualification method demonstrating method by means of examples, 33, 38, 66, 95, 102, 117, 122, 129, 138, 155, 163, 189, 218, 253 ‘our method’ (Wittgenstein’s conception of philosophy), 30, 33–7, 38, 40–1, 44–6, 85–6, 91n.45, 95–6, 103–4, 119, 120, 122, 124–5, 131, Essay 8 passim, Essay 9 passim, 207, 213, 219, 238, 241, 250–2, 253, 256n.34, 257n.35, 258n.59, 269, 277, 279–80, 287, 290–2; see also argument; ‘the nature of philosophy’ modal qualifications (e.g. ‘cannot’, ‘might’, ‘must’), 34, 40, 63,

309

70n.2, 71n.10, 75, 77, 82, 83, 85, 88n.29, 94, 97–8, 99, 101–2, 103, 104n.12, 107n.36, 124, 135, 136, 147, 151, 152, 153, 157, 163, 164–5, 175n.35, 175n.37, 180, 181, 184, 186, 188, 191, 199, 202n.20, 210, 212, 214, 243, 244, 248, 249, 251, 252, 256n.29, 261, 265, 266, 273, 282, 285, 286, 287; see also essence; metaphysics; possibility morality, 198–200, 222n.45; see also freedom Morris, K.J., 257n.49 motive, 85, 91n.52, 103, 154, 163, 184, 185, 198, 199, 202n.16, 208, 209, 210, 213, 214, 215, 217, 220n.14, 242, 244, 277, 287, 290; see also psychoanalysis nature see essence negotiation see dialogue Nietzsche, F., 222n.43, 277, 292n.4 object of comparison, 31–6, 42, 48n.15, 69, 81, 82, 97, 116, 123, 128, 130, 133, 137, 146, 149, 150, 156, 162, 171n.3, 188, 189, 190, 240, 251, 283, 286, 288, 291; see also analogy obsession see craving order see pattern pattern (order, etc.), 58, 74, 84, 90n.39, 123, 131, 147, 158, 161, 169, 170, 174n.25, 187, 189, 190, 193, 253, 259n.72, 264, 268, 280, 282, 284, 286, 291–2; see also aspect disorder, impression of, 166, 183, 212, 216 person-relativity (of Wittgenstein’s method), 68, 85, 131–2, 138, 146, 147, 151, 152, 158–9, 162–5, 171, 173n.13, 175n.5,

310 Index person-relativity (cont’d) 181–4, 187, 192–3, 196, 197, 198, 200, 203n.33, 203n.43, 204n.46, 208, 209–15, 217, 218, 220n.14, 221n.21, 252; see also acknowledgement; freedom; therapy ¨ bersicht perspicuous representation see U persuasion (propaganda), 44, 52, 68, 72n.23, 102–3, 145, 147, 149, 151, 163, 165, 168, 170, 188, 219, 243, 272, 282, 286, 287, 289, 290, 292; see also argument; conversion; ‘our method’ philosophy; see also argument; ‘our method’; persuasion ‘the nature of philosophy’, 26, 38, 72n.23, 85, 93, 121, 122, 146, 150, 207, 241, 277 Wittgenstein and ordinary-language philosophy, 74–5, 87n.16, 92–5, 103, 105n.14, 105n.16, 260 Wittgenstein’s conception of philosophy see ‘our method’ picture, 24, 33, 34, 38, 39, 41, 47n.1, 55, 83–4, 85, 88n.26, 103, 115, 118, 125, 132, 135, 136–8, 148, 149, 154, 157–8, 160, 162, 164, 171, 175n.32, 177n.50, 181, 184–90, 192–3, 199, 200, 202n.19, 202n.21, 208, 209, 210–11, 213, 214, 217, 218, 220n.15, 237, 243, 249, 251, Essay 12 passim, 286–7, 290, 291; see also analogy; aspect; Augustine’s picture; conception possibility, 30, 33–5, 41, 45, 46, 48n.15, 68, 70n.4, 90n.39, 101–2, 103, 107n.35, 113, 116, 121, 123, 124, 125, 126, 128, 134, 137, 149, 151, 154, 162–3, 175n.35, 188–91, 196, 197, 203n.36, 210, 213, 217, 219,

232, 244, 247, 249, 251, 253, 259n.72, 266, 267, 268, 280, 282, 283–4, 286, 288–90; see also conception; modal qualification pluralism, 180, 210, 268, 280, 281, 284 preconception see prejudice prejudice (preconception); see also dogma; grammatical illusion; superstition examples, 60, 112, 116, 126, 128, 133, 150, 166–71, 177n.56, 188, 195, 204n.55, 290 as targets of Wittgensteinian therapy, 38, 43, 85, 94, 103, 118, 122, 132, 133, 135, 137, 139, 145, 146–7, 148, 151, 152–4, 157–8, 160, 164, 166, 169–70, 171, 174n.27, 183, 184–5, 191, 199, 200, 204n.56, 217–19, 244, 251, 264, 267, 277, 287, 290, 292 proof see argument propaganda see persuasion psychoanalysis (censorship, repression, resistance, sublimation), 49, 68, 132, 136, 140, 140n.7, Essay 8 passim, 179, 180, 181, 182, 185, 187, 188, 190, 192, 199, 200, 201n.12, 202n.22, 202n.24, Essay 10 passim, 265, 277, 278n.6; see also acknowledgement; anxiety; craving; motive; obsession; unconscious psychotherapy see psychoanalysis QED. see argument quietism, 163 reductio see argument refusal see voluntariness refutation see argument relativism, 40, 43, 51, 269, 289 repression see psychoanalysis

Index resistance see psychoanalysis Rundle, B., 105n.15 Russell, B., 39, 68, 71n.9, 77, 79, 84, 87n.12, 107n.31, 110, 114, 117, 138, 156, 173n.12, 177n.53, 179, 182, 184, 202n.17, 211, 220n.14, 255n.19, 257n.45 Ryle, G. (categories, categorymistakes, combinatorial possibilities (or rules), logical geography), 23, 24, 25, 26, 39, 52, 68, 70n.4, 70n.8, 74, 76–8, 81, 86n.3, 86n.4, 87n.8, 87n.9, 87n.10, 87n.11, 91n.51, 93, 94, 95, 105n.13, 109–12, 115–17, 121, 123, 125, 130, 131, 134, 135, 139n.4, 145, 152, 165, 171, 176n.45, 177n.52, 179, 181, 192, 193, 197, 199, 217, 221n.21, 221n.26, 252, 260, 272, 276 Satzklang, 75–7, 80 Schlick, M., 110, 114, 138, 145, 150, 179, 180, 239, 257n.48 scholasticism, 194, 202n.26, 231, 285, 286 simile see analogy speech see language Stern, D., 87n.14 stipulation (convention, decision), 39, 96, 99, 139n.2, 149, 150, 159, 164, 166, 179, 181, 190, 191, 192, 194, 196, 197, 203n.43, 221n.18, 266, 273, 288, 289, 290 Strawson, P.F., 52 sublimation see psychoanalysis suffering see anxiety superstition, 39, 132, 151, 161, 185, 218, 247, 262, 264, 267, 277, 287; see also dogma; prejudice ¨ bersicht surveyability see U ¨ bersicht synoptic view see U

311

temptation see craving therapy, 26, 31, 33–4, 36, 37, 38, 49n.19, 58, 67–8, 93–4, 104n.10, 111, 114, 117, 118, 130, 131–2, 136, 137–8, 140n.7, Essay 8 passim, 179, 181–4, 187–9, 190, 191–2, 199–200, 201n.6, 201n.9, 202n.16, 202n.23, 204n.46, 204n.49, Essay 10 passim, 268, 277; see also Freud; ‘our method’; psychoanalysis thrall see tyranny torment see anxiety troubled state of mind see anxiety tyranny (captivity), 30, 31, 33, 34, 36, 83, 84, 85, 157, 158, 162, 170, 185, 188, 190, 201n.13, 204n.46, 211, 218, 240, 252, 264, 268, 286, 287 ¨ bersicht (perspicuous representation, U vision), Essay 1 passim; 53, 56–9, 69, 72n.23, 74, 85, 91n.48, 122, 128, 150, 158, 167, 169, 191, 192, 193, 197, 200, 216, 217, 232, 235, 236, 245, 255n.20, 255n.25, 258n.59, 260, 269, 271, 277, 280, 290–2; see also conception; pattern unacknowledged see unconscious unconscious (bring to consciousness, unacknowledged), 34, 38, 49n.19, 83, 85, 91n.52, 136, 138, 147–9, 153, 154, 157, 158, 160, 162, 164, 171, 173n.19, 174n.21, 174n.27, 175n.33, 184, 185, 186, 187, 188, 189, 191, 192, 196, 199, 200, 202n.22, 202n.23, 206–10, 214, 217, 218, 243, 251, 265, 267, 286–7, 290; see also psychoanalysis unease see anxiety unrest see anxiety urge see craving

312 Index ¨ bersicht vision see U voluntariness (refusal), 46, 59, 93, 113, 124, 149, 150, 154, 159, 164, 165, 168, 170, 176n.44, 177n.50, 188, 190–2, 196, 200, 251, 253, 267–70, 276, 281, 282, 284, 288, 291, 293n.11; see also freedom; morality; will how one wants to see things, what one wants to say 46, 68, 82, 91n.52, 152–3, 163, 173n.16, 183, 191, 198, 199, 203n.32, 204n.46, 206, 272, 276, 289; see also prejudice Waismann, F., 47n.11, 49n.15, 49n.16, 49n.17, 50n.21, 50n.25, 51n.30, 51n.31, 86n.5, 88n.22, 90n.38, 90n.41, 104n.4, 106n.25, Essay 8 passim, Essay 9 passim, 206, 217, 219n.6, 221n.19, 221n.23, 221n.24, 221n.33, 222n.44, 239, 253n.3, 256n.32, 258n.54, 258n.59, 272, 283, 288, 293n.7 way of seeing things/of looking at things see aspect

way of thinking see conception Weltanschaaung, 22, 51, 147, 255 Whitehead, A., 76n.9 will (‘rather than intellect’), 50n.25, 136, 163, 190, 196, 269, 275; see also freedom; morality; voluntariness worry see anxiety Some philosophical ‘problems’ and their ‘treatments’ ‘Does the nothing noth?’, 207–11, 214–16 ‘The problem of colour-exclusion’, 165–6 ‘The problem of identity in difference’, 30–1, 211–12 ‘The problem of indescribability’, 60–2 ‘The problem of intentionality’, 62–5 ‘The problem of measuring time’, 164–5 ‘What lies behind the expression of a thought?’, 166–71 ‘The problem of the connection between language and reality’, 287–90 ‘The problem of the triviality of identity’, 212