History of Analytic Philosophy Series Editor: Michael Beaney Titles include: Stewart Candlish THE RUSSELL/BRADLEY DISPU...
63 downloads
1124 Views
1MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
History of Analytic Philosophy Series Editor: Michael Beaney Titles include: Stewart Candlish THE RUSSELL/BRADLEY DISPUTE AND ITS SIGNIFICANCE FOR TWENTIETH-CENTURY PHILOSOPHY Annalisa Coliva MOORE AND WITTGENSTEIN Scepticism, Certainty and Common Sense Omar W. Nasim BERTRAND RUSSELL AND THE EDWARDIAN PHILOSOPHERS Constructing the World Nuno Venturinha (editor) WITTGENSTEIN AFTER HIS NACHLASS Pierre Wagner (editor) CARNAP’S LOGICAL SYNTAX OF LANGUAGE Forthcoming: Andrew Arana and Carlos Alvarez (editors) ANALYTIC PHILOSOPHY AND THE FOUNDATIONS OF MATHEMATICS Rosalind Carey RUSSELL ON MEANING The Emergence of Scientific Philosophy from the 1920s to the 1940s Giusseppina D’Oro REASONS AND CAUSES Causalism and Non-Causalism in the Philosophy of Action Sébastien Gandon RUSSELL’S UNKNOWN LOGICISM A Study in the History and Philosophy of Mathematics Anssi Korhonen LOGIC AS UNIVERSAL SCIENCE Russell’s Early Logicism and its Philosophical Context Sandra Lapointe BERNARD BOLZANO’S THEORETICAL PHILOSOPHY An Introduction Douglas Patterson ALFRED TARSKI Philosophy of Language and Logic
Erich Reck (editor) THE HISTORIC TURN IN ANALYTIC PHILOSOPHY Graham Stevens THE THEORY OF DESCRIPTIONS
Moore and Wittgenstein Scepticism, Certainty and Common Sense Annalisa Coliva University of Modena and Reggio Emilia, Italy
© Annalisa Coliva 2010 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6–10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The author has asserted her right to be identified as the author of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2010 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN 978–0–230–58063–3
hardback
This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Coliva, Annalisa, 1973– Moore and Wittgenstein: scepticism, certainty and common sense/Annalisa Coliva. p. cm. — (History of analytic philosophy) Includes bibliographical references and index. ISBN 978–0–230–58063–3 1. Moore, G. E. (George Edward), 1873–1958. 2. Wittgenstein, Ludwig, 1889–1951. 3. Scepticism. 4. Certainty. 5. Common sense. I. Title. B1647.M74C65 2010 192—dc22 2010023767 10 9 8 7 6 5 4 3 2 1 19 18 17 16 15 14 13 12 11 10 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne
Contents Preface to the English Edition
vii
Series Editor’s Foreword
viii
Acknowledgements
xi
Abbreviations of cited works
xiii
Introduction 1
2
3
1
G. E. Moore: Scepticism, Certainty and Common sense
13
1 ‘A defence of common sense’
14
2 ‘Proof of an external world’
25
3 Malcolm: ‘common sense’ and ‘ordinary language’
28
4 Clarke and Stroud: At the origins of contextualism
37
5 Moore and Humean scepticism: Wright’s interpretation of the proof
42
6 Moore’s comeback: Pryor’s dogmatist interpretation
44
7 Having knowledge and being able to prove that one does
47
Wittgenstein: Belief, Knowledge and Certainty
55
1 The philosophical use of ‘to know’. On the misleading assimilation of ‘to know’ and ‘to believe’
57
2 The language game with ‘to know’ and ‘I know’. A perspicuous description
60
3 The grammatical use of ‘I know’
74
4 Wittgenstein and the ‘assertion fallacy’
90
5 Coda. Wittgenstein and semantic contextualism
100
Wittgenstein: Doubts and the Nonsense of Scepticism
103
1 The language game with ‘to doubt’: A perspicuous description
104
2 Some philosophical consequences: Why idealism and scepticism in general are nonsensical
111
3 Two classical sceptical arguments and Wittgenstein’s replies
118
v
vi Contents
4
Wittgenstein: Hinges, Certainty, World-Picture and Mythology
149
1 Only apparently empirical propositions: Examples and preliminary considerations
152
2 Propositions that characterize a method. The hinges around which all other propositions rotate
161
3 On the groundlessness of the foundations
166
4 World-picture
179
5 Different world-pictures and epistemic relativism?
188
6 Propositions that might be part of a kind of mythology
203
Conclusion: Moore and Wittgenstein on Epistemology and Language. A Synopsis
208
1 Truisms, hinges, common sense, world-picture and knowledge
208
2 Meaning, use and philosophical contexts
209
3 Scepticism
209
4 Certainty
210
5 Epistemic foundationalism and epistemic relativism
210
Notes
211
Bibliography
234
Index
241
Preface to the English Edition This book is a revised and much enlarged version of my Italian volume ‘Moore e Wittgenstein: scetticismo, certezza e senso comune’, published in 2003 (Padova, Il Poligrafo). In 2004 and 2005 several important works on On Certainty, but also on Moore’s epistemology and, in particular, ‘Proof of an external world’, appeared. That by itself was reason enough to recommend a revision and extension of the original book. But, meanwhile, I also came to a rather different view about the anti-sceptical potential of Wittgenstein’s notes in On Certainty as well as to a much deeper understanding of their bearing on the problem of epistemic relativism. The present book bears testimony to such changes and developments. So, in more detail, the new parts are in Chapter 1, sections 6 and 7; in Chapter 2, sections 3.4 and 5; in Chapter 3, sections 2, 2.1, 2.2, 3.2, 3.2.1, 3.2.2 and 3.3; and, finally, in Chapter 4, sections 3.2, 5, 5.1.1, 5.1.2, 5.1.3, 5.2, 5.2.1, 5.2.2 as well as the Introduction and the Conclusion. Some themes, already present in the Italian edition, have received a slightly different treatment throughout the text. In particular, I have insisted more than before on what I take to be Wittgenstein’s view(s) on the notion of nonsense, also in consequence of recent ‘therapeutic’ interpretations of his thought in general, and of On Certainty in particular. I have paid more attention to the issue of Wittgenstein’s conception of propositions and rules, as well as to possible connections with Kant’s views on synthetic a priori judgements. Finally, I have also enlarged on why I personally think Wittgenstein didn’t put forward a new form of epistemic foundationalism.
Credits In Chapter 4, the entire section 5 is taken from my paper ‘Was Wittgenstein an epistemic relativist?’, Philosophical Investigations 33/1, 2010.
vii
Series Editor’s Foreword During the first half of the twentieth century, analytic philosophy gradually established itself as the dominant tradition in the Englishspeaking world, and over the last few decades it has taken firm root in many other parts of the world. There has been increasing debate over just what ‘analytic philosophy’ means, as the movement has ramified into the complex tradition that we know today, but the influence of the concerns, ideas, and methods of early analytic philosophy on contemporary thought is indisputable. All this has led to greater self-consciousness among analytic philosophers about the nature and origins of their tradition, and scholarly interest in its historical development and philosophical foundations has blossomed in recent years, with the result that history of analytic philosophy is now recognized as a major field of philosophy in its own right. The main aim of the series in which the present book appears, the first series of its kind, is to create a venue for work on the history of analytic philosophy, consolidating the area as a major field of philosophy and promoting further research and debate. The ‘history of analytic philosophy’ is understood broadly, as covering the period from the last three decades of the nineteenth century to the start of the twentyfirst century, beginning with the work of Frege, Russell, Moore, and Wittgenstein, who are generally regarded as its main founders, and the influences upon them, and going right up to the most recent developments. In allowing the ‘history’ to extend to the present, the aim is to encourage engagement with contemporary debates in philosophy, for example, in showing how the concerns of early analytic philosophy relate to current concerns. In focusing on analytic philosophy, the aim is not to exclude comparisons with other – earlier or contemporary – traditions, or consideration of figures or themes that some might regard as marginal to the analytic tradition but which also throw light on analytic philosophy. Indeed, a further aim of the series is to deepen our understanding of the broader context in which analytic philosophy developed, by looking, for example, at the roots of analytic philosophy in neo-Kantianism or British idealism, or the connections between analytic philosophy and phenomenology, or discussing the work of philosophers who were important in the development of analytic philosophy but who are now often forgotten. viii
Series Editor’s Foreword ix
The present book, by Annalisa Coliva, offers an account of remarks that Ludwig Wittgenstein (1889−1951) wrote in the last year and a half of his life, remarks that were edited, translated, and published posthumously as On Certainty in 1969. Wittgenstein is arguably the most influential philosopher of the twentieth century, and is undoubtedly a central figure in the analytic tradition. His two major works are the Tractatus Logico-Philosophicus, which appeared in German in 1921 and in its first English translation in 1922, and the Philosophical Investigations, which was also edited, translated and published posthumously, appearing shortly after Wittgenstein’s death, in 1953. The Tractatus is one of the canonical texts of analytic philosophy, while the Philosophical Investigations subjected some of its key ideas to criticism. In both works, Wittgenstein also criticizes the views of other philosophers, most notably, Frege and Russell. On Certainty contains further criticisms of all these earlier views, but its target, in particular, are ideas Wittgenstein found in two of Moore’s most famous papers, ‘A defence of common sense’, published in 1925, and ‘Proof of an external world’, published in 1939. It is Wittgenstein’s response to these two papers that Coliva sets out to elucidate and assess. On Certainty has been influential in analytic epistemology ever since it first appeared in 1969, but its interpretation has been enormously controversial. Part of the reason for this is that Wittgenstein did not live long enough to revise and polish his remarks to the extent that he had done in the case of the Philosophical Investigations. Its interpretation also requires understanding of Moore’s own papers, which have generated their own share of dispute. The problem is exacerbated by the fact that Wittgenstein’s (re)reading of Moore’s papers in the last two years of his life was inspired and mediated by discussions with Norman Malcolm, who has made his own contribution to analytic epistemology. There is a complex history here, from Moore’s original papers to the most recent interpretations of Moore’s and Wittgenstein’s views, and it is one to which Coliva is sensitive in the account she offers. Moore had been concerned in his papers to show that scepticism about the external world is misguided. While sympathetic to Moore’s goal, Wittgenstein was critical of his approach and the specific claims he made. In his engagement with Moore’s views, Wittgenstein raised questions about the notions of knowledge, belief, and certainty involved both in philosophical disputes and in our ordinary discourse, and about the status of key propositions that he called ‘hinge’ propositions. Coliva brings out the similarities and differences between Moore’s and Wittgenstein’s approaches and conceptions, locating her account in
x
Series Editor’s Foreword
current debates about scepticism and knowledge. Over recent years, there has been a blossoming of interest in Moore’s attempt to refute scepticism and in Wittgenstein’s On Certainty. Coliva’s book provides a fine example of how work in history of analytic philosophy not only sheds light on the historical development of themes in analytic philosophy but also, in doing so, advances the philosophical debates themselves. Michael Beaney April 2010
Acknowledgements This work has been with me, in one form or another, for about fifteen years. Hence, acknowledgements will be rather long. First of all, I would like to thank the late Roberto Dionigi, who taught theoretical philosophy at the University of Bologna, for first getting me interested in Wittgenstein’s philosophy and for teaching me how to work with and on a philosophical text. Second, I would like to thank Eva Picardi, who was my supervisor for my tesi di laura on On Certainty, for unfailing support throughout the years. Third, I would like to thank Crispin Wright, who partly supervised that work, when I was an undergraduate student visiting St Andrews with an Erasmus exchange programme; and who, later on, made me appreciate how Wittgenstein’s ideas in On Certainty could become relevant also to contemporary epistemology. Fourth, I would like to thank Danièle Moyal-Sharrock, for her lovely friendship and support. If this book is coming out in English, it is essentially due to her encouragement. Also Duncan Pritchard has been a continuous support while writing this book. Both his and Danièle’s comments to earlier versions of this work have proved fundamental in improving it towards its final version. I should also like to thank Mike Beaney, not just for having the book in his Palgrave series in the history of analytic philosophy, but also for very stimulating discussions on what it means to do history of analytic philosophy, originally occasioned by our co-participation as invited speakers in the Second SIFA Graduate Conference, held in Bologna (October 29–31 2009). Moreover, I would like to thank all my colleagues at COGITO Research Centre in Philosophy; so, once again, Eva Picardi, Paolo Leonardi, Sebastiano Moruzzi, Elisabetta Lalumera, Giorgio Volpe, Walter Cavini, as well as the PhD and Master students Delia Belleri, Alessia Pasquali, Alessia Marabini, Sara Neva, Fabio Minocchio, Andrea Marino, Iryna Sivertsava, Filippo Ferrari, Eugenio Orlandelli, Michele Palmira, Beatrice Collina and Giovanni Mascaretti for their attendance to a series of seminars on Moore and Wittgenstein held in Fall 2009, which helped me a lot to fix several details of the manuscript. But, most of all, I would like to thank them for their support and enthusiasm for our Research Centre. Finally, I would like to thank my husband, Marco Panza, and my son Leonardo. I hope Marco won’t get offended if this book is dedicated only to our xi
xii Acknowledgements
child. I promise that the next one on the shelf will be his, for as our song says, ‘Myself and I seem to agree that the best thing for you would be me.’1 This one, however, is for our frugolo, for the two of them have slowly come to light together.
Abbreviations of cited works Moore C
‘Certainty’, 1959, reprinted in PP, pp. 227–251.
DCS
‘A defence of common sense’, 1925, reprinted in PP, pp. 32–59.
FFS
‘Four forms of scepticism’, 1959, reprinted in PP, pp. 196–226.
LM
‘Letter to Malcolm’, 1949, reprinted in SW, pp. 213–216.
MP
‘Moore’s paradox’, 1944 (?), reprinted in SW, pp. 207–212.
PEW
‘Proof of an external world’, 1939, reprinted in PP, pp. 127–150.
PP
Philosophical Papers, London, George Allen & Unwin, 1959.
RMC
‘A reply to my critics’, in P. A. Schillp (ed.) 1942 The Philosophy of G. E. Moore, Evanston and Chicago, Northwestern University.
SW
Selected Writings, London-New York, Routledge, 1993.
TF
‘Truth and falsity’, 1901–2, reprinted in SW, pp. 20–22. Wittgenstein
CV
Culture and Value, Oxford, Blackwell, 1980.
NF
Notes on Frazer’s ‘The Golden Bough’, Synthese 17, pp. 233–53, 1967. Reprinted in Philosophical Occasions 1912–1951, Oxford, Blackwell, 1993, pp. 115–155. Page reference to the latter.
OC
On Certainty, Oxford, Blackwell, 1969.
PI
Philosophical Investigations, Oxford, Blackwell, 1953.
RFM
Remarks on the Foundations of Mathematics, Oxford, Blackwell, 1967/1978.
RPP I, II
Remarks on the Philosophy of Psychology, Oxford, Blackwell, 1980.
TLP
Tractatus Logico-Philosophicus, London, Routledge & Kegan Paul, 1961.
Z
Zettel, Oxford, Blackwell, 1967. xiii
xiv Abbreviations of cited works
Others A/B
Kant, I. 1787 Kritik der reinen Vernunft, English trans. by Norman Kemp Smith, London, Macmillan, 1929.
AT
Descartes, R. 1641 Meditationes de prima philosophia, English transl. by E. S. Haldane and G. R. T. Ross, The Philosophical Works of Descartes, Cambridge, Cambridge University Press 1911, vol. I. Page reference to the C. Adam and P. Tannery edition of Œuvres des Descartes, Paris, Vrin/CNRS 1964–76.
SA
Searle, J. 1969 Speech Acts: An Essay in the Philosophy of Language, Cambridge, Cambridge University Press.
Introduction
This book focuses predominantly on the notes that Ludwig Wittgenstein wrote, mostly in the last eighteen months of his life, which were edited by the literary executors G. E. M. Anscombe and G. H. von Wright in 1969 under the title of Über Gewißheit – On Certainty. This ‘text’ is, arguably, Wittgenstein’s third and last masterpiece. It has been defined as a ‘gem in the rough’.1 That it is a philosophical gem is clear at first encounter, for it contains a wealth of extraordinarily rich observations on G. E. Moore, idealism and scepticism, the nature of knowledge, belief, doubt, certainty and the role of reasons or justifications with respect to them. Yet it is clear, upon second encounter with it, that it is also ‘in the rough’ because different threads emerge, often in tension with one another. Let us review some of the main difficulties afforded by On Certainty (OC). Consider the incipit of the work ‘If you do know that here is one hand, we’ll grant you all the rest’ (OC 1). Here Wittgenstein is referring to G. E. Moore’s celebrated ‘Proof of an External World,’ published in 1939. The ‘Proof’ is generally considered a puzzling philosophical performance and it has stirred considerable interest upon different generations of scholars and philosophers, since its first appearance, up to present times. But reference to Moore’s work is not confined to his 1939 paper; it extends also to a previous one – ‘A Defence of Common Sense’ – published in 1925. This earlier paper is clearly different from the later one, as it does not aim to offer a proof of a given metaphysical thesis, but simply to characterize a series of truisms which collectively make up our common sense picture of the world, according to Moore. It then endeavours to clarify what kind of philosophical implications that might have, against those who have denied or held views incompatible with the metaphysical and epistemological truths allegedly contained in Moore’s truisms. 1
2
Moore and Wittgenstein
This immediately poses the problem of understanding Moore’s work per se and then of seeing how Wittgenstein was struck by it. Neither of these two tasks is straightforward, though. For generations of interpreters have offered substantially different readings of Moore’s 1939 paper and this bears testimony to the extreme difficulty of making sense of it. Yet the earlier paper, too, is by no means easy to understand. Still Wittgenstein clearly thought there was something extremely important to it. Furthermore, he became interested in these two papers mainly as a consequence of his visit to his former pupil and friend, Normal Malcolm, in Ithaca, in 1949. Malcolm was working on them at the time and, through conversations with him, Wittgenstein started to ponder on their philosophical significance. This is historically important because it means that Wittgenstein’s reception of Moore’s two articles was mediated by the reading of them afforded by one of Moore’s first commentators, viz. Malcolm. Yet Malcolm’s reading of Moore itself was mediated by his earlier attendance of Wittgenstein’s lectures (1938/1939) and, as we shall see, by Wittgenstein’s treatment of epistemic expressions in connection with first-person avowals of mental states, whether sensations or propositional attitudes, as well as by further exposure to Wittgenstein’s influence in 1946–7. This complex entanglement poses a number of historically difficult questions. From a methodological point of view I have therefore decided to devote an entire chapter to Moore’s two papers (with occasional incursions also into other epistemological works by him, whose later publication, however, makes them less relevant to the task of understanding Wittgenstein’s reception of Moore’s work). After a presentation of them, I have then reviewed the main interpretations to which they have given rise, with special attention to Malcolm’s reading of ‘Proof of an External World’, in particular, as well as to his private correspondence with Moore on the topic. This correspondence clarifies in many respects Moore’s own understanding of his later paper, and the differences between him and Malcolm, as well as between him and Wittgenstein, on three interrelated issues: first, the conception of meaning at work in their respective philosophical perspectives; second, their different notions of nonsense – either more semantically oriented, in Malcolm and Wittgenstein, or more epistemically oriented in Moore; finally and connectedly, their contrasting conceptions of the legitimacy of philosophical scepticism. However, it soon becomes clear that Malcolm’s reading of Moore is indeed very partial and can hardly be considered a faithful and illuminating rendition of the latter’s work. Hence, I have also decided to deepen the understanding of Moore’s epistemology per se by looking
Introduction
3
at a number of influential interpretations of it that have been offered since, up to the present day. Thus, I have focused on Thompson Clarke’s and Barry Stroud’s first attempts at reading Moore as a somewhat protocontextualist philosopher; as well as on Crispin Wright’s influential reading of ‘Proof of an External World’ as – ironically – a sceptical argument in disguise, up to a recent, more favourable, appraisal of it proposed by Jim Pryor. For different reasons, I claim that all these readings are somewhat wanting and that they don’t do justice to the complexity of Moore’s reflections. I therefore close by proposing my own interpretation of Moore’s work, which basically sees it as a form of proto-externalism in epistemology. Yet, I must stress, if my reading is on the right track at all, Moore’s externalism is really just in nuce. For it depends on the basic move of divorcing the obtainment of knowledge from the possibility of claiming it, and does not contain any analysis of the concept of knowledge which might sustain such a separation; nor does it contain any clear hint of how the sceptical challenge and its ultimate legitimacy can be dealt with from such a perspective. This passage from the mere reconstruction of the historiography around Moore’s work to a new interpretation of it marks the passage from a simply historiographical to a properly historical work, conceived as a form of rational reconstruction, whose validity, however, can only be tested against textual evidence. That is to say, it is my firm conviction that when engaged in doing history of philosophy – analytic or otherwise – as opposed to doing philosophy as such, we are all first and foremost involved in a process of interpretation of more or less remote texts. However, such interpretations can’t be assessed merely on the basis of their own specific philosophical interest and merits, if they are meant as exegeses. Rather, they must always be evaluated with respect to the text of which they are purported interpretations, as well as to other historical evidence pertaining to what we may generically call the ‘context’ in which the work was created. I believe there is a strong tendency, especially in analytic philosophy, to forget such a methodological requirement and thus to present more how such and such a text struck one, as it were, rather than what that text itself tried to say, in light of its author’s intentions and working context. However, rational reconstructions without textual evidence, or else supported by very little textual evidence, seem to me to be ‘empty’ – to use a piece of Kantian terminology. I have therefore endeavoured to avoid falling prey to such a tendency as much as possible. Whether I have succeeded or not is for the reader to decide. Yet, as an analytic philosopher, I myself am sensitive to what I think are, upon close examination, just rational reconstructions. Let me offer
4
Moore and Wittgenstein
but one example. Take for instance Saul Kripke’s work on Wittgenstein’s remarks on rule-following and the so-called ‘private language argument’.2 Needless to say, its exegetical accuracy is extremely debatable – indeed, as a historian, I would rank myself among those who think that ‘Kripkenstein’ is clearly not Wittgenstein. Yet, there is no denying that that reading is intellectually engaging and philosophically fruitful. How to resolve the tension then, between, as it were, rational reconstructions that are historically empty and historical reconstructions that might end up being philosophically ‘blind’ – to pursue our Kantian comparison? That is to say, how can one resolve the pressure between, on the one hand, offering interpretations that may be accurate from a textual point of view, yet not philosophically germane to future investigation and, on the other, presenting rational reconstructions that may end up having little to do with the text they aim to be exegeses of? I have, as a rule, divorced my activity as a historian from my work as a philosopher. So, if you happen to be interested in ‘Colivenstein,’ as it were, you will not find it here. Not that I have ever been tidy and disciplined enough to pursue these two activities seriatim. My initial interests in Moore and Wittgenstein were as a historian, since my first approach to philosophy in the late 1990s. Then I have abandoned the issue for a number of years, by working essentially on the self and self-knowledge and on my own attempt at retaining as much of Wittgenstein’s anti-realist epistemological outlook as possible, without embracing his view that the basic propositions that form the framework of our epistemic practice are rules. Meanwhile, I have occasionally gone back to my activity as a historian, especially to complete the Italian edition of this book, which came out in 2003. As I became interested in the issue of relativism and wrote a book on the topic in Italian, I also worked on Wittgenstein’s alleged relativism, and the final sections of this book bear testimony to that. Yet I think I have been fortunate to have become interested in two major philosophical contributions: Moore’s epistemological works, on the one hand, and Wittgenstein’s On Certainty, on the other. To try and bring out their own respective philosophically rich insights, without superimposing my own attempt to develop them in directions which may seem fruitful, when engaging in purely theoretical work, is – I think – both historically and philosophically interesting. For it helps better to see why and in which possible directions they may then be expanded, without themselves already containing these later developments, but only their buds, as it were. In other words, I have been lucky enough always to be interested in philosophical classics, which do indeed deserve that name
Introduction
5
because they remain the best and richest examples of philosophy in the making, and have been inspirational to generations of philosophers. It is indeed with this ideal in mind – namely, the ideal to let these texts talk for themselves, while signalling if and when they elicit points that are relevant to contemporary debates in epistemology as well as to a rising way of doing philosophy of language, which pays close attention to the role played by context in determining the proposition expressed by a given sentence – that I have written this book. I hope that this way of proceeding will have another important effect, especially with respect to the perception of Wittgenstein’s work. For, although in recent years there has been something of a Moorean renaissance, mostly fostered by new and detailed reconstructions of his celebrated proof, it is clear that Wittgenstein’s work both in philosophy of language and in epistemology is often considered of interest only to Wittgensteinians nowadays. That is to say, to a varying number of scholars who, while mostly engaged in exegesis, tend to think that Wittgenstein had the last word on pretty much everything. On the contrary, I hope this book will succeed in showing the opposite, viz., that Wittgenstein (and indeed Moore, when it comes to epistemological externalism) had, if anything, often the first word; that his philosophical genius consisted frequently in sensing, and sometimes in developing, philosophical problems and broad outlooks which, before him, were either ignored or not as pressingly and deeply felt as he perceived them, or else, were simply unavailable. If so, then, reading Wittgenstein (as well as Moore), keeping him, as it were, on our philosophical bedside tables, will remain a source of inspiration and will often lead us to think for ourselves. It will then be a powerful propellant for starting doing philosophy, rather than ending it. This, I think, is the basic lesson I have learned from Crispin Wright, although, being more historically minded than him, I tend to sharply distinguish between a sort of ars inveniendi – the activity of tracking down certain themes and problems in Wittgenstein’s (and Moore’s) texts – and a sort of ars inventionis, which consists in further developing those themes and in proposing solutions to those problems, in turn often devised either by analogy or by contrast with suggestions that may be found in the texts. Another advantage of keeping the two activities separate becomes apparent when one considers Wittgenstein’s conception of his own way of doing philosophy as essentially therapeutic rather than constructive. No doubt there may well have been elements of self-deception in this sort of self-appraisal. Yet, one can’t simply dismiss this quite explicit attitude of Wittgenstein. Still, the real question is this: why should one want to do so? Well, essentially for two reasons, I take it. First,
6
Moore and Wittgenstein
because as remarked, Wittgenstein wasn’t always consistent with his metaphilosophical pronouncement and at least at places he seems to be engaged in some kind of constructive philosophical activity after all. Second, because if one is interested in Wittgenstein mostly as an inspirational source for one’s own way of thinking, then one will have the more or less conscious tendency to play down those remarks, as taking them at face value would presumably prevent or deter one from developing his and/or one’s own ideas in a constructive manner. But if one feels free to give Wittgenstein his due, while divorcing one’s own thinking from the idea of thereby bringing to light something which was already present in Wittgenstein, as opposed to merely inspired by him, one will also feel free to acknowledge, on the one hand, the therapeutic elements in Wittgenstein’s own thought, while, on the other, ignoring them in one’s own philosophical activity, if one sees that fit. So much for an explanation of the methodology employed in this work. Let me now go back to a description of its content. I said before that On Certainty is a difficult text. We have been reviewing one reason for such an assessment, namely the difficulty of seeing clearly the relationship with its official opponent, viz. G. E. Moore. But this is just one reason. There are at least three more that are somehow interconnected. The first one is that On Certainty revolves around the idea that Moore’s truisms and first premise of his proof – ‘Here is my hand’ – as well as propositions similar to them, are like ‘hinges’ that must stay put in order for our language and epistemic practices to be possible. Yet as soon as we look at them, the truth of this claim is far from obvious. For instance, it is perfectly conceivable that, on occasion, ‘Here is my hand’ be a genuinely empirical proposition, subject to verification and control; and indeed, one of Wittgenstein’s favourite examples, ‘Nobody has ever been on the Moon,’ simply appears false to us. By contrast, things look much better when one considers ‘The Earth has existed for a very long time,’ which obviously must stand fast if we are to pursue our usual epistemic practices, such as history or geology; or else, ‘There are physical objects,’ which, pace some interpreters, is a hinge for Wittgenstein. For, if we gave that up, how could we keep having a conceptual scheme of mindindependent objects and how could we possibly take testimonies and perceptual experiences to bear upon them? Yet, it remains to be clarified in what sense, if any, hinges are hinges. From a structural point of view, Wittgenstein repeatedly claims that hinges are (i) neither true nor false; (ii) neither justified nor unjustified;
Introduction
(iii) (iv) (v) (vi)
7
therefore, they are neither known nor unknown; hence, they cannot sensibly be called into doubt; furthermore, they aren’t either rational or not rational; thus, finally, for these very reasons, they aren’t empirical propositions but rules.
This set of claims poses a series of difficulties, both because they don’t seem unequivocal, if one goes through On Certainty, and because, even supposing they constituted the backbone of Wittgenstein’s official view – known, in the literature, as the framework reading of On Certainty – they would seem problematical. For instance, it is extremely weird to say that ‘Here is my hand’ is neither true nor false, when I am holding my hand up in front of my face in what appear to be perceptually and cognitively optimal conditions. Similarly, we take ourselves to have an overwhelming body of evidence in favour of the fact that, for instance, the Earth has existed for a very long time, up to the point that, if we know anything at all, we know at least that much and, surely, to believe such a thing is at any rate entirely justified and therefore rational. Moreover, among hinges, as we have seen, there is ‘Nobody has ever been on the Moon,’ which simply appears false to us, let alone dubitable. Furthermore, if they are rules, supposing one could make sense of this claim, what would they be rules of? The Philosophical Investigations, as well as Wittgenstein’s middle-period writings, made us familiar with the idea of rules of grammar; roughly, with the idea of meaning constitutive rules, albeit rules determined by linguistic use itself and therefore dependent on it. But ‘The Earth has existed for a very long time’ could only be regarded as meaning constitutive on a rather holistic picture of linguistic meaning, which is neither clearly Wittgenstein’s official view, nor obviously correct as such. So, in what sense, if any, could it be a rule of grammar? Moreover, if hinges are rules because they are neither true nor false, as well as neither epistemically assessable, in what sense, if any, could they still be regarded as propositions? That is, if bipolarity is the mark of propositionality, how could Wittgenstein claim that they are hinge-propositions? Wouldn’t it be more correct to say that they are strings of sounds or graphemes that don’t express any proposition at all? And, if so, wouldn’t that explain one of Wittgenstein’s further claims – tentative as it might be – that with respect to them, all of our propositional attitude verbs – for example, ‘to assume,’ ‘to believe,’ ‘to know,’ ‘to be certain,’ etc. – seem to be inappropriate? Yet, if so, what on earth could the certainty we have with respect to them – whence the title of the book – be? Finally, if hinges are rules, isn’t there a danger of embracing
8
Moore and Wittgenstein
epistemic relativism, or at least of making it intelligible? After all, we are familiar with the idea of different, even formal games, defined by different sets of rules or axioms. Or else, was Wittgenstein somehow proposing a new and sui generis form of epistemic foundationalism, where at the foundations lie rules, rather than immediately verified propositions? A further reason why understanding On Certainty is difficult depends on the complexity of its anti-sceptical strategy. For, if Moore is clearly one opponent, On Certainty is by no means a sceptical treatise. Yet, when one tries to pin it down as far as its anti-sceptical potential is concerned, one will run into several exegetical troubles. Just to give the reader a sense of the difficulty of making sense of the anti-sceptical strategy contained in that book, consider that, to the best of my knowledge, at least four different families of interpretations of this issue have been proposed. They can be graded according to whether they attribute to Wittgenstein a view of scepticism as nonsensical, in two senses to be further elucidated; or whether they allow that, even for Wittgenstein, scepticism did indeed make sense; and, furthermore, that it could actually be answered. So, to clarify, the therapeutic interpretation, due essentially to James Conant, has it that, in On Certainty, scepticism is deemed utterly and radically nonsensical, as sceptical doubts are expressed outside any context of use and, given the connection between use and meaning, can’t be endowed with sense. Sceptical doubts are therefore meaningless much like the combination of signs ‘Ab sur um.’ The framework reading, in contrast, has it that hinges are rules which, as such, can’t be subject to epistemic appraisal. Hence, scepticism, which doubts them, raises a doubt where it cannot rationally be sustained. Therefore, sceptical doubts are nonsensical not because they are meaningless but because they are irrational. The naturalist reading, mainly proposed by Peter Strawson, holds that, for Wittgenstein, sceptical doubts are neither meaningless nor irrational, just unnatural, since they are raised with respect to propositions that we find natural to take for granted, given our upbringing within a community that collectively holds them fast. Finally, the epistemic reading, championed by Crispin Wright and Michael Williams, but somehow connected to earlier work by Thomas Morawetz, claims that, for Wittgenstein, hinges are really not rules but propositions that, at least in context, cannot be evidentially justified. Yet, if we broaden our conception of warrants so as to include also non-evidential ones – call them ‘entitlements’ – we can actually see how, even for Wittgenstein, hinges can in point of fact be justified, contrary to what sceptics maintain. Furthermore, if we recognize that Wittgenstein didn’t tie either truth to evidential justification,
Introduction
9
or knowledge to a subject’s ability to provide warrants for one’s claims, and actually held a somewhat externalist conception of knowledge, we can go so far as to argue that he himself maintained the view that hinges are known. Either way, on this interpretation, Wittgenstein would actually recognize the full intelligibility and appropriateness of the sceptical challenge, and in fact devise a rather sophisticated line of response to it. Finally, going from one extreme to the other of this taxonomy, the overall appraisal of On Certainty would change, as it would go from taking it as a merely therapeutic to a robustly theoretically committed text. One final reason why making sense of On Certainty is difficult has to do with the issue which, provocatively, one could put like this: why that title? Now, it is a well-known fact that it was chosen by the editors, yet was their choice utterly misguided or deeply motivated? Again, given the number of interpretations of On Certainty available nowadays, answers to these questions vary considerably. For, on the one hand, if one holds the therapeutic reading, according to which Wittgenstein didn’t put forward any substantive view of hinges but merely aimed to cure ourselves of the temptation of either doubting them or of insisting that we knew them, it is simply not clear whether certainty would have a definite object, and what kind of attitude it would be. On the other hand, if one held a naturalist reading of hinges, they would be certain because, given our practices, they are actually exempt from doubt. Moreover, our attitude with respect to them would be of a somewhat instinctive kind; that is to say, we would take them for granted because we have been brought up within a community that behaves that way with respect to certain propositions. If, in contrast, one maintained an epistemic reading, the category of certainty as a propositional attitude would simply be inappropriate, though there would be certainties, viz. propositions for which doubts would be ungrounded and for which entitlements could be provided. For, to put it succinctly, we would either have non-evidential warrant for them, hence justified belief in them, or actually knowledge of them. Certainty as a propositional attitude would thus drop out of consideration on this view. Finally, on the framework reading, hinges are certain because of their normative role that exempts them from doubt. Moreover, as to our attitude with respect to them, if we held the ineffabilist conception of hinges, whereby, failing to be propositions, they could neither be said, nor made the object of any propositional attitude, our certainty with respect to them would have to be thought of as non-propositional, non-conceptual and therefore of a merely ‘animal’ kind. Or else, if we maintained that hinges, while being rules, are still propositions – on a fairly relaxed notion of proposition – we
10
Moore and Wittgenstein
could claim that, as such, they can be the object of a propositional attitude of certainty which parallels the kind of attitude we bear to our most wellentrenched, yet fully explicit rules. That is to say, we do accept them and hold them fast, we behave in accord with them, or else accept criticism for not doing so, and, finally, pass them on to our children. As will become apparent, I personally maintain one specific version of the framework reading, whose elements of novelty with respect to other ones are mainly as follows. First, in my view, hinges are not just meaning constitutive rules, but also, and in fact in most cases, rules of evidential significance. Hence, I side with those interpreters of Wittgenstein who actually hold that after the Philosophical Investigations Wittgenstein came to broaden both the notion of rule – such as Wright and Moyal-Sharrock – and that of grammar, so as to include this kind of epistemic rules. However, I don’t know whether this would justify the claim that there actually is a ‘third Wittgenstein’, after the one of the Tractatus and of the Philosophical Investigations. Ultimately, I don’t think the counting is that important; all that matters is to be clear about the basic notions at play here and the developments to which they were subject. Second, I don’t embrace the ineffabilist claim that since hinges aren’t bipolar, they can’t be propositions and that, therefore, they remain, as certainties, always beyond the possibility of being said. On the contrary, I hold the view that while failing at bipolarity, they are still propositions, albeit with a normative function, rather than a descriptive one, and that we do indeed express them on various occasions: either to teach them to someone who ignored them or to remind someone of them were they to violate them, as philosophers such as Moore and a sceptic do. Contrary to other framework interpreters, most notably Moyal-Sharrock, I do think that at least the former context in which hinges are actually said is a genuine language game, by Wittgensteinian lights. Hence, hinges do indeed figure as such within language games, yet not as part of epistemic ones, where empirical propositions are put forward and assessed. Third, I do think that the anti-sceptical strategy presented in On Certainty actually depends on an entailment between the irrationality of raising doubts with respect to rules which, as such, cannot be subject to epistemic evaluation, and the meaninglessness of those doubts. Hence, to my mind, there is something important to learn from the therapeutic reading. What supporters of the therapeutic reading are mistaken about, I think, is to hold that the primacy of use prevents us from attributing Wittgenstein the view of hinges as rules. For, granted that primacy, I actually think that rules are what we can recognize ex post as held fixed by our actual practices. Fourth, I hold that
Introduction
11
the title On Certainty is entirely appropriate for two different reasons. First, because this text deals with hinges, that is to say, with propositions whose status of certainties depends on their role of basic rules of our conceptual scheme and epistemic practices. Hence, their certainty is of a ‘grammatical’ (or even ‘logical’) nature (provided ‘logic’ and ‘grammar’ are taken as synonyms), not of a psychological, or animal one. Yet, there is also a sense in which our attitude with respect to them is rightly characterized as certainty, as opposed to knowledge, or (non-evidentially) justified belief or whatever you have. For we do accept them as rules, in a way which knows no doubt – we hold them fast – and, in view of that, behave in accord with them, accept reproach for not doing so, and transmit them to future generations by teaching them a given language and to take part in our epistemic practices. Hence, I maintain that certainty as an attitude is a form of acceptance not of ungrounded presuppositions but of rules, conceived of, in their turn, as normative propositions. Thus, I contrast, at once, both epistemic readings such as Wright’s, whereby we would accept these truth-apt propositions, without any evidential warrant to do so; as well as other framework readings, such as Moyal-Sharrock’s, according to which certainty with respect to hinges could never be a propositional attitude, as its objects wouldn’t be propositions, and should thus be regarded as some kind of ‘animal’ certainty. Finally, I conclude by showing why the framework reading need not be committed either to interpreting Wittgenstein as an epistemic foundationalist, or to attributing him some form of epistemic relativism. For, on the one hand, Wittgenstein’s hinges are such that we bear no epistemic relation to them and, without that much, it seems to me there really is no room for foundationalism. On the other, rules are entrenched in a practice that, quite clearly, could have been metaphysically different from what it in fact is. Yet, from the inside of such a practice, as we in fact are, we can’t really conceive of its alternatives. Apparent examples to the contrary are then defused by appealing also to textual evidence, both in On Certainty and in the Notes on Frazer’s ‘The Golden Bough’ and in Culture and Value. Hence, I conclude, somewhat against generations of interpreters, that also by Wittgenstein’s lights epistemic relativism remains a mere metaphysical possibility, but that it is not a description of our actual epistemic situation, nor of an epistemic condition we can really make sense of. So, if epistemic relativism has a grip on us at all, it remains ineffectual, as we could not really conceive of it in detail. This, in broad outline, is the reading of On Certainty I defend. Yet, exactly because I think, in agreement with ‘therapists’ as well as with many other
12
Moore and Wittgenstein
commentators, that in the later Wittgenstein the observation of linguistic use and epistemic practices is the starting and fundamental point, my analysis and eventual defence of the view just presented proceed in three stages, corresponding to each of the last three chapters of this book. The first and the second chapters pay close attention to Wittgenstein’s observations respectively about the use of ‘I know’ in everyday contexts, in philosophical ones and in a grammatical sense; and to his remarks about the use of ‘I doubt’, both in ordinary contexts and in philosophical ones. Those analyses allow me to clarify, on the one hand, Wittgenstein’s view of knowledge, which I regard as firmly internalist, and to defuse appearances to the contrary, as essentially due either to a conflation between propositional and practical knowledge or to the oblivion of the grammatical use of ‘I know’. They also allow me to further clarify Wittgenstein’s conception of meaning and nonsense, as well as to investigate the complexity of his anti-sceptical strategy and thus to oppose, in particular, naturalist and epistemic readings of it. In the fourth and last chapter, in contrast, I take up the issue of the status of hinges, their many different kinds, their propositional nature, their allegedly foundational role, and the kind of propositional attitude we bear to them, as well as the issue of whether they open the way to epistemic relativism. In passing, I contrast my interpretation with other prominent relatives inside the family of so-called ‘framework readings’ of On Certainty. Finally, in the Conclusion, I signal the main similarities and differences between Moore and Wittgenstein with respect to epistemology and philosophy of language. Along this lengthy and winding road, I have placed a number of signposts to indicate the various junctions at which both contemporary philosophy of language and epistemology may be seen as having their departure points. As said, both Moore’s and Wittgenstein’s works are classics, and their being so is, in my view, tantamount to their having had the first word on many issues which, nowadays, we all perceive as philosophically important. That should also make us curious to read them, in order to deepen our self-understanding, by giving us a sense of our place in the context of a philosophical tradition. Still, it is only to be expected that by further reading them, someone will be inspired to develop some of their suggestions in yet unforeseen directions. My wish, in effect, is ultimately this. Yet, if so much happens, I also hope people will give credit to their inspirational sources. This, to my mind, is by no means a sign of lack of originality.
1 G. E. Moore: Scepticism, Certainty and Common Sense
In this chapter we will discuss two papers by George Edward Moore which are, in many respects, the focus of Wittgenstein’s criticism in On Certainty: ‘A defence of common sense’, and ‘Proof of an external world’, published in 1925 and 1939 respectively.1 As I emphasized in the Introduction, the aim of this chapter is threefold: on the one hand, to understand and evaluate Moore’s work per se; on the other, to comprehend better the way in which Wittgenstein received it; and, finally, to present and discuss the most relevant interpretations of Moore’s work, since its first appearance, some of which are of special interest also to contemporary epistemology. In between Moore’s two papers there are fourteen years and substantial differences. The former is a characteristic instance of that appeal to common sense, against idealist or sceptical philosophical theses, of which Moore is rightly considered to have been one of the main supporters throughout the history of Western philosophy. To appeal to common sense means to oppose those who, in doing philosophy, deny the existence, or the possibility of knowing that there are physical objects, the self with its own mental states, other minds, space and time. A philosopher of common sense holds that all these theses are not only paradoxical, but altogether false. For, if they were true, it would then be impossible, in his view, to formulate all the propositions we commonly use in everyday life, that are either about an external world, or about one’s own and others’ bodies and minds, or about space and time. A philosopher of common sense takes advantage of such a consequence to impugn the legitimacy of idealist and sceptical philosophical theses. He doesn’t think, however, that a diagnosis of what he deems to be the mistakes made by either idealists or sceptics should be provided. Rather, it is enough for him to denounce them. He then counters them 13
14
Moore and Wittgenstein
by holding the opposite theses, for example, that there are physical objects and minds, that time exists, and that we know all this and can’t sensibly doubt it. All these theses are either expressed or implicit in what Moore presents as common sense truisms. That is to say, those evidently true empirical propositions, accepted and known by all speakers of a language which, for this very reason, express a common sense view of the world. Namely, the view of the world that all people who possess common sense hold on to. Thus, the appeal to common sense turns into a defence of the metaphysical and epistemological theses that are taken to be contained in it. In ‘Proof of an external world’, in contrast, Moore tried to prove – against idealism – that one of the metaphysical theses implicit in his truisms – namely, that there is an external world – is true. Thus he somehow rejected his earlier position, according to which that thesis couldn’t be proved. This latter paper attracted the attention of Norman Malcolm, a pupil and friend of Wittgenstein’s and of Moore’s himself. He discussed and criticized Moore’s article at length and shared his ideas with Wittgenstein during the latter’s stay at Cornell in 1949. For Wittgenstein who, up to that point, had always ignored epistemological issues, those discussions were the occasion to develop the reflections which occupied him during the last 18 months of his life and were published posthumously in On Certainty. But Moore’s proof has fostered a number of different interpretations since it was first presented. Some of them are particularly relevant for the ongoing debate in epistemology, either because they are the object of that debate or because they have somehow inspired it. As we shall see, however, though interesting and thought provoking, all of them are somewhat wanting as purported renditions of Moore’s performance and overall strategy. In closing I will then put forward a different interpretation of what Moore was really up to, which, interestingly, is significant for contemporary epistemology as well.
1 ‘A defence of common sense’ In 1925 Moore published ‘A defence of common sense’ (henceforth DCS), which Wittgenstein considered one of Moore’s best papers. In DCS Moore lists a number of truisms – that is to say, empirical propositions in the first person mostly about his own mind and body and maintains that (1) he knows with certainty that they are true when they are about himself;
G. E. Moore 15
(2) everyone else knows that they are true when they are about himself or herself; (3) he knows, just like everyone else knows of him and of everyone else, that everyone knows with certainty that they are true when they are about each of them. For this very reason Moore’s truisms, although trivial, express a kind of common sense knowledge. Here are some examples:2 (1.a) There exists at present a living human body, which is my body. (1.b) This body was born at a certain time in the past, and has existed continuously ever since, though not without undergoing changes. (1.c) Ever since it was born, it has been either in contact with or not very far from the surface of the earth. (1.d) At every moment since it was born there have also existed many other things, having shape and size in three dimensions, from which it has been at various distances. (1.e) Also, there have existed some other things of this kind with which it was in contact. (1.f) Among the things which have, in this sense, formed part of its environment there have, at every moment since its birth, been large numbers of other living human bodies, each of which has, like it, at some time been born, continued to exist from some time after birth and been, at every moment of its life after birth, either in contact with or not far from the surface of the earth; but many of these bodies have already died and ceased to exist. (1.g) The earth had existed also for many years before my body was born. (1.h) For many of these years, a large number of human bodies had, at every moment, been alive upon it; and many of these bodies had died and ceased to exist before it was born. (1.i) I am a human being. (1.j) I have, at different times since my body was born, had many different experiences of many different kinds. (1.k) For instance, I have often perceived things, observed facts. (1.l) I have had expectations and beliefs, both true and false. (1.m) I have imagined things; had dreams and feelings of many different kinds. (1.n) Very many of other human bodies which have lived upon the earth, each has been the body of a different human being, who has, during the lifetime of that body, had many different experiences of these (and other) different kinds.
16
Moore and Wittgenstein
With respect to (2), here is what Moore writes:3 In the case of very many (I don’t say all) of the human beings belonging to the class (which includes myself) defined in the following way, i.e. as human beings who have had human bodies, that were born and lived for some time upon the earth, and who have, during the lifetime of those bodies, had many different experiences of each of the kinds mentioned in (1), it is true that each has frequently, during the life of his body, known, with regard to himself or his body, and with regard to some time earlier than any of the times at which I wrote down the propositions in (1), a proposition corresponding to each of the propositions in (1), in the sense that it asserts with regard to himself or his body and the earlier time in question (namely, in each case, the time at which he knew it), just what the corresponding proposition in (1) asserts with regard to me or my body and the time at which I wrote that proposition down. Finally, (3) follows from (1) and (2): just as Moore knows that some propositions that are about himself are true – (1) – similarly everyone else knows that those propositions are true when they are about each of them. But since (2), in its turn, is a common sense truism that Moore claims to know, then not only Moore, but also everyone else knows that (2) is true. Hence, each of us knows that any other person knows that the propositions in (1) are true when they are about himself or herself – (3). 1.1
Ordinary meaning and the analysis of meaning
According to Moore, his truisms are not only known with certainty by (almost) everyone, but they are also perfectly well understandable to any competent speaker of English. However, this is far from saying that the analysis of their meaning is known.4 Such a claim clarifies how, for Moore, there is, on the one hand, an ordinary meaning of words and sentences, and, on the other, a legitimate philosophical activity of analysis of that meaning. In his view, the former must always be kept in mind in order to avoid formulating theses which, once translated into ordinary language, could not be maintained, as they violate or forbid the ordinary usage of linguistic expressions. The latter, in contrast, allows us to understand the logical structure and the epistemological and metaphysical bases of our ordinary language. Despite the fact that the distinction between the understanding and the philosophical analysis of ordinary meaning is largely plausible, it
G. E. Moore 17
may sound surprising that Moore, in Part IV of DCS, where he goes back to the issue of the analysis of meaning of those truisms that are about or assume the existence of an external world, attempts an analysis of them in terms of sense data – albeit recognizing its failure. For, as we will see, this is problematical for a philosopher of common sense. According to Moore, the proposition ‘At present I am perceiving a human hand’ derives from the conjunction of two more basic propositions: ‘I am perceiving this’ and ‘This is a human hand.’ Now, the problem that Moore himself recognizes as pressing is to understand what kind of entity is the one which is said to be perceived and whether it is identical to what is usually taken to be a hand. According to him, it is certain that ‘this’ in the first conjunct refers to a sense datum and that ‘this’ in the second conjunct does not refer to a hand but to what is representative of it.5 In Moore’s view, however, it remains a problem to determine what a sense datum really is. In his opinion, there are at least three possible explanations, although he thinks that none of them is undoubtedly correct. Here they are: (i)
The sense datum itself is part of the surface of the hand. That is to say, it is identical to that part of the surface of the hand I am now actually perceiving. (ii) There is only one thing (or set of things), for which it is true both that it is part of the surface of a human hand and that this sense datum is a manifestation of it; where the relation it bears both to the surface of the hand and to the sense datum which is a manifestation of it, isn’t further analysable. (iii) Physical objects are permanent possibilities of sensation. Therefore, my current perception is a perception of a hand if I am able to formulate (a finite?) conjunction of propositions about sense data and about the circumstances of their occurrence like the following ones. If the sense datum I am now having is representative of a human hand, then I ought to have had such-and-such a sense datum if I had perceived it from this side, and from that side; from below, and from above, etc. According to Moore, there are objections to all these possible explanations. Against (i) it can be objected that when one has a double image, there are two sense data. It is thus impossible that both of them be identical to the only part of the surface of a hand, which is actually given to one. Hence, sense data are only representative of physical objects and can’t be identical to their perceived surfaces.6 As far as (ii) is
18
Moore and Wittgenstein
concerned, it remains an open question what this alleged entity, which should stand in a no further analysable relation both to the surface of the hand and to the sense datum, really is and how it can be known to us.7 Finally, against (iii) it can be objected that the various relations among different sense data, which should determine that what is being perceived is a human hand, can be substituted by relations among (parts of) material objects, for example, ‘If what I am now perceiving is a human hand, then from this point of view I should see five fingers, whereas from another point of view, I should see only three, etc.’ Moreover, it is not obvious at all that there is a clear and known relation among the sense data caused by a given material object on certain circumstances and those caused by it on different ones, which would enable a sure-fire identification of the physical object in question.8 The impossibility or, at the very least, the incapability, acknowledged by Moore himself, of clarifying what a sense datum is and what relation it should bear to the physical object which is its supposed cause, opens up a gap between knowing something about physical objects themselves and knowing something about the way they appear to us. This, in turn, creates a problem for the alleged commonsensical knowledge that there are material objects. Moore could obviously reply that the appeal to sense data, from his own point of view, is meant only to give an analysis of the meaning of propositions about material objects and that, whatever its result might be, it won’t prevent physical objects from being part of the common sense picture of the world and thus of our shared knowledge. However, Moore’s position remains problematical also if taken this way. As we have already seen, the proposition ‘At present I am perceiving a human hand’ derives, in his view, from two simpler propositions: ‘I am perceiving this’ and ‘This is a human hand’. Now, ‘this’ in the first conjunct refers to a sense datum; and, in the second conjunct, what ‘this’ refers to is predicated to be a hand. Moore is very careful to say that the meaning of ‘hand’ isn’t the sense datum to which ‘this’ refers; and that ‘to be a hand’ is predicated of the sense datum insofar as the latter is representative of a hand. Yet, even so, it remains that one would be using a public term – ’hand’ – to predicate something of a private object (as long as by ‘sense datum’ one does not mean (i)). Still, this way, it would become problematical to secure that people predicated ‘hand’ of the same kind of object. Moreover, as is well known, in the Philosophical Investigations, Wittgenstein forcefully objected to the idea of a language putatively used to talk about private entities, on the grounds that that would
G. E. Moore 19
annihilate the difference between seeming right and being right, which is fundamental to linguistic meaning.9 Finally, although, as Moore stresses, ‘hand’ isn’t the name of the sense datum one is having, but refers to a three-dimensional object with many parts that are simply invisible to the naked eye (such as the bones), why suppose that the analysis of the meaning of the proposition ‘This is a hand’ should proceed by appeal to sense data? Couldn’t one simply hold that ‘to be a hand’ is predicated of the physical object currently perceived, though, necessarily, only some parts of it are actually visible and presently seen? Hence, to sum up, Moore distinguishes between the ordinary meaning of a linguistic expression and the analysis of it. This in turn makes it seem prima facie plausible to attempt an analysis of the meaning of his truisms in terms of sense data. However, the problematical nature of the latter – which Moore himself acknowledged – makes this attempt difficult with respect to both its epistemological and semantic consequences. As we have seen, the semantics it would entail may well be incoherent and the epistemology it would lead to would cast doubts on the fact that the common sense picture of the world and, in particular, the belief in the existence of an external world, could really be certain and beyond doubt. 1.2
The defence of common sense
In defence of common sense Moore maintains that the common sense picture of the world is absolutely true10 and not just because it is accepted and shared by everyone, but because it consists of evidently true propositions, which are obvious and indubitable. For, otherwise, our language and actions would be unintelligible. However, these truisms stand opposed to idealist and sceptical theses. It therefore follows that both idealism and scepticism are wrong. Yet, the relevant point is that the latter aren’t proved to be wrong. Rather, their falsity is simply (and mostly, as we shall see) taken to be entailed by the fact that truisms of common sense are true, where, in turn, their truth isn’t proved but somehow assumed, also as a precondition of meaningful thought and action. According to Moore, an idealist denies the reality of material objects, of space and time and of the self.11 As a consequence, he can’t deem true the proposition ‘There is at present a human body which is my body’, or any other proposition among those listed by Moore, which entail the reality of space, time, material objects and the self. However, according to him, an idealist maintains something false, as he can uphold his position just because he is a human being (albeit a philosopher) who
20
Moore and Wittgenstein
can’t deny either to have a body, or that he has had thoughts and memories.12 Furthermore, when such a philosopher attributes to other philosophers the mistaken, by his lights, view that there are material objects, space, time and the self, he is implicitly assuming to know that there are other human beings – that is to say, the philosophers he is countering – for which each of the propositions in (1) is true. So he can’t consistently deny the existence of material objects, space, time and the self which are entailed by (1).13 Moreover, if he thinks that the belief in the existence of an external world, space and time and the self entailed contradictory propositions – as Kant and Bradley respectively maintained in the Transcendental Dialectic of the Critique of Pure Reason and in Appearance and Reality14 – and was thus to be rejected, he would simply be wrong, according to Moore.15 For – he argues – that there is an external world, for instance, is a true proposition; thus, it can’t entail a contradiction. Finally, Moore claims that the negations of the propositions entailed by his truisms aren’t self-contradictory, but, nevertheless, they are false, since truisms are true.16 In contrast, in Moore’s view, a sceptic, who denies that one can know that there are material objects, or other people, maintains a selfcontradictory position. For, according to Moore, he recommends to consider the existence of material objects and of all other people as beliefs (as opposed to knowledge) which make up the common sense picture of the world. By so doing, he holds that propositions such as ‘There are physical objects’ or ‘There are other people’ are only believed, rather than known to be true, yet shared by most people, since they belong to ‘common sense’. However, in Moore’s opinion, a sceptic is thereby committed to maintaining, among other things, a proposition such as ‘No human being has ever had knowledge of the existence of other human beings’, which is not just about the sceptic himself, but also about all other human beings, whose existence he is calling into doubt. To put it otherwise, when a sceptic says, ‘No human being has ever had knowledge of the existence of other human beings’ he is thereby saying, according to Moore, ‘There have been many human beings, myself included, and none of them (myself included) has ever had knowledge of other human beings’. This latter proposition is in fact self-contradictory since it assumes the existence of what it denies to be known to exist, that is to say, other human beings.17 However, in Moore’s view, a sceptic wouldn’t be better off if he claimed, ‘The belief that there are material objects and other people is a belief of common sense, but it doesn’t amount to knowledge.’ This would be
G. E. Moore 21
tantamount to claiming, ‘There have been many more human beings beyond myself who have shared these beliefs (since these are beliefs of common sense, hence they are shared by numerous people), but neither I nor anyone else has ever known them to be true’. For, according to Moore, this latter proposition is, once more, self-contradictory, since it assumes the existence of common sense beliefs and hence, in particular, of other human beings, but only to deny that it is known that they exist. Clearly Moore’s arguments can’t be taken to be a confutation of either idealism or scepticism. For, on the one hand, an idealist doesn’t take the category of physical objects as fundamental and denies that the correct analysis of propositions about the so-called external world can be in terms of objects which exist independently of the minds of those who perceive and make judgements about them. Hence, he need not deny that there is an external world, but only to offer a different analysis from common sense of what that means. In response to this rejoinder, Moore, in section II of DCS, makes two claims: namely, that he neither thinks that physical facts such as the existence of the earth for a long time in the past are logically dependent on some mental facts – that is to say, that (at least some) physical facts entail mental ones,18 or, equivalently, that they couldn’t conceivably have been the case, unless there had been some mental facts, as Berkeley held; nor that they are causally dependent upon (either human or divine) mental states.19 Now, this response is much more convincing than Moore’s mere appeal to common sense and his standard way of defending the latter by claiming that the denial of the metaphysical theses implicit in it would commit one to holding something false given that common sense is true. Yet, as is apparent, it is an entirely philosophical move, which rightly objects to strong forms of idealism on the grounds that there doesn’t seem to be any contradiction in thinking, for instance, that the earth exists independently of our own or God’s existence. Of course this is no conclusive argument against such a form of idealism, but at least shows that the latter is far from being mandatory. Moreover, to deny that the earth could – causally – have existed (and did exist) before our own appearance on the face of it, would entail either retroactive causation at the macrolevel or a contradiction.20 Hence, Moore was quite right to object to this kind of idealism, even though these considerations couldn’t carry over to establish the falsity of a form of ‘causal’ idealism whereby a divine mind would be the cause of physical facts.21 On the other hand, it is quite obvious that a sceptic can avoid claiming ‘No human being has ever had knowledge of the existence of
22
Moore and Wittgenstein
other human beings’ and confine himself to suspend judgement about any alleged knowledge about empirical propositions, such as that there are other human beings. Moore too anticipates this last objection. However he claims that a sceptic is committed to maintaining propositions which entail, on the one hand, that he knows with certainty that other human beings exist or have existed – because he would talk of these beliefs as belonging to common sense, and hence as belonging to many human beings – while denying, on the other, that anyone may ever know with certainty that, for instance, other human beings exist or have existed.22 But, once more, it seems that a sceptic may simply hold that insofar as one – perhaps only himself at an earlier stage in his life – believes propositions in (1), one can’t really have knowledge of them, despite the fact that those beliefs may seem as true and certain as ever. Be that as it may, what is more interesting in Moore’s position is his unfailing defence of the common sense view of the world, rather than whether it is ultimately cogent or entirely successful. Such a defence contains various threads which are sometimes in tension with one another. However, it is worth surveying with care because it has many elements which will be taken up – either critically or approvingly – in Wittgenstein’s On Certainty. First of all, as we have seen, for Moore the truth of his truisms seems to him evident. Now, what does this really mean? It could mean that their truth is self-evident, and therefore known immediately and with no inference, because a given proposition is self-verifying. But, of course, while ‘I am thinking’ enjoys such a property, when entertained by each of us, it is more difficult to maintain this view with respect to, say, ‘The earth has existed for a very long time.’ So, it may be that by ‘evident’ Moore meant that his truisms were known immediately, namely, on the basis of no inference. But that again seems problematical, for what could it mean to say that we know them immediately? Certainly we don’t have an intuitive grasp – whether or not we could make sense of such a notion – of the truth of what, by Moore’s lights, are empirical and contingent propositions mostly about physical objects. If intuition can be called upon at all, it would have to be appealed to in connection with different kinds of propositions, such as logical-arithmetical ones, or first-person psychological ones. One could then suggest, that by ‘evident’, Moore meant ‘indubitable’. But ‘indubitable’ in what sense? Apparently for a subject who seems psychologically incapable, given his overall beliefs, to call them into question and feels therefore certain of their truth and of knowing it.
G. E. Moore 23
No matter what interpretation we chose, then, Moore’s knowledge of the truth of his truism would seem to be a function of the kind of mental state he is in, and is aware of, rather than dependent on there being objective grounds in their favour and of having access to them. As we will see in the next chapter, Wittgenstein will strongly attack the model of knowledge as a mental state, introspectively known to its subject, to favour instead a model according to which claims to knowledge are meaningful just in case public criteria of control are satisfied and knowledge itself depends on their being objective grounds in favour of a given proposition (together with the fact that that proposition be true and believed). It must be noticed, however, that in DCS Moore asks himself: But do I really know all the propositions in (1) to be true? Isn’t it possible that I merely believe them? Or know them to be highly probable? And answers: In answer to this question, I think I have nothing better to say than that it seems to me that I do know them, with certainty. It is, indeed, obvious that, in the case of most of them, I do not know them directly: that is to say, I only know them because, in the past, I have known to be true other propositions which were evidence for them. If, for instance, I do know that the earth had existed for many years before I was born, I certainly only know this because I have known other things in the past which were evidence for it. And I certainly do not know exactly what the evidence was. Yet all this seems to me to be no good reason for doubting that I do know it. We are all, I think, in this strange position that we do know many things, with regard to which we know further that we must have had evidence for them, and yet we do not know how we know them, i.e. we do not know what the evidence was.23 Here Moore is saying that one’s knowledge of these propositions depends on one’s further knowledge of some other propositions which entail them, despite the fact that we no longer remember which known propositions served as bases for the ones comprised in (1). In fact, according to him, we must have had this kind of evidence for them, if we can now sensibly say that we know them. Moore’s position can seem ironical: he is saying that he knows these propositions because he has known other propositions which were
24
Moore and Wittgenstein
evidence for them, though – alas – he can’t produce such evidence. I think most of us would, in contrast, quite unhesitatingly say that, for instance, we know that there is, at present, a human body which is our body because we see and feel it; or that we know that the earth has existed for a very long time before our birth because of testimonies about our ancestors, and historical and geological studies. Yet I think that Moore’s apparent simple-mindedness conceals three important elements. The first one concerns the temporal order in which, allegedly, we would have got evidence for these propositions: for, according to Moore, if there has been evidence for them at all, it must have been given to us quite early on. The interesting aspect of this claim is that what allegedly makes our knowledge secure with respect to them is not the product of philosophical reflection, and is somehow as old as ever. As we shall see in Chapter 2, Wittgenstein will repeatedly stress these points too. The second interesting feature of Moore’s position is that he thinks that it is because we know other propositions that we do know both propositions in (1) and some very general propositions entailed by them, such as that there are an external world, space, time and the self. This, again, will be a point on which Wittgenstein will elaborate at length in On Certainty. In particular, he will stress the fact that we have somehow ‘swallowed’ all these propositions by being taught other ones, as the result of having been brought up within a community that has long been having practices and language games of which Moore’s truisms appear to be the ungrounded presuppositions. The third and last element of interest in Moore’s position, in contrast, has to do with the fact that even if we can’t recall how we got to know the propositions which collectively make up our common sense picture of the world, and thus say how we know them, we do know them nonetheless. This is in fact the crucial move in Moore’s strategy, which will reappear time and again in his work – though with significant changes – and which, as we shall see, raises some very interesting conceptual issues (§7). Still, it remains that Moore’s position in DCS is difficult: on the one hand, if the truth of his truisms is evident – whatever that might turn out to mean – it seems to be known immediately, without inference; on the other, he claims that we know such truisms because we must have known some other propositions which entail them. However, either Moore’s truisms are known by inference from other propositions, but then they can’t play the role of foundations for all the rest of our knowledge, or else, they play the latter role, but then their truth cannot be known by inference from other propositions which we would, allegedly, know. As we shall see, the difficulty that Wittgenstein will have
G. E. Moore 25
to overcome is precisely that of embracing the second horn of this dilemma without thereby falling back onto the model of knowledge as a mental state introspectively known to us.
2 ‘Proof of an external world’ Moore’s ‘Proof of an external world’ is almost always inexorably presented without mentioning its original context and, moreover, as if it was directed against scepticism about an external world. The dialectical setting which is most of the times taken for granted comprises two characters: a sceptic about the existence of an external world and Moore himself in his capacity of philosopher of common sense. But things aren’t that simple. For ‘Proof of an external world’ (henceforth PEW) is a long essay divided into two parts. In the first and much longer one, Moore takes his lead from Kant’s famous observation, in the Critique of Pure Reason: It still remains a scandal to philosophy … that the existence of things outside of us … must be accepted merely on faith, and that, if anyone thinks good to doubt their existence, we are unable to counter his doubts by any satisfactory proof.24 Moore claims that Kant hasn’t been able to give a successful proof of the existence of things outside of us and that his own proof will remedy this situation. However, before presenting his proof, Moore introduces a series of terminological distinctions, which are meant to clarify the meaning of the expressions ‘things outside of us’. According to Moore, the philosophical tradition in general, and Kant in particular, erroneously believe that the following expressions are equivalent: (A) (B) (C) (D) (E)
‘things outside of us’; ‘external things’; ‘things which are external to our minds’; ‘things which can be met in space’; ‘things presented in space’.25
According to Kant, all these locutions are synonymous because they make reference to phenomena as opposed to noumena.26 The former are necessarily presented in space, as space is a pure form of sensibility which makes it possible for us to perceive outer things. In contrast, according to Moore, these expressions can’t be equivalent because he doesn’t buy into Kantian transcendentalism, either
26
Moore and Wittgenstein
about empirical objects, or about space. Moore’s discussion is then very convoluted and at times not entirely coherent. But, given our purposes, it is enough to notice the following. First, that according to him, (E) doesn’t entail (D), although (D) entails (E). For there may be things which are presented in space and yet can’t be met within it. A case in point is that of pains and itches, that are presented in a part of one’s body, yet can’t be met in space. Moreover, according to Moore, (C) doesn’t entail (D), although (D) entails (C). For example, animals’ pains are external to our minds, yet can’t be met in space. Finally, with respect to (A) and (B), if they are taken to be equivalent to (C), the point just made would hold in their case too. But they could also be taken to be synonymous with (D). In which case both points just made would hold for them too. Be that as it may, since for Moore ‘physical object’ means ‘an object which exists independently of being perceived by us (human beings)’, he thinks that by giving a proof of the existence of physical objects he will ipso facto prove that there are things which can be met in space and that are external to us, no matter how (A) and (B) are read. Now, this latter claim is obviously problematical, at least prima facie. For Moore himself points out that animals’ pains are external to our human minds as we, human beings, can’t perceive them. But it doesn’t follow from this that they are what we intuitively regard as physical objects. In response to such an objection, one could say that a physical object is everything that we could perceive, which, however, exists independently of the fact that we actually perceive it. With this clarification in hand, let us now turn to the proof itself. By holding up his hands in front of himself and in clear view, Moore makes a gesture with the right hand and says, (1) ‘Here is a hand.’ then, making the same gesture with the left hand, he says, (2) ‘Here is another’; he then concludes, (3) ‘There are at present two human hands.’ Since the conclusion concerns the existence of objects which can be met in space, Moore claims that (3) entails (4) ‘There are physical objects.’ and hence, that he has proved (5) ‘There is an external world.’ It must be stressed that up to this point Moore’s proof is directed against an idealist who denies that there is an external world, as he denies that there
G. E. Moore 27
are objects that exist independently of the fact that we actually perceive them. Furthermore, the proof is clearly based on the idea of presenting two instances of physical objects (but notice that one would do just as well), in order to support the claim that there is an external world. Thus, it proceeds just like a ‘proof’ of, say, there being misprints in a book – that is, by presenting specific instances of the category of misprint. Still, it is clear that an idealist could concede both premises – (1) and (2) – and the conclusion – (3) – and yet deny that that entails that there are physical objects – (4) – if by ‘physical object’ one meant objects existing independently of being perceived by us (human beings). Hence, an idealist wouldn’t take Moore’s performance to show the truth of (5).27 Moreover, it must be noticed that up to this point nothing has been done to show that the premises are known to be true and aren’t merely assumed to be such. Similarly, nothing has yet been done to show that the conclusion of the argument is known. Hence, up to now, Moore’s proof has no bearing whatsoever against scepticism. In fact, some years later, Moore himself maintained, in response to his critics,28 that his proof was directed merely at an idealist and not against a sceptic. For, in his opinion, in order to take issue with a sceptic he should have proved that he knew with certainty its premises. In particular, he should have proved that he wasn’t dreaming. But Moore himself candidly acknowledged that he couldn’t have proved such a thing. For all his evidence would have been compatible with the hypothesis that he might be dreaming of it.29 The interesting question is therefore the following. How come most readers of Moore’s paper have taken his proof as directed at a sceptic?30 Furthermore, given Moore’s explicit pronouncements, is this reading legitimate? In order to answer both these questions, we have to take into consideration the sequel of Moore’s paper where he claims that his proof is a rigorous one because (a) the premises are different from the conclusion; (b) they are known to be true and aren’t merely believed to be true; (c) the conclusion really follows from the premises. For, given (b) and the fact that the inference is valid, it follows that also the conclusion of the argument is known.31 Hence, if it is true that Moore knew that there were two hands, it follows that he also knew that there was an external world and this is clearly an anti-sceptical thesis. This, however, raises the following issue: how could Moore maintain that he knew that his premises were true, from which it follows that he
28
Moore and Wittgenstein
also knew the conclusion of his argument, while holding that he was unable to prove that he knew them,32 and that that was necessary in order to convincingly oppose scepticism? I think the most charitable interpretation of Moore’s claim, which can also explain the interest Moore’s work stirred in Wittgenstein, is as follows: if one is a philosopher of common sense then it doesn’t matter how much a sceptic can press one to give a justification for one’s claims to knowledge. Hence, it doesn’t matter if one doesn’t know how one knows that here there are two human hands, or, more precisely (cf. §7), if one can’t prove that one knows it. For such ignorance is entirely consistent with the fact that one does know such a thing. To put it differently, a philosopher of common sense tends to distinguish between knowledge and the conditions of its obtainment, on the one hand, and what he sees as the challenge raised by scepticism, on the other: namely, the challenge of proving to know what one legitimately takes oneself to know. In support of this interpretation consider what Moore, in effect, writes in PEW: I certainly did at the moment [in which the proof was given] know that which I expressed by the combination of certain gestures with saying the words ‘Here is one hand and here is another’. I knew that there was one hand in the place indicated by combining a certain gesture with my first utterance of ‘here’ and that there was another in the different place indicated by combining a certain gesture with my second utterance of ‘here’. How absurd it would be to suggest that I did not know it, but only believed it, and that perhaps it was not the case!33 I can know things, which I cannot prove; and among things which I certainly did know, even if (as I think) I could not prove them, were the premises of my (…) [proof].34 We shall now look at some of the most influential interpretations and assessments of Moore’s proof that have been presented since its first appearance up to present days. As we shall see, they are all interesting both as attempts to make sense of it and for their conceptual relevance for subsequent and contemporary epistemology. Yet – I will argue – they all somehow fail properly to take the measure of – to my mind – this very interesting and rich-in-implications feature of Moore’s strategy.
3 Malcolm: ‘common sense’ and ‘ordinary language’ In 1942, in a series of talks on G. E. Moore’s philosophy, Norman Malcolm presented ‘Moore and ordinary language’. Seven years later, he
G. E. Moore 29
published ‘Defending common sense’ in which he presented a radical criticism of Moore’s use of the verb ‘to know’ in PEW. In the same year (1949) Wittgenstein was Malcolm’s guest at Ithaca for a few months. As Malcolm writes in his biography of Wittgenstein, the two of them had long conversations on Moore and, in particular, on PEW and DCS. But the discussions that were of most value to me that summer were a series that took place between Wittgenstein and me, our topic being Moore’s ‘Proof of an external world’ and also his ‘Defence of common sense’. In particular, we talked about Moore’s insistence that it is a correct use of language for him to say, when holding one of his hands before him, ‘I know that this is a hand’, or to say, while pointing at a tree a few feet away, ‘I know for certain that this is a tree!’ In a published article [Malcolm 1949] I had maintained that this was a senseless use of ‘know’, and Moore had made a spirited reply to me in a letter [partially published in Malcolm 1977]. Wittgenstein and I discussed these matters in a number of conversations, he making many observations of the first importance about the concept of knowledge.35 Malcolm then gives a summary of those exchanges, from which it emerges how the first hints of what would have become Wittgenstein’s reflections in On Certainty were already present. It thus seems possible to hold that the genealogy of Wittgenstein’s text dates back to these conversations with Malcolm in 1949. Moreover, if that was the case, the debate between Malcolm and Moore, which took place at different times between 1942 and 1949, could also be seen as an indirect exchange between Wittgenstein and Moore, as Malcolm’s and Wittgenstein’s criticisms of Moore’s use of the verb ‘to know’ in relation to his truisms are rather similar. However, in the same biography Malcolm writes: In 1939, G. E. Moore read a paper to the Moral Science Club on an evening when Wittgenstein did not attend. Moore was attempting to prove in his paper that a person can know that he has such and such a sensation, e.g. pain. This was in opposition to the view, originating with Wittgenstein, that the concepts of knowledge and certainty have no application to one’s sensations (see Philosophical Investigations, §246). Wittgenstein subsequently heard about Moore’s paper and reacted like a war-horse. He came to Moore’s home, on the following Tuesday. G. H. von Wright, C. Lewy, Smythies and myself
30
Moore and Wittgenstein
were there, and perhaps one or two others. Moore re-read his paper and Wittgenstein immediately attacked it. He was more excited than I ever knew him to be in a discussion. He was full of fire and spoke rapidly and forcefully.36 I think that this testimony should be taken as the common antecedent to all three philosophers regarding the issue of the use of the verb ‘to know’. Hence, I submit that the most appropriate historical reconstruction is as follows. In 1939 Wittgenstein criticized Moore’s use of ‘to know’ in relation to sensations, on the basis of the ideas he himself was developing on that issue. Malcolm was in attendance and assimilated Wittgenstein’s argumentative line. Later he applied the gist of Wittgenstein’s lesson in the context of a discussion of Moore’s use of ‘to know’ in relation to the truisms of common sense. In 1949, Wittgenstein who was at Malcolm’s in Ithaca, discussed that issue with him and then went over these ideas during the last 18 months of his life, developing his reflections well beyond the criticism of Moore’s use of ‘to know’ made by Malcolm. But let us now look closer at Malcolm’s assessment of Moore’s position by presenting and discussing his two papers in chronological order. 3.1
‘Moore and ordinary language’
In his 1942 paper, which remains the forefather of an authoritative interpretative line of PEW,37 Malcolm aims to show, contrary appearances notwithstanding, how Moore’s ‘Proof of an external world’ is both a perfectly true confutation of scepticism and a good response as such. According to Malcolm the real essence of Moore’s strategy consists in clarifying how sceptical doubts, once made explicit, go against ordinary language. Malcolm aims to show in what sense sceptical doubts go against ordinary language and how an argument which proves that a given thesis goes against ordinary language is a confutation of that thesis. According to Malcolm, when a sceptic says, ‘It cannot be known with certainty that physical objects exist’ or ‘We can’t know with certainty that statements about physical objects are true’, he isn’t expressing an empirical judgement. Rather, these are grammatical statements. For they don’t say that it is a contingent empirical fact that sometimes, when certain statements about material objects are made and prefixed with the verb ‘to know’, what is said is false. Rather, they say that such statements are always false. This is so because any empirical proposition is liable to an infinite number of verifications that can’t (logically) be exhausted. Hence, our knowledge of empirical claims can only
G. E. Moore 31
be probable and not certain. According to Malcolm, a sceptic then proposes a revision of our ordinary language which consists in forbidding any expression of the form ‘I know with certainty that p’, where p is an empirical proposition and in substituting it with an expression such as ‘It is highly probable that p’.38 Moore’s response to scepticism, on Malcolm’s reading of it, is thus as follows. Let us consider a paradigmatic case of sure-fire knowledge, such as that this, that I hold up in front of me in good lighting conditions, while I am cognitively lucid, is my hand. It would be nonsense to say ‘It is highly probable that there is a hand here.’ If a child who were learning the language said such a thing, we would correct him by saying ‘It is certain that there is a hand here and it is not just merely probable.’ According to Malcolm, Moore’s answer to the sceptic appeals to our language sense, reminding him of the fact that there is an ordinary use of the locution ‘to know with certainty’ in which it gets applied to empirical statements. But then if ‘I know with certainty that p’, where p is an empirical statement, is an expression of our ordinary language – that is to say, it has a perfectly accepted usage within our linguistic community – it can’t be maintained, with a sceptic, that it is self-contradictory. For, otherwise, it couldn’t be used to describe any kind of situation and, thus, it couldn’t have the ordinary use it does in fact have, and that Moore reminds us of. The only partial criticism that Malcolm raises against Moore is that it fails to convince the sceptic of his mistake because he doesn’t clarify that his argument is a logico-linguistic one, rather than an epistemicempirical one and also because it fails to make explicit the origin of the sceptical mistake. Malcolm’s understanding of the sceptical position is surely contentious, though. For a sceptic usually takes for granted our ordinary use of ‘to know’ and ‘to know with certainty’. He does so in order to show how it seems to be a necessary condition upon the correct usage of those expressions that one should be able to give reasons in favour of what one claims to know (with certainty). He then maintains that, with respect to beliefs about physical objects, it may be shown – in different ways depending on which sceptical argument is at stake – that one doesn’t really have a justification to believe what one takes to know. Hence, assuming the classical tripartite definition of knowledge as justified true belief, a sceptic concludes that one doesn’t know what one claims to know. As we saw, however, Malcolm bases his interpretation of scepticism on Ayer’s understanding of it. So – it may be argued – his considerations may well be effective against at least that particular kind of sceptic. We will come back to this counter in a moment. What must be stressed for
32
Moore and Wittgenstein
now, however, is, first, that there is no textual evidence that Moore’s sceptic is whom Malcolm (and Ayer) take(s) him to be. Second, that even if it were, Moore’s response, on Malcolm’s interpretation of it, would leave out a great number of sceptical positions and would engage with one of the least interesting ones. For Malcolm’s sceptic’s position depends on the finitude of human cognitive capacities. So it leaves it open in principle that a creature without such cognitive shortcomings, yet exercising the same kind of cognitive faculty, could have sure-fire knowledge of propositions about material objects. However, the best sceptical arguments purport to establish a stronger result: namely, that by exercising the very same cognitive capacities we usually employ – however freed from all defects and limitations ours might have – nobody could get to know with certainty a proposition about physical objects. Just briefly to exemplify, according to Cartesian scepticism nobody would be able to tell, just on the basis of one’s evidence, whether one is dreaming or not. Supposing, for the sake of discussion, that the Cartesian argument were correct in reaching the conclusion it takes to follow from that observation, namely that in those conditions one could not know that one isn’t dreaming, it would then show that nobody who were not endowed with means other than sensory evidence to form beliefs about physical objects could have knowledge of such truths. Similar considerations could be advanced in the case of another, powerful form of scepticism, namely, Humean scepticism (§5). As we will see at length in the following, a Humean sceptic notices that if one tried to prove the existence of an external world in Moore’s way, the argument would be circular. For, in order to have an ordinary experiential warrant for its first premise, one would already need to have warrant for its conclusion. Since no such independent warrant could be gotten either by assembling evidence or through a priori reasoning, at least not by beings like us, a Humean sceptic concludes that such a belief remains unjustified. Hence neither the general belief in the existence of an external world nor any of our specific beliefs about given physical objects could amount to knowledge. So, as this detour should make apparent, by following Ayer, Malcolm is assuming that the sceptical position is much weaker than it in fact is. Moreover, Moore’s proof, on Malcolm’s understanding of it, could hardly be effective against its opponent, even as Malcolm represents him, because it would be based on a petitio principii. For Moore would be trying to counter a sceptic by saying that he does have certain knowledge of an empirical truth and that it would be nonsense to claim that it is only probable. This, however, would mean to assume in the premise precisely what one should prove.
G. E. Moore 33
Moore’s position, however, is a bit subtler than that. As we saw before, in PEW Moore sharply distinguishes between knowledge and the conditions of its obtainment, on the one hand, and the possibility of proving that they are indeed satisfied. He also recognizes that this is what needs to be done in order to confront scepticism. However, Moore also thinks that acknowledging the impossibility of meeting this challenge doesn’t impugn the fact the he knows the premises of his proof and hence its conclusion. So, first of all, Moore proposes an interesting gambit, which, if successful, would considerably diminish the impact of the sceptical argument. We will look at it at much greater length in the following (§7). However, just to anticipate a little, if one can somehow stop the inference from the impossibility of proving that one knows that p, to the fact that one doesn’t know that p, one would have greatly lessened the force of sceptical arguments which make play with such an inference. Second, he also puts forward some considerations to try and meet the sceptical challenge. For he argues that the sceptical – evidently Cartesian – argument which appeals to the hypothesis of dreaming in order to cast doubt on the fact that we may be able to prove that we are not dreaming, and thus on the fact that we have knowledge of some ordinary empirical propositions, isn’t reasonable. For either there are absolutely no reasons to think that we might be dreaming in the circumstance of Moore’s proof;39 or they are indeed weaker than the reasons we have to think that we are not dreaming.40 Hence, admitting such a hypothesis wouldn’t be sensible. Now, I don’t mean to suggest that Moore’s gambit and counter to the sceptic are successful.41 Yet he had at least the merit of devising an interesting move, which, if successful, would greatly weaken the force of the sceptical challenge. On Malcolm’s reading of him, in contrast, he would simply be begging the question and do so in an utterly obvious way. 3.2
‘Defending common sense’
The criticisms that Malcolm raises against Moore in his 1949 paper are totally different. As anticipated, he focuses on the use that Moore makes of the verb ‘to know’ in relation to his truisms. According to Malcolm, the correct use of the expression ‘I know (with certainty) that p’ requires that (i) There be an open question and a doubt to be removed; (ii) That the person who makes the assertion be able to produce reasons in favour of his claim to knowledge; (iii) That it be possible to make an enquiry that could determine whether p is the case.42
34
Moore and Wittgenstein
According to Malcolm, none of these features is respected by Moore’s use of that expression. The first one isn’t because when Moore says ‘I know that there is a hand here’, there is no doubt that it be so. An objection that Malcolm takes into account is that Moore is here responding to a philosophical kind of doubt. The sceptical question is ‘How do you know that there is a hand here?’ and Moore is responding to it. Clearly, however, Moore would merely believe to be answering the sceptical question, for it can only be an ironic response to the question ‘How do you know it?’ to answer ‘Because I know it’. This, however, according to Malcolm, shows that a sceptic raises a doubt where there is no reason to do so. According to Malcolm, then, the second feature of the grammar of ‘to know’ is violated by the fact that Moore claims to know that he has a hand, but he is unable to give reasons in favour of that, or indeed a proof of it. Finally, according to Malcolm, Moore goes against the third feature of the correct use of ‘I know’ for there is no enquiry that could determine that that really is a hand. For touching and observing it would make it plain that one had misunderstood the nature of sceptical doubts, which can’t be silenced by ordinary empirical investigations. Several things are worth noticing, if we wish to assess Malcolm’s paper as an interpretation of Moore’s PEW. First, regarding Malcolm’s contention that Moore’s use of ‘to know’ hasn’t respected (i) – that there be a doubt to be removed – it has to be observed that of course there was no ‘real’ or ‘ordinary’ doubt. Indeed, Moore would very much agree on that. If, then, one considers a sceptical doubt, it should be kept in mind that its removal wasn’t Moore’s aim in PEW. Moreover, he was perfectly aware that he couldn’t have responded to such a kind a doubt – which depends on asking for a proof of the fact that he knew the premises of his proof – just by saying ‘because I know it’. Second, with respect to the allegation (ii) that Moore didn’t give reasons in favour of his claim to knowledge, it should be noticed that he repeatedly said that he knew there were two hands because of his perceptual evidence.43 Hence, an enquiry could have settled the ‘ordinary’ issue of whether that was the case – (iii). Surely, however, that wouldn’t have settled the philosophical issue of giving a proof of the fact that he really knew that there were two hands. But, as we have repeatedly seen, Moore never thought or claimed to have done so. Finally and more generally, one could object, and Moore himself did so, that his use of ‘I know’ was peculiar but not mistaken. As he wrote to Malcolm in 1949: You wanted then [Malcolm 1949], and want now, to say that my use of that expression [‘I know that p’] was a ‘misuse’ & ‘incorrect’; but
G. E. Moore 35
the only reason you give for saying so is that I used it under circumstances under which it would not ordinarily be used, e.g., under the circumstances that there neither was at the moment nor had been just previously any doubt whether it was a tree or not. But that I used it under circumstances under which it would not ordinarily be used is no reason at all for saying that I misused it or used it incorrectly (…). I was using it in the sense in which it is ordinarily used (…). The argument I have just given is an argument designed to show that I was using it in the ordinary sense, though not under any ordinary circumstances.44 According to Moore, the use he makes of ‘I know’ violates the rules of its ordinary employment, yet is still meaningful. This, in turn, clarifies how Moore’s and Malcolm’s conceptions of meaning are utterly different. In Moore’s view, an expression maintains its meaning even if it is used in different circumstances from the ones in which it is typically employed. In Malcolm’s view, in contrast, just like for the later Wittgenstein, the meaning of words is given by the rules which govern the various circumstances of their use. The same difference may be noted by considering Moore’s and Malcolm’s debate over philosophical doubts. Malcolm maintains that a philosophical doubt arises where there is no real doubt about the fact that, for instance, there is a hand here. One may then think that it would be correct to say ‘There is no doubt that there is a hand here’ and hold that that would be equivalent to ‘I know that there is a hand here’ or even to ‘It is certain that there is a hand here.’ From this, it would follow that Moore would be right in saying ‘It is certain that there is a hand here.’ But, according to Malcolm, a sceptic’s mistake isn’t so much that of doubting where one doesn’t usually do so, but, rather, of doubting where a doubt cannot be raised, on pain of nonsense. Hence, in such a context, ‘There is no doubt that there is a hand here’ isn’t equivalent to ‘It is certain that there is a hand here’; rather, it means ‘To doubt that there is a hand here would be nonsense.’ Thus, it is clear how, in Malcolm’s reading of Moore, Moore’s mistake would consist in failing to see how the impossibility to meaningfully raise doubts in such a context goes together with the impossibility of making any claim to knowledge. This, from a logico-linguistic point of view, both for Malcolm and Wittgenstein, boils down to the view that where one cannot say that something isn’t known, one cannot say that it is known either.45 In his letter to Malcolm, Moore discusses also this objection and points out how the crucial difference between him and Malcolm (and
36
Moore and Wittgenstein
hence between Wittgenstein and him) is, as we have already seen, a different conception of meaning and of the legitimacy of philosophical doubts. If so [namely, if Moore’s use of ‘I know’ has the same sense as it has in its usage in normal circumstances, though it isn’t made in those circumstances], it follows that you were wrong not only in saying that mine was a misuse & incorrect, but also in saying that I was using the words in such a way that they ‘did not make sense’. It seems to me that you have been misled into saying this latter partly at least through having failed to notice an ambiguity in our use of ‘senseless’. If a person, under circumstances in which everybody would see quite clearly that a certain object was a tree, were to keep repeatedly pointing at it & saying ‘That’s a tree’ or ‘I know that’s a tree’, we might well say that that was a senseless thing for him to do, & therefore, in a sense, a senseless thing for him to say (…) – meaning (…) that it was a sort of thing which a sensible person wouldn’t do, because, under those circumstances, it could serve no useful purpose to say those words. But this is an entirely different thing from saying that the words in question don’t, on that occasion, ‘make sense’, if by this is meant something which would follow from the proposition that they were not being used in their ordinary sense. It is perfectly possible that a person who uses them senselessly, in the sense that he uses them where no sensible person would use them because, under those circumstances, they serve no useful purpose, should be using them in their normal sense & that what he asserts by so using them should be true. Of course, in my case, I was using them with a purpose – the purpose of disproving a general proposition which many philosophers have made; so that I was not only using them in their usual sense, but also under circumstances where they might possibly serve a useful purpose, though not a purpose for which they would commonly be used.46 For Moore, contrary to Malcolm and Wittgenstein, the philosophical doubt is nonsensical because it goes against the common sense picture of the world. But it is not nonsense because it violates some linguistic norm. Yet he thinks he can oppose it47 by maintaining that the common sense picture of the world is indeed certain, despite a sceptic’s claims to the contrary. Thus, to conclude, this latter paper of Malcolm’s has many things in common with the discussion of the use of ‘to know’ in On Certainty
G. E. Moore 37
which we will consider in Chapter 2. From a historical point of view, I do think it has been the origin of Wittgenstein’s interest in the topic of certainty. From a conceptual point of view, I think there are many similarities between Malcolm and Wittgenstein, such as an identical conception of meaning as use and the observations on the rules which govern the employment of ‘to know’ and of ‘to doubt’. Yet these ideas had been already amply developed by Wittgenstein in the 1930s, as we shall see in Chapter 2, and were fully known to Malcolm. The latter, however, has had the merit of being the first to apply those considerations to the epistemological problem par excellence – that is, the problem of scepticism about the existence of an external world – and to Moore’s response to it. It remains, however, that those accusations often miss the real gist of Moore’s strategy, which, to repeat, mostly consisted in divorcing knowledge and the conditions of its obtainment, from the ability, or even the possibility of proving that they do obtain.
4 Clarke and Stroud: At the origins of contextualism In this section we will consider two influential interpretations of Moore’s proof, provided by Thompson Clarke and Barry Stroud, respectively, in 1972 and 1984. They bear deep similarities to one another; hence I will postpone critiquing them until I have presented both of them. Apart from being interesting in their own right, these two interpretations can be seen as the forefathers of contemporary contextualist strategies in epistemology, which we will take up again in Chapter 2 (§5). 4.2
Clarke’s ‘The legacy of scepticism’
In the paper ‘The legacy of scepticism’ (1972) Thompson Clarke introduced a distinction between ‘plain talk’ and ‘philosophical talk’. In his view, the former is what is produced within all our usual linguistic practices, with their characteristic embedment within non-linguistic activities. The latter, in contrast, is what is produced while doing philosophy. Philosophical talk extrapolates from any ordinary practice and from non-linguistic activities, to consider language in its own right. While in plain talk the conditions of meaningful discourse are subject to pragmatic constraints – such as relevance, and other forms of appropriateness48 – in philosophical talk all these limitations are removed and words are considered as such. Any well-formed sentence of natural language can be subject to philosophical analysis. A philosophical question, like the sceptical question about the foundations of our knowledge, is formulated within philosophical talk and, according to Clarke, it is perfectly legitimate
38
Moore and Wittgenstein
since it satisfies a deep intellectual need that isn’t fulfilled by any of its counterparts in plain talk.49 Still, in Clarke’s view, it should be inquired under which conditions philosophical discourse and questions are meaningful. He proposes the following three: (a) It must be possible to sever the concept at issue and also the entire conceptual scheme from our plain talk. (b) It must be possible to imagine a domain of items, separate from the corresponding concepts, which are their referents. (c) It must be possible to represent oneself a certain way: as a pure observer, detached from concepts and items and capable, mostly by means of one’s senses, to verify that the latter satisfy the conditions determined by the former.50 According to Clarke, the satisfaction of these conditions is necessary and sufficient to have access to questions and, possibly, to answers that are absolutely objective. By contrast, within plain talk, questions and answers have only a relatively objective character. To exemplify this distinction: that there is an external world can be verified by kicking a chair and hurting oneself, or by touching and looking at Moore’s hands. But it cannot be ascertained in the same way that during those very verifications our senses aren’t deceiving us or that we aren’t merely dreaming of kicking a chair and of hurting ourselves, or of touching and seeing Moore’s hands. In the former case, we answer the question ‘Is there an external world?’ from the inside. That is to say, from within those epistemic practices which we usually employ to determine whether there are physical objects in our vicinities. By contrast, in the latter case, we are questioning, from the outside of our ordinary epistemic practices, the legitimacy of the ungrounded assumption that having had a certain experience, it has in fact been produced by the causal interaction with physical objects. Moreover, from the outside, we may wonder how such an issue can be settled and can’t but realize that our sensory experience can be of no avail in this connection. For our senses testify from the inside that there is a chair that we have kicked. But, at the same time, they can be deceived, either by a dream or by the use of a drug. Hence, they can’t be used to prove an absolutely objective claim to knowledge such as ‘There is an external world that exists independently of our perceiving it and that we can know of without help from the usual kind of sensory evidence that commonly certifies its existence.’51 According to Clarke, Malcolm’s interpretation of Moore (as well as Wittgenstein’s) conflates the peculiarity of Moore’s use of ‘to know’
G. E. Moore 39
with its alleged lack of sense. Clarke then points out that if Moore’s use is taken to be part of plain talk, it reveals a ‘philosophical lobotomy’, since Moore means to oppose philosophical theses. If, in contrast, it is taken as part of philosophical talk, then it is dogmatic, since it doesn’t face the sceptical challenge of explaining how he knows that there are two hands where he seems to see them (and consequently, that there is an external world) and simply counters ‘Because I do.’ 4.3 Stroud’s ‘G. E. Moore and scepticism: “internal” and “external”’ As anticipated, Stroud’s interpretation of Moore’s proof closely resembles Clarke’s, which is explicitly mentioned as its inspirational source. However, according to Stroud, it is quite clear that Moore’s proof is given within plain talk. If so, it is a good proof, but, obviously, it can’t have any anti-sceptical bearing, since it doesn’t even face the sceptical challenge. From the inside, it is a good proof because it appeals to the greater degree of certainty possessed by the premises of Moore’s argument over the degree of certainty possessed by those premises which, within plain talk, would be necessary in order to maintain that he may only be dreaming of having two hands. Furthermore, according to Stroud, Moore’s use of the expression ‘I know (with certainty)’ is perfectly legitimate. However, all these considerations do nothing to forbid a sceptic to think that a doubt about Moore’s knowledge of his premises might be possible. Hence, the proof fails to address the philosophical issue and, for this reason, when it is considered from the outside, in the way in which confronting the sceptical challenge requires, it can’t be a success. Hence, the ‘internal’/’external’ dialectic is as follows: it is possible to doubt from the inside only when there are actual reasons to doubt that p, or only when there are stronger reasons to doubt of it than the ones one can produce in favour of one’s claim to know that p. Since, however, with respect to the premises of Moore’s proof there are no such reasons – or at any rate, they aren’t stronger than the ones in favour of holding those premises – any form of scepticism appears, from the inside, totally misguided. From the outside, in contrast, any doubt is legitimate, inasmuch as it is possible or conceivable. Thus, it is perfectly right to doubt the fact that the premises of Moore’s argument be known, because one can raise hypotheses, such as the one from dreaming, that would call into question any supposed instance of sensory-based knowledge regarding physical objects. Therefore Moore, by failing to show why a merely
40
Moore and Wittgenstein
possible doubt – like the Cartesian one – is illegitimate or, in effect, no doubt at all, doesn’t address the sceptic, whose challenge is raised at the purely philosophical level. According to Stroud, Moore’s strategy is wanting because it provides no account of why our knowledge from the inside is legitimate and because it offers no diagnosis and solution (or dissolution) of what, in his view, would be the sceptical mistake. This way, Stroud introduces a sort of ‘evaluative lobotomy’ that, however, is useful to him, because it can help throw light on the relationship between ordinary and philosophical discourse. This, in turn, is, in Stroud’s view, what is really at stake in the debate over scepticism and common sense. From such a perspective, Moore’s proof is interesting because – its failure notwithstanding – it contradicts scepticism while being compatible with it. It contradicts scepticism because it claims that we do have knowledge both of its premises and of its conclusion. Still, it is compatible with scepticism because such a claim is made in a context other than the sceptical one. Similar considerations would, moreover, apply in the other direction too – that is, scepticism contradicts common sense, yet it is compatible with it. Intuitively, however, this claim is dubious. For, if ‘knowledge’ is context-sensitive in the way proposed, there would be no real contradiction between scepticism and common sense, because, in order to have a contradiction, p should be both known and not known in the same context. Be that as it may, in Stroud’s view, Moore’s mistake consists, therefore, in failing to appreciate that if this is the right description of the relationship between ordinary and philosophical discourse, then his negation of the sceptical thesis can’t be a confutation of scepticism. Hence ‘the price of philosophical scepticism’s immunity (…) would be the corresponding immunity of all our ordinary assertions to philosophical attack’.52 Clarke’s and Stroud’s interpretations somehow connect with contemporary contextualist as well as relativist positions in epistemology, although, as we shall see in Chapter 2 (§5), the latter also develop some of Wittgenstein’s remarks in On Certainty as well as some of his reflections in the Philosophical Investigations. For our present purposes, however, it is enough to say that, according to contextualism, there are different contexts, determined by different standards about what must be the case in order for knowledge to be obtained. So, what may be known in a context may turn out not to be so in a different one. This would in fact be what happens in the passage from ordinary contexts to philosophical ones. Moore’s proof could thus be correct in the ordinary context, yet fail as an anti-sceptical weapon. Contemporary contextualism therefore provides only a strategy of damage limitation against scepticism, more than a rebuttal of it. For, as we saw by presenting Stroud, the price to pay
G. E. Moore 41
to philosophical scepticism, in order to make our ordinary knowledge immune to it, is to make it turn right in its own context. A corollary of this view is that, contrary appearances notwithstanding, ‘knowledge’ is a context-sensitive term (or concept), like ordinary indexical terms such as ‘I’ and ‘yesterday’, which pick out different people and days, respectively, according to the context of their utterance. Accordingly, also the extension of ‘knowledge’ varies according to shifts in the context of utterance (or of assessment, on the relativist variant of contextualism). Let us now consider Clarke’s and Stroud’s interpretations from a historical point of view – that is to say, let us now turn to the issue of whether their readings tally with Moore’s own explicit pronouncements. As we saw before, in his letter to Malcolm, Moore clearly states that he is using the verb ‘to know’ according to its ordinary meaning. Furthermore, Moore claims that his use of ‘I know’ in relation to the premises of his proof is meant to engage with his philosophical opponents, contrary to what Stroud maintains.53 Hence, on the one hand, according to Moore, it is with an ordinary conceptual repertoire that one enters philosophical discourse, contrary to what Clarke claims. On the other, in his view, there is no separation between philosophy and common sense, in such a way that statements respectively made in those different contexts could contradict each other yet be somehow compatible with one another, as opposed to what Stroud holds. These two claims together boil down to the view that, according to Moore, the concept of knowledge simply isn’t context-sensitive54 nor is there any hint that he may have favoured a relative view of truth, in such a way that opposite knowledge ascriptions could both turn out to be true, when assessed on the basis of different standards of evaluation.55 Finally, and most importantly, in Moore’s own understanding of scepticism, the latter arises when the issue of showing how one can know what one claims to know, and thus prove that one does indeed know it, is raised. To ask such a question, however, does not produce a change of context according to Moore. Nor does it show the context-sensitivity of knowledge, let alone the relativity of knowledge ascriptions to different standards of evaluation. On the contrary, it depends on making use of an invariant concept and yet asking a different question. Not whether it is known that p, but, rather, how one can prove that p is known (if indeed it is). Moore, moreover, agrees that he isn’t able to answer such a question. Yet, as we have repeatedly seen, he refuses to agree with the sceptic that because one can’t answer that question, it follows that one doesn’t know that p. This, however, isn’t simply a dogmatic and unphilosophical position, as Clarke claims.56 Rather, it depends on a specific conception of the relationship
42
Moore and Wittgenstein
between the conditions of knowledge and their obtainment and the possibility of proving that they are in fact the case.57 Thus, to conclude, Clarke’s and Stroud’s interpretations of Moore’s proof – though interesting in their own right, as well as in light of contemporary debates on semantic contextualism and relativism – are deeply at odds with Moore’s own understanding of his proof. Thus they can’t be taken as faithful interpretations of what Moore was up to.
5 Moore and Humean scepticism: Wright’s interpretation of the proof In a paper titled ‘Facts and certainty’ (1985) Crispin Wright has put forward another reading of Moore’s proof, which is widely debated nowadays, although mostly from a conceptual point of view, rather than as an interpretation of Moore’s actual thought. According to Wright, it is important to make explicit the grounds on which Moore claims to know the premises of his proof in the circumstances in which the proof was offered. As is apparent, in those circumstances, his assertion ‘Here is a hand’ was based on his sensory evidence. Indeed, Moore himself made clear, though not in PEW, that his claim was based on those very grounds.58 That Moore was basing his claim on his available evidence may appear obvious, but it isn’t within the historiography of the proof. For it has been denied that Moore made explicit the criterion on the basis of which he could assert his premises.59 In response to such a claim, it must be noted that Moore certainly said he couldn’t prove that he knew his premises, but, as we have repeatedly seen, he equally denied that that entailed the fact he didn’t know them in those very circumstances – that is to say, when he was holding up his hands in front of him, in clear view, and could thus perfectly well see them. Hence, Moore didn’t mean to deny that he was justified in holding his premises. Rather, he denied that the inability to prove, against a sceptic, that his grounds were sure-fire could impugn his knowledge – on whatever bases it might have been achieved. According to Wright, then, if the grounds on which Moore’s knowledge of his premises is based are made explicit, it becomes immediately evident why the proof is a failure. He reconstructs Moore’s proof as follows: (I)
A given proposition describes the salient aspects of my experience at the time in question; (II) I have a hand; (III) Therefore, there is an external world.
G. E. Moore 43
In fact (I) amounts to saying that that there is a proposition which correctly describes the relevant aspects of Moore’s experience in the circumstances in which his proof was given. For instance, ‘I am perceiving (what I take to be) my hand.’ According to Wright, (II) is then inferred from (I) and (III) follows from (II) since a hand is a physical object. Moreover, given that the premises are known, according to Moore, so would be the conclusion.60 However, Wright points out how the transition from (I) to (II) is problematical: for, if Moore was merely dreaming his hand, (II) would no longer follow from (I). More generally, in Wright’s view, it is clear that (I) can ground (II) only if one can already take it for granted that one’s experience is being produced by causal interaction with physical objects. Hence, any sensory experience can warrant a belief about empirical objects only if it is already taken for granted that there is an external world. Thus, one needs already to have a warrant for (III) in order to justifiably go from (I) to (II). This, however, makes it clear how the proof is epistemically circular (or question begging). For antecedent and independent warrant for (III) is needed in order to have warrant for (II) in the first place,61 given one’s current sensory experience, as described in (I). Ironically enough, then, Moore’s proof – on Wright’s understanding of it – rather than being a response to scepticism instantiates the template of a powerful form of scepticism, that Wright calls ‘Humean’, as opposed to ‘Cartesian’. The difference between these two forms of scepticism resides in the fact that while the latter makes play with uncongenial scenarios, such as the hypothesis that one may be the victim of a sustained and lucid dream, whereby one would be systematically unable to tell whether one is dreaming or not, the former doesn’t. From such a starting point, the Cartesian sceptic then claims that for any specific empirical proposition we take ourselves to know on the basis of our sensory experience, it is metaphysically possible that it be produced in a non-standard way. From the impossibility to exclude that this is the case, he takes it to follow that we don’t know any such empirical proposition. Consequently, that we don’t know (III) – that there is an external world.62 Humean scepticism, in contrast, merely draws on the kind of epistemic gap between having a certain kind of evidence, and warrantedly forming a belief about a domain which goes beyond the one immediately testified by one’s experience, presented by inferences such as the (I)-(II)-(III) argument just offered, or indeed inductive inferences – whence the title of ‘Humean’ for this form of scepticism. To repeat, in order to warrantedly go from the first premise, which is about one’s sensory experience,
44
Moore and Wittgenstein
to the second, which is about an object whose existence is independent of one’s experience, warrant for the conclusion of the argument – that there is an external world – must be independently available.63 Since, however, by sceptical lights, there is no way of getting such an independent warrant, the argument fails to provide warrant for its conclusion. Wright calls this phenomenon ‘failure of transmission of warrant’. For the need of an antecedent warrant for the conclusion in order to warrantedly go from (I) to (II) prevents the warrant one may after all have for (I) – if somehow the sceptic was wrong in claiming that independent warrant for (III) could not be attained – to be transmitted to (II) and thus to (III). That is to say, Moore’s proof cannot either give one a first warrant to believe (III), or further epistemic support for the warrant one might already have to believe it.64 Now, if Wright’s reading of Moore’s proof were correct, it would be devastating. For, regardless of Moore’s insistence that despite being unable to prove that he knew his premises he knew them nonetheless and notwithstanding the fact that – given the principle of closure – one would know the conclusion, he may have indeed had the former piece of knowledge, without in fact being in a position to acquire knowledge of the conclusion, or indeed somehow enhance it, by running his proof.65 So, interpreted this way, his proof would simply be no proof whatever of (III) – namely, that there is an external world. For, characteristically, proofs are means which allow us to extend our knowledge from their premises to their conclusions and thus provide us with reasons for first believing them, or else with reasons which should enhance our epistemic support for believing them. In Wright’s reading of it, in contrast, Moore’s ‘proof’ would dramatically fail to do so. Recently, however, it has been challenged that the conditions for having knowledge of the premises of Moore’s proof are as Wright thinks of them. If this were true, it would certainly come as a relief for the proof’s prospects of success. Hence, we must now turn to a discussion of this rejoinder.
6 Moore’s comeback: Pryor’s dogmatist interpretation Jim Pryor in ‘What’s wrong with Moore’s argument?’ (2004) has recently claimed that Wright’s diagnosis of Moore’s proof’s failure is wrong. In Pryor’s view, the proof is perfectly fine, from an epistemological point of view. In particular, it isn’t true that the proof is epistemically circular or question begging. That is to say, according to Pryor, it exhibits no failure of warrant transmission. For, on his understanding of the structure
G. E. Moore 45
of perceptual warrants, it suffices, in order to possess such a warrant, merely to have a certain course of experience while lacking any reason to doubt that there may be an external world, at least when what is at stake is the justification of what Pryor takes to be perceptually basic beliefs.66 Hence, there is no need, in his view, to possess an independent warrant for the conclusion of that argument, namely, that there is an external world.67 Pryor, however, thinks that although Moore’s proof is perfectly in order from an epistemic point of view, it is not successful against a sceptic. In particular, it is dialectically ineffective for, according to him, a sceptic will think it is (likely) false68 that there is an external world. For such a reason he will not consider Moore’s experience of a hand in front of him as sufficient for warrant for the corresponding empirical belief. Starting off with an unwarranted premise – at least by a sceptic’s lights – the proof won’t be able to confer warrant upon its conclusion and hence to convince a sceptic that there is an external world. Pryor, however, doesn’t think that the sceptical doubt is legitimate. He actually says that it is a ‘disease’ we should not catch, or else should cure ourselves of.69 Yet this judgement is slightly at odds with the claim that Moore’s proof would fail for dialectical reasons when propounded against a sceptic. For reflect: if Pryor is right in thinking that the proof is epistemically correct, by being presented with it one should – if rational – give up one’s disbelief in the existence of an external world. Hence, in Pryor’s account of the proof, it could be dialectically ineffective only against a – as it were – stubborn kind of sceptic.70 That is to say, it would be ineffective only against a sceptic who resolutely – and by Pryor’s lights, irrationally – denied that just by having a certain course of experience one would thereby get a warrant for a belief about a specific material object like ‘Here is a hand’, and could thus discard hypotheses that are incompatible with it – for example, the dreaming hypothesis – while also acquiring a warrant for the conclusion of the argument, that is, ‘There is an external world.’ Although such an outcome would be very sympathetic to Moore’s proof, a few things must be noticed, if Pryor’s story were meant to be a historically faithful rendition of the proof. First, that Moore never explicitly argued for the view about perceptual warrants that Pryor sees as the key to the proof’s epistemic success. No doubt he made explicit the grounds of his proof – that is, the fact that he believed its premises on the basis of his perceptual evidence. But he never said that that would be sufficient, by itself, to give one a warrant – or indeed knowledge – of certain propositions about physical objects such as ‘Here is a hand.’
46
Moore and Wittgenstein
On the contrary, we saw the main tenets of Moore’s philosophy of perception (§1.1) and his appeal to sense data, whose nature, however, was somehow left indeterminate. Now, it merits emphasis that any account of perception in terms of sense data would entail a conception of perceptual warrant (and therefore knowledge) as dependent on some extra element beside the occurrence of the sense datum itself – at least unless the latter were taken to be identical with some part or other of a physical object. Such an extra element would presumably be the (warranted) assumption that the experience one is having be produced by causal interaction with physical objects. Indeed in ‘Four forms of scepticism’ Moore explicitly said that he agreed with Russell that propositions about specific material objects in one’s surroundings aren’t known immediately, but always on the basis of some ‘analogical or inductive argument’.71 Thus, it is almost certain that he didn’t endorse in PEW72 the conception of the structure of empirical warrants (or of knowledge) that Pryor has recently put forward and which has served as basis for his interpretation of Moore’s proof. Moreover, by saying in PEW that in order to counter scepticism he should have been able to prove he wasn’t dreaming of having a hand, Moore seems to have come close to saying that simply to have evidence – that is an experience with the representational and phenomenal content of a hand – wouldn’t by itself be enough to give one a warrant for the corresponding empirical belief, capable of silencing the sceptic’s quest of a justification for believing that there was indeed a hand where Moore seemed to see it. This would suggest that Moore himself thought that in order to have a perceptual warrant, one would already need to know that one’s experience had not been produced in a dream, but, rather, by causal interaction with physical objects, contrary to what Pryor would make him hold. Yet I myself believe this latter interpretation would be contentious. For it seems to accord better with Moore’s overall strategy to think that he simply raised the issue of having to prove that he wasn’t dreaming because he realized that that was what stood in the way of his opponent’s recognition that the premises of his proof were indeed known; rather than to think that he did raise that issue because he thought that such a proof was necessary in order to really have knowledge of the premises of his proof. If so, however, and this brings us to the second point worth noticing, Pryor’s and Moore’s understanding of what would be needed to confront a sceptic would dramatically diverge. For, in Pryor’s account it would suffice simply to remind him that in order to possess a perceptual warrant for ‘Here is a hand’ it is enough to have a certain course of experience when there is in fact no reason to doubt that there is an
G. E. Moore 47
external world and that the conditions are such that no doubt of that kind is reasonable. In Moore’s view, in contrast, it should be proved that, in those circumstances, such a doubt would be unreasonable. Indeed Moore thought he could make some gesture in that direction, and accordingly said that the grounds for thinking that he might have been dreaming of his hand were weaker than the ones available to him to claim knowledge of his premises, or even totally non-existent.73 Now, this move is unsuccessful, especially if coupled with other things Moore does say. For, if it is legitimate to distinguish between knowledge and the conditions of its obtainment and the ability to prove that they are satisfied, as Moore claims, and if, as he argues, the sceptical challenge concerns such a proof, then merely to insist that one does know that there are hands and that, given the evidence, the hypothesis that one may be dreaming is less supported, or even unsupported by evidence, doesn’t exclude the metaphysical possibility that it be the case, after all. Hence, Moore didn’t succeed in proving that he wasn’t dreaming, contrary to what he thought should have been done to meet the sceptical challenge. Accordingly, he also failed to prove that he knew the premises of his proof, once more contrary to what he thought should have been done in order to counter a sceptic. Now, it remains to be assessed whether such a failure could be compatible with his insistence that he did after all know both the premises of his proof and that he wasn’t dreaming. Still, it is clear that he saw and characterized the sceptical position not as a ‘disease’, but as a genuine challenge, that arises out of asking a kind of question, which can’t be answered by simply exposing the structure of empirical warrants. It is to such a question and to its legitimacy that we shall now turn.
7 Having knowledge and being able to prove that one does Before closing this chapter, we need to consider the issue of the relationship between having knowledge and the conditions of its obtainment, on the one hand, and the possibility of proving that those conditions do in fact obtain and thus that one really has that knowledge one takes oneself to have, in such a way as to be able rationally to claim or redeem it. The ‘How do you know?’ kind of question that the sceptic typically asks is meant to raise that issue. Hence, as we have repeatedly seen, that question, when voiced by a sceptic, isn’t a simple request of what the grounds of one’s knowledge are. Rather, it is meant as a request for something stronger: namely, of a proof of what one claims to know. To exemplify once more
48
Moore and Wittgenstein
with the premises of Moore’s proof, Moore can say that his ground for holding that there is a hand in front of him is his current visual experience. A sceptic, however, is precisely questioning the fact that Moore be able to prove that his experience is indeed veridical in those circumstances. Now, according to Moore, the sceptic holds the following view of the relationship between having knowledge and being able rationally to redeem it: (1) If you can’t show how you know that p – that is, prove that you do know it – and can’t, therefore, rationally redeem your claim to knowledge, you don’t know that p. Moore, however, thinks that this entailment doesn’t hold. Why? One possible explanation is to say that he was in fact endorsing a somewhat externalist conception of knowledge. The caveat is apposite because he never proposed anything which would suggest his leaning towards one of the views that, later on, would have been qualified as externalist – let it be reliabilism, relevant alternative theories, counterfactual analyses, etc. All he seems to have done is to introduce a certain kind of move, which would then become the typical externalist manoeuvre: namely, the move according to which one can know that p even if one is unable to prove that one does. In other words, Moore denied (1). The evidence in Moore’s writings that he would deny (1) is plenty.74 For instance, in PEW, he quite explicitly said that neither him nor anyone else may have been able to prove the truth of his premises: [If what is required is a general proof for the existence of physical objects], (t)his, of course, I haven’t given; and I do not believe it can be given: if that is what is meant by proof of the existence of external things, I do not believe that any proof of the existence of external things is possible.75 Yet he was adamant that such an impossibility wouldn’t have impaired the fact that he did know them. To repeat the relevant quotation: I can know things which I cannot prove; and among the things which I certainly did know, even if (as I think) I could not prove them, were the premises of my (…) [proof].76 In the recent literature on scepticism it is often remarked that if one makes this kind of move, one will then face some ‘crisis of intellectual
G. E. Moore 49
conscience’ or ‘angst’.77 That is to say, it is conceded that one may know that p, even if one is unable to prove it. Yet one wouldn’t be able to reassure oneself that one does. Scepticism, in this reading, wouldn’t directly challenge one’s knowledge but, rather, one’s ability rationally to redeem it. Hence, scepticism wouldn’t be characterized by (1), but by (2), which says: (2) If you can’t show how you know that p – that is prove that you know it – and can’t therefore, rationally redeem your claim to knowledge, you may know p, but you can’t reassure yourself of it. Thus, scepticism brings about an ‘intellectual crisis’, at least prima facie, because, even supposing that through a gift of nature, as it were, we knew that p, we would also feel the intellectual need to be able to prove that we really do. The problem then becomes that of explaining the sources and legitimacy of this intellectual need. The issue is subtle, yet, I believe, of the utmost epistemological and, as we shall see, meta-epistemological significance. So I will proceed by considering four possible reactions to (2), each of which will depend on a specific understanding of the notion of knowledge. I will then connect each of them to Moore to see whether any one of them would allow us to give a coherent interpretation of what he was up to. So, first, one may say that we have just an externalist notion of knowledge – which makes us hold that we know that p even if we can’t say how – and then a sort of self-reflective spontaneous attitude which forces us to look for an explanation of how that knowledge may have come about. Scepticism would thus merely be due to such a deeply rooted human attitude. Yet it would have no bearing whatever on whether we possess knowledge. Stories may be told or invented to explain why, adaptively, it would be useful for us to have such an attitude. For instance, it forces us to make inquiries, which sometimes show that we don’t really know what we thought of knowing. Hence, that attitude prevents us from indulging in mistakes. Moreover, such inquiries into the sources of our knowledge may contribute to increase it, precisely by providing an explanation of how it comes about. A case in point, which exemplifies both aspects, is that of chicken sex tellers. For, by inquiring into how they come to have the ability to know the sex of chicken, it was discovered that it is based on the operations of the sense of smell and, thereby, the incorrect belief that it depended on sight and touch was removed. One may then say that the price to pay, its advantages
50
Moore and Wittgenstein
notwithstanding, is that at times such a self-reflective attitude gets in the way of our placid recognition of our real epistemic status and makes us worry when there is no need to. The circumstances in which Moore claimed to know that there was a hand where he seemed to see it would just be an example of this downside of our self-reflective attitude. For it is as clear as ever that he knew that there was a hand in what appeared to be cognitively optimal conditions. But, of course, if our self-reflective attitude kicks in and demands that we provide a proof of the fact that those are indeed optimal conditions, how on earth could we accomplish such a task? We could only set into operation the same cognitive faculties in the same kind of optimal external conditions for which the problem of how we could prove them to be indeed optimal was raised. Hence, we would be caught up in a circle which would prevent us from being able to prove that we know what we take ourselves to know. Yet an externalist should simply insist that despite our worries, we do know what we take ourselves to know and that that is all we need to care about. Furthermore, he should add the recommendation that we had better tame our self-reflective attitude whenever its setting into motion would get in our way by raising challenges that can’t, in principle, be met. This kind of explanation would also have a metaepistemological consequence: the traditional epistemological project that takes seriously the sceptical challenge of proving that one knows what one takes oneself to know is rooted in a kind of attitude that, though natural, isn’t always legitimate. Hence, we should simply give up that project.78 Second, one may maintain that we have just an internalist notion of knowledge, according to which knowledge is intrinsically such that, in order to have it, we should also be in a position to be able to prove that we do have it. On such an understanding, those cases which are usually listed as examples of the fact that we may divorce knowledge from being able to prove that we have it should be redescribed in such a way as to make it apparent how proof is needed in general and in principle, though one specific subject or even all subjects at one particular moment in history are unable to provide it. So, for instance, a subject may know that the sum of the internal angles of a triangle is 180° even if he can’t himself prove it, only by courtesy of the fact that someone else in the community can do it. Similarly, the chicken sex-teller would know the sex of chicken because there is a method, whose characteristics have been ascertained by experts only recently, that explains how that is possible.79 The fact, then, that in the absence of such a proof – in principle and not just as a matter of contingency – one could express
G. E. Moore 51
oneself by saying ‘I do know that p’ is simply taken as an expression of the kind of psychological conviction one has with respect to p’s truth, and not as an expression of the fact that one bears a genuinely epistemic relation to p. Thus, in the case of Moore’s argument, the impossibility to prove that one’s experience is reliable, or of giving an independent proof of the fact that there is an external world, impugns the fact that one really knows the former and the latter respectively. On such an interpretation, scepticism raises a much more serious problem. For it would really show that if one can’t answer the ‘How do you know?’ question and thus justify one’s claim to knowledge, one simply doesn’t know that p in the first place. Thus, on an internalist account of knowledge, the sceptical challenge is in fact correctly captured by (1) and not by (2). Therefore, Moore’s proof would be a case where although we have all the psychological confidence in the world that we know the premises to be true, epistemically no such optimism is justified. The crisis of intellectual conscience would then depend on realizing – perhaps only implicitly – that one is equivocating between actually having knowledge and merely having the psychological conviction that what one takes oneself to know is in fact the case. Therefore, from a meta-epistemological point of view, scepticism would be successful at showing that at least when it comes to certain very fundamental assumptions in our conceptual scheme, we don’t know them as we can’t prove that we do. Yet that teaches us something important (that we will explore at length in Chapter 4): namely, that our relation to these fundamental assumptions isn’t epistemic at all. That is to say, once properly understood, the sceptical challenge shows that we can’t prove how we know that, for instance, there is an external world, that our sense organs are mostly working reliably and that we aren’t currently dreaming. That, in turn, entails that we don’t know any of these things, given the internalist conception of knowledge. For a sceptic that would have disastrous consequences, as we would end up not knowing all sorts of empirical propositions we ordinarily take ourselves to know. As we shall see, the kind of epistemological lesson that can be elicited from Wittgenstein’s reflections in On Certainty, which proceeds from taking scepticism quite seriously, will capitalize on the sceptical point – namely, that our relationship with respect to certain fundamental assumptions isn’t epistemic – while putting forward considerations which would block the unwanted consequences of that view: namely, that we can’t know any of the ordinary empirical propositions we usually take ourselves to know.
52
Moore and Wittgenstein
As we have seen, however, Moore didn’t think at all of having conflated knowledge with psychological conviction, as an internalist would charge him of; or, for that matter, that the challenge posed by scepticism were somehow illegitimate, just as a full-blown externalist would hold. One might then interpret such a predicament as an expression of one’s commitment to the view that there actually are two strands or aspects to our unique concept of knowledge:80 one externalist – which makes us hold that we know that p, even if we can’t say how – and one internalist – that pushes us towards looking for such an explanation. For, otherwise, we would feel that our knowledge isn’t secure and that it may just be a pretence. On such an understanding, scepticism would be the expression of this second strand to our notion of knowledge, whose real significance would be to show that these two aspects are in conflict. As soon as we assume to have knowledge acquired by means of what we take to be our best epistemic methods, we realize that if we can’t prove to ourselves that we are indeed in such favourable circumstances – and, as we saw before, in the relevant cases we can’t do it – we are forced to acknowledge that we don’t really have it. An important symptom of this twofold nature of our concept of knowledge would be the fact that these conflicts would always be irresoluble, for each party will be rooted in an aspect of our conceptual repertoire we can’t really get rid of. Equivalently, it may be held that if this were the nature of knowledge then conflict will always have to be resolved by legislating in favour of one of these strands, perhaps by adducing pragmatic considerations to support one’s choice.81 So, for instance, if our aim is to engage in everyday activities, we should go by the externalist strand and trust that we know that there is a hand where we seem to see it, even if we can’t prove it (and we may even draw out the consequences of such a piece of knowledge, for example, that we know that there is an external world). If, in contrast, we are doing epistemology, we may go by the internalist strand and end up denying that we know that there is a hand where we seem to see it or that there is an external world.82 On such an understanding, however, the deep significance of scepticism – which is seldom appreciated – would be to show that our concept of knowledge is incoherent, for it would lead, via the acceptance and the denial of (1), to holding, for a suitable p, that one both knows and doesn’t know it. Meta-epistemologically, that would entail the rather shocking conclusion that there is no sensible epistemological project to be pursued in connection with knowledge and the conditions of its obtainment, just as there isn’t any sensible epistemological project to be pursued with respect to square circles. For, if the concept
G. E. Moore 53
is contradictory, it can’t have instances and there would be no point in trying to establish whether one has knowledge with respect to any proposition p whatever.83 Be that as it may, it is quite clear that Moore didn’t think either that our concept of knowledge is incoherent, or that there is no sensible epistemological project to be pursued in its connection. Finally, one may suggest that rather than there being two strands to one single concept of knowledge, knowledge is a genus, as it were, that can come in two species. We can define ‘knowledge’ minimally at least as a propositional attitude, which is factive (one might add other features, if needed).84 One can then say that such a single genus manifests itself in two different species. On the one hand, there is knowledgee that consists, for instance, in having a true, reliably formed belief which is safe.85 On the other, there is knowledgei, according to which, in order to have knowledgei of p, one must be able to prove that one knows that p. So, to exemplify with the premises of Moore’s proof: one can knowe that there is a hand and, if closure holds, knowe what is entailed by that, for example, that there is an external world or that one’s sense organs are working reliably. However, one would not knowi that there is a hand, because one wouldn’t be able to knowi that there is an external world or that one’s sense organs are working reliably, as any ‘proof’ one might give of that would already depend on taking for granted that one’s sense organs are working reliably and that one is causally interacting with a world populated by physical objects. If we take this account seriously then there wouldn’t be any issue of ‘intellectual integrity’, as it were. For one may have one ‘epistemic good’, so to speak – namely, knowledgee – but not some other ‘epistemic good’ – namely, knowledgei. If these are really two different ‘epistemic goods’, one may feel a little disappointed at realizing that one has the one but not the other; or even at realizing that one has a different ‘epistemic good’ than the one one thought of having. But there would be no issue of ‘intellectual crisis’, or ‘angst’. For these reactions would be appropriate just in case one hadn’t lived up to one’s intellectual duties. However, if it is true that it can’t be proved that there is knowledgei (for a suitable p) one can’t blame oneself for not obtaining such an ‘epistemic good’. Yet, in turn, if that is our epistemic situation, the traditional epistemological project in which scepticism seems to be a challenge, would be legitimate only if knowledgei were obtainable at least in principle. But it isn’t. Hence there is no legitimate epistemological project in that connection. While surely the account we are considering would have the merit of making sense of some of the things Moore says, when he claims, for
54
Moore and Wittgenstein
instance, to know what he is unable to prove, there are several problems with it, as a possible interpretation of his overall view. For he never suggested that scepticism and the project of showing how we may have ‘knowledgei’ are illegitimate, while they would be on the present account of knowledge. Furthermore, there is no hint in his writings that he thought that knowledge is a single genus which comes in two species. On the contrary, as we saw in our discussion of Clarke’s and Stroud’s readings of him, Moore thought that we have one single and unified notion of knowledge and that, on the basis of that single and unified concept, he could try to object to scepticism. Now, it may well be that there is no single coherent reading of what Moore was up to. That is to say, a reading that could reconcile all the things he said and the strategies he actually employed in his writings. My hunch is that he somehow anticipated, in many respects, the kind of approach that epistemic externalists have developed after him. It remains, however, that most of his contemporaries, and in particular Wittgenstein, were firmly rooted in an internalist conception of knowledge and, accordingly, thought that Moore was altogether missing the point. Equally, it is true that Moore did not develop that ‘externalist’ strategy and, in particular, did not offer a diagnosis of why the sceptical challenge would be illegitimate by the lights of an externalist epistemologist. These two things together may well explain both why he was criticized by his contemporaries and why, also nowadays, quite independently of one’s epistemological preferences, which may even go in that very same direction, one may find Moore’s strategy somewhat unconvincing.86 Yet it is obvious that he had the great merit of individuating a series of propositions, in PEW and especially in DCS, that go well beyond the ones traditionally investigated in epistemology, for which it is a genuine challenge to understand whether we bear an epistemic relation to them. Moreover, he had the merit of expressing, perhaps inappropriately, the commonsensical intuition that no matter how unprovable these propositions turn out to be, we would never give them up.
2 Wittgenstein: Belief, Knowledge and Certainty
In this chapter we will begin to explore Wittgenstein’s position with respect to scepticism and common sense. We will start by considering his account of the use of the verb ‘to know’, as well as of other doxastic expressions, such as ‘to believe’ and ‘to be certain’. Many of the entries in On Certainty (henceforth, OC) are outright criticisms of Moore’s position. Yet, the relationship between the two is more complex than it first appears. For instance, we know that Wittgenstein greatly admired ‘A defence of common sense’ (DCS), even if he had some reservations about it, while he was very critical of ‘Proof of an external world’. The reasons behind such an assessment had mainly to do with the fact that Wittgenstein thought that, in the first essay, Moore had spotted something extremely interesting, from a philosophical point of view. Namely, that there are a number of everyday propositions, mostly about physical objects, which are certain, while having nothing to do with the propositions traditionally considered beyond the possibility of doubt in philosophy, viz. logical necessities and self-verifying propositions. By contrast, according to Wittgenstein, in his later work Moore had embarked on an utterly misleading philosophical project, viz. the project of giving a proof of a philosophical proposition like ‘There is an external world’. It is in fact a trait of all of Wittgenstein’s philosophy – or, better, metaphilosophy – that he thought that philosophical theses were nonsense and hence unprovable for principled reasons. As noted, however, despite his interest in DCS, Wittgenstein was also critical of it. In particular, he strongly criticized Moore’s passage from the appeal to common sense to the defence of what he took to be the metaphysics and the epistemology contained in it, without going through a diagnosis of the idealist and the sceptical mistakes. For, according to Wittgenstein, it is legitimate to appeal to common sense in order 55
56
Moore and Wittgenstein
to somehow resist idealists and sceptics. Still, it can’t be maintained that simply to remind them that we use propositions about material objects, as well as about our own and other minds; or else, about space and time and that we say that we know them, is enough to prove that their theses are false. On the contrary, one will have to provide a diagnosis of their error, and to avoid the mistake, parallel to the one made by idealists and sceptics, of thinking that the appeal to common sense may show fundamental metaphysical and epistemological truths. Here I will concentrate on Wittgenstein’s remarks on the use of ‘to know’ and the expression ‘I know’, trying to clarify the two-fold aim of his observations. For if, on the one hand, there is a pars destruens whose end is to criticize a philosophical conception of knowledge which, according to Wittgenstein, is common to many philosophical positions and is instantiated also by Moore’s, on the other, it is clearly present a pars construens. The latter doesn’t take the form of a theory of knowledge – an effect that wouldn’t have been in keeping with Wittgenstein’s overall conception of the aim of philosophy – but, rather, of a perspicuous description of our language game with ‘to know’ and with the expression ‘I know’ and of a clarification of the grammatical role that these expressions can play with respect to Moore’s truisms. This latter aspect, as we shall see at much greater length in the last chapter, will be put at the service of a redefinition of the relationship between subjects and these truisms in non-epistemic terms. Finally, Wittgenstein’s remarks against Moore’s use of the expression ‘I know’ are clearly influenced by Malcolm, as they share many of the latter’s misgivings. Yet, three things are worth noticing. First, as we saw in the Chapter 1, Malcolm’s criticisms of Moore’s use of ‘to know’ in the first person present are greatly indebted to Wittgenstein’s own views about the correct description of the employment of that expression, already developed in relation to psychological avowals, which we will rehearse in the course of this chapter. Second, Wittgenstein’s own treatment of the issue has a deeper philosophical twist. For, as we shall presently see, he wants to attack an entire philosophical conception of knowledge. Such a wide perspective and deeper aim are entirely missing in Malcom’s work. Finally, as On Certainty proceeds, Wittgenstein tends to move on from this kind of critique and to offer a different and more complex diagnosis of Moore’s mistake. In particular, while Malcolm’s analysis mainly stresses the anomalous use of ‘I know’ made by Moore in relation to his truisms, but does not put forward anything in the way of a positive proposal, Wittgenstein’s analysis, in contrast, develops a new conception of the role which Moore’s truisms and other
Belief, Knowledge and Certainty 57
propositions similar to them actually play in our lives.1 This, in turn, will be Wittgenstein’s entirely original contribution to an understanding of their nature and role, and therefore to epistemology, whose proper treatment will have to be deferred until the last chapter.
1 The philosophical use of ‘to know’. On the misleading assimilation of ‘to know’ and ‘to believe’ In OC 90 Wittgenstein writes, ‘I know’ has a primitive meaning similar to and related to ‘I see’ (‘wissen’, ‘videre’). And ‘I knew he was in the room, but he wasn’t in the room’ is like ‘I saw him in the room, but he wasn’t there’. ‘I know’ is supposed to express a relation, not between me and the sense of a proposition (like ‘I believe’) but between me and a fact. So that the fact is taken into my consciousness. (Here is the reason why one wants to say that nothing that goes on in the outer world is really known, but only what happens in the domain of what are called sense-data.) This would give us a picture of knowing [my emphasis] as the perception of an outer event through visual rays which project it as it is into the eye and the consciousness. Only the question at once arises whether one can be certain of this projection. And this picture does indeed show how our imagination presents knowledge, but not what lies at the bottom of this presentation. The model of vision at the basis of this picture of knowledge entails that just as one can’t fail to see what is present in one’s visual field, similarly one can’t fail to know what is given to one’s conscience. However, on this model, the external world isn’t the primary object of knowledge; rather, sense data are. Hence, when I say ‘I see red’ – where ‘red’ is the name of the sense datum – I can’t be mistaken (bar forms of linguistic impropriety or of inattention). Of course I can be mistaken about the colour of the physical object responsible for my sense datum, but not about the latter. If, then, knowledge is modelled after the paradigm of vision, it will have to preserve such a certainty and infallibility. Hence, on the one hand, the content of one’s knowledge must be transparent to the subject and infallibly known to him and, on the other, one’s awareness of knowing such a content will have to be immediate. Wittgenstein notices that this is only one among the various representations of knowledge we have, but that it doesn’t say what grounds it. We can say that such an image has somehow seduced us and keeps
58
Moore and Wittgenstein
us prisoner, but, at careful analysis, it fails to account of our ‘highly specialized’ use of that concept (OC 11). First, it doesn’t account for our use of ‘to know’ with respect to all empirical propositions we say we know. For if, strictly speaking, we know only what is given to our conscience, but nothing can be said to be known with respect to the objects or state of affairs, which should stand in a causal relation with the sense datum, then our claims to knowledge with respect to empirical objects and states of affairs will all be inappropriate. Second, such a representation of knowledge can’t account for our expression ‘I thought I knew’ (OC 12). For if knowledge is a mental state, which is infallibly known to the subject who has it, then there can’t be mistakes in recognizing it. Hence, I can’t be mistaken about the fact that I know that p. Moreover, if such a mental state guarantees that what is known is the case, then p can’t be false. According to Wittgenstein, this means to assimilate ‘I know’ to the avowal [Äußerung] ‘I’m in pain’. For in the latter case, I can’t fail to know that I am in pain and my sincere avowal suffices to guarantee that I am in pain (OC 178). However, for Wittgenstein, the mental state of the subject who claims to know that p is in is neither a necessary nor a sufficient condition for knowledge. It is not sufficient because I could claim to know that p and yet not know it. It is not necessary either for one could know that p and still not claim it. Nor is the mental state one is in when one claims to know that p is a necessary and sufficient condition to guarantee p’s truth. For, clearly, p can be true without one’s claiming to know it and the mere fact of claiming to know that p is true isn’t enough to guarantee that it is. Hence, for Wittgenstein, both the assertion ‘p’ and ‘I know that p’ need to conform to publicly available criteria and can’t be legitimated simply by the mental state the subject is in when he makes such claims (OC 21). The use made by Moore of ‘to know’ and ‘I know’ is, according to Wittgenstein, a particular instance of the image of knowledge described so far, because Moore’s truisms have (mostly, though not invariably, as we saw in the Chapter 1) the form of empirical propositions on specific physical objects and thus differ from propositions about sense data. However, as we saw in Chapter 1, Moore claims both that their truth is evident and that he couldn’t doubt to know them with certainty. On Wittgenstein’s view, Moore therefore conflates the psychological impossibility of doubting them with their logical – that is to say, objective – certainty and treats both doubting and knowing with certainty as two mutually exclusive mental states. That is to say, from the fact that he can’t doubt his truisms,
Belief, Knowledge and Certainty 59
Moore infers that he knows them with certainty. Hence, from realizing that he isn’t – in fact that he finds it psychologically impossible to be – in the mental state of doubting them, Moore concludes that he thereby knows them with certainty. By so doing, he conceives of knowledge along the lines of the misleading image that, in Wittgenstein’s view, is so captivating. For one’s own knowledge will then be a mental state, which is introspectively available to one. According to Wittgenstein, an important indication of the fact that Moore conceives of knowledge that way is the fact that he enumerates what he knows ‘straight off’ (OC 6). For he seems to be assenting to each of his truisms, thereby assuming them within his conscience, then exclude the fact that it would be possible for him to doubt them and finally conclude that he therefore knows each of them with certainty. For Wittgenstein, as we shall see, not only is this model of knowledge problematical, but, in particular, the certainty of the propositions enumerated by Moore depends on the role that they collectively – as opposed to one by one – play with respect to our language games. Furthermore, Wittgenstein claims that When one hears Moore say ‘I know that that’s a tree’, one suddenly understands those who think that that has by no means been settled. The matter strikes one all at once as being unclear and blurred. It is as if Moore had put in it the wrong light. (OC 481) Moore’s view really comes down to this: the concept ‘know’ is analogous to the concepts ‘believe’, ‘surmise’, ‘doubt’, ‘be convinced’ in that the statement ‘I know…’ can’t be a mistake. And if that is so, then there can be an inference from such an utterance to the truth of an assertion. (OC 21) The wrong use made by Moore of the proposition ‘I know…’ lies in his regarding it as an utterance as little subject to doubt as ‘I am in pain’. And since from ‘I know it is so’ there follows ‘It is so’, then the latter can’t be doubted either. (OC 178)2 The ‘wrong light’ shed by Moore on the propositions he claims to know is the one that gets thrown onto them by a conception of knowledge
60
Moore and Wittgenstein
as a mental state accessible to the subject through introspection. Such a state would be infallibly recognized by the subject. Moreover, given the factivity of knowledge, from the sincere and authoritative claim ‘I know (with certainty) that here is one hand’, it follows that there is. Hence, ‘Here is one hand’ would owe its certainty – its lying beyond doubt – to the mental state of a subject who claims to know it, rather than to its role within our language games and epistemic practices. Thus, according to Wittgenstein, Moore unduly assimilates the concepts of knowledge and belief because a subject’s sincere avowal that he believes that p cannot be doubted, nor does it make sense to ask him ‘How do you know that you believe thus-and-so?’, for the only possible answer would be a repetition of his avowal. Obviously the reasons of his belief may be questioned, but this would merely show that he doesn’t have good reasons to believe what he does, not that he doesn’t believe it. In contrast, a subject’s sincere avowal that he knows that p can both be questioned and probed by asking ‘How do you know it?’, where the repetition of one’s claim would not be regarded as a good answer to that question. But if a subject is unable to offer cogent grounds in favour of his assertion, that shows that he doesn’t know what he claims to know (but that he merely believes it, for instance). Hence, while belief is subjective – that is to say, it doesn’t need to satisfy public criteria of justification – knowledge is objective: the fact that a subject knows something must be decidable on the basis of intersubjective criteria.3 We can then see that, according to Wittgenstein, the philosophical use of ‘to know’, of which Moore’s is but one instance, puts the expression ‘I know that p’ beyond the possibility of being mistaken. This would be indeed a remarkable, or super fact, not an ordinary, empirical one. It would actually be an odd metaphysical fact. Now, Wittgenstein writes that he wants to bring words back from their metaphysical (philosophical) use to their ordinary employment and, in particular, that he would like ‘to reserve the expression ‘I know’ for the cases in which it is used in normal linguistic exchange’ (OC 260).4 We shall now therefore turn to the perspicuous description of our habitual use of the expression ‘I know’.5
2 The language game with ‘to know’ and ‘I know’. A perspicuous description First of all, let me put forward some general considerations on the significance of the description of a language game in Wittgenstein’s later works. The observation of how words are used in ordinary linguistic
Belief, Knowledge and Certainty 61
exchanges has, for Wittgenstein, a therapeutic effect with respect to the use and abuse made of our words while doing philosophy, or, to borrow an expression of Stroud’s, in ‘philosophical discourse’. The description of the main features of the language game is meant to make explicit the criteria of application of concepts which occur in them, and thereby show their constitutive conditions. The criteria of application of concepts delimit the language game but aren’t part of it. That is to say, they aren’t moves within it. We may say that they are the (often implicit) rules which make a language game the language game it in fact is and that they make concepts what they actually are. The fact that the participants of the language game behave in accord with those rules, which they may well know just implicitly, allows us to consider them as competent speakers and practitioners of the language game. As remarked, rules may not be acquired explicitly and may be learnt by simply being trained to take part in the game, or by merely participating in it. Moreover, the rules don’t pre-exist the language game and aren’t the result of a previous explicit convention. Rather, they depend on our actual linguistic practice – on how the language game is in fact played. Yet, it is an essentially philosophical task to bring them back to the surface and make them explicit. Once that task is accomplished, rules will, per force, appear to competent speakers like well-known trivialities. They won’t say anything new or surprising. In fact, properly speaking, they won’t, according to Wittgenstein, say anything whatever, since they are not empirical and informative propositions, which describe states of affairs; but, rather, manifestations of the conditions under which our concepts are the concepts they in fact are, whose status is merely logical, that is to say, grammatical.6 Nevertheless, their being made explicit and their clarification can teach philosophers something. In particular, not to ignore and abuse them by taking them for other than what they are. That is to say, for genuinely descriptive propositions that seem to state necessary, metaphysical truths, or that can, alternatively, be challenged while still making sense, when, in fact, in both cases what is being produced would be mere nonsense.7 2.1
‘I know’ and reasons
In OC 11 Wittgenstein remarks that ‘we just do not see how very specialized the use of “I know” is’. On the one hand, this means that different language games can be played with ‘I know’; on the other hand, this remark suggests that the language game with ‘to know’ is complex and can be played only when one already possesses a high level of competence with language. Let us look at it more closely.
62
Moore and Wittgenstein
According to Wittgenstein, here are some examples of the correct use of ‘I know’: ‘I know that that’s a tree’ – this may mean all sorts of things: I look at a plant that I take for a young beech and that someone else thinks is a black-currant. He says ‘that is a shrub’; I say it is a tree. – We see something in the mist which one of us takes for a man, and the other says ‘I know that that’s a tree’. (OC 349) Someone with bad sight asks me: ‘do you believe that the thing we can see there is a tree?’ I reply ‘I know it is; I can see it clearly and am familiar with it’. – A: ‘Is N. N. at home?’ – I: ‘I believe he is.’ – A: ‘Was he at home yesterday?’ – I: ‘Yesterday he was – I know he was; I spoke to him.’ – A: ‘Do you know or only believe that this part of the house was built on later than the rest?’ – I: ‘I know it is; I got it from so and so’. (OC 483) In these cases, then, one says ‘I know’ and mentions how one knows [das Grund], or at least one can do so. (OC 484) Let us consider the second of Wittgenstein’s examples from the abovementioned quotation. The fact that I talked to NN yesterday and offer it as my ground for saying that I know he was at home is a criterion for the correct assertion ‘I know that…’. A criterion is neither a necessary nor a sufficient condition for knowledge that p. For, clearly, I could have talked to NN’s twin brother, and NN could have been at home while I had no conversation with him. Criteria, that is, aren’t necessary and sufficient conditions meant to define concepts. That a definition of knowledge, or of any other concept, was not what Wittgenstein was after, was clear since the passages on the concept game and the notion of family resemblance in the Philosophical Investigations.8 However, the fact that in general our use of ‘to know’ in the first person present be made on the basis of grounds – whatever they contextually happen to be – is a criterion for saying that that use of ‘I know’ was meaningful though, perhaps, mistaken. Thus, if I had talked to NN’s twin brother this would (probably) falsify the claim that NN was at home and would (certainly) show that I did not know, but only believed to have talked to him. My claim ‘I know that NN was at home yesterday’ would thus be
Belief, Knowledge and Certainty 63
false, but my use of ‘I know’ would not be nonsensical. By contrast, if I claimed to know that NN was at home yesterday but if I were unable to offer reasons in favour of such a claim, that would show that I misused the expression ‘I know’.9 That, in turn, would determine my exclusion from the language game with that expression. My words would be no move within the language game and would thus lack a meaning.10 Hence, the correct – that is to say, meaningful – employment of ‘I know that p’ entails that my belief that p be based on avowable reasons, which may yet turn out not to be conclusive, thereby showing that my claim to knowledge was, after all, false. If I conform to the criteria for the use of the expression ‘I know’ my interlocutor will then acquire good reasons to believe that things are as I have said them to be. Hence, the specific grounds offered, that are criteria for my correct use of ‘I know’, are then evidence [Evidenz] of the fact that I claim to know. My interlocutor could thus conclude – as defeasibly as that might be – from my claim to know that p to ‘p’. Thus, in this case, the claim to know that p and the simple assertion of ‘p’, accompanied by giving reasons in its favour, can play the same role within the language game.11 In some other cases, in contrast, ‘I know that p’ says something different from ‘p’. ‘In the first sentence a person is mentioned, in the second, not’ (OC 587). However, in the same entry, Wittgenstein also writes, ‘but that doesn’t show that they have different meanings’. However, it is clear that, in context, the fact that the first sentence mentions a person makes it unsuitable as a replacement of ‘p’. Think of ‘a zoo [where there might be] a notice “this is a zebra”; but never “I know this is a zebra”’ (OC 588). Hence, what’s the difference between the two assertions? Wittgenstein suggests that ‘‘‘I know” has meaning only when it is uttered by a person. But, given that, it is a matter of indifference whether what is uttered is “I know…” or “That is…’’’ (ibid.). That is to say, if I have good reasons to say that the animal in the pen is a zebra I can equally say that I know that that is a zebra or ‘That is a zebra’. If someone asked me how I knew it, I could give my grounds for either of my assertions. But, obviously, nobody could ask a notice how it knows that that is a zebra. This, however, doesn’t yet show that there is a difference between asserting the two. For, after all, the notice doesn’t assert anything whatsoever. It is only because we have been trained to take it as stating the name for the animal in the pen, that we can use it as a ground for our assertion ‘That is a zebra’ or, even, ‘I know that that is a zebra’. Hence, while it is obvious that ‘I know that p’ makes sense only when it is a person to utter it, that doesn’t yet show its difference
64
Moore and Wittgenstein
with respect to the mere assertion of ‘p’ when, again, the latter is made by a person. However, Wittgenstein hastens to add: Even when one can replace ‘I know’ by ‘It is …’ still one cannot replace the negation of the one by the negation of the other. With ‘I don’t know…’ a new element enters our language games. (OC 593) For, clearly, to deny ‘p’ isn’t in any way equivalent to the denial that one knows it. The former should entail that one knows that not-p, or, at any rate, that one believes that not-p with good grounds, while ‘I don’t know that p’ denies that one has knowledge of p, while entailing, perhaps only conversationally, that one doesn’t know that not-p either. So ‘I don’t know that p’ says something about the subject – that he doesn’t have knowledge of ‘p’ – in a way in which ‘not-p’ doesn’t.12 Hence, the main difference between the language game of making assertions and of making positive or negative claims to knowledge consists, according to Wittgenstein, in the fact that all of a sudden the subject comes in and may come to play a role when he denies that he himself has a given piece of knowledge. Yet, for Wittgenstein, also when ‘I don’t know’ is at stake, nothing is said about one’s own mental states. Rather, by asserting ‘I don’t know’ one declares one’s inability to play the role, within the language game, of a reliable witness or informant with respect to ‘p’. One thus declares one’s impossibility to play that language game further and, accordingly, will expect his interlocutor not to probe him with other questions about whether ‘p’ is the case. Hence, to conclude, the language games of asserting ‘p’ and of claiming ‘I know that p’ get very close and overlap in the affirmative case, while they deviate from one another in the negative case. 2.2
‘I know’ and ‘I am certain’
Writes Wittgenstein: The difference between the concept of ‘knowing’ and the concept ‘being certain’ isn’t of any great importance at all, except where ‘I know’ is meant to mean: I can’t be wrong. In a law-court, for example, ‘I am certain’ could replace ‘I know’ in every piece of testimony. We might even imagine its being forbidden to say ‘I know’ there. [A passage in Wilhelm Meister, where ‘You know’ or ‘You knew’ is used in the sense ‘You were certain’, the facts being different from what he knew]. (OC 8)
Belief, Knowledge and Certainty 65
The certainty Wittgenstein is talking about in this remark – not to be conflated with the certainty due to the impossibility of error, or objective certainty (that we will discuss in §3) – is the one produced by the correct development of the language game. That is to say, when one’s grounds are particularly strong, one could say ‘I am certain that NN was at home yesterday because I talked to him’, rather than ‘I know that NN was at home’. The fact that NN might have an identical twin brother is not relevant to the language game and it is not required that I be in a position to exclude such an event to be entitled to say, ‘I am certain/ know that NN was at home yesterday’. Hence, in this case, the language games with ‘to know’ and ‘to be certain’ overlap. For in both cases the ability to give reasons [Grund], justifications [Rechtfertigung] and evidence [Evidenz, Beweis] in support of one’s claim is presupposed. According to Wittgenstein, Moore’s use of ‘I know’ and ‘I am certain’ in relation to his truisms goes against this criterion. As we saw at length in the Chapter 1, in PEW Moore claims that he knows with certainty that there is his hand where he seems to see it, but that he doesn’t know how he knows it. ‘I know that that’s a tree’. Why does it strike me as I did not understand the sentence? Though it is after all an extremely simple sentence of the most ordinary kind? It is as if I could not focus my mind on any meaning. Simply because I don’t look for the focus where the meaning is. As soon as I think of an everyday use of the sentence instead of a philosophical one, its meaning becomes clear and ordinary. (OC 347) ‘I know that that’s a tree’ is something a philosopher might say to demonstrate to himself or to someone else that he knows something that is not a mathematical or a logical truth. Similarly, someone who was entertaining the idea that he was no use any more might keep repeating to himself ‘I can still do this and this’. If such thoughts often possessed him one would not be surprised if he, apparently out of all context, spoke such a sentence loud. (But here I have already sketched a background, a surrounding, for this remark, that is to say, given it a context). But if someone, in quite heterogenous circumstances, called out with the most convincing mimicry: ‘Down with him!’, one might say of these words (and their tone) that they were a pattern that does indeed have familiar applications, but that in this case it was not even clear what language the man in question
66
Moore and Wittgenstein
was speaking. I might make with my hand the movement I should make if I were holding a hand-saw and sawing through a plank; but would one have any right to call this movement sawing, out of all context?– (It might be something quite different!). (OC 350) As we have seen, however, the point in Moore’s strategy is not that he somehow refuses to produce the grounds for his claim. For he acknowledges that he knew that there was a hand in front of him on the basis of his sensory evidence. Yet, Wittgenstein insists that Moore’s use of ‘I know’ in relation to the premises of his proof, as well as with respect to his truisms, is nonsense. Wittgenstein’s point here is that Moore’s grounds aren’t stronger than what they are supposed to justify (OC 245). The idea then seems to be that one’s visual experience can justify the belief that there is a hand where one seems to see it just in case it is generally reliable. But in order to know that it is, it would have to be tested in circumstances broadly similar to those in which Moore’s proof was delivered, in order to check whether it would give rise, in a given subject, to the belief ‘Here is a hand’. Yet this means that, in those circumstances, that there is a hand there is taken for granted and used, in its turn, to test one’s sight. Hence, in the circumstances of Moore’s proof, one’s visual experience can’t be a justification for ‘Here is a hand’, for that there is a hand there is no more certain before or after one has looked at it (OC 243, 245, 250).13 This doesn’t at all mean that one can’t imagine different contexts in which the claim ‘I know this is my hand’, based on one’s visual experience, would be meaningful – that is to say, where it would be a move within the language game. But this would make that claim totally philosophically unsurprising. This could happen after a car accident, for example, when one suddenly opens one’s eyes after having been unconscious, looks around, sees various objects, doesn’t feel one’s limbs, yet can see one’s hand at the end of one’s arm and says ‘(Oh, good) here is my hand’. Writes Wittgenstein: But now it is also correct to use ‘I know’ in the contexts which Moore mentioned, at least in particular circumstances. (Indeed, I do not know what ‘I know that I am a human being’ means. But even that might be given a sense.) For each one of these sentences I can imagine circumstances that turn it into a move in one of our language-games, and by that it loses everything that is philosophically astonishing. (OC 622. Cf. OC 347)
Belief, Knowledge and Certainty 67
For what is philosophically surprising in the contexts in which Moore claims that he knows? Precisely the fact that the opposite seems inconceivable. That is to say, it is inconceivable that Moore (just like anyone else) may not know, in those circumstances, that what he holds up in front of himself is his hand. According to Moore, as we have seen, such an inconceivability makes the proposition ‘Here is my hand’ certain and indubitable, so that it makes sense, for Moore, to preface it with ‘I know’ and to use it to oppose idealists and sceptics. However, if the same proposition were used in a different context where ‘Here is my hand’ were really a hypothesis to be confirmed and one had good grounds to say ‘I know here’s my hand’, one’s claim to knowledge wouldn’t have any philosophical implication and could be questioned on the basis of a thorough examination of one’s grounds for asserting it. But, in such an event, it wouldn’t cross our minds that that claim could be used to oppose scepticism and idealism. Writes Wittgenstein, Moore has every right to say that he knows there’s a tree there in front of him. Naturally he may be wrong. (For it is not the same as with the utterance ‘I believe there is a tree there’.) But whether he is right or wrong in this case is of no philosophical importance. If Moore is attacking those who say that one cannot really know such a thing, he can’t do it by assuring them that he knows this and that. For one need not believe him. If his opponent had asserted that one could not believe this and that, then he could have replied: ‘I believe it.’ (OC 520) Moore’s mistake lies in this – countering the assertion that one cannot know that, by saying ‘I do know it’. (OC 521) If Moore’s use of ‘I know’ were meaningful, one could believe him or not, depending on the cogency of his reasons; his claim, however, wouldn’t have any philosophical significance. Yet, Moore’s claim that he knows that p wouldn’t be enough to show or guarantee that he really does. As we have seen, ‘I know’ is used differently from ‘I believe’, and, in particular, in such a way that one’s subjective assurance is irrelevant and proves nothing. This, however, doesn’t mean at all that Wittgenstein would agree with those sceptics who would claim that it is always impossible to have knowledge of genuinely empirical propositions such as ‘Here is my hand’ (in context). For, according to him, once good grounds in their favour are available, that often suffices for one to have knowledge of them.
68
Moore and Wittgenstein
2.3
‘I know’ and finding out
Wittgenstein writes, So if I say ‘I know that I have two hands’, and that is not supposed to express just my subjective certainty, I must be able to satisfy myself that I am right. But I can’t do that, for my having two hands is not less certain before I have looked at them than afterwards. (OC 245) According to Wittgenstein, it belongs to our language game with ‘to know’ that there may be a doubt or an open question with respect to ‘p’ and that they can both be answered by an enquiry that would justify the claim that one knows that p. As we have already seen, this is impossible with the propositions Moore claims to know, in the very circumstances in which he makes his claims. For, on the one hand, there is no doubt or question that there is his hand in the circumstances of his proof. On the other, there is no enquiry whose methodological characteristics would be stronger of what they are supposed to prove. Writes Wittgenstein, Only in certain cases it is possible to make an investigation ‘is that really a hand?’ (or ‘my hand’). For ‘I doubt whether that is really my (or a) hand’ makes no sense without some more precise determination. One cannot tell from these words alone whether any doubt at all is meant – nor what kind of doubt. (OC 372) If I don’t know whether someone has two hands (say, whether they have been amputated or not) I shall believe his assurance that he has two hands, if he is trustworthy. And if he says he knows it, that can only signify to me that he has been able to make sure, and hence that his arms are e.g. not still concealed by coverings and bandages, etc. etc. My believing the trustworthy man stems from my admitting that it is possible for him to make sure. But someone who says that perhaps there are no physical objects makes no such admission. (OC 23) It is quite rare to have doubts as to whether one still has one’s hands, but of course it might happen, as we saw in §2.2. In such an event it would be possible to find out and, accordingly, one’s claim to knowledge would be a move within the language game whereby one would declare that
Belief, Knowledge and Certainty 69
one has made such an enquiry and that it was successful. By contrast, Moore is not in these unusual circumstances where there is a doubt as to whether there is his hand and, connectedly, an enquiry which could answer it. As we have already seen, while giving his proof, Moore has the evidence of his senses. But that evidence can function as a justification for ‘Here is my hand’ just in case (i) the existence of an external world isn’t called into doubt and (ii) the reliability of one’s senses is taken for granted.14 But, then, (ii) can be taken for granted only if, once it is held fast that, in a given circumstance, there are specific physical objects, it is ascertained that subjects would form the appropriate beliefs. The circumstances of Moore’s proof and the fact that there is a hand there would just serve that purpose. For reflect, how would that be ascertained? Obviously by presenting subjects with specific physical objects in optimal conditions and by testing their responses. Now, the circumstances of Moore’s proof and its premises would perfectly fit the bill. If so, however, it must be taken for granted that there is a hand in the circumstances of Moore’s performance, in order to be entitled to take one’s sensory experience as reliable. Yet that was what, in its turn, was required in order for one’s experience really to ground one’s belief ‘Here is a hand’. Hence, sight can’t non-circularly ground one’s belief in the premises of Moore’s proof. That, according to Wittgenstein, suffices to show that there is no non-circular enquiry that could ground Moore’s claim that he knows that there is his hand where he seems to see it. This, in turn, is tantamount to saying that there is no enquiry that could show that Moore really knows what he claims to know. Hence, on Wittgenstein’s view, his claim isn’t a move within the language game with ‘I know’ and, for this reason, it becomes doubtful that we can really make sense of it – that is to say, that it really has a meaning, contrary appearances notwithstanding. 2.4
‘I know’ and relevance
The philosophical use of ‘I know’ (§1) is characterized, according to Wittgenstein, by the fact that it isn’t a move within our usual language game with that expression. It might then be suggested, following Moore’s response to Malcolm, cited in Malcolm [1977] (chapter 1, §3.2), that to assert ‘I know this is my hand’ is a superfluous claim in most cases, yet that it is meaningful and actually true. Wittgenstein’s rejoinder concedes the first part of this response, but not the second.15 For surely, in most cases, we wouldn’t for a moment think of calling into doubt such a thing and we would also be disposed to say that it is true that Moore knows that that is his hand.16 As Wittgenstein writes, ‘Thus it seems to me that I have known something the whole time, and yet there is no meaning in
70
Moore and Wittgenstein
saying so, in uttering this truth’ (OC 466, cf. 464). Hence, he would agree with Moore that the latter’s assertion is weird and superfluous. It would be so for it is just a triviality, which is common knowledge, and that wouldn’t add anything new or inform one of something. As Wittgenstein notes, ‘The truths Moore says he knows, are such as, roughly speaking, all of us know, if he knows them’17 (OC 100). But Moore would in turn agree with Wittgenstein on this much. After all, he calls his truths, ‘truisms of common sense’, for he thinks that they belong to a shared world-view and are entirely obvious trivialities. Yet Wittgenstein asks himself: Why doesn’t Moore produce as one of the things that he knows, for example, that in such-and-such a part of England there is a village called so-and-so? In other words: why doesn’t he mention a fact that is known to him and not to every one of us? (OC 462) The reason why Moore doesn’t mention something that is known only to him is that he actually needs to mention trivial and uninformative truths as he wishes to characterize common sense propositions, which are known to everyone and could thus serve as foundations of all the rest of our knowledge. In contrast, it would certainly be possible to doubt the fact that there is a village called so-and-so in a certain region of England. Hence, one’s claim to possess such a piece of knowledge couldn’t express a commonsensical certainty and would thus be philosophically pointless. (It should be recalled that Moore’s aim is to list a number of obvious and certain truths which he thinks would entail the falsity and/or absurdity of idealism and scepticism). However, according to Wittgenstein, the claim to know what is obvious isn’t a move – no matter how pragmatically improper – within the language game with ‘to know’.18 To hold otherwise would, in his opinion, annihilate the distinction between meaningful and meaningless uses of that verb. For, in Wittgenstein’s view, to give a relevant piece of information isn’t a maxim of the logic of conversation (pace Grice). That is to say, it doesn’t pertain to the pragmatic of language. Rather, it is a requirement of the grammar of our language game with ‘to know’. That language game, which by being what in fact is determines the rules for the correct employment of ‘to know’, simply doesn’t contemplate the use of ‘I know’ in connection with trivial and well-known truths. For reflect: if someone came up with ‘I know this is my hand’ in the middle of a conversation where there was no issue about one’s having a hand, we would have trouble in making sense of that claim. We could either take it as a symptom of a (perhaps brief
Belief, Knowledge and Certainty 71
and suddenly overcome) lack of memory with respect to the meaning of a word, which could make sense if, for instance, our interlocutor’s mother tongue weren’t English, or if we knew he had recently had some brain injury which had somehow impaired his linguistic competence. Alternatively, we would perhaps take it as a sign of mental illness (OC 467). For to claim knowledge of the obvious without reason, that is to say, when there is no doubt or question as to whether p is the case, makes us see that behaviour as an assurance given one’s previous uncertainty. But that there may be an uncertainty with respect to obvious trivialities, which, according to Wittgenstein, constitute the background against which we have learnt to make judgements and to distinguish between truth and falsity, makes it doubtful that the subject who makes such a claim to knowledge is actually ‘judging in conformity with mankind’ (OC 4). Hence, his behaviour will be a symptom of deviance from rationality and common sense. Yet, it would not give rise to an alternative language game. That is to say, although we may at first attribute a meaning to his claim to knowledge and, by so doing, take it as a symptom of mental illness, we wouldn’t then extend our language game to include his use of ‘I know’. Rather, if the subject persisted in making those assertions, we would soon lose interest in his words, and would cease to attribute them a meaning – that is to say, we would soon take them as mere mumblings produced by a feeble mind. To recap: to claim knowledge of the obvious could be meaningful, according to Wittgenstein, only in certain circumstances – viz. after a temporary lack of linguistic memory due to various reasons – but, bar such a kind of scenario and contextualization, we couldn’t make sense of such a claim. We could perhaps take it as a sign of mental illness, but that would not represent a direction for a possible development of our language game with ‘to know’. Thus, this shows, in Wittgenstein’s view, that it belongs to the grammar of our use of ‘to know’ that what one claims to know be not a triviality to which everyone is party. Since, according to Wittgenstein, philosophy simply isn’t a new and different language game, where the rules which govern our ordinary ones can freely ‘go on holiday’,19 while still producing sense, it follows that, in his view, Moore’s use of ‘I know’ isn’t just superfluous or weird, but altogether meaningless or nonsensical.20 2.5
‘I know’ and ‘I don’t know’
Writes Wittgenstein, ‘I know that I am a human being.’ In order to see how unclear the sense of this proposition is, consider its negation. At most it might be
72
Moore and Wittgenstein
taken to mean ‘I know I have the organs of a human’. (E.g. a brain which, after all, no one has ever yet seen.) But what about such a proposition as ‘I know I have a brain’? Can I doubt it? Grounds for doubt are lacking! Everything speaks in its favour, nothing against it. (OC 4) In §2.1 we have seen how, in order to claim knowledge that p, one needs to meet criteria of application for that concept. However, as we have remarked, the reasons produced in favour of one’s claim are evidence in favour of ‘p’ and make it a well-grounded hypothesis. Yet, that falls short of entailing its certainty. The proposition that p remains a hypothesis even when it is confirmed by further evidence. Hence, in order for ‘p’ to be a hypothesis its negation must be possible – that is to say, conceivable; and be so in the same circumstances in which ‘p’ is in fact confirmed by evidence. So, it must in fact be possible not to know that p. This is tantamount to saying that if ‘p’ is a hypothesis, then it must be possible both to know and not to know it. Thus, the requisite of sense for a proposition – that is to say, its bipolarity21 – holds also for ‘I know that p’. As a consequence, in order for ‘I know that p’ to be meaningfully asserted, it must in principle be possible to be mistaken both about ‘p’ and about one’s claim to knowledge.22 In such a way that the remark ‘I thought I knew’ (OC 12) could always make sense. However, when Moore’s truisms are at stake, it wouldn’t make sense to say that one doesn’t know them. If I doubted of the fact that I am a human being, or if I said I didn’t know it, it would become doubtful that I knew the meaning of the expression ‘human being’; or else, my assertion would be taken as a symptom of mental illness. For, if I know the meaning of that expression, I must be able to apply it to one of its paradigmatic instances, for example, myself. If I wanted to doubt whether this is my hand, how could I avoid doubting whether the word ‘hand’ has a meaning? So that is something I seem to know after all.23 (OC 369) If, in contrast, I were able to apply it to other instances of that concept and yet claimed not to know that I am a human being, I would show to be affected by some mental disorder. Moore’s mistake thus consists in failing to realize that where it doesn’t make sense to doubt or to claim not to know, then it doesn’t make sense to claim that one does know either. For, first, one’s claim to knowledge
Belief, Knowledge and Certainty 73
wouldn’t be based on the criteria that govern the correct employment of ‘I know’; and, second, because what is claimed to be known wouldn’t be a hypothesis which may be confirmed or disconfirmed. As we have repeatedly seen, Moore, according to Wittgenstein, doesn’t base his claim to know that there are his hands where he seems to see them on reasons that are stronger than what they are supposed to ground. Hence, his claim, in Wittgenstein’s view, doesn’t conform to the criteria of the correct use of ‘I know’. Moreover and connectedly, it is Wittgenstein’s opinion that ‘Here is my hand’, said in the circumstances of Moore’s proof, isn’t a hypothesis that can actually be confirmed (or infirmed) by the experience. For if that were in question, then either it couldn’t serve its alleged philosophical purpose, or else, it would become doubtful, in those circumstances, either that one could trust one’s own senses to try and find out whether one’s own hands are where they appear to be, or that one knew the meaning of the word ‘hand’. Thus, in the context of Moore’s proof, ‘Here is my hand’ is not an empirical proposition, but, as we shall see further in the following (§3.2), either a rule for the correct use of language, or a norm of evidential significance. That is to say, it is what has to stand fast in order to gather evidence in favour of other, genuinely empirical propositions. So far we have seen the criteria that govern the use of the verb ‘to know’ and the expression ‘I know’ within our ordinary language game and we have repeatedly remarked that, according to Wittgenstein, the only guarantee of sense of one’s utterances is their belonging to a language game. The case of claims to knowledge is no exception. Hence, to sum up, we have identified five criteria: • claims to knowledge must be based on reasons (§2.1); • and, in fact, on reasons stronger than what the latter are supposed to ground (§2.2); if this criterion is respected, one can then equally say ‘I know that p’ or ‘I am certain that p’ (§2.2); • claims to knowledge are in order only where it is possible to make an enquiry to verify that things are thus-and-so (§2.3); • furthermore, it makes sense to say ‘I know’ only when that is relevant within the communicative exchange (§2.4); • finally, it makes sense to claim ‘I know that’ only where it is equally possible that one may not know, or doubt that p (§2.5). These criteria make explicit the grammar of the expression ‘I know’ and they are learnt as one learns to take part in the language game with such an expression. Of course this does not require that these criteria
74
Moore and Wittgenstein
be made explicit, as one is trained to participate in the language game. Yet, to anyone who masters it, they will appear as obvious. Furthermore, since these criteria say something about language and concepts and not about the world, their status isn’t empirical, but grammatical. They aren’t descriptions but rules which characterize when a given use of ‘I know’ is to count as meaningful, though, perhaps, false. As long as our language game with ‘to know’ is as it is, these rules are in place and can be denied only on pain of annihilating the distinction between correct and incorrect – that is, meaningful and nonsensical – uses of that verb. Of course it is possible that our language game might change,24 if our linguistic practice with that expression got altered. Accordingly, the new language game will be at least implicitly yet constitutively governed by different rules. Still, this wouldn’t show that the ones so far outlined were false or mistaken, but only that the meaning of ‘to know’ and therefore the concept of knowledge have somehow changed. For it is a tenet of Wittgenstein’s philosophy that concepts are a function of how words are used in our language game.25 Hence, all the criteria evinced so far aren’t just rules for the meaningfulness of claims to knowledge, which would say nothing, however, on the very concept of knowledge, or that, at the very least, would have no clear bearing on the latter.26 Rather, by characterizing the correct employment of ‘to know’ and ‘I know’, those criteria determine the very concept at issue, albeit without providing a definition of it. Accordingly, we can see that for Wittgenstein the concept of knowledge is constitutively tied to the ability to give reasons in favour of one’s belief that p and to the fact that such reasons be stronger than, and not circularly dependent on, what they are supposed to ground.27 Furthermore, we can now appreciate how, for him, the concept of knowledge, while factive, would not entail either (what philosophers other than Wittgenstein would call) certainty – that is to say, indefeasibility – or transparency (or what other philosophers would call ‘luminosity’). Hence, any genuine claim to knowledge must always be revisable, at least in principle, in such a way that the claim ‘I thought I knew’ could always be sensibly made. We shall now turn to a different employment of the verb ‘to know’, especially in the first person present, that doesn’t conform to the previous criteria and which, in fact, isn’t governed by any criterion at all (OC 574). We will thus consider the ‘grammatical use’ of that expression.
3 The grammatical use of ‘I know’ Wittgenstein, as we saw in §2.2, thinks that ‘the difference between the concept “knowing” and the concept of “being certain” isn’t of any
Belief, Knowledge and Certainty 75
great importance at all’ except, however, ‘where “I know” is meant to mean: I can’t be wrong’ (OC 8). When we say that we are certain ‘we express complete conviction, the total absence of doubt, and thereby we seek to convince other people. That is subjective certainty’ (OC 194, cf. 245, 415). But – asks Wittgenstein – ‘when is something objectively certain?’ (ibid. my italics); and answers, ‘When a mistake is not possible’ (ibid.). That is to say, when the possibility of being mistaken is ‘logically excluded’ (cf. ibid. my italics).28 At another place, however, he writes, ‘Knowledge’ and ‘certainty’ belong to different categories. They are not two ‘mental states’ like, say ‘surmising’ and ‘being sure’ (…). What interests us now is not being sure but knowledge. That is, we are interested in the fact that about certain empirical propositions no doubt can exist if making judgments is to be possible at all. Or again: I am inclined to believe that not everything that has the form of an empirical proposition is one. (OC 308) But, as we saw in §2.2, sometimes ‘I am certain’ can occur in place of ‘I know’, and yet, Wittgenstein now says, ‘I know’ doesn’t express a subjective conviction. This makes it puzzling to see how for Wittgenstein ‘I am certain’ can express a subjective conviction and thus a mental state. In the contexts in which ‘I know’ can be substituted by ‘I am certain’, I suggest to interpret both these claims as in fact equivalent to the following one: ‘for all I know, things are thus-and-so’. For this last assertion makes clearer how my certainty is here a function of the information, evidence and grounds I have in favour of a given hypothesis. This, however, doesn’t guarantee that things are as I – with good right – claim them to be. Thus, one can say that my certainty is subjective, as it is a function of a subject’s state of information. This doesn’t mean that the claim ‘I am certain’ would somehow be arbitrary and whimsical. Simply, my certainty doesn’t exclude the possibility of error: I could have overlooked or misinterpreted some evidence and wrongly taken it to support ‘p’; or else, I could fail to have some decisive piece of evidence against it. So, we can make sense of Wittgenstein’s claim that, in this case, one’s certainty would be subjective. Yet, can we also make sense of his claim that it would then be a mental state? After all, it won’t differ from knowledge and Wittgenstein, as we saw in §2.1, stressed the fact that knowledge isn’t a mental state. He writes, One may for example call ‘mental state’ what is expressed by tone of voice in speaking, by gestures etc. It would thus be possible to speak
76
Moore and Wittgenstein
of a mental state of conviction, and that may be the same whether it is knowledge or false belief. (OC 42) I think that the kind of certainty Wittgenstein dubs ‘subjective’ isn’t itself a mental state, just as knowledge isn’t, for him, but that it can be accompanied by a particular ‘mental state of conviction’, that can be manifested in the tone of voice and in one’s manners. That there be such aspects of ‘tone’ and that there be also a particular mental state of conviction – a special feeling, as it were – is entirely plausible, from a phenomenological point of view. This, however, doesn’t exclude the possibility of error. I may feel as certain as ever of having taken my keys this morning when I left home, and yet find out later that I didn’t. A particular feeling is not a sufficient condition (nor a necessary one) for the truth of the hypothesis one claims to be certain about, as it could occur both when one really knows that p and ‘p’ is thus the case and when one doesn’t know it because in fact ‘p’ isn’t the case. So one’s claim to be (subjectively) certain and its accompanying feelings of total conviction don’t guarantee the impossibility of being mistaken. Wittgenstein, therefore, isn’t interested in this kind of certainty, nor, in this case, in the use of ‘I know’ within our usual language games, where it can be substituted by ‘I am certain/sure’. Rather, he wishes to clarify that peculiar use of ‘I know’ which is a synonym of ‘I can’t be wrong’ and which expresses objective certainty (OC 569). A certainty which, in its turn, doesn’t depend on the infallibility of a subject, but on the logic of our language and of our practices of enquiry. Such a certainty, therefore, will be somehow grammatical and normative, rather than psychological. Let us explore it in detail. 3.1
In the grammatical use of ‘I know’ the ‘I’ is unimportant
Wittgenstein writes, If ‘I know etc’ is conceived as a grammatical proposition, of course the ‘I’ cannot be important. And it properly means ‘There is no such thing as a doubt in this case’ or ‘The expression “I don’t know” makes no sense in this case’. And of course it follows from this that ‘I know’ makes no sense either. (OC 58) Let us consider the first part of this remark: in its grammatical use ‘I know’ means ‘There is no such thing as a doubt in this case’. As we
Belief, Knowledge and Certainty 77
saw in §1, the impossibility of doubt and mistake is often seen, in philosophy, as a consequence of the fact that a subject be in an epistemically privileged position, capable of ruling out doubts and mistakes in his own case, or from his own point of view. Clearly, if, strictly speaking, I know only what is given to my conscience and if I can’t be wrong or have doubts in this case, the impossibility of mistakes or of doubts, on my side, depends on the fact that I am in such an epistemically privileged position. That is to say, a position such that the object of knowledge is private to its owner and its nature transparent to him. Moore’s propositions, however, aren’t about (putatively) private objects, given to a subject’s conscience; but about ordinary physical ones. If so, however, in order to show ‘someone that we know truths, not only about sense data but also about things (…) it can’t be enough (…) to assure us that [one] knows this’ (OC 426). For in such a case the subject clearly isn’t in any epistemically privileged position. As Wittgenstein puts it: ‘The train leaves at two o’clock. Check it once more to make certain’ or ‘The train leaves at two o’clock. I have just looked it up in a new time-table’. One may also add ‘I am reliable in such matters’. The usefulness of such additions is obvious. (OC 444) But if I say ‘I have two hands’, what can I add to indicate reliability? At the most that the circumstances are the ordinary ones. (OC 445) Nor can it be enough that a mistake be impossible because of either my good faith or scrupulousness. For they would account, at most, only for my subjective certainty. According to Wittgenstein, then, Moore’s mistake would depend on having explained in a psychologistic way his own certainty with respect to his truisms and the premises of his proof. He thus failed to see that the certainty with respect to, for instance, ‘Here is my hand’, depends on the role that that proposition plays in our language. In particular, on the fact that, unless there is a genuine question as to whether an object visually presented to one is one’s hand, that proposition can only be used either to teach or remind one of the meaning of the word ‘hand’, or as a ‘unity of measure’ against which one can test other propositions, for example, whether one’s sense organs are working correctly. Thus, contrary to what Moore seems to hold, one’s own (objective) certainty
78
Moore and Wittgenstein
is not and can’t be a function of one’s own subjective impossibility to call that proposition into doubt, in those circumstances, as if that might depend on one’s own privileged epistemic position, or on one’s own personal conviction that things may not be otherwise. 3.2 In its grammatical use ‘I know’ means ‘There can’t be a doubt in this case’ or ‘I can’t be wrong’ As Wittgenstein claims in the second part of OC 58, mentioned in the previous section, if ‘I know’ is used grammatically, or if it is a ‘logical insight’, as he puts it in OC 59, then it means ‘There can’t be such thing as a doubt in this case’. Hence, the assertion ‘I don’t know’ is senseless as well as the claim ‘I know’. What we need to clarify, though, is that the impossibility of doubts and mistakes doesn’t depend on a subject’s epistemically privileged status, but on the role that the propositions that are said to be (grammatically) known play in our language and practices. Writes Wittgenstein: I want to say: propositions of the form of empirical propositions, and not only propositions of logic, form the foundation of all operating with thoughts (with language). – This observation is not of the form ‘I know…’. ‘I know…’ states what I know, and that is not of logical interest. (OC 401) In this remark the expression ‘propositions of the form of empirical propositions’ is itself thoroughly bad; the statements in question are statements about material objects. And they do not serve as foundations in the same way as hypotheses which, if they turn out to be false, are replaced by others (…). (OC 402. Cf. 308) First of all, according to Wittgenstein, the expression ‘propositions of the form of empirical propositions’, which refers to Moore’s truisms and the like, is inappropriate because it suggests that the role of a proposition in a language game can be fathomed from its form, viz. from the fact that it mentions physical objects, for instance. In contrast, according to Wittgenstein, from their form one can only evince that Moore’s truisms mention material objects. This, however, doesn’t entail that they play the role that empirical propositions usually play in our language.29 Contrary to genuinely empirical propositions, they are not subject to verification and checking. Their ‘twofold status’ – that is to
Belief, Knowledge and Certainty 79
say, their being about material objects and yet playing a role other than that of ordinary empirical propositions – is their peculiarity and the source of their interest, from a philosophical point of view.30 As we have seen time and again, when I say ‘I know this is my hand’ (in circumstances such as those of Moore’s proof) could I sensibly fail to know such a thing? If I did, it would show either that I don’t know the meaning of the word ‘hand’, or that I am suddenly affected by pathological doubts. But, then, in either case, I could not hold it up in front of me to claim ‘I know that this is my hand’. Hence, the use of ‘I know’ in relation to ‘Here/This is my hand’ is grammatical (cf. OC 58-9) – that is to say, normative – as it wouldn’t make sense to doubt or say that one doesn’t know such a thing. Furthermore, if I really doubted that the one that I am holding up right in front of me is my hand, this would impair my actions. How could I possibly still grab the glass of water that I have been asked to bring for someone, for instance, if I doubted of the existence of my hand? As Wittgenstein put it: I KNOW that this is my foot. I could not accept any experience as proof to the contrary – That may be an exclamation; but what follows from it? At least that I shall act with a certainty that knows no doubt, in accordance with my belief. (OC 360) Now I would like to regard this certainty, not as something akin to hastiness or superficiality, but as a form of life. (That is very badly expressed and probably badly thought as well). (OC 358) But that means I want to conceive it as something that lies beyond being justified or unjustified, as it were as something animal. (OC 359) In the last chapter, we will come back to the relationship between certainty in action and (grammatical) knowledge that this is my hand (or foot), and on their dependency on a form of life. Although I think it is already evident that such a certainty, more than a positive, explicit attitude towards a proposition, manifests itself in action negatively, as it were – as a lack of doubt and even of consideration for its various objects. Be that as it may, what is relevant to present purposes is the fact that, when I say that this that I hold up in front of myself is my hand, ‘I could not accept any experience as proof to the contrary’. There is a twofold explanation of this claim.
80
Moore and Wittgenstein
First, an experience can be taken by a subject to disprove an empirical proposition only if he understands it. Hence, just in case he knows the meaning of the terms used to express it. But if one doubted, in circumstances such as those of Moore’s proof, of the fact that what one holds up in front of oneself is one’s hand, that would cast doubt on the fact that one really understands the meaning of that word. Second, if I doubted the fact that there is my hand in circumstances such as those of Moore’s proof – that is in circumstances in which the object is clearly in view, where one is cognitively lucid and so on – how could I possibly rely on my experience in order to rationally find out whether it is there after all? Therefore we can now better appreciate why we can’t be wrong with respect to Moore’s truisms or about the premises of his proof. This is not due to the peculiar sense of certainty a subject might have towards such things; or to his psychological incapability of entertaining doubts about them. Rather, it is due to the fact that when taken in those very circumstances, Moore’s propositions don’t have an empirical role, but a normative one. In the twofold sense that they are both linguistically paradigmatic judgements – that is to say, judgements that must be agreed on if the meaning of the words in them is to be determined – and epistemically paradigmatic ones.31 That is to say, judgements that can’t be called into question on the basis of contrary evidence, for they must stay put – however in context that might be – in order for a subject to be in a position rationally to gather any kind of evidence for other, genuinely empirical propositions. Hence, one can’t be wrong about them because of the grounding role they play, with respect to the very possibility of there being language games about empirical objects, where evidence for or against actual empirical propositions is collected.32 Now, the interesting point is that these are judgements and yet they have a normative role and are exempt from doubt, just like rules, that can’t be called into doubt, but only abandoned or revised. How can that be? Well, for Wittgenstein this is no great mystery. As he already put it in the Philosophical Investigations: If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgments. This seems to abolish logic, but does not do so. – It is one thing to describe methods of measurement, and another to obtain and state results of measurement. But what we call ‘measuring’ is partly determined by a certain constancy in results of measurement. (PI 242)
Belief, Knowledge and Certainty 81
So, ‘Here is my/a hand’, said in the circumstances of Moore’s proof, is a judgement; yet that there be agreement in so judging – that is to say, that all of us, who can speak English, judge that there is a hand there – contributes to the determination of the meaning of ‘hand’ in English. And in On Certainty Wittgenstein writes, Not only rules, but also examples are needed for establishing a practice. Our rules leave loop-holes open, and the practice has to speak for itself. (OC 139) We do not learn the practice of making empirical judgments by learning rules: we are taught judgments and their connexion with other judgments. A totality of judgments is made plausible to us. (OC 140) (…) ‘That I regard this proposition as certainly true also characterizes my interpretation of experience’. (OC 145) My judgments themselves characterize the way I judge, characterize the nature of judgment. (OC 149) How does someone judge which is his right and which is his left hand? How do I know that my judgment will agree with someone else’s? How do I know that this colour is blue? If I don’t trust myself here, why should I trust anyone else’s judgment? Is there a why? Must I not begin to trust somewhere? That is to say: somewhere I must begin with not-doubting; and that is not, so to speak, hasty but excusable: it is part of judging. (OC 150) I should like to say: Moore does not know what he asserts he knows, but it stands fast for him, as also for me; regarding it as absolutely solid is part of our method of doubt and inquiry. (OC 151) Quotations could be multiplied ad libitum.33 But the general thrust is clear enough already. To judge thus-and-so, in certain circumstances,
82
Moore and Wittgenstein
is said, in OC, to be part of ‘our method of doubt and enquiry’. Now, everything which is part of method does in fact belong to ‘grammar’ and ‘logic’ too, in the extended, not merely linguistic sense that those terms came to have for Wittgenstein at the time of On Certainty.34 To judge that there is a hand in the circumstances of Moore’s proof therefore belongs to the logic of our epistemic practices because it is what must stand fast if we want to test other things – for example, the reliability of our senses – which we do in turn need to trust in order to go about forming specific empirical judgements in circumstances where it is an open question whether what we see in front of us is really a hand. So, that judgement is itself part of logic and therefore comes to have a normative role, rather than a genuinely empirical one. Its staying put, or standing fast, is what in turn enables the possibility of really acquiring empirical evidence. It is not itself the statement of a rule like ‘If the traffic light is red, stop’, or ‘Patience is played alone’. But it nevertheless plays a normative role, rather than a descriptive one. It is perhaps also for this very reason that, as we have already seen,35 Wittgenstein is careful to warn us about the fact that the very notion of proposition ‘is not a sharp one’ (OC 320). But it also merits note that that very judgement is, in addition, constitutive of meaning, just like Wittgenstein held in PI. For if it were called into doubt that, in the circumstances of Moore’s proof, what is clearly visible to everyone is a hand, it would then be doubtful what we mean by ‘hand’. Now, it would not be enough to settle this issue simply to be given a definition, like ‘A hand is a part of a human body, which lies at the end of one’s arms and is constituted by five fingers, etc’. For we need also examples of application of that rule to fix the meaning of the word ‘hand’. So, ‘Here is a hand’ in the circumstances of Moore’s proof remains a judgement and not the statement of a rule. Yet, that very judgement plays a normative role because it contributes to the determination of the meaning of the word ‘hand’ and allows us, in its turn, to use that very object to give someone an ostensive definition of that term, which, in contrast, is an explicit formulation of a rule.36 Here, we may in fact see how close Wittgenstein gets to Kant’s idea of synthetic a priori judgements.37 That is to say, of judgements, not definitions or statements of rules, which, however, have a normative and therefore a priori status. Of course, there is nothing like Kant’s transcendentalism in Wittgenstein – that is to say, the appeal to universal and practice-transcendent categories and conditions of possibility of experience, such as space and time. There is, however, an even more explicit recognition of the normative and thus non-empirical role of
Belief, Knowledge and Certainty 83
propositions which are about physical objects and their properties, both for the determination of meaning and for the very possibility of our various epistemic practices. That is to say, for those practices in which we gather evidence for or against genuinely empirical, verifiable propositions or, if one so wishes, judgements, for which it does make sense to claim knowledge or to raise doubts. Which judgements and propositions do in fact play such a normative role is, however, entirely determined by practice, for Wittgenstein. That is to say, by how our language games are in fact played. Small surprise then if many commentators have found Wittgenstein’s form of ‘Kantianism’ to be much more ontologically respectable than the original one.38 3.3 What I can’t be wrong about depends on what I have been drilled and trained to Let us look at some of Wittgenstein’s examples: We teach a child ‘that is your hand’, not ‘that is perhaps (or ‘probably’) your hand’. That is how a child learns the innumerable language-games that are concerned with his hand. An investigation or question, ‘whether this is really a hand’ never occurs to him. Nor, on the other hand, does he learn that he knows that this is a hand. (OC 374) The appropriate behaviour of a subject who is learning the meaning of the word ‘hand’ by means of ostensive explanations is in many ways passive: there is simply no room, at that stage, for disagreement and criticism. As Newton Garver remarks, The most important thing to say about the response appropriate to grammatical remarks is negative: it is not appropriate to argue whether they are true or false. This does not mean that it is morally wrong to do so: there is no room for it in the language-game. If my nephew were to object that my remark about patience is false, he would thereby be rejecting my remark as a grammatical remark: he would implicitly be claiming that he knows full well how to use the word ‘patience’, and that it designates a range of card games which include some for two players. He would be challenging me all right; but by the very challenge he turns my remark from (as I thought) a grammatical remark into an empirical remark. (…) But if a remark is grammatical, its truth cannot be challenged.39
84
Moore and Wittgenstein
The child gets instructed about the correct use of the word ‘hand’ in English and the correct behaviour towards the ostensive definition of that term (like of any other) is not to call it into question or doubt it. The opposite behaviour would make it impossible to learn the meaning of that word and, connectedly, would make it impossible to participate in the various language games in which that word is used or that, at any rate, presuppose one’s total confidence in the fact that one has two hands and that those that one can see in front of oneself, by holding them up, are indeed one’s hands. Still, it is true that the child doesn’t learn that he knows that he has hands, or that those that someone is holding up in front of him are hands. We teach him the meaning of the word ‘hand’ and the fact that he has grasped it is shown in his subsequent use of that word and in the ability to explain its meaning to others. Moreover, apart from explicit explanations of the meaning of the word ‘hand’, the child is drilled to take part in language games, such as to obey commands, whose execution entails or depends on the use of his hands. But, then, does the child know that those are his hands? Wittgenstein’s answer is as follows: The child, I should like to say, learns to react in such-and-such a way; and in so reacting it doesn’t so far know anything. Knowing only begins at a later level. (OC 538) A dog might learn to run to N at the call ‘N’, and to M at the call ‘M’, – but would that mean he knows what these people are called? (OC 540) But is it wrong to say: ‘A child that has mastered a language-game must know certain things’? If instead of that, one said ‘must be able to do certain things’, that would be a pleonasm, yet this is just what I want to counter the first sentence with. – But ‘a child acquires a knowledge of natural history’. That presupposes that it can ask what such and such a plant is called. (OC 534) With respect to what we have been seeing so far, the order of priorities in the phase of one’s learning of language must be clarified. First of all,
Belief, Knowledge and Certainty 85
the child, like an animal, is drilled to react in various ways to given orders, or to specific situations. For instance, he gets drilled to react by giving his hand to his mother, when she tells him ‘Give me your hand’, before crossing a street. Then his linguistic competences evolve and he learns to answer the question ‘What’s the name of this (pointing to his hand)?’. At this stage the child masters the language game because he knows how to do several things. This way of putting things may sound pleonastic, however, because the language game is, by its very nature, an ensemble of linguistic and non-linguistic practices. Thus, to say ‘The child masters the language game’ is equivalent to saying, in a more humdrum way, ‘The child knows how to do this and that’.40 But, according to Wittgenstein, this kind of knowledge is different from knowledge tout court. Part of the difference can be expressed by saying that one is practical knowledge – a know how – while the other is propositional – a know that. The former, however, on Wittgenstein’s view, doesn’t exclude at least some linguistic abilities. He writes, A child can use the names of people long before he can say in any form whatever: ‘I know this one’s name; I don’t know that one’s yet’. (OC 543. Cf. 564–6) That is to say, the child may perfectly be able to use names, even proper names, well before possessing the concepts of being called thus-and-so and of knowledge. Therefore, he can use (proper) names without being able to say ‘I know that this (person) is called thus-and-so’. It is only when he becomes able to use such an expression that one can say that he knows that such an object, or person, is called thus-and-so.41 But, in this case, beside mastering the use of the concept of being called thus-and-so, he will have to be able to give reasons in favour of his claim that he knows that an object, or a person, has a given name. For it is only by being able to do so that he will show his mastery of the concept of knowledge. But, then, knowledge that this is my hand is acquired very early or very late? In some respects – that is to say, as far as one’s practical knowledge, which shows itself in one’s ability to participate in simple language games, is concerned – it is acquired very early. By contrast, one’s knowledge that this, standing in front of oneself, is one’s hand and that this object is called thus comes later. That is to say, when one is able to ask ‘What’s the name of this (showing one’s hand)?’ or to explain to someone else the meaning of that word, when they ask the same question.
86
Moore and Wittgenstein
If that is right, then, it shows that the ultimate justification of one’s claim to (grammatical) knowledge that that is a hand is simply given by one’s appeal to what one has been drilled or trained to, while learning one’s mother tongue. But it must be noticed that the role of such a justification is peculiar. For, strictly speaking, it doesn’t exhibit a reason that legitimates my claim to know that that is my hand. As we have seen in the first part of this chapter, reasons can be produced, for Wittgenstein, only within the epistemic language game. Rather, the kind of justification I mention in favour of my claim that I know that that is my hand exhibits the cause of my ability to participate in the various language games with the word ‘hand’, which may be epistemic or otherwise. Such a cause consists in the fact that I have been trained to use language a certain way and I therefore master the use of certain words.42 Hence, this once more shows how the grammatical use of ‘I know’ doesn’t conform to the criteria that govern the ordinary and genuinely epistemic use of those very words. To conclude, • The grammatical use of ‘I know’ in relation to propositions like Moore’s truisms is not based on criteria; • thus, it doesn’t express knowledge at all, but objective certainty. • Hence, far from stating an epistemic relation occurring between a subject and a proposition or a fact, it means ‘Here I can’t be wrong’ or ‘Here a mistake/doubt is logically impossible’, or, even, ‘I could not admit any experience as proof to the contrary’ (§3.2). • The fact that I can’t be wrong in some particular cases doesn’t depend on my privileged epistemic status (§3.1), but, rather, on the role that the propositions that I claim (grammatically) to know play in our language games. • These very propositions – like ‘This is my hand’, ‘There are physical objects’, ‘I am a human being’, ‘The earth is very old’, etc. – have a normative function, even when they are the content of a judgement and not the explicit statement of a rule (§3.2). But, of course, they can also, in context, be used as explicit formulations of rules, for instance when we use ‘This is a hand’ as an ostensive definition of the meaning of the word ‘hand’ in English. In either case, they contribute to the determination of meaning and to the possibility of acquiring evidence for or against genuinely empirical – that is verifiable – propositions and, thus, to the very possibility of our epistemic practices. That is why they play both a linguistically and an epistemically normative role.
Belief, Knowledge and Certainty 87
• The way in which we come to have them depends on what we have been drilled to and taught, either implicitly or explicitly, in the process of being drilled to and taught language, or to take part in our various epistemic practices. That is why, ‘I know’, in connection with them can also be glossed as ‘I have been drilled to/taught thusand-so’ (§3.3). • This, however, exhibits the cause of our mastery of the use of a given word and not an epistemic reason which could sustain a genuine claim to knowledge (§3.3). 3.4
Nonsense and the grammatical use of ‘I know’
According to Wittgenstein, the grammatical use of ‘I know’ in relation to Moore’s truisms and the like isn’t based on the criteria that govern the ordinary use of ‘I know’. Therefore it is senseless. As a corrective to the tendency, displayed for instance by Moore, of thinking that ‘I know’ is being used appropriately in those contexts, he proposes to substitute it with the expressions ‘It is certain’ or ‘It stands fast’. What is (objectively) certain, or stands fast, holds for me as well as for all others and is a function of the role it plays within our language and epistemic practices.43 Hence, the grammatical use of ‘I know’, which doesn’t conform to the ordinary use of that expression explored in the first part of this chapter, is – as I have said – senseless (or sinnlos), but it is not nonsensical (unsinnig) – to resume the well-known distinction already present in the Tractatus. It would be nonsense if it were meant to be a move within the usual language game with ‘I know’. In such a case, it would qualify as incorrect by the standards of our ordinary language game with that expression, since it would appear like a description of what would seem to be an incontrovertible fact, known to all that are party to the language game. In short, it would be nonsense if it were taken to mean what, traditionally, philosophers (of a realist persuasion) have taken it to mean before On Certainty: the expression of a genuinely epistemic relation between a subject and an incontrovertible fact. But that is not what the grammatical use of ‘I know’ is meant for, according to Wittgenstein. For, it is meant, in context, to remind us of the fact that there are propositions we can’t sensibly doubt, on pain of no longer being competent speakers of English, or practitioners of specific epistemic language games. Whenever ‘I know’ plays that role, it can be replaced by different expressions, as we have seen in the second part of this chapter. This shows that the use of ‘to know’ isn’t essential, because it is not knowledge what those words are meant to convey. What they are meant to express, rather, is objective certainty which is
88
Moore and Wittgenstein
tantamount to the logical/grammatical impossibility of being mistaken or to raise doubts. Insofar as the proper – call it ‘grammatical’ – use of ‘I know’ with respect to them is understood, it is not nonsense, but, as we have seen, an expression of certainty and of the logical/grammatical fact that a given proposition lies beyond doubt, just like ‘I know what I am feeling’, or ‘I know what I intend’, when one is experiencing great pain, or has formed an intention, in Wittgenstein’s view of the matter.44 ‘I know what I am feeling’ would in fact be a very useful reminder against any annoying interlocutor who, our great pain and display of characteristic symptoms notwithstanding, tried to question the reality or intensity of our current inner experience. Similarly, it would be helpful against a possible sceptic about our own minds, if such a sceptic ever existed. For it would remind him that ‘I am in pain’ is not an empirical proposition – a description of a state of affairs – based on grounds, and that could thus be wrong or challenged. But, rather, an avowal – viz. an immediate expression of one’s ongoing pain – which makes no sense to doubt – whence the authority we accord subjects over it – just as it wouldn’t make sense to doubt the fact that an instinctive cry were caused by pain, when we saw someone being injured. By analogy, ‘I know here is my hand’, meant as a grammatical remark, would be valuable both against a weird interlocutor – perhaps a child – who went on raising idle doubts with respect to the fact that there really is a hand there or that that is my hand, as well as against those philosophers, such as a sceptic or Moore, who either tried to doubt what can’t sensibly be called into question, or who, realizing such an impossibility, concluded that our relation to these truisms is a genuine and certain form of knowledge. In this latter case, pointing them out that ‘I know’ is meant grammatically and not empirically and does not therefore express knowledge, but a certainty which is a function of the fact that our language game excludes from doubt and error certain judgments, would have – clearly – a highly therapeutic effect. Now, after the raise of so-called therapeutic or resolute readings of the Tractatus, it has become common place to admit only of one notion of nonsense and to take ‘sinnlos’ and ‘unsinning’ as in fact interchangeable. I think that this has only caused confusion. Clearly a problem with keeping that distinction after the Tractatus has to do with the fact that in that work tautologies and contradictions (and therefore all propositions of logic) were said to be sinnlos because, contrary to ordinary empirical ones, they could not be bipolar and had to occupy only one of the two poles – either truth or falsity. Grammatical propositions, in contrast, are neither true nor false, for Wittgenstein. Moreover, many of
Belief, Knowledge and Certainty 89
what in the Tractatus were considered unsinnig combinations of signs, viz. philosophical propositions such as ‘Every effect has a cause’, were later considered as grammatical. That would point in the direction of equating grammatical propositions with nonsensical ones. But, contrary to such a tendency, I think it is important to note that although grammatical statements aren’t moves within the language game, they can have a point – in fact a use – which is either that of instructing someone about the use of a word, or about the rules of one of our epistemic practices; or of reminding someone – perhaps a philosopher – of their proper nature and role, against a tendency to take them for other than what they are.45 Compare with the rules of chess: they aren’t moves within the game, but what determines permissible and impermissible moves within the game, which indeed bear saying and, in fact, need being said either to instruct one about how to play the game or to remind a player of them, when he displays a perhaps unintentional tendency to violate them. Indeed, as already remarked, they figure in normative language games, though not in descriptive or epistemic ones. So, statements of rules, or grammatical remarks, are not absurd combinations of sounds like ‘Ab sur um’ or even, appearances notwithstanding, like ‘I know this is my hand’, said by Moore. One might of course say that there are two kinds of nonsense, here. However, I prefer to avoid such a bifurcation and reserve ‘nonsense’ for nonsense – for what is unsinnig because it is a mistaken combination of words since it does not have a use not even as an instruction or as a reminder of how symbols are used in our language – and ‘senseless’ for those grammatical propositions, which, like tautologies and contradictions in the Tractatus, are said by Wittgenstein to belong to the ‘scaffolding’ of (the world and therefore of) thought (Cf. TLP 6.124 and OC 211).46 That is to say, for those propositions which, while not being moves within the language game (compare the logical space of the Tractatus), nevertheless lie at its limits and make it possible, by determining what moves are permitted or forbidden within it. Again, the distinction between what lies within a language game and what constitutes the limits of thought and language is reminiscent of Kant. But I think it is only to be acknowledged that there is a lot in Wittgenstein’s entire philosophical production that is much more in consonance with Kant than with empiricism. The main difference between Kant’s and the later Wittgenstein’s positions, as already remarked, is that while for the former the conditions of possibility of thought are given transcendentally – that is to say, independently of human practice – for the latter aren’t. So, for the later Wittgenstein the conditions of possibility of thought are not determined prior to the actual
90
Moore and Wittgenstein
development of language games and independently of our actual form of life. This, however, doesn’t stand in the way of recognizing them ex post. After all, for Wittgenstein, this is just the aim of philosophical activity, when properly conducted.
4 Wittgenstein and the ‘assertion fallacy’ In Speech Acts (henceforth, SA) John Searle accuses Wittgenstein, as well as other philosophers, of making what he dubs ‘the assertion fallacy’. This fallacy, according to Searle, consists in ‘confusing the conditions for the performance of the speech act of assertion with the analysis of the meaning of particular words occurring in certain assertions’.47 According to him, such a mistake has a twofold origin: on the one hand, the ‘failure to base particular linguistic analyses on any coherent general approach to or theory of language’;48 on the other hand, the adherence to the slogan ‘meaning is use’, which constitutes the only methodological perspective in absence of a more general theory of language.49 In particular, as far as Wittgenstein is concerned, Searle takes into account Philosophical Investigations I, 246: In what sense are my sensations private? – Well, only I can know whether I am really in pain; another person can only surmise it. – In one way this is wrong, and in another nonsense. If we are using the word ‘to know’ as it is normally used (and how else are we to use it?), then other people very often know when I am in pain. – Yes, but all the same not with the certainty with which I know it myself! – It can’t be said of me at all (except perhaps as a joke) that I know I am in pain. What is it supposed to mean – except perhaps that I am in pain? Other people cannot be said to learn of my sensations only from my behavior – for I cannot be said to learn of them. 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. Wittgenstein’s mistake, for Searle, would be to hold that the meaning of an expression is given by its use and therefore ask when, for instance, we would make assertions such as ‘I know I am in pain’. Then, according to Searle: The philosopher notices that it would be very odd or bizarre to say certain things in certain situations; so he then concludes for that
Belief, Knowledge and Certainty 91
reason that certain concepts are inapplicable to such situations. For example, Wittgenstein [PI I, 246] points out that under normal conditions, when I am in pain, it would be odd to say, ‘I know I am in pain’. (…) But then [he] conclude[s] that these are points about the [concept] of knowing (…); that [this concept is] only applicable under certain conditions.50 According to him, Wittgenstein, by noticing that in normal circumstances it would be very odd to assert ‘I know I am in pain’, while it would be appropriate to assert it in aberrant ones, infers that the latter are the conditions of applicability of the concept of knowledge and that, for such a reason, they must figure in the analysis of that concept.51 In order to counter what Searle takes to be Wittgenstein’s argument, he considers the following two sentences: ‘He’s breathing’ and ‘He has five fingers’. According to Searle, it would be weird to assert them in normal circumstances, while it would be appropriate to do so in aberrant ones. Yet, it wouldn’t be very plausible to say that the oddity of these assertions in normal circumstances would be due to the concepts that figure in them, viz. the concept of fingers and of breathing. Therefore, Searle aims to offer a more general explanation of the reasons why it would be odd to make such assertions in normal circumstances. It goes as follows. There are standard or normal situations. People normally remember their own names, know whether or not they are in pain (…), breathe and have five fingers per hand. In general, it is inappropriate to assert of a particular standard or normal situation that it is standard or normal unless there is some reason for supposing (…) that it might have been non standard or abnormal. The explanation, then, has nothing to do with the analysis of particular words; it lies in explaining what it is to make an assertion. The assertion – for example, that I remember my own name – is just pointless unless the context warrants it in some way. But that pointlessness has nothing to do with the concept of remembering but with the concept of what it is to make an assertion. The general character of the assertion fallacy, then, is to confuse the conditions for making non-defective assertions with the conditions of applicability of certain concepts.52 According to Searle, then, the oddity of an assertion depends on the violation of those general criteria that govern the illocutionary act of
92
Moore and Wittgenstein
making assertions and that the general study of speech acts has brought to light. The general form of such criteria is: if one doesn’t want the assertion of a proposition ‘p’ to be defective, then ‘p’ must not be too obvious for both the speaker and the listener.53 Searle adds three further objections against those who hold that from the oddity of an assertion it follows that that assertion is nonsensical. The first one is, ‘He has five fingers’ and ‘I know I am in pain’ are false in normal circumstances, otherwise the conditions identified as normal would be abnormal. Hence ‘He doesn’t have five fingers’ and ‘I don’t know I am in pain’ are true in normal circumstances, since the negation of a false proposition is true. Thus, according to Searle, those who maintain that the assertion of these sentences as well as of their negation is nonsensical – and therefore neither true nor false – are wrong. For otherwise we could no longer distinguish between normal and abnormal conditions, while, such a distinction, in Searle’s view, plays a crucial role in determining the conditions under which an assertion is correctly executed. The second objection, in contrast, takes into account the fact that, in more complex sentences or in ones containing the verb ‘to know’ in forms other than the first person present, such as ‘I knew I was in pain, yet I wanted to finish the race’, the concepts that in the simpler statements are said to be employed in a nonsensical way seem to be used in a perfectly kosher way. Hence, if an analysis of the meaning of a word (or morpheme), in order to be adequate, ‘must be consistent with the fact that the same word (or morpheme) can mean the same thing in all the grammatically different kinds of sentences in which it can occur’,54 the alleged conditions of application individuated by considering the simpler contexts of use can’t be taken to be constitutive of the meaning of the word in question. For, otherwise, it wouldn’t be possible to apply it correctly in contexts which violate the conditions of applicability thus individuated. Finally, Searle reformulates the issue of the assertion fallacy as follows: The character of the mistake I am citing is that it confuses conditions of assertibility with presuppositions of concepts. Most concepts do indeed have presuppositions which determine the scope of their intelligible applicability. (…) But the fact that [an] assertion is odd except in abnormal or aberrant situations is not sufficient to show that aberrance or abnormality is a presupposition of the applicability of the concept (…).55
Belief, Knowledge and Certainty 93
This formulation differs from the first one proposed by Searle because it recognizes as integral to the meaning of an expression the presuppositions (criteria) of its application, which did not figure among Searle’s more general exposition of the assertion fallacy. I now wish to assess whether Wittgenstein is really guilty of the assertion fallacy, as far as the analysis of expressions such as ‘I know I am in pain’ is concerned; whether he is making such a mistake when he is considering the use of ‘I know’ in relation to Moore’s truisms; and, finally, to compare Wittgenstein’s observations in On Certainty with Searle’s counter to the problem of the assertion fallacy in general. Wittgenstein’s position about ‘I know I am in pain’ is roughly as follows:56 in many cases this statement doesn’t say anything more than ‘I am in pain’, because ‘I know’ is applied on the basis of no evidence: we can’t either observe our own mental states or infer them from our behaviour. Of course my assertion ‘I am in pain’ is a criterion for others to say that I am in pain or that they know I am. But this can’t be a criterion for me, as it would imply the implausible view that in order to know what I am feeling I have to wait for my pronouncements. Hence, on the one hand, we take occurrent mental states such as ongoing sensations to be transparent to their owners. That is to say, occurrent sensations are, by their very nature, something which we are aware of and, given our conceptual mastery, something we may immediately express in the form of an avowal. On the other hand, it belongs to the grammar of ‘I am pain’ that, if used by a competent and sincere speaker, those words are accorded a distinctive authority: if subjects say they are in pain, we take it for granted that they are. For just as we wouldn’t doubt that someone who screams and moans after having been injured is in pain, so we don’t doubt that the assertion ‘I am in pain’, made in similar circumstances, should be taken at face value.57 This exemption from doubt is accorded by the language game to subject’s pronouncements over their own mental states and is not at all the result of some especially secure cognitive achievement on a subject’s part. An important cue that so-called self-knowledge is no real knowledge at all but, rather, a feature of grammar, is the fact that questions such as ‘How do you know you are in pain?’ or ‘Are you sure you are in pain?’ wouldn’t find any other answer but ‘Of course I do/am – I am in pain!’. This kind of reply shows how, in the attempt to give one’s grounds for one’s claim, one couldn’t but repeat it. Thus, one’s original avowal wasn’t based on reasons and grounds and that makes it nonsense to say ‘I know I am in pain’, if that were meant as a genuinely epistemic statement. Neither claims to knowledge nor doubts find a place in the language game with
94
Moore and Wittgenstein
avowals of sensations and this is simply a function of how our language game of making psychological avowals is played. So, ‘I know I am in pain’ is either nonsense – if it were meant as a genuine assertion of knowledge which could allegedly respond to some sensible doubt about the fact that I am in pain58 – or else, it is a mere grammatical statement. Just as the reminder ‘Patience is played alone’,59 so ‘I know I am in pain’ can be meant as a reminder of the fact that in most circumstances a subject’s sincere avowal can’t be doubted, or called into question. One brief detour is necessary though. I have stressed how for Wittgenstein ‘I am in pain’ is authoritative in most cases, from which it follows that ‘I know I am in pain’ could only be nonsense or, alternatively, a grammatical remark, in those very circumstances. Yet in some cases the former may not be authoritative and the latter may not be nonsense. This would be so if somehow the former statement was based on criteria and was thus a genuine empirical statement, rather than an avowal – viz. a mere expression of one’s ongoing pain – and if the latter was then, in its turn, a genuinely empirical statement. We could in fact imagine cases in which one might realize of being in pain – perhaps more easily in cases of psychological rather than physical distress – by observing and reflecting on one’s own behaviour. In that case, then, there would be no difference between first-person and third-person attributions of pain and, as a consequence, first-personal claims to knowledge, as well as doubts, would be in order.60 But let us go back to the majority of cases in which ‘I am in pain’ is said on the basis of no criteria and is thus, according to Wittgenstein, an avowal and not a genuinely assertoric statement. In all these cases, in Wittgenstein’s view, ‘I know I am in pain’ wouldn’t be an odd assertion, but, rather, an altogether nonsensical one. It would be so not because of reasons which pertain to the analysis of speech acts or to the pragmatics of conversation, but because those who said those words would only suppose to be using terms endowed with meaning when things are otherwise. The reason for such a startling claim is this. If only words which belong to a language game – that is to say, that have a stable use – are meaningful, since ‘I know I am in pain’ (in the case we are considering) would be no move in the language game, it would have no meaning either, contrary appearances notwithstanding. That is to say, we shouldn’t be fooled by the fact that we do understand the single words figuring in that sentence and by the fact that they have been put together in a syntactically correct way into thinking that taken together they do have a meaning – viz. a use. As we have seen, there is
Belief, Knowledge and Certainty 95
no room in the language game with avowals of sensations for asking/ giving grounds of one’s statements and this in its turn entails the fact that there is no room either for genuine claims to knowledge on a subject’s part or for real doubts from a third party. We have seen how, for Wittgenstein, in other circumstances, things may be different, but this doesn’t change the fact that in most cases ‘I am in pain’ is an avowal and not a description and ‘I know I am in pain’ nonsense.61 But, of course, for Wittgenstein, all these remarks aren’t simply observations on the use of ‘I am in pain’ and of ‘I know I am in pain’. They are always therapeutically relevant against a specific ‘disease’: the typically philosophical attitude of taking certain forms of expressions, such as ‘I (know I) am in pain’, to express super facts. For philosophers have long thought that from the fact that those words lie beyond doubt, in most cases, as we have just seen, it follows a kind of sure-fire knowledge with respect to the fact they would allegedly describe. Accordingly, they have long sought after a plausible explanation of a such a special kind of knowledge,inventing, for instance, the mythology of a privileged and extremely secure access to one’s own mental states, conceived of, in their turn, as objects luminously presented in one’s own mental arena. Furthermore, they have taken these facts to imply that the only certain knowledge, or perhaps the only knowledge we really have, is the one about our own mental states. Wittgenstein’s cure, against such an amount of metaphysics, is, as we have seen, to remind us first that ‘I am in pain’ is, in most cases, applied without criteria and is therefore no genuine empirical statement, but a mere avowal – a simple expression of one’s own occurrent pain, that we have learnt to voice in substitution of some form of instinctive behaviour, such as moaning or crying. Second, that for this very reason, ‘I know I am in pain’ isn’t a move in the language game with avowals of sensations and is thus nonsense. Insofar as philosophers take it to express a specially secure epistemic fact, then, they themselves are simply speaking nonsense. Yet, that expression may also be meant as a simple grammatical remark. Grammatical remarks aren’t moves within (descriptive or epistemic) language games, but only ways of making explicit their underlying rules: and it is a rule of our language game with avowals of sensations that they are the immediate expression of an on-going sensation and that we accord subjects authority over their own pronouncements. We do so, moreover, not because we grant them a private and infallible kind of cognitive contact with their own mental states, which should allegedly make it plausible to speak of their relation to their own mental states both as a genuine kind of knowledge and as particularly secure; but, simply, because our
96
Moore and Wittgenstein
language game is played thus. This is where we reach ‘the bedrock’ and our ‘spade is turned’ (PI I, 217) – that is to say, where Wittgenstein’s socalled quietism shows itself. As far as Wittgenstein’s observations about the use of ‘to know’ and ‘I know’ in relation to Moore’s truisms and alike are concerned, I already stressed in section 2 how, by observing the usual language game with those expressions, Wittgenstein individuated a number of criteria of application of the concept of knowledge, which aren’t operative when ‘to know’ and ‘I know’ are used in connection with propositions à la Moore. From this it follows, in parallel with the case of ‘I know I am in pain’, that ‘I know this is my hand’, for instance, is nonsense and therefore oddly asserted. Hence, also in this case, the oddity of the assertion isn’t a sufficient condition, for Wittgenstein, to say that those words are nonsensical. Nor does saying that an assertion is odd explain why it is. The fact that we find an assertion odd is merely a cue, in Wittgenstein’s overall strategy, that prompts us to look for the reasons why we form that impression. By carefully examining the language game with ‘to know’ and especially its use in the first person present, Wittgenstein came to realize that when taken either in relation with avowals of sensations, or with truisms such as Moore’s, that impression depended on a misuse and abuse of language – in short, that it depended on the nonsensical application of the concept of knowledge in those contexts. Just as he did notice in the Philosophical Investigations that in some contexts ‘I know I am in pain’ is a move in a descriptive language game and is therefore endowed with meaning, so he did stress in On Certainty, that in some contexts ‘I know this is my hand’ does make sense. This, however, is neither the case when a solipsist, say, asserts ‘(Only) I do know what I am feeling’, nor when Moore says ‘I know this is my hand’. Again, the explanation is similar in both cases: their assertions are groundless and therefore what they mistake as an indubitable empirical fact – that in some cases we allegedly have infallible knowledge – is a mere product of how our language game is played: we accord subjects authority over their own occurrent sensations and over judgements such as ‘Here is my hand’, when made in the relevant contexts. Thus, ‘I know I am in pain’ or ‘I know that here is my hand’ can only be taken as statements of rules of grammar: ways of making explicit the fact that our language games are like that – that in both cases subjects are endowed with authority over their pronouncements, or, equivalently, that there is no room for doubt in such cases. What is new in On Certainty, therefore, isn’t so much the host of considerations about the use of ‘I know’, which was already largely present in the Philosophical Investigations; rather,
Belief, Knowledge and Certainty 97
it is their extension to the case in which that expression is applied in connection with propositions not about one’s own mental states but about physical objects, as well as the kind of explanation provided of such an analogy. Thus, contrary to Searle’s first objection, I think it clear that, for Wittgenstein, it isn’t the infelicity of the speech act, which manifests itself in the impression of oddity it creates, that makes the statement nonsensical. Rather, it is the other way round: the use of a word irrespective of the criteria that govern the language game and confer that word its meaning makes its application nonsensical and its employment sound odd. For, albeit implicitly, we sense that no move within the language game has been made with it at all. Hence, Wittgenstein can’t be guilty of the assertion fallacy for, in his view, it isn’t the felicity of the execution of the ‘speech act’ of making an assertion that renders the application of those words meaningful; but vice versa: it is the meaningful employment of words – which takes place when it conforms to the criteria that rule the language game – that makes the ‘speech acts’ thereby executed felicitous. To put it otherwise, it is only within the language game, which is the empirical datum we can’t change or explain, that words find what we call their meaningful employment. The language game is implicitly governed by rules, following which guarantees the felicity of the ‘speech act’ thereby executed. Moreover, as we have seen, the fact that in most cases a present-tense first-personal psychological statement is an avowal, which can’t, therefore, be the object of a genuine first-person claim to knowledge, doesn’t preclude the fact that in different contexts things may be otherwise; or that past-tense psychological statements be descriptions, based on memory and personal recollections, which can sensibly be prefaced by ‘I know’. So the general condition of adequacy of any analysis of meaning identified by Searle – viz. that an expression should always have the same meaning in any context of use – is spurious by Wittgensteinian lights. It would depend on abstracting away from what gives signs their meaning, as if they could carry it on their sleeves, independently of the actual features of their use. Finally, the reader will recall Moore’s reply to Malcolm, mentioned in Malcolm [1977], that resembles closely Searle’s way of characterizing what he calls ‘assertion fallacy’ and what the latter considers the right solution to the problem of the oddity of, for instance, the assertion ‘I know I am pain’. To sum up that discussion and compare it with Wittgenstein’s position, we could say that, for both Moore and Searle, the starting point is that there are sentences, which, as such, that is to
98
Moore and Wittgenstein
say, outside of any specific context of use, are meaningful, insofar as they contain meaningful words assembled in a syntactically correct way. Being meaningful, they are semantically evaluable and can be used to make a speech act, whose felicity conditions are in turn independently fixed. Let us suppose that the speech act one wants to make is an assertion. Among other things, it can be maintained that in order for that speech act to be correctly made, the information conveyed ought not to be too obvious (in Gricean terms, one could say that it should be relevant) for the parties involved in the communicative exchange. Obviously, that would depend on further contextual features, which need not be specified here. Given all this, both Moore and Searle will agree that in normal circumstances the assertion ‘I know this is my hand’ (or ‘I know I am in pain’) is infelicitous because it is too obvious (Searle) and therefore useless (Moore). But, Moore will also add that, in the context of his utterance of that sentence, that very utterance wasn’t either too obvious or useless, for it was meant to oppose both sceptical and idealist theses. Furthermore, both Searle and Moore will agree that the sentence in question is, as such, meaningful and either true or false. Hence, if the analysis of its meaning doesn’t depend on the observation of the speech acts that are made by using a given sentence, it cannot be concluded that, on a given occasion of its use, which may violate the norms for the correct execution of the speech act in question (viz. the assertion), that sentence is, after all, meaningless. This is so because, by varying the conditions of execution of that very speech act, the assertion of the sentence would be perfectly in order (as Wittgenstein himself acknowledges). Consequently, the expression ‘I know’ should mean something different in this latter case. This, however, would be in contrast with the condition of adequacy of any analysis of meaning individuated by Searle, which, as we saw, consists in the fact that any such theory should respect the fact that the same word can mean the same in all kinds of phrase, albeit grammatically different, in which it can occur.62 Therefore, the assertion of, for instance, the sentence ‘I know this is my hand’ can never be nonsensical, but, at most, simply odd. This, in turn, would never be due to an abuse of its sense – that is to say, to a semantic impropriety – but only to the inadequacy of the contextual factors in which it is used and, thus, merely to pragmatic reasons. One thing should be obvious though: Searle’s (and Moore’s) analysis of the oddity of the assertion ‘I know this is my hand’ (or ‘I know I am in pain’) doesn’t explain in the least why that statement sounds like a hopeless response to philosophical scepticism. For, after all, we don’t merely find it an odd assertion, but, in context, a pointless move
Belief, Knowledge and Certainty 99
against external world scepticism. Hence, without going any further in the discussion of speech acts’ theory, it is quite clear that down that route there is no hope of finding some philosophically illuminating explanation of either Moore’s or the sceptical mistake. Wittgenstein’s analysis, in contrast, differs from the start because it considers the clarification or, in fact, the dissolution of the philosophical problem the end to which the conceptual elucidation should aim. The first move always consists in the observation and description of the relevant language games. They will therefore appear to be governed, albeit implicitly, by certain rules, which will demarcate meaningful from meaningless employments of the concepts at issue. It will then be shown how the formulation of certain philosophical theses doesn’t conform to the rules which govern the language games with those very concepts. By so doing, the analysis will expose, on the one hand, the fact that philosophers often think of being stating some very important and deep truth, while they are in fact speaking nonsense. On the other, it will also be shown how the allegedly metaphysical truth stated by philosophers does in fact depend on taking a grammatical fact – like our authority over our own mental states, or the immunity to doubt of certain judgements about objects in specific contexts of their use – for a factual truth, characterized by some extraordinary degree of certainty or even by necessity. Once the grammatical nature of certain statements, like ‘I know what I am feeling’ or ‘I know this is my hand’ is exposed, it thus appears obvious that what is nonsense aren’t those statements when correctly taken for what they are – viz. norms – but, rather, their philosophical misinterpretation. That is to say, to repeat, the fact of considering them as expressing factual truths, marked by a special kind of certainty or even by necessity. After all, Wittgenstein never wanted to deny or reform our actual use of language and it is a fact that sometimes, when in fact there are no real doubts or questions at stake, we do say ‘I know what I am feeling’ or ‘I know this is my hand’. When do we do such a thing? Well, when someone is in fact probing us with non-pertinent questions. When we do reply with those words, what we are doing is – whether we acknowledge it or not – to remind them of the grammar of the relevant concepts. Hence, we remind them that we accord sincere and competent speakers authority over their own psychological avowals and that doubts about judgements about empirical objects, in certain contexts, are simply not up for grabs.63 The point of Wittgenstein’s analysis is to make us – philosophers – aware of the real function of those remarks, and thus contrast the misleading interpretation that – as philosophers – we are prone to give of them. By so doing, that is to say,
100
Moore and Wittgenstein
by seeing the function of those words aright, the alleged philosophical problems of our knowledge of our own minds, of common sense certainty and, correlatively, of external world scepticism, should be dissolved. It is only a by-product of such a dissolution that the oddity of the original philosophical claims will in turn be explained.
5 Coda. Wittgenstein and semantic contextualism In my interpretation of Wittgenstein’s criticism of Moore’s use of ‘I know’ in relation to his truisms and the premises of his proof, and in my defence of his position against Searle’s allegations, I have greatly relied on the decisive role of use in the determination of meaning. I have therefore come very close to so-called therapeutic interpretations of his thought, according to which Moore, and more generally philosophers, often tend to fall prey to an illusion of meaning: they think they are making sense, when they employ certain words, and that they are therefore discovering deep philosophical truths (or problems), while they are in fact speaking nonsense. I don’t wish to retract any of that here, although I think therapeutic interpretations are too partial. For they fail to acknowledge the positive and purposive elements in Wittgenstein’s overall philosophy and particularly in On Certainty. More specifically, they fail to see that while use comes first and determines meaning, it does so, for Wittgenstein, because it fixes the rules for the correct employment of signs and for the actual deployment of our epistemic practices. So there is no opposition, in my view, between stressing the importance and primacy of use in the later Wittgenstein and his insistence on rules and grammar. In my view, this goes together with the fact that Wittgenstein was primarily engaged in conceptual clarification, rather than in simply exposing nonsense and that, to such an end, he too put forward some positive suggestions – in the form of grammatical remarks. Be that as it may, here I do wish to signal my distance from semantic contextualist readings of, in particular, On Certainty and of Wittgenstein’s observations on Moore’s use of ‘I know’. This is the right place to take up this matter because knowledge claims have been the main topic of this chapter. Moreover, I think it appropriate to address this issue now because semantic contextualists do join ‘therapists’ and myself in stressing the fact that Wittgenstein’s insistence on use would have deep philosophical consequences. In particular, it would entail that most philosophical employments of, for example, ‘I know that p’ would actually fail to make sense.64 This, however, may suggest that
Belief, Knowledge and Certainty 101
there are more similarities between semantic contextualist readings of Wittgenstein and the present one, than it is in fact the case. Charles Travis, in his The Uses of Sense (1989), has argued that in On Certainty Wittgenstein made out that ‘A knows that p’ can express many different thoughts – viz. various truth-conditions – on different contexts of its use. He draws most of his inspiration, however, from one passage of OC where Wittgenstein is in fact discussing an altogether different example, that is, ‘I am here’. Just as the words ‘I am here’ have a meaning only in certain contexts, and not when I say them to someone who is sitting in front of me and sees me clearly, – and not because they are superfluous, but because their meaning is not determined by the situation, yet stands in need of such a determination. (OC 348) so too, the comparison goes, do the words ‘I know that that’s a tree’ uttered when one is clearly in view of a tree (OC 347). Travis’ idea is that, unless we look at context, the truth-conditions of a given utterance of those words are underdetermined. Hence, it is only by considering the context that we can actually find out what thought they express and establish whether they are used to express a truth or a falsity. I find this suggestion odd for a number of reasons. First of all, I think Wittgenstein’s view is more radical than Travis makes of it. For in OC 348 Wittgenstein is saying that outside a specific context of use, those words have no meaning. I take that to entail that, using a piece of classical semanticist terminology, they do not even have a role or character. Therefore, they naturally fail to determine a thought, that is, specific truth-conditions. Travis, on the contrary, seems to think that they may actually have the former, while the context has to be invoked in order for them to come to express the latter, because, as a matter of fact, those truth-conditions are context-sensitive (just like the ones of utterances of sentences containing indexical elements). Second, it seems to me that the only kind of ‘contextualism’ envisaged by Wittgenstein is about the fact that in different contexts ‘I/A know(s) that p’ could actually mean different things. In particular, either an epistemic relationship between the subject and ‘p’; or else, a grammatical remark where no epistemic property is attributed to a subject. Finally, those words could fail to mean anything at all, when they are in fact used by philosophers such as Moore, irrespective of all criteria that govern their mundane employment. Only in the first case could those
102
Moore and Wittgenstein
words express a Fregean thought, that is, something which is susceptible of being either true or false. For in the second case, the role of the ascription would be that of expressing a rule, hence nothing which could either be true or false; and, in the third, those words would fail to have a meaning altogether and would then express no Fregean thought at all. Third, contrary to Travis,65 I find no suggestion in Wittgenstein that, depending on various factors, such as our practical interests and worldinvolving conditions, an utterance of the first kind could actually express different truth-conditions. Let us illustrate this idea with an example, which has by now become a classic in the literature on the topic: there is a subject who needs to deposit a cheque in the bank. Now, depending on whether it is a matter of urgency or not and, therefore, on whether certain defeaters are salient to the case, an ascription of knowledge to him (either made by the subject himself or by a third party), in those very circumstances, could either be true or false. However, as remarked, given the textual evidence at our disposal, I find no hints of this kind in On Certainty. So, I actually think that, if it makes sense to describe Wittgenstein’s position in these anachronistic terms, he was rather an invariantist with respect to knowledge ascriptions. Finally, one may think that though a semantic invariantist with respect to knowledge ascriptions, Wittgenstein was after all an epistemic contextualist. For he did point out that, for example, ‘Here is my hand’ could be the object of a genuine claim to knowledge in some cases and could thus be supported by grounds, while it would fail to be so in different ones. Yet, I find this suggestion too quite misleading. For, in Wittgenstein’s view, in the latter case ‘Here is my hand’ would be a ‘hinge’ and hence, as we shall see at much greater length in Chapter 4, a rule. If so, it would simply fail to be in the business of epistemic justification and assessment.66 Therefore, at the end of the day, Wittgenstein did not even argue that the structure of justification varies depending on context. Rather, he argued for the view that, in certain cases, propositions which, in different circumstances, may really be subject to verification and control, would fail to be so, as their status and role would be different in those different contexts. To repeat, in some contexts they would not be empirical propositions but rules and would thus be unsuitable for epistemic support and appraisal.67
3 Wittgenstein: Doubts and the Nonsense of Scepticism
In this chapter I first present Wittgenstein’s description of our language game with ‘to doubt’. The criteria for the correct employment of ‘I doubt’ thus identified turn out to be systematically violated in the sceptic’s use of that expression. Hence, for Wittgenstein, sceptical as well as idealist doubts are nonsensical. I then turn to two classical arguments in favour of scepticism about the external world – viz. Descartes’ argument from dreaming and the Humean argument against the general reliability of the senses. I consider what I take to be Wittgenstein’s quite different answers against them, which I propose to dub ‘the linguistic argument against Cartesian scepticism’ and ‘the transcendental argument against Humean scepticism’ respectively. Finally, I look at two influential readings of Wittgenstein’s last argument – Crispin Wright’s and Michael Williams’ – and show why they don’t tally with the text and actually depend on failing to appreciate adequately Wittgenstein’s radical anti-epistemicism. Yet, it remains that the two broad antisceptical arguments which can be found in On Certainty deliver two different verdicts as to why scepticism would be nonsense. According to the latter, scepticism is nonsense because it is irrational, as it cannot be supported by reasons and actually depends on a misunderstanding of the conditions of possibility of there being epistemic justifications for or against any empirical claim. In contrast, according to the former, scepticism is nonsense because it violates those criteria which make a specific utterance of ‘I may be dreaming right now’ meaningful. Hence, it is nonsensical in a much more radical way: because the sceptical hypothesis of a lucid and sustained dream cannot be meaningfully expressed, or, for that matter, judged or even entertained in thought. I claim that while the latter crucially depends on a specific view of meaning and, therefore, of philosophical ‘context’ or ‘discourse’, which may not be universally 103
104
Moore and Wittgenstein
shared, the former doesn’t and can actually be brought to bear on the sceptical argument from dreaming as well. Hence, it is less contentious and deserves full consideration also in the current debate on scepticism in contemporary epistemology, which generally rejects Wittgenstein’s conceptions of meaning and of philosophical ‘discourse’.
1 The language game with ‘to doubt’: A perspicuous description In the case of our concept of doubt, like in the case of our concept of knowledge, Wittgenstein thinks that it is the concept it is only in virtue of the language game to which it belongs. By observing the way the latter functions, it is possible to identify the criteria of the correct employment of ‘to doubt’ that are constitutive of that concept. Hence, according to Wittgenstein, if the idealist and the sceptic raise doubts without respecting them, their doubts are nonsense because they aren’t real moves within the language game with that expression. But, as we saw in the previous chapter with respect to Moore’s use of ‘I know’ (§2.4), such apparent moves cannot be integrated within our language game, on pain of annihilating the distinction between meaningful and meaningless applications of ‘I doubt’ and hence the very existence of the corresponding concept. 1.1 Doubts manifest themselves only in certain circumstances and must have consequences in practice Wittgenstein writes, Imagine someone who is supposed to fetch a friend from the railway station and doesn’t simply look the train up in the time-table and goes to the station at the right time, but says ‘I have no belief that the train will arrive, but I will go to the station all the same’. He does everything that the normal person does, but accompanies it with doubts or with self-annoyance, etc. (OC 339) The situation imagined by Wittgenstein can, in its turn, be incorporated in different contexts, which would make the subject’s behaviour look either sensible or paradoxical. He might have been informed by his friend about which train the latter will travel on, but he may know of a strike and hence doubt that that train will arrive. This doubt notwithstanding, he goes to the railway station because he is scrupulous and
Doubts and the Nonsense of Scepticism 105
isn’t certain that the train won’t arrive either. Yet while going to the railway station for a sense of duty and hospitality, he is doubtful and irritated with himself. If that were the case, the words and actions of the subject would be perfectly in order, but they wouldn’t be philosophically interesting. Let us suppose instead that that person did not have any reason to doubt of the fact that his friend’s train will arrive at a given time. Are then the manifestation of a doubt and a sense of irritation enough to say that that man doubts that the train will arrive, even if he goes to the railway station? As always, according to Wittgenstein, a mere sensation or feeling isn’t enough to guarantee the correct application of a concept, namely, that of doubt. At most, a mental state or a given feeling can accompany the manifestation and the behaviour typical of doubt, but aren’t what the expressions ‘to doubt’ or ‘I doubt’ mean or even refer to. If, then, a characteristic mental state isn’t enough to guarantee the correct employment of ‘I doubt’, the manifestation of a doubt will have to conform to public criteria of use and verification. First of all, as we will see (§1.2), it will have to be possible to exhibit the reasons behind one’s doubt. Second, in conformity with the criteria for rational behaviour, it will have to be accompanied by characteristic actions (ceteris paribus). However, in the case described by Wittgenstein in the previous quotation, none of this happens. For we know that the subject doesn’t have reasons to doubt the fact that the train will arrive; nor does he behave in accord with his supposed doubt. Thus, he doesn’t conform to public criteria for the correct employment of that expression. Can we nevertheless describe his attitude as typical of those who actually doubt? On the one hand, we could say that he shows some of the characteristic manifestations of doubt but, on the other, his behaviour seems weird and peculiar given the circumstances. As Wittgenstein puts it, Doubting has certain characteristic manifestations, but they are only characteristic of it in particular circumstances. If someone said that he doubted the existence of his hands, kept looking at them from all sides, tried to make sure it wasn’t ‘all done by mirrors’, etc., we should not be sure whether we ought to call that doubting. We might describe his way of behaving as like the behaviour of doubt, but his game would not be ours. (OC 255 my emphasis) On the other hand a language-game does change with time. (OC 256)
106
Moore and Wittgenstein
In the first passage, the subject shows a behaviour which is similar to the one of those who doubt, even if he doesn’t draw its consequences. But in order for that behaviour to be considered as typical of doubt, also the circumstances in which it is manifested must be appropriate. He therefore seems to imitate or mimic the behaviour which is typical of doubt. Yet we couldn’t really say that he is doubting, for that will impugn the differences between the circumstances in which he (seems to) doubt and those which are in fact typical of our language game with ‘to doubt’ (as we shall see, those circumstances crucially depend on there being reasons for one’s doubting). If that distinction is obliterated, also our concept of doubt would change or would cease to exist. For it would be applied both when there is a kind of behaviour which takes place in characteristic circumstances and when there is a kind of behaviour that resembles the former but which manifests itself in totally different circumstances. That is to say, in circumstances in which our actual or, at least, present concept of doubt doesn’t get applied at all. That, as it might happen, a language game could change through time is tantamount to saying that a concept of ours can either disappear or change. Hence, There are cases such that, if someone gives signs of doubt where we do not doubt, we cannot confidently understand his signs as signs of doubt. I.e.: if we are to understand his signs of doubt as such, he may give them only in particular cases and may not give them in others. (OC 154) On the basis of what we have seen so far, a doubt is characterized by a typical behaviour, which must take place in typical circumstances (OC 255–6, 334, 339). What is more relevant, though, is that a doubt should have practical consequences (OC 120, 524–5). The subject in our first example doesn’t behave in accord with his supposed doubt. While it is constitutive of rational behaviour that, ceteris paribus, be in accord with the beliefs and doubts one has. That man’s behaviour – viz. his going to the railway station anyway – looks thus incongruous and out of kilter, as it doesn’t differ from the behaviour of a subject who – sensibly in those circumstances – doesn’t doubt at all that the train will arrive. But if a doubt has the same practical consequences of a lack of doubt, what is the point in saying that one is doubting that p? If the concept of doubt is applied also where the accompanying behaviour is the same as the one typical of a lack of doubt, what difference may there be
Doubts and the Nonsense of Scepticism 107
between the concept of doubting and that of not doubting? And if that difference is obliterated, then ‘to doubt’ seems to lose its meaning. Doubt isn’t, therefore, a particular mental state, even if such a mental state can accompany the manifestation of doubt. It can sensibly take place only in specific circumstances and it must be followed by behavioural manifestations which ought to have practical consequences. To put it otherwise, what matters is ‘what (…) practical effects (…) this belief [or doubt would have]’ and it would be useless to say ‘That’s not the point. A belief [or a doubt] is what it is whether it has any practical effects or not’, while perhaps thinking that ‘it is the same adjustment of the human mind [Geist] anyway’ (OC 89). However, if we consider the proposition ‘This is a hand’, how would a doubt about it manifest itself in practice? One can easily imagine the consequences of doubting that a train will arrive at a given time. But it is difficult to imagine the consequences of doubting that this, which I hold up in front of me, be my hand. If I doubted of that proposition in those circumstances, it would be doubtful that I really knew the meaning of the word ‘hand’. Moreover, if that doubt manifested itself in practice, my actions would look odd and be clear symptoms of some mental illness. Be that as it may, as we saw while discussing Thompson Clarke’s and Barry Stroud’s positions (Chapter 1, §4), philosophical doubts about the premises of Moore’s proof wouldn’t have any consequences in practice. In Wittgenstein’s view of the matter, this is a clear indication of the fact that they would thus be nonsense. 1.2
Doubts must be based on grounds
In OC 458 Wittgenstein writes, ‘One doubts on specific grounds’.1 Let us then imagine a specific situation. A friend tells me ‘I know that Marco is at home’ and I reply ‘I doubt it, because I’ve called him several times and had no reply’. Another time, I’m skiing with a friend and he tells me, in the middle of a snowstorm, ‘That thing over there must be a restaurant’. I say: ‘I doubt it, there are no restaurants on this side of the mountain. Perhaps it’s a shepherds’ hut’. With respect to someone’s claim to knowledge, I have reasons to doubt it if I am aware of facts, or circumstances, that speak against what they claim to know. The grounds of my doubt are the criteria of my claim and are evidence against what the other person declares to know. Being criteria of the application of the concept of doubt, my grounds are sufficient for my claim, but they don’t guarantee that I am right and my interlocutor wrong. For, clearly, Marco could have been in the shower and thus may have failed to hear
108
Moore and Wittgenstein
the telephone ring, and, unbeknownst to me, a restaurant may have been recently opened on that side of that mountain. But, in the case of Moore’s proof, what grounds could there ever be to doubt that what Moore holds up in front of himself is his hand? If I had that doubt in those circumstances, where perceptual conditions are optimal and I am cognitively lucid, I should in fact have to doubt of the deliverances of my eyesight. Hence, I should have to doubt that those are in fact optimal conditions. But that runs contrary to the nature of the case. If I did nevertheless doubt it, that would either show that I don’t know the meaning of the word ‘hand’ or that I am affected by some mental illness. Once more, the sceptical doubt, which is raised irrespectively of the usual criteria that govern the language game with ‘to doubt’ is, in Wittgenstein’s opinion, nonsensical. Because, to repeat, it is his view that philosophy is no further an independent language game where our ordinary language can go ‘on holiday’. Rather, it often depends on a misuse of our ordinary language – of the only language we have got – which produces a mere appearance of sense. 1.3
Doubts are possible only within a language game
What we are interested in now is to explore the features of the language game in which it makes sense to raise doubts about the existence of a physical object. Consider the following situation. One might say ‘Perhaps this planet doesn’t exist and the light-phenomenon arises in some other way’ (OC 56); or else, one might claim that a given historical figure never existed. These are perfectly legitimate doubts which characterize the methodology of scientific and historical investigations. But in order to find out whether a given planet or a historical figure exist(ed), as well as to doubt it, we can’t call into question the existence of the instruments that give us the data on the basis of which we formulate our hypotheses and doubts; nor can we doubt the fact that the earth has existed for a very long time, if we still want to be able to use a finding or a source as evidence for or against a given historical hypothesis. We can thus see that our doubts about the existence of physical objects and people are subject to methodological restrictions, which guarantee the very possibility of raising those doubts. For, otherwise, we would no longer know what could speak for or against a given hypothesis.2 To put it differently, If someone doubted whether the earth had existed a hundred years ago, I should not understand, for this reason: I would not know what such a person would still allow to be counted as evidence and what not. (OC 231)
Doubts and the Nonsense of Scepticism 109
On the one hand, our system of historical and geological evidence speaks in favour of the fact that the earth has existed for a very long time.3 On the other, what we consider our system of justification can exist just as long as the extended existence of the earth isn’t called into question. For, otherwise, we could no longer consider a fossil, or a historical document, as such. They would just be queer rocks or ambers and useless pieces of paper or parchment. Hence, if, on the one side, we don’t have any evidence that speaks against the fact that the earth has existed for a very long time; on the other, to call that into question would destroy what we ordinarily regard as grounds either to assert or to doubt, within history, geology and other disciplines, of the existence of objects, people, facts and events.4 Hence, there can’t logically be reasons, internal to one of our disciplines, to doubt of the very long existence of the Earth.5 As a consequence, any doubt we might have with respect to that fact would only have the appearance of a doubt, but wouldn’t be real. Or else, it would simply be pathological (OC 452–4). Yet, once more, given that sceptical doubts are raised irrespectively of those methodological restrictions which are constitutive of the possibility of putting forward sensible doubts about the existence of a given physical object, they are nonsensical by Wittgenstein’s lights. 1.4
Doubts presuppose certainty
According to Wittgenstein, it is only on the basis of certainty that we can raise doubts. As he writes, ‘The game of doubting itself presupposes certainty’ (OC 115). Hence, certainty precedes doubts and makes them possible. For, first of all, we must be certain about the meaning of the words we use to express our doubts, if we really want to doubt something. If I don’t know that, how do I know if my words mean what I believe they mean? (OC 506) We say: if a child has mastered language – and hence its application – it must know the meaning of words. It must, for example, be able to attach the name of its colour to a white, black, red or blue object without the occurrence of a doubt. (OC 522) And indeed no one misses doubt here; no one is surprised that we do not merely surmise the meaning of our words. (OC 523)
110
Moore and Wittgenstein
If I doubted the fact that this object that I hold up in front of myself is my hand, it would then be doubtful that I knew the meaning of that word. For, as we have seen at length in the previous chapter (§3.2), if I were in doubt about the application of that word, on those circumstances, I would show that I don’t really know the meaning of that term. Another way of seeing why this would be so can be evinced from PI I, 560, where Wittgenstein writes, ‘The meaning of a word is what is explained by the explanation of the meaning’. Hence, in order to know the meaning of a word, one must be able to explain it. But if I doubt the fact that this is my hand, how could I still explain the meaning of that word? I could, at most, give some sort of verbal definition of it, which, however, I would be unable to apply to its worldly referent. So I would then be unable to use it, thereby showing that I don’t really know its meaning. But if one didn’t know the meaning of the words one is using, what sense would one’s words make? And what sense would one’s doubt make? Once more, it would not be a real doubt but a mere appearance of doubt. The absence of uncertainty or doubt is constitutive of one’s knowledge of the meaning of the words one is using, which, in its turn, is a necessary condition – in fact a presupposition – for raising any meaningful doubt. Second, the absence of doubt – hence an attitude of trust 6 – is constitutive of the possibility of acquiring a language. For, as we saw in the previous chapter (§3.3), the correct attitude towards an ostensive definition is not to call it into question, and not to doubt the existence of the objects used as samples to formulate that definition. For, otherwise, one would be ignoring the normative role of that utterance and would be taking it to be stating an empirical state of affairs, which may not be the case. By so doing, however, one could never learn the meaning of that word and could never really come to express a genuine doubt about the existence of its referent. A pupil and a teacher. The pupil will not let anything be explained to him, for he continually interrupts with doubts, for instance as to the existence of things, the meaning of words, etc. The teacher says ‘Stop interrupting me and do as I tell you. So far your doubts don’t make sense at all’. (OC 310) Imagine that the schoolboy really did ask ‘and is there a table even when I turn round, and even when no one is there to see it?’ Is the teacher to reassure him – and say ‘Of course there is!’?
Doubts and the Nonsense of Scepticism 111
Perhaps the teacher will get a bit impatient, but think that the boy will grow out of asking such questions. (OC 314) That is to say, the teacher will feel that this is not really a legitimate question at all (…). The teacher would feel that this was only holding them up, that this way the pupil would only get stuck and make no progress. – And he would be right. (…) This pupil has not learnt how to ask questions. He has not learnt the game that we are trying to teach him. (OC 315) Finally, it is only on the basis of taking for granted certain things, such as the very long existence of the earth, that we can take part in the various language games within which it makes sense to doubt whether a given historical figure really existed, or whether a specific historical event actually took place. (Cf. §1.3 and OC 160–2, 310–318, 327). Hence, to sum up and conclude, here are the main features of our language game with ‘to doubt’, according to Wittgenstein: Doubts can be raised only in certain circumstances (§1.1); they must make a difference in practice (§1.1); in particular, they must be based on grounds (§1.2); they can therefore be raised only within specific language games where certain things stay put and make it possible to gather evidence which speaks (for and) against a given hypothesis (§1.3); • they therefore presuppose certainty and in fact that certain things be exempt from doubt (§1.4); • thus making it possible to have and learn a language (§1.4). • • • •
2 Some philosophical consequences: Why idealism and scepticism in general are nonsensical Let us now explore some philosophical consequences of Wittgenstein’s remarks about our language game of doubting. We will first consider their bearing on idealism, then on scepticism at large. 2.1
The idealist doubt about the existence of an external world
Let us consider the significance of Wittgenstein’s considerations regarding the possibility of raising doubts about existence only within a language game (§1.3) with respect to the idealist doubt about the existence
112
Moore and Wittgenstein
of an external world. As we saw at length in Chapter 1, idealism differs from scepticism: the former actually denies the existence of physical objects – or, possibly, that what we call ‘physical objects’ are actually mind-independent entities, rather than mere collections of sense data. The latter, in contrast, suspends judgement about such an existence, for it can’t find reasons either to assert or to deny that physical objects – conceived as mind-independent entities – exist. Thus, to speak of ‘idealist doubts’ may be somewhat misleading and possibly cause a conflation between idealism and scepticism. Yet, this is how Wittgenstein often puts the issue (OC 24, 37). Still, it is quite obvious that the question ‘What right have I not to doubt the existence of [physical objects]?’ (OC 24, my emphasis) is the leading question behind idealism. For, assuming that the direct object of perception be sense data, then it becomes doubtful that there really are self-standing objects which exist unperceived. As a consequence, instead of suspending judgement about such an issue, one could go so far as to deny their existence,7 or, at least, to claim that what we call ‘physical objects’ are just collections of sense data. Wittgenstein’s counter against idealism is quite complex. First, he claims, Someone who asks such a question [viz. the idealist question about the existence of an external world] is overlooking the fact that a doubt about existence only works in a language-game. Hence, that we should first have to ask: what would such a doubt be like? and don’t understand this straight off.8 (OC 24) Wittgenstein’s point here is that we simply don’t have a language game in which the existence of all physical objects is called into question at once. As we saw (§1.3), we only have language games in which the existence of specific physical objects, such as planets, is called into question, on the basis of grounds which take for granted the existence of other physical objects. Given that for Wittgenstein it is only within a language game that signs come to have a meaning, raising a doubt outside any of our language games has only an appearance of sense. Furthermore, it goes against Wittgenstein’s conception of philosophy to think that philosophical doubts, such as those purportedly expressed by an idealist, could belong to a peculiar language game – that of philosophy – and could thus be meaningful. Given that in philosophy words tend to be employed against all criteria that govern their use in ordinary language, they simply fail to make sense – to be meaningful – according to him.9
Doubts and the Nonsense of Scepticism 113
Secondly, Wittgenstein radicalizes this claim by pointing out the following: But can’t it be imagined that there should be no physical objects? I don’t know. And yet ‘There are physical objects’ is nonsense. Is it supposed to be an empirical proposition? – And is this an empirical proposition: ‘There seem to be physical objects?’ (OC 35) ‘A is a physical object’ is a piece of instruction which we give only to someone who doesn’t yet understand either what ‘A’ means, or what ‘physical object’ means. Thus it is an instruction about the use of words, and ‘physical object’ is a logical concept. (Like colour, quantity, …) And that is why no such proposition as: ‘There are physical objects’ can be formulated. Yet we encounter such unsuccessful shots at every turn. (OC 36) Now, ‘physical object’ is a logical – categorial – concept, like ‘colour’ or ‘quantity’, which can be predicated of specific instances. Those instances, in their turn, must exist if we are to be able to exhibit and use them to explain the meaning of ‘physical object’, or ‘colour’, or ‘quantity’. From which it obviously follows that there are physical objects (or colours or quantities. Cf. OC 57). Yet, Wittgenstein tells us that ‘There are physical objects’ is nonsense, if it is meant to be an empirical proposition.10 Why? Well, because we can’t take ourselves to have proved the existence of physical objects – as mind-independent entities – just by noticing that the expression ‘physical object’ is used in our language. Just like the existence of numbers – as mind-independent abstract objects, let’s say – can’t be proved by noticing that we say ‘Five is a (natural) number’ and that we can thus conclude ‘So there are (natural) numbers’. But to point this out is ‘an adequate answer to the scepticism of the idealist, or the assurances of a realist’? (OC 37). ‘For them after all it is not nonsense’ (ibid.). Answers Wittgenstein: ‘It would, however, be an answer to say: this assertion, or its opposite is a misfiring attempt to express what can’t be expressed like that’ (ibid.) What, then, are the realist and the idealist trying, misleadingly, to express? As to the realist, Wittgenstein’s answer is, So one might grant that Moore was right, if he is interpreted like this: a proposition saying that there is a physical object may have the same logical status as one saying that here is a red patch. (OC 53)
114
Moore and Wittgenstein
Hence a realist like Moore is right to point out that we aren’t objectively certain just of propositions about sense data and subjective sensations, or even of propositions of arithmetic (OC 447–8, 455, 651–5, cf. 656–76), but also about propositions regarding what we categorize as physical objects, such as hands, tables and chairs, at least in certain circumstances, like those paradigmatically exemplified by Moore’s proof.11 An idealist, however, is right to insist that we haven’t thereby proved the mind-transcendent existence of objects. Yet, he fails to notice that ‘There are physical objects’ is and can only be a grammatical statement, not an empirical one. If so, its negation can’t be taken to state a deep metaphysical truth either. Hence, ‘There aren’t physical objects’, as well as its opposite, are just nonsense, if interpreted in the metaphysical way in which both the realist and the idealist tend to interpret them. Wittgenstein’s point, I think, is therefore that ‘There are (not) physical objects’ can only make explicit some basic feature of grammar, or, equivalently, of our conceptual scheme, which may (or may not) countenance, within the fundamental fabric of the world, mind-independent objects; or else, see them as mere constructions out of sense data. In fact, Wittgenstein’s answer to idealism, up to this point, sounds ultimately as follows: we simply do have a conceptual scheme – which is a product of our language – that comprises fundamental reference to physical, mindindependent objects, contrary to what idealists would have us believe. That we do have such a conceptual scheme is shown by our linguistic and epistemic practices and its objective certainty – not truth – which, for Wittgenstein, is always a function of the role certain propositions play in our overall picture of the world, is manifested by the fact that the hypothesis […] that all the things around us don’t exist […] would […] be like the hypothesis of our having miscalculated in all our calculations. (OC 55) [And it is not] conceivable that we should be wrong in every statement about physical objects; that any we ever make are mistaken. (OC 54) Why, then, is it inconceivable that we should always be mistaken in our statements about physical objects (or about our calculations)? Because, according to Wittgenstein, as we have already repeatedly seen, the very meaning of our words doesn’t depend just on there being an agreement
Doubts and the Nonsense of Scepticism 115
in definitions, but also in judgements. Now, it is a fact that we do agree in judging of certain objects, which may not be presently perceived, or that may pre-date our existence, that they are physical objects. That is why the hypothesis that there be no mind-independent physical objects boils down to the hypothesis that we may always have been mistaken and that hypothesis, in its turn, seems to make no sense, for it would deprive the expression ‘physical object’ of what is – at least currently – its meaning. Once more, Wittgenstein’s counter against idealism seems an elaboration on the very simple claim that our grammar and conceptual scheme comprise mind independent physical objects, contrary to what idealism suggests. The philosophically relevant aspect of such a claim, however, is not so much the mere observation about that feature of our conceptual scheme. Rather, I think, it is the insistence on the grammatical nature – at most – of ‘There are physical objects’, which places both the realist and the idealist metaphysical interpretations of that statement beyond the bounds of sense.12 To put it otherwise: the realist/idealist dispute is hollow, or, to use a Wittgensteinian term, ‘nonsensical’, for if ‘There are physical objects’ is a grammatical statement and it is a fact that we predicate ‘physical object’ of unperceived and pre-existent objects, then it is a mere empirical fact that our conceptual scheme comprises mind-independent physical objects. Yet, this neither proves realism right nor idealism wrong, for it only shows how ‘physical object’ is used within our language. Neither realism nor idealism can be proved right or wrong for they both misconceive the nature of statements such as ‘There are physical objects’, which is not empirical and yet incontrovertible – that is to say, metaphysical – but grammatical. And it can be empirically ascertained how certain words are actually used in our language and thus what features our conceptual scheme does in fact have. 2.2
Why scepticism in general is nonsensical
We have seen in §1.4 that, for Wittgenstein, doubts can only come after certainty, in the twofold sense of coming after an attitude of trust and of being necessarily based on taking for granted certain things. We know, however, that in classical epistemological projects, since Descartes’ Metaphysical Meditations, doubt has been considered the source of certainty; and it has been thought that only by calling into question any opinion, we could then determine what is actually known with certainty. This way, certainty would come after doubt and any conceivable doubt would therefore be legitimate – that is to say, intelligible and meaningful. Furthermore, methodological scepticism was meant to
116
Moore and Wittgenstein
be global, because it called everything into question to see if something could be certain after all. For instance, Descartes writes, Several years have now passed since I first realized how numerous were the false opinions that in my youth I had taken to be true, and thus how doubtful were all those that I had subsequently built upon them. And thus I realized that once in my life I had to raze everything to the ground and begin again from the original foundations, if I wanted to establish anything firm and lasting in the sciences. (…) Accordingly, I have today suitably freed my mind of all cares, secured for myself a period of leisurely tranquility, and am withdrawing into solitude. At last I will apply myself earnestly and unreservedly to this general demonstration of my opinions. (AT 17–8) This very common philosophical attitude is completely mistaken, according to Wittgenstein. In order to see why, it is necessary, in his view, to examine its consequences. By so doing, one will see that global scepticism would annihilate the very possibility of language and of its learning, as well as of participating in our usual epistemic practices. As Wittgenstein put it, ‘If you tried to doubt everything you would not get as far as doubting anything’ (OC 115, cf. 450, 519, 625). For, as we saw (§1.4), the very existence of language and the possibility of learning it depend on a general attitude of trust and on not calling into question certain things. Similarly, doubts are subject to methodological restrictions that depend on the features of specific language games where some things must stay put (§1.3). Furthermore, not every possible doubt is meaningful for Wittgenstein (Cf. OC 302, 392, 606). Only those doubts that are grounded in reasons (§1.2) and that can make a difference in practice (§1.1) are. These doubts, in their turn, presuppose that something be not called into doubt.13 There is a common reply to this Wittgensteinian attitude of deeming scepticism about the existence of an external world nonsensical. It runs as follows: let us assume that in order to have language and our ordinary epistemic practices, doubt must come after certainty (§1.4). But, once having acquired knowledge, one can then raise any kind of doubt, while still using words meaningfully, and thus call into question the very foundations of our ordinary epistemic practices. Another frequent kind of objection takes the following form: let us concede that in order to have language, doubts are possible only on the basis of taking for granted certain things (§1.4). Yet, philosophical doubts
Doubts and the Nonsense of Scepticism 117
aren’t essentially linguistic. They could occur just in the mind and they would simply require the possession of the relevant concepts.14 Both these objections need to be addressed by a Wittgensteinian, in order to clarify the strength of his position. As to the former, the reader should recall what we saw in the previous chapter about the use of ‘I know’ outside our ordinary language game (§§3.4, 4). The idea that words could still have a meaning even when used outside of it, is, according to Wittgenstein, a bad mistake, which depends on thinking of meaning as something that signs can carry on their sleeves independently of their use and of thinking that the philosophical employments of words are a genuine language game, though always irrespective of the criteria that govern our ordinary ones. Both assumptions are deeply at odds with Wittgenstein’s conception of meaning as use and with his conception of philosophy. To repeat two points already made several times: the meaning of a word is, for Wittgenstein, the use that that word has in our ordinary language, and, in philosophy, our ordinary language tends to ‘go on holiday’. Therefore, most philosophical pronouncements have only an appearance of sense. Concerning the second objection, it has to be kept in mind that Wittgenstein thought of concepts as dependent on language (OC 65) and of their mastery as constitutively tied to the ability to apply them in judgements, characteristically (though not inevitably) expressed by assertions. So, on a Wittgensteinian perspective, there can’t be a language just of thought. Nor can concepts be internalistically individuated. For the application to their referents is essential to their determination, as Wittgenstein’s insistence on the importance of agreement in judgements beside agreement in definitions, in order for words to have their meanings, made clear since the Philosophical Investigations (PI I, 241, cf. OC 126, 268, 306).15 Hence, for a full-fledged Wittgensteinian, sceptical doubts are really nonsense essentially because philosophical contexts, or, as Stroud would put it, ‘philosophical discourse’, are spurious – they are no real context or discourse at all, as they go against the criteria that govern ‘ordinary’ contexts or discourses – that is, the only contexts and discourses we really have. Nor can the retreat to thought help the cause of a sceptic, since such a recoil, by Wittgenstein’s lights, is in fact always parasitic on their being a publicly shared language. To sum up: • Idealism is not wrong, but, in fact, nonsensical (§2.1). For it tends to interpret ‘There are physical objects’ as the realist does – viz. as expressing an empirical proposition, which can thus be meaningfully
118
Moore and Wittgenstein
negated – while in fact it merely states a grammatical ‘truth’. Namely, that we have a language and, therefore, a conceptual scheme, which comprises mind-independent physical objects. • Wittgenstein’s considerations about the conditions of meaningfulness of doubts run contrary to the epistemological tradition according to which certainty comes after doubt (§2.2, cf. §1.4); • such a tradition is totally mistaken, according to Wittgenstein, and so-called methodological doubts – that call everything into question, aren’t based on grounds, nor do they make any difference in practice – completely nonsensical (§2.2, cf. §1.1–1.3). • Appearances to the contrary are in fact a product of a mistaken conception of meaning, as well as of philosophy and, finally, possibly of the idea that thought could be independent of language and of its applications (§2.2).
3 Two classical sceptical arguments and Wittgenstein’s replies In this section I will consider two classical sceptical arguments and will present what I take to be the two radically dissimilar styles of response which can be evinced from On Certainty. I will dub the former ‘the linguistic argument against Cartesian scepticism’ and the latter ‘the transcendental argument against Humean scepticism’. I will then present and critically discuss some recent interpretations of Wittgenstein’s second argument – viz. Wright’s and Williams’ – and show why they all fail to take proper measure of his radical anti-epistemicism. On the basis of this second Wittgensteinian counter to external world scepticism, we will see how the latter is nonsensical because it is in fact not rational, in a way which will have to be further specified and clarified. By contrast, by considering Wittgenstein’s counter to the argument from dreaming, we will see how, for him, that argument is nonsensical because it is in fact meaningless – it can’t even be meaningfully formulated. Hence, two quite different notions of nonsense with respect to sceptical hypotheses will emerge. Though, no doubt, Wittgenstein thought that somehow the first entailed the second, that conclusion can be reached only on a quite specific view of meaning – the conception of meaning as use – which may not be unanimously shared. To be sure, it is widely rejected nowadays, and its rejection characterizes much of contemporary epistemology, where there is no issue that sceptical doubts can be meaningfully raised and expressed. So, although that isn’t Wittgenstein’s ultimate
Doubts and the Nonsense of Scepticism 119
view, it is important to see what kind of damage the transcendental argument, taken in isolation from its (alleged) semantic consequences, can make to external world scepticism. That will in turn allow us to assess its significance also for the contemporary debate in epistemology. Finally, I will show how that style of response can be extended, in particular, also against the argument from dreaming. Hence, I will claim that the extensive version of the transcendental argument deserves full consideration within the ongoing debate in epistemology, given its background presuppositions, and isn’t merely of historical interest. 3.1
The linguistic argument against Cartesian scepticism
The argument from dreaming is traditionally considered one of the most powerful arguments in favour of external world scepticism. Let us look at how it proceeds by presenting Descartes’ classical formulation of it in the First Meditation (though without any presumption of exegetical accuracy). But perhaps, even though the senses do sometimes deceive us when it is a question of very small and distant things, still there are many other matters concerning which one simply cannot doubt, even though they are derived from the very same senses: for example, that I am sitting here next to the fire, wearing my winter dressing gown, that I am holding this sheet of paper and the like. But on what grounds could one deny that these hands and this entire body are mine? Unless perhaps I were to liken myself to the insane, whose brains are impaired by such an unrelenting vapor of black bile that they steadfastly insist that they are kings when they are utter paupers, or that they are arrayed in purple robes when they are naked (…). But such people are mad, and I would appear no less mad, were I to take their behaviour as an example for myself. (AT 19) Descartes is perfectly aware of the fact that given the circumstances he couldn’t sensibly doubt of having a body, of wearing a gown and so forth. It is therefore a paradigmatic case in which knowledge of numerous empirical facts is obtained. But, then, isn’t there any way of sensibly doubting of having a body, of wearing certain clothes and of having a hand? Descartes’ move in order to call into question even those apparent certainties is as follows: This would all be well and good, were I not a man who is accustomed to sleeping at night, and to experiencing in my dreams the very same
120
Moore and Wittgenstein
things, or now and then even less plausible ones, as these insane people do when they are awake. How often does my evening slumber persuade me of such ordinary things as these: that I am here, clothed in my dressing gown, seated next to the fireplace – when in fact I am lying undressed in bed! But right now my eyes are certainly wide awake when I gaze upon this sheet of paper. This head which I am shaking is not heavy with sleep. I extend this hand consciously and deliberately, and I feel it. Such things would not be so distinct for someone who is asleep. As if I did not recall having been deceived on other occasions even by similar thoughts in my dreams! As I consider these matters more carefully, I see so plainly that there are no definitive signs by which to distinguish being awake from being asleep. As a result I am becoming quite dizzy, and this dizziness nearly convinces me that I am asleep. (AT 19)16 If in a dream I have sensations and see images that are very similar to those which I feel and see while awake, what criterion could there be for saying that the ones I am having and seeing now are real, and not produced in a dream? According to Descartes, the quality of our perceptions while awake or dreaming is usually different and this ordinarily guarantees that they can be distinguished. In order to give up this usual opinion, and be able more easily to call into question the reality of his current sensations and perceptions, Descartes formulates the hypothesis of the existence of an evil genius that makes him have dreams which are qualitatively indistinguishable from real-life situations. I will suppose not a supremely good God, the source of truth, but rather an evil genius, supremely powerful and clever, who has directed his entire effort at deceiving me. I will regard the heavens, the air, the earth, colors and shapes, sounds and all external things as nothing but the bedeviling hoaxes of my dreams, with which he lays snares of my credulity. (AT 22) Two observations are in place. First, there can’t (logically) be reasons to think it is probable, or even possible, that there be an evil genius, whose pastime is to deceive us. Hence, the first thing Wittgenstein would remark is that since such a hypothesis has nothing in its favour, it is irrational and, therefore, nonsense. However, we know that a sceptic could reply by saying that the mere logical-metaphysical hypothesis that there be such a genius is enough
Doubts and the Nonsense of Scepticism 121
for his purposes. As we saw in Chapter 1 (§4), while presenting Clarke’s and Stroud’s interpretation of Moore’ proof, sceptical doubts are universal, theoretical and methodical and don’t need reasons in their support. We have seen in §2.2 of this chapter how Wittgenstein is deeply contrary to deem legitimate a merely possible doubt. However, I now wish to consider in more detail the argument which can be evinced from On Certainty against the dreaming hypothesis as such, even if, being totally unmotivated, it would already be irrational and therefore nonsensical, by Wittgenstein’s lights. Second, it has been noted that the argument from dreaming is a challenge or a serious threat to the possibility of having knowledge of an external world on the assumption that such a kind of knowledge be based on perceptual grounds.17 For if in a dream I had the same sensations and perceived the same images which I usually think I have and see while awake, then I couldn’t say, on the basis of the testimony of the senses, that what I am now perceiving is real. In order to do so, I should be able to exclude that I am now dreaming. But, to such an end, I should have a test which could determine whether, right now, I am awake or asleep. However, since any test, in order to prove anything, should actually be carried out, I should already have a way of determining whether I am actually running the test, or merely dreaming of doing so. But this would clearly produce an infinite regress.18 Hence, since we can’t exclude of being in dream, we can’t know that our current perceptions are actually produced by causal interaction with physical objects around us. We can’t therefore base our knowledge of an external world on the deliverances of the senses. But if perception is the only source of our knowledge of the existence of an external world, we can’t but suspend judgement about the latter. Hence, we can’t but be sceptical about it.19 Wittgenstein’s counter to the dreaming argument is meant to show that that argument can’t even be meaningfully formulated. Let us look at it in detail. Suppose I am not really seeing the computer screen I am thinking of seeing in front of me right now, and that I am merely dreaming of it. If I said, ‘This computer screen is on’ while, ex hypothesi, I am dreaming; if someone heard me, they wouldn’t take my words as an assertion. For I can’t possibly have the intention of informing them of something. Nor could they take them to be a description of a state of affairs, even if, by chance, it happened to be true that the computer screen in front of me is on. Similarly, if the words ‘I am dreaming’ were uttered while I am dreaming, they couldn’t be taken as an assertion and
122
Moore and Wittgenstein
hence as a description of a state of affairs, which, ex hypothesi, is in fact the case. Writes Wittgenstein, ‘But even if in such cases I can’t be mistaken, isn’t it possible that I am drugged?’ If I am and if the drug has taken away my consciousness, then I am not now really talking and thinking. I cannot seriously suppose that I am at this moment dreaming. Someone who, dreaming, says ‘I am dreaming’, even if he speaks audibly in doing so, is no more right than if he said in his dream ‘it is raining’, while it was in fact raining. Even if his dream were actually connected with the noise of the rain. (OC 676) Hence, from a third-person point of view, there being criteria which can determine whether I am asleep – having closed eyes, being almost immobile, breathing regularly, being insensitive to soft sounds, etc. – prevent any utterance I could make from being considered an assertion. Nobody would verify my alleged assertions, nor would they consider them informative. That is to say, they wouldn’t decide on the basis of my testimony either to switch off the computer or to take an umbrella before leaving home, even if, of course, my words could causally prompt them to perform those very actions, after having ascertained on the basis of their perceptual experience that the computer is in fact on or that it is actually raining. But that all this could happen just on the basis of the information one takes to have received from another subject is constitutive of making assertions, or of taking certain pronouncements as assertions. Hence, given our actual practice of making assertions and of interpreting other subjects’ pronouncements, there is no reason to hold that if a subject said ‘I am dreaming’, while actually being asleep, he would be making an assertion, even if it can happen that he is in fact dreaming. Let us therefore concede that when one is asleep, one cannot be making assertions. Yet, one may wonder whether, while asleep, one may truly judge ‘I am dreaming’. According to Wittgenstein, however, if one were really dreaming one would also be dreaming that the phrase ‘I am dreaming’ has a meaning. The argument ‘I may be dreaming’ is senseless for this reason: if I am dreaming, this remark is being dreamed as well – and indeed it is also being dreamed that these words have any meaning. (OC 383. Cf. RPP I, 1057)
Doubts and the Nonsense of Scepticism 123
This is an extremely strong claim. As a matter of fact, in section §2.2, we have already seen how, for Wittgenstein, judgements are in fact unspoken assertions. So, one would have expected Wittgenstein to claim simply that even if one dreamt the words ‘I am dreaming’, they could not be used to make a judgement about one’s current situation, since their out-loud counterpart could not be used to make an assertion either. For it could never be intentionally uttered to describe a current state of affairs. But Wittgenstein’s claim here is stronger, because it consists in denying that those words, while occurring in the dream, could have any meaning at all. To put it differently: Wittgenstein isn’t simply denying that the mental counterpart of the speech-act of making an assertion could not be performed. Rather, he is denying that the alleged act of judgement could have a content at all. To see why Wittgenstein is led to such a strong conclusion it is important to keep in mind a point we have been seeing since Chapter 2. Namely, that in order for our words to have a meaning – whether they are actually uttered or merely entertained in one’s mind – they must belong to a language game. That is to say, they must occur in specific circumstances where their utterance can serve the recognized aim of the communicative exchange at issue. If they occur out of the blue or out of any attested context of use, they will retain an appearance of meaning, but they won’t really have one. To think otherwise, as we have repeatedly remarked, would mean to fall pray to a conception of meaning as, ultimately, a thing which words would possess independently of their being part of a real language game – independently, that is, of their being actually used. Wittgenstein’s point about the impossibility of the assertion/judgement ‘I am dreaming’, while one is in fact asleep and dreaming, therefore, carries over to establish that if those words did occur in one’s dream, they would just have an appearance of meaning but wouldn’t really have one. To be more precise, they could not mean, within the dream, that one is in fact dreaming. To put it shortly, they could not have their ‘descriptive’ meaning. Rather, they could mean what they would normally mean if they were said out loud while awake. For instance, the utterance ‘I am dreaming’ could be used to express one’s bewilderment or intense joy. Hence, they could only have their ‘expressive’ meaning. Alternatively, in one’s dream one could be dreaming that certain images or experiences one is having are themselves part of a dream. This, however, would not show that one may be aware of being dreaming while in fact being in a dream but only that one may be dreaming of being in a dream with content thus-and-so. Hence, the phrase ‘I am dreaming’
124
Moore and Wittgenstein
in the context of such a dream would have a descriptive meaning, but, notice, it would in fact be applied on the basis of third-personal criteria. In some sense, one would be seeing oneself being dreaming and could thus report on that much in the same way as one could report on someone else’s being ostensibly dreaming. Of course, the funny thing about such a dream would be the kind of immediate ‘knowledge’ of its content, which is unparalleled in the third-person case. Yet, it remains that ‘I am dreaming’ cannot be meaningfully used or judged, if it were meant as a description of one’s occurrent mental state. Descartes’ supposition is therefore meaningless, and in this semantic sense, nonsense, contrary appearances notwithstanding. One final attempt to rescue the meaningfulness of Descartes’ supposition could consist in insisting that it is, precisely, a supposition, or a hypothesis, not a description of a state of affairs or a judgement. The Wittgensteinian counter would predictably run as follows: in order for suppositions or hypotheses to be meaningful it should be possible for them to be true and for the words used to express them to be employed to make an assertion (or to make a judgement in thought). Failing such a possibility, on the grounds we have explored thus far, also the insistence of the fact that Descartes was merely making a supposition, or entertaining a hypothesis, and wasn’t trying to make any assertions, would lose its bite. To conclude and recap, • According to Wittgenstein the doubt or even the hypothesis ‘I might be dreaming right now’ is nonsense because those words (either spoken or entertained silently in one’s mind) could never be a move in the language game. • If I were dreaming, I would also be dreaming of making an assertion or a judgement, for there could be no intentional connection between those words and the fact they are supposedly describing. • That, in turn, would impair the fact that those words could have their ‘descriptive’ meaning, contrary appearances notwithstanding (while they may retain their ‘expressive’ one); • or be used other than on the basis of third-personal criteria, to describe the peculiar content of one’s dream, viz. that one is dreaming such-and-so. • If I am awake, in contrast, my doubt could only be sensibly expressed in the past – for example ‘Did I just dream thus-and-so?’ – and could be dissipated by a principle of coherence, viz. by the fact that the content supposedly dreamt and one’s current perceptions and memories fit together, as we shall see in the next section.
Doubts and the Nonsense of Scepticism 125
3.1.1 A detour. The language game with ‘(to) dream’. A perspicuous description In this section we will look at Wittgenstein’s remarks on the actual use of ‘(to) dream’. As will become apparent, he showed great concern with the issue and put forward a number of illuminating suggestions. I think this bears testimony both to his long-lasting interest in the topic and to his characteristic way of dealing with philosophical problems. Namely, to the fact that in order to clarify their nature and dissolve them – rather than solve them – one should always look at the actual employment of words, which makes our concepts the concepts they in fact are, contrary to the typically philosophical attitude of misunderstanding them; and, thus, of thinking of making sense, while in fact speaking nonsense. As we have seen, there isn’t a descriptive or assertoric use of ‘I am dreaming’, according to Wittgenstein (barring the kind of atypical situation described at the end of the previous section). That is to say, in his view, there is no language game in which ‘to dream’ is used descriptively in the first person present. However, there is obviously a descriptive use of the verb ‘to dream’ in the third person present. For instance, if I saw a person asleep and mumbling, I could meaningfully say ‘He is dreaming’. Moreover, there is a perfectly meaningful use of the verb ‘to dream’ in the first person past, like in the locutions ‘I was dreaming’ and ‘I dreamt that such-and-so’, uttered when one is abruptly woken up, or when one is asked whether one had dreams the previous night. When one utters those words, those assertions are criteria for others to say that one was dreaming or did dream such-and-so. The interesting issue is how we learn to use those forms of expression, if we never make a descriptive use of the sentence in the first person present. Wittgenstein’s suggestion is as follows: People who on waking tell us certain incidents (that they have been in such-and-such places, etc.). Then we teach them the expression ‘I dreamt’, which precedes the narrative. Afterwards I sometimes ask them ‘did you dream anything last night?’ and am answered yes or no, sometimes with an account of the dream, sometimes not. That is the language-game. (PI II, vii, 184) Hence, according to Wittgenstein, we often learn how to use the expression ‘I dreamt such-and-so’ by telling, after waking, peculiar and
126
Moore and Wittgenstein
fantastic things, supposedly occurred to us, and by being told by our interlocutor ‘It was just a dream’, or ‘You were dreaming’. In this way we learn to preface future narrations of dreams with ‘I dreamt that…’, which becomes a sort of ‘operator’ that allows us to distinguish between actual facts and events and the description of merely dreamt facts and events. We will therefore learn that ‘I dreamt that I had climbed the Everest’ doesn’t mean to relate what has actually happened while one was asleep – for example, that one did climb the Everest while in fact asleep – but to tell what one dreamt of, but did not actually occur. According to Wittgenstein, we acquire the concept of dream in a similar way. A child wakes up frightened and tells his mother ‘I saw a man dressed like a pirate going around me’ and she will tell him ‘Don’t worry, it was just a bad dream’. We thus learn the concept of dream from the narration we make of it. This doesn’t mean, however, that ‘to dream’ means ‘to tell a dream’ or ‘I dreamt’ or ‘I told a dream’ (RPP I, 101: Z 530). Rather, the point is that since nobody could (logically) tell a dream while one is dreaming it, the acquisition and use of the expressions which make reference to such an event can only come after it. According to Wittgenstein, these expressions, like any other, need public criteria of use which will regulate their application. The narration of the dream can thus be one of the manifestations of the dream which makes it possible to acquire the relevant concepts.20 Writes Wittgenstein, Of course I don’t want to give a definition of the word ‘dream’; but still I want to do something like it: to describe the use of the word. My question then runs roughly like this: If I were to come to a strange tribe with a language I didn’t know, and the people had an expression corresponding to our ‘I dream’, ‘He dreams’ etc. – how should I find out that this was so; how should I know which expressions of their language I am to translate into the expressions of ours? (RPP I, 374) Hence, for Wittgenstein, it isn’t the case that ‘if someone does not tell a dream, it is false to say he had it’, for ‘he may well have had the dream even when he doesn’t report it’ (RPP I, 364). After all, Wittgenstein was no behaviourist. Rather, the possibility of dreaming without then telling anyone depends on the fact that one could dream and let others know (RPP I, 364–5). For it is mostly, though not inevitably (note 20), by telling others (or being told by them) what happened in a dream that we acquire the concept of dream and it is only by possessing that
Doubts and the Nonsense of Scepticism 127
concept that we can know that we did have a dream even if we aren’t disposed to tell anyone its content. Hence, [b]ack to the language game of telling dreams: Someone says to me one day: ‘What I dreamt last night I will tell to no one.’ Now does that make sense? Why not? Am I supposed to say, after what I have said about the origin of the language-game, that it makes no sense – as the original phenomenon just was the dream narrative? Absolutely not! (RPP I, 371) However, if the assertion ‘I dreamt such-and-so’ is followed by a narration and, as a matter of fact, the very possibility of mastering the use of expressions such as ‘dream’ and ‘I dreamt’ depends on such a narration, one may wonder whether the latter corresponds to what was actually dreamt. To put it differently, the assertion ‘I dreamt thus-and-so’ could be followed by the question ‘Did you really dream thus-and-so?’, ‘Are you sure you’re remembering correctly?’. For if the memory of the dream was unreliable, there would be a gap between the dream and its narration. Yet, according to Wittgenstein, this kind of question doesn’t belong to the usual language game of telling one’s dreams. Now must I make some assumption about whether people are deceived by their memories or not; whether they had these images while they slept, or whether it merely seems so to them on waking? And what meaning has this question? – And what interest? Do we ever ask ourselves this when someone is telling us his dream? And if not – is it because we are sure his memory won’t have deceived him? (And suppose it were a man with a quite specially bad memory?) Does this mean that it is nonsense ever to raise the question whether dreams really take place during sleep, or are a memory phenomenon of the awakened? It will turn on the use of the question. (PI II, vii, p. 184. Cf. RPP, I, 369) In the usual language game of relating one’s dreams it doesn’t make sense to ask whether the subject is correctly remembering his dream, because all we can (logically) know about the dream depends on the narration he makes of it. We can’t compare his words with the dream in order to determine whether the former correspond to the latter. Hence, the issue of the truth of the narration of one’s dreams doesn’t even arise.
128
Moore and Wittgenstein
This, however, shouldn’t induce one to believe that a person could say (aloud or silently) ‘I am dreaming’ in his dream, wake up, remember his words and then conclude that he made a true assertion. For those words belonged to the dream anyway and couldn’t be about a real fact or event, not even the dream one was in fact having. Hence, with respect to the narration of one’s dreams, the only issue we can sensibly raise is not one about truth – about the correspondence between a subject’s words and his dream – but one about veridicality – viz. about the fact that a subject is being sincere and is relating what he remembers about his dream and isn’t merely deliberately21 fabricating it. Therefore, his narration becomes the criterion to determine (at least) the (apparent22) content of his dream and there is no further content of the dream which could be independently accessed and compared with a subject’s account. Writes Wittgenstein, Assuming that dreams can yield important information about the dreamer, what yielded the information would be truthful accounts of dreams. The question whether the dreamer’s memory deceives him when he reports the dream after waking cannot arise, unless indeed we introduce a completely new criterion for the report’s ‘agreeing’ with the dream, a criterion which gives us a concept of ‘truth’ as distinct from truthfulness. (PI II, xi, 222–3) Finally, it should be kept in mind that sometimes we express a genuine doubt as to whether an event has just really taken place or was merely dreamt of. However, whenever we do so, the only sensible expression of it is in the past, by saying ‘Did I just dream thus-and-so, or was it real?’. As Norman Malcolm put the point in his essay Dreaming: A man can be deceived by a dream when he awakes from sleep, not when he sleeps. When one awakes with a certain impression and is in doubt whether it belongs to a dream or to reality, one can indeed consider whether this impression fits in with what one remembers or presently perceives. Thus coherence has a sensible application to the question ‘Was I dreaming?’ but none at all to the question ‘Am I dreaming?’ (1959, 113) For, as we have seen, one can’t entertain a doubt when one is dreaming and, correlatively, it is a precondition of sensibly doubting about
Doubts and the Nonsense of Scepticism 129
the reality of an ongoing experience that one be awake. So, if one wants meaningfully to use the expression ‘to dream’ actually to refer to a possibly dreamt content, one can do so only by using it in the past, thus wondering whether what was just experienced was dreamt or real. To summarize and conclude, • While, for Wittgenstein, there is no descriptive use of ‘to dream’ in the first person present; • there is a perfectly attested use of ‘to dream’ both in the third person present and in the first person past. • The latter necessarily occurs after the dream; • and it is learnt by being taught to preface one’s narrations of one’s dreams with ‘I dreamt that’. • Furthermore, there is no issue of truth with respect to a dream’s narration, but only of its veridicality, viz. of its being a sincere narration and not a (deliberate) fabrication. 3.2
The transcendental argument against Humean scepticism
As we saw in Chapter 1, there is a kind of sceptical argument that can be evinced from Hume’s considerations about induction and extended to other domains.23 It proceeds by noticing that our empirical, evidential warrants depend not just on having some course of experience, but also on collateral assumptions, such as that our sense organs aren’t currently deceiving us, and that, therefore, they don’t always do so; that we aren’t currently dreaming; and that there is an external world. It then goes on to notice that those assumptions can’t, in their turn, be supported by evidential grounds, as the latter depend on taking the former already for granted. It thus concludes that at the bottom of all our practices there are ungrounded assumptions which can’t rationally be held. Now, the disquieting effect of such a conclusion is that, in the hands of a sceptic, it is taken to imply that those assumptions are totally arbitrary and, at least in principle, as legitimate as their opposite. Hence, we take ourselves to have specific specimens of knowledge regarding empirical facts, for example, that here there is my hand, or that I am currently seeing a computer screen in front of me, but none of them can be granted, for they depend on assumptions which are not rationally grounded. It is important to keep in mind that while in On Certainty there are remarks explicitly addressed to the dreaming hypothesis, what I call the ‘transcendental argument against Humean scepticism’ can
130
Moore and Wittgenstein
only be evinced from the text and, in particular, from those passages where Wittgenstein discusses the fact that our inquiries are all based on presuppositions which can’t sensibly be doubted on pain of annihilating the possibility of raising rational doubts and questions at all (§1.3); as well as from those passages where he points out that while we can sometimes be mistaken in our empirical judgements or calculations, we can’t always have gone or go astray (OC 217, 232, 519). The most interesting applications of this style of argument have to do with what Fred Dretske calls ‘heavy-weight assumptions’,24 such as ‘There is an external world’, ‘I am not currently deceived by my senses’ and thus ‘Our sense organs don’t always deceive us’, ‘I am not now dreaming’, etc. I am therefore taking the liberty of applying Wittgenstein’s considerations to Humean scepticism regarding the reliability of the senses, first, to see how they would fare in this connection. Afterwards, we will see how the transcendental argument can in fact be extended also against Cartesian scepticism, thus targeting the assumption ‘I am not now dreaming’, as well as against the Humean kind of sceptical argument already introduced in Chapter 1 (§5), which concerns the assumption of the existence of an external world. In order to enquire whether it is possible to doubt that our senses ever function properly, we should ask the following question: what could make us have such a doubt? Now, a possible first pass could be this: on occasion, we have found that our senses had deceived us. Indeed this is how Descartes motivates his scepticism about the senses. As he writes (AT 18), Surely whatever I had admitted until now as most true I received either from the senses or through the senses. However, I have noticed that the senses are sometimes deceptive; and it is a mark of prudence never to place our complete trust in those who have deceived us even once. Yet, one should then notice that if we have been in a position to realize the occasional unreliability of our sense organs, it is because further observations have made that manifest to us. For instance, further observations have made it clear that a given object wasn’t of the colour it first appeared to us to have. Alternatively, we have been able to find out that our senses had deceived us because other people told us so. In both cases, however, it is the testimony of the senses that certifies to us the truth of (or is part of our justification for) ‘My senses deceived me’. Hence, the argument for scepticism about the reliability of the
Doubts and the Nonsense of Scepticism 131
senses which proceeds from noticing that sometimes our senses have in fact deceived us is self-defeating for it presupposes the falsity of its conclusion, or, at the very least, that the negation of the conclusion be taken for granted. Still, it seems difficult to find any other sensible motivation for general scepticism (beside supposing the existence of an evil genius) about the reliability of the senses. If so, however, such a doubt appears to be totally ungrounded. Yet we know (§1.2) that a doubt which is not based on any grounds at all – in fact, that, like in this case, cannot be supported by any kind of grounds, for the latter would presuppose what the argument tries to call into question – is no real doubt at all. It would just be an illusion of doubt.25 In particular, it will be irrational 26 and in this non-linguistic, or semantic, sense nonsensical. Of course, as we saw in the previous section, from such a kind of nonsense Wittgenstein would draw the consequence that that hypothesis is entirely meaningless, as it depends on a violation of the criteria that govern the correct employment of ‘to doubt’. In particular, it would violate the requirement that doubts can meaningfully be raised only where there are reasons or grounds for them (§1.2). The latter are in their turn dependent on the fact that doubts be raised within a language game (§1.3) where certain things must stay put (§§1.3–1.4) – for example, that our sense organs don’t deceive us now (and hence don’t always mislead us) – if we are to be able to carry out certain empirical investigations – for example, to find out that they did deceive us on a previous occasion of their use. Yet, such a conclusion draws, in its turn, on a specific view of meaning as use and on a certain conception of philosophical doubts and questions as never autonomous from, or independent of, the criteria that govern the raising of doubts in ‘ordinary discourse’, which may not be commonly accepted. It would remain, though, that the sceptical argument just considered goes against the conditions of possibility of raising a sensible doubt because it is somehow self-defeating, for its very formulation depends on the falsity of its conclusion, or at least on taking for granted that things aren’t as the argument purports to show.27 Even so, from a Wittgensteinian perspective, we can neither claim nor have knowledge of the fact that our senses don’t always deceive us. For there can’t be a non-circular justification of ‘My senses aren’t deceiving me now’ from which we could then gain justification for ‘Our senses don’t always deceive us’. Moreover, we know that for Wittgenstein circular justifications aren’t justifications at all. Hence, both these propositions appear to be the ungrounded presuppositions of all our going about gathering perceptual evidence for, or against, empirical
132
Moore and Wittgenstein
propositions. If so, however, they can’t be called into question either, for, in order sensibly to do so, they would have to be taken for granted. Once more in a Kantian fashion, we may say that their being taken for granted is a condition of possibility of all our perceptual investigations. In fact Wittgenstein remarks that it belongs to the logic of our investigations (OC 342, cf. 56, 82, 628) and to the method of our inquiries (OC 151, 318) that certain things be indeed exempt from doubt and not called into question. For that would actually amount to conceiving of the possibility that all our judgements (or calculations) be always wrong, a hypothesis which, in its turn, could – logically – have nothing in its favour. Such a hypothesis would thus be irrational and, in this nonlinguistic sense, nonsensical. The example of ‘My sense organs aren’t deceiving me now’ and thus of ‘Our sense organs don’t always deceive us’ and of their role with respect to specific perceptual judgements is just one possible illustration of this train of thought. So a Humean sceptic is right to notice that certain heavyweight assumptions can’t be supported by reasons. Yet, he fails to appreciate that also doubt with respect to them could not be supported by any grounds and would thus be irrational. For rationality requires both claims to knowledge and doubts to be supported by reasons. A point worth noticing is that the insistence on the reason-bounded nature of doubt is not an observation just about our practice which may suggest the idea that we don’t or even can’t raise doubts about the presuppositions of our epistemic investigations because that would be pragmatically impractical, as it would deprive us of those very practices – like forming beliefs on the basis of perceptual and testimonial evidence – which, after all, serve us reasonably well. Rather, it is a point about the logic of any epistemic enquiry, as we have just seen.28 Of course, this is suggestive but far from clear, as it is. In order to clarify a bit the sense of this observation one has to keep in mind that whenever the later Wittgenstein talks about logic, he is in fact introducing the idea of a norm. Now, according to his later views, norms, even those of evidential significance, depend on the actual features of our language games. As he repeatedly stressed, ‘everything descriptive of a language-game is part of logic’ (OC 56, cf. 82, 628). Hence, first of all, if doubts were raised against what makes it possible for us to conduct our epistemic enquiries, this would deprive us of those very practices and, therefore, of all epistemic norms. As a consequence, it would deprive us, and a sceptic, of the very notion of epistemic rationality, which constitutively depends on those very practices, for Wittgenstein (OC 65). Second, if the notion of epistemic rationality depends on those practices
Doubts and the Nonsense of Scepticism 133
and it is a feature of them that doubts be raised on the basis of reasons, to raise them where there can’t, logically, be grounds for them, far from being the most rational move, would in fact betray our very notion of epistemic rationality. Finally, we already know, and will see at much greater length in the following chapter, that whenever a proposition appears to be beyond doubt and justification, its status is not empirical but normative, on Wittgenstein’s view of the matter. If so, a Humean sceptic’s mistake consists, according to Wittgenstein, in raising doubts about heavyweight assumptions failing to notice that whenever a judgement is beyond justification and doubt, its nature will not be empirical, but normative. This, in turn, suffices to place it beyond the possibility of being sensibly doubted. For norms can’t be subject to epistemic evaluation, hence they can’t either be justified or doubted. Of course, they might in principle be changed or abandoned. Yet this won’t be the result of their being epistemically assessed. Rather, it may be a product of a decision or, simply, of contingent facts which may force us to abandon a given norm and perhaps substitute it with another one.29 But, even more radically and, ultimately, independently of viewing heavyweight assumptions as rules, to raise doubts against propositions when there can’t logically be reasons for such doubts betrays our very notion of epistemic rationality. It does so both because it goes against its features which, for Wittgenstein, are dependent on how our epistemic practices are in fact conducted; and because it would deprive us of it, as it would annihilate those practices which, in turn, constitute it. Ironically, then, a Humean sceptic, far from being a champion of epistemic rationality as he is traditionally portrayed to be, would simply go against it, by failing to see that epistemic rationality mandates that whatever plays the role of its condition of possibility be in fact beyond doubt. Finally, it is a consequence of Wittgenstein’s transcendental argument that, although the conditions of possibility of our epistemic practices lie equally beyond doubt and justification, it is a fact that within those practices we do produce justifications for specific empirical propositions which, when true, amount to knowledge. So, the worrying conclusions reached by Humean scepticism – that we never really have knowledge of ordinary empirical propositions – is blocked. Surely, it is always knowledge within a system of justification and therefore by courtesy of some assumptions. But it is knowledge nevertheless. In fact, it shows how knowledge is not boundless or absolute, but always delimited by the kind of epistemic system which gives rise to it. Yet, by Wittgensteinian lights, this is simply what our concept of knowledge is like.30
134
Moore and Wittgenstein
To sum up: • Any reason to call into doubt the presuppositions of perceptual judgements, like ‘My sense organs aren’t deceiving me now’, or ‘Our sense organs don’t always deceive us’, would depend on taking them for granted. • Hence, there can’t be reasons to call them into question. • However, a doubt that isn’t motivated by reasons is irrational, as it would go against the characteristic traits of our epistemic practices which are in turn constitutive of our notion of epistemic rationality. • It would thus be nonsensical, albeit in a non-semantic sense. • Its occurrence, moreover, would deprive us of our (perceptual) epistemic practices and, with that, of our very notion of epistemic rationality, which constitutively depends on them. • Furtheremore, it would be raised against propositions whose failure to be subject both to doubt and justification, makes them normative in nature, and thus beyond the possibility of being epistemically called into question. • Humean scepticism, therefore, far from exposing some fundamental feature of our epistemic practices and, thus, the ultimate nature of epistemic rationality, represents its double betrayal. For, to repeat, it would depend on a misconception of the status of ‘heavyweight’ assumptions and of the self-defeating implications of its doubt, which go against our shared notion of epistemic rationality. • Finally, to acknowledge that ‘heavyweight’ assumptions are beyond justification (and doubt) is consistent with the fact that within those epistemic practices of which they are the conditions of possibility, justifications be produced for or against specific and genuine empirical propositions, which, when true and justified, will thus amount to knowledge. Now, Wittgenstein’s claim that it belongs to the ‘logic’ of our epistemic enquiries that certain propositions be in fact not doubted has given rise to two influential readings of his anti-sceptical views; namely, Crispin Wright’s and Michael Williams’. They both share the intuition that Wittgenstein’s appeal to ‘logic’ should be taken epistemically, rather than normatively, contrary to what I have suggested. That is to say, as eliciting the idea that, after all, for Wittgenstein, the ungrounded presuppositions of our epistemic practices would be somewhat justified, if not even known. It is now time to investigate them.
Doubts and the Nonsense of Scepticism 135
3.2.1
Wright’s rational entitlements
Recently, in a paper titled ‘Warrant for nothing (and foundations for free)?’ (2004a), Crispin Wright has developed this interpretative line and has introduced the term ‘rational entitlement’ in order to characterize the kind of epistemic relationship which, in his view, we bear to a kind of presupposition such as ‘Our senses don’t always deceive us’. Wright has proposed the notion of rational entitlement as a development of the Wittgensteinian ideas just presented.31 But, as a matter of fact, it seems to me that there are several crucial differences between Wittgenstein’s position and Wright’s, at least insofar as the former is taken literally and isn’t elaborated on. According to Wright, any cognitive enquiry, like forming empirical beliefs on the basis of one’s perceptual experience, presents an element of risk. That is to say, if we wish to carry out that enquiry, we can’t do any better than taking it for granted its presuppositions. This, in turn, gives us a rational entitlement, which is a non-evidential kind of warrant for, in our case, ‘Our sense organs aren’t deceiving/don’t always deceive us’. Hence, according to him, a sceptic is wrong when he claims that we aren’t rationally entitled to assume that much. On the contrary, we are, for we have a non-evidential warrant for that presupposition. However, nonevidential warrants such as entitlements fall short of grounding belief and, therefore, knowledge, in Wright’s view. They only make it rational to assume certain propositions that have to stay put if we then want to go on to gather perceptual evidence, which, together with collateral heavyweight assumptions, will give us proper warrant, and in some cases even knowledge, of ordinary empirical propositions such as ‘Here is my hand’, in circumstances broadly analogous to those in which Moore produced his proof. So, while a sceptic is right to insist that we don’t have either evidential warrants for, or knowledge of the most basic presuppositions of our cognitive inquiries, he is wrong to conclude that those presuppositions aren’t rationally held. For a sceptic wrongly takes the category of warrant to be exhausted by evidential warrants, while also rational entitlements should be encompassed. Consequently, he is also wrong to conclude that we can never justifiably believe or know specific empirical propositions, for, given that our assumptions are non-evidentially warranted and that we do have a certain course of experience, then we can warrantedly believe or even know that, for instance, here is one hand. Now, several things are worth noticing, if this is meant to be an interpretation of Wittgenstein’s broadly transcendental argument as evinced from On Certainty. First of all, the element of risk which, in Wright’s view, characterizes our cognitive enquiries is totally absent from Wittgenstein’s
136
Moore and Wittgenstein
considerations. On the contrary, Wittgenstein insists on the tranquillity of our certainty (OC 174, 278, 357, 425), viz. of our taking for granted, or trusting, what reflectively appear to be the presuppositions of our cognitive inquiries. This, however, isn’t due to ‘hastiness’, or ‘superficiality’ (OC 358), in Wittgenstein’s view. Nor is it something that we do in light of the practical and beneficial consequences of behaving that way (OC 131, 338, 474), for example, to be able to pursue in our ordinary cognitive activity of forming perceptual beliefs. Finally, it isn’t something we do simply because of our ‘animal’ nature (OC 358–9, where emphasis should be placed on Wittgenstein’s caveat that the ‘animal’ nature of our certainty and our belonging to a ‘form of life’ aren’t what he thinks can ultimately ground our certainty). Rather, it belongs to the ‘logic’ or ‘method’ of our investigations (OC 342, 151, cf. 56, 82). What this means is that there is no option but to take certain things on trust, if we are to pursue our cognitive activities. But if there is no option, then there is simply no room for risk. After all, one runs a risk whenever it is possible to chose among different options, none of which is completely safe. Although it might perhaps be a practical (Oblomovian?) option not to carry out our ordinary cognitive practices, it would never be an epistemically rational one. After all, as we have seen, the kind of activity we are considering is what paradigmatically contributes to the determination of the notion of epistemic rationality. For that notion crucially depends on the possibility of gathering evidential, perceptual warrants for or against genuinely empirical propositions. So, the real and deeply anti-sceptical point Wittgenstein is making is that sceptics, far from being the champions of epistemic rationality, as they have been traditionally portrayed to be, raise doubts with respect to presuppositions against which justifications can’t logically be gotten and which can’t be doubted on pain of annihilating the very notion of epistemic rationality. The latter, as we have seen, crucially depends on the possibility of gathering perceptual evidence for and against ordinary empirical propositions. Thus, on a Wittgensteinian perspective, sceptical doubts aren’t merely based on too narrow a conception of warrant, which somehow ignores non-evidential ones. Rather, they are altogether nonsensical – that is, irrational; for far from being rational, they can’t be based on grounds and actually undermine the very notion of epistemic rationality. Moreover, it is important to note how the very notion of non-evidential warrant – of entitlement – is totally at odds with Wittgenstein’s views on warrant and properly epistemic notions. It is in fact a characteristic trait of On Certainty to hold that no epistemic category can be applied to what stands fast for us. For these presuppositions of enquiry are
Doubts and the Nonsense of Scepticism 137
said to be ‘neither true nor false’ (OC 196–206), neither (epistemically) ‘grounded’ nor ‘ungrounded’ (OC 110, 130, 166), neither ‘rational’ nor ‘irrational’ (OC 559). Wittgenstein in fact shares the sceptical view that warrants, in this area, are just ordinary perceptual ones and that epistemic rationality requires warrants – that is to say, perceptual justifications. Still, in his view, the presuppositions that make epistemic rationality possible don’t lie outside the scope of it, but, rather, at its limit. That is why they are neither epistemically rational – that is, perceptually justified – nor epistemically irrational – viz. held against perceptual evidence. Like anything which lies at the limit, in the deeply Kantian way in which Wittgenstein has always been concerned with that notion [Grenze] since the Tractatus, they are what makes possible the determination of an ‘inside’ and an ‘outside’. Out of metaphor, in this case the limits are what makes it possible to have perceptual warrants for or against a given empirical proposition and makes it possible to speak either of a rational or of an irrational belief in it, depending on whether such a belief is supported by reasons or goes against them all. Hence, these presuppositions belong to the logic of our investigations not because they are somewhat specially – non-evidentially – warranted, but precisely because they are its unwarranted and unwarrantable conditions of possibility. This, of course, may sound oxymoronic to a somewhat empiricist ear. But this is what a transcendental argument would always purport to show: a condition of possibility of thought and enquiry which, as such, cannot be subject to the constraints that arise only within thought and enquiry. Now, there is no denying that all this is conceptually odd and I think it was one of Wittgenstein’s greatest achievements, within the whole history of philosophy, to try and reinterpret such a notion of ‘limit’ in a conceptually respectable way. It is in fact important to keep in mind that anything which, in a Kantian terminology, would belong to the conditions of possibility of either judgement or enquiry, has, in the later Wittgenstein’s hands, a normative status. Let it be a norm of grammar, hence of language and thought; or else, a norm of evidential significance, like some apparently empirical propositions like ‘Our sense organs don’t always deceive us’, ‘The Earth has been existing for a very long time’, etc. Being normative in nature, these presuppositions of thought and enquiry are simply not in the business of semantic and epistemic evaluation. Hence we can thus finally appreciate how the recourse to normativity was Wittgenstein’s distinctive way out of the problem of making transcendental moves in philosophy respectable. The normative option was thus mandated, for Wittgenstein, by two equally important
138
Moore and Wittgenstein
philosophical sentiments: by his adherence to a broadly Kantian, as opposed to either empiricist or rationalist, view of philosophy and by his rejection of any form of human transcendent category. In Wright’s reading, in contrast, it would appear as a mistake motivated by too hasty an attempt to find an easy way out of scepticism. Finally, it should be kept in mind that Wittgenstein had a rather different view than Wright of what may count as a presupposition of enquiry, or as a ‘hinge’, as we shall see at much greater length in the following chapter. For, it was his view that not just very general presuppositions like ‘Our sense organs don’t always deceive us’ but that also very specific and ‘local’ ones belonged to the logic of our investigations. In particular, that in the case of Moore’s proof, its first premise too wasn’t really something we could get to know or believe in on the basis of grounds. For he held the view that that very judgement – as specific as it is – had to stay put and was in turn a ‘unity of measure’ against which we could ‘measure’ other propositions, perhaps of a very general nature, such as ‘Our sense organs don’t always deceive us’ (OC 250). Hence, Wittgenstein was much more inclined to some form of local holism,32 as far as justifications and the network of presuppositions on which they are based are concerned, than Wright, whose final picture looks far more foundationalist than Wittgenstein’s. This latter aspect of On Certainty has been repeatedly stressed and elaborated on by Michael Williams, whose rendition of the anti-sceptical thrust of Wittgenstein’s considerations it is now time to present. 3.2.2
Williams’ epistemological contextualism
Michael Williams, in a series of papers that appeared between 2004 and 2005, and connected to his Unnatural Doubts (1991), has argued for the view that Wittgenstein ultimately proposed a somewhat epistemic counter to scepticism, whose fundamental features are the adherence to contextualism and the rejection of what Williams calls ‘epistemological realism’. Namely, the view according to which there is a predetermined and absolute structure of justification which is independent of human practice and interests. Williams’ preliminary move (notes 9–13) consists in denying that for Wittgenstein the very general statement ‘There are physical objects’ is a grammatical one. Rather, he argues, contrary to this tenet of the so-called framework reading of On Certainty in general, and of that proposition in particular, that it is just plain nonsense. Hence, that neither it nor its negation can be proved, as there is absolutely nothing to prove (Williams 2004a, 79, 86–7).
Doubts and the Nonsense of Scepticism 139
Williams then goes on to claim that with respect to specific statements or judgements about material objects, such as ‘Here is my hand’, in circumstances like those of Moore’s proof, Wittgenstein held the view that they are certain. But he thinks of certainty as an epistemic category – as an epistemic concept which, however, does not depend on the availability of evidential justifications for the propositions which fall under it (Williams 2004b, 255–6). Indeed, Williams argues that with respect to some specific judgements about material objects, our certainty is as secure as the one we place in judgements about our occurrent sensations, feelings and other mental states, and does not depend on prior knowledge of sense data. Further on, Williams goes so far as to argue that such an epistemic certainty is in fact knowledge, once it is acknowledged that Wittgenstein allowed for the propositions believed to be true, though in some deflationary sense (Williams 2004b, 250, §5), and for the existence non-evidential justifications for them (Williams 2004b, 258, 280). For, like Wright, Williams thinks that this is what Wittgenstein is alluding to in those passages of OC where he writes that the fact that certain things stand fast for us belongs to the ‘logic’ or ‘method’ or our empirical investigations and that, in contrast, I take to be the locus of his ‘transcendental argument’. Whatever plays such a role with respect to our epistemic practices is thus unavoidable and, consequently, in Williams’ view, ‘reasonable’ – that is to say, justified. Yet, contrary to Wright, Williams seems to give an externalist twist to such non-evidential warrants, when he writes that there isn’t ‘anything wrong with the view that there are lots of things we know that we can’t defend by showing, in an everyday citing-evidence, displaying credentials sort of way, how we know them’ (Williams 2004b, 281). In Williams’ view, this is the radically anti-sceptical thought Wittgenstein put forward in On Certainty, which was already somehow envisaged also by Moore (Williams 2004a, 90–2).33 For a sceptic aims to deny that we have knowledge of any of our ordinary empirical judgements, essentially because they are based on the deliverances of the senses, whose reliability can be called into question either through the argument from dreaming, or through the argument about the general unreliability of the senses. Or else, because ‘My sense organs are working reliably (right now)’ appears to be a necessary yet ungroundable assumption of all our judgements about specific empirical objects, whose unwarrantedness reverberates through all judgements ultimately based on it. By contrast, in Williams’ interpretation, Wittgenstein would be maintaining the view that ‘such [specific empirical] judgements are not ordinarily treated as either supportable by evidence or open to
140
Moore and Wittgenstein
question’ (Williams 2004a, 83), yet they are epistemically secure, once we open up our eyes to the fact that knowledge may proceed from non-evidential warrants and get the stamp of certainty by the fact that its objects play a fundamental role in grounding our scientific and empirical investigations at large. Still, in Williams’ view, it is important to note that ‘Here is my hand’ enjoys such an epistemic security only on certain circumstances, while it loses it in different ones, when it may be an open question whether what one is seeing next to oneself, for instance, is really one’s hand. In Williams’ reading, what this shows is that, for Wittgenstein, evidential justification is context-bound – that it depends on a series of ‘methodological necessities’, as he calls them (Williams 2004b, 276) – which are typical of our language games or epistemic practices in general (note 2). Furthermore, the role of methodological necessities (which are known and certain, though unsupported by evidence) is in its turn determined by our needs and choices. As he writes (Williams 2004a, 83), ‘Wittgenstein is well aware (…) that our indulgent attitude [whereby some specific empirical judgements are exempt from doubt] is merely a reflection of our practical exigencies’. This way, Williams rallies Wittgenstein round his own cause – that is the cause of defending epistemological contextualism versus epistemological realism. In particular, he too would have emphasized the role of human interests in determining what must stay put and thus be epistemically secure and what, in contrast, is subject to control and verification. This is what Wittgenstein wants us to do [viz. to break with epistemological realism]. The alternative to epistemological realism is a pragmatic view of norms. The normative structure of doubting and justifying is implicit in practices of enquiry which, as human institutions, are subject to change. There is no immutable order of reasons for the sceptic to discover or exploit. Reminders to the effect that the rules of our language-games, thus the epistemic constraints implicit in those games, are not wholly beyond our control recur in Wittgenstein’s refutation of idealism.34 (Williams 2004a, 95) Finally, in the way of an exposition of Williams’ views, it is important to note that he thinks, and takes Wittgenstein to think, that ‘philosophical context’ is such that practical considerations are set aside, [and] we can put ourselves into an epistemic [justificatory] relation with the most banal everyday
Doubts and the Nonsense of Scepticism 141
certainties. Indeed we can come to appreciate that we always stand in such a relation, even though for practical purposes we may ignore the epistemic [justificatory] demands that this relation imposes. (Ibid.) So, the view is that, as a matter of fact, we would have to enquire even into our most ordinary empirical judgements, yet we don’t do so because of our practical concerns. This way, however, philosophical doubts, at least with respect to specific propositions about physical objects, would appear to be meaningful and intelligible, yet unnatural – whence the title of Williams’ book – because they would be raised irrespectively of those constraints which characterize our ordinary way of doubting. Let us now turn to a critical examination of Williams’ view. First of all, as we saw in §2.1 his preliminary move – viz. to consider ‘There are physical objects’ as plain nonsense, rather than a grammatical statement – is contentious. I argued, in contrast, that it depends on failing to acknowledge the fact that, for Wittgenstein, it is ‘bad’ philosophy which produces nonsense, by misconceiving the nature of certain statements. In this case, it does so by taking this statement to be an empirical one, which, however, should reveal some fundamental aspect of reality. In short, it takes it to be a metaphysical statement. Yet, one should always keep in mind Wittgenstein’s warning: ‘A whole cloud of philosophy [is] condensed in a drop of grammar’ (PI II, xi, 222, italics mine). Bad philosophy produces clouds – that is, confusions and nonsense – by failing to see aright the grammar of language in general or, like in this case, the grammatical nature of specific statements. Moreover, the failure to acknowledge the grammatical status of ‘There are physical objects’ may be due to an exaggerated view of the primacy of use over norms, which, I think, characterizes both Williams’ and so-called resolute or ‘therapeutic’ readings of Wittgenstein’s thought.35 Surely norms do not precede use and depend on it. Nor do we need to follow them self-consciously. We can of course follow them purely practically, by having been trained to play the (language) game, and thus to act in accord with them. Yet use, for Wittgenstein, is not random or such that anything would go. Wherever it happens to draw the line between correct and incorrect employments of words, there does use establish a rule too. Of course the place where the line is drawn may change over time. Yet, to acknowledge the normative nature of certain statements does not run contrary to Wittgenstein’s insistence on the primacy of use; on the contrary, it means properly to recognize its implications. If that is right, it means that, for Wittgenstein, there isn’t really a sharp difference between ‘There are physical objects’, ‘The Earth has existed for a very long time’, or ‘Here is my hand’ uttered in circumstances broadly
142
Moore and Wittgenstein
similar to those of Moore’s proof. They are all ‘hinges’ and thus normative in nature. What, however, they regulate can either be language and therefore thought, or else practices of enquiry and ways of going about gathering evidence for or against genuine empirical propositions. Furthermore, just like Wright, Williams too fails to take proper measure of the fact that, for Wittgenstein, no epistemic category applies to ‘hinges’. Hence, they aren’t warranted, albeit in a non-evidential way. In fact, both Wright and Williams fall short of seeing that Wittgenstein’s argument here is transcendental, in the peculiar sense in which Wittgenstein’s way of thinking can be deemed so. To repeat, at the end of enquiry we discover its conditions of possibility. They aren’t mind- and human-transcendent universal categories, though. Rather, they depend on human language and practices. Moreover, their nature is not synthetic a priori in Kant’s sense of that term, but, as Alberto Coffa (1991) rightly pointed out, though only for Wittgenstein’s intermediate phase, grammatical, or, more generally, normative. To clarify: the important point here, which, according to Danièle Moyal-Sharrock (2004, 163–5), marks the passage from the ‘second’ to the ‘third’ Wittgenstein,36 is that there are judgements, mostly about empirical objects, that we may ordinarily pass on various circumstances, which, however, in context, play a normative role. ‘Here is my hand’, ‘The Earth has existed for a very long time’, ‘My name is AC’ are all cases in point. That they do play a normative role, while also being judgements, can be evinced from the fact that they constitutively contribute to the determination of what would count as, for instance, normal conditions of perception, evidence for or against historical or geological empirical judgements, normal conditions of human functioning and so on. They do so in such a way that, if they were given up, one could no longer count on one’s perceptions, memories, or on apparent testimonies regarding the age of the Earth, to form evidence which could, in its turn, disprove those very propositions. What is more, Williams’ epistemic reading of Wittgenstein’s argument, according to which the latter would end up maintaining that hinges are, in fact, known, is due, in my view, on several important misunderstandings. Exegetically, I think it is mostly based on failing to acknowledge that, for Wittgenstein, there is a grammatical use of ‘I know’ (Ch. 2, §3), which, as weird as it may seem, does not express an epistemic relation between a subject and a proposition. So, many of the passages in which Wittgenstein seems to be saying that after all hinges are known should in fact be read as saying the opposite. Of course, we do find it natural to express ourselves by saying ‘I know’. Yet that twist of phrase conceals the fact that it is not knowledge but the
Doubts and the Nonsense of Scepticism 143
impossibility of being mistaken with respect to ‘p’, which, in its turn, is a function of how the relevant language game is played, we are thereby expressing. Were one unpersuaded of that much, one should always go back to Wittgenstein’s considerations on the use of ‘I know’ in relation to psychological avowals, we explored at length in the previous chapter (§4), where it is utterly clear that whenever ‘I know’ occurs, it is not meant to express an epistemic relation. Another possible textual source of confusion is the fact that sometimes Wittgenstein considers practical knowledge or ‘know-how’ and, in that case, it is obviously unnecessary to be able to offer reasons which would prove that one knows. A know-how, however, isn’t propositional in nature and it in fact amounts to the possession of a given ability. Be that as it may, it should be clear that, for Wittgenstein, when proper propositional knowledge is at stake, it is always dependent on reasons and justifications and on the ability (at least in principle) to offer them as one’s grounds for one’s claim. Hence, as we have already seen in the previous chapter (§2.5), Wittgenstein’s conception of knowledge is internalist through and through. Thus, not only does it make no sense, in his view, to claim knowledge of the hinges, as Williams would have it; but it is altogether false that we nevertheless bear an epistemic relation to them. For we can’t – logically – produce grounds in their favour. To miss that much, amounts to conflating Wittgenstein’s position with Moore’s, in my opinion. For, as I have argued in Chapter 1 (§7), it was in fact Moore who insisted on depicting our relationships to his truisms and the premises of his proof as knowledge, even if he was unable to give a proof of them, or of giving grounds for those propositions which were stronger than them. Hence, it was Moore, not Wittgenstein, who somehow anticipated externalism about knowledge (if not also about justification). If I am right, then, Williams’ reading of On Certainty is doubly mistaken. For, like Wright’s, it wrongly takes Wittgenstein to be proposing a wider conception of warrants, whereby also non-evidential ones should be countenanced, at the expense of properly acknowledging that the normative nature of hinges, for Wittgenstein, makes them unsuitable as possible objects of any kind of epistemic relation. In addition, Williams’ interpretation, contrary to Wright’s, turns non-evidential warrants into externalist ones and takes them as ingredients of a somewhat externalist conception of knowledge, which should ultimately characterize our relationship to hinge-propositions. But this would betray not only Wittgenstein’s conception of the non-epistemic relationship we bear to those propositions, but also his overall account of knowledge.
144
Moore and Wittgenstein
One further sign of how contentious Williams’ reading really is has to do with the fact that, in his attempt to turn Wittgenstein into an externalist about knowledge (and justification), he is forced to play down (at the very least) the fact that in On Certainty, wherever Wittgenstein talks about truth, he seems to characterize it in epistemic terms – as constitutively tied to there being a justification for ‘p’. In Williams’ view, in contrast, Wittgenstein would be endorsing a deflationary view of truth – in fact almost a redundantist one.37 The reasons behind such a reading I think are obvious. If truth were tied to justification, and, as Williams holds, knowledge is, even for an externalist, justified – albeit non evidentially – true belief, knowledge would after all end up being dependent on there being evidential reasons in favour of ‘p’. But once the idea that Wittgenstein might have been an externalist about knowledge is dissipated, much of the motivation for reading him as a supporter of a deflationary conception of truth evaporates. In addition, I think Williams rightly points out one aspect of Wittgenstein’s view of the role of some propositions. Namely, that depending on several features of the context of their use, they may either be exempt from doubt, enquiry and verification; or else, be subject to them. I also think he is right to stress that such a characteristic is clear for ‘Here is my hand’, ‘My name is AC’ and the like, while it isn’t so obvious for ‘There are physical objects’, ‘The Earth has existed for a very long time’ and similarly general or subject- and contextindependent propositions. In the former case we can easily imagine a context where those propositions may be subject to verification and control; while in the latter I think we can’t. In fact, Williams takes this to be a sign of the fact that only those propositions which fall into the first category are properly called ‘hinges’, while the ones falling into the second (save for the problematic case of ‘The Earth has existed for a very long time’) aren’t. But, even so, is this enough to make of Wittgenstein an epistemological contextualist? I think the answer very much depends on what one means by ‘epistemological contextualism’. But if one thinks it means that the structure of justification varies with contextual factors, he wasn’t, first appearances notwithstanding. For, when there is no possibility of giving evidential warrants for ‘p’, there is no issue of there being any epistemic relationship between subjects and ‘p’ either, according to Wittgenstein. So, it is simply not the case that, for him, in certain circumstances we get one kind of justification for ‘p’, while, in others, we get a different kind of justification for it. Rather, in one, probably less frequent kind of case, we do, or at least, can have justifications for ‘p’, while in others,
Doubts and the Nonsense of Scepticism 145
we can’t – logically – have any. So, the changes of context envisaged by Wittgenstein do not so much affect the structure of justifications for ‘p’. Rather, they affect the nature of that proposition, which turns from empirical – and thus justifiable and knowable – to normative – and therefore unsuitable for any kind of epistemic embedding. One might then suggest that at least it was Wittgenstein’s opinion that there could be different overall systems of justification. This would be testified by those remarks in which he envisaged different people who would believe that a king might make rain, or that one could go to the Moon tele-transported, as it were, etc. (OC 92, 106–8, 132, 264, 286). The structure of justification might thus appear to be culturally determined and variable, as a function of human interests, history and so forth. In fact, epistemic contextualism would thereby end up in epistemic relativism. We will have to come back to this issue in the next chapter. For now, suffice it to notice that, first, even for Williams Wittgenstein wasn’t an epistemic relativist;38 second, that on proper examination, the relevant entries in On Certainty have much less relativistic implications than it might first seem;39 and, finally, that there is no hint of such a kind of pragmatic view of epistemic norms in Wittgenstein’s work.40 More specifically, epistemic norms originate in human practice and aren’t mind-transcendent all right. This, however, does not entail that they may be changed at will, or as a function of our practical interests, once they have been established. So, as I read him, Wittgenstein is more inclined to epistemological ‘realism’ than Williams would have us believe. Of course the structure of justifications isn’t independent of human practice – from a genealogical point of view – and it might at least partially change over time, if, for instance, the development of a given discipline imposes new constraints on and practices of verification of specific empirical claims. Yet, the fact that some propositions exempt from doubt and enquiry might become susceptible to them and vice versa, would only be a consequence of a change in their role within our Weltbild – viz. of their losing or acquiring a genuinely empirical status – thereby becoming subject to verification and control or cease to be so. Yet when they turn into verifiable propositions, they become subject pretty much to the same kind of justification or proof that all ordinary empirical propositions are susceptible of – that is to say, perceptual and/or testimonial and/or instrumental evidence. Conversely, when they cease to be empirical propositions, they do not get liable to a different kind of justification – one of a non-evidential nature; rather, they altogether stop being in the business of justification (and knowledge) tout court.
146
Moore and Wittgenstein
Finally, as Moyal-Sharrock perceptively remarks (2004, 157–63), Wittgenstein’s views about the legitimacy of sceptical doubts are totally different from Williams’. In fact, according to the transcendental interpretation proposed, sceptical doubts are irrational and thus illegitimate and nonsensical, at least in the merely epistemic sense which this last term can have. Hence, they are not just ‘unnatural’ – that is to say, contrary to our ordinary way of raising doubts, yet legitimate given the specific features of philosophical context. On the contrary, their being ungrounded makes them irrational tout court, according to Wittgenstein, and therefore, illegitimate and nonsensical, albeit in a way which may fall short of entailing their being incomprehensible strings of sounds. Of course they do reveal something right: that justification and knowledge are bounded by certain propositions which must stay put, though not being justifiable in their turn. Yet sceptical doubts can do nothing – rationally and not merely practically speaking – to shake our confidence in them, for they are raised against and not just independently of any reason. For they are ultimately oblivious to the normative status of those presuppositions which, by being interpreted as eventually empirical, they try, unsuccessfully, to call into question.41 As I hope to have shown, both Wright’s and Williams’ interpretations of some anti-sceptical remarks in On Certainty are in fact deeply at odds with textual evidence. Yet, it should be clear that, as long as they are intended more as developments of some moves one can trace back to that text, than as exegeses of it, I think they are perfectly legitimate and worthy of consideration. The fact that they are very much discussed nowadays may even be a powerful tool to make On Certainty and Wittgenstein alive in contemporary philosophy. A fact that, personally, I can only welcome. Yet it remains that, when looked at through the lenses of Wittgensteinian scholarship, they seem more promising as possible pretexts to go back to that text, to unravel its gems and difficulties such as they are, than as faithful interpretations of it. 3.3
The transcendental argument extended
As already remarked, Humean scepticism can be extended in various ways, to target the numerous ‘heavyweight’ assumptions on which our epistemic practices do in fact rest. Its advantages over Cartesian scepticism, dialectically speaking, are clear. For it does without raising sceptical hypotheses, such as the evil genius or the brain in the vat one, that appear frankly odd and difficult to take seriously. The Humean sceptic would point out that since any test we may run to determine whether we are dreaming would depend on the fact that we
Doubts and the Nonsense of Scepticism 147
are not actually dreaming; there is no way of providing a justification for ‘I am not now dreaming’ as its being either true or at least taken for granted is a precondition of executing any test at all. Since we can’t have evidential warrants either for that assumption, or for its negation, we had better suspend judgement on the issue and realize that it is an ungrounded assumption of all our empirical inquiries, which, in principle, is as legitimate as its negation. Of course there is a point of contact between Cartesian and Humean scepticism insofar as they both place ‘I am not now dreaming’ beyond the possibility of being justified, although they do so by following two different strategies. Yet, clearly, the Humean one relies on fewer presuppositions, in its turn, and is therefore stronger. Thus, were it possible to counter Humean scepticism about dreaming, it would also be possible to defy its Cartesian counterpart. The transcendental argument promises that much. Hence, it is extremely powerful, if successful at all. However, it should always be kept in mind that Wittgenstein did not think of extending it to Cartesian scepticism, or indeed to a kind of Humean scepticism which targeted the very assumption that there is an external world, or, equivalently, that there are physical objects. As we saw in the previous sections, he put forward semantic considerations against the very intelligibility of the dreaming hypothesis (§3.1) and seemed more concerned with pointing out the semantically normative nature of ‘There are physical objects’ (§2.1), rather than its normative role with respect to our epistemic practices. It remains, though, that it can be extended to these assumptions, by noticing how they are presupposed by any cognitive enquiry which would, positively, try to justify, or else, negatively attempt to raise doubts against them. If so, they would both, in a Wittgensteinian spirit, appear to be conditions of possibility of enquiry, whose being beyond both doubt and justification could ultimately be explained by conceiving of them as normative in nature. Therefore, any attempt to raise doubts against them would, first, run contrary to our practice of doubting, and, second, have the potential implication of depriving us of all our epistemic procedures. Either way, a doubt as to whether one might be currently dreaming about the existence of an external world would go against our very notion of epistemic rationality or even divest us of it. Thus, far from being the most rational move one could make, it would in fact either run contrary or destroy the very possibility of epistemic rationality. It would thus be irrational and in this, epistemic sense, nonsensical.
148
Moore and Wittgenstein
I think the transcendental argument is worth considering in the contemporary debate in epistemology. The fact that it can be detached from considerations of meaningfulness, thus endowing sceptical hypotheses with (linguistic) sense, even if they ultimately turn out to be deeply problematical, makes it suitable for consideration in an intellectual climate where (nearly) all philosophers are sympathetic to such a view. Yet, most theorists are equally seeking a way out of scepticism, capable of diagnosing of why sceptical paradoxes are ultimately rationally powerless to establish their intended conclusions, while appearing intelligible and even captivating. The transcendental argument would potentially deliver such an explanation as it would expose the reasons why scepticism is ultimately rationally ineffective as well as the reasons why it keeps us in its grip. Scepticism would thus turn on the difficulty we have of seeing clearly both the nature and the implications of our notion of epistemic rationality. Of course several passages in the transcendental argument, as I have reconstructed it, can be attacked. In particular, one may recoil from conceiving of ‘Here is my hand’, at least in some circumstances, or of ‘I am not now dreaming’, as normative. Yet, I think this is a deep point – viz. that a normative role or function can in fact be played by statements which concern empirical objects and which do not superficially present themselves as rules. For, as weird as it may seem, this is more commonly the case than we may first think. After all, even ‘Full wins over a couple’ presents itself with the appearances of a description, yet plays the role of a rule at poker. Still, as said, this is a claim which clearly invites discussion and it would be worth exploring whether much of the gist of the transcendental argument could be retained by giving that up. That, in its turn, would presumably require tampering with the idea of whether at least heavyweight assumptions could be seen as lying within epistemic rationality, since they make it possible, even if they aren’t, nor can they be, warranted in any way – let it be evidential or otherwise. This would in turn invite embracing a wider conception not of warrants (contra Wright and Williams), but of epistemic rationality itself, as extending to both evidentially warranted empirical propositions and to those assumptions which make it possible to gather such evidence, no matter how unwarrantable they in fact turn out to be.42 Yet, all this would simply testify to the interest that the transcendental argument put forward in On Certainty should raise in ongoing debates in epistemology, when taken as such, or only with minor variations, and not as an epistemic argument in disguise, perhaps with an externalist flavour to it.
4 Wittgenstein: Hinges, Certainty, World-Picture and Mythology
In the second chapter I have analysed Wittgenstein’s considerations on the ordinary use of ‘to know’ and on the grammatical use of ‘I know’ mostly in relation to the premise of Moore’s proof ‘Here is my hand’. No doubt, Wittgenstein’s criticism of the Moorean use of that sentence is a leitmotif throughout the whole of On Certainty, but it does not always have the same end. In the first three sections of the text, the emphasis is on Moore’s anomalous use of ‘I know‘, which goes against the criteria of its ordinary employment in our common language games and is meant to expose the fact that the philosophical use of that expression is nonsensical. In the last and major section of OC, Wittgenstein focuses instead on the role that Moore’s truisms in A defence of common sense, and propositions similar to them play in our lives and in our ‘system of judgements’ (OC 141–2).1 Wittgenstein’s analysis becomes more and more tentative and is full of second-thoughts and doubts, up to the following remark: Haven’t I gone wrong and isn’t Moore perfectly right? Haven’t I made the elementary mistake of confusing one’s thoughts with one’s knowledge? Of course I don’t think to myself ‘The earth has already existed for some time before my birth’, but do I know it nevertheless? Don’t I show that I know it by always drawing its consequences? (OC 397) I don’t think Wittgenstein is really retracting what he has been saying up to this point with respect to both the ordinary and the Moorean use of ‘to know’, but, rather, that he is engaging in a continuous elaboration of the role that propositions like Moore’s play in our language. Moreover, he is ever more conscious of the erroneous representations 149
150
Moore and Wittgenstein
one can make of it and of the fact that he himself is fighting to find out what he really wants to say on the issue. Interestingly, just a little after the previous quote, he writes, ‘Here I am inclined to fight windmills, because I cannot yet say the thing I really want to say’ (OC 400). I also believe that Wittgenstein, in the end, is still holding the view, with respect to the role of these propositions, he put forward in OC 91–9. Namely, that we don’t bear to them a genuinely epistemic relation, such that one could meaningfully claim to know them; though he becomes more and more wary of describing such a relation in many other ways, such as ‘assuming’, ‘believing’ and ‘being certain’. For the point Wittgenstein wants to stress, in the end, is that as a result of our upbringing within a community that shares a language and a form of life, certain propositions ‘of the form of empirical propositions (…) form the foundation of all operating with thoughts (with language)’ and that ‘the statements in question are statements about material objects. And they do not serve as foundations in the same way as hypotheses which, if they turn out to be false, are replaced by others’ (OC 401–2). Hence, all verbs of propositional attitude are somewhat misleading. For they suggest the idea that we may believe these propositions because we have evidence for them, or that we assume them, but they may turn out to be false; or else, finally, that our certainty with respect to them is of a psychological nature and a function of the fact that we find it psychologically impossible to call them into doubt. Moreover, as soon as one claimed to be certain of them, that would have the effect of inserting them within the ordinary language game with ‘I am certain’, which, as we saw in Chapter 2 (§2.2), requires the ability to give cogent reasons in favour of what one claims to be certain of and the possibility of there being a way of verifying these propositions. By contrast, according to Wittgenstein, it is precisely the idea that these propositions could be semantically assessed by means of a method of comparing them with reality which is deeply misleading. For they themselves do determine the method(s) by means of which we can establish whether ordinary empirical propositions are true or false (OC 151). Many interpreters have been baffled by such a view. Since at least Françoise Armengaud’s (1983), up to Danièle Moyal-Sharrock’s reading of On Certainty (2005a), through a number of other interpretations such as Elizabeth Wolgast’s (1987) and Luigi Perissinotto’s (1991), it has been felt as a paradox, or at least, as a problem, that certainties should lose their status as soon as they are spoken and that we may find it impossible to express the relationship we bear to them. Moreover, many interpreters, who have been concerned with this aspect of OC,
Hinges, Certainty, World-Picture and Mythology 151
have tried to connect it with the saying/showing dichotomy exposed in the Tractatus.2 My view is that much of the puzzlement disappears if one keeps in mind that certainties can be said as such at least when we may need to remind ourselves or others of their role and that we can perfectly well preface them with ‘I know’, provided that expression is meant as a grammatical remark, and not as an expression of a genuinely epistemic relation. In such a case, ‘I know’ could also be replaced by ‘it stands fast (for me and many others)’ (OC 116); or else, by ‘I can’t be wrong’ or ‘A doubt is logically excluded’ as to whether p, where ‘p’ is substituted by one of Moore’s truisms and propositions alike them (Ch. 2, §3.2). Yet, no doubt it is problematical to realize that there is a risk in trying to describe our relationship to these propositions by means of our ordinary epistemic and cognitive vocabulary. So, small surprise that Wittgenstein too, as he realizes that danger, becomes more and more tentative and wary of his own forms of expression. It should not be forgotten that Wittgenstein, like all of us, is immersed in the language and system of thought whose fundamental features he is trying to clarify. It is simply an effect of that proximity that as soon as we use certain forms of expression, ordinarily meant to express certain kinds of relationships between, in this case, subjects and Moore’s truisms and propositions alike them, we are immediately and mostly inadvertently misled into taking them as we ordinarily and rightly do. Much of Wittgenstein’s philosophical insight has always been to feel uneasy within his own language, when used as a tool to clarify itself and our own way of thinking and to sense that various forms of expression could disguise and travesty things. Yet this does not really point to, or ends up in, a form of ineffability with respect to certainties and our relationship to them, as if, by having noticed that our ordinary ways of expressing things could betray them, we were forced merely to contemplate them in silence. As I have already said, I think Wittgenstein put forward a precise view with respect to the role and nature of Moore’s truisms and alike, as well as with respect to the kind of relationship we bear to them. Moreover, I believe that he thought all this could be said in our language, though paying attention to the peculiar characteristics of certain forms of expression. While many interpreters missed or partially overlooked this fact, many others have actually been completely misled by Wittgenstein’s use of ‘I know’, and have failed to appreciate its grammatical role, in context. They have then wound up thinking that Wittgenstein was, in the end, somewhat concurring with Moore in thinking that we do know these propositions, though – mysteriously – we could never claim such knowledge. In Chapter 3, we have already seen this epistemic reading, originally due to
152
Moore and Wittgenstein
Thomas Morawetz (1978) and lately developed, in particular, by Crispin Wright, Michael Williams and Duncan Pritchard, by reference to its antisceptical implications. In the context of the present chapter, we will have to take it into account once more, in order to criticize it as a possible explanation of Wittgenstein’s positive view with respect to the role of Moore’s truisms and of our relationship to them. We will also have to consider a number of other interpretations of so-called hinge propositions (OC 341, 343), that all somehow run contrary to what has become known as ‘the framework reading‘, which is the one I myself support, though with some caveats and developments with respect to more traditional versions of the same interpretation. So we will take into account the naturalist reading of On Certainty put forward mainly by Peter Strawson, but also by Wolgast and Gertrude Conway, in the middle of the 1980s, as well as the therapeutic reading, advanced during the last decade, mostly by interpreters such as James Conant, Edward Minar, Alison Crary and Ruper Read. Finally, by exposing the version of framework reading I personally favour, we will have to address two further issues. Namely, whether Wittgenstein did in the end propose a somewhat new form of foundationalism, as Avrum Stroll (1994) and more recently Danièle MoyalSharrock (2005a) have maintained; and whether, with respect to hinges, his views lead to some form of epistemic relativism.
1 Only apparently empirical propositions: Examples and preliminary considerations Writes Wittgenstein, I want to say: propositions of the form of empirical propositions, and not only propositions of logic, form the foundation of all operating with thoughts (with language). – This observation is not of the form ‘I know…’. ‘I know…’ states what I know, and that is not of logical interest. (OC 401) In this remark the expression ‘propositions of the form of empirical propositions’ is itself thoroughly bad; the statements in question are statements about material objects. And they do not serve as foundations in the same way as hypotheses which, if they turn out to be false, are replaced by others (…). (OC 402)
Hinges, Certainty, World-Picture and Mythology 153
First of all, Wittgenstein insists on the fundamental role of some propositions which have the form of empirical propositions. Then he denounces his conceptual imprecision by saying that this expression – ‘propositions of the form of empirical propositions’ – is bad, that is to say, misleading. It is confusing because it alludes to the fact that the status of a proposition can be evinced from its form, while – on this view – it can only be evinced from the role it plays in a language game.3 Wittgenstein goes on to deny that the role these propositions play is that of hypotheses, viz. of propositions which could be abandoned if the comparison with facts should falsify them. Yet, it should be noticed that to assimilate them to hypotheses might have seem promising, because in scientific disciplines hypotheses determine what counts as a proof within the theory; they aren’t normally questioned, at least not in the same way as, say, the results of a measurement or an experiment may be; and, finally, they constitute a model of interpretation of phenomena and a paradigm for the construction of those propositions, within a theory, which must be confronted with reality.4 So, why did Wittgenstein deny that these propositions are like scientific hypotheses? I surmise he did so because hypotheses can somewhat be compared with facts, at least insofar as we may get enough evidence to disconfirm them. By contrast, these propositions, which look like empirical ones but aren’t, can’t be confirmed or disconfirmed, in his view. What this suggests is that if we do abandon them, perhaps in light of new findings, it is not because this new evidence has proved them false. In other words, the propositions Wittgenstein is talking about are such that, in principle, we could decide to retain them in the face of ‘contrary’ evidence. Yet we may decide not to do so, because that would simplify things for us, or allow for similar explanations of different phenomena, and so forth. This idea, I think, has nowadays become commonplace through the work of those philosophers and historians of science, like Thomas Kuhn and Paul Feyerabend, who have held the view that the passage from one scientific paradigm to the other, such as the one from a Ptolemaic universe to a Copernican one, isn’t decided on the basis of evidence which proves wrong some central tenet of the former and confirms some equally fundamental tenet of the latter. After all, even the apparently backward movements of the planets could be explained by making the original Ptolemaic theory more complex, through the introduction of the notion of epicycles. Rather, at some point people found it easier and more expedient to adopt the Copernican theory. Hence, ‘The sun goes round the Earth’ was never disconfirmed, only abandoned to uphold its contrary, in light of the practical benefits of doing so.
154
Moore and Wittgenstein
Now, although I think this comparison with a certain conception of science and its development can help the contemporary reader better to understand Wittgenstein’s point, I believe it is important to keep in mind, first, that Wittgenstein never talked of scientific theories and their hypotheses; on the contrary, as we have seen, when it comes to science and its hypotheses he seems to hold a much more realist view of their nature and role. Second, that, as we shall see in the following, he never really stressed the seemingly pragmatist idea that, depending on our needs or wants, we could abandon these fundamental propositions. Yet, with these caveats in mind, I think the analogy with some trends in the history and philosophy of science can help at least see why Wittgenstein denied that these propositions could be likened to hypotheses. To stress, they aren’t like scientific hypotheses for him, because they can’t either be confirmed or disconfirmed. Moreover, Wittgenstein denies that they are empirical propositions, or propositions of experience (Erfahrungssätze). More specifically, he denies that they play the same role, or have the same function as empirical propositions, within the ‘system’ of our empirical judgements (OC 136–7); even if, between these two classes of propositions there isn’t a sharp boundary. That is to say, there isn’t any rule whose application can determine whether a given proposition is or isn’t genuinely empirical. We have already seen, in the way of an elucidation of this point, how ‘(I know) here is my hand’ can either be genuinely empirical, say after a car accident, or else non empirical in circumstances such as those of Moore’s proof. Still we can’t provide ourselves with a rule which could decide the issue, a priori, as it were; rather, we can only observe the role that proposition plays with respect to our language games. These two latter points together amount to the view that a given sentence – that is to say, a particular string of words of a natural language such as English – can express different propositions depending on the context of its use. So, for instance, ‘(I know) here is my hand’ can either express an empirical proposition, given certain circumstances, or a non-empirical one, in different conditions. Similarly – and this is a contextualist point less familiar to contemporary readers – one of Wittgenstein’s favourite examples, that is, ‘Nobody has ever been to the Moon’, used to express a non-empirical proposition; though, now, the same string of words no longer does. So, it is not the case that, in a Quinean spirit, the analytic/synthetic divide can’t be drawn. That is to say, that for any sentence we can never determine whether it is ‘analytic’ or ‘synthetic’, that is, whether it expresses a proposition which is subject to verification and control or exempt from it. Rather, the point
Hinges, Certainty, World-Picture and Mythology 155
is contextualist. Depending on the context of its both synchronic and diachronic use, the same string of words can come to express two different kinds of proposition: an empirical one, or a non-empirical one. In this connection, Danièle Moyal-Sharrock has introduced the idea of doppelgangers.5 ‘I know (here) is my hand’ has a doppelganger and they can each figure in different contexts, so that, in the circumstances of Moore’s proof, a non-empirical proposition is expressed by that sentence, while, after a car accident, say, an empirical one is expressed by its doppelganger. One may doubt whether we need the idea of a doppelganger to capture the point Wittgenstein is making in these passages. After all, one might have thought that we just have different tokens of the same sentence type – that is to say, of the same string of symbols interpreted as part of English – which may express two different propositions in different contexts of their use. But, as a matter of fact, introducing the idea of doppelgangers may be useful in order to make these ideas compatible with Wittgenstein’s further claim, we have seen at length in the previous two chapters, that a given string of phonemes or graphemes, while apparently having a meaning and thus seeming to be a sentence of some natural language, may in fact fail to have one, if employed in ways which run contrary to its various actual uses in different language games. If so, we could no longer really talk of sentences’ types, whose different tokens could be used to express different propositions in different contexts. Hence, the idea of doppelgangers could come handy to stress this point. Another way of putting it is, I think, the following: we can’t determine either the sentences or the propositions expressed by a string of phonemes or graphemes before looking at their actual use. And it may turn out that identical strings of phonemes or graphemes could be used to express different propositions, in different contexts, or even nothing at all, in such a way as to amount or fail to amount to a meaningful sentence of a given natural language, contrary appearances notwithstanding. As we have seen in Chapter 2, for instance, ‘I know here is my hand’ could exemplify all these possibilities, according to Wittgenstein. It could be used to express an empirical proposition, say after a car accident, when one is laying in bed and looking at one’s limbs; it could be used to express a non empirical one when used simply to make a grammatical and/or normative point; or else, finally, it could just be sheer nonsense when uttered by a philosopher such as Moore. Still, it is important to notice that while in On Certainty the possibility for a string of phonemes or graphemes of expressing different propositions in different contexts of use, as well as the possibility of their amounting or failing to amount to a meaningful
156
Moore and Wittgenstein
sentence of a natural language, depending on the context of use, are both present and perhaps run together, in general, they need not be. Yet this is the kind of outcome a radical interpretation of the slogan ‘meaning is use’ should in fact lead to. So, I think that in point of exegesis, Moyal-Sharrock’s recourse to the idea of a doppelganger can actually be extremely illuminating. By contrast, I don’t think it helpful to follow Moyal-Sharrock all the way down to her claim that only empirical propositions are in fact propositions for Wittgenstein. As we saw in Chapter 2 (note 30), Moyal-Sharrock insists that bipolarity – that is to say, the possibility of being both true and false – is the mark of propositionality. Since, in Wittgenstein’s view, Moore’s truisms and alike are neither true nor false, they aren’t propositions either. If they aren’t, they can’t be the object of propositional attitudes and this suggests that the kind of certainty Wittgenstein is endeavouring to characterize in his book is of a non propositional nature too. It is, in Moyal-Sharrock’s view, some kind of animal certainty we mostly display in acting with confidence – with a confidence that doesn’t know doubt – with respect to objects and people in our surroundings. While we will have to come back to the notion of certainty, for now it is enough to repeat a point already emphasized. Namely, that in On Certainty ‘proposition‘ (Satz) is an umbrella term, which is explicitly claimed not to have sharp boundaries (OC 320) and thus to be applicable both to genuinely empirical propositions and to non-empirical ones. This claim would really be odd if Satz were meant as ‘sentence‘, as Moyal-Sharrock suggests. For reflect: what does it mean to say that a sentence doesn’t have ‘sharp boundaries’? It could mean, at most, that before being used, we can’t say whether a string of phonemes or graphemes does belong to a given natural language and is thus a sentence of it, or not. But, clearly, when read in context, the remark at OC 320 is meant to express an altogether different idea. Let us quote the relevant set of passages in full. ‘The question doesn’t arise at all.’ Its answer would characterize a method. But there is no sharp boundary between methodological propositions and propositions within a method. (OC 318) But wouldn’t one have to say then, that there is no sharp boundary between propositions of logic and empirical propositions? The lack of sharpness is that of the boundary between rule and empirical proposition. (OC 319)
Hinges, Certainty, World-Picture and Mythology 157
Here one must, I believe, remember that the concept ‘proposition’ itself is not a sharp one. (OC 320) Isn’t what I am saying: any empirical proposition can be transformed into a postulate – and then becomes a norm of description. But I am suspicious even of this. The sentence is too general. One almost wants to say ‘any empirical proposition can, theoretically, be transformed …’, but what does ‘theoretically’ mean here? It sounds all too reminiscent of the Tractatus. (OC 321) The point Wittgenstein seems to be making here is that given attested uses in our language, certain strings of phonemes or graphemes can either express empirical propositions or rules. In both cases, they would amount to sentences of the language, since signs occurring in them would have a meaning, as they would have a use. Yet, even when they express rules, they do express propositions, but propositions that aren’t subject to verification and control, that is, propositions which aren’t empirical but normative. The problem is what one wants to make of this claim. There is a lightweight reading of it, which goes as follows: don’t make too much fuss over the term ‘proposition’, the important point is the distinction between empirical propositions – which are subject to semantic evaluation and epistemic control – and rules – one may call them ‘normative propositions’ – which are neither semantically evaluable nor epistemically assessable. There is, however, also a much stronger reading of that claim, which says that some empirical propositions, because of their attested use, can in fact turn or be turned into norms. Here the emphasis is on the fact that what is a description can become so remote from verification and control, given how in fact our language and epistemic practices have developed, that it is then turned into a rule, or a normative proposition, in fact into a norm of description (OC 167, 321). That is to say, that as the result of use, its role is like that of a norm. What this would suggest is that Wittgenstein is here fighting with the Kantian notion of synthetic a priori judgements. That is to say, he seems to be talking of propositions which while being about objects and also passed in judgement as descriptions, like ‘Here is my hand’, even in the circumstances of Moore’s proof, come to be exempt, in those very circumstances, from the kind of procedures to which we subject ordinary empirical
158
Moore and Wittgenstein
ones, and thus become like rules or norms which are in fact free from semantic and epistemic evaluation. Similar things could be said for ‘The Earth has existed for a very long time’ and many other truisms: they can be the content of judgements, that describe standing or passing states of affairs, yet be exempt from the kind of semantic and epistemic evaluations our ordinary empirical judgements are answerable to. If, then, we must acknowledge that there are propositions that can partake at once of normativity and of descriptivity, it becomes understandable why Wittgenstein insists on the lack of sharpness in the boundary between rules and empirical propositions. To wit, one and the same proposition can have elements of both. If so, the very concept of proposition is intrinsically vague, stretching across descriptivity and normativity. It is difficult, in my opinion, to decide once and for all what Wittgenstein meant. Whether he pointed out that the same sentence type can be used sometimes to express an empirical proposition, sometimes a rule, though there is no rule which can decide the issue a priori; or whether he struggled with the idea that one and the same proposition, while being a description, could be hardened into (something like) a rule; or, finally, whether he meant that unique strings of phonemes and graphemes could be used in such a way as to express different propositions, and even nothing at all, depending on the context of their use, thus being or failing to be a sentence of a given natural language. At least the first two readings are compatible with another famous set of passages in On Certainty, viz. those, we will look at more thoroughly in the following, in which Wittgenstein claims that ‘it might be imagined that some propositions, of the form of empirical propositions, were hardened [into a rule] and functioned as channels for such empirical propositions as were not hardened but fluid’ and that ‘this relation altered with time’ (OC 96, my emphasis), and yet he insisted that while it is wrong to think that ‘logic too is an empirical science’, it is right to say that ‘the same proposition may get treated at one time as something to test by experience, at another as a rule of testing’ (OC 98, my emphasis). While the third reading is, as we have seen, what would help better make sense of Wittgenstein’s claim that philosophers’ uses of words are often nonsensical. I think all these readings are compatible with this or that passage in On Certainty and that we can’t really hope to give one single interpretation which could fit them all. However, I think that once we take seriously the idea that meaning is use, these three views can be held together. For use determines if and when a string of phonemes or graphemes is
Hinges, Certainty, World-Picture and Mythology 159
meaningful and hence a sentence of a natural language; if and when it expresses an empirical proposition or a rule; and it is again use which establishes if and when a given proposition which, perhaps originally, was meant just as a description, becomes exempt from verification and control in such a way as to function as a rule (or vice versa). But none of these interpretations seems to me to force us to give up the idea that ‘proposition’ (Satz), at least in On Certainty, is an umbrella term, which may be applied to genuinely empirical propositions, actual rules and to propositions which, while originally descriptive, have been hardened into rules and behave like norms, though they may, in time, return to be treated just as descriptions and no longer as norms of description. Another important reason to take seriously the idea that it is use which always determines the meaningfulness and function of signs is Wittgenstein’s denial that what he is trying to convey is the idea that ‘any empirical proposition could theoretically be turned into a rule’ (Wittgenstein actually says ‘postulate’, but postulates are rules for him). The latter would be a claim a Quinean would endorse, while Wittgenstein is insisting that it is only the actual development of our language games that can determine the role and function of meaningful sentences, which then express propositions. Moreover, it is equally an anti-Quinean point to deny that we can no longer distinguish between logic, in the extended sense this term has for the later Wittgenstein, and empirical sciences. For, according to Wittgenstein, by observing the actual use of those meaningful sentences in our language we can determine whether they are used to express empirical propositions, linguistic rules or norms of description. Although their function can change in time, and even if the very notion of proposition may be blurred up to the point of exhibiting, in some cases, both normative and descriptive aspects, it remains that, at any given moment we can, by looking at how we actually play our language games, determine the function specific strings of words have with respect to them. Let us now turn to Wittgenstein’s positive characterization of Moore’s truisms and propositions alike. In OC 42, he tells us that they are about material objects. However, if we look at the various examples in the text, we can easily realize that this isn’t entirely true. For instance, we have seen that there are propositions that are about categorial concepts, such as ‘There are physical objects’, or ‘A is a physical object’; there are exemplifications of the principle of induction, like ‘In the same circumstances a given substance A will always react in the same way with a given substance B’; there are, moreover, paradigmatic propositions which determine unities of measure, like ‘Water boils at 100°C’.
160
Moore and Wittgenstein
Of course there are many propositions which are about material objects, such as ‘The Earth has existed for a very long time’, ‘Cats don’t grow on trees’, ‘The Sun isn’t a hole in the vault of heaven’, etc. Nevertheless, it is clear that not all instances of propositions whose role is other than (merely) descriptive, are about material objects. Failing a global characterization of these propositions on the basis of their content – of their being about material objects – as well as on the basis of the role that each of them invariably plays within our various language games (at least on a certain view of what Wittgenstein meant by Satz); we can only formulate some general considerations on the role they have with respect to our language games, when it so happens that they function as norms of representation, or, equivalently, as propositions which characterize a method, as opposed to genuine descriptions and to propositions within a method. That is to say, when it so happens that they belong to the foundations – and hence, for Wittgenstein, to the grammar – of our language games. We have already, at least partially, seen some of their general characteristics in the last two chapters. For instance, that, according to Wittgenstein, it would be nonsensical to doubt them, both because such a doubt would be irrational and because any alleged doubt with respect to them would actually be meaningless. Moreover, that it would be nonsensical to claim to know them, because in fact all reasons we could produce in their favour would actually presuppose them. Connectedly, that these propositions are such that we somehow uphold them, even if they have never really been subject to verification and control. Furthermore, that even to mention them would usually be nonsensical, since they are trivialities all of us are party to, and that to say them would produce, at most, either a comic or curious effect. All these features make them unlikely to be embedded in language games with epistemic expressions such as ‘to know’ and ‘warrantedly to believe’, but also with those expressions of propositional attitude that are somehow connected with evidence and grounds, such as ‘to believe’, ‘to be certain’ and alike. Finally, as we have already seen, Wittgenstein’s suggestion is to use, in their connection, other twists of phrase, such as ‘it stands fast for me (and many others) that…’, ‘Here a mistake is logically excluded’, ‘I can’t be wrong about them’, etc. To sum up: • There are three possible interpretations of hinges. According to the first, the same string of phonemes/graphemes can express either an empirical proposition or a rule, or fail to express anything at all, depending on context.
Hinges, Certainty, World-Picture and Mythology 161
• According to the second, the same sentence of a given natural language can express either an empirical proposition or a rule, depending on context. • According to the third, the very same proposition can play either an empirical or a normative role, depending on context. • While different, these three interpretations can be made compatible with one another, by taking seriously the idea that meaning is determined by use. • In any event, there is no reason to abandon the view that when hinges are at stake, we are dealing with propositions which, however, play a normative role. • Finally, not all hinges are in fact propositions about objects, though most of them are. Let us now turn to the development of these initial characterizations.
2 Propositions that characterize a method. The hinges around which all other propositions rotate Let us look at scientific practice. Running experiments presupposes that there are questions to be answered and hypotheses to be verified. However, no one doubts of the existence of the instruments one is using in order to settle those issues. As we have already seen in the last chapter (§1.3), it is crucial that not everything be called into doubt, for, otherwise, we could no longer answer our legitimate questions (or even pose them, for that matter). Similarly, if I do a calculation, I can doubt the fact that I have done it correctly, but not the fact that figures on paper change while I don’t look at them. For, otherwise, how could I discover my possible mistake? (OC 337). Again, ‘I might suppose that Napoleon never existed, and is a fable, but not that the earth did not exist 150 years ago’ (OC 186), for to doubt the long existence of the earth would impair all historical pieces of evidence, even those of which ‘I cannot say (…) that [they are] definitely correct’ (OC 188). To repeat, a point by now familiar, if the earth hadn’t existed 150 years ago, everything we now regard as historical evidence within that discipline – documents, findings, testimonies, etc. – would lose its status. It would just amount to a set of pieces of paper and objects devoid of any historical significance. We can then say that in this very specific sense ‘The Earth has existed for a very long time’ is an assumption [Annahme] or a presupposition [Voraussetzung] (OC 411, 168) of doing history. This, however, doesn’t mean to say that that assumption/presupposition gets formulated or is explicitly entertained while we do history (cf. OC 411, 153); nor that
162
Moore and Wittgenstein
it is the result of an arbitrary decision on historians’ part,6 or a hypothesis which may be falsified (OC 402). Rather, that proposition, and many others similar to it, have been inherited with our being drilled to take part in the various practices – language games – which belong to our form of life, as part of a shared world-picture (OC 93–94).7 Symmetrically, however, what we do regard as historical evidence tells us about events which have happened in a remote past and therefore prove that the earth has existed for a very long time. As Wittgenstein writes in OC 190: What we call historical evidence points to the existence of the earth a long time before my birth; – the opposite hypothesis has nothing on its side. (OC 190, cf. 89, 93, 117, 119, 144–5, 153, 182–3, 185, 191, 203, 312, 421, 594) If we then add geological data to historical evidence, with which we can reach a reasonable assessment of the age of the earth, we can also say that our entire system of proof speaks in favour of the very long existence of the earth, and nothing against it. But, then, wonders Wittgenstein: If everything speaks for an hypothesis and nothing against it, is it objectively certain? One can call it that. But does it necessarily agree with the world of facts? At the very best it shows us what ‘agreement‘ means. We find it difficult to imagine it to be false, but also difficult to make use of it. What does this agreement consist in, if not in the fact that what is evidence in these language games speaks for our propositions? (Tractatus logico-philosophicus) (OC 203) This passage is one of the most controversial entries in On Certainty. Wittgenstein himself had deleted the first part of it in the manuscript and it isn’t clear how one should interpret the reference to the Tractatus which occurs in the end: whether as a sign of having changed his mind with respect to that earlier work, or as meaning ‘it sounds all too reminiscent of the Tractatus’. Nor is it clear why, in this context, Wittgenstein refers to ‘The Earth has existed for a very long time’, as a hypothesis. Let us take care of these difficulties in reverse order. To accommodate this last point we could chose two different interpretative lines. We could say that Wittgenstein’s use of ‘hypothesis’ here
Hinges, Certainty, World-Picture and Mythology 163
is sloppy. Alternatively, we could hold – and I am leaning towards such an interpretation – that if that proposition occurred as a hypothesis, for instance in the questions of a child, we would appeal to our system of proof in history and geology, which we accept, in its turn, as a result of what we have been taught (OC 162), to teach that child the ‘truth’ of that presupposition. However, as we have seen, everything we have learnt as a system of proof in history and geology presupposes the ‘truth’ of ‘The Earth has existed for a very long time’ and such a presupposition has been ‘swallowed’, as it were, together what we have been explicitly taught (OC 143). Accordingly, what we have learnt explicitly gives support to what then appears to be one of its presuppositions, without which what we have been taught would make no sense. We will have to come back to this interconnection between what is presupposed by a system of proof and that very system in connection with the issue of whether Wittgenstein proposed a (perhaps new) form of foundationalism in On Certainty. However – in order to address the first issue on our present agenda – even when an entire system of proof speaks in favour of a proposition, which, in its turn, plays a foundational role with respect to that system, does it mean that that presupposition unconditionally agrees with the facts? Not really, as OC 205 makes clear: ‘If the true is what is grounded, then the ground is not true, nor yet false’. For if, on the one hand, what lies within our method of verification can be said to be true when what we consider as evidence speaks in favour of it (and false otherwise); on the other, what is itself necessary in order for that system of evidence to exist as such, can’t be said to be true or confirmed on the basis of what in effect presupposes it. Nor can we (logically) find reasons or evidence to think that presupposition false. Yet, even if our entire system of evidence speaks in favour of ‘The Earth has existed for a very long time’, we can’t say that it is a fact, or that it is unconditionally true that things are the way that propositions describes them as being. Of course, we could say – thus running the risk of generating some confusion – that it is true and that it agrees with the facts. But this would only in effect show what ‘being true’ and ‘agreeing with the facts’ really means for us (but not for the author of the Tractatus). Namely, that what we regard as evidence speaks in favour of a given proposition which we therefore regard as ‘true’ and that facts are whatever propositions are confirmed (as fallibly as it might be) by our system of evidence.8 What, then, appears to us to be objectively certain is not a fact or a true proposition (in the sense required for instance by a correspondence
164
Moore and Wittgenstein
theory of truth) which ‘strikes’ us as immediately true (OC 204). ‘It is not a kind of seeing [a fact, or the truth of a proposition] on our part; but our acting – [that is to say, judging] –, which lies at the bottom of the language-game’ (ibid.). Hence, the propositions which belong to the ‘ground’, to the ‘foundations’, of all out thinking and acting aren’t themselves grounded in, and made objectively true by mindtranscendent facts. Rather, they are held fast by everything which rotates around them – our system of judgement – and yet owes its status of system of evidence to their being held fast. As Wittgenstein put it: The questions that we raise and our doubts depend on the fact that some propositions are exempt from doubt, are as it were hinges [die Angeln] on which those turn. (OC 341) That is to say, it belongs to the logic of our scientific investigations that certain things are in deed [in der Tat] not doubted. (OC 342) But it isn’t (my emphasis) that the situation is like this: We just can’t investigate everything, and for that reason we are forced to rest content with assumption. If I want the door to turn, the hinges must stay put. (OC 343) My life consists in my being content to accept many things. (OC 344) Hence, propositions like ‘The Earth has existed for a very long time’ are like hinges that must stay put in order for a system of proof to exist. However, according to Wittgenstein, if it is true that hinges are kept fixed by what rotates around them, like the fact that in deed we do have historical and geological investigations,9 one should not think that the latter are made possible by some kind of explicit decision among historians or geologists not to call into question certain propositions, for example, ‘The Earth has existed for a very long time’. Therefore, ‘we can’t investigate everything’ isn’t a pragmatic constraint, due to a practical choice, or to the realization that we would never be able to exhaust an infinite number of tests and controls. Rather, it is a logical – viz. a grammatical – constraint. For otherwise, as we saw in the previous chapter (§3.2), we would no longer have a system of verification which
Hinges, Certainty, World-Picture and Mythology 165
could speak for or against a given hypothesis, and, with that, we would lose the very notion of epistemic rationality. Thus, not to call everything into question isn’t a pragmatic choice, rather, it is a way of acting (viz. of judging) that is itself constitutive of epistemic rationality. Wittgenstein extends these considerations, originated in the analysis of scientific enquiry, to all our practice of judging, raising questions and doubting. He is not interested in the genealogy of those propositions which function as stable hinges for the other ones that rotate around them, although they are indeed connected to our form of life.10 Actually, he puts forward the hypothesis that once upon a time they might have been subject to verification and control (OC 211), but that, since ‘unthinkable ages’, they have belonged ‘to the scaffolding of our thought’ (ibid.), to our ‘method of doubt and inquiry’ (OC 151) and have formed ‘the foundation of all our operating with thoughts (language)’ (OC 401) and of action (OC 411). This is an important point, also in contrast to what Moore said, in ‘A defence of common sense’, about the origin of our knowledge of his truisms.11 Contrary to what Moore maintained, Wittgenstein doesn’t think that we were ever given a proof of them, perhaps at the origins of our ability to take part in our various linguistic practices. Rather, when we acquired that ability we came to master those linguistic and non-linguistic practices, characteristic of our form of life, and thereby absorbed or ‘swallowed’ what lies at their foundations since their very beginning, which has never been proved, nor produced by some kind of reasoning. To put it otherwise, by being drilled and taught to speak our language and to take part in the practices that belong to our form of life, we have inherited what, through the history of human kind, has come to play the role of a foundation for our thought and language. Namely, those propositions that ‘lie apart from the route travelled by inquiry’ (OC 88), are ‘removed from the traffic’ and, as it were, are ‘shunted onto an unused siding’ (OC 210), of which Moore’s truisms are just some examples. Others may be ‘My name is A.C.’, ‘Every human being has parents’, ‘This is blue’, said of an object in clear view, ‘The Earth has existed for a very long time before my birth’, ‘Cats don’t grow on trees’ and ‘Nobody has ever been on the Moon’ (said before 1969). For it is part and parcel of our ability to speak a language that we know the name of usual colours and objects in our surroundings; just like it is an ingredient in our ability to take part in the language game with people’s proper names that each of us, as a rule, knows his/her own name. If we deviated from these regularities, people would try to find the causes of our, perhaps,
166
Moore and Wittgenstein
momentary lack of memory, such as having ingested some drug, having been injured and caused to be amnesiac or being senile. Equally, it is part and parcel of being taught a lot of other things, such as history, geology and the like, that we take it for granted that the earth has existed for a very long time before our birth; as is equally a function of what people were taught in the way of their scientific upbringing, that, before 1969, ‘Nobody has ever been on the Moon’, functioned as a hinge. Finally, propositions like ‘Cats don’t grow on trees’ and ‘Every human being has (two biological) parents’ belong again to our shared world-picture, partially determined by certain observed regularities, as well as about the kind of common view about how animal and human life function, given at least what we now know to be physically possible. Hence, were a person not to hold these hinges, we could try to find out whether he knows something we all ignore. Still, failing that, we would think him mad, or at least that he belongs to a very distant culture, which we would not take as a serious alternative to ours, not, at any rate, on these issues.12
3 On the groundlessness of the foundations Avrum Stroll in his Moore and Wittgenstein on Certainty (1994) and,13 more recently, Danièle Moyal-Sharrock in her Understanding Wittgenstein’s On Certainty (2005a)14 have maintained that Wittgenstein developed an extremely original kind of foundationalism in his last work. A sort of foundationalism which, in particular, would be coherent with his own methodological dictum that description should replace explanation. Here I will concentrate mostly on Stroll’s defence of this interpretation. In his view, Wittgenstein’s foundationalism would distinguish itself from other forms of foundationalism, which have been put forward throughout the history of philosophy, in positing a categorial difference between what grounds and what is grounded. In particular, according to Stroll, Wittgenstein would have maintained that what grounds a language game isn’t part of it and in fact lies outside of it. Moreover, according to Stroll, foundations can be external to the language game either in an absolute or in a relative way. They are relatively external to it when one of the propositions which belong to the foundations of one or more language games, is part of some other language game. By contrast, they are absolutely external, when the relevant propositions aren’t part of any language game. ‘Here is my hand’ and ‘There are physical objects’, for instance, would exemplify each of these two categories respectively.
Hinges, Certainty, World-Picture and Mythology 167
According to Stroll, the originality of Wittgenstein’s form of foundationalism would be preserved by both analyses of it he offered: one according to which certainty is propositional in nature and one according to which it is not, and is rather displayed in our instinctive behaviour which knows no doubt. Clearly – claims Stroll – human behaviour is categorically different from language games with epistemic expressions and propositional attitude verbs. However, also the propositions which serve as foundations can’t be said to be propositions in the ordinary sense, as they aren’t subject to truth and falsity, or to verification and control, thus failing to be part of our language games with ordinary empirical propositions. Although there is no denying that in On Certainty Wittgenstein used a lot of foundational images,15 I personally think that he did not propose any new and sui generis form of foundationalism, based on the heterogeneous nature of what grounds and what is grounded. The reason why I find Stroll’s thesis misleading is that Wittgenstein didn’t participate in the epistemological, and therefore philosophical enterprise of establishing the solid foundations of all our knowledge. Now, it seems to me that outside such a project, it really makes little sense to talk of foundationalism – albeit sui generis. Wittgenstein’s ‘foundations’ are simply what is kept fixed by our actual way of playing different epistemic language games and this shows, on the one hand, that what appears to be fixed is so not because of its intrinsically stable status, but because of what ‘rotates’ around it; and, on the other, that foundations aren’t a product of philosophical reflection but of our actual epistemic practices. All philosophy does is to attract attention onto them and, in fact, on their ungrounded nature. As Wittgenstein put it, ‘The difficulty is to realize the groundlessness nature of our believing’ (OC 166). Hence, I therefore fully agree with Michael Williams who, in ‘Why Wittgenstein isn’t a foundationalist‘ (2005), forcefully argues that although Wittgenstein used a lot of foundational images in On Certainty he can’t be considered a real foundationalist. For foundationalism is more than simply a structural thesis and requires meeting further conditions that aren’t actually satisfied by Wittgenstein’s hinge propositions. These constraints, for Williams, are (i)
holding that there are universal and clearly specifiable foundations; while Wittgenstein’s hinges are often local; (ii) maintaining that there are criteria to tell apart grounds from what is grounded. Yet, as we have seen, for Wittgenstein there aren’t a priori rules which can determine, for any given proposition, whether it is going to belong to the foundations or not. It is only
168
Moore and Wittgenstein
the actual development of our language games that determines the role a given proposition comes to play. Furthermore, hinges are very heterogeneous in nature, for they can either be mathematical or psychological propositions as well as propositions about physical objects, laws of nature, and so forth; (iii) holding that grounds and grounded propositions are independent of each other; while, clearly, Wittgenstein maintained that hinges are like axes, which are kept fixed by what rotates around them;16 (iv) that the passage from foundational to non-foundational propositions doesn’t require any appeal to non-basic propositions. In contrast, Wittgenstein held the view that as a result of changes in our practices what appeared to be fixed – for example, ‘Nobody has ever been on the Moon’ – could cease to look that way.17 Nevertheless, contrary to, for instance, Joachim Schulte (2005) and other scholars, I don’t think Wittgenstein proposed a form of coherentism either. For, again, coherentism is an epistemological thesis about the nature of justification and knowledge (often accompanied by a parallel theory of truth). Wittgenstein, however, was mainly interested in certainty, rather than in knowledge and justification and construed it as an utterly non-epistemic notion. So, once more, although there is no denying that he also used a lot of coherentist images,18 he did not put forward a coherence theory of certainty, for that would have required to think of it as an epistemic notion, which he denied. So, in my view, being extraneous to traditional epistemological projects, Wittgenstein proposed neither a novel form of foundationalism, nor a new kind of coherentism.19 Rather, he endeavoured to clarify the notion of certainty taken to be other than epistemic in nature. Such a clarification, if successful, is indeed relevant to epistemology. For, in fact, it would speak to two of its major concerns, viz. the nature of those propositions which have to stay put in order for us to have any epistemic practice and our relationship to them. Yet it is the characteristic feature of such a conceptual clarification that these propositions are actually excluded from semantic and epistemic appraisal and that our relationship to them is deemed other than epistemic. There are, however, two theses Stroll and Moyal-Sharrock put forward, that I find particularly worth discussing. First, that in On Certainty Wittgenstein distinguished between relative and absolute foundations – that is, between relative and absolute hinges – and, second, that he proposed two different characterizations of certainty: one propositional and one non propositional. We will examine them in this order.
Hinges, Certainty, World-Picture and Mythology 169
3.1
Absolute and relative foundations
According to Stroll, Wittgenstein held that there are relative foundations, constituted by propositions which may be set aside the route on which some of our language games travel, while being part of it in some other language games; and absolute ones, comprising propositions which are never part of our language games. In my view, this is partly true and partly false. It is true insofar as it is correct to say: (a) that there are propositions, like ‘Here is my hand’ and ‘My name is A.C.’, etc. that can both belong to the ground and to our ordinary and present epistemic language games; (b) that there are propositions like ‘Nobody has ever been on the Moon’ that have played the role of hinges, but that no longer do so; (c) that there are propositions like ‘The Earth has existed for a very long time’, or the principle of induction, for which we can’t, given our present situation, conceive of how they may be called into question or doubted. If so, it is correct to distinguish between (a) synchronically relative hinges; (b) diachronically relative ones; and (c) contingently absolute hinges. It is false, in contrast, insofar as Stroll’s claim may be taken to suggest that there are necessarily absolute hinges. For, even if it may sound oxymoronic, also absolute hinges are, on Wittgenstein’s view, only contingently so. That is to say, we can’t, at present, given our world-picture, imagine credible, non-fictitious circumstances which would make these propositions play a role other than that of hinges. Accordingly, they appear to us as absolute. But nothing excludes that something really unheard of could happen, which could force us to revise our entire system of empirical judgements. Thus, what now appear as absolute hinges would no longer be so.20 Moreover, Wittgenstein repeatedly warns us that there is no rule which can determine a priori the role a proposition can have with respect to our language games and epistemic practices. Therefore, it is only use which can decide the issue, and use may evolve in unpredictable ways. Thus, the distinction between absolute and relative hinges is itself the product of a specific world view and doesn’t have any metaphysical grounding or the stamp of necessity. So, to put it in an altogether different way, the analytic/synthetic distinction, which for Wittgenstein is a distinction between grammatical and empirical propositions, is not given once and for all. Moreover, it can change depending on context and time. Yet, contrary
170
Moore and Wittgenstein
to what a Quinean would hold, it can always be traced at one particular moment or with respect to one particular context of use. So, to sum up and conclude, • Hinges are kept fixed by what rotates around them and aren’t themselves grounded in mind-independent facts (§2). • Wittgenstein didn’t propound any new form of foundationalism; nor, for that matter, a novel kind of coherentism (or a sui generis mixture of both). For he didn’t participate in the epistemological project of giving a theory of knowledge or justification. Rather, he was interested in certainty and gave a thoroughly non-epistemic account of it (§3). • Among hinges, he implicitly distinguished between (a) synchronically relative ones; (b) diachronically relative ones; and, finally, (c) contingently absolute ones (§3.1). • All these distinctions can never be drawn a priori, by appealing to some rule or other, but only by looking at our actual linguistic and epistemic practices (§3.1). 3.2 Propositional and non-propositional certainty. Is certainty ineffable? According to Stroll, in On Certainty there are two different characterizations of that very notion in tension with each other: one of a propositional kind, and one of a non-propositional, but pragmatic or even animal kind. The former belongs to the earlier parts of the work, the latter to the later ones. The former characterizes all those passages where Wittgenstein discusses Moore and explicitly talks of the propositions which play the role of hinges of our thinking and acting. The latter, in contrast, is advanced in passages such as Giving grounds, however, justifying the evidence, comes to an end; – but the end is not certain propositions’ striking us immediately as true; i.e. it is not a kind of seeing on our part; it is our acting, which lies at the bottom of the language-game. (OC 204, cf. 110) However, I think both textual evidence and further considerations should in fact lead us to a different conclusion. First, regarding textual evidence, it is important to keep in mind that Wittgenstein had serious misgivings with respect to his own characterization of certainty as
Hinges, Certainty, World-Picture and Mythology 171
‘a form of life’ or as something ‘animal’. As he wrote in OC 358–9 (my emphasis): Now I would like to regard this certainty, not as something akin to hastiness or superficiality, but as a form of life. (That is very badly expressed and probably badly thought as well). But that means I want to conceive it as something that lies beyond being justified or unjustified; as it were, as something animal. The caveat in the first passage, as well as Wittgenstein’s qualified forms of expression in the second suggest that here he is more alluding to some features of certainty which could be more easily grasped if we likened it to animal certainty, than straightforwardly saying that certainty, for him, is of a non-propositional, but merely instinctual or animal nature. Second, in OC 204, the acting is seen as the ground of our language games, yet again this quote doesn’t straightforwardly speak in favour of ‘animal certainty’, as it were. For, surely, actions are neither true nor false. But, what is to be stressed is that they are simply not in the business of semantic evaluation. Hence, it would be a somewhat hollow remark to point out that they are neither true nor false. Moreover, it is simply unfortunate that Stroll doesn’t recognize that also judging, for Wittgenstein, is an action. As he put it in OC 232: ‘Our not doubting [certain propositions taken as a whole] is simply our manner of judging, and therefore of acting’. So, I think there is no real contrast in Wittgenstein between saying that certainty is displayed in action and saying that it is displayed in the way we judge, for instance, that this object standing in front of me in clear view is my hand and we would not admit anything as proof to the contrary. Finally, and more generally, there would be little of philosophical significance in Wittgenstein’s last work if he were just trying to bring into focus animal certainty. For, surely, nothing would be achieved – philosophy-wise – by saying that we instinctively behave in ways which are categorically immune from doubt, as they would not be guided by propositional attitudes. That has never been the issue for sceptics like Descartes or Hume, who would happily concede that much. The point would be, rather, whether such animal certainty could ever be enough to counter sceptical doubts. For, if it turned out that it is just a habit, however ingrained it may be, this would precisely show what all sceptics have been claiming for a long while: that our epistemic practices aren’t rationally grounded. Although we may find it inescapable to act in a given way, this doesn’t show at all that what, reflectively,
172
Moore and Wittgenstein
will appear as its ungrounded presuppositions – like the existence of an external world, the fact that we aren’t currently dreaming, that the earth has existed for a very long time and so forth – are rationally supported in such a way as to be able to sustain rational scrutiny. As a consequence, far from presenting any way out from scepticism, On Certainty would just play in its hands. In more detail, On Certainty would be hopeless with respect to Humean scepticism, contrary to what Peter Strawson has maintained in his influential reading of it,21 because Humean scepticism is just, in the end, the acknowledgement of the fact that we take for granted certain things in our acting and thinking, with no rational support for doing so. As is familiar, Strawson argued that in On Certainty Wittgenstein put forward a naturalist conception of hinges, which somehow parallels Hume’s. The main point of difference, in his view, is that while Hume appeals to Nature, Wittgenstein appeals to the role of our upbringing within a community that shares a language and a form of life. In virtue of that upbringing we couldn’t help accepting certain hinges, while lacking reasons and grounds in their favour. Finally, Strawson claims that for Wittgenstein scepticism cannot be rebutted by argument, but simply recognized as idle and unreal, because it would call into doubt what we can’t help believing, given our shared form of life. As I argued in the previous chapter, however, Wittgenstein isn’t merely exposing sceptical doubts as unnatural, but as nonsensical in the twofold sense of either being irrational or even altogether meaningless. So, the naturalist reading of hinges would, on the one hand, simply embrace the outcome of Humean scepticism, and, on the other, counter it with claims which are much weaker than Wittgenstein’s. Nor would the naturalist interpretation of hinges do any better with respect to Cartesian scepticism. For, again, it would be an outcome of that kind of scepticism that of course we act and think with no doubt as to whether we might be dreaming, but we can neither rationally support the corresponding belief, nor rationally dismiss the dreaming hypothesis save by saying that it would be unnatural. Finally, it is important to keep in mind that neither Descartes nor Hume deny the ineluctability of the ‘animal’ and instinctive certainty we display in our acting and thinking. Hence, a naturalist reading of certainty would just end up endorsing the consequences of traditional scepticism. I therefore think that Wittgenstein’s definitive view on the nature of certainty wasn’t that it is of an animal, non-propositional nature. But, of course, there is no denying that, at least at places,22 he also talked of this kind of certainty. Hence, the question is: how do the propositional
Hinges, Certainty, World-Picture and Mythology 173
and the non-propositional account of certainty go together, if they do? In order to answer this question, we will do well to take up MoyalSharrock’s distinction between certainty as an attitude, to be further specified, and certainties, which are propositions that are certain for us.23 Hence, we should distinguish further between certainty of an instinctive, animal kind and certainty with respect to hinge propositions. Let us start with the former. As already stressed, it is of very little philosophical significance, for no real sceptic – such as Descartes in his sceptic guise and Hume – ever denied it. Furthermore, being nonpropositional in nature, it could not be doubted, because doubt necessarily concerns a given proposition. Anyway, we can see, in the way of its description, that it concerns objects in our surroundings but also – and this differentiates us from other animals – the testimony and teaching of other people. So we act with respect to objects around us with no doubt, like when we go and fetch a chair from another room, without thinking for a moment it might have disappeared while we weren’t looking at it. Similarly, we take in what other people and school books tell us, without thinking for a second that they may be lying. This way of acting would, however, be incorrectly described as a trust, in my view. For trust presupposes a positive attitude to be contrasted with one of mistrust. Wittgenstein’s point, however, is that especially at the beginning of our upbringing we are drilled to react in certain ways to given orders, or explanations of the meaning of words, or to believe in what we are taught. At that stage, there is no positive attitude of trust either in objects, or in certain patterns of behaviour of objects or people, or in the latter’s words, for there simply can’t be an attitude of mistrust. We instinctively act, mostly as we are expected to, in a way which knows no doubt, and which is not motivated by an explicit trust in objects and people.24 Surely from a third-personal point of view we often find it expedient to describe it as a trust. But we should be wary of not projecting this form of description back into infants’ and children’s minds that would often lack the concepts which would be necessary to entertain the propositions which would be the objects of this alleged trust. Now, as a result of this upbringing, we find ourselves saddled with a great number of certainties, of a very heterogeneous nature, which we can only recognize reflectively when we either think about the issue or are confronted with challenges to them. But when we do reflectively recognize these certainties, we do entertain them in thought as propositions, perhaps for the very first time. Yet, in Wittgenstein’s view, their being indeed removed from our ordinary epistemic language games
174
Moore and Wittgenstein
should make us realize that, while being propositions, they in fact play the role of norms or rules. Finally, it remains to characterize our relationship to these propositions, whose role is that of norms. As Moyal-Sharrock and other interpreters25 have emphasized, it is thoroughly non epistemic and thus unlike belief, knowledge or even psychological certainty, for Wittgenstein. What they have failed to notice, in my view, however, is that this doesn’t entail, nor supports the view that it is animal or instinctual, and hence utterly non-propositional. After all, it has propositions – albeit of a normative kind – as its objects. So, what kind of propositional attitude may it be? I think an analogy would be useful here. Think of our rules of conduct, for example, we ought to stop at traffic lights when red. In fact we can and do say that we accept such a norm. That is to say, we behave in accord with it and would either reproach those who don’t, or accept to be reproached ourselves if we didn’t. Furthermore, we teach our children to obey it. This pragmatic sense of accepting a proposition whose nature is normative, to be contrasted with an epistemic sense of accepting empirical propositions for which we don’t have (enough) evidence,26 and with respect to which it would make utterly no sense to talk of epistemic evidence as it has norms as its objects, is, I submit, what Wittgenstein is alluding to when he talks of certainty with respect to hinges qua hinge-propositions.27 Now, the interesting aspect of our acceptance of hinges is that while being a pro-attitude, it is itself constitutive of epistemic rationality, and not merely a pragmatist acceptance due to an evaluation of its expected practical utility. Moreover, it is worth stressing that not all pragmatic acceptances of propositions need be explicit. It may well be that we have either been taught them once or that we were taught other things which entailed them and have been taught or drilled to obey them. By then being used to behaving in accord with them, we may well no longer give a thought to these norms of, as it were, epistemic conduct, when we do in fact make empirical judgements, or subject empirical propositions to verification and control. Yet, to attribute to ourselves and others such an acceptance helps make sense of our behaviour. Such an attribution, in its turn, is correct because, contrary to what would happen in the case of a corresponding attribution of a propositional attitude to subjects who still have no or very little concepts, it would not put into their minds more thoughts and propositional attitudes than there might possibly be. Yet, this pragmatic acceptance of rules tallies with many of Wittgenstein’s remarks which allude to our certainty as displayed in our way of acting. In particular, to acknowledge this sense of ‘accepting’
Hinges, Certainty, World-Picture and Mythology 175
allows us not to go all the way down the path of mere ‘animal certainty’ and is compatible with saying that we accept, often implicitly and as a result of our upbringing, norms both of language and of evidential significance, and ‘instinctively’ behave in conformity with them, in ways which show no doubt. It must be stressed, moreover, that such an instinctive behaviour is not the one of the animal, but that of the human animal. That is to say, of the creature who has been drilled and taught to use language and to take part in our various epistemic practices and is thus already fully conceptual and a competent epistemic agent. If, then, there is still room for instinct here, it is one concerning our so-called second nature, not our merely first and animal one. Be that as it may, what, in my view, Wittgenstein is really interested in in On Certainty is the characterization of certainties more than the characterization of certainty – let it be animal or pragmatic yet relative to hinge propositions. As we have seen at length, when he talks (very rarely) of hinge propositions as certainties, they are ‘certain’ because of the role they play, qua norms, with respect to our language games. That is to say, they are certain in the sense of being removed from doubt and such that any possible doubt with respect to them would be nonsense in either of the two senses – that is, irrational or meaningless – elucidated in the previous chapter. Being so removed from the traffic, as it were, they appear to be norms and it is their being norms that makes them immune to doubt and therefore certain. So, ‘animal certainty’ is really not the focus of On Certainty, though it is a precondition for being drilled to take part in our various language games. What On Certainty centres on are rather those propositions which, as a result of our upbringing within a community that shares a language and a form of life, we can reflectively identify as being exempt from verification and control as well as from doubt. This makes them play the role of norms, either of language or of evidential significance. Moreover, with respect to them, we do practically (as opposed to epistemically) accept them – that is to say, we behave in accord with them, in ways which know no doubt – where our doing so is, in its turn, constitutive of epistemic rationality. Hence, to go back to our initial question, raised by Stroll’s claim that Wittgenstein gives two different characterizations of certainty in tension with one another, I think we can conclude the following. That there would be a tension between Wittgenstein’s different ways of characterizing certainty if we thought he was interested in animal certainty on the one hand, and in hinge propositions on the other. For the two are categorically different. If we acknowledge instead that his focus was
176
Moore and Wittgenstein
on hinges and on our practically accepting them, albeit often implicitly, and therefore on our acting in accord with them, in a way which shows no doubt, the tension, I submit, disappears. Thus, to sum up: • We can distinguish between certainty and certainties (or hinges); • as well as between animal certainty and certainty with respect to hinges. • Animal certainty is non-conceptual and instinctive; it concerns our actions with respect to objects and people in our surroundings, as well as our reaction to their testimony. Though interesting in its own right, it isn’t the focus of On Certainty. • Certainty regarding hinges, in contrast, can be characterized as an acceptance of norms, which can’t be subject to any epistemic constraint. We may call it a ‘pragmatic acceptance’, as opposed to an ‘epistemic’ one. • Certainties are in fact those hinges – viz. those normative propositions – which are the objects of that pragmatic acceptance that characterizes our non-animal certainty. • Hence, pragmatic acceptance, as we may call it, is conceptual, even if it can be and it often is merely implicit and tacit. Yet, we can attribute it to subjects (when and) because they do in fact have the concepts necessary to entertain the relevant propositions and do behave in accord with them. An interesting consequence of this understanding of certainties and certainty with respect to hinges is that it allows us to avoid the ineffabilist reading,28 supported, for instance, by Moyal-Sharrock (2005a) and Wolgast (1987). In their view, as soon as our certainties are said, they become doubtful and open to epistemic scrutiny.29 This is so because, in their opinion, hinges are entirely non propositional and displayed, as such, merely in action. So, no wonder that as soon as we speak them, we inevitably insert them into our ordinary epistemic language games and make them open to epistemic enquiry. Connectedly, as soon as we remove them from their proper environment – that of unreflective and in fact non-conceptual action – also their characteristically animal certainty is lost and can no longer be purported by our various forms of expression, not even by saying that we are certain of them, or that we accept them. Hence, they conclude that both hinges qua hinges and certainty as animal certainty – which, for them, is the only certainty there really is – are ultimately ineffable. The most we can do, with respect to hinges, according to Moyal-Sharrock, is to give heuristic formulations of them, which
Hinges, Certainty, World-Picture and Mythology 177
may be used to drill someone to take part in our language games or to remind philosophers of the real – viz. normative – status of hinges. By contrast, by defending the view that hinges are propositions, though normative rather than descriptive in nature, and that our certainty with respect to them is a kind of acceptance which displays itself in our acting in accord with them, we make both certainties and certainty with respect to them effable. Of course, I think Moyal-Sharrock is right to claim that when we utter hinges qua hinges we mostly do so with heuristic purposes in mind of the kind she correctly identifies. Yet, by acknowledging their propositionality hinges become indeed sayable as such – that is to say, as norms.30 Moreover, one can better appreciate how their being said can serve a recognized communicative purpose: either that of instructing someone about the norms of our language or epistemic practice, or that of reminding someone – usually philosophers – of their very nature against their tendency often to misconceive it. For, reflect, if hinges qua hinges were ineffable, how could one really motivate the idea that their being spoken in certain circumstances could serve a communicative purpose, however heuristic that might be? So, to conclude, I think supporters of ineffabilism are right to notice that Wittgenstein was utterly unhappy with any kind of rendition of certainty by means of an epistemic and cognitive vocabulary, ranging from ‘accepting/assuming’ in the epistemic sense, to ‘being certain’, through ‘believing’ and ‘knowing’; and that he was worried about a certain way of speaking certainties which would make them look like empirical propositions to be inserted into our descriptive and epistemic language games. Yet, this does not force us to abandon the view that, for him, hinges are propositions, albeit normative rather than descriptive in nature, that they can be said as such, though essentially with heuristic purposes in mind; and, finally, that our certainty with respect to them consists in our (pragmatically) accepting them and therefore behaving, in our thinking and inquiring, in accord with them, in ways which know no doubt. If so much is right, we can finally assess also two other influential interpretations of hinges. The epistemic one, initiated by Morawetz (1978) and further pursued by Wright (2004b), Williams (2004a, b) and Pritchard (2010); and the therapeutic reading, maintained by Conant (1998) and additionally developed by Minar (2005), Crary (2005) and Read (2005). As already repeatedly stressed,31 the former’s mistake consists in ignoring that, in Wittgenstein’s view, there is a grammatical use of ‘I know’, which serves no epistemic purpose and that may well take as its objects propositions which aren’t empirical or descriptive in
178
Moore and Wittgenstein
nature, but normative. Hence, somewhat like the ineffabilist reading, the supporters of the epistemic reading fail to see that propositions, for Wittgenstein, can either be empirical/descriptive or normative; or else, to put it differently, that the term ‘proposition’ can cover both norms and descriptions. By ignoring this, they then go on to think that any use of ‘I know’ in On Certainty is descriptive of an epistemic relationship between a subject and a proposition, and that it must therefore take as its object propositions which, connectedly, can’t but be empirical. We have already seen in the previous chapter how contentious as a rendition of Wittgenstein’s anti-sceptical strategies this reading would be. But, given the focus of the present chapter on hinges, it should be pointed out how strongly it would go against Wittgenstein’s insistence that hinges are (like) rules. In section 1, I have already explained the sense of the latter qualification – to repeat: it may well be that Wittgenstein was at times (more or less consciously) attracted to the view that judgements mostly about objects – and thus having as contents empirical propositions – could actually play the role of norms of description, in given contexts. But even if one were inclined to such a Kantian reading of hinges, I don’t think it would follow that, for Wittgenstein, in those contexts, those propositions would have merely an empirical – that is to say, a descriptive – function. To clarify the point: ‘Here is my hand’, stated or judged in circumstances such as those of Moore’s proof, would be a description of a state of affairs which, however, would play a normative role in that context, for it could in turn be used to test one’s eye sight, one’s linguistic competence and so forth. So, to conclude, even on their most empirically oriented, yet still supported by textual evidence interpretation, hinges could not be the straightforwardly empirical propositions the epistemic reading would make of them. However, also the therapeutic reading ultimately makes a similar mistake. For it ignores all the positive – I dare say, constructive – elements in On Certainty both in the way of a description of the role of hinges and on the grammatical use of ‘I know’. Scholars of a therapeutic persuasion rightly insist on the fact that, when spoken by Moore and a sceptic, ‘I know’ and ‘to doubt’ would literally make no sense for Wittgenstein, as we saw in Chapters 2 and 3 respectively. However, they fail to notice that, for him, there is also a grammatical use of ‘I know’, which can be appealed to in order correctly to characterize the kind of non-epistemic relation we bear to certain propositions. Furthermore, when it comes to these propositions, Wittgenstein wasn’t merely appealing to our ‘sensibilities’ and linguistic sense to make us realize that they stand fast
Hinges, Certainty, World-Picture and Mythology 179
for us.32 Rather, he put forward several suggestions – in fact one single suggestion that could be declined in rather different ways – about what their status really is and therefore about what it means that they stand fast for us. Namely, that, at least in context, they are (like) rules – viz. norms of language and thought as well as of evidential significance, we have been brought up to behave in accord with and which we could not give up, given that they themselves determine what is to count as evidence. Therefore, doubts about them can’t be supported by reasons and are thus irrational (and therefore nonsensical).
4 World-picture In OC 93 Wittgenstein introduces the expression ‘world-picture’ (Weltbild) for the first time. He writes, The propositions presenting [darstellen] what Moore ‘knows’ are all of such a kind that it is difficult to imagine why anyone should believe the contrary. E.g. the proposition that Moore has spent his whole life in close proximity of the earth. – Once more I can speak of myself here instead of speaking of Moore. What could induce me to believe the opposite? Either a memory, or having been told. – Everything that I have seen or heard gives me the conviction that no man has ever been far from the earth. Nothing in my picture of the world [Weltbild] speaks in favour of the opposite. Wittgenstein introduces the expression Weltbild with respect to Moore’s truisms: these propositions are peculiar because it would be difficult for any competent speaker to imagine their opposite. More specifically, memories and testimonies all speak in favour of them. Since memories and testimonies constitute the best part of our system of proof, we can’t imagine proofs that might induce us to believe the opposite. We haven’t learnt from experience that we can trust memories and testimonies. Hence, in this respect, our system of proof isn’t of a hypothetical nature. On the contrary, we have learnt how to verify and control information and hypotheses on the basis of memories and testimonies, which therefore characterize the method of our inquiries and verifications and which, as such, can’t always be called into doubt. Since these elements of my world-picture speak in favour of the fact that I (like Moore) have never been too far away from the surface of the earth, in order to imagine the opposite, I should in fact conceive of a different system of verification and control. This may
180
Moore and Wittgenstein
sound surprising, but reflect, it is essential to the very idea of memories and testimonies, as opposed to, say, fantasies and dreams, that we can distinguish between correct and incorrect instances of them. Hence, the possibility that the content of all of them could in fact be wrong, which we should entertain in order sensibly to call into doubt ‘I have never been too far away from the surface of the earth’, would in fact annihilate our very method of enquiry. If so, in order really to doubt of ‘I have never been too far away from the surface of the earth’ there should be a different method of verification and control, other than the one we currently have, based on memories and testimonies that speaks in favour of that proposition. As Wittgenstein himself puts it, Might I not believe that once, without knowing it, perhaps in a state of unconsciousness, I was taken far away from the earth – that other people even know this, but do not mention it to me? But this would not fit into the rest of my convictions at all. Not that I could describe the system of these convictions. Yet my convictions do form a system, a structure. (OC 102) All testing, all confirmation and disconfirmation of a hypothesis takes place already within a system. And this system is not a more or less arbitrary and doubtful point of departure for all our arguments: no, it belongs to the essence of what we call an argument. The system is not so much the point of departure, as the element in which arguments have their lives. (OC 105) Suppose some adult had told a child that he had been on the moon. The child tells me the story, and I say it was only a joke, the man hadn’t been on the moon; no one has ever been on the moon; the moon is a long way off and it is impossible to climb up there or fly there. – If now the child insists, saying perhaps that there is a way of getting there which I don’t know, etc., what reply could I make to him? (…) But a child will not ordinarily stick to such a belief and will soon be convinced by what we tell him seriously. (OC 106) Hence, from within a given world-picture that determines what counts in favour of a hypothesis and the methods of verification and control, as well as of judgement and description, it is impossible to imagine a
Hinges, Certainty, World-Picture and Mythology 181
different one. That is to say, it is impossible to conceive of different methods of representation and control because it is on the basis of the ones we have got that we know what it means to describe and verify how things are. But it is also impossible to conceive of a radically different world, so to speak, viz. of contents other than those that characterize our actual memories and testimonies, our judgements and descriptions, for the latter and the corresponding concepts depend on the possibility of distinguishing between right and wrong instances of them. According to Wittgenstein, we have learnt to draw such a distinction on the basis of examples, which have been taught to us as instances of correct (or incorrect) memories, testimonies, judgements and descriptions. In OC 94–99 Wittgenstein then offers a metaphorical characterization of the notion of world-picture, that is worth citing in full. But I didn’t get my picture of the world by satisfying myself of its correctness; nor do I have it because I am satisfied of its correctness. No: it is the inherited background [der überkommene Hintergrund] against which I distinguish between true and false. The propositions describing this world-picture might be part of a kind of mythology. And their role is like that of rules of a game; and the game can be learnt purely practically, without learning any explicit rules. It might be imagined that some propositions, of the form of empirical propositions, were hardened and functioned as channels for such empirical propositions as were not hardened but fluid; and that this relation altered with time, in that fluid propositions hardened, and hard ones became fluid. The mythology may change back into a state of flux, the river-bed of thoughts may shift. But I distinguish between the movement of the waters on the river-bed and the shift of the bed itself; though there is no sharp division of the one from the other. But if someone were to say ‘So logic too is an empirical science’ he would be wrong. Yet this is right: the same proposition may get treated at one time as something to test by experience, at another as a rule of testing. And the bank of the river consists partly of hard rock, subject to no alteration or only to an imperceptible one, partly of sand, which now in one place now in another gets washed away or deposited. We shall now examine the crucial elements of this highly metaphorical characterization of the notion of world-picture one by one.33
182
Moore and Wittgenstein
4.1
Our world-picture is inherited
According to Wittgenstein, as we have already remarked several times, it is typical of the fact that we have been drilled to the use of language in a community that shares many linguistic and non linguistic practices, that some ‘facts’ and the apparently empirical propositions which ‘describe’ them get swallowed within such a training. We have also seen already how this might happen. We are drilled to act both linguistically and non-linguistically in certain ways and we are credited as competent members of our community when we share with others the same kind of confidence in the use of language and in taking part in our usual epistemic practices. Such an upbringing, however, doesn’t consist merely of passing definitions on to children, or of merely methodological characterizations. Rather, it is mostly made of examples of judgements, based on our methods of verification and control and of calculations, etc. These examples, in their turn, function as paradigms of correct judgements and calculations and contribute to our learning what it means to calculate or to judge. Thus, they help determine our concepts of correct/incorrect calculation, or of true/false judgement. We can therefore imagine that from such a primitive mastery of language, judgement and calculus more and more complex forms of judgement, verification and control, as well as of calculus evolve, which, in their turn, will be passed on to future generations. Now, as we have already seen, some ‘facts’ expressed by some propositions – viz, by our hinge propositions – appear to be presupposed by everything we have been drilled and taught to do both linguistically and non linguistically, for example, that there are physical, mind-independent objects; that the earth has existed for a very long time, that objects and figures on paper don’t appear and disappear of their own accord, etc., even if these very propositions have hardly ever been made explicit during that drill. We have already seen how this might happen. To repeat, we have been drilled to act in ways which presuppose the ‘truth’ of these propositions, by being asked to go and fetch objects currently unperceived; by being taught to ask who had moved them, in case they weren’t where we thought they would be; by being asked to verify the formula written on paper to control the result of a given calculation; or, finally, by being taught geology or history and thus of objects and events which go back to a distant past and that have always been or have taken place on the (surface of the) earth. What, however, is important to keep in mind is that none of these hinges, according to Wittgenstein, depends on some mythical agreement between human beings, nor are they empirically grounded in
Hinges, Certainty, World-Picture and Mythology 183
facts. Rather, whatever their precise origin and evolution have been, they arise out of our actual human history, as part of the life of our species, and haven’t been determined by some kind of – perhaps philosophical – reasoning. 4.2 Our world-picture is neither true nor false, neither grounded nor ungrounded, neither rational nor irrational According to Wittgenstein, our world-picture is neither true nor false34 because it consists of propositions which aren’t descriptive but normative. Hence, they simply aren’t in the business of semantic evaluation. We have inherited it and it determines our form of representation as well as our methods of enquiry. It therefore determines what is true or false – a correct or an incorrect description which is, in its turn, supported or disconfirmed by evidence. For this very reason, it can’t itself be true (or false). For, in order to be assessed as true or false, we should dispose of another form of representation and of assessment of empirical evidence which could determine whether our present ones are true or false. Yet this is simply not attainable for Wittgenstein, who, throughout his entire philosophical production, has always been opposed to the idea of a metalanguage and, we may now add, of meta epistemic methods. The removal from semantic evaluation of hinges, however, isn’t in contradiction with those passages in OC where Wittgenstein’s says that the truth of some of our empirical judgements belongs to our system of reference, hence to our world-picture.35 For, in context, it is clear that he is not thereby affirming that those propositions are true. Rather, he is denying that anything could move him to regard them as false. Yet, whenever this would happen, it wouldn’t show that the proposition at issue is thereby true, but only that it is removed from semantic evaluation and is therefore a rule. With this clearly in mind, we might in fact – somewhat confusingly – choose to express this idea by saying that, for us, the ‘truth’ of it is incontestable. Yet, things might change in time, or given a switch of context. So, for instance, ‘Here is my hand’ could be false, in context, just as ‘Nobody has ever been on the Moon’ is no longer true. Now, does this mean, therefore, that hinges are, after all, subject to semantic evaluation? The answer must be in the negative because, as a matter of fact, if ‘Here is a hand’ is (or can be) false, this simply means that, in the context of its utterance (or of its entertainment in thought), it is no hinge. Yet this is compatible with the fact that the same sentence could still express a hinge in a different context and thus be exempt from semantic evaluation. (Or else, if one preferred the Kantian interpretation, that, depending on the
184
Moore and Wittgenstein
context of use, the proposition expressed by that sentence could either play an entirely empirical role and therefore be subject to semantic evaluation, or else, play the role of a norm of description, and be exempt from it). Similarly, the fact that we no longer except ‘Nobody has ever been on the Moon’ from semantic evaluation – and in fact deem it false – simply shows that that sentence (or string of phonemes or graphemes) no longer plays a normative role for us, but an empirical one, and that we have good evidence against it. (Or else, that the same sentence (or string of phonemes and graphemes) was once used to express a hinge and is now used to express an empirical proposition). Moreover, as we have already seen (§3.1), our world-picture is neither grounded – that is, supported by reasons – nor ungrounded – that is, held against reasons.36 For it itself determines what counts as empirical evidence, which, in its turn, can’t in fact go against it. So, nothing speaks against our world-picture because nothing could do so, since it owes its status of evidence to keeping fixed what thus appear to be its ungrounded foundations. Nor is our world-picture ‘rational’ (vernünftig) or ‘irrational’ (unvernünftig) (OC 559). For, as we saw in Chapter 3 (§3.2), for Wittgenstein only evidential grounds could function as justifications in this context. Lacking those, he therefore thinks that the hinges that form our world-picture are simply alien to epistemic support of any kind.37 They therefore must be regarded as norms we do accept, in the pragmatic sense outlined in §3.2, and that are in turn constitutive of our notion of epistemic rationality, while being neither rational nor irrational, but merely the conditions of possibility of rationality. 4.3
Channels and rivers
In the way of a description of our world-picture, Wittgenstein introduces two famous images, at OC 96–7, 99: that of the channels and that of the river. The first goes as follows: ‘it might be imagined that some propositions, of the form of empirical propositions, were hardened and functioned as channels [Leitung] for such empirical propositions as were not hardened but fluid; and that this relation altered with time’. As already remarked, this is the passage which is most congenial to a Kantian interpretation of hinges as propositions partaking of both descriptivity and normativity. The fact that propositions of an empirical form are in deed removed from verification and control makes them help determine what counts as verification and control and thus makes them play a normative role. This role, however, may be changed again both synchronically and diachronically,38 as an effect of use and of new
Hinges, Certainty, World-Picture and Mythology 185
empirical facts. The different roles of ‘Here is my hand’ and ‘Nobody has ever been on the Moon’ would be cases in point.39 In a similar vein, Wittgenstein metaphorically describes the propositions which belong to our world-picture as part of the river-bed and introduces the idea of two different kinds of movement: the one of the waters, which may occasionally remove part of the riverbed and make it become part of the flux, and the one of the riverbed itself. OC 97: ‘(…) the river-bed of thoughts may shift. But I distinguish between the movement of the waters on the river-bed and the shift of the bed itself; though there is no sharp division of the one from the other’. We have already illustrated the first, while the second is harder to exemplify. However, we can get a glimpse of what Wittgenstein might have had in mind by thinking of formal systems and their different axioms. So the idea would be that we might decide to change one or more of our hinges not because of the effect of use or of some contingent state of affairs, like Armstrong’s landing on the Moon, but for different reasons, which may attain to considerations which have nothing to do with coping with new evidence. Again, this idea has become familiar through the work of Kuhn and Feyerabend, who have argued that paradigmatic changes in the history of science had little or even nothing to do with coping with the available evidence and were rather motivated by considerations of an altogether different kind, such as aesthetic or ideological ones. So, for instance, Kepler was moved to embrace the heliocentric system because of his religious ideas and, for a long time, all the available evidence was compatible with both the heliocentric and the geocentric systems of the universe. Even so, however, we can appreciate how such a change of hinges could produce a ‘movement’ of the waters. For, one could have no longer endorsed the statement ‘The change of seasons is due to the movement of the Sun around the Earth’. Yet, given that our world-picture isn’t a formal system, and it is discussable to what an extent it comprises scientific elements,40 it may in fact be very rare to witness such a kind of ‘movement’. Be that as it may, Wittgenstein is also suggesting that the movement of waters may produce a movement in the riverbed and vice versa. This is in fact tantamount to saying that empirical facts could turn hinges into non-hinges and that new or different hinges would produce different empirical descriptions. Yet, he hastens to say that there is no sharp division between these two kinds of movement. Again, the Kantian interpretation is the most suitable in this context. For if hinges are in fact propositions of an empirical form, actually removed from traffic, and thus turned into norms, it is easy to see why there would be no sharp boundary between the
186
Moore and Wittgenstein
movement of waters and that of the riverbed itself. For the riverbed itself is constituted by propositions of an empirical form. Yet, Wittgenstein warns us, ‘if someone were to say “So logic too is an empirical science” he would be wrong’ (OC 98). Why? Because the lack of a sharp boundary doesn’t mean that there is no boundary or, better, that it can’t be drawn, at least in context, or relative to a temporal slice. As Wittgenstein puts it, ‘Yet this is right: the same proposition may get treated at one time as something to test by experience, at another as a rule of testing’ (ibid.). As already pointed out, this would be the case also on the Kantian interpretation, of for example, ‘Here is my hand’ and ‘Nobody has ever been on the Moon’, which are propositions of an empirical form that, depending on use and empirical, contingent facts can either be treated as genuinely empirical, or as rules of testing. Finally, the degree of normativity possessed by hinges, which is to say how deep they sit in the riverbed, which makes them removed from the flux of waters, isn’t always the same, according to Wittgenstein. For, he writes: ‘And the bank of the river consists partly of hard rock, subject to no alteration or only to an imperceptible one, partly of sand, which now in one place now in another gets washed away, or deposited’ (OC 99).41 Once more, the Kantian view of hinges would make this passage better comprehensible. For normativity would be a function of how a proposition of an empirical form is in fact used, or indeed of how much it is removed from ordinary use. This is something which admits of degrees, like ‘Here is one hand’, which is not always removed from the traffic, or ‘No human being lives on a planet of the solar system other than the Earth’, which – arguably – is a hinge for us but, at the same time, we can start seeing how it may in fact become an empirical proposition. By contrast, ‘There are physical objects’ or ‘The Earth has existed for a very long time’ are always removed from traffic. 4.4
World-picture and mental disturbance
According to Wittgenstein, a mistake can be made only within a language game, for instance because one hasn’t collected enough evidence in support of one’s claim, or because one has ignored or misunderstood it. So it is possible to make a mistake only if one already ‘judge(s) in conformity with mankind’ (OC 156, 74). If so, however, there are deviant forms of behaviour which would make it doubtful that one be capable of judging in conformity with the rest of us. For instance, if one got one’s own name repeatedly wrong, or if one said that 2 + 2 equals 5, one would think that that person doesn’t yet know his own name, and that he doesn’t know how to calculate. All this would be
Hinges, Certainty, World-Picture and Mythology 187
understandable in the case of a child, where we would say that he hasn’t yet learnt his own name or how to calculate, as well as in the case of a subject who suffered from various forms of amnesia, which would explain the causes of his poor performance. But, if such traceable explanations of why he may fail at those tasks are absent, when we met a subject who stubbornly declared, against all evidence, that his name is Napoleon, or that 2 + 2 = 5, we would take his pronouncements as symptoms of mental disturbance.42 Again, we might find causes43 for such a deviant behaviour, but we could not take his pronouncements as simple mistakes. For, by making them, he would simply show that he is no party to our ordinary language games with proper names and simple arithmetical calculations. However, it is only of someone who can take part to them that we can find reasons for their deviant responses, which could make them be ranked as errors.44 So, for instance, it is only of someone who can already calculate and who showed that by holding, for instance, that 2 + 2 = 4, that we can say that they have made a mistake in calculating that 125 + 329 = 444, because they haven’t added 1 to 2 + 2. Characteristically, someone who were already calculating in conformity with the rest of us would then accept correction and therefore rectify their judgement. Not so in the case of mental disturbance: even if presented with the correct answer to the calculation or with their real names, subjects would refuse to revise their judgement. Yet, we are familiar with cases of different people believing things that are totally at odds with our world-picture. After all, there are tribes who think that a king might make rain, or that one could simply fly to the Moon (OC 92, 106, 132, 264). Are they mad, mistaken or neither? Trying to answer this question will lead us to considering the difficult issue of whether Wittgenstein was some kind of epistemic relativist. Yet, to sum up what we have been seeing so far: • Our world-picture is inherited (§4.1); • it is neither true nor false, neither grounded nor ungrounded, neither rational nor irrational (§4.2); • it is constituted by propositions which can possess different degrees of normativity as an effect of their actual role with respect to our language and epistemic practices, up to the point of losing it altogether over time (§4.3). • Deviations from our world-picture that aren’t otherwise explainable should be taken as symptoms of madness rather than as mistakes. While the latter have grounds, the former has only causes (§4.4).
188
Moore and Wittgenstein
5 Different world-pictures and epistemic relativism? Relativists and anti-relativists alike are nowadays mostly united in considering Wittgenstein an epistemic relativist.45 Accordingly,46 there could be, either in principle or as a matter of fact, different epistemic systems, none of which would be intrinsically correct; each of them would be, from a metaphysical point of view, as good as any other one, and would certify as (true and) justified different propositions. As a consequence, knowledge – that is, justified true belief – if and when attained, would always be situated: what counts as knowledge within one system of justification may not be so within another. Moreover, should alternative epistemic systems compete with each other, the choice couldn’t be based on rational considerations, for it is only within each system that reasons and justifications are produced. Hence, the passage from one epistemic system to another would always be a form of conversion or persuasion, reached through a-rational means. Relativist readings of Wittgenstein’s thought base their interpretation mostly on his claim, in On Certainty, that at the foundations of our language games and, in particular, of our epistemic ones – those in which we provide reasons for and against certain propositions or theories, and are interested in assessing their truth – lie propositions which are neither true nor false; grounded or ungrounded; rational or irrational (cf. §4.2). If, however, our language games are neither supported by grounds nor can be said to be true, they can’t be rationally held and it would be possible, at least de jure, to have alternative ones, which would be as legitimate as ours. Hence, it would be possible, at least in principle, to have different world-pictures (Weltbilder) (OC 93–7, 162, 167, 233, 262). Another set of passages that have usually been taken to support a relativist interpretation of Wittgenstein’s later ideas are those where he claims that should we find someone who doesn’t comply with our system of justification, we could only persuade or convert them to adopt ours, by appealing not to grounds and reasons – as there are none that could support one system over the other – but to altogether different considerations – in effect to aesthetic ones – such as its simplicity and symmetry, viz. the fact that certain kinds of explanations are less complex than others and can be extended from one domain to other, different ones. Let us quote in full. However, we can ask: May someone have telling grounds for believing that the earth has only existed for a short time, say since his own birth? – Suppose he had always been told that, – would he have any good reason to doubt it? Men have believed that they could make rain; why should not a king be brought up in the belief that
Hinges, Certainty, World-Picture and Mythology 189
the world began with him? And if Moore and this king were to meet and discuss, could Moore really prove his belief to be the right one? I do not say that Moore could not convert the king to his view, but it would be a conversion of a special kind; the king would be brought to look at the world in a different way. Remember that one is sometimes [my emphasis] convinced of the correctness of a view by its simplicity or symmetry, i.e., these are what induce one to go over to this point of view. One then simply says something like: ‘That’s how it must be’. (OC 92) I can imagine a man who had grown up in quite special circumstances and been taught that the earth came into being 50 years ago, and therefore believed this. We might instruct him: the earth has long … etc. – We should be trying to give him our picture of the world. This would happen through a kind [my emphasis] of persuasion. (OC 262) Is it wrong for me to be guided in my actions by the propositions of physics? Am I to say I have no good ground for doing so? Isn’t precisely this what we call a ‘good ground’? (OC 608) Supposing we met people who did not regard that as a telling reason. Now, how do we imagine this? Instead of the physicist, they consult an oracle. (And for that we consider them primitive.) Is it wrong for them to consult an oracle and be guided by it? – If we call this ‘wrong’ aren’t we using our language-game as a base from which to combat theirs? (OC 609) And are we right or wrong to combat it? Of course there are all sorts of slogans which will be used to support our proceedings. (OC 610) Where two principles really do meet which cannot be reconciled with one another, then each man declares the other a fool and heretic. (OC 611) I said I would ‘combat’ the other man, – but wouldn’t I give him reasons? Certainly; but how far do they go? At the end of reasons comes persuasion. (Think what happens when missionaries convert natives). (OC 612)
190
Moore and Wittgenstein
In the following I will try to motivate the view that once we look closer at these and other passages, the first impression of dealing with an epistemic relativist should be dramatically revised. As I will suggest, there is room for maintaining that Wittgenstein was merely an antifoundationalist: he believed that our world-picture is ungrounded and that it isn’t a mere reflection of a (totally) mind-independent reality. But anti-foundationalism is a long way short from relativism, let it be ‘factual’ – the view according to which there actually are different, incompatible epistemic systems that are all equally valid, or merely ‘virtual’ – that is, the view according to which there could be, at least in principle, different, equally valid, and incompatible epistemic systems, all in fact conceivable from our own standpoint.47 For simply to say that our world-picture is ungrounded doesn’t entail either that there actually are different ones, or – more contentiously – that there could intelligibly be other ones, at least in principle. 5.1 5.1.1
Was Wittgenstein a factual epistemic relativist? Different hinges and theories
It is a striking feature of On Certainty we have already hit upon several times, that among the various hinges Wittgenstein listed ‘Nobody has ever been on the Moon’, as this proposition is no longer a hinge and appears to our minds as on a par with ‘The Sun moves around the Earth’, which is simply a false scientific statement. But are different hinges and scientific theories enough to ground the idea that we are dealing with different systems of justification which can’t be argued for or against on rational grounds? It seems plausible to think that both Ptolemy and Wittgenstein would have been rationally persuaded to change their views had they had all the evidence available to us: pictures taken from satellites, in the former case, and the images of Armstrong and associates landing on the Moon in 1969 in the latter case. So, it is perfectly conceivable that within just one system of justification – call it ‘Science’ for short – people may have different beliefs, depending on the quantity and quality of the evidence available to them. Moreover, it is a striking feature of Science that as evidence piles up people engaging in it would tend to converge on certain beliefs. This suggests that there is just one system of justification – Science – which evolves and develops over time, where certain propositions and theories may be outdated by others, because new information comes in and actually proves certain beliefs or theories false, or calls for a new kind of explanation. But even in this latter case, where a change of ‘paradigm’ occurs, it should be kept in mind that it is the presence of new evidence, which is recalcitrant to being incorporated
Hinges, Certainty, World-Picture and Mythology 191
within a given theory, that calls for a new explanation and hence for a new scientific model. So, at these critical moments in the history of science, one may be more slowly persuaded to adopt the new theory than another, being perhaps more cautious or just biased. All the same, sooner or later scientists will converge on the new theory because of evidence in its favour and of its greater explanatory and predictive power. Of course one day the new theory may fall into disrepute, but, if it does, it will do so because further new evidence will have come in. Moreover, subjects are rationally persuaded to change their views only when specific empirical questions, which up to some point had no answer, are replied. So, scientific change is, after all, driven by rational considerations (which is not to say that there may not be other, non-rational ones: simply, the latter aren’t as decisive as the former). Interestingly, there are various passages at which Wittgenstein’s considerations support precisely this view. For instance: ‘But is there then no objective truth? Isn’t it true, or false, that someone has been on the moon?’ If we are thinking within our system, then it is certain that no one has ever been on the moon. Not merely is nothing of the sort ever seriously reported to us by reasonable people, but our whole system of physics forbids us to believe it. For this demands answers to the questions ‘How did he overcome the force of gravity?’ ‘How could he live without an atmosphere?’ and a thousand others which could not be answered. But suppose that instead of all these answers we met the reply: ‘We don’t know how one gets to the moon, but those who get there know at once that they are there; and even you can’t explain everything.’ We should feel ourselves intellectually very distant from someone who said this. (OC 108) What we believe depends on what we learn. We all believe that it isn’t possible to get to the moon; but there might be people who believe that that is possible and that it sometimes happens. We say: these people do not know a lot that we know. And, let them be never so sure of their belief – they are wrong and we know it. If we compare our system of knowledge with theirs then theirs is evidently the poorer one by far. (OC 286) Thus Wittgenstein held it impossible that someone had been on the Moon because there was no evidence in favour of that hypothesis and
192
Moore and Wittgenstein
because several empirical questions hadn’t yet had an answer. Conversely, he would presumably have been open to change his mind if reliable evidence had come in and if those empirical questions had had an answer. Moreover, that – and not just considerations pertaining to simplicity and symmetry – would have made him revise his position.48 Hence, it can be maintained that according to how Wittgenstein actually develops his points, rather than according to his prima facie more obvious pronouncements, it is only when entirely rational considerations militate against ‘Nobody has ever been on the Moon’ that that proposition would cease to be a ‘hinge’ and would become a genuinely empirical one, subject to truth and falsity. 5.1.2
Different methods I
Perhaps, in order to get a grip on the idea that there may actually be different systems of justification, we should look at cases where some people use Science and others use a seemingly different system of justification, such as Divination, where subjects form beliefs by using methods that don’t rely on evidence, on the formation of hypotheses and on their testing. For instance, in order to know whether it is going to rain, some people cast oracles by inspecting animals’ interiors. With respect to these cases, Wittgenstein seems to have been a relativist, as he claimed that no rational conviction could be possible, only persuasion. (OC 262, 612). Yet, I think this would be a poor case in favour of relativism. For looking at the way animals’ interiors deteriorate could be an indicator of the humidity present in the air, hence a piece of empirical evidence suitable to the purpose at hand – no matter how poor and unreliable it may turn out to be. If that is right then Divination, once deprived of all its symbolic and religious elements, would just be a piece of, as it were, ‘primitive’ Science. Hence, it can’t be used to support the idea that we are actually confronted with a different system of justification, but only to maintain the view that our system of justification may evolve through time. This is, however, what Wittgenstein’s following remark in the Notes on Frazer’s ‘The Golden Bough’ (NF hereafter) also points out: The nonsense here is that Frazer represents these people as if they had a completely false (even insane) idea of the course of nature, whereas they only possess a peculiar interpretation of the phenomena. That is, if they were to write it down, their knowledge of nature would not differ fundamentally from ours. Only their magic is different. (NF 141)
Hinges, Certainty, World-Picture and Mythology 193
To say that their knowledge of nature isn’t fundamentally different from ours means to say that they are responsive to empirical evidence in forming their beliefs about the weather. Their ways of acquiring and treating such evidence may look ‘primitive’ to us. But this would only support the claim that their empirical methods are less developed than ours, not that their methods of formation of belief in this area are totally different from ours. One might then object that leaving out the symbolic and religious elements is illegitimate. Wittgenstein too in the Notes stressed their importance and criticized Frazer’s methodology because he had tried to explain certain religious practices as forms of primitive and erroneous science. Yet, I think that this observation, though correct, can’t lend support to relativist interpretations of Wittgenstein’s thought. For it is one thing to say that in order to understand different religious practices we can’t ignore their symbolic elements and must, therefore, connect them, through family resemblance, to what we do, so that, by so doing, we can familiarize ourselves with them and abandon a grossly judgemental attitude, which would actually compromise anthropological research (we will come back to this in the following). Still, it is a totally different thing to say that, while doing epistemology, thus engaging in a (at least partially) normative enterprise, we can’t judge their epistemic practices, taken as such, and deem them erroneous, or, at any rate, ‘primitive’, if compared to ours, should they so be.49 A case in point, which can help illuminate the contrast, is that of the Azande, described by Edward E. Evans-Pritchard (1937), who admit that termites can consume the legs of granaries so that they sometimes suddenly fall down and thus injure people sitting underneath, while also invoking witchcraft to explain the calamity. Evans-Pritchard himself50 already noted that the Azande distinguish between how a certain phenomenon happens and why it does. Peter Winch comes close to seeing that this may be due to the fact that two different notions of causality are at play, when he notices that our own ‘concept of causal influence is by no means monolithic’.51 But he then goes on to write as if reasoning by ‘efficient causes’ and by ‘final’ ones – as we might put it – were mutually exclusive. Arguably, however, these two ways of reasoning are compatible, as they respond to different concerns human beings can have. This opens up the possibility of saying that our scientific attitude is more advanced insofar as our account of efficient causes is better than theirs; while also maintaining that their magical practices aren’t a sufficiently strong ground to motivate the idea that they embrace a different epistemic system. It remains, however, that we have either an unlike ‘magic’ or that we have altogether given up on it, and on its attendant enquiry into final causes.
194
Moore and Wittgenstein
In effect, what we witness in the two cases just presented is simply the fact that Science is characterised by the indifference to final causes. For to explain, in Science, is to find out merely the efficient causes of phenomena.52 Yet, human beings seem naturally prone to look for the final causes of what happens to them, in the attempt to find the meaning and significance of what they experience throughout life. This, in effect, is what characterizes the religious attitude and to try and reduce it to a primitive scientific one is thus a categorial mistake, viz. the mistake of conflating efficient and final causes and their attendant motivations. Moreover, that these attitudes are, at bottom, independent can explain the Azande’s faith in oracles. As Evans-Pritchard remarks,53 among the Azande there is widespread scepticism about the powers claimed by the witch doctors. Nevertheless, they are routinely consulted to explain why something happened, or to see whether a given course of action might be fraught with mystical dangers and should thus be avoided. In contrast, many of us today have totally abandoned the religious attitude. As I read Wittgenstein’s Notes, however, all these elements are compatible with his views. Some telling passages, which are worth citing in full, are these: [N]o phenomenon is in itself particularly mysterious, but any of them can become so to us, and the characteristic feature of the awakening mind of man is precisely the fact that the phenomenon comes to have meaning for him. One could almost say that man is a ceremonial animal. That is, no doubt, partly wrong and partly nonsensical, but there is also something right about it. (NF 129, italics mine) But then it is nonsense for one to go on to say that the characteristic feature of these actions is the fact that they arise from faulty views about the physics of things. (Frazer does this when he says that magic is essentially false physics or, as the case may be, false medicine, technology, etc.). (Ibid.) Rather, the characteristic feature of ritualistic action is not at all a view, an opinion, whether true or false, although an opinion – a belief – can itself be ritualistic and part of a rite. (Ibid., but see also NF 121, 123) Here Wittgenstein is distinguishing between opinions and theories, on the one hand, and the symbolic and religious elements of a ritual, on the
Hinges, Certainty, World-Picture and Mythology 195
other. He claims that there is continuity between the opinions and theories of the ‘primitives’ and ours, for, presumably, they evolve as possible answers to the same kind of questions. Yet, the symbolic and ritual elements differ and might actually disappear for us. The disenchantment of the world brought about by Science’s indifference to final causes was, arguably, what Wittgenstein opposed in the spirit of his (our?) time. For he felt that Science couldn’t provide an answer to the fundamental questions of human beings’ lives – those questions which had troubled him throughout his life and that were openly addressed in the final sections of the Tractatus (6.4–6.7). That, however, would have been acceptable to him, as long as those questions remained part of our ‘form of life’, perhaps receiving at least partial answer from art and meditation, especially on the example of the lives of saints.54 But Wittgenstein sensed that the affirmation of Science would have had the effect, in the long run, of making those questions disappear from our lives. This perceived threat to what, in his opinion, made us really human, motivated much of his attack on Science.55 Still, to expose and criticize this possible outcome of embracing Science – viz. to end up holding a form of ‘scientism’ – is nowhere near to maintaining the postmodernist mantra that Science is just one way of knowing the world on a par with, say, Zuni’s and Christians’ creationist epistemology. 5.1.3
Different methods II
Another example often discussed in this connection concerns the allegedly different logic used by the Azande. This too would be a case of an alternative epistemic method, since logic is what is used to connect beliefs and to draw conclusions from them. According to Evans-Pritchard,56 the Azande believe that witchcraft transmits patrilineally. In particular, it is an ‘inherited physical trait, consisting of a substance in the belly called “witchcraft-substance”’.57 Sometimes, however, they would deny that the son of a witch doctor is a healer in his turn. According to Winch (1964), this shows that the Azande don’t accept modus ponens and have thus a different logic. Such a conclusion, however, isn’t mandatory. Evans-Pritchard himself argues that the ‘Azande do not perceive the contradiction as we perceive it because they have no theoretical interest in the subject and those situations in which they express their belief in witchcraft do not force the problem upon them’.58 Hence, the Azande’s attitude in this particular case is in fact compatible with their acceptance of modus ponens in general, that is, in all other cases. Hence, it cannot be appealed to claim that they have a different logic from ours.
196
Moore and Wittgenstein
A similar remark would be apposite also in the case of Catholics’ acceptance of certain dogmas, such as the dogma of the Holy Trinity, which goes against the hinge that something can’t be one and three things at once. Now, I think we would resist the idea that Catholics have a different logic than ours and that we would rather say that certain issues remain insulated from the employment of logic. In relation to such dogmas and in a similar spirit, Wittgenstein writes, [Catholic] dogma is expressed in the form of an assertion, and is unshakable, but at the same time any practical opinion can be made to harmonize with it; admittedly more easily in some cases than in others. It is not a wall setting limits to what can be believed, but more like a brake which, however, practically serves the same purpose; it’s almost as though someone were to attach a weight to your foot to restrict your freedom of movement. This is how dogma becomes irrefutable and beyond the reach of attack. (CV 28) Here Wittgenstein is actually saying that when it gets to Catholic dogmas subjects who believe in them aren’t free to apply logic and their usual epistemic methods, despite their behaviour in everyday life. Still, the two spheres – the religious and the mundane one – are sufficiently insulated from one another not to give rise to problems. Be that as it may, Boghossian lists Wittgenstein among those who would have been sympathetic to the relativistic claim that the Azande have a different logic from ours.59 But, in fact, more recent studies have shown that a mistake was made in translating them.60 For they actually believe that only those sons of a witch doctor who are ‘hot’, and thus not only have the potential for being witch doctors but actually realize it, are witch doctors. That is why the Azande may perfectly rationally deny that a witch doctor’s son, who happens to be ‘cold’, is a healer, despite the fact that he has the witchcraft substance in his belly. Remarkably, however, the conclusion that we should revise our translation is exactly the one Wittgenstein reaches in the case he actually discusses and that Boghossian quotes,61 without noticing that it is in keeping with his own opposition to epistemic relativism. Namely, the case of a community of people who ‘sell timber by cubic measure’ (RFM 148, cf. 143). In such a case, we would have a different epistemic method, where quantities are measured in an odd way and prices are determined differently. How could we show them that – as I should say – you don’t really buy more wood if you buy a pile covering a bigger area? – I should,
Hinges, Certainty, World-Picture and Mythology 197
for instance, take a pile which was small by their ideas and, by laying logs around, change it into a ‘big’ one. This might convince them – but perhaps they would say: ‘Yes, now it’s a lot of wood and costs more’ – and that would be the end of the matter. – We should presumably say in this case: they simply do not mean the same by ‘a lot of wood’ and ‘a little wood’ as we do; and they have a quite different system of payment from us. (RFM 150) So Wittgenstein seems to be saying that we can imagine such a community, whose way of measuring and paying for wood is different from ours. But clearly he thinks that if we had to deal with them, we would try to convince them to measure wood by weight and to pay for it accordingly, by using an entirely rational procedure, viz. by showing them that the quantity of wood remains the same although its area and volume are changed. If that succeeded, the case just presented wouldn’t constitute an instance of an alternative epistemic method, but, rather, a case in which people hold a false belief, which leads them to using an unreliable procedure to measure (and pay for) wood. In contrast, if rational persuasion failed, Wittgenstein suggests that we should suppose that they don’t mean what we do by ‘a lot of wood’ and ‘a little wood’. Hence, we should conclude that we have made a mistake in translating (even homophonically) those people’s words with what we mean by ‘a lot of wood’. We should then revise our translation because we can’t go against the hinge that one thing – a certain amount of wood – is identical to itself, no matter how we arrange it. Hence, Wittgenstein turns out to be in complete agreement with Boghossian in thinking that just by conceiving of a possibly different community, which is recalcitrant to rational persuasion, we wouldn’t have ipso facto discovered a different epistemic system alternative to ours. 5.2
Was Wittgenstein a virtual relativist?
In order to assess whether Wittgenstein was a virtual relativist, we need to consider what he maintained about the possibility of having, at least in principle, different yet perfectly legitimate and intelligible worldpictures.62 Undoubtedly, Wittgenstein repeatedly imagined different communities, in which hinges we take for granted seem not to be held fast. The example of the community that measures wood by the area is clearly a case in point. The conceivability of these communities would then appear to support Wittgenstein’s virtual relativism – the idea that it is conceivable that there could be people with altogether different world-pictures and, consequently, conceptual schemes. Hence, the
198
Moore and Wittgenstein
issue of whether Wittgenstein was a virtual epistemic relativist connects with the problem of whether he was a conceptual relativist. Conceptual relativism, in turn, can be defined as the view according to which it is possible that there be different conceptual schemes. They would return incompatible descriptions of the world, none of which is intrinsically correct, and yet which are intelligibly describable from a theorist’s point of view.63 Had one to choose among them, the choice couldn’t be based on the fact that one scheme would be more true to facts than the other ones, but only on pragmatic and even aesthetic reasons, if not entirely on sociological ones.64 However, Wittgenstein seems also to suggest that if we failed to persuade them of their mistake, we should revise our translation: the seeming refusal to correct themselves would cast doubt on the fact that they actually mean what we do by the words they use. Conversely, the meaning of a word seems to be a function of at least some central inferences we accept. Hence, did it turn out that, by our lights, the latter aren’t accepted, we should conclude that, in fact, their words have a different meaning than the one we first attributed to them. Thus, according to Wittgenstein, radically different epistemic practices wouldn’t really be conceivable, just as Barry Stroud maintains.65 Whenever we might be led to thinking otherwise, we should conclude that we have misunderstood them, because they don’t actually mean what we do by, for example, ‘a lot of wood’. Yet, it remains that Wittgenstein repeatedly asked us to imagine these different communities. Was he utterly incoherent? How could he escape this dilemma? 5.2.1
Family resemblance?
Before addressing this impasse, let me consider a suggestion made by Diego Marconi in order to defuse it.66 He thinks that it would be complicated to make sense of a radically different community for whom a given quantity of wood changes depending on how it is spatially arranged, because we would have to make a lot of other adjustments to let them turn out to be coherent; yet, we could extend our concepts through family resemblance. Hence, ‘little wood’ or ‘a lot of wood’ would mean something only slightly different from what we mean by those words, which we could understand by analogy with what we mean. Though ingenious, I don’t find this suggestion convincing. The notion of family resemblance explains how we apply our concepts to cases that don’t present a definite set of common features, so long as we can find some analogy between them and something else which we regard as an instance of the concept. Now, the problem is that since
Hinges, Certainty, World-Picture and Mythology 199
the examples are under-described, we simply don’t know if we could find those analogies which would allow us to apply the idea of family resemblance. But, in general, there seem to be only two options: if, in the other areas, their use of the relevant words is like ours, we would think that when it comes to measuring wood, they mean exactly what we do and simply have wrong beliefs. Hence, this community wouldn’t embrace a conceptual and an epistemic system that are real alternatives to ours. If, in contrast, in those other areas there were substantial differences, we couldn’t find the ‘intermediate steps’ we need in order to understand their use of those words as just partially different from, yet intertwined with ours. One may protest that I have been unfair to Marconi’s suggestion regarding the first horn: if a difference were sufficiently local, it could still be taken to be a partial conceptual difference, which we would understand thanks to its family resemblance with our concepts. But we need to be careful here: if they mean something only slightly different from what we mean by ‘there is a lot of wood’, then we can’t really translate them with those words. To see why, consider the following case: we may try to translate the Portuguese word ‘saudade’ with ‘nostalgia’. But this isn’t really what ‘saudade’ means, as many Portuguese would confirm. Nor is it the case that through an enormous amount of empathy non-Portuguese could understand what saudade is about, while acknowledging the lack of words that could translate it into their own language. What we should admit rather is that, if saudade is really anything at all over and above nostalgia, we don’t truly know what it is (or what ‘saudade’ means). Hence, should we have to translate that word, we probably had better borrow the Portuguese expression itself adding a partly explanatory footnote, which could give an idea of the semantic area in which that word should be located. Thus, family resemblance can’t help us preserve conceptual differences while making them really intelligible. Insofar as these differences persist, they remain unbridgeable, at least at bottom. On reflection, however, this should come as no surprise. For the appeal to family resemblance can explain how our concepts work and how we can extend them to new cases. Here, however, we weren’t trying to extend our concept of nostalgia to comprise also cases of saudade; rather, we were trying to use our own concept – just as it is – in an attempt to get to grips with that of saudade, while maintaining their difference. That is why the appeal to the idea of family resemblance couldn’t help us in the end: insofar as it works, it abolishes the differences by partially transforming the original concept. Hence, if differences have to be preserved, as relativism requires, the idea of family resemblance can’t be applied.67
200
Moore and Wittgenstein
Another suggestion may be as follows: suppose that this community just uses wood to cover areas, as it might happen if they only employed it to make floors. So, if they were presented with one single log of weight x, given their purposes, it would be less wood than if they were presented with two juxtaposed logs amounting to the same weight. This, however, wouldn’t be enough to conclude that they have a different epistemic practice of measuring (and paying for) wood. At least not insofar as they were prepared to agree that the weight and hence the quantity of wood remains the same and thus to say: ‘Of course we can see that the weight is the same, still, in the second case there is more wood’. Nor would this be enough to conclude that their concept of more (wood) and its cognates is different from ours (as Marconi seems to suggest and relativism requires). For what they really mean when they say ‘There is more wood here’ is simply ‘There is a bigger area of wood here’, and obviously, given their purposes, they are prepared to pay more for it, as it allows them to cover a bigger extension. Hence, everything else being equal, we should accordingly revise our translation. Indeed if the only thing they do with wood is to cover areas, then it is obvious why they don’t disambiguate between the weight of wood and the area it covers and hence why they keep using the same words they employ in connection with quantities of ice-cream and flour, which they measure by their weight, just as we do. Thus, we haven’t found any clear sense in which thinking of this deviant community should force us to admit that they have either a different epistemic practice or a different concept of a lot (and its cognates). All they have is, at most, a language where, for the reasons just explained, they don’t need to distinguish between weight and areas covered by wood and can thus use the words ‘a lot of’ and their cognates across different language-games in these apparently irreconcilable ways. But let us go back to Marconi’s suggestion. Another reason why the notion of family resemblance wouldn’t help us make sense of the deviant community is that if we tried to extend our concept of a lot to comprise their use of it in connection with wood, on the assumption that they really meant ‘a lot’ as we do, we would introduce a contradiction between the use of that concept with respect to, for example, ice-cream and flour, on the one hand, and in connection with wood, on the other. The notion of family resemblance, in contrast, explains the functioning of our concepts and the possibility of applying them to new cases, only when such inclusions do not introduce overt contradictions.68 Thus, the suggestion that we may really understand Wittgenstein’s imagined communities, while maintaining the differences between us
Hinges, Certainty, World-Picture and Mythology 201
and them, by mobilizing the notion of family resemblance, is spurious: if differences are local, we would regard their use either as a mistaken application of the same concepts we use, or as a use of a different concept, which would call for a more careful translation of their words; or else, as the use of a different concept, which, however, we can’t quite grasp, in such a way that we would end up not finding them intelligible, or to find them only partially intelligible but on the background of a largely similar conceptual scheme. If, in contrast, differences between our use and theirs aren’t local, we couldn’t really make sense of such a community and, insofar as we may want or need to interpret them, we would have to revise our translations altogether, in such a way as to re-establish widespread agreement. 5.2.2 Alternative world-pictures as mere metaphysical possibilities: on the ungroundedness and yet universality of our own world-picture Our dilemma was: how could Wittgenstein both ask us to imagine allegedly radically different communities and maintain that, if we tried to think them through, we couldn’t really make sense of them and should thus revise our translations? Or, to put it differently, if these radically different communities aren’t really conceivable from our own point of view – which, after all, is the only one we have got – what is left to the idea that, at least in principle, there might be possibly different communities, with their own epistemic systems and conceptual schemes? All it seems left to this idea is the purely metaphysical possibility69 that if certain ‘facts of nature’ (PI II, xii, 230) had been totally different, or if ‘something really unheard of’ (OC 513) were to happen, there could be creatures who don’t believe in the existence of material objects, who don’t reason as we do, for whom 2 + 2 isn’t equal to 4, who don’t use their senses to gather evidence, who don’t think that the Earth has existed for a long time before they were born, etc. Hence, all it remains to it is the metaphysical possibility that all this might happen, or might have been the case. Still it is a possibility that we can’t really conceive of in detail, given the kind of creatures we are and the fact that our concepts are what they are also because of some very general facts about us and about nature. That is to say, our own way of thinking, when applied to the task of imagining substantially different communities, works in such a way that it screens them off, by suggesting the various verdicts, adjustments and revisions we have been seeing. Notice, however, that this wouldn’t be a form of virtual relativism. For virtual relativism requires that radically different world-pictures were at least conceivable, that is to say, imaginable in detail and understandable
202
Moore and Wittgenstein
from the point of view of those who are conceiving them. If the preceding is right, this isn’t the case, according to how Wittgenstein actually elaborates on his own attempts at imagining different communities, his explicit pronouncements notwithstanding.70 So, the question arises of what was the point of reminding us of the fact that things might have been or might be very different from what they are. The value of what in the end appears to be an entirely notional exercise is, I believe, to make us aware of the ungroundedness of our own world-picture – that is to say, of its contingency.71 Hence, the point of this exercise in imagination is not to prove that altogether different world-pictures are really conceivable, but to make us better aware of the ungroundedness of the only world-picture we have and can mobilize to understand apparent and, when really conceivable, merely local departures from it. At a reflexive level, moreover, this exercise teaches us something about our own way of thinking, namely that it doesn’t screen off the metaphysical possibility that there might be radically alternative world-pictures, but, rather, their conceivability in detail from our own point of view. This, in turn, entails that even if they existed, we wouldn’t be able to recognize them as such, because if we got so far as to be in a position to interpret them, we would inevitably do so by the lights of our own world-picture and conceptual scheme. Thus, our world-picture and conceptual scheme may well be metaphysically contingent, still they are unavoidable for us and, therefore, universal, if only from our own point of view. If so, a negative answer must be returned to the question ‘Was Wittgenstein a virtual epistemic relativist?’. Since, however, we haven’t found any decisive evidence that Wittgenstein was a factual epistemic relativist either, we must return a negative answer also to the general question of whether he was an epistemic relativist at all. All he was, it seems, is simply an anti-foundationalist. That is to say, he believed that we have just one world-picture; that, in principle, it could have been dissimilar, had certain facts of nature been very different from what they are, or that it could dramatically change if something really unheard of were to happen. But despite the fact this world-picture is contingent and can partially evolve through time (OC 96–99), it remains one. And where there is just one picture of the world – no matter how ungrounded it is – there simply is no room for relativism. Hence, to sum up and conclude, • Despite appearances to the contrary, Wittgenstein held that it is more evidence than persuasion that can induce us to abandon some of our hinges in favour of different ones (§5.1.1), not by showing them false, as such, but by forcing us to turn them into empirical propositions.
Hinges, Certainty, World-Picture and Mythology 203
• Different tribes and people, on his view, present more different religious and symbolic elements than a fundamentally different account of nature (§5.1.2). • Yet, explanations of natural phenomena may evolve over time, but this simply marks the development of one shared world-picture (§5.1.2). • Also cases of apparently deviant kinds of logic are treated by Wittgenstein in ways which are not congenial to relativism; viz. by proposing confinement strategies or translation’s revisions (§5.1.3). • Similarly, he invoked a revision of translation whenever our current ones would make us and other people diverge too starkly over hinges (§5.2.1). • Hence, more than an epistemic and conceptual relativist (of either a factual or virtual brand), Wittgenstein was in fact an anti-foundationalist and an anti-realist, who wanted us to realize the metaphysical ungroundedness of our conceptual and epistemic systems, as well as their ineluctability for us (§5.2.2). • Hence, they would always screen off the intelligibility of conceiving of radically different ones, thus making relativism simply incomprehensible from our own standpoint. All it remains is thus the idea that it is metaphysically possible that there be creatures with radically different conceptual and epistemic systems, though we can’t really understand the ways in which they would deviate from ours (§5.2.2).
6 Propositions that might be part of a kind of mythology As we saw in §4, Wittgenstein introduces the term ‘mythology’ in connection with our world-picture. This is an undoubtedly puzzling remark, which we need to clarify. Doing this will have, in my opinion, rich and somewhat surprising implications. Let us quote the relevant passages in full again. The propositions describing this world-picture might be part of a kind of mythology. And their role is like that of rules of a game; and the game can be learned purely practically, without learning any explicit rules. (OC 95) The mythology may change back into a state of flux, the river-bed of thoughts may shift. But I distinguish between the movement of the waters on the river-bed and the shift of the bed itself; though there is no sharp division of the one from the other. (OC 97)
204
Moore and Wittgenstein
The passages are confusing because in its usual meaning mythology is eminently discursive and it contains religious elements as well as imaginative explanations of facts and events. Here, however, Wittgenstein is explicitly denying that we have got our mythology mostly discursively and it is clear in context that he is not talking of religious elements and of imaginative explanations of facts and events. Rather, he is concerned with the most fundamental aspects of our shared world-picture. So, in my view, this is enough to show that we need to dig deeper in order to understand Wittgenstein’s use of ‘mythology’ in these passages.72 6.1
Mythology chez Wittgenstein
In order to clarify Wittgenstein’s use of ‘mythology’ (and myth) it is useful to keep in mind that this term occurs in several of his works. For instance, in the Notes on Frazer’s ‘The Golden Bough’ he writes, ‘An entire mythology is stored within our language’ (NF 133); and in Zettel we find the following claim: ‘In philosophy one is in constant danger of producing a myth of symbolism or a myth of a mental process. Instead of simply saying what anyone knows and must admit’. However sketchy these remarks might be, it is clear that there are at least two fundamental senses in which ‘mythology’ occurs chez Wittgenstein. On the one hand, a mythology is deposited in our language and every speaker can’t but inherit it. On the other, a mythology is produced by doing philosophy – in fact bad philosophy, on Wittgenstein’s view of the matter. Let us now look at the first of these two notions of mythology. The context of Wittgenstein’s remark in NF is roughly as follows: Wittgenstein criticizes Frazer’s interpretation of certain rituals present among various people, especially in Great Britain.73 In particular, he opposes Frazer’s attempt to explain them in a causal way, which makes them seem dependent on a plethora of false beliefs, of a pre-scientific nature. By contrast, Wittgenstein suggests that rituals don’t (always) depend on belief or opinion. Rather, it belongs to human beings – conceived as ‘cerimonial animal(s)’ (NF 129) – to create rites around everything which is felt as important for their lives and the lives of their communities – let it be birth or death, or the passage from childhood to adulthood, or else the rain, or hunting animals and fighting enemies. Thus, when confronted with maybe alien rituals, one should not so much try to explain them in a causal fashion, by appealing to mistaken and pre-scientific beliefs, as to understand them in light of what those people deem important for their lives. Hence, the comprehension of a ritual is necessarily connected with an understanding of the life of the
Hinges, Certainty, World-Picture and Mythology 205
community that performs it.74 Therefore, the hermeneutic activity of an anthropologist, according to Wittgenstein, should aim at exhibiting a net of images, in fact of intermediate links, that, once put one next to other, will give us a perspicuous representation – that is a representation that will allow us to stop looking for other interpretations – of the ritual under scrutiny. An example will clarify the point. Cannibalistic rituals seem very far from us and almost incomprehensible. But this impression will disappear (at least to some degree) if we juxtapose them with other rituals in which, for instance, a lamb is eaten as a symbol of the body of Christ, or a piece of bread with the same symbolic elements gets consumed, or finally, a piece of bread, which is believed to have become the very body of Christ is eaten. Frazer’s mistake, therefore, is not that of according pride of place to science over magic – which is something Wittgenstein would endorse too (as we saw in §5) – but that of holding that rituals be motivated by erroneous, pseudo-scientific opinions and that the only model of interpretation of reality, even of religious practices, be that of scientific explanation, where priority is given to causal explanations. Furthermore, according to Wittgenstein, Frazer doesn’t realize that when he gives an explanation like ‘That these observances are dictated by fear of the slain seems certain’ (NF 130), he himself is using a term – ‘ghost’ – that is itself a superstitious expression. That is to say, a word itself full of semantic sedimentations which connect to rituals and myths, but that, interestingly, belongs to our ordinary vocabulary and that we seem to understand perfectly well. Thus, concludes Wittgenstein, Frazer doesn’t realize either ‘our kinship with those savages’ (NF 133), or the fact that, ‘an entire mythology is stored within our language’ (ibid.). Hence, it seems clear that, according to Wittgenstein, in our own language there are elements which originally belonged to a mythological and magic way of conceiving of reality. As a consequence, Frazer’s paternalistic attitude with respect to the communities he describes is particularly shortsighted, by Wittgenstein’s lights. For Frazer fails to realize how we are still very close to them, at least insofar as we freely use words and expressions which are intimately compromised with those mythological and magical views of the world Frazer is so keen on condemning. Now, although the reference to mythological an magical conceptions of the world isn’t very useful in order to understand OC 95, 97, claiming that an entire mythology is deposited in our language can help us better understand Wittgenstein’s remarks in On Certainty. For a mythology is rooted in a specific view of the world and gives expression to it. But
206
Moore and Wittgenstein
it can also be deposited in our language and in our deepest ways of thinking and of coping with reality, in such a way that who happens to be brought up within a community that shares it, inherits it as implicitly as it might be. So, each of us, whether consciously or unconsciously, is the bearer of a tacit mythology, which is itself the expression of a specific world-picture. Now, one may wonder whether there is a way in which such a mythology can be made explicit. According to Wittgenstein, it is in fact philosophy that allows the mythology deposited in our language and our deepest ways of thinking and coping with reality to be made explicit. 6.2 Philosophy as a kind of mythology: Beyond the saying/showing dichotomy In order to clarify the relationship between mythology and philosophy in Wittgenstein it is important to keep in mind that, in this connection, he has two different attitudes. On the one hand, as the passage from Zettel makes clear, he thinks that sometimes it is philosophy itself that produces myths. That is to say ‘false pictures’75 of the workings of our language produced by the philosophical tendency to obliterate all the different uses of language and by focusing on just one of them, and by the habit of extending to all others, considerations which, if correct at all, would work only for that particular case. So, for instance, it is philosophy that produces the myth of meaning as a sui generis kind of thing, or of the mind as again a peculiar sort of thing and of mental states as objects that inhabit it, like tools inside a box. From such a perspective, we can see why, for instance, Wittgenstein wrote that ‘Frazer doesn’t notice that we have before us the teaching of Plato and Schopenhauer’ (NF 141).76 The claim, therefore, is that bad philosophy produces myths, and becomes itself a mythology, by misunderstanding the actual grammar of our language. But, on the other hand, good philosophy eradicates the myths produced by its bad counterpart and exhibits, in particular, the world-picture which is deposited in our language and deep-seated ways of thinking and coping with reality and that remain mostly implicit and tacit. How does it do that? By manifesting those linguistic rules and epistemic norms that characterize our world-view, whose normative nature is often ignored or misconceived by bad ways of doing philosophy. Thus, when Wittgenstein talks of mythology, in OC 95, 97, with respect to those propositions which describe our world-picture he has in mind, I think, the fact that these propositions make explicit what belongs to our world-view, which usually shows itself in our ways of
Hinges, Certainty, World-Picture and Mythology 207
acting, speaking and coping with reality, and thus remains mostly implicit. Yet, when our world view is said, we realize that what is thereby said can’t really have a descriptive content which can then be assessed as either true or false. For if the propositions which belong to a kind of mythology exhibit our methods of representation and of epistemic enquiry they do not describe any facts and can’t, therefore, be subject to semantic and epistemic evaluation. Thus, they won’t be true empirical propositions known with certainty by all speakers and people sharing common sense, as Moore thought of them. Nor can they be denied or called into question, like idealists and sceptics do, because they themselves determine which sensible and meaningful doubts can be raised and how propositions about existence can be denied. If we took them this way, we would be producing a mythology of symbolism, by giving a false picture of their real nature. Yet we do put them into words and thus make them explicit. So, are we perhaps trying to say what can only be shown? My final contention – supported also by the defence of the propositionality of hinges developed in §3.2 – is that we should answer this question in the negative. For there is an alternative in between the warning (OC 501) that logic – viz. grammar and everything which is essential to the very possibility of there being a given language game – can’t be described,77 and the invitation to look at it in its showing itself in practice (ibid.). Namely, there are propositions – ‘hinges’ – whose nature is normative, which can and are put into words and exposed by doing proper philosophy, thereby clarifying their status and function. Thus, we can summarize the relevant features of Wittgenstein’s talk of ‘mythology’ as follows: • Our world-picture is like a mythology implicitly and tacitly deposited in our language and various epistemic practices (§6.1); • philosophy constantly runs the risk of producing myths – viz. false pictures – of our conceptual and epistemic systems (§6.2), • however, when properly conducted, it can actually bring to the surface the deep-seated elements of our world-picture and let us see their nature and status aright (§6.2).
Conclusion: Moore and Wittgenstein on Epistemology and Language. A Synopsis
Let us conclude this book with a synopsis of Moore’s and Wittgenstein’s views on the themes which we have been focusing upon. As should by now be apparent, they held deeply contrasting views, although they also had some points of convergence. I will summarize the main differences and analogies by presenting them in the order in which they came up in the previous chapters.
1 Truisms, hinges, common sense, world-picture and knowledge Moore and Wittgenstein agree on recognizing that there is a body of propositions that play a special role in our conceptual and epistemic systems. For Moore they can be fairly clearly stated and are said to constitute the common-sense picture of the world. Interestingly no scientific or religious proposition is taken to be part of it. For Wittgenstein, their identification is more complex and in fact wider. Moreover, for Moore they clearly have an empirical nature, are true, supported by evidence and are known even if we are unable to prove how. In contrast, for Wittgenstein, they play a normative role. Thus, they are like ‘hinges’, or rules, that must stay put, if we want to pursue our ordinary epistemic practices. Therefore, they are neither true nor false, neither justified/able nor unjustified/able. Hence, although everything we do regard as evidence would speak in their favour, they are neither known nor unknown. Connectedly, these different views on hinges’ knowability bear testimony to the fact that Moore and Wittgenstein ultimately disagreed on the very notion of knowledge. While Moore can be seen as a proto-externalist, Wittgenstein is deeply rooted in an internalist conception of knowledge, according to which, in order to know that p, 208
Conclusion
209
it must be possible, at least in principle, to make explicit one’s warrants for – supporting grounds, as opposed to mere evidence – for ‘p’.
2 Meaning, use and philosophical contexts By going through Moore’s correspondence with Malcolm, we have seen how there are deep disanalogies also between the former and Wittgenstein on the nature of meaning, its relationship with use and, consequently, regarding their views about the meaningfulness of philosophical contexts. Also the attack launched by Searle against Wittgenstein on the issue of the assertion fallacy can be seen as depending on a Moorean conception of meaning, use and context. Indeed, even contemporary discussions on semantic contextualism can be traced back to either a Moorean or a Wittgensteinian conception of meaning. In outline, while Moore thinks that any well-formed sentence of a given natural language has a meaning, although it may be used in contexts were its utterance serves no recognized aim of the communicative exchange, Wittgenstein thinks that only attested uses of signs can confer them meaning. So, seemingly wellformed sentences can in fact lack it, when uttered in contexts in which they serve no specifiable communicative purpose. They appear to have a meaning because we are naturally prone to take them as belonging to one or the other of their ordinary contexts of use. But this is in fact a delusion. Yet, for Wittgenstein, use can evolve through time and it may thus happen that signs which lacked a meaning acquired one or vice versa. Connectedly, while Moore thinks that when uttered in philosophical contexts, certain combinations of signs do make sense, provided they are well-formed sentences of a given natural language, Wittgenstein deems the opposite. For, in his view, in philosophical contexts words are used independently of the criteria that govern their ordinary uses. Hence, they only appear to have a meaning, while they actually lack it. They are therefore pieces of semantic nonsense. By contrast, Moore thinks that they are semantically meaningful, yet that they may be used to express or say something false, or paradoxical, and thus in some ways nonsensical, as it happens with sceptics and idealists, yet also something true, when used by him, for instance.
3 Scepticism In consequence of what we have just been reviewing, while Moore thinks that scepticism is meaningful yet either paradoxical (DCS) or unsupported by reasons or indeed false (C and FFS), Wittgenstein thinks
210
Moore and Wittgenstein
it is ultimately semantically nonsensical. Still, also in Wittgenstein this conclusion is reached, when dealing with Humean scepticism, by first exposing its being epistemically nonsensical. It is irrational because it is based on no reasons and would ultimately deprive sceptics and non-sceptics alike of the notion of epistemic rationality. Since, for Wittgenstein, a doubt which is raised against all reasons, or with no reasons at all is no real doubt, he then sees the irrationality of scepticism as entailing its being a piece of semantic nonsense. When dealing with Cartesian scepticism, in contrast, Wittgenstein deems it semantically nonsensical and does so by relying upon considerations which pertain to the actual and possible uses of ‘I am/may be dreaming right now’ either in speech or in thought.
4 Certainty With respect to the category of certainty in connection with truisms, arguably Moore saw it as a consequence of the fact that each of us, by being party to common sense, finds them indubitable. Certainty would thus ultimately be a psychological notion. By contrast, according to Wittgenstein, hinges are certain because of the normative role they play in our language and epistemic practices. Furthermore, in my reading of On Certainty, he held the view that we do accept them and hold them fast, accept criticism for occasionally failing to do so and pass them on to future generations by teaching them language and to participate in our epistemic practices.
5 Epistemic foundationalism and epistemic relativism Arguably, Moore was after a foundationalist project in epistemology, according to which, at bottom of all our knowledge would lie his truisms of common sense, viz. propositions which are true and known with certainty yet that can’t be proved. By contrast, Wittgenstein, in my reading of him, did not put forward any form of foundationalism, not even of a sui generis kind. For, in his view, hinges, being rules, lie beyond epistemic appraisal. Yet, I don’t think he ever proposed a form of epistemic relativism either. Despite appearances to the contrary, he defended the view that radically different hinges would be unintelligible. He thus seemed more interested in letting us see their ungroundedness yet also their unavoidability for us. So, I think we can safely conclude that he was in fact defending a form of epistemic anti-realism, rather than epistemic relativism, properly so viewed.
Notes Acknowledgements 1. Irving Berlin, ‘The best thing for you would be me’.
Introduction 1. 2.
Moyal-Sharrock and Brenner, 2005, 1. Kripke 1982. For an assessment of Kripke’s work with respect merely to its historical significance, see Glock 2008, 883–4.
1 G. E. Moore: Scepticism, Certainty and Common Sense 1. We will also take into account ‘Four forms of scepticism’ (FFS), written between 1940 and 1944, but published only in 1959 and ‘Certainty’ (C), written in 1941, but published in 1959, with strong reservations on Moore’s part (PP, Preface). We will use these papers to try and clarify Moore’s position, from time to time, but they won’t be the focus of our discussion as their later publication makes them less relevant to the debate between Moore and Wittgenstein. 2. DCS 33–4. 3. DCS 34. 4. DCS 37, 53. 5. DCS 54–5. 6. DCS 56–7. For a recent appraisal of Moore’s defence of (i), see Snowdon 2007. 7. DCS 57. 8. DCS 57–8. 9. PI I, 258. 10. These propositions are absolutely true for Moore, and not simply partially true, like an idealist à la Bradley, for whom truth admits of degrees and isn’t ever fully instantiated by our ordinary judgements, would have maintained. 11. According to Moore, an idealist can deny one or more, or even all the facts just listed, but is anyway mistaken. DCS 38–42. 12. DCS 40. 13. DCS 41. 14. But notice that Moore doesn’t explicitly mention either Kant or Bradley and talks of an unspecified philosopher who upholds these views. 15. DCS 41. 16. DCS 42. 17. DCS 42–3. This interestingly shows that Moore was here, at least implicitly, committed to the view that knowledge is the norm of assertion. For it is only 211
212
18. 19. 20.
21.
22. 23. 24. 25. 26. 27.
28.
Notes on such a basis that there would be a contradiction, since from the assertion of the first conjunct it would follow that one knows it, contrary to what is asserted in the second. In ‘Certainty’ (C 221) he goes so far as to argue that by asserting propositions similar to his truisms he implied to know them with certainty. In ‘Moore’s paradox,’ after claiming that by uttering a sentence assertively one implies to believe its content, he goes so far as to argue that people often imply to know it, and to know it with certainty (Moore’s paradox (MP) 211). So, either he in fact held that there are two norms of assertion or perhaps that by complying with the knowledge-norm one would be complying also with the belief-norm. DCS 50–1. DCS 45–52. It is seldom clear whom Moore is attacking, though. All we know in this particular case is that the first thesis is directly attributed to Berkeley (DCS 51). Although I think Moore is right in his criticism of what I have called ‘strong’ forms of idealism, there may be ‘weak’ ones (such as the ones subsequently put forward by Nelson Goodman and Richard Rorty), according to which there is no commitment to the logical and causal dependence of physical facts onto mental ones, but only to the thesis that insofar as we can sensibly speak of physical facts this is due to the application of our representational means – essentially concepts – and that in this, representational sense physical facts depend upon mental ones. It may be an open question whether and to what extent representational dependence would suffice to qualify a position as idealist. Be that as it may, Moore has no argument against such a view. It is also interesting to note that, in section III of DCS, Moore excludes from common sense truisms ‘God exists’ and ‘We shall live after the death of our bodies’. It is therefore clear that his notion of common sense, far from being a mere rendition of what most people believe (or believed at least at his time), is quite peculiar and mediated by rational and philosophical considerations. DCS 43. DCS 44. B xxxix. Moore translates ‘vorgestellet’ as ‘presented’, though the usual translation is ‘represented’. A 373. Alternatively, if by ‘physical objects’ one meant something compatible with idealist theses, such as objects that afford the possibility of occurrent perceptions, (4) would follow, but not (5), if, again, by ‘external world’ one meant a world populated by objects that exist independently of being perceived by us. Obviously, since Moore has painstakingly defined how ‘physical object’ and ‘external world’ should be understood, the alternative reading just presented is not the intended reading of Moore’s proof, as he himself made clear in ‘A reply to my critics’ (RMC 669–70). RMC 668: ‘I have sometimes distinguished between two different propositions, each of which has been made by some philosophers, namely (1) the proposition “There are no material things” and (2) the proposition “Nobody knows for certain that there are any material things”. And in my last published writing, my British Academy lecture called “Proof of an external
Notes
29.
30. 31.
32.
33. 34. 35. 36. 37. 38.
213
world” (…) I implied with regard to the first of these propositions that it could be proved to be false in such a way as this; namely, by holding up one of your hands and saying “This hand is a material thing; therefore there is at least one material thing”. But with regard to the second of these two propositions, which has, I think, been far more commonly asserted than the first, I do not think I ever implied that it could be proved to be false in any such simple way; e.g., by holding up one of your hands and saying “I know that this hand is a material thing; therefore at least one person knows that there is at least one material thing.”’ PEW 149: ‘How am I to prove now that “Here’s one hand and here’s another”? I do not believe I can do it. In order to do it, I should need to prove for one thing, as Descartes pointed out, that I am not now dreaming. But how can I prove that I am not? I have, no doubt, conclusive reasons for asserting that I am not now dreaming; I have conclusive evidence that I am awake: but that is a very different thing from being able to prove it. I could not tell you what all my evidence is; and I should require to do this at least, in order to give you a proof.’ In fact the problem is not that one could not mention all of one’s available evidence in favour of ‘I am not now dreaming’, contrary to what Sosa 2007 claims. Rather, it lies in the fact that, for a Cartesian sceptic, all that evidence would be compatible with the hypothesis that one were merely dreaming of it. A notable exception is Sosa 2007, 52. Unless one denied the principle of closure for knowledge – according to which, if you know that p and you know that p entails q, you know that q – as Dretske 1970 and Nozick 1981 did. Moore, however, never proposed such a thing. Moore often speaks simply of proving or being able to prove his premises, and hence the conclusion of his argument, rather than of proving that he knows them. However, a proof is a procedure which should allow one to reflexively know what it is thereby proved. Lacking a proof of p, one then can’t show how one knows, and reassure oneself of knowing it, though one may of course, with luck, in fact know that p. Moreover, the connection between proving that there is one’s hand where one seems to see it and the dreaming hypothesis (cf. note 29) suggests that the proof Moore is lacking is a proof about his having knowledge of ‘Here is my hand.’ PEW 146. PEW 150. Malcolm 1958, 70–1. Ibid. 30. See Ambrose 1942 and Lazerowitz 1942. Malcom’s reconstruction of the sceptical position is based on what Alfred J. Ayer 1940, 44 maintains: ‘We do indeed verify many such propositions [i.e. propositions which imply the existence of material things] to an extent that makes it highly probable that they are true; but since the series of relevant tests, being infinite, can never be exhausted, this probability can never amount to logical certainty. […] It is an analytical proposition that one cannot run through all the members of an infinite series. […] Accordingly, what we should say […] is not that we can never be certain that any of the propositions in which we express our perceptual judgments are true, but
214
39.
40.
41. 42. 43. 44. 45. 46.
Notes rather that the notion of certainty does not apply to propositions of this kind. It applies to the a priori propositions of logic and mathematics, and the fact that it does apply to them is an essential mark of distinction between them and empirical propositions’. FFS 222: ‘I don’t see any reason to abandon my view that I do know for certain (…) that I am not dreaming now. And the mere proposition, which I admit, that percepts of the same kind in certain respects do sometimes occur in dreams, is, I am quite certain, no good reason for saying: this percept may be one which is occurring in a dream’. See also PEW 149: ‘I have, no doubt, conclusive reasons for asserting that I am not now dreaming; I have conclusive evidence that I am awake: but that is a very different thing from being able to prove it’. In FFS 220, Moore writes, ‘It seems to me more certain that I do know that this is a pencil and that you are conscious, than that any single one of these four assumptions [from which it would follow that he didn’t know that] is true, let alone all four. (…) Of no one of [them] do I feel as certain as that I do know for certain that this is a pencil. Nay more: I do not think it is rational to be as certain of any one of [them], as of the proposition that I do know that this is a pencil. And how on earth is it to be decided which of the two things it is rational to be most certain of?’. See also C 247: ‘I agree, therefore, with that part of this argument which asserts that if I don’t know now that I’m not dreaming, it follows that I don’t know that I am standing up, even if I both actually am and think that I am. But this first part of the argument is a consideration which cuts both ways. For, if it is true, it follows that it is also true that if I do know that I am standing up, then I do know that I am not dreaming. I can therefore just as well argue: since I do know that I’m standing up, it follows that I do know that I’m not dreaming; as my opponent can argue: since you don’t know that you’re not dreaming, it follows that you don’t know that you’re standing up. The one argument is just as good as the other, unless my opponent can give better [my italics] reasons for asserting that I don’t know that I’m not dreaming, than I can give for asserting that I do know that I am standing up.’ Lycan 2007 shows how this line of reasoning has been present in Moore’s writings since very early on. However, I personally don’t share his confidence in deeming this a good anti-sceptical argument. As we shall see, a sceptic need not claim that believing the conclusion of his arguments is more or even as rational as believing the opposite (given one’s collateral beliefs). All he needs to show is that it is a metaphysical possibility one cannot rule out on the basis of one’s available evidence. Notice, moreover, that Moore himself never denied such a thing and that is why, I believe, he was ultimately dissatisfied with FFS and C. We will consider them more fully in the following, §§6–7. In Malcolm 1977 the number of conditions on the correct use of ‘I know (with certainty) that p’ rises up to 12. Cf. note 58. Moore’s letter to Malcolm, in Malcolm 1977, LM 173–4. Reprinted in SW 213–6. OC 58. LM 215–6 (reference to Malcolm’s version).
Notes
215
47. Moore is in fact trying to resist the inference from being unable to prove that one knows that p, to one’s not knowing that p, by insisting that he does know that p, even if he can’t prove it. In this sense, his contention in this letter written in 1949 that he was using ‘I know’ with the ‘purpose of disproving a general proposition which many philosophers have made’ isn’t in contrast with his pronouncements in PEW and in RMC written in 1942, that he didn’t aim to refute scepticism in that paper, as that would have required to prove that he knew the premises of his conclusion – something which he was unable to do. 48. See Grice 1957. 49. Clarke 1972, 292. 50. Clarke 1972, 293–4. 51. However, according to Clarke, our concept of dream doesn’t satisfy the conditions for its purely philosophical use, for any of its ordinary applications requires that it be possible to employ it with respect to any x if and only if it can be known that x belongs to the real world, even if one doesn’t actually know that it does. Descartes’ argument from dreaming, however, denies such a possibility. Hence, according to Clarke, it violates the conditions of the correct employment of such a concept. It seems to me that Clarke’s way of putting the point may be confusing. For, according to his first condition of meaningfulness for philosophical talk, a concept must be severed from its ordinary counterpart. I therefore think that Clarke’s conclusion would be more secure if grounded on his third condition for a purely philosophical usage of ‘dream’. For, within Descartes’ hypothesis, one could never know whether a certain item is a dream or a veridical experience. 52. Stroud 1984, 127. 53. For a similar objection to Stroud’s interpretation of Moore, see McGinn 1989, Ch. 3. 54. Travis 1989, 165–6, who is sympathetic to Moore’s claim that he knows the premises of his proof as well as his truisms, and who supports contextualism about knowledge ascriptions, clearly recognizes this point and actually deems it the source of the failure of Moore’s anti-sceptical strategy. 55. Moore seems to have favoured what is now called ‘the identity theory of truth’. See his ‘Truth and falsity’, TF 20–3. 56. McGinn 1989, 49–53, rightly notes that Moore is directly engaging with scepticism and isn’t confining his claims to knowledge merely to ordinary contexts. Yet she herself shares the view that he is being dogmatic and, hence, somewhat unphilosophical, by failing to diagnose what is wrong with scepticism. In contrast, and as will become apparent in the following (esp. §7), I think Moore is proposing a subtler, though not obviously, successful move. 57. According to Travis 1989, 192–6, however, Moore appreciated the contextsensitivity of proofs, though he confined it to the premises of his own proof of an external world. So, on Travis’ reading of PEW, Moore agreed with a sceptic that he could not prove ‘Here is one hand’ and ‘Here is another.’ Yet he claimed that he knew them nonetheless, because, according to Travis, he thought that, in given contexts, they simply couldn’t be proved. Still, he held that on different ones they might well be. From a textual point of view, however, Travis’ interpretation of Moore as a proto-contextualist
216
Notes
about proof has nothing on its side. For he never explicitly claimed that, in different contexts, such a proof could in fact be given. As we shall see in the following, this was in fact one of Wittgenstein’s intuitions, not Moore’s. Nor did Moore ever claim that he lacked evidence for holding the premises of his proof (note 58). Hence, if he thought that no proof of the premises of PEW could be given, he thought so while having in mind a rather special notion of proof, whose absence was compatible with his having evidence and grounds for those very premises. Thus, I think we can safely conclude that Moore was neither a contextualist about knowledge nor about proof. 58. In support of this reading of Moore, see C 243: ‘A third characteristic which was common to all those seven propositions [namely, propositions about material objects in Moore’s surroundings] was one which I am going to express by saying that I had for each of them, at the time when I made it, the evidence of my senses (…). In other words, in all seven cases, what I said was at least partly based on ‘the then present evidence of my senses’. 59. See Malcolm 1949 and Stroll 1994, Ch. 4, 50–2. 60. In his 2002 paper Wright puts forward another possible rendition of Moore’s proof where (II) isn’t inferred from a proposition about one’s experience such as (I). Rather, it is simply grounded in one’s experience. In Wright’s view that makes no difference to the eventual diagnosis of the proof because also in this case one’s perceptual warrant would depend not just on one’s available experience but also on having independent warrant that there is an external world. On this latter reading, the proof would simply be an instance of modus ponens of the following form: Here is a hand. If there is a hand here, then there is an external world. Therefore, there is an external world. 61. Warrant for (II) is a necessary condition for having knowledge of (II), on the tripartite conception of knowledge. Since nowhere to my knowledge does Moore impugn the tripartite account, lack of such a warrant would impugn one’s alleged knowledge of (II), also by Moore’s lights. So, the fact that Wright is talking about warrant while Moore talks of knowledge, though certainly inaccurate from a historical point of view, makes no substantial difference from a conceptual one – or so it seems to me. 62. Cf. Wright 1985, 2004a, b; Coliva 2008, 2010a. 63. Or else, in the case of induction, in order to go from a premise about certain regular patterns of events witnessed in the past, to a premise about the fact that the same pattern of events will take place in the future, one already needs to have warrant for the conclusion – namely, that there are uniformities of nature. 64. These qualifications are important in order to clarify the difference between ‘transmission failure’ and the failure of the principle of closure for epistemic operators (note 31). There is an ongoing discussion about whether, once granted the specific conception of the architecture of empirical warrants recommended by Wright, Moore’s proof could at least enhance the previous warrant one might have for (III), once the principle of closure for evidential
Notes
65.
66.
67. 68.
69. 70. 71. 72. 73. 74. 75. 76. 77. 78.
217
warrant is retained. Be that as it may, there is substantial agreement that, given that conception of empirical warrants, the proof couldn’t give one a first evidential warrant to believe its conclusion. Things might be different, though, if Wright’s attack were meant to impugn merely Moore’s ability to redeem his knowledge of (III) – namely, the ability of proving that he did really have it – as Wright’s more recent discussions of Moore’s proof seem sometimes to suggest (Wright 2004a, 167, 210–1; Id., 2007); and if this de-coupling could be matched by endorsing an externalist notion of knowledge (and/or warrant). In such a case Moore’s proof could establish that its conclusion is known – since its premises would be – and yet, just as Moore held, fail at proving, against scepticism, that either the premises or the conclusion be known. We will come back to this issue in §7. ‘Here is a hand’ would be such a perceptually basic belief. If one found this claim odd, Pryor would allow substituting it with ‘Here is a pinkish expense.’ I am not sure whether Moore himself would be happy with that substitution. Bet that as it may, the important point (for both) is that perceptually basic beliefs would be about physical objects. Pryor in fact considers Wright’s later rendition of Moore’s proof (note 60). This is indeed a contentious rendition of the sceptical position, for a sceptic is no idealist! Rather, on the basis of philosophical arguments he holds an agnostic position and, in particular, that it can’t be warrantedly believed either that there is an external world, or that there isn’t. But, as already noticed, Cartesian and Humean sceptics reach this position for slightly different reasons. The former think that since all our perceptual evidence is compatible with the hypothesis that we might be dreaming of it, no single belief about specific physical objects can be warranted and therefore known. From that, they conclude that our belief about the fact that the whole category of physical objects isn’t empty is equally unwarranted and unknowable. Humean scepticism, in contrast, directly shows the latter belief to be unwarrantable and unknowable because any warrant one may produce for it would in turn depend on already having warrant for it, see Coliva (2008, 2010a). Cf. Pryor 2004, 368. See Coliva 2010a. FFS 226. But it should be kept in mind that Moore wrote FFS between 1940 and 1944, thus after PEW. See notes 39–40. Sosa 2007, 50 concurs with this appraisal. PEW 149. PEW 150. The former phrase is Wright’s (2004a, 167, 210–11); the latter is Pritchard’s (2005a); similar remarks have been made by Stroud 1994. I think Sosa (1994) concurs with this judgement, although he thinks – wrongly to my mind – that there is some kind of logical flaw in the demands placed by traditional epistemology, which would lead to scepticism about the possibility of having a satisfactory understanding of human knowledge. I think, in contrast, that the necessary circularity involved in trying to meet those demands simply points out that we cannot meet them. But this is by itself no proof of their incoherence.
218
Notes
79. On such an account what would be necessary and sufficient in order to have an internalist notion of knowledge is that the justification be available in general and surveyable, at least by experts, and not that it be actually at the disposal of one particular subject at a particular time. If that sounded too compromised with externalism, one could hold that the subject who should have been able to survey such a justification is the ideal epistemic agent from whose perspective epistemology is usually done. 80. See Wright 2004a, 210–11 and Pritchard 2005a and my earlier writings, though I don’t hold this view anymore. 81. Cf. Pritchard 2005a, 204. 82. It is opaque to me whether this strategy would legitimate contextualism. For the latter aims at a form of compatibilism which isn’t the outcome of the position we are considering here. For, as it will be apparent, the outcome of the view under consideration would be that the very concept of knowledge we possess is, after all, incoherent. Yet contemporary contextualism assumes that we have a unique and coherent concept of knowledge, and then makes play with the idea that the standards for knowledge attributions may differ depending on differences in context (of utterance or of assessment, on its relativist variant). 83. Of course, unless one were happy to think that there are true contradictions. This, however, is a problematic view in its own right and, to the best of my knowledge, has never been proposed with respect to knowledge. 84. This is important in order to be able sensibly to maintain that there may be two different kinds of knowledge. If, in contrast, one simply said that we have two different concepts of knowledge – one externalist and one internalist – it would thus become arbitrary why they should both count as knowledge or, equally, why, depending on one’s choice, only one would. 85. Roughly, a belief that p is safe if p isn’t false in nearby possible worlds. 86. Personally I endorse an internalist view, whereby justification must be available and surveyable at least in principle in order for a subject to have knowledge. On that view, Moore would fail even to hear the Cartesian sceptical challenge, which precisely targets the fact that he can claim to know the premises of his proof – and hence really know them – as I have argued in my 2008 article.
2 Wittgenstein: Belief, Knowledge and Certainty 1. This, again, does not take the form of a theory, on Wittgenstein’s view of the matter, but of an elucidation of the role of these propositions with respect to our language games and epistemic practices. 2. OC 424, 389. 3. Williams 2004a, 92–3 makes a similar point. For an ample discussion of the difference between ‘to know’ and ‘to believe’, see Armengaud 1983. The author compares the use of ‘to know’ made by Moore and Wittgenstein’s observations on this issue in OC. In particular, Moore’s use of ‘I know’ is mistaken, according to Armengaud, because of an erroneous assimilation of the concepts of knowledge and belief. In general, notes Armengaud, there are three kinds of belief: (a) belief-conviction; (b) belief-conjecture;
Notes
219
(c) belief-faith. (a) consists of belief in what one knows and in knowledge of what is believed, so that the difference between belief and knowledge disappears. (b), in contrast, consists of belief in what isn’t known, but could be known, in such a way that the difference between belief and knowledge is only contingent. Finally, (c) consists in believing what isn’t known and can’t be known, whence the difference between belief and knowledge is de jure and not merely de facto. Armengaud then compares the two concepts following Wittgenstein’s suggestions in OC and stresses how, for Wittgenstein, (i) ‘I believe’ expresses the relation between a subject and sense of a proposition, while ‘I know’ expresses the relation between a subject and a fact (OC 90). (ii) One can say ‘A believes that p, but it isn’t the case that p’, but not ‘A knows that p, but it isn’t the case that p’ (OC 42). (iii) One can believe something without being able to answer the question ‘Why do you believe it?’, but one can’t know something without being able to answer the question ‘How do you know it?’ (OC 550). (iv) ‘I believe’ expresses a subjective truth, while ‘I know’ doesn’t (OC 179). (v) The use of ‘I know’ entails the ability to give cogent grounds in favour of the truth of the proposition that is said to be known; hence the reasons adduced in its favour can’t be weaker than the proposition which is claimed to be known. Therefore, according to Armengaud, Wittgenstein objects to Moore for having assimilated ‘I know’ to the subjectivity of ‘I believe’ and for having meant the former as the description of a mental state, with respect to whom a subject can’t be mistaken. A subject who claimed to know that p couldn’t thus make a mistake with respect to his knowledge, from which it would follow the truth of ‘p’ and thus of his very claim. This, however, is obviously false. 4. OC 482: ‘It is as if “I know” did not tolerate a metaphysical emphasis’. 5. For an interesting and complex defence of the idea that knowledge is a mental state and not just belief accompanied by having a justification for ‘p’, while ‘p’ is true, see Williamson 2000. At any rate, Williamson thinks that one’s mental state isn’t transparent to the subject and hence that one can be mistaken with respect to it and thus either ignore it or falsely claim a piece of knowledge one in fact lacks (Williamson 2000, 23–7). 6. I therefore agree with what I take to be the gist of Moyal-Sharrock’s discussion about the distinction between saying and speaking (2005a, 43–7). In her view, rules can be voiced or spoken but not said. I would however express the basic point made by Wittgenstein thus: rules, whenever they are stated, spoken, voiced, or even said, if we use this word with no special connotation, simply to refer to the act of formulating them, and not to what is actually done by formulating them, do not describe any state of affairs. Yet, they can be voiced either to give one an instruction – like ‘This is a hand/ what we call “hand” in English’ – or to remind (especially) philosophers of the real function of those propositions, given the former’s tendency to conflate them with descriptions of (indubitable) states of affairs. In either of the last two cases, the statement isn’t part of what we may call a ‘descriptive language game’ – that is to say, of a language game in which words are used to describe states of affairs. Rather, it is either an instruction or a reminder about the former’s preconditions. Hence, such instructions and reminders are perfectly proper uses of language, or, equivalently, language games, that we may call, in contrast with the former, ‘normative language games’ – that
220
7.
8. 9.
10.
11.
12.
Notes is to say, those language games in which words’ meanings are defined, either ostensively or verbally, or reminded to subjects. I disagree, therefore, with Moyal-Sharrock’s insistence on the ineffability of such rules (and, as we shall see in Chapter 4, of Wittgenstein’s hinges). Rules can be stated for either of the purposes we have just seen, though, of course, that doesn’t mean that by so doing words occurring in them are being used to describe states of affairs and, therefore, given Wittgenstein’s ‘technical’ use of ‘saying’, for ‘saying’ something. In short, rules are effable, though their being said or spoken is never a description of a state of affairs. I therefore entirely agree with Conant’s 1998 claim that, when spoken by Moore or a sceptic, for instance, certain combinations of words altogether fail, for Wittgenstein, to have a meaning. See the rest of this chapter for a development of this point. PI I, 65–77. What I should have said, rather, is that I believed or was convinced that he was at home, though I had no evidence in favour of my belief/conviction. Similarly, no one would say that I knew such a thing if I were unable to give reasons in favour of my claim that NN was at home yesterday and they would correct my assertion ‘I know NN was at home yesterday’ by saying that I only believed it or was convinced of it, but that I didn’t really know it. This shows, once more, the objectivity of knowledge (as opposed to the subjectivity of belief) in that the use of ‘to know’ in the first and third person is always governed by criteria. In particular, that a subject’s ability to offer his grounds is a criterion both for his correctly claiming ‘I know that p’ and for someone else’s proper assertion ‘S knows that p’, though both assertions remain defeasible. For the meaning of an expression is its use (PI I, 43) and thus the moves made with it in a language game. If, then, a string of words is uttered while being no move in the language game, they would lack a meaning, contrary appearances notwithstanding. Compare PI I, 432 where Wittgenstein writes, ‘Every sign by itself seems dead. What gives it life? – In use it is alive’. There will be more on this in the following. Wittgenstein deemed it important to be able to imagine a language in which our concept of knowledge didn’t exist (OC 562). If that were the case, what could replace it? In some cases, the mere assertion of ‘p’ followed by the exhibition of reasons in its favour. In other cases this, however, wouldn’t suffice. We will presently come back to this issue in the main text. Be that as it may, this shows that for Wittgenstein the norm of assertion – that is, what makes a speech act as an assertion, that may, however, be false – was to assert only what one believes (to be true, of course) on the basis of grounds, which may or may not suffice for knowledge. For a different account of the norm of the assertion, see Williamson 2000, ch. 11, according to whom that norm must make reference to knowledge, rather than to merely truth or warranted belief. There are also cases in which with ‘I know’ one wants to put emphasis on the subject, perhaps because one wants to underline the fact that one has a piece of information unavailable to others. But even here the emphasis on the first person pronoun isn’t a symptom of the fact that one is describing one’s own mental state (OC 588–90).
Notes
221
13. Compare OC 125: ‘If a blind man were to ask me “Have you got two hands?” I should not make sure by looking. If I were to have any doubt of it, then I don’t know why I should trust my eyes. For why shouldn’t I test my eyes by looking to find out whether I see my two hands? What is to be tested by what? (Who decides what stands fast?)’. 14. We won’t consider (i) here. 15. See Conant 1998 for a similar assessment, developed in opposition to McGinn (1989, 106–9), who wants to see relevance as a merely pragmatic requirement over claims to knowledge as opposed to a grammatical one. 16. But we will have to come back to the real meaning of these words for Wittgenstein (§3). 17. Again, notice the qualification – ‘roughly speaking’ – made by Wittgenstein. We will come back to it in due course (§3). 18. And, for the reasons given in §2.1, also simply to assert the obvious, unaccompanied by a claim to knowledge, isn’t a move in the language game of making non-epistemic assertions (OC 348, 441, 460, 468–9). 19. Cf. PI I, 38. Compare also Travis 1989, 153 on this point. 20. There will be more on this issue in §4. 21. Cf. TLP 3.144, 3.221, 6.111–6.126. 22. McGinn (1989, 113–7) too stresses the possibility of being mistaken as one of the characteristic features of the grammar of the expression ‘I know’. 23. Again, notice the qualification ‘something I seem to know’. We will clarify the sense of such remarks in §3. 24. OC 65, 256, 336. 25. OC 65: ‘When language-games change, then there is a change in concepts, and with the concepts the meanings of words change’. 26. This is Morawetz’s 1978 view, repeated in Morawetz 2005 and approvingly mentioned in Pritchard 2010. Such a reading would make Wittgenstein’s position unbelievably close to externalism about knowledge. A thesis this which has no textual evidence in its favour (cf. Moyal-Sharrock 2005a, 15), and that, I surmise, is based on textual misinterpretation and on the failure to acknowledge the so-called grammatical use of ‘I know’, which we shall review in the following. Furthermore, it may be based on failing to appreciate that in a set of passages where Wittgenstein is wondering whether, after all, Moore knows what he claims to know (OC 527–49), he comes to the conclusion that practical knowledge of, for example, that those are one’s hands, can be attributed to subjects also when they aren’t capable of making sensible claims to knowledge, supported by grounds. Yet, in OC 550–1, propositional knowledge is clearly said to require the ability to produce one’s grounds for it. That Wittgenstein adhered to a classical internalist and tripartite conception of knowledge, according to which knowledge consists in justified true belief, can also be evinced from the fact that he deemed knowledge to entail belief (OC 42) and that he tied it to the possession of available reasons and grounds for the proposition known (cf. the following fn for a set of passages where the availability of justifications is said to be required for knowledge). 27. OC 18, 91, 243, 245, 250, 307, 438, 483–4, 504, 555, 563–4. 28. Compare PI II, xi, 225 (my italics): ‘“While you can have complete certainty about someone else’s state of mind, still it is always merely subjective certainty, not objective certainty” – These two words betoken a difference between
222
29.
30.
31.
32.
33.
Notes language-games’. Interestingly the issue of subjective/objective certainty occurs in connection with (so-called) knowledge of one’s own and others’ mental states. As we will see in the following (§4), Wittgenstein’s remarks about Moorean certainties interestingly parallel, in many ways, those about self-‘knowledge’. One could put the point simply by saying that for Wittgenstein the fact that a sentence involves reference to physical objects isn’t a sufficient condition for its expressing an empirical proposition. That is to say, a proposition that describes a state of affairs and that can thus be both true and false. We will come back to this issue in the last chapter. For now, let me simply note that Moyal-Sharrock (2005a, esp. Chs. 2, 4) argues that Wittgenstein did not want to consider these ‘propositions’ as propositions at all, because they aren’t bipolar – they can’t be both true and false – and because bipolarity is the mark of propositionality throughout Wittgenstein’s entire philosophical production. I beg to disagree. The real point, for Wittgenstein, in OC, is the distinction between Erfahrungssätze – empirical propositions – and Regeln – rules. That is to say, Wittgenstein’s concern is with the fact that propositions which look like genuinely empirical ones and that, at least in some contexts, can be used to make empirical statements – that is to say, to make genuine descriptions or hypotheses that can be both true and false – are in fact sometimes turned into rules or norms of description and are therefore no longer empirical propositions. So the point of contrast is between the empirical/factual/descriptive and the regulative/normative role of Sätze – propositions – in the language. That ‘proposition’ – Satz – be such an umbrella term, at least in OC, can be evinced from, especially, OC 320 (see also 319 and 321): ‘Here one must, I believe, remember that the concept ‘proposition’ [Satz] itself is not a sharp one’. This passage, taken in its proper context, suggests that Wittgenstein was inclined to view both ‘statements about material objects’ which aren’t in fact empirical propositions, but, rather, akin to rules, and genuine empirical propositions as propositions, though he wouldn’t have likened their role and characteristic features. The former being normative and lacking bipolarity, the latter being descriptive and bipolar. There will be more on this issue in Ch. 4, §1. OC 231: ‘If someone doubted whether the earth had existed a hundred years ago, I should not understand, for this reason: I would not know what such a person would still allow to be counted as evidence [my emphasis] and what not’. Cf. OC 670. For an in-depth discussion of the notion of error and its impossibility in OC, see Soles 1982, where the difference between mistakes and mental disturbance is analysed. According to Soles, a mistake when, as a norm, one would say ‘I can’t be making a mistake in this case’, will be a symptom of mental illness or, at any rate, will call into question a subject’s cognitive capacities. OC 114, 126, 158, 268, 306, 369–70, 456, 506–7. Compare the Remarks on the Foundation of Mathematics, 324–5, where Wittgenstein writes, ‘It is as if we had hardened empirical propositions into a rule. And now we have, not a hypothesis that gets tested by experience, but a paradigm with which experience is compared and judged. And so a new kind of judgment. (…) An empirical proposition hardened into a rule (…) It is thus withdrawn from being checked by experience, but now serves as a paradigm for judging experience’. The point seems to be that certain judgements cease to be hypotheses
Notes
34.
35. 36.
37.
38. 39. 40. 41.
42. 43. 44.
45.
223
and become paradigmatic in such a way as to acquire a normative, rule-like function, for determining the veridicality of experience and hence the truth or falsity of genuinely empirical judgements. Interestingly, the final twentyeight entries of OC are devoted to the impossibility of being mistaken about some propositions, their rule-like function and their comparison with mathematical ones (OC 447–8, 455). Such an extension of ‘grammar’ to not merely linguistic norms is what, according to Moyal-Sharrock 2005a, 163–5, marks the passage from the ‘second’ to the ‘third’ Wittgenstein. See note 33. Moyal-Sharrock 2005a talks, in this connection, of ‘judgement’ as a ‘transit term’, occasionally employed by Wittgenstein before settling for the view that its alleged content is just a norm. I think, in contrast, that here Wittgenstein is using the right word, from his point of view. Namely, the point of view according to which to judge that here is a hand in circumstances such as Moore’s is the right thing to do; it contributes to the determination of the meaning of the word ‘hand’; and, finally, fixes what is beyond doubt and enquiry in that context. Hence, while being a judgement, it happens to play a normative role, as opposed to a genuinely empirical one. An early remark which can be interpreted as pointing in the same direction is Malcolm’s (1952, 187; quoted in Moyal-Sharrock 2005a, 135): ‘In a certain important respect some a priori statements and empirical statements possess the same logical character. The statement that 5 × 5 = 25 and that here is an ink-bottle, both lie beyond the reach of doubt. On both my judgment and reasoning rests. If you could somehow undermine my confidence in either, you would not teach me caution. You would fill my mind with chaos!’ See Brenner 2005 and Rudd 2005. Moyal-Sharrock’s remarks (2005a, Ch. 6) point in the same direction, though less explicitly. Garver 1984, 212–3. For a discussion on the notion of a form of life, see von Wright 1972, Wolgast 1987, Johannessen 1988, and Garver 1994. OC 560: ‘And the concept of knowing is coupled with that of the language-game’. OC 541: ‘“He only knows what this person is called – not yet what that person is called”. That is something one cannot, strictly speaking, say of someone who simply has not yet got the concept of people’s having names’. Moyal-Sharrock 2005a, Chs 5–6 stresses the same point. OC 116, 125, 144, 151–2, 234–5. Cf. PI I, 246–7, where Wittgenstein marks the contrast between empirical, grammatical, and philosophical uses of ‘to know’ in relation to psychological avowals. Moyal-Sharrock 2005a talks of merely heuristic formulations of rules that have no real use in the language. I think there may be a conflation here between two possible senses of ‘use’. Surely formulations of rules have a use in our language. Indeed we can even say that there are language games – the normative ones (cf. note 6) – which precisely consist in formulating these rules explicitly, in order to instruct someone about the use of the words in the language. Similarly we do remind subjects (and on Wittgenstein’s view of the matter, philosophers) of such rules, when they display a tendency to violate or misunderstand them. But of course words occurring in these
224
46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.
57. 58. 59. 60.
61. 62. 63.
64.
65. 66.
67.
Notes language games aren’t used, but mentioned and defined or explicated. In contrast, they are used in the non-normative language games in which they occur (cf. note 6). In OC the idea of the scaffolding of the world is dropped while it is retained with respect to thought. SA 141. SA 131. SA 146. SA 141. SA 142–3. SA 144. SA 149. SA 137. SA 145–6. PI I, 246. But also PI I, 247 is important. Writes Wittgenstein, ‘“Only you can know if you had that intention.” One might tell someone this when one was explaining the meaning of the word “intention” to him. For then it means: that is how we use it. (And here “know” means that the expression of uncertainty is senseless)’. PI I, 244. Cf. PI II, xi, 222: ‘I can know what someone else is thinking, not what I am thinking’. Cf. PI I, 248: ‘The proposition ‘Sensations are private’ is comparable to ‘One plays patience by oneself’. It is therefore important to note how Wittgenstein isn’t discriminating between normal and abnormal conditions, but only between different language games, some of which may be more common than other ones or less. This pre-empts Searle’s third objection listed in the main text. Thus, I am perfectly in agreement with Conant 1998 on this much. SA 137. On the illusion or appearance of sense of Moore’s claim to knowledge or of sceptical doubts, cf. OC 10, 19–20, 24, 35–7, 53–7, 237, 347–55, 372, 406–8, 412–4, 423, 433, 451, 467, 477–80, 526, 553–4, 622. Travis 1989, 153. As will become apparent in the rest of this section, however, they mean something different by that from what therapists and myself do mean. See, for instance, Travis 1989, 156–66. Not surprisingly, Travis (1989, 234–5) denies that hinges are normative in nature, according to Wittgenstein, though citing no passage in OC to support his claim. There will be more on this issue in Ch. 3, §3.2.2.
3 Wittgenstein: Doubts and the Nonsense of Scepticism 1.
OC 122–3, 221, 247, 322–3, 458, 516. On this topic, see also Bouveresse 1976/1987, 608, where he writes, ‘There must be reasons to doubt. This doesn’t mean simply that doubt isn’t a normal or usual attitude. This would be such a classic objection as to be powerless against philosophical scepticism.
Notes
2.
3. 4.
5. 6.
7. 8.
9.
10.
225
Wittgenstein wants to say that a doubt that isn’t positively motivated is hardly understandable, given what we normally call “to doubt” and “to believe”’. Compare the analysis of the methodological constraints to which our doubts are subject proposed in Williams 1991, Ch. 3, 117 and §3.6. Williams notices how any enquiry, far from being unrestricted – contrary to what a sceptic claims epistemological analysis to be – is subject to several constraints. Williams individuates three kinds of constraint: (a) topical/disciplinary constraints: for instance, not to call into question whether the earth has existed for a very long time is a precondition for doing history. These disciplinary constraints fix the domain of the admissible questions and doubts. (b) Dialectical constraints: these are limitations imposed to legitimate questions and doubts by the state of the art in a given discipline. So to doubt that the earth is roughly spherical isn’t admissible, given the present state of the art in astronomy. (c) Situational constraints: these constraints are imposed by the kind of situation or context one finds oneself in. For instance, in many cases it would not be sensible to raise a doubt about the present existence of one’s hand, yet, clearly, things may be otherwise, for example, after a car accident. OC 190: ‘What we call historical evidence points to the existence of the earth a long time before my birth; – the opposite hypothesis has nothing on its side.’ OC 185: ‘It would strike me as ridiculous to want to doubt the existence of Napoleon; but if someone doubted the existence of the earth 150 years ago, perhaps I should be more willing to listen, for now he is doubting our whole system of evidence. It does not strike me as if this system were more certain than a certainty within it.’ We will come back to the relationship between a system of evidence and assumptions in the next chapter. Moyal-Sharrock 2005a, Ch. 3 rightly distinguishes between the attitude of trust, which characterizes certainty, and the objects of certainty, which she calls ‘certainties’. Contrary to what a sceptic would recommend. In this passage and in OC 25 Wittgenstein is actually assessing the question ‘What right do I have not to doubt the existence of my hands?’ and OC 25 makes clear that while a doubt about the existence of one’s hands is possible in certain circumstances, it is not always possible. In particular, it isn’t possible when one is holding one’s hands up in clear view and is cognitively lucid. I have modified the quote in order to address directly the fundamental issue of doubts about the existence of an external world in general. Williams 2004a, 79 too stresses that for Wittgenstein sceptical doubts are nonsensical and concludes, ‘If the scruples of the sceptic or the idealist are incoherent, then so are the reassurances of the realist. No proof is possible because there is nothing to prove. This means that a response to scepticism cannot be dialectical: that is, it cannot take the form of showing that the sceptic is wrong, proving what he doubts. Rather, it must be diagnostic and therapeutic. It must identify the conceptual misunderstanding that gives rise to the illusion of the sceptical doubts; and it must explain why the sceptic fails to see the illusion for what it is’. I agree with the spirit of Williams’ considerations, but I propose a different account of Wittgenstein’s diagnostic and therapeutic strategies. I think that Williams 2004a, 86–7 crucially misses this point. Namely, that ‘There are physical objects’ is nonsense if it is meant as an empirical statement.
226
11.
12.
13.
14. 15. 16.
Notes Yet, it is not nonsense if it is seen for what it is, viz. ‘a piece of instruction’, that is to say as a statement which makes explicit a feature of grammar. Hence, as a matter of fact, ‘There are physical objects’ is nonsense if and only if it is interpreted – like most philosophers interpret it – as a descriptive, empirical claim, which admits of a sensible negation, yet is supposed to say something about the ultimate structure of reality, and is thus turned into a metaphysical statement. It is this crucial misunderstanding that leads Williams to claim, ‘“There are physical objects” is not a hinge proposition: it is nonsense’ (2004a, 87). On the contrary, I think Wittgenstein is claiming that it is a hinge proposition, precisely because it is a statement that makes explicit a feature of the grammar of our language and, when interpreted this way, it is not nonsense, but, as we saw in the previous chapter (§3.4), at most sinnlos. That is to say, neither susceptible to semantic evaluation nor to epistemic grounding. I thus agree with Williams 2004a who stresses that Wittgenstein is concerned with denying absolute epistemic priority to judgements about sense data. However, I do so for reasons radically different from Williams’, whose aim is to contrast the so-called framework reading of these passages. Accordingly, Wittgenstein would be suggesting that both realists and idealists are mischaracterizing ‘There are physical objects’ by failing to realize its grammatical nature. According to Williams, in contrast, and as we saw in the previous note, ‘There are physical objects’ is just nonsense. As we shall see at much greater length in the following (§3.2.2), Williams thinks that Wittgenstein ultimately holds the view that specific judgements about empirical objects can, in context, be as epistemically certain as judgements about sense data, so that the former should not be epistemically based on the latter (Williams 2004a, 90–2). Williams 2004a, b reaches a similar conclusion but he denies that for Wittgenstein ‘There is an external world’ is a grammatical proposition. For him, it is just plain nonsense. Hence, neither realism nor idealism can use it (or its negation) to make a meaningful statement. Williams 2004a, 80–3 somewhat stresses the same point. Yet he doesn’t really connect it with Wittgenstein’s overall view of meaning as use, nor with his idea that philosophical ‘contexts’ and, specifically, ‘doubts’ are nonsensical precisely because they go against the ordinary use of our words, which is the only use words have for Wittgenstein (see in particular, p. 83: ‘[C]omments on the logic of ordinary doubting and knowledge claiming will cut no ice if we are in the grip of the illusion that there is a special kind of philosophical doubt, purporting to call epistemic ordinary procedures into question’). It is on the basis of such a failure to make that connection that Williams is led to thinking that Wittgenstein did actually purport to provide some kind of counter to sceptical doubts that did not hinge on the development of semantic considerations against the very meaningfulness of sceptical doubts. We will look at Williams’ reading of Wittgenstein in §3.2.2. It should be recalled, for instance, that Descartes’ doubts are raised in thought. Hence, we can see how a Wittgensteinian view of concepts runs contrary to the two tenets of Jerry Fodor’s 1975 original hypothesis of a language of thought. Williams 1991, §4.1, 137 points out the difference between Descartes’ argument from dreaming and Sestus Empiricus. The latter accords real and
Notes
227
independent existence to images and dreams, hence there is no reason to regard the senses as the only sources of knowledge about real things. 17. Cf. Stroud 1984, ch. 1; Wright 1990, 91–2 and fn. 17; Williams 1991 all stress this very point. In Wright 1990, fn. 17, the argument gets formulated thus: I have no way of determining that not: H; If H, then I’m not perceiving; hence I have no way of determining that I am perceiving; hence I am not entitled to any of what I normally regard as perceptually grounded beliefs. Notice that most recent arguments against the dreaming hypothesis tend to claim that we can have non-evidential warrants (and perhaps even knowledge) of the fact that that we aren’t currently dreaming. Cf. Wright 2004a; some trends in Williams 1991 could be developed in such a direction too. 18. For a development of this argument, see Stroud 1984, 20–3. 19. Stroud 1984, Ch. 1 notices how Descartes’ argument depends on the thesis according to which (1) if one doesn’t know whether one is dreaming, one can’t have knowledge of an external world. Hence it doesn’t rely on the more trivial thesis that (2) if one is dreaming, one can’t know anything about the external world. Stroud (1984, Ch. 2) wonders whether (1) is legitimate. He notices, following Austin (1961), that in real life we never require one to be in a position to exclude that one is dreaming, in order to grant one knowledge of an external world. However, Stroud doesn’t agree with Austin that by imposing (1) as a condition over the obtainment of knowledge, a sceptic is thereby changing our concept of knowledge. According to Stroud, the different perspective onto knowledge of an external world that common sense and scepticism command, can be reconciled by noticing that in ordinary life we are only concerned with the legitimacy of making assertions about the external world, not with their truth. That is why (1) holds while doing epistemology and (2) in real-life situations. For the satisfaction of (1) is necessary in order to have real knowledge, as opposed to merely making appropriate knowledge claims. Hence, according to Stroud, an anti-sceptical argument can’t rely on a supposed difference between a sceptic’s notion of knowledge and an everyday one. Nor can it be based on an analysis of how we ordinarily use the expressions ‘I/to know’ and ‘I/to doubt’. For this would simply betray a misunderstanding of the objective or absolute nature of epistemological enquiry. In Chapter 1 (§4) I have already expressed dissatisfaction with respect to such a position. Nevertheless, it remains that Stroud is right to stress that Descartes’ argument relies on (1) and not merely on (2). Yet this, in my view, would have a bearing – if the argument from dreaming succeeded – also on our ‘everyday’ knowledge and knowledge claims. 20. In RPP I, 375 Wittgenstein notices that not necessarily does a child learn how to use the verb ‘to dream’ in the way described so far, viz. ‘by first merely reporting an occurrence on waking up, and our then teaching it the words “I dreamt”. For it is also possible that the child hears the grown-up say he has dreamt and now says the same thing of itself and tells the dream. I am not saying, the child guesses what the grown-up means. Suffice it that he uses the word, and uses it under the circumstances under which we use it’. Hence,
228
21.
22. 23. 24. 25. 26. 27.
28. 29.
30. 31. 32. 33.
34. 35. 36.
37.
Notes ‘the proper question is not: “How does he learn the use of the word?” but rather “How does it come out that he does use it as we do?”’ (RPP I, 376). This qualification is important because there could be unconscious fabrications with respect to the content of one’s dreams. Yet, by their very nature, they couldn’t count as cases of lying. Psychoanalysis would in fact read into one’s dreams a different, deeper and in some sense truer content. Wright 1985 was probably the first to point out the potential generality of Humean scepticism. Dretske 2005. See also Moyal-Sharrock 2005a, Ch. 8. Pritchard 2010, 5–6 calls it ‘incoherent’. And it may be a common view that self-defeating arguments are meaningful yet hopeless – viz. rationally impotent – precisely because they are self-defeating. Pritchard 2010, 10 stresses the same point. This, interestingly, was what happened with one of Wittgenstein’s favoured ‘hinges’, viz. ‘Nobody has ever been on the Moon’. We will have to come back to this issue in the following chapter. We will postpone a treatment of the potentially relativistic implications of such a view until the last chapter. Wright 2004a, b. Cf. also Williams 2004b, 253. Although, ultimately, Moore’s mistake was that of upholding the wrong conception of knowledge as a kind of mental state introspectively accessible to one. But notice that, in my reading of Moore, he was much more of an externalist than Wittgenstein. Notice that Williams does not distinguish the anti-idealist from the antisceptical arguments in On Certainty. Cf. Williams 2004a, 84–6; Williams 2004b, 263–5, 278–9; Conant 1998; Minar 2005; Crary 2005 and Read 2005. I am not entirely convinced of the utility of classifying Wittgenstein’s production into a specific number of phases, as there are continuities as well as changes throughout all his work. The counting will thus depend on what one focuses on. So, if one considered the relationship between norms, use, definitions and judgements, there may be even four Wittgensteins: the Tractarian, the Middle-Phase, the Philosophical Investigations, and the PostInvestigations ones. The first would concentrate on ostensive definitions as ways of determining the meaning of signs and therefore the rules of their employment; the second would hold the primacy of rules, which may or may not be ostensive ones, in determining use; the third would invert this picture; and, finally, the fourth would recognize that also apparently empirical judgements may, in context, play a normative role. Cf. Williams 2004b, 265–76. I do agree with Williams that whenever Wittgenstein refers to hinges by calling them ‘truths’, his use of that word is emphatic. Yet, whenever he seriously addresses the issue of truth in OC, he does give an epistemic twist to it. Only contingent propositions can in fact be true or false because they are the only ones which can be confirmed or disconfirmed by empirical evidence.
Notes
229
38. 39. 40. 41.
Williams 2004b, 249–51, 281. As I have argued in Coliva 2010b. OC 131, 338, (422), 474. For further criticisms of Williams’ reading, somewhat more general and less based on the textual evidence provided by On Certainty, see Pritchard 2010, 17–21. Pritchard (ivi, 24–6), however, seems to endorse much of Williams’ externalism, although he isn’t sympathetic to his contextualism. I have already explained in the main text why I think this would betray Wittgenstein’s views. 42. This is in fact the line that I myself have developed in Coliva 2010c, d, e.
4 Wittgenstein: Hinges, Certainty, World-Picture and Mythology 1. As the editors of On Certainty remark, the text is a collection of notes that Wittgenstein wrote during the last eighteen months of his life. They were written at different times and collected in different notebooks. In the edited text, the passage from one section to the other is indicated by a line: the first one comprises §§1–65; the second §§66–192; the third §§193–299; the fourth; one, finally, §§300–676. Kienzler 2006, 134 argues that in fact John Henry Newman’s Grammar of Assent was more inspirational to Wittgenstein than Moore’s DCS and PEW. He also claims (ivi, fn. 53) that Wittgenstein chose Moore’s examples only because of their conciseness. I beg to disagree: in OC continuous reference is made to Moore’s premises in PEW and to his truisms in DCS. This can hardly be regarded as merely a pragmatic choice. Yet, the relevance of Moore’s work for Wittgenstein’s reflections in OC is entirely compatible with there being other influences on him, such as Newman’s sample of mundane certainties. 2. Perissinotto 1991, 239. 3. Schulte 1988 too insists on this point. 4. I thus agree with Frongia 1983, Ch. 5. 5. Moyal-Sharrock 2005a, Ch. 7. 6. In OC 49, 146, 362, 368 Wittgenstein is indeed suggesting that at bottom of our enquiries lie propositions we have somehow decided not to call into question. However, it is also clear in context that such a decision wasn’t arbitrary but belonged to the logic of the investigation at hand and, consequently, it was constitutive of epistemic rationality. 7. For an illuminating discussion of the role of assumptions/presuppositions in On Certainty, see Hudson 1978. For an analysis of the concept of worldpicture, see the following sections of this chapter. 8. So, contrary to Williams 2004b, I think that throughout On Certainty Wittgenstein is concerned with an anti-realist conception of truth. 9. In OC 152 Wittgenstein uses another image to stress the same point, viz. that of an axis kept fixed just by the movement of a body around it. As we shall see in the following, this is a even less foundationalist image, because there is no sustaining wall that keeps these propositions fixed, contrary to what the image of hinges suggests. 10. Thus, contrary, for instance, to the hermeneutic tradition, Wittgenstein’s analysis remains totally non historical, both because it doesn’t delve
230
11. 12. 13. 14. 15. 16. 17.
18. 19. 20.
21.
22. 23. 24.
25. 26.
27. 28.
29. 30.
Notes into the social and historical factors which may have determined which propositions would belong to the ‘horizon’ within which our investigations and practices of judgement are conducted, and because it doesn’t engage in a philological reconstruction of certain concepts (words). See Chapter 1, §1.2. We will consider the issue at much greater length in the following. Stroll 1994, Ch. 9. Moyal-Sharrock 2005a, 75–80; 172–3. He repeatedly talks of grounds and foundations, as well as of hinges, as we have seen. Schulte 2005 too rightly insists on this crucial point, yet to claim that Wittgenstein was rather a coherentist. If one doesn’t like this way of putting the point one could, rather equivalently to my mind, say that the hinge ‘Nobody has ever been on the Moon’ was abandoned and its empirical doppelganger introduced in our language games. See, in particular, OC 139–52. So he was no ‘foundherentist’ either, Haack 1993, Ch. 1; Lehrer 1990, 398. In Moyal-Sharrock’s suggested reading, in contrast, it would turn out that there are sentences which, at present, can both be used to express empirical propositions and rules; which can do so diachronically; and which, finally, we can’t at present conceive of how they may come to express an empirical proposition. Strawson 1985, Ch. 1. Other supporters of the naturalist reading are Wolgast 1987, Conway 1989, partly shared also by Stroll 1994 and Moyal-Sharrock, although the latter, in particular, proposes an interesting blend between the framework and naturalist readings. For instance those in which he explicitly talks of animal behavior like OC 287, 475. Moyal-Sharrock 2005a, Ch. 3, although she does not consider certainties as propositions. OC 143, 152–3, 159, 283, 286, 310–6, 472–3, 476. Moyal-Sharrock 2005a, 193–9 talks in this connection of ur-trust, whose violation is not mistrust but pathology. Moyal-Sharrock 2005a, Chs. 2–3, 9. Cf. Wolgast 1987, Conway 1989, Stroll 1994, Ch.9. Which is not what Wittgenstein is concerned with, see OC 343, 411. Such an ‘epistemic’ notion of acceptance is what in fact characterizes Cardinal Newman’s conception of assent. It is thus only a misfortune, from my point of view, that Kienzler 2006 takes it to be the common element between Newman’s Grammar of Assent and Wittgenstein’s notion of certainty. He often characterizes our attitude with respect to hinges in terms of acceptance, see OC 196, 344, 399. An ineffabilist reading claims that as a matter of principle hinges and certainty can’t be said. Hence, it is not to be conflated with the much weaker thesis that usually or even mostly they are not said. Armengaud 1983, 2, fn. 4 calls this the ‘paradox’ of On Certainty. As already noted in Chapter 2 (note 6) Moyal-Sharrock is of course right to claim that they cannot be said in the ‘technical’ sense in which Wittgenstein
Notes
31. 32. 33. 34. 35. 36. 37. 38. 39.
40.
41. 42. 43. 44. 45.
46.
47.
48.
49. 50. 51.
231
sometimes uses ‘saying’, viz. as equivalent to ‘stating an empirical proposition’. For if they are norms, they can’t be said in this technical sense. But there is also a much more humdrum notion of ‘saying’ which doesn’t entail that one is thereby describing states of affairs, according to which norms can be said as such. Chapter 2, §3 and Ch. 3, §3.2.2. Crary 2005, 287–9, 296. Other passages on the notion of Weltbild are OC 146–7, 162, 167(–73), 209, 233, 262, (292). OC 162, (163), 196–206, 222, 404, 500. OC 83, 163, 206, 426, 514–5. OC 110, 130, 166. Contrary to what Wright and Williams maintain, cf. Ch. 3 §§3.2.1–2. See OC 336 on diachronic changes. See also OC 162, 167: ‘The propositions describing [our world-picture] are not all equally subject to testing’ and, especially, ‘It is clear that our empirical propositions do not all have the same status, since one can lay down such a proposition and turn it from an empirical proposition into a norm of description’. Surely one of Wittgenstein’s favourite examples, namely, that we have a picture of the earth as a ball (OC 146–7), was originally taken from science (in the pre-Modern sense of this term). Yet, by now, it belongs to common sense. OC 162, quoted in note 39. OC 71–5, 155, 217, (219–23), (254), 257, 281, (323–7), (334), 420. For a discussion of this theme, see Soles 1982 and Perissinotto 1991, Ch. 5, §5. OC 74. OC 74, 322–4. Here I have in mind especially Rorty 1979 and Boghossian 2006. But see also Phillips 1977, Chs. 4–5, Lukes 1982, 281, Hintikka and Hintikka 1986, 21 and Haller 1995. Winch 1964 was inspirational to most relativist interpretations of Wittgenstein. A partially dissenting voice is Marconi 1987, with whom I am much in agreement. Another one is Williams 2004b. Luckhardt 1981 points out how Wittgenstein wasn’t an ethical relativist. This characterization of relativism draws on Boghossian 2006 and Coliva 2009. See also Glock 2007, 378. It is expressed both in actual and hypothetical terms for reasons which will become apparent in the following. These labels can be found in Marconi 1987, 122–4. I disagree with Marconi in considering anti-foundationalism a ‘mild’ form of relativism. Relativism is committed at least to the possibility of there being different and incompatible, though still intelligible systems of justification, whereas antifoundationalism isn’t. Compare also the telling qualifications in OC 92 and 262. ‘Remember that one is sometimes [my emphasis] convinced of the correctness of a view by its simplicity or symmetry’; and ‘We should be trying to give him our picture of the world. This would happen through a kind [my emphasis] of persuasion’. It is notable that in Culture and Value (CV 72) Wittgenstein himself talks of superstition as a ‘false science’; cf. OC 286. Evans-Pritchard 1937, 72. Winch 1964, 38.
232
Notes
52. Of course modern science has importantly revised the notion of efficient cause, since Aristotle’s formulation. We no longer look for whom caused a certain phenomenon, but for a physical law that can explain and predict its occurrence. 53. Evans-Pritchard 1937, 194. 54. von Wright 1978. 55. CV 5–8, 56, 79. 56. Evans-Pritchard 1937, 24. 57. Bloor 1976, 123. 58. Evans-Pritchard 1937, 25. 59. See Boghossian 2006, 70–2, 108–9. 60. See Bloor 1976, 123–30. 61. Boghossian 2006, 108–9. 62. While there are scholars, like Marconi 1987, 128, who wholeheartedly support the view that Wittgenstein was a virtual relativist, others, like Stroud 1965, deny it. The latter develops his position in opposition to Dummett’s 1959 radically conventionalist reading of Wittgenstein’s notes in RFM. Dummett’s reading is now generally in disrepute and, at any rate, could hardly carry over to On Certainty, where there is little room for conventionalism. 63. Glock 2007, 392: ‘Whistling in the dark, as Quine remarks somewhere, is not the proper method in philosophy. Conceptual relativism would be supremely toothless if it amounted to the claim that we cannot exclude the possibility of alternative conceptual schemes which, ex hypothesi, remain forever unrecognisable’. 64. This characterization of conceptual relativism draws on my 2009. It partially overlaps with Glock’s 2007, 381–2. An oft-mentioned example in favour of relativism is Putnam’s case (1990, 96) of the two communities which describe a little world composed of the circles x, y, z either as a world populated by three objects, or by seven (mereological) ones. As Boghossian 2006, 36–7 remarks, however, the two descriptions aren’t really incompatible, for there is no inconsistency between saying that in a world there are three objects or seven, if in the two descriptions the word ‘object’ means ‘individual’ and ‘mereological object’ respectively. (Similar considerations can be advanced against other alleged instances of conceptual relativism presented in Glock 2007, 394). Nor is it enough, in order to rank Wittgenstein among conceptual relativists, simply to point out that he believed that grammar is arbitrary, as opposed to what Glock 2007, 382 maintains. For that by itself doesn’t license the conclusion that different and incompatible conceptual schemes would be intelligible. Taken as such, the idea that grammar is arbitrary merely hints at the view that our language and therefore, in a Wittgensteinian perspective, our concepts are not grounded in a ‘putative essence or form of reality, and that [they] cannot be correct or incorrect in a philosophically relevant way’ (ivi). 65. See also Zettel (Z 350) and Broyles 1974. 66. Marconi 1987, 132–3. 67. See Winch 1964, 33. 68. Even Winch (1964, 30) emphasizes that possible extensions have to be intelligible in light of previous uses. 69. See also Broyles 1974.
Notes
233
70. In PI II, xii, 230 for instance, he writes that if one imagines that certain facts of nature were different from what they are, ‘the formation of concepts different from the usual ones will become intelligible to him‘ (italics mine). I think that here Wittgenstein is merely alluding to the fact that, if certain facts of nature were different from what they are, the possibility of forming different concepts would become intelligible to us. 71. That the point of this exercise is to clarify our own concepts, world-picture and their ungroundedness can be elicited from PI II, xii, 230 OC 513 and CV 74. 72. On this topic, see Schulte 1988, 1992, Ch. 6; Perissinotto 1991, Ch. 5 and Hilmy 1995. 73. On Wittgenstein’s contribution to anthropology, see, in particular, Bouveresse 1975 and Andronico 1997. 74. As Wittgenstein notices (NF 121) it would be foolish to believe that those tribes that perform the rituals of the rain didn’t know that sooner or later it will rain anyway, whether they dance or not. Hence, it would not be an illuminating anthropological explanation to say that they dance because they believe that will make rain. Rather, we can better understand them if we say that they have those rituals because the rain is considered extremely important for the life of their community. 75. Hilmy 1995, 238. 76. Cf. Baker and Hacker 1980/1983, 273. 77. Or ‘said’, in the ‘technical’ sense that this term had especially for the early Wittgenstein where only what could be both true and false could, properly speaking, be said.
Bibliography Aldrich, V. C. 1973, ‘Something really unheard of’, Philosophy and Phenomenological Research 44, 528–38. Ambrose, A. 1942, ‘Moore’s “Proof of an External World”’, in P. Schillp (ed.), 397–417. ———1978, ‘The defence of common sense’, Philosophical Investigations 1, 1–11. Andronico, M. 1997, ‘Giochi linguistici e forme di vita’, in D. Marconi 1997 (ed.), 241–86. Andronico, M., Marconi, D. and Penco, C. 1988 (eds), Capire Wittgenstein, Genova, Marietti. Armengaud, F. 1983, ‘Moore et Wittgenstein: “Je crois que”/“Je sais que”’, in Parret, H. 1983 (ed.), 31–47. ———1986, ‘Moore et le sens commun don Quichotte au secours de Sancho Pança’, Revue Internationale de Philosophie 158, 304–312. Atkins, J. 1992, ‘Common-sense or non-sense’, Philosophical Investigations 15, 346–56. Austin, J. L. 1961, Philosophical Papers, London, Oxford University Press. Ayer, A. J. 1940, The Foundations of Empirical Knowledge, London, Macmillan. ———1971, Russell and Moore. The Analytical Heritage, Cambridge (Mass.)-London, Harvard University Press. ———1974, ‘Wittgenstein on certainty’, in G. Vesey 1974 (ed.), 226–245. Baker, G. P. and Hacker, P. M. S. 1980/1983, Wittgenstein. Meaning and Understanding, Oxford, Blackwell. Baldwin, T. 1992, G. E. Moore, London, Routledge. Bearn, G. C. F. 1993, ‘Wittgenstein and the uncanny’, Soundings 76, 29–58. Bennett, P. W. 1980, ‘Wittgenstein’s theory of knowledge in On Certainty’, Philosophical Investigations 3, 38–46. Black, C. 1974, ‘Taking’, Theoria 40, 66–75. ———1982, ‘Doubts’, Philosophical Investigations 5, 205–14. Bloor, D. 1976/1991, Knowledge and Social Imagery, London, Routledge. Boghossian, P. 2006, Fear of Knowledge, Oxford, Oxford University Press. Bonjour, L. 1993, ‘Externalism/internalism’, in Dancy, J. and Sosa, E. 1993 (eds), 132–6. Bouveresse, J. 1975, ‘Wittgenstein antropologo’, in the Italian translation of NF, 59–92. ———1976/1987, Le mythe de l’intériorité, Paris, Les Éd. du Minuit. Brenner, W. H. 2005, ‘Wittgenstein’s Kantian solution’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 122–41. Brown, F. 1989, ‘Wittgenstein and Searle’s assertion fallacy. A second pass’, Philosophical Investigations 12, 154–61. Broyles, J. E. 1974, ‘An observation on Wittgenstein’s use of fantasy’, Metaphilosophy 5, 291–7. Burnyeat, M. F. 1977, ‘Examples in epistemology: Socrates, Theaetetus and G. E. Moore’, Philosophy 52, 381–98. 234
Bibliography
235
Caraway, C. 1978, ‘Is Wittgenstein’s view of the relationship between certainty and knowledge consistent?’, Philosophical Investigations 1, 16–22. Churchill, J. 1988, ‘Wittgenstein: The certainty of world pictures’, Philosophical Investigations 11, 28–48. Clarke, T. 1972, ‘The legacy of scepticism’, reprinted in M. Williams 1993a (ed.), 287–302. Coady, C. A. J. 2007, ‘Moore’s common sense’, in S. Nuccetelli and G. Seay 2007 (eds), 100–18. Coffa, A. 1991, The Semantic Tradition from Kant to Carnap: to the Vienna Station, Cambridge, Cambridge University Press. Coliva, A. 1997, ‘Moore: scetticismo e senso comune’, Lingua e stile 32, 167–88. ———1999, ‘Wittgenstein: certezza versus scetticismo e naturalismo’, in D. Marconi 1999 (ed.), 169–87. ———2008, ‘The paradox of Moore’s proof’, The Philosophical Quarterly 58/231, 234–43. ———2009, I modi del relativismo, Roma-Bari, Laterza. ———2010a, ‘Moore’s Proof and Martin Davies’ epistemic projects’, Australasian Journal of Philosophy 88/1, 101–16. ———2010b, ‘Was Wittgenstein an epistemic relativist?’, Philosophical Investigations 33/1, 1–23. ———2010c, ‘Liberals and conservatives. Is there a (Wittgensteinian) third way?’, in A. Coliva (ed.), Knowledge, Meaning and Mind: Themes from the Philosophy of Crispin Wright, Oxford, Oxford University Press (forthcoming). ———2010d, ‘Moderatism, transmission failures, closure and scepticism’, MS. ———2010e, ‘Varieties of failure’ (of warrant transmission – what else?!), MS. Coliva, A. and Picardi, E. 2004 (eds), Wittgenstein Today, Padova, Il Poligrafo. Conant, J. 1998, ‘Wittgenstein on meaning and use’, Philosophical Investigations 21, 222–50. Conway, G. D. 1989, Wittgenstein on Foundations, Atlantic Highlands, N.J., Humanities Press. Cook, J. W. 1980, ‘Notes on Wittgenstein’s “On Certainty”’, Philosophical Investigations 3, 15–37. ———1985a, ‘The metaphysics of Wittgenstein’s “On Certainty”’, Philosophical Investigations 8, 81–119. ———1985b, ‘Hanfling on Moore’, Philosophical Investigations 8, 287–94. Coope, C. 1974, ‘Wittgenstein’s theory of knowledge’, in G. Vesey 1974 (ed.), 246–67. Crary, A. 2005, ‘Wittgenstein and ethics: a discussion with reference to On Certainty’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 275–301. Dancy, J., and Sosa, E. 1993 (eds), A Companion to Epistemology, OxfordCambridge (Mass.), Routledge. Davidson, D. 1974, ‘On the very idea of a conceptual scheme’, in Davidson 1984, Inquiries into Truth and Interpretation, Oxford, Oxford University Press. Davis, R. 1978, ‘I may be dreaming now: another dip into the Cartesian well’, Philosophical Investigations 1, 54–8. De Martelaere, P. 1986, ‘Wittgenstein, critique de Moore: de la certitude sans savoir’, Revue Philosophique de Louvain 84, 208–28. Dionigi, R. 1997, La fatica di descrivere. Itinerario di Wittgenstein nel linguaggio della filosofia, Firenze, Vallecchi.
236
Bibliography
Dretske, F. 1970, ‘Epistemic operators’, Journal of Philosophy 67, 1007–23. Dretske, F. 2005, ‘Is knowledge closed under known entailment? – The case against closure’, in M. Steup and E. Sosa 2005 (eds), Contemporary Debates in Epistemology, Oxford Blackwell, 13–26. Dummett, M. 1959, ‘Wittgenstein’s philosophy of mathematics’, Philosophical Review 68, 324–48. ———1993, Origins of Analytical Philosophy, Cambridge (Mass.), Harvard University Press. Dunlop, C. E. M. 1977 (ed.), Philosophical Essays on Dreaming, Ithaca, Cornell University Press. Egidi, R. 1995 (ed.), Wittgenstein: Mind and Language, Dordrecht, Kluwer. Evans-Pritchard, E. E. 1937, Witchcraft, Oracles and Magic Among the Azande, Oxford, Oxford University Press. Fodor, J. 1975, The Language of Thought, Cambridge (Mass.), Harvard University Press. Frongia, G. 1983, Wittgenstein. Regole e Sistema, Milano, Franco Angeli. Gargani, A. G. 1965, ‘Linguaggio e società in G. E. Moore e nell’ultimo Wittgenstein’, Giornale Critico della Filosofia Italiana 19, 98–118. ———1978, ‘Scienza, filosofia e senso comune’, in the Italian translation of OC, vii–xxx. Garver, N. 1984, ‘Neither knowing nor not knowing’, Philosophical Investigations 7, 206–24. ———1994, This Complicated Form of Life, Chicago La Salle, Open Court. Giacconi, A. 1980, ‘Certezza e senso comune negli appunti dell’ultimo Wittgenstein’, Rivista di Filosofia neo-scolastica 2, 274–89. Gill, J. H. 1974a, ‘Saying and showing: radical themes in Wittgenstein’s On Certainty’, Religious Studies 10, 279–90. ———1974b, ‘On reaching bedrock’, Metaphilosophy 5, 277–90. Gillett, G. 1990, ‘An anti-sceptical fugue’, Philosophical Investigations 13, 304–21. Glock, H. J. 2007, ‘Relativism, commensurability and translatability’, Ratio 20, 377–402. Glock, H. J., 2008, ‘Analytic philosophy and history: a mismatch?’, Mind 117/468, 867–97. Goldstein, L. 1988, ‘Wittgenstein’s late views on belief, paradox and contradiction’, Philosophical Investigations 11, 49–73. Graff, E. 1984, ‘Wittgenstein’s “world-picture” in On Certainty’, Kinesis 13, 21–37. Grice, P. 1957, ‘Meaning’, Philosophical Review 66, 377–88. ———1989, Studies in the Way of Words, Cambridge (Mass.)-London, Harvard University Press. Gullvåg, I. 1988, ‘Remarks on Wittgenstein’s Über Gewißheit and a Norvegian Discussion’, Inquiry 31, 371–85. Haack, S. 1993, Evidence and Inquiry: Towards Reconstruction in Epistemology, Oxford, Blackwell. Haller, R. 1988, ‘Justification and praxelogical foundationalism’, Inquiry 31, 335–45. ———1995, ‘Was Wittgenstein a relativist?’, in R. Egidi 1995 (ed.), 223–31. Hanfling, O. 1982, ‘On the meaning and use of “I know”’, Philosophical Investigations 5, 190–204. ———1989, Wittgenstein’s Later Philosophy, New York, MacMillan Press.
Bibliography
237
Harris, N. G. E. 1987, ‘Wittgenstein and Searle’s assertion fallacy’, Philosophical Investigations 10, 134–41. High, D. M. 1981, ‘Wittgenstein on doubting and groundless believing’, Journal of the American Academy of Religion 49, 249–66. Hilmy, S. S. 1995, ‘Wittgenstein on language, mind and mythology’, in Egidi, R. 1995 (ed.), 235–48. Hintikka M. B. and Hintikka J. 1986, Investigating Wittgenstein, Oxford, Blackwell. Hottois, G. 1974, ‘Philosophie du sens commun et dissolution de la philosophie’, Logique et Analyse, 253–75. Hudson, W. D. 1977, ‘Wittgenstein on fundamental propositions’, Southwestern Journal of Philosophy 8, 7–21. ———1978, ‘Language games and presuppositions’, Philosophy 53, 94–9. Hyman, J. 1991 (ed.), Investigating Psychology, London, New York, Routledge. Johannessen, K. S. 1988, ‘The concept of practice in Wittgenstein’s later philosophy’, Inquiry 31, 357–69. Kenny, A. J. P. 1973, Wittgenstein, London, Penguin Press. Kienzler, W. 2006, ‘Wittgenstein and John Henry Newman on certainty’, Grazer Philosophische Studien 71, 117–38. King-Farlow, J. 1980, ‘“Common sense” and “certainty”: Earlier Moore, later Moore, and later Wittgenstein’, Philosophical Investigations 3, 73–85. Klemke, E. D. 1969a (ed.), Studies in the Philosophy of G. E. Moore, UrbanaChicago-London, University of Illinois Press. ———1969b, ‘G.E. Moore’s proof of an external world’, in E. D. Klemke 1969a (ed.), 276–87. Kober, M. 2005, ‘“In the beginning was the deed”: Wittgenstein on knowledge and religion’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 225–50. Kölbel, M. and Weiss, B. 2004 (eds), Wittgenstein’s Lasting Significance, London-New York, Routledge. Kripke, S., 1982, Wittgenstein on Rules and Private Language, Oxford, Blackwell. Lazerowitz, M. 1942, ‘Moore’s paradox’, in P. Schillp (ed.), 371–93. Lehrer, K. 1990, ‘Externalism and epistemology naturalized’, reprinted in E. Sosa and J. Kim 2000 (eds), Epistemology: an Anthology, Oxford, Blackwell. Leonardi, P. 1994, ‘L’analisi delle verità del senso comune’, in C. Penco and G. Sarbia 1996 (eds), Alle radici della filosofia analitica, Genova, Erga. LeRoy Finch, H. 1975, ‘Wittgenstein’s last word: ordinary certainty’, International Philosophical Quarterly 15, 383–95. ———1977, Wittgenstein. The Later Philosophy. An Exposition of the Philosophical Investigations, Atlantic Highlands N.J., Humanities Press. Linville, K. 1991, ‘Dialogue and doubt in Descartes’ “First Meditation”’, Philosophical Investigations 14, 115–30. Lloyd, A. C. 1979, ‘Wittgenstein’s last epistemology’, Philosophical Inquiry 1, 195–204. Luckhardt, G. C. 1978a, ‘Wittgenstein on paradigms and paradigm-cases’, in Wittgenstein und sein Einfluss auf die gegenwärtige Philosophie (Akten des 2 internationalen Wittgenstein-Symposium 1977 in Kirchberg), 379–84. ———1978b, ‘Beyond knowledge: paradigms in Wittgenstein’s later philosophy’, Philosophy and Phenomenological Research 39, 240–52.
238
Bibliography
———1981, ‘Wittgenstein and ethical relativism’, in R. Haller and W. Grassl (eds), Sprache, Logik und Philosophie, Akten des 4.Internationalen Wittgenstein Symposium 1979, Wien, Hölder-Pichler-Tempsky, 316–20. Lukes, S. 1982, ‘Relativism in its place’, in M. Hollis and S. Lukes (eds), Rationality and Relativism, Oxford, Blackwell, 261–305. Lycan, W. G. 2007, ‘Moore’s anti-skeptical strategies’, in S. Nuccetelli and G. Seay 2007 (eds), 84–99. McManus, D. 2004 (ed.), Wittgenstein and Scepticism, London-New York, Routledge. Malcolm, N. 1942, ‘Moore and ordinary language’, in P. Schillp 1942 (ed.), 345–68. ———1949, ‘Defending common sense’, in E. D. Klemke 1969a (ed.), 200–19. ———1958/1984 Ludwig Wittgenstein: A Memoir, Oxford, Oxford University Press. ———1959/1962, Dreaming, London, Routledge and Kegan Press. ———1964, ‘George Edward Moore’, in Knowledge and Certainty, Prentice-Hall, Inc. Englewood Cliffs, 163–83. ———1977, ‘Moore and Wittgenstein on the sense of “I know”’, in Thought and Knowledge, Ithaca, Cornell University Press, 170–98. ———1982, ‘Wittgenstein. The relation of language to instinctive behaviour’, Philosophical Investigations 5, 3–22. ———1986, Nothing is Hidden, Oxford, Blackwell. ———1988, ‘Wittgenstein’s “scepticism” in On Certainty’, Inquiry 31, 277–93. Marconi, D. 1987, L’eredità di Wittgenstein, Roma-Bari, Laterza. ———1997 (ed.), Wittgenstein, Roma-Bari, Laterza. ———1999 (ed.), Naturalismo e naturalizzazione, Vercelli, Edizioni Mercurio. McDowell, J. 1994, Mind and World, Cambridge (Mass.)-London, Harvard University Press. McGinn, M. 1989, Sense and Certainty. A Dissolution of Scepticism, Oxford, Blackwell. Minar, E. 2005, ‘On Wittgenstein’s response to scepticism: the opening of On Certainty’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 253–74. Monk, R. 1990, Ludwig Wittgenstein. The Duty of Genius, London, Jonathan Cape. Morawetz, T. 1974, ‘Wittgenstein and synthetic a priori judgments’, Philosophy 49, 429–34. ———1978, Wittgenstein and Knowledge: The Importance of On Certainty, Atlantic Highlands N.J., Humanities Press. ———1980, ‘Understanding disagreement and conceptual change’, Philosophy and Phenomenological Research 41, 46–63. ———2005, ‘The contexts of knowing’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 165–88. Mounce, H.O. 2005, ‘Wittgenstein and classical realism’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 103–21. Moyal-Sharrock, D. 2005a, Understanding Wittgenstein’s On Certainty, Basingstoke, Palgrave Macmillan. Moyal-Sharrock, D. 2005b, ‘Unravelling certainty’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 76–100. Moyal-Sharrock, D. and Brenner, W. H. 2005 (eds), Readings of Wittgenstein’s ‘On Certainty,’ Basingstoke, Palgrave Macmillan.
Bibliography
239
Nozick, R. 1981, Philosophical Explanations, Cambridge (Mass.), Harvard University Press. Nuccetelli, S. and G. Seay 2007 (eds), Themes from G. E. Moore. New Essays in Epistemology and Ethics, Oxford, Oxford University Press. Parret, H. 1983 (ed.), On Believing. Epistemological and Semiotic Approaches, Berlin-New York, V. de Gruyter. Perissinotto, L. 1991, Logica e immagine del mondo, Milano, Guerini Studio. Phillips, D. 1977, Wittgenstein and Scientific Knowledge, London, Macmillan. Pritchard, D. 2005a, ‘Scepticism, epistemic luck and epistemic Angst’, Australasian Journal of Philosophy, 83/2, 185–205. ———2005b, ‘Wittgenstein’s On Certainty and contemporary anti-scepticism’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 189–224. ———2010, ‘Wittgenstein on scepticism’, in M. McGinn, O. Kuusela 2010 (eds), Oxford Handbook to Wittgenstein, Oxford, Oxford University Press, forthcoming. Pryor, J. 2004, ‘What’s wrong with Moore’s argument’, Philosophical Issues 14, 349–78. Putnam, H. 1990, Realism With a Human Face, Cambridge (Mass.), Harvard University Press. Quine, W. V. O. 1951, Two Dogmas of Empiricism, reprinted in Quine 1953/1961 From a Logical Point of View, Cambridge (Mass.)-London, Harvard University Press, 20–46. Read, R. 2005, ‘“The first shall be the last and the last shall be the first …”: a new reading of On Certainty 501’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 302–21. Rorty, R. 1979, Philosophy and the Mirror of Nature, Princeton, Princeton University Press. Ross, J. J. 1978, ‘Rationality and common sense’, Philosophy 53, 374–81. Rudd, A. 2005, ‘Wittgenstein, global scepticism and the primacy of practice’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 142–62. Schillp, P. A. 1942 (ed.), The Philosophy of G. E. Moore, Evanston-Chicago, Northwestern University Press. Schulte, J. 1987, ‘Moore e Wittgenstein’, in R. Simili 1987 (ed.) L’epistemologia di Cambridge, Bologna, Il Mulino, 289–304. ———1988, ‘World-picture and mythology’, Inquiry 31, 323–34. ———1992, Wittgenstein. An Introduction, Albany, State University of N.Y. Press. ———1993, Experience and Expression. Wittgenstein’s Philosophy of Psychology, Oxford, Clarendon Press. ———2005, ‘Within a system’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 59–75. Shiner, R. A. 1974, ‘Wittgenstein and Heraclitus: Two river-images’, Philosophy 49, 191–7. ———1978, ‘Wittgenstein and the foundations of knowledge’, Proceedings of the Aristotelian Society 78, 103–24. Snowdon, P. 2007, ‘G. E. Moore on sense data and perception’, in S. Nuccetelli and G. Seay (eds), 119–41. Soles, D. 1982, ‘Some ways of going wrong: On mistakes in On Certainty’, Philosophy and Phenomenological Research 42, 555–71. Sosa, E. 1994, ‘Philosophical scepticism and epistemic circularity’, Aristotelian Society Supplementary Volume 68, 263–90.
240
Bibliography
———2007, ‘Moore’s proof’, in S. Nuccetelli and G. Seay 2007 (eds), 49–61. Strawson, P. F. 1985, Scepticism and Naturalism: Some Varieties, London, Methuen. Stroll, A. 1987, ‘Foundationalism and common sense’, Philosophical Investigations 10, 279–98. ———1994, Moore and Wittgenstein on Certainty, New York, Oxford University Press. ———2005, ‘Why On Certainty matters’, in D. Moyal-Sharrock, W. H. Brenner 2005 (eds), 33–46. Stroud, B. 1965, ‘Wittgenstein and logical necessity’, Philosophical Review 68, 504–18. ———1984, The Significance of Philosophical Scepticism, New York, Clarendon Press. ———1994, ‘Scepticism, “externalism”, and the goal of epistemology’, Aristotelian Society: Supplementary Volume 68, 291–307. Travis, C. 1989, The Uses of Sense, Oxford, Oxford University Press. Vesey, G. 1974 (ed.), Understanding Wittgenstein, Ithaca, Cornell University Press. Voltolini, A. 1998, Guida alla lettura delle Ricerche Filosofiche, Roma-Bari, Laterza. White, A. R. 1969, G. E. Moore. A Critical Exposition, Oxford, Blackwell. ———1986, ‘Common sense: Moore and Wittgenstein’, Revue Internationale de Philosophie 158, 313–30. Williams, B. 1974, ‘Wittgenstein and idealism’, in G. Vesey 1974 (ed.), 76–95. Williams, M. 1991, Unnatural Doubts, Cambridge (Mass.), Blackwell. ———1993a (ed.), Scepticism, Aldershot, Dartmouth. ———1993b, ‘The Death of epistemology’, in J. Dancy and E. Sosa 1993 (ed.), 88–91. ———2004a, ‘Wittgenstein’s refutation of idealism’, in D. MacManus 2004 (ed.), 76–96. ———2004b, ‘Wittgenstein, truth and certainty’, in M. Kölbel and B. Weiss 2004 (eds), 249–84. ———2005, ‘Why Wittgenstein isn’t a foundationalist’, in D. Moyal-Sharrock and W. H. Brenner 2005 (eds), 47–58. Williamson, T. 2000, Knowledge and its Limits, Oxford, Oxford University Press. Winch, P. 1964, ‘Understanding a primitive society’, reprinted in Winch 1972, Ethics and Action, London, Routledge, 8–49. ———1988, ‘True or false?’, Inquiry 31, 265–76. Wolgast, E. 1987, ‘Whether certainty is a form of life’, The Philosophical Quarterly 37, 161–5. Wright, C. 1985, ‘Facts and certainty’, Proceedings of the British Academy 71, 429–72. ———1991, ‘Scepticism and dreaming: imploding the demon’, in Williams 1993a (ed.), 139–68. ———2004a, ‘Warrant for nothing (and foundations for free)?’, Aristotelian Society Supplement 78/1, 167–212. ———2004b, ‘Wittgensteinian certainties’, in D. McManus 2004 (ed.), 22–55. (Abridged version, titled ‘Scepticism, certainty, Moore and Wittgenstein’, in A. Coliva and E. Picardi 2004 (eds), 371–95; also in M. Kölbel and W. Weiss 2004 (eds), 228–48). ———2007, ‘The perils of dogmatism’, in S. Nuccetelli and G. Seay (eds), 25–48. Von Wright, G. H. 1972, ‘Wittgenstein on certainty’, in Problems in the Theory of Knowledge, The Hague, 47–60. ———1978, Wittgenstein and His Times, Oxford, Blackwell.
Index acceptance, 11, 52, 174, 176–7, 195–196, 230 action(s), 19, 79, 105, 107, 122, 165, 171, 176, 189, 194 agreement, 11, 80–81, 114, 117, 162, 182, 197, 201, 217, 224, 231 Ambrose, A., 213 Andronico, M., 233 animal(s), 26, 173, 192 Armengaud, F., 150, 218–219, 230 assertion, 33, 40, 58, 60, 62–64, 70–70, 75, 78, 94–96, 113, 117, 121–125, 127–128, 196, 212, 220–221, 227 fallacy, 90–93, 97–98, 209 norm of, 211, 220 assumption(s), 38, 46, 51, 117, 121, 127, 129–130, 139, 161, 164, 200, 214, 225, 229 heavyweight, 132–135, 146–148 attitude(s), 5, 49–50, 79, 95, 105, 125,140, 193–195, 205–206, 224, 230 of trust, 110, 115–116, 173, 225 propositional, 2, 7, 9–12, 53, 150, 156, 160, 167, 171, 174 pro-, 174 Austin, J. L., 227 axes, 168 Ayer, A. J., 31, 32, 213 background, 65, 71, 119, 181, 201 Baker, G. P., 233 bedrock, 96 belief, 1, 9, 11, 20, 21–22, 32, 43, 45–46, 49, 53, 57, 60, 63, 66, 69, 74, 79, 104, 107, 132, 135–137, 172, 174, 180, 189–195, 199, 212, 214, 218–220 false, 76, 197, 204 in the existence of an external world, 19–20, 32 of common sense, 20–21
perceptually basic, 45, 217 true, 31, 144, 188, 221, 227 belief that, 20, 49, 63, 66, 74, 104, 118, 218 bipol(arity), 7, 10, 72, 88, 156, 222 Bloor, D., 232 Boghossian, P., 196, 197, 231–232 boundary, 154, 156, 158, 185–186 bounds of sense, 115 Bouveresse, J., 224, 233 Bradley, F. H., 20, 211 Brenner, W. H., 211, 223 calculation(s), 114, 130, 132, 161, 182, 187 cause(s), 18, 21, 86–87, 112, 117, 140, 165, 187, 193–195, 231 certainty / certainties, 7, 9–11, 14–16, 22–23, 27, 29–33, 37, 39, 57–60, 65, 72, 74–77, 79–80, 88, 90, 99, 109, 111, 115–116, 118–119, 136, 139–141, 150–151, 156, 167–168, 170–177, 207, 210, 212, 214, 221–222, 225, 229–230 animal, 9, 11, 136, 156, 171–172, 175, 176 as an attitude, 11, 173 commonsensical, 70, 100 epistemic vs. non–epistemic, 139, 168, 170 grammatical / logical, 11 indubitable, 67, 210 ineffable, 170, 176 logical, 58, 213 objective, 65, 88, 76–77, 86, 88, 114, 221–222 propositional vs. nonpropositional, 9, 168, 170–173 psychological, 150, 174, 210 subjective, 68, 75–77, 221–222 un-, 71, 110, 224 Clarke, T., 3, 37–42, 54, 107, 121, 215 Coffa, A., 142 241
242
Index
coherence / coherentism, 124, 128, 168, 170, 230 Coliva, A., 121, 216–217, 229, 231 commitment, 52, 212 common sense, 13–22, 24–25, 28–30, 33, 36, 40–41, 55–56, 70–71, 100, 207–208, 210, 212, 227, 231 community, 8–9, 24, 31, 50, 172, 175, 182, 196–201, 205–206, 233 Conant, J., 8, 152, 177, 220–221, 224, 228 concept(s), 3, 29, 38, 41, 52–54, 58–62, 64, 72, 75, 85, 91–92, 96, 99, 104–107, 113, 117, 125–128, 133, 139, 158–159, 173–174, 176, 181–182, 193, 198–201, 212, 215, 218–223, 226–230, 232–233 conceptual scheme(s), 6, 11, 38, 51, 114–115, 118, 197–198, 201–202, 232 context(s), 40–42, 63, 65–67, 73, 75, 80, 86–87, 91–92, 96, 98, 100, 104, 117, 123 ordinary, 12, 40, 117, 209, 215 philosophical, 117, 146, 209 contextual (ism / ist / ization), 37, 40–43, 71, 97–98, 100–2, 104, 123, 215, 218, 229 context of use, 8, 97–98, 101, 123, 156, 170, 184 epistemological, 138, 140, 144, 145 semantic, 42, 100, 209 conversion / persuasion, 188–189, 192, 197, 202 conviction, 51–52, 75–76, 78, 179–180, 192, 219–220 Conway, G. D., 152, 230 Crary, A., 152, 177, 228, 231 deed, 164, 168 Descartes, R., 103, 115–116, 119–120, 124, 130, 171–173, 213, 215, 226 doubt(s) / doubting, 9–12, 14, 19–20, 23, 25, 33–37, 39–40, 45–47, 55, 58–60, 68–73, 75–84, 87–88, 90, 94–96, 99, 103–112, 115, 118–119, 121, 124, 128, 130–134, 136, 140–141, 144–147, 149–151,
155–156, 160–161, 164–165, 167, 171–173, 175–180, 188, 194, 198, 210, 213–214, 220, 223–227 behaviour of, 105 impossibility of, 58 meaningful, 110, 207 methodological, 118 ordinary, 34, 141, 226 philosophical, 35–36, 107, 112, 116, 131 real, 35, 95, 99, 110, 210 sceptical, 8, 30, 34, 45 108–109, 117–118, 121, 136, 146, 171–172, 225–226 dream(ing), 27, 32–33, 38–39, 43, 45–47, 51, 119–129, 146–148, 172, 180, 210, 220, 228 (Descartes) argument from, 103–104, 118–119, 121, 139, 215, 226–227 descriptive use of the verb to -, 125, 129 hypothesis, 45, 121, 129, 143, 172, 213, 227 Dretske, F., 130, 213, 218 Dummett, M., 232 earth, 6–7, 15–16, 21–24, 50, 86, 108–109, 111, 120, 137, 141–142, 144, 149, 153, 158, 160–166, 169, 172, 179–180, 182, 185–186, 188–190, 201, 214, 222, 225 entitlement, 8, 9, 135–136 epistemological realism, 138, 140 error(s) (mistake(s)), 56, 65, 75–76, 88, 187, 222 evidence, 63, 65–66, 69, 72–73, 75, 80, 82–83, 86, 93, 102, 107–109, 111, 131–132, 135–137, 139, 140, 142, 145–146, 148, 150, 153, 160–164, 170, 174, 178–179, 183–187, 190–193, 201 existence, 20–21, 25–26, 38, 44, 48, 79, 104–105, 108–116, 120–121, 130–131, 139, 147, 161–162, 172, 201, 207, 213, 225, 227 of an external world, 17, 19–20, 25, 32, 37, 45, 69, 111–112, 116
Index experiment, 153, 161 expression(s), 2, 16, 19, 25, 31, 33–35, 39, 51–52, 55–56, 58, 60–61, 63, 69, 72–74, 76, 78, 85, 87–88, 90, 93–98, 103–105, 113, 115, 125–129, 149, 151–153, 160, 167, 171, 176, 179, 199, 205–206, 220–221, 224, 227 Evans-Pritchard, E. E., 193–195, 231–232 family resemblance, 62, 193, 198–201 form of life, 79, 90, 136, 150, 162, 165, 171–172, 175, 195, 223 foundation(s), 8, 24, 37, 70, 78, 116, 135, 150, 152, 160, 164–169, 184, 188, 230 foundational(ism / ist), 12, 138, 152, 163, 166–168, 170, 229 anti -, 190, 202–203, 231 epistemic, 8, 11, 210 Frongia, G., 229 Garver, N., 83, 223 Glock, H. J., 211, 231–232 grammar, 10–11, 34, 70–71, 73, 82, 93, 99–100, 114–115, 137, 141, 160, 206, 207, 221, 223, 226, 232 grammatical, 30, 56, 61, 74, 83, 89, 98–99, 115, 141–142, 155, 221 grammatical / logical, 11–12, 61, 88, 164 nature, 99, 115, 141, 226 remark(s), 83, 88–89, 94–95, 100–101, 151 role, 56, 151 statement(s), 30, 89, 94, 114–115, 141 ‘truth’, 118 ground(s) / grounded, 23, 42, 45, 47–48, 63–67, 73–75, 88, 93, 95, 124,138, 163–164, 166–168, 170, 172, 182, 184, 187–188, 190, 193, 209, 215–216, 219–222, 230–232 and knowledge, 42, 47, 57, 60, 62–63, 67, 102, 112, 143 evidential, 129, 184
243
for doubting, 72, 107–109, 111, 118, 131–133, 136 of mistake, 187 perceptual, 121 stronger than, 39, 66, 73–74, 143 groundless(ness), 96, 166–167 Grice, P., 70, 98, 215 Haack, S., 230 Haller, R., 231 hand(s), 1, 6–7, 17–19, 26–29, 31, 34–35, 38–39, 42–43, 45–48, 50, 52–53, 60, 65–70, 72–73, 77, 79–86, 88–91, 96, 98–99, 102, 105– 107–108, 110, 114, 119–120, 129, 135, 137, 139–142, 144, 148–149, 154–155, 157, 166, 169, 171–172, 175, 178, 183, 185–186, 213, 215–217, 219–221, 223, 225 history, 6, 13, 50, 84, 109, 137, 145, 154, 161, 163, 165–166, 182–183, 185, 191, 225 Hilmy, S. S., 233 Hintikka, M. B., 231 Hintikka, J., 231 holism local, 138 Hudson, W. D., 229 Hume, D., 129, 171–173 I know grammatical use of, 12, 74, 76, 78–79, 86–88, 142, 149, 177–178, 223 I am in pain, 58–59, 88, 90–96, 98 this is my hand, 66, 69–70, 72, 79, 85–86, 89, 96, 98–99, 110 nonsensical use of, 63, 71, 74, 87, 94, 96 idealism, 1, 14, 19, 21, 67, 70, 111–112, 114–115, 117, 140, 212, 226 illusion, 226 of doubt, 131, 225 of meaning, 100 of sense, 224 imagination, 57, 202 implicit(ly), 14, 20–21, 51, 61, 74, 83, 87, 97, 99, 140, 170, 175–176, 188–189, 192, 201 (see also tacit)
244
Index
impossibility of mistake(s), 77 indexical(s), 41, 101 induction, 129, 159, 169, 216 ineffability, 151, 220 instinct, 175 instinctive(ly), 9, 88, 95, 167, 171–173, 175–176 instinctual, 171, 174 Johannessen, K. S., 233 judgment(s), 21–22, 45, 71, 82, 86, 96, 99, 112, 115, 117, 121, 123–124, 132–133, 137–139, 147, 149, 178, 180–182, 187, 217, 222, 226 empirical, 30, 82–83, 130, 139–142, 154, 158, 169, 174, 183, 223, 228 normative role of, 80, 82–83, 142 ordinary, 211 perceptual, 132, 134 synthetic a priori, 82, 157 justification(s), 8, 28, 31, 45–46, 60, 65–66, 69, 86, 102–103, 130–131, 133–134, 136–138, 143, 144–147, 168, 170, 184, 188, 218–219, 221 evidential / non-evidential, 8, 139, 140 system of, 103, 109, 188, 190, 192, 231 Kant, I., 20, 25, 82, 89, 142, 211 Kienzler, W., 229–230 know-how (see also practical knowledge), 143 knowledge, 1, 3, 9, 11, 15, 18–25, 28–35, 37–42, 44–54, 56–60, 62–64, 67–68, 70–79, 83–89, 91, 93–97, 104, 107, 110, 119, 121, 124, 129, 131–135, 139–140, 142–146, 149, 151, 165, 167–168, 170, 174, 188, 191–193, 208, 210, 211–213, 215–222, 224, 226–228 ascriptions, 41, 102, 215 claim(s), 100, 226 externalist notion of, 49, 217 internalist notion of, 50, 218
practical, 12, 85, 143, 221 propositional, 143, 221 self-, 4, 93 tripartite conception of, 216, 221 Kripke, S., 4, 211 Lazerowitz, M., 213 language game(s), 10, 24, 56, 59–61, 63–64, 68, 73–74, 76, 78, 80, 83–90, 93–97, 99, 103–106, 108, 111–112, 18, 116–117, 123–125, 127, 131–132, 140–141, 143, 149–150, 153–155, 159–160, 162, 164–171, 173, 186–189, 202, 207, 218–221, 223–224, 230 descriptive, 95–96, 177, 219 epistemic, 86–87, 95, 167, 169, 173, 176–177, 188 normative / non-normative, 89, 219, 224 law(s) logical, 217 of nature, 168 physical, 232 Lehrer, K., 230 logic(al), 11, 16, 58, 61, 75–76, 78, 80, 82, 86, 88, 113, 132, 134, 136, 181, 186, 195, 196, 203, 207, 217, 223 -linguistic, 31, 35 necessities, 55 of our investigation, 132, 137–139, 164, 229 (see also grammatical and method) Luckhardt, G. C., 231 Lukes, S., 231 Lycan, W. G., 214 Malcolm, N., 2, 14, 28–38, 41, 56, 69, 97, 128, 209, 213–214, 216, 223 Marconi, D., 198–200, 231–232 McGinn, M., 215, 221 meaning, 2, 12, 19, 25, 35, 63, 65, 69, 71–72, 80–84, 90, 92–94, 96–98, 100–3, 107, 109–110, 112–115, 117, 122–123, 127, 155, 162, 173, 194, 198, 204, 206, 209, 220–221, 223, 224, 228 analysis of, 16–18
Index appearance of, 123 as/ and use, 8, 90, 117–118, 131, 156–158, 161, 209, 226 conception(s) of, 35–7, 104, 118, 123, 209 constitutive rules, 7, 10 descriptive, 123–124 linguistic, 7, 19 of the word ‘hand’, 73, 77, 79, 82–84, 86, 107–108, 223 ordinary, 16, 19, 41, 65 primitive, 57 meaningful(ness), 19, 23, 35, 37–38, 62–63, 66–67, 69–72, 74, 77, 79, 94, 96–99, 103–104, 112, 115–118, 121, 124–125, 129, 131, 141, 148, 150, 155, 159, 209, 215, 226, 228 meaningless(ness), 8, 10, 70–71, 98–99, 104, 118, 124, 131, 160, 172, 175 memory, 71, 97, 127–128, 166, 179 method(s), 81–82, 136, 139, 150, 160–161, 163, 165, 192 epistemic, 52, 183, 195–197 of our enquiry(is) / verification(s), 132, 146, 179–180, 182–183 of representation, 181, 207 (see also logic and grammatical) methodological necessities (see Williams), 140 restrictions, 108–109, 116 (see also scepticism) Minar, E., 138, 141, 152, 177, 228 mistake(s), 13, 31, 35, 40, 49, 55–56, 58–59, 67, 72, 75, 77–78, 86, 90, 92–93, 96, 99, 117, 133, 138, 149, 160–161, 177–178, 186–187,194, 196–198, 201, 204, 205, 222, 228 Moore, G. E., 1–6, 8, 10, 12–56, 58–60, 65–73, 77–82, 86–89, 93, 96–101, 104, 107–108, 113–114, 121, 135, 138–139, 142, 151–152, 154–157, 159, 165–166, 170, 178–179, 189, 207–208, 221, 223–224, 228–229 truism(s), 1, 6, 14–17, 19–20, 22–24, 29–30, 33, 56, 58–59, 65–66, 70, 72, 77–78, 80, 86–89, 93, 96, 100,
245
143, 149, 151–152, 156, 158–159, 165, 179, 208, 210, 212, 215, 229 A defence of common sense, 1, 13–14, 55, 149, 165 Proof of an external world, 1–3, 13–14, 25, 29, 30, 55, 112, 215 Morawetz, T., 8, 152, 177, 221 Moyal-Sharrock, D., 10–11, 142, 146, 150, 152, 155–156, 166, 168, 173–174, 176–177, 211, 219–223, 225, 228–230 mythology, 95, 141, 181, 203–207 name(s), 4, 19, 57, 63, 85, 91, 109, 142, 144, 165, 169, 186–187, 223 nature, 11, 19, 34, 46, 49, 52, 57, 77, 81, 85, 89, 93, 108, 115, 125, 132–134, 136–138, 141–143, 145, 147–148, 150–151, 154, 167–168, 171–175, 177–179, 192–193, 201–204, 206–209, 216, 224, 226–228, 232–233 necessity, 99, 169 Newman, J. H., 229–230 nonsense (unsinnig), 2, 12, 31–32, 35–36, 55, 61, 66, 87–90, 93–96, 99–100, 103–105, 107, 113–114, 117–118, 120, 124–125, 127, 131, 138, 141, 155,175, 192 194, 209–210, 225–226 nonsensical, 8, 36, 63, 71, 74, 87, 89, 92, 94, 96–98, 103, 108–109, 111, 115–118, 121, 131–132, 134, 136, 146–147, 149, 158, 160, 172, 179, 194, 209–210, 225–226 norm(s), 98–99, 132, 133, 140–141, 157–160, 174–179, 184–185, 206, 212, 222, 228, 230–231 epistemic, 132, 145, 206 linguistic, 36, 223 of evidential significance, 73, 137 of grammar, 137 (see also rule and proposition) Nozick, R., 213 oracle(s), 189, 192, 194
246
Index
Perissinotto, L., 150, 229, 231, 233 Phillips, D. L., 231 Philosophical Investigations, 7, 10, 18, 29, 40, 62, 80, 90, 96, 117, 228 philosophy, 3–5, 12–13, 25, 28, 37, 41, 46, 55–56, 61, 71, 74, 77, 100, 108, 112, 117–118, 137–138, 141, 146, 154, 166–167, 171, 204, 206–207, 232 physical object(s), 6, 13– 14, 17–22, 26–27, 30–32, 38–39, 43, 45–46, 38, 53, 55, 57–58, 68–69, 78, 83, 86, 97, 108–109, 112–115, 117–118, 121, 138, 141, 144, 147, 159, 166, 168, 186, 212, 217, 222, 225–226 planet(s), 108, 112, 153, 186 practice(s), 9–10, 24, 107, 111, 118, 122, 129, 132, 138, 145, 147, 162, 165, 168, 207, 229 cognitive, 136 epistemic, 4, 6, 11–12, 38, 60, 81, 83, 86–87, 89, 100, 114, 116, 133–134, 139–140, 146–147, 157, 167–171, 175, 177, 182, 187, 193, 198, 200, 207–208, 210, 218 linguistic / non-linguistic, 37, 61, 74, 85, 165, 182 of enquiry, 76, 140, 142 religious, 193, 205 pragmatic(s), 37, 52, 70, 94, 98, 132, 140, 145, 164–165, 170, 184, 198, 221, 229 acceptance, 174, 176 presuppositions, 11, 24, 92, 93, 110, 119, 130–132, 134–138, 146–147, 161, 163, 172, 229 primitive, 57, 182, 189, 192–195 Pritchard, D., 152, 177, 193–195, 217–218, 221, 228–229 proof, 1, 25–26, 34, 39, 44, 46–47, 50, 53, 55, 79, 86, 143, 145, 153, 162–165, 171, 179, 213, 215–217, 225 of an external world, 5–6, 14, 25–28, 32–34, 37,39–48, 50–51, 53, 55, 66, 68–69, 73, 77, 79–82, 86, 100, 107–108, 114, 121, 135, 138–139, 142–143, 145, 149,
154–155, 157, 178, 212–213, 215–218, 225 of ‘Here is one hand’, 28, 215 proposition(s), 4–5, 7–9, 11, 13, 16– 24, 32, 36, 42–43, 45–46, 53–59, 61, 67–68, 70–72, 77–80, 82–83, 85–89, 92, 96–97, 102, 107, 113– 114, 131, 133–135, 138–139, 141– 146, 149–151, 153–171, 173–174, 176, 178–190, 206–208, 210–216, 218–219, 221–224, 226–231 descriptive role of / function, 10, 61, 89, 159, 178, 183, 222 empirical, 6, 7, 10, 14, 22, 30–31, 33, 43, 51, 58, 67, 73, 75, 78–80, 88, 102, 113, 117, 133–137, 142,145, 148, 150, 152–161, 167, 169, 174, 177–178, 181–182, 184, 186, 202, 207, 214, 222, 230–231 hinge(s), 6–12, 102, 138, 142–144, 149, 152, 160–161, 164–170, 172–178, 182 182–187, 190, 192, 196–197, 202–203, 207–208, 210, 220, 224, 226, 228–230 methodological, 156 normative role of, 9, 83, 86, 147–148, 161, 184, 208, 210, 222 of logic / logical, 78, 88, 152, 156, 168, 214 Pryor, J., 3, 44–46, 217 Putnam, H., 232 Quine, W. V. O., 232 Read, R., 152, 177, 228 reading of On Certainty epistemic, 8–9, 11–12, 142, 151, 178 framework, 7–12, 138, 152, 226, 230 naturalist, 8–9, 152, 172, 230 therapeutic, 9–10, 141, 152, 177–178 transcendental, 146 realism, 115, 226 epistemic /anti-epistemic, 145, 210 reason(s), 1, 23, 31, 33–34, 39, 44–46, 55, 60, 63, 65, 67, 72–74, 85–87, 91, 93–94, 96, 98, 101, 103, 105–109, 112, 116, 120–122, 131–134, 137, 140, 143–144, 146, 148, 150,
Index 160–161, 163, 172, 179, 184–185, 187–189, 198, 200, 209–210, 213–214, 217, 219–222, 224 reasonable, 33, 47, 139, 162, 191 relativism conceptual, 198, 232 epistemic, 8, 11–12, 145, 152, 188, 196, 210 factual, 190, 202–203 virtual, 190, 197–198, 201–203, 232 reliability of the senses, 69, 82, 103, 130–131, 139 river-bed, 181, 185, 203 Rorty, R., 212, 231 Rudd, A., 223 rule(s), 4, 7–11, 35, 37, 61, 70–71, 73–74, 80–82, 86, 89, 95–97, 99–100, 102, 133, 140–141, 148, 154, 156–161, 165, 167, 169–170, 174, 178–179, 181, 183, 186, 206, 208, 210, 214, 219–220, 222–223, 228, 230 epistemic, 10 -following, 4 of grammar, 7 (see also norm and proposition) Russell, B., 46 say(ing), 35, 62, 85, 120, 171, 175–176, 214, 219–220, 230–231 saying / showing dichotomy, 151, 206–207 scaffolding, 89, 165, 224 scepticism, 1–2, 8, 19, 21, 25, 27–28, 30–33, 37, 39–43, 46, 48–49, 51–55, 67, 70, 98–100, 103–104, 111–113, 116, 130–131, 138,148, 172, 194, 209–211, 215, 217, 224–225, 227–228 Cartesian, 32, 103, 118–119, 130, 146–147 172, 210 external world, 99–100, 118–119 Humean, 32, 42–43, 103, 118, 129–130, 133–134, 146–147, 172, 217 methodological, 115 Schulte, J., 168, 229–230, 233 science(s), 154, 190, 192–195, 205, 231–232
247
empirical, 158–159, 181, 186 history of, 185, 191 scientific, 108, 140, 153, 161, 165–166, 185, 190–191, 193–194, 205, 208 hypotheses, 153–154 investigation(s), 108, 140, 164 pre-, 204 pseudo-, 205 theories, 154, 190 Searle, J., 90–93, 97–98, 100, 209, 224 sense-datum / data, 17–19, 46, 57–58, 97 sense organs, 51, 53, 77, 129–132, 134–135, 137–139 senseless (sinnlos), 29, 36, 78, 87, 89, 122, 224 sentence(s), 5, 16, 37, 63, 65–66, 84, 91–92, 94, 97–98, 101, 125, 149, 154–159, 161, 183–184, 209, 212, 222, 230 Snowdon, P., 211 Sosa, E., 213, 217 speech act(s), 92, 94, 97–99 stand(ing) fast, 6, 73, 82, 139, 178–179 Strawson, P. F., 8, 152, 172, 230 Stroll, A., 152, 166–171, 175, 216, 230 Stroud, B., 3, 37, 39–42, 54, 61, 107, 117, 121, 198, 215, 217, 227, 232 system(s) epistemic, 133, 188, 190, 193, 197, 199, 201, 203, 207–208 of evidence, 163–164, 225 of judgments, 149, 164 of proof, 162–164, 179 tacit, 176, 206–207 (see also implicit) test(s) / testing, 3, 66, 69, 77, 82, 121, 146–147, 158, 164, 178, 180–181, 186, 191, 213 221, 231 the linguistic argument against Cartesian scepticism, 103, 118–119 (see also Cartesian scepticism) the transcendental argument against Humean scepticism, 103, 118–119, 129–130, 133, 135, 137, 139 (see also) Humean scepticism extended, 146–148
248
Index
transcendentalism, 25, 82 Tractatus logico-philosophicus, 10, 87–89, 137, 151, 157, 162–163, 195 Travis, C., 101–102, 215, 221, 224 tribe(s), 126, 187, 233 true, 6–7, 13–17, 19–23, 27, 30–31, 36, 41, 44, 51, 53–54, 56, 58, 69, 81, 83–84 89, 92, 98, 102, 116, 121, 124, 128, 130, 133–134, 137, 139, 144, 147, 150, 156, 159, 163–164, 168–171, 181–183, 187–188, 191, 194, 198, 207–211, 213–214, 218–222, 228, 233 trust, 52, 68, 73, 81–82, 136, 173, 179, 221, 225, 230 attitude of, 110, 115–116, 173, 225 mis-, 173, 230 truth(s), 6, 8, 11, 19, 22–24, 27, 32, 41, 48, 51, 58–59, 70–71, 76–77, 83, 89, 99–101, 120, 127–130, 144, 163–164, 167–168, 182–183, 188, 191–192, 211, 219–220, 223, 227–229 -conditions, 101–102 epistemological, 1, 56 factual, 99 logical, 65 metaphysical, 61, 99, 114 ungrounded(ness), 9, 11, 24, 38, 129, 131, 134, 137, 146–147, 167, 172, 183–184, 187–188, 190, 201–203, 210, 233 unheard-of, 169, 201–202 unreasonable, 47
useful(ness), 36, 40, 49, 77, 88, 155, 174, 204–205 warrant, 8–9, 11, 32, 43–47, 91, 129, 135–137, 139–140, 142–144, 147–148, 209, 216–217, 220, 227 Williams, M., 8, 103, 118, 134, 138–146,148, 152, 167, 177, 218, 225–229, 231 Williamson, T., 219–220 Winch, P., 193, 231–232 Wittgenstein, L., 1–2, 4–14, 18, 22–24, 28–30, 35–38, 40, 51, 54–71, 73–91, 93–105, 107–118, 120–147, 149–160, 162–175, 177–188, 190–198, 200–211, 216, 218–232 Wolgast, E., 150, 152, 176, 223, 230 world-picture (Weltbilt), 162, 166, 169, 179–181, 183–185, 187–188, 190, 197, 201–204, 206–207, 231, 233 von Wright, G. H., 1, 29, 223, 232 Wright, C., 3, 5, 8, 10–11, 29, 42–44, 103, 118, 134–135, 137–148, 139, 142– 143, 146, 148, 152, 177, 216–218, 223, 227–228, 231–232 wrong, 19–20, 36, 44, 59, 64, 67, 75–78, 80, 83–84, 86, 88, 90, 92, 107, 114–115, 117, 132, 135, 139, 149, 151, 152, 158, 160, 180–181, 186, 189, 191, 194, 199, 215, 225, 228 (see also error and mistake)