The Textual Genesis of Wittgenstein's Philosophical Investigations [1 ed.] 9781136179983, 9780415640688

Citation preview

The Textual Genesis of Wittgenstein’s Philosophical Investigations

Sixty years after its first edition, there is an increasing consensus among scholars that the work posthumously published as Philosophical Investigations represents something that is far from a complete picture of Wittgenstein’s second book project. G. H. von Wright’s seminal research on the Nachlass was an important contribution in this direction, showing that the Wittgenstein papers can reveal much more than the source of specific remarks. This book specifically explores Wittgenstein’s Philosophical Investigations from the different angles of its originary conceptions, including the mathematical texts, shedding new light on fundamental questions in twentieth-century philosophy. Leading authorities in the field focus on newly published or hitherto unpublished sources for the interpretation of Wittgenstein’s later work. A Wittgenstein typescript, translated for the first time into English, is included as an addendum. Nuno Venturinha is a senior research fellow in Philosophy and teaches at the New University of Lisbon, Portugal.

Routledge Studies in Twentieth-Century Philosophy

For a full list of titles in this series, please visit www.routledge.com

10 Hilary Putnam Pragmatism and Realism Edited by James Conant and Urszula M. Zeglen 11 Karl Jaspers Politics and Metaphysics Chris Thornhill 12 From Kant to Davidson Philosophy and the Idea of the Transcendental Edited by Jeff Malpas 13 Collingwood and the Metaphysics of Experience A Reinterpretation Giuseppina D’Oro 14 The Logic of Liberal Rights A Study in the Formal Analysis of Legal Discourse Eric Heinze 15 Real Metaphysics Edited by Hallvard Lillehammer and Gonzalo Rodriguez-Pereyra 16 Philosophy After Postmodernism Civilized Values and the Scope of Knowledge Paul Crowther

17 Phenomenology and Imagination in Husserl and Heidegger Brian Elliott 18 Laws in Nature Stephen Mumford 19 Trust and Toleration Richard H. Dees 20 The Metaphysics of Perception Wilfrid Sellars, Perceptual Consciousness and Critical Realism Paul Coates 21 Wittgenstein, Austrian Economics, and the Logic of Action Praxeological Investigations Roderick T. Long 22 Ineffability and Philosophy André Kukla 23 Metaphor and Continental Philosophy From Kant to Derrida Clive Cazeaux 24 Wittgenstein and Levinas Ethical and Religious Thought Bob Plant

25 The Philosophy of Time Time Before Times Roger McLure

31 Philosophy and the Vision of Language Paul M. Livingston

26 The Russellian Origins of Analytical Philosophy Bertrand Russell and the Unity of the Proposition Graham Stevens

32 The Analytic Turn Analysis in Early Analytic Philosophy and Phenomenology Edited by Michael Beaney

27 Analytic Philosophy Without Naturalism Edited by Antonella Corradini, Sergio Galvan and E. Jonathan Lowe 28 Modernism and the Language of Philosophy Anat Matar 29 Wittgenstein and Other Minds Rethinking Subjectivity and Intersubjectivity with Wittgenstein, Levinas, and Husserl Soren Overgaard 30 Russell vs. Meinong The Legacy of “On Denoting” Edited by Nicholas Griffin and Dale Jacquette

33 The Existentialism of Jean-Paul Sartre Jonathan Webber 34 Heidegger and the Romantics The Literary Invention of Meaning Pol Vandevelde 35 Wittgenstein and Heidegger Pathways and Provocations Edited by David Egan, Stephen Reynolds, and Aaron James Wendland 36 The Textual Genesis of Wittgenstein’s Philosophical Investigations Edited by Nuno Venturinha

This page intentionally left blank

The Textual Genesis of Wittgenstein’s Philosophical Investigations Edited by Nuno Venturinha

First published 2013 by Routledge 711 Third Avenue, New York, NY 10017 Simultaneously published in the UK by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2013 Taylor & Francis The right of the editor to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data The textual genesis of Wittgenstein’s philosophical investigations / edited by Nuno Venturinha. pages cm. — (Routledge studies in twentieth century philosophy) Includes bibliographical references and index. 1. Wittgenstein, Ludwig, 1889–1951. Philosophische Untersuchungen. I. Venturinha, Nuno. B3376.W563P532765 2013 192—dc23 2013003825 ISBN: 978-0-415-64068-8 (hbk) ISBN: 978-0-203-08258-4 (ebk)

Contents

List of Illustrations Acknowledgments

ix xi

Introduction: A Composite Work of Art

1

NUNO VENTURINHA

PART I Argumentative Uses 1

Religious Language as Paradigmatic of Language in General: Wittgenstein’s 1933 Lectures

19

GABRIEL CITRON

2

Wittgenstein’s Blue Book: Reading between the Lines

37

JONATHAN SMITH

3

Wittgenstein and His Audience: Esotericist or Evangelist?

52

JAMES C. KLAGGE

4

The Philosophical Investigations and Syncretistic Writing

65

ALOIS PICHLER

5

The Role of the Big Typescript in Wittgenstein’s Later Writings JOACHIM SCHULTE

81

viii

Contents

PART II The Significance of Logic and Mathematics 6 Logic and Ideality: Wittgenstein’s Way beyond Apriorism, Empiricism and Conventionalism in the Philosophy of Logic

93

OSKARI KUUSELA

7 Kant and Wittgenstein: The Matter of Transcendental Arguments

120

P. M. S. HACKER

8 Wittgenstein in the Mid-1930s: Calculi and Language-Games

135

ANDREW LUGG

9 Wittgenstein on Rules in Language and Mathematics

155

SEVERIN SCHROEDER

10 The Form of Proofs: Wittgenstein vs. Principia Mathematica

168

ANDRÉ MAURY

Addendum: A Wittgenstein Typescript

177

EDITED BY NUNO VENTURINHA, WITH AN ENGLISH TRANSLATION BY JAMES M. THOMPSON

Notes Contributors Index

193 217 221

Illustrations

1. Facsimiles of pages 11v–12r from the Thouless Blue Book 2. Facsimiles of pages 32r and 33r from the Thouless Blue Book

44 48

This page intentionally left blank

Acknowledgments

I owe thanks to many colleagues and institutions without whose help and support this book would have never seen the light of day. It falls within the scope of the research project “Wittgenstein’s Philosophical Investigations: Re-Evaluating a Project”, funded by the Portuguese Foundation for Science and Technology (FCT), of which I was the principal investigator. The project was hosted by the Institute for Philosophy of Language (IFL) at the Faculty of Social and Human Sciences (FCSH) of the New University of Lisbon (UNL) in collaboration with the Wittgenstein Archives at the University of Bergen (WAB). I wish to express my gratitude to António Marques, Director of IFL and member of the project, for his continuous interest and support. I am also deeply grateful to Alois Pichler, Director of WAB and also a member of the project, for his permanent encouragement. I would like to extend my appreciation to Alberto Arruda, who has joined the team at the beginning. Many of the chapters included here were originally presented in meetings organized by the project between 2010 and 2012, and I thank the authors for their contribution. The volume includes a Wittgenstein typescript which is published by kind permission of the Master and Fellows of Trinity College Cambridge, holders of the Wittgenstein copyright. I especially wish to thank David McKitterick and Jonathan Smith for their hospitality during a visit to Trinity College in October 2012 for the completion of the manuscript. Passages from Wittgenstein’s Nachlass: The Bergen Electronic Edition are quoted by permission of Oxford University Press. Last but not least, I am most grateful to Vanessa Boutefeu for her invaluable help. Nuno Venturinha

This page intentionally left blank

Introduction A Composite Work of Art Nuno Venturinha

The title of this Introduction comes from an idea advanced by Denis Paul in his lifelong study of Wittgenstein, according to which “the entire 1929–1951 corpus was a composite work of art”.1 To say this is too strong a claim. The later Wittgenstein is essentially known as the author of several publications edited by his trustees and, notwithstanding the criticism of some editorial decisions, none of these publications would seem to fall under the category of an artistic work. However, Paul does not ignore the fact that there are many items in this part of the Nachlass that do not hang together, nor does he look at Wittgenstein as an artist. His suggestion is that behind the various philosophical texts left by Wittgenstein is a literary project which occupied him for more than twenty years. Paul is keen to emphasize that “Wittgenstein did not plan this composite work of art in advance”, but that “[i]t just grew”.2 Wittgenstein wrote indeed about a multitude of subjects and in many formats, from diary-like entries to extensively revised statements. Even when his remarks went through complicated stages of revision, plenty of them were eventually abandoned and others retrieved. It is from these breakthroughs and setbacks that the Philosophical Investigations emerge as the nearest outcome of Wittgenstein’s longing to write a second book after the Tractatus Logico-Philosophicus. In order to understand the Investigations along the lines suggested by Paul, it is absolutely essential to bear in mind that Wittgenstein’s alleged chef-d’œuvre is a posthumous publication and that only what we have as Part I can legitimately be seen as the result of a maturation. But one also needs to differentiate the Investigations from other posthumous publications of key philosophical works. Kant’s Opus postumum, Nietzsche’s Will to Power and Heidegger’s Contributions to Philosophy, to mention only a few, were each edited after the authors’ death. The editorial problems they pose are, however, of a very different nature. In Kant’s case, the question lies in establishing a coherent text out of the drafts he made at the end of his life, the original sequence of which has been lost. What should be the accomplishment of the system is thus an unfinished piece, with fundamentally an instructive role. In fact, as Eckart Förster writes, “[u]nlike his published

2

Nuno Venturinha

works, which only present the reader with the polished end product of his labors, the Opus postumum therefore shows Kant at work over a number of years, providing us with a unique insight into the genesis of a major text”.3 The fate of Nietzsche’s Will to Power is well known and the state of the text bears similarities with Kant’s. What we really have is a large amount of material intended for a final book, which its author was not able to complete in his lifetime. Giorgio Colli and Mazzino Montinari’s edition of the Nachgelassene Fragmente has returned the Will to Power to Nietzsche’s notebooks, where it belongs, as a collection of sketches.4 This “non-book” was interestingly discussed by Babette Babich in connection with Heidegger’s Contributions to Philosophy. According to Babich, “Heidegger deliberately composed the Beiträge zur Philosophie in order to keep it an unpublished work throughout his life” and this “on the model of the publishing phenomenon that could be called Nietzsche’s ‘Contributions to Philosophy’, had Nietzsche only written his own Will to Power, as, of course, he did not”.5 Babich’s view is that “[t]he Beiträge would then work to limit Heidegger’s vulnerability to editorial manipulation inasmuch as it could be claimed as Heidegger’s authoritative legacy”.6 This is a thought-provoking interpretation of the aim of Heidegger’s main work after Being and Time, but for our purposes here what must be pointed out is that his Contributions to Philosophy constitutes merely a problematic book in regard to the order of its divisions, the posthumous arrangement of which Babich criticizes. Clearly this issue means a lot to Babich since “[s]hifted from its original position as the section following the prelude, Seyn, now makes up the eighth section as the final division of the text”, and “[t]he shift radically changes the disposition of the work”.7 Yet what makes Wittgenstein’s Investigations philologically tricky is not so much the unfinished character of the text,8 for it is undoubtedly more complete than Kant’s or Nietzsche’s texts, nor even its organization, like in Heidegger’s Contributions. Actually, apart from the troubling inclusion of Part II, it can be said that Part I of the Investigations “is in order as it is”, to employ an expression that appears in §98.9 If there is a posthumous book in the history of philosophy truly comparable with Wittgenstein’s Investigations, this book can only be Pascal’s Thoughts. In an influential paper, Georg Henrik von Wright writes that “[b]etween Wittgenstein and Pascal there is a trenchant parallelism which deserves closer study”.10 This is discussed by Maurice O’Conner Drury, another former student of Wittgenstein’s, in an equally influential paper. He avers that “[c]ertainly Pascal’s intensity, his seriousness, his rigorism, these find a parallel in Wittgenstein” and goes on to stress the “concern” and “unrest” that characterize their writings.11 But Drury also identifies some differences between these two authors. The most important has to do with seeing Wittgenstein’s Investigations and Pascal’s Thoughts as “a haphazard arrangement of remarks and aphorisms”. In Drury’s opinion, this can be true in Pascal’s case, who would certainly have arranged his book in a different manner had he lived longer, but not in Wittgenstein’s case. He says that “[w]e know that Wittgenstein was constantly rearranging the material found in his book, that

Introduction

3

he spent a lot of time and thought in obtaining the precise order we now have” in such a way that “[t]o grasp the significance of the Investigations it is essential to see the order of development of the thoughts”.12 These claims must be handled with care. First, there are reasons to believe that the fragmentation of Pascal’s Thoughts corresponds to the very nature of the subject of the book itself and that therefore there could never be an ordered string to be established. Second, there is some evidence that the Investigations represents a consciously unfinished work, with other items in the Nachlass offering different arrangements of much the same material. The Thoughts and the Investigations cannot indeed be compared with other posthumous texts. Both projects constitute singular attempts in the western tradition to radically conciliate philosophy with writing. Let us take a closer comparative look. Since the first appearance of the book in 1670, eight years after Pascal had passed away, several doubts have been raised regarding the correct order of the fragments that form the Thoughts. Pascal’s executors were well aware of the efforts he made during the last years of his life to conceive an apology of Christianity. Yet what was left of this project was far from being something that could even be regarded as an unfinished enterprise out of which a definite structure could be found. Pascal’s annotations were in such a chaotic state that it seemed impossible to extract from them a book where the order is crucial. It was this order that the first editors tried to determine, putting together the more apologetic papers in thematic series and correcting them as needed. The following editions have added other fragments making the Thoughts more real but at the same time more formless since the themes blend into each other, be they autobiographical, philosophical or concerning the preparation of other works. The puzzle faced by the editors of Pascal nonetheless kept the author much more alive through his corrections, cancellations, additions, etc., than if there were a final copy, purified of all these indecisions—a result and not the process. As a matter of fact, this was a privileged situation given that at that time an author’s papers were not seen as possible objects of study of their literary practices, with manuscripts being invariably lost after publication of the works. However, it would only be in the twentieth century that, through the editions of Zacharie Tourneur, Louis Lafuma and, almost definitively, Michel Le Guern,13 readers would be able to approach the Thoughts as close as to how Pascal left them. In effect, these editors have all followed the so-called “First Copy”, prepared by Pascal’s family shortly after his death from the arrangements he himself had made of many fragments into labelled chapters, corresponding to about half of those forming the most recent edition of the Thoughts. Other textual sources are the “Recueil original”, which contains the autograph fragments collected only some decades after Pascal had died, and the “Second Copy”, a copy of the first, which even presents important differences as regards these “papers of a deadman”, to use Le Guern’s phrase.14 Interestingly, Le Guern’s view somehow coincides with Drury’s. Were we not dealing with “the papers of a deadman”, they would surely present a coherent and linear form, with their fragmentary, sometimes contradictory,

4

Nuno Venturinha

argumentation vanishing. This is certainly the general idea one has of the Thoughts and that is why so many editorial initiatives have tried to organize the text in accordance with a certain logic, which could overcome the disconnected, inclusively arbitrary, image that it purports. But was this what Pascal had in mind? There is a fragment belonging to the series “That the Law was Figurative” that illuminates the procedure to be followed in the work. It reads: Contradiction. It is not possible to give a good physiognomy save by bringing all contraries into harmony, and it is not enough to dwell upon a series of accordant qualities, without reconciling the contraries. To understand the sense of an author we must harmonize all the contrary passages. Thus, to understand Scripture, we must find a sense in which all the contrary passages are reconciled; it is not enough to have one which agrees with many consonant passages, but we must find one which reconciles even dissonant passages. Every author has a sense in which all the contradictory passages agree, or he has no sense at all. The latter cannot be said of Scripture and the prophets, which assuredly abound in good sense. We must then seek for a sense which harmonizes all contraries.15 The nexus established by Pascal between what he calls “the sense of an author” and the Bible makes evident what he himself aims to carry out in his book: to convey a vision resulting from the contradictory, of what is incompatibly expressed because it ought to be so, but whose synthesis of contradictoriness produces an agreement and not a disagreement. The obscurity of the Old and New Testaments must also be a feature of Pascal’s apology, with the interpretative success depending on an understanding not of what is uttered in each passage but of the authorial sense in which it falls. Hence there is a twofold dimension of reading: one concerning the text itself and another lying in the understanding of the author and not merely of what he articulates. This notion of a sense that must be indirectly communicated undermines conventional literary production, where the interpretation depends solely on the grasping of the contents and not of the form through which the contents are put forward. In Pascal’s Thoughts, it is the fragmentary, the disorder that correspond to the real expression of what is at stake, with this book being one that its author recognizes as impossible to establish—since it is impossible to establish by means of meaningful language and in a well-organized sequence the thoughts we have about this domain. Thus he affirms: Prin. I shall here write my thoughts without order, yet not perhaps in undesigned confusion, that is true order, which will always denote my object by its very disorder.

Introduction

5

I should do too much honour to my subject if I treated it with order, because I wish to show that it is incapable of it.16 Pascal’s allusion to Descartes’ Principles of Philosophy cannot consequently be taken as if the overcoming of rationalist dogmatism were simply at issue. One of Pascal’s goals is definitely to overcome it, but we will not understand the Thoughts if we do not recognize that what they aim at can never find completeness, an ending. That is where the fragmentation comes from as a result of human finitude. Pascal’s Thoughts are truly a kind of diary, a set of reflections which makes sense first and foremost to the author. But what about the Philosophical Investigations? In their revised edition of the posthumous publication, Peter Hacker and Joachim Schulte comprehensively debate the origin of the text.17 The two editors start by emphasizing that Wittgenstein’s earliest project to write a philosophical book after publishing the Tractatus has materialized in the Big Typescript (TS 21318), composed in 1933 on the basis of the studies he began when he returned to Cambridge and took up philosophy again in 1929. This typescript was revised at length in the following years but never reached the definition Wittgenstein wanted to achieve. The second project mentioned by Hacker and Schulte derives from the so-called Brown Book, dictated in the academic year 1934–35. Wittgenstein revised and translated the text into German in 1936, already calling it Philosophische Untersuchungen, but again this initiative was not considered good enough.19 What has been ignored is that the Brown Book itself seems to have been regarded by Wittgenstein as worthy of publication. Ray Monk claims that “[t]here is no indication that Wittgenstein considered publishing the Brown Book”.20 Yet the draft of a letter from G. E. Moore to the British Academy and its reply dated 18 February 1935, as well as a letter from J. M. Keynes to Moore of 6 March 1935, show that Wittgenstein envisaged publishing a hundred-page book in the Proceedings of the Academy.21 Although susceptible to the proposal, the suggestion made by the Academy to Wittgenstein to seek publication with Cambridge University Press probably concurred with his preparation of the manuscript containing approximately the first 188 sections and the beginning of §189 of the Investigations we know (MS 142). This, as Hacker and Schulte accurately note, would be finished in 1937, giving rise to a new typescript (TS 220) still in the same year. The typescript, together with a synopsis of his later logical and mathematical writings (TS 221) and a preface (TS 225), would be offered to the publisher in 1938 with the working title Philosophical Remarks.22 It is important to mention that the German text was to appear with an English translation, with the Wittgenstein papers preserving not only a partial translation by Rush Rhees of the first half of the book in TS 226 but also a translation of the preface prepared by Wittgenstein himself with the help of Theodore Redpath.23 Hacker and Schulte then refer in their study that “[s]ometime between late 1939 and 1943, Wittgenstein revised the Early Draft [of the

6

Nuno Venturinha

Investigations]” and that “[o]ne of the typescripts of TS 220 was extensively revised by hand (TS 239)”.24 They go on to say that “TS 221 was reworked, cut up and rearranged” and that “[t]he subsequently dictated typescript, TS 222, has been printed as Part I of the Remarks on the Foundations of Mathematics”.25 To be more exact, what “was reworked, cut up and rearranged” was one of the three copies of TS 221, not having been a “subsequently dictated typescript”.26 But the most important thing is that, as Hacker and Schulte highlight, the work on the philosophy of logic and mathematics had not been put aside when “Wittgenstein again approached the Syndics of the Press in September 1943, proposing publication of a book with the title Philosophical Investigations, to be printed together with a new impression of the Tractatus”.27 It is a matter of speculation why the publication, after it had been accepted by the publisher, was once more aborted by Wittgenstein. The fact is that he did not publish the book until the end of his life in 1951 and what came to be the Philosophical Investigations was something quite different. This work, as edited by Anscombe and Rhees in 1953, is made out of a much extended typescript of the original first half (TS 227) and of another one summing up later work chiefly addressing the philosophy of psychology (TS 234). Conscious of the arbitrariness of this decision, Hacker and Schulte have titled the second half “Philosophy of Psychology—A Fragment”, but even so they keep the text as an integral part of Wittgenstein’s opus magnum. The legitimacy of the second half of the Investigations has given rise to a good deal of discussion28 and it is not this particular point that motivates the confrontation with Drury’s reading, to which we shall now return. As mentioned above, Drury accentuates Wittgenstein’s painful work of revision in order to arrive at “the precise order we now have”. Given that the reorganizations of the materials included in the second half are incomparably fewer in number and substance to those that gave rise to the establishment of the first half, it is likely that Drury’s observation primarily regards what Hacker and Schulte deem to be the Investigations as such, that is, the first half. As they remark: Whatever Wittgenstein’s final intentions were, the fact is that the closest he ever came to completing the Philosophical Investigations is the current text consisting of §§1–693. It is, we believe, this text that should be known as Wittgenstein’s Philosophical Investigations.29 The assertion made by Drury is absolutely sound if we see work on philosophy as consisting of structured lines of reasoning which should lead to a resolution of clearly defined problems. This “precise order” is what commentators of the Investigations, from Garth Hallett30 to Gordon Baker and Peter Hacker,31 not forgetting Eike von Savigny,32 have attempted to explain. But the Wittgenstein papers can leave us in a quandary as to this “precise order”. In their historical-philological presentation, Hacker and

Introduction

7

Schulte bring up for consideration another typescript, labelled by Wittgenstein “Remarks I” (TS 228), “which consists of 698 numbered remarks, some 400 of which he then incorporated into the final draft of the Investigations”, this TS 227 being “probably dictated in the course of the academic year 1945–6”.33 The difficulty is that, apart from having prepared “Remarks I”, Wittgenstein also produced a typescript labelled “Remarks II” (TS 230). This constitutes 542 numbered remarks, the majority of which come from TS 228. According to von Wright, these “Bemerkungen II” must have been prepared “[s]ometime after the completion of Bemerkungen I, but before the commencement of the writings for Part II of the Investigations”.34 Von Wright wonders why Wittgenstein prepared this typescript, rejecting the hypothesis of “Remarks II” being the second half of the book at this time—given the repetition of the material in relation to TS 227—but recognizing that “[w]ith some justification Bemerkungen II may be regarded as an independent and final work by Wittgenstein”.35 In this regard, a paper by Paul, also posthumously published, sheds new light on the matter. He reports an episode in which, showing him “different copies” of TSS 228 and 230 “housed in two identical box-files”, Anscombe said “Wittgenstein wanted them to show how philosophical ideas could lead on to each other in different orders”.36 It is not without relevance that the very last section in “Remarks II”, corresponding to §140 of “Remarks I” and to a fragment placed at the end of §133 of the Investigations, states that “[t]here is not a single philosophical method, though there are indeed methods, different therapies, as it were”.37 In The Concept of Irony, Kierkegaard spoke about the “Socratic possibility of beginning anywhere, actualized in life”.38 Much the same takes place in the later Wittgenstein, for whom philosophy should be able to begin at any point, each one manifesting its whole context.39 As in Plato’s initial dialogues, if we accept Schleiermacher’s distinction between “dialogical” and “constructive dialogues”, as Kierkegaard did,40 the outcome could hardly be positive.41 The intriguing dialogical structure of many of Wittgenstein’s remarks, to mention only the most evident of his argumentative techniques,42 specifically aims to point towards the illusion of arriving at definite and encompassing answers by following a specific path. What Wittgenstein’s “irony” does is to expose the negative of approaching the same fields by different routes, which never come to an end. It principally teaches a mode of philosophizing that is at odds with one-sided conclusions. The 1945 preface to the published Investigations, which forms part of TS 227, exposes in a very clear way the reasons for their failure—and here lies the paradox of an eventual publication. Wittgenstein says that “it seemed to [him] essential that in the book the thoughts should proceed from one subject to another in a natural, smooth sequence” but that “[he] realized that [he] should never succeed” and that “[t]he best that [he] could write would never be more than philosophical remarks”. He goes on to say: “my thoughts soon grew feeble if I tried to force them along a single track against their natural inclination”.43

8

Nuno Venturinha

There is a very similar remark of Wittgenstein’s in code dating from 15 September 1937, when work on the prewar version of the Investigations was under way. It runs as follows: If I am thinking just for myself without wanting to write a book, I jump about all round the topic; that is the only way of thinking that is natural to me. Forcing my thoughts into an ordered sequence is a torment for me. Should I even attempt it now?? . . .44 A first conclusion to draw from these passages is that Wittgenstein’s intent had always been the publication of a book that should not proceed artificially or dogmatically but have instead “a natural, smooth sequence”. This is extraordinarily difficult to conceive and impossible to realize, as Wittgenstein came to recognize, for when we try to isolate this or that aspect, our understanding proceeds analytically, i.e., through a specific focus, abstracting from the synthetic acquaintanceship with reality we normally have. What Wittgenstein looked for was exactly to grasp the world, making visible what we presuppose when we analyze such and such. The great difficulty lies, therefore, in analyzing what all the time escapes us, what language, or thought, inexplicitly projects under the form of a comprehensive synthesis. That is the reason why the organization of the original remarks can promote a fertile recontextualization but at the same time an inevitable decontextualization. It thus becomes evident that the projected book did not fail simply because Wittgenstein was not able to establish it, because the “natural inclination” of his thoughts was dispersive; it was bound to fail in virtue of what it should portray: reality itself as linguistic horizon. That this is so becomes even clearer in a parenthetical note of 6 January 1949, in which Wittgenstein declares: To piece together /Piecing together/ the landscape of these conceptual relationships out of their individual fragments /out of their individual fragments, as language reveals them to us,/ is too difficult for me. I can make only a very imperfect job of it.45 The idea of giving “an idea of the landscape”, like a “draughtsman”, is already in the preface to the Investigations46 and echoes earlier remarks, for instance when Wittgenstein states that “a thinker is very similar to a draughtsman”, since the latter “wants /would like/ to represent all the interconnections”.47 He also identifies himself with “a painter” in a note penned on 17 March 1949, adding that he is “often a very bad painter”.48 All this helps us to understand Paul’s notion of “a composite work of art” when applied to the Philosophical Investigations. One way of putting it is to say that the remaining documents belonging to the projected Investigations form this “composite work of art”. But we can go further. We usually think of a work of art as being total, not composite. That is precisely how we look at a

Introduction

9

painting. The multiple details that form the work, which become evident as soon as we look closer, only make sense in virtue of its unity. This is analogous to what happens, for example, when we listen to a symphony or read a novel. Our image of the work is constructed along with the perception of new elements until it turns out to be a complete image. Even when the parts of the work are not entirely surveyable, as is the case with a work of architecture, we know that all the parts are there and that they are necessary for the whole. What is important is that there is an idea—the idea preconceived by the artist—to be grasped, all components being subordinated to this idea. The trouble with a composite work of art is that it lacks this defined idea. It is merely an accretion of related pieces according to a general conception. The real problem of the Investigations is exactly their never having achieved a clear definition. We can now see that Drury’s judgment also misses the point in regard to the Investigations. What was posthumously published under that title does not really correspond to the book Wittgenstein had in mind, one that can only be guessed at from the various drafts he made for it. What is at stake is a way of philosophizing quasi-coincident with our way of thinking, and that is something that cannot but result in, as Gérard Guest so well dubbed it, an “impossible book”.49 The Textual Genesis of Wittgenstein’s Philosophical Investigations is divided into two parts. The first, “Argumentative Uses”, helps us understand how Wittgenstein put into practice the conviction he had held since at least the beginning of 1931 “that [his] book [was] meant for only a small circle of people”,50 something that he forcefully reaffirms in the last paragraphs of the 1945 preface to the Investigations.51 In his “Religious Language as Paradigmatic of Language in General: Wittgenstein’s 1933 Lectures”, Gabriel Citron looks into Moore’s unedited notes of Wittgenstein’s first courses in Cambridge, particularly the notes focusing on religious questions, and shows how this material is important for grasping the sort of discourse he wanted to communicate after beginning to lecture in 1930. What is more, Citron finds in these notes the embryos of key remarks in the Investigations, namely §§371 and 373, where it is said that “[e]ssence is expressed in grammar” and that “[g]rammar tells what kind of object anything is”, with Wittgenstein adding parenthetically “Theology as grammar”. The two appear in consecutive notes written after May 1945,52 and the first is also found in a list of chapter titles prepared at around the same time.53 If this should be no. 125, no. 121 reads: “Causality in reality and in the image of reality. (Theology)”.54 There is also a remark of this period, published in Zettel, that goes in the same vein: “How words are understood is not told by words alone. (Theology.)”55 Equally interesting is the fact that the equivalent to §373 of the Investigations in “Remarks II” is preceded by this illuminating observation, which closes Zettel, testifying to the continuous use of this analogy: “You can’t hear God speak to someone else, you can hear him only if you are being addressed”.—That is a grammatical remark.56

10

Nuno Venturinha

Jonathan Smith’s chapter, “Wittgenstein’s Blue Book: Reading between the Lines”, analyzes in great depth the sense in which this first dictation for students is also one of the “preliminary studies for the Philosophical Investigations”.57 Composed at the same time as the Big Typescript, the Blue Book consists of more than a simple set of notes for the use of students after the sessions. As Smith plainly demonstrates through the study of an earlier version of the text given by Wittgenstein to Robert Thouless, the Blue Book possesses all the characteristics of a developed work, even if intended for only a restricted audience. Its “genetic dossier”, to employ the terminology of Genetic Criticism,58 which Smith broadly uses, manifests Wittgenstein’s laborious efforts to arrange the text in ways that make it an indispensable forerunner to the Investigations. Smith’s discussion of the parallels between a passage belonging to the issue of “our craving for generality” in the Blue Book59 and §73 of the Investigations is paradigmatic, with the German text originally appearing in the midst of Wittgenstein’s first “Umarbeitung” of the Big Typescript.60 It can be argued that the pedagogic linearity of the Blue Book was to some extent responsible for the breakdown of the Big Typescript itself, with Wittgenstein then moving on to the preparation of the Brown Book with actual publication in sight, which should constitute the culmination of his fellowship.61 The third chapter of this first part, “Wittgenstein and His Audience: Esotericist or Evangelist?”, approaches the textual genesis of the Investigations from a wider perspective. James C. Klagge extends the interpretation of what it is like to write for a selected group of readers, seeing Wittgenstein’s literary activity evolving from an esoteric attitude to an exoteric, evangelical one, and from this back to esotericism again. Klagge’s account of similar procedures in the history of philosophy, notably those of Plato and Aristotle, who wrote for both private and public assemblies, contributes to a better view of Wittgenstein’s aims. One of the main points made by Klagge is that during the 1930s, “[i]n stark contrast to Aristotle, who would only lecture to students whose habits were already trained to respond to reason, Wittgenstein found himself lecturing to students whose habits of thought were resistant to his ways of thinking” and so “[h]e would no longer proceed esoterically”. It is noteworthy that almost all of Wittgenstein’s writings from the academic year 1935–36 are in English, with a clear intention of exposing the topics he was addressing in class in the most accessible way. It is also significant that in the following summer it was the Brown Book that he took up again to prepare his first German work with the title Philosophische Untersuchungen. In fact, Klagge meticulously identifies numerous passages in what would come to be the Investigations in which Wittgenstein specifically constructs his discourse in order to change mentalities not used to his way of philosophizing. That is the meaning of the evangelism Klagge ascribes to him, a stance that can actually be extended to spheres not directly covered by the Investigations, including that of religion. Wittgenstein’s recognition that his exotericism was bound to fail is, according to Klagge, the result of

Introduction

11

disappointment with his teaching. The refuge he sought after resigning his professorship in 1947 made the book return to its esoteric origin, with Wittgenstein realizing that what he could send to the world would never be a finished work but, as Klagge brilliantly put it, “more likely something else”. Alois Pichler’s “The Philosophical Investigations and Syncretistic Writing” continues the analysis of Wittgenstein’s argumentative strategies in his conception of the book. Pichler focuses on Hanspeter Ortner’s classification of writing strategies and claims that the Investigations have both a syncretistic origin and form. This is explained by Pichler through various considerations on textual gestalt, in particular how the work is really to be interpreted as “an album”, as added to the 1945 preface.62 What is addressed in this chapter is a decisive question for Wittgenstein interpretation: should one take Wittgenstein’s literary production as the result of a “puzzlewriting”, as Ortner himself takes it, or should it be understood purposely as an open enterprise? Pichler thoroughly scrutinizes the multifarious aspects of this openness and its importance for the philosophical text in general. His provocative association of the “poetic” to the “syncretistic” was acknowledged by Wittgenstein more than once, with some of his remarks going as far as to imply that poetry would be the only right way of presentation in philosophy—something that reminds us not only of Nietzsche, whom he explicitly mentions, but also of the later Heidegger.63 In “The Role of the Big Typescript in Wittgenstein’s Later Writings”, Joachim Schulte concentrates on what a textual “revision” (Umarbeitung)64 means for Wittgenstein, with this being decisive to understand the Investigations. The way Schulte describes how the Big Typescript was continuously revisited, with a special emphasis on its employment in MS 116 and TS 228 (the “Remarks I”), is illustrative of the extremely sophisticated processes used by Wittgenstein to conceive his book, which go far beyond the simple polishing of previous drafts. Schulte’s insightful analyses of the recontextualization that the remarks ultimately selected for the Investigations assume, with each of them playing different argumentative roles in their different genetic places, conclude the first part of this volume. As Brian McGuinness brightly put it in discussing “the indeterminate literary form of [Wittgenstein’s] writings after the Tractatus”: “Nothing has to be where it is, so anything can be moved to a better place”.65 The second part of the collection, entitled “The Significance of Logic and Mathematics”, explores the central place occupied by Wittgenstein’s logical and mathematical work in the genesis of his new book. It is customary to find a separation in Wittgenstein commentary between his work in the Investigations and his work in the Remarks on the Foundations of Mathematics. The fact is that, as previously mentioned, Part I of the latter, including the appendices, belonged in one way or another to his plans for publication with Cambridge University Press in the 1930s and 1940s. In the first chapter, “Logic and Ideality: Wittgenstein’s Way beyond Apriorism, Empiricism and Conventionalism in the Philosophy of Logic”, Oskari Kuusela journeys

12

Nuno Venturinha

through various manuscript volumes and notebooks from the 1930s in which Wittgenstein originally distances himself from the Tractarian philosophy of logic. Kuusela’s mapping of the germination of Wittgenstein’s new logical views allows one to take him seriously when he mentions in the various versions of the preface to the Investigations that the book is also about “logic”,66 something that is obscured not only by the specificity of his novel account of the discipline but also because, as already said, the Remarks on the Foundations of Mathematics alone seem to cover that field of inquiry. The connections traced by Kuusela make us see that the disciplinary divisions often made by scholars between philosophy of language, of logic, of mathematics, of psychology, and so on and so forth, do not really apply in the case of Wittgenstein, who was committed to a single program: that of a “natural history of human beings”.67 P. M. S. Hacker’s chapter, “Kant and Wittgenstein: The Matter of Transcendental Arguments”, widens the examination carried out by Kuusela by viewing Wittgenstein’s mature philosophy as a dismantling of Kantianism. Hacker surveys the available textual evidence, namely in manuscripts and typescripts from the 1930s, laying emphasis on a remark written down in the so-called Koder diary, only discovered some years ago. Here, in 1931, Wittgenstein talks about apriorism in “cultural” terms68 and, like Kuusela, Hacker highlights the relevance for the Investigations of removing the “nimbus” surrounding the concept of a priori, as sketched out in 1937.69 The result is a real logic, as “an interwoven network of rules”, which is better captured by the term “grammar”. This, in Hacker’s view, finds its most vivid expression in the Investigations when Wittgenstein designs series of arguments that speak for themselves rather than positively affirming how things are in reality, displaying, for example, the incoherence of scepticism. The final sentence of §303 alluded to by Hacker (“Just try—in a real case—to doubt someone else’s fear or pain!”) provides a good picture of how rhetorical and stylistic devices are used by Wittgenstein to portray something that our logic rules out. In his chapter “Wittgenstein in the mid-1930s: Calculi and Languagegames”, Andrew Lugg exposes in detail the important relationship between the Philosophical Investigations and Part I of the Remarks on the Foundations of Mathematics, concentrating on the core manuscripts that led to both of Wittgenstein’s proposals to Cambridge University Press. In so doing, Lugg evaluates the alleged repudiation by Wittgenstein soon after resuming philosophy of a “calculus model of language” in favour of the “language-game model” as found in the prewar and later versions of the Investigations. Providing an almost line-by-line commentary on MS 142 and MSS 117–19, Lugg makes the bridge between several remarks that are seen in a completely different light when viewed in their original relationship. It is symptomatic that two dozen remarks from MSS 117 and 119 were ultimately selected for Part I of the Investigations,70 half of them appearing in TS 221 and ten in TS 222, with these consequently overlapping in Part I of the Remarks.71 They include §§189–90 of the Investigations, the exact

Introduction

13

continuation of TS 220, as well as the following §§191–7, together with §415. Not appearing in the Remarks are §§214–16, 309, 317, 359, the series consisting of §§466–70, §518 and finally §§607–8. There is also a later manuscript, MS 124, used for the compilation of the Remarks that made an important contribution to Part I of the Investigations.72 Severin Schroeder’s chapter, “Wittgenstein on Rules in Language and Mathematics”, takes up many of the same issues dealt with by Lugg but approaches them from different angles. Schroeder makes use of additional documents, such as MS 156a and the later MSS 165 and 179, which complement the views presented in the earlier versions of the Investigations. It is Schroeder’s conviction that there are more differences than similarities between the materials overlapping in the Investigations and the Remarks on the Foundations of Mathematics due to the “directions” they pursue. “The most striking difference is this”, avers Schroeder: “Mathematics is rule-governed, in the full sense of the word, whereas language is not”. This is a consequence of Schroeder’s alignment with the reading, challenged by Lugg, that “the calculus conception” gave its place to the “language-game conception”. The final chapter, “The Form of Proofs: Wittgenstein vs. Principia Mathematica”, considers an interesting discrepancy in a quotation from Russell and Whitehead’s Principia between the first two editions and the third edition of the Remarks on the Foundations of Mathematics. As André Maury reveals, only the latter reproduces verbatim the text of Wittgenstein’s typescript. The interpretative implications of what is at stake, the rule-following in mathematics and ordinary life, are tied together with observations on other texts and even unpublished editorial correspondence between Rhees and von Wright. Maury also gives a living account of what it meant for him to collaborate in von Wright’s research when work on the so-called “Helsinki Edition” of the Investigations, the basis for Schulte’s Kritischgenetische Edition, was under way. The volume closes with an “Addendum”, entitled “A Wittgenstein Typescript”, which consists of an edition of a short text made up of extracts and cuttings from a copy of TS 221, the second half of the prewar Investigations. The document, which contains handwritten corrections by Wittgenstein, was never given a von Wright number and is to be found at Trinity College Library, Cambridge. This edition includes a historical and philological introduction of mine as well as an English translation prepared by James M. Thompson.73

REFERENCES Babich, Babette. 2007. “Heidegger’s Will to Power”. Journal of the British Society for Phenomenology 38, no. 1: 37–60. Baker, G. P. 2004. “Italics in Wittgenstein”. In Wittgenstein’s Method: Neglected Aspects, edited by Katherine Morris, 224–59. Oxford: Blackwell Publishing.

14

Nuno Venturinha

Baker, G. P., and P. M. S. Hacker. 2005a. Wittgenstein: Understanding and Meaning. Vol. 1 of An Analytical Commentary on the Philosophical Investigations—Exegesis §§1–184. Second edition. Edited by P. M. S. Hacker. Oxford: Blackwell Publishing. ———. 2005b. Wittgenstein: Understanding and Meaning. Vol. 1 of An Analytical Commentary on the Philosophical Investigations—Essays. Second edition. Edited by P. M. S. Hacker. Oxford: Blackwell Publishing. ———. 2009. Wittgenstein: Rules, Grammar and Necessity. Vol. 2 of An Analytical Commentary on the Philosophical Investigations—Essays and Exegesis of §§185–242. Second edition. Edited by P. M. S. Hacker. Oxford: Wiley-Blackwell. Biggs, Michael, and Alois Pichler. 1993. Wittgenstein: Two Source Catalogues and a Bibliography. Catalogues of the Published Texts and of the Published Diagrams, each Related to Its Sources. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen. Cann, Kathleen. 1995. “The Papers of GEORGE EDWARD MOORE (1873–1958) Cambridge University Library Add. MSS 8330 and 8875”. Wittgenstein Studies 1/95. http://sammelpunkt.philo.at:8080/444/1/18–1-95.TXT. Conant, James. 2011. “Wittgenstein’s Methods”. In The Oxford Handbook of Wittgenstein, edited by Oskari Kuusela and Marie McGinn, 620–45. Oxford: Oxford University Press. de Biasi, Pierre-Marc. 2004. “Toward a Science of Literature: Manuscript Analysis and the Genesis of the Work”. In Genetic Criticism: Texts and Avant-textes, edited by Jed Deppman, Daniel Ferrer and Michael Groden, 36–66. Philadelphia: University of Pennsylvania Press. Drury, Maurice O’Connor. 1996. “Some Notes on Conversations with Wittgenstein”. In The Danger of Words and Writings on Wittgenstein, edited by David Berman, Michael Fitzgerald and John Hayes, 75–96. Bristol: Thoemmes Press. Förster, Eckart. 1993. “Introduction”. In Immanuel Kant, Opus postumum, edited by Eckart Förster, translated by Eckart Förster and Michael Rosen, xv–lv. Cambridge: Cambridge University Press. Guest, Gérard. 2003. Wittgenstein et la question du livre: Une phénoménologie de l’extreme. Paris: Presses Universitaires de France. Hacker, P. M. S. 1990. Wittgenstein: Meaning and Mind. Vol. 3 of An Analytical Commentary on the Philosophical Investigations. Oxford: Basil Blackwell. ———. 1996. Wittgenstein: Mind and Will. Vol. 4 of An Analytical Commentary on the Philosophical Investigations. Oxford: Basil Blackwell. Hacker, P. M. S., and Joachim Schulte. 2009. “The Text of the Philosophische Untersuchungen”. In Ludwig Wittgenstein, Philosophische Untersuchungen / Philosophical Investigations, edited by P. M. S. Hacker and Joachim Schulte, translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte, xiii–xxiii. Oxford: Wiley-Blackwell. Hallett, Garth. 1977. A Companion to Wittgenstein’s “Philosophical Investigations”. Ithaca: Cornell University Press. Kierkegaard, Søren. 1992. The Concept of Irony with Continual Reference to Socrates. In Kierkegaard’s Writings, vol. II, second printing, edited and translated by Howard V. Hong and Edna H. Hong, 1–329. Princeton: Princeton University Press. Le Guern, Michel. 2000a. “Notice”. In Blaise Pascal, Œuvres complètes, vol. II, edited by Michel Le Guern, 1296–304. Paris: Gallimard. ———. 2000b. “Note sur le texte”. In Blaise Pascal, Œuvres complètes, vol. II, edited by Michel Le Guern, 1204–13. Paris: Gallimard. Maury, André. 1981. “Die Manuskriptquellen der Bemerkungen im Teil I der PU: Ein Verzeichnis (auf Englisch)”. In Ludwig Wittgenstein, Philosophische Untersuchungen I (TS 227), edited by G. H. von Wright and Heikki Nyman, “Anhang I”. Unpublished.

Introduction

15

———. 1994. “Sources of the Remarks in Wittgenstein’s Philosophical Investigations”. Synthese 98: 349–78. McGuinness, Brian. 2002. “Manuscripts and Works in the 1930s”. In Approaches to Wittgenstein: Collected Papers, 270–86. London: Routledge. ———, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. Monk, Ray. 1990. Ludwig Wittgenstein: The Duty of Genius. London: Jonathan Cape. Montinari, Mazzino. 1999. “Vorwort”. In Friedrich Nietzsche, Kommentar zu Band 1–13: Kritische Studienausgabe, vol. 14, new edition, edited by Giorgio Colli and Mazzino Montinari, 7–17. Berlin: Walter de Gruyter. Pascal, Blaise. 1885. The Thoughts of Blaise Pascal. Edited by M. Auguste Molinier. Translated by C. Kegan Paul. London: Kegan Paul, Trench & Co. ———. 1977. Pensées. Edited by Michel Le Guern. Paris: Gallimard. ———. 2000. Pensées. In Œuvres complètes, vol. II, edited by Michel Le Guern, 541–1082 and 1296–609. Paris: Gallimard. Paul, Denis. 2007. Wittgenstein’s Progress 1929–1951. Edited by Alois Pichler. Bergen: Publications from the Wittgenstein Archives at the University of Bergen. ———. 2009. “Wittgenstein Passages”. In Wittgenstein’s Enduring Arguments, edited by Edoardo Zamuner and D. K. Levy, 283–92. London: Routledge. Savigny, Eike von. 1994. Wittgensteins “Philosophische Untersuchungen”: Ein Kommentar für Leser. Band I: Abschnitte 1 bis 315. Second edition. Frankfurt am Main: Vittorio Klostermann. ———. 1996. Wittgensteins “Philosophische Untersuchungen”: Ein Kommentar für Leser. Band II: Abschnitte 316 bis 693. Second edition. Frankfurt am Main: Vittorio Klostermann. Somavilla, Ilse, ed. 2012. Begegnungen mit Wittgenstein: Ludwig Hänsels Tagebücher 1918/1919 und 1921/1922. Innsbruck: Haymon. Venturinha, Nuno. 2010a. “A Re-Evaluation of the Philosophical Investigations”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 143–56. Basingstoke: Palgrave Macmillan. ———, ed. 2010b. “Wittgenstein’s 1938 Preface”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 182–8. Basingstoke: Palgrave Macmillan. ———. 2011. “Introdução”. In Linguagem e Valor: Entre o Tractatus e as Investigações, edited by Nuno Venturinha, 9–16. Lisbon: IFL. von Wright, G. H. 1982a. “Ludwig Wittgenstein: A Biographical Sketch”. In Wittgenstein, 13–35. Oxford: Basil Blackwell. ———. 1982b. “The Origin and Composition of the Philosophical Investigations”. In Wittgenstein, 111–36. Oxford: Basil Blackwell. ———. 1993. “The Wittgenstein Papers”. In Ludwig Wittgenstein: Philosophical Occasions 1912–1951, edited by James C. Klagge and Alfred Nordmann, 480– 506. Indianapolis: Hackett. ———. 1997. “Wittgenstein Materials kept with the Department of Philosophy, Helsinki University”. Unpublished. http://www.helsinki.fi/wwa/Wittgenstein%20 Materials.pdf. von Wright, G. H., Rush Rhees and G. E. M. Anscombe. 1978. “Editors’ Preface to the Revised Edition”. In Ludwig Wittgenstein, Remarks on the Foundations of Mathematics, third edition, edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe, translated by G. E. M. Anscombe, 29–33. Oxford: Basil Blackwell. Wittgenstein, Ludwig. 1969. Preliminary Studies for the “Philosophical Investigations”: Generally known as The Blue and Brown Books. Second edition. Edited by Rush Rhees. Oxford: Basil Blackwell. ———. 1977. Remarks on Colour. Edited by G. E. M. Anscombe. Translated by Linda L. McAlister and Margarete Schättle. Oxford: Basil Blackwell.

16

Nuno Venturinha

———. 1981. Zettel. Second edition. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. ———. 1984. “Eine Philosophische Betrachtung”. Edited by Rush Rhees. In Werkausgabe, vol. 5, 117–237. Frankfurt am Main: Suhrkamp. ———. 1998. Culture and Value: A Selection from the Posthumous Remains. Edited by G. H. von Wright in collaboration with Heikki Nyman. Revised edition of the text by Alois Pichler. Translated by Peter Winch. Oxford: Blackwell. ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt am Main: Suhrkamp. ———. 2003. Movements of Thought: Diaries 1930–1932, 1936–1937. Edited by Ilse Somavilla. Translated by Alfred Nordmann. In Ludwig Wittgenstein: Public and Private Occasions, edited by James C. Klagge and Alfred Nordmann, 3–255. Lanham: Rowman & Littlefield. ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell.

Part I

Argumentative Uses

This page intentionally left blank

1

Religious Language as Paradigmatic of Language in General Wittgenstein’s 1933 Lectures Gabriel Citron

1. INTRODUCTION Sometime between 1929 and 1932, Maurice O’Conner Drury began attending G. E. Moore’s lectures at Cambridge. Drury later recalled that “[a]t the commencement of his first lecture Moore read out from the University Calendar the subjects that his professorship required him to lecture on; the last of these was ‘the philosophy of religion’. Moore went on to say that he would be talking about all the previous subjects except this last, concerning which he had nothing to say”.1 Drury was unhappy with this, and when he saw Wittgenstein he expressed his dissatisfaction: “I told Wittgenstein that I thought a professor of philosophy had no right to keep silent concerning such an important subject”.2 Wittgenstein agreed and reassured Drury, saying: “I won’t refuse to talk to you about God or religion”.3 In fact, more than simply not refusing to talk about God or religion to his students, in the May Term of 1933 Wittgenstein announced in one of his lectures: “I have always wanted to say something about [the] grammar of ethical expressions, or e.g. of ‘God’” (MWL, 8, 74). And he spent the rest of the lecture discussing God, the soul, belief in idols, the afterlife, and prayer. Ironically, the best notes that we have of this lecture were taken by G. E. Moore. 2. G. E. MOORE’S NOTES OF WITTGENSTEIN’S LECTURES, 1930–334 Wittgenstein gave his first lecture in Cambridge in January 1930,5 about a year after his return to the university; and he continued to lecture, with a few interruptions, until the end of the academic year in 1947. Moore— who had been elected Professor of Philosophy five years earlier—attended most of Wittgenstein’s lectures from that January 1930 until the end of the academic year in 1933. And he took copious notes, as he later recalled: “At the lectures . . . I took what I think were very full notes, scribbled in note-books of which I have six volumes nearly full” (PO, 49–50). These six

20

Gabriel Citron

notebooks—which now reside in the G. E. Moore Collection of Cambridge University Library—provide us with both the most comprehensive and the most accurate record that we have of those first three seminal years of Wittgenstein’s lecturing. I will briefly say a word about each of these qualities—comprehensiveness and accuracy—in turn. The relative comprehensiveness of Moore’s notes can be seen by a simple word count. Covering the period of Moore’s notes we also have the published notes of John King and Desmond Lee for 1930– 32,6 and the published notes of Alice Ambrose for 1932–33.7 Together these amount to approximately 50,000 words. At 80,000 words, however, Moore’s notes are longer than King’s, Lee’s and Ambrose’s by more than half again.8 This ratio is even more exaggerated when it comes to Wittgenstein’s remarks on religion from 1933. It is not just that Moore’s notes are more detailed, but rather, they contain entire discussions which are missing from King, Lee and Ambrose. The greater comprehensiveness of Moore’s notes itself makes them more accurate records of Wittgenstein’s lectures than the other existing notes. But there are further reasons for taking them to be more accurate. The first major difference between Moore’s notebooks and the published notes by King, Lee and Ambrose is that all the latter have undergone later editorial neatening-up and even sometimes extensive rearrangement, whereas with Moore’s notes we have an unedited and unaltered record of what Wittgenstein said, in exactly the order in which, and in much the way that, Wittgenstein actually said it. As Moore himself says, “I tried to get down in my notes the actual words he used” (PO, 50). However, there are reasons to believe that Moore’s notes are more accurate than even the original, unedited versions of King’s, Lee’s and Ambrose’s notes9—for Moore probably had a better understanding of what Wittgenstein was doing in his lectures and was therefore in a better position to record it accurately. First of all, Moore was a professor of philosophy at the time, whereas King, Lee and Ambrose were all at one stage or another of their student careers. Moreover, it seems that the lectures were, in a sense, directed at Moore. Karl Britton, one of the students who attended the lectures, said: “We felt that Wittgenstein addressed himself chiefly to Moore, although Moore seldom intervened and often seemed to be very disapproving . . . we had the impression that a kind of dialogue was going on between Moore and Wittgenstein, even when Moore was least obviously being ‘brought in’”.10 Wittgenstein himself wrote to one of his correspondents in 1932 that he was glad that Moore was attending his discussion classes, given the poor quality of his other students: “My audience is rather poor—not in quantity but in quality. I’m sure they don’t get anything from it and this rather worries me. Moore is still coming to my classes which is a comfort”.11 And most significantly, Moore later reported Wittgenstein to have told him that “he was glad I was taking notes, since, if anything were to happen to him, they would contain some record of the results of

Religious Language as Paradigmatic of Language in General

21

his thinking” (PO, 50). Thus, Moore was the attendee of the lectures who was best placed to understand what Wittgenstein was saying; Wittgenstein recognized this fact, directed his lectures at Moore, and considered Moore’s notes to be a “record of the results of his thinking”. In general, Wittgenstein’s attitude towards the status of his lectures seems to have been somewhat conflicted. On the one hand, as we have just seen, Wittgenstein viewed Moore’s notes of his early lectures as an authoritative record of his thought. In keeping with this positive attitude towards his lectures, Norman Malcolm and others report Wittgenstein to have said, on a number of occasions, “that he had always regarded his lectures as a form of publication”.12 Casimir Lewy recalled that “Wittgenstein once said to me that ‘to publish’ means ‘to make public’, and . . . therefore lecturing is a form of publication”.13 On the other hand, however, in his letter to Mind—apropos R. B. Braithwaite’s 1933 article purporting to summarize Wittgenstein’s recent thought14—Wittgenstein wrote: “I . . . have not published any of my work . . . had I published . . . in print I should not trouble you with this letter” (PO, 156–7). Wittgenstein thus identified “publishing” with “publishing in print”, thereby discounting his lectures as true publications. Furthermore, Wittgenstein once stopped a student from taking notes in his lectures, saying: If you write these spontaneous remarks down, some day someone may publish them as my considered opinions. I don’t want that done. For I am talking now as freely as my ideas come but all this will need a lot more thought and better expression.15 It is therefore unclear to what extent Wittgenstein considered notes of his lectures to officially represent his thought—perhaps it depended on the particular lectures, or lecture series. However, even if Wittgenstein did not consider his lectures to be quite on a par with print publications, and even if he thought that they did not fully represent his considered opinions, they are nonetheless often extremely important progenitors of that text which was considered to the highest degree, and which was intended by him for publication—namely, the Philosophical Investigations. In this essay I look at Wittgenstein’s remarks on religion from his 1933 lectures as progenitors of some of the key ideas of the Philosophical Investigations. For there are indications that Wittgenstein’s 1933 grammatical investigations into religious language may have acted as a seedbed for the development of ideas which he later came to see as applying with much more general scope to both religious and nonreligious uses of language. Religious language was able to play this seminal role because it is, in many ways, paradigmatic of language more generally—by which I mean that religious language has certain characteristics which are shared by many other realms of language, but which are more pronounced or more clearly seen in the religious context than in others. I illustrate the way in which religious language is paradigmatic of language in general by means of two examples—both of

22

Gabriel Citron

which emerge from Wittgenstein’s 1933 remarks on religion, and both of which touch on core Wittgensteinian insights. In section 3 I discuss Wittgenstein’s idea that grammar expresses essence, and in sections 4 and 5 I discuss Wittgenstein’s notion of the “messiness” of grammar. In both instances I argue that the case of religious language highlights these Wittgensteinian insights particularly clearly—thus possibly explaining why Wittgenstein paid so much attention to religious language in the lectures he was giving while he was developing these themes in his later thought. 3. ESSENCE, GRAMMAR, THEOLOGY AND PI §373 Near the middle of the Philosophical Investigations is a short remark which, according to Peter Hacker, “crystallises a leitmotif of W[ittgenstein]’s later philosophy”.16 Wittgenstein writes: “Essence is expressed in grammar” (PI, §371). Two remarks later, Wittgenstein expands on this terse pronouncement by repeating the point in slightly different words: “Grammar tells what kind of object anything is” (PI, §373). The point is that a thing’s essential properties are not the kind of properties which it could have failed to instantiate—rather, they are properties which it would not make sense to deny of the thing in question. Since the grammar of a given word is the collection of rules governing its meaningful use, it follows that an overview of the grammar of a given word informs us what it would and would not make sense to say of the referent of that word. The grammar of a given word thereby expresses the essence of its referent. Thus: essence is expressed in grammar. Wittgenstein summarized the point in a 1936 lecture by saying that the essential “nature of . . . objects . . . is not determined by properties which we can attribute to them truly as opposed to those which we can’t. It is determined by the grammar of the word which denotes it” (PO, 307). This is a fundamental insight of Wittgenstein’s later philosophizing. What is interesting about this, for our purposes, is that Wittgenstein appended to the claim that “Grammar tells what kind of object anything is” (PI, §373), a cryptic bracketed note: “(Theology as grammar.)” Without help from elsewhere in Wittgenstein’s corpus, this bracketed remark would almost certainly have remained impossibly obscure. Fortunately, however, Wittgenstein’s remarks on religion from his 1933 lectures come to our aid. Moore’s notes read as follows:17 Luther said: “Theology is Grammar of word of God”. / This might mean: An investigation of idea of God is a grammatical one. / Now (a) suppose “god” means something like a human being; then “he has two arms” & “he has four arms” are not grammatical propositions. but (b) suppose someone says: You can’t talk of god having arms, this is grammatical. / Austrian general: “I will remember you after death, if this is possible”. / That so & so is ridiculous (as this is), or blasphemous, shows grammar. (MWL, 8, 74–8)

Religious Language as Paradigmatic of Language in General

23

The claim that Wittgenstein attributes to Luther provides the key to understanding his bracketed remark—“(Theology as grammar)”—in PI §373. Reports differ, however, as to precisely what claim Wittgenstein actually attributed to Luther. Moore’s original notes, just quoted, record Wittgenstein to have claimed that “Luther said: ‘Theology is Grammar of word of God’”; whereas Alice Ambrose’s published notes have Wittgenstein claiming that “Luther said that theology is the grammar of the word ‘God’” (AWL, 32). There are good reasons—over and above the general ones mentioned in the previous section—for thinking that Moore’s notes are the more accurate in this case.18 Nonetheless, even Moore seems to have been in two minds as to whether his original notes recorded what Wittgenstein actually meant— for though Moore originally wrote “Luther said: ‘Theology is Grammar of word of God’”, he later put a circle around the second “of” in the sentence and marked it with a question mark.19 What may explain the conflicting evidence is the possibility that Wittgenstein himself vacillated between the two claims, which are not actually as different as they may seem at first glance. For—given the role that God plays in scripture—a grammar of the word of God would, in the end, necessarily also be a grammar of the word “God”. This is borne out by looking further into the very example that Wittgenstein gives of a theological-grammatical statement, namely: “You can’t talk of god having arms”. At the close of the fourth century, a controversy erupted in the Christian world over just this matter: whether or not God had a corporeal form. Cassian reports that the archbishop of Alexandria, in his festal letter of 399 CE, wrote, “a long refutation of the absurd heresy of the Anthropomorphites”.20 The archbishop’s claim—as a deacon called Photinus elaborated at one public reading of the letter—was that God’s “immeasurable, incomprehensible, invisible majesty cannot be limited by a human frame or likeness. His nature is incorporeal, uncompounded, simple, and cannot be seen by human eyes nor conceived adequately by a human mind”.21 The anti-anthropomorphic letter proved explosive: “Nearly all the monks in Egypt . . . received this with bitterness and hostility: and . . . decreed that the bishop was guilty of a grave and hateful heresy, because (by denying that Almighty God was formed in the fashion of a man, when Scripture bears clear witness that Adam was created in his image) he seemed to be attacking the text of Holy Scripture”.22 Thus the debate was equally and simultaneously about the correct grammar of the word “God”, and about the correct grammar of those parts of the scriptural word of God which state that man was created in the image and likeness of God.23 Thus perhaps it would be best for us to speak of Wittgenstein’s attribution to Luther of the claim that “[t]heology is [the] grammar of [the] word (of) God”—with brackets around the second “of”. Wittgenstein fleshed out the thought that “[t]heology is [the] grammar of [the] word (of) God” by saying that “[a]n investigation of [the] idea of God is a grammatical one”. This latter claim is simply the application to the limited field of religion of the general principle we have just seen in PI §373, that “Grammar tells what kind of object anything is”. As this general principle

24

Gabriel Citron

has innumerable particular instantiations, it is interesting to ask why, in his bracketed note, Wittgenstein chose to illustrate his general remark in PI §373 by means of the example of religion and theology in particular. I think the explanation for this is that—as I have said—the realm of religious language is paradigmatic of language in general. That is, there are grammatical qualities which are important to Wittgenstein which—though shared by both religious and nonreligious uses of language—are highlighted especially clearly in the realm of religious language. As Wittgenstein said in the lecture passage just quoted: “[t]hat so & so is ridiculous . . . or blasphemous, shows grammar”. This is because—as pointed out above—utterances which contravene grammatical rules do not produce false statements, but ones that make no sense; and sometimes ridiculousness is a sign of just this kind of failure to make sense. Often a person will not be consciously aware of the existence of a certain grammatical rule until he hears how ridiculous—or funny—it sounds when it is contravened.24 But when it comes to the realm of religious language—unlike almost any other realm of language—there already exists an historical discipline which is largely dedicated to clarifying and highlighting precisely what the rules for the proper use of religious terms are. Namely, the discipline of theology. Furthermore—unlike almost any other realm of discourse—the realm of religious language has special designations for certain kinds of contravention of the grammatical rules governing its key terms, namely: “blasphemy”, “heresy” and the like.25 And to top it off, due to the religious importance of avoiding blasphemy and heresy, theologians and religious people generally are highly alert to their occurrences. So it is not hard to see why the realm of religious discourse struck Wittgenstein as paradigmatic for the purposes of illustrating his general methodological principle that “Grammar tells what kind of object anything is” (PI, §373)—for it comes with a pre-attached discipline which is largely engaged in precisely the kind of grammatical boundary-drawing that Wittgenstein sees as central to the practice of the form of philosophy that he is advocating. For Wittgenstein, the realm of religious discourse was much like other realms of discourse, only more so. It highlights what is important—but less clearly visible—in other realms of discourse. 4. THE VARIETIES OF RELIGIOUS BELIEFS AND UTTERANCES26 In the remainder of this chapter, I would like to look at one further example— emerging from Wittgenstein’s remarks on religion in his 1933 lectures—of a way in which the realm of religious discourse may have highlighted for Wittgenstein, in a particularly clear way, a grammatical characteristic that is present but not obvious in discourse more generally. The characteristic in question is what I call the “messiness” of grammar—and we can begin by looking at Wittgenstein’s stress on the grammatical and logical varieties of

Religious Language as Paradigmatic of Language in General

25

religious utterances and beliefs in the part of Wittgenstein’s lecture from 1 May 1933 which leads up to his above-quoted remark about Luther:27 I have always wanted to say something about grammar of ethical expressions, or e.g. of “God”. / . . . / Now: use of such a word as “God”. / It has been used in many different ways: e.g. sometimes for something very like a human being—a physical body. / Cf. “soul” which has sometimes been described as something “gaseous”.28 / But others haven’t meant by “soul” anything like this. / . . . / If I restricted use of “soul” to such phrases as “His soul is at rest” or “His soul is easily stirred”, you might say I’m denying that there is any soul: but you may mean by “Men have souls” simply that such propositions are true: / Fact that “I hope to God” is used by a man, may mean that he does believe in God in a sense. / When a man worships idols / (1) One possibility is that he believes idol is alive & will help him. / Then man must have forgotten that he made it: but this can happen. / (2) In millions of cases, this will not happen, but e.g. / (a) God dwells in the statue. / But what does “dwells” mean? / By asking what he would say, & what he wouldn’t, you can get at how he uses the word. / Haeckel said “God is a gaseous vertebra[te]”,29 meaning that that’s what people meant. / This is like saying “Soul is a gaseous human being”; & answer is sometimes people so use this word, but sometimes not at all. / . . . / Luther said: “Theology is Grammar of word of God”. / This might mean: An investigation of idea of God is a grammatical one. / Now (a) suppose “god” means something like a human being; then “he has two arms” & “he has four arms” are not grammatical propositions. but (b) suppose someone says: You can’t talk of god having arms, this is grammatical. / . . . / There are many controversies about meaning of “God”, which could be settled by saying “I’m not using the word in such a sense as that you can say———” (MWL, 8, 74–8) In this passage Wittgenstein discusses the words “God”, “soul” and “idol”, sentences which include them, and the beliefs expressed by those sentences. In each case his point is the same—namely, to remind us that the given word or sentence can have a variety of different grammars, and that the given beliefs can have a variety of different logics. The first statement that he makes in his first organized discussion of religious topics in his later period is to say that the word “God” “has been used in many different ways”. He reiterates this when he says that “There are many controversies about [the] meaning of ‘God’, which could be settled by saying ‘I’m not using the word in such a sense as that you can say———’”. The point of both of these statements is to make clear that the word “God” does not have a uniform grammar— and the same would apply to sentences which include the word “God”. He makes the same point with regard to the word “soul”: “‘soul’ which has sometimes been described as something ‘gaseous’. But others haven’t meant

26

Gabriel Citron

by ‘soul’ anything like this”. Again, his point is to remind us of the variety of ways in which the word is used, and the variety of roles that the belief “that people have souls” may play. His response to the claim that the “Soul is a gaseous human being” could almost be taken to be paradigmatic of Wittgensteinian grammatical remarks: “sometimes people so use this word, but sometimes not at all”. That is, Wittgenstein accepts that words such as “God” and “soul” are often used “scientifically” (broadly speaking) as names referring to entities in the world about which we may theorize; but he tries to remind us that they are not only used like that. As well as pointing out the logical and grammatical variety of religious words and beliefs, Wittgenstein highlights the fact that religious practices can have a variety of different natures, noting that there are different kinds of idol worship. Idol worship could be a form of pragmatic-instrumental behaviour, i.e., a practice which the worshiper believes will help him attain his practical ends: “One possibility is that he believes [the] idol is alive & will help him . . . this can happen”. But, then again, “[i]n millions of cases, this will not happen”, and the worship may be something quite different, such as an expression of awe and reverence, or a matter of instinct with no aim at all, or any number of other things which are not pragmatic or instrumental in nature. These remarks stress the logical and grammatical variety that can be found in religious beliefs, utterances and practices—but they are rather programmatic, and they do not provide us with any fleshed-out examples of the different forms that these beliefs, utterances and practices could take. In the next part of his lecture, Wittgenstein begins to remedy this by looking in a little more detail at two forms of the belief that God answers prayers for guidance. He says:30 If parents believe that they have an answer to prayer, then in a sense they have an answer from God. / . . . / The son of these parents began to have doubts as follows. He wanted to go to a theatre. Parents told him to pray. He prayed & got “Yes”; they prayed & got “No”. He was wrong to doubt on this ground: or rather his religion was merely scientific. / If you interpret your experience on this basis, it is possible that different people should get different answers. / Such difficulties can be solved by “We didn’t pray in the right way”. / Disagreement of 2 answers can quite properly produce a religious conflict; but this need not be religion of that kind which takes religious statements scientifically (as the son did). / . . . / if “Yes” does lead* to bad plays*, I shall give up praying; but it never does. / if it does, it must be sent me by God for some particular purpose. (MWL, 8, 78–9) As is often the case with Wittgenstein, this is a very dense passage, and it does not follow a single, linear thought. However, by picking out and ordering its various threads, we can extract from it two opposed examples of the belief that God answers prayers for guidance. On the one hand,

Religious Language as Paradigmatic of Language in General

27

Wittgenstein describes a boy who gives up believing in a God who guides us in prayer (“the son . . . began to have doubts as follows”), either (a) because he seemed to be given guidance opposed to that which his parents received in their prayers on the same matter (“He prayed & got ‘Yes’; they prayed & got ‘No’”), or (b) because the guidance he received seems to be somehow inappropriate (“if ‘Yes’ does lead* to bad plays*, I shall give up praying”). And on the other hand, Wittgenstein describes a believer who, regardless of the results of his prayers, does not give up his belief in a God who guides us in prayer, for if (i) the guidance seems to conflict with the guidance that someone else receives, he will say that perhaps one of them did not pray in the right spirit (“Such difficulties can be solved by [saying] ‘We didn’t pray the right way’”), and if (ii) the guidance seems to be inappropriate, he will say that perhaps God sent him such apparently inappropriate guidance for a special reason unknown to him (“it must be sent me by God for some particular purpose”). In these two descriptions Wittgenstein has provided us with two very different examples of possible relationships to prayer, and two very different examples of possible beliefs in a God who guides us in prayer. Namely, one example of a belief in prayer which seems to be a thesis which is constantly subjected to testing and held open to possible falsification by experience; and another, in which people are clearly not open to any form of falsification of their belief at all, for whatever happens, they always have a way of understanding and assimilating it within their belief system.31 In short, he has described two varieties of belief in a God who guides us in prayer: one which is logically “scientific” and the other which is logically “nonscientific”. A further example of Wittgenstein’s stress on the fact that religious beliefs, utterances and practices can come in a variety of different forms can be found in Wittgenstein’s discussion of Frazer’s Golden Bough in his 1933 lectures. In the lecture of 5 May 1933, Moore records Wittgenstein to have said: Frazer talks of Magic performed with an effigy, & says primitive people believe that by stabbing effigy they have hurt the model. / I say: Only in some cases do they thus entertain a false scientific belief. / It may be that it expresses your wish to hurt. / Or it may be not even this: It may be that you have an impulse to do it, as when in anger you hit a table; which doesn’t mean that you believe you hurt it, nor need it be a survival from prehuman ancestors. / Hitting has many sides. / Frazer also talks of festivals in which effigy of a human being is killed; & explains all as down to fact that once this was done to a man. / This may be so; but it’s not true that it must. / The experience of making an effigy & throwing in water has a peculiarity which may be satisfactory for its own sake: like tearing a photograph of our enemy. (MWL, 9, 5) Wittgenstein says that Frazer takes rituals—such as the stabbing of an effigy—to be instrumental-pragmatic attempts to effect something in the

28

Gabriel Citron

world, grounded in certain scientific theories about how the world works. Then—rather than denying that Frazer’s scientific description ever applies— Wittgenstein simply insists that it should not be taken to apply universally: “Frazer . . . says primitive people believe that by stabbing effigy they have hurt the model. / I say: Only in some cases do they thus entertain a false scientific belief”. Wittgenstein then suggests two other possible explanations for a person’s participation in a ritual such as the stabbing of an effigy, which would make it quite different from an instrumental-scientific action. Namely: “It may be that it expresses your wish to hurt. / Or . . . It may be that you have an impulse to do it”. Wittgenstein wants to remind us about the variety of games that we might be playing when involved in ritual and religious behaviour: “Hitting has many sides”, i.e., aspects. Wittgenstein’s comment towards the end of the above passage—“[t]his may be so; but it’s not true that it must”—sums up his grammatical approach well. That is, Wittgenstein’s position is that religion can, and sometimes does, come in scientific and informational varieties, but that it can also, and sometimes does, come in nonscientific and noninformational varieties. The final example of Wittgenstein’s sensitivity to the wide grammatical and logical variety in religion in his 1933 lectures comes from a little earlier in the year. Wittgenstein had been discussing topics surrounding the nature of mathematical proof for a number of lectures. Then, in the middle of his lecture of 3 February 1933, he expresses the now familiar Wittgensteinian sentiment that if you: “[s]how me your proof; I can then understand what it proves, & therefore what you mean. / What existence means depends on what sort of proof you have” (MWL, 7, 77).32 He then proceeds to illustrate this principle with a nonmathematical example—namely, the example of proofs for the immortality of the soul. Moore records him to have said: Consider: Proofs of immortality of soul. / This is one of cases where a word is used in two entirely different senses without people realising it. / This question has meant a lot to many people, & these33 the greatest. Hence* not mere bosh. / But yet, when you hear of proofs of it, something will smile inside you. / Now one might think: These people have just made a scientific mistake. / But a mere mistake never means so34 much to anyone. / This question, & many others, have always been treated in 2 ways, a hot, & a cold: & that when treated cold, an absurdity results, does not prove that when hot they’re absurd. / Once I was woken from profound sleep by a severe cramp, & for 2 hours I felt as if my soul had not quite come back: I was in great state of excitement. / The feeling that there was something not quite coincident with me, was irresistible to me; though I don’t habitually think in terms of body & soul, etc. / Or, in case of person dying, one feels that he has gone very far away. / Suppose you have developed a theory: & talked about measuring distance between soul & body, there bosh would begin. / Certain analogies may be irresistible, & yet only hold in a very restricted field.

Religious Language as Paradigmatic of Language in General

29

/ Read Oliver Lodge, you’ll find pseudo-experiments, & bosh—false scientific statements. / But, I read in Kierkegaard, about duty, that his father when he was a boy, gave him a task to do, & he felt that even death could not rid him of the duty: & then he said: this was a proof of immortality.35 / This struck me as giving one meaning of immortality which made it important. / . . . / Now: If you want to know what a proof proves, look at the proof: shows that Lodge & Kierkegaard don’t mean the same by “immortality of soul”. (MWL, 7, 78–9) In this passage Wittgenstein stresses the variety of forms that belief in the immortality of the soul takes. Sentences about the immortality of the soul can be used, on two different occasions, with “two entirely different senses”—one scientific and the other not. According to Wittgenstein, the nature of a belief is largely constituted by its grounds, its ramifications and the roles that it plays—that is, by its connections to a range of beliefs and behaviours. As Wittgenstein says in the Big Typescript: “the proof is part of the grammar of the proposition” (BT, 121, 426), and, “the proof belongs to the sense of the proved proposition, i.e. determines its sense” (BT, 121, 428). Oliver Lodge, the eminent English physicist and spiritualist, used “experiments . . . and . . . scientific statements” to try to furnish proof of the immortality of the soul. The belief that he thereby tried to prove was therefore a scientific hypothesis—though, in Wittgenstein’s opinion, a very bad one. Kierkegaard’s proof, however, was not a scientific one at all—but rather, had more to do with a certain moral attitude—and the belief that he thereby tried to prove was therefore not a scientific hypothesis. “Now: If you want to know what a proof proves, look at the proof: [this] shows that Lodge & Kierkegaard don’t mean the same by ‘immortality of [the] soul’”— by noting that both scientific and nonscientific proofs have been given for the immortality of the soul, Wittgenstein again reminds us that religious beliefs can be scientific and nonscientific.36 Wittgenstein’s concern to highlight that religious beliefs can come in a variety of logical kinds becomes very clear in the middle of the above passage. Talking about the question of whether or not the soul is immortal, he says: “This question has meant a lot to many people . . . / But yet, when you hear of proofs of it, something will smile inside you. / Now one might think: These people have just made a scientific mistake. / . . . This question, & many others, have always been treated in 2 ways, a hot, and a cold: & that when treated cold, an absurdity results, does not prove that when hot they’re absurd”. Wittgenstein imagines someone who dismisses “scientific religion” as laughably bad science, but who does not recognize that religious beliefs are not always scientific. This person therefore judges all proofs of immortality by the same measure—that of scientific reasoning—and he finds that they are all absurd. Using the terms “hot” and “cold”, Wittgenstein distinguishes between religious beliefs which are scientific hypotheses and those which are not;37 and he warns us that just because some versions

30

Gabriel Citron

of belief in immortality really are absurdly bad science, this does not mean that all varieties of that belief are—for some may not be attempts at science at all, and so cannot be judged as such, or found wanting as such. 5. LOGICAL AND GRAMMATICAL “MESSINESS” GROUNDED IN LOGICAL AND GRAMMATICAL VARIETY The basic virtue of Wittgenstein’s recognition of both scientific and nonscientific forms of religious beliefs, utterances and practices in the 1933 lectures is that it reflects a diversity that actually exists. Both the scientistic view that all religion is “scientific”, and the antiscientific view that no religion is “scientific”, are equally crude distortions of the complex and multiform reality of religion. It is important, however, to note the kind of logical and grammatical variety that Wittgenstein highlights in these lectures. Let us focus, for the moment, on religious utterances. Wittgenstein does not point out merely that some religious utterances are grammatically scientific while other— completely unrelated ones—are nonscientific. Rather, he points out that given two religious utterances of the very same sentence, one of them may be grammatically scientific while the other may be grammatically nonscientific. For example, one utterance of the sentence “People have immortal souls” could be scientific and another nonscientific—depending on the kinds of things that the respective utterers think could ground their utterance, or could be grounded by it, and other such aspects of the broad roles that the respective utterances play. An important consequence of recognizing that different utterances of the same sentence can have radically different grammars is that it opens the way to recognizing the existence of a further level of grammatical complexity that religious utterances often exhibit. Namely, because utterances of a given sentence can be of two different grammatical kinds, it is also possible for utterances to be grammatically indeterminate between those two grammatical kinds, or for utterances to mix together different elements of those two grammatical kinds, or for a sequence of utterances as uttered by a given person to move fluidly from being of one grammatical kind to being of another. Often this kind of grammatical mixedness, indeterminacy and fluidity go unnoticed, because these “messy” qualities are masked by the fact that the utterance is always of the same unchanging and stable sentence. In his discussion of the varieties of ways in which the word “soul” is used, Wittgenstein makes the very astute observation that it “is one of [the] cases where a word is used in two entirely different senses without people realising it” (MWL, 7, 78; my italics).38 This observation is important because it is often the failure to notice that a given sentence has a variety of possible grammars which allows for the flourishing of grammatical indeterminacy, mixedness and fluidity between the unnoticed varieties in the first place.

Religious Language as Paradigmatic of Language in General

31

Indeterminacy, mixedness and fluidity can be found in religious beliefs and practices, as well as in religious utterances. For—as Wittgenstein illustrated—given religious beliefs (such as belief in the immortality of the soul) and practices (such as idol worship) also come in different varieties, and can therefore sometimes also be indeterminate, mixed or fluid between those varieties. I will illustrate, in a little more detail, what I mean by indeterminacy, mixedness and fluidity, using Wittgenstein’s example of belief in the immortality of the soul. Consider, for instance, a religious person who uses Lodge’s scientific proofs of the immortality of the soul, in conjunction with Kierkegaard’s emotional or moral proof, and who insists that they both help to bolster his belief in immortality. Or consider someone whose belief in immortality is grounded in Lodge’s scientific proofs, but who refuses to countenance any scientific attempts to disprove his belief, saying that they are not relevant to the kind of belief that he holds. These people’s beliefs in immortality each seem to contain elements unique to both the scientific and the nonscientific versions of the belief: we might describe their beliefs as having a mixed logic; and if they expressed their beliefs by saying that the soul is immortal, their utterances would be grammatically mixed. Next consider a religious person who was brought up with a belief in immortality, and who takes comfort in his belief that his parents live on in the afterlife—but other than giving him that comfort, his belief plays little role in his life, and he has never given much thought to the precise nature of the immortality in which he believes. He has, for example, never thought about what kinds of argument or evidence, if any, would count as grounding his belief, or what he would count as refuting it. This person’s belief in immortality could perhaps best be described as logically indeterminate—it is simply not sufficiently fleshed-out to count as being of one determinate kind rather than another; and if he expressed his belief by saying that he believed in immortality, this utterance would be grammatically indeterminate. Perhaps if we asked the believer probing questions, or if he was confronted with some possibly relevant evidence, he would be prompted to develop his belief in one way or another, and—depending on how he answered or reacted—his belief would come to be determinately scientific or nonscientific. In this case his belief will have moved fluidly from having been logically indeterminate between being scientific and being nonscientific, to being determinately one or the other. Lastly, to illustrate fluidity, consider a religious person who believes in the immortality of the soul, for the kind of scientific reasons that Lodge puts forward. But, when confronted with stronger scientific arguments the other way, rather than reject his belief in immortality, he concludes that he should deepen his understanding of what immortality is. He may then develop an understanding of immortality which is more akin to Kierkegaard’s, and which is unfalsifiable. This new understanding of immortality is radically different from his old one, but he does not consider it to be a

32

Gabriel Citron

new belief that has replaced his old refuted one—rather, he considers his old belief in immortality to have developed into a deeper and more mature form of the same belief. The nature of his belief has changed—it was logically fluid; and if he had expressed his belief at its different stages, saying that the soul is immortal, then this sequence of utterances could be described as “grammatically fluid”. In each of the cases illustrated above, the believer could not be said to be playing any one particular game—scientific or nonscientific; rather, his belief mixes different games together, or is indeterminate or fluid between them. The recognition of the possibility of logical and grammatical indeterminacy, mixedness and fluidity in beliefs and utterances in general is a crucial insight of Wittgenstein’s later period. In fact, he himself saw it as one of the pivotal differences between his earlier and later thought. As he said in the Big Typescript: Now wasn’t it a mistake of mine (for that’s what it strikes me as now) to assume that whoever uses language always plays a particular game? For wasn’t that the sense of my remark that everything about a proposition—no matter how off-the-cuff it is—“is in order”? (BT, 58, 198) That is, we should not assume that a language user is neatly playing one particular language-game when he makes a given utterance, because its grammar might actually be hovering indeterminately between two different language-games, or it may be a mixture of elements of two different language-games; and we should not assume that he is neatly playing one particular game with a given sequence of utterances of a sentence, because its grammar might well move fluidly from one language-game to another (and perhaps back again). This quotation from the Big Typescript is embedded right in the middle of an early draft of what later became PI §§81–3. Wittgenstein begins PI §81 with the insight that when people use language they are not necessarily playing a game “with fixed rules”.39 This is an eschewal of the position that “if anyone utters a sentence and means or understands it, he is thereby operating a calculus according to definite rules”. Wittgenstein thinks that games and play can provide an illuminating analogy for language and language use. Sometimes language use is analogous to playing official football matches or basketball games. In these cases a definite single game is being played, and it is clear which game it is and what rules govern the play (though the rules do not, per impossible, cover every conceivable circumstance). But in PI §83 Wittgenstein points out that language use is not only analogous to the playing of such formal and regimented games, but also sometimes to the way in which people play around with a ball in the park: they may begin by simply throwing the ball about, then move on to

Religious Language as Paradigmatic of Language in General

33

play a particular game, then another, or they may be undecided between games, or any combination of the above and similar things. Wittgenstein’s ball-playing analogy of PI §83 may be expanded as follows: we could imagine people playing with a ball in such a way that (a) the rules governing their play are a mix of rules from a number of different games (though each individual rule involved in governing their play is from a particular game, some are from one particular game, and some are from another); (b) their play is objectively indeterminate between a number of different games (there are not yet enough individual rules governing their play—or the individual rules they have are not yet sufficiently fleshed out—for there to be an answer to the question of which of a number of possible games they are playing); or (c) the rules governing their play are fluid (some of the individual rules involved in governing their play change over time, or rules are added or taken away). Thus it is a key insight of Wittgenstein’s thought in the Philosophical Investigations that we often mix together rules from incompatible language-games, and often hover indeterminately between, or move fluidly between, them. It will be instructive to look briefly at an example of Wittgenstein’s that illustrates how the use of a sentence may be indeterminate or mixed between two different grammars. The possibility of indeterminacy and mixedness in the following example derives from the variety of grammars that can be had by different utterances of a sentence such as “I am afraid”: If I tell you “I have been afraid of his arrival all day long”—I could, after all, go into detail: Immediately upon awakening I thought. . . . Then I considered. . . . Time and again I looked out of the window, etc., etc. This could be called a report about fear. But if I then said to somebody, “I am afraid . . .”—would that be as it were a groan of fear, or an observation about my condition?—It could be either one, or the other: It might simply be a groan of fear; but I might also want to report to someone else how I have been spending the day. And if I were now to say to him: “I have spent the whole day in fear (here details might be added) and now too I am full of anxiety”—what are we to say about this mixture of report and statement? Well what should we say other than that here we have the use of the word “fear” in front of us? (RPP II, §156) The sentence “I am afraid” can be used exclusively to express the emotion of fear, or exclusively to report information about my current state—the sentence thus has at least two possible grammars. Sometimes utterances of the sentence will determinately and purely play one or other of the possible grammatical roles. But sometimes utterances of the sentence will be indeterminate between the two distinct grammatical varieties, or some kind of mixture of them. There needn’t be anything wrong with grammatical

34

Gabriel Citron

indeterminacy, mixedness or fluidity in themselves—but philosophical confusion and distortion is very likely to ensue from failing to recognize the existence of such grammatical “messiness”. I said above that sensitivity to grammatical mixedness, indeterminacy and fluidity is dependent on sensitivity to the logical and grammatical variety that given beliefs, utterances and practices can have—because the mixedness, indeterminacy and fluidity will be between those varieties. I suspect that it is no coincidence that in precisely the period during which Wittgenstein was developing his ideas about the “messiness” of grammar, he made detailed remarks about the logical and grammatical variety of religious beliefs, utterances and practices in his lectures. For—similarly to what we saw in section 3 with the idea of grammar expressing essence—the religious realm of discourse is paradigmatic of other realms of discourse when it comes to logical and grammatical “messiness”. This is because logical and grammatical “messiness” are both more common in religious beliefs and utterances than in other realms of language and life, and because they are more liable to be noticed in religious contexts than in others. The reason for this brings us back to our discussion in section 3; for most religious beliefs—such as belief in God, or belief in the immortality of the soul—can be held in orthodox forms or in heretical forms, and therefore the sentences which express those beliefs will be able to be uttered with at least two different grammars. This dual logical possibility for so many religious beliefs, and this dual grammatical possibility for so many religious sentences, makes logical and grammatical variety far more common in the realm of religion than elsewhere; and the pervasiveness of logical and grammatical variety allows, in turn, for the pervasiveness of the second-order characteristics of logical and grammatical “messiness”: indeterminacy, mixedness and fluidity between the different varieties. Furthermore, as I said in section 3, the religious importance of the distinction between orthodox and heretical forms of given beliefs means that theologians, and religious people more generally, are highly sensitive to the existence of this variety—making logical and grammatical “messiness” not only more common in the religious realm, but also more likely to be picked up on. So once again it turns out that Wittgenstein’s remarks on religion in his 1933 lectures touch on certain qualities in the realm of religious discourse which are shared more generally, but which can be seen more clearly in the religious context. Wittgenstein’s key insight that “whoever uses language does not always play a particular game” is true in many areas of grammar, but nowhere truer than with religion. Perhaps it was the particularly illustrative nature of religious language and life which made it so fruitful for Wittgenstein to give it such attention in his 1933 lectures, at that seminal period in the development of his later thought on the “messiness” of grammar and the method of grammatical elucidation.40

Religious Language as Paradigmatic of Language in General

35

REFERENCES Bengel, John Albert. 1858. Gnomon of the New Testament, vol. I. Translated by Andrew R. Fausset. Edinburgh: T & T Clark. Braithwaite, R. B. 1933. “Philosophy”. In University Studies: Cambridge 1933, edited by Harold Wright, 1–32. London: Ivor Nicholson & Watson. Britton, Karl. 1999. “Portrait of a Philosopher”. In Portraits of Wittgenstein, vol. 2, edited by F. A. Flowers III, 205–11. Bristol: Thoemmes Press. Cassian, 1958. The Conferences of Cassian. In Western Asceticism, edited and translated by Owen Chadwick, 190–289. London: SCM Press. Drury, Maurice O’Conner. 1984a. “Some Notes on Conversations with Wittgenstein”. In Recollections of Wittgenstein, edited by Rush Rhees, 76–96. Oxford: Oxford University Press. ———. 1984b. “Conversations with Wittgenstein”. In Recollections of Wittgenstein, edited by Rush Rhees, 97–171. Oxford: Oxford University Press. Hacker, P. M. S. 1997.Wittgenstein: Meaning and Mind. An Analytical Commentary on the Philosophical Investigations, Vol. 3. Oxford: Blackwell. Haeckel, Ernst. 1929. The Riddle of the Universe. Translated by Joseph McCabe. London: Watts & Co. Hamann, Johann Georg. 1956. Briefwechsel, vol. II (1760–1769). Edited by Walther Ziesemer and Arthur Henkel. Wiesbaden: Insel-Verlag. ———. 2007. “Miscellaneous Notes on Word Order in the French Language”. In Writings on Philosophy and Language, edited and translated by Kenneth Haynes, 20–32. Cambridge: Cambridge University Press. Lewy, Casimir. 1976. Meaning and Modality. Cambridge: Cambridge University Press. Malcolm, Norman. 2001. Ludwig Wittgenstein: A Memoir. Oxford: Oxford University Press. McGuinness, Brian, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. Monk, Ray. 1991. Ludwig Wittgenstein: The Duty of Genius. London: Vintage. Moore, G. E., and Norman Malcolm. 2003. “George Edward Moore/Norman Malcolm: Correspondence (1937–1958)”. Edited by Josef G. F. Rothhaupt, Aidan Seery and Denis McManus. Wittgenstein-Jahrbuch 2001/2002: 245–95. Wittgenstein, Ludwig. 1966. “Lectures on Religious Belief”. In Lectures and Conversations on Aesthetics, Psychology and Religious Belief. Compiled from Notes taken by Yorick Smythies, Rush Rhees and James Taylor, edited by Cyril Barrett, 53–72. Oxford: Blackwell. (LRB) ———. 1978. Remarks on the Foundations of Mathematics. Third edition. Edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe.Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RFM) ———. 1979. Wittgenstein’s Lectures: Cambridge, 1932–1935. From the Notes of Alice Ambrose and Margaret Macdonald. Edited by Alice Ambrose. Oxford: Basil Blackwell. (AWL) ———. 1980. Wittgenstein’s Lectures: Cambridge, 1930–1932. From the Notes of John King and Desmond Lee. Edited by Desmond Lee. Oxford: Basil Blackwell. (LWL) ———. 1993. Ludwig Wittgenstein: Philosophical Occasions 1912–1951. Edited by James C. Klagge and Alfred Nordmann. Indianapolis: Hackett. (PO) ———. 1998a. Culture and Value: A Selection from the Posthumous Remains. Edited by G. H. von Wright in collaboration with Heikki Nyman. Revised edition of the text by Alois Pichler. Translated by Peter Winch. Oxford: Blackwell. (CV)

36

Gabriel Citron

———. 1998b. Remarks on the Philosophy of Psychology, vol. II. Edited by G. H. von Wright and Heikki Nyman. Translated by C. G. Luckhardt and M. A. E Aue. Oxford: Blackwell. (RPP II) ———. 2003. Ludwig Wittgenstein: Public and Private Occasions. Edited by James C. Klagge and Alfred Nordmann. Lanham: Rowman & Littlefield. (PPO) ———. 2005. The Big Typescript: TS 213. Edited and translated by C. Grant Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell Publishing. (BT, chapter number, page number) ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte.Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. (PI) ———. Forthcoming.Wittgenstein’s Lectures: Cambridge 1930–33. From the Notes of G. E. Moore. Edited by Gabriel Citron, David Stern and Brian Rogers. Cambridge: Cambridge University Press. (MWL, notebook number, page number)

2

Wittgenstein’s Blue Book Reading between the Lines Jonathan Smith

INTRODUCTION When Rush Rhees produced his edition of the Blue and Brown Books he called the volume “Preliminary Studies for the Philosophical Investigations: generally known as the Blue and Brown books”.1 The title was, no doubt, deliberately chosen to emphasize the fact that these works were the first attempts by Wittgenstein to publicize in print, albeit to a controlled group, ideas such as family resemblance and language-games that became familiar to a much wider audience through the publication of Philosophical Investigations. Some of these concepts were already quite well developed in the form that they appear in the Blue Book, the earlier of the two works in Rhees’ volume, but others still had to undergo some development before they reached the state in which they were first published by the trustees in the Philosophical Investigations. By Rhees’ title we should not ascribe to the Blue Book the role of a draft for Philosophical Investigations. It does not, for example, appear in the Kritisch-genetische edition of the Philosophical Investigations, and quite rightly so. However, the relationship between the two works is genetic, both in the sense of the development of concepts in Wittgenstein’s philosophy, but also, in spite of the different languages used, in the development of a text and of specific terms to communicate those concepts to the wider world. Perhaps because of the relationship with the Brown Book in print, the singularity of the Blue Book is rarely acknowledged. Yet the work published in 1958 is unique amongst Wittgenstein’s surviving corpus. It differs from other material published from the Nachlass in a number of ways. Firstly it is the only dictation whose primary purpose was to assist Wittgenstein’s students, rather than to clarify his own thoughts and to incorporate them into his attempts at publication (though we should not ignore this latter aspect as an important secondary consequence). Secondly, unlike the majority of Wittgenstein’s literary remains, it is not written in the familiar remarkbased style. And thirdly it was published from a duplicated typescript that Wittgenstein was happy to circulate to a group of people that included, in addition to his pupils, established thinkers such as Russell, Moore and Sraffa. From this I think we can presuppose that the text was regarded by

38

Jonathan Smith

Wittgenstein as sufficiently finished to circulate, and that therefore it carries with it a certain amount of authority. The existence of such an authoritative text gave the trustees cause for different considerations than for the other volumes that they produced from the Nachlass. Unlike the other “works” published by the trustees, little of the preparatory material, which might be used to help us understand its development, had apparently survived. The introduction by Rhees to the first edition does nothing to suggest that the text as printed was not the verbatim text dictated by Wittgenstein during his classes. Indeed, by omission it positively encourages this belief. As a result the Blue Book has not been subject to the level of textual criticism as the other works published posthumously and in terms of its textual development it is something of a “black box”. However, we can now hope that considerable light will be shed on the process by which the Blue Book was composed thanks to the generous donation to Trinity College Library of the copy given by Wittgenstein to Robert Thouless.2 Trinity already has several copies of the Blue Book, so the acquisition of another copy might seem unremarkable. However, what is immediately obvious about the Thouless copy, even after the briefest examination, is that it is formed of earlier drafts of the text that differ markedly from that with which we are familiar. Though there is some complexity in its construction, it bears considerable evidence of amendments in the hand of Wittgenstein on almost every page and in many ways it is what might be called a proto-Blue Book. EVIDENCE FOR THE COMPOSITION OF THE BLUE BOOK Before examining the Thouless copy of the Blue Book, we should first revisit the documentary evidence for its composition. Such evidence that we have comes both from Wittgenstein himself and from those to whom he dictated it, though it is neither necessarily complete nor straightforward. Let us start with Wittgenstein himself, who gives us the most immediate description of the process. On 11 December 1933, he wrote to W. H. Watson thus: I am lecturing a good deal and have adopted a method which I think is the right one for me. I explain things to my pupils and then dictate to them short formulations of what we’ve been discussing and of the results. These are then typed and duplicated so that each man can get a clear copy. In case you should be interested to have a copy . . . just let me know.3 Two years later Wittgenstein told Russell a little more of his intentions in a letter published in Rhees’ introduction to the Blue and Brown Books: two years ago I held some lectures in Cambridge and dictated some notes to my pupils so that they might have something to carry home

Wittgenstein’s Blue Book

39

with them, in their hands if not in their brains. And I had these notes duplicated. I have just been correcting misprints and other mistakes in some of the copies and the idea came into my mind whether you might not like to have a copy. . . . (I think it is very difficult to understand them, as so many points are just hinted at. They were meant only for the people who heard the lectures.)4 Further contemporary evidence comes from Alice Ambrose. Writing to C. L. Stevenson, the American philosopher, she describes the sessions with Wittgenstein thus: This year he’s been in proper antics. The class started off too large, and remained so. After about two weeks, he decided he’d meet 5–7 people of his own choosing, twice a week. . . . At present I am one of the chosen who go and wait for the oracle to speak. Mrs. Braithwaite, Skinner, Goodstein, Coxeter (mathematician of Trinity) and I came first and later Mrs Knight and a chap named Martineau were allowed to join.5 . . . this is what he does in the inner sessions. He discusses with us what he wants to say (he’s talked about language games, as before) and then he dictates. And he makes an attempt to give a connected discussion. I should say that when the year is over we’ll have the equivalent of a semi-official deliverance. At least he takes great pains with the English and one of us types off the notes for him, and they are corrected. Each time we meet with him we stay from 2½-3 hours, and we’re all limp but him. That’s twice a week and then there’s the other two hour lecture. The last week of term, he wasn’t finished, so we met once more for 2½ hours. And he still wasn’t done so we met again for 3¼ hours. His criticisms and his analyses are as fine as ever.6 Ambrose tells a similar story nearly thirty years later when approached by O. K. Bouwsma, who was in the process of reviewing Rhees’ edition. To him she also indicated that the Blue Book was mimeographed “but not a few pages at a time”, and though she does not remember exactly when the Blue Book was assembled, she supposed it to be sometime after the year was over.7 It is also worth recording that during the Christmas vacation—that is sometime between the tenth and eleventh dictations—Wittgenstein sought a grant from the Moral Sciences Faculty to cover duplicating of the typescripts, and had an exchange with Coxeter after he had informed Wittgenstein that he had made a trial of a duplicator by duplicating the first six dictations.8 Wittgenstein’s reaction was characteristically sharp; he was no doubt angry at the thought that the texts might be circulated before he had placed his imprimatur on them, even claiming that he would “go as far as resigning my lectureship if it turned out that as a lecturer I cannt [sic] keep my teaching clean, and free from a connection with journalism”.9 Alice Ambrose communicated to Bouwsma that this was the reason Coxeter left the

40

Jonathan Smith

inner circle and did not attend the second term of dictations.10 However, shortly afterwards Wittgenstein wrote to Coxeter, insisting that there was no ill-feeling on his part and indicating that he might use his duplicates and pay him for them if the Moral Sciences Faculty gave him a grant.11 The typescript that Wittgenstein produced for his pupils consisted of two volumes bound in blue, hence the name by which it became commonly known. On the rear of each volume is the legend “Duplicated by Darking and Arch, 59 Bridge Street, Cambridge”. Why he split the work in two is uncertain—it may have been for ease of binding—but the first volume is shorter than the second. Later typescript copies were produced. Von Wright’s copies of this later version of the Blue and Brown Books were donated to Trinity College before his death. With certain caveats, I think from this evidence it is possible to construct a reasonably accurate outline view of the stages that took place in the compilation of the work that we now know as the Blue Book: 1. 2. 3. 4.

Wittgenstein dictated a session to a number of his students. The content of this dictation was typed up. Wittgenstein then edited the typescripts. The edited typescript of each session’s dictation was circulated among his students. 5. At some point later he gathered the material together and reedited it to form what we now know as the Blue Book. 6. Wittgenstein had it duplicated for circulation among his pupils. So much, I think, can be deduced from the quotations above. However, there is little evidence of the process either in the known duplicated copies of the Blue Book or in the published version. In both the text is presented continuously, with no indication of the place where each of the sessions of dictation began or ended. There is also little evidence of any textual amendments, with the exception of a few small manuscript changes made by Wittgenstein himself in some of the duplicated copies, as explained in his letter to Russell and in an editorial note in the introduction to the second edition. For a better idea of the development of the Blue Book we need to find physical evidence of the various steps in its production outlined in the letters quoted above. Such evidence can be found in an examination of the Thouless Blue Book taken within the context of the compositional process of which it forms a part. THE THOULESS BLUE BOOK In a letter to Jim Klagge,12 David Thouless, the donor of the new discovery, wrote that his father went through his papers in the last year of his life. Amongst the more interesting things that he came upon were the

Wittgenstein’s Blue Book

41

conversations with Wittgenstein that Klagge and Nordmann later published in Public and Private Occasions13 and a “heavily edited copy of the Blue Book”. How Thouless senior came to have this particular version is something of a mystery. Robert Thouless entered Corpus Christi College Cambridge in 1912 to study Natural Sciences. After the First World War he took his PhD and moved to the Universities of Manchester and Glasgow to teach psychology, returning to Cambridge in 1938, first as lecturer and later as reader in educational psychology. It is thus highly unlikely that he was present in Cambridge when the Blue Book was being dictated. However, we know from Wolfe Mays14 that he attended Wittgenstein’s lectures in 1940— showing once more the attraction of these lectures not only for students but for established academics. We also know from the other manuscript Thouless rediscovered in 1983, that in 1941 he had regular discussions with Wittgenstein in the Fellows Garden at Trinity on topics not unrelated to those covered by the Blue Book. It is understandable that Wittgenstein may have wanted to explain his thought to Thouless by giving him a copy of the Blue Book, but why did he give him this particular copy? It may well be that he found that the only copy that he had in his possession was this working copy that he had used for editorial purposes, by this time deemed surplus to his requirements, and so Wittgenstein gave it to Thouless. In doing so he may have ensured its survival. In common with all archival material, the Thouless version of the Blue Book yields evidence of the purpose of its creation and the methodology of its creator. It consists of 141 pages in a buff-coloured cardboard binding, which, though now detached, bears the name and address of Thouless and appears to be original. The paper used is 26 × 20 cm—the same size as the duplicated version—and the typing is for the most part either a top-copy or carbon-copy page typed on the same or a very similar typewriter as the copies of the Blue Book that were later circulated. The exception to this is a section formed by dictations 2 to 6, which is a duplicated version containing numerous textual alterations in the type of indelible pencil that Wittgenstein was known to favour. Throughout most of the text there are amendments by Wittgenstein, usually in pen but sometimes in pencil, and a few in the hand of Skinner, though these latter changes were probably not made on his own authority but that of Wittgenstein, perhaps as a result of him reading the text back to him. Most of these amendments consist of refinement of language rather than complete changes of sense and concentrate on making the language of Wittgenstein’s arguments more precise. Yet this is of no little importance as the process may have a causal relationship with the crystallization of his ideas as well as indicating the importance of the aesthetics of the text. As there are several layers of amendment, for the purpose of understanding the textual development of the complete work, we need to consider the typescript not only as a single piece, but additionally both as a series of individual dictations and as an amalgam of at least three sections of text. These

42

Jonathan Smith

sections are formed by dictation 1 alone, dictations 2 to 6 and, much the larger part, dictations 7 to 39. The layout of the typescript in general is different from the version with which we are familiar in that it was originally split into the individual dictations, headed “Wittgenstein 1”, “Wittgenstein 2”, etc., up to “Wittgenstein 39”. This aspect not only gives the document the feeling of a work composed of short chapters, but also shows that each dictation was not particularly lengthy, allowing space within the allotted lecture periods for the discussions of which Ambrose speaks. In the process of editing, these headings have been for the most part deleted in order to allow the text to run from one dictation to another, aided where necessary by the introduction of short transitional passages, usually consisting of only a few words. This was probably done at a later stage of editing after the dictations had been edited individually and, if we follow Ambrose, must have been part of the process of compiling the individual dictations into the Blue Book, which she supposes to have happened at the end of the academic year. However, it is clear that at this point she is far from sure of her evidence and it may be the case that this happened in two parts, as the duplicated version circulated by Wittgenstein falls into two volumes. Although they are not dated, we know from another source15 that the first dictation, beginning with the familiar “What is the meaning of a word?” was begun on 8 November 1933, and thus corroborates the view that is generally held that the decision to dictate the Blue Book took place some time after the beginning of Michaelmas term, when Wittgenstein realized he did not wish to teach a large class. From the same source we also can give Monday 22 January 1934 as a terminus post quem when the first volume of the duplicated version of the Blue Book was brought together, this being the date of dictation 12, the last to be included in that volume. The first distinct section of text is made up purely of dictation 1 which, with the probable exception of some retyped bridging parts in section 3, is the latest part of this particular document to be produced. This is a clean text without manuscript amendments. Its most notable textual feature is the fact that in it the phrase “we try to find a substance for a substantive”, which appears both in the duplicated version and the first edition of the printed work, is replaced by the phrase “a substantive makes us look for a thing which corresponds to it”. The latter version was not introduced until the second edition of the published work.16 The authority for such a textual change was supplied by a holograph amendment by Wittgenstein in Sraffa’s copy of the Blue Book that had originally belonged to Francis Skinner, Sraffa imparting the information to von Wright in a letter as early as 27 August 1958.17 The second section of text is formed by dictations 2 to 6. This consists of a number of duplicated sheets that were at one point stapled together. It is quite likely that this is the result of Don Coxeter’s experiments with a duplicator that so aroused Wittgenstein’s ire when he told him about it and which he offered to use in a subsequent letter. Coxeter claimed to have duplicated dictations 1 to 6.18 If this section is indeed the work of Coxeter, the

Wittgenstein’s Blue Book

43

conjecture that this formed the shorter volume of the Blue Book,19 which we can now see is made up of the first twelve dictations, no longer holds true. It also suggests that Alice Ambrose’s memory may have been at fault when she claimed Coxeter duplicated the first term’s dictations.20 Although the nature of the deletions in this section, made first with indelible pencil and then using thick pen strokes, is designed to make the excised text impossible to read, there are a considerable number of manuscript changes in the text. These correspond almost exactly to those in the published version. In some cases there are multiple levels of amendment and it is clear that there are several layers present in the text. This can be seen in an example taken from dictation 4. Here we see not only the typescript text, but on the reverse of the previous page an attempt by Wittgenstein to improve one of the sentences, which he then edits and reintroduces into the main text. Evidence of an additional layer of amendment is shown a little farther down on the same page by a change made in Skinner’s hand. The fact that Wittgenstein heavily edits the text of a dictation given on 29 November on a duplicate that he offered to use in late January may be suggestive that he was not originally aware of how much he would attempt to improve the text of the work. It certainly indicates that any such improvement might not have taken place immediately. The third (and, for the most part, earliest) section consists of the remainder of the text, that is dictations 7 to 39, and is made up of 31 discrete dictations, dictation 21 having been completely removed and dictations 30 to 31 being treated as one. Each dictation has individually numbered pages. Several of these show signs of careful folding, suggesting that they may have been taken away one at a time for editing on the move, perhaps slipped inside one of the small-format pocket notebooks that Wittgenstein was known to carry with him. This certainly is in accord with the implication in Ambrose’s comments that the dictations were not edited in their totality at the first attempt. Most of the amendments made in this section were incorporated into the version of the Blue Book that was duplicated and circulated (and thence into the published version), but not all. It is noteworthy that some of the amendments appear to be the result of misreadings or, as is possible given the nature of the work as a dictation, mishearings. For example “two single” has been corrected to “too simple”. However these errors arose, they are suggestive the original typescript text in this section is, with the exception noted above, closer to the text that Wittgenstein actually dictated. Indeed it may well be that this section consists of the original typescript copies made of the dictations by the privileged members of his class. A notable exception to the omnipresence of Wittgenstein’s annotations is a short section of two typescript pages, inserted within section 3 to replace the latter part of dictation 12 and all of dictation 13. From this amendments are absent. These appear to be the last pages added to the document and may have been created to replace parts of the typescript where the legibility had been compromised simply by the amount of alterations. However,

Figure 2.1 Facsimiles of pages 11v–12r, reproduced by kind permission of the Master and Fellows of Trinity College Cambridge.

Wittgenstein’s Blue Book

45

there may be a more interesting explanation. It is surely not insignificant that one of these sections coincides with the start of the second volume of the original duplicated Blue Books. In most surviving copies of this version, the first eight pages are missing, presumably removed at the insistence of Wittgenstein. However, in at least one copy, owned by Reuben Goodstein,21 two pages are tipped in at the beginning of the text. These pages are identical with the pages inserted at this point in the Thouless copy. That these sections were added so late in the enterprise, when the duplication of both parts of the Blue Book had been completed, leads me to think that these short sections were added not because the text had been edited out of recognition, but because Wittgenstein wished to suppress ideas which he realized that he no longer held or which he felt were too badly expressed for circulation. Unfortunately, the original texts were not retained. CHRONOLOGY OF THE DEVELOPMENT OF THE BLUE BOOK I have so far described the Thouless copy of the Blue Book and the three sections that make up the text. I would now like to consider the chronological relationship between these three sections of the typescript, and to explain why I have come to the decisions on the relative chronology of the sections that I have. This is crucial to our understanding of the text as a whole. The reconstruction of the development of ideas on temporal lines using variant manuscripts does rely on our ability to date them, or at least make an accurate appraisal as to the order in which the manuscripts were composed. Without this, our understanding of the variant readings would be severely compromised. An important aid in the case of the Blue Book is the comparison of the Thouless document with a surviving manuscript of part of the work in Skinner’s hand, held by the Wittgenstein Archive in Cambridge.22 If we can be sure at which stage of composition of the Blue Book this manuscript was created, a comparison between the text of this and the Thouless Blue Book would be significant. At first sight, purely because it is manuscript in format, this document might be assumed to be Skinner’s copy of the dictation taken down during Wittgenstein’s classes, but this is not, I believe, either its origin or function. Although it is tempting to take the simplistic view that manuscript texts represent the earliest stage of composition of a work, while the move to typescript indicates a later, more finished period of composition, this is not always the case. There are indications that this manuscript in Skinner’s hand might be a fair-copy manuscript produced at a certain phase of the composition of the Blue Book, made presumably at Wittgenstein’s behest, possibly as copy for a typist. It is written in a fair hand with very few corrections, which suggests that it may have been produced at leisure with the aid of an exemplar. Another indication that it was not taken down in class is that it breaks off in mid-sentence. It is hard to believe that this would have been the case when

46

Jonathan Smith

dictation was in full-swing, and it is noteworthy that it breaks off during the passage in section 3 that required complete retyping mentioned earlier in this piece. As we have observed, this may well suggest some unhappiness that Wittgenstein had with this section of text. Crucially, nearly all of the corrections made in Wittgenstein’s hand in the typescripts that make up section 3 are included in the body of the text of the Skinner manuscript. And though the manuscript does include what appear to be the dates of the individual dictations, these could easily be supplied from elsewhere, probably Skinner’s original notes taken from the dictation. We must tread carefully in the case of the Blue Book, where there are likely to have been parallel texts taken by each of the listeners present. We could concoct a possible scenario where the dictations were taken down incorrectly by other students and where the resulting typescripts had to be corrected from a “perfect” text taken down by Skinner. However, it seems to me far more likely that the typescripts that together form the third section of the Thouless Blue Book constitute a version of the text that predates the Skinner manuscript. In addition, the fact that the text of the manuscript is very similar to the original unaltered text of section 2 of the Thouless document, which itself has numerous corrections, warrants the conclusion that the Skinner manuscript represents the immediate subsequent stage of textual construction and section 2 of the Thouless Blue Book the next after that. Dictation 1, with its amendment that is only introduced into the printed version in the second edition, was the next part to be created. The uncorrected typed pages inserted into section 3 being in all probability the last substantial corrections to the text. As a whole the Thouless Blue Book is indeed a complex document. If we consider that there are some amendments in Skinner’s hand, and that the removal of the headings seems to suggest that Wittgenstein made a second pass over the whole text, the Thouless Blue Book includes fractions of at least five stages of the text. These stages reside between the earliest drafts of ideas in the C-series notebooks23 and the duplicated version of the Blue Book with which we are familiar. This is also the case for the Skinner manuscript, which might have been produced dictation-by-dictation after Wittgenstein had corrected it as a matter of record. Yet I believe it is much more likely that it was created at one time, perhaps as a proposed copy-text brought to a premature end by indecision on Wittgenstein’s part about the quality of dictations 12 and 13. TEXTUAL AND PHILOSOPHICAL IMPLICATIONS: AN EXAMPLE Having established the order in which the sections of the Thouless Blue Book were created, we now have a means of better understanding the phases of its composition, which will, I trust, act as a guide to anyone intending a scholarly examination of the work. But I also hope that it will be of real significance to anyone making a study of the Blue Book and of the development of

Wittgenstein’s Blue Book

47

Wittgenstein’s thought in general, by helping us to understand the process which brought his ideas into print. To judge the significance of this, it is also important to consider the purpose for which the Blue Book was created and its position within Wittgenstein’s corpus. From the first, there was no consensus of its importance. To Rhees it was just a set of lecture notes,24 yet to Alice Ambrose it “contains the most revolutionary ideas in the work of Wittgenstein”.25 Two very different estimates of the work are implicit in these two statements, however they are not incompatible; they simply reflect two different aspects of the complex identity of the Blue Book. What is certain is that the Blue Book made public within a small semi-controlled group, a number of the most famous ideas that we associate with the later Wittgenstein. What then are Wittgenstein scholars to make of this addition to the Nachlass? Certainly the more philologically minded will find in the work more than a little call for rejoicing. Although the method Wittgenstein employs differs from much of his other work, and his failure to use his familiar remark-based style does not allow for complex rearrangement and development of individual remarks, the typescript does include a number of levels of amendment. And as numerous scholars have shown, there is something to be understood from the comparison between the different stages of construction of a paragraph, sentence or phrase. With this in mind I would like to give just one example of where I believe that changes of Wittgenstein’s phraseology are significant enough to suggest a refinement of his ideas. Let us consider a section from dictation 10, where Wittgenstein uses the example of different types of leaf to illustrate the idea of family resemblance. According to the original typescript in the Thouless Blue Book, section 3, which we have concluded is the closest part of the text to Wittgenstein’s original dictation, he said the following: There is a tendency, noted in our forms of expression to think that the man who has learnt to understand a general term, say the term “leaf”, has thereby come into the possession of a kind of general leaf as opposed to special leaves. Wittgenstein, however, is still developing the idea, and he alters the text so that his subject comes “to possess a kind of general picture of a leaf as opposed to a picture of particular leaves”. The double insertion of “picture” here not only instantly gives Wittgenstein’s point another dimension and makes it more readily intelligible, but also suggests that his thoughts were in flux and had not settled into their final form. By the time we reach the typescript of Part I of the Investigations, the leaf scenario has become part of a more complex exposition of family resemblance. In remark 73 Wittgenstein writes: So if I’m shown various leaves and told “this is called a ‘leaf’”, I get an idea of the shape of a leaf, a picture of it in my mind.—But what does

Figure 2.2 Facsimiles of pages 32r and 33r, reproduced by kind permission of the Master and Fellows of Trinity College Cambridge.

Wittgenstein’s Blue Book

49

the picture of a leaf look like when it does not show us any particular shape, but rather ‘what is common to all shapes of leaf’? So from that first change in the composition of the Blue Book, the picture analogy is retained through to the Philosophical Investigations.26 Such examples not only show the textual as well as the conceptual relationship between the Blue Book and the Philosophical Investigations, but also demonstrate the importance of the survival of the Thouless Blue Book. “POLISHED”, “PRECISE”, “FINISHED” In whatever ways the Thouless copy of the Blue Book elucidates our understanding of the development of Wittgenstein’s philosophy, one fact that remains worthy of remark is that it is clear that Wittgenstein went to some lengths to mould the text in the form that he wanted, as Alice Ambrose informs us. Thanks to the Thouless document a much clearer picture emerges of the process it entailed. As we have seen, Wittgenstein may have carried sections of it around so that he could edit it at an appropriate moment. He did not edit it once, but clearly went through the text a number of times, introducing several layers of amendments, at times with the assistance of Francis Skinner. He changed the format from a series of short dictations to a continuous text and in the process made it look more book-like. Once the text had been duplicated, it appears that he may well have substantially altered the text at the beginning of the second volume and also went through the duplicated copies, making small manuscript changes. And when Coxeter duplicated the text before Wittgenstein felt ready to circulate it, he fell out vehemently with him. It seems clear that he is keen to produce not just an aesthetically pleasing polished text, but a precise one which transmitted his ideas in the most accurate way possible. This may well be part of his nature—the man who had a ceiling in the house that he designed for his sister raised 3 cm. However, I think there is another important aspect to this question, one which relates to the idea of the “finished” text in Wittgenstein’s Nachlass. Many people have commented on aspects of Wittgenstein’s process of developing a text to its finished state. Katherine Morris, for one, points out: “Few philosophers polish their texts as thoroughly as Wittgenstein did in those works that he actually published”.27 Though the Blue Book is not remark-based, the spirit of Schulte’s third criterion, that for a text to be a “work” by Wittgenstein, we should establish “whether the text has undergone a certain amount of stylistic polishing and rearranging of individual remarks showing that there has been some improvement in the direction of enhanced readability and intelligibility”,28 also holds true. Brian McGuinness hints at an indivisibility of style and content: “literary character is not something from which the arguments can be regarded as detachable”.29 In the light of these comments

50

Jonathan Smith

the importance of a document such as the Thouless typescript is further highlighted. It reveals a process of editing and reediting the Blue Book until it was polished and a sufficiently precise exposition of his philosophical views to circulate. So was it “finished”? It seems to me that there is a certain misconception when we talk about a “finished” work by Wittgenstein. It is clear that “finished” and “most developed” are, in most cases, not synonymous, and that a text may not be the most perfect development of a particular idea or ideas and yet be finished according to the purpose that its author has for it. If we now come back to the purpose of the Blue Book and to Ambrose’s and Rhees’ opinions of it, we can see that, while it includes some of the more radical ideas that we associate with the later Wittgenstein, its purpose is that of a pedagogical work, pitched at the level of Wittgenstein’s students. It is not aiming to be a work as deep as the Philosophical Investigations, yet it is a work that Wittgenstein concluded was not only fit for purpose but was also happy to circulate amongst pupils and colleagues alike. To all intents and purposes it was finished. CONCLUDING THOUGHTS The Thouless copy of the Blue Book, then, is a multilayered draft containing numerous amendments that does more than simply corroborate the evidence that we already know about the creation of the work. It not only brings vital new evidence to our understanding of how the text was created but, by allowing us access to the record of the creative process itself, shows how that very text developed. By revealing the actual amendments Wittgenstein made to his text, it helps us elucidate the development of his thought at a crucial time in his life. In doing so, the conceptual and textual relationship between the Blue Book and Philosophical Investigations is highlighted. However, it is extremely important to remember what the purpose of the Blue Book was. It was not intended to be an early version of the Investigations, but to present in a clear form aspects of Wittgenstein’s emerging ideas to his students—“something to carry home with them, in their hands if not in their brains”. The evidence that we now have showing the efforts he made to edit the text of the Blue Book might lead us to conclude that, if it is not the most developed example of his philosophy, it is certainly one of Wittgenstein’s most finished works. Perhaps we might conclude that it is the closest thing to a finished major work after the Tractatus itself.30 REFERENCES Ambrose, Alice. 1977. “The Yellow Book Notes in Relation to the ‘Blue Book’”. Critica: Revista Hispanoamericana de Filosofia 9, no. 26: 3–23. Bouwsma, O. K. 1961. “The Blue Book”. Journal of Philosophy 58: 141–62.

Wittgenstein’s Blue Book

51

Klagge, J. C., and Alfred Nordmann, eds. 2003. Ludwig Wittgenstein: Public and Private Occasions. Lanham: Rowman & Littlefield. Mays, Wolfe. 1999. “Recollections of Wittgenstein”. In Portraits of Wittgenstein, vol. III, edited by F. A. Flowers III, 129–36. Bristol: Thoemmes Press. McGuinness, Brian. 2006. “Wittgenstein: Philosophy and Literature”. In Wittgenstein: The Philosopher and his Works, edited by Alois Pichler and Simo Säätelä, 367–81. Frankfurt: Ontos. ———, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. Morris, K. J., ed. 2004. Wittgenstein’s Method: Neglected Aspects. Essays on Wittgenstein by Gordon Baker. Oxford: Blackwell Publishing. Schulte, Joachim. 2006. “What is a Work by Wittgenstein?” In Wittgenstein: The Philosopher and His Works, edited by Alois Pichler and Simo Säätelä, 397–404. Frankfurt: Ontos. Wittgenstein, Ludwig. 1958, 1969. Preliminary Studies for the “Philosophical Investigations”: Generally known as The Blue and Brown Books. Edited by Rush Rhees. Oxford: Basil Blackwell. ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt: Suhrkamp. ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell.

3

Wittgenstein and His Audience Esotericist or Evangelist? James C. Klagge

Whom was Wittgenstein addressing in his work, what was he trying to accomplish, and how did he try to accomplish it? To engage the Philosophical Investigations it is important to consider these questions. And once we consider the questions, it becomes clear that the answers changed over time. To appreciate the ways they changed over time, we must trace the development of Wittgenstein’s work both as a writer and as a lecturer. While much attention has rightly focused on the genesis of the Investigations in the manuscripts, I contend that we must also appreciate Wittgenstein’s work as a lecturer to fully understand his changing conception of his audience and his task, which then impacts his writing after all. The terms “esotericist” and “evangelist” concern matters of audience. They have not generally been used in characterizing Wittgenstein, but by using them I want to place this discussion in a broader context that draws on a wider range of authors and audiences. The English word “esoteric” comes directly from the Greek word esoterikos, which means: belonging to an inner circle. An early use of the term applied it to the Pythagorean cult of fifth-century Greece, in which followers of the mathematician Pythagoras (c. 570–495 BCE) held certain metaphysical beliefs about the soul in common, and withheld them from the public. Some few scholars (e.g., the so-called Tübingen School) hold that Plato wrote his dialogues for public (exoteric) consumption, but reserved his own, esoteric, beliefs for sharing only with a limited group of followers, perhaps members of his academy, and only orally, since he allegedly did not trust things written down.1 The distinction seems to be employed by Aristotle when he alludes to (his) exoteric writings (exoterica: NE 1102a27, 1140a3; and cf. enkuklika [popular]: 1096a4), which may be his legendary lost dialogues. What is interesting about these allusions is that they imply that what we are reading by Aristotle are in fact his esoteric writings. And this fits with the requirement, which he sets out early in the Nicomachean Ethics (1095a3–12), that his teachings are meant “for those who accord with reason in forming their desires and in their actions”. He specifically declines to instruct the youth (whether young in age or in experience), who “lacks experience of the actions in life”, “tends to follow his feelings”, and “gets no benefit from

Wittgenstein and His Audience

53

his knowledge”. And near the end of the Ethics he returns to this theme (1179b27) when he emphasizes that “the soul of the student needs to have been prepared by habits of enjoying and hating finely, like ground that is to nourish seed”. His articulation of this point in terms of the proper “student” supports the theory that what we are reading are (close to, or based on) notes that Aristotle used at his Lyceum when literally lecturing to students. Similar issues of audience arise in the sayings of Jesus and the gospel writings that recount his life. Jesus regularly taught by telling parables. Concerning the parable of the sower (Luke 8: 9–10; and cf. Mark 4: 10–12 and 34): “His disciples asked Jesus what this parable meant, and he answered, ‘The knowledge of the secrets of the Kingdom of God has been given to you, but to the rest it comes by means of parables, so that they may look but not see, and listen but not understand.’ . . .” This seems rather surprising, and doesn’t make much sense, since parables seem designed precisely for their accessibility. But the idea of a secretive message was reinforced by the discovery in 1945 (at Nag Hammadi, in Egypt) of ancient writings about Jesus that described “the secret words which Jesus spoke” (Gospel of Thomas, opening line). This and other recently discovered writings have come to be called the “Gnostic” gospels, and have provided fragments of views of Jesus rather different (sometimes surprisingly so) from the picture found in the four canonical gospels of the Christian Bible. So the idea of an esoteric message seems to be connected with the notion that certain things should or can only be shared with certain people who are especially deserving or able to appreciate the message. The clearest illustration of this is Aristotle’s lectures, meant for the non-youth whose habits and emotions have been trained to be able to use reason in forming his desires and actions. Only he will benefit from, that is—be able to use, the knowledge gained in the lectures, for “the end is action, not knowledge” (1095a5).2 The clearest indication that Wittgenstein was an esotericist comes in a draft “Zu einem Vorwort [towards a Foreword]” written on 6 or 7 November 1930: This book is written for those who are in sympathy with the spirit [Geist] in which it is written. This spirit is, I believe, different from that of the prevailing European and American civilization. The spirit of this civilization the expression of which is the industry, architecture, music, of present day fascism & socialism is a spirit that is alien [fremder] and uncongenial to the author. This is not a value judgement.3 Important parts of Wittgenstein’s thought here are connected with his views, derived from Oswald Spengler, about the difference between culture and civilization.4 Wittgenstein continues: . . . I contemplate the current of European civilization without sympathy, without understanding its aims if any. So I am really writing for friends who are scattered [verstreut] throughout the corners of the globe.

54

James C. Klagge It is all one to me whether the typical western scientist understands or appreciates my work since in any case he doesn’t understand the spirit in which I write. . . . So I am aiming at something different than are the scientists & my thoughts move [Denkbewegung] differently than do theirs.5

Wittgenstein later comments on these friends (18 January 1931): If I say that my book is meant for only a small circle [kleinen Kreis] of people (if it can be called a circle) I do not mean to say that this circle is in my view the elite of mankind but it is a circle to which I turn . . . because they form my cultural circle, as it were my fellow countrymen in contrast to the others who are foreign [fremd] to me.6 Who were among the small circle of friends Wittgenstein had in mind? Likely he was thinking of friends in Olmütz from the Great War—Paul Engelmann, Heini Groag, Fritz and Max Zweig. In fact, when Engelmann himself described the friends in Olmütz in 1916, he described the group precisely as a “kleinen Kreis”.7 Perhaps also Moritz Schlick, Adolf Loos, Karl Kraus and Wittgenstein’s sister Hermine, in Vienna. Among more current friends, perhaps Piero Sraffa and Nicholas Bachtin in England. These were people Wittgenstein considered to be friends, peers, and sympathetic to him in some sense. While they were not “scattered throughout the corners of the globe”, they were at least scattered throughout the corners of Europe. In contrast, when he speaks of the “typical western scientist”, I think he has in mind a certain spirit that he imagines “the prevailing European and American civilization” to hold—what we might call a sort of “scientism”: overvaluing the role and importance of science in society. Von Wright has plausibly conjectured that Wittgenstein had Rudolf Carnap, and especially Carnap’s “Vorwort” to his Logische Aufbau der Welt, in mind when drafting his own “Vorwort”.8 So Wittgenstein sees himself in the early 1930s as writing for a select group of people who would share with him a certain spirit that he imagines most people would not share. This shared spirit would allow them to understand him, unlike the general public, who would not. He concludes with a reflection on the Foreword: The danger in a long foreword is that the spirit of the book has to be evident in the book itself & cannot be described. For if a book has been written for only a few readers that will be clear just from the fact that only a few understand it.9 Wittgenstein famously insisted, over and over, that his work would not generally be understood. This, I think, makes him an esotericist. This esotericism did not just take hold in the early 1930s. His Tractatus was prefaced with similar thoughts: “Perhaps this book will be understood

Wittgenstein and His Audience

55

only by someone who has himself already had the thoughts that are expressed in it—or at least similar thoughts.—So it is not a textbook.—Its purpose would be achieved if it gave pleasure to one person who read and understood it” (TLP, 3). And the sheer difficulty of understanding the Tractatus made it natural to suppose that there was some esoteric intent. A book on the Tractatus by Alexander Maslow, written in 1933 but not published until 1961, includes the following warning in its Introduction: “. . . the Tractatus is in many of its passages so obscure that it would be impossible, I believe, to gather the fundamentals of Wittgenstein’s view without some help from people who have been initiated into it directly by the author himself”.10 That is the essence of esotericism. This need for special insight, of a different sort, was asserted by another commentator, Roy E. Lemoine: The Tractatus is probably the most significant philosophical document since the Critique of Pure Reason, from which it is in some ways derivative; but it is much harder to read. Even Wittgenstein, as he stated in his foreword, was aware that perhaps only those who had thought similar thoughts would understand him. It may be that my contribution to the study of the Tractatus comes from the fact that my own background is different from that of most scholars and has some similarity to Wittgenstein’s. I have been both a line officer and a chaplain, and I also served in a great war.11 This experience is supposed to account for the fact that his book “departs radically from the traditional interpretations of the Tractatus”. After the end of the Great War, and the publication of his book, Wittgenstein trained as an elementary school teacher and taught in rural Austria for six years. He then lived in Vienna, assisting Paul Engelmann in the design and construction of a house for Wittgenstein’s sister. During this time, 1927–1928, he had occasional meetings with members of the “Vienna Circle”, especially Moritz Schlick, Friedrich Waismann and occasionally others, such as Carnap and Herbert Feigl. Wittgenstein was somewhat reluctant to have such meetings, but eventually consented to some. The circle members were interested to learn more about the Tractatus, which they had studied carefully, but, according to Feigl: “only on relatively rare occasions could we get him to clarify one or another of the puzzling or obscure passages in his work. . . . On occasion, he would read poetry to us (e.g., that of Rabindranath Tagore)”.12 When relating this latter fact McGuinness adds: “usually sitting with his back to the audience”.13 It is hard to imagine a clearer expression of an esoteric attitude. Wittgenstein’s attendance at a talk by the intuitionistic mathematician L. E. J. Brouwer, in March 1928, sparked his interest in discussing philosophy again,14 and led to continued meetings between Wittgenstein and Schlick, with Waismann present to record Wittgenstein’s expositions or elaborations of his thoughts. Carnap was excluded, presumably because his approach to the

56

James C. Klagge

issues was so different: “Although the difference in our attitudes and personalities expressed itself only on certain occasions, I understood very well that Wittgenstein felt it all the time and, unlike me, was disturbed by it. He [Wittgenstein] said to Schlick that he could talk only with somebody who ‘holds his hand’”.15 At Schlick’s urging, Waismann’s notes were meant to be shared with the other circle members as expositions of Wittgenstein’s thoughts. Over the next few years there were continued meetings between Wittgenstein and Waismann connected with plans for Waismann to cooperate with Wittgenstein in writing a book setting out Wittgenstein’s ideas. But despite a great deal of effort on Waismann’s part, these plans came to nothing.16 When Wittgenstein returned to Cambridge in 1929 and began teaching in 1930, his classes were small, and he published almost nothing (save for the “Remarks on Logical Form”). It may have seemed from the outside as though his students were insiders, and his classes took on a reputation of being only for the select few. This secretive reputation seems well-captured by the account in Alan Turing’s biography of how Turing managed to get into Wittgenstein’s class in 1939: There were about fifteen in the class . . . and each had to go first for a private interview with Wittgenstein in his austere Trinity room. These interviews were renowned for their long and impressive silences, for Wittgenstein despised the making of polite conversation. . . . after they had talked about some logic, Wittgenstein . . . said that he would have to go into a nearby room to think over what had been said.17 But I do not accept the implicit suggestion that Wittgenstein allowed only certain people into his classes. Wittgenstein’s courses were almost always publicly announced in the Cambridge Reporter. The only exceptions of which I am aware were classes in Lent and Easter 1938. Redpath reports that “. . . Wittgenstein had started lecturing that week, but didn’t want too many people to come, and so the lectures were not ‘open lectures’, but for people Wittgenstein had ‘decided on’ to attend if they wished”.18 And then also there is the famous case of the Blue Book, which was dictated to a select group of students in 1933–34 from a cancelled class for mathematicians that had grown too large. Two announcements of his classes, for Lent 1930 (the first class), and Michaelmas 1931, in the Cambridge Reporter noted that times were “to be arranged to suit the convenience of students, who are requested to call upon Dr Wittgenstein” at a specified place and time before the start of the term. While this could be the mysterious “private interview”, it sounds more like a mere formality to facilitate scheduling. The only requirement of which I am aware that Wittgenstein placed upon students attending his classes was that they attend for the whole term.19 He did not want casual visitors. As far as visiting professors from abroad, Malcolm mentions that Morris Lazerowitz had sought permission from Wittgenstein to attend his lectures

Wittgenstein and His Audience

57

in 1947–48 and Wittgenstein wrote back to grant it (though, in the end, he resigned and never gave these lectures).20 Perhaps such permission was appropriate for nonstudents, but there is no reason to suppose that Wittgenstein granted or withheld it selectively. As Gasking (who attended lectures in 1939) and Jackson (who attended lectures in 1946–47) wrote: “anyone was welcome who seriously wanted to learn philosophy (and not just to hear Wittgenstein)”.21 While the classes Wittgenstein taught were not large, he did teach at least one class each term, beginning in Lent 1930 and ongoing through the Easter term of 1936. (And then again from Lent 1938 though Michaelmas 1941; a reduced schedule for Lent 1942 through Lent 1943; and a full schedule again from Michaelmas 1944 through Easter 1947.22) These classes tended to have over a dozen attendees (including some dons) and, as I have argued above, the attendees were self-selected, not selected by Wittgenstein. This is an important fact, because it means that Wittgenstein was addressing people that he had not chosen. It is also important that Wittgenstein’s classes were primarily discussions. They were sometimes announced as “lecture & conversation class” and sometimes simply as “conversation class” or “informal discussions”. Although the discussions were often actually monologues, the reality is that Wittgenstein was faced with regular feedback—either in the form of questions from students, or unanswered questions posed by Wittgenstein, or the silence of incomprehension. In any case, Wittgenstein learned how attendees responded, or failed to respond, to his thoughts. This put him in a very different situation from the one he imagined in the draft foreword quoted previously. Far from writing for the “small circle” of cultural friends scattered around Europe, he was now faced with dozens of students, term after term, who were “foreign” to him. He came to know how these others thought, and how that affected the issues he wanted to address. He began to identify those thought patterns that ran contrary to his own (or, at any rate, to those he preferred), and he began to address them. In stark contrast to Aristotle, who would only lecture to students whose habits were already trained to respond to reason, Wittgenstein found himself lecturing to students whose habits of thought were resistant to his ways of thinking. He would no longer proceed esoterically.23 This transformation is a conjecture on my part, and the evidence I have marshaled is somewhat limited. For example, in notes by students from his lectures it is not until Easter 1931 term that he mentions being “tempted” or the “resignation of temperament”.24 Then in 1932 mentions of temptation and other kinds of noncognitive factors become increasingly common. What Wittgenstein is doing is beginning to appreciate and engage with differing attitudes. Exactly how and where this transformation can be evidenced is research that I have not done, but that there was such a transformation is clear.25 The Philosophical Remarks, a manuscript from 1930 to which a later version of the above-quoted foreword was attached, hardly even alludes

58

James C. Klagge

to issues of temperament. Wittgenstein only once mentions what “we are tempted to say”.26 And there are perhaps half a dozen other such confessions in Philosophical Remarks. Josef Rothhaupt argues, however, that these prefatory remarks were never intended for the Philosophical Remarks text to which they were attached by editors. Rather, they are more likely intended for a selection of remarks that Rothhaupt labels the “Kringel-Buch”.27 A survey of the remarks intended for the Kringel-Buch, however, shows the same result—that they do not presuppose or address wayward temptations.28 But when Wittgenstein comes to write the foreword for the Philosophical Investigations, in 1945, he is explicitly no longer writing for the select few who think like he does. About his thoughts: “I make them public with misgivings. It is not impossible that it should fall to the lot of this work, in its poverty and in the darkness of this time, to bring light into one brain or another—but, of course, it is not likely”.29 Wittgenstein is pessimistic about how successful he might be, but his aim now is clearly to engage ways of thinking different from his own. So it is that in Part I of the Philosophical Investigations, Wittgenstein provides us with a running commentary (of well over a hundred points) on what produces the philosophical confusions we get into, and what the problems are with appreciating his resolutions. A survey of Part I of the Investigations shows us that philosophical problems arise or remain because of: 1. What . . . forces itself on us, holds us captive, demands an answer, must be, leads us, we can’t help, or no one would say (14 times); 2. What we are . . . tempted, seduced, bewitched, or dazzled by (19); 3. What . . . suggests itself, strikes us, occurs to us, or impressions we are under (7); 4. How things look to us (2); 5. What we find . . . surprising, convincing, senseless, ludicrous, sensible, or matter-of-course (8); 6. Our . . . compulsions, needs, urges, wants, tendencies, inclinations, expectations, or prejudices (28); 7. What we . . . notice, can get ourselves to think, can be satisfied with, only think of, overlook, don’t realize, fail to see, or forget (14); 8. What we would like (6); 9. What we . . . are committed to, choose, decide, allow, or refuse (6); and 10. How we . . . look at, or represent things (5). These noncognitive tendencies in us can apply to a great variety of issues, leading to many different philosophical problems. The sum of such tendencies could be said to constitute a temperament—a spirit of the times. I conjecture that it was Wittgenstein’s teaching in the 1930s that brought him to face and engage these differences, and led him to try to address them.30 He would no longer approach things esoterically. Instead, he was trying to

Wittgenstein and His Audience

59

figure out how to bring about the changes needed to appreciate a different way of viewing philosophical issues.31 He had become an evangelist.32 Of course, “evangelism” has religious connotations, which I do not mean to invoke here. But the term still seems to me to be appropriate, because what Wittgenstein intended to bring about was not simply a change in beliefs. He saw the changes needed to go more deeply. In the opening lines of his chapter on “Philosophy” in the so-called Big Typescript that he compiled in 1933, he wrote: The difficulty of philosophy [is] not the intellectual difficulty of the sciences, but the difficulty of a change of attitude. Resistances of the will must be overcome. As I have often said, philosophy does not lead me to any renunciation, since I do not abstain from saying something, but rather abandon a certain combination of words as senseless. In another sense, however, philosophy requires a resignation, but one of feeling and not of intellect. And maybe that is what makes it so difficult for many.33 What he has “often said” can’t refer to anything other than his lectures, and the “many” for whom this is “so difficult” can only be his students. In Lee’s notes of Wittgenstein’s lectures in Easter term 1931, we find the line: “Doing philosophy may perhaps mean resignation of temperament, but never of intellect”.34 His work, then, in the Philosophical Investigations, is to see how he might bring this resignation about. Since his goal was articulated by him in the 1930 draft as pertaining to the sort of “spirit” one has, it seems appropriate to use a term like “evangelism” after all. How Wittgenstein went about his evangelizing is a question I will not try to address extensively. I have emphasized that he was concerned with the noncognitive aspects of temptation and other attitudes towards the views in question. I believe that this was motivated largely by his encounters with students. In the Philosophical Investigations, and the drafts leading up to it, Wittgenstein often expressed these temptations and other wayward approaches in quotation marks, or between dashes. Stanley Cavell has written: “The voice of temptation and the voice of correctness are the antagonists in Wittgenstein’s dialogues”.35 And further research has claimed to identify a “commentator”—a third “ironic” voice—in addition to the voices variously identified as “narratorial”, and “interlocutory”.36 There is no uniformity to how Wittgenstein expresses these voices, sometimes invoking “you”, sometimes “I”, sometimes “us”. Jane Heal writes: “sometimes it is part of a dialogue, in that it is directed at the interlocutor, while at other times it represents simply the flow of Wittgenstein’s own ideas. . . . Some stretches can be read either way”. But it is natural, once one has thought of the possibility, to read the conversation as one imitating a classroom discussion. Even if the voice of temptation had once lived in Wittgenstein’s own head, Heal argues, he often “represents himself . . . as no longer impelled to say

60

James C. Klagge

those things but rather as recognising sympathetically the impulse which another is there represented as experiencing”.37 Thus the dialogical character of the Investigations seems plausibly derived from the classroom setting, and aimed at diagnosing and treating the temptations of the wide variety of those present in that setting.38 How we conceive of the dialogical character of the Investigations could well depend on what sort of picture of Wittgenstein holds us captive. If one is captivated by a picture of Wittgenstein alone at his desk, agonizing over a subject, then it is natural to think of the voices largely as expressions from within himself. But if one thinks instead of Wittgenstein in front of a classroom of students, then it may seem natural to think of the voices as arising from the students. Heal concludes that Wittgenstein “presents himself, pretty much throughout the Investigations, as having, to some extent at least, succeeded in escaping from the false pictures . . . and from which he hopes also to release his reader”. This is how we tend to think of ourselves in our role as teacher. One of Wittgenstein’s students from 1938—James Taylor—went on to graduate school in philosophy at the University of California at Berkeley. They corresponded, and in one of his letters back to Wittgenstein reporting on the department he wrote (24 September 1938): “I haven’t done any missionary work”. This would seem to imply that there was a common sense that something could now be accomplished, though surely the term “missionary” was a jest, whether it originated from Wittgenstein or from Taylor. But the term “missionary” does clearly convey a sense of being among those who are quite different in important ways—and wanting to do something about it. When Taylor mentioned the missionary work, he confessed “am quite aware I’m not good enough to”.39 What did Wittgenstein think that he could accomplish, as the missionary, preaching against the idols? Discussing in 1946 an upcoming Joint Session of the Mind Association and the Aristotelian Society meeting, Karl Britton described how Wittgenstein “railed against professional philosophers, mourned the present state of philosophy in England and asked: ‘What can one man do alone?’”40 By now it is clear that Wittgenstein finally felt that he had failed. The sense of failure is already evident in the 1945 preface quoted above, and even in a 1938 draft preface: “I don’t dare to hope that it should fall to the lot of this inadequate work to throw light into this or that brain, in our dark age”.41 But he is, according to these prefaces, trying to evangelize. During the break before Easter term, 1947, what would turn out to be his last term teaching, Wittgenstein reflected on the difficulty of trying to change people’s philosophical views by writing or arguing (13–14 April 1947): “It is as though I wanted to change men’s and women’s fashions by talking”. Perhaps recalling his 1931 strategy to deal with common philosophical problems by erecting “signposts . . . to help people past the danger points”, he now reflects: “my warnings are like the posters at the ticket

Wittgenstein and His Audience

61

offices at English railway stations ‘Is your journey really necessary?’ As if anyone reading that would say to himself ‘On second thoughts, no’”. The philosopher says “Look at things like this!”—but first, that is not to say that people will look at things like this, second, he may be altogether too late with his admonition, & it’s possible too that such an admonition can achieve absolutely nothing & that the impulse towards such a change in the way things are perceived must come from another direction. . . . The diagnosis he proposes sounds positively political: “It is not by any means clear to me that I wish for a continuation of my work by others, more than a change in the way we live, making all these questions superfluous. . . .”42 Wittgenstein finally quit teaching in 1947. He explained his resignation to Bouwsma as necessary to finish his book, but also as a result of pessimism about his role as a teacher.43 He told Drury in 1949: “My thinking is not wanted in this present age, I have to swim so strongly against the tide. Perhaps in a hundred years people will really want what I am writing”.44 And in a draft of yet another prefatory comment (8 January 1948) he wrote: “With repugnance I hand over the book to the public. The hands in which it will fall are mostly not the ones in which I like to imagine it. May it, I wish, soon become entirely forgotten by the philosophical journalists, and thus perhaps remain preserved for a better kind of reader”.45 Here we find him jettisoning his evangelism and returning to esotericism—his audience is not “the public” or the “philosophical journalists”, but “a better kind of reader”, who presumably understands what he is up to. He awaits “a change in the way we live, making all these questions superfluous”. When Frank Ramsey met with Wittgenstein in 1923, after the publication of the Tractatus, to discuss the book with him, he discovered this same sort of orientation to the future—still esoteric, but without a currently existing inner circle. Ramsey wrote to his mother (20 September 1923): “His idea of his book is not that by reading it anyone will understand his ideas, but that some day someone will think them out again for himself, and will derive great pleasure from finding in this book their exact expressions”.46 My conjecture is that the writing he did once he lost faith in and gave up his teaching may show a move away from addressing the noncognitive aspects of temperament as they bear on philosophical puzzles.47 So my answer to the question posed in the title of this paper is: “Both— first one, then the other, and then the first again”. If I am right about the transitions I have outlined, then this would provide a basis for another way of talking about stages in Wittgenstein’s philosophy. Initially scholars distinguished between an early and late Wittgenstein, based primarily on the differences of method and of substance in the Tractatus and then the Investigations. That approach has been criticized from two directions, some adding a third, “middle”, Wittgenstein (and even a fourth, post-PI), and others wanting to reestablish a unity all along. The grounds for making or denying

62

James C. Klagge

stages have been somewhat unclear and, indeed, variable. My suggestion is not wholly separate from these, but asks us to focus on whom Wittgenstein takes himself to be addressing and how, and on what he is trying to accomplish. From this point of view, the Tractatus and the work of the early 1930s is esoteric; the work from about 1932 or so becomes evangelical until he loses confidence in his approach before or around the time he quit teaching in 1947, and then again becomes esoteric. These stages do not involve sharp dividing lines, nor do they fit with previous maps of the stages, but they do focus attention on issues of importance to us and to Wittgenstein. And in particular they draw on Wittgenstein’s own ways of characterizing what he was doing. Finally, they offer an agenda for looking at Wittgenstein’s work in ways that have not so far received much attention. In that respect I hope that this may be a fruitful idea for future work. One implication of my stages is that most of what is familiar to us in Wittgenstein’s later philosophy, especially (Part I of) the Philosophical Investigations, falls into the evangelical stage. While this is a new term of description, it helps to focus our attention on what Wittgenstein was trying to do, and how he was trying to do it. I think it is important to see that Wittgenstein took evangelism seriously, even while he did not see how to succeed and eventually admitted failure. It is important because proponents of Wittgenstein’s views too often proceed as though his views and arguments should “take hold” just as a result of being presented, and resistance can be addressed by louder or clearer restatement of the view. But that is not at all how Wittgenstein saw it. Wittgenstein writes (13–14 April 1947): Quite different artillery is needed here from anything I am in a position to muster. Most likely I could still achieve an effect in that, above all, a whole lot of garbage is written in response to my stimulus & that perhaps this provides the stimulus for something good. I ought always to hope only for the most indirect of influences.48 Of course then we would want to ask who it is that is writing the “garbage”. Perhaps that was all the Wittgensteinian publications! And then, what is the “something good” that may come from it? Something itself written? No— more likely something else. But what?49

REFERENCES Anscombe, G. E. M. 1981. Metaphysics and the Philosophy of Mind: The Collected Philosophical Papers of G. E. M. Anscombe, vol. II. Minneapolis: University of Minnesota Press. Baker, Gordon, ed. 2003. The Voices of Wittgenstein: The Vienna Circle. New York: Routledge. Bouwsma, O. K. 1986. Wittgenstein: Conversations, 1949–1951. Edited by J. L. Craft and Ronald E. Hustwit. Indianapolis: Hackett.

Wittgenstein and His Audience

63

Britton, Karl. 1999. “Portrait of a Philosopher”. In Portraits of Wittgenstein, vol. 2, edited by F. A. Flowers III, 205–11. Bristol: Thoemmes Press. Carnap, Rudolf. 1999. “Wittgenstein and the Vienna Circle”. In Portraits of Wittgenstein, vol. 2, edited by F. A. Flowers III, 172–6. Bristol: Thoemmes Press. Cavell, Stanley. 1976. “The Availability of Wittgenstein’s Later Philosophy”. In Must We Mean What We Say?, 44–72. New York: Cambridge University Press. Drury, Maurice O’Connor. 1984. “Conversations with Wittgenstein”. In Recollections of Wittgenstein, edited by Rush Rhees, 97–171. New York: Oxford University Press. Engelmann, Paul. 1967. Letters from Ludwig Wittgenstein: with a Memoir. Oxford: Basil Blackwell. ———. 2006. Wittgenstein—Engelmann: Briefe, Begegnungen, Erinnerungen. Edited by Ilse Somavilla. Innsbruck: Haymon Verlag. Feigl, Herbert. 1969. “The Wiener Kreis in America”. In The Intellectual Migration: Europe and America, 1930–1960, 630–73. Cambridge, MA: Belknap Press of Harvard University Press. Gasking, D. A. T., and A. C. Jackson. 1999. “Ludwig Wittgenstein”. In Portraits of Wittgenstein, vol. 4, edited by F. A. Flowers III, 141–6. Bristol: Thoemmes Press. Grattan-Guinness, I. 1992. “Russell and Karl Popper”. Russell 12, no. 1: 3–18. Heal, Jane. 1995. “Wittgenstein and Dialogue”. In Philosophical Dialogues: Plato, Hume, Wittgenstein, edited by Timothy J. Smiley, Proceedings of the British Academy 85: 63–83. Hodges, Andrew. 1984. Alan Turing: The Enigma. New York: Simon & Schuster. Klagge, James C. 2003. “The Wittgenstein Lectures”. In Public and Private Occasions, edited by James C. Klagge and Alfred Nordmann, 331–72. Lanham: Rowman and Littlefield. ———. 2011. Wittgenstein in Exile. Cambridge, MA: MIT Press. Krämer, Hans Joachim. 1990. Plato and the Foundations of Metaphysics: A Work on the Theory of the Principles and Unwritten Doctrines of Plato With a Collection of the Fundamental Documents. Edited and Translated by John R. Catan. Albany: SUNY Press. Lemoine, Roy E. 1975. The Anagogic Theory of Wittgenstein’s “Tractatus”. The Hague: Mouton. Lewy, Casimir. 1976. Meaning and Modality. New York: Cambridge University Press. Malcolm, Norman. 1984. Ludwig Wittgenstein: a Memoir. New York: Oxford University Press. Maslow, Alexander. 1961. A Study in Wittgenstein’s Tractatus. Berkeley and Los Angeles: University of California Press. McGuinness, Brian. 1979. “Editor’s Preface”. In Wittgenstein and the Vienna Circle, 11–31. Oxford: Basil Blackwell. ———. 2002a. “Relations with and within the Circle”. In Approaches to Wittgenstein: Collected Papers, 184–200. New York: Routledge. ———. 2002b. “In the Shadow of Goethe: Wittgenstein’s Intellectual Project”. European Review 10, no. 4: 447–57. ———, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. ———. 2011. “Waismann: The Wandering Scholar”. In Friedrich Waismann: Causality and Logical Positivism, edited by Brian McGuinness, 9–16. New York: Springer. Popper, Karl. 1945. The Open Society and Its Enemies, Vol. II: The High Tide of Prophecy. Princeton: Princeton University Press. Redpath, Theodore. 1990. Ludwig Wittgenstein: A Student’s Memoir. London: Duckworth.

64

James C. Klagge

Rothhaupt, Josef. 2010. “Wittgenstein at Work: Creation, Selection and Composition of ‘Remarks’”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 51–63. Basingstoke: Palgrave Macmillan. Stern, David. 2004. Wittgenstein’s Philosophical Investigations: An Introduction. New York: Cambridge University Press. Venturinha, Nuno, ed. 2010. “Wittgenstein’s 1938 Preface”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 182–8. Basingstoke: Palgrave Macmillan. von Wright, G. H. 1993. Myten om Framsteget. Helsinki: Söderströms. Wittgenstein, Ludwig. 1958. Preliminary Studies for the “Philosophical Investigations”: Generally known as the Blue and Brown Books. Edited by Rush Rhees. Oxford: Basil Blackwell. (BB) ———. 1961. Tractatus Logico-Philosophicus. Translated by D. F. Pears and B. F. McGuinness. London: Routledge and Kegan Paul. (TLP) ———. 1973. Letters to C. K. Ogden. Edited by G. H. von Wright. Oxford: Basil Blackwell. (LCO) ———. 1975. Philosophical Remarks. Edited by Rush Rhees. Translated by Raymond Hargreaves and Roger White. Oxford: Basil Blackwell. (PR) ———. 1979. Wittgenstein and the Vienna Circle. Translated by Joachim Schulte and Brian McGuinness. Oxford: Basil Blackwell. (WVC) ———. 1980a. Culture and Value. Edited by G. H. von Wright in collaboration with Heikki Nyman. Translated by Peter Winch. Chicago: University of Chicago Press; 1998. Culture and Value. Edited by G. H. von Wright in collaboration with Heikki Nyman. Revised edition of the text by Alois Pichler. Translated by Peter Winch. Oxford: Blackwell. (CV) ———. 1980b. Wittgenstein’s Lectures: Cambridge, 1930–1932. From the Notes of John King and Desmond Lee. Edited by Desmond Lee. Totowa, NJ: Rowman & Littlefield. (LWL) ———. 1993. Ludwig Wittgenstein: Philosophical Occasions 1912–1951. Edited by James C. Klagge and Alfred Nordmann. Indianapolis: Hackett. (PO) ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. (PI) ———. Forthcoming. Kringel-Buch. Edited by Josef Rothhaupt. Berlin: Walter de Gruyter.

4

The Philosophical Investigations and Syncretistic Writing Alois Pichler

SYNCRETISTIC WRITING In Chapter IV of his Schreiben und Denken, the Austrian linguist Hanspeter Ortner distinguishes and describes ten writing strategies (“Schreibstrategien”). One of them is “syncretistic writing”.1 A simple application of Ortner’s definition and description of syncretistic writing to the genesis of the Philosophical Investigations (PI) makes clear that the PI can be said to be of syncretistic origin.2 Wittgenstein’s writing of the PI3 can be characterized by Ortner’s eight features of syncretistic: his writing (1) hops all over the place (“Sprunghaftigkeit”); (2) combines disparate elements from his writings (“Verbindung von weit Auseinanderliegendem”); (3) is semantically open, under-determined and under-determining (“Unterdeterminiertheit und semantische Offenheit”); (4) postpones gestalt-formation/ elaboration (“Aufschub der Gestaltbildung”); (5) invites and offers many opportunities for creative ideas (“viele Chancen für und Einladungen an den kreativen Einfall”); (6) gives freedom to choose the points of departure and reference (“Freiheit bei der Wahl des/der Startpunktes/e und des/der Gesichtspunktes/e”); (7) is hierarchically under-determined (“hierarchische Unterbestimmtheit”); (8) works side-by-side with the already “finished” and the newly begun, which implies long text building processes and parallel operations (“lange ‘Bauzeit’ und Nebeneinander von Fertiggestelltem und Neubegonnenem”).4 In the following, I will try to show in more detail how the genesis of the PI is characterized by these eight features. First, the writings that constitute the PI’s genesis are characterized by a strong discrepancy between the sequence of remarks in their textual order and the sequence of remarks in their physical order. Texts are put together from chronologically and argumentatively dispersed units. One example is Wittgenstein’s rearrangement of remarks from an earlier dictation (TS 208) into a new text in 1930 (TS 209, published by Rush Rhees as Philosophical Remarks). In this new text, he abandoned both the original argumentative order and the chronological order and did not necessarily obey the criteria of consistency and coherence, not even on linguistic levels such as demonstrative reference. The work that

66

Alois Pichler

emerges is seen by many as an unordered agglomerate of remarks, although I have argued that this view can be challenged.5 The second example is the revision and rearrangement of the so-called Big Typescript (TS 213) in 1933–34, which is paradigmatic in its triple use of (1) the text in the typescript, (2) the handwritten revisions of it in the typescript, and (3) text in other manuscripts. In his edition of the Philosophical Grammar (1969), Rush Rhees has tried to take this complicated network of revisions into account and to follow it painstakingly and faithfully; by looking at the manuscript sources for this edition6 one realizes how much “hopping all over the place” was going on in the originals. Thirdly, MS 142, the “Urfassung” of the PI, was produced in 1936–37 from remarks stemming from different places in manuscripts and typescripts and various loci of discourse. MS 157b, 13v, contains a list of references to pages in TS 213 from which parts of the text were to be taken to write the “philosophy chapter” of this first PI version; other sources include MS 140 (last page), MS 152, MS 156a, MS 156b and MS 157a, all yielding materials, lists and drafts for the text of MS 142. The final example is TS 228: in the later stages of the PI genesis, Wittgenstein selected about 400 remarks from this typescript to include in TS 227, the typescript used as the printer’s copy for the PI. The presence of these features alone, hopping all over the place and combining disparate elements, would not be sufficient to establish syncretistic origin. However, the PI genesis is also marked by other features of the syncretistic such as openness and semantic under-determinacy. This includes forms of textual variation: when writing the PI, Wittgenstein makes heavy use of (diachronic and synchronic) variants (the latter, typically occurring within one and the same remark, I call “alternatives”). Three examples, all taken from the genesis of the first sections of the PI, may suffice. In July 1931, Wittgenstein embarks on a long-lasting discussion of “the mistake” Augustine made in the description of how he had learnt language (Confessiones I/8). But his account of what this mistake actually consisted of varies from text to text. Even in the PI it is not clear what the mistake was that Augustine made: one part states that Augustine’s description of learning language does not pay sufficient attention to the fact that different word classes exist since it suggests that there is only one word class, namely the class of names (PI, §1). Other parts attribute to Augustine the belief that all words are names (PI, §6). PI §6 states that Augustine does not even describe language correctly in relation to names; in contrast to this, however, §§3–4 concede that Augustine does correctly describe language with regard to names. Again other parts suggest that Augustine’s description reveals a mistake of attitude rather than a mistake of reasoning: neither a wrong position nor a faulty generalization lies at the heart of Augustine’s description, but rather an attitude of neglect or of drawing our attention to only one aspect (PI, §1). The second example concerns the issue of how, in different variants, Bedeutung (meaning) is related to Hindeuten (pointing):7 in the first versions, Hindeuten seems to be the basic notion from which Bedeutung is derived,

The Philosophical Investigations and Syncretistic Writing

67

but the PI “Urfassung” suggests the idea that Bedeutung is the primitive form. The third example again relates to Augustine’s mistake: What lies behind it? Is it such a simple thing as a wrong description of language learning, or is the wrong description already the expression and result of a wrong, “primitive” conception of the world? Different answers are given in the genesis of the PI. Wittgenstein’s investment in producing textual variants is impressive.8 Both the diachronic variants at different points in the history of the texts and the alternatives within a remark document and bear semantic under-determinacy and openness. They can introduce new perspectives on a subject or open up different lines for proceeding further into it. And, as we have seen, not even the “final” PI is free from semantic openness, nor from either textual or conceptual ambiguity.9 As the exact meaning of remarks is often left open, or as the meaning of the text is not always fixed but multiplied and varied, so too is the form of the work: the genesis of the PI is, in some places, characterized by conscious enrichment with further gestalts, and generally a postponing of the fixation of gestalt. It is clear that after his return to Cambridge in 1929 Wittgenstein aimed to produce a new book that was to be published,10 but that he did not come up with a satisfactory vision of the form of this book until November 1936. Even after having gained such a vision, he did not succeed in bringing the work to publication in his lifetime. One reason for this was surely that he never had the feeling of having put the absolutely right content into the absolutely right form. His biggest problem was to find a way to avoid both the Scylla of a collection of aphorisms and the Charybdis of a closed hierarchy and taxonomy. In the Tractatus he had achieved cohesion and focus through strong ties of textual connectors (with a is b, b is c, c is d . . . as a fundamental gestalt principle)11 and a hierarchical tree-structure12—but the PI would become an “album” of fragments.13 We find a few other attempts at structuring that lie in between these two forms, among them two more academic ones: the hierarchical one in TS 213 (1933–34) and the attempt at rigid step-by-step linear structuring in the Brown Book complex (1935–36). Wittgenstein abandoned both these attempts at more standard academic forms, the latter with harsh criticism in late autumn 1936 (see MS 115, 292). What came after the Brown Book? The album form. But is the PI album a form whose gestalt principles are easy to recognize? Although I find it very difficult, I will later in this chapter make an attempt to see and understand the form principles of the PI, considering the fact that Wittgenstein calls his PI an “album”. The PI preface clearly tells us that Wittgenstein had at different times held different views regarding the form his book should take. It is also clear that he had a vision from early on: “in the book the thoughts should proceed from one subject to another in a natural, smooth sequence” (PI, 3). The author of the PI abandoned that vision in late 1936, or he reconceived what “natural, smooth sequence” should mean. What he ended up with was a form that permitted him “to travel criss-cross in every direction over a wide

68

Alois Pichler

field of thought” (PI, 3). Thus, he settled on a form which not only permitted him to proceed with very little closed gestalt but also demanded further gestalt-shaping from the reader. The preface makes an important point: it is the philosophical investigation itself which compels us to such philosophical crisscross travelling. The PI’s nonlinear album form is thus a response to a requirement. But neither the postponing of gestalt-formation nor the choice of an open gestalt such as the album prevents or frees the writer from at least sometimes drafting text and content arrangement and sequence. Therefore, lists of remarks and content tables find their natural place in the PI genesis also after 1936. An album does not lack a gestalt, but rather has a very specific gestalt, a gestalt that is crucially different from a textbook. A form which develops and promotes openness and crisscrossing is not a form without structure, but rather a form with a structure that does exactly that: it develops, maintains and promotes openness and crisscrossing. This form could not be a hierarchical one. While the genesis of the Tractatus may be seen as crucially involving attempts at diminishing hierarchical underdeterminacy, the genesis of the PI is characterized by a departure from fighting hierarchical under-determinacy. Therefore, the album form is not to be regarded as a shortcoming in achievement, or as the achievement of no form, but as an achievement of a special kind. The alternative to hierarchy and linearity is not no form, but just a very different form which shall enable both the author and the reader to do in better ways what they can also do with hierarchy and linearity, and, in addition, to do other things which they cannot do with hierarchy and linearity. Clearly, invitations to and opportunities for creative ideas as well as giving (to both the author and the reader) freedom to choose the points of departure and reference needed are better taken care of by a crisscross album form than by a hierarchical and linear form.14 The PI genesis is characterized by the coexistence of the already “finished” and the newly begun. While indeed some of the remarks that made it into the PI underwent little textual change since their first drafts and were thus fully formed on their first building, others were subjected to a long process of construction and reconstruction: perhaps they were first erected in one form and then taken apart and combined anew with others—all this over many years. Let us look again at the first sections of the PI, more specifically §§1–4, which have subsisted as a unit since the beginning of 1931. None of the sentences of the “final” TS 227 version had been there from the beginning in 1931, but some of them had still been “sort of there” and underwent relatively little change thereafter.15 In terms of changes on conceptual levels, the first idea of bringing in Augustine’s description was to use it as a positive object of comparison: Augustine’s description is too simple a description of (how we learn) language, but still a correct description of a language simpler than ours. This use of and perspective on Augustine’s description is also present in the last version. From 1936 onwards,16 however, although the earlier approach was not given up, a different and clearly

The Philosophical Investigations and Syncretistic Writing

69

negative tone becomes more dominating: Augustine’s description is the expression and documentation of a primitive view of language (or even the world) which is at the basis of much philosophical confusion and leads to many problems. Augustine’s description and the views it embodies become one of the main targets, if not the key targets, of the PI. A different case, however, is the history of PI §§89–133. While one can say that most of PI §§1–4 existed from the beginning as a discursive unit, this never seems to have been so for the remarks from the philosophy chapter before 1937 (and it is disputed whether they even are actually in the PI17). Some of this “chapter’s” remarks are first found in separate and dispersed places and were only bundled together in the making of TS 213.18 They did not, however, pass over from TS 213, the Big Typescript, to the PI without significant change, but in the spring of 1937 underwent a thorough revision which focused on what particular conception of philosophy they (and the PI “Urfassung” as a whole) should promote. Clearly, the author believed that they—in their TS 213 version—had carried a wrong conception. In the course of this revision, some remarks and passages were left out while new ones emerged and were included.19 In this section I have tried to show that it makes good sense to look at the origin of the PI as being strongly syncretistic in Hanspeter Ortner’s terms. However, whether one is actually willing to acknowledge the features of the PI genesis described here as characteristic for the writing and text work which led up to Wittgenstein’s PI may depend on one’s view of the question as to whether the function of syncretistic writing can be more than just preliminary. Additionally, it may also depend on one’s view and evaluation of Wittgenstein’s philosophical programme. Ortner himself classifies Wittgenstein as a “puzzle-writer”.20 I think he does so, firstly, on the basis of attributing to Wittgenstein a vision of philosophy which cannot permit the syncretistic to be more than a stage to be overcome, and, secondly, on the basis of his (Ortner’s) own specific view of what both philosophy and scholarly writing are or should be. While Ortner may admit that the genesis of the PI bears marks of the syncretistic, he tries to find in this genesis something, in his view, “better” than merely syncretistic writing, precisely because he considers the syncretistic as something to be overcome and as perceived by Wittgenstein himself as something to be overcome. In contrast to this, I hold that an analysis of the PI genesis which is “ideologically” unbiased to the largest possible degree will end up classifying the author of the PI as a syncretistic writer. I additionally want to argue that the syncretistic— in the sense introduced by Ortner, but applied not only to text production (writing) but also to text forming (the making of works)—is in tune with Wittgenstein’s own understanding and description of what philosophy and philosophical investigation is. Consequently, I not only regard Wittgenstein as the author of the PI as a syncretistic writer, but I also look at the PI’s philosophy as “syncretistic philosophy”: a philosophy which defends and promotes the features of the syncretistic as natural and necessary aspects

70

Alois Pichler

of doing philosophy. This is indeed in contrast to Ortner’s normative conception of writing which embodies the ideal of linearly ordered writing and composition. The opposition between the syncretistic on the one hand and Ortner’s ideal on the other is brought up by Ortner himself; as he explicitly says, the syncretistic is the opposite of “discursive linearity”.21 Thus, both Ortner and the PI author see a contrast between the two: Ortner values the second higher than the first; interestingly, the author of the PI, however, welcomes and requires, even for the purpose of work forming, the first and dismisses the second. While Ortner himself regards Wittgenstein as a “puzzle-writer”, he still develops a concept of the syncretistic which is fully applicable to the PI genesis. As said, I disagree with Ortner’s classification of Wittgenstein as a “puzzle-writer” and assume that our disagreement stems from different views regarding what syncretistic writing and philosophy, as conceived by the Wittgenstein of the PI, can and should achieve, and how. At the same time, I stand entirely on the shoulders of Ortner’s work and use his concept of syncretistic writing not only for the study of the PI genesis but also of its form. Looking at the relation between the genesis and the form of a work, we are faced with a potential asymmetry: the fact that the genesis of a work is syncretistic does not imply that the work itself is syncretistic too. Many writers work syncretistically,22 but their results will still not be syncretistic. The genesis of some of Wittgenstein’s Tractatus was definitely syncretistic—but the final work was not. Therefore, the fact that the PI has not only a syncretistic origin, but also a syncretistic form deserves considerable attention. In the next section, I will first elaborate on the claim that the PI really is of syncretistic form. Then I will (try to) give further reasons why the PI has received this form. I will, however, in this chapter not be able to do more than just scratch the surface of the big question concerning what we should make of the syncretistic form in philosophy, especially in the context of today’s academic philosophy. SYNCRETISTIC FORM “Syncretistic” literally means “combining”: the combining of beliefs and views, but also practices (especially in the field of “Weltanschauung”), in ways that can even be incoherent or inconsistent. A work combining a wide range of different and varied methodological approaches and stylistic features such as the PI23 is syncretistic in this nontechnical sense. Ortner acknowledges this use of “syncretistic”.24 In addition, “syncretistic” has for Ortner a specific sense: lacking linear-discursive order.25 This is an issue explicitly addressed by the PI as well and clearly expressed in its preface. The preface stresses two points: not only does the PI lack linear-discursive order, but it even opposes it; and it does so with a fundamentum in re, not only in homine. The PI does not literally speak of “syncretistic”, but of “album”. How do “album” and the syncretistic relate to each other? Does “album”

The Philosophical Investigations and Syncretistic Writing

71

add something which is not contained in “syncretistic”? In the following section, I will first reflect upon what an album is and then see how this reflection can relate to the syncretistic. I start by identifying characteristic features of an album (e.g., a photo album): • Albums are collections of units assembled for later inspection. • Albums assemble what already exists in its own right before the album; what makes an album specific and new is the particular arrangement or composition of the items. • Albums do not need to be complete and comprehensive in order to be albums. • There can be different ways of arranging the items in an album (the chronological one, the thematic one . . .), and different ways of reading and looking at the album. • Albums can give different and diverging views of one and the same object. No statement needs to be made about whether a particular view is correct, and the crisscross relations between these views is often under-explicated and under-documented. • Albums can contain representations of items from a wide range of different situations without needing to make something out of this diversity. In what sense can the PI be an album? This I understand to be a question about the form of the PI. So, to answer it, let us first try to identify central features of the form of the PI. Then, in a subsequent step, we can see whether PI features of form match album features: • The PI text is composed of remarks (mostly fragments, not aphorisms).26 • The PI text is the result of a careful arrangement of what has been elsewhere before. • In the PI, not all the topics which one might expect to be discussed are in fact discussed, and the topics discussed are not discussed comprehensively. • In the PI, the same topic is often dealt with in many different places and from different perspectives—and one and the same remark often addresses several topics. • The PI displays incoherence, inconsistency, ambiguity and textual openness; the internal structure of cross-reference is under-explicated and under-documented in the PI. • The PI contains both real-life and invented cases/samples.27 Every analyst of the PI’s form should be able to agree with this list of form features of the PI. While one need not think of an “album” at all in order to come up with this list, I think that many will still be able to agree that the items in the list can be regarded as exemplifications of the more general album features. We can thus say that the two lists match well and that there

72

Alois Pichler

seems to be an intimate relationship between the form of the PI and what we understand by “album”. But can we also say that the PI’s album form stands in an intimate relationship to the syncretistic? Is there something which all three, the PI, albums and the syncretistic, have in common? According to Ortner’s own understanding, it is nonlinearity which characterizes the syncretistic, and, according to Wittgenstein, it is the nonlinear crisscross form which also characterizes his PI “album”. I do not think that albums by their very nature are nonlinear, but the album form definitely invites nonlinear composition and also reception. It seems to me now that what I have earlier called “album in the wide sense”,28 and what includes stylistic multiplicity and polyphony, is better described in terms of the syncretistic. Yet there is another feature of the PI which is well captured by “syncretistic”: the heuristic and creative functions of its form. Again, Ortner confines this function of syncretistic writing to writing, the process of text production only,29 but it seems to me that we need not and should not confine it in this way. It can be a feature also of the resulting work and the way it shall be read: “I should not like my writing to spare other people the trouble of thinking. But if possible, to stimulate someone to thoughts of his own” (PI, 4). In my view, the syncretistic has an epistemic-heuristic and creative role both in the production and in the reception of the text.30 At first sight it may seem rather unimportant and irrelevant to try to find out why Ortner classifies Wittgenstein as a puzzle-writer rather than a syncretistic writer. It may also seem unimportant to ask why he restricts the syncretistic to text production and consequently does not speak of syncretistic works. But trying to answer these questions actually helps us to gain a better understanding not only of what is at stake with Ortner’s assessment, but also of what was at stake for Wittgenstein when he left behind the Brown Book complex and began in the autumn of 1936 what today is called the PI “Urfassung”. An important question for him at this time must have been whether philosophy can benefit more from nonlinear and crisscross writing and composition than from work oriented towards linearity and textbook systematicity: could it even be imperative that he adopts nonlinear crisscross procedures for the treatment of philosophical problems? The PI preface answers this question clearly in the affirmative. Most philosophers and academics, however, would answer no. So too does Ortner: first of all, he does not apply the notion of the syncretistic to anything other than text production; the syncretistic is a feature of text genesis only. Secondly, even if the syncretistic was also a feature of works, for Ortner it could not have a distinctive positive value. For Ortner, as for most of us, the syncretistic is only “on the way to” something better, and it is to be overcome by that something better: • “Syncretism is a way of working which the tentative and experimenting intellect applies if it cannot cope with a way of working that is more elaborated, or if it does not yet have a way of working that is more elaborated. . . .”

The Philosophical Investigations and Syncretistic Writing

73

• “The heyday of syncretism is the incubation period. . . .” • “With syncretism, the writer wants to expand the status-quo in order to coherently and consistently integrate new parts of the world into the already developed knowledge base.” • “Syncretistic writing is writing which searches to establish higher (= better) gestalts, and for that purpose continuously has to revise and break down lower gestalts.”31 But is this necessarily so? Can we not accept that the syncretistic may be an author’s preferred method, even if s/he is capable of the other, “better” methods and strategies? “The heyday of syncretism is the incubation period”: can the syncretistic not have an equally important function at the time of completion and reception? Was the PI’s syncretistic form not consciously chosen as the better one, with a focus on both the problems to be treated and the reader to be inspired and helped—chosen by someone whom we consider to be one of the twentieth century’s most important philosophers? Consequently, would the value of syncretistic writing not only be a means of incubation, and therefore not only an author’s writing device, but would have a function for forming the work? Could it have a function for the reader too? When Ortner says that with syncretistic writing the writer actually seeks to expand the status quo to more coherent and consistent higher gestalts, is he really still talking of syncretistic writing—or is he rather ascribing to syncretistic writing a function which it in fact does not have according to his own definition, and which it should not have either? Here we find ourselves not only at the centre of normative writing theory, but also at the centre of the difficulties of Wittgenstein scholarship: can we utilize Wittgenstein well by trying to do crisscross philosophy in the PI spirit rather than by first identifying and then applying discursive-linear strands of argument in his work? Can we do it for academic philosophy? Wittgenstein himself seems to have been rather pessimistic about the influence he might have. Like Ortner, most of us seem to be held captive by the picture (PI, §115) that progress in writing and thinking is progress in respect to the following points: gestalt-elaboration, continuous integration, improved consistency and coherence, linearity. In this model, syncretistic writing can never be more than an element and a phase of text production which is to be overcome by something better, and—since syncretistic forming is by definition something to be avoided— it is even more unlikely ever to be considered a positive principle of work forming. One is inclined to agree with Ortner:32 syncretistic writing should typically be practised by inexperienced writers only, or, when used by experienced writers, overcome when they are about to produce the work to be published. Nevertheless, we must face the fact that an eminent thinker and writer such as Wittgenstein thought differently and challenged our view with his PI. It is in fact precisely this dominant view of ours that the PI opposed, and to which it tried to present an alternative.

74

Alois Pichler

The PI is usually seen as a continuation of the Brown Book. Thus, the similarities of the two texts are emphasized more than their differences. This blurs fundamental differences between the two in form and method: while the Brown Book focuses on linearity, step-by-step procedure and systematicity, the PI employs a crisscross procedure. The very method which was central to the Brown Book and was indeed its backbone—the linear language-game method—was “de-linearized” and “fragmentized”33 in the transition to the PI.34 According to Rush Rhees’ editorial note to Eine Philosophische Betrachtung, Wittgenstein had dismissed the Brown Book enterprise on the basis of wrong method: Moore told Rhees that “Wittgenstein had told him [Moore] that in the Brown Book he had followed the wrong method, but in this manuscript [TS 220] he had applied the correct method”.35 What was the Brown Book’s wrong method; which is the right one? In the 1938 PI preface drafts, Wittgenstein says: “I begin these publications with the fragment of my last attempt to arrange my philosophical thoughts in an ordered sequence. This fragment has perhaps the advantage of giving comparatively easily an idea of my method”.36 There has been some discussion about which “fragment” this was which Wittgenstein wanted to begin his publication with in 1938. I suggested it was MS 142 (or its typed version TS 220), which he began immediately after the abandonment of the German translation of the Brown Book in MS 115.37 The 1938 preface drafts make it clear that Wittgenstein wants to start his publication(s) with a) the “fragment” of his last attempt at ordering his philosophical thoughts in a series, and that this fragment b) shows his method. The same drafts clearly described which method this is: it is the method of crisscrossing rather than the method of linearity. Now, whereas the Brown Book complex is an example of the method of linearity, in MS 142 it is the method of crisscrossing which is exemplified. Wittgenstein says in both the preface drafts and the final PI preface that this method was an adequate response to the nature of the investigation; the crisscross method was responding to the demands of the subject area, and was thus responding to a requirement. From this, I think, we should be able to infer not only that “fragment” refers to MS 142 / TS 220, but also that in 1936 the crisscross method had been consciously upheld and transferred from the writing of the PI remarks (1929–1936) to their forming into the PI work (1936): it had been transferred from the process to the product.38 In this chapter I have so far tried to show that the PI has a syncretistic genesis and that the principles of the syncretistic were also employed in the late autumn of 1936 for the creation of the PI form: the PI is not only of syncretistic origin but also of syncretistic form. That in the PI the principles of genesis were also made the principles of form I considered significant. I have drawn attention to the PI preface which documents and defends this move with reference to the nature of the philosophical investigation required. I will conclude this chapter by trying to give my analysis of Wittgenstein’s choice of syncretistic form yet more substance by relating it to some

The Philosophical Investigations and Syncretistic Writing

75

additional reflections about possible reasons for making the PI the way he did. After all, if the Tractatus was (at least partly) of syncretistic origin and did in fact not end up in syncretistic form, why should the PI have ended up as being syncretistic? Why was the syncretistic chosen for the PI? One of the reasons may have been that the syncretistic was more in tune with the nature of the writer than any other form; the syncretistic may thus have been chosen out of an ideal of authenticity and sincerity. The PI product should mirror and be in harmony with the PI text production, and with the author’s own writing nature.39 Moreover, the syncretistic may have been chosen out of opposition to a philosophy of “progress”: the syncretistic PI form was to promote or preserve an alternative to scientific philosophy as defended by Bertrand Russell or Rudolf Carnap, or at least a space for something which is needed in addition to scientific procedure. This is most strongly expressed in a draft to a preface six years earlier, where for the first time crisscross procedure is positively described and opposed to linear procedure: Each sentence that I write is trying to say the whole thing, that is, the same thing over and over again & it is as though they were [alternative: they are as it were] views of one object seen from different angles. . . . One movement orders one thought to the others in a series, the other keeps aiming at the same place. One movement constructs & takes (in hand) one stone after another, [alternative: picks up one stone after another] the other keeps reaching for the same one. (CV, 9–10)40 Compare this with what Rudolf Carnap says in the preface to his The Logical Structure of the World: “. . . in slow careful construction insight after insight will be won. . . . Thus stone will be carefully added to stone and a safe building will be erected at which each following generation can continue to work”.41 These two reasons cannot of course be separated from the PI’s conception of philosophy and its conception of the nature of philosophical problems themselves. We may also want to relate the fact that the PI received a syncretistic and crisscrossing album form, and retained it until its last version, to the role the PI was to have as a work in contrast to the Tractatus, as is also stated in the PI preface. There is one aspect to this which has been strongly emphasized recently by James Conant: PI §133 defends that in philosophy we have to use many methods rather than only one. While the Tractatus had followed basically one method, the PI introduces, follows and defends many methods (see PI, §133). Conant thinks there is a “distinction in philosophical conception between the methodological monism of the early Wittgenstein (who seeks to present the method of clarification) and the methodological pluralism of a later Wittgenstein (who seeks to present an open-ended series of examples of methods—a series that can be continued in both unforeseen and unforeseeable ways—and that can be broken off at any point). . . . This transition from a definite article (‘the’ logic of

76

Alois Pichler

our language) in Early Wittgenstein to a plurality (‘grammars’) in Middle Wittgenstein presages and prepares the ground for a subsequent transition, yet again from a definite article (‘the’ method) in Middle Wittgenstein to a further plurality (‘methods’) in Later Wittgenstein”.42 Not only the “early” but also the “middle” Wittgenstein had, in contrast to the Wittgenstein who made the PI, aspired to produce a work with one method only.43 The PI holds that philosophy needs many methods. But what are these methods to be derived from? From philosophy’s own problems—the “many” methods shall be derived from the ways one struggles or has struggled with one’s philosophical problems. And this was, of course, connected with the very nature of the investigation. For it compels us to travel criss-cross in every direction over a wide field of thought. (PI, 3) The nature of philosophical problems is such that they are multilayered and multi-rooted.44 Their treatment compels us to philosophical crisscross travelling; Wittgenstein had been travelling in this way since 1929—but sometimes (e.g., when producing the Brown Book) he had forgotten that he was doing this, and should be doing precisely this. It is syncretistic writing more than any other type of writing that permits one to develop and utilize many different methods. It is the syncretistic as a form of the work which again asks the reader to engage in the many different methods and perspectives. This way the syncretistic can find its powerful way into the PI: The forms of creativity required for the discovery of fruitful methods in philosophy and the forms of creativity required for the fruitful application of such methods to particular problems of philosophy are recognized by Later Wittgenstein as two aspects of a single task, each of which requires an unending cultivation of the other.45 At the beginning of Chapter IV of his Schreiben und Denken, Hanspeter Ortner also discusses differences between writing that shares knowledge, writing that expands knowledge and writing that creates new knowledge.46 What is syncretistic writing best at? Probably not knowledge sharing, but rather the extension of existing knowledge and the creation of new knowledge.47 Syncretistic writing is actually a strategy central to epistemic-heuristic writing, the writing which has the creation of new knowledge as its primary cognitive function. Is it even the case that through syncretistic writing and syncretistic form we can gain knowledge which is not available otherwise? Gottfried Gabriel thinks that not only science and logical reasoning, but also literature and the poetic have a cognitive value.48 In a recent interview he states: We cannot equate the academic form of philosophy with philosophy in general. Being scientific is not a necessary condition for cognition.

The Philosophical Investigations and Syncretistic Writing

77

Arts and literature also have a cognitive value. . . . That’s why I think that what we need is a reconciliation of logical and poetic discourse. Both forms of discourse are fully justified, depending on what the concrete aim of articulation is. They do not necessarily oppose one another. . . . just as there are smooth transitions in the color spectrum, so that you can go from red to green, there are also intermediate forms in philosophy. The thesis that the different forms of philosophy are complementary concerns not only the extremes, but also these in-between cases. Ultimately, the issue at stake is a reconciliation of analytic and continental philosophy by means of analyzing the transitions among the different forms of philosophy.49 According to Gabriel, both analytic and continental philosophy (where a place for the poetic is retained) are required for the creation of philosophical knowledge and expertise. The poetic and literary on the one hand and the analytic on the other are complementary and add both to cognition and truth. Wittgenstein’s PI represents an intermediate form in philosophy, mediating between the analytic and the continental. I have tried to show in this chapter that the Wittgenstein of the PI saw the syncretistic (which belongs under the poetic and literary) as required for philosophical engagement. In a notebook entry from December 1933 Wittgenstein had stated that “one should write philosophy only as one writes a poem” (CV, 28).50 This statement stands in remarkable contrast to Carnap’s opposition to “Dichtung” in philosophy, as Carnap had expressed it in his preface to the Aufbau: “Consequently they have taken the strict and responsible orientation of the scientific investigator as their guideline for philosophical work, while the attitude of the traditional philosopher is more like that of a poet. . . . This requirement for justification and conclusive foundation of each thesis will eliminate all speculative and poetic work from philosophy”.51 We can say that Wittgenstein—if the Tractatus was his first work of “Dichtung” in philosophy (something Gabriel would assert)—achieved with the PI “Urfassung” his second. While the Brown Book had been poor in terms of the poetic, in MS 142 the dialogical, the metaphorical, the simile, the analogical and the means which I have described above as elements of the syncretistic are given philosophical functionality.52 One example is the PI §18: it invites the one who philosophizes in Carnap’s spirit to look at language in a way which is guided by a liberating metaphor. Bouwsma writes that Wittgenstein appreciated (at least in September 1950) Plato’s “allegories, the myths”—but not his arguments.53 This appreciation of the poetic in Plato fits well with the choice of syncretistic form for the PI. Is there an opposition between poetic and syncretistic philosophy on the one hand and analytic philosophy on the other? The Wittgenstein of the 1930 preface drafts and also of the PI seems to have seen a deep opposition between the two. The PI’s form and philosophy were intended as

78

Alois Pichler

alternatives and in opposition to forms of analytic philosophy, which Wittgenstein found in his time was becoming more and more dominant. But independently of whether one will agree with Wittgenstein on whether there is such an opposition, one will have to recognize and acknowledge that the Wittgenstein of the PI seems to have considered the syncretistic a required part of philosophical writing and also of philosophy as such—the syncretistic has to be a feature of both the process and the work. This already seems to be enough of a challenge, a challenge not only to analytic philosophy but to our academic standards in general.54 REFERENCES Baker, G. P., and P. M. S. Hacker. 1983 (1980). Wittgenstein: Meaning and Understanding. Essays on the Philosophical Investigations, Vol. 1. Oxford: Basil Blackwell. Bazzocchi, Luciano. 2008. “On Butterfly Feelers: Some Examples of Surfing on Wittgenstein’s Tractatus”. In Wittgenstein and the Philosophy of Information: Proceedings of the 30th International Wittgenstein Symposium, Vol. 1, edited by Alois Pichler and Herbert Hrachovec, 125–40. Heusenstamm: Ontos. Biggs, Michael, and Alois Pichler. 1993. Wittgenstein: Two Source Catalogues and a Bibliography. Catalogues of the Published Texts and of the Published Diagrams, each Related to its Sources. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen. Binkley, Timothy. 1973. Wittgenstein’s Language. The Hague: Martinus Nijhoff. Bouwsma, O. K. 1986. Wittgenstein: Conversations 1949–1951. Edited by J. L. Craft and R.E. Hustwit. Indianapolis: Hackett. Carnap, Rudolf. 1961 (1928). Der Logische Aufbau der Welt. Leipzig: Felix Meiner. ———. 1967. The Logical Structure of the World. Translated by R. A. George. London: Routledge. Conant, James. 2011. “Wittgenstein’s Methods”. In The Oxford Handbook of Wittgenstein, edited by Oskari Kuusela and Marie McGinn, 620–45. Oxford: Oxford University Press. Erbacher, Christian. 2010. Zur philosophischen Bedeutung der sprachlichen Gestaltung von Wittgensteins Logisch-philosophischer Abhandlung. Bergen: Universitetet i Bergen [PhD thesis]. Gabriel, Gottfried. 1991. Zwischen Logik und Literatur: Erkenntnisformen von Dichtung, Philosophie und Wissenschaft. Stuttgart: Metzler. ———. 2012. “Analytical Philosophy and its Forgetfulness of the Continent? Gottfried Gabriel in Conversation with Todor Polimenov”. Nordic Wittgenstein Review 1: 155–75. Gibson, Arthur. 2010. “The Wittgenstein Archive of Francis Skinner”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 64–77. Basingstoke: Palgrave Macmillan. Gorlée, Dinda L. 2012. Wittgenstein in Translation: Exploring Semiotic Signatures. New York: Walter de Gruyter. Gründler, Hana. 2011. “‘Eine Menge von Landschaftsskizzen.’—Zur Bedeutung des Zeichnerischen in Ludwig Wittgensteins Spätphilosophie”. In Image and Imaging in Philosophy, Science and the Arts: Proceedings of the 33rd International

The Philosophical Investigations and Syncretistic Writing

79

Wittgenstein Symposium, Vol. 1, edited by Richard Heinrich, Elisabeth Nemeth, Wolfram Pichler and David Wagner, 1–22. Heusenstamm: Ontos. Hilmy, S. Stephen. 1987. The Later Wittgenstein: The Emergence of a New Philosophical Method. Oxford: Basil Blackwell. Janik, Allan. 2006. Assembling Reminders: Studies in the Genesis of Wittgenstein’s Concept of Philosophy. Stockholm: Santérus Academic Press Sweden. Keicher, Peter. 2000. “Aspekte musikalischer Komposition bei Ludwig Wittgenstein: Studienfragmente zu D 302 und Opus MS 114ii/115i”. In Das Verstehen des Anderen, edited by Katalin Neumer, 199–255. Frankfurt am Main: Peter Lang. ———. 2004. “‘Ich wollte, alle diese Bemerkungen wären besser als sie sind.’— Vorworte und Vorwortentwürfe in Wittgensteins Nachlass”. In Essays on Wittgenstein and Austrian Philosophy: In Honour of J. C. Nyíri, edited by Tamás Demeter, 275–309. Amsterdam: Rodopi. ———. 2011. “‘So ist also dieses Buch eigentlich nur ein Album.’ Bemerkungen zu Wittgensteins Photoalbum”. In Wissenschaftstheorie, Sprachkritik und Wittgenstein: In memoriam Elisabeth und Werner Leinfellner, edited by Sascha Windholz and Walter Feigl, 127–34. Heusenstamm: Ontos. Moreno, Arley R. 2009. “Como ler o Álbum?”. In Como ler o álbum, edited by Arley R. Moreno, 131–82. Campinas: CLE. Ortner, Hanspeter. 2000. Schreiben und Denken. Tübingen: Niemeyer. Pichler, Alois. 1994. Untersuchungen zu Wittgensteins Nachlass. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen 8. ———. 1997. Wittgensteins Philosophische Untersuchungen: Zur Textgenese von PU §§1–4. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen 14. ———. 2004. Wittgensteins Philosophische Untersuchungen: Vom Buch zum Album. Amsterdam: Rodopi. ———. 2007. “The Interpretation of the Philosophical Investigations: Style, Therapy, Nachlass”. In Wittgenstein and His Interpreters: Essays in Memory of Gordon Baker, edited by Guy Kahane, Edward Kanterian and Oskari Kuusela, 123–44. Oxford: Blackwell Publishing. ———. 2009. “Wittgenstein’s albums: ‘Philosophical Investigations’ and ‘Philosophical Remarks’ as alternatives to the ‘spirit of progress’ in philosophy”. In Como ler o álbum, edited by Arley R. Moreno, 57–97. Campinas: CLE. Rothhaupt, Josef. 2011. “Der Komplex ‘MS140(I) + MS114(II) + MS115(I)’ als Wittgensteins Buch ‘Lsrpmhmlsrhxsv Yvoviqfntvn’”. In Epistemology: Contexts, Values, Disagreement. Papers of the 34th International Wittgenstein Symposium, edited by Christoph Jäger and Winfried Löffler, 247–50. Kirchberg am Wechsel: ALWS. Savigny, Eike von. 1991. “No Chapter ‘On Philosophy’ in the Philosophical Investigations”. Metaphilosophy 22, no. 4: 307–19. Schobinger, Jean-Pierre. 1991. “Weshalb sind die Philosophischen Untersuchungen Wittgensteins nur ein Album?”. Zeitschrift für philosophische Forschung 45: 249–56. Stenius, Erik. 1969. Wittgensteins Traktat: Eine kritische Darlegung seiner Hauptgedanken. Frankfurt am Main: Suhrkamp. Stern, David. 2006. “How many Wittgensteins?” In Wittgenstein: The Philosopher and his Works, edited by Alois Pichler and Simo Säätelä, 205–29. Heusenstamm: Ontos. Venturinha, Nuno, ed. 2010. “Wittgenstein’s 1938 Preface”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 182–86. Basingstoke: Palgrave Macmillan. .

80

Alois Pichler

von Wright, G. H. 1993. Myten om Framsteget. Helsinki: Söderströms. Wittgenstein, Ludwig. 1961. Tractatus Logico-Philosophicus. Translated by D. F. Pears and B. F. McGuinness. London: Routledge and Kegan Paul. (First published in 1921 in German and 1922 in English) (TLP) ———. 1984. “Eine Philosophische Betrachtung”. Edited by Rush Rhees. In Werkausgabe, vol. 5, 117–237. Frankfurt am Main: Suhrkamp. (EPB) ———. 1998. Culture and Value: A Selection from the Posthumous Remains. Edited by G. H. von Wright in collaboration with Heikki Nyman. Revised edition of the text by Alois Pichler. Translated by Peter Winch. Oxford: Blackwell. (CV) ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. (PI)

5

The Role of the Big Typescript in Wittgenstein’s Later Writings Joachim Schulte

In this paper, I wish to make two central points. At first blush, these points may seem a little too obvious to be worth stating. But as their true import is often overlooked or ignored, I shall here make an attempt at explaining and illustrating them by way of giving a sketchy account of the development of some of Wittgenstein’s later writings. But before I get down to giving this account, I shall briefly state my two points: (1) It is important to distinguish between a given manuscript or typescript as a material object and the same manuscript or typescript as, for example, described in terms of a more comprehensive project specifiable by reference to documented intentions, variously interpretable plans, sketches, etc. (2) In giving an account of the development of Wittgenstein’s writings, it is essential to pay attention to the meaning of the descriptive terms used to characterize this development. This is especially urgent if these terms have conventional meanings that do not really fit the known history of the writings concerned. Examples of expressions I have in mind are the words “revision” and “work”. Let me explain a couple of minor questions, so we can get them out of the way: first, in general I shall use the words “manuscript” and “typescript” in their most literal meanings, viz. in the sense of strings of remarks written by hand or by typewriter, respectively. Occasionally, perhaps, the word “manuscript” will be used as a general term signifying items of Wittgenstein’s Nachlass, i.e., his papers, irrespective of whether they were written by hand or typed. Of course, questions might be raised concerning the status of small or large handwritten insertions in typescripts or pasted-in snippets of typescript in manuscript volumes. But these questions need not detain us for the time being, as they do not present us with problems that could not be resolved by means of a more exact description.

82

Joachim Schulte

Second, as regards point two, I mentioned a couple of examples, viz. the words “revision” and “work”. Further examples could be mentioned, but here I shall concentrate on certain aspects of the term “revision”. The word “work” has been discussed by me in several earlier papers or chapters.1 And even though its use continues to produce misleading effects, I shall here confine myself to the following short comment: While the word “work” may sometimes be employed in fairly neutral fashion, signifying little more than the results of an author’s labours, in most cases its use will be perceived as implying something more ambitious. So it will, for example, imply the fulfilment of relatively high standards of order, retraceability of steps of reasoning, interconnectedness and sign-posting. In view of Wittgenstein’s habitual way of writing short remarks, such standards are difficult to fulfil anyway. In the case of his writings, they may arguably be satisfied by certain passages of a small number of manuscripts or typescripts. But on the whole, I should want to claim that there are only two texts among the whole of his oeuvre that would deserve the title “work” by any reasonable standards. These are, first, the Logisch-philosophische Abhandlung (better known under the title of its English edition: Tractatus Logico-Philosophicus) and, second, his posthumously published masterpiece, Philosophische Untersuchungen or Philosophical Investigations. This is not the occasion to support this claim by giving additional reasons. Instead, I shall hasten to mention a few facts about the text we are used to calling the chief work of Wittgenstein’s later philosophy, his Philosophical Investigations. And these facts are at the same time meant to throw some light on the first of my two points. If we look at the Investigations as a material object, certain features of the case are quite striking. (Naturally, in inviting you to look at this text as a material object, I am not asking you to forget that its sentences can be read as making this or that sense and its words as having certain meanings. I am basically asking you to pay particular attention to the actual sheets of paper bearing Wittgenstein’s words, and their history.) Now, the first striking fact is this: that the typescript from which the first edition of the Investigations was printed has been lost. In other words, it is likely that the compositor was the last person to set eye on this typescript before it disappeared—according to one story, in the Royal Mail. However, we know at least one thing about this typescript: it contained a number of corrections and additions in Wittgenstein’s hand. We know this from a letter to Georg Henrik von Wright (6.6.47), to whom Wittgenstein had lent this copy.2 In the same letter he mentions the existence of two further copies of his work and says that they do not contain these additions (or some of them, at any rate). What we do not know is whether the corrections and additions contained in this copy at this time were complete, as it were, or whether more material was added at a later stage. There is some circumstantial evidence that Wittgenstein tinkered with the typescript in 1949 or 1950, because it was at that time that Georg Henrik von Wright was given a number of pages

The Role of the Big Typescript in Wittgenstein’s Later Writings

83

forming part of an earlier version of the Investigations which may have been kept together with the last version and were removed at this late stage of the process of putting the typescript in order. This does not mean, however, that Wittgenstein made any other changes at that time. The text of the Investigations as we know it may have been complete by 1947,3 but it cannot be excluded that further changes were made later than that. We simply do not know. But if further changes were made, they cannot have been very numerous. After all, their possible extent can be estimated on the basis of the difference between the text as we know it and the clean typescript. Of this typescript we know that it cannot have been finished before some time in the second half of 1945, and that it must have been completed before the autumn of 1946, when Wittgenstein lent a copy to Norman Malcolm as a basis for discussion. Given various remarks in Wittgenstein’s correspondence, I should think it likely that the typescript was finished towards the end of 1945 or early in 1946 at the latest. So, we need to distinguish between different material objects here: the clean typescript of the Philosophical Investigations as completed sometime around 1945–46, on the one hand, and the supplemented typescript of the Investigations as known to us from its posthumous publication, on the other. The extent of this supplementation can also be gauged, although it is not exactly matched, by the two extant copies of the typescript of the Investigations which contain, apart from additions in Wittgenstein’s own hand, some material inserted by other hands or in typewritten form. Let me give just two reasons for emphasizing the distinction between the material-object aspect of the typescript we have been talking about and other forms of describing it: the first reason is the fact that in his preface to the Investigations Wittgenstein gives a list of topics to be treated in his book which goes clearly beyond what we actually find in the text. This means that Wittgenstein intended to produce a more widely ranging book of which the text as we know it may be regarded as a part, perhaps even a large part. So there is a sense in which the Investigations are not complete at all, viz. in the sense of various projects Wittgenstein entertained at different times and whose realization would have gone beyond the book as we know it. My second reason is connected with this one. Occasionally (for instance in a letter to his former pupil Redpath [22.7.48]) Wittgenstein insists even after 1946 that his book has not yet found its conclusion. But if we take this sort of statement at face value, he is clearly not referring to the text of the Investigations in one of its material manifestations but rather to a more comprehensive though unrealized project. After all, in at least one other letter from the same time he treats the typescripts of his “stuff” (as he calls it, to Malcolm, 6.11.48) as records of work completed some time ago.4 Now let us look at another case which, as we shall see, displays some surprising parallels. The case I mean is that of the so-called Big Typescript, also known as TS 213 in accordance with its number in Georg Henrik von Wright’s catalogue of Wittgenstein’s papers.5 In a way this typescript is a

84

Joachim Schulte

summary or, as one may want to put it, the quintessence of Wittgenstein’s labours during the time between his return to Cambridge and philosophy in 1929 and the typing of this huge record in 1933. Wittgenstein’s thoughts were first written down in large manuscript notebooks, sometimes drawing on drafts in smaller and less finished notebooks. From these big manuscript notebooks he extracted a large number of remarks and had them typed, mostly in the same order in which they occurred in the manuscripts, normally introducing only minor changes and additions. Copies of these typescripts were cut up into smaller units and rearranged, thus producing an enormous bulk of material, which was grouped into roughly 140 sections and 19 chapters. This order was the basis for what, in the summer of 1933, became the Big Typescript with its 768 pages. In view of the orderly, one might even say, conventional, appearance of this neat pile of pages with its systematic-looking table of contents, people have wondered why this typescript was never published by the trustees of Wittgenstein’s papers. Well, some of these people may have looked not at the Big Typescript in its strict and proper sense, but at its carbon copy or even a retyping of that. For the Big Typescript in the strict sense is a pretty mess: most of the pages of its first half or so are full of Wittgenstein’s handwritten scribbles, amending or supplementing the typed text, and a great number of verso pages are also covered with Wittgenstein’s notes and jottings. Many of these additions must have been written shortly after the completion of the typescript, and some may have been written a little later. If serious work has been done on distinguishing different strata of this additional material, it has not come to my attention. At any rate, only a few months of this kind of work passed before Wittgenstein embarked on a new stage of dedicating himself to the remarks collected in roughly the first half of the Big Typescript: he used his last big manuscript notebook to copy and rearrange revised versions of selected remarks from the Big Typescript (including handwritten additions in margins or on verso pages). And as soon as this manuscript notebook was full, he took a new one to continue his work. As the first page of the new manuscript book bears a date, we know that it was begun in December 1933, and we can be sure that it took Wittgenstein several weeks at any rate until he stopped on page 117 of this manuscript book. Perhaps it was at this point that he became dissatisfied with at least the early stage of this rearrangement, for he wrote another version of the beginning on a set of 39 sheets of large-size paper. The first rearrangement—the one written into the manuscript book—was called Umarbeitung; the new version on loose sheets was called the Zweite Umarbeitung. Wittgenstein’s term Umarbeitung, I think, is satisfactorily translated by the English word “revision”. The question, however, is: what are these revisions revisions of? And to answer this question it is surely helpful to concentrate on the material aspect emphasized before. What Wittgenstein revises in the course of his first revision is not—or at any rate: not only—the collection

The Role of the Big Typescript in Wittgenstein’s Later Writings

85

of earlier manuscript remarks assembled and rearranged in the clean typescript, but the collection as modified by an enormous number of handwritten corrections and additions. And the basis for the second revision is surely not the modified typescript, the Big Typescript, but the text as presented in the manuscript volume containing the first revision. Taken in this sense, our use of the word “revision” is quite unproblematic. But in many cases its use carries a certain implication which needs to be made explicit. What I mean is that speaking of a revision here tends to suggest that the Big Typescript is more than a modified collection of remarks— more than a quarry (as it has aptly been called6), where Wittgenstein would from time to time return to dig up material that might come in handy. And this suggestion can be misleading. To make this clearer think, for example, of a poet who revises a poem. In talking this way one implies that there is a more or less finished version of the poem which is then modified to meet the poet’s standards. One would not so easily speak of a “revision” if all the poet was working on was a sequence of lines with gaps—a sequence, moreover, laid out in uncertain order. Using this analogy, we may differ about whether the material we are talking about was a “version” of something that could be “revised” in the more ambitious sense just explained. But even if we do differ, it must be recognized on all sides that a fairly convincing case can be made for holding that while Wittgenstein was working on the Big Typescript, he was not revising a text that would really deserve to be called a “version” of a projected work. The shape the material collected and modified in the Big Typescript was in after the second Umarbeitung (or revision) can be studied by reading the first half of the book Philosophical Grammar, edited by Rush Rhees and later translated into English by Anthony Kenny.7 Kenny actually became one of the most vocal critics of Rhees’ editorial decision to publish the 1934 text, that is, the second Umarbeitung, rather than the clean typescript. It would be interesting to look into various aspects of the matter, but here I want to concentrate on the idea of a revision to clarify the notion a little further. In order to do so we need to take another step on the path of the development of Wittgenstein’s writings and move on to the autumn of 1937. By this time Wittgenstein had not only dictated the Blue and Brown Books, but also written the early version of the first third of the Investigations and a part of those manuscripts that became a typescript forming the basis of Part I of the Remarks on the Foundations of Mathematics.8 It was at this time of his second consecutive year in Norway that Wittgenstein returned to the Big Typescript. More precisely, on 23 October 1937, he decided to select and copy remarks from his “old typescript” (as he now called it). The result of this process, which does not seem to have occupied him for an unduly long time, can be found on the first ninty-five pages of a large manuscript book that has played a very controversial role in discussions of the genesis of Wittgenstein’s writings. Wittgenstein himself makes it quite explicit in a kind of diary kept in a different manuscript from the same

86

Joachim Schulte

period that he uses this “old typescript” as nothing more than a storehouse from which possibly useful remarks can be drawn. He describes his work as an effort to separate the wheat from the chaff and states that he finds the majority of these remarks indifferent or mediocre. He writes that those which simply express a problem are the best ones. So, in a way, the Big Typescript is now regarded as a mere ragbag of forgotten remarks, only a few of which appear to be worth preserving. And the process of dealing with this material is characterized as one of satisfying his curiosity about his earlier thought in the hope of not being crushed by the experience and of learning something (25.10.37). On the day before writing down this note, he compares his activity with that of a person who is supposed to furnish a flat by means of objects pulled out of a rubbish heap, taking pains to examine and clean them (24.10).9 One of the difficulties posed by the relevant part of manuscript 116 was that it contains no dates and, because of its content, people were tempted to date it early while some external evidence seemed to speak against this. So they were prone to look in the wrong place and failed to identify the diary remarks I have just drawn on as referring to Wittgenstein’s work on his selection of remarks in manuscript 116. This failure is chiefly due to a mistaken description of the material bearers of Wittgenstein’s ideas and the development leading from one manuscript or typescript to the next. Many passages could be cited as evidence for this judgement, among them, unfortunately, one or two from my own papers. Here I shall quote just one passage from an article by Anthony Kenny, which even today is often relied on. Kenny says: Volume XII [manuscript 116] is extremely difficult to fit into the history of the revision of the Big Typescript. Both von Wright and Rhees believe that it is substantially later than the Zweite Umarbeitung. . . . The reader who approaches Volume XII in this belief is astonished to find that the revision it contains is, at least at the beginning, very much closer to the original text of the Big Typescript than are either of the earlier revisions. This fact alone, if the accepted chronology is correct, surely casts doubt on any claim of the Zweite Umarbeitung to be the final revision of the typescript: it means that after the second revision Wittgenstein came to the conclusion that his book would do better to stick closer to the original text of the typescript.10 This, I suppose, suffices to show clearly what is wrong about this way of looking at the matter and the chain of reasoning that goes together with it. No distinction is made between the Big Typescript as a material vehicle carrying a collection of remarks and their modifications, on the one hand, and the Big Typescript as an alleged version of a work that has not only been revised twice in the months immediately following its “completion” but a third time in manuscript 116 (and its successor 117iv).

The Role of the Big Typescript in Wittgenstein’s Later Writings

87

But the picture drawn in this quotation is completely wrong and misleading—quite apart from the fact that its implications are extremely unfair to the work done by Rush Rhees as editor of Philosophical Grammar. No matter if the Big Typescript is regarded as a clean typescript or as the result of this plus Wittgenstein’s handwritten modifications, it is wrong to describe it as a version of a work by Wittgenstein. It was never more than a superficially ordered and subdivided collection of remarks selected from earlier typescripts and hence from even earlier manuscripts. As such it is of great interest to the student of Wittgenstein’s writings, but it must not be placed in one of the ill-fitting categories offered in much of the literature. The Umarbeitungen may fairly be described as “revisions” in one sense of the word: they were attempts at transforming an unwieldy mass of remarks into something that might one day become the “natural and smooth” summary of his ideas that Wittgenstein was striving for.11 The remarks from the Big Typescript to be found in manuscript 116 (and 117), however, are not a revision of anything. They are a selection of remarks from the first 250 pages or so of what he regarded as his “old typescript”. He made this selection partly to see whether any of these remarks could be used for present or future purposes and partly to find out about his former self and his former way of thinking. In my view, he was successful in both respects. He did learn something about himself and his former way of thinking, but to establish the correctness of this impression would be a completely new and very speculative story. As regards the other respect, viz. the potential usability of the selected remarks, it can be put on record that many of them were selected when in 1945 Wittgenstein assembled a typescript (called Bemerkungen I = TS 228) which later served as the basis of the dictation of the relevant parts of his last typescript of the Investigations (TS 227). With the mention of Bemerkungen I, I am approaching the end of my tale. This typescript comprises nearly 700 remarks, and the material taken from the first part of manuscript 116 is not the only set of remarks that can also be found in the Big Typescript. What Wittgenstein did when he dictated Bemerkungen I was that he took no typescript but a number of manuscript books (in particular 114–17 and 129–30) to select from them usable remarks in the order he found them. In this way a great number of remarks that had passed through the Big Typescript found their way into Bemerkungen I and from there into the Investigations. But no one would want to call Bemerkungen I a “revision” of anything. It isn’t one. It’s a pretty random collection, and if any work of revision had gone into earlier writings containing these remarks, all traces of this were lost at this stage and played no role in Wittgenstein’s further work aimed at producing the text of the Investigations. We can be sure that a good deal of effort went into this work, but apart from a list (manuscript 182) that must have been the (or a) basis for the dictation of the final typescript, we have no clear evidence of how he proceeded.12

88

Joachim Schulte

However, as I tried to bring out before, it is obvious that Wittgenstein basically relied on Bemerkungen I in producing his final typescript of the Investigations, and all earlier versions and revisions did not come into this process. The reasons mentioned and the developmental story told contribute to showing that it is quite wrong to claim (as has often been the case in the literature) that the existence of a great number of remarks from the Big Typescript that can also be found in Philosophical Investigations demonstrates something about the truth of the thesis that much of the later thought was anticipated in the Big Typescript or its manuscript sources or its revisions. What is indubitably true is that there is an impressive overlap of remarks between the Big Typescript and the Investigations. But what this fact shows about anticipations and similarities of ideas and ways of thinking needs to be established independently. In particular, we need better and more comprehensive accounts of what Wittgenstein was up to at the time he wrote and reused those remarks. Only then can we begin to think about what conclusions to draw from their occurrence in both earlier and later writings. On the whole, I suppose Rhees was right when he observed that “When Wittgenstein writes a paragraph here [in the Big Typescript or its Umarbeitungen] that is also in [earlier writings], this does not mean that he is just repeating what he said there. The paragraph may have a different importance, it may belong to the discussion in a different way”.13 Actually, I think this insight can safely be generalized so as to apply to other writings by Wittgenstein and their interconnections. REFERENCES Biggs, Michael, and Alois Pichler. 1993. Wittgenstein: Two Source Catalogues and a Bibliography. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen. Hilmy, S. Stephen. 1987. The Later Wittgenstein. Oxford: Basil Blackwell. Kenny, Anthony. 1984. “From the Big Typescript to the Philosophical Grammar”. In The Legacy of Wittgenstein, 24–37. Oxford: Basil Blackwell. Kienzler, Wolfgang. 2005. “Die Stellung des Big Typescripts in Wittgensteins Werkentwicklung”. In Wittgensteins “große Maschinenschrift”, edited by Stefan Majetschak, 11–30. Frankfurt: Peter Lang. Klagge, James C., and Alfred Nordmann, eds. 1993. Ludwig Wittgenstein: Philosophical Occasions 1912–1951. Indianapolis: Hackett. McGuinness, Brian, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. Rhees, Rush. 1974. “Note in Editing”. In Ludwig Wittgenstein, Philosophical Grammar, edited by Rush Rhees, 487–90. Oxford: Basil Blackwell. Schulte, Joachim. 1989. Wittgenstein: Eine Einführung. Stuttgart: Reclam; 1992. Wittgenstein: An Introduction, translated by W. H. Brenner and J. F. Holley. Albany: State University of New York Press.

The Role of the Big Typescript in Wittgenstein’s Later Writings

89

———. 2005. “What is a Work by Wittgenstein?” In Wittgenstein: The Philosopher and His Works, edited by Alois Pichler and Simo Säätelä, 356–63. Bergen: Working Papers from the Wittgenstein Archives at the University of Bergen. Wittgenstein, Ludwig. 1974. Philosophical Grammar. Edited by Rush Rhees. Translated by Anthony Kenny. Oxford: Basil Blackwell. ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt am Main: Suhrkamp. ———. 2011 (2004). Gesamtbriefwechsel: Innsbrucker elektronische Ausgabe. Second release. Edited by Brian McGuinness, Monika Seekircher and Anton Unterkircher. Innsbruck: InteLex.

This page intentionally left blank

Part II

The Significance of Logic and Mathematics

This page intentionally left blank

6

Logic and Ideality Wittgenstein’s Way beyond Apriorism, Empiricism and Conventionalism in the Philosophy of Logic Oskari Kuusela

INTRODUCTION This chapter seeks to elucidate the novel account of the status of logic developed by Wittgenstein in his later philosophy, distinct from the alternatives of apriorism, empiricism and conventionalism.1 Wittgenstein’s reasons for developing such an account are multiple but connected. Most of all, he becomes disappointed with the limited capacity of his early Fregean-Russellian logico-philosophical method to do work of philosophical clarification, especially when the modes of language use to be clarified are complex and fluid, as characteristic of everyday language. A key reason for this shortcoming, according to Wittgenstein, is the conception of language as an ideal abstract entity assumed by his early method, which he later rejects, emphasizing that language is a spatial and temporal phenomenon. But when one rejects the conception of language as an abstract entity that underlies the multifaceted phenomena of language we know from everyday life, a host of difficulties arise. Above all, it seems that any attempt to understand logic as a study of the actual multiplicity of the forms of language and thought forces one to construe logic as an empirical discipline. For evidently this multiplicity is an empirical phenomenon. Yet, as far as Wittgenstein is concerned, collapse into empiricism would mean logic’s demise. According to him, empiricism makes it impossible to explain logical necessity and the nontemporality of logical statements but also, equally seriously, to account for the rigour of logical descriptions, definitions, and so on. For if actual uses of language are typically vague, fluid and complex, how can the discipline of logic describe them without falsification in exact and simple terms, and thus meet its ideals of rigour and clarity? It is in order to resolve such problems, I argue, that Wittgenstein develops his later conception of the status of logic, and an associated account of idealization in logic. Part of the proposed interpretation of Wittgenstein’s later philosophy of logic is an account of how he comes to abandon the Fregean anti-psychologism assumed in the Tractatus that forced him to deny categorically the relevance of facts of nature and natural history for logic. Certainly, Wittgenstein’s

94

Oskari Kuusela

abandonment of this form of anti-psychologism doesn’t mean that he would embrace psychologism or any other form of empiricism. But his renouncing of Tractarian anti-psychologism is an integral part of his rejection of the Tractatus’ apriorism that construes logic as concerned with a priori principles and forms that govern thought or language in the sense of an abstract entity. By contrast, Wittgenstein’s later conception enables him to recognize the relevance of the facts of nature for logic, without commitment to empiricism. This opens the doors for arguably more fruitful approaches to logico-philosophical clarification than possible within an aprioristic framework, and for the development of novel logical methods for philosophical purposes. Wittgenstein’s new account of logic, as I will explain, amounts to what one might call “non-empiricist naturalism”. Or as Wittgenstein characterizes his goal: “Not empiricism and yet realism”, describing this as the most difficult thing in philosophy and as constituting a criticism of Ramsey (RFM, VI, §23, see MS 164, 67). The connection with Ramsey will become clearer in section 4. The exegetical method of this chapter involves as an important element the employment of Wittgenstein’s notebooks, manuscript and typescripts as an aid in the interpretation of his remarks on logic in the Philosophical Investigations. In so doing the chapter throws light also on specific aspects of the textual genesis of the Investigations. Especially relevant for my discussion are certain notebooks from 1936–38 (MSS 152, 157a, 157b, 183) that contain multiple drafts of some of the remarks on logic in Investigations §§89–133. To a lesser extent I also make use of Nachlass material from 1932–34 (MSS 114, 115, 140 and TS 211), for example, in interpreting Investigations §81. Some still earlier Nachlass material is employed in discussing Wittgenstein’s early conception of logic. This is where I begin.2 1. THE IDEALITY AND PURITY OF LOGIC: WITTGENSTEIN’S STARTING POINT In the Tractatus and into the 1930s Wittgenstein assumed a conception of logic as an a priori investigation whose object of study, the logical principles governing thought and language, is pure and devoid of anything empirical and ideal in this and certain related senses. (I will discuss the meanings of “ideal” below. Frege and Russell, share a similar conception.) An important part of this view of the purity of logic is the idea that comprehension of logic must be something prior to knowledge of the facts of empirical reality, including the facts of psychology and anthropology. Because all thinking that takes empirical reality as its object and aims at a true understanding of it must already abide by and presuppose relevant principles, it is not possible to derive the principles of logic from any such facts or justify them with reference to facts. On this account, the scope of the discipline of logic, whose business it is to clarify logical principles, then far exceeds that of any special

Logic and Ideality

95

empirical sciences, and logic possesses unique foundational significance. As Wittgenstein describes this view retrospectively in the Investigations, rather than interested in how things happen to be contingently, or with “facts of nature”, logic is motivated by “an urge to understand the basis, or essence, of everything empirical”. In this sense it seems then to have “peculiar depth” and “universal significance” (PI, §89). Thus, rather than concerned with contingencies, logic investigates what is possible or thinkable in a logical, not psychological sense, and what is necessary, i.e., what pertains to the essence or identity of things and underlies their possibility of being contingently thus and so. From this point of view any particular facts of nature or natural history are excluded from having any relevance to logic. Wittgenstein writes in (January) 1930 about natural historical considerations pertaining to language: “But the natural history of the use of a word can’t be any concern of logic” (PR, §15; see WA 2, 168 (MS 107, 234)/TS 209, 6). In addition to such empirical considerations about the use of language, human natural history for Wittgenstein includes any considerations pertaining specifically to human beings that concern their constitution as living beings. Rather than studying human thought in particular, however, logic is concerned with thought in a more abstract and general sense: “What makes a thought a thought for us can’t be anything human, something that had to do with the constitution and nature of humans, but something—purely logical—that obtains independently of the natural history of a living being” (WA 2, 286 (MS 108, 217); see TS 210, 50).3 A related way in which Wittgenstein characterizes logic is this: rather than having anything to do with mental processes or anything characteristically human, “[l]ogic is the geometry of thought” (WA 2, 299 (MS 108, 242); see TS 210, 58/TS 236, 11). Part of this comparison of logic with geometry presumably is that logic is independent of how thought is actualized or realized in the empirical world, just as geometry is independent of such considerations. As the study of geometry is not concerned with physical triangles, and their properties have no relevance to geometry, so the peculiarities of human makeup, their psychology and physiology or their environment are not relevant to the discipline of logic. But if neither discipline is concerned with the actualizations of their objects of investigation in the empirical world, what are they concerned with? Here the view suggests itself that they are concerned with something ideal that lies beyond the physical world subject to time and change; that their objects of investigation are something nontemporal, as opposed to objects of the empirical sciences. This is one sense in which one might understand logic as concerned with something ideal. Although this view is readily amenable to a Platonistic interpretation that assigns the objects of logical investigation a status as entirely independent of the empirical world (as Frege maintained), it may also be understood in a more Aristotelian fashion. According to the latter, the discipline of logic investigates a priori logical forms, but those forms are not regarded as having an existence independent of the empirical world.

96

Oskari Kuusela

Instead, the empirical phenomena are the bearers of forms. It is natural to understand the Tractatus’ view of logic in the latter way, in the sense explained in the next paragraph (cf. TLP, 5.552–5.5562). According to the Tractatus, rather than concerned with actual languages or thought processes in an empirical sense, the task of logic is to reveal the underlying common essence of language or thought that is shared by all their possible forms, or is characteristic of language or thought as such. Logical investigation therefore must abstract away anything specifically human. What is relevant to logic, in this view, is not any particular linguistic expression, but the possibility of their expressing a sense or meaning; essential to an expression is what makes it possible for it to express a sense or meaning. Accordingly, we can, so to speak, zoom in on an expression’s essential features by determining what is common to all different expressions that can express the same sense or meaning, and in this way abstract from their accidental characteristics (TLP, 3.34, 3.341, 3.3421). This is most likely what Wittgenstein means in the later remark just quoted, when he says that logic is concerned with thought as something purely logical, and that anything natural historical is irrelevant to it. For example, that humans use spoken and written expressions and express themselves in a variety of different natural languages exemplify the kind of natural historical facts that logic aims to abstract from. Any particular modes of linguistic expression are of interest to logic only insofar as they reveal something about the underlying common essence of thought and language. This conception of logic and the associated method of logical abstraction—the purification of the objects of logical investigation from anything empirical—finds an interesting expression also in how the Tractatus explains the relation between psychology and philosophy or logic.4 At first sight its view seems quite straightforward: “Psychology is no more closely related to philosophy than any other natural science” (TLP, 4.1121; cf. 4.111). Wittgenstein’s formulation suggests that this is to be understood in the sense explained earlier: Empirical scientific investigations, including psychological ones, already assume and rely on the principles of logic and therefore can’t clarify them. Thus understood, the Tractatus’ anti-psychologism would be a particular instance of its more general demarcation between the empirical sciences and logic/philosophy. However, in the same remark Wittgenstein also acknowledges that his “study of sign-language” is not entirely safe from the risk of entanglement in “unessential psychological investigations” such as is often witnessed in the philosophy of logic, according to him. The risk is there, he says, because there is a study of thought processes that corresponds to his investigation just as it does to the investigations of those other philosophers whom this correspondence has misled into irrelevant psychological investigations. Thus, “. . . with my method too there is an analogous risk” (TLP, 4.1121). Wittgenstein’s acknowledgement of this risk brings into view certain complexities in his view of the relation between logic/philosophy and psychology.

Logic and Ideality

97

What Wittgenstein means by the psychological study that corresponds to his logical investigation would be a study of how thought or language, as characterized by logic, is realized or implemented in human beings. He refers to such studies in a letter to Russell, responding to a query about the Tractarian elements of thought and their relation to what thought represents: “I don’t know what the constituents of a thought are but I know that it must have such constituents which correspond to the words of language. Again the kind of relation of the constituents of thought and of the pictured fact is irrelevant. It would be a matter of psychology to find it out” (CL, 125). His view here seems to be that there are, on the one hand, purely logical questions about the nature of thought and language that can be answered without reference to anything particularly human. We can know on purely logical grounds that thought must have certain kinds of constituents that stand in a relation of reference to the constituents of reality and that the possibility of true/false linguistic representation depends on there being such constituents (cf. TLP, 6.124). But what those constituents of thought are, or how it actually comes about that they refer to the constituents of reality, is not relevant to logic. Thus, the Tractarian account of thought and language as true/false representation doesn’t yet tell us anything about how humans actually manage to represent states of affairs. Such questions would be answered through empirical studies concerning human beings, in particular, psychology. Nevertheless, as regards the relation between logic and psychology, it is important that, assuming that human thought is the only example of thinking available for us to study, in the end both logicians and psychologists are investigating the same phenomena, namely human thought and/or language use. Certainly, Wittgenstein maintains that there is a crucial difference between the perspectives of the logicians and psychologists. How that difference is to be understood can be explained on the basis of the preceding: in the Tractatus’ view the task of logic is to abstract away from the accidental, contingent features of language in order to clarify its common underlying essence. Rather than seeking to find generalities that pertain to empirically given cases and may be merely accidental, logic’s aim is to grasp something universal. What, in this view, is therefore meant to prevent the collapse of the logical investigation into a psychological one is its strict focus on what is necessary and essential, i.e., that which the possibility of particular cases presupposes, and what in this sense underlies those cases and is common to them. Thus, while psychology establishes generalities (laws governing human thought), logic is concerned with abstraction and what is universal (thought as such, not merely human thought). However, as Wittgenstein comes to realize later, this account of logic and the irrelevance of empirical facts for logic, is problematic. It may be that the Tractatus’ account of logic as abstraction is less mysterious than knowledge of logic is from a Platonistic point of view which leaves our access to Platonic forms unexplained. Nevertheless, the Tractatus’ abstractionist

98

Oskari Kuusela

apriorism suffers from severe difficulties of its own, as I will explain. Consequently, this way of explaining the nonempirical status of logic must be rejected. 2. IDEALITY AND SUBLIMATION: TURNING AWAY FROM CONCRETE CASES As the later Wittgenstein explains the problem with his early philosophy of logic, logic’s search for the underlying common essence of language and thought leads to a problematic sublimation of logic’s objects of investigation. Rather than understanding by a proposition anything concrete, a written or spoken sign used by someone somewhere sometime, a proposition is here conceived as an abstract, ideal entity of which actual instances of language are only impure manifestations. This turning away from concrete instances of language use, however, affects negatively logic’s capacity to clarify thought and language. Let’s look more closely into how this problem arises. Wittgenstein writes in 1932: In reflecting on language and meaning we can easily get into a position where we think that in philosophy we are not talking of words and sentences in a quite common-or-garden sense, but in a sublimated and abstract sense.—As if a particular proposition wasn’t really the thing that some person utters, but an ideal entity (the “class of all synonymous sentences” or the like). (PG, 121; see BTE, MS 114, 109/BTE, TS 213, 71v; TS 228, 139/TS 230, 2) Later in the 1930s he puts the point also like this: “When we ask: ‘What is a word?’, ‘What is a rule?’, etc. we would like to find, so to speak, entities more pure than what we commonly denote by these expressions. The latter [diese] we want to present as a kind of impurification of the former [jener]” (BTE, MS 152, 90). The way of thinking described in these quotes captures quite accurately the Tractatus’ conception that what signifies in a symbol, and what we should really understand by a name or a proposition, is something common to the different symbols that denote the same object or express the same sense. In this view, names and propositions mark classes of synonymous expressions, just as the first quotation says (TLP, 3.341–3.3411, 3.344). Here names and propositions are treated as abstract entities that exist independently of the use of any particular signs to express them, just as one might envisage the chess king as an abstract ideal entity that an actual chess figure only stands in for as a dummy representative. But by turning away from concrete instances of thought and language, logic enters a realm of hypothesized logical entities. It assumes the form of abstract theorizing whose difference from mere speculation becomes

Logic and Ideality

99

unclear. Wittgenstein describes this development in a notebook from 1937: “The ‘sublime conception’ forces me to move away from the concrete case since what I say doesn’t fit it. I now move into the ethereal region, talk of the real sign, of rules that must exist (even though I can’t say where & how)— and find myself on ‘thin ice’” (PPO, 173 (MS 183, 164); see MS 142, 88, 89; BTE, MS 152, 83, 84; PI, §§94, 107). By “the real sign” he means here signs in a sense that would satisfy logic’s account of the possibility of the expression of sense. For example, if logic requires that propositions must always have a determinate truth-value, as the Tractatus’ account does require, then propositions must also have a precise sense. Otherwise their truthvalue would not be determinate. But given that the propositions of everyday language are not precise in the required sense, apparently language use must involve such precise propositions at a deeper level. In this way the conception of propositions and names as abstract entities then generates further high-level theoretical questions about their identity. For even if it would be left for psychology to explain what actually happens when someone gives expression to or grasps a proposition, or uses a sign as a name, it would still be part of logic’s task, as understood in the Tractatus, to explain what this involves in principle. As the Investigations comments on the question about the identity of the “real sign”: “. . . Is it perhaps the idea [Vorstellung] of the sign? Or the idea at the present moment?” (PI, §105) The latter view that what really signifies is the sign mentally entertained at a particular moment seems to be how Wittgenstein thought of propositions at the time of the Tractatus. In any case, he seriously considered it during the book’s composition, as testified by the Notebooks. According to this account, although the meaning of a sign may be imprecise and ambiguous, it has a precise sense at the moment when a speaker says it and means it in a particular way (NB, 66–8). In this view, “the real sign” is the sign as interpreted by the mind and in this sense its “idea at the present moment”.5 Accordingly, by the “rules that must exist (even though I can’t say where & how)”, Wittgenstein means the kind of precise rules that the Tractatus postulated as underlying the use of everyday language. That there must be such rules we supposedly know on purely logical grounds relating to the possibility of the expression of sense. The Notebooks then outlines a possible answer to the “where and when” question: such precise rules are fixed in the mind of the speaker, when she means a proposition in a particular way. This explains how the early Wittgenstein was led away from natural languages and their actual uses in search for the ideal “real sign”. As a result, however, he found himself, as he puts it, on “slippery ice, where there is no friction” so that in a sense “the conditions are ideal”. This, however, also makes walking (see below) impossible. To walk “we need friction” and therefore must get: “Back to the rough ground!” (PI, §107) Interestingly, in a notebook from the period of the composition of relevant remarks of the Investigations, Wittgenstein also formulates the point about getting back to the rough ground as a matter of getting back to real or concrete examples (BTE,

100

Oskari Kuusela

MS 152, 84). Rather than assuming the form of abstract theorizing, logic needs to stay in contact with concrete examples and actual language use. Thus he rejects the account of logic’s objects of study as something sublime, as “concerned with sublime objects” (BTE, MS 152, 82). The point that we should not lose contact with concrete cases explains the somewhat opaque metaphor of walking in §107. The problem with the Tractarian approach is that, removed too far from concrete cases, logic ultimately loses its capacity to do real clarificatory work. It becomes unclear how the ideal linguistic items that logic is concerned with relate to the uses of everyday language, and how actual instances of language use are to be clarified in terms of the ideal concepts of the discipline of logic (MS 157a, 53v). Consequently, in these conditions of ideality we can’t do the very thing that logicians are supposed to do: to clarify thought and language. It is as if something as fundamental to logic as our ability to walk in our everyday lives were compromised. The problem with the Tractatus’ approach can be summed up as follows. The pursuit of common, underlying essential features and the purification of logic’s objects of investigation from anything empirical leads logic away from concrete cases of language use and thought. In this way the discipline loses contact with its object of investigation in the substantial sense of “the spatial and temporal phenomenon of language” (PI, §108). But herewith the investigation is on a path to become a mere “pursuit of chimeras” (PI, §94), speculation about a “non-spatial, non-temporal phantasm” (PI, §108). Ultimately, this loss of contact with concrete instances of language use compromises logic’s clarificatory capacity. In this way Tractarian apriorism as an explanation of logic’s distinctness from empirical investigations ultimately makes logic incapable of doing what its task is to do. If we accept Wittgenstein’s description of the problem, what sort of a solution does his later philosophy offer to it? In particular, if we regard logic as the study of concrete actual uses of language, doesn’t this force upon us a conception of logic as an empirical investigation? Are we obliged to choose between apriorism that regards logic as concerned with something sublime and ideal and empiricism? 3. THE NEW ROLE OF THE IDEAL AND “OUR REAL NEED” In order to comprehend Wittgenstein’s solution to problems with the Tractatus, it is important to understand the motivation for his early view. As explained in section 1, this view was partly motived by a need to explain the special status of logical principles that seem to be always already presupposed and relied on in any empirical investigation. However, there are also other relevant considerations. For the Tractatus’ (philosophy of) logic was a response to a need for clarity whose legitimacy Wittgenstein continues to

Logic and Ideality

101

recognize in his later philosophy. Accordingly, rather than simply rejecting his earlier conception, he intends his solution to constitute a correction to his earlier view, thus acknowledging its goals in certain respects. How Wittgenstein sought to satisfy the following two ideals of logic crucially contributed to the sublimation of language and thought in his early work. 1) Logic ought to be exact. The striving for exactness in the discipline of logic is connected with its aspiration for clarity in that greater exactness can remove confusions and misunderstandings, although this doesn’t mean that exactness is identical with clarity. As Wittgenstein notes, the aspiration for exactness can also create confusions. An example is striving for greater exactness than possible in a certain connection, consequently coming to misrepresent uses of language as more exact than they are. He writes about the notion of exactness in relevant notebooks: “Exactness is an ideal & connected with the removal of confusions” (BTE, MS 152, 90). And: The complete purity and clarity [insertion: crystal purity] of logic: it should be a crystal, nothing amorphous in it. . . . For the knowledge of the amorphous doesn’t interest us here (that is correct). And we also have to do with crystal systems. That is, we have to do with exactness. For (a) greater exactness removes certain misunderstandings. But there are also misunderstandings that easily come about when one strives for greater exactness. But it isn’t so that we could bargain something out of this crystal clarity! . . . (MS 157a, 66v-67r; cf. 48r) The latter quote is particularly interesting due to Wittgenstein’s acknowledgment that he is still in his later philosophy concerned with exactness, and that this striving for exactness and clarity isn’t something that could be given up. (I’ll return to this shortly.) This means that a novel way must be found to explain the sense in which logic is concerned with exactness and logical “crystal systems” that doesn’t lead to logic’s sublimation. For what drove Wittgenstein to theorizing about language in a sublimated and abstract sense was the seeming conflict between logic’s aspiration for exactness and the actual vagueness of everyday language. To meet the requirement of exactness it had to be assumed that beneath the vagueness of everyday language there must be a level at which language use is perfectly exact, governed by clear and precise rules. 2) Logic ought to be simple. This ideal, too, is connected with the discipline’s aspiration for clarity, as exemplified by the goal of reducing logic to the fewest basic notions and principles in order to make it maximally perspicuous. A case in point is the Tractatus’ conception of the general propositional form as the sole logical constant (TLP, 5.47; cf. 5.54 ff.). However, simplicity also seems a presupposition of the conception of logic as knowable a priori, at least in the strong sense that certain logical principles

102

Oskari Kuusela

underlie all experience. For to work out such an account it must be possible to determine the principles of logic once and for all, independently of whatever modes of language use or thought we might come across in future experience. In other words, it is not possible to reach the goal of reducing logic overall to a few basic principles, without assuming that a limited set of principles applies everywhere.6 This requires logic to be something simple in comparison to natural languages of whose uses it is impossible to have an overview, without reducing their uses to a limited number of simpler, allegedly more fundamental modes, such as referring and representing. With the latter, however, we are already on the path of sublimation (see note 5). Accordingly, as Wittgenstein puts it in the Investigations, the Tractatus’ view of logic as something “utterly simple” and “prior to all experience” presents logic as it were “of the purest crystal” (PI, §97). That everyday language doesn’t meet this requirement of simplicity thus leads us towards the view that logic speaks of language in a sublimated and abstract sense: “We have a tendency to see in logic a certain kind of ideal (simplex sigillum veri). And this causes it that we don’t want to understand by a word a sign of flesh and blood //printer’s ink// but something sublimated” (BTE, MS 152, 81; cf. TLP, 5.4541). Or, as Investigations §97 continues: “We are under the illusion that what is peculiar, profound, essential, in our investigation, resides in its trying to grasp the incomparable essence of language. . . .” When thinking about logic in the latter terms, what is profound in logic—the discipline’s uncompromising search for clarity—is interpreted as a matter of its disclosing to us something extraordinary: the underlying essence of language that meets the ideals of simplicity and exactness. This, however, is an illusion that involves as a crucial component what Wittgenstein believes is a misunderstanding of the role of the ideals of exactness and simplicity in logic. The situation can be described thus: a misunderstanding regarding the role of the ideals of exactness and simplicity pushed Wittgenstein towards a sublimated account of logic as something pure and devoid of anything empirical and language as an abstract entity with an existence independent of its particular impure manifestations (cf. MS 157a, 49r, 50r, 63r ff.). For although the aspiration for simplicity and exactness are perfectly legitimate as such, there is a problem with how the Tractatus sought to satisfy them. As Wittgenstein describes the mistake, he misunderstood the role that the ideals play in our talk about language and thinking in logic. He assumed that meeting these ideals would require that language and thinking actually satisfy the characterization of them in logic’s terms of simplicity and exactness, so that the ideal somehow must be found in reality. And since everyday language in its vagueness and complexity doesn’t meet the ideals of simplicity and exactness, he reasoned, it must meet them at an underlying level. However, this way of thinking, so to speak, reifies the ideals. What is essential to the investigation—that it presents us with something exact and simple—is now turned into an essential feature of logic’s object

Logic and Ideality

103

of investigation, i.e., language or thought, whereby they are now claimed to possess the characteristics of exactness and simplicity (see also section 6). Wittgenstein writes in the Investigations: . . .—But I want to say: we misunderstand the role played by the ideal in our language. . . . We want to say that there can’t be any vagueness in logic. The idea now absorbs us that the ideal ‘must’ be found in reality. Meanwhile we do not as yet see how it occurs there, nor do we understand the nature of this “must”. We think it must be in reality; for we think we already see it there. The strict and clear rules for the logical construction of a proposition appear to us as something in the background—hidden in the medium of the understanding. I already see them (even though through a medium), for I understand the propositional sign, I use it to say something. (PI, §§100–2) The question now is: is there another way to satisfy the ideals of simplicity and exactness? Is it possible to account for their role in logic without sublimating thought and language? Arguably, the methodological shift that constitutes Wittgenstein’s response to these questions is the key to his later philosophy of logic. Without a proper grasp of it, the latter is bound to seem a watering down of logic as anthropology of language, with Wittgenstein’s insistence on distinguishing between the logical and the factual dropped, and logic’s rigour lost. Rightly understood, however, Wittgenstein’s solution leads us beyond the opposition between empiricism and apriorism without landing us in conventionalism, as the next two sections explain. 4. TURNING THE EXAMINATION AROUND: IDEALIZATION IN LOGIC As Wittgenstein elucidates his methodological shift in earlier drafts of the Investigations and relevant notebooks, rather than to be gotten rid of, the ideal of crystalline purity is to be given a new position as part of logical investigation. This is achieved, as Wittgenstein says, by turning the whole examination around, where the metaphor describes a methodological repositioning that results in a novel way of understanding the role of the ideal. This conception is designed to prevent the ideal from figuring in our thinking as a dogmatic preconception about what language or thought must be,7 and leading to the sublimation of logic. Wittgenstein continues the explanation quoted at the beginning of section 3: “. . . The preconception that lies in [the ideal of crystalline purity] can only be eliminated by turning around our whole examination; and thereby positioning that purity in a new place”

104

Oskari Kuusela

(MS 157a, 67r-v, my square brackets; cf. MS 142, 90, 102; MS 157b, 2v; TS 220, 76; TS 239, 76). The Investigations explains: We see that what we call “sentence” and “language” has not the formal unity that I imagined, but is the family of structures more or less related to one another.—But what becomes of logic now? Its rigour seems to be giving way here.—But in that case doesn’t logic altogether disappear?— For how can it lose its rigour? Of course not by our bargaining any of its rigour out of it.—The preconceived idea of crystalline purity can only be removed by turning our whole examination round. (One might say: the axis of reference of our examination must be rotated, but about the fixed point of our real need.) (PI, §108) We can start unpacking this as follows.8 Assume that we have reached the conclusion that language can’t be claimed to possess the kind of underlying common essence and unity the Tractatus thought it had. That is, the view can’t be upheld that every proposition is a true/false representation and that language use is governed by strict and precise rules at an underlying level. If so, it looks as if the ideals of the simplicity and exactness in the discipline of logic can’t be met: the logic of language can’t be described in exact or simple terms. Firstly, now the option is no longer available that we explain the possibility of logic’s exact descriptions of language by postulating an underlying structure of fixed and precise rules. That is, we can’t explain away the conflict between the vagueness and fluidity of everyday language and the ideal of exactness by claiming that logic’s exact characterizations apply at a deeper level rather than to languages as we know them from everyday life. (See BTE, MS 115, 60, quoted below.) Secondly, if language doesn’t possess a common essence or formal unity, but it is a“family of structures more or less related to one another”, then its logic can’t be claimed to be anything simple. Now the complexity and diversity of the uses of everyday language can’t be explained away by claiming them to be reducible to more fundamental underlying modes of language use. The discipline of logic therefore appears to be losing its rigour: the concept of language and its uses can’t be defined rigorously in simple and exact terms, and no criteria can be given in advance of experience for deciding whether a given case belongs to language. To explain Wittgenstein’s solution to the problem of how to satisfy the ideals of simplicity and exactness, and to account for the possibility of the rigour of logic, let’s first note two related, according to him, unworkable conceptions. 1) To try to construe logic as a natural or empirical science would leave no place for the ideals (BTE, MS 152, 82–3). This is because, assuming that underlying linguistic/thought structures are not postulated, the actual complexity and vagueness of everyday language makes it impossible to satisfy the ideals of simplicity and exactness without compromising the empirical accuracy of the logical account of language. To assert that actual language uses are more exact and simple than they are would be to state

Logic and Ideality

105

something false.9 2) Similarly, it would be a mistake to infer that, because there is no underlying level of language at which the ideals are satisfied, we can only approximate them by reforming our extant language or constructing an ideal one. “But it isn’t so either that we could say: ‘The preconception is that the ideal exists, while we only bring reality closer to it.’ For therein too lies a misunderstanding. The role of the ideal is not grasped in that way” (MS 157a, 67v; cf. PR, 52; PI, §98). Problematically, this approach would still give us less than originally desired from logic. For it couldn’t clarify to us thought and language as we have been operating with them up to this point, i.e., our present concepts and inferences in natural language. Thus, our “real need” for clarity would not be met. Crucially, even if we did succeed in developing an ideal language or languages that meet the ideals of logic and started using them for the tasks of thinking in the future, the problem about the employment of such languages to clarify thoughts expressed in everyday language would persist. For, if everyday language is complex, vague and fluid, how can it be clarified in the exact and simple terms of an ideal language without misrepresenting it?10 However, it would be wrong to conclude from the failure of options 1) and 2) that the ideals of logic can’t be met, but at best approximated, and to accept that the discipline can’t have the desired rigour. Ultimately, the problem with the conclusion is this: logic can’t lose its rigour—its ability to clarify thought or language use in simple and exact terms, thus making whatever is said/thought crystal clear—without losing its identity as a discipline. The aspirations for exactness and simplicity are so central to the identity of logic that their yielding would mean logic’s demise. Hence, its rigour can’t be bargained away, for example, in the sense of logicians moving on to describing the uses of everyday language in their actual fluidity and vagueness. Not only is this to give up on the two ideals of logic, but it is to seek to justify the dumbing down of logic on grounds highly problematic from the point of view of the discipline. For how could empirical facts about the vagueness and fluidity of actual language use decide the faith of logic? Consequently, as Wittgenstein notes, one can always respond that logic in this dumbed-down sense is of no interest to one (MS 157a, 48r). In effect, such a dumbed-down logic would be a different discipline. Hence, a new way of understanding the ideals must be found that satisfies logic’s aspirations and “our real need” for clarity. This is what “turning our whole examination around” is meant to achieve. Essentially, to turn the examination around is to abandon the approach that treats the ideals of simplicity and exactness as requirements that thought and language must meet, and to adopt an alternative approach that understands exactness and simplicity as part of logic’s mode of examination, and as characteristic of its methods of clarification. A crucial issue is how to avoid falsely claiming that language uses are more exact or simple than they are. According to Wittgenstein, this falsification arises from presenting the idea that language is governed by strict and precise rules, as if

106

Oskari Kuusela

making an assertion about actual language use. But this description of the problem already hints at its solution. The problem can be avoided by putting forward the strict and precise rules, not as a claim about language, but as a particular way in which logic, for its purposes, seeks to describe the uses of language (see PG, 77; BTE, MS 140, 33). For it is one thing to say that, for the specific purposes of certain discussions, we can regard the use of expression “x” as governed by certain fixed, precise rules, and another thing to assert that it is actually thus used. While the former claim may be understood as abstracting from complexities that are not relevant for the purposes of the particular discussion at hand, or as idealization, the latter case can’t be thus conceived. It is simply a false claim about language. This can be elucidated with reference to a contrast Wittgenstein draws between idealization in science and in logic. As explained in section 2, according to Wittgenstein, logic’s object of investigation is not language as an ideal entity. The Investigations makes plain his misgivings with this view by emphasizing that in the philosophy of logic we speak of “sentences and words” in just the same sense as in everyday life when for example distinguishing a decoration from actual writing. In this sense the philosophy of logic is then concerned with “the spatial and temporal phenomenon of language”, not a “non-spatial” and “non-temporal phantasm” (PI, §108).11 That logic’s object of investigation is not language in an ideal or idealized sense distinguishes it from natural sciences that sometimes do abstract away features of reality—as when assuming a perfect vacuum or leaving friction out of account. Here what is said holds strictly speaking for reality in the ideal sense, not actual reality.12 By contrast, Wittgenstein maintains that, although idealization in logic may indeed involve the employment of idealized means of (re)presenting thought or language, such as calculi according to precise and fixed rules, this doesn’t mean that what is said holds only for an ideal/idealized language. Rather, in logic we compare language with calculi or games according to fixed and precise rules, and in this sense employ ideal/ ized means to speak about language, to represent and clarify its uses and to bring into sharper focus their specific aspects. The Investigations explains the contrast between idealization in science and logic as follows: . . . in philosophy we often compare the use of words with games and calculi which have fixed rules, but cannot say that someone who is using language must be playing such a game.—But if someone says that our languages only approximate to such calculi he is standing on the very brink of a misunderstanding. For then it may look as if what we were talking about were an ideal language. As if our logic were, so to speak, logic for a vacuum.—Whereas logic does not treat of language— or of thought—in the sense in which natural science treats of a natural phenomenon, and the most that can be said is that we construct ideal languages. . . . (PI, §81)

Logic and Ideality

107

Earlier versions of the remark put the point as follows: “One could at most say: ‘We construct ideal languages, in contrast to, say, ordinary language; but not, we say something that would only hold of an ideal language’” (PG, 77; see BTE, MS 140, 33). The object of logic’s statements, in other words, is not an ideal language so that what logic says would be correct for such a language only. Logic isn’t, in this sense, logic for an idealized reality—“logic for a vacuum”. Rather, logic constructs ideal languages that function as idealized modes of (re)presenting actual language, whereby simple and exact languages are employed to describe the logically nonideal, messy, complex and fluctuating reality of actual language use. Hence, in this sense: “. . . it must be said that this ‘ideal’ interests us only as an instrument for the approximate description of reality” (BT, 394; see TS 211, 490/BTE, TS 212, 727/BTE, TS 213, 253r). On Wittgenstein’s account, descriptions in terms of ideal languages then are no more than approximate descriptions of everyday language in all its complexity, fluidity and multiplicity of uses. Nevertheless, ideal languages with fixed and precise rules may still be employed to describe and clarify aspects of the use of everyday language that they do capture, or capture up to a point, while at the same time acknowledging that actual language is not used according to precise and fixed rules, and is not ideal in this sense. This is what Wittgenstein means by comparing the use of words with calculi or games with fixed rules in §81. The point is to employ a means of representation whose features are clear and under our control to render perspicuous particular characteristics of a much messier object of investigation, without pretending that the latter is simpler than it is. For, according to Wittgenstein, many of the philosophical problems we battle with arise from describing our concepts as simpler than they are (BB, 25–6; cf. Z §467; RC, §§3–4). This outlines Wittgenstein’s account of idealization in logic and philosophy. Idealization serves here, just as in science, a strictly methodological purpose (cf. MS 157b, 10v). For example, when physics abstracts from friction, this idealization involves no ontological claims in that, although a feature of reality is omitted from the physicist’s account, this is done without claiming falsely that reality wouldn’t include this feature. Consequently, it remains open for physics to include this feature in its accounts at another time. We can say that in such a case the physicist is talking about reality in an idealized sense, and her account only approximates actual reality. However, as Wittgenstein emphasizes, logic differs from natural science in this regard. In his view, the goal of logic is not to progress beyond idealized accounts and ultimately offer a comprehensive nonidealized one. For logic’s accounts in idealized terms are not merely approximate clarifications in the absence of more proper clarifications. Rather, clarification by means of ideal languages constitutes a particular method for resolving philosophical problems. This is possible in philosophy, because here one needs to take into account only those aspects of language use that are relevant to the solution. Not every feature of use is always relevant in this way, and in the case of different philosophical problems relating to a particular concept, different aspects

108

Oskari Kuusela

of the concept may require attention. As Wittgenstein explains, the role of words in language is described only as far as is required for the dissolution of philosophical difficulties (MS 121, 59v; cf. PI, §182).13 We are now in a position to understand the idea of turning the examination around. Yet another formulation sums it up like this: “It looks as if logic would lose the essential in it; its rigour. As if it had been bargained out of it. [New paragraph] But it now only plays another role. It has turned from a preconception concerning reality into a form of presentation. [New paragraph] . . . It has become a form of presentation and nothing besides that” (MS 157b, 5r). Thus, rather than prejudicially demanded of language as logic’s object of investigation, simplicity and exactness are to be recognized as characteristic of how the discipline of logic, for its specific purposes, seeks to (re)present thought and language. In search of clarity, logic does so in exact and simple terms. Rather than features of logic’s objects of investigation, simplicity and exactness are therefore to be understood as part of the form of examination of the discipline of logic. The ideals are principles that partly determine logic’s approach to the study of thought and language, and ought not to be construed as requirements thought and language must satisfy in order for “our real need” for clarity to be met (see CV, 31; MS 157b, 15v). Hence, instead of claiming that thought or language actually function like calculi according to precise and fixed rules, logical calculi should be recognized as particular modes of describing and clarifying thought and language. In this way we can abstract and idealize in logic without committing to any claims about the abstractness and ideality of its objects of investigation, and without making false, simplistic claims about language or thought. The point regarding the role of calculi is explained instructively from a slightly different angle in a remark on the application of logic and the relation of logical calculi to examples of actual language use: It is of the greatest importance that we always think of an example with a calculus of logic to which it has an application, and that we don’t give examples and say they are not really the ideal ones, which we don’t yet have. This is the sign of a false conception. (Russell and I have in different ways laboured on it. Compare what I say in the Tractatus about elementary propositions and objects.) If I can apply the calculus at all, this is the real application and the application that matters. . . . [O]ne can’t await for a dreamed up application. . . . The mistake lies in not allowing the calculus its real application, but promising it for an ideal case [replaced alternative: a foggy distance]. (BTE, MS 115, 55–6; cf. WA 4, 60 (MS 111, 118); MS 157b, 13r) The intended sense in which Russell worked under the false conception has to do with how his logic is ultimately a logic for a language in which we speak about the epistemologically privileged objects of acquaintance, not what we seem to be speaking about. It is at this level that logic finds its proper application for Russell, just as for the Tractatus logic is really logic

Logic and Ideality

109

for speaking about states of affairs made up of logically simple objects. But to thus conceive the application of logic is to hypothesize—to dream up—the application for it, and to lose contact with concrete real examples of thought and language (cf. section 2). Instead, the later Wittgenstein maintains, it is to concrete examples of language that logic is to be applied, if it is to have an application at all. Such concrete cases are what matters, and they should not be brushed off as not really proper examples because they show logical multiplicity which a logical calculus can’t accommodate. As he says in the previous quote in a sentence I left out: “. . . one doesn’t want to acknowledge the example as the real one, because one sees in it a multiplicity that the calculus can’t account for” (BTE, MS 115, 56). Crucially, despite shortcomings of this kind, the calculus might still capture what is essential about the function of relevant expressions from the point of view of the particular problems to be addressed and in the context of the particular discussion at hand. But if so, the inability of the calculus to account for those other not presently relevant features of the use of words doesn’t constitute a shortcoming. This is not a shortcoming, insofar as the calculus does capture whatever is essential to the clarificatory task in question. As explained, logic can idealize and abstract, thus focusing on what is relevant or essential for its purposes, and its having such foci is clearly distinguishable from saying something false. To illustrate, it is part of Wittgenstein’s later outlook, for example, that the Tractatus’ account of propositions as constituting a truth-functional calculus can in principle be used to clarify certain matters about the concept of a proposition, because it does capture particular aspects of how we use the word “proposition”, and part of the grammar of the connectives “and”, “or” as well as the word “no” (WA 2, 158 (MS 108, 52); WA 3, 255 (MS 110, 164)/BT, 67; BTE, TS 213, 83; WA 4, 54 (MS 111, 107)). Similarly, implication as defined in Russell’s logic, although it doesn’t, according to Wittgenstein, accurately capture the uses of “if . . . then” in everyday language, may still be regarded as helpfully capturing certain aspects of the use of this linguistic device (RPP I, §§269–72; RPP II, §§102–4; LW I, §§4–7). What is important is not the correspondence of logic’s concepts and principles with those of everyday language or the language targeted for clarification. What matters is that the logical account achieves the specific clarificatory aims it was designed for, and that we have a clear grasp of how its exact and simple terms relate to the locutions to be clarified. 5. BEYOND THE APRIORISM-EMPIRICISM-CONVENTIONALISM TRICHOTOMY The preceding account of idealization in logic supplies us with a way to describe the confusion of the Tractatus and that of other philosophers who have taken logic’s object of investigation to be something ideal. The mistake is, as briefly outlined at the end of section 3, to reify the ideals of simplicity

110

Oskari Kuusela

and exactness in logic. What ought to be seen as characteristic of logical investigation and its mode of (re)presenting thought and language is turned into a claim about the nature of thought or language as logic’s objects of investigation. In this way what is characteristic of logic’s mode of (re)presentation or examination is projected onto its objects of study. The objects of logical investigation thus become, so to speak, an embodiment of the ideals of logic, an idol of logic’s aspirations that is postulated because this seems required in order for the discipline to attain its goals. Here it is worth observing a parallel between this characterization of the Tractatus’ confusion and Wittgenstein’s account of the problem with Platonism. For this parallel offers a way to extend the description of the Tractatus’ confusion regarding the ideals of exactness and simplicity to cover certain kinds of accounts of the nontemporality, necessity and exceptionless generality (universality) of logical statements that partly motivate views about the ideality of logic’s objects of study. The problem with Platonism can be explained thus. A peculiar character of statements about essences (which they share with mathematical and geometrical statements) is their nontemporality. Unlike empirical or factual statements whose truth is relative to time and place, statements about essences involve no reference to time and place but remain valid, as it were, always and everywhere, without any exceptions, about relevant objects. Problematically, however, Platonism interprets this feature of the nontemporality of statements of essence as indicating the peculiar character of the objects spoken of, i.e., that the statements concern something nontemporal or eternal (forms or ideas) beyond the reach of time and the corrosion that reigns in the physical world.14 Thus, what could be more economically understood as a particular feature of the use of relevant kinds of statement, i.e., that they involve no reference to time, is explained in an ontologically costly way by reference to a strange characteristic of their objects. The point is that just this same problematic move is made when logic is taken to speak of language in the sense of an ideal entity. What is characteristic of the statements of logic, i.e., that they speak of thought or language in ideal terms, is interpreted as indicating the peculiar ideal nature of logic’s objects of investigation. As yet another illustration of what is at bottom the same confusion, consider a metaphysical interpretation of Frege’s idea of function-argument analysis of thought or language. Such an interpretation would consider the possibility of function-argument analysis as disclosing something about the essence of thought or language themselves, i.e., that it is of their nature to be made up of such elements and have a relevant kind of structure. But as in the case of the conception of the ideality of logic’s objects of investigation, here too a particular mode of (re)presenting (analysing) thought or language is projected onto the objects of investigation and turned into a claim about their nature. Consequently, function and argument now appear as two fundamental logical categories that are part of the makeup and nature of thought or language themselves.

Logic and Ideality

111

What is at issue, and what these examples are meant to illustrate, is Wittgenstein’s identification of a confusion pertaining to the notion of a priori, i.e., that of an a priori form or a priori order of things which could be the object of an a priori knowledge claim that expresses an a priori truth. He writes: “The ‘order of things’, the idea of form(s) of imagination/phenomena [Vorstellung] and therefore of the a priori is itself a grammatical illusion” (MS 157b, 1r-v). His view of the problem with the notion of a priori has already been outlined: it is a confusion to take what is a necessary, constitutive or characteristic feature of the philosopher’s/logician’s mode of (re)presentation to be a necessary, constitutive or characteristic feature of the object of (re)presentation. Certainly, a statement regarding a defining feature of a mode of (re)presentation does give the impression of providing us with a priori knowledge of the object of (re)presentation which is independent of the contingencies of experience. This is so because, when a particular mode, form or norm of (re)presentation is assumed, the possibility is excluded that anything in experience should conflict it. To fix a mode, form or norm of (re)presentation is precisely to organize what is encountered in experience in a given way, in accordance with the mode, form or norm of (re)presentation. Nevertheless, it is still a mistake to confuse a feature of the form of (re)presentation with a feature of the object. This confusion, as Wittgenstein notes, is entirely comparable to an optical illusion where one perceives objects as having the colour of the tinted glasses one is wearing (MS 157a, 57v-58r; MS 157b, 9v). Or as he puts the point in the Investigations, having just compared the logical ideal to a pair of glasses and referring to the Tractatus’ conception of language as his example: “We predicate of the thing what lies in the mode of representation. We take the possibility of a comparison, which impresses us, as the perception of a highly general state of affairs” (PI, §104; cf. 103, 114). What is meant by the possibility of a comparison here is, for example, the possibility of comparing language with a calculus, a comparison which makes possible quite impressive clarifications of aspects of language use, whereby the calculus functions as a mode of (re)presenting actual language use (see section 4). To “predicate of the thing what lies in the mode of representation”, however, is to make the further unlicensed assertion that language really is a calculus. This description of the confusion indicates a more down to earth alternative that doesn’t involve any presumed a priori claims about the necessary features of logic’s objects of investigation. Herewith Wittgenstein abandons the conception of logic as an a priori investigation whose goal is to discover the underlying a priori principles or forms that govern thought and/ or language. Nevertheless, as I will explain, this doesn’t mean that he would therefore settle for either of the two available alternatives, empiricism and conventionalism. He explains the switch away from apriorism as follows: “The a priori must become a form of examination [Betrachtung; deleted alternative: (re)presentation (Darstellung)]. That is, this concept must be

112

Oskari Kuusela

deprived of its nimbus. A proposition a priori arises through dressing up a proposition about the mode of (re)presentation in the form of a proposition about the (re)presented objects” (MS 157b, 3v; my square brackets; cf. Z, §442). The point should be clear on the above basis. Statements of so-called a priori necessities are not to be interpreted—dressed up—as statements about necessities pertaining to the objects of (re)presentation. Rather, necessary principles, such as every thought having a function-argument structure or every possible proposition possessing a certain general form, are to be recognized as what they are: as constitutive of the logician’s/philosopher’s mode of (re)presentation or examination. Hence, the apparent a priori statements about necessities are to be understood, not as statements about the objects of investigation, but as expressions of rules constitutive of modes of (re)presenting them. To illustrate, a statement of necessity, or a rule, such as “To be a proposition is to be a true/false representation”, is better understood, by Wittgenstein’s lights, as follows: not as stating an alleged truth about the essence of propositions which every proposition must fit, but as a principle determining a logician’s/philosopher’s mode of examination or (re)presentation. To adopt this mode of (re)presentation is to regard every proposition as a true/ false representation; from this point of view nothing counts as a proposition unless it is a true/false representation. However, when the statement is taken as a principle that determines a mode of (re)presentation, rather than as a philosophical/metaphysical claim about propositions, the need disappears to insist misleadingly and dogmatically that reality must conform to it, i.e., that every proposition must fit the account. For, importantly, understood as constitutive of a mode of (re)presentation, not as a claim about propositions, the statement is not falsified by counter-examples that don’t fit it. Such examples only indicate the limits of the applicability of the mode of (re)presentation, while the scope of its application may still remain very broad, and the account’s generality need not be damagingly threatened by the cases that don’t fit it. What matters is that the account can clarify the cases that it is used to clarify, not every possible case. Consequently, on Wittgenstein’s later account there is no longer any need to postulate an underlying level behind appearances where, it is claimed, the account does hold. (Neither is there any need for any other gerrymandering operations purported to make the phenomena fit the account.) After all, now the account doesn’t function as a philosophical/metaphysical thesis that can in principle be shown to be false by a single counter-example. In this way the account is deprived of its metaphysical nimbus, and the nature of logic’s statements of necessity explained in a more down to earth way. Still, because none of the clarifying power of the account of propositions is lost, but only the need to claim for it a greater generality than it has, Wittgenstein seems right to observe: “The ideal loses none of its dignity if it is posited as the principle determining the form of the approach [Betrachtungsform]” (CV, 31; see MS 157b, 15v; cf. WA 4, 60 (MS 111, 119–20); MS 157b, 56–7).

Logic and Ideality

113

Nevertheless, contrary to how it might appear, Wittgenstein’s alternative to apriorism in logic is not conventionalism in the sense of an account that explains logical necessity as dependent on conventions. For Wittgenstein’s alternative involves no claim about conventions (regarding modes or norms of (re)presentation) as the source of logical necessity. Rather, his view is only that linguistic conventions or so-called grammatical rules are the proper expression of logical necessity. Or as his view may be explained, he holds that statements about essences, or necessary characteristics constitutive of essences, should be understood, not as true/false factual statements about the objects they concern, but as statements of grammatical rules. As he puts the point, the correlate in language to a natural necessity—the sole thing we may hope to extract from a natural necessity into a proposition—is a grammatical rule (see PI, §372). Here the key to understanding Wittgenstein’s distinction between the two statement types is that statements of grammatical rules, unlike empirical or factual statements, can’t be justified by reference to facts spoken about. This point can be summed up thus: any claim about the in/ correctness of a mode or norm of (re)presentation that appeals to facts already assumes some mode of (re)presenting those facts. Consequently, such an appeal can’t provide an independent justification for the in/correctness of a mode or norm of (re)presentation (PR, 55; PG, 184–6). Still, this doesn’t mean that grammatical or logical rules couldn’t be understood as reflecting factual necessities pertaining to human beings, their environment or both. In this sense the source of necessities recorded in logic or grammar could sometimes be said to be empirical or factual regularities. An example is the exclusion of reddish-green from our experiential range and colour vocabulary, which Wittgenstein characterizes at the same time as: i) not a result of our linguistic habits, ii) as presumably correspondent to something physical (such as the makeup of the human brain), and iii) as geometrical, i.e., as having the status of a logical necessity whose proper expression is a logical/ grammatical rule (Z, §§351–9). It is in this rather complex sense that Wittgenstein characterizes essence as finding its expression or being articulated in grammar (see PI, §371). But to emphasize, this doesn’t mean that essences would be a grammatical creation or construction (see PI, §§370, 372–3).15 What emerges here is Wittgenstein’s view of the intertwinedness and interdependence of the logical and the factual, which is the final issue to be explained to complete my account of his overcoming of the apriorismempiricism-conventionalism trichotomy. Certainly, the later Wittgenstein insists just as adamantly as the Tractatus that logical statements can’t be understood as empirical or factual. As explained, according to him, this would make it impossible to account for the rigour of logic and to satisfy its ideals (section 4). Moreover, he maintains (see below), empiricism makes it impossible to explain logical necessity, leaving also inexplicable the nontemporal character of logical statements. Finally, as just outlined, in his view the principles or rules of logic can’t be justified by reference to facts. But although Wittgenstein therefore evidently is no empiricist, his rejection of apriorism

114

Oskari Kuusela

makes his later philosophy of logic fundamentally different from his early account in that it involves a recognition of the relevance of the empirical to logic absent from the Tractatus, as can now be explained. A central component of Wittgenstein’s rejection of apriorism is, as emphasized, his abandoning the conception that logic speaks of language as an ideal abstract entity, and his insistence that language is to be regarded as a spatial and temporal phenomenon (PI, §108). But here a problem arises. If language or its logic is not an ideal abstract entity whose existence is independent of the existence of particular actual languages, and logic isn’t simply to be stipulated either, it seems that we are forced to recognize language and its logical principles as somehow upheld by the speakers of language and dependent on actual language use. But if so, logic seems to collapse into anthropology or psychology (and so on), and to become part of human natural history. For how language users actually use language is an empirical matter, and insofar as the statements of logic are statements about speakers’ actual language use, they seem empirical. Thus, the admittance of logic’s dependence on actual language use threatens to make logical necessity unexplainable. As Wittgenstein remarks: “What you say seems to amount to saying that logic belongs to the natural history of humans and that is not compatible with the hardness of the logical must” (RFM, VI, §49; see MS 164, 149–50). This reasoning about philosophy’s collapse into empiricism is mistaken, however. The problem and Wittgenstein’s answer are explained in §242 of the Investigations (quoted below), where he discusses the consequences of the observation that the possibility of linguistic communication requires the speaker’s agreement, not only on the meanings of words (definitions), but also on judgments. By the latter he means agreement in language use in the sense of an agreement on facts spoken about or what is judged to be the case (cf. PI, §241; RFM, VI, §§39, 49; see MS 164, 149–51). And indeed, the possibility of language as a means of communication does seem to presuppose such an agreement in that, if we didn’t agree in our judgments in an overwhelming number of cases (especially basic ones, such as that the object before us is an apple and its colour green), communication would be thrown into disarray. On the face of disagreement about obvious facts, the question would immediately arise whether we mean the same thing by our words at all (RFM, VI, §39). Thus, agreement in definitions in effect amounts to and manifests itself as an agreement in judgments. (Even in a case where a language was never used, and no judgments were ever made to be agreed upon, the counter factual would apparently still have to hold that if the language were used, there would have to be an agreement on judgements.) Accordingly, as the point might also be put, a broad agreement in judgments is a precondition for the possibility of disagreement about other judgments that are less obvious to decide. The problem, however, now is that our agreement on judgments is a contingent empirical fact. Hence the existence of logic, in the sense of principles governing language and thought, seems to depend on a contingent empirical fact. And as noted, if the possibility of language is in

Logic and Ideality

115

this way dependent on how speakers as a matter of fact use it, doesn’t this mean that statements about the logical principles governing language use are empirical statements? Here is Wittgenstein’s remark: Part of communication by means of language is, not only agreement in definitions, but also (odd as it may sound) agreement 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) The solution to the problem, outlined in the penultimate sentence of the quote, unfolds from Wittgenstein’s account of the status of the statements of logic (or grammatical statements) discussed in the preceding sections. Let’s work our way back to it from this remark. To describe a method of measurement is to say what one must do to obtain a correct result of measurement, or for one’s activity to qualify as measuring at all. It is, in other words, a matter of stating rules or principles that govern measuring. But this means that such statements are not empirical true/false statements. That is, although it is certainly possible to make empirical statements about methods of measurement (to study them historically, anthropologically, and so on), this is not the function of statements of the methodology of measuring that explain how measuring ought to be done. Unlike empirical statements about practices of measuring, such statements of methodology are not falsified by the fact that people deviate from them, for example.16 The role of statements of logic resembles descriptions of method in the sense of methodological statements. They are not empirical statements about relevant linguistic practices but play a different role. As Wittgenstein notes in Investigations, having just said that philosophy of logic is concerned with “the spatial and temporal phenomenon of language”, the way we speak about language in the philosophy of language can be compared with how we speak about “chess when we are stating the rules of the game” (PI, §108). On Wittgenstein’s account, the statements of logic are therefore to be understood as stating rules for the use of language, whereby the rules governing the function of signs in a calculus, for example, can be straightforwardly compared with the rules of a game governing the function of its pieces. (A calculus can always be understood as a game according to rules—albeit not vice versa.) However, an important disanalogy between logic and statements of the methodology of measuring or those found in a rule book for chess is that logical descriptions, as explained in section 4, may involve idealization and abstraction. Hence, the rules of logic need not capture actual use exactly as it is, and consequently need not involve or imply any assertions about how one ought to use the expressions to use them correctly. This is so especially when the descriptions concern expressions with complex and fluid uses. For although logical rules can be given a normative function of

116

Oskari Kuusela

prescribing how we ought to speak/think, as in the case of syllogistic rules of inference, they may also be used for merely clarificatory purposes, as when analysing a calculus in terms of another one. Or as explained earlier, on Wittgenstein’s account, logical rules (or systems thereof) employed to clarify the function of expressions in language are to be understood as constituting modes of (re)presenting the function of those expressions, or models, with which actual use is compared (section 4). It is not the purpose of such comparisons to prescribe how language ought to be used, but to render perspicuous aspects of its use. However, it should now be clear that the statements of logic in this capacity are not empirical statements. They are not true/ false statements about the rules governing an actual linguistic practice, but statements of a rule constitutive of modes of (re)presenting actual language use employed in logic or constitutive of logical models. Wittgenstein’s view is summed up in the Investigations §130, where it is explained with reference to the employment of simple language-games to model actual, fluid and complex uses of language, and contrasted with the use of idealized models in science.17 (However, as just noted, a logical calculus can always be understood as a game according to rules, so that the same point applies also to calculi as models for language use.) As Wittgenstein explains, his “clear and simple language-games” do not constitute “preliminary studies for a future regimentation of language”, as if they were “first approximations, ignoring friction and air-resistance”. On the contrary, language-games are “objects of comparison” whose purpose is to clarify the characteristics of language by means of both “similarities and dissimilarities” (PI, §130). Hence, Wittgenstein’s account of logic doesn’t lead to logic’s collapse into anthropology, psychology or the history of the use of language, but the role of logical statements is clearly delineated from empirical statements. His conception of logic thus enables us both to regard language as a spatial and temporal phenomenon, not an ideal abstract entity with a problematic ontological status, and to account for logic’s ideality and exceptionless necessity. 6. NEW PERSPECTIVES: BEYOND ANTI-PSYCHOLOGISM To conclude, an important advantage of Wittgenstein’s later philosophy of logic, in comparison to the Tractatus, is how it makes it possible to avoid excessive abstraction that threatens logic’s ability to clarify language or thought. As explained (section 3), the Tractarian approach that regards language as an abstract entity ultimately leaves it unclear how the tools of logic (such the Tractatus’ truth-functional calculus) are to be employed to analyse concrete instances of thought or language use. Part of how Wittgenstein is led to this impasse is his early anti-psychologism: the way he tries to guard logic from empiricism or from falling into “unessential psychological investigations”. As explained (section 1), the Tractatus seeks to uncover a priori

Logic and Ideality

117

logical forms assumed to underlie the empirical phenomena of language by means of a method of abstraction, thereby treating actual linguistic expressions as merely secondary manifestations of underlying forms that are logic’s real interest. This abstraction culminates in the notion of general propositional form which is meant to set a framework for future applications of logic to analyse language and thought. But ultimately those applications, to which the Tractatus is a prolegomenon, don’t ever materialize, a failure for which the Tractatus’ apriorism is to blame. By contrast, Wittgenstein’s later conception of the role and status of logical models makes it possible to bring such models into intimate connection with concrete instances of actual language use. The purpose of logical models used as objects of comparison is precisely to make possible the clarification of actual uses of language as encountered in life and experience. This new account and method then enables Wittgenstein to acknowledge the multiplicity of the forms of language and thought—as he says, “Thinking is not a single thing [eines], but many different things [vieles Verschiedene]” (BTE, MS 152, 83; cf. RFM, I, §§133–4)—while at the same time protecting logic from empiricism. In this way the Tractatus’ view that logic must be understood as concerned with language or thought as such, an ideal, non-spatio-temporal phenomenon, is left behind. The account also provides Wittgenstein with an explanation of the possibility of logic’s exact and simple descriptions, i.e., how to satisfy the ideals of exactness and simplicity in logic. Hence, the rigour of logic is protected against the philosophical conclusion that it is impossible to satisfy it once the vagueness and fluidity of actual language use is admitted. Herewith new possibilities are opened up for logic beyond Wittgenstein’s early anti-psychologism that deems facts of nature or natural history as irrelevant to logic (section 1). Thus we arrive at a novel kind of a position that might be characterized as nonempiricist naturalism in the philosophy of logic. I conclude by briefly outlining this view, whose detailed examination must be left to another time, but whose possibility the preceding sections explain. From the point of view of Wittgenstein’s later conception it is, in a certain sense, irrelevant how a logical model for the functioning of expressions is arrived at or where its origin is. Accordingly, the origin of such a model might well be, for example, in natural historical observations relating to human beings, as in the case of Wittgenstein’s characterization of the role of first-person expressions of pain as an extension of primitive pain behaviour, such as crying and moaning. Like crying and moaning, so too certain (though not all) conventional first-person expressions of pain can be understood as manifestations of pain rather than descriptions of it (PI, §244). Clearly, it is an empirical statement about human natural history to say that language learning involves replacing natural, primitive expressions of pain with conventional expressions that still play a role similar to the natural expressions. Crucially, however, it is entirely possible to base a logical

118

Oskari Kuusela

model for the functioning of linguistic expressions on natural historical consideration without the model itself therefore constituting an empirical assertion about human natural history. Rather, such a model constitutes a mode of (re)presenting the function of relevant expressions, and it is to be employed as an object of comparison just as a calculus would, when put to work as a model for the functioning of linguistic expressions (see section 4). In this sense it is immaterial where the ideas for a logical model originate. Stipulated logical rules, actual and invented natural history are all equally possible bases for logical models (cf. PI, II, xii). What matters from the point of view of the philosophy of logic is the envisaged function or role of the model, and that when used as an object of comparison, the function of a natural history based model is not to make an empirical statement about natural history. This is evident, for instance, in how the generality of such a model differs from the generality of an empirical statement. If understood as an empirical claim, it would be clearly illegitimate to extend the account of the function of pain expressions in Investigations §244 to expressions of other sensations too, without offering further empirical evidence to support the extension. Nevertheless, such an extension is possible in the case of a logical model. The question here is merely whether the model can make comprehensible the function of those other expressions and help to resolve relevant philosophical difficulties. As this point about the generality and justification of logical models shows, their use differs from empirical assertions. Nevertheless, such models can still be characterized as naturalistic because of how they can take into account considerations pertaining to human natural history, and especially how such models can clarify modes of thought and language whose existence is part of human natural history. This is, in broad outline, what Wittgensteinian nonempiricist naturalism in the philosophy of logic comes to.18 REFERENCES Carnap, Rudolf. 1967. The Logical Syntax of Language. London: Routledge & Kegan Paul. Frege, Gottlob. 2007. The Foundations of Arithmetic. New York: Pearson Education. Kuusela, Oskari. 2008. The Struggle Against Dogmatism: Wittgenstein and the Concept of Philosophy. Cambridge, MA: Harvard University Press. ———. 2012. “Carnap and the Tractatus’ Philosophy of Logic”. Journal for the History of Analytic Philosophy 1, no. 3: 1–25. ———. Forthcoming. “The Method of Language-games as a Method of Logic”. Philosophical Topics. Russell, Bertrand. 2007. “The Philosophy of Logical Atomism”. In Logic and Knowledge: Essays 1901–1950, edited by R. C. Marsh, 175–283. Nottingham: Spokesman. Stokhof, Martin, and Michiel van Lambalgen. 2011. “Abstractions and Idealizations: The Construction of Modern Linguistics”. Theoretical Linguistics 37, nos. 1/2: 1–26.

Logic and Ideality

119

Wittgenstein, Ludwig. 1951. Tractatus Logico-philosophicus. Translated by C. K. Ogden. London: Routledge & Kegan Paul. (TLP) ———. 1958. Preliminary Studies for the “Philosophical Investigations”: Generally known as the Blue and Brown Books. Edited by Rush Rhees. Oxford: Basil Blackwell. (BB) ———. 1961. Notebooks 1914–1916. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (NB) ———. 1967. Zettel. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (Z) ———. 1974. Philosophical Grammar. Edited by Rush Rhees. Translated by Anthony Kenny. Oxford: Basil Blackwell. (PG) ———. 1975. Philosophical Remarks. Edited by Rush Rhees. Translated by Raymond Hargreaves and Roger White. Oxford: Basil Blackwell. (PR) ———. 1978. Remarks on Colour. Edited by G. E. M. Anscombe. Translated by Linda L. McAlister and Margarete Schättle. Oxford: Basil Blackwell. (RC) ———. 1980a. Remarks on the Philosophy of Psychology, vol. I. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RPP I) ———. 1980b. Remarks on the Philosophy of Psychology, vol. II. Edited by G. H. von Wright and Heikki Nyman. Translated by C. G. Luckhardt and Maximilian A. E Aue. Oxford: Basil Blackwell. (RPP II) ———. 1982. Last Writings in the Philosophy of Psychology, vol. I. Edited by G. H. von Wright and Heikki Nyman. Translated by C. G. Luckhardt and Maximilian A. E. Aue. Oxford: Basil Blackwell. (LW I) ———. 1992. Last Writings in the Philosophy of Psychology, vol. II. Edited by G. H. von Wright and Heikki Nyman. Translated by C. G. Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell. (LW II) ———. 1994. Wiener Ausgabe, vol. 2. Edited by Michael Nedo. Vienna: Springer. (WA 2) ———. 1995a. Wiener Ausgabe, vol. 3. Edited by Michael Nedo. Vienna: Springer. (WA 3) ———. 1995b. Wiener Ausgabe, vol. 4. Edited by Michael Nedo. Vienna: Springer. (WA 4) ———. 1995c. Ludwig Wittgenstein: Cambridge Letters, Correspondence with Russell, Keynes, Moore, Ramsey and Straffa. Edited by G. H. von Wright and Brian McGuinness. Oxford: Blackwell. (CL) ———. 1997. Philosophical Investigations. Second edition. Edited by G. E. M. Anscombe and Rush Rhees. Translated by G. E. M. Anscombe. Oxford: Blackwell. (PI) ———. 1998. Remarks on the Foundations of Mathematics. Edited by G. H. von Wright, Rush Rhees and G. E. M Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RFM) ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2003. Ludwig Wittgenstein: Public and Private Occasions. Edited by James C. Klagge and Alfred Nordmann. Lanham: Rowman & Littlefield. (PPO) ———. 2005. The Big Typescript: TS 213. Edited and translated by C. Grant Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell Publishing. (BT) ———. 2009. Wittgenstein Source: Bergen Text Edition. Edited by Alois Pichler, in collaboration with H. W. Krüger, D. C. P. Smith, T. M. Bruvik and V. Olstad, http://www.wittgensteinsource.org. (BTE, MS & TS)

7

Kant and Wittgenstein The Matter of Transcendental Arguments P. M. S. Hacker

There can be no doubt that careful study of Wittgenstein’s Nachlass sheds great light upon the meaning of many of his otherwise baffling remarks in the Investigations. In numerous cases, it definitively settles important interpretative controversies concerning what he meant by some critical expression (e.g., “Praxis”) or what he meant by some individual remark (e.g., philosophy “leaves everything as it is” (§124)). It also sheds much light upon the development of Wittgenstein’s ideas over the sixteen years of work on his second masterpiece. For not everything changed at once, and progress in achieving clarity developed now on one front, now on another. Moreover, in some cases different ideas were tried out and then abandoned (e.g., the symptom/hypothesis relation, the methodological solipsism of the Philosophical Remarks, the distinction between “‘I’ as subject” and “‘I’ as object” in the Blue Book), and various routes through the landscape were found to be dead-ends. The availability of Wittgenstein’s Nachlass enables us to trace the genesis of many of the ideas in the Investigations, and that in turn illuminates their significance and implications. Sometimes it also enables us to pinpoint influences (positive and negative alike) on the development of his thought. In the following essay I shall explore the question of the relationship between Wittgenstein’s philosophy and Kant’s. Is there any sign that Kant influenced Wittgenstein, either in the Tractatus or in the Investigations? Is there any sign of a Kantian influence, perhaps transmitted through others, such as Schopenhauer? And if the answer to both these questions is negative, are there any affinities between Kant’s philosophical methods and Wittgenstein’s? 1. TEXTUAL EVIDENCE It has sometimes been suggested that both in the Tractatus and in the Investigations, Wittgenstein advances what amount to forms of transcendental arguments in a broadly Kantian spirit. After all, does he not argue in the

Kant and Wittgenstein

121

Tractatus that there must be simple objects if representation by means of language is to be possible? Or that there must be facts—the obtaining or non-obtaining of states of affairs—if propositions are to be possible? These are conditions of the possibility of logic and language—of thinking (reasoning) and representing. And is that not a transcendental argument? Similarly, in the Investigations, does he not argue that a community of speakers must exist as a condition of the possibility of a language? Or that there must be other subjects of experience as a condition of the possibility of selfascription of experience? And are these not transcendental arguments?—It depends on what is to be called a transcendental argument; and on what exactly Wittgenstein was arguing. We shall see. In this paper, I shall give a schematic overview of similarities and differences between Kant and Wittgenstein. There are important similarities. Nevertheless, I shall argue, if we are to take the term “transcendental argument” seriously, then there are no transcendental arguments in Wittgenstein. If we take the term loosely, in a manner that would have been objectionable to Kant, then one can make out a case for characterizing the Tractatus as employing transcendental arguments. But no such case can be made out for the Investigations. First, let me briefly detail what we know of Wittgenstein’s acquaintance with Kant’s work and his attitude towards Kant’s thought. In the pre-Tractatus notebooks, Kant is mentioned once (NB, 19.10.14): the theory of tautologies, Wittgenstein remarks, will shed light on Kant’s question “How is pure mathematics possible?” In Tractatus 6.36111, Kant’s problem of right- and left-hand incongruence is discussed. Wittgenstein read the Critique of Pure Reason with his friend Ludwig Hänsel in prisoner of war camp at Cassino in 1919. As far as I know, there is no information about how much of the book they read together, or what Wittgenstein made of it. In Wittgenstein’s post-Tractatus Nachlass, Kant is mentioned in but two remarks. In one, he writes that the limits of language show themselves in that the only way to specify the fact that agrees with a true proposition is to repeat the proposition. Here, he notes, “we’re concerned with the Kantian solution to the problems of philosophy” (TS 211, 173). In MS 107, 183 (= BT, 672), he remarked, “Isn’t what I’m saying here what Kant meant by saying that 5 + 7 = 12 is not analytic but synthetic a priori”. In TS 209, 45, he elaborated: that an equation cannot be reduced to a tautology explains what Kant meant by claiming that propositions of arithmetic are not analytic but synthetic a priori. In his students’ notes and memoirs I have found only two comments on Kant. The important one was made to Desmond Lee apropos Broad’s classification of methods in philosophy. The Transcendental Critical Method, Broad had said, is Kant’s, but without the peculiar applications Kant made of it. To this Wittgenstein responded with some enthusiasm: “This is the right sort of approach. Hume, Descartes and the others had tried to start with one proposition such as ‘Cogito ergo sum’ and work from it to others.

122

P. M. S. Hacker

Kant disagreed and started with what we know to be so, and went on to examine the validity of what we suppose we know” (LWL, 73).1 The catch from this trawl is meagre. It does not suggest any significant Kantian influence on Wittgenstein or even any evident Kantian inspiration. In 1931, he made a careful list of influences upon him (MS 154, 16r). Although Schopenhauer was mentioned (obviously with Tractatus 5.6-s and 6.4-s in mind), Kant was not (although, of course, there may be a “Kantian influence” via Schopenhauer). Nevertheless, there are striking affinities between certain aspects of Kant’s philosophy and Wittgenstein’s, both early and late. These do not necessarily betoken influence—perhaps merely a partial convergence of route through the jungles of philosophy. But the convergence is certainly of interest, and worth spelling out. i. Metaphilosophy: No other philosophers in the history of the subject have been so preoccupied, and so fruitfully preoccupied, with the nature and status of philosophy itself. Both agreed that philosophy (or, as Kant put it, “pure philosophy”) is not continuous with the natural or mathematical sciences. Both argued that it is a second-order, reflective discipline. ii. Dialectic: No other philosophers in the history of the subject were so preoccupied with the Dialectic of Reason—the logic of conceptual illusion. Both agreed that there are more or less systematic patterns to philosophical error, and that clarification of the sources of conceptual confusion is of capital importance. Moreover, both concurred that the most important way of so doing is to identify the unquestioned assumptions underlying philosophical controversies and challenging them. iii. The bounds of sense: Both were concerned with characterizing the bounds of sense. Kant endeavoured to do so by a “deduction” (a justification of a right by reference to its sources) of a priori concepts, which investigates the a priori conditions of their use in judgements and limits their intelligible application to possible experience. Wittgenstein took a different route, namely by investigating the conditions for the meaningful use of language. iv. Rationalism and empiricism: Both philosophers had a highly critical attitude towards the rationalist and empiricist traditions. Both repudiated foundationalist epistemology of the Cartesian or Lockean kind—our knowledge of how things are in the world around us is not inferred from how things sensibly seem to us to be. Both rejected Cartesian as well as Humean conceptions of the mind—the mind is neither an immaterial substance nor a bundle of perceptions. v. The nature of necessity: Both were preoccupied with clarifying the nature of necessity. Both denied that there are any de re a posteriori necessities to be discovered from experience. Kant held that the necessary truths of logic are “entirely without content”; Wittgenstein argued that the tautologies of logic are “senseless (i.e. have ‘zero sense’)”. They both denied that the necessary propositions of arithmetic and geometry are analytic, and agreed that pure arithmetic is concept-construction. More generally, both

Kant and Wittgenstein

123

located the roots of nonlogical necessity in us—albeit for different reasons and in a very different sense. vi. Rational theology: Both philosophers repudiated rational theology. Doubtless other points of convergence can be found. Nevertheless, these suffice for our purposes. They are striking and important. But when examined proximately, the sense of convergence changes—for two reasons. First, the negative, critical consensus is not matched by constructive agreement. Secondly, the world-view, the philosophical Weltanschauung, that informs their thought is utterly different. The most fundamental source of disagreement turns on the deepest roots of Kant’s inspiration. The master-problem of the Critique of Pure Reason is: How are synthetic a priori judgements possible? (B 19). And this, putting aside the investigation of the possibility of the synthetic a priori truths of mathematics, is glossed as “How is metaphysics as a science possible?” (B 22). His critical first step towards resolving this question was his so-called “Copernican Revolution” (cf. B xvi; xxiia). All previous systems of metaphysics, he thought, had assumed that all our knowledge must conform to objects. But they proved impotent to explain the possibility of synthetic a priori knowledge of nature, e.g., that substance must persist over time, that objects must stand in reciprocal causal relations, or that every event must have a cause. Consequently, Kant proposed that the investigation be turned around. We should suppose that nature, insofar as it can be known a priori, must conform to the conditions of our sensible and cognitive constitution— to the a priori forms of intuition and the a priori categories of understanding. His radical conclusion was that knowledge cannot transcend the limits of possible experience.2 We can know synthetic a priori truths about nature (the world as we experience it), but we cannot attain knowledge of the existence of God, of the immortality of the soul, or of things as they are in themselves. Synthetic a priori knowledge of nature is possible, Kant thought, because the mind imposes structural principles on nature as a condition of possible experience. “Nothing in a priori knowledge can be ascribed to objects save what the thinking subject derives from itself” (B xxiii), for “we can know a priori of things only what we ourselves put into them” (B xviii). Synthetic a priori judgements can be shown to be possible if we relate the formal conditions of a priori intuition, the synthesis of the imagination, and its necessary unity in a transcendental apperception to a possible cognition of experience in general, and say: The conditions of the possibility of experience in general are at the same time conditions of the possibility of the objects of experience, and on this account have objective validity in a synthetic judgement a priori. (A 158/B 197) The mind makes the scaffolding of nature. It is on this scaffolding that appearances are perceived. The scaffolding is described by the synthetic a

124

P. M. S. Hacker

priori propositions of metaphysics. And it is in virtue of this scaffolding that empirical knowledge of nature is possible. Although he did not hesitate to use the term “a priori”, Wittgenstein, after his return to philosophy in 1929, held that its employment in the philosophical tradition was riven with the deepest of misconceptions. He wrote (in 1931): A characteristic of theorists of the past cultural era was wanting to find the a-priori where there wasn’t one. Or should I say it was characteristic of the past cultural era to form //to create// the concept, or nonconcept, of the “a priori”. For it would have never created the concept if from the start it had seen things //the situation// the way we see it. (A great—I mean, significant— error would then have been lost to the world.) But in reality one cannot argue like that, for that //this// concept was founded in culture itself //in the whole culture//. (MT, 89 (MS 183, 81)) This is an important observation with ramifying implications. Wittgenstein did indeed use the term “a priori” in his later philosophical writings, but there is a gulf separating his use of the expression from the Kantian conception of a priori knowledge. To be sure, he thought that the propositions of mathematics and logic are a priori. Like Kant he thought them vacuous. He thought that apparently synthetic a priori propositions such as “nothing can be red and green all over” or “time-travel is impossible” are a priori. But he did not think that to know one of these (nonanalytic) a priori propositions was correctly characterized as knowing the truth of a description of how things necessarily are in nature. Such propositions are not a priori descriptions of the scaffolding of the world. Rather, they are norms of description. The world has no scaffolding—neither original (traditional metaphysics), nor constructed and imposed (Kantian metaphysics of experience). Such (apparently synthetic a priori) propositions constitute the scaffolding FROM which we describe the world. So, such knowledge is knowledge of rules of representation. To know that red is darker than pink, for example, is precisely to know that if anything is red, then, without looking to see, one may infer that it is darker than anything pink. This “synthetic a priori proposition”, this apparent necessary truth about nature, is no more than an inference-rule in the guise of a description. It is an inference-rule that is partly constitutive of the meaning of the colour predicates involved. Consequently, Kant’s question: “How is synthetic a priori knowledge of nature possible?” crumbles in Wittgenstein’s hands. Kant, he would surely have argued, was mistaken to think that synthetic a priori propositions correctly describe how things necessarily are. For, in Wittgenstein’s view, what appear to be necessities of nature, and what Kant argued to be a priori principles that the understanding imposes upon intuitions to constitute nature, are no more than shadows cast upon nature by the grammar of our language. I shall elaborate below.

Kant and Wittgenstein

125

Nevertheless, Wittgenstein undertook his own Copernican Revolution. In reaction to his early philosophy of the Tractatus, he declared that “the inquiry must be turned around, but on the pivot of our real need” (PI, §108).3 We must turn from the sublime to the mundane, from Wesensschau to grammar, from the essence of the world to the “quiet weighing of linguistic facts” (Z, §447). The calculus of logic is not a mirror image of the a priori order of the world, nor is it the depth-grammar of any possible language.4 It does not lie hidden beneath the surface grammars of human languages; it is a grid to be placed upon them for checking the deductive validity of certain kinds of inference. It is a form of re-presentation. Wittgenstein’s Copernican Revolution involved repudiating the very idea of a priori knowledge of the world as understood effably in the seventeenth to nineteenth century and as conceived ineffably in the Tractatus. The young Wittgenstein had thought that the ineffable truths of the Tractatus concerning the essential nature of things that is shown by the well-formed sentences of a language were conditions of the possibility of symbolic representation and thought (reasoning). The later Wittgenstein, by contrast, held that what appears to be the metaphysical order of things—effable or ineffable—is an illusion. There is no metaphysical order of things. The a priori, he said, must have its nimbus removed (MS 157b, 3v). It seems to describe adamantine necessities informing the world we experience. But although such a priori propositions look like descriptions, they are actually expressions of norms of representation. The a priori also seems to exclude possibilities—for example, disembodied minds and time-travel. But logico-metaphysical impossibilities are not possibilities that are impossible. What these a priori propositions exclude are not possibilities; what they exclude are not intelligible impossibilities either. They exclude forms of words and forms of inference. Forms of words are excluded as nonsense. And forms of inference (e.g., from “x is white” and “y is black”, infer “x is darker than y”) are excluded as incoherent. 2. ARE THERE TRANSCENDENTAL ARGUMENTS IN THE TRACTATUS? It is clear enough why one might think that the metaphysics and ontology of the Tractatus are in effect established by a form of transcendental argument. For does Wittgenstein not argue that the propositions of logic “presuppose that names have meaning and elementary propositions sense” (TLP, 6.124)? But the meanings of names are the simple objects for which they stand (TLP, 3.203). Simple objects are the substance of the world—that which subsists independently of what is the case (TLP, 2.024). If there are simple objects, then there are states of affairs—possible configurations of objects. An actual configuration of objects is a fact. Facts are the obtaining or non-obtaining of states of affairs. The world is the totality of facts (TLP, 1.1). Does this not appear to be a transcendental argument—from the existence of propositions of logic to the essential nature of the world?

126

P. M. S. Hacker

Furthermore, one can run the argument a different way with equally dramatic conclusions.—If there were no simple objects, then whether a proposition had sense would depend upon whether another proposition was true (TLP, 2.0211). That would generate an infinite regress. It would also make sense depend upon the facts. But sense must be independent of what is actually the case. So there must be simple objects that are the meanings of simple names. And the logical form of simple names (their combinatorial possibilities in accord with logical syntax) must be the same as the metaphysical combinatorial forms of the objects that are their meanings. What is possible in language must neither exceed nor fall short of what is possible in reality. Language and reality must have the same logical multiplicity. That is a condition of the possibility of representation. Again, unless there were simple objects, then sense would not be determinate (TLP, 3.23). But if the senses of sentences were not determinate, then the law of the excluded middle would not apply. But bipolarity is constitutive of being a proposition with a sense. So unless there is an a priori guarantee of determinacy of sense, there could be no propositions—hence no representation, and no logic either. But we do represent things to ourselves, and there are propositions of logic. So there are simple objects. So there are states of affairs. So there are facts. And these are not contingent truths. Logic, Wittgenstein averred, is transcendental (TLP, 6.13). It is a condition of the possibility of thought (reasoning). But we can think (reason). So whatever is presupposed by logic must be the case. Now, is this not a battery of transcendental arguments that prove how things must be in reality from considerations that pertain to things we can do and know we can do—namely reason, think, represent things to ourselves? This is a tempting conclusion. But I think it is at best misleading. It depends on how we are to conceive of a transcendental argument. One common current way of construing the general form of a transcendental argument is: We can V (or: We do V); Unless things are thus and so in reality, then we would not be able to V; so, Things are thus and so in reality. Or, even more schematically—a straightforward modus tollens (and one could throw in a couple of modal operators and the axiom “p ⊃ ◊p”): p if ~q, then ~p ∴q Kant did indeed hold that a transcendental argument is a form of argument, but he did not hold that it is a formal argument. To be sure, if this is

Kant and Wittgenstein

127

the general form of a transcendental argument, then the Tractatus offers us transcendental arguments. And so indeed does any philosophy that argues from the indisputable character of logic, thought, language or experience to how the world must be. However, there are three objections to this common construal. First, very many philosophers have argued in this vein. But it is not obviously fruitful or illuminating to extend this Kantian term of art thus. Historians of philosophy have not found it helpful to characterize Plato’s or Aristotle’s metaphysics (not to mention all the others who have argued from the nature of language or thought to the nature of the world) as involving transcendental arguments. Secondly, so to construe transcendental arguments renders them, as it were, an a priori form of scientific arguments to the best explanation. Astronomers argued thus: If there is a deviation in the orbit of a planet, the best explanation is the presence of another planet exerting gravitational force upon it; There is a deviation in the orbit of planet X; so, The best empirical explanation is that there is a planet Y exerting gravitational force upon planet X. Such empirical arguments to the best explanation are confirmable (Uranus) or infirmable (Vulcan) in experience. Now, one might argue that many forms of metaphysics employ arguments to the best a priori explanation, e.g., arguments for the existence of Platonic Ideas as the best explanation for the possibility of predication, of a First Mover as an explanation of the ultimate cause of all motion, of Leibnizian monads, and so forth. But such arguments, unlike those in the sciences, are neither empirically verifiable nor falsifiable. Moreover, such meta-physicists commonly confused a contingent conclusion’s (necessarily) following with a necessary conclusion’s following, i.e., conflating “Necessarily, if [if p then q] and p; then q”, with “If [if p then q] and p, then necessarily q”. For to be sure, they typically wished to demonstrate necessities, not contingencies, in reality. In this manner, one might unkindly argue, metaphysics mimics the methods of science to its own detriment. But what Kant called a “transcendental proof” does not really look like this at all. He was not concerned with inferences to the best explanation (as Locke was, in inferring the existence of objective material particulars to explain the character of our ideas). In the “Analogies” his concern is with demonstrating that we can and do know synthetic a priori truths of nature, such as the Principle of Causality. In the “Transcendental Deduction” he tries to demonstrate the a priori validity of the categories—that the categories necessarily apply to any possible experience. And even in the “Refutation of Idealism”, what is shown is that inner experience presupposes

128

P. M. S. Hacker

outer experience of objects as a condition of its possibility—not that the existence of objects is the best explanation of inner experience. Nor was he concerned, as rationalists had often been, with transcendent, nonempirical, conditions of experience and its objects, since he denied the applicability of the categories beyond the domain of possible experience. Thirdly, this construal distorts Kant’s idea of a transcendental proof in two ways. First, precisely because he is concerned with proving synthetic a priori truths of metaphysics, i.e., propositions the concepts of which are not analytically connected, he argued that they are connected by a “third thing”, namely possible experience. Reference to possible experience is an essential ingredient in a transcendental proof, but missing in the above (modern) construal of its form. Secondly (and consequently), Kant quite clearly thought that any such proof demands the truth of transcendental idealism (A 130)—it is “the sole means for solving [the] problem [of synthetic a priori knowledge]” (Prolegomena, 377). So, if we construe “transcendental argument” as Kant himself did, then the Tractatus is not engaged in transcendental argumentation. For (i) It is not an attempt to vindicate the possibility of synthetic a priori knowledge. In fact, it denies that there is any such thing, since it holds that all expressible necessity is logical necessity. Insofar as there is metaphysical necessity, it is inexpressible in propositions with a sense, and so cannot be the content of propositional knowledge. (ii) The Tractatus does not link independent concepts by reference to possible experience and the a priori conditions of its possibility. (Rather, it links material concepts with the formal concepts that are in effect variables of which the meanings of the material concepts are values.) (iii) The Tractatus does not attempt to prove that the world of appearances (outer experience) is a condition of the possibility of inner experience. That objects exist and that there are states of affairs is a presupposition of logic, not of experience. Of course, some may think that transcendental idealism is an unacceptable doctrine, but nevertheless hold that the bare idea of a transcendental argument can be salvaged from it. And the form in which it can be salvaged is the above cited form in which we argue from something we can do (and indisputably know that we can do) to how things must be in reality for it to be possible for us to do what we can do. If so, they may argue, then there are, in this diluted sense, transcendental arguments in the Tractatus. Certainly, in the Tractatus, Wittgenstein did argue from the fact that we can do certain things (represent things to ourselves, reason), that the world must have a certain scaffolding—a necessary structure and necessary forms. (Of course, he also went on to argue that such putative metaphysical statements about how things necessarily are transgress the bounds of sense in the

Kant and Wittgenstein

129

attempt to say what can only be shown.) So, if we construe a transcendental argument to be an attempt to establish how things necessarily are in reality from considerations concerning what we do and know we can do, and if we disregard the ineffability of the metaphysical theses of the Tractatus, then, in a watered-down sense, one may say that the Tractatus makes use of transcendental arguments. It argued that things must be thus and so in reality, otherwise we would not be able to represent things and to reason validly from one propositional representation to another. Nevertheless, it is clear that Wittgenstein adamantly repudiated even this in the 1930s. For it is this that he referred to as dogmatism (cf. MSS 111, 87, 119 ff.; MS 115, 57; MS 130, 53). It is the dogmatism of projecting features of our method of representation onto the objects represented and then insisting that they must be thus and so, otherwise we wouldn’t be able to . . . 3. ARE THERE TRANSCENDENTAL ARGUMENTS IN THE INVESTIGATIONS? When Wittgenstein initiated his Copernican Revolution in the early 1930s, he abandoned the thought that necessary, but ineffable, truths of metaphysics were presupposed by logic and the possibility of symbolic representation. In the first place, as we noted above, what looked like ineffable metaphysical truths that are shown but cannot be said, are—insofar as they are correct —rules of representation. To know that red is darker than pink, for example, is no more than to know an inference-rule: that if A is red, and B is pink, one may infer that A is darker than B. This rule of inference is partly constitutive of the meanings of the words “red”, “pink” and “darker than”. Hence to know that red is darker than pink is not to know a synthetic a priori proposition about nature, but rather to know a rule. So too, to know what Kant thought of as synthetic a priori propositions of metaphysics, for example that every event has a cause, would (if Kant were right) be no more than to have mastered the grammar of discourse concerning events. If what seemed to be synthetic a priori propositions that are conditions of the possibility of empirical cognition are no more than norms of representation, then the question of how it is possible for us to have knowledge of such truths simply fades away. For there is no great mystery here. (There are no mysteries in philosophy—only mystifications.) For our knowledge of such propositions is just knowledge of the rules of our language—knowledge of our own form of representation. If the law of causality is indeed a part of our form of representation, then to know that every event has a cause is just to know that if something is described as being an event, it may be inferred that it had a cause. If the inference from “E is an event” to “E was caused” is a priori legitimate, that would be because it is partly constitutive of the meaning of “event”. What Kant thought to be synthetic a priori truths describing necessary constraints upon reality are merely what Wittgenstein

130

P. M. S. Hacker

called “grammatical propositions” seen through a glass darkly. A grammatical proposition is a rule of representation in the guise of a description of how things necessarily are. Since grammar, in the sense in which Wittgenstein used the term, is an interwoven network of rules that are partly constitutive of the meanings of expressions in our language, our knowledge of—or perhaps: our grasp of— grammatical propositions is manifest in our linguistic practices. We use the words of our language in accordance with these norms of representation. We draw inferences in accordance with them. If anyone reasons “A is red, B is pink, so A is lighter than B”, we would not understand them. We should correct them, or ask what they mean. For we employ such grammatical propositions as standards for the correct uses of words and as measures of valid inferences. We invoke them in teaching our children (“No, no! If it’s red all over, it can’t also be green all over”, we may say). And we refer to them in explaining to foreigners how expressions in our language are used (“If something is one foot long, then it’s not (can’t be) shorter than twelve inches”, we might explain). It is a pair of cardinal (grammatical) insights of Wittgenstein that there is no such thing as inferring facts about the world from corresponding rules of grammar, or of inferring rules of grammar from corresponding facts about the world. This is part of what he meant by his insistence upon what he called “the autonomy of grammar”.5 No fact, no contingent fact-stating proposition, entails the content of a rule of representation. Conversely, the content of a rule of representation does not entail that things are thus and so in reality. That red is a colour is a rule of inference (i.e., “A is red ⊢ A is coloured”). That red is a colour does not entail that some object or other is red or coloured. This rule of inference—the grammatical proposition in the guise of a description—is not made true by the fact that red is a colour. For it is not a fact that red is a colour. If it were a fact (a matter of fact) that red is a colour, we should know what it would be for red not to be a colour. But we do not—for that is a nonsensical form of words. The autonomy of grammar, however, is perfectly compatible with the idea, which Wittgenstein certainly held, that the existence, employment and usefulness of certain grammatical forms presupposes a large variety of general facts about us, about the world we live in and about our engagement with it. It presupposes a host of regularities in the world, as well as regularities in our nature and behaviour. But it would be absurd to suppose that it is the task of philosophy to prove the existence of such regularities from the existence of specific norms of representation. (As if one might infer, or even want to infer, the existence of human beings from the existence of the rules of chess!) In the first place, we already know the relevant truths and do not need to infer them from anything. Secondly, Wittgenstein—unlike Kant— would have repudiated the very idea that it is or could be part of the task of philosophy to prove the existence of anything (least of all of the “external world”). What it makes no sense to prove (that I exist, that the “external

Kant and Wittgenstein

131

world” exists) cannot stand in need of a proof. What philosophy has to do is examine the challenges launched by sceptics of one kind or another, and show why the reasons advanced for thinking that we do not know what we take ourselves to know in these respects (e.g., that there are “material objects” in the room, that there are “other minds” in the Common Room) are misconceived.—But didn’t Moore try to prove the existence of the external world? And, more recently, didn’t Davidson try to prove the existence of events? Indeed they did. As Friedrich Waismann remarked: what can one say, save that they are great provers before the Lord. One might object: does Wittgenstein not prove that there cannot be a private language? Does he not prove that there cannot be private ostensive definitions? Does he not prove that one man, just once in his life, cannot follow a rule? To be sure, these are negative existential statements—but they are existential statements. What is proved is a truth about the world, indeed—a necessary truth, even if a merely negative one!—Not so. As remarked above (§1), such necessary negative existential statements do not exclude possibilities that are impossible, but forms of words that have no use within our language. Despite their appearance, they are not statements about reality, but grammatical statements concerning the use (and uselessness) of words. Philosophy moves within grammar. It clarifies the network of concepts and conceptual connections of our conceptual scheme. It is not its business to prove that this, that or the other thing or kind of thing exists in the world—only to show that it is incoherent for the sceptic to claim that we cannot know whether there is a table in the next room, or whether Jack or Jill has a headache. And equally incoherent of him to suppose that while he knows how things are subjectively with him, he cannot know how things are with others. Some philosophers may object to this description of Wittgenstein’s procedures in the Investigations. To be sure, there are no transcendental arguments in Kant’s sense of the phrase. There is no attempt to prove the truth of any synthetic a priori propositions by reference to the a priori conditions of the possibility of experience. Nevertheless, in the sense in which one can argue that the Tractatus invokes a watered-down form of transcendental argument, does the Investigations not do so too? For it has seemed to many that the discussion of the possibility of a private language is a transcendental argument from either (i) the conceptual (grammatical, a priori) conditions of the possibility of knowledge of our own subjective experiences, or (ii) the conceptual (grammatical, a priori) conditions of the possibility of self-ascription of subjective experience, to: our knowledge (and hence the existence) of other subjects of experience. Equally, it has been suggested that Wittgenstein’s detailed examination of what is involved in following a rule shows that a conceptual (grammatical, a priori) condition of the possibility of anyone’s following a rule is that there exist others who likewise follow the rule. For, it is suggested, Wittgenstein showed that the criteria for following a linguistic (meaning-constituting) rule are to be found in community

132

P. M. S. Hacker

practice—in what is generally called “following the rule” by members of a linguistic community. So my following a rule presupposes that there are others who follow the rule I follow. And is that not, in a loose sense, a transcendental proof of the existence of a community of language-users? This, I think, is mistaken. It misconstrues what is severally shown by the private language argument and the investigations into following rules, and it misconstrues the method by which it is shown. First, the examination of the idea of a logically private language does not show or even try to show (i) that knowledge of how things are with oneself, or the possibility of self-ascription of experience, is possible only if (ii) one knows that there are other subjects of experience. (i) Wittgenstein denies that it makes sense to speak of knowing or of not knowing that one is in pain or that things sensibly seem to one to be thus and so. So there is (in Wittgenstein’s later work) no investigation of the conditions of the possibility of self-consciousness as construed by Descartes, since it is an illusion that there is any such thing. A cogitatio cannot be both a form of consciousness and an object of consciousness. Nor is there an investigation of the conditions of the possibility of self-consciousness (transcendental self-consciousness) as construed by Kant. The “I think” is not capable of accompanying all my representations. Although “It sensibly (visually, auditorily) seems to me that things are thus and so” makes perfectly good sense, that is not a statement of a “representation of a representation”, but a qualification on a statement of a representation. But the possibility of qualifying the statement of a representation presupposes the possibility of unqualified statements of representations. The language-games of “It seems to me that things are thus and so” presuppose antecedent mastery of the language-games of “Things are thus and so”. One may grant that, and nevertheless insist that what Wittgenstein does is to infer the existence and knowledge of the existence of other minds from the bare conditions of the possibility of self-ascription of experience. But that too is mistaken: (ii) Wittgenstein does not try to show against a sceptic about other minds that we do know that there are other subjects of experience. Indeed, he brushes such scepticism brusquely aside on two explicit grounds. First, “if we are using the word ‘know’ as it is normally used (and how else are we to use it?)”, then we very often know how things are with other people (cf. PI, §246). Secondly, “just try—in a real case—to doubt someone else’s fear or pain!” (PI, §303). Wittgenstein was not interested here in refuting the sceptic, any more than Kant was in the “Transcendental Deduction” (as opposed to the “Refutation of Idealism”). What he aimed to show is that a condition of the possibility (intelligibility) of groundless avowals and averrals of experience is the recognition of the logical criteria for the ascription of experience to others. So what he is doing is showing a connection between the possibility of immediate application of concepts of experience to oneself and mastery of their conditions of application to others. He is

Kant and Wittgenstein

133

not showing that we know that there are other subjects of experience. (“Do you need a proof?” he might mock one.) Now this does have bearing on the sceptic—but not by proving that we know something that he denies we can know. For the sceptic about other minds thinks it intelligible that he should know how things are with him, or at the very least, be able to say how things are with him, while simultaneously denying the adequacy of the behavioural criteria for the ascription of experience to others. And, if Wittgenstein’s argument holds, that makes no sense.6 I shall deal more briefly with the construal of Wittgenstein’s discussion of following rules as a transcendental argument for the existence of a community of followers of shared rules—of a language-using community, for I have discussed it extensively elsewhere.7 The so-called “community view” interpretation of Wittgenstein’s discussion of following a rule holds (roughly) that Wittgenstein showed that it is, logically speaking, only possible for an act to be an instance of following a given rule if it accords with what other members of a community of rule-followers do when they count themselves as following that rule. Following a rule, Wittgenstein remarked, is a practice—and a practice is (it is alleged) by definition a social practice. So a condition of the possibility of my following a rule to V (to expand the series of even integers, to apply a word in a language in accordance with a rule for its application) is that there be a community of rule-followers who engage in a common practice of following this rule. I have shown elsewhere that this is a misinterpretation of Wittgenstein’s observations. In his use of the expression “Praxis” in German, it is no pleonasm to say that a pattern of action is a social practice.8 It is not a criterion for correctly following a given rule that one do the same as others do when they follow that rule. Rather, if others are correctly following the same rule as one is correctly following oneself, then they will do the same as what one does oneself. It is the rule that determines what counts as following it—given that there is a practice (which may or may not be shared), i.e., a regularity of behaviour, which is recognized as a uniformity, and the employment of that uniformity as a canon of correctness manifest in critical normative behaviour (e.g., of self-correction, of explaining, of teaching others should the occasion arise). It would, I believe, be an egregious misrepresentation of Wittgenstein’s argument concerning following rules to suggest that it is even a watered-down transcendental argument from the conditions of the possibility of my following a rule to the existence of a linguistic community. What the discussion of following a rule shows is not that an a priori condition of following a rule is the existence of a social practice of following a rule, but rather that there can be no such thing as following a rule which cannot in principle be followed by others. And equally, that there can be no such thing as following a rule in the absence of a practice, indeed a normative regularity, of following the rule. For someone who follows a rule must not only exhibit a regularity of behaviour, he must also see that regularity as a uniformity, and treat it as a standard. But there is here no vestige of a transcendental argument, not

134

P. M. S. Hacker

even in a watered-down sense—only an array of connections within the web of concepts.9 REFERENCES Baker, G. P., and P. M. S. Hacker. 2009a. Wittgenstein: Understanding and Meaning. Vol. 1 of An Analytical Commentary on the Philosophical Investigations—Essays. Second edition, pb. ed. Edited by P. M. S. Hacker. Oxford: Blackwell Publishing. ———. 2009b. Wittgenstein: Rules, Grammar and Necessity. Vol. 2 of An Analytical Commentary on the Philosophical Investigations—Essays and Exegesis of §§185–242. Second edition. Edited by P. M. S. Hacker. Oxford: Wiley-Blackwell. Drury, Maurice O’Connor. 1981. “Some Notes on Conversations with Wittgenstein”. In Ludwig Wittgenstein: Personal Recollections, edited by Rush Rhees, 91–111. Oxford: Basil Blackwell. Hacker, P. M. S. 1986. Insight and Illusion: Themes in the Philosophy of Wittgenstein. Oxford: Clarendon Press. Kant, Immanuel. 1929. Critique of Pure Reason. Translated by Norman Kemp Smith. London: Macmillan and Co.; 1998. Critique of Pure Reason. Edited and translated by Paul Guyer and Allen W. Wood. Cambridge: Cambridge University Press. Kant, Immanuel. 2002. Prolegomena to any future metaphysics that will be able to come forward as science, translated by Gary Hatfield. In Theoretical Philosophy after 1781, edited by Henry Allison and Peter Heath, 29–169. Cambridge: Cambridge University Press. (Prolegomena) Strawson, P. F. Individuals. London: Methuen, 1959. Wittgenstein, Ludwig. 1961. Tractatus Logico-Philosophicus. Translated by D. F. Pears and B. F. McGuinness. London: Routledge and Kegan Paul. (TLP) ———. 1979. Notebooks 1914–1916. Second edition. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (NB) ———. 1980. Wittgenstein’s Lectures: Cambridge, 1930–1932. From the Notes of John King and Desmond Lee. Edited by Desmond Lee. Oxford: Basil Blackwell. (LWL) ———. 1981. Zettel. Second edition. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (Z) ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2003. Movements of Thought: Diaries 1930–1932, 1936–1937. Edited by Ilse Somavilla. Translated by Alfred Nordmann. In Ludwig Wittgenstein: Public and Private Occasions, edited by James C. Klagge and Alfred Nordmann, 3–255. Lanham: Rowman & Littlefield. (MT) ———. 2005. The Big Typescript: TS 213. Edited and translated by C. Grant Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell Publishing. (BT) ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. (PI)

8

Wittgenstein in the Mid-1930s Calculi and Language-Games Andrew Lugg

Wittgenstein’s writings from 1929 to 1936/1937, aptly dubbed “transitional”, show him struggling to shake the influence of the Tractatus and groping towards the viewpoint pioneered in the Investigations. Only after much twisting and turning, and more than a few stumbles, did he mange to free himself from what he took to be his earlier errors and discover the path he would travel for the rest of his life. During these years he shifted his ground on many matters, some small, some big. Thus he modified and eventually dropped the claim, central to the Tractatus, that elementary propositions are independent, substituted the idea of a motley of languages for the idea of a single language, came to construe meaning in terms of rules rather than in terms of naming, straightened out his view of colloquial language as completely in order, and turned the spotlight on understanding, intending and thinking, topics that received short shrift in the Tractatus. Yet not everything changed, and some of what is generally believed to have changed remained intact. In particular Wittgenstein did not shift his view of language in the mid-1930s and renounce “the calculus model of language” for “the language-game model”. Contrary to conventional wisdom, he continued to treat language in calculus-like terms, and it is a mistake to regard the notion of a language-game as a mainstay, never mind the mainstay, of his later philosophy. What exactly the calculus and language-game models involve and what is supposed to have prompted the switch is by no means easy to say. There is, however, reason to question how the tale is usually told and to conclude that the common view of the development of Wittgenstein’s ideas in the mid-1930s hampers a proper understanding of his philosophy. This is especially clear when work from the period is put under the microscope. Had Wittgenstein embraced a new view of language, one would expect to find him touting it and underlining its superiority in the remarks he composed in 1936–1937, just after the shift is said to have occurred. It would be a surprising turn of events if this shift in view did not surface in the most important material from the time, MS 142, a manuscript that ended up practically unchanged as §§1–189a of the Investigations, and MSS 117–19, manuscripts eventually published—in a reorganized and lightly revised form—as

136

Andrew Lugg

Part I and an accompanying appendix of Remarks on the Foundations of Mathematics.1 In neither set of remarks, however, is there any indication that Wittgenstein modified his thinking about language substantially and took himself to be moving off in a fundamentally new direction. He does not mention, even in passing, that he has come to favour a different model but speaks of language in terms of both calculi and language-games. None too surprisingly, he opts for the idiom that best serves the purposes at hand and what he is attempting to explain. The picture of Wittgenstein as revising his view of language in the mid1930s is variously expressed in the secondary literature. P. M. S. Hacker writes: “The notion of a language as a system of interlocking Satzsysteme (propositional systems) . . . was gradually replaced by that of a motley of language-games. . . . A language is something used in speech and writing, in human activities which take place and have significance only against complex contexts of human forms of life and culture” (1986, 132).2 S. Shanker avers that “the calculus model would soon disappear [after the introduction of Satzsysteme in 1930], and out of these Satzsysteme would evolve ‘language games’” (1987, 9). D. Stern holds that Wittgenstein “came to distance himself from the calculus model of language” and moved away during the first half of the 1930s “from a conception of language as a formal system . . . towards the view that mastery of rules depends on a background of shared practices” (1995, 103).3 H.-J. Glock discerns a “gradual abandonment of the calculus view” and holds that “[t]he game analogy gradually replaces the calculus analogy” (1996, 70 and 194). And J. Medina sees Wittgenstein as repudiating around 1934–35 “the view of language as ‘an autonomous calculus of rules’” and eventually adopting a conception centred on “a practice-based view of normativity” (2002, 139). What such interpretations share is the thought that Wittgenstein replaced one conception of language with a second conception, a view of the development of his philosophy that has become one of the firmer fixtures of Wittgenstein studies.4 This line of interpretation of Wittgenstein’s developing philosophy places considerable weight on the assumption that he dropped the idea of language as a system of rules that stands apart from speech and writing and came to promote the idea of it as a collection of linguistic activities. The core contention is that Wittgenstein viewed language during the calculus period as a symbolism with a grammar, the primary purpose of which is to represent the world, and during the language-game period identified it with reporting truths, issuing orders, expressing desires, telling jokes and so on, activities comparable to playing chess, boiling an egg and driving a car. He is to be regarded as no longer thinking of language as a structure of the sort treated by logicians, one entirely detached from practice, but thinking of it instead as a practice, something foreign to formal logic. What is new in his later philosophy is the role of language in human discourse and conduct. Out is the conception of language as a “formalism” unconnected with everyday life; in

Wittgenstein in the Mid-1930s

137

is the conception of it as a “form of life”. Indeed this change of conception is regularly taken to be what divides Wittgenstein’s late philosophy from his early philosophy. Characterizing the calculus and language-game models in terms of symbolisms and games, while helpful, gets us only so far. The main hitch is that symbolisms and games are both naturally described by adumbrating the rules that govern them, and Wittgenstein typically refers to language-games as well as calculi as rule-governed. The predicate calculus, a symbolism if ever there was one, is explicable in game-like terms, while chess, Wittgenstein’s prime example of a game, is explicable in calculus-like terms. What primarily differentiates the two models would seem better captured by the answers they provide to the question of what a codification of language depends on. As far as standard thinking about the development of Wittgenstein’s philosophy goes, the key contrast is that the calculus model takes language to be rule-governed one way, the language-game model a second way. It is central to the calculus model that the relevant rules are codifiable without reference to its application, while central to the language-game model that these rules cannot be so codified. This suggestion, rough as it is, places language where the two models would have it placed—in the case of language construed as a calculus in company with mathematical algorithms, systems of musical notation and rule books for a game; in the case of language construed as a practice in company with proving a theorem, singing a song, playing a game and mixing paint. With this clarification, the view of the development of Wittgenstein’s philosophy in the mid-1930s under discussion takes him to have traded a view of language as governed by rules specifiable without reference to linguistic practice for a view of language as governed by rules that are not specifiable without essential reference to such practice. The claim is that the conception of language as a symbolism with a geometry was replaced by the conception of it as a matter of use and application. In other words, as Hacker states it, Wittgenstein rejected his 1929/1930 view that “[a] language . . . is a calculus of signs” and accepted that “[t]he use of a sign in discourse is an operation in exactly the same sense as writing down the result of a multiplication is carrying out an operation” (1986, 131–2). “The idea that language is a calculus of signs came to seem far-fetched”, and his concern with “the ‘geometry’ of a symbolism” fell by the wayside. In his eyes, “the fruitfulness of the comparison [of language with a calculus] ceased to be very impressive”, “the dangers and temptations of the comparison [being] arguably greater than the illumination” (132).5 As to whether Wittgenstein shelved the concept of “the geometry of a symbolism”, it is neither here nor there that he began to speak of language as a motley rather than a monolith. While he was at pains during the period to stress the multifaceted character of language, this lends very little, if any, support to the view that he set aside the calculus model and promoted the language-game model. For one thing, he rejected the view of language

138

Andrew Lugg

as a monolith soon after returning to philosophy in 1929, i.e., during the so-called “calculus period”. And for another it is beyond belief that he saw himself as having to concede that inasmuch as language is a motley, it is wrongly viewed in calculus-like terms. The notion of a motley of calculi is no less intelligible than the notion of a motley of language-games, and language conceived in terms of (many) calculi is hardly less plausible than language conceived in terms of (many) language-games. Just as mathematics is sensibly held to comprise the calculus of adding and subtracting, the calculus of classes, the differential and integral calculi, etc., so language is sensibly held to comprise a calculus of spatial terms, a calculus of temporal terms, a calculus of colour terms, etc. In fact, in the mid-1930s Wittgenstein speaks of language in terms of a motley of calculi. What he wrote off was the idea of language as a single calculus, what has been dubbed “the universal calculus model” (Shanker 1987, 8). Equally questionable is the claim that Wittgenstein substantially began to think differently about language because he realized it comprised a set of interconnected items rather than a set of disconnected ones, language-games in the one case, calculi in the other. Without doubt, many calculi are autonomous in the sense that they are cut off from other calculi: the calculus of up/down, front/back and similar pairs of terms, for instance. And without doubt many language-games are parasitic on other language-games: the language-game of joking and asserting, for instance, the humour of a fair number of jokes depending on putative facts being conveyed with a straight face. But the converse is also true. Many calculi are interdependent and many language-games independent. The calculi of colour and position are interconnected (in the case of transparency) while the language-games of reporting and telling tall stories are mostly disconnected. (Indeed it is commonly, if erroneously, maintained that Wittgenstein took language-games to be self-contained.) The safest assumption would seem to be that the calculus conception and the language-game conception are on a par regarding their standing and falling together. Both calculi and language-games can be dependent or independent. Nor is Wittgenstein’s repeated reference to language-games (and the use of language) after 1936/1937 a good reason to regard him as having shunned the model of language as a calculus. There can be no denying that in the mid-1930s his interests changed and post-1935 he devoted considerably more attention to language use. Hacker is right that “as Wittgenstein probed a battery of questions about rule-following . . . his interest shifted from the ‘geometry’ of a symbolism (whether a language or a calculus) to its place in human life, its use in human behaviour and discourse” (1986, 132), and Glock right that Wittgenstein’s view of language-games as “‘ways of using signs’ which are simpler than those of everyday language . . . evolve[d] into the idea of a language-game as a ‘system of communication’ by which children ‘learn’ or are ‘taught’ their native language” (1996, 194).6 What is debatable is whether Wittgenstein shifted his view as well as his attention.

Wittgenstein in the Mid-1930s

139

The important question is whether he revised his thinking about language in the first half of the 1930s, not whether he began to take a different interest in it. To say that he focussed on the acquisition of language and the role it plays in our lives is not to say that he eschewed one model of language in favour of another model. It is also irrelevant to the question of the development of Wittgenstein’s views about language that he began around mid-decade to speak of words and sentences as intertwined with their application in everyday life. Once again the premise does not establish the conclusion. True, Wittgenstein concentrated more on the application of words and sentences in his post-1936 remarks than in his pre-1933 remarks. But application does not differentiate the two models. To describe language as a calculus is not to deny it is applied, still less to suggest it is entirely separate from application. And to describe language as a practice is not to say language-games are always applied, or even that they always have meaningful application. On the one hand, the +1 calculus has application and the +100003 calculus is applicable whether or not it has ever been applied. On the other hand, the language-game of shouting “Hello” 100003 times has probably never been fully carried out and the language-game of trisecting an angle using just a ruler and compass is provably without application. Less confusing would be to treat linguistic calculi as (mostly) associated with language-games, and to treat language-games as (mostly) associated with calculi, the language that they, by definition, involve being normally expressible in the form of a calculus. Certainly Wittgenstein is more charitably saddled with this way of regarding calculi and language-games than with the view that language-games differ essentially from calculi in point of their applicability. Turning now to MS 142 bearing these preliminaries in mind, we find no evidence for concluding that Wittgenstein drastically shifted his view of language just prior to 1936. In this manuscript he frequently refers to language-games but not in any way at variance with the conception of language as a calculus. Consider, to begin with, the section in which language-games make their first appearance. After mentioning a couple of ways in which children learn their native language, Wittgenstein says he will refer to these ways as “language-games” and will sometimes call primitive languages language-games (§8b in MS 142; §7b in the Investigations).7 While this is useful terminology and doubtless important in the context of the later Wittgenstein’s discussions of language, the remark scarcely endorses the superiority of the language-game model over the calculus model. Far from characterizing language as a practice (or offering a characterization of language), Wittgenstein merely indicates how he intends to use the term “language-game”. The notion of a game is tightly associated with the notion of an activity, but Wittgenstein does not commit himself to regarding language as a practice, let alone promote the language-game model. Rather the opposite, he would surely have agreed that primitive languages are comparable to calculi and the games by which children learn language are as

140

Andrew Lugg

rule-governed as other games. Nobody can deny the “game” of a teacher calling out “block” and a child fetching a block, which Wittgenstein alludes to, is describable in calculus-like terms. And there is no evidence either in Wittgenstein’s next paragraph that he embraced the language-game model. In suggesting that the business of “naming and repeating” what a teacher says may also be referred to as a language-game and drawing attention to “the use of words in games”, he is not saying language is, or is like, a game (§8c/§7c). He is concerned with the activity of learning a language, not with what is learnt when it is learnt; i.e., he is concerned with the process rather than the result of languageacquisition. Having mentioned in the previous paragraph processes involving language, he now augments his list with other processes, processes that have counterparts in the learning of calculi. As he imagines it, learning the names of building stones is not so very different from learning the first three numerals, the words in both cases being typically inculcated by what he calls “ostensive teaching” (§7/§6; also compare §10/§9). And it is of no consequence either that ring-a-ring-a-roses requires the participants to fall down at the appropriate moment. The game itself is specifiable, rather simply, as a rule-governed calculus, one that says what the players have to say and do and at what point they have to say and do it. The final paragraph of §7 in the published version of the Investigations (TS 227) does not appear in MS 142—it was added in 1944/1946. It is worth considering briefly, however, if only because it is often thought to provide—along with the other three paragraphs—nigh-on irrefutable evidence that Wittgenstein substantially shifted his view of language in the mid-1930s. While it is not at all clear what Wittgenstein means by “the whole” when he says he will “also call the whole, consisting of language and the actions into which it is woven, the ‘language-game’”, he is implausibly interpreted as championing the language-game model of language as a practice. Whether he is taking “the whole” to cover language-like processes, systems of communication such as the builders’ language in §3/§2, primitive language, natural languages like English and German or language as such, he is clearly of the view that language-games comprise two things, not one. His thought is that “the whole” consists partly of language, partly of actions—a view perfectly consistent with the conception of language as a symbolism. In fact he can be read without too much difficulty as picturing language-games as consisting of calculi interwoven with action rather than as picturing them as consisting solely of practice. The remarks in §24/§23, another section often taken to show Wittgenstein came to embrace the language-game model of language, are similarly inconclusive (and remain so whichever version of §24 in MS 142 is considered). In the first paragraph of the section, Wittgenstein mentions language-games but only in passing and it would not have required much fixing to be about calculi. He is not criticizing the idea of language as a system of symbols when he says there are “countless [unzählige] different kinds of

Wittgenstein in the Mid-1930s

141

use of what we call ‘symbols’, ‘words’, ‘sentences’”, and there is nothing to be concluded from the fact that he goes on to refer to “kinds of use” (§24a/§23a). If anything, his remarks square better with the conception he is generally believed to have defended in the early 1930s and subsequently shunned. Indeed he is most naturally taken to be thinking about calculi since he goes on to observe that “new types of language, new language-games as we may say, come into existence, and others become obsolete and get forgotten”. It is hardly insignificant that he immediately adds (within parentheses) that mathematics provides “a rough picture” of how language changes, such changes involving the introduction of new calculi, as occurred, for example, when complex numbers were concocted. In the following paragraph, §24b/23b, Wittgenstein says something else that is likely to be seen—actually has been seen more than once—as signalling a shift in his view of language, namely that the phrase “language-games” highlights “the fact that the speaking of language is part of an activity, or of a form of life”. Despite the reference to language-games, however, there is nothing in the passage to suggest that the language-game model is being plugged and language equated with practice. The calculus conception is not antithetical to the idea of speech as “part of an activity”, and Wittgenstein comes nowhere close to stating that a specification of the rules governing spoken language necessarily involves how words are learnt and used. What he says is altogether different and much tamer. Rather than advance a conception of language, he takes the phrase “language-game” to emphasize that speech is “part of an activity”. What concerns him is “the speaking of language”, not language itself, and not without reason he underscores “speaking”. Like the rest of us, he takes language to be one thing, speaking a language another. As for the final paragraph of the section, §24c/23c, this may seem on first sight to show fairly conclusively that Wittgenstein had at the time embraced the language-game model. It is tempting to read him as endorsing this model when he provides a list of examples of language-games and intimates that language as it is “used” is far more varied that “logicians” presume (§24c/23c). On closer consideration, however, it is clear that he has not shifted his view of language fundamentally. To recognize the multiplicity of language-games is to recognize a point foreign to his former self, but it is not to disparage the calculus conception as such. Wittgenstein neither states nor implies that language cannot be regarded as a calculus insofar as it is multifaceted. The language-games he lists—singing catches, making jokes, cursing, greeting, praying and the like—are rule-governed. Presumably the language in which catches are sung, jokes made, curses issued and so on is no less credibly regarded as codifiable as a calculus than the language in which describing and reporting are cast. While the author of MS 142 recognizes far more uses of language than the author of the Tractatus, he seems happy enough with the idea of language as a multitude of calculi. His use of the phrase “structure of language [Bau der Sprache]” in the paragraph,

142

Andrew Lugg

moreover, strongly suggests that he is still thinking of language as codifiable (in appropriate circumstances) as a calculus or series of calculi. A similar story can, I would argue, be told about other references to language-games in MS 142. The references at the beginning of the work are not atypical, and Wittgenstein’s subsequent references are just as readily reexpressed in calculus-like terms. What is more telling, however, is that the conception of language as a calculus is still in play in the manuscript (and in the Investigations §§1–189a). Instead of committing himself to a view of language as a practice, Wittgenstein treats it as a symbolism. It is no accident that he refers to the system of communication used by a tribe of builders (introduced in §3/§2) as a language as well as a language-game. In his view the builders’ system of signs is a rule-governed system, each of their four words being correlated with a different kind of building-stone, indeed so strictly correlated that the tribe is regularly described (and criticized) as comprising mechanical signalling machines. Moreover Wittgenstein seems just as committed to the conception of language as a calculus when in §47/§48 he applies what he says at §3/§2 to the view of language adumbrated in the Theaetetus (§43/§46). Wittgenstein specifies a language-game (he also calls it a language) to which what he says there applies, an account that has all the appearance of a calculus, one that can, as his own specification makes clear, be codified without saying how it is embedded in its users’ lives. Further along in the manuscript Wittgenstein allows that he has not characterized “the essence of a language-game, and hence of language” (§66/§65). He does not, however, say or indicate that he has had a change of heart. He just acknowledges that he has mentioned many kinds of language-games without saying why they all count as “language or parts of language” and goes on to combat the idea that they all share something in common (and by extension the idea of a universal calculus conception). This allows for the view of language as a set of calculi as well as the possibility of many sorts of “word and sentence” (§24c/§23c). It is, Wittgenstein is noting, a fool’s errand to attempt to specify the essence of language—the reason that various kinds of language are all called language is that they are variously interrelated. This is not to say something true only of language-games. As Wittgenstein observes, what counts as language constitutes a family in pretty much the same way that different sorts of numbers constitute one (§64/§67); i.e., they are related in a manner similar to the way in which the integers, rational numbers, real numbers and complex numbers—each of which can be codified in terms of a calculus—are related to one another. In pointing this out, Wittgenstein is not jettisoning the calculus conception of language on the grounds that our words are vague. He is challenging the idea that language has a single specifiable essence. If anything is wrongly held to be vague, he would have noted, it is a numerical calculus. Later still in MS 142 Wittgenstein treats calculi and language-games in tandem and without any apparent misgivings. He notes that when he wrote the Tractatus it had not dawned on him that while philosophers are apt to “compare” language use with playing games (and calculi with “fixed

Wittgenstein in the Mid-1930s

143

rules”), they cannot “say” language use is a matter of “playing a game” (§78/§81). In stressing this, he is not announcing that he has come to a new view of language and endorsing another model. Rather than turning his back on the comparison of a language with a calculus (or calculi) and discounting the comparison of using a language with operating a calculus, he is noting that we tend to misconstrue what using a language involves. The problem is that we have a shaky grasp of concepts like that of “meaning” (not, notice, a shaky grasp of the concepts of a practice, playing a game, and using words). It is, Wittgenstein would have us appreciate, a mistake to suppose whoever means and understands something is “operating a calculus with definite rules”, a trap he sees himself as having once fallen into. As long as we take the notion of a calculus under advisement and do not burden it with additional assumptions regarding what “operating a calculus” involves, there is, he apparently thinks, no harm and possibly much good in treating language in calculus-like terms. What is problematic is not the picture of language as a calculus but the picture of “something that lies beneath the surface”, something that philosophers can prise out (§90/§92). Wittgenstein does not believe it wrong to try to capture the essence of language by providing a calculus; he only thinks it is a mistake to think its essence is lying in wait to be uncovered. As he says, he too is out to understand the “function” and “structure” of language, albeit not as something understood as buried in the recesses of the mind, behaviour or linguistic practice. The difference between his approach and that of many other philosophers is that he seeks what lies on the surface and is “surveyable by rearrangement”. For him surveyability is allimportant, and it is a surveyable account of language, not something revealed by digging, that is required to understand its essence. Philosophy is like arithmetic inasmuch as the philosopher, no less than the mathematician, aims to provide a calculus (or calculi). To answer the question of the structure of language—and its function to the extent that this concerns its essence—is to provide a grammar or, what amounts to the same thing, a calculus free of philosophical add-ons. Wittgenstein could hardly be clearer about the importance of surveyable rearrangement. In his view philosophers tie themselves in knots because they “do not command a clear view of the use of our words”, a failure he takes to be remediable by supplying a “perspicuous representation” of our grammar, i.e., a calculus that condenses the relevant principles in a compendious form (§115a-b/§122a).8 The thought here is clearer when considered along with the example that he provides in Philosophical Remarks, the source manuscript. In this work he mentions the colour octahedron, a kind of calculus, as “a rough representation of colour-space”, “a grammatical representation”, “a representation [that] gives us a bird’s eye view of the grammatical rules” (1975, §1). (In §39 of the same work he adds: “The colour octahedron is grammar since it says that you can speak of a reddish blue but not of a reddish green, etc.”.) Such a representation, he writes in MS 142 (and the Investigations), is crucially important for us—Wittgenstein has in mind

144

Andrew Lugg

philosophers who think like him—since it “earmarks the form of account we give” (§115c/§122b). As important as language-games are in his later philosophy, then, they are not a be-all and end-all. He is not committed to regarding language exclusively or essentially as a language-game, only to regarding language as perspicuously representable—for some purposes —by means of principles, diagrams and devices, each of which defines or constitutes a calculus. How we are to understand “the form of account we give” is clearer later in the manuscript. Taking up the question of what counts as a proposition, Wittgenstein says: “[W]e call something a proposition when in our language we apply the calculus of truth functions to it” (§137/§136). This is noteworthy because it shows that Wittgenstein does not abjure the language of calculi altogether. But it is perhaps even more striking that he would treat the concept of a proposition as necessarily, rather than contingently, related to the concepts of truth and falsity (and hence as codifiable as a grammatical rule). In his view, as he goes on to observe, what counts as a proposition is variously “determined by the rule of sentence formation (in English for example), and . . . by the use of the sign in the language-game”. Either way, the concepts of truth and falsity “belong” to our concept of a proposition rather than “fit” it; i.e., they are logically or conceptually, not empirically, connected to it. And given this it is but a small step to the conclusion that the concepts are perspicuously representable in the form of a calculus. It is not a problem that the material on perspicuous representation in MS 142 was composed earlier in the decade during the so-called calculus period. Wittgenstein gives every impression of being satisfied with his remarks— they also turn up virtually unchanged in the Investigations—and had he come to think he had misspoken, he would have in all probability modified or suppressed what he had written. Glock is on firm ground when he notes that Wittgenstein “recycle[d] material from the early 1930s which compares language with a calculus in order to bring out [that sentences are dead apart from a system and that grammar is not a causal mechanism]” and “these same points are expressed in the comparison of language with a game, in particular that of chess” (1996, 67). It is, however, another question whether Wittgenstein set aside “the calculus model” as opposed to retaining it and widening the scope of his investigation. Wittgenstein may have suggested in the Blue Book (1933/34) that viewing language as “a calculus proceeding according to exact rules” is “one-sided” and have observed that “we [do] not think of the rules of usage . . . while using language [and] when we are asked to give such rules, in most cases we aren’t able to do so” (1958, 25). But he would also have allowed that the language-game conception is onesided, and noting there are “rules of usage” is not to say the language users have the rules in their heads and can say what they are. So much for MS 142. When we examine MSS 117–19, we find—not unexpectedly given that the manuscripts were composed around the same time—that Wittgenstein works with much the same conception of language.

Wittgenstein in the Mid-1930s

145

When discussing mathematics he regards it in calculus-like terms and takes it, not unnaturally, to comprise so many calculi. For him mathematics is a system of representation, not a substantive science, and he mostly ignores the activities of mathematicians and the use of mathematics in everyday life. M. Wrigley is right, it seems to me, when he judges the continuities in Wittgenstein’s philosophy of mathematics to outweigh the discontinuities and portrays him, early and late, as opting for the language of calculi when discussing mathematics (1993, 73 and 82). As Wrigley notes, Wittgenstein did not reject the calculus conception when he came to see mathematics as a motley but remained wedded to the conception (broadly understood) and treated mathematical propositions in the manner he had treated them earlier in the decade (74). Throughout the period, Wrigley observes, Wittgenstein takes mathematics, as he had taken it in Philosophical Remarks, to be representable by means of a notation “[r]oughly in the same way as chemical compounds are represented by means of such names as ‘triethylamide’ etc.” (1975, §162, 192). This way of regarding Wittgenstein’s philosophy of mathematics has not been universally accepted, however, but mostly ignored or discounted, and the remarks from the mid-1930s are still taken to differ substantially from his earlier remarks. Thus S. Gerrard argues that at this time Wittgenstein came to treat mathematics in terms of language-games and to regard it as covered by the view canvassed in the Investigations, a view that is “to say the least, well known” (1991, 140). On Gerrard’s reckoning, there are “two chief post-Tractatus accounts of mathematics” and Wittgenstein moved from regarding mathematics as a calculus to believing that “meaning and truth can be accounted for only in the context of a practice” and thus to favouring the idea of mathematics as “a nexus of language-games” (126). More succinctly, “[i]n the middle 1930s Wittgenstein’s views changed [and] he began to look at mathematics as a motley of language-games” (Gerrard 1996, 174). One snag with this is that Gerrard takes the calculus conception (for mathematics) to be equivalent to claims that are, to my way of thinking, ancillary to the conception rather than constitutive of it. The calculus conception, sympathetically understood, does not presuppose that calculi are self-contained (and invulnerable to external critique), or that every revision, however minor, results in a new calculus, or that calculi and application are sharply separated, or that there are requirements on calculi in general (1991, 133).9 Still it remains to be seen whether Wittgenstein examines mathematics during the period, as Gerrard contends, with an eye to the “special role it plays in our lives and its special relation to other language-games” (132). While barely a section of MS 142 goes by without some reference to language-games, Wittgenstein refers to them just five times in MSS 117–19. He holds that philosophers need to consider what sort of “language-game inferring is” (MS 117, 20; Remarks on the Foundations of Mathematics, I.17); observes that in this “language-game” propositions are uttered or written down one after another (MS 117, 15/I.19); asks how propositions

146

Andrew Lugg

are “used in the language-game” (MS 118, 108v/III-6); compares the “language-game” in which someone says: “I am lying” with “thumb-catching” (MS 118, 112r/III-12) and notes that thinking of the future use of words as buried in its present use appears only when it is taken to figure in an out-ofthe ordinary “language-game” (MS 119, 42/I.126). (The section numbers, added by the editors of Remarks on the Foundations of Mathematics, refer to material in Part I or Appendix III of the work.) These remarks, while valuable, do not show that post-1936 Wittgenstein embraced the language-game model of language. In MS 117 the references to language-games are to the procedure of inferring, in MS 118 to linguistic activity, and in MS 119 yet again, to activities having to do with language rather than language itself. In the first remark Wittgenstein points out that we would do well, given how philosophers think of inference, to consider what inferring involves. In the second remark he raises the question of how in the language-game of inferring a basic logic law justifies inferences. In the third remark he highlights the difference between asserting a proposition and “the utterance of [a] sentence, e.g. as practice in elocution”. In the fourth remark he calls attention to the fact that a certain proposition is “unusable”. And in the fifth remark he comments on the idea that the present use of a word determines its future use, zeroes in on the activities of asserting and uttering words and suggests that it pays to consider different language-games since this can show the use of a sentence is peculiar (and correlatively clarify how it is normally used). The notion of a language-game invoked in these five remarks is the notion, introduced in the final version of the Investigations, of “the whole, consisting of language and the actions into which it is woven”. Linguistic activity is assumed throughout to involve language, something with a logic (and a grammar). Thus in the first passage Wittgenstein allows that the truth of inferred propositions is vouchsafed by the truth of other propositions, the object of his attack being common thinking about what goes on in us when we logically infer something from something else. In the second passage he accepts that it makes sense to say one proposition follows from a second one and a person is justified in inferring one from another, the only trouble being how philosophers use the verbs “to follow” and “to infer”. In the third and fourth passages he limits himself to observing that a sentence may express different propositions in different contexts (as, for instance, in the case of “The door is closed”, which can be a comment or command, even a question). Finally in the fifth passage he directs his fire against a tempting view of what assertion involves, one that leaves untouched the question of whether or not it is properly understood in calculus-like terms. In none of the five remarks, then, is language equated with practice. To be sure, Wittgenstein occasionally seems to think that understanding the role of mathematics in our lives is key to understanding the sort of business it is. Near the beginning of MS 117 he writes (in response to a question about counting being “only a use”): “The truth is that counting

Wittgenstein in the Mid-1930s

147

has proved to pay” (4/I.4), and later in the same manuscript he stresses that going through a proof is something that “we do”, a “fact of our natural history” (84/I.63; also MS 118, 65v). Moreover, in MS 119 he speaks of himself as providing “remarks on the natural history of man” (1/I.142). On a closer reading, however, there is nothing indicating that Wittgenstein came round to embracing the language-game model of language and to identifying meaning in mathematics with use. The facts of our natural history are important since they alert to us how things are and could be with us. And in noting that counting pays, Wittgenstein is reminding us that, as he puts it in the first remark, “the series of natural numbers” can no more be said to be true than language can be. It can only be said to be “usable and, above all, [that] it is used”. Though the word “calculus” appears only infrequently in the material presently under discussion, its few appearances are revealing. In MS 118, Wittgenstein avails himself without any sort of apology of the concept of “geometry (or kinematics)” and refers to “the application of the calculus” (82v/I.120). Moreover, in the same manuscript (97v–98r), after noting that there is nothing wrong with the words “endlos [endless]” and “unendlich [infinite, indefinite]”, only the spirit in which they are used, he refers to the “hocus-pocus” with which mathematicians accompany their “calculi/ calculations [Kalküle]” (TS 221, 217–18; Wittgenstein 2001, 400). Such remarks do not amount to an endorsement of the language-game model and are much harder to parlay into a thesis about language than Wittgenstein’s observation a few years earlier that “[t]he meaning is the role of the word in the calculus” (Wittgenstein 1974, 63). To the contrary, given he continues to speak of mathematics (and logic) in calculus-like terms, he is far more easily regarded as continuing to endorse a calculus model of language. His acknowledgement of the existence of symbolic representations, his suggestion that it is indisputable that calculi can be applied and his criticism of the tendency of mathematicians to encumber calculi with hocus-pocus can hardly be dismissed as anomalous. Wittgenstein has the calculus conception even more clearly in mind when he discusses “essences”, something that in the version of the material published in Remarks on the Foundations of Mathematics he describes mathematicians as creating (I.32). In MS 117 he refers to the “mark of a concept” as “essential” (94/I.73) and goes on to suggest that reference to an essence is tantamount to reference to a convention (96/I.74). At a stretch, I suppose, this can be read as concerning language-games, but it is more charitably read as concerning language as a symbol system, the language of mathematics included. For Wittgenstein essences are conceptual, logical, grammatical, and as such to be found in language, sometimes in association with forms of life, sometimes not. Nor should it be overlooked that in MS 118 he says that logical propositions show “the essence, the technique, of thinking” (87r/I.133). Whatever he is driving at exactly, he is not suggesting that creating essences is a matter of creating new kinds of mathematical

148

Andrew Lugg

activity, new uses of language as opposed to new kinds of language, new symbolisms, new calculi. It is one of Wittgenstein’s major themes that nontemporal propositions about essences are categorically different from temporal propositions about the way things are, and in MS 117 he emphasizes more than once that mathematical propositions are nontemporal. Thus he contrasts the nontemporal use of “consist” in some sentences about numbers of objects (he mentions: “The 100 apples in this box consist of 50 and 50”) and adds that it is “the characteristic mark of ‘internal properties’ [that] they persist always, unalterably” (25/I.101-I.102). Here the focus is, once again, on concepts, not use. We are being reminded that a mathematical proposition records an internal property and is part of a calculus, rather than an empirical finding (32–3/I.103). Furthermore, by way of ramming the point home, Wittgenstein observes that proving the equality of the number of strokes with the number of angles mathematically is different from the physical process of correlating strokes with the angles of a figure; the result of the first “game” being “non-temporal”, the result of the second “temporal” (36–7/I.27). Especially significant in this context is Wittgenstein’s discussion in MS 119 of what he calls “the machine-as-symbol” (29–31/I.122 and I.125), the bulk of which is recycled in the Investigations (§§193–4). With an eye on what he judges to be a major error in philosophy—the conflation of the logical with the empirical—he notes that machines understood as impervious to outside forces are very different from—and all-too-often mistaken for—machines subject to breaking and bending. There is nothing in this material to suggest he deprecates the notion of a machine-as-symbol, the object of the exercise being to warn against mistaking such machines for machines-in-use, a point that applies, mutatis mutandis, to language-as-a-calculus and language-inuse. As Wittgenstein says in MS 118 (after noting that there is no such thing as absolute rigidity) there is a difference between “an indication as to mathematical method” and “an indication about the application of the calculus” (82v-83r/I.120). As its name implies, the machine-as-symbol is a symbol, something that has application, and Wittgenstein has no compunction about helping himself when discussing this issue to the idea of a “‘geometry’ of a symbolism”. Langue is not parole, and the conception of language as grammar has not been supplanted by the conception of language as practice. Three other areas in MSS 117–19 in which the idea of language as a calculus manifests itself deserve attention. First, in MS 117 Wittgenstein discusses what is referred to in the table of contents of Remarks on the Foundations of Mathematics, appropriately enough, as “the logic of colour” (5). After observing that a certain sort of proof is nontemporal (32–3/I.103), he avers that “White is lighter than black” is likewise nontemporal and grammatical (33/I.104). While one surface might be lighter than a second surface today, darker yesterday—it may have been painted in the meantime—the proposition itself “expresses the connexion of the words ‘white’, ‘black’ and ‘lighter’ with a paradigm” (33–5/I.105). Normally the sentence does not

Wittgenstein in the Mid-1930s

149

provide information but functions as a grammatical rule, one that anyone with a modicum of English will already have assimilated. In fact, as far as its logical status goes, it is on a par with the likes of “3 is greater than 2” and “No bachelor is married”. It is part of the grammar or logic of colour, a sentence that, along with other sentences, constitutes a calculus. Secondly, in the course of discussing various abnormal practices, Wittgenstein refers to linguistic activities that for all the world seem describable in calculus-like terms. In MS 117 he imagines people paying for wood according to the formula: cost equals length times breadth times height, a formula that is clearly expressible as a calculus (46/I.143). As he notes, wood-buyers who calculate how much to pay for wood take their calculation to show what they have to pay (47/I.147), i.e., they possess a system of calculation, a calculus that can be specified without referring to their behaviour. And likewise for a people who sell wood by weight and a people who give it away. If wood were bought by how much area it covers, the only difference from our own practice would be that the buyers have a different way of determining what to pay (48–9/I.150). Yet another example would be a people who use coins differently from us and pay what they want for goods (50–1/I.153). However unusual their calculations, the specification of the calculus they operate with no more requires reference to their practice of paying than the specification of the calculus we operate with requires reference to ours. Thirdly, albeit somewhat less clearly, Wittgenstein’s remarks in MS 119 about mathematical propositions and mathematical investigation also indicate that he is still working with the calculus conception. After observing that experience tells us the results of a calculation, he asks—doubtless expecting “No” for an answer—whether “Experience says the calculation comes out in such and such a way” is a “mathematical statement” (94– 5/I.164–5). He is noting that mathematics is not about experience, about what mathematicians discover, for the simple reason that a mathematical statement expresses a point of grammar, an “essence”, something specifiable apart from “the actions into which it is woven”. In his view mathematical propositions are rules of the most solid sort and mathematical proofs, like proofs in logic, proceed from rules to rules. As he sees it, mathematicians are forever devising “new rules” (95/I.166), indeed “always inventing new forms of description” (95–6/I.167). Change in the network of rules—dare I say “calculus”?—is like change in language and no less grammatical in nature, whatever happened to motivate it (compare Wittgenstein 1953, §35, and consider the change in language that occurred when “he or she” came to be thought preferable to “he”). In MSS 117–19 Wittgenstein explores the foundations of rule-governed enterprises, not least mathematics and logic, more or less from scratch. He allows, even insists, that mathematics and logic are rule-governed calculi but takes issue with ways in which this anodyne idea is construed. Thus in MS 117 he rejects the conception of logic as “a kind of ultra-physics”,

150

Andrew Lugg

perceivable by means of “a kind of ultra-experience” and criticizes Frege for thinking of straight lines as existing before they are drawn (8/I.8 and 17/I.21), while in MS 118 he warns against picturing “the logical machine” as an “ethereal mechanism” (90v/I.119). Such remarks—these are just a few among many—are pitted against the philosophical myth of rules and calculi as ghostly entities graspable by pure thought or intuition. As Wittgenstein observes in MS 119, we have “no model of this superlative fact” but “are seduced into using a super-expression” (36/I.124). He is deploring what he describes in MS 142, when deprecating our “tendency to sublime the logic of our language” (§35/§38). It is significant too that in his discussion of the misunderstandings regarding the apparent inexorability of rule-following, Wittgenstein takes it for granted that rules are followed and calculi are applied. The gist of his argument is that rules and calculi do not in and of themselves force us to do anything, never mind compel us to conform with what they demand. In MS 117, in remarks concerning the +2 series, he accepts that 2006, not 2008, follows 2004 and asks how we know it (19/I.3). This is not to dispute the existence of a +2 calculus, only to dispute a view regarding its “peculiar inexorability” (3/I.4). For him the “inexorability” here is like the inexorability of “logical inference” (5/I.5) and the “process” is not at all “occult” (6/I.6). Once the notion of a calculus is disentangled from the theoretical assumptions with which it is regularly loaded, it is, Wittgenstein is saying, entirely unproblematic. There is no mystery, and philosophers’ theorizing about calculi is as unnecessary as it is indefensible. The remarks about mathematical belief in MS 117 are equally revealing (72–8/I.106–12). In this discussion Wittgenstein takes it to go without saying that 13 x 13 = 169, not 196, just as the multiplication table, or equivalent calculus, states. What concerns him is the sense in which someone rightly believes that 13 x 13 = 169 or wrongly believes that 13 x 13 = 196. He stresses that believing a mathematical calculation is not “a mysterious act” (73/I.106). There is no “underground connection with the correct calculation”, only the multiplication table, and it is absurd to suppose my believing that 13 x 13 = 169 involves more, indeed involves anything in addition to my ability to multiply by 13. Wittgenstein is not challenging the obvious fact that one can believe 13 x 13 = 169, even believe 13 x 13 = 196, in the everyday sense of “believe”. He is pointing out that these sentences are not believable in the sense that informative (true or false) sentences are believable, i.e., as corresponding or failing to correspond to the way the world is. The issue at stake is how the notion of a calculus is typically understood by philosophers, not whether the notion itself is suspect and mathematics is correctly or incorrectly regarded as a calculus (or series of calculi). In neither MS 142 nor MSS 117–19 is there, as far as I can see, any hard evidence of Wittgenstein defending or promoting a model of meaning. To oversimplify two highly complex investigations, he explores a range of philosophical problems and isolates a number of philosophical confusions,

Wittgenstein in the Mid-1930s

151

actual and potential. He seeks a standpoint from which to approach the problems and confusions, one from which the problems can be perceived to be more apparent than real and the confusions perceived to be just that— confusions. Much later, in Remarks on Colour (1950), in a passage I take to cover what he says in MS 142 and MSS 117–19, if not all his remarks, he says: “In philosophy we must always ask: ‘How must we look at this problem in order for it to become solvable?’” (1977, I.11). Taking philosophers to be their own worst enemies when it comes to understanding the nature and operation of calculi, Wittgenstein attempts to show them how to look at the problem so it becomes solvable (and no longer philosophically troublesome). This is not to advance a philosophical theory or model of language, nor is it to ridicule the philosopher’s worries. One might say Wittgenstein is engaged in a therapeutic exercise, but this merely puts a name to an exceedingly complicated business. It is overly simple to suppose he is a philosophical therapist if he is not a philosophical theorist.10 I trust it will not be thought blindingly obvious, given Wittgenstein’s aversion to “philosophical pictures”, that he would have had no truck with models of language, the calculus model and the language-game model included. There can be no disputing that he disparaged philosophical speculation and explanation and counselled philosophers to confine themselves to exploration and criticism. But it is by no means clear what philosophy limited to exploration and criticism comes to in practice when it comes to language and meaning. Even allowing for the possibility that Wittgenstein unwittingly defended or presupposed a philosophical theory of language and meaning, one before the mid-1930s and another one afterwards, it is altogether too pat to suppose he did not offer pictures of his own. Worse, there is the difficulty that he was not against pictures of every description but held, as he put it in Remarks on Colour (within parentheses): “The wrong picture confuses, the right picture helps” (III.20). To rule out the suggestion that he worked with one picture of language at one time and another later, pictures reasonably spoken of as models, it has to be argued, not presupposed, that his thinking cannot be summarized as commonly attempted in the literature, just what I have been attempting. If nothing else, it should be clear that there is no quick way of proving that Wittgenstein radically shifted his view in the mid-1930s. Remarks that seem to point to this conclusion have to be handled carefully remembering that in MS 142 and MSS 117–19 he continued to avail himself of the idiom of calculi as well as the idiom of language-games. Consider, for example, the remarks about our use of the word “oh” as a sigh in the so-called Grosses Format (1934), material posthumously published in Philosophical Grammar (Wittgenstein 1974, 67). It is tempting to regard Wittgenstein as favouring the language-game model when he asks: “[W]hat corresponds now to the calculus, the complicated game we play with other words?” and adds: “In the use of the words ‘oh!’, or ‘hurrah’, or ‘hm’, there is nothing comparable”. This, however, settles next to nothing. Leaving aside the fact

152

Andrew Lugg

that he identifies meaning of a word with its “role . . . in the calculus”, there is the awkward fact that he does not imply that the meaning of “oh” is the part it plays in our lives as opposed to how it functions in sentences. “If we were asked about it”, he writes, “we would say ‘“oh”! is a sigh; we say, for instance, things like “Oh, it is raining again already”’”, i.e., he would mention the part it plays in the calculus. Getting straight about Wittgenstein’s thinking about calculi and language-games is especially important since both the calculus and the language-game conceptions were at play from the start, though not under these names. In the Tractatus he not only refers to language as a symbolism, he also expresses thoughts most naturally regarded as having to do with use and application. Thus he says: “What does not get expressed in the sign is shown by its application. What the signs conceal, their application declares” (1933, 3.262), and he adds: “In order to recognize the symbol in the sign we must consider the significant use. The sign determines a logical form only together with its logical syntactic application” (3.326–3.327).11 Moreover, in the 1933-version of the Big Typescript, written before the shift to language-games was supposed to have occurred, he maintains that the understanding we have of a language is like “the understanding we have of a calculus when we come to know its existence or its practical application” (2005, 3; also with small changes in the 1937 version of the work).12 And he is reported to have said in 1939, i.e., well after the shift is alleged to have taken place, that “when the expression ‘imaginary numbers’ was introduced . . . the phrase . . . was used to join up [a] new calculus with the old calculus of numbers” and even to have said: “‘300’ is given its meaning by the calculus—that meaning which it has in the sentence ‘There are 300 men in this college’” (1976, 15 and 249–50).13 More importantly still, if Wittgenstein is taken to have relinquished the calculus conception by the time he wrote MS 142 and MSS 117–19, it is, to put it mildly, hard to do justice to the remarks he penned at the end of his life. In Remarks on Colour (1950), for instance, there are remarkably few references to language-games, mostly perfunctory, and rather than accepting “the language-game model”, Wittgenstein focuses on the grammar of colour concepts. While in the final arrangement of the material Wittgenstein begins by inviting the reader to consider a couple of language-games, he does so only to orient the discussion, specifically to draw attention to the possibility of sentences of the form “X is lighter than Y” being used to express either an internal or an external relation (1977, I.1/III.131). For the most part, he clarifies how we think and speak about colours (and the relations among them). Thus regarding the fact that certain colours are darker than others, he says: “Here we have a sort of mathematics of colour” (III.3) and writes: “Among the colours: Kinship and Contrast. (And that is logic)” (III.46). Nor is it beside the point that he explores the idea of a “people having a geometry of colours different from our own” (III.86 and I.66/III.154) and says: “We do not want to find a theory of colour (neither a physiological nor

Wittgenstein in the Mid-1930s

153

a psychological one), but rather the logic of colour concepts” (I.22/III.188). As Rush Rhees, who knew Wittgenstein well during the period, not least in 1950, observes: “Wittgenstein came to the views he did because he was serious about logic” (1967, 77).14 REFERENCES Gerrard, Steven. 1991. “Wittgenstein’s Philosophies of Mathematics”. Synthese 87: 125–42. ———. 1996. “A philosophy of mathematics between two camps”. In The Cambridge Companion to Wittgenstein, edited by Hans Sluga and David G. Stern, 171–97. Cambridge: Cambridge University Press. Glock, Hans-Johann. 1996. A Wittgenstein Dictionary. Oxford: Blackwell. Hacker, P. M. S. 1986. Insight and Illusion: Themes in the Philosophy of Wittgenstein. Revised edition. Oxford: Oxford University Press. ———. 1996. Wittgenstein’s Place in Twentieth-Century Analytic Philosophy. Oxford: Blackwell. Hilmy, S. Stephen. 1987. The Later Wittgenstein: The Emergence of a New Philosophical Method. Oxford: Basil Blackwell. McGuinness, Brian. 2002. Approaches to Wittgenstein: Collected Papers. London: Routledge. Medina, José. 2002. The Unity of Wittgenstein’s Philosophy. Albany: SUNY Press. Rhees, Rush. 1967. Contribution to “Ludwig Wittgenstein: A Symposium”. In Ludwig Wittgenstein: The Man and His Philosophy, edited by K. T. Fann, 74–8. New York: Dell. Shanker, Stuart G. 1987. Wittgenstein and the Turning-Point in the Philosophy of Mathematics. Albany: SUNY Press. Stern, David G. 1995. Wittgenstein on Mind and Language. New York: Oxford University Press. ter Hark, Michel. 1990. Beyond the Inner and the Outer: Wittgenstein’s Philosophy of Psychology. Dordrecht: Kluwer. Wittgenstein, Ludwig. 1933. Tractatus Logico-Philosophicus. Translated by C. K. Ogden. London: Routledge & Kegan Paul. ———. 1953. Philosophical Investigations. Edited by G. E. M. Anscombe and Rush Rhees. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. ———. 1958. Preliminary Studies for the “Philosophical Investigations”: Generally known as The Blue and Brown Books. Edited by Rush Rhees. Oxford: Basil Blackwell. ———. 1974. Philosophical Grammar. Edited by Rush Rhees. Translated by Anthony Kenny. Oxford: Basil Blackwell. ———. 1975. Philosophical Remarks. Second edition. Edited by Rush Rhees. Translated by Raymond Hargreaves and Roger White. Oxford: Basil Blackwell. ———. 1976. Wittgenstein’s Lectures on the Foundations of Mathematics: Cambridge 1939. Edited by Cora Diamond. Chicago: Chicago University Press. ———. 1977. Remarks on Colour. Edited by G. E. M. Anscombe. Translated by Linda L. McAlister and Margarete Schättle. Oxford: Basil Blackwell. ———. 1978. Remarks on the Foundations of Mathematics. Third edition. Edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. ———. 1979. Notebooks 1914–1916. Second edition. Edited by G. H. von Wright and G. E. M. Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell.

154

Andrew Lugg

———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt: Suhrkamp. ———. 2005. The Big Typescript: TS 213. Edited and translated by C. Grant Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell Publishing. ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. Wrigley, Michael. 1993. “The Continuity of Wittgenstein’s Philosophy of Mathematics”. In Wittgenstein’s Philosophy of Mathematics: Proceedings of the 15th International Wittgenstein-Symposium, edited by Klaus Puhl, 73–84. Vienna: Hölder-Pichler-Tempsky.

9

Wittgenstein on Rules in Language and Mathematics Severin Schroeder

Following a rule is an important topic in Wittgenstein’s later philosophy. In this paper I shall first give a brief account of some significant changes in Wittgenstein’s views on rules in language. Then I shall offer some remarks on the role of the rule-following discussion in the Philosophical Investigations in the light of those changes. Finally, I shall point out some noteworthy differences between Wittgenstein’s discussion of following a rule in the Investigations and his remarks on rules in the context of his philosophy of mathematics, which originally was also intended to become part of the book. 1. In the Tractatus Logico-Philosophicus Wittgenstein offered a mathematical picture of language: language as a calculus. The essence of language, the general form of the proposition was given by a simple formula (TLP, 6). Like a calculus, language was claimed to be governed by syntactic rules: (i) formation rules about the licit combination of names to form elementary propositions; (ii) formation rules about the licit combination of elementary propositions to make complex propositions; and finally (iii), truth-table rules: rules that enable us to identify logical truths and entailments. Notoriously, the existence of the first type of rule remained a postulate. As no examples of actual names were given, the rules governing their use could of course not be presented either. Moreover, Wittgenstein insisted rather perversely that no syntactic rule could be meaningfully stated. When later Wittgenstein grew dissatisfied with his earlier view of language, it was not the idea that language was essentially rule-governed that he found fault with. On the contrary, that idea he held on to emphatically, only correcting his account of what those rules were and how they functioned. In fact, it would not be much of an exaggeration to say that Wittgenstein’s break with his early philosophy was largely due to a careful reconsideration of the role of rules in language. Roughly speaking, while the author of the

156

Severin Schroeder

Tractatus thought that rules could work in secret and that their workings had to be discovered and analysed (just as one has to discover and analyse invisible chemical processes), the Wittgenstein of the 1930s realized that language is an artefact and to the extent to which it is governed by rules, those rules must be made and applied by us (BB, 27–8). Hence the idea that those rules could to a large extent be unknown to competent speakers and waiting to be unearthed by future logicians must be absurd; or “a hellish idea”, as Wittgenstein now called it in conversation with Friedrich Waismann (WVC, 129–30); just as absurd as the idea that nobody yet knows exactly what the rules of football are. Indeed the comparison between language and a game becomes one of the leitmotivs of Wittgenstein’s later philosophy, as suggested by his concept of a language-game. The title of Chapter 45 of the Big Typescript makes his new view explicit: Language functions as language only by virtue of the rules we follow in using it, just as a game is a game only by virtue of its rules. (BT, 196) Apart from his rejecting his earlier idea of “subterranean” rules, not known to those who follow them, his 1930s account of linguistic rules differs in at least two more respects from the Tractatus account: First, his concern shifted from syntactic rules to semantic rules. Secondly, in his later discussions he no longer insists on a complete system of exact rules; indeed, he becomes increasingly critical of the idea that language is rule-governed, entering a number of important reservations. I shall say a little more about both points. First, in the Tractatus account there is no mention of semantic rules. The connection between a name and its meaning, that is, the object named, is not fixed by a rule, but by a mental act of meaning (NB, 104, 130; PG, 97). (Hence, Wittgenstein will spend a lot of time in his later philosophy exorcizing this idea that linguistic meaning depends on mental processes of meaning something.) By contrast, Wittgenstein’s 1930s concern with linguistic rules is mainly focused on semantic rules, explanations of the meanings of words. Even ostensive explanations, such as “This colour Ö . . . is called ‘red’”, are now regarded as rules (BT, 199, 234; PG, 88). An ostensive explanation can perhaps be called a “rule” if it involves a canonical sample (such as the standard metre in Paris). Ordinary ostensive explanations, however, that explain a word by pointing to whatever suitable instance of the concept is at hand, are more plausibly regarded as explanations by example than as rules. The difference between these two types of explanation is that statements of rules, unlike instantiations, are not themselves “moves in the game”. Thus, to explain the waltz by giving a list or diagram of the correct steps can be called: giving a rule. The explanation is not itself a performance of the dance. But one can also teach the waltz by giving the learner a demonstration of it. This would not be a rule, but an explanation by example. For the instructor’s teaching is itself an instance

Wittgenstein on Rules in Language and Mathematics

157

of dancing the waltz, a “move in the game”. Similarly, a casual ostensive explanation of the word “purple” by pointing at and naming the colour of a violet is simply an instance of a correct application of the word (“This flower is purple”); and as such it is already a move in the language-game (teaching by doing).1 Secondly, over the years Wittgenstein came up with a number of important qualifications and reservations about his idea of language being rule-governed, like a game: (i) Linguistic rules are not normally part of the game (“a tool of the game itself” [PI, §54; cf. §82]) in the way they are in some of Wittgenstein’s simplified language-games. He imagines, for example, that someone is given a table correlating letters and movements and then a sequence of letters as an order to move in a certain way (EPB, 139). To carry out the order one consults the table, that is, the rule. By contrast, if someone directs me in English (“Right. Straight on to the traffic lights, then second to the left.”), I do not consult a rule to derive my movements from his words. (ii) We have not been taught our mother tongue by rules (cf. PI, §54). And this is an important point for Wittgenstein, for he identifies as a typical philosophers’ mistake an inclination to describe our mastery of language in a way that already presupposes linguistic competence, like the learning of a foreign language (PI, §32). Therefore, he emphasizes that at the basic stages language teaching cannot be explaining (let alone giving rules),2 but must be training or drill (“Abrichtung”) (PI, §6). And even at later stages, language is for the most part learnt by picking up a sufficient number of examples of correct usage, rather than by rules or explicit explanations. “One learns the game by watching how others play it” (PI, §54). (iii) Hence linguistic rules for natural languages are not necessary for the actual use of a language, they are only summarizing descriptions of that use (PI, §54). “Grammar”, Wittgenstein remarks, “is a description of language ex post” (WA 3, 223). It is of course true that linguists’ systematic descriptions of a language, in dictionaries and grammar books, will subsequently have a normative and stabilizing impact on certain aspects (usually fairly subtle aspects) of the use of language by educated speakers; but this is a feature of a highly advanced and sophisticated literary culture. It is clearly not essential to the phenomenon of language and, from a historical point of view, would not have been relevant for most speakers of most languages for most of the time. (iv) Taking the points made so far into consideration, Wittgenstein withdraws his earlier claim that “language functions as language only by virtue of the rules we follow in using it, just as a game is a game only by virtue of its rules” (BT, 196). In a handwritten note he comments: That is not correct, in so far as no rules have to have been laid down for language; no more than for a game. But one can look at language (and a game) from the standpoint of a process that uses rules. (BT, 196)

158

Severin Schroeder

The rules of a language, he now suggests, are a useful fiction, like that of a social contract: “Contrat social”—here too, no actual contract was ever concluded; but the situation is more or less similar, analogous, to the one we’d be in, if . . . And there’s much to be gained in viewing it in terms of such a contract. (BT, 196v) It is, after all, a fact that language is based on normativity. We do not only use certain expressions, we regard them as correct, and others which we regard as incorrect we criticize and reject. This important feature of language is rightly emphasized by comparing language to a rule-governed game, even if the comparison does not give an accurate picture of the way linguistic normativity is conveyed and implemented. (v) However, there are further qualifications to be entered. While the author of the Tractatus had insisted on the determinacy of sense and on perfect linguistic precision, appearances to the contrary notwithstanding, the later Wittgenstein rejected this ideal as a prejudice and declared that our semantic rules, or norms, were often vague. He now states explicitly that there is no logical calculus underpinning our language, although we may conveniently use such a calculus as an object of comparison: . . . remember that in general we don’t use language according to strict rules—it hasn’t been taught us by means of strict rules, either. We, in our discussions on the other hand, constantly compare language with a calculus proceeding according to exact rules. (BB, 25; cf. PI, §81) (vi) The rules that could be written out to reflect the meanings of our words are not only vague, our lists of such rules would also remain incomplete: The game which we play with words “is not everywhere bounded by rules” (PI, §68). For example, we have no rules by which to decide how the word “chair” is to be applied to chair-like objects that keep disappearing like hallucinations (PI, §80). (vii) Wittgenstein’s reservations go further still. Acknowledging that linguistic rules need not actually be consulted, but may be nothing but summary descriptions of people’s use of words, he writes (in 1933–34): But what if observation does not clearly reveal any rule, and the question [put to a speaker as to what rule he follows] brings none to light?—For he did indeed give me an explanation when I asked him what he meant by “N”, but he was prepared to withdraw this explanation and alter it.—So how am I to determine the rule according to which he is playing? He does not know it himself. Or, more correctly: what is left for the expression “the rule according to which he proceeds” to say? (PI, §82)3

Wittgenstein on Rules in Language and Mathematics

159

Wittgenstein seems to envisage two reasons why in many cases no rule can be given: One is illustrated in the following section by an analogy with games: Doesn’t the analogy between language and games throw light here? We can easily imagine people amusing themselves in a field by playing with a ball like this: starting various existing games, but playing several without finishing them, and in between throwing the ball aimlessly into the air, chasing one another with the ball, throwing it at one another for a joke, and so on. . . . And is there not also the case where we play, and make up the rules as we go along? And even where we alter them—as we go along. (PI, §83) The idea seems to be that a variety of different language-games can be played with the same words and that (unlike the builders of §2 of the Investigations, who always use the same words in the same way) we tend frequently to move between them, mix them up, and introduce new variations.4 Picking out one of those uses, we might well be able to produce something like a rule, but then it will not be applicable to what we do with the same expression in another context. So we’d have to “withdraw” and “alter” our explanation; and will do so again when considering yet other uses. Our grip on linguistic normativity is essentially piecemeal, and the explanations that we can give manifesting our linguistic competence are always just provisional, read off from some language-game which could easily on another occasion shade into, or be developed into, a slightly different one. (viii) Another reason why it may not always be possible to state rules for the way we use language is given by the considerations that introduce the idea of a family resemblance concept. Wittgenstein suggests that when we try to formulate a rule, that is, a definition, for the use of the word “game”, we draw a blank (PI, §66). Here, as in many other cases, what is subsumed under a given concept cannot be derived from a general rule, for it has not been decided once and for all, but case by case, according to the way the concept was first introduced with only some applications in mind, and then applied or not applied to new kinds of cases as people saw fit. Thus, for example, tennis is a game, but judo is not, although they are both competitive, rule-governed sports. (ix) Finally, the view that language is rule-governed is not only qualified; not only weakened to the suggestion that it is at any rate illuminating to compare language to a rule-governed game; at one point Wittgenstein goes so far as to denounce his earlier view as the main source of philosophical puzzles (BB, 25–6).5 We are inclined to try to achieve a better understanding of concepts such as “time” or “knowledge” by finding a definition; but any one that comes to mind leads to paradoxical results. So at this point Wittgenstein

160

Severin Schroeder

suggests not only that the search for linguistic rules may not always be successful, but that it is likely to be misleading to embark on it at all! 2. If by the mid 1930s Wittgenstein had so many reservations about regarding language as rule-governed, why should he spend so much time in the Philosophical Investigations discussing what it is to follow a rule? First of all, it has to be said that Wittgenstein’s use of the word “rule” (and similar terms) is sometimes a bit careless. This may partly be a hangover from an earlier period when he did believe rules to be operative in every use of signs (cf. MS 156a, 58v). For example, in §143 a language-game is introduced in which “when A gives an order, B has to write down series of signs according to a certain formation rule [Bildungsgesetz]”. But then—and this is very important—the first of these series (the only one discussed in that section) is the series of natural numbers, which is not taught or written down by following a formation rule. Sometimes Wittgenstein considers regular continuations of patterns of dots (e.g., in PI, §208). In some sense this can perhaps be called “rule following”, but it is noteworthy that the instruction is given by a paradigm, not by a rule in the proper sense of the word. In §237 Wittgenstein implies that if you draw a line parallel to another, you would use the latter as a rule. Again, “paradigm” or “exemplar” would perhaps be better terms. In §85 Wittgenstein compares a rule to a signpost (Wegweiser), and in §198 he considers a signpost as the expression of a rule. That’s unfortunate, for a signpost is just a piece of geographical information, equivalent to a statement, for example, that Shipton under Wychwood lies 2 miles in this direction. However, the most important point about Wittgenstein’s use of the word “rule” in the Investigations is explained in the Brown Book. There he distinguishes between two kinds of (what he calls) rules, namely: semantic rules and instruction rules (BB, 96, 98; cp. EPB, 140, 143). A semantic rule gives the meanings of signs. For example, in the form of a table: a b c d

move to the right move to the left move forward move backward.

Using those signs to give somebody an order (e.g., “cada”) would rely on the previously given rule, but it would not itself be a rule. However, where such an order is meant to be followed again and again, say, in drawing a

Wittgenstein on Rules in Language and Mathematics

161

continuing ornamental pattern, Wittgenstein is happy to call it a “rule”, too. It is what I call an instruction rule: In this case I think we should say that “cada” is the rule for drawing the design. Roughly speaking, it characterizes what we call a rule to be applied repeatedly, in an indefinite number of instances. (BB, 96) This, I think, helps us to understand what is going on in the rule-following discussions in the Investigations. What I presented above were Wittgenstein’s misgivings about the assumption of semantic rules, but what occupies him in the Investigations under the title “following a rule” is something different: It is not semantic rules, but instruction rules; that is, orders to continue doing something in a regular manner (e.g., writing down series of numbers).6 Why then should Wittgenstein be interested in such orders?—Because his main concern in those parts of the book is the concept of understanding. And orders and their execution provide the most straightforward and most perspicuous example of semantic understanding (cf. MS 165, 30). If you make a statement, giving me a piece of information, my understanding may or may not show in my behaviour. Telling me that there is beer in the fridge may make me go and open the fridge or not, depending on whether I would like any just now. In that way, most linguistic understanding has no direct behavioural manifestation. Orders given to somebody assumed to be willing to comply are rather different. Understanding or misunderstanding show immediately in what is done. At this point language comes into direct contact with action. Hence Wittgenstein’s preference for language-games of ordering, such as buying apples (PI, §1) or directing an assistant to pass building material (PI, §2). Moreover, Wittgenstein had a particular interest in an understanding that goes beyond a particular occasion, such as the understanding of a system (PI, §143). For one thing, because such an understanding of complex, possibly even infinite contents seems particularly puzzling, especially as it can happen in a flash. We are inclined to think of understanding as having mental representations; yet how can a highly complex system be represented instantaneously in one’s mind?7—For another thing, linguistic understanding is obviously systematic. We do not just understand single utterances (tokens); we understand types of words and expressions, such that we know how to apply them or how to respond to them again and again, on an endless number of occasions. So, there is no tension between Wittgenstein’s reservations about the importance of semantic rules and his intensive discussion of following rules, because the well-known examples of rule-following in the Investigations (continuing arithmetical series) are not concerned with semantic rules. They are just cases of carrying out orders with an endless applicability. As such they are just variations of the builders language-game: “Write down the

162

Severin Schroeder

series of even numbers!” is like “Keep bringing me slabs!”8 The deviant pupil of §185 could just as well have been presented as a deviant builder’s assistant who, when given the order “slab” for the fifth time, brings a block &c.—So it’s not that all linguistic understanding is a form of rule-following; rather: rule-following is just one simple type of linguistic understanding, one language-game. There is, however, more to be said about the relation between Wittgenstein’s misgivings about semantic rules and his discussion of rule-following. As I said, his concern with understanding how to continue an arithmetical series is that, in this case, an act of understanding must cover an infinity of instances. How is that possible? How can an infinity be grasped in an instance? The most natural answer is that such understanding can only be achieved by means of a general rule or formula, which although easily grasped in a moment can determine an infinity of instances. So, what those discussions in the Investigations are meant to investigate is the possibility of an endlessly applicable understanding. That is the explanandum, the common phenomenon that needs to be clarified. Rules or formulas are only considered as an explanans, as a suggested solution—which is shown not to work. Our understanding cannot be based on rules. No rule can guarantee understanding: we still have to know how to apply it in any given case. Thus the puzzle about the infinity of instances resurfaces as a puzzle about the infinity of applications of a rule. Of course it is possible to learn how to apply a rule to indefinitely many cases. But ultimately, our mastery of rules has to be grounded in an ability to continue in what we regard as a regular manner that is not guided by any rule, but can only be taught by examples (MS 136, 124a; Z, §§300–1).9 Hence, although Wittgenstein was certainly no sceptic about the possibility of following a rule, it can be said that his reservations about the importance of linguistic rules dovetail neatly with his discussions of continuing an arithmetical series. While earlier he argued that rules are not necessary for understanding general notions (cf. Z, §295), now he shows that rules are not sufficient for such understanding either. Even if there are linguistic rules, ultimately they cannot be the foundation of linguistic normativity. What is held up both as an alternative to and as a basis for rules—is examples and training. 3. By 1937 Wittgenstein had completed about the first 189 sections of the Philosophical Investigations (the so-called “Proto-Investigations”: MS 142). His plan was to continue with a discussion of rule-following in mathematics and the nature of mathematical proof. First attempts at this continuation are TS 221 and its revision TS 222 (later published as Part I of Remarks on the Foundations of Mathematics). Then, however, he

Wittgenstein on Rules in Language and Mathematics

163

changed his mind, and in 1944 he produced an entirely different continuation in which, after an extended discussion of following a rule (i.e., an instruction rule or open-ended order), he moved into the philosophy of mind (“Intermediate Version”). In the end, his work on the philosophy of mathematics was not included in the Philosophical Investigations, and indeed abandoned. In it we find considerations of rule-following that to some extent overlap with the remarks in the Investigations,10 but to some extent go in different directions. That is not surprising in light of the obvious differences between ordinary language and mathematics. The most striking difference is this: Mathematics is rule-governed, in the full sense of the word, whereas language is not. To Wittgenstein, language appeared to be strictly rule-governed only while he was in the grip of a mathematical picture of language: as a calculus. As argued, the piecemeal normativity of language may sometimes be conveniently described in terms of rules, but speakers do not actually follow these rules. They figure in ex post descriptions of language, not in its actual practice. Mathematics, by contrast, is a calculus, or better: an open-ended network of calculi, which one can master only by learning its rules. Children do not pick up maths the way they pick up language; they have to be drilled to learn their multiplication tables, which are (as Wittgenstein emphasizes) transformation rules that together with some further rules (e.g., “Negative × negative = positive”; or the BEDMAS rule for the order of operations) allow one to work out more complicated calculations. At a slightly higher level one has to learn rules like Pythagoras’ Theorem, the Cosine Rule or the Quadratic Formula. Thus, rule following, which was introduced only as a postulate in Wittgenstein’s philosophy of language, is a common and indubitable fact in mathematics. According to Wittgenstein, the concept of rules is essential to mathematics in yet another way. He argues that mathematical propositions—all mathematical propositions—function as rules, as norms of representation, when mathematics is applied to empirical propositions. So he regards “3×3 = 9” as a rule in two ways: First, uncontroversially, as a calculation rule that is applied in the process of carrying out more complicated multiplications, such as: 31,973 × 33. This long multiplication would not, in this sense, serve as a rule. We do not memorize it for ready application in other calculations, as we do with our elementary times tables.—But then, secondly, Wittgenstein holds, “3×3 = 9” functions not just as a rule of maths, but also as a rule of language, that is to say: as a rule for transforming empirical number statements. It licences me, for example, to derive from “I have given to three children three apples each” the proposition: “I have given out nine apples”.

164

Severin Schroeder

Yet this normative function is a feature of all mathematical propositions, not just those that are commonly reapplied inside mathematics. Thus, “31,937 × 33 = 1,053,921” can be applied as a grammatical rule to the proposition: “I own 31,937 shares each worth £33” to derive: “I own shares worth £ 1,053,921.” However, this key idea in Wittgenstein’s philosophy of mathematics (that mathematical propositions are grammatical norms) appears to involve another philosophical and extended use of the term “rule”, for large sums and nonelementary mathematical propositions (which are not memorized for repeated application) would not normally be called “rules”. Wittgenstein acknowledges that when in his 1937 lectures he suggests, more carefully, that it is illuminating to compare mathematical propositions to rules.11 Elsewhere he observes that a “mathematical proposition has the dignity of a rule” (RFM, 99; cf. 363); or that it is “akin to a rule” (RPP I, §266); and: “What I say amounts to this: mathematics is normative” (RFM, 425). Therefore, when it comes to the issue of actually following rules in the full sense of the word, we had better focus on the many straightforward and uncontroversial instances in mathematics, where rules are conventionally taught and learnt for repeated application. There are two important themes in Wittgenstein’s discussion of rulefollowing in mathematics that do not play much of role in the considerations pursued in the Investigations: First, the hardness of the logical must; and secondly, the possibility of rules’ having unexpected results. I shall say a little more about each point. (i) The hardness of the logical must. The discussion in the Investigations brings up the question, “How am I able to follow a rule?” (PI, §217). That is to say: How can I ever know what the correct application of a rule, or an order, is at a given point? (PI, §198). The problem is presented in epistemological terms: of knowing what is right. In TS 222, however, the same problem is given a slightly different twist, or shifted in a different direction (only touched upon, but not developed in the Investigations). Namely: a lot of emphasis is put on the idea of compulsion. It’s not only that writing 2002 after 2000 is the correct application of the rule or order “+2”; it is also a step which seems to be forced on us. We must continue in this way, or so it appears. Whence this extraordinary authority of a rule? Whenever we have such an experience of absolute necessity, of a link of inexorable hardness, Wittgenstein remarks, this is due to a grammatical norm (RFM, 88; cf. AWL, 16). And the inexorability is ultimately just our own attitude towards the rules we’ve given (RFM 82, 170, 430; Z, §299). Because the practice of counting, for example, is so important to us: “we

Wittgenstein on Rules in Language and Mathematics

165

learn to count as we do: with endless practice, with merciless exactitude; that is why it is inexorably insisted that we shall all say ‘two’ after ‘one’, ‘three’ after ‘two’ and so on” (RFM, 37). It is not surprising that this aspect of rule-following goes with the discussion of mathematical rules, and not so much with the quasi or ex post rules of language, because we are typically not as rigorous and inexorable about linguistic norms. In analytic philosophy this difference is often obscured by the prevalence of the mathematical calculus model of language. Ordinary language is regarded as, basically, formal logic with an interpretation and a bit of pragmatics. But it isn’t really like that. For instance, “p . ~p” is, according to the rules of the calculus, dismissed as a necessary falsehood, from which anything would follow. But faced with a corresponding contradiction in ordinary language, we are usually able and willing to make sense of it. “It’s raining and it’s not raining”—meaning there are a few drops now and then, but you can’t make up your mind whether to take an umbrella. “I believe it and I don’t believe it.” “I’m free and I’m not free.” These are fairly common locutions, used to express ambivalence. You are inclined to say different things depending on which aspects of the situation you focus on. Again, for a logician, like Frege, “This lamp is different from itself” is a contradiction, impossible to hold true; but in fact it is not so difficult to find some descriptive meaning for it. As Wittgenstein observes, we can quite well use that sentence to describe a certain way in which the lamp shimmers in sunlight (RFM, 89–90). The inexorable rules of mathematics and formal logic are transformation rules to which we allow no exceptions (RFM, 40–1, 324–5). In ordinary language, on the other hand, absolutely neat semantic equations are relatively rare. A pint is 0.565 litre, and a vixen is a female fox; but most words are not definable in such a tidy way. Even the applications of mathematical equations to ordinary objects and operations such as adding or taking away allow for some deviation. 12 + 3 = 15; but if I put first twelve and then another three apples into a basket, it is not utterly inconceivable that the total number of apples in the basket is only fourteen. So, “12 + 3 = 15” is an inexorable rule of arithmetic; but the corresponding calculation of the number of apples in a basket hasn’t got the same normative status. It is not inconceivable that the operation of adding three apples to twelve may occasionally lead to a different result. Indeed, there are familiar cases where the result of adding up numbers of actual things or units of substances cannot be calculated arithmetically: One quart of alcohol and one quart of water yield only about 1.8 quarts of vodka; and three tablespoons of water and one tablespoon of salt do not make four tablespoons.12 In TS 222 (RFM, I), Wittgenstein gives a couple of examples of real-life inferences: “The stove is smoking, so the stove pipe has been moved out of position again” (RFM, 40, translation modified). Or: “All who are taller than five foot six are to join the . . . section. . . . ‘N.N. [is] five foot nine.’ ‘So N.N. to the . . . section’” (RFM, 43). In both cases we can easily imagine

166

Severin Schroeder

how these inferences might be rejected. The stove could be smoking for other reasons; and N.N. comes with a special recommendation from the general, so he is exempt from the rule and to go straight to HQ. (ii) In mathematics rules can have unexpected results. Wittgenstein illustrates that with the analogy of a game: I give the rules of a game. The other party makes a move, perfectly in accord with the rules, whose possibility I had not foreseen, and which spoils the game, that is, as I had wanted it to be. . . . It was surely possible, for example, for me not to have foreseen that a quadratic equation need have no real root. Thus the rule leads me to something of which I say: “I did not expect this pattern: I imagined a solution always like this. . . .” (Z, §293; cf. PI, §125) This leads to Wittgenstein’s discussion of the concern for consistency and the search for consistency proofs in mathematics: the worry that there may be a hidden contradiction in our calculus (LFM, 209 ff.). Again, this is an issue about rules that does not fit the rule-following consideration in the Investigations and their concern with ordinary linguistic understanding. For ordinary language has not been devised on the drawing board, like a calculus or a card game. Special scientific or legal contexts apart, nobody has given linguistic rules for our words in advance of their use. So in ordinary language we cannot encounter that kind of surprise or dismay at finding that the game doesn’t work as intended. However, interestingly, something analogous to a mathematician’s coming across unexpected consequences of his rules exists in philosophy. This relates to Wittgenstein’s diagnosis in the Blue Book that our striving for rules or definitions is a major source of philosophical confusion (BB, 25–6). We try to derive general rules from the observation of our use of certain words, only to find later that those rules are in conflict with aspects of common usage or with each other. So one could say that in philosophy we stipulate the rules for a language-game with the word “knowledge”, say, or “free will”, only to find later that the resulting game is not the one we do in fact play with those words, and what’s more, it isn’t one we would want to play.13 REFERENCES Kline, Morris. 1980. Mathematics: The Loss of Certainty. Oxford: Oxford University Press. Schroeder, Severin. 2001. “Elucidation and Ostensive Explanation”. In Wittgenstein Studies. Special Issue: From the Tractatus to the Tractatus, edited by Gianluigi Oliveri, 69–79. Frankfurt am Main: Peter Lang. ———. 2006. Wittgenstein: The Way Out of the Fly-Bottle. Cambridge: Polity.

Wittgenstein on Rules in Language and Mathematics

167

Wittgenstein, Ludwig. 1958. Preliminary Studies for the “Philosophical Investigations”: Generally known as The Blue and Brown Books. Edited by Rush Rhees. Oxford: Basil Blackwell. (BB) ———. 1961. Tractatus Logico-Philosophicus. Translated by D. F. Pears and B. F. McGuinness. London: Routledge & Kegan Paul. (TLP) ———. 1967. Zettel. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (Z) ———. 1974. Philosophical Grammar. Edited by Rush Rhees. Translated by Anthony Kenny. Oxford: Basil Blackwell. (PG) ———. 1976. Wittgenstein’s Lectures on the Foundations of Mathematics: Cambridge, 1939. Edited by Cora Diamond. Hassocks, Sussex: Harvester Press. (LFM) ———. 1978. Remarks on the Foundations of Mathematics. Third Edition. Edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RFM) ———. 1979a. Notebooks 1914–1916. Second edition. Edited by G. H. von Wright and G. E. M Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (NB) ———. 1979b. Ludwig Wittgenstein and the Vienna Circle: Conversations recorded by Friedrich Waismann. Edited by Brian McGuinness. Translated by Joachim Schulte and Brian McGuinness. Oxford: Basil Blackwell. (WVC) ———. 1979c. Wittgenstein’s Lectures: Cambridge, 1932–1935. From the Notes of Alice Ambrose and Margaret Macdonald. Edited by Alice Ambrose. Oxford: Basil Blackwell. (AWL) ———. 1980. Remarks on the Philosophy of Psychology, vol. I. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RPP I) ———. 1984. “Eine Philosophische Betrachtung”. Edited by Rush Rhees. In Werkausgabe, vol. 5, 117–282. Frankfurt am Main: Suhrkamp. (EPB) ———. 1995.Wiener Ausgabe, vol. 3. Edited by Michael Nedo. Vienna: Springer. (WA 3) ———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2005. The Big Typescript: TS 213. Edited and translated by C. Grant Luckhardt and Maximilian A. E. Aue. Oxford: Blackwell Publishing. (BT) ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell. (PI)

10 The Form of Proofs Wittgenstein vs. Principia Mathematica André Maury

In a simple proof you move from one proposition to another, from p to q. The proof is right if p is known, and q indeed follows from p. The move— transition or Übergang as Wittgenstein calls it—is then alright. This pattern is used when we say p, so q. The form in itself—the “so” here meaning follows—does not guarantee that you are right. The “so” may be unjustified and in fact q does not follow though you think it does. Jealous husbands, for instance, engage in such dubious proofs! You may also make a mistake as to the truth of the premiss p. The topic we shall discuss here is the place of the rule “q follows from p”, the backbone of the proof. Let us call conclusive arguments proofs. Wherever you find a purported conclusive argument—in language-games or logic—you see the form “p, so q”. In a conclusive argument you have proved your point; that is, you have come up with a proof. Note that the basic structure is extremely simple and the proof is effected in one move. It is curious that the most important feature of thought is such a simple device. Despite the simplicity of the proof-form, as I shall call it, Principia Mathematica and Wittgenstein manage to say a great deal about it. Wittgenstein involved himself in a controversy with Russell about the status of rules in inferences, and thus also in proofs. In explaining the matter I rely on the distinction between showing and saying, which is familiar from the Tractatus although Wittgenstein does not later use it himself. However, I hope to show that he writes in the spirit of that very early invention of his. For Wittgenstein the proof-form shows a rule of inference, which is not mentioned. In PM, on the other hand, the rule is a (third) proposition. It is hidden away, but it is still meant. “Regelfolgen” is then first a feature of inference. The rules work on propositions and effect relations between them. When Wittgenstein eventually removes this from the Philosophical Investigations and concentrates, or so it would seem, on rules of meaning, he changes the topic. However, the change of emphasis builds on a distinction made by Russell as early as in The Principles of Mathematics. Here, in §15, a material implication is one between propositions proper. A “formal” implication is one between propositional functions, that is, it basically predicates. Formal implications display that feature of logic which is seen as their

The Form of Proofs

169

hallmark—generality. Connections between predicates are obviously bound up with their meaning. The very term “material implication” is not originally meant as a contrast to latter-day “entailment”. Rather it is that feature of logic which is displayed in language-games (also a latter-day term). My main reference is to the second part of the early version of the PI, TS 221 in von Wright’s list. Most of it was later published in the Remarks on the Foundations of Mathematics, the title of which was, I believe, chosen by the editors Anscombe, von Wright and Rhees. The early version of PI consisted of two parts. The first one, TS 220, consisted of the material now published as §§1–189. The second, TS 221, consisted of material published, with changes, as Part I of Remarks on the Foundations of Mathematics. Before the final version, there was an intermediate version from 1944, as explained by von Wright in “The Origin and Composition of the Philosophical Investigations”.1 In that version Wittgenstein takes out the material on mathematics and replaces it with new material, primarily on the topics of “Regelfolgen” and private language. This meant that all references to Russell disappeared, which was, incidentally, quite late. There is an extensive discussion of logical rules in 221 with explicit references to Russell, in many cases actually to PM. Wittgenstein’s discussions of rules underwent a change during the war. In the notebooks 122 and 123 rules are discussed intensively, but the references to Russell are dropped. The topic is still logical rules, but there is a marked change. The discussion concerns rules in general and very few examples are used. In the final text around §200 there is a shift from logical rules to rules of meaning. This might very well be the reason why the discussion of logical rules, and Russell, was taken out. However, it is worth noting that the general issue of “Regelfolgen” grew out of the discussion of logical rules, which is our topic. As an editor of Wittgenstein’s texts G. H. von Wright was exemplary. He worked on the texts for almost 40 years, on a daily basis as far as I can tell. Rhees and Anscombe were involved too, but to me it very much seems as if von Wright was the primus motor. I might be somewhat biased though, having worked with him on the papers in the late 1970s. To my mind it is not wrong to say that von Wright helped the mature Wittgenstein in a way which resembles Russell’s help to young Ludwig. Be that as it may, it is curious that von Wright was not really under his influence philosophically. As far as the Investigations is concerned, von Wright initiated the work which eventually resulted in the “Helsinki edition” of PI. Here he was greatly assisted by Mr. Heikki Nyman. The publication of the volume with the four versions of PI was in the end carried out and overseen by Joachim Schulte and Eike von Savigny.2 The book is a landmark in the editing of philosophical texts. Wittgenstein’s references to Russell are mainly in fact to Principia Mathematica by Whitehead and Russell. PM was published in 1910 so Wittgenstein’s discussion in TS 221, from 1938, is a late return to an early acquaintance. There is no great difference between 221 and my quotes from

170

André Maury

the published text, with one exception which I will discuss later, so the references are to the published text, mainly section 19. The controversy seems to be of a minor sort. The proof-form is, according to PM, just the surface form of proofs. The real form is slightly more complicated. When we say p, so q, we mean: p, q follows from p, so q. Of course we do not say this, but we mean it (PM, 9). This seems innocent enough. However, according to Wittgenstein, it is a bad mistake to put things in this way. His point seems to be that you cannot make the form of the proof part of the proof. I think Wittgenstein is right in this. As far as I know, Wittgenstein’s criticism of PM on this point has been almost completely neglected. One reason is that the opening sections of RFM get all the attention. The epistemology of working out the series “add 2” steals the show. The discussion of PM, around section 19 in RFM, is neglected although the key to an understanding of the opening sections can be found here. At least, that is what I shall argue. Another reason seems to be that the topic of “Regelfolgen” is nowadays, as noted, seen as belonging to the theory of meaning. However, the topic was introduced by Wittgenstein within logic. He asked what following a logical rule involves, given an actual inference or proof. This discussion is in no way superseded by later developments. Elizabeth Anscombe used to say that she did not understand Wittgenstein. A case in point was the “add 2” example. If a pupil develops a series adding 2, then he may at some later point do “whatever he likes”, she said. I do not understand this, she added, and looked very concerned indeed. In fact Wittgenstein denies in section 3 that the pupil can do whatever he likes. There still remains, though, a nagging feeling that there is something amiss in Wittgenstein’s discussion. He seems to be setting his face against our ordinary mathematical practice by raising the very question of how the pupil knows how to go on. This is atypical. Is it not one of the points in Wittgenstein that the fundamental thing, the first thing, is the very practice? As it turns out I have not been able to answer Anscombe’s concern. However, our discussion of PM will hopefully throw some light on the problem. Let us start with an example of an ordinary proof, a language-game that is. Suppose you say that a particular man is in Lisbon. I am surprised and ask how you know that. You say that you saw him. Having read some philosophy I act surprised. You say you saw him. That is a fact about you. You are quite irresponsible in holding that a fact about you could prove anything about other people and their whereabouts! Most people would be quite taken aback by this. Let us imagine, against all odds, that you have read PM, the prose parts at least. So you say: Logic backs me up! I saw him and, in the words of PM, anything implied by a true premiss is true (13). Seeing him here implies his being here, so he is here! Now I am taken aback, for a moment at least. But I see a way out. I point out to you that you invoke some kind of logical law or rule involving an “implication”. But in what logic book is the implication about him said to be a logical law? It is certainly not said in PM! As far as the ordinary language-game is

The Form of Proofs

171

concerned, you are quite right. We move from one proposition to another one. No logical rule is explicitly mentioned. We say: p, so q—no more. For Wittgenstein this is crucial. There is also another thing. The example shows what a logical rule is like, given a particular proof. In PM the point of mentioning the rule is to back up the move from p to q. The rule mentions the very same propositions which make up the proof. The difference between the rule and the proof lies elsewhere. The logical connection expressed by the rule can be expressed in various ways: “q follows from p”, “p implies q” or “p ⊃ q”. The definition of the horseshoe in PM, ~p ∨ q, brings out the truth-functional character of the rule, no more. Here, the fact that in a proof you move from one proposition to another is suppressed. That feature is preserved in the common expression “if p, then q”, which also expresses the rule. Since the rule mentions both the premiss and the conclusion of a proof, it is obvious that our normal language-game of moving from one proposition to another shows the rule invoked, used or adhered to. There is a parallel to this in the structure of action. Indeed, inferring is an action. We see the intention in the action. We also see the rule in the proof. Below is the passage in PM which is the main target of Wittgenstein’s criticism: Inference. The process of inference is as follows: a proposition “p” is asserted, and a proposition “p implies q” is asserted, and then as a sequel the proposition “q” is asserted. The trust in inference is the belief that if the two former assertions are not in error, the final assertion is not in error. Accordingly whenever, in symbols, where p and q have of course special determinations, “├ p” and “├ (p ⊃ q)” have occurred, then “├ q” will occur if it is desired to put it on record. . . . for the sake of drawing attention to the inference which is being made, we shall write instead “├ p ⊃├ q,” which is to be considered as a mere abbreviation of the threefold statement “├ p” and “├ (p ⊃ q)” and “├ q”. Thus “├ p ⊃ ├ q” may be read “p, therefore q”, being in fact the same abbreviation, essentially, as this is; for “p, therefore q” does not explicitly state, what is part of its meaning, that p implies q. An inference is the dropping of a true premiss; it is the dissolution of an implication. (PM, 8–9) If we start out instead from the proof-form, as we have done, it should be noted that in setting out the “hidden” proposition that p implies q, we do not thereby justify the inference or show that the proof thereby is right. PM only spells out, in its own way, what we mean. But, of course meaning it does not make it so. The question of justification is raised in PM on a

172

André Maury

general level. In listing the primitive propositions of the logic used in the book the first Pp. might surprise modern-day readers: Anything implied by a true premiss is true. This is the rule which justifies inference, it says (13). As noted in the preceding quote, this rule, which captures the very “process of inference”, cannot be put in a formula. The rule gives us the most general form of any proof: what follows from a true proposition is true. Even if the rule cannot be put in a formula it is, indeed, “shown” by any proof. Wittgenstein’s harsh criticism of PM, which I shall discuss below, is surprising. After all, the saying-showing distinction is quite clear in PM. Here, of course, “saying” means that a thing can be expressed in a formula. The difference between the general rule and a formula is also noted later in the book. On page 94 the rule is given as “Anything implied by a true elementary proposition is true”. The comment is: It is not the same as “if p is true, then if p implies q, q is true”. This is a true proposition, but it holds equally when p is not true and when p does not imply q. It does not, like the principle we are concerned with, enable us to assert q simply, without any hypothesis. We cannot express the principle symbolically, partly because any symbolism in which p is variable only gives the hypothesis that p is true, not the fact that it is true. Later the Pp. is given as “What is implied by a true premiss is true” (9.12, 132). The comment is: “That is to say, given ‘├ . p’ and ‘├ . p ⊃ q’ we may proceed to ‘├ . q’”. What is here put in symbols is an instance of the rule, not the rule itself —though this may be suggested by Wittgenstein (see below). The rule works for “p, so q”. Thus the thesis that an inference is the “dissolution of an implication” is not right in cases like the one above where you do not start out with a rule. There is nothing to “dissolve”. Wittgenstein alludes to this when he suggests that the intransitive verb “follows” might suggest that there is a ready-made connection between premiss and conclusion. Section 19 in RFM I reads: One is often in the dark about what following and inferring really consists in; what kind of fact, and what kind of procedure, it is. The peculiar use of these verbs suggests to us that following is the existence of a connexion between propositions, which connexion we follow up when we infer. This comes out very instructively in Russell’s account (Principia Mathematica). That a proposition ├ q follows from a proposition ├ p ⊃ q . p is here a fundamental law of logic: ├ p ⊃ q . p . ⊃ .├ q Now this, one says, justifies us in inferring ├ q from ├ p ⊃ q . p. But what does ‘inferring’, the procedure that is now justified, consist in? Surely in this: that in some language-game we utter, write down (etc.),

The Form of Proofs

173

the one proposition as an assertion after the other; and how can the fundamental law justify me in this? However, the formula given by Wittgenstein is not presented as a primitive proposition or a fundamental law of logic in PM. What Wittgenstein offers as a criticism is in fact quite in line with the position of PM, as we can see from the quotes above. If the language-game mentioned by Wittgenstein is one of going from p to q, then the formula mentioned does not justify the transition. The formula only tells us what is going on. It describes the form of this particular language-game. In the early editions of RFM the editors replaced Wittgenstein’s formula by the correct Pp., which is discussed below. In the next section Wittgenstein continues to take Russell to task for mistaken views of justification: Now Russell wants to say: “This is how I am going to infer, and it is right”. So he means to tell us how he means to infer: this is done by a rule of inference. . . . Quite correct, the rule is really only a piece of information that in this book only this transition from one proposition to another will be used (as it were a piece of information in the index); for the correctness of the transition must be evident where it is made; and the expression of the ‘fundamental law of logic’ is then the sequence of propositions itself. (RFM, I, §20) What Wittgenstein is in effect saying here has already been mentioned. First, adding “q follows from p” to the proof-form does not justify the move. It is no more than a description of it. That the move has to be right “where it is made” is somewhat bleak. The move is correct on the basis of the very “sequence of propositions”. The correctness of the move from p to q obviously depends on the structure of the proposition that p. In the early editions of RFM the formula mentioned by Wittgenstein was replaced by the quote from PM 9.12: “What is implied by a true premiss is true”. According to PM that rule in words justifies inference. The tautological inference mentioned by Wittgenstein does not, since it is an inference in itself. In PM, the formula is mentioned only as an instance of how the rule works, as we have seen. Indeed Wittgenstein’s discussion is better seen as a comment on the passage on page 9 in PM, quoted above. The inference is said to work as follows: a proposition “p” is asserted, and a proposition “p implies q” is asserted, and then as a sequel the proposition “q” is asserted. Here the proof-form is then seen as an abbreviation. When this passage is taken as the target, Wittgenstein’s criticism seems more to the point. Wittgenstein’s original formula was restored after a request by Rush Rhees. In a

174

André Maury

letter dated 4 April 1973 to G. H. von Wright, who was working on Part I for a new edition, Rhees said that the change back would be desirable. G. H. von Wright wrote “thank you!” in the margin of the letter (kept in the Wittgenstein/von Wright Archives at the University of Helsinki) and eventually the change was made. However, in the new edition there is still a footnote giving—perhaps misleadingly—a reference to 9.12. The insight that a proof has to stand on its own, to be correct “where it is made”, is a criticism of the view of an abbreviation, the view that there is a hidden proposition in the proof. In fact you cannot make that proposition, the logical rule which spells out the meaning of “so”, a premiss in the proof. By adding p ⊃ q as a premiss, you get a different proof. The proposition that q follows from p and p, is a conjunction which says more than the simple proposition that p. Since the rule involves the full premiss, the new rule will be different too. It is obvious then that we cannot state the form of the proof in that very proof—we come too late. Lewis Carroll famously showed that if you demand that the rule used in a proof be made explicit by adding it as a premiss, you end up in a regress. The new proof is, as noted, different and it shows a new rule. That has to be made a premiss, and so on.3 Indeed, the ensuing regress is a fine example of Wittgenstein’s early principle: What is shown cannot be said. Of course we have to add “cannot be said in the proof itself”. We can say it by indicating the rule on the side of the proof, as is done in logic books and in PM. The rules given alongside the proof are really only descriptions of what goes on—“a piece of information in the index”, as Wittgenstein puts it. Since the rule is shown by the proof, is it really necessary to assume that it occurs there as a proposition at all, though hidden? The radical view would be that there is no extra proposition. There are just two propositions, not three. The “so” just expresses an operation on p. What happens in the proof “p, so q” is that you spell out a truth-condition of p. Is a (third) proposition really needed here? Let us now return to Anscombe’s bewilderment. Let us see if any light is thrown on the “add 2” case, starting from the conception of a rule mentioning both premiss and conclusion. This is like asking why an action was done so we assume that the task of working out a series has already been done. In section 3 of RFM I Wittgenstein asks: How do I know that in working out the series + 2 I must write “20004, 20006” and not “20004, 20008”? Note that Wittgenstein does not say that the pupil could write 20008. Consequently, Anscombe’s reaction was unfounded. So, where does the problem lie? If there is one, it must lie in the relation between the rule and the step taken as these are the elements offered for consideration. A comparison with a simple proof may throw some light on this. Is there a gap between

The Form of Proofs

175

the rule and the step, or is the problem the opposite one of there not being enough of a gap to raise certain epistemological questions about the relation? The usual answer seems to be that there is a gap—the rule does not determine its own application. But then there would seem to be no great point in having rules. So let us for a change take the other line. In a simple proof the rule used is shown. The problem is that in making the rule a premiss, any sequence of propositions will be a proof, at least it will have the form of one. In working out + 2 the rule is likewise shown. If you make the actual rule used the justification for the step taken, you move in a circle. The rule is in the step taken. There is no gap and thus no room for a question of justification. The step is the criterion for the rule having been used. Later in RFM Wittgenstein puts the point thus: “Thus the truth of the proposition that 4 + 1 makes 5 is, so to speak, overdetermined” (319). The move is overdetermined, so the epistemologist must be more than satisfied. There is no room for the question “how do I know”, after all. So what is all the fuss about? Taking our lead from PM again, there might be a problem in what a rule really states. The rule “4 + 2 = 6” seems to state just what it says. PM raises some doubts here: [R]ules are not premisses, since they assert any instance of themselves, not something other than their instances. (98) What Wittgenstein might have had in mind is the (possible) illusion that a rule goes beyond its instances. If this is right, then Wittgenstein’s way of expressing it is very roundabout, to say the least. Many of the principles used in PM are already present in the early Principles of Mathematics. Indeed, the implicit references to Russell may very well be primarily references to the Principles. In fact, the fundamental distinction between material and formal implication is there, as is the problem of what the generalized rule states and its relation to actual instances. Yet it is curious that Wittgenstein accepts Russell’s position up to a point and then criticizes Russell! However, the idea of a “hidden” proposition in proofs is not in the Principles. It emerges only in PM. But the emphasis on the primacy of the “language-game” is already expressed in the Principles: 45. The fundamental importance of formal implication is brought out by the consideration that it is involved in all the rules of inference. This shows that we cannot hope wholly to define it in terms of material implication, but that some further element or elements must be involved. We may observe, however, that, in a particular inference, the rule according to which the inference proceeds is not required as a premiss. . . . it is just as easy, and consequently just as legitimate, to perceive immediately the implication in question as to perceive that it is implied by one or more of the rules of inference. Wittgenstein would most certainly not object to this.

176

André Maury

REFERENCES Carroll, Lewis. 1895. “What the Tortoise said to Achilles”. Mind 4: 278–80. Russell, Bertrand. 1964 (1903). The Principles of Mathematics. London: George Allen & Unwin. von Wright, G. H. 1982. “The Origin and Composition of the Philosophical Investigations”. In Wittgenstein, 111–36. Oxford: Basil Blackwell. Whitehead, A. N., and Bertrand Russell. 1927. Principia Mathematica, vol. I. Second edition. Cambridge: Cambridge University Press. (PM) Wittgenstein, Ludwig. 1958. Philosophical Investigations. Second edition. Edited by G. E. M. Anscombe and Rush Rhees. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (PI) ———. 1978. Remarks on the Foundations of Mathematics. Third edition. Edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell. (RFM) ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte, in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt am Main: Suhrkamp.

Addendum A Wittgenstein Typescript Edited by Nuno Venturinha, with an English Translation by James M. Thompson

In his catalogue of Wittgenstein’s literary remains, G. H. von Wright makes a joint comment on TSS 222–4, the ultimate sources for Part I of the posthumous Remarks on the Foundations of Mathematics. He explains that “[t]he reason for distinguishing these as three different items, although they all stem from 221, is that Wittgenstein himself separated 223 and 224 from the main body of cuttings comprising 222”.1 Immediately before, von Wright writes that “[o]f TS 221 two copies exist more or less intact”, but “[n]either of them is a top copy”, the reason being that “[t]he top copy was evidently used for composing 222”.2 In another study, von Wright expands this commentary saying that “[t]he cuttings he clipped together in a large number of bunches” and that “[t]here are numerous changes and annotations in handwriting on the cuttings”.3 Further pieces of information are given in the preface to the Remarks by von Wright, Rush Rhees and G. E. M. Anscombe, part of which deserves to be quoted at length: The last section of the rearranged collection [TS 222, 136–47 (256–66 in Wittgenstein’s pagination)] consisted of papers that had not been cut up, though there were many manuscript additions, and it is not quite clear whether Wittgenstein regarded them as belonging with the preceding text. This section deals with the concept of negation . . . . Its content occurs in great part in the “Investigations” §§547–568. The editors left it out of the first edition, but have included it here as Appendix I to Part I. The collection had two further appendices. They come from the same typescript as the second half of the (earlier) “Investigations”; nevertheless they were separated from the rest of the collection of clippings. The first [TS 224] deals with “mathematical surprise”. The second [TS 223] discusses among other things Gödel’s theories of the existence of unprovable but true propositions in the system of “Principia Mathematica”. In the first edition we included only the second appendix, but here both are published (Appendices II and III).4

178 Addendum It becomes clear from these passages that the editors of the Remarks were not sure about the insertion of the “appendices” in this work, the 1956 edition including only TS 223. Like pages 256–66 of TS 222, both TS 223 and TS 224 were clipped together, as can be seen from the marks on the paper, but similar marks also appear on other pages or cuttings in TS 222, a document that is described by von Wright, Rhees and Anscombe precisely as a “collection of clippings”. The particular place occupied by TSS 223 and 224 is therefore problematic. The first forms a continuous block from page 246 to page 255, none of these having been cut up. Some pages include corrections by the author but others are clean. TS 224, in its turn, is made of pages 205–7 and 213 as well as cuttings from pages 204, 208, 212 and 214–15, most of them with corrections. What is published here for the first time is an uncatalogued bunch made up of pages 237–9 and 244–5 as well as cuttings from pages 236, 240 and 243, which is housed at Trinity College Library, Cambridge.5 The fact that Wittgenstein corrected two mistakes on pages 237 and 244, not doing the same on the corresponding pages of TS 221(i and ii), and that he cut up pages 236 (doing away with the remainder of the remark begun on page 235), 240 (keeping only the remainder of the remark begun on page 239) and 243 (taking only the last remark which continues on page 244), gives to this document the same status as that of TSS 223 and 224.6 If we add to this that its last pages immediately precede those of TS 223, it is unclear why the editors of the Remarks have not included it as another appendix to the work. The reason for the fate of this fragment seems to have been its passing unnoticed or at least almost unnoticed. The copy of TS 221 published in the Bergen Electronic Edition contains on the left-hand side of the upper margin of page 236—the corresponding page from which the first cutting of our text was taken—this indication in German: “belongs to the original TS (GHvW)”.7 This typescript, which follows the pagination of TS 220, was at some time separately paginated by hand on the right-hand side of the upper margin of all pages with the exception of this one. As a matter of fact, whereas page 235 bears the number 100, page 237 bears the number 101. Documents kept at the von Wright and Wittgenstein Archives at the University of Helsinki help to explain this mystery.8 Two typed pages containing comments by von Wright on TS 221, dated 20 November 1986 and 11 January 1990, each give indications that page 236, then in Helsinki, should clearly form part of one of the two typescripts housed at Trinity College Library. The page must have been subsequently returned to Cambridge, but the question remains as to how it came to be separated from TS 221. The first of these sets of comments give us some preliminary hints. Von Wright reports the existence of an additional page in TS 221 numbered 248, comprising a total of four pages taking into account its appearance in the other copy of 221 and in TS 223. The collection of the Wittgenstein Papers at Trinity College Library actually holds an item with

Addendum

179

the reference 221(iii), which corresponds precisely to this single page.9 On the basis of a comparison between the pages and not without some hesitation, von Wright takes page 248 in TS 223 as not belonging, like the others, to the main copy of TS 221, even though he cannot explain its retyping. As the facsimile of item 221 in the Bergen Electronic Edition shows, the pages are definitely different. However, (a carbon copy of) the original page 248 is included in TS 223, not in 221, which features a clearly retyped page, copies of which are also found in items 221(i) and (iii). This can be perceived not only from the line breaks but also from the page number itself, “248” instead of “– 248 –”, the model we find in most of TS 221, in parts of TS 222, in TSS 223–4 and in our typescript. I say in most of TS 221 because from page 138, where it begins, to page 154 the pagination is by hand at the centre of the pages on the top, and from page 256 to the end, on page 271, we again find the numbers but without dashes, this being the result of different typing. But I also say in parts of TS 222 because this contains pages and cuttings of pages from these three series of typed texts: the first one without page numbers, these having been inserted like on the opening pages of TS 221; the second one with page numbers with dashes; and the third one with page numbers without dashes. What was published as Appendix I of the Remarks, pages 256–66, together with another block made of pages 267–71, represent the latter. In his genetic study on the Investigations, von Wright identifies those last pages with the production on another typewriter of a “second part of TS 221”, which includes extracts from MS 115 and MSS 121 and 162a, the latter dating from “the turn of the year 1938 to 1939”.10 But if this “second part” begins only on page 256, how to explain the triple existence of page 248 without the dashes? In their preface to the so-called Helsinki Edition of the prewar Investigations, von Wright and Heikki Nyman suggest that page 248 may actually have the same origin as pages 256–71, this also applying to one of the pages numbered 158.11 The facsimile of TS 221 included in the Bergen Electronic Edition does indeed contain three pages with this number, two with exactly the same content but with different line breaks, the second being identical with the one cut up for TS 222, and another with different content.12 Surprisingly, they are numbered by hand on the right-hand side of the upper margin 21–3. The conjecture about page 158 is also made in von Wright’s 1986 notes. Those from 1990 only indicate the duplication of the number on the pages with different content. In both sets of comments, von Wright also mentions that a number of pages from the first fraction of one of the copies of TS 221, numbered 142– 55, did not form part of the document he brought to Helsinki which he intended to examine. Von Wright informs us that this material was located among Anscombe’s papers, but only after a considerable period of time had elapsed. Though von Wright does not indicate when he brought the typescript from Cambridge, a cover sheet included with the Trinity folder containing TS 235—a list of 163 headings—gives us the answer.13 We are

180 Addendum informed that there were three items together with this list and an exemplar of “Remarks II” (TS 230) that von Wright took with him on 21 October 1951; namely, what we now have as TS 221, more specifically the fulllength copy to judge from its description as consisting of pages 138–271; a couple of incomplete exemplars of TS 221, nonspecified parts having suffered cuts, one of them probably lacking pages 142–55 and the other corresponding to TS 222 as clippings with handwritten insertions are mentioned; and plenty of additional Zettel and remarks written by hand related to the themes covered by TS 221. It would have been from these materials, gathered together just a few months after Wittgenstein’s death, that Part I and Appendix I of the first edition of Remarks on the Foundations of Mathematics were established some years later. It comes as no surprise that, like pages 142–55, other bits could have subsequently been discovered—and others rediscovered. Quite astonishingly, the document published as Appendix I to the 1956 edition of the Remarks, TS 223, was itself lost for some time. It appears in the original version of von Wright’s catalogue as “Missing”.14 The year of publication of the catalogue, 1969, was also the year when the papers in possession of the three trustees were transferred to Trinity College Library.15 Yet, as von Wright explains in his later version of the catalogue, much was still waiting to be found. He reports, for instance, that “[i]n 1976 and 1977 further typescripts came to light . . . listed here as items 235–245 of the catalogue”.16 One may wonder why TS 235 had not been listed before since it had already been described in October 1951, but more intriguing is where it went after that. In his 1986 notes, von Wright comments on another typescript, 240, whose two pages and one cutting also derive from TS 221, expressing the belief that they came from Anscombe. The fragment that is published here, however, came directly from Rhees. We learn from an undated note written in von Wright’s hand that the document was discovered by the archivist Jonathan Smith in the Rhees collection of Wittgenstein-related documents, which came to Trinity College Library shortly after Rhees’ death in 1989. This note makes reference to item 2 in box 2 of the collection, which is described in some detail by von Wright in a 15-page typescript entitled “The Rhees Material” and dated November 1990.17 We would not go very far with only this information, but fortunately there is more to be learned from von Wright’s archive. In his inventory, von Wright lists two duplicates made by Rhees of duplicates he himself had made from other copies of 220 and 221 which Wittgenstein possessed.18 The discrepancies von Wright specifically spots between Rhees’ copy of 220 and TS 220 are in line what von Wright writes elsewhere: It is known that Wittgenstein had a typescript with him in Norway in the autumn of 1937, which most probably was a typescript of what he had written one year earlier. But there is some evidence which points to it not being the typescript (TS 220) which has been preserved of the first

Addendum

181

half of the early version. I would therefore conjecture that the entire “early version”, that is, both TSS 220 and 221, were made at Cambridge in 1938—and that the concluding part of 221 was added some time after January 1939.19 In his biography of Wittgenstein, Ray Monk reports that in the summer of 1937 “[a]t Cambridge he undertook work that, presumably, Francis [Skinner] could help him with” and “he dictated a typescript of the remarks, written the previous winter, that now form the first 188 paragraphs of Philosophical Investigations”.20 Denis Paul takes this indication by Monk to mean “a typescript dictated to Skinner” that “has not survived” and shares with von Wright the view that TS 220 was produced in 1938.21 For Michael Nedo, however, what was dictated in the summer of 1937 was TS 220 itself.22 Joachim Schulte, in turn, proposes to date pages 1–65 to the end of 1936 and pages 66–137 to May of the following year.23 Another view is offered by Brian McGuinness, who claims that what was dictated in Cambridge in the summer of 1937 was only a new introduction to TS 220 (pages i–iii), for the first two parts of the typescript (pages [1]–45 and 46–6524) must have already been produced in Vienna in May, whereas the dictation of the third part of it (pages 66–13725) would have had to wait until the next visit to Vienna at the end of the year.26 It is possible that von Wright’s assumption is at least partly based on a handwritten letter he got from Rhees dated 11 December 1973, which is preserved with the other materials in Helsinki together with two typed letters dated 13 and 31 [sic] November 1973. It is in these letters that Rhees explains how he came to do the duplicates of what correspond to items 220 and 221. In the two earlier letters, Rhees seems to address only the first document, in connection with the revision made in one of its copies, which constitutes TS 239 (part of which Rhees also copied at the turn of the year 1942 to 1943, mentioning some pages and cuttings in his possession27), but it is possible that a unique text is envisaged due to the continuous pagination mentioned above which is retained in this copy. Incidentally, Rhees refers in the first letter to 1939 as the year of Wittgenstein’s loan and his subsequent duplication, but the second and the third letter agree with 1938 for the loan of both parts, that is TSS 220 and 221. The reproduction made by Rhees 35 years later and sent to von Wright not only includes but a part of the revisions we find in 220, but also contains some differences.28 It would be tempting to identify the Rhees text with that used for his 1938–9 incomplete translation of the prewar Investigations (TS 226), but this is not the case. There is evidence to assume that the translation was made from a subsequent stage of revision which can be detected in TS 239.29 What the Rhees typescript helps us do is to demarcate what Wittgenstein had revised up to that time. As for the replica of 221—besides the fact that Rhees has omitted Wittgenstein’s cancellations and included his handwritten corrections,30 which

182 Addendum show that the revision of the text as we have it must have been finished by then—the typescript stops at a particularly interesting point: the equivalent to page 237 of the original (whose last remark would continue on the next page). This is neither more nor less than the first complete page of our fragment. Even more interesting, the corresponding page 236, the same that came to be among von Wright’s papers and that Wittgenstein cut up for our bunch,31 was not typed by Rhees. As a matter of fact, there is a note in his hand at the end of the penultimate page of the typescript, which corresponds to page 235 of TS 221, manifesting some puzzlement at the lack of it. Another interesting detail is the copy on page 195 of a parenthetical indication in English from page 184 of the original typescript which reads “insertion from p 241 etc.”, with Rhees leaving the rest of the page blank. Although page 241 does not form part of our fragment, it belongs to the same thematic area. Looking for it in TS 222, we find a cutting consisting of the remainder of the last remark from the previous page, which forms another cutting.32 We also find another cutting from page 241 that contains its last three remarks,33 this one continuing on page 242, which is preserved in toto and has a continuation on page 243, a fragment of which, up to the end of the second complete remark, is also preserved.34 Note that what we have of page 243 in our bunch is the last remark,35 which continues on page 244, with the selection ending on the next page. It is easy to see that the “insertion from p 241 etc.” Wittgenstein refers to on page 184 of TS 221 constitutes the first of these two cuttings, viz. the one that follows the previous page. In fact, whereas the remark of 221 that precedes that indication is numbered in margin 281, the remark of the cuttings from pages 240–1 together with the remark of the cutting from page 222 bear the marginal numbers 281.1 and 281.2.36 But there is another parenthetical note by Wittgenstein in TS 221, which is absent from the Rhees text, that must be taken into consideration. It occurs on page 204 and runs as follows: “I have not got cuts from this point”. According to Schulte, this “[m]eans probably that no more cuttings from TS 221 existed” for, as he emphasizes, “[i]n TS 222 [pages 62–3] the numeration of the remarks stops just at this point (see remark 316)”.37 With the exception of a couple of corrections, TS 221 does not in fact contain further revisions from page 204 onwards. It is also worth remembering that TS 222 plus TSS 223–4 do not include cuttings from the pages following our fragment. Even if one excludes the possibility that the revision as we have it of pages 240–1 (and 222) of TS 222 occurred at that time, Wittgenstein must already have had it in mind since both the extensional numeration and the cuttings fit into the note inserted on page 184 of TS 221. Moreover, the circumstance that Rhees’ duplicate passes over page 236—again, the first that has been cut up for our fragment and of which an integral copy would later appear in Helsinki—and ends precisely on the next page suggests that our bunch may have been arranged soon after Wittgenstein’s loan of TS 221. Relying on Rhees’ assertion that

Addendum

183

this loan took place in 1938, as his December letter to von Wright indicates, we can also explain why the copy did not include material from the so-called “second part of TS 221”, which in all probability would have been finished only at the beginning of 1939.38 Taking into account too that Rhees’ translation in TS 226 is sometimes closer to TS 239 than to TS 220, and that another “Translation of Some Remarks of Wittgenstein’s Philosophical Investigations (preliminary version)” done by Yorick Smythies has TS 239 itself as its basis,39 von Wright seems to be absolutely right when he conjectures “that the cutting up of TS 221 into ‘Zettel’ took place in connection with a revision of the entire early version”.40 Yet, although von Wright reminds us that “[i]t is noteworthy that pages 120–135 [256–71 in the typescript pagination] were not cut up” and that “[t]his might indicate that the cutting up happened before these pages were even dictated”,41 he leans towards “the end of 1942 or beginning of 1943” for the establishment of TS 239 and 1944 for that of TS 222.42 Schulte also expresses the view that it was “sometime between 1939 and 1944” but, following von Wright, most probably at the “end of 1942 or beginning of 1943” that TS 239 was reworked, stating that “the second part of the early version would be subjected to an even more radical revision, cut up in several Zettel, many times modified and again reordered”.43 The date 1942–43 is based on Rhees’ information in his first November letter, but what it suggests is that the revision Wittgenstein undertook at that time was mostly from page 77 to page 93, as TSS 237 and 238 testify.44 With respect to TS 222, von Wright himself just mentions a couple of corrections that by no means indicate any decisive reworking other than what must have occurred when the cuttings were “again reordered”, as Schulte stresses.45 The pieces of evidence discussed here show that Wittgenstein’s revisions of both TSS 239 and 222–4 to some extent took place earlier than has commonly been assumed, including by myself.46 There can be little doubt that the survival of this fragmentary typescript presents us with a number of issues that help to cast some light on this particular phase of Wittgenstein’s work on the Investigations. Some riddles remain, but it is hoped that they can eventually be solved. As for the edition, I have limited myself to reproducing the text as closely to the original as possible. For the sake of simplicity, words spaced out in the typescript are given in italics, words typewritten or handwritten above the linear ones are shown as such with Wittgenstein’s ink corrections appearing in bold, and words underlined with dashes are simply underlined. Intralinear variants marked by double bars are given within single bars. I have numbered the remarks using square brackets. I have not indicated the manuscript sources of the remarks nor their variations from the typescript as this has already been done by Schulte in his edition of the full TS 221.47 The material comes from MSS 118 and 119 and was written in a short period of time, between 17 and 28 September 1937. This also attests to the unity of the work.48

184 Addendum REFERENCES Hacker, P. M. S., and Joachim Schulte. 2009. “The Text of the Philosophische Untersuchungen”. In Ludwig Wittgenstein, Philosophische Untersuchungen / Philosophical Investigations, edited by P. M. S. Hacker and Joachim Schulte, translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte, xiii–xxiii. Oxford: Wiley-Blackwell. McGuinness, Brian. 2002. “Manuscripts and Works in the 1930s”. In Approaches to Wittgenstein: Collected Papers, 270–86. London: Routledge. ———, ed. 2008. Wittgenstein in Cambridge: Letters and Documents 1911–1951. Oxford: Blackwell Publishing. Monk, Ray. 1990. Ludwig Wittgenstein: The Duty of Genius. London: Jonathan Cape. Nedo, Michael. 1993. Einführung / Introduction. In Ludwig Wittgenstein, Wiener Ausgabe / Vienna Edition. Vienna: Springer. Paul, Denis. 2007. Wittgenstein’s Progress 1929–1951. Edited by Alois Pichler. Bergen: Publications from the Wittgenstein Archives at the University of Bergen. Schulte, Joachim. 2001a. “Einleitung”. In Ludwig Wittgenstein, Philosophische Untersuchungen: Kritisch-genetische Edition, edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright, 12–50. Frankfurt am Main: Suhrkamp. ———. 2001b. “Historisch-philologische Nachbemerkungen”. In Ludwig Wittgenstein, Philosophische Untersuchungen: Kritisch-genetische Edition, edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright, 1089–112. Frankfurt am Main: Suhrkamp. Venturinha, Nuno. 2010a. “A Re-Evaluation of the Philosophical Investigations”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 143–56. Basingstoke: Palgrave Macmillan. ———, ed. 2010b. “Wittgenstein’s 1938 Preface”. In Wittgenstein After His Nachlass, edited by Nuno Venturinha, 182–8. Basingstoke: Palgrave Macmillan. von Wright, G. H. 1969. “The Wittgenstein Papers”. The Philosophical Review 78, no. 4: 483–503. ———. 1982. “The Origin and Composition of the Philosophical Investigations”. In Wittgenstein, 111–36. Oxford: Basil Blackwell. ———. 1993. “The Wittgenstein Papers”. In Ludwig Wittgenstein: Philosophical Occasions 1912–1951, edited by James C. Klagge and Alfred Nordmann, 480– 506. Indianapolis: Hackett. ———. 1997. “Wittgenstein Materials kept with the Department of Philosophy, Helsinki University”. Unpublished. http://www.helsinki.fi/wwa/Wittgenstein%20 Materials.pdf. von Wright, G. H., and Heikki Nyman. 1979. “Vorwort der Herausgeber”. In Ludwig Wittgenstein, Philosophische Untersuchungen: Frühversion 1937–1938. TS 225, TS 220, TS 221 mit MS 141 als Anhang, edited by G. H. von Wright and Heikki Nyman, 1–9. Unpublished. von Wright, G. H., Rush Rhees and G. E. M. Anscombe. 1978. “Editors’ Preface to the Revised Edition”. In Ludwig Wittgenstein, Remarks on the Foundations of Mathematics, third edition, edited by G. H. von Wright, Rush Rhees and G. E. M. Anscombe, translated by G. E. M. Anscombe, 29–33. Oxford: Basil Blackwell. Wittgenstein, Ludwig. 1979. Philosophische Untersuchungen: Frühversion 1937– 1938. TS 225, TS 220, TS 221 mit MS 141 als Anhang. Edited by G. H. von Wright and Heikki Nyman. Unpublished.

Addendum

185

———. 2000. Wittgenstein’s Nachlass: The Bergen Electronic Edition. Edited by the Wittgenstein Archives at the University of Bergen. Oxford: Oxford University Press. (MS & TS) ———. 2001. Philosophische Untersuchungen: Kritisch-genetische Edition. Edited by Joachim Schulte in collaboration with Heikki Nyman, Eike von Savigny and G. H. von Wright. Frankfurt am Main: Suhrkamp. ———. 2004. Gesamtbriefwechsel: Innsbrucker elektronische Ausgabe. Edited by Monika Seekircher, Brian McGuinness and Anton Unterkircher. Charlottesville: InteLex. ———. 2009. Philosophical Investigations. Edited by P. M. S. Hacker and Joachim Schulte. Translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell.

186 Addendum [1.] Wann sagen wir denn: der Vollzylinder passt in den Hohlzylinder? Da gibt es viele verschiedene Fälle; aber ein wichtiger ist der: wir stecken sie zusammen, probieren, ob sie passen. Passen sie dann, so sagen wir, sie passen; d.h., auch dann,1 wenn sie wieder getrennt sind—nämlich unter bestimmten Bedingungen. Probieren wir nun wieder, und sie passen nicht,—wann, sollen wir sagen, haben sie zu passen aufgehört.2 Diese Frage wird manchmal so beantwortet: der Zeitpunkt der [Ä]nderung3 sei der einer anderen [Ä]nderung4 (als wir den Zylinder erhitzten, da hat er aufgehört, zu passen). Wenn wir aber kein solches Kriterium für diesen Zeitpunkt haben; wenn wir—sozusagen—gar[] nicht5 wissen, was in dem Intervall6 zwischen den Proben mit den Dingen geschieht?7: passen sie nun8, oder passen sie nicht.9 [2.] Man sagt: “Es ist schwer zu wissen, ob diese Medizin wirklich hilft,10 oder nicht, weil man nicht weiss, ob der Schnupfen länger gedauert hätte,11 oder ärger gewesen wäre, wenn man sie nicht genommen hätte.” Wenn man dafür wirklich keinen Anhaltspunkt hat, ist es dann bloss schwer zu wissen? Denke, ich hätte eine Medizin erfunden; ich sage: diese Medizin, einige Monate hindurch genommen, verlängert das Leben jedes Menschen um einen Monat. Hätte er sie nicht genommen, so wäre er einen Monat früher gestorben. “Man kann nicht wissen, ob es wirklich die Medizin war; ob er nicht ohne sie ebenso lang gelebt hätte.”—Ist diese Ausdrucksweise nicht irreführend? Sollte es nicht besser heissen: “Es heisst nichts, von dieser des Satzes Medizin zu sagen. sie verlängere das Leben; wenn eine Prüfung der Behaupist tung in dieser Weise ausgeschlossen wurde.” Nämlich: wir haben hier zwar einen richtigen deutschen Satz nach Analogie oft gebrauchter Sätze gebildet, aber Du bist Dir nicht klar über den grundlegenden Unterschied in den Verwendungen dieser Sätze. Diese zu überblicken, ist nicht leicht. Der Satz liegt Dir vor Augen, aber nicht eine übersichtliche Darstellung der Verwendung. Mit “Es heisst nichts. . . .” will ich also sagen /will also gesagt werden/: dies sind Worte, die Dich irreführen, sie spiegeln einen Gebrauch vor, den sie nicht haben. Sie rufen wohl auch12 eine Vorstellung hervor (der Verlängerung des Lebens, etc.), aber das Spiel mit dem Satz ist so eingerichtet, dass es die wesentliche[n]13 Pointe nicht hat, die dem Spiel mit ähnlich gebauten Sätzen seinen keinen14 Nutzen gibt. (Wie der ‘Wettlauf’ zwischen dem Hasen und dem Igel’ zwar aussieht wie ein Wettlauf, aber keiner ist.) [3.] Du musst Dich fragen: was nimmt man als Kriterium dafür, dass eine Medizin geholfen hat? Es gibt verschiedene Fälle. In welchen Fällen sagt man: “Es ist schwer zu sagen, ob sie geholfen hat”. In welchen Fällen ist die Redeweise als sinnlos zu verwerfen: “Man kann natürlich nie sicher sein, ob es die Medizin war, die geholfen hat”. [4.] Wann nennen wir zwei Körper gleich schwer? Wenn wir sie gewogen haben, oder nur während wir sie wägen?

Addendum

187

[1.] When do we ever say: the solid cylinder fits in the hollow cylinder? Here, there are many different cases; but an important one is this: we put them together, test whether they fit. If they fit, then we say they fit; that is to say, even then when they are separated again—namely under specific circumstances. If we now try again, and they do not fit,—when should we say that they have stopped fitting. This question is sometimes answered thusly: the point in time when the change took place was another change (when we heated the cylinder, then it stopped fitting). If we do not have such a criteria for this point in time; if we—so to speak—do not know what happened in the interval between the two tests with the things: do they fit now or do they not fit. [2.] One says: “It is difficult to know whether this medicine actually helps or not, because one does not know whether the runny nose would have lasted longer or would have been worse, if one had not taken it.” If one really does not have a point of reference, is it then simply hard to know? Think if I had invented a medicine; I say: this medicine, taken for several months, increases the life of anyone by one month. If he had not taken it, he would have died one month earlier. “One cannot know whether it was actually the medicine; whether without it he would have lived just as long.”—Is this expression not misleading? Would it not be better to say: “It doesn’t of the sentence mean anything to say that the medicine extends life when a test of the claim is was in this way closed off.” Namely: we have here indeed constructed a proper English sentence according to the analogy of often used sentences, however, you are not clear about the fundamental difference in the use of these sentences. To get an overview of them is not easy. The sentence lies before your eyes, but not a perspicuous representation of the use. By “It doesn’t mean anything. . . .” I want to say /wants to be said/: these are words which are misleading you; they suggest a use, which they do not have. They call forth an idea (the extension of life, etc.), but the game with the sentence is arranged such that it does not possess the essential points, its which lends the game with similarly constructed sentences no utility. (Like the ‘race between the rabbit and the hedgehog’ indeed looks like a race but isn’t one.) [3.] You must ask yourself: what does one count as a criterion that a medicine helped? There are various cases. In what cases does one say: “It is difficult to say whether it helped”. In what cases is this way of speaking to be rejected as senseless: “One can of course never be certain whether it was the medicine that helped”. [4.] When do we say that two bodies are equally heavy? When we have weighed them or only while we are weighing them?

188 Addendum Wenn Wägen das einzige Kriterium für das Gewicht wäre,—wann hat nun ein Körper sein Gewicht geändert, wenn er bei einer Wägung mehr wiegt, als bei der vorhergehenden? Der Sprachgebrauch könnte so sein: der Körper hat das und das Gewicht, bis er beim Wägen ein anderes zeigt; auf die Frage: “wann hat er sein Gewicht geändert?” gibt man den Zeitpunkt dieser Wägung an.—Oder: man sagt: “Man kann nicht wissen, wann er sein Gewicht ändert, wir wissen nur: bei der ersten Wägung hatte er dieses, bei der zweiten jenes Gewicht.”—Oder: “Es ist sinnlos, zu fragen, wann er sein Gewicht geändert hat; man kann nur fragen, wann sich die Gewichtsänderung gezeigt hat”. [5.] “Aber der Körper hatte doch zu jeder Zeit irgend ein Gewicht, also war doch die Antwort die richtige: wir wüssten nicht, wann er es geändert habe.”—Und wie, wenn wir sagten, ein Körper habe gar kein Gewicht, ausser dann, wenn es sich irgendwie zeigt, oder, er habe kein bestimmtes Gewicht, ausser, wenn es gemessen wird? Könnten wir nicht auch dieses Spiel spielen? Denke, wir verkaufen ein Material ‘nach dem Gewicht’ und das Herkommen ist so: Wir wägen das Material alle fünf Minuten und berechnen dann den Preis nach dem Resultat der letzten Wägung. Oder ein anderes Herkommen: Wir berechnen den Preis auf diese Weise nur, wenn das Gewicht bei der Wägung nach dem Kauf das gleiche ist, hat es sich dann geändert, so berechnen wir den Preis nach dem arithmetischen Mittel der beiden Gewichte. Welche Art der Preisbestimmung ist die richtigere?— (Wenn sich der Preis einer Ware von gestern auf heute geändert hat, wann hat er sich geändert? Wie hoch stand er um zwölf Uhr Mitternacht, als niemand kaufte?) Resultat: Die Verbindung der Ausdrücke: “der Körper hat jetzt das Gewicht. . . .”, “der Körper wiegt jetzt ungefähr. . . .”, “ich weiss nicht, wieviel er jetzt wiegt”,—mit den Ergebnissen der Wägung /mit dem Ergebnis einer Wägung/ ist keine ganz einfache, hängt von diversen Umständen ab, wir können mit der Wägung, und also mit diesen Sätzen, uns leicht verschiedene Spiele gespielt denken./ /wir können uns leicht verschiedene Rollen denken, die die Wägung in den Verrichtungen des15 Lebens spielen könnte. und also verschiedene Rollen für die Ausdrücke, die das Wägspiel begleichten./ Und das Gleiche gilt von der Rolle des Wortes “passen” in unsern Sprachspielen. [6.] Wir sagen: “es passt, ich habe es probiert”, nicht nur: “es hat gepasst, ich habe es probiert.”. Und ebenso: “er wiegt 50 kg, ich habe ihn gewogen,”, “16er ist 1 m lang, ich habe ihn gemessen”, und auch: “ich kann es, ich habe es probiert.” Wir sagen: “der Schuh passt”, auch wenn wir ihn nicht anhaben.

Addendum

189

If weighing were the only criterion for weight,—when did the body then change its weight, if at the weighing it weighs more than the previous one? The use of language could thus be: the body has this and this weight until during a weighing it shows something else; regarding the question: “when did it change its weight?” one gives the time of the weighing.—Or: one says: “One cannot know when it changed its weight, we only know: during the first weighing it had this [weight], during the second this weight.”—Or: “it is senseless to ask when did it change its weight; one can only ask when did the change of weight show itself”. [5.] “But the body clearly has a weight at any given moment, thus the right answer was indeed: we did not know when it changed.”—And what about when we say, a body has no weight except when it somehow shows or it does not have a specific weight, except when it is measured? Could we not also play this game? Think [for example], we sell a material ‘according to its weight’ and the procedure would be: We weigh the material every five minutes and then calculate the price according to the result of the last weighing. Or another procedure: We calculate the price in this way only if the weight is the same at the weighing after the sale; if it has changed, then we calculate the price according to the arithmetic middle of both weights. Which kind of price determination is the more correct?— (If the price of the ware changed from yesterday to today, when did it change? How high was it [the price] at twelve midnight when no one was buying?) Result: The connection of the expressions: “the body now has this weight. . . .”, “the body weighs now roughly. . . .”, “I don’t know how much it now weighs”,—has to do with the results of the weighing /with the result of a weighing/ is not very easy, depends upon diverse circumstances; with the weighing and, thus, with these sentences, we could easily think of various games played. /we could easily think of different roles which weighing could play in carrying out life and thus different roles for the expressions that accompany the weighing-game./ And the same applies for the role of the word “to fit” in our language-game. [6.] We say: “it fits, I have tried it”, not just: “it did fit, I have tried it”. And likewise: “he weighs 50 kg, I have weighed him”, “he is 1m long, I have measured him”, and also: “I can do it, I have tried it”. We say: “the shoe fits”, even when we do not have it on.

190 Addendum [7.] Was für einen Sinn hätte es, anzunehmen, dass ein Stück Stahl, wann immer man seine Festigkeit gerade nicht prüft, /gerade auf keine Weise beansprucht,/ es seine Festigkeit verliert, oder sie sich ändert. Das hängt mit der Idee zusammen, dass die Körper um uns nur solange existieren, als sie wahrgenommen werden. Das ist wirklich: das Zifferblatt mit dem Zeiger kuppeln; denn hier kuppelt man in der Grammatik die Aussage eines Tatbestandes mit der Bedingung der Nichtkontrollierbarkeit. Man hat die d Annahme dadurch zu einem leerlaufenden Rat17 der Sprache gemacht; und sie stört nun den Mechanismus der Sprache jedenfalls nicht. [8.] Hat es einen Sinn, zu fragen: “Hat dieses Stück Eisen nur dann diese Festigkeit, oder Elastizität, wenn sie geprüft wird, oder auch sonst?” Dies ist doch eine gut deutsche Frage! Und, dass wir empfinden: “das heisst ja nichts!”—weist uns darauf hin, den eigentlichen Gebrauch des Ausdrucks “eine Festigkeit haben” aufzusuchen, seinen Zusammenhang mit den Erfahrungen, die uns die Festigkeit zeigen. (Von so einer Frage sagen wir: ‘sie treibe das Problem auf die Spitze’.) [9.] Vergleiche denn ‘Zustand’, diese Festigkeit zu haben, mit dem ‘Zustand’, diese Farbe zu haben. Und wieder mit dem ‘Zustand’, diese Länge zu haben: a) die gesehene, b) die gemessene.—Und nun mit dem ‘Zustand’, einer bestimmten Fähigkeit: z.B., der Fähigkeit dieser Feder, diesen Druck auszuüben; meiner Fähigkeit, dies Gewicht zu stemmen; meiner Fähigkeit dies Gedicht aufzusagen; meiner Fähigkeit diese Reihe fortzusetzen. [10.] Unter welchen Umständen sagt man: “x ist in diesem Zustande”. D.h., wie, unter welchen Umständen, braucht man hier die Gegenwart des Verbums; unter welchen, die andern Zeitformen? Es gibt Umstände, unter denen man sagt: “ich kann es”, während man es nicht gerade tut.—Andere, unter denen man in so einem Fall sagt: “ich glaube, ich kann es”. Es gibt auch Fälle, in welchen man nur dann sagt: “ich kann es”, wenn man es gerade tut. [11.] Was sind unsere Kriterien dafür, ‘dass wir etwas können’? Dass wir es früher getan haben; dass wir etwas anderes (etwa ‘Schwereres’) früher getan haben; dass wir es jetzt tun; dass wir jetzt gerade etwas getan haben, was als Probe der Fähigkeit gilt. (Sich das Gedicht leise vorsagen, als Probe dafür, dass man es laut aufsagen kann.)

Addendum

191

[7.] What sense would it make to assume that a piece of steel, whenever one is not testing its strength, /just [when it] is in no way stressed,/ it loses its strength or it has changed. This hangs together with the idea that the bodies around us only exist as long as they are being perceived. That really means: the combining of the watch face with the hands; for here one combines in grammar the statements of a factual case with the conditions of uncontrollability. One has thereby turned the assumption into a wheel of language running in neutral; and it does not at all disrupt the mechanism of language. [8.] Does it make sense to ask: “Does this piece of iron only have this rigidity or elasticity when it is tested or also otherwise?” This is clearly a good German question! And that we feel: “that does not mean anything!”— reminds us to look for the actual use of the expression “to have a rigidity”, its connection with the experiences that show us the rigidity. (About such a question we say: ‘you are taking the problem to the extreme’.) [9.] Compare the ‘state’, to have this rigidity, with the ‘state’, to have this colour. And again with the ‘state’, to have this length: a) the seen, b) the measured.—And now with the ‘state’ a specific ability: e.g., the ability to apply this pressure to this spring; my ability to lift this weight; my ability to recite this poem; my ability to continue this series. [10.] Under what conditions does one say: “x is in this state”. This means, how, under what conditions, does one use here the present tense of the verb; under what [conditions] other forms of time? There are circumstances under which one says: “I can do it”, while one is not at that moment doing it.—Others under which one says in such a case: “I believe I can do it”. There are also cases in which one only says: “I can do it”, when one is at that moment doing it. [11.] What are the criteria for ‘that we can do something’? That we did it earlier; that we did something else (something ‘more difficult’) earlier; that we are doing it now; that we just did something that counts as a test of the ability. (To recite the poem quietly so that one can say it out loud.)

This page intentionally left blank

Notes

NOTES TO THE INTRODUCTION 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16. 17. 18.

Paul 2007, 23. Ibid., 24. Förster 1993, xxiv–v. See in this regard Montinari 1999. The “Kommentar zu Band 6” in this volume of Nietzsche’s works, focusing on the 1885–8 papers, is also worth consulting. Babich 2007, 38. Ibid., 39. Ibid., 40. A discussion of the unfinished state of the Investigations, namely the relationship between Parts I and II, can be found in Venturinha 2010a. Cf. Wittgenstein 2009. The most apparent oddity in the text of Part I is possibly the repetition of almost the same remark in §§386b and 594c. von Wright 1982a, 33. Drury 1996, 92–3. It should be noted that Drury does not recall any conversation in which Wittgenstein mentioned Pascal. Ibid., 93. Pascal 1977. A new version, including Pascal’s variants, was included in Le Guern’s edition of the Œuvres complètes (Pascal 2000). Cf. Le Guern 2000a, which informs the preceding account. Detailed information about the surviving manuscripts and previous editions, as well as about the rationale of this one, is given in his 2000b. Pascal 2000, fr. 241; I have slightly amended C. Kegan Paul’s translation of the text (Pascal 1885, 167). One of these amendments is the rendering of “physionomie” literally by “physiognomy”, which will be a key concept for Wittgenstein. See also Pascal 1885, 158, for the translation of fr. 455, which is directly related to 241. It is very interesting that the recently published diary notes by Wittgenstein’s fellow captive Ludwig Hänsel written in Monte Cassino include references to Pascal’s Provincial Letters, in particular to Letter XVIII, where the same topic is dealt with (cf. Somavilla 2012, 70 ff., esp. 73). Pascal 2000, fr. 472; following Le Guern’s edition, I have replaced “Scepticism” with “Prin.” in Kegan Paul’s translation (Pascal 1885, 108). Hacker and Schulte 2009. Allusions to Wittgenstein’s Nachlass are according to von Wright 1993. Almost all the items that make up this later version of the catalogue are published in Wittgenstein 2000.

194

Notes

19. At the end of MS 115, 292, in the second half of which Wittgenstein drafted this “Umarbeitung”, he wrote: “this entire ‘attempt at a revision’ from page 118 up to here is NOT WORTH ANYTHING”. (I use the translation given in von Wright 1993, 493.) This work was published in Wittgenstein 1984. 20. Monk 1990, 346. 21. The documents referred to belong to the collection of the G. E. Moore Papers housed at the Cambridge University Library, described by Kathleen Cann in her 1995 catalogue, and bear the references 8330–8K/1/1a, 8330–8K/1/1 and 8330–8K/3/13, respectively. I would like to thank the library staff for their kindness when I consulted these materials in 2010. To the best of my knowledge, Brian McGuinness is alone in mentioning this possible publication by the British Academy, making reference to Keynes’ letter. However, McGuinness does not conjecture the Brown Book, reporting only “the fact that Wittgenstein in the spring of 1935 had been discussing with Keynes his plans of publishing the book on which he was then working” (McGuinness 2008, 244, note). 22. For a detailed account, see von Wright 1982b, 120–1. 23. A thorough study of this matter, in connection with the edition of the English version of the preface—not catalogued by von Wright—can be found in Venturinha 2010b. Early plans for a translation are evidenced by a letter from Nicholas Bachtin to Wittgenstein of 22 November 1936, in which he says: “I am very glad you are well and work is going on—and very impatient too, to begin my attempts at translating it” (McGuinness 2008, 258). I have tried to locate these “attempts” at the Cadbury Research Library at the University of Birmingham, where the Bachtin Papers are kept, but without success. My thanks to the library staff for their kindness when I researched the collection in 2010. 24. Hacker and Schulte 2009, xix. See also the Addendum to this volume. 25. Hacker and Schulte 2009, xix. 26. See apropos of this von Wright 1993, 499, as well as 1982b, 118. See also Venturinha 2010a, 153, note 7, and the Addendum to this volume. 27. Hacker and Schulte 2009, xx. They concur with von Wright that the rereading of the Tractatus referred to in the preface to the Investigations which motivated this plan of publication was made with Bachtin (von Wright 1982b, 121; 1993, 496, note 14). Among the Bachtin Papers there are no traces of such conversations, with the possible exception of Bachtin’s numerous notes on Plato’s Cratylus and the origin of language, a topic that may well have been discussed with Wittgenstein in connection with his rereading of the Tractatus. Some notes date from January 1943. 28. See note 8 above. 29. Hacker and Schulte 2009, xxiii. 30. Hallett 1977. 31. Baker and Hacker 2005a, 2005b and 2009; Hacker 1990 and 1996. 32. Savigny 1994 and 1996. 33. Hacker and Schulte 2009, xx-xxi. 34. von Wright 1982b, 133. 35. Ibid. This “independent and final work” would, however, dispense with roughly the first 200 sections of the Investigations we know, arguably the most important part. On the other hand, it includes additional sections on mathematical topics such as “contradiction” (§§133–40) and Hardy’s view that “the finite cannot understand the infinite” (§363). As for the remarks on “contradiction”, which are prefaced by considerations on Russell’s theory of types and rule-following (§§131–2), all deriving from MS 130, 11–15, it is

Notes

36.

37.

38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.

195

of some interest that cuttings from a copy of TS 228 inserted between pages 88 and 89 of TS 227b contain the equivalent to §§140, 132–3 and 137, these forming now, following an instruction in English on page 89, §125 of the Investigations. Cuttings from all the remainder are to be found in TS 233b, 60, with the remark on Hardy being found in TS 233a, 56. They were published in Wittgenstein 1981, §§273, 685–9 and 692. Paul 2009, 290. Paul’s allusion to “different copies” of these two typescripts is something of a mystery since, in addition to those included in the Bergen Electronic Edition, only two other versions of TS 230 seem to have survived. It is noteworthy that one of them was under Moore’s custody when Wittgenstein died. Together with TS 209, it explains his reference to “my type-scripts” in a letter to Moore of 16 December 1948 (see McGuinness 2008, 435; I thank Gabriel Citron for reminding me of this letter). The two further versions of “Remarks II” are alluded to by von Wright (1982b, 133), but they were not included in his 1993 catalogue. They nonetheless appear in his personal catalogue (von Wright 1997, 18). The originals are all at Trinity College Library, Cambridge. They reveal interesting variations; the most remarkable are perhaps the insertion in TS 230(iii) of an unnumbered typed page between pages 36 and 37, which would be replaced by page 36a in TSS 230(i) and (ii), and the absence in TS 230(iii) of page 128a of TSS 230(i) and (ii)—the former of these being the typescript published in the Bergen Electronic Edition. I am grateful to Trinity’s archivist Jonathan Smith for his kindness when I examined these typescripts in detail in 2010. Wittgenstein 2009, with the emphasis added according to the German original. In a recent paper, Conant (2011) discusses this remark at length, seeing it as the motto of a fundamental “turning point” (Kehre) which occurred in 1937. Conant does not forget that its original version dates from 23 February 1938 (in MS 120, 85r), but he passes over this telling occurrence in TS 230 in silence. As I have indicated elsewhere (Venturinha 2011, 11), the fact that Wittgenstein has replaced “but there are only methods” with “though there are indeed methods” (sondern nur wohl aber gibt es Methoden) in that manuscript is also revealing of his effort to avoid yet again a unified perspective. Kierkegaard 1992, 17. See, for example, Wittgenstein 1998, 11 (MS 110, 10). Cf. Kierkegaard 1992, 53–5. For Wittgenstein’s own evaluation of the Socratic dialogues, see Wittgenstein 1998, 21 (MS 111, 55). See also Wittgenstein 1998, 64 (MS 133, 95r-95v). For a detailed study of other techniques, see Baker 2004. Wittgenstein 2009, 3. Wittgenstein 1998, 33 (MS 118, 94v). See also a remark of 6 March 1937 in Wittgenstein 2003, 219 (MS 183, 210–11). Wittgenstein 1998, 90 (MS 137, 141a). Wittgenstein 2009, 3–4. Wittgenstein 1998, 14 (MS 153a, 90v). See in addition a note dated 24 November 1946 in Wittgenstein 1998, 64 (MS 133, 42r). Wittgenstein 1998, 95 (MS 138, 31a). Cf. Guest 2003, 363 ff. Wittgenstein 1998, 12 (MS 110, 18). See Wittgenstein 2009, 4. See MS 116, 340–1, the date appearing on page 316. See also TS 228, 185, §§697–8. See TS 235, 7.

196

Notes

54. Ibid.: my translation. “(Theology)” was a later addition. See MS 130, 9–11, where these ideas are developed after a direct reference to the theologian Karl Barth. This takes place a few pages before Wittgenstein sketches the list of headings, with the notes in MS 130 immediately following on page 11 containing the mathematical remarks mentioned in note 35. 55. Wittgenstein 1981, §144 (TS 233a, 30). This remark originally comes from MS 129, 189, having made its way into TS 228, 100, §354 (a copy of which is the source of the cutting in TS 233a), as well as into TS 230, 40, §153. 56. Wittgenstein 1981, §717 (TS 233b, v). It should be mentioned that the placement of this remark does not follow the arrangement in the typescript made for the posthumous publication. The remark originally derives from MS 130, 7, a copy of §568 of TS 228, 158, having been used for TS 233b, and corresponds to §340 of TS 230, 92. See in addition §589 of the Investigations, which stems from the opening remark of MS 130, 1, having been incorporated in TS 228, 157, §563, and in TS 230, 135, §484, as well as a later note, again with a reference to Barth, in MS 173, 92v–93r, included in both Wittgenstein 1998, 97, and Wittgenstein 1977, III, §317. 57. It was Rhees who labelled the text in this way (Wittgenstein 1969). 58. See, in particular, de Biasi 2004, 62–3. 59. Wittgenstein 1969, 17–18 (TS 309, 27–8). 60. See MS 114, 45. The title “Umarbeitung” appears on page 31v, or 1 in Wittgenstein’s pagination, the phrase “Second revision on large sheets” (Zweite Umarbeitung im grossen Format), a reference to MS 140, having been added later. As von Wright observes, “[t]he second part of 114 . . . may be regarded as of one piece with the first part (pp. 1–117) of 115 [see note 19 above] and with MS 140 on large sheets (‘Grosses Format’)” (1993, 493). He writes a bit further on: “The revisions and ‘jumps’ back and forth between the manuscripts mean that the reading of them is not easy. From the point of view of their content, however, they form a close unity. It is clear that Wittgenstein is here attempting to write a book, to give a consecutive and coherent statement of his philosophical position at the time. It is a plausible conjecture that this piece of writing dates from the academic year 1933–34 and that it is at least partly contemporary with the dictation of the so-called Blue Book”. The archaeology of §73 allows us to corroborate this “conjecture”. 61. This is clear from the aforementioned letter from Keynes to Moore of 6 March 1935 (see note 21 above).Wittgenstein’s intentions are straightforwardly announced in a letter to Keynes dated 30 June 1935: “to go to Russia . . . in September . . . and . . . to get a suitable job there” or “to return to England and if possible study Medicine” (McGuinness 2008, 244). Wittgenstein would return to England but to continue his work on philosophy. See a postcard sent to Moore from Russia on 18 September 1935 (ibid., 249). It is relevant that, as can be seen from two undated letters, what he did at the beginning of the new academic year was to send a copy of the Blue Book to Bertrand Russell and then to demonstrate his satisfaction at his “reading [his] M.S.” (cf. ibid., 250 and 252). Another copy of the Blue Book was sent to W. H. Watson together with a letter of 19 October 1935 in which Wittgenstein again says that his “intention is now to have something publishable ready by the end of this academic year”, with his following plans eventually being “to study medicine” (cf. ibid., 251). 62. See MS 130, 22, a manuscript firstly dated, on page 147, 26 May 1946, and Wittgenstein 2009, 4. 63. See, principally, MS 146, 50; MS 115, 30; and MS 120, 145r.

Notes

197

64. See notes 19 and 60 above. 65. McGuinness 2002, 284. 66. Cf. MS 159, 34; MS 117, 110, 116 and 120; TS 225, 1; and Venturinha 2010b, 187. Cf. also MS 128, 42; MS 129, 1v, 5r and 8v; TS 243, 1; and Wittgenstein 2009, 3, the latter reproducing the preface of TS 227. 67. Cf. Wittgenstein 2009, §415, this notion appearing in various sources. 68. Cf. Wittgenstein 2003, 89 (MS 183, 81). 69. Cf. MS 157b, 3v. 70. See von Wright 1982b, 130, and Maury 1994, 351–68, an earlier version of the latter already being found in Maury 1981. 71. See the edition of TS 221 in Wittgenstein 2001, 329–446. 72. See von Wright, Rhees and Anscombe 1978, 31–2. See additionally Maury 1994, 355–9, and Biggs and Pichler 1993, 19. 73. Some paragraphs of this Introduction have appeared in an earlier form in my “Wittgenstein, Pascal e o Livro Impossível”, in Filosofia e Literatura 1, edited by Humberto Brito (Lisbon: IFL, 2010), 41–55. I am grateful to the participants at the 1st Luso-Brazilian Colloquium of Wittgensteinian Studies, which took place at the Institute for Philosophy of Language at the New University of Lisbon in 2012, for their comments on a test-tube version of the first half of the text.

NOTES TO CHAPTER ONE 1. Drury 1984a, 89. A candid expression of Moore’s views on religion and belief in God can be found in his letter to Norman Malcolm, responding to news of Malcolm’s conversion to Christianity: “It seems to me that no rational person could believe that there is a God . . .” (Moore and Malcolm 2003, 286 [letter of 20 May 1957]). 2. Drury 1984a, 89. 3. Ibid., 90. 4. I owe much of the material in the first part of this section to Brian Rogers and David Stern, with whom I am editing Moore’s lecture notes. 5. See PPO, 340, and Monk 1991, 289. However, in a letter to John Wisdom—written in 1947—Wittgenstein recalled that he “started lecturing in 1929, though then without any University post” (Wittgenstein to Wisdom, 28.3.47; in McGuinness 2008, 408). I do not know which lectures he is referring to; perhaps he was thinking of his one-off “Lecture on Ethics” (PO, 36–44), which was given in November 1929. 6. LWL. 7. AWL, 1–40. 8. In fact, because Moore’s notes are not written in full sentences, they actually cover even more extra ground than indicated by this simple word-count comparison. 9. King’s and Lee’s original notes can be consulted at the Wren Library in Trinity College, Cambridge. Ambrose’s original notes seem to have been sold into private hands. 10. Britton 1999, 205. 11. Wittgenstein to W. H. Watson, 4.11.32, in McGuinness 2008, 203. 12. Malcolm 2001, 48. 13. Lewy 1976, xi, note 1. 14. Braithwaite 1933.

198

Notes

15. Drury 1984b, 141. This took place in Wittgenstein’s 1938 Lectures on Aesthetics. 16. Hacker 1997, 437. 17. Moore did not write in full sentences, and when I quote from his notes I have left them as he wrote them. He also made very frequent paragraph breaks (sometimes almost every sentence), which I have indicated by forward slashes. 18. For example, in a diary entry from 1937, Wittgenstein wrote: “I read somewhere, Luther had written that theology is the ‘grammar of the word of God’, of the holy scripture” (PPO, 211). Wittgenstein’s gloss of the phrase “the word of God” as “holy scripture” shows that he must have meant the word of God specifically, rather than merely the word “God”. Furthermore, the diary entry indicates that Wittgenstein read Luther’s alleged comment in a secondary source, and I have found no more likely source than a letter written by J. G. Hamann—one of Wittgenstein’s favourite authors—in which he writes about John Albert Bengel’s book, Gnomon of the New Testament. Hamann writes: “In the Preface, the author cites a very noteworthy saying by our Luther, which testifies to his philosophical spirit: ‘The science of Theology is nothing else, but Grammar, exercised on the words of the Holy Spirit’. This explanation is sublime, and adequate only to the high notion of true Theology” (Hamann 1956, 10 [letter from 19 February 1760]; see also Hamann 2007, 22). If this was indeed Wittgenstein’s source, it would further confirm Moore’s original version of the attributed claim. 19. The fact that according to Moore’s notes Wittgenstein continued by saying “This might mean: An investigation of idea of God is a grammatical one”, seems to provide further support for Ambrose’s version of the remark. 20. Cassian 1958, 234. 21. Ibid., 10, 3, 235. 22. Ibid., 10, 2, 234; and see also 10, 5, 236. 23. Namely, Genesis 1, 26 and 9, 6. 24. It is for this reason that Wittgenstein thought that “a serious and good philosophical work could be written that would consist entirely of jokes” (Malcolm 2001, 29; see also PI, §111, and BT, 90, 311). For “grammatical joke[s]” (PI, §111) are funny precisely in their flouting of grammatical rules in certain ways—thereby highlighting the existence and nature of the grammatical rules that are flouted. 25. As we saw in the above quotations regarding the Anthropomorphite Controversy, the terms “blasphemy” and “heresy” are indeed applied to the uses of the word “God” which each faction took to be absurd. In fact, Cassian even called the anthropomorphite understanding of God an “absurd heresy”—thereby bringing together one of the generally applied designations (“absurd”) and one of the specifically religious designations (“heresy”) of contraventions of the grammatical rules governing the use of a given word. 26. Some parts of this and the following section also appear—in one form or another—in my article “Simple Objects of Comparison for Complex Grammars: an Alternative Strand in Wittgenstein’s Later Remarks on Religion”, Philosophical Investigations 35, no. 1 (2012): 18–42. Despite this occasional overlap, each paper covers much ground not covered by the other. 27. I use ellipses to indicate that I have skipped out short sections of the notes; and as I usually skip only whole “paragraphs”, these ellipses are usually surrounded by forward slashes on either side. 28. Ernst Haeckel describes such a belief in The Riddle of the Universe (from which Wittgenstein quotes later in this lecture). See Haeckel 1929, 163–4 (chap. XI).

Notes

199

29. See ibid., 235 (chap. XV). 30. An asterisk by a word indicates that—due to Moore’s difficult handwriting— the reading is likely but not certain. 31. In a different context Wittgenstein says: “[T]hese statements were not scientific statements, not corrected by experience” (PO, 440). 32. See also, for example, BT, 121, 425–9, and RFM, III, 25. 33. Perhaps: them 34. Perhaps: as 35. See also LRB, 70. 36. Interestingly, in 1947 Wittgenstein again made an analogy between mathematical proof and proofs for religious beliefs, and again used Kierkegaard as an example of someone furnishing a nonscientific proof (see PPO, 404–5). 37. Wittgenstein returns to these terms in 1946 and 1947, and again, Kierkegaard is his example of a person with “hot”—i.e., nonscientific—religious beliefs (see CV, 60 and 64). 38. Wittgenstein was possibly referring to the word “immortality” rather than the word “soul”—but the same point would apply. 39. In his discussion of our language use often not having a fixed and definite grammar, Wittgenstein sometimes talks about words, and sometimes about sentences or utterances. What he says of each applies to the other, because the grammars of words are internally related to those of the sentences or utterances which include them. 40. I would like to thank Rachel Bayefsky, David Citron, David Egan, Peter Hacker, Elisha Mallard, Adrian Moore, Stephen Mulhall, and Nuno Venturinha for their helpful comments on earlier versions of parts of this paper (though I should note that some of them still disagree with much of what I have said). I would also like to thank Barbara Vetter for her help with translating Hamann; Thomas Baldwin for his permission to quote from G. E. Moore’s manuscript notes of Wittgenstein’s lectures; and Brian Rogers and David Stern for their permission to use material derived from our joint work in editing Moore’s notes.

NOTES TO CHAPTER TWO 1. 2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13.

Wittgenstein 1958. Trinity College Library: Add.Ms.a.522. McGuinness 2008, 216. Ibid., 250. Margaret Masterman Braithwaite, Sidney George Francis Guy Skinner (usually known as Francis), Reuben Louis Goodstein, Harold Scott MacDonald Coxeter (usually known as Don), Helen Knight and possibly Robert Martineau. Quoted in McGuinness 2008, 219. Bouwsma 1961, 141. McGuinness 2008, 219. Ibid., 220. Bouwsma 1961, 141. Letter from Wittgenstein to Coxeter of 22 January 1934, transcribed in a letter from G. K. Plochmann to G. H. von Wright in the von Wright/Wittgenstein Archive in Helsinki. Information supplied by Jim Klagge. Klagge and Nordmann 2003.

200

Notes

14. Mays 1999. 15. Partial manuscript copy of the Blue Book in the hand of Francis Skinner. For more information on this manuscript see below. 16. Wittgenstein 1969. 17. Trinity College Library, SRAFFA I21. 18. McGuinness 2008, 219. 19. Ibid., 220. 20. Bouwsma 1961, 141. 21. Property of the Mathematical Association. Currently on loan to Trinity College Library. 22. This being the Wittgenstein Archive headed by Michael Nedo. Not the Wittgenstein archive in Trinity College Library. 23. The “C-series” manuscripts are those large-format softcover notebooks used by Wittgenstein in the mid-1930s which someone has numbered C1, C2, C3, etc. They are numbered MSS 145–52 in von Wright’s catalogue. 24. Wittgenstein 1958, v. 25. Ambrose 1977, 3. 26. Wittgenstein 2009, 39–40. 27. Morris 2004, 3. 28. Schulte 2006, 402. 29. McGuinness 2006, 370. 30. Many friends and colleagues have been kind enough to offer suggestions when faced with drafts of this chapter or on hearing a version of it delivered in Lisbon. I would particularly like to thank Arthur Gibson, Adam Green, Kirsten Lamb, Brian McGuinness, David McKitterick, André Maury, Bernt Österman, Alois Pichler, Joachim Schulte, Christopher Stray and Michael Nedo for their help and guidance.

NOTES TO CHAPTER THREE 1. E.g., Krämer 1990. 2. The sense of “esoteric” that I am concerned with in this paper is not material that is meant to be kept secret from others, as with the gospel message, but rather material that others are not in a position to appreciate, as with Aristotle’s lectures. Thomas Wallgren urged this clarification. 3. CV, 6/8. In the pagination I employ, the first number refers to the page in the 1980 University of Chicago edition, and the second number refers to the page in the 1998 revised Blackwell edition. 4. I will not get into that here, though it is discussed extensively in my 2011, especially chaps. 2 and 6. 5. CV, 6/9. 6. CV, 10/12–13. 7. Engelmann 2006, 90. 8. See von Wright 1993, 97 ff. Carnap’s book was published in 1928. Herbert Feigl reports of meetings between Carnap and Wittgenstein in 1927–1928: “I recall Wittgenstein, on one occasion, precipitating a quarrel with Carnap, which . . . was mainly an expression of diametrically opposed personalities” (Feigl 1969, 638). 9. CV, 7/10. 10. Maslow 1961, x. 11. Lemoine 1975, 9. I have argued that awareness of Wittgenstein’s wartime experiences is crucial to the understanding of certain paragraphs in the Tractatus, in chap. 1 of Klagge 2011. Brian McGuinness goes so far as to

Notes

12. 13. 14. 15.

16. 17. 18. 19. 20.

21. 22. 23.

24. 25.

26. 27. 28. 29. 30.

201

say (2002b, 448): “no one else could have written his Tractatus Logicophilosophicus. Above all, no one who had not been through that War could have written it”. Engelmann claims (1967, 94) that Wittgenstein’s basic thoughts in the Tractatus “are incomprehensible” to the ordinary reader “owing to the absence of the psychological conditions from which alone such thinking can spring and which must exist, though to a lesser degree, in the reader’s mind as well”. Feigl 1969, 638. McGuinness 1979, 15. Personal communication from Feigl seems to be the source of this information: McGuinness 2002a, 189. Feigl 1969, 639. Carnap 1999, 175. Engelmann adds to this (1967, 118): “Wittgenstein found Schlick a distinguished and understanding partner in conversation, all the more so because he appreciated Schlick’s highly cultured personality—something which Wittgenstein found essential in his intellectual contacts with others”. The full account of this sad story has yet to be told. But the outlines can be found, among other places, in Baker 2003, xviii-xxi. Hodges 1984, 152–3. I recently queried Hodges about the source of this story, and he supposed he had gotten it from the late Robin Gandy, but could not remember any details. Redpath 1990, 46. McGuinness reports that the lectures in Lent (if there were any) and Easter terms of 1938 were unpaid, so they were not even official university classes (2011, 13). Wittgenstein’s letter to von Wright, 9 March 1939, in PO, 459–60. Malcolm 1984, 53–4 and 103. (Lazerowitz is not named in the published version of the letter.) Also cf. Britton 1999, 205, who “wrote to Wittgenstein asking his permission to attend his . . . discussion class [in 1931–32]” and got it. Gasking and Jackson 1999, 143. For a full account of Wittgenstein’s teaching at Cambridge, see my 2003, 331–72. The fact that Wittgenstein did not consider his lectures to be occasions for sharing his views only with like-minded disciples is shown by Malcolm’s report (1984, 48) that Wittgenstein “said that he had always regarded his lectures as a form of publication”. And Casimir Lewy recalled: “Wittgenstein once said to me that ‘to publish’ means ‘to make public’, and that therefore lecturing is a form of publication” (1976, xi). LWL, 60 and 63. Dated and published notes that we have from this era are LWL. Editing work is being done (by David Stern, Gabriel Citron and Brian Rogers) on G. E. Moore’s original chronological notes from this era toward publication, and these would be relevant sources of evidence as well. A survey of Waismann’s notes of the conversations with Wittgenstein from December 1929 to July 1932 show forms of the word “temptation” occurring only four times and really no occurrences of other noncognitive characterizations of philosophical issues (WVC, 72, 73, 157 and 185). PR, 85. Rothhaupt 2010, 58–9. Rothhaupt’s construction of the Kringel-Buch will appear in Wittgenstein forthcoming. PI, 4. The Blue Book, which Wittgenstein dictated to selected students in 1933–34, already includes discussion of what we are tempted by (14 times), what we crave, incline to, tend toward, or are fascinated or preoccupied by (10 times),

202

Notes

what we are dissatisfied with or contemptuous of (3 times), and what will break the spell. This shows a concern for how to address those who think differently. Elizabeth Anscombe recalls Wittgenstein’s response in lectures in the fall of 1944: “Let me think what medicine you need”, and reports that “the ‘medicine’ was effective, and the story illustrates Wittgenstein’s ability to understand the thought that was offered to him” (1981, viii-ix). 31. In 1946, one of Wittgenstein’s pupils from 1938, James Taylor, reflected back on his experience of Wittgenstein’s classes (McGuinness 2008, 394): If the question is just “Did I feel at that time (or at any other time) that I had been cheated?”, then the answer is the one I gave when I last saw you after thinking about it, simply “No”. (I have at various times discussed your teaching at Cambridge with other former pupils of yours who in some cases did think they had been cheated; in such discussions I have disagreed in all cases with the other person.) However the original question is not far distant from the question “Did you feel at that time that you had been misled?” I should have to answer to this that at that time, & in fact now, I did feel & feel that I was misled. . . . I found that I had very painfully to accustom my eyes to look at other things in a new way. Two things are clear from this—there was quite a mixed reaction among students to Wittgenstein’s teaching, and Taylor experienced Wittgenstein’s teaching as an attempt to change his way of looking at things. 32. Oddly, Karl Popper accused Wittgenstein of cultivating “esoter[ic]ism” in his infamous 1946 talk to the Cambridge Moral Sciences Club (Klagge 2003, 398). What Popper meant by this is clarified somewhat in a letter to Bertrand Russell, who was also at the meeting, written shortly after the meeting (published in Grattan-Guinness 1992, 13–15). In this letter of 27 October Popper wrote: “philosophical activity in Wittgenstein’s sense . . . is not exoterically arguable. It cannot, and does not, consist of more than clever guesses about various intended meanings. It leads to a series of ‘He may have meant . . .’, but it does not lead to any assertion which can be open to argument. This fact completely destroys any link with the rationalist tradition in philosophy and must lead to esotericity”. It is hard to know what Popper has in mind here, given that none of Wittgenstein’s later work had been published in 1946. But consider Popper’s criticism of Wittgenstein in volume II of his 1945 The Open Society and Its Enemies (20): From this point of view we may criticize a doctrine like that of Wittgenstein’s [in the Tractatus], who holds that while science investigates matters of fact, it is the business of philosophy to clarify the meanings of terms, thereby purging our language, and eliminating linguistic puzzles. It is characteristic of the views of this school that they do not lead to any chain of argument that could be rationally criticized; the school therefore addresses its subtle analyses exclusively to the small esoteric circle of the initiated. Popper goes on in a note to this text (note 52 to chap. 11, 299): It appears that irrationalism in the sense of a doctrine or creed that does not propound connected and debatable arguments but rather propounds

Notes

203

aphorisms and dogmatic statements which must be “understood” or else left alone, will generally become the property of an esoteric circle of the initiated. And, indeed, this prognosis seems to be partly corroborated by some of the publications that come from Wittgenstein’s school. (I do not wish to generalize; for example, everything I have seen of F. Waismann’s writing [a one-time expositor of Wittgenstein’s views] is presented as a chain of rational and exceedingly clear arguments, and entirely free from the attitude of “take it or leave it”.)

33. 34. 35. 36. 37. 38.

39. 40. 41. 42. 43. 44. 45. 46. 47.

48. 49.

This suggests that Wittgenstein is not interested in engaging with the thoughts of others who might disagree with him. The fact that he had not published anything, plus the perception that his students were selected by him, may have contributed to that image. But apparently what kept Wittgenstein from publishing was the very sense that he had not yet found the best way to engage with those who might disagree with him. See his 1933 letter to the editor of Mind, renouncing Braithwaite’s attempted summary of his views (PO, 157). The fact that he had still not published anything thirteen years after that might have provoked Popper to suspect that he had no intention of sharing his views publicly. But, in fact, as of 1945 anyway, Wittgenstein was still hard at work trying to publish something that would engage with those who did not share his preferred ways of thinking. He was still trying to figure out how to evangelize. And the preface to the Investigations, dated 1945, makes it clear that he wished to publish his work precisely because versions of his ideas “were in circulation, frequently misunderstood and more or less watered down or mangled” (PI, 4). PO, 161. LWL, 63. Cavell 1976, 71. Stern 2004, 22. Heal 1995, 64–71. Already in the Blue Book we find Wittgenstein inserting other voices into his reflections: “Now you might ask . . .” (BB, 3), and itself in quotation marks: “‘But surely the word “I” in the mouth of a man . . .’” (BB, 67). The Blue Book makes for a nice work to test this idea, since it was dictated, but in something very like, or substituting for, a classroom setting. In McGuinness 2008, 282. Britton 1999, 210. Venturinha 2010, 188. CV, 62/71, 18/25, and 61/70. Bouwsma 1986, 9. Drury 1984, 160. CV, 64/73. LCO, 78. In Chapter 8 of my book I trace the evolution of Wittgenstein’s thought on a certain issue concerning the relationship between physiological phenomena and mental phenomena, and claim that this change in purpose, from evangelical back to esoteric, helps explain what he says. But again, I consider research on this issue to be very preliminary. CV, 62/71. A few sentences in this paper have been reprinted from Wittgenstein in Exile, by James C. Klagge, published by the MIT Press, with permission of the publisher. The paper was improved by discussions following presentations at the University of Bergen and the University of Helsinki, in March 2012.

204

Notes

NOTES TO CHAPTER FOUR 1. Ortner 2000, 505 ff. 2. Ortner’s complete list of writing strategies includes: (1) writing in one go; (2) writing of one idea to one text version; (3) writing of one idea to several text versions; (4) writing, through several text versions, of one idea to several new ideas; (5) planning with subsequent writing out of the plan; (6) writing down of text-externally elaborated results; (7) linear step-by-step writing; (8) syncretistic writing; (9) writing of parts; (10) puzzle-writing (“PuzzleSchreiben”). Ortner’s German labels and short descriptions are as follows: (1) “(Scheinbar) nicht-zerlegendes Schreiben. Schreiben in einem Zug, Schreiben im Stil der pensée parlée, écriture automatique. Typ des Aus-demBauch-heraus-( = Flow)Schreibers”; (2) “Einen Text zu einer Idee schreiben. Typ des Einzigtext-, des Einen-Text-zu-einer-Idee-Schreibers”; (3) “Schreiben von Textversionen zu einer Idee. Typ des Mehrversionenschreibers, des Versionenneuschreibers”; (4) “Herstellen von Texten über die redaktionelle Arbeit an Texten (Vorfassungen), von verbesserten Versionen durch Arbeit am vorliegenden Text. Typ des Text-aus-den-Korrekturen-Entwicklers”; (5) “Planendes Schreiben (Plan = eine Version in Kurzschrift). Typ des Planers”; (6) “Einfälle außerhalb eines Textes weiterentwickeln. Konzeptuell extralingual + niederschreibend. Typ des Niederschreibers”; (7) “Schrittweises Vorgehen—der Produktionslogik folgend. Typ des Schritt-für-SchrittSchreibers”; (8) “Synkretistisch-schrittweises Schreiben. Typ des Synkretisten”; (9) “Moderat produktzerlegend. Das Schreiben von Produktsegmenten. Typ des Textteilschreibers”; (10) “Schreiben nach dem Puzzle-Prinzip. Extremproduktzerlegend. Typ des Produkt-Zusammensetzers”. My view is that Wittgenstein’s writing contains exemplifications of all of these strategies, but that the Wittgenstein of the PI is best classified and described as a syncretistic writer (strategy 8). Ortner, however, classifies Wittgenstein as a puzzle-writer (strategy 10). 3. I distinguish two main phases: from 1929 to 1943, with the “Urfassung” (1937) and “Pre-war” (1938), and from 1943 to 1951, with the “Intermediate” and “Final” versions as the high peaks. 4. Translations of terminology and quotations from Ortner into English are mine. 5. Pichler 2009, 57–97. 6. See Biggs and Pichler 1993, 50 ff. 7. For more details on this and the following issue see Pichler 1997, 38 ff. 8. See more on this in Pichler 1994, 91 ff. For examples stemming from the genesis of PI §§1–4 see Pichler 1997, sect. 2. 9. Allan Janik connects Wittgenstein’s writing of alternatives with Hertz’s concern “for showing us how alternative modes of presentation and representation can dissolve philosophical problems” (2006, 60). 10. Pichler 2004, 78 ff. 11. Also the deviation from this principle is significant: see the transition from TLP 2 to TLP 3 and from TLP 3 to TLP 4: “logisches Bild” / “logical picture” is the only term exempted from the a is b—b is c—c is d schema (see Stenius 1969, 18 ff., and Erbacher 2010, 82). 12. See Bazzocchi 2008, 125–40. 13. Pichler 2004. 14. According to Janik, Wittgenstein inherited the ideal of “Lückenlosigkeit” from Frege, but saw eventually that “only such a seeming potpourri of fragments could fittingly express his message” (2006, 108).

Notes 15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25. 26. 27.

28. 29. 30.

31. 32. 33. 34. 35.

36.

37. 38.

205

Pichler 1997. See ibid., 78 ff. See for example Savigny 1991, 307–19. See esp. TS 212. It has been standard to present the story as though the so-called PI philosophy chapter had been more or less taken over from the Big Typescript (see esp. Hilmy 1987). In my opinion, this view has, however, been successfully challenged by Stern (2006, 205–29), Pichler (2007, 123–44) and, more recently, Conant (2011, 620–45). Ortner 2000, 543 ff. “Ich verwende den Begriff synkretistisch als Gegenbegriff zu linear geordnet-fortschreibend-diskursiv” (Ortner 2000, 496). For examples see Ortner 2000, 491 ff. See Pichler 2004, 199 ff. Ortner 2000, 497. Ibid., 496. I do not claim that Wittgenstein wrote no aphorisms; most of the remarks published in Vermischte Bemerkungen / Culture and Value are in my view aphoristic. The two lists are based on material published earlier (Pichler 2009; taken, with revisions, from Wittgenstein: Como ler o álbum? Organized by Arley R. Moreno. Published by Coleção CLE, Campinas, Brazil. Used by permission of the publisher). Ibid. “Ein weiterer Ertrag dieser Strategie: Gedanken—solange die Tätigkeit des Schreibens dauert, Gedanken bis zum letzten Augenblick, Gedanken auch, die erst beim Schreiben entstehen” (Ortner 2000, 529). So far there has been relatively little discussion of Wittgenstein’s “album”. Exceptions include Binkley (1973), Schobinger (1991), Pichler (2004), Moreno (2009), Gründler (2011), Keicher (2011, with a focus on Wittgenstein’s photo album) and, most recently, Gorlée (2012). Moreno calls the process which led up to the PI, Wittgenstein’s “Big Diary”; the PI “album” is one of several possible outcomes from that process. Gorlée provides a comprehensive investigation of Wittgenstein’s “album” and related notions such as “fragment” and “fragmentariness” from a semiotic point of view. Ortner 2000, 533, 535–6. Ibid., 493. Pichler 2004, 136. One should expect that a study of the Skinner Wittgenstein materials (see Gibson 2010) would shed additional light on the move from the Brown Book to the PI. My translation; the original German reads: “Wittgenstein, sagte Moore weiter, habe ihm erklärt, er sei im Brown Book der falschen Methode gefolgt, in diesem Manuskript dagegen habe er die richtige Methode angewandt. Moore gab zu, er wisse nicht, was Wittgenstein damit meinte” (EPB, 12–13). I quote from a Wittgenstein Nachlass typescript which contains the English translation of the German “Pre-war version” of the PI preface and was edited and published by Venturinha (2010). This typescript was discovered only recently and is kept in the Austrian National Library under Cod. Ser. N. 39,544. The original German is in TS 225. Pichler 2004, 62. Josef Rothhaupt has recently suggested the fragment was the compound of MS 140, MS 114 and MS 115 (2011, 249); Peter Keicher has earlier

206

39. 40. 41.

42. 43.

44. 45. 46. 47. 48. 49. 50.

Notes suggested it is the first remark of MS 116 (2000, 225 ff.). But these items can only be the “fragment” inasmuch as they can be shown to exemplify the crisscross method. In my view, it is MS 142 and TS 220 which do this much better than any other piece up to 1938. On this see more in Pichler 2004, 222 ff. The Nachlass source is MS 109, 200–1, from November 1930. For a brief general study of all of Wittgenstein’s (drafts for) prefaces, see Keicher 2004. Carnap 1967, XVII. The original German from Der logische Aufbau der Welt (1928)—which Wittgenstein must have alluded to—reads: “. . . es wird in langsamem, vorsichtigem Aufbau Erkenntnis nach Erkenntnis gewonnen . . . So wird sorgsam Stein zu Stein gefügt und ein sicherer Bau errichtet, an dem jede folgende Generation weiterschaffen kann” (1961, XVIII–XIX). I discuss the contrasting relation between Carnap’s 1928 preface and Wittgenstein’s 1930 preface draft more comprehensively in Pichler 2009. The view that Wittgenstein’s 1930 preface draft might refer to Carnap’s preface to his Aufbau was to my knowledge first defended by Georg Henrik von Wright (1993, 97 ff.). On Wittgenstein’s similar relation to Russell’s conception of philosophy see Baker and Hacker 1983, 259 ff. Conant 2011, 632 and 640. Conant puts strong emphasis on the year 1937 and that it was in that year that Wittgenstein awoke to a new conception of philosophy: “It is this conception of what he seeks, in seeking the method of philosophy, that Wittgenstein finally came to abandon in Norway in 1937” (ibid., 642). I think that if a date is stressed, it should be November 1936 when Wittgenstein embarked on the syncretistic “Urfassung” of the PI in MS 142. If the PI’s philosophical methods are tied to the syncretistic and the album form (as I have tried to show), then Wittgenstein’s new conception of philosophy is practised from November 1936 onwards. Wittgenstein’s spring 1937 work on the PI’s “chapter on philosophy”, which is the focus of Conant’s claim, is definitely a crucial element in expressing this new conception of philosophy, but it is just that: an expression of an old conception abandoned and a new one already found and practised. Thus, I see the 1937 “meta-philosophical” remarks as a reflection upon and stock-taking and documentation of a philosophical practice since November 1936 which sought expression and documentation, rather than as the mark of a new beginning. Naturally, as Conant himself states (ibid., 624), any talk of terminus post or ante quem can be misleading, but if one wants to invoke a date (as Conant himself does), it is November 1936 rather than spring 1937. Stressing the later date blurs the point which Conant himself wants to emphasize in the end: with the shift to the PI, a change in methods and practice took place rather than only a change in the conception of methods and practice. Conant’s emphasis on 1937 could, however, be defended if one could show that Wittgenstein’s spring 1937 work on the “chapter on philosophy” had a significant retroactive effect, so that the MS 142 material written before underwent renewed editing along with, or subsequent to, the meta-philosophical work. Pichler 2007, 130. Conant 2011, 643. Ortner’s terminology is: “Wissen wiedergebend”, “Wissen erweiternd” and “Wissen schaffend” (2000, 348). Ibid., 537. See for example Gabriel 1991. Gabriel 2012, 170. The Nachlass source is MS 146, 25v, with its fair copy version in MS 115, 30.

Notes

207

51. Carnap 1967, XVI–XVII. The original German reads: “Das hat zur Folge, dass die strenge und verantwortungsbewusste Grundhaltung des wissenschaftlichen Forschers auch als Grundhaltung des philosophisch Arbeitenden erstrebt wird, während die Haltung des Philosophen alter Art mehr der eines Dichters gleicht . . . Aus dieser Forderung zur Rechtfertigung und zwingenden Begründung einer jeden These ergibt sich die Ausschaltung des spekulativen, dichterischen Arbeitens in der Philosophie” (Carnap 1961, XVIII–XIX). 52. In Pichler 2004 I have made an attempt at explicating this functionality in some detail. 53. Bouwsma 1986, 60–1. 54. This paper has come about through presentations at a number of conferences and meetings: in Aachen (2010, org. K. Herrmann, D. van Hulle and A. Gellhaus), Innsbruck (2011, org. A. Janik, U. Lobis and J. Wang), Marifjøra (2011, org. K. S. Johannessen and S. Säätelä) and Paris (2012, org. A. Soulez). I want to thank these meetings’ organizers and participants for helpful discussion. For valuable comments I also want to thank H. W. Gabler, Gottfried Gabriel, D. L. Gorlée, J. Klagge, A. Moreno, D. Smith (who also helped me with translating parts of Ortner 2000 into English), S. Szeltner and N. Venturinha.

NOTES TO CHAPTER FIVE 1. See, for example, chapter 1, section 3 of my 1989, trans. 1992; and my 2005. 2. Letters are quoted from the Gesamtbriefwechsel (2004, 2011). The letter referred to above can also be found in McGuinness 2008, no. 366; and in Klagge and Nordmann 1993, 463. 3. The motto was entered in MS 134 (77r) in April 1947. 4. See McGuinness 2008, nos. 391 and 393. 5. Reprinted in Klagge and Nordmann 1993, 480–510. 6. By Wolfgang Kienzler in his 2005 instructive article, 11. 7. Wittgenstein 1974. Cf. Biggs and Pichler 1993, 50–8. 8. The first version of this typescript (TS 221) is printed, as part of the Frühfassung, in my Kritisch-genetische Edition of Philosophische Untersuchungen (2001), 329–446. 9. Cf. MS 119, 79r ff.: “Fing an, meine alte Maschinschrift anzusehen und den Weizen von der Spreu zu sondern; wenn sie nur reiner zu sondern wären! . . . Mir ist beim Prüfen meiner alten Bemerkungen, als sollte ich den Hausrath einer Wohnung herstellen, indem ich Gegenstände aus einem Kehrrichthaufen ziehe und sie umständlich prüfen und zu säubern versuche. . . . Lese meine alten Bemerkungen. Die große Mehrzahl ist mir recht gleichgültig; viele, viele sind flau. Am besten sind die, die einfach ein Problem aussprechen. . . . Es ist viel Denken hinter diesen Bemerkungen. Aber brauchbar für ein Buch sind doch nur wenige ohne Umarbeitung. . . .” (Spelling and punctuation slightly modified.) 10. Kenny 1984, 36. Cf. Hilmy 1987, 233, note 95: “. . . Rhees, when compiling Philosophical Grammar, not only was familiar with the ‘revision’ of TS 213 in the early pages of MS 116, but also was of the opinion that the latter revision was carried out in 1936, long after the revisions (in MSS 114, 115 and 140), which he chose to publish in Philosophical Grammar as the definitive ones. Why Rhees chose to displace TS 213 by Philosophical Grammar, when he was aware that there was an even later revision of TS 213 material, is a mystery”. Two things are to be said about this. First, as far as I can see,

208

Notes

Rhees nowhere claims to have published the “final” (Kenny) or “definitive” (Hilmy) revision of the Big Typescript. Second, it seems that Rhees (in contrast with Kenny and Hilmy) was aware of the fact, if it is a fact, that MS 116i does not contain a “revision” of an earlier version of a work by Wittgenstein. 11. The quoted words are of course taken from the preface to Philosophical Investigations, where he says that “die Gedanken von einem Gegenstand zum andern in einer natürlichen und lückenlosen Folge fortschreiten sollten”. 12. Possibly the evidence is there, but we have not succeeded in interpreting it. Thus, it may very well be that a close study of certain numerical indications added to one copy of Bemerkungen I would help us understand Wittgenstein’s procedure. It seems clear that these numbers refer to a list (or lists) of topics which may have played a role in organizing the material. However, no such list is extant. But even if it should prove possible to reconstruct this list (or lists), it will be a difficult and time-consuming job. It may be that taking into consideration what Wittgenstein did in organizing parts of Bemerkungen I in such a way that he was able to produce Bemerkungen II (TS 230) would be helpful in getting this job done. 13. Rhees 1974, 489.

NOTES TO CHAPTER SIX 1. As explained later, by the first I mean a conception of logic as concerned with a priori principles and forms governing thought or language, by the second a view of logic as an empirical investigation of the principles and forms governing thought or language, and by the third the account that logical principles and forms are a matter of convention. 2. For discussion of the principles of the method of interpretation employed here, see also Kuusela 2008, 13–15. 3. This remark is from summer 1930. Where no published translation exists, the translation is mine. I have also occasionally amended existing translations. 4. Given that philosophy for Wittgenstein is the activity of logical clarification (TLP, 4.112), the terms “logic” and “philosophy” can be used interchangeably when talking about the contrast of Wittgenstein’s investigations with psychology and the empirical sciences. 5. This view may have been influenced by Russell’s conception of “this” and “that” as the only real names, and is perhaps to be regarded merely as an extension of this Russellian view onto propositions (Russell 2007, 201). Wittgenstein talks about Russell in this regard, alluding to a sublimation of logic in Investigations §38. Wittgenstein also characterizes Frege’s and Russell’s logical notations as involving sublimation in that they seek to reduce many different forms of expression into only a few forms. Although Wittgenstein recognizes that such a unification might have advantages, he regards the consequent loss of grammatical/logical distinctions a greater disadvantage. Another example he gives of a sublimated conception of logic is Frege’s view of logic as the class of the most general truths (BTE, MS 152, 82; BTE, MS 154, 89v-91v). 6. Less ambitious projects of reduction and simplification that are relative to particular logical or technical languages or limited to particular domains of language can be undertaken without the assumption of the uniformity of logic. For Frege’s explanation of the requirement of the uniformity of thought, see Frege 2007, 12. 7. The sense in which Wittgenstein’s methodological shift constitutes a response to the problem of dogmatism is discussed in my 2008. See especially chap. 3.

Notes

209

8. An alternative interpretation by Baker and Hacker of the sense in which logic might be losing its rigour is discussed and problematized in Kuusela 2008, 132 ff. 9. According to Wittgenstein, Ramsey’s goal was to explain logic on the model of a natural science, whereby Ramsey, however, didn’t see the problem of explaining the relation between logical exactness and the vagueness of actual language use (MS 162b, 9v). If this is correct, Ramsey would have been blind to just the problem Wittgenstein sought to address. “I always wanted to say (against Ramsey): Logic can’t become an empirical science. But how we use words is certainly an empirical matter [Empirie]” (BTE, MS 152, 193–4). Wittgenstein’s view of Ramsey is presumably based on discussions with him. 10. Carnap, whose approach conforms to option 2), acknowledges this problem. While the task of logic for him is the construction of exact languages for the purposes of philosophy and science, he admits that the clarification of the syntactical rules of natural languages would be too complicated a task to undertake due to their unsystematic and logically imperfect structure (Carnap 1967, 2, 312). Consequently, the direct analysis of such languages must “inevitably fail” (ibid., 8). Still, Carnap’s conclusion is not that the method of logical syntax, i.e., clarification by spelling out syntactical concepts and rules, cannot be applied to natural word-languages, but he insists that it may in principle “also be applied to the analysis of the incredibly complicated word-languages” (ibid., 8). This reveals certain limitations of Carnap’s method that Wittgenstein’s approach is designed to overcome. For a reconsideration of the relation between the Tractatus’ and Carnap’s philosophies of logic, see Kuusela 2012. 11. The remark is made in a note inserted between manuscript pages of the Investigations in the context of §108. 12. Possibly one might understand idealization in science similarly to how Wittgenstein proposes to think of it in logic. In that case, his account of idealization in logic simply contrasts with a particular way of understanding idealization in science. For a discussion of the notions of abstraction and idealization in science, see Stokhof and van Lambalgen 2011. 13. For fuller quotation and discussion, see Kuusela 2008, 79 ff. 14. See LW II, 48; MS 169, 71v; TS 302, 5; and Kuusela 2008, chap. 5.4 for discussion. 15. This interpretation of Wittgenstein’s conception of essence as expressed in grammar contrasts with various idealist, constructivist and conventionalist readings. A detailed argument for the proposed reading is given in Kuusela 2008, chap. 5. 16. For a parallel contrast between mathematical and anthropological propositions, see RFM, III, §65. For discussion of the distinction between factual statements and statements of a rule, see Kuusela 2008, 113–16. 17. For an account of Wittgenstein’s later method of language-games as a method of logic that extends logic beyond calculus-based methods, see Kuusela forthcoming. 18. I’m grateful to Angela Breitenbach, Alberto Emiliani, Rupert Read, Davide Rizza and NunoVenturinha for comments on earlier versions of this chapter.

NOTES TO CHAPTER SEVEN 1. The second comment was made in conversation with Drury. Wittgenstein remarked that Kant and Berkeley were “very deep thinkers”, whereas with Schopenhauer “I seem to see to the bottom very quickly” (Drury 1981, 95).

210

Notes

2. One must constantly keep in mind the fact that the concept of experience, as Kant deploys it, is a weighty one. It is “a cognition that determines an object through perceptions” (A 176/B 218). 3. For detailed examination of this transformation, see Baker and Hacker 2009a. 4. Did he think there was when he was writing the Tractatus? After all, in 5.634 he wrote: “There is no a priori order of things”. That remark is misleading when taken out of context. In its context it embroiders on 5.63, “I am my world”. The fact that one encounters no subject in one’s experience (as Hume had insisted) is connected with the fact that “no part of our experience is at the same time a priori”, since everything we experience, and what we can describe in words, could be other than it is. For “There are no pictures that are true a priori” (TLP, 2.225). But the logical form of reality, which cannot be described in words, but is shown by true and false descriptions of the empirical world (TLP, 4.121), is, of course, a priori. 5. For detailed discussion of these principles, see Hacker 1986, chap. VII. 6. This, incidentally, is why Barry Stroud’s famous criticism of Strawson’s purported transcendental argument in Individuals (1959, chapter 3) is mistaken. Stroud argued that the most Strawson’s transcendental argument can establish is how we must believe things to be. But what Strawson’s argument establishes is that subjective experience is possible only on condition of the recognition of the adequacy of logical criteria for other-ascription of experience. Satisfaction of such criteria, in the absence of defeating conditions, suffices for knowledge (not mere belief) of how things are with another. But whether such criteria are, as a matter of fact, satisfied, is not for philosophy to say—that is a matter for experience to determine. And we all know how it determines it. 7. See Baker and Hacker 2009b, in the essays entitled “Following Rules, Mastery of Techniques and Practices” (135–56), and “Private Linguists and ‘Private Linguists’—Robinson Crusoe sails Again” (157–68). 8. Ibid., exegesis of §202. 9. I am grateful to Hanjo Glock, Edward Kanterian, Adrian Moore, Hans Oberdiek, Herman Philipse and especially Daniel Robinson for their helpful comments on an earlier draft of this paper. An earlier version of this paper was published in French in “Kant et Wittgenstein, le problème des arguments transcendentaux”, in Grammatical ou Transcendental?, edited by Arley R. Moreno and Antonia Soulez, Cahiers de Philosophie de Langage, vol. 8 (Paris: L’Harmattan, 2012), 17–44.

NOTES TO CHAPTER EIGHT 1. MS 142 is reproduced in Wittgenstein 2001, 51–204, while MSS 117–19 serve as the basis for TS 221, cuttings from which were used to produce TS 222, the source for material in Wittgenstein 1978. TS 221 is also reproduced in Wittgenstein 2001, 329–446, with references to the manuscript sources. I have consulted the sequence of the remarks in MSS 117–19 in Wittgenstein 2000, to which the page numbers I provide refer. I use Wittgenstein 1953 for translations of remarks from MS 142 and Wittgenstein 1978 for translations of remarks from MSS 117–19 except in one case. I supply my own translation of a remark in TS 221, not reproduced in TS 222. 2. Elsewhere Hacker refers to Wittgenstein more guardedly as denying that language is “au fond” a calculus with “strict rules” (1996, 86).

Notes

211

3. For Stern there is “no sharp transition” between the two models but in later work the calculus model is mostly deployed “to bring out the disanalogies between ordinary language and a calculus” (1995, 120). 4. This is not to deny dissenting voices. S. Hilmy observes that from 1930 on Wittgenstein used the terms “calculus” and “language-game” “virtually interchangeably” and when revising his remarks, replaced the former with the latter without noticeable fanfare (1987, 98, Hilmy’s emphasis). Whether this undermines “the standard views that during the 1930s Wittgenstein relinquished the ‘calculus conception’ of language and replaced it with a new ‘language-game model’” is, however, another matter (99), as is ter Hark’s claim that Hilmy has “convincingly shown that the shift from ‘calculus model’ to ‘language-game model’ was no more than terminological” (1990, 7). 5. Also compare Glock (1996, 194): “[B]y turning to language-games, Wittgenstein switched attention from the geometry of a symbolism (whether a language or a calculus) to its place in human practice”. 6. In this connection Glock cites Wittgenstein 1958, 17 and 81. Not without reason he also takes Wittgenstein to have introduced the term “language-game”, “to draw attention to various similarities between language and games, just as the calculus analogy highlighted similarities between language and formal systems” (1996, 193). 7. To budget for a small difference between the two texts, I modify the English translation of the remark in Wittgenstein 1953 slightly. From now on I do not flag such minor modifications. 8. In their translation of the Investigations Hacker and Schulte have Wittgenstein lamenting that “we don’t have an overview” and extolling “surveyable representation” (Wittgenstein 2009). 9. Gerrard also holds, I believe wrongly, that there is no accounting for the growth of mathematics if mathematical language is reducible to “mere syntactical rules . . . divorced from human use or purpose” (Gerrard 1991, 132). 10. Compare McGuinness (2002, 201): “Wittgenstein was not a completed revelation, but a questioning philosopher, not a book but a man”, one “concerned with individual problems of philosophy”. 11. Also note that Wittgenstein wrote as early as 1914: “In order to recognize the sign in the sign we have to attend to use” and in 1916 wrote: “The way in which language signifies is mirrored in its use” (1979, 18 and 82). 12. The phrase “language-games [Sprachspielen]” occurs in MS 113, a manuscript compiled in 1931/1932, but only incidentally. Compare ter Hark, 1990, 7 and 12. 13. Also far from equating meaning with use, Wittgenstein apparently said: “If the meaning is represented by the use of a symbol, it is no use saying, ‘The use is different, therefore the meaning is different’” (1976, 81). 14. I have benefited from Paul Forster’s comments on an early draft and am indebted to Lynne Cohen and Nuno Venturinha for help preparing the final version of the paper.

NOTES TO CHAPTER NINE 1. Cf. Schroeder 2001. 2. Cf. MS 179, 7v: “To follow a rule presupposes a language”. (Einer Regel folgen setzt eine Sprache voraus.) 3. Cp. BB, 25: “. . . not only do we not think of the rules of usage—of definitions, etc.—while using language, but when we are asked to give such rules,

212

4. 5.

6. 7. 8. 9. 10. 11.

12. 13.

Notes in most cases we aren’t able to do so. We are unable clearly to circumscribe the concepts we use; not because we don’t know their real definition, but because there is no real ‘definition’ to them”. Cf. BB, 28: “there are words . . . which . . . are used in a thousand different ways which gradually merge into one another”. “Why then do we in philosophizing constantly compare our use of words with one following exact rules? The answer is that the puzzles which we try to remove always spring from just this attitude towards language” (BB 25–6; cf. AWL, 3). “Now it is clear that this problem about the concept of time asks for an answer given in the form of strict rules. The puzzle is about rules.—Take another example: Socrates’ question ‘What is knowledge?’ Here the case is even clearer, as the discussion begins with the pupil giving an example of an exact definition, and then analogous to this a definition of the word ‘knowledge’ is asked for. As the problem is put, it seems that there is something wrong with the ordinary use of the word ‘knowledge’. It appears we don’t know what it means, and that therefore, perhaps, we have no right to use it. We should reply: ‘There is no one exact usage of the word “knowledge”; but we can make up several such usages, which will more or less agree with the ways the word is actually used’. The man who is philosophically puzzled sees a law in the way a word is used, and, trying to apply this law consistently, comes up against cases where it leads to paradoxical results. . . . ‘What is time?’ This question makes it appear that what we want is a definition” (BB, 26–7). Note how often he uses the word “order” together with or instead of the word “rule”, e.g., PI, §§186–9, 206, 212; cf. MS 165, 40, 84. Cf. Schroeder 2006, 181–97. Cf. RFM, 341c–f, where “How can I follow a rule?” and “How do I know what to do in response to the order ‘Slab!’?” (translation modified) are treated as on a par. Cf. Schroeder 2006, loc. cit. TS 222 contains variants of PI, §§185, 189, 190, 191, 192, 193, 194, 196, 197. “I have no right to want you to say that mathematical propositions are rules of grammar. I only have the right to say to you, ‘Investigate whether mathematical propositions are not rules of expression, paradigms—propositions dependent on experience but made independent of it. . . .’” (LFM, 55) Kline 1980, 93. I am indebted to David Dolby, Sebastian Greve and Peter Hacker for comments on earlier drafts of this paper.

NOTES TO CHAPTER TEN 1. See von Wright 1982. 2. See Wittgenstein 2001. 3. See Carroll 1895.

NOTES TO THE ADDENDUM (INTRODUCTION) 1. von Wright 1993, 499–500. 2. Ibid., 499. These copies are kept at Trinity College Library, Cambridge, bearing the reference 221(i) and (ii). It is noteworthy that in the first version of

Notes

3. 4. 5.

6.

7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

213

the catalogue von Wright only mentions one complete copy: “The existing copy of 221 is a second copy. The first copy was evidently used for composing 222” (1969, 500). von Wright 1982, 118. von Wright, Rhees and Anscombe 1978, 30. I saw this document for the first time in 2009 during a research visit to Trinity College Library and reexamined it in 2012. I owe a great debt of thanks to Jonathan Smith for his help and stimulating encouragement and extend my gratitude to the other staff of the Wren Library. I should point out that the current order of the pages and cuttings clearly reveals two mistakes. The two small cuttings are placed between pages 238 and 239, but the cutting from page 240 must follow page 239 whereas the one from page 243 must precede page 244. Their correct place in the document is therefore between pages 239 and 244. TS 221, 236, my translation. This is TS 221(ii). The documents, which include a photocopy of our fragment (with the pages and cuttings in their correct order), accompany the copies of items 220 and 221, a summary of which is provided by von Wright himself in his 1997 unpublished inventory (16–17). I express my gratitude to André Maury, Thomas Wallgren and Bernt Österman for their kindness when I visited the Archives in 2009. A reference to this page as well as to the uncatalogued bunch of pages and cuttings is already made in Venturinha 2010a, 153, note 7. See von Wright 1982, 118–19. See also Schulte 2001b, 1102–5. See von Wright and Nyman 1979, 3. See also note 38 below. It is noteworthy that the first page numbered 158 in TS 221(i) is identical with the first version of that page in TS 221(ii). The only other page with this number in 221(i) is that with a different content. I have summarized this important cover sheet for the first time in Venturinha 2010a, 151 and 156, note 37. von Wright 1969, 493. See ibid., 483, this coinciding with von Wright 1993, 482. von Wright 1993, 481. There is a new catalogue of these papers, prepared by Jonathan Smith, according to which the items are now organized at the Wren Library. I direct the reader again to von Wright 1997, 16–17. von Wright 1982, 119. Cf. Monk 1990, 373. Cf. Paul 2007, 25. See Nedo 1993, 37. See Schulte 2001b, 1098. These two sets of pages come from a different typewriter but they follow the same mode of pagination and section numbering, viz. “#.”. In this set of pages both the pagination and the section numbering do not include a full stop, viz. “#”. See McGuinness 2002, 278–9. McGuinness mentions indeed two letters that lend support to this view, namely one from Francis Skinner to Wittgenstein of 14 October 1937, in which Skinner writes that “[he] asked [Fania Pascal] about a typist for [Wittgenstein]” but that “[at] the moment neither she nor [Roy] Pascal can think of any-one in Cambridge who could do the job” (Wittgenstein 2004), and another from Margarete Stonborough to Wittgenstein of 1 December 1937, in which she talks about a previous dictation with her brother and suggests that a new one will take place (cf. ibid., where it is, however, conjectured in a note that the result of this dictation may have been TS 221).

214

Notes

27. Rhees’ letter of the end of November indicates that the pages and cuttings are those that form TS 238, a revision of which was retyped by Rhees upon Wittgenstein’s request to fill in that space of TS 239 (pages 77–93), as becomes clear from the other November letter (see also von Wright 1997, 21). The resulting typescript was nevertheless rearranged once again by Wittgenstein, with fragments of Rhees’ typing being pasted in TS 239. Since the unrevised pages of both TSS 238 and 239 match those of TS 220, the former shows, as Schulte notes, “that at least from pages 66–137 there must have been a third exemplar” (2001b, 1101, my translation). Other portions from pages 80–92 of TS 220 are to be found in TS 237. The Rhees papers at Trinity College include a typescript made by him that covers pages 84–100 of TS 239. It is followed by a photocopy of 222 beginning with the current page 17. This folder bears the reference E3. 28. Compare, for example, the revised last paragraph of §7 and the revised first paragraph of §10 of TS 220 with the same sections in what von Wright called TS 220b, the corresponding material in TS 239 reading like in the unrevised 220. 29. See in this regard Schulte 2001b, 1100, as well as McGuinness 2002, 279–80 and 284. 30. In certain cases, Rhees has also omitted remarks not crossed out by Wittgenstein but simply put in parentheses (compare, for example, page 173 of TS 221 with page 183 of that TS 221b, its next sheet containing a handwritten copy, apparently in von Wright’s hand, of the two parenthetical remarks not crossed out). Rhees has also followed Wittgenstein’s editorial instructions (compare page 195, together with 209 and 214, of TS 221 with pages 204–5 of 221b). 31. The beginning of page 236 which continues a remark from the previous page is not to be found in TS 222, which does not include any other content from page 235. 32. See TS 222, 112–13, the document following with the famous section about “the natural history of human beings” (114), from page 222 of TS 221, which would make its way into the Investigations (see Wittgenstein 2009, §415). The other remark of page 240, if we exclude the opening line in our bunch, is not to be found in TS 222. 33. The other remark of page 241 is not to be found in TS 222. 34. See TS 222, 135, the document following with pages 256–66, i.e., Appendix I of Remarks on the Foundations of Mathematics. 35. The other remark of page 243 is to be found in TS 222, 109. 36. The first remark actually bears a second marginal number in parentheses, 271, apparently from a later revision. In his edition of TS 221, Schulte accurately points to this remark as the “insertion from p 241 etc.” (see Wittgenstein 2001, 370, note 1). See also the editors’ note to §254 (260) of TS 221 in Wittgenstein 1979. 37. Wittgenstein 2001, 387, note 3, my translation. The remark also bears other marginal numbers in parentheses, 309 and 279. For a similar conclusion, see von Wright and Nyman 1979, 4, and their note to §295 (301) of TS 221 in Wittgenstein 1979. 38. As von Wright observes, the German preface “dated ‘Cambridge, August 1938’” that constitutes TS 225 “was typed on the same typewriter as the second part of TS 221”. He concludes that “[t]his is an indication . . . that the first and main part of TS 221 was typed not later than June or July 1938” (1982, 120). For McGuinness, though, the typing of “the bulk of 221” will have taken place at the beginning of November (cf. 2002, 284). It is possible that

Notes

39.

40. 41. 42. 43. 44.

45.

46. 47. 48.

215

what Wittgenstein dictated at that time had already to do with “the second part of TS 221”, whose first 11 pages come from an earlier source, MS 115. Smythies’ translation would appear to begin at §35 of page 20 of TS 239 and continue—with some gaps—until §141 (see Notebook XIII of his papers, a microfilm of which is available at Trinity College Library under the reference 00J00003C, here Box 3). I first pointed this out in Venturinha 2010b, 183 and 186, note 3. TS 239 was for some time in the possession of Smythies (see Paul 2007, 45–6, note 19) and therefore his translation can have been done at any time. The only thing we really know is that a revision of TS 226 was made by him and Wittgenstein, who writes to G. E. Moore on 2 February 1939 saying that they were at a point corresponding to §19 of TS 220 and §23 of TS 239, both on page 13 (see McGuinness 2008, 291). von Wright 1982, 119. Ibid. See ibid., 119–22. Schulte 2001a, 22, my translation. See in addition 32. See also Hacker and Schulte 2009, xix, where it is said that it was “[s]ometime between late 1939 and 1943” that “Wittgenstein revised the Early Draft” (my emphasis). An additional piece of evidence for the earliness of some revisions in TS 239 is a reference to MS 121, more specifically to the remarks that would later form §182 of the Investigations, at the end of the document (page 133), remarks that must have been written between 5 September and 25 December 1938—MS 121 having also been used for the final part of TS 222. See Schulte 2001b, 1102, as well as his edition of TS 239 in Wittgenstein 2001, 558, note 6. A cover sheet preserved with TS 222, probably by Anscombe, presents this item as a reorganization of Wittgenstein continuing the first half of the prewar Investigations. The cover sheet of TS 221(ii) is similar, highlighting only that this was the version originally prepared. See, in particular, Venturinha 2010a, 146. See Wittgenstein 2001, 416–19 and 422–4. I would like to thank the audience at a meeting organized by the Wittgenstein Archives at the University of Bergen in 2011 where I first presented this original, and Joachim Schulte for helpful comments on a draft of the introductory text.

NOTES TO THE ADDENDUM (TEXT) 1. The comma is missing from the Bergen edition of TS 221. 2. A quotation mark appears in place of the full stop in the Bergen edition of TS 221. 3. “Anderung” in the typescript. Both the Bergen edition and Schulte’s edition of TS 221 correct the mistake but they do not indicate the original. 4. See note 3. 5. The typescript reads “garnicht”. This is corrected to “gar nicht” in the normalized transcription offered by the Bergen edition of TS 221, but the original is preserved in Schulte’s edition of that typescript. 6. This is erroneously transcribed as “Interval” in the Bergen edition of TS 221. 7. Immediate corrections of punctuation or orthography are not reproduced in Schulte’s edition of TS 221. 8. This is erroneously transcribed as “dann” in the Bergen edition of TS 221.

216

Notes

9. A quotation mark appears in place of the full stop in the Bergen edition of TS 221. No punctuation mark appears in Schulte’s edition of that typescript. 10. The comma is missing from the Bergen edition of TS 221. 11. The comma is missing from the Bergen edition of TS 221. 12. This is erroneously transcribed as “ach” in the Bergen edition of TS 221. 13. Both the Bergen edition and Schulte’s edition of TS 221 correct “wesentlichen” to “wesentliche”, but no indication of the original is given. 14. This is the first of Wittgenstein’s ink corrections in our typescript. It is already made in the normalized transcription offered by the Bergen edition of TS 221 as well as in Schulte’s edition of that typescript, the latter with reference to the manuscript source. 15. This is erroneously transcribed as “unseres” in the Bergen edition of TS 221. 16. The opening quotation mark is missing from the Bergen edition of TS 221. 17. This is the second of Wittgenstein’s ink corrections in our typescript. It is also already made in the normalized transcription offered by the Bergen edition of TS 221 as well as in Schulte’s edition of that typescript, but the latter does not indicate the original. Translator’s Note: Due to the “work in progress” nature of the text, I have attempted to remain as close to the original German as possible. Often this meant sacrificing a degree of readability for a more precise conveyance of Wittgenstein’s writing process. However, given the inherent limitations as to how “faithful” a translation can be and the fact that Wittgenstein’s use of language is sometimes unusual or even eccentric, I hope that this translation strikes a useful balance between conveying the struggling and unfinished character of his writing and comprehensibility.

Contributors

Gabriel Citron is a junior research fellow at Worcester College, Oxford. He has published articles on Wittgenstein’s thought and the philosophy of religion, and is co-editing Wittgenstein’s Lectures: Cambridge 1930–33, from the notes of G. E. Moore. P. M. S. Hacker is Emeritus Research Fellow at St John’s College, Oxford. He is author of the four-volume Analytical Commentary on the Philosophical Investigations, the first two volumes co-authored with G. P. Baker (1980–96), of Wittgenstein’s Place in Twentieth-century Analytic Philosophy (1996) and of Wittgenstein: Connections and Controversies (2001). He has written extensively on philosophy and the neurosciences, most recently Philosophical Foundations of Neuroscience (2003), and History of Cognitive Neuroscience (2008) co-authored with M. R. Bennett. He is currently working on a three-volume work on human nature, the first volume of which, Human Nature: the Categorial Framework, was published in 2007. The second volume entitled The Intellectual Powers: a Study of Human Nature will be published later this year. James C. Klagge received his PhD from UCLA in 1983, and has taught at Virginia Tech since 1985. He has co-edited two collections of primary material on Wittgenstein with Alfred Nordmann—Philosophical Occasions 1912–1951, published in 1993; and Public and Private Occasions, published in 2003. He edited a collection of essays on the relationship between Wittgenstein’s life and work—Wittgenstein: Biography and Philosophy, in 2001; and recently published a book, Wittgenstein in Exile, in 2011. He has served as chair of his department since 2007, and was Guest Professor at the Wittgenstein Archives in Bergen, Norway, in March 2012. Professor Klagge is currently working on a collection of background materials that influenced Wittgenstein in his two main books. Oskari Kuusela is Senior Lecturer at School of Philosophy, University of East Anglia. He is the author of The Struggle Against Dogmatism:

218

Contributors

Wittgenstein and the Concept of Philosophy (2008), and a co-editor of Wittgenstein’s Interpreters: Essays in Memory of Gordon Baker (2007) and The Oxford Handbook of Wittgenstein (2011). Andrew Lugg is Emeritus Professor of Philosophy at the University of Ottawa. He is the author of Wittgenstein’s Investigations 1–133: a Guide and Interpretation (2000) and is presently writing a book on Wittgenstein and colour. He lives and works in Montreal. André Maury, Docent and sometime Professor, teaches at the University of Helsinki. He has worked with G. H. von Wright on the Wittgenstein papers. His research interests lie in the philosophy of language and Wittgenstein. Alois Pichler is Associate Professor at the Philosophy Department at the University of Bergen and the Head of its Wittgenstein Archives. He publishes mainly in the fields of Wittgenstein research and digital humanities. He is the editor of Wittgenstein Source, an Open Access online platform which makes available 5,000 pages of the Wittgenstein Nachlass in both text and facsimile editions (www.wittgensteinsource.org) and one of the editors of the newly established journal Nordic Wittgenstein Review. He is the author of Wittgensteins Philosophische Untersuchungen: Vom Buch zum Album (2004) and co-editor of Wittgenstein: The Philosopher and His Works (2006). Severin Schroeder is Reader in Philosophy at the University of Reading. He is the author of three books on the philosophy of Wittgenstein: a monograph on the private language argument (Das Privatsprachen-Argument, 1998), Wittgenstein: The Way Out of the Fly-Bottle (2006), and Wittgenstein lesen (2009). He is the editor of Wittgenstein and Contemporary Philosophy of Mind (2001) and Philosophy of Literature (2010). He is currently working on Wittgenstein’s philosophy of mathematics. Joachim Schulte teaches at the University of Zürich. He has published a number of articles and four books on the philosophy of Wittgenstein. He is co-editor of critical editions of Wittgenstein’s Tractatus LogicoPhilosophicus (1989) and Philosophical Investigations (2001). In recent years, he has chiefly worked on Wittgenstein’s middle period. Jonathan Smith has been archivist at Trinity College Cambridge since 1991 where he is responsible for a large part of the Wittgenstein Nachlass. His academic interests include enquiries into the nature of archives produced by scholars and the various ways they are subsequently used for research purposes. Recent articles include “Circuitous processes, jigsaw puzzles and indisputable results: making best use of the manuscripts of

Contributors

219

Sraffa’s Production of Commodities by Means of Commodities” (2012) and “Surprise in the archive” (forthcoming). James M. Thompson is a senior research fellow at the Graduate School Society and Culture in Motion at the Martin-Luther-University HalleWittenberg (MLU). He is currently a guest professor at the K. U. Leuven in Belgium and has been teaching philosophy as senior lecturer for many years at the department of philosophy (MLU). Moreover, he has researched and lectured at the University of Zürich (Switzerland) and Southern University of Illinois, Carbondale (USA). He is the author of Wittgenstein on Phenomenology and Experience (2008) and co-editor of Wittgenstein and Ancient Thought, with Ilse Somavilla (2012). Among his other research interests he is working on the connections between Wittgenstein, American pragmatism, phenomenology and practice theory. NunoVenturinha is a senior research fellow at the Institute for Philosophy of Language and teaches at the Faculty of Social and Human Sciences of the New University of Lisbon. He has held visiting teaching positions at the universities of Lisbon and São Paulo, and has been a visiting researcher on various occasions at the universities of Bergen, Innsbruck, Oxford, Cambridge and Helsinki. His research interests concentrate on the philosophy of language and Wittgenstein. He is the author of Lógica, Ética, Gramática: Wittgenstein e o Método da Filosofia (2010) and the editor of Wittgenstein After His Nachlass (2010).

This page intentionally left blank

Index

Adam 23 additions 3, 82–5, 178; see also amendment(s); corrections; revision(s); variants Ambrose, A. 20, 23, 39, 42–3, 47, 49–50, 197–8, 200 amendment(s) 38, 40–3, 46–7, 49; see also additions; cancellations; corrections; revision(s); variants analogical (the) 77 analogy (uses of) 9, 32–3, 49, 136, 159, 166, 187, 199, 211 analytic philosophy 77–8, 165 Anscombe, G. E. M. 6–7, 169–70, 174, 177–80, 197, 202, 213, 215 anthropology 94, 114, 116; of language 103 anti-psychologism 93–4, 96, 116–17 aphorisms 2, 67, 71, 203, 205; see also fragment(s) apriorism 12, 93, 103, 109, 111, 113–14; Tractarian 94, 97–8, 100, 117 argumentation 3–4, 128 argumentative: order 65; roles, strategies 11; techniques 7; see also order; writing: strategies argument(s): chain of, connected and debatable 202; clear 203; conclusive 168; discursive-linear strands of 73; empirical or of metaphysics 127; function and 110; private language 132; scientific 31, 127; series of 12; transcendental 120–1, 125–9, 131, 133, 210; see also language: private; linear; linearity Aristotle 10, 52–3, 57, 127, 200 audience 10, 20, 52–3, 55, 61 Augustine 66–9

Babich, B. 2, 193 Bachtin, N. 54, 194 Baker, G. P. 6, 194–5, 201, 206, 209–10 Baldwin, T. 199 Barth, K. 196 Bayefsky, R. 199 Bazzocchi, L. 204 belief(s) 19, 59, 210; esoteric or metaphysical 52; in inference 171; mathematical 150; religious 24–7, 29–32, 34, 199; scientific 27–8; see also esoteric(ism); inference; mathematical; metaphysical; religion Bengel, J. A. 198 Berkeley, G. 209 Bible 4, 53; Genesis 198; see also Scripture Biggs, M. A. R. 197, 204, 207 Big Typescript (TS 213) 5, 10–11, 29, 32, 59, 66–7, 69, 83–8, 98, 107, 109, 121, 152, 156–8, 198–9, 205, 207–8 Binkley, T. 205 Blue Book 10, 37–50, 56, 107, 120, 144, 156, 158–9, 166, 196, 200–1, 203, 211–12 Bouwsma, O. K. 39, 61, 77, 199–200, 203, 207 Braithwaite, M. M. 39, 199 Braithwaite, R. B. 21, 197, 203 Breitenbach, A. 209 Brito, H. 197 Britton, K. 20, 197, 201, 203 Broad, C. D. 121 Brouwer, L. E. J. 55 Brown Book 5, 10, 37–8, 40, 76–7, 85, 160–1, 194, 205; complex 67, 72, 74

222

Index dictation(s) 10, 37, 39–43, 45–7, 49, 65, 87, 181, 196, 213 discourse 9–10, 24, 34, 66, 136–8; grammar of 129; logical and poetic 77; see also grammar; logic(al); poetic disorder 4 dogmatic: preconception 103; procedure 8, 112; statements 203 dogmatism 129, 208; see also rationalism Dolby, D. 212 Drury, M. O’C. 2–3, 6, 9, 19, 61, 193, 197–8, 203, 209

calculus: mathematical 165–6; model of language 12–13, 32, 111, 135–52, 155, 158, 163, 210–11; of logic 108–9, 115–16, 118, 125, 209; see also language; logic(al); mathematical; mathematics cancellations 3, 181; see also amendment(s); corrections; deletions; revision(s) Cann, K. 194 Carnap, R. 54–5, 75, 77, 200–1, 206–7, 209 Carroll, L. 174, 212 Cassian 23, 198 Cavell, S. 59, 203 Christianity 3, 197 Citron, D. 199 Citron, G. 9, 195, 198, 201 Cohen, L. 211 Colli, G. 2 Conant, J. 75, 195, 205–6 consistency 65, 73; proofs in mathematics 166; see also mathematical: proofs; mathematics continental philosophy 77 contradiction 4, 165, 194; hidden 166 conventionalism 93, 103, 109, 111, 113 corrections 3, 13, 45–6, 82, 85, 178, 181–3, 215–16; see also additions; amendment(s); deletions; revision(s); variants Coxeter, D. 39–40, 42–3, 49, 199 creativity 76; creative functions 72; creative ideas 65, 68; creative process 50 crisscross: method or procedure 74–5, 206; nonlinear 72; philosophy 68, 73, 76; relations 71; see also method(s)

Egan, D. 199 Emiliani, A. 209 empiricism 93–4, 100, 103, 109, 111, 113–14, 116–17, 122; see also argument(s): empirical or of metaphysics Engelmann, P. 54–5, 200–1 Erbacher, C. 204 esoteric(ism) 10–11, 52–5, 61–2, 200, 202–3 essence(s) 22, 34, 95–8, 102, 104, 110, 112–13, 142–3, 147–9, 155, 209; of the world 125 evangelism 10, 59, 61–2; evangelical attitude 203 exactness 101–5, 108, 110, 117, 209 exoteric(ism) 10, 52 expression(s) 4, 12, 21, 26, 53, 55, 60–1, 67, 69, 96, 98, 112, 115–18, 158–61, 173, 189, 206, 208, 212; ethical 19, 25; forms of 47; meanings of 130; of logical necessity 113; of norms of representation 125; of sense 99; super-expression 150; see also logic(al): necessity

Davidson, D. 131 de Biasi, P.-M. 196 decontextualization 8 deletions 43; see also amendment(s); cancellations; corrections; revision(s) Descartes, R. 5, 121, 132 dialogue(s) 20, 52, 59, 195; dialogical and constructive 7; see also Philosophical Investigations: dialogical character of the

Feigl, H. 55, 200–1 finitude 5 fluidity 30–2, 34, 104–5, 107, 117; see also indeterminacy, vagueness formless 3 fragment(s) 3–4, 6–8, 53, 67, 71, 74, 178, 180, 182, 204–6, 213–14; see also aphorisms fragmentariness 205 fragmentary (the) 4 fragmentation 3, 5

Index Frazer, J. G. 27–8 Frege, G. 94–5, 110, 150, 165, 204, 208 Förster, E. 1, 193 Forster, P. 211

223

Heal, J. 59–60, 203 Heidegger, M. 1–2, 11 Herrmann, K. 207 Hertz, H. 204 Hilmy, S. S. 205, 207–8, 211 Hodges, A. 201 Hume, D. 121, 210

Gabler, H. W. 207 Gabriel, G. 76–7, 206–7 Gandy, R. 201 Gasking, D. A. T. 57, 201 Gellhaus, A. 207 Genetic Criticism 10 Gerrard, S. 145, 211 gestalt(s) 67–8, 73; gestalt-formation 65, 68; textual 11 Gibson, A. 200, 205 Glock, H. J. 136, 138, 144, 210–11 God 9, 19, 22–3, 25–7, 34, 53, 123, 197–8 Gödel, K. 177 Goodstein, R. L. 39, 45, 199 Gorlée, D. L. 205, 207 grammar(s) 9, 12, 19, 22–5, 29–30, 32– 4, 76, 109, 113, 124–5, 129–31, 136, 143–4, 146, 148–9, 152, 157, 191, 198–9, 209; rules of 130, 212; theology as, is 9, 22–3, 25, 198; see also grammatical; rule(s); theology grammatical 22, 24–6, 28, 30, 32–4, 130–1, 143, 147–9, 208; creation or construction 113; illusion 111; investigation(s) 21–3, 25, 198; joke(s) 198; norm(s) 164; remark(s) 9, 26; rule(s) 24, 113, 143–4, 149, 164, 198; statement(s) 23, 115, 131; see also grammar(s); normativity; rule(s) Grattan-Guinness, I. 202 Green, A. 200 Greve, S. 212 Groag, H. 54 Gründler, H. 205 Guest, G. 9, 195

Kant, I. 1–2, 120–4, 126–32, 209–10; Critique of Pure Reason 55 Kanterian, E. 210 Keicher, P. 205–6 Kenny, A. J. P. 85–6, 207–8 Keynes, J. M. 5, 194, 196 Kienzler, W. 207 Kierkegaard, S. 7, 29, 31, 195, 199 King, J. 20, 197 Klagge, J. C. 10–11, 40–1, 199–203, 207 Kline, M. 212 Knight, H. 39, 199 Koder, R. 12 Krämer, H. J. 200 Kraus, K. 54 Kuusela, O. 12, 208–9

Hacker, P. M. S. 5–6, 12, 22, 136–8, 193–4, 198–9, 206, 209–12, 215 Haeckel, E. 25, 198 Hallett, G. 6, 194 Hamann, J. G. 198–9 Hänsel, L. 121, 193 Hardy, G. H. 194–5

Lafuma, L. 3 Lamb, K. 200 language: acquisition, learning 66–9, 117, 139–40, 157; origin of 194; philosophy of 12, 115, 163; private 131–2, 169; religious 21– 2, 24, 34; see also anthropology:

idol(s) 19, 25–6, 31, 60, 110 implication 109, 168–72, 175; see also inference(s) indeterminacy 30–4; see also fluidity; vagueness inference(s) 105, 125, 127, 130, 146, 150, 165–6, 168, 170–3, 175; inference-rule, rule(s) in, of 116, 124, 129–30, 168, 173, 175; see also implication; rule(s) intelligibility 49, 132 irony 7 Jackson, A. C. 57, 201 Janik, A. 204, 207 Jesus 53 Johannessen, K. S. 207 justification 77, 113, 118, 122, 171, 173, 175

224

Index

of language; calculus: model of language; language-game(s); logic(al): of our language; mathematical: language, picture of language; mechanism: of language; natural: language(s) language-game(s) 12–13, 32–3, 37, 39, 132, 135–42, 144–7, 151–2, 156–7, 159–62, 166, 168–73, 175, 189, 209, 211; method 74; see also method(s) Lazerowitz, M. 56, 201 Lee, D. 20, 59, 121, 197 Le Guern, M. 3, 193 Lemoine, R. E. 55, 200 Lewy, C. 21, 197, 201 linear: form 3, 68; procedure 75; stepby-step writing 204; structuring 67; see also argument(s): discursive-linear strands of; writing linearity 10, 68, 72–4; discursive 70 literary (the) 77; activity 10; character 49; culture 157; practices 3; production 4, 11 literature 76–7 Lobis, U. 207 Locke, J. 127 Lodge, O. 29, 31 logic(al) 12, 24–6, 28–32, 34, 56, 93–118, 121–2, 124–9, 132, 136, 146–50, 152–3, 155, 158, 164–5, 168–74, 204, 208–10; necessity 93, 113–14, 123 (nonlogical), 128; of our language 75–6, 150; philosophy of 6, 12, 93, 96, 98, 100, 103, 106, 114–18; see also discourse: logical and poetic; reasoning: logical Loos, A. 54 Lugg, A. 12–13 Luke 53 Luther, M. 22–3, 25, 198

picture of language 155, 163, 211; practice 170; proof(s) 28, 149, 162, 199; proposition(s) 145, 148–9, 163–4, 209, 212; sciences 122; statement(s) 110, 149; surprise 177; see also calculus: mathematical; consistency: proofs in mathematics; proof(s); rule(s) mathematics 6, 13, 123–4, 138, 141, 145–7, 149–50, 152, 162–6, 169, 211; philosophy of 12, 145, 155, 163–4; pure 121 Maury, A. 13, 197, 200, 213 Mays, W. 41, 200 McGuinness, B. 11, 49, 55, 181, 194–7, 199–203, 207, 211, 213–15 McKitterick, D. 200 mechanism: causal 144; ethereal 150; of language 191; see also language Medina, J. 136 metaphorical (the) 77 metaphysical 110, 112, 125–6, 128–9; see also belief(s): esoteric or metaphysical metaphysics 123–5, 127–9 method(s) 38, 47, 61, 73–4, 93–4, 96, 105, 107, 115, 117, 120–1, 132, 205; mathematical 148; of grammatical elucidation 34; of logic 209; of representation 129; of science, 127; plurality of 7, 75–6, 195, 206; see also crisscross: method or procedure Monk, R. 5, 181, 194, 197, 213 Montinari, M. 2, 193 Moore, A. 199 Moore, G. E. 5, 9, 19–23, 27–8, 37, 74, 131, 194–9, 201, 205, 215 Moreno, A. R. 205, 207, 210 Morris, K. J. 49, 200 Mulhall, S. 199 multilayered: draft 50; philosophical problems 76

machine-as-symbol 148 Malcolm, N. 21, 56, 83, 197–8, 201 Mallard, E. 199 Mark 53 Martineau, R. 39, 199 Maslow, A. 55, 200 mathematical: activity 147–8; algorithms 137; calculation 150; equations and rules 165; investigation 149; language,

names 26, 66, 98–9, 125–6, 140, 145, 155, 208 natural: history 12, 93, 95, 114, 117–18, 147, 214; inclination 7–8; language(s) 96, 99, 102, 105, 140, 157, 209; numbers 147, 160; science(s) 41, 96, 104, 106–7, 122, 209; sequence 7–8, 67, 87 naturalism 94, 117–18

Index Nedo, M. 181, 200, 213 Nietzsche, F. 1–2, 11, 193 nonlinear(ity) 68, 72 Nordmann, A. 41, 199, 207 normativity 136, 158–9, 162–3; see also grammatical: norm(s); rule(s); rule-following Nyman, H. 169, 179, 213–14 Oberdiek, H. 210 openness 11, 66–8, 71 order 2–7, 32, 65, 82, 85; lineardiscursive 70; of the world, of things 111, 125, 210 order(s) 136, 157, 160–4, 212; see also rule(s) Ortner, H. 11, 65, 69–70, 72–3, 76, 204–6 Österman, B. 200, 213 Pascal, B. 2–5, 193 Pascal, F. 213 Pascal, R. 213 past cultural era, 124 Paul, C. K. 193 Paul, D. 1, 7–8, 181, 195, 213, 215 Philipse, H. 210 Philosophical Investigations: as a composite work of art 1, 8–9; as a diary 205; as an album 11, 67–8, 70–72, 75, 205–6; as literary project 1; as unfinished 2–3, 193; dialogical character of the 60, 77; see also dialogue(s): dialogical and constructive Photinus 23 phraseology 47 physiognomy 4, 193 Pichler, A. 11, 197, 200, 204–7 Plato 7, 10, 52, 77, 127, 194; Platonic forms 97; Platonic Ideas 127; Platonism 110; Platonistic interpretation 95, Platonistic point of view 97; Theaetetus 142 Plochmann, G. K. 199 poetic (the) 11, 76–7 poetry 11, 55; poem 77, 85, 191 polyphony 72; see also voice(s) Popper, K. 202–3 proof(s) 28–9, 31, 127–8, 131–3, 147–9, 166, 168, 170–5, 199; see also mathematical: proof(s) psychologism 94; see also antipsychologism

225

psychology 41, 94–7, 99, 114, 116, 208; philosophy of 6, 12 Pythagoras 52, 163 Ramsey F. P. 61, 94, 209 rationalism 122; rationalist dogmatism 5; rationalist tradition 202; see also dogmatism Read, R. 209 reality 8, 12, 97, 102–3, 105–8, 112, 124, 126–31; causality in 9; empirical 94; logical form of 210 reasoning 66, 82, 121, 125–6; logical 76; scientific 29; structured lines of 6 recontextualization 8, 11 Redpath, T. 5, 56, 83, 201 religion 10, 19–24, 26, 28–30, 34, 197; philosophy of 19; see also belief(s): religious (re)presentation(s) 97, 104, 107, 110–13, 121, 124–6, 129–33, 143–5, 147, 161, 163, 187, 204, 211 revision(s) 1, 6, 11, 66, 69, 81–2, 84–8, 162, 181–3, 194, 196, 207–8, 214–15; see also additions; amendment(s); cancellations; corrections; deletions; variants Rhees, R. 5–6, 11, 13, 37–9, 47, 50, 65–6, 74, 85–8, 153, 169, 173–4, 177–8, 180–3, 196–7, 207–8, 213–14 Rizza, D. 209 Robinson, D. 210 Rogers, B. 197, 199, 201 Rothhaupt, J. G. F. 58, 201, 205 rule(s) 22, 24, 32–3, 98–9, 101, 103–8, 112, 115–16, 135–7, 139–44, 149–50, 155–66, 168–9, 171–5, 209–12; BEDMAS and Cosine 163; of representation 124, 129–30; see also grammar(s): rules of; grammatical: rule(s); inference(s): inference-rule, rule(s) in, of; rule-following rule-following 13, 131–3, 138, 150, 155, 157, 160–6, 168, 194, 211–12; see also normativity Russell, B. 13, 37–8, 40, 75, 94, 97, 108–9, 169, 172–3, 175, 194, 196, 202, 206, 208; Principia Mathematica 177

226

Index

Säätelä, S. 207 Savigny, E. von 6, 169, 194, 205 Schleiermacher, F. 7 Schlick, M. 54–6, 201 Schobinger, J. P. 205 Schopenhauer, A. 120, 122, 209 Schroeder, S. 13, 211–12 Schulte, J. 5–7, 11, 13, 49, 169, 181–3, 193–4, 200, 207, 211, 213–16 Scripture 4, 23, 198; see also Bible sense of an author 4 simile 77 Shanker, S. 136, 138 Skinner, F. 39, 41–3, 45–6, 49, 181, 199–200, 205, 213 Smith, D. 207 Smith, J. 10, 180, 195, 213 Smythies, Y. 183, 215 Socrates 212 Somavilla, I. 193 soul 19, 25–6, 28–32, 34, 52–3, 123, 199 Soulez, A. 207, 210 Spengler, O. 53 Sraffa, P. 37, 42, 54 Stenius, E. 204 Stern, D. G. 136, 197, 199, 201, 203, 205, 211 Stevenson, C. L. 39 Stokhof, M. 209 Stonborough, M. 213 Strawson, P. F. 210 Stray, C. 200 Stroud, B. 210 style 37, 47, 49 stylistic: devices 12; features 70; multiplicity 72; polishing 49 symbolism 136–8, 140, 142, 148, 152, 172, 211 systematicity 74; textbook 72 Szeltner, S. 207 Tagore, R. 55 Taylor, J. 60, 202 ter Hark, M. 211 text production 69, 72–3 theology 9, 22–5, 196, 198; rational 123 theory of types 194 Thomas 53

Thompson, J. M. 13 Thouless, D. 40 Thouless, R. 10, 38, 40–1, 45–7, 49–50 Tourneur, Z. 3 Tractatus Logico-Philosophicus 1, 5–6, 11, 50, 54–5, 61–2, 67–8, 70, 75, 77, 82, 93–4, 96–102, 104, 108–11, 113–14, 116–17, 120–2, 125–9, 131, 135, 141–2, 145, 152, 155–6, 158, 168, 194, 200–2, 204, 209–10 Tübingen School 52 Turing, A. 56 under-determinacy 66–8; see also openness vagueness 101–6, 117, 209; see also fluidity; indeterminacy van Hulle, D. 207 van Lambalgen, M. 209 variants 193; diachronic and synchronic (alternatives) 66–7, 204; intralinear 183; see also additions; amendment(s); corrections; revision(s) Venturinha, N. 193–5, 197, 199, 203, 205, 207, 209, 211, 213, 215 Vetter, B. 199 voice(s) 59–60, 203; see also polyphony von Wright, G. H. 2, 7, 13, 40, 42, 54, 82–3, 86, 169, 174, 177–83, 193–7, 199–201, 206, 212–15 Waismann, F. 55–6, 131, 156, 201, 203 Wallgren, T. 213 Wang, J. 207 Watson, W. H. 38, 196–7 Whitehead, A. N. 13, 169; Principia Mathematica 177 Wisdom, J. 197 Wittgenstein, H. 54 Wrigley, M. 145 writing 3, 52, 61, 106, 136; process 216; strategies 11, 65, 204; syncretistic 11, 65, 69–78, 204, 206 Zweig, F. 54 Zweig, M. 54