209 80 627KB
English Pages 150 Year 2010
Eric Lemaire, Jesús Padilla Gálvez (Eds.) Wittgenstein: Issues and Debates
APORIA Apori/a HRSG. VON / EDITED BY Jesús Padilla Gálvez (University of Castilla-La Mancha) Alejandro Tomasini Bassols (National Autonomous University of Mexico) ADVISORY BOARD Pavo Barišić (University of Split) Michel Le Du (Université de Strasbourg) Guillermo Hurtado (National Autonomous University of Mexico) Lorenzo Peña (Spanish National Research Council) Nuno Venturinha (New University of Lisbon) Nicanor Ursua Lezaun (University of the Basque Country) Pablo Quintanilla (Pontifical Catholic University of Peru)
Aporia is a new series devoted to studies in the field of philosophy. Aporia (Aπορία) means philosophical puzzle and the aim of the series is to present contributions by authors who systematically investigate current problems. Aporia (Aπορία) puts special emphasis on the publication of concise arguments on the topics studied. The publication has to contribute to the explanation of current philosophical problem, using a systematic or a historic approach. Contributions should concern relevant philosophical topics and should reflect the ongoing progress of scientific development.
Band 3 / Volume 3
Eric Lemaire, Jesús Padilla Gálvez (Eds.)
Wittgenstein: Issues and Debates
Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.
North and South America by Transaction Books Rutgers University Piscataway, NJ 08854-8042 [email protected] United Kingdom, Ireland, Iceland, Turkey, Malta, Portugal by Gazelle Books Services Limited White Cross Mills Hightown LANCASTER, LA1 4XS [email protected]
Livraison pour la France et la Belgique: Librairie Philosophique J.Vrin 6, place de la Sorbonne; F-75005 PARIS Tel. +33 (0)1 43 54 03 47; Fax +33 (0)1 43 54 48 18 www.vrin.fr
2010 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm www.ontosverlag.com ISBN 978-3-86838-083-5 2010 No part of this book may be reproduced, stored in retrieval systems or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use of the purchaser of the work Printed on acid-free paper FSC-certified (Forest Stewardship Council) This hardcover binding meets the International Library standard Printed in Germany by buch bücher.de
CONTENTS
ERIC LEMAIRE and JESÚS PADILLA GÁLVEZ Wittgenstein: Issues and Debates
7
MICHEL LE DU Tacit Knowledge and Action
11
SABINE PLAUD Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology
31
ERIC LEMAIRE Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
47
AYùEGÜL ÇAKAL What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
65
ALEJANDRO TOMASINI BASSOLS Wittgenstein and the Myth of Hinge Propositions
83
LARS HERTZBERG Hacker on Concepts
117
JESÚS PADILLA GÁLVEZ Wittgenstein’s Criticism against Gödel’s Project of Metalogic
127
Abbreviations
149
Wittgenstein: Issues and Debates Eric LEMAIRE and Jesús PADILLA GÁLVEZ The title of the book refers to the philosopher Wittgenstein, and on issues on which he was working. Through the publication of his writings of the middle period, we have gained a new perspective of both, the Tractatus and the Philosophical Investigations. In summary, one could say that Wittgenstein has corrected some aspects that he had later considered as mistakes in the Tractatus. He was very receptive towards the contemporary discussions and addressed all the relevant issues in philosophy. Consequently, his arguments became more and more sophisticated. The book attempts to present a new view of Wittgenstein’s works. Philosophy can give rise to a certain malaise. On the one hand, in philosophy we ask important or fundamental questions about the nature of human beings, the existence or inexistence of God, values we should follow in our life, the limits of our knowledge, and so on. We want that these questions do not remained unanswered while, on the other hand, as P. van Inwagen notices, that they are no established facts or theories in philosophy, no normal philosophy. Philosophical investigations do not give rise to wide and long-standing consensus among philosophers, contrary to what usually happens in natural sciences or mathematics. Why philosophers never attain definitive solutions to philosophical problems? During the history of philosophy, this fact has given to philosophers (like Hume or Kant or the logical positivists for example) the occasion to raise doubts on our abilities to pursue philosophical knowledge and has received several explanations. One could say that philosophical problems are too difficult for us, that our epistemic equipment is not suitable to solve them. One could as well think that we could attain knowledge if we reform philosophy, its methods, its ambitions, and its object. One could also believe that we should be more patient while others assume this is a brutal characteristic of philosophy itself. Wittgenstein’s answer, which is the foundation of his conception of philosophy, is probably the most disconcerting aspect of his works and maybe the most radical of all answers ever given. According to Wittgenstein, philosophers should not try to attain established theories because philosophical problems are only Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ). Ontos Verlag, Frankfurt a. M., 2010, 7-10.
8
Wittgenstein: Issues and Debates
apparent problems grounded in misuses of language, diseases that analysis has to cure. Philosophy should not be theoretical; it should let everything as it is. Philosophical propositions are pure non-sense. It has give rise to incredulous reactions from distinguished philosophers like Russell or Ramsey. It appears to be strongly opposed to traditional conceptions of the discipline, in which the philosopher aimed at discovering objectively true and widely accepted propositions. It remains until now a very interesting topic, which is at the heart of many works from scholars. But there are numerous disagreements between them about the nature of this conception and its implications for philosophy itself. Is it consistent for a philosopher to hold that philosophical propositions are non-sense? How should we cure our understanding from philosophical diseases? What is the correct methodology to pursue this therapy? Is this cure purely destructive or not? Was Wittgenstein’s practice really faithful to his aims? Was his philosophy really opposed to tradition? Works presented here try to discuss some of these issues. The work of L. Wittgenstein addresses a huge variety of topics. The spectrum ranges from mathematics to the analysis of ethical problems. These issues have generated many important philosophical discussions and the aim of this book is to examine a broad range of philosophical problems. It contains relevant issues and debates on Ludwig Wittgenstein’s writings that make reference to social sciences, morphology, metaphysics, private language and the so-called hinge prepositions. In his article entitles ‘Tacit Knowledge and Action’, Michel Le Du investigates the problems and solutions proposed by L. Wittgenstein in his philosophy of social sciences. M. Le Du analyzes the grammar of the concept of understanding. L. Wittgenstein’s remarks on this topic are scattered in many different places and are therefore difficult to catch in one single grasp. Nevertheless, the author tries to sum up these remarks and examine their implications for the social sciences. The paper presented by Sabine Plaud deals with the issue of synoptic views vs. primal phenomena concerning Wittgenstein’s remarks on Goethe’s Morphology. S. Plaud showed that L. Wittgenstein did not always Goethe the appreciation that he would have deserved for having traced phenomena to their origins. Nevertheless, L. Wittgenstein took up Goethe’s thoughts and consequently tried to re-evaluate Goethe’s approach of primal phenomena so as to make it compatible with his own conceptions. She will focus on a parallelism that might be drawn between
Tacit knowledge and Action
9
the “morphological method” supported by Johann Wolfgang von Goethe in his scientific writings on the one hand, and Wittgenstein’s concern for “synoptic views” on the other hand. Eric Lemaire’s article contains critical remarks on Wittgenstein’s anti-metaphysical readings and asks whether Wittgenstein had actually destroyed metaphysics. The author argues that if this were true it has to be made clear that Wittgenstein’s philosophy cannot actually be qualified as metaphysics. Moreover, he had underlined that metaphysical propositions are nonsensical. By examining Wittgenstein’s texts, the author raises two main questions, first whether there is an anti-metaphysical reading of the second Wittgenstein. Second, E. Lemaire asks whether Wittgenstein’s works actually offer a plausible explanation for the assumption that metaphysical propositions are nonsense. Ayşegül Çakal enquires what the repudiation of private language means in Wittgenstein’s Philosophy. By challenging, the dichotomy of “inner” and “outer” in the Tractatus Wittgenstein rejects private language by means of reducing it to something meaningless. However, in the Investigations Wittgenstein expelled meaningfulness from mere internal to external by eliminating the possibility of an internal way of grasping the meaning of linguistic expressions. According to his view, the meaning of a linguistic expression can thus only be learned by an ‘act of meaning’ within the society in which one lives through its use. Therefore, the radical change in Wittgenstein’s position on meaning can be found in his private language argument. She in this paper, I will examine his rejection of the traditional distinction between reality and language and evince the effect of it to his later understanding of meaning in language. To this aim, she will first give a background of his understanding of language and meaning in general. A. Çakal present the private language argument in parallel with its effects on his later understanding of meaning in order to be able to see how it renders private (internal) ostensive definition unintelligible and why it is a rejection of identifying referential and representational attitudes and the way of asking questions of traditional approaches. Alejandro Tomasini Bassols focuses on Wittgenstein and the myth of the so-called ‘hinge propositions’. From the point of view of the history of ideas, the effects of this myth just cannot be ignored. Indeed, based on this myth it now has become a sort of unavoidable fashion to speak of a “third Wittgenstein”, a thinker who apparently would have superseded or overcome his own magnum opus, i.e., the Philosophical Investigations. A.
10
Wittgenstein: Issues and Debates
Tomasini Bassols starts out from the following point: if we contemplate his work as a whole, how many Wittgensteins is it reasonable to think there are? His paper is devoted to answering this question. Lars Hertzberg focuses on P.M.S. Hacker’s thoughts on concepts and argues that it is through speakers’ responses to utterances that distinctions between correct and incorrect enter into language. The misunderstanding we need to guard against is that we could base a description of correct use on neutral observations of the linguistic behaviour of the members of a speech community. We need what Hacker calls an internal point of view. However, if that is granted Quine’s reason for questioning the analytic/synthetic distinction dissolves. On the other hand, the author has some disagreements with the way Hacker construes that point of view. Jesús Padilla Gálvez discusses Wittgenstein’s criticism against Gödel’s project of metalogic. The author makes reference to a chapter in the Big Typescript entitled ‘Metamathematics does not exist’ and that presented a criticism of Gödel’s metalogic point of view. The author examines the frame within which the meta-mathematical discussion took place in Vienna at that time. He wants to analyse the historical background of the meta-mathematical discussion in Vienna. He will deal with the purpose of the calculus and he will try to define what a rule is. Finally, he will focus on the problem of completeness and internal consistency in Wittgenstein’s critical arguments against Gödel’s proposal. Wittgenstein’s arguments create a new point of view in this field as he has shown that many equivocal mathematical results originate from being based on ambiguous philosophical terminology. In this book, the editors Eric Lemaire and Jesús Padilla Gálvez have undertaken to sum up the papers presented at the Congress on Wittgenstein ‘Philosophie et pratique de la philosophie’ in Nancy in 2007. Additionally, some papers from specialist on Wittgenstein were included in the publication. We would like to thank all those colleagues who accepted our invitation to both participate in the congress and to contribute to the book. We are indebted to the public institutions that have financially supported the congress. On this occasion, we benefited not only from the continued and generous support of the Université Nancy, the Archives Henri Poincaré (UMR 7117), the Conseil général de Meurthe et Moselle, the Conseil scientifique de l’Université Nancy 2, the UFR Connaissance de l’Homme, the University of Castilla-La Mancha and the Faculty of Law and Social Sciences in Toledo.
Tacit knowledge and Action Michel LE DU Introduction In this paper, I try to estimate the importance and relevance of a wittgensteinian legacy (themes, problems and perhaps solutions) in recent philosophy of social sciences. The general idea underlying my talk is that this legacy is never as visible as when the concept of understanding is beeing questioned. In other words, my purpose here is to examine the grammar of this concept ; however, Wittgenstein’s remarks on this topic are scattered in many different places and difficult to catch in one single grasp and that’s the reason why I shall try to apprehend them through their consequences for social sciences. During the last decades, a shared point of view (foreign to Wittgenstein’s influence and resulting in fact from debates internal to the social sciences) has consisted in saying that the structures govern the actions and are, at the same time, generated by the actions. Different versions of this point of view can be seen in the works of Pierre Bourdieu, Anthony Giddens and many others.1 I wish to label this theoretical schema the double status of structures and to question its relevance. The main difficulty, in so doing, comes from the fact that the concept of structure has been intrepreted, with the passing times, in very different ways. So, one need first to clarify its use. As I suscribe to a form of methodological separatism between social an natural sciences, one way to undertake this clarification is to make explicit the differences in the way the concept is used in the two domains. When one speak of a physical structure (the structure of a water molecule, for instance), one has normally in mind a system of objective relations which can’t be seen by the lay man: to get to know such a structure is one and the same thing as to get to know the molecule. In the social sciences, on the other hand, the concept of structure is applied to a universe filled with agents who have elaborated their own understanding of it and act in accordance with this understanding. In addition, the understanding of his environment by the social agent, apart 1
See, for instance: Bourdieu, 1979; Bourdieu, 1987; Giddens, 1979; Giddens, 1986.
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ). Ontos Verlag, Frankfurt a. M., 2010, 11-30.
12
Wittgenstein: Issues and Debates
from the fact it has its own legitimacy, is very well exercised without the help of the stuctural explanation of the same topic a sociologist can supply. That’s why it’s simply impossible to identify the knowledge of the social world with the knowledge of such an explanation: on the contrary, the concept of stucture, when applied to the social world, has to be combined with an account of the agents’ spontaneous understanding and of the rules they follow. Moreover, in the natural sciences one might legitimitely waver between a realistic interpretation of structures, conceiving them as spatiotemporal realities, as the framework (endowed with causal powers) of the observed phenomenas and an anti-realistic interpretation, seeing them as abstract characteristics. One of the aspects of the thesis defended in this paper consists in saying that, in the social sciences, it’s almost impossible to substantiate the realistic interpretation because it doesn’t fit well the idea that agents have intentions and understand their own actions. In other words, I do agree with Giddens when he insists that structures are abstract features. Let us take a linguistic example. The present state of a language (English, French) can be refered to as a structure: such a characterization implies that, in the description of the linguistic material, one doesn’t take into account the idiosyncrasies, the tendency of a few speakers to set themself apart or even to secede, in order to concentrate on the convergences and the big macroscopic regularities that appear as the balances of a language’s life. The speaker’s largely implicit linguisitic capacity can then be described as a rather confused knowledge of those regularities and his practice as their continuation. Such a continuation supposes the speaker to have, through his tacit linguistic knowledge, an access to the normativity of the linguistic institution. So this paper will be largely dedicated to the examination of this access to social normativity and that’s why I’ll not be able to save myself a grammatical examination of the concept of understanding: it’s both important and illuminating to specify what is meant by phrases saying that an agent understands a rule or a social norm. This is also the point where our argumentation will separate from Giddens’, both because his exegesis of Wittgenstein’s thought on rule is inaccurate and because it inclines him to a misleading interpretation of what an implicit rule is supposed to be. The general idea I’ll defend on that topic will be that one needs to free oneself from the belief that an implicit rule is a rule which is subject to an implicit grasp: normally, a rule is implicit precisely because it doesn’t have to be grasped anymore. The regularities we alluded to need, in order to be
Tacit knowledge and Action
13
perpetuated, the intervention of the agent’s actions (they must not be identified with forces carrying weight on them and bending their will) but these normed actions don’t need to be initiated by a specific mental act or event. 1.
Giddens on rules
In a chapter of his book Social theory and modern sociology, Giddens notices that the concept of structure has elicited very few comments from the english speaking sociologists and philosophers (in this respect, the fate of the concept of function has been quite different). He comments upon that point in the following way: The reason is probably that most English-speaking social scientists have a clear idea of how the concept of structure should be understood. When they talk of structure, or of the “structural properties of institutions”, they have in mind a sort of visual analogy. They see the structural properties of institutions as like the girders of a building, or the anatomy of a body.2
In such a perspective, the structures appear as patterns or types of relations one can oberve in different social contexts and the concept of structure seems to be naturally connected with what has been called the “objectivist” approach in social sciences. As Giddens emphasizes, the difficulty raised by such a notion is that “the structure then appears as a constraint which is ‘external’ to action.” 3 Giddens’ own conception of structures is set out as an alternative to the interpretation in terms of constraint. But precisely because this is his aim, his argument also has to include the idea that, unlike the patterns we have just been speaking of, the structures are abstract features with no location in space and time. So, according to this last interpretation (1) societies, institutions have abstract structural proprerties resulting from the ongoing of human action (2) agents are able to proceed in their everyday business in virtue of their capability to instantiate such structural properties. In other words, those properties have no action on the agents (unlike the wind who has an action on the tree to the point of bending it): on the contrary, they are abstract descriptions of the result of convergent “upstream” actions by numerous agents. But they are perpetuated only because those very same agents act spontaneously in a rule-guided way. This is why the agents, seen 2 3
See: Giddens, 1984, pp. 52-72 and especially p. 60. Giddens, 1984, p. 61.
14
Wittgenstein: Issues and Debates
collectively, can be said to “instantiate structures” in their interactions. If this line of argumentation confirms our hopes, the articulation between agency and stuctures seems to involve a circle (a virtuous, not a vicious one, indeed) in the close of which the capacities of the agents have a keyrole. However, as the structures so conceived are identified with abstract properties, the notions of necessity and causality do not play any part in this account (and this is the main difference with the conception of structures as patterns). Giddens’ conception is insightful enough to avoid undesirable reifications, but involves, as has already been said, a few confusions of its own, most of them having to do with his interpretation of Wittgenstein. We are now going to study this interpretation for itself, in the first place through an examination of the relations between Giddens and a major author both of the wittgensteinian tradition and of the philosophy of social sciences, the british philosopher Peter Winch. 4 This study of Giddens’ views will lead us to a direct examination of the notion of disposition. Our goal will not be to reconstruct an historical sequence of the discussions about social sciences in british philosophy, but to underline several permanent epistemological issues. We will concentrate on them, not on the history of ideas. 2.
Giddens and Winch
According to Giddens, the structures, understood as abstract features, are corollaries of the scientific outlook. As such, they might be seen as the result of action as well as governing agency (in so far as “governing” is not taken in a causal sense). The structures are not, strictly speaking, parts of the implicit knowledge mastered by the agents, unlike the rules. This reference to rules is where a wittgensteinian legacy is visible in Giddens’ writings. As we will see, this very same legacy is understood differently by
4
As one knows, Peter Winch’s controversial essay (Winch, 1958) has become a wittgensteinian locus classicus on the erklären/verstehen distinction. Winch defends a form of methodological separatism partly on the basis of arguments inspired by the author of the Philosophical Investigations (Winch’s other main source of inspiration, as one can guess from the title of his book, is Collingwood’s Idea of history). Another important essay by Winch on kindred topics is “Understanding a primitive society” in: Winch, 1964.
Tacit knowledge and Action
15
Theodore Schatzki.5 This difference deserves an explanation, but we need first to make explicit what this heritage consists in. In the paper we have already mentionned, Giddens emphasizes several times the traditional opposition between the “hermeneutic” and the “objectivist” perspectives, and one can see very clearly that he favors the first one. He underlines, by the way, that phenomenologists, oxonians and Wittgenstein are, among philosophers, those who have most insisted that a kind of mundane description was unavoidable and have taken a direction he labels “sociological” in philosophy, whose main feature is an effort to reconquer what matters for daily routines and mundane life. 6 He characterizes more specifically this redirection by its reevaluation of common sense, which should not be disparaged as mere inertia and habit. When he deals with this topic, Giddens favors the expression mutual knowledge. Infortunately, this terminology is puzzling, because it might make one think of a knowledge A has about B and B about A. In reality, what he has in mind is a knowledge of rules and social conventions agents must have in common if they are to give a meaning to what they do, as well as to what others do. This mutual knowledge is involved in what he calls the “stretching” of everyday activities. It has to do, he says, with ... the methods used by the lay actors to generate the practices which are constitutive of the tissue of social life.7
These methods define the skilled and knowledgeable agent, whose activities are entangled in everyday social life and whose knowledge is discernable through his practices (indeed, this form of knowledge is not, most of the time, directly available for a linguistic mastery). In this context, the first idea Giddens borrows from Winch is that we need to take into account the understanding agents have of what they do and, moreover, the idea that this understanding is not contingently related to what they undertake but is constitutive of this undertaking. Let us pay attention to an example from Winch: “My behavior is governed ... by my concept of myself as a member of a belligerent country. The concept of war belongs essentially to my behavior. But the concept of gravity does not belong essentially to the behavior of the
5
See: Schatzki, 1992, p. 280-295. Giddens, 1984, p. 65. 7 Giddens, 1984, p. 66. 6
16
Wittgenstein: Issues and Debates falling apple in the same way ; it belongs rather to the physicist’s explanation of the apple behavior.”8
This famous passage is both convincing and unsatisfactory. The unsatisfactory point is that it seems to suggest that one must be able to apply the concept of war to oneself in order to participate in a war. It’s perfectly clear, indeed, that a soldier wearing a uniform would necessarily justify some, at least, of his actions by relating them to this concept, but a situation might easily be imagined that one would feel inclined to describe “from the outside” as a war situation even if the belligerent agents showed reluctance to describe themselves as warriors. So, the exact idea seems to be that when a relation does exist between the concept of a member of a belligerent country and the behavior of a combatant, then this relation is internal (in an event such as an apple fall, on the contrary, we are confronted to an external relation between the fall and the concept of gravitation).9 Winch has such an internal relation in mind when he makes use of the phrase behavior governed by a concept. What his argument eventually shows, is that if you are a member of a belligerent country and have the corresponding concept at your disposal, then this concept essentially participates in your behavior. Giddens draws from the previous considerations the consequence that the notion of a mutual knowledge should be extended to the relation
8
Cf.: Winch, 1958, pp. 127-128. The notion of internal relation is of an utmost importance here and requires explanation.Let us take an elementary psychological example. Wittgenstein’s idea consists (unlike the mentalist’s) in saying that pain is not only a subjective event and (unlike the behaviorist’s) in saying that it can’t be identified with the pain behavior: normally, the pain behavior is part of the pain, in other words to have pain expressions is part of what is meant by the phrase to be in pain. Accordingly, the pain behavior is a criteria of pain and the relation between the two is internal. Of course, counterexamples are always imaginable (a feigned pain played by a pretender ; a stoïcal guy curbing all his pain expressions), but without such a criteria, the learning of the concept of pain would simply be impossible. Winch invites us to establish a similar connexion (though more complex) between x’s behavior as a soldier and x’s possession of the concept of a soldier: x’s behavior as a soldier is the criteria that leads us to attribute him not only the concept of a soldier but the fact that he applies to himself this very same concept and justify himself with its help. This relation of justification, in turn, leads us to the conclusion that the concept of a soldier is part of the soldier’s behavior and, in more general terms, to the idea that having the concept of a soldier is part of what it is to be a soldier. 9
Tacit knowledge and Action
17
between the agent and the observer, but the way he puts things remains somewhat vague. “... in order to generate valid descriptions of social life, the sociological observer must employ the same elements of mutual knowledge used by the participants ...”10
The difficulties are indeed concentrated in the phrase “the same elements of mutual knowledge” which has to be taken cum grano salis. In fact, neither Winch, nor Giddens really think that sociologists are committed to limit themselves, in the course of doing their job, to concepts that social agents employ on their own initiative. In order to figure this out, the best thing to two is to pick out two examples from Winch. The first of those two examples illustrates the situation in which the social actor and the observer do share the very same concept ; the second one illustrates the situation in which the actor’s behavior shows his mastery of concept(s) connected to the one (or those) used by the observer. When confronted to such a case, one is forced to specify the nature of this connection. Example 1: an observer says of agent N that he has voted for the Labor Party because he thought it would be able to preserve the industrial peace. The strength of such an explanation lies in the fact that the concepts it uses have to be grasped by N as well.11 Example 2: an observer explains the behavior of a group of businessmen by using the keynesian concept of liquidity preference. Such a concept doesn’t belong to the set of cognitive tools mastered by those businessmen ; however, Winch notices, “... it is logically tied to concepts which do enter into business activity, for its use by the economist presupposes his understanding of what it is to conduct a business, which in turn involves an understanding of such business concepts as money, profit, cost, risk etc. It is only the relation between his account and these concepts which makes it an account of economic activity as opposed, say, to a piece of theology.”12
Those two examples are set out in two different sections of the book and illustrate in two different ways the idea according to which the agent and the observer share the same elements of mutual knowledge. As far as I know, very few commentators have paid attention to this difference. Its significance will become clear if we start from the following remark: if the explanation by the observer of the agent’s behavior with the help of a concept is to be valid, it’s not enough for the agent to have the capacity to 10
Giddens, 1984, p. 66. See: Winch, 1958, p. 46. 12 Winch, 1958, p. 89. 11
18
Wittgenstein: Issues and Debates
acquire the concept in question. By saying he has such a capacity one simply says that it’s not impossible for him to learn it. In the two previous examples, the agents do not simply have the possibility to learn the concept. They have something more, and it’s important to us to discover, in each of these two cases, what this something more consists in. In the first example, the agent masters indeed the concept discussed (industrial peace) and is theorically able to use it if, for instance, he is asked to justify his vote. One can see in this mastery a kind of capability, and, when using this word, we normally denote more than a mere capacity: indeed the fact that it’s up to the agent to speak in terms of such a concept, even if he might, for some reason, decide not to do so. And it’s because the agent is supposedly able to express such a concept that it is thought to be part of the mutual knowledge he shares with the observer. In the second example, the most important point is that the expression liquidity preference, belongs to the technical terminology of economy and not to the businessmen’s vocabulary. 13 Accordingly, they tend, at best, to hover round this notion when they are asked to explain what they do, but do not achieve to express it. Nevertheless, when supplied with the appropriate expression, they would say they agree with the description of their reasons to act couched in such words. Such an agreement is normally accompanied by characteristic expressions of approval (“this is what I was trying to say all along” ; “you’ve said it better than I could” etc.). 14 This situation illustrates the way the phrase logically related is to be understood in the present context: the agents do not use the concerned expression when they speak spontaneously of their activities, but avow they would have used it, if only they had come to know it before.15 Accordingly, an observer is fully 13
The liquidity preference, as one knows, is inversely proportional to the interest rate. If interest rates are low, the economic agents, expecting better days, will keep their money and, accordingly, the liquidity preference will be maximum ; on the contrary, if interest rates becomes high, the liquidity preference diminishes because making investments becomes the best thing to do with one’s money . 14 This situation is, of course, quite different from the case of a discussion partner helping a speaker to find a word he has on the tip of his tongue. In such a case, the speaker does have the capability to use the word he is looking for, but this capability is temporarily impeded by an external factor (tiredness, absent-mindedness etc.). 15 Some of the most important reflexions on the notion of awoval in the context of Wittgenstein’s thought can be found in his Bemerkungen über die Philosophie der Psychologie, see for instance Wittgenstein, 1984, vol. 1, § 836 and vol. 2, § 63. His idea, roughly speaking, is that we are not, as far as our own mental states are concerned, in the position of an very privileged witness: we do not get an information
Tacit knowledge and Action
19
justified to sum up, with the help of a single concept, what the agents express by using a plurality of concepts. It’s also worth noticing that the wittgensteinian concept of internal relation plays indeed to different roles in the previous observations. First, it occurs between the agent’s behavior and the concept said to be “essential” to it ; secondly it occurs between the concept used by the agent and the concept used by the observer. In this second case, we are confronted to the agent’s propension to employ such a concept (when asked to jusfify his demeanour) rather than to his capability to do so. In fact, the word “propension” is appropriate here because without an external help, the agent remains unable to achieve a full mastery of a concept which happens to be used, in other respects, to describe his deeds. Obviously, a speaker who only has the propension to adopt a way of speaking doesn’t necessarily have its full mastery. He might need, to say the least, a little help to gain this full mastery, and such an help might well never occur. The small conversational dramas Pierre Bourdieu is prone to describe very often involve a divorce between the agent’s own capability to speak in a certain way and his propension (in specific social situations) to speak in another way he takes to be more gratifying.16 Unfortunately, Bourdieu has obscured this interesting observation by labelling habitus sometimes the speaker’s competence to talk in a certain way, sometimes his propension to adopt norms of expression he cannot fit and sometimes the two things simultaneously. So far, the conclusion of the story seems to be that the assertion according to which a sociological explanation has to entertain an internal relation with the one the agent could give doesn’t correspond to an unique kind of situation because the modal verb “can” in the present context, covers different kinds of situations. The previous conclusion is a good reason to give a maximum extension to the concept of (which could happen to be false) about our own mind’s content. This is where his idea that what he calls psychological verbs (to belief, to think, to understand etc.) in the singular, first person, present tense are expressions (Ausserung, Ausdrück) comes from. He also thinks that those very same verbs have a descriptive use as soon as they are employed at other tenses and / or other persons. In other words, the avowal, by an agent, of his reason to act in a certain way or to say a certain thing is not the result of a cognitive contact he entertains with his own mind, but an expression. Becoming aware that one’s own acts proceed from a certain reason has nothing to do with forming an hypothese. The concept of awoval is, of course, also a central feature in Wittgenstein’s reflections on Freud. See Wittgenstein, 1972. 16 See: Bourdieu, 1980. For a full examination of the relations between capacity, capability and propensity, see; Scheffler, 1983.
20
Wittgenstein: Issues and Debates
mutual knowledge: it would be impoverishing to identify the so-called double hermeneutics posture17, that the two authors we discussed take to be characteristic of the social sciences, with the idea that only concepts the agents master fit for explanatory purpose.18 3.
Understanding the rules: Giddens reading Wittgenstein
The discussion concerning the way the continuity of human action is to be explained, has clearly its origin in Wittgenstein’s seminal remarks on rulefollowing in the Philosophical Investigations. Nevertheless, the now common assertion according to which our relation to rules is basically a practical one turns out to be, after a careful scrutiny of Wittgenstein’s pronouncements, the superimposition of two clearly different ideas. Let us start with the best known of those two ideas, even if the less recognized is the most important to us in the present context. Wittgenstein has established a celebrated distinction betweenone believing than he is following a rule and one following a rule. In connection with this distinction, he also has suggested a comparison between following a rule and obeying an order.19 The central problem involved in this comparison is the one raised by the transition between the general formulation of the rule 17
Sociologists interpret a world which already contains interpretations. Most probably, the reconciliation with common sense wished by the thinkers we have been discussing cannot be summarized by the idea that social scientists have to get back to concepts the agents master, or whose explanatory use they might approve. A vast range of investigations in recent cognitive psychology is dedicated to the task of explaining the efficiency and rapidity of common sense thought. In the course of doing so, those investigations tap notions such as mental model or prototype, sometimes echoing Wittgenstein’s own reflections on the blurredness of numerous everyday concepts which prove to be unfit for definitions in terms of necessary and sufficient conditions (one might think, for instance, of Eleanor Rosch’s seminal work on categorization). It’s clear that prototype is anhypothetical concept (= a prototype is attributed to a speaker on the basis of tests he is subjected to). When A explains B ‘s demeanour by saying that a prototype structures its thought, A doesn’t normally expect B’s approval (unlike the economist in the liquidity preference example). B might eventually say that A hypothesis is insightfull, but such an approval is not an avowal: first person authority is relevant in the businessmen case, not in B’s case. The moral of this story seems to be that one needs to distinguish the descriptions of the common world mobilizing elements from our mutual knowledge and descriptions of the common world which are, indeed, theoretical descriptions of this knowledge. Obviously, the concept of prototype has to do with the second kind of undertaking. 19 Cf.: Wittgenstein, PU, 1984, § 206. 18
Tacit knowledge and Action
21
and its application in particular cases. The agent receives the rule just like he receives an order, at least if he isn’t already in possession of the text of the rule, and is supposed to “translate” this text into a determinate action. The idea, cherished by Wittgenstein, that following a rule is a practice contrasts with the hypothesis of an subjective act of interpretation, thought as responsible for the transition between the pronouncement of the rule and its application, and supposed to be the place where the criterium separating correct application from infringement is located. If someone suscribes to such a subjectivist mythology, he finds himself deprived of all means to distinguish an effective application of the rule from the mere thought that he is following the rule. The second, and less noticed idea on this very same topic consists in the recognition of the fact that, in most cases, the rules are implicit in our practice: they remain unformulated. In this respect, the problem is not the one raised by the transition between the general rule and a particular action, but the question of the explicitation. In other words, how does the transition occurs between a rule inscribed in the agent’s tacit knowledge, of which he doesn’t currently think (and of which he might even never have effectively thought of) and the very same rule verbalized (or even interpreted) ? Giddens is prone to assert that most rules are unformulated, but he also maintains that every formulation of the rule is an interpretation. 20 When he says that social agents “draw on rules”, he certainly doesn’t mean that they follow, consciously or unconsciously, already formulated rules.21 He wants to say that their actions are governed by a “tacit grasp” of unformulated rules. The word “grasp” seems to denote an actuality, a psychological act or event -and this should alert us. Should not we say, on the contrary, that an implicit rule is not subject to any grasp instead of saying it’s subject to an implicit one ? The distinctive feature of a familiar rule is that it does not requires from the applicator, when he proceeds to a new application, an act of recognition anymore, though the way he acts gives us evidence that he has knowledge of the rule. The object of my argumentation in the rest of this paper can be deduced from the content of the previous remarks: I shall try to show that it is impossible to give any substantial content to the idea of a “tacit grasp” as it is used by Giddens.
20 21
See: Giddens, 1986, chap. 1. See on this point Schatzki’s remarks (Schatzki, 1992, p. 291).
22
Wittgenstein: Issues and Debates
A famous passage of the Philosophical Investigations questions the possible senses of the phrase the rule by which he proceeds.22 Wittgenstein is interested in this passage, among other things, by the situation in which the agent consults the rule. His goal is obviously to show that the ruleconsultation model cannot be generalized: we most of the time apply rules without prior consultation.This model however allows us an acceptable (but banal) interpretation of the phrase “tacit grasp of the rule”: one can proceed to the consultation in foro interno, without the rule being uttered or read anywhere. However, in Giddens hypothesis, the grasp is supposed to be about unformulated rules and, for that reason, doesn’t fit the tacit rule-consultation model because, in this model, some kind of relation with the text of the rule is still required from the agent, even if this relation is entertained in an internal way (the best description of such a situation might consist in saying that the grasp is both tacit and explicit). In fact, an other case might legitimately be described in terms of an “implicit grasp”: when the agent tends to express the rule but doesn’t have the appropriate linguistic resources, or simply enough time, to achieve this task. As this last situation is obviously not what Giddens has in mind either, the conclusion is that his notion of a tacit grasp of the rule remains unclear and mysterious. In other words, unlike the notion of an implicit rule, which is understandable and relevant far beyond the previous cases, the idea of a tacit grasp results in a muddle as soon as it is employed beyond those same cases. Our next task will be to make explicit the ins and out of this muddle. The confusion we are dealing with here has much to do with the one between the “grammar of conscious events” and the grammar of “dispositions” 23 since (1) we are supposed to be confronted to an analogon of a conscious grasp of the rule, minus the formulation of this rule (2) the hypothesis of such a mental token dissimulates the fact that our ordinary knowledge of rules (as soon as it is not actualized through the process of consultation) is dispositional in nature and, accordingly, unlike an act or an event, not open to clock-check.24 In order to fully clear-up this confusion, we need to explicate a little more the concept of disposition. In the Philosophical Grammar25, Wittgenstein asserts that there are two different interpretations to the assertion that one must understand a proposition in 22
Wittgenstein, PU, 1984, § 82. Wittgenstein, PU, 1984, § 149. 24 Wittgenstein often says that it’s pointless to ask someone when he is able to play chess. 25 Cf.: Wittgenstein, PG, 1984, 1, § 8. 23
Tacit knowledge and Action
23
order to act according to it. The first of those two interpretations is chronic (= understanding is an act or an event entertaining a temporal relation with action) ; the second interpretation is grammatical. The first interpretation sees understanding as an event numerically distinct from the action. On the contrary, the second interpretation sees understanding as part of the action. This is eventually what was involved in the industrial peace example: Winch’s idea was certainly not to say that this notion, essential to the agent’s behavior, was to cross his mind at the moment he casts his vote. One way to draw the consequences of the second interpretation is to notice that the word “actualization” in the phrase actualization of a capacity might in turn be taken in to differents ways. (1) A capacity (such as the understanding a concept) can be actualized in the Erlebnis of the speaker (as when one sees a halo of connexions between the word expressing the concept and other elements of our linguistic universe and accordingly comes to see this same word under an aspect). 26 (2) A capacity can be actualized in the agent’s behavior, for instance through the fact that the speaker uses an expression in an appropriate way. Obviously, the linguistic behavior is, to some extent, part of the subject’s Erlebnis, but the important point, in this last case, is that no conscious state, no mental event intervenes, strictly speaking, in the understanding. And that’s why the only thing to say is that the understanding is part of the action. It must be noted that on a few occasions, Wittgenstein has said that the verb to understand could be used to mean a mental event (as when one speaks of a sudden insight), but this illustrates in the first place the vagueness of the concept. Anyway, the extension of this concept, taken as a whole, has not to be modelled on this specific portion. 27 The conclusion is that, most of the time, the understanding of a proposition according to which one acts is not an event concomitant with the action, and we eventually are to think of form of understanding which is not a grasp and this is where the notion of disposition enters the scene.28 In this perspective, the main criticism to be adressed to Giddens is not the imprecision of his reading of Wittgenstein, but the fact that he invents a syncretic mixture of mental tokens and dispositions. But even the recognition of the dispositional nature of the 26
Cf.: Wittgenstein, BPP, 1984, vol. 1, § 192-202. See: Wittgenstein, The Cambridge Lectures 32-35, 1979, p. 114. 28 One must remember that the word disposition is to be understood in two different ways. Some dispositions are such as their actualization is automatic (like beeing irritable). Others need to be exercized (our knowledge of elementary arithmetics, for instance). Dispositions who need to be exercized are often called capacities. 27
24
Wittgenstein: Issues and Debates
implicit understanding of rules is a conceptual progress, it doesn’t shed light on the other side ot the problem we are interested in, the relation between this understanding and the continuity of action. And this what the last section of this paper will be dedicated to. 4.
To follow a rule and to follow a concept
In fact, very few activities in our everyday life can be said to proceed from the ongoing application of the same rule, or of the same body of rules. In a paper I have already mentioned, Schatzki refers to several remarks on language by Wittgenstein in order to back up this view.29 Understanding a concept, he says, is not identical with understanding a rule. Many of our concepts are blur-edged and made of family resemblances, in other words of multiple threads among which multiple transitions can be found. Even if the speaker is able to give afterwards a rule, in order to justify himself, this is not a sufficient basis to say that to understand this concept and to understand this rule are one and the same thing. In fact, No set of rules ... formulated either by an actor or by an observer on the basis of a finite number of uses of a concept, will be able pre facto to cover all possible acceptable uses.30
Each time a speaker discovers a new application of a concept or extends willingly its use, he might justify this additional extension by a new rule. But one cannot expect from a rule that it explains both the new application of the concept and the continuity of its use along time. Again, the notion of a tacit grasp contributes to occult this difference and Schatzki conludes that Giddens “errs in equating knowing how to go on with knowing a rule.” 31 After reading this analysis, however, one feels like asking the following question: the concept of rule not beeing univocal, would not Wittgenstein have accepted to include in it what Schatzki, in his criticism, calls “concept” and opposes to rules ? My answer to that question is “yes”. Giddens is wrong because, in spite of the fact that he foresees the importance of implicit rules, he persists, with his idea of a tacit grasp, in conceiving them through the prism supplied by the explicitly grasped rules. But Schatski is also wrong when he restricts the notion of 29
Schatzki, 1992, p. 290-293. The paper quotes a passage from The blue book & the brown book, where Wittgenstein dicusses the notion of a general idea. See: Wittgenstein, 1958, p. 18. 30 See: Schatzki, 1992, p. 291. 31 Schatzki, 1992, p. 291.
Tacit knowledge and Action
25
rule either to the idea of a formula supposed to cover all the possible cases or to the justification after the event. By so doing, he occults (1) that what he calls “concept” has its place inside the spectrum of different things we call “rules” (2) that the dispositional notion of implicit rule here mobilized is the best safeguard one can find against the conceptual syncretism illustrated by the notion of tacit grasp. Blur-edged concepts are windows opened on what we try to grasp through the notion of “implicit rule”. More precisely, if one pays attention to everyday’s langage’s life, one can easily see that a concept can be used by speakers who prove unable to interpret it (= to establish synonymy relations in which this concept enters, to express what it “means”).32 The speaker might well have the know-how allowing him to apply correctly a concept and be deprived of the power to give the rules justifying this very same application.33 The rule, in such a case, is implicit not in the (weak) sense of being expressible though not expressed, but in the (strong) sense of not being expressed because the speaker is deprived of the power to express it. X can apply correctly the concept C 1 and the concept C 2, and can be, at the same time, unable to give the rule R distinguishing the contexts in which C 1 can be used and not C 2 and vice versa. Advancing such a rule would be interpreting C 1. X’s cognitive performances here are in relation with R, but does not proceed by an application of R. John Searle has noticed that many institutional rules are like the one distinguishing C 1 from C 2. He explains that we are endowed with abilities which make us sensitive to the normativity inherent in the institutions.34 However, those abilities are not on an equal footing with conceptual capacities we have been previously talked about. The main reason is that the practical knowledge Searle is dealing with has to be figured out as functionally equivalent to complete statements, permitting of forbiding things. In 32
I agree with Quine’s idea that in asserting what a term “means”, one doesn’t refer to an entity: there are no substantive meanings. In fact, when I explain to someone what a word or a phrase means, I reformulate it in terms I think to be more accessible to my listener. Between the two expressions a sameness of meaning does exist, but this does not mean, to repeat myself, that they express the same abstract entity or share the same attribute. See: White, 2002, chap. 6 (for an illuminating state of the art on this topic) and Quine, 1953, pp. 1-19 (especially p. 11). 33 The distinction between tun and können is recurrent in Wittgenstein’s remarks on psychology. For an application of this distinction to calculation, see: Wittgenstein, BPP, 1984, vol. 1 §§ 651, 655, 735. On being aware of one’s lie as a können, see § 263. 34 See: Searle, 1996, chap. 6.
26
Wittgenstein: Issues and Debates
addition to that, in the case of concepts, even if a speaker is unable to interpret them, he is able to term them, whereas the implicit rules, in the strong sense, whose normativity is often illustrated by the functioning of our institutions, are, by definition, to remain not only uninterpreted, but unformulated. Besides, Giddens’ identification between formulation of the rules and interpretation of the rules is puzzling. Such an identification is understandable only if one hypothesizes a grasp directed torwards unformulated rules, and compared to which the mere formulations appears as a interpretation. We have seen what one must think about such an idea. In fact, there is interpretation only when a formulation is subsituted for another formulation. The basic formulation is not an interpretation but rather what deserves an interpretation. This basic formulation is precisely what lacks in the situation described by Searle: no one is in position to express it. That’s why there is no interpretation of the rule, not even an application of it. The relevant distinction is then between the implicit rules the agent is, in principle, able to formulate and those he cannot possibly express. This last case must be distinguished from the one where the rules have become a second nature, although the agent has started by learning their formulation. Searle seems to think that both cases deserve the same treatment, and this doesn’t make the point in stake clearer. In order to illustrate the last situation, we can however pick up one of his examples. A chess master has no effort to supply anymore to apply the rules. With this allusion to effort, Searle probably wants to mean that the good player doesn’t have towant to apply the rules anymore, but this boils down to saying that he won’t be giving them as reasons for his moves. A good description of such a situation consists in saying that the use of a rule has become automatized. However, the predicate “automatic”, in turn, might be understood in two different ways: (1) in the sense of beeing applied mechanically (2) in the sense of being substituted for an automatism. With the first interpretation, we stay in the vicinity of the rule blindly applied, as the Philosophical Investigations (§ 219) say, the new thing being that forgetting is here added to blindness. The second interpretation is recurrent in cognitive sciences: a cognitive mechanism replaces the rule and we get out here of the realm of the agent’s deeds. In other words, in the first case, it still makes sense to speak of the rule as being applied ; in the second case, the very notion of application goes out of the picture. Occurrences of those two cases might well be difficult to distinguish. Only routine rules are prone to be automatized and become a second nature. This distinction
Tacit knowledge and Action
27
between routine rules and the so-called strategical rules35 is much more relevant than Searle’s pseudo-distinction between constitutive and regulative rules.36 One could say of someone, if it came to it, that he is good in his knowledge of chess rules or in buttoning up his uniform in the army, but no one would say of an agent that he is excellent or brilliant in such activities. Such predicates might, without hesitation, be applied to an insightful players, equiped with high tactical capacities. Unlike strategical rules, routine rules are acquired through a process of repetition (of the same sequence of actions) ; the application of strategical rules involves the consideration of the quality of the deeds, and their mastery requires, from a player for instance, that he participates in a great number of games with different players. In The constitution of society, Giddens explains that it’s important to distinguish, conceptually speaking, two different dimensions of the rules: the one having to do with the constitution of meaning, on the one hand ; the one having to do with the sanction of social conduct, on the other hand. He seems to think that Winch has proved unable to make such a distinction. Giddens has in mind, when he deals with this theme, the difference between constitutive and regulative rules, and his goal is to show that it can’t be understood as a difference between types of rules. That’s why he insists that such a distinction must be thought as conceptual. The rule defining the checkmate is clearly “constitutive” (by so doing it contributes to the definition of what a chess game is) and the rule forcing the employees of a company to clock in is regulative. Nevertheless, he notices, the first one can also be seen, in some respect, as regulative (it refers to an element of the game which is supposed to be respected and involves the possibility of a sanction) and the second one as constitutive (it doesn’t defines work as such but enters in the definition of more circumscribed phenomenas). If the difference is conceptual indeed (and I 35
The predicate “strategical” is used here because the application of those rules involves a judgment. See on this point: Scheffler, 1965, chap. 5. Scheffler distinction concerns types of skills rather than types of rules, but it’s tempting to extend his perspective. 36 See: Searle, 1969, p. 34-35. The so-called constitutive rules are supposed to define a specific game. This means, among other things, that the physical appearance of a checker doesn’t matter at all (it might be a trouser button), the essential thing being the rules governing the moves. Breaking a constitutive rule is either committing an infringement or practising a game other than the one apparently played. Violating a regulative rule is normallly no infringement.
28
Wittgenstein: Issues and Debates
think Giddens is perfectly right on that point), the sinn he accuses Winch of is a venial one. Anyway, Winch probably thought that rules might fulfil numerous different roles, the previous distinction being between two of those many roles. In addition to that, there is no sufficient textual evidence that this distinction, no matter how it is understood, exists as such in Wittgenstein’s writings. The difference between routine rules and strategical rules, on the contrary, is to be interpreted as a real one (no rules can be both routine and strategical at the same time) and this assertion sheds light on what is meant by the observation according to which some rules are such that the agent doesn’t need to want their application when he follows them and on the extension to be attached to the word to understand. This verb might denote a posture involving the exercise of judgment but might also denote the mastery of a routine. Following a routine rule to such an extend that it is so to speak forgotten is not to be identified with an action deprived of any comprehension. In that respect, the agent applying a protocol without any judgment remains different from the thermostat turning up the heat when the outside temperature drops. Conclusions What are we left with after this process of decantation? We have propensions orienting the agent’s actions in our social furniture and capabilities making him able to know what he is to do in contexts that can well be unexpected. We have understood that some of those capabilities are know-how while others have a more complex structure and include the capacity to give one’s justifications for what one has done. In addition to that, we have taken into account routines we have identified with a second nature. In the ongoing of human life we see that is involved a mixture of propensions and capacities. Understanding is on the side of capacities. Many of those capacities are blur-edged (as we have seen through our discussion of concepts) and the formulation of their content seems to root them to the spot, except that the application of a formulated rule, as Wittgenstein says, always leaves ways-out open. Even if rules are really abstract features, it still makes sense to say that the agent’s actions perpetuates them: this only means that a different stretching of actions would have instantiated different abstract features. So the double status is not really a problem.
Tacit knowledge and Action
29
When a agent is in a ruled universe like an eye-ball in his socket, the word “structure” seems nothing more than a mean to denote the systematic body of rules organizing this universe. From this observation, as we have seen, cannot automatically be deduced that this same agent, even he looks normed in his actions, proceed by a knowledge of those rules. If now we undertake a description of the material conditions of action, we might use the terminology mentioned at the beginning of this paper and say that patterns are being brought out. One can think moreover that some of those patterns at least can, in turn, be described in terms of rules. This means that the term is taken in the sense of being a model in the mind of the observer. By doing so, one gets out of mutual knowledge and adopt the objective posture. Bibliography Bourdieu, P., 1979: Le sens pratique, Minuit, Paris. ---
1987: Choses dites, Minuit, Paris.
---
1980: Ce que parler veut dire, Fayard, Paris.
Collingwood, R. G., 1956: Ithe Idea of history, Oxford University Press, Oxford. Giddens, A., 1979: Central problems in social theory, Macmillan, London. ---
1984: The social sciences and philosophy”, in: Social theory and modern sociology, Stanford University Press, Stanford.
---
1986: The constitution of society, second edition, Polity Press, Cambridge.
Quine, W.V.O., 1953: On what there is”, in: From a logical point of view, Harvard University Press, Cambridge Mass., pp. 1-19. Schatzki, Th., 1992: Do social structures govern actions?, Midwest studies in philosophy, vol. 17, pp. 280-295. Scheffler, I., 1965: Conditions of knowledge, Scott & Foresman, Glenview. ---
1983: Of human potential, Routledge, London.
Searle, J.R., 1969: Speech acts, Cambridge University Press, Cambridge. ---
1996: The construction of social reality, Penguin Books, London.
White, M., 2002: A philosophy of culture: the scope of holistic pragmatism, Princeton University Press, Princeton.
30
Wittgenstein: Issues and Debates
Winch, P., 1958: The idea of a social science and its relation to philosophy, Routledge, London. ---
1964: Understanding a primitive society, American Philosophical Quaterly, vol. 1, 307-324.
Wittgenstein, L., 1958: The blue book & the brown book, Blackwell, Oxford. ---
1972: Conversation on Freud, in: Lectures & Conversations, University of California Press, Los Angeles.
---
1979: The Cambridge Lectures 32-35, Blackwell, Oxford.
---
1984: Philosophische Untersuchungen, Werkausgabe, Band 1, Suhrkamp, Frankfurt a.M.
---
1984: Philosophische Grammatik, Werkausgabe, Band 4, Suhrkamp, Frankfurt a.M.
---
1984: Bemerkungen über die Philosophie der Psychologie, Werkausgabe, Band 7, Suhrkamp, Frankfurt a.M.
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology Sabine PLAUD In a famous passage of his Philosophical Investigations, Ludwig Wittgenstein expresses his opposition to the idea of a “scientific philosophy”, and he stresses therefore the following distinction between the methods of philosophy and those of natural sciences: [O]ur considerations could not be scientific ones. And we may not advance any kind of theory. There must not be anything hypothetical in our considerations. We must do away with all explanation, and description alone must take its place1.
In other words, the difference between science and philosophy lies in the fact that science needs explanation, whereas philosophy needs description; that scientific inquiries are empirical, whereas philosophical investigations are grammatical; that science produces hypotheses and theories, whereas philosophy “leaves everything as it is”2. Yet, these very features of philosophy may, in some other respect, be viewed as marking a deep similarity between philosophy and at least some aspects of natural sciences: for besides its explicative aspect, science also involves a whole descriptive part, aiming at a classification or at a taxonomy of the given. And science, intended as a classificatory device, is thus no longer opposed to philosophy, even in its Wittgensteinian description. Such a convergence between natural classifications and Wittgenstein’s philosophical methods is what I wish to investigate here. I will focus on a parallelism that might be drawn between the “morphological method” supported by Johann Wolfgang von Goethe in his scientific writings on the one hand, and Wittgenstein’s concern for “synoptic views” on the other hand. Goethe’s morphological method is in fact a descriptive and comparative one, whose purpose is to display the essential structure which is common to all natural species, notwithstanding 1 2
Wittgenstein, 2001, PI, § 109. Wittgenstein, 2001, PI, § 124.
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ). Ontos Verlag, Frankfurt a. M., 2010, 31-46.
32
Wittgenstein: Issues and Debates
their variety. Morphology is thus regarded as a key to the unity of nature. As we shall see, Wittgenstein’s notion of synoptic views, in the Philosophical Investigations, sounds like an echo to this approach, with only one main difference: namely that this method is no longer used to investigate nature, but rather to investigate language and its grammar. Such a comparison between Goethe and Wittgenstein has its justification in the fact that the former is known to have exerted a strong influence on the latter. Such an influence might have been relayed by Oswald Spengler, whose heritage Wittgenstein himself explicitly acknowledged. 3 Spengler’s main work on The Decline of the West 4 purports in fact to be a morphological study of World History, resting upon an appeal to Goethe’s methodology as applied to cultural rather than natural phenomena. What Wittgenstein does in the Investigations is quite analogous: he uses a method Goethe had introduced in natural sciences as a tool for an analysis of language and social practices. Yet, Wittgenstein’s appeal to Goethe’s method in his own inquiries is not unproblematic. In fact, morphology as it is introduced by Goethe cannot be severed of another aspect of his scientific research, namely of his concern for “primal phenomena” which, in the case of botany, would lie in some kind of “primal plant”. Wittgenstein, on the contrary, seems utterly critical towards such an aspiration to “primal phenomena”, and he regularly denounces the damages caused by “primal pictures” or “paradigms” (Urbilder). Hence a paradox: in Goethe’s work, the morphological method is essentially directed towards the search for the primal phenomenon; in Wittgenstein’s work, this same morphological method appears as weapon directed against the quest of the primal phenomenon. The aim of this study will be to resolve this paradox. In this purpose, I will try to reevaluate the idea of an Urphänomen, by claiming that this notion can actually be interpreted in two different ways, one of which is incompatible with Wittgenstein’s positions, whereas the other is compatible with them. The first of these understandings of the primal phenomenon is a realistic one, regarding the Urphänomen as a common ancestor to all species: this reading is unambiguously rejected by Wittgenstein. The second of these readings is a methodological one, which 3
Cf.: “That is how Boltzmann, Hertz, Schopenhauer, Frege, Russell, Kraus, Loos, Weininger, Spengler, Sraffa have influenced me” Wittgenstein, 1980, CV, p. 19. 4 Spengler, 1991.
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 33
thinks of the Urphänomen as a kind of “prototype”, as a mere idea setting the phenomena in order. In this second reading, a search for the ‘primal phenomenon’ can doubtlessly be fitted into Wittgenstein’s later methodology, whose purpose is precisely to elaborate a typology of our linguistic practices. Consequently, in order to decide whether Wittgenstein’s positions are compatible with Goethe’s, I will have to face the following dilemma: -
Either Goethe regarded the primal phenomenon as a historical fact. In this case, a) his positions are incompatible with Wittgenstein’s conceptions, and b) Wittgenstein’s attitude involves a contradiction, since it implies both an appeal to Goethe’s morphology and a rejection of the very finality of this method.
-
Or Goethe regarded the primal phenomenon merely as a methodological tool. In this case, a) his search for Urphänomena is utterly coherent with Wittgenstein’s concern for schematic presentations of linguistic practices, and b) there is no contradiction in Wittgenstein’s assumption of Goethe’s morphological method.
In order to face this dilemma, my strategy will start by an exposition of Goethe’s morphological method, thereby clarifying its connexion with the idea of a ‘primal phenomenon’. I will then introduce the way Wittgenstein, in his later philosophy, would appeal to this method as a therapy against “primal pictures”. Lastly, I will lay stress on an evolution of Goethe’s conception of the primal phenomenon from a realistic to a methodological reading, and I will claim that, under this second interpretation, his quest for primal phenomena is quite similar to what Wittgenstein does by setting forth his method of “language-games”5. 1.
Morphology as a key to Primal Phenomena in Goethe’s scientific writings
1.1. The primal plant The circumstances in which Goethe came to the idea of a “primal plant” are well known. During his journey to Italy, he had a chance to visit the Botanic Garden of Palermo, where he was struck both by the diversity and by the unity of the plants he discovered. It suddenly appeared to him that 55
For further developments on Goethe’s influence on Wittgenstein, see: Nordmann, 2003, 91-110, and Genova, 1995.
34
Wittgenstein: Issues and Debates
all of these different kinds, despite their differences, must display a common structure, that they must be the manifestations of a single common essence. That is how he formulated the hypothesis of an “original identity” of all plants, an original identity which he located in a “primal plant” (Urpflanze): Seeing such a variety of new and renewed forms, my old fancy suddenly came back to my mind: among this multitude, might I not discover the Primal Plant?6
Under this expression of a “primal plant”, Goethe may have meant four different (and complementary) things: 1.
In its most obvious sense, such a hypothesis can be regarded as a chronological one, resting upon the transformist postulate of a common ancestor out of which all the natural kinds would have evolved.
2.
A second presupposition of this concept is the idea of an essential structure common to each particular plant, which may itself be a remnant of this very ancestor.
3.
Such a structure may in turn be intended not only in a static, but also in a dynamic way: more than a mere skeleton, it can be defined as a model of development.
4.
Lastly, Goethe’s idea of a primal plant may have the sense of a single organ which, in each plant, would serve as a basis for its formation, and which is to be detected in a certain kind of leaf, as suggested by the following passage: While walking in the Public Gardens of Palermo, it came to me in a flash that in the organ of the plant which we are accustomed to call the leaf lies the true Proteus who can hide or reveal itself in vegetal forms.7
All of these definitions are not rival understandings of Goethe’s concept of a primal plant, for they all have one thing in common: all of them imply that any account of the development of plants has to be directed towards the Urpflanze. Accordingly, botany itself turns out to be a study of the “metamorphosis” of plants, i.e. of these transformations that bear witness of the existence of a supra-empirical primal plant within the real, empirical plants: [T]he process, by which one and the same organ presents itself to our eyes under protean forms, has been called the Metamorphosis of Plants.8 6 7
Goethe, 1992, Italian Journey, April 17th, 1787. Goethe, 1992, Italian Journey, July 31th, 1787.
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 35
Such a botanical project is thus not only a scientific project, but also a romantic one: by its endeavor to bring back all the particular plants to a single, primal one, its purpose is to exhibit the synthesis of unity and variety displayed by the “beautiful garden of the universe”. 1.2. Primal Plant, Primal Phenomenon, Primal Picture Such a quest for the origins is, in Goethe’s work, far from being limited to the realm of botany. In fact, the notion of an Urpflanze may be generalized so as to gain applications in all of the realms of natural sciences. Such a generalization is essentially linked to the general concept of a ‘primal phenomenon’ (Urphänomen), which is displayed by Goethe as the most general aim to be reached by any natural science. As a matter of fact, every science has to investigate its own primal phenomenon. In the case of botany, the primal phenomenon is of course the Urpflanze. In the case of optics, this phenomenon lies in the opposition between brightness and obscurity, an opposition which is the key to the production of the various colors. Hence, for example, this passage of his Theory of colours, where Goethe is quite explicit on his ambition to discover the primal phenomenon: From henceforth everything is gradually arranged under higher rules and laws, which, however, are not to be made intelligible by words and hypotheses to the understanding merely, but, at the same time, by real phenomena to the senses. We call these primordial phenomena, because nothing appreciable by the senses lies beyond them9.
In the case of geology, the primal phenomenon is in turn to be detected in the granite, since granite is the real Urgestein, the primal rock which entails the other ones and which can be described as the “cradle of earth”. In the case of zoology, the Urphänomen lies in a kind of Urtier (“primal animal”), which would serve as a common ancestor to the animals in general and to the insects in particular. Lastly, in the case of anatomy, the primal phenomenon is located by Goethe in the intermaxillary bone, a bone whose existence he discovered in the human skull itself, thus proving the morphological unity of men and animals. In a nutshell, then, the notion of a primal plant leads one to the notion of a primal phenomenon; but these two notions themselves are in 8 9
Goethe, 1790, §4. Goethe, 1982, §175.
36
Wittgenstein: Issues and Debates
turn essentially connected with a third one: namely with the notion of a “primordial picture” or “paradigm” (Urbild). This idea of a “primordial picture” means in fact two things. On the one hand, the Urbild amounts to the general pattern out of which plants and animals happen to be formed10; in this sense, the notion of an Urbild is almost identical to Goethe’s Urphänomen. However, in a broader sense, the Urbild can also be read as the representation science is likely to give of such a pattern. In his Writings on Morphology, Goethe claims for example that the main purpose of natural science is to: present this primordial picture, if not to senses themselves, at least to the mind; and to elaborate our descriptions on its basis, as it were a norm.11
In other words, the very purpose of natural sciences is to set forth a clear view (an Urbild) of the primal phenomenon that pertains to each particular field of nature. Only thus can they set the natural phenomena in order. 1.3. The morphological method This method aiming at a clear representation of the organization of phenomena on the basis of the primal phenomenon is precisely what Goethe refers to when he refers to his “morphology”. This concept of morphology is in fact the key-concept in Goethe’s natural sciences. Goethe defines seems to have elaborated this term himself, and he defines it as “the science of the form, formation and transformations of living organisms”.12 And since the form (Gestalt) and the formation (Bildung) of phenomena are but a manifestation of the Urphänomen, then morphology itself is but a study of the Urphänomen. Accordingly, morphology hunts for the connections, for the “divine order” that underlie phenomena. In this respect, this method is essentially a descriptive one, resting upon analogies. A characteristic feature of Goethe’s morphology is besides its overt phenomenism. To claim that the final aim of morphology is the primal phenomenon is to maintain that, even though science is concerned with a primordial phenomenon, still it isn’t supposed to go beyond the phenomena. This is what Goethe recalls in the following passage, where he condemns any attempt to go beyond the (primal) phenomenon: 10
Cf. Goethe, 1962, “Hefte zur Morphologie”, Erster Band, p. 472. Goethe, 1962, “Entwürfe zu einem osteologischen Typus”, p. 270. 12 Goethe, 1962, “Hefte zur Morphologie”, Erster Band, Introduction. 11
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 37 But when even such a primordial phenomenon is arrived at, the evil still is that we refuse to recognize it as such, that we still aim at something beyond, although it would become us to confess that we are arrived at the limits of experimental knowledge. Let the observer of nature suffer the primordial phenomenon to remain undisturbed in its beauty.13
Yet, Goethe’s morphological method is not devoid of ambiguity: for despite this vindicated phenomenism, one may regard his very endeavor towards origins and primacy as a mere transgression of the phenomenist principle he assumes. I will now examine how such a reproach was formulated by Wittgenstein. 2.
Wittgenstein and Goethe: morphology as a tool against Urbilder
As I have tried to show, Goethe’s concept of an Urphänomen, as well as the morphological method correlated with it, are essentially linked with the concept of a picture, and more precisely with the concept of a paradigm or of a ‘primal picture’. Now, another philosopher who was quite concerned with the concept of picture is Ludwig Wittgenstein. In his later philosophy, Wittgenstein displays in fact a rather ambiguous attitude towards pictures. On the one hand, he expresses what may be called an “iconophobia”, which leads him to denounce our fascination by pictures as a source of philosophical confusions: A picture held us captive. And we could not get outside it, for it lay in our language and language seemed to repeat it to us inexorably.14
But, on the other hand, the remedy Wittgenstein intends to apply to such philosophical diseases rests itself upon a certain use of pictures, namely on the appeal to “synoptic representations”. My claim, here, will be twofold: first, I will show how Wittgenstein’s iconophobia is directed against Goethe’s hunt for an Urbild. Second, I will show how Wittgenstein’s remedy against Urbilder is nonetheless inspired by Goethe’s morphology. 2.1. Wittgenstein’s rejection of Goethe’s Urbilder In some of his later works, Wittgenstein displays the strongest criticism towards Goethe’s aspiration to discover an original image. In his view, 13 14
Goethe, 1982, § 177. Wittgenstein, 2001, PI, § 115.
38
Wittgenstein: Issues and Debates
such an idealization of “primordial pictures” prevents one from grasping the specificity of singular objects: But since we confuse prototype [Urbild] and object, we find ourselves dogmatically conferring on the object properties which only the prototype necessarily possesses15.
In other words, the Goethean search for an Urbild is, at least in some cases, the mark of a dangerous essentialism which may divert one out of reality. This happens, for example, in the philosophy of language, where classical theories of meaning typically regard referentiality as the Urphänomen of meaning in general, thus leaving aside a major part of linguistic devices. But philosophy of language is not the only field in which our captivation by primal pictures has proved to be detrimental. Psychoanalysis is another example, since, in Wittgenstein’s opinion, some of its initially clever intuitions have been corrupted by the prejudice that all psychical phenomena are governed by one and the same law, or that all dreams are wish-fulfillments: The “primary phenomenon” is, e.g., what Freud thought he recognized in simple wish-fulfillment dreams. The primary phenomenon is a preconceived idea that takes possession of us16.
In other words, the “primal phenomenon” Goethe would regard as the final aim of any scientific research is here characterized as a mere illusion we are the victims of. 2.2. Synoptic views as an echo to Goethe’s morphology Now, what about the therapy Wittgenstein offers to such illusions? This is the point where the paradox arises: for I will now claim that, although the illusions to be cured were a product of Goethe’s conceptions, the therapy itself is in turn also inspired by Goethe’s method. When one suffers a disease, one must in fact be provided with the convenient remedy. In the case of the philosophical disease which originates in a fascination for Urbilder, the convenient remedy is, according to Wittgenstein, the elaboration of synoptic views: A main source of our failure to understand is that we do not command a clear view of the use of our words. – Our grammar is lacking in this sort of perspicuity17. 15 16
Wittgenstein, 1980, CV, p. 14. Wittgenstein, 1977, § 230.
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 39
In other words, our philosophical confusions result of our incapacity to view language in its variety because of a captivation by preconceived ideas; therefore, the right therapy against these confusions is to exhibit with the utmost clarity the connections that characterize the different aspects of language. That is exactly the reason why: The concept of a synoptic representation is of a fundamental significance for us. It earmarks the form of account we give, the way we look at things18.
But this method of perspicuous or synoptic representations is clearly an echo to Goethe’s morphology. One may indeed point to four similarities between Goethe’s and Wittgenstein’s respective methods: 1.
In both cases, we are confronted with a descriptive approach, whose purpose is merely to shed light upon (natural or linguistic) phenomena. This is, for example, the reason why Wittgenstein characterizes his philosophical method as a device meant to change perspective on things: I wanted to put that picture before him, and his acceptance of the picture consists in his now being inclined to regard a given case differently (…). I have changed his way of looking at things.19
Such a switch of perspective is quite similar to what Goethe intended to do in the realm of natural sciences. 2.
The second of these similarities has to do with the common phenomenism that underlies both of these approaches. I have in fact stressed the idea that, in Goethe’s morphology, science was not supposed to go beyond phenomena. This is a requirement Wittgenstein himself imposes on philosophy when he claims that: Philosophy simply puts everything before us, and neither explains nor deduces anything. – Since everything lies open to view, there is nothing to explain. For what is hidden, for example, is of no interest to us.20
3.
17
Both Wittgenstein and Goethe support an essentially a comparative approach of phenomena. Just like Goethe, Wittgenstein intends in fact to display analogies, connections, similarities and dissimilarities among our practices. This is for example what he does in his famous analysis of games:
Wittgenstein, 2001, PI, § 122. Wittgenstein, 2001, PI, § 122. 19 Wittgenstein, 2001, PI, § 144. 20 Wittgenstein, 2001, PI, § 126. 18
40
Wittgenstein: Issues and Debates Consider for example the proceedings that we call “games”. (…) – [I]f you look at them, you will not see something that is common to all, but similarities, relationships, and a whole series of them at that. To repeat: don’t think, but look!21 There again, such an exhibition of the phenomena in their varieties could certainly be described as a typically “morphological” one, and regarded as an echo to Goethe’s search for connections among living organisms.22
4.
Lastly, I have introduced above the idea that Goethe’s project was to lead one to see unity in variety. Just the same goes with Wittgenstein’s method, since his work itself is nothing like a demonstrative speech, but consist of a series of sketches whose unity has to be drawn afterwards: After several unsuccessful attempts to weld my results together into such a whole, I realized that I should never succeed. (…) – The remarks of this book are, as it were, a number of sketches of landscapes which were made in the course of these long and involved journeyings. (…) Thus this book is really only an album.23
3.
The paradox and its solution
If, as I have just argued, Wittgenstein’s method of synoptic representations does assume most of the devices Goethe had introduced in his morphology, then there is doubtlessly a tension as for the relationship between Wittgenstein and Goethe. I will sketch this tension as follows: a)
In Goethe’s work, morphological method and search for the primal phenomenon are bound altogether.
b)
Wittgenstein rejects Goethe’s search for the primal phenomenon.
c)
Therefore he ought to reject Goethe’s morphology as well.
d)
Yet he does adhere to Goethe’s morphological method, and
e)
He even uses it as a weapon against Goethean trends towards Urbilder. I will now examine how such a tension can be resolved.
21
Wittgenstein, 2001, PI, § 66. See also: “A perspicuous representation produces just that understanding which consists in “seeing connections”. Hence the importance of finding and inventing intermediate cases”. Wittgenstein, 2001, PI, § 66. 23 Wittgenstein, 2001, PI, Preface. 22
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 41
3.1. Two readings of primal phenomena - Goethe’s controversy with Schiller As I have tried to show, the main reason why Wittgenstein rejects Goethe’s notion of an Urbild is that he regards such a notion as an essentialist, and therefore metaphysical or illusory one. To put it in another way, if the primal phenomenon is interpreted as an essentialist notion, then Wittgenstein necessarily rejects it. But maybe such an interpretation is not the only possible interpretation. Maybe the search for the primal phenomenon, even as it is intended by Goethe himself, can be read as a merely heuristic or “grammatical” one: if it is so, then Goethe’s approach is no longer incompatible with Wittgenstein’s own views. And my claim is that it is so: or at least, that it is so in Goethe’s later attitude towards primal phenomena. When he first formulated the hypothesis of the Urpflanze, Goethe was indeed certainly, up to some extent, a defender of a transformist approach to natural kinds, and he thus doubtlessly regarded the Urpflanze as a common ancestor to all the actual plants. Yet, even his early characterizations of the Urpflanze turn out to display an underlying idealistic background. One may examine, for example, the way Goethe would demonstrate the necessity of this primal plant: There certainly must be one. Otherwise, how could I recognize that this or that form was a plant, if all were not built on the same model?24
In this passage, Goethe’s key-argument is that one would be unable to recognize a particular plant as a particular plant lest one had the idea of what a plant is, in general. But in those conditions, the Urpflanze itself should rather be regarded as a type or as a mere idea, in a sense which would be quite similar to Plato’s ideas. Such a switch towards an idealistic reading of primal phenomena is again manifest in some other texts, where Goethe, instead of assigning the Urpflanze an empirical existence, rather describes it as a mere rule for the construction of possible plants: The primal plant is going to be the strangest creature in the world, which nature itself shall envy me. With this model and the key to it, it will be possible to go on forever inventing plants and know their existence is logical.25
24 25
Goethe, 1992, Italian Journey, April 17th, 1787. Goethe, 1992, Italian Journey, May 17th, 1787.
42
Wittgenstein: Issues and Debates
The primal plant, here, has nothing to do with a genuine or existing plant: it appears much more as a “model”, as a methodological and regulative idea allowing the scientist to find out what is possible in nature and what is not, what is “logical” or consistent and what is not. But such an idealistic turn will be achieved in particular under the influence of Schiller, as related by Goethe himself in his text by the name of “An Happy Event” (“Ein Glückliches Ereignis”). Goethe had in fact, in 1793, a chance to give an enthusiastic account of his hypothesis of the Urpflanze to the young Schiller, and he even went so far as to sketch such a plant on a piece of paper. Schiller, of course, was not exactly skeptical towards Goethe’s drawing; yet he answered, with no less enthusiasm: “That is not an experience, it is an idea!”. As might have been expected, Goethe was initially quite irritated by such a remark. Nevertheless, he was to regard, a few years later, this event as an “happy event” indeed: for it had paved the way for a more lucid approach of the Urpflanze, understood as mere theoretical archetype. 4.
Wittgenstein’s language-games as another kind of primal phenomena
My last step will now be to demonstrate that this understanding of primal phenomena which Goethe had gradually acknowledged is utterly coherent with Wittgenstein’s own tenets, and in particular with his method of language-games. Assuming that Goethe does not support a realistic account of primal phenomena, then Wittgenstein’s target, in his criticisms of dogmatic approaches to primal phenomena, is not Goethe’s approach: it’s rather a misinterpretation of it. As a matter of fact, Wittgenstein clearly draws a distinction between two possible uses of Urbilder: a correct one, and an improper one. Whereas the improper use of Urbilder is their dogmatic use, the correct one is a strictly methodological and regulative use of Urbilder as a prototype, devoid of any dogmatic or historical purport. Such a distinction between two different uses is quite obvious in Wittgenstein’s Remarks on Frazer’s Golden Bough. In this text, one may read the following remark: “And all this points to some unknown law” is what we want to say about the material Frazer has collected. I can set out this law in an hypothesis of
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 43 development. (…) But equally I might see the hypothesis of developement as nothing but a way of expressing a formal connection.26
Such a passage entails a direct reference to Goethe’s morphology. The very idea that “all this points to some unknown law” is in fact a quotation of Goethe, drawn from the Hymn to Nature by which he concludes his Metamorphosis of Plants: And all this points to some unknown law, To a sacred enigma. The “unknown law” Wittgenstein alludes to is nothing but the law governing the development of plants out of the Urpflanze. This law is exactly what James George Frazer and Oswald Spengler have tried to generalize, by applying it to the development of societies. And such a generalization is, in Wittgenstein’s opinion, the point where misunderstandings arise: for these two authors have missed the fact that Goethe’s hypothesis of evolution is a regulative, not a historical one. It might have been justified to appeal to Goethe’s Urphänomen: but it cannot be justified to forget that the Urphänomen is nothing but a methodological tool. If, however, one avoids such a mistake, then one is allowed, even in a Wittgensteinian perspective, to make of the search for the primal phenomenon a central device in philosophy. This is the reason why, despite the criticisms he had explicitly directed towards the improper use of Urbilder, Wittgenstein claims, no less explicitly, to be a follower of Goethe: What we are doing here runs parallel to some extent with Goethe’s ideas about the metamorphosis of plants.27
Note the restriction Wittgenstein introduces with the expression “to some extent”: this restriction means, in my opinion, “as far as the Urphänomen in not intended in a realistic way”. But such a realistic interpretation is something that Wittgenstein, at least in this text, is not willing to detect in Goethe’s work. Quite the contrary: he wants to give Goethe’s conceptions an anti-realistic reading, and he wants to assume that the dogmatic interpretations of this method such as defended by Spengler or Frazer are misled: 26 27
Wittgenstein, 1979, RFGB, p. 8. Wittgenstein, 2003, Ch. 3, p. 311 (my emphasis).
44
Wittgenstein: Issues and Debates From Goethe stems the conception of the ‘primal plant’; yet surely he saw in it only an idea, not something real. (…) Goethe’s aphorism ‘All the organs of plants are leaves transformed’ gives us a schema by which we group the organs of plants according to their degrees of similarity, as it were around a central case.28
In those conditions, what Wittgenstein does himself is indeed quite similar to what Goethe does in the realm of botany, except that he does it in the realm of language. Goethe’s purpose was in fact to view the leaf “in its natural surroundings of forms” 29 ; Wittgenstein’s purpose is to view social practices in their linguistic surroundings of forms: And this is exactly what we are doing: we situate a linguistic form in its surroundings, we see the grammar of our language against a background of similar and related games, and that banishes disquiet.30
Such a linguistic turn of Goethe’s morphology will obviously be achieved in his later philosophy with the so-called method of languagegames. The method of language-games is introduced by Wittgenstein in his Blue Book31, and becomes a key device of the Philosophical Investigations. In a broad acception, a language-game can in fact be defined as a determinate set of linguistic and social practices. But in a more specific acception, a language-game is also a theoretical and primitive reconstruction of linguistic practices, whose function is to bring clarity among the complexity of our real performances. This is what Wittgenstein means when he claims that he “will sometimes speak of a primitive language as a language-game” 32 . His language-games thus provide the reader with primitive and over-simplified representations of linguistic practices, as may be adopted, for example, by children or primal tribes33. Yet, such a primacy has to be devoid of any psychological, historical or anthropological purport; for this is the condition that, alone, may guarantee the legitimacy of this search for linguistic primal phenomena: Our mistake is to look for an explanation where we ought to look at what happens as a ‘proto-phenomenon’. That is, where we ought to have said: this language-game is played.34 28
Wittgenstein, 2003, Ch. 3, p. 311, Wittgenstein, 2003, Ch. 3, p. 311. 30 Wittgenstein, 2003, Ch. 3, p. 311. 31 Cf. Wittgenstein, 1958, BBB, p. 18. 32 Wittgenstein, 2001, PI, § 7. 33 On these issues, see: Hilmy, 1987. 34 Wittgenstein, 2001, PI, § 656. 29
Synoptic Views vs. Primal Phenomena: Wittgenstein on Goethe’s Morphology 45
Language-games, in other words, are in fact the Wittgensteinian equivalent to Goethe’s primal phenomena. On that respect, Wittgenstein’s method in philosophy should definitely be regarded as an echo to Goethe’s method in natural sciences. Conclusion In a passage of his Farbenlehre, Goethe formulates the following judgment on his own achievements: We believe we merit thanks from the philosopher for having traced phenomena to their origins (…). Further, it will be gratifying to him that we have arranged the appearances in an easily surveyed order, even should he not altogether approve of the arrangement itself.35
What I have tried to show in this paper is that, although Wittgenstein was certainly a philosopher, he did not always give Goethe the thanks he merited for having traced phenomena to their origins. Nevertheless, it was certainly very gratifying to him to have had the appeareances arranged in an easily surveyed order; so gratifying indeed that he made of this easily surveyed order the key to his own method, and that he consequently tried to reevaluate Goethe’s approach of primal phenomena so as to make it compatible with his own conceptions. Bibliography Genova, J., 1995: Wittgenstein, A Way of Seeing. Routledge, New York. Goethe, Johann Wolfgang von, 1790: Versuch die Metamorphose der Pflanzen zu erklären. Carl Wilhelm Ettinger, Gotha. Goethe, Johann Wolfgang von, 1962: Schriften zur Morphologie. “Hefte zur Morphologie”, Erster Band, (Ed. D. Kuhn). Deutscher Klassiker Verlag, Frankfurt a.M. Goethe, Johann Wolfgang von, 1982: Theory of Colours. (Tras. Charles Lock Eastlake). The M.I.T. Press, Cambridge, Massachusetts. Goethe, Johann Wolfgang von, 1992: Italian Journey. (trans. W.H. Auden and Elizabeth Mayer), Penguin, London. Hilmy, S., 1987: The Later Wittgenstein : The Emergence of a New Philosophical Method. Blackwell, London. 35
Goethe, 1982, Introduction.
46
Wittgenstein: Issues and Debates
Nordmann, A., 2003: “’I have changed his way of seeing’: Goethe, Lichtenberg and Wittgenstein”, in: F. Breipthaut, R. Raatzsch And B. Kremberg (ed.): Goethe and Wittgenstein. Seeing the World’s Unity in its Variety, Wittgenstein-Studien, 5, Lang, Frankfurt a.M., 91-110. Spengler, Oswald, 1991: The Decline of the West. (Ed. Arthur Helps, and Helmut Werner. Tr. Charles F. Atkinson). Oxford UP, New York. Wittgenstein, Ludwig, 1958: The Blue & Brown Books: Preliminary Studies for the Philosophical Investigations. Blackwell, Oxford. Wittgenstein, Ludwig, 1979: Remarks on Frazer’s Golden Bough. (Ed. R. Rhees, trans. A. C. Miles). Brymill, Humanities Press, Atlantic Highlands, NJ. Wittgenstein, Ludwig, 1980: Culture and Value. (Ed. G. H. von Wright, tr. P. Winch). Blackwell, London. Wittgenstein, Ludwig, 1977: Remarks on Colour. (Ed. G.E.M. Anscombe, trans. L. McAlister and M. Schätle). Blackwell, London. Wittgenstein, L., 2001: Philosophical Investigations. (Trans. G.E.M. Anscombe). Blackwell, London. Wittgenstein, Ludwig, 2003: Dictation for Schlick, in The Voices of Wittgenstein. (Ed. G. Baker, trans. G. Baker, M. Mackert, J. Connolly and V. Politis). Routledge, London.
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein Eric LEMAIRE Introduction Has Wittgenstein destroyed Metaphysics? Wittgenstein scholars often hold an affirmative answer as if it were obvious. If it is true, it has to be made clear that, first, Wittgenstein produced a philosophy, which cannot be qualified as Metaphysics; and second, he really established that metaphysical propositions are nonsensical. There are several ways to discuss these problems. The most direct consists in examining Wittgenstein’s texts. Another one, the one that will be followed here, consists in examining the readings proposed by his interpreters. These strategies are not incompatible. In a longer study, it would be possible to combine them. We shall raise mainly two questions here. In the first place, is there an anti-metaphysical reading of the second Wittgenstein? Secondly, assuming that the readings, which we are going to study are true to the second Wittgenstein’s works, do they offer a plausible defense in favor of the thesis according to which metaphysical propositions are nonsense? Specifically, are they consistent? Following R. Fogelin, D. Stern proposed a general classification of the readings of Wittgenstein: “The principal fault line separating Wittgensteinians is over a question of philosophical method: whether or not a radical philosophical change – putting an end to philosophy – is possible. Robert Fogelin draws a helpful distinction between ‘Pyrrhonian’ readings of the Investigations, which see the book as informed by a quite general skepticism about philosophy and so as aiming at bringing philosophy to an end, and ‘Non-Pyrrhonian’ readings, which construe the book as a critique of certain traditional theories in order to do philosophy better.[…] Another way of putting the distinction is to say that Pyrronhian Wittgensteinians believe philosophy, properly conducted, should not result in any kind of theory, while Non-Pyrhonian Wittgenstenians maintains that Wittgenstein’s criticism of traditional Philosophy leads us to a better
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ). Ontos Verlag, Frankfurt a. M., 2010, 47-64.
48
Wittgenstein: Issues and Debates philosophical theory, albeit not the kind of theorizing we find in the philosophical tradition.”1
We find anti-metaphysical readers among pyrrhonians as among their opponents. Among non-pyrrhonian readers, P. Hacker is the emblematic author whose works are often the target of pyrrhonians’ attacks. On the pyrrhonian side, C. Diamond and J. Conant are also important authors. We will concentrate widely on these three authors. Before studying what they have said, let us begin by clarifying what we mean by “Metaphysics”. 1.
What is metaphysics?
In order to know if the second Wittgenstein has showed that metaphysics is illegitimate, we have to ask first what metaphysics is. At least two types of definition can be offered. V. Descombes distinguishes between dogmatic and useful definitions of Metaphysics 2 . According to him, a useful definition of “Metaphysics” tries to be theoretically neutral. In contrast with this, a dogmatic definition favors certain philosophical theories. It presupposes that certain theories are true. In a dogmatic definition, the identity of Metaphysics depends on a set of metaphysical thesis but not on the metaphysical problems themselves. On the contrary, when we look for a useful definition, the identity of Metaphysics depends strongly on problems, which are studied in Metaphysics. That a dogmatic definition be not suitable to our purpose is clear. It would indeed be self-defeating to use a dogmatic definition to undermine the possibility of Metaphysics. It would be as mistaken as to think that physicists overcame physics since they abandoned Aristotle’s physics. This first requirement can be named the principle of neutrality. However, there is a risk of emptying the concept “Metaphysics”. Metaphysics is a very old discipline. It was developed in quite different ways by several people and thus has received many definitions. We have a second requirement: the principle of distinction and unity. A good definition has to account for the intuitive fact that Metaphysics differs from natural sciences, economy, religion, arts, etc. These two principles are vague. They are as vague as, e.g., an article of the Bill of rights or the French constitution. They can be interpreted in various ways. It happens 1 2
Stern (2004), pp. 34-6. Descombes (1995), p. 113, also see: pp. 111-9.
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
49
that they conflict with each other. And what they order us to do is not always obvious. According to the French constitution, freedom and intellectual property are two examples of inalienable rights. But the constitution does not say explicitly how to solve the problem of illegal downloading. According to the second principle, Metaphysics, mathematics, physics, economy and many other disciplines differ, but nothing is said about how much they do. The vagueness of a constitutional right is not a problem in itself. On the contrary, it is necessary a constitution be able to face unexpected situations. When such unexpected situations happen, it is essential that we strike a balance between conflicting rights. Our present situation is similar. What is Metaphysics? According to us, the best definition is the following: Metaphysics is the study of the most general and/or central structure of reality or of our conceptual scheme, considered as a whole. Certain characteristics of this definition need to be clarified in order to avoid possible misunderstandings: 1. Metaphysics is vague and open. 2. Our definition has to account for the various forms of antirealism. 3. Our definition is neutral from the point of view of the nature of the survey, its methodology and its ambitions.
Let us briefly examine these three points. To claim that Metaphysics is vague means that there is no sharp limit between the concepts, which it studies and those studied in other branches of philosophy (like philosophy of language, of arts, of religion, etc.)3. Some concepts, like “table”, “star”, “rock”, seem far too particular to be studied in metaphysics. Others, such as “causality”, “identity”, “necessity”, “possibility”, “concrete particular”, “abstract entity”, “property”, “change”, “event”, “fact”, “reality”, “time”, “space”, “possible worlds”, etc. seem to satisfy our demand of generality and/or of centrality. Other concepts, such as “God”, “propositions”, “meaning”, appear to be borderline cases. But if what can be said about Metaphysics can be generally said about Philosophy, what will the difference be between Metaphysics and the other branches of Philosophy such as epistemology or philosophy of language, etc.? If it is true that Metaphysics is vague, then the distinction between Metaphysics and the other branches of Philosophy is vague too. In fact, the frontiers of Metaphysics seem to be somewhat arbitrary and conventional: it could be 3
We found the idea that “metaphysics” is a vague concept in Van Inwagen (1998).
50
Wittgenstein: Issues and Debates
decided to subsume every branch of philosophy under the name “Metaphysics”. There are several possible classifications of philosophical problems. Sometimes, the mind/body problem is presented as a part of philosophy of mind. Sometimes, it is presented as a problem of Metaphysics of mind. Is the problem of God’s existence a problem for Metaphysics or Philosophy of religion? Are the question whether propositions exist and the problem of their nature, parts of Metaphysics or Philosophy of language? If we wonder if fictional entities, such as Madame Bovary, exist, do we do Philosophy of art or Metaphysics? And what about the question “Does some knowledge exist?”. Is it a metaphysical or an epistemological one? After all, “knowledge” is a rather general term and the question of its existence is raised. Let us examine the second point. Can we really say that anti-realisms are forms of metaphysics? This is a controversial point. Roughly speaking, realism is the idea that what we call “reality” is not built or constructed (even partially) by our conceptual activity. By “antirealism” here, we mean the negation of the latter idea. Some philosophers think that realism only is a metaphysical theory whereas others deny it4. In our view, both realists and their opponents are metaphysicians. They argue about the nature of reality taken as a whole. If antirealism proved to be correct, this could only show that we have misunderstood the scope of metaphysical discourse. But we would keep on raising and addressing metaphysical problems. Philosophers such as Goodman, Quine, or Putnam, addressed the problem of universals and other metaphysical problems. They argued with other, realistic, metaphysicians. To agree that antirealism is metaphysical does not seem to entail practical consequences concerning the way we philosophize. If antirealism proved to be correct, Metaphysics would not become identical to physics, economy or another intellectual discipline. In fact, it would not change its position in the architecture of human knowledge. It would remain the first philosophy, at least in the sense that it would still have the highest degree of generality (even M. Dummett agrees with it). It could still maintain the same standard of epistemological ambitions, use the same methods, and aim to play an important and foundational role towards other disciplines. It is possible to make use of “Metaphysics” in a less open way. However it would be necessary to create a new word (as Postmodern Philosophy, for example). Moreover, 4
See Putnam (1998) and Van Inwagen (2002), chapter 5, for those who deny that antirealisms are metaphysical; and see Loux (2006), chapter 1 and Lowe (1998), chapter 1 for those who, to the contrary, accept it.
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
51
this could easily create misunderstandings by suggesting that using this new word entails a radical break. One can wonder whether Aristotle’s physics is closer to quantum mechanics than a realist metaphysics to an antirealist metaphysics*. Finally, we have insisted on the necessity for our definition of Metaphysics to be neutral with regards to its nature, its methodology and its ambitions. It is important to clarify this point because a certain picture has long dominated the history of metaphysics. And Metaphysics should not be mistaken for this picture. Though it has not been accepted by all philosophers, it nevertheless seems to have been prevailing at least in the modern period. This picture depicts Metaphysics as a completely a priori, absolutely incredulous or self-critical, science which aims, 1. at discovering absolute truths about the necessary structure of reality and, 2. at providing particular sciences and possibly common sense with an indestructible foundation, in order to preserve them from skepticism. This picture has been put into question. Some philosophers reject the idea that Metaphysics could discover the actual structure of reality by a purely a priori investigation or that it could achieve absolutely certain knowledge5. And today, the strong foundationalist ambitions of Metaphysics toward sciences do not seem plausible anymore. Inasmuch there are disagreements about these points, it seems preferable to be neutral here. 2.
Hacker’s interpretation
On P. Hacker’s view, Wittgenstein has showed that philosophy, correctly understood, differs radically from Metaphysics as it was conceived. Let us see first what he means by “Metaphysics”: “In metaphysics, defenders of the metaphysical enterprise conceived of true metaphysical propositions as descriptions of necessary, essential relations between simple natures, or Platonic Ideas, or Universals. Metaphysics was conceived as a super-physical investigation of the most general features of the universe and its ultimate constituents, which would yield a description of the necessary structural features not merely of the world but of any possible world. Unlike truths of physics, metaphysical truths were held to be descriptions not of contingent facts about reality, but of necessities (including transcendent necessary truths about God and the immortality of the soul). Kant’s Copernican 5
See: Lowe (1998), chapter 1, (2002) (a), chapter 1 and (b), and (2007). See also Armstrong (1978), chapter 1 and (1997), chapter 1. We regret that we cannot develop these points, which would need more attention.
52
Wittgenstein: Issues and Debates revolution attributed the deliverances of metaphysics to insight not into mindindependent necessities, but rather into necessities of thought, structural features which the mind imposes upon the raw data of intuition in the process of yielding conceptualized experience. Kant rejected transcendent metaphysics, limiting metaphysical knowledge to synthetic a priori truths which describe the conditions of the possibility of experience.”6
Then, he affirms that Wittgenstein has established that these conceptions of Metaphysics rest on a false assumption: “Wittgenstein rejected the common assumption that what are conceived of as metaphysical truths are descriptions of anything, that the “necessary truths” of metaphysics are truths about objects in reality at all. Rather, what we conceive of as true metaphysical propositions are norms of representation, rules for the use of expressions in the misleading guise of descriptions of objects and relations. They are not synthetic a priori truths about the conditions of possible experience (as Kant argued), but rules of grammar; not nonsense (as Hume regarded), but conventions for the description of things.”7
Wittgenstein is supposed to have rejected the presupposition on which these conceptions are based, namely the idea that metaphysical propositions are descriptions. Although metaphysical propositions appear to be factual descriptions, they are indeed only arbitrary conventions, i.e., norms of representation or grammatical rules. These rules determine the conditions according to which the expressions of language are meaningful. These norms are neither true nor false. In other words, they are not bipolar, but nonsensical. Grammatical rules are arbitrary, but nevertheless necessary. They are not the products of a lively imagination. We cannot choose them of our own free will. According to P. Hacker, Wittgenstein’s Philosophy is not a radical anti-essentialist pamphlet 8 . The latter would have simply held that essence should be conceived differently. It should no longer be conceived as a characteristic of things themselves, but rather as a convention. Essence is expressed by the grammatical rules which determine the meaning of words9. Propositions such as “Material objects cannot be located in two places at the same time” or “Each event has a cause” or “Time travel is impossible” look like descriptions (of the structure of reality or the structure of our thought). But no reality justifies them. These propositions only determine the meaning of expressions as 6
Hacker (1996), pp. 101-2. Ibid. 8 Hacker (1996), p. 119: “All talk of essences is talk of conventions, and what seems to be the depth of the essences is actually the depth of our need for the conventions.” 9 Hacker (1996), p. 113. 7
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
53
“material objects”, “event”, etc. They describe no metaphysical fact10. If this is true, we have no chance of discovering any new and important facts. We will never discover that material objects do not persist in time or that they are not spread out in space. These pseudo-discoveries are only meaningless propositions11. They are nonsense because they break the rules of grammar. In metaphysics words having an ordinary use defined by rules are used outside the boundaries circumscribed by these rules12. If all this is true, Philosophy does not increase our knowledge13. Its function lies rather in improving our understanding of the grammar of language. A better understanding of grammar enables us to identify metaphysical nonsense. For this purpose, we need to recollect the ordinary use of expressions involved in metaphysical problems. Wittgenstein has always held that philosophical speculations are attempts to cross the limits of language. According to him, the philosopher’s task is to show this in revealing nonsense. This short summary could incite one to conclude that Wittgenstein’s project is purely destructive. But P. Hacker insists on the fact that it has two aspects 14 . On the one hand, it consists of a therapeutic cure of the 10
Hacker (1996), p. 119. Hacker (1996), p. 112 and 120. 12 Hacker (2001), pp. 361-2, for further details. 13 Hacker (1996), p. 110: “If one has to choose one a single fundamental insight from the whole corpus of Wittgenstein’s later work, it might well be argued that it should be the insight that philosophy contributes not to human knowledge, but to human understanding; that there is not, and cannot be, a body of established philosophical propositions, a corpus of philosophical truths to which successive generations may add to constitute a body of ever growing philosophical knowledge on the model of empirical sciences.” Hacker (1996), p. 114: “Philosophy can make no claims that are testable in experience or subject to falsification by the discovery of new facts. New facts may lead to the formation of new scientific theories, which may in turn involve new conceptual articulations which give rise to philosophical puzzlement (e.g. in quantum mechanics); that is grist for the philosopher’s mill, not a verification or falsification of anything he may legitimately assert.” 14 Hacker (1996), p. 111: “On the one hand, philosophy is characterized as a quest for a surveyable representation of the grammar of a given problematic domain, which enable us to find our way around when we encounter philosophical difficulties. On the other hand, philosophy is characterized as a cure for diseases of the understanding. These different aspects correspond to the difference between connective analysis and therapeutic analysis […].” Hacker (1996), p. 37: “Side by side with his demolition of philosophical illusion in logic, mathematics, and philosophy of psychology, he gives us numerous overviews of the logical grammar of problematic concepts, painstakingly 11
54
Wittgenstein: Issues and Debates
understanding’ diseases, which leads us to cross the bounds of sense. The main cause of philosophical diseases is our lack of synoptic representations. Usually, we use words without questioning their meaning. This can easily lead us to misunderstandings. On the other hand, it is necessary to analyze the grammatical connections between expressions of language in order to cure these diseases. This analysis is called ‘connective’ by opposition to an in-depth analysis (of tractarian type). It builds synoptic representations of our grammar and reminds us of the real use of words, that is, the real function of words. It establishes links, because it reminds of the possibilities of use of concepts involved in the philosophical knots. In order to build a synoptic representation we have to remember the use of concepts, not to look for new information. We need to rearrange what we all know without seeing it, that is, things always lying “under our eyes”. As a result, the map of our grammar gradually appears15. This is the first constructive, non destructive, side of Wittgenstein’s project. In Wittgenstein’s second philosophy, the question of the nature of language is still at the heart of his thoughts and his criticism of Metaphysics. His conception of language is another constructive aspect. Although the Tractatus presented language as a strictly defined calculus, Wittgenstein progressively came to compare it to a set of heterogeneous games, no longer reducible to a set of elementary propositions (logical combinations of names referring to absolutely simple objects) which the analyst would have to discover a posteriori. Language-games are internally connected with various shared forms of life. They are rooted in a common practice. Understanding a word does not consist in grasping some internal representation: it rather is the capacity to use it in accordance with rules fixed by the language-users. Words are like tools. They have functions. Not all words are referring devices. Language is not taught by ostensive definitions presupposing no language. And so on16. tracing conceptual connections that we are all prone to overlook.[…] Providing such a perspicuous representation of some segment of our language, elucidating the conceptual forms and structures of some domain of human thought that is philosophically problematic, is a positive, constructive achievement that is complementary to the critical and destructive task of shattering philosophical illusion, destroying philosophical mythology, and dispelling conceptual confusion.” 15 Hacker (2001), p. 37, and Hacker (2000), pp. 18-9. 16 Hacker (1996), p. 125: “The critical destructive work is counterbalanced by the constructive account. In place of the conception of language as a calculus of rules, we
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
55
We shall now come back to our two initial questions. Is this an antimetaphysical reading of the second Wittgenstein? And does this plausibly show that metaphysical propositions are nonsense? Both of these shall receive a negative answer. First of all, P. Hacker attributes metaphysical believes to Wittgenstein. He develops at length Wittgenstein’s antiscientism17. The core idea is that mankind occupies a particular place in nature. In other words, human beings are a kind of being different from stones, rivers, stars, tables, computers, etc. This non reductionist view is the negation of a metaphysical thesis (which takes various forms) about the nature of the human beings. At least, it means that human being are not only physical or biological mechanisms. P. Hacker would reply that this is not a metaphysical thesis. He could insist on the fact that this is neither a description of reality as it is in itself nor of our conceptual scheme. However, he does not seem perfectly coherent on this. This shows it: “It is striking that all of these characterizations, most of which are true, are dependent upon a more fundamental feature – namely, that mankind is unique in nature in possessing a developed language. The languages of mankind enable us to describe the world we experience, to identify and reidentify objects in a spatio-temporal framework and to distinguish the objects we experience from our experiences of those objects. Knowledge of truths of reason is knowledge of the norms of representation, and of the propositions of logic correlative to the inferential rules, of the conceptual scheme constituted by a language.”18 are offered a conception of a language as a motley of language-games. Language is indeed rule-governed – in the more or less loose manner in which games are rulegoverned. Indeed, speaking a language is comparable to playing a game, and a language to motley of language-games. The use of language is interwoven with the life and practice of language-users. Speaking a language is part of an activity, which is partly constitutive of a form of life. Training and, later, teaching underpins the mastery of language, and these presuppose shared reactive and behavioural propensities within a linguistic community. Words are comparable to tools, and the diversity of their employment is as great as that of different tools – hence masked by conceiving of their essential function as naming. Even declarative sentences are used for endlessly diverse purposes, of which describing is only one, and non-declarative sentences are misrepresented by conceiving of them as containing a force-indicator and a descriptive, truth-value-bearing sentence-radical. Moreover, the concept of describing is itself not uniform; for describing is not uniform; but describing a scene is (grammatically or logically) altogether unlike describing a dream, describing the impression something made is quite different.” 17 Hacker (2001), chapter 2: The autonomy of human understanding. 18 Ibid, p. 56.
56
Wittgenstein: Issues and Debates
What is this “knowledge of our conceptual scheme”? Is this not in a kind metaphysical knowledge? As we have seen, he said that Metaphysics can be conceived as the study of our conceptual scheme. He also said that Wittgenstein rejected this conception. But it is difficult to understand how a grammatical rule differs from a description of our conceptual scheme. How can it be that metaphysical propositions (understood as description of our conceptual scheme) describe nothing at all? One can believe that “A move takes place in space” is just an arbitrary convention, which constitutes our conceptual scheme rather than describes it. However, the proposition ““A move takes place in space” is just an arbitrary convention.” is not a convention, but a metaphysical description. Knowing grammar cannot mean “stipulating or deciding what the norms of representation are”. Wittgenstein did not reject reductionist grammars because he didn’t like them. He argued against this. And his arguments tried to show that in metaphysics of mind are doubtful. If all this is true, P. Hacker avoids the problem of metaphysical propositions. His reading does not show that the second Wittgenstein destroyed Metaphysics. 3.
Pyrrhonian’s interpretation
According to the pyrrhionian readers, P. Hacker misunderstands Wittgenstein’s anti-metaphysical conception of Philosophy. They think P.Hacker is wrong to take seriously the constructive aspects of Wittgenstein’s Philosophy. All theses, they say, are just examples of philosophical nonsense. These constructive aspects are only illusions. Wittgenstein has intended to reject them too. For Pyhrronians, nothing should remain! Their main argument against P. Hacker is stated by J. Conant. It extends our previous remarks. We can sum it up as follows19. 19
Conant (2002), pp. 74-5. His paper was published in French: “L’idée que le but de Wittgenstein est de nous montrer qu’il y a quelque chose que nous ne pouvons pas faire va de paire avec la tendance à le lire comme quelqu’un qui, à chaque étape de son évolution, a cherché à indiquer à l’avance les conditions de tout discours pourvu de signification – que l’on situe celles-ci dans la structure logique du langage, dans les règles de grammaire, ou dans la forme de notre pratique linguistique. Chaque fois, Wittgenstein aurait eu pour but de montrer au philosophe que ce qu’il dit est du nonsens parce qu’il a violé certaines conditions. L’intégrité de sa méthode de critique philosophique semble, au bout du compte, exiger de Wittgenstein qu’il soit en mesure, en premier lieu, de spécifier quelles sont les conditions en question. Cette exigence, à son tour, donne corps à l’idée qu’il est tenu de souscrire, à l’exigence d’un ensemble de quasi-vérités – nommées principes de la syntaxe logique, règles de la grammaire,
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
57
J.Conant’s target is an argument, which aims at showing that metaphysical propositions are nonsense. This is the following: (1)
A proposition can be empirical or a priori. A priori propositions either are analytic or synthetic.
(2)
Certain “propositions” p, q, r,…,n bound the conditions of intelligibility of language. They show that synthetic a priori propositions are nonsense.
(3)
Metaphysical propositions are synthetic a priori.
(4)
If (2) and (3) are correct, then Metaphysical propositions do not fit the conditions defined by p, q, r,…,n, and are nonsense.
According to J. Conant, we are faced with difficulties here. What about p, q, r,…, n? The most plausible answer is: (5)
p, q, r, n are synthetic a priori too.
We are now caught on the horns of a dilemma: (6)
Either, p, q, r,…, n are nonsense or these propositions are true in new and mysterious sense.
J. Conant thinks that none of these solutions is comfortable. He concludes that we should reject (2). No proposition bounds the conditions of intelligibility of language. If P. Hacker was right, accounting for the fundamentally therapeutic and anti-metaphysical aspect of Wittgenstein’s works would be impossible. For Philosophy to be a therapy requires a radical criticism of philosophical propositions, not their replacement by mysterious “quasi-truths”. Pyrrhonians intend to defend the therapeutic side at all costs.
ou jugements constitutifs du cadre de notre pratique – qui marquent les limites du sens et révèlent le point exact au-delà duquel le philosophe s’est égaré. Ces quasi-vérités doivent se tenir hors du domaine des vérités empiriques, tout en se tenant juste en deçà de la limite au-delà de laquelle commence le domaine du non-sens philosophiquement prohibé. Mais sitôt qu’on soulève la question de savoir quelle sorte de vérité attribuer à ces quasi-vérités, commencent verbiages et tergiversations. Car la conception du nonsens qui émerge de la spécification d’un tel ensemble de conditions de possibilité de tout discours pourvu de signification requiert toujours une distinction entre deux sortes de propositions : les propositions empiriques ordinaires, et les propositions logiques, grammaticales, ou appartenant au cadre – les conditions qui pèsent sur les premières, et le problème étant que celles-ci finissent toujours par laisser transparaître, à leur propre aune, leur échec à honorer les conditions de signification qu’elles cherchent elles-mêmes à articuler.”
58
Wittgenstein: Issues and Debates
How do they account for it? According to D. Stern, Wittgenstein’s method aims explicitly at reversing the Socratic methodist method. 20 Socrates dismisses concrete uses of words. What really matters in Philosophy is to find the common element to all the particular cases, the pure essence of the thing. However, accounting for the essence of knowledge, for example, does not entail a detailed examination of concrete uses. Wittgenstein’s second philosophy can be read as a defense of the first answer of Theaetetus to Socrates: “The best we can do in answering questions about the essence of a word like “knowledge” is to give examples in order to show that what Socrates is talking nonsense, and so return words from their metaphysical use to their ordinary use.”21
Wittgenstein has rejected essential definitions and has insisted on the family resemblance instead. That is why, in the Investigations, Wittgenstein did not try to present his thoughts more theoretically and systematically. In the Investigations, the main methodological strategy is called “the method of the paragraph 2” by D. Stern. Its first stage consists of the statement of a philosophical position “which usually emerges out of an exchange with another voice”22. In § 1, for example, we find the idea that to have a meaning each word has to be directly correlated with an object. Then, Wittgenstein describes a set of circumstances in which this idea is correct. In § 2, for example, Wittgenstein writes: “let us imagine a language for which the description given by Augustine is valid.” After 20
Stern (2004), pp. 13-15. Contrary to what he claims, his reading is a version of pyrrhonianism: “In the Philosophical Investigations, one of the principal reasons for Wittgenstein’s opposition to systematic philosophical theorizing is that our use of language, our grasp of its meaning, depends on a background of common behaviour and shared practices – not on agreement in opinions but in ‘form of life’ (§ 241). But to say this so quickly is potentially misleading, for a great deal turns on how one understands the ‘agreement’. Most readers take it to be a gesture towards a positive theory of practice or the place of community in a theory of meaning. I shall be proposing that we take Wittgenstein at his word when he tells us that the work of the philosopher ‘consists in assembling reminders for a particular purpose’ (§ 127) – that the remarks about common behaviour, shared practices, and agreement in opinions are intended as reminders of what we ordinarily do, reminders assembled for the purpose of helping his readers see the shortcomings of certain theories of knowledge, meaning, and the like.” 21 Stern (2004), p. 14. 22 Ibid, p. 10.
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
59
describing these circumstances he proposes the language game with the two builders. The third stage consists of “the deflationary observation that the circumstances in question are quite limited, and that once we move beyond them, the position becomes inappropriate.” 23 . In §2 and §3, Wittgenstein asserts that Augustine’s description of language is not exhaustive. It does not represent all that is called ‘language’. If we do not acknowledge this fact, we run the risk of sublimating our ordinary concepts such as “name”, “proposition”, “object”, etc. In the Investigations, Wittgenstein faces several paradoxes. The paradox of ostensive definition says that any ostensive definition can be variously interpreted in every case (§ 28, 26-31). The paradox of explanation says that an explanation hangs in the air unless it is supported by another one (§ 87). According to the paradox of rule-following, any rule can be interpreted in a number of mutually incompatible ways, so that no rule can determine by itself what we make or shall make. All these paradoxes can be reduced to the following: no expression (ostensive definition, explanation, rule) is intrinsically meaningful because any determination of meaning depends on an interpretation. Our future actions cannot be determined by a unique act (whatever it is) because the question of its interpretation always arises. To dissolve the paradoxes we have to take care of the context in which our expressions are used. According to D. Stern, Wittgenstein has rejected all philosophical theses because:
23 24
1)
They create paradoxes.
2)
These paradoxes can lead us to sublimate our understanding of ordinary concepts. It can lead us to believe that appearances are totally misleading and reality lies behind them. For example, real names - which establish the foundations of language - are not those we ordinarily think about, but “this” or “it”, that is, words supposedly always denoting an object, contrary to ordinary names refer to destructible concrete objects such as ‘Nothung’.
3)
We never fall into paradoxical situations when words are not used in metaphysical discussions. In their ordinary context, expressions create no paradox because we clearly understand them24. Usually if the role played by a word in language is already clear, teaching its meaning by means of ostensive definitions do not raise problem. Concerning explanation, one can recall that unless we intend to avoiding
Ibid. Ibid, p. 20.
60
Wittgenstein: Issues and Debates misunderstandings, no explanation needs to be supported. The paradox of rules has a similar solution.
Most of the readers of the Investigations see it as a dialogue between two voices. One is the “voice of the temptation”, asserting erroneous philosophical theses. The other is the “voice of the correction”, Wittgenstein’s one, which refutes and presents correct philosophical positions. According to D. Stern, this last voice is not Wittgenstein’s. He thinks a third voice exist whose role is to make an ironic comment concerning the other voices’ exchanges, consisting mainly in reminding banalities about ordinary use of language they both neglected. This third voice does not commit to any philosophical opinion. This voice of Ordinary looks like what C. Diamond called “Realistic Spirit “. Let us see now what “Realistic Spirit” is. C. Diamond distinguishes “Realistic Spirit” from Metaphysics25. She seems to think of Metaphysics as a kind requirement. This requirement can be ontological when constraints about what must or can exist are imposed. One can say, for example, that only simple and indestructible objects possibly exist. It also can be logical or conceptual when one asserts that a real proposition has to possess certain ideal characteristics such as a logical form, or when one requires for authentic concepts to be sharply bounded. On this view metaphysical requirement is a kind constraint aiming at appraising the truth of ordinary beliefs. For example, understanding the nature of propositions requires the general essence of proposition be grasped. What is “Realistic Spirit”? According to C. Diamond, “realism” refers in Philosophy to views underlying the importance of mind/experience-independence 26 . But “realism” to which “realistic spirit” refers to is taken by her in its usual meaning 27 . She points to different characteristics of the ordinary use of “realism”. But I confess not to be completely sure to grasp them. Nonetheless, they appear to be the followings. Sometimes we use ‘realism’ to speak of people or literary fictions. In the former case, “realist” refers to someone who does not believe in spite of obvious facts. Someone can be said to lack realism if, basing himself on a theory or pious wishes, he pretends to know what reality must be and nothing in reality could weaken his faith. In the latter case, “realist” qualifies fictions in which certain things do not happen. People do not travel through time, apples do not 25
Diamond (2004), pp. 27-8. We had no access to the English version of Diamond’s book. So we indicate references of the French edition. 26 Ibid, p. 56. 27 Ibid, pp. 56-8.
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
61
speak French, material objects such as books do not suddenly evaporate, persons do not have multiple locations at the same time, are examples. Finally, she insists on “realistic coherence”. She seems to mean that in a realist world, consequences of action and causal relations between events should not be unbelievable or impossible. “Realistic Spirit” is a kind of attitude towards reality It aims at showing that philosophical theories are nonsense. Correctly understood, it should lead us to give up our need to find Metaphysical explanations or foundations to ordinary ways of thinking28. It should as well lead us to renounce Metaphysical problems such as “Is there external material objects?” Do Pyrrhonians offer an anti-Metaphysical reading of the second Wittgenstein? We think a negative answer we should be given. The “Realistic Spirit” has ontological commitments. These appear to be rather close to common sense’ commitments. In other words, C. Diamond’s “Realistic Spirit” looks like a common sense Metaphysics program, that is, a kind of metaphysics, which aims at defending ordinary beliefs about the general structure of reality. As we have seen, “Realistic Spirit” is an attitude towards reality. C. Diamond seems to believe that there are facts, causal relations, change, person and other objects with persistence conditions, etc. Her world is not made of elementary particles! Moreover, its program looks rather dogmatic. “Realistic Spirit” plays the role of an absolute standard by mean of which philosophical theories are assessed. One can ask for “which function is supposed to play the Realistic Spirit?” And I think the answer should be: it fixes the conditions of intelligibility of language. Pyrrhonians will probably deny it. But, then, how could they justify that metaphysical propositions are nonsense? In such a case, this latter view would become arbitrary. Pyrrhonians need a distinction between correct (ordinary) and incorrect (metaphysical) uses. “Use” should refer to a normative concept. If “Use” were a descriptive concept, no critics of Metaphysics were possible. Metaphysical concepts trivially have use. One cannot acknowledge that philosophers use concepts, propositions, aim at building, defending or criticizing theories. Then, it is not surprising to wonder, as P. Hacker does, whether there is a difference between correct and incorrect uses of words. The latter question would give birth to a properly philosophical elaboration. This would raise problems belonging to Philosophy of language as well as Metaphysics. Indeed, a look at Philosophy of language shows that it raises metaphysical problems. The 28
Ibid, p. 100.
62
Wittgenstein: Issues and Debates
problem of apparent reference to nonexistent and the problem of negative existential are examples. One could say that this elaboration would be useless because 1. we intuitively know that philosophical propositions are nonsense, and 2. we have an innate and infallible capability, to identify philosophical nonsense. One could also say we only need a grasp of our ordinary concept of nonsense29. For Pyrrhonian, a philosophical nonsense is gibberish. But 2500 years of Metaphysics show how implausible would be this hypothesis. Furthermore, a philosophical proposition has nothing in common with gibberish. It is made of meaningful items, has a grammatically correct structure, and logical inferences can be derived from it. None of these can be said about gibberish. Conclusion These two readings seem to be partially correct. But, as D. Stern writes, they build Wittgenstein as more doctrinaire than he really is30. His works seem to be inhabited by a tension between two strengths among which one tends to give a definitive answer to the problems of Philosophy, whereas the other tries to eliminate them. The second Wittgenstein seems to have tried to minimize the constructive aspects of his Philosophy. How eloquent is his use of the concept of family-likeness, in § 65 of the Investigations and elsewhere 31 . He has acted as this concept exempted him from constructing theories. It is used as a theory-deflation tool. But, as it is shown by P. Hacker and others, we find constructive aspects, such as the conception of meaning as use. This tension is not virtuous, contrary to D. Stern’s opinion. It plausibly indicates Wittgenstein fell into a trap. Did he show that Metaphysics is illegitimate? If what has been said is true, it is not the case. None of these readings expounds anti-metaphysical philosophy. None of these is really consistent. Of course, it does not follow for a certainty that Wittgenstein was a metaphysician. It nevertheless shows the necessity to minimize anti-metaphysical claims repeatedly made by Wittgenstein. In other words, it is time to give up our picture of Wittgenstein as the sworn enemy of Metaphysics. Moreover, metaphysical researches are very active today; his writings could give us resources to face its problems. 29
Baier (1967). Stern (2004), pp. 36-7. 31 See the last page of the Blue Book. 30
Critical Remarks on Anti-Metaphyscial Readings of Wittgenstein
63
Bibliography Armstrong, David, 1978: Nominalism and Realism, Universals and Scientific Realism. Vol. I. Cambridge University Press, Melbourne. ---
1997: A world of States of affairs. Cambridge University Press, Melbourne.
---
2004: Truth and Truthmakers. Cambridge University Press, London.
Baier, Anet, 1967: Nonsense. In: McMillan Encyclopedia of philosophy. 1, Vol 5, pp. 520-522. Conant, James, 2002: Le premier, le second et le dernier Wittgenstein. In: Wittgenstein, Dernières pensées. ed by J. Bouveresse, S. Laugier, J. J. Rosat. Agone, Paris, pp. 49-88. Crary, A. and R. Read (eds.), 2000: The New Wittgenstein. Routledge, London. Descombes, Vincent, 1995: La denrée mentale. Les éditions de Minuit, Paris. Diamond, Cora, 2004: L’Esprit Réaliste, Wittgenstein la philosophie et l’esprit. (french trans by E. Halais and J. Y. Mondon). PUF, Paris. Hacker, P.M.S., 1972: Insight and Illusion, Wittgenstein on Philosophy and the Metaphysics of Experience. Oxford University Press, New York. ---
1986: Laying the Ghost of the Tractatus. In: Wittgenstein Critical Assessments. Vol. 1 (ed. by S. Shanker). Routledge, London, pp. 7692.
---
1986: Wittgenstein Today. (with G. Baker) In: Wittgenstein Critical Assessments. Vol. 2 (ed. by S. Shanker). Routledge, London, pp. 2436.
---
1996: Wittgenstein’s place in the twentieth-century analytic philosophy. Blackwell, Oxford.
---
2003: Wittgenstein, Carnap, and the new American wittgensteinans. The Philosophical Quarterly, 2003. Vol. 53, No. 210, pp. 1-23.
---
2000a: Wittgenstein. (French trans by J. L. Fidel). Seuil, Paris.
---
2000b: Wittgenstein, Mind and Will. Blackwell, Oxford.
---
2001: Wittgenstein: connections and University Press, Clarendon, Oxford.
controversies.
Oxford
64
---
Wittgenstein: Issues and Debates
2005: Wittgenstein: Understanding and Meaning. 2nd ed., Blackwell, London.
Loux, J. Mickael, 2006: Metaphysics: A Contemporary Introduction. 3rd ed., Routledge, Oxon. Lowe, E. Jonathan, 1998: The possibility of metaphysics, Substance, Identity, and Time. Oxford University Press, New York. ---
2002a: A Survey of Metaphysics. Oxford University Press, New York.
---
2002b: La connaissance métaphysique. Revue de Métaphysique et de Morale. October-December 2002, No 4, pp. 423-441.
---
2007: La métaphysique comme science de l’essence. (French transl. by R. Pouivet). In: Métaphysique Contemporaine. (ed. by F. Nef), Vrin, Paris, pp. 84-120.
Putnam, Hilary, 1998: After metaphysics, what? In: Metaphysics: The Big Questions. (ed. by P. Van Inwagen and D. Zimmerman). Blackwell, Malden, pp. 388-392. Stern, David, 2005 : Comment lire les Recherches Philosophiques? Philosophie, Nº 86, pp. 40-61. ---
2004: Wittgenstein’s Philosophical Investigations, an Introduction. Cambridge University Press, Cambridge.
Van Inwagen, Peter, 1998: The nature of Metaphysics. In: Contemporary Readings in the foundations of metaphysics. (ed. by C. McDonald, S. Laurence). Blackwell, Oxford, pp. 11-21. ---
2002: Metaphysics. (2nd ed). Westview Press, Cambridge Ma.
Wittgenstein Ludwig, 1965: Cahier bleu et Cahier brun. (french. Trans. by G. Durand). Gallimard, Paris. ---
1980: Grammaire philosophique. (french. Trans. by M. A. Lescourret). Gallimard Folio Essais, Paris.
---
2004: Recherches philosophiques. (french. Trans. by E. Rigal, D. Janicaud, F. Dastur.). Gallimard, Paris.
---
2005: Philosophica IV. (french. Trans. by J. P. Cometti, E. Rigal). TER, Mauvezin.
---
2006: De la certitude. (french. Trans. by D. Moyal-Sharrock). Gallimard, Paris.
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy? Ayúegül ÇAKAL1 Introduction In a paper of mine – How to Make Opposite Ends Meet 2– I contrasted Wittgenstein’s understanding of meaning in his two periods in several respects. One respect was his position on language. I argued that Wittgenstein has two opposite approaches to meaning in his first and second periods: Meaning as representation (viz. picture theory of meaning) and meaning as use. In his first period, where he adopted a theory of meaning as representation, he considers language, the sentences of which are pictures of reality, and which is the vehicle of thought, as independent from reality, standing for the objects or states of affairs in the world. However, in the second period, where the use conception of meaning is put forward he conceives language as something dependent on the human agents that employ it, hence he conceives it as something into which actions are interwoven. In this sense, Wittgenstein, in his second period, can be thought to consider language as an action in the world, thence consider it as in and part of reality, when it comes to the semantic connectedness between the two.3 In that sense, Wittgenstein can be thought 1
I am indebted to Prof. Peter Hacker for his very careful reading of my paper, his invaluable remarks and suggestions for the refinement of the paper, and for his mindopening detailed explications at some points in my paper as well as on my peculiar questions on the issue. I am also grateful to Prof. Ali Karatay who contributed on the very emergence of the ideas in this paper through our long-lasting discussions and through his useful comments on some key points. Without their respectable contributions, this paper would not have appeared. 2 Cakal, 2006. 3 For in the latter view praxis as a part of reality is included in language; since language and reality interwoven (as the speaking of language is part of an activity or of a form of life –hence language is woven in a form of life). The meaning of linguistic expressions is not independent of humans as it was in Tractatus. In the Investigations it is the way they are used; it depends on how they are used (by humans –in a society). Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ), Ontos Verlag, Frankfurt a. M., 2010, 65-81.
66
Wittgenstein: Issues and Debates
to provide us with a monistic picture of the world. My claim, in this paper, is that the gist of this radical change on the semantic connectedness between language and reality can be found in the private language argument given by Wittgenstein in Philosophical Investigations (hereafter the Investigations). For, bearing in mind that this is a semantic claim, I think, the repudiation of private language in Wittgenstein’s philosophy amounts to the repudiation of the meta-logical connection or the representational relation between language and world –where they can be presupposed to be distinct and independent entities from each other; and hence, the traditional dichotomy of mind-body or inner-outer, which is inherited from Descartes. This, I think, also means the rejection of the traditional distinction of language (as a vehicle of thought) and world as two different entities, where skepticism is possible. In parallel with the rejection of this dichotomy, in this paper, by focusing on Wittgenstein’s understanding of meaning by which his overall philosophy is structured, I will examine his rejection of the representational theory of meaning put forward in the Tractatus and will evince the effect of it to his later understanding of meaning. For this, I will first mention the traditional (dualistic) understanding of the world very briefly; and then, give a synoptic view of Wittgenstein’s conception of language and meaning in general to be able to see what the radical change in his understanding of meaning is. Then, I will present the private language argument within its effects on his later understanding of meaning in order to be able to see why it is a rejection of identifying representational approaches concerning the notion of meaning, and thence, renders private (internal) ostensive definition unintelligible and mere internal part of the dualistic view something meaningless. Finally, I will come to the conclusion that the gist of the radical change in his view on meaning and language can be found in his private language argument. In the traditional view, language is usually taken to be a vehicle of thought and thought is usually correlated with mind, which is, by Descartes, considered to be one of two substances, viz., matter and mind. ‘Matter’ is thought to be defined by extension, and is the totality of “spacematter” –i.e., the plenum (individual material things are modes of matter), Language is inseparable from humans –human actions. While in the Tractatus the statement that the proposition that p is made true by the fact that p is a meta-logical connection between language and world, in the Investigations, it is an intragrammatical relation, viz., ‘the proposition that p’ is equal to ‘the proposition made true by the fact that p’.
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
67
whereas ‘mind’ is considered as something by which perceptual and intellectual processes are activated. In this sense, mind is, in general, seen as something that provides us with understanding through mental faculties as well as that which perceives.4 According to Descartes, mind is a being “which affirms, which denies, which wills, which rejects, which imagines also, and which perceives”.5 So, the domain of sense perceptions is in a sense under the control of mind [for a sense perception is a mode of thought, the vehicle of which is language]. So, in this view, it is the mind which sees, smells, feels, and intends to do things. Moreover, in Meditation VI, Descartes argues that the mind, that is, a thinking thing, can exist apart from its extended body. And therefore, the mind is a substance distinct from the body, a substance whose essence is thought. The consequence of this view is, apparently, to contrast the public character of the body with the private status of the mind by drawing a distinction between ‘inner’ and ‘outer.’ In this view, while the operation of mind is considered as ‘internal’, the operation of body is considered as ‘external.’ Now, this dualistic view of the world, when taken to the extreme – that is, to the idea that mental processes happen in isolated fields and human beings are completely confined within the mental happenings of their own so there is in fact no connection between happenings of the mind and those of the other- surely brings about the problem of skepticism; and this problem is labeled by Wittgenstein as something not irrefutable in Tractatus-Logico Philosophicus (hereafter the Tractatus). 6 It seems, Wittgenstein’s bugbear is the problem of skepticism. In the Tractatus, thence, to tackle the problem, as Kenny also remarked, Wittgenstein might have set out the programme of refuting skepticism, with the following lines: Skepticism is… obviously nonsensical, when it tries to raise doubts where no questions can be asked. For doubt can exist only where a question exists, a question only where an answer exists, and an answer only where something can be said.7
It is possible to read Wittgenstein’s programme then undermining skepticism by showing that it is nonsensical. Now, if skepticism is nonsensical -as Wittgenstein maintains- and raises doubts, then what must 4
Cf. Descartes, 1951, 23-33. Descartes, 1951, 27. 6 Wittgenstein, 2004, TLP 6.51. 7 Kenny, 1986, 180. 5
68
Wittgenstein: Issues and Debates
be done, admittedly, is to abate that doubt. His programme that had already been set out in the Tractatus, in this sense, must be to pick out the element that is responsible for skepticism and to refine the representational theory of meaning, which he put forward in Tractatus, in accordance with it. And as the result of this refinement we’ll see there will emerge the use conception of meaning, where there is no room for skeptical scenarios, and which is the very opposite of the Tractarian representational theory of meaning. It can be thought that the doubtful connection between the two components of the dualistic approach can be said to be the culprit for skepticism as the connection between happenings of the mind and those of the other can lead to skeptical problems when they are thought to be separated and when taken to the extreme -as it is noted above. If so, then Wittgenstein must either revise the semantic connection he established between the two (language and reality) in his representational theory in such a way that it is not doubtful anymore so that what he says would be something can be said or in order not to encounter this problem at all, he must eliminate this doubtful connection in his theory as a whole so that there is no need to speculate on where something “cannot be said.” It seems to me that Wittgenstein, in his later philosophy, follows the latter way and eliminates the problem of skepticism by wiping out one component of the duality, namely ‘(mere) internal’8 –or what belongs only to mind. Note here that the internal part of dualism is associated with mind, and mind was associated with language as a vehicle of thought. And I think that Wittgenstein does this [wiping out one component of dualistic view (which is the mere internal)] through his private language argument. In the argument, by rendering bare ostensive definition unintelligible, he apparently eliminates the possibility of having a meaningful and merely internal language, a language the words of which can only be known to the person speaking. So, in his later philosophy, it is not possible for a linguistic expression to be known merely internally (through mere mental faculties); hence, to have a meaning only internally –or by bare (internal)
8
As Descartes did not make the subtle distinction between mind and soul, but Wittgenstein did, in order to provide a conflux and congruence in the terminologies may be used or partly used for both philosophers, I will use the term ‘mere internal’ to be standing for ‘private’ in the sense of incommunicability (not inalienability) in Wittgenstein’s philosophy, and for ‘mind’ or ‘soul’ (which is not ‘matter’) in Descartes’ philosophy.
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
69
ostensive definition. As, for him, ‘mere internal’9 or what belongs only to mind has no meaning. In this sense, his “programme,” can be said to avoid skepticism by rendering the internal (or mental part of dualistic view) meaningless by means of reducing it to what cannot be said, in Tractarian way of approaching the issue, so that there will remain only the talk of external where there is no room for skeptical scenarios. Note here that what cannot be said, in the Tractatus, is considered as something nonsensical or something that lacks meaning.10 So, what is merely internal or private must be proved to be what cannot be said if it is meaningless. And I think he does so through his private language argument by denying the meaningfulness of linguistic expressions that have no criterion for their correctness; and this will explain why one cannot give a(n) (ostensive) definition of a private sensation by focusing on the sensation and on the symbol: It has no meaning as it has no criterion for its correctness. But then the question arises: How about thinking? As a “mental activity” is it also meaningless? Is there no such thing as thinking as the internal part of dualism? Wittgenstein claimed in Blue and Brown Books that thinking is ‘the activity of operating with signs’. 11 As, there, he maintains: …thinking is performed by the hand, when we think by writing; by the mouth and larynx, when we think by speaking; and if we think by imagining signs or pictures, I can give you no agent that thinks.12
Although Wittgenstein later on leaves this strong behaviorism, the idea will bring about, in the Investigations, the idea that words and actions are interwoven. Language games, which are introduced in the Investigations, in this sense, will be thought to be rule-governed activities. Thus for Wittgenstein “teaching of language” will be considered as training, not explanation. 13 Language will be considered as something taught and learned by means of practices; not through one’s own mere
9
Caveat: the sensation names which we use are not ‘mere internal’, according to Wittgenstein. 10 For my purpose, I will not consider the subtle difference between meaning and sense Wittgenstein made in the Tractatus. I will rather use the term ‘meaning’ for both terms in order to provide a congruity in terminology in overall paper. 11 Wittgenstein, 1958, BBB, p. 6 [my italics]. 12 Wittgenstein, 1958, BBB, p. 6 [my italics]. 13 Cf. Wittgenstein, 2001, PI § 5.
70
Wittgenstein: Issues and Debates
internal faculties –like focusing on the sensation and on the symbol –by bare ostensive definition. 1.
A Brief Background: Wittgenstein’s Understanding of Language and Meaning (What is the radical change the gist of which could be found in the private language argument?)
Wittgenstein’s theory in the Tractatus represents an approach to meaning based on word-world semantic connection. What is meaningful according to the Tractarian notion of meaning is the representative of its referent: Language represents reality. In addition, this representative relation depends on the method of projection, which is the method of correlating name and nominatum. In this respect, the thinking subject introduces correlations between entities within a realm independent of it. So, one can grasp the meaning of linguistic expressions through one’s ability to make correlations between words and entities. Thereby, one can understand the meaning of linguistic expressions by oneself through putative acts of meaning or one’s own mere internal faculties. Moreover, if one can grasp the meaning of linguistic expressions given through the representation relation above-mentioned, then one could think that (Cartesian) dualism is embedded in the picture theory as mind provides us with understanding the word-world semantic connection.14 14
At this point one may object that since in Wittgenstein’s account it is a world of objects and facts; it’s a world which lacks selves, values, cognitive relations (such as belief), and God, the world Tractatus presents is monistic, not dualistic. Yet, I think that it is dualistic. For: (i) In the Tractatus, it is said “We picture facts to ourselves.” [Wittgenstein, 2004, TLP 2.1, my italics] So, as Kemerling notes (Philosophy Pages, http://www.philosophypages.com/hy/6s.htm, 24.04.2007). Human beings are aware of the facts by virtue of our mental representations or thoughts, which are most fruitfully understood as picturing the way things are.” Tully also touches upon the same point and writes that Tractatus is not concerned with determining the conditions of good representation or with deciding when a fact must forfeit its claim to be a picture. For him, it is enough that there are many non-failures; we do “picture facts to ourselves.” (ii) As Tully writes in his article called Tractarian Dualism, “Wittgenstein seems to have the activity of thought in mind although he elaborates very little beyond remarking that “a propositional sign applied and thought out is a thought” [Wittgenstein, 2004, TLP 3.5] and immediately after that “a thought is a proposition with a sense”.” [Wittgenstein, 2004, TLP 4.0] For Tully, “Nothing
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
71
As to Wittgenstein’s later views on meaning –especially in the Investigations, for him, what linguistic expressions mean is not the representation relation between world and language -not the objects or states of affairs that correspond to reality, as it was in Tractatus, but it is rather, the use of those linguistic expressions in a language game by humans, and which is “the whole, consisting of language and the actions into which it is woven.” 15 Therefore, according to Wittgenstein’s later view, uttering something is considered as an action in the world; when one uses a term in the language game it belongs, actions are also inherent in the use of the term. Therefore, Wittgenstein does not view language as a fixed and timeless framework but rather as a vulnerable and changeable aspect of human life. In this sense, he denies the semantic connectedness between words and world, which gave words their meanings. Thus, we cannot look outside the linguistic practice to find, that which governs it; we cannot “sublime the logic of our language”.16 What we have is after all language and the forms of life, which arise from language and make it possible. It is this idea that brings about the idea that words and actions are interwoven. would be a Tractarian picture if we did not cast some fact into a representational role, and no propositional sign would project any possibility at all if it were not our expression of that possibility. The semantical circuit, according to Tully, between a proposition and its sense is not closed until a further component is introduced, for what also finds its expression in a proposition is a thought.” For, as Wittgenstein noted, “The method of projection is to think the sense of the sentence.” [Wittgenstein, 2004, TLP 3.11] Hence, for Tully, in order for propositions have any sense at all, they have to be applied and thought. He thinks since Tractarian propositions points us towards the world, they add a dimension to the world. The terms Wittgenstein uses like “display” (display logical form [Wittgenstein, 2004, TLP 4.121]) or “show” (they show their sense [Wittgenstein, 2004, TLP 4.022]), according to grammatical terminology, do take indirect objects. That is, they are 3-term relations. And subject, the “I” to whom the sense of a proposition might be shown is, according to Tully, not mentioned explicitly up until later in Tractatus. For him, till then only grammar is served by this move. So, by following up especially Tully, who defends the view that the Tractatus points away from monism towards dualism and that Wittgenstein’s concepts of thought, sense, and understanding are an essential part of its structure. As he also remarks, although for some philosophers it seems ironic that Wittgenstein defends dualism in the Tractatus, one can see that in fact he does so in the only form in which it could be defended. 15 Wittgenstein, 2001, PI § 7. 16 Wittgenstein, 2001, PI § 38.
72
Wittgenstein: Issues and Debates
Language games, in this sense, are rule-governed activities. Therefore, for Wittgenstein, “teaching of language is not explanation, but training. 17 Language is taught and learned by means of practices. So, one can know what a word means not by oneself through their representations in one’s mind by establishing the meta-logical connection between word and world (or by picturing the facts), but from the society in which one lives –through its use in public practices. In this respect, in Wittgenstein’s later views, the uses of words are something agreed upon (under the form of life –or grammar). The rules of language (grammar) are analogous to the rules of games; meaning something in language is thus analogous to making a move in a game. In that sense, in playing language games, following rules, and obeying them, as Baker and Hacker emphasize, there must be public criteria.18 If someone satisfies the criteria for giving orders, framing rules, and playing them, asking questions, then she is correctly said to play these (language) games. Note that, in this view, since the meaning of linguistic expressions is formulated as the use of them in the language game they belong; expressions that are not used in a language game –hence have no public criteria for their rightness or wrongness- are not meaningful at all. Private language, we will see, is such a language. 2.
Wittgenstein’s Private Language Argument
Through the emphasis on the inseparability of language from human activities, Wittgenstein argues for the impossibility of a private language by, I think, implementing his new ideas on meaning to emphasize that the meaning of linguistic expressions is not somewhere accessible only by an individual (through one’s own mental capabilities), but it is their uses in public practices. Thus the idea of a private language, which is closely connected with a certain (Cartesian) picture of language explaining how words are defined, learned or how they obtain meaning criticized by Wittgenstein. Wittgenstein defines private language as a language the words of which ‘refer to what can only be known to the person speaking: to his immediate private sensations.’19 If a private language is a language whose 17
Wittgenstein, 2001, PI § 5. Baker and Hacker, 1985, 176. 19 Wittgenstein, 2001, PI § 243. 18
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
73
words refer to the speaker’s immediate private sensations, then one may ask: how is it that words refer to sensations? As, for Wittgenstein, we talk about sensations everyday and give them names, for him, this question should be asked as “how the connection between the name and thing named is set up?” This is according to him the same as asking the question: “how does a human being learn the meaning of the names of sensations?”20 He believes, “...words are connected with the primitive, the natural, expressions of the sensation and used in their place”.21 Sensation names are thus the replacement of the natural behavioral expressions of the sensation by the verbal expression. The sensation-word, however, does not describe (but replaces) the natural expression of sensation; for, according to him, language cannot be used to step in between sensation and its expression.22 In a private language, Wittgenstein argues, words that stand for sensations cannot be “tied up with” the “natural expressions of sensation” because the words of this language have no natural expression.23 As he thinks that all sensations do have their natural expressions and they are publicly observable, any sensation word which has its natural expression for him is not private. For example, our (sensation) word ‘pain’ is not “private” because it can be understood or known by other people through its expression (as it is shared publicly). A specimen: Suppose a person who has a kind of toothache, calls her sensation as an ‘X’ secretly. When she feels that sensation, which she calls “X”, for Wittgenstein if she shows ‘the natural expression’ of her sensation, in our case, say, she holds her cheek in a sulk, “X” is not a word in a private language. For it has its (publicly sharable) natural expression. Now the question arises: What if one does have the sensation but does not show any natural expression for it? In such a case, Wittgenstein avers, her sensations remain unexpressed and no one else can know that she has the sensation. However, from this one cannot infer that X would be meaningless because it could be expressed and its expression would be a criterion for ascribing X to that person. Then, one can further ask: what if human beings do not show outward signs of sensations at all? If that were the case, Wittgenstein asserts, by pointing out the connection between sensation and the expression of sensation in human life, that if human 20
Wittgenstein, 2001, PI § 244. Wittgenstein, 2001, PI § 244. 22 Wittgenstein, 2001, PI § 245. 23 Wittgenstein, 2001, PI § 256. 21
74
Wittgenstein: Issues and Debates
beings did not show outward signs of sensations, which amounts to saying that if sensations were to lack expression, then it would be impossible to teach the use of sensation-words.24 The remarks offered so far concerning private language as inventing names are obviously not enough to show Wittgenstein’s claim that private language is impossible. The impossibility claim is made in PI § 258. As a private language is something impossible, from now on I will talk about it as putative private language. Now, let us be clear about what is it that is impossible: The thing which is impossible questioned in the argument by Wittgenstein is whether one can name25 one’s own sensations and if so, whether one could by oneself understand and use those names in one’s private language. He investigates this through his celebrated diary example in PI § 258: Imagine someone who wants to keep a diary about the recurrence of a certain sensation, but in accordance with the putative private language, viz., the words of which can only be known to the person speaking. In order that one may do so, Wittgenstein thinks, one has to “associate” one’s sensation with the sign ‘S’ and write this sign in a calendar every time one has the sensation. Such a diary in fact cannot be kept according to Wittgenstein due to lack of any criterion of correctness in recollecting what ‘S’ means. So, in a (putative) private language, there is no difference between ‘something’s being right’ and ‘thinking that it is right’. In this sense, ‘S’ lacks criteria for its correct use. And as one cannot know whether what one says is meaningful, one cannot know what one means by it according to Wittgenstein. Therefore, one cannot name one’s own (putative) private sensation. Since, for Wittgenstein, one learns the meanings of words from society through their use in a social environment, which determine their senses, to understand a word is to know how to use it, the meaning of a word is the use that we make of that word; the difference between correct and incorrect use is manifest in publicly observable behavior. Then one may ask: What is the criterion for a public language, which is not a private one? The use of language, for Wittgenstein, is fundamentally rule following and rule following is a practice, which is the result of custom and training within a social milieu. A language is thus based on an agreement for him. It is agreement in form of life; and it is what makes linguistic 24
Wittgenstein, 2001, PI § 257. By ‘name’ he means ‘word whose meaning is learned by bare ostensive definition. (cf., Wittgenstein, 1958, BBB, p. 82) 25
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
75
expressions meaningful. As in PI § 242, Wittgenstein argues, in order for a language to be a means of communication, there must be agreement not in definitions but also in judgments. The reason, in the diary case, for the impossibility of naming a private sensation, for Wittgenstein, is that a definition for the private sensation ‘S’ cannot be formulated (due to the lack of any criterion of correctness in recollecting what ‘S’ means). Moreover, this claim is the gateway to the arguments that private ostensive definition is unintelligible and private language impossible. A definition for the private sensation ‘S’ cannot be formulated because one cannot know what one means by it. For, in order to mean something, it must satisfy the condition for its meaningful use. In order for it to be meaningful, it must have grammar –or rule- with which it can be checked for correctness of its use. So in order to define the sign ‘S’, a grammar or a ‘stage setting’ under which the sign can be used correctly or incorrectly is needed. Now, since ‘S’ is the putative name of a private sensation, which is impossible for another person to know or have, hence which it has no grammar to check for its correct use, a criterion for its correct use cannot be given. Therefore, it does not mean anything. Then, the question arises: can I myself not determine the standard of the correct use of the sign ‘S’? Or one may ask this question in another way: Can I not give myself an ostensive definition –a private mental ostensive definition- that would function as a criterion for the right or wrong use of ‘S’? For Wittgenstein, it cannot be given either; because the definition of a name serves to establish its meaning. That is, the function of a definition is to provide the meaning of a sign and establish a criterion for the correctness of its future use. And one cannot achieve this by simply concentrating on the sensation as speaking or writing ‘S’ and thereby memorizing the relation between the sign and the sensation. For one thing, there is no criterion of correctness in recollecting what ‘S’ means, in other words, one cannot remember whether ‘S’ means this –or the sensation one has. In our (public) language, there is no such question; because we follow rules in the course of the practices of everyday life. But in this case there is no criterion (rule, grammar, or stage setting) in whether one is right or wrong in recollecting what ‘S’ means. Then again one may ask what if there is such a person who remembers the connection by means of concentrating her attention on the sensation through a mental pointing or “pointing inwardly”, i.e., a private
76
Wittgenstein: Issues and Debates
mental ostensive definition. Even then, Wittgenstein would reject it. His rejoinder would be the criticism of the way of reasoning. He censures this way of reasoning by labeling it as an “idle ceremony”.26 Remember that early in the Investigations Wittgenstein argued against the idea that ostensive definitions constitute the foundations of language. 27 He considered the example of someone pointing to two nuts while saying “This is called two” and asked: How does it come about that the listener associates this with the number of items, rather than the type of nut, their color, or even a compass direction? One conclusion of this, for Wittgenstein, was that to participate in an ostensive definition presupposes an understanding of the process and context involved, of the form of life.28 Another was: “an ostensive definition can be variously interpreted in every case”. 29 Therefore, without a stage-setting –a determinate grammar– we already cannot talk about an ostensive definition. We cannot, by ourselves, connect a name and the thing named without an appropriate stage-setting. In the diary case, there is no such setting, which will function as a criterion for the correctness of the use of ‘S’; so, ‘S’ lacks meaning. And if we grant that the definition of a name serves to establish its meaning, then the putative private ostensive definition by reference to a putative memory sample cannot function as a definition. That is to say, ‘S’ conceived as the name of a “private” sensation, which is both privately known (or incommunicable, viz., no one else can know it) and privately owned (or inalienable, viz., no one else can have it) is unintelligible. In this sense, it can be thought that Wittgenstein’s private language argument eliminates the possibility of merely internal meaningful language. Briefly, in a (putative) private language since the subject would not have any criterion of correctness in recollecting what ‘S’ means, hence to decide whether the sign ‘S’ is used correctly or not, it would lack meaning. In this scene, as Kenny emphasizes, when next I call something “S” I do not know what I mean by “S”, thence, according to commentators of Wittgenstein, ‘I do not say anything’30 or such a use is ‘empty’.31
26
Wittgenstein, 2001, PI § 258. Wittgenstein, 2001, PI §§ 23-28. 28 Wittgenstein, 2001, PI § 23. 29 Wittgenstein, 2001, PI § 28. 30 Ayer, 1971, p. 52; Rhees, 1970, p. 66. 31 Strawson, 1954, p. 28; Hacker, 1972, p. 234. 27
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
3.
77
What is not a private language?
Now, having seen in general what the private language argument is, let’s see what is not a private language: From the above mentioned remarks we can see that what Wittgenstein means is not a contingently private language which no one else can understand, but rather, it is a purely private language that it is logically impossible for any one to understand. The words of such language refer to the speaker’s immediate private sensations and inner experiences. So, the question for Wittgenstein does not simply concern the possibility of someone’s inventing a name for her own sensation. For this is something very well possible in Wittgenstein’s philosophy, as it is something knowable by any other person by means of its natural expression or it can be translated into an “ordinary” (public) language. To open this up a bit, let’s consider a specimen: Suppose, now and then, I feel a sensation of throbbing my head twice and then tingling my ear for about 3 seconds right after that and suppose I call it, say, “füútürük.” When it happens, I say, “this is ‘füútürük’ again,” but nobody knows that I call it ‘füútürük.’ Now, according to Wittgenstein, is this, i.e., calling my sensation ‘füútürük’ secretly, a private language? The answer is obviously not. For here, what is the case is only inventing a name for a sensation that which is not unknowable by others; it can be expressed and understood –therefore knowable- by any other person. Thus, it is neither a private sensation nor a private language. Inventing a name for a sensation is not something impossible in his philosophy. So, let’s not confuse it with an invention of a name for a sensation. 4.
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
Wittgenstein’s repudiation of private language whose words acquire meaning simply by being linked to private experiences can be assessed as the rejection of the traditional distinction and accompanying philosophical approaches or principles on such as meaning, mental states, sensations, subject-object duality, and so on. In this respect, although his discussion of private language can be related to many sub-disciplinary issues of philosophy and discussed from various perspectives in philosophy, since his revolutionary character of philosophy is structured by his notions of meaning, here, I have focused on the semantic connection between word and world; and thereby his rejection of the traditional distinction between ‘inner’ and ‘outer’ -language (as the vehicle of thought) and reality-
78
Wittgenstein: Issues and Debates
through his rendering the mere ‘internal’ meaningless, hence rendering bare ostensive definition unintelligible. In the picture theory of meaning, since language and world are thought to be distinct entities, the connection between them has always been in question. It is doubtful, and as the doubtful connection between them brings about questions concerning skepticism, to avoid the threat of skepticism this connection had to be explained indubitably. This, I think, is the reason why it renders problematic the dualistic way of approaching the issue. So, by being very much aware of this threat of skepticism on his theory, Wittgenstein, in the Tractatus, must have set the programme of coming up with a new theory of meaning which avoids the problem of skepticism. Thence, in his later views, by wiping out the doubtful (semantic) connection between language and reality as the culprit of skepticism, he might have wanted to eliminate the distance between word and world. And if we consider bare mental activities consisting in the internal part of Cartesian dualism, it can be said that Wittgenstein’s elimination of the doubtful connection between the two is through reducing the mere internal part meaningless and hence rendering only the talk of external part meaningful; so there is no need to establish a doubtful speculative connection between these two, viz. inner and outer or internal and external. Thereby, he can be thought to exempt himself from being expected to establish a speculative (doubtful) connection between two different realms or entities. Only external would be proper and sufficient to give an account for a meaningful language. Thus, by melting language and human action in one bowl (language games), he can be thought to render them as a unit consisting of reality and language under the external. After all, he repudiated the meta-logical connection between language and world and adopted an intra-grammatical relation instead. It is obviously the private language argument, which assures the inseparability of language from human activities, and hence provides an explanation for an external account of meaning. In the private language argument, Wittgenstein, argued for the impossibility of private language. Private language, as a language knowable by only the speaker herself, is not possible. For, according to him, one cannot assign a meaning to a word by oneself. Neither by an “act of meaning” nor by merely internally (like one’s picturing facts to oneself), which could be thought to consist in the internal part of Cartesian dualism. In this view, it can be thought that private language implies what belongs to mere internal and what is knowable merely internally. The rejection of private language, in this
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
79
sense, is the rejection of an essential part of inner or internal component of dualistic approach (when it comes to the semantic connectedness between word and world). It is the rejection of the method of correlating name and nominatum, which consists in meaning. As “the inner” is only conceptually linked with “the outer” through behavior, it is the rejection of ‘mind’ part of (yet not soul) dualism through the reduction of it something meaningless. Therefore, it is the rejection of picture theory of meaning, in which the Cartesian dichotomy is embedded. Through private language argument, I think, it is abated the dichotomy by rendering mere internal nonsensical –through the denial of the possibility of the semantic connection between words and world which gave words their meanings. Since such a view does eliminate the possibility of establishing the semantic connection between language and reality or picturing facts to ourselves, one can say that in Wittgenstein’s later philosophy there is no such thing as (mere) internal -or private in the sense of incommunicability.32 So, it can be thought that Wittgenstein, in his later philosophy, reduces the internal-external dual into a unit, which consists only of external. That is to say, by private language argument, he denies the method of correlating name and nominatum which consists in meaning, hence the possibility of mere internal (the notion of privacy in the sense of incommunicability) on the grounds that it is meaningless. In this view, so, by focusing on the sensation and on the symbol, it cannot be given an ostensive definition. As it wouldn’t be a meaningful expression. 5.
Conclusive Remark
Wittgenstein’s “programme” of undermining the dichotomy of “inner” and “outer” set in Tractatus seems to be undermined with his repudiation of private language by means of reducing it to something meaningless (on the grounds that there is no criteria for the correctness of it). Now if there can be no such thing as private language whose words acquire meaning simply by being linked to private experiences, then, it is impossible for one to assign meaning to a linguistic expression by oneself – neither by an act of meaning nor in a merely internal way. But in the Tractatus, the latter was possible because in there, “we picture facts to 32
Incommunicability, as Kenny introduces the term, connotes that the language used or the expression uttered is knowable by any other person.
80
Wittgenstein: Issues and Debates
ourselves”. 33 We picture facts by virtue of the meta-logical connection between words and world, which gave words their meanings. Yet, now in the Investigations, by eliminating the possibility of mere internal way of grasping the meaning of linguistic expressions (through establishing a connection by mere mental or internal faculties –such as having ability to picture facts to ourselves in order to assign a meaning to a word), Wittgenstein expelled meaningfulness from mere internal to external. The meaning of a linguistic expression, for him, anymore can only be learned or taught through an ‘act of meaning’ -from the society in which one lives, through its use. So, since it is the impossibility of private language which excludes the mere internal from the meaningful area, the gist of the radical change in Wittgenstein’s position on meaning can be found in his private language argument. Bibliography Ayer, A. J., 1971: ‘Could language be invented by a Robinson Crusoe?’. In: The Private Language Argument. (ed. O. R. Jones), Macmillan, London. Baker, G. P., Hacker, P. M. S., 1985: Wittgenstein, Rules, Grammar and Necessity. Basil Blackwell, New York. Baç, M., 2001: ‘Wittgenstein ve Anlamın Ortalıkta Olması’, Felsefe Tartıúmaları, 28. Kitap, Bo÷aziçi Üniversitesi Yayınevi, Istanbul. Cakal, A., 2006: ‘How to Make Opposite Ends Meet’. Papers of the 29th International Wittgenstein Symposium. (Eds. G. Gasser, C. Kanzian, E. Runggaldier), Vol. XIV, Kirchberg am Wechsel. Descartes, R., 1951: Meditations on First Philosophy. Bobbs-Merill Educational Publishing, Indianapolis. Grandy, R. E., 1976: ‘The Private Language Argument.’ Mind, New Series, Vol. 85, No. 338, pp. 246-250. Hacker, P. M. S., 1972: Insight and Illusion. Wittgenstein on Phylosophy and the Metaphysics of Experience. Oxford University Press, Oxford. Kenny, A., 1986: Wittgenstein. The Penguin Press, Allen Lane.
33
Wittgenstein, 2004, TLP 2.1, my italics.
What Does the Repudiation of Private Language Mean in Wittgenstein’s Philosophy?
81
Pitcher, G., ed., 1966: Wittgenstein. Garden City, NY, Anchor Books, Doubleday, pp. 22-64. Rhees, R. 1970: Discussions of Wittgenstein. Routledge and Kegan Paul, London. Sidiropoulou, C., 2004: Sp. Top: Wittgenstein. Unpublished Lecture Notes, Fall. Soames, S., 2005: Philosophical Analysis in the Twentieth Century Vol. 2: The Age of Meaning. Princeton University Press, Princeton. Strawson, P., 1954: ‘Review of Philosophical Investigations’ Mind, 63: reprinted in Tully, R. E. Tractarian Dualism. Retrieved over: http://www.bu.edu/wcp/Papers/Meta/MetaTull.htm (28.02.2010). Wittgenstein, L, 1958: Blue and Brown Books. Blackwell, Oxford. ---
2001: Philosophical Investigations. (The German Text, with a Revised English Translation, 3rd edition). Blackwell, Oxford.
---
2004: Tractatus-Logico Philosophicus. (Trans. D.F. Pears, B.F. Mc Guiness), Routhedge and Kegan Paul, London.
Wittgenstein and the Myth of Hinge Propositions Alejandro TOMASINI BASSOLS I.
Historical-Philosophical Considerations
My main goal in this paper is to examine critically, in order to put an end to it once and for all together with all it has brought about, what from my point of view is a very harmful myth, ascribed to Ludwig Wittgenstein by several distinguished scholars, 1 viz., the myth of the so-called ‘hinge propositions’. From the point of view of the history of ideas, the effects of this myth just cannot be ignored. Indeed, based on this myth it now has become a sort of unavoidable fashion to speak of a “third Wittgenstein”, a thinker who apparently would have superseded or overcome his own magnum opus, i.e., the Philosophical Investigations. According to those who support this reading of Wittgenstein’s work, he would have left behind his own phase of grammatical analysis in order to enter a new field of philosophical speculation, a new kind of therapy (whose nature, by the way, has never been made quite clear), much more similar to the sort of rational adventure which is practiced in traditional metaphysics than to the kind of analysis which Wittgenstein painfully managed to developed and practice and which, last but not least, brought him so many good results. Now in order to count with more elements to carry out our examination it’ll be convenient to have at our disposal a simple but neat and well argued panorama, a global but a convincing one too, of the philosophical background in which Wittgenstein has recourse to the expression ‘hinge’, which so many headaches it has already caused. We are thus forced to exercise ourselves in a task of reconstruction of ideas. Our starting point, therefore, can only be the following: if we contemplate his work in toto: how many Wittgensteins is it reasonable to think there are? To answering this question I’ll devote the first part of this essay. 1
Among the most important that should be mentioned are the following ones: Baker 2004; Harré, 2008 and 2009; Cook, 1994; Moyal-Sharrock, 2004 and Moyal-Sharrock (Ed.), 2004; Stroll, 2002.
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ), Ontos Verlag, Frankfurt a. M., 2010, 83-116.
84
Wittgenstein: Issues and Debates
To begin with, I’d like to present my approach: I face the issue of how many Wittgensteins there are not as a matter of exegesis of isolated texts, but rather as a matter of global understanding of the philosophical activity he deployed. So in accordance with the principle of charity, our main goal must be to offer a reconstruction which will make our philosopher first of all a coherent one. So it’ll be natural to think that if his ideas are so reconstructed that the outcome is a picture full of tensions and contradictions, then that picture of his writings is essentially wrong and should be rejected. Now my stance is that this is precisely what happens with the reading of those who tend to speak of a “third Wittgenstein”. I also hold that my general presentation of the systems of ideas built up by Wittgenstein is not vitiated by that defect. Obviously, I also feel confident that apart from our rendering Wittgenstein’s work a coherent one, my interpretation of Wittgenstein’s contribution will be convincing. It is relatively clear nowadays that we can find in the results of Wittgenstein’s philosophical activity both constant as well as variable elements. In spite of huge differences and even radical discrepancies that can be found in the writings belonging to his main philosophical periods, it is plausible to hold that there nevertheless is something like a backbone, that there are some communicating vessels, certain threads of a skein which cross it from one end to the other. Thus I maintain that typical, constant or even defining features of all of Wittgenstein’s philosophies are at least the following ones: 1) his anti-foundationism, 2) the conviction that philosophical problems are not genuine but rather the outcome of some kind or form of misunderstanding, 3) the idea that philosophy is radically different from science, and 4) the idea that philosophy is first and foremost an activity. Accordingly, my task will consist first in showing that as a matter of fact those “elements” can be traced throughout Wittgenstein’s work. I’ll try to make my point without quoting Wittgenstein for every thing I say, except in the cases for which I think it is indispensable. a) Anti-foundationism. That an anti-foundationist perspective pervades the Tractatus is something which would be difficult to put into question. For instance, in open opposition to the Russellian approach, in the Tractatus Wittgenstein rejects in toto the Fregean and Russellian project of founding mathematics on logic. For Wittgenstein, mathematics
Wittgenstein and the Myth of Hinge Propositions
85
are simply a “logical method” that shows the logical structure of reality through its equations. Mathematics cannot be given any kind of foundation for the simple reason that they do not express thoughts, since they do not contain names. Mathematics are systems of identities which are worked out by means of the method of substitution. But mathematics are in need of foundations no more than logic is or, as Wittgenstein puts it, they take care of themselves. The truths of mathematics can be established without having to appeal to set theory. In the same vein, the idea of founding language is totally alien to the Tractatus. That every possible language divides itself into propositions which in some sense are the last residue of analysis, that is, into their respective elementary propositions, is not a matter of founding language on anything, but of the logic of language. There are no elementary propositions which would be more fundamental than others. In the philosophy that Wittgenstein develops from 1929 onwards, the foundationist approach is not only absent, but is openly attacked. In the conception of language that Wittgenstein puts forward in the Investigations there is no place for a pyramidal view. Of course, there are languagegames more primitive than others, more refined or sophisticated than others, but the development of language is accomplished in relation with practices and not as a purely symbolic extension, as when we expand a calculus by demonstrating new theorems starting from some axioms by means of a couple of rules of inference. Human practices need no justification at all. We reach the end of explanations when we acknowledge that a certain language-game is actually played and there is nothing else to say. Since definitions and explanations end at some stage and cannot go on ad infinitum, we reach the bottom when we are faced not with any kind of special truths but with spontaneous human reactions, with human action, i.e., with praxis, for what could be more fundamental than that? Now, since there is no direct causal connection between spontaneous reactions and language-games, since grammar is not conditioned by any reality, the limit we speak about is not a foundational one, in the sense that starting from it we could logically derive the rest, but simply a platform on which language grows. Naturally, it could be objected with respect to foundationism that I am just begging the question and that I am simply denying that which others have argued. However, and awaiting to say a bit more about it later on, I’d like to point out that the only thing I’ve done so far is simply to maintain in a purely abstract way that we don’t have the least reason to
86
Wittgenstein: Issues and Debates
think that so far as the issue of founding a practice is concerned, Wittgenstein modified his initial attitude. It is senseless to try to give language, music, religion, knowledge and so on a foundation. I hope this view will be confirmed or reinforced as we go forward. b) The non existence of genuine philosophical problems. I really don’t believe this could be a topic which would require much textual justification. From his very first confrontation with professional philosophy, Wittgenstein had the insight that there was something odd, lacking a proper articulation, something ill-built in a philosophical problem and that “insight” was kept untouched till the end. The only thing that changed was the diagnostic, which on the other hand is perfectly understandable. In the Tractatus, the explanation of the emergence of philosophical problems is given in terms of speakers’ misunderstanding the logic of our language, while what we are informed in the second period of Wittgenstein’s working in philosophy is that such problems arise when language “goes on holidays”, that is, when it is utterly idle, unplugged from all real, socially recognized as such activity (form of life). For the second Wittgenstein, a philosophical problem appears when a rule of depth grammar of the language-game being considered is violated, just as for the first or young Wittgenstein it happened when it was a rule of logical syntax, i.e., of the Picture Theory, which was not respected. At any rate, in both cases it was essential to the Wittgensteinian way of thinking the rejection of any philosophical problem, in any branch of philosophy whatsoever, as something more than the result of a profound misunderstanding. Contrary to what partisans of a supposed “third” Wittgenstein hold, On Certainty does exemplify what I’ve just said. Regardless of there being or not throughout the book some ideas which would induce us into thinking about a new and spectacular third philosophical project, the fact is that the central preoccupation of the conglomerate of notes collected under the form of a book (incidentally, a book of a remarkable unitary character) is the demonstration that both the skeptic (say, Bertrand Russell) and he who literally accepts what he holds and tries to refute him (for instance, G. E. Moore) are prey of a grave and profound error, that both of them make the same conceptual mistakes and that both of them make and discuss assertions which are just senseless. To put it briefly, the theory of knowledge too contains nothing else than pseudo-problems. Naturally, I’ll return to this in the second part of the essay, but for the moment I think that it suffices to point out that not even in his last writings did
Wittgenstein and the Myth of Hinge Propositions
87
Wittgenstein seriously think of abandoning his mission as a destructor of philosophical myths. If apart from that he set upon himself new goals, that is something which demands to be shown. c) Philosophy as something radically different from science. The Tractatus explicitly states that philosophy cannot be viewed as something which could eventually compete with science. Metaphorically: it’s something which is either over or under science, but not at the same level. This is so because it is not the business of philosophy to describe reality, to discover new facts, regardless of their nature (mental, logical, physical, etc.). In his second and great period, Wittgenstein reinforced and radicalized this point of view. It is also true that from his new perspective the only thing we cannot do in philosophy is to theorize. There are no philosophical theses. The only thing we can (and should) do is to carry out grammatical analyses of philosophical assertions: to make the relevant reminders, to raise the adequate questions and so on in such a way that it becomes manifest that the philosophical assertion in question arouse out of a background of nonsense not acknowledged as such. It is evident that the methods for philosophical research developed by the second Wittgenstein are much richer, varied, ingenious, etc., than what is suggested in the Tractatus, but this is explained by the fact that the views about language, thought, reality, knowledge, etc., that Wittgenstein developed after 1929 are considerably more complex than those we find in his first book. Nevertheless, essentially the same stance is advocated or argued for. If this is so, then our suspicion that something must be radically wrong in the interpretation which culminates in the invention of a “third” Wittgenstein, a neo-philosophical Wittgenstein, must be correct. Such an interpretation would be tantamount to holding that at his very last moment Wittgenstein got rid of what had so far been his fundamental insight. The least we can say is that this way of looking at his work is rather paradoxical and highly implausible d) Philosophy is above all an activity. In the Tractatus Wittgenstein explicitly ascribes to philosophy the task of logically clarifying our thoughts. A way of achieving it is by pointing out what the limits of meaningfulness are, something that from the first Wittgenstein’s perspective could only consist in the tracing of references, given the theory of meaning advocated in the book (meaningfulness and factuality coincide). Concerning his second philosophical period, it is equally undeniable that philosophy for Wittgenstein continued to be conceived above all as an activity. The latter is understood as a kind of therapy. The
88
Wittgenstein: Issues and Debates
therapy in question is not mental in character but rather, to call it some way, ‘praxiological’, 2 since it is seen as an investigation about the application of words, the tracing of uses, with all it involves. Never, therefore, did Wittgenstein set for himself the goal of transforming philosophy into something different, that is, into a sort of research that would aim at the establishing of truths, of whatever kind they could be. What philosophy aims at is the kind of understanding which comes from the descriptions of the uses of words and from the stating of their respective rules of depth grammar, something that requires as well a description of the contexts in which words are employed. What I maintain is that it is impossible to perceive such a change in Wittgenstein’s writings after 1929. Although no doubt in a rather superficial way, nevertheless I think I’ve offered some reasons to hold that there are lines of thought which were kept alive from the beginning till the end of Wittgenstein’s philosophical activity. Obviously, these links didn’t prevent Wittgenstein from enduring an evolution and that slowly but firmly he articulated a new “way of thinking”. In the Preface to the Philosophical Investigations, Wittgenstein suggests that for a better understanding of his book it should be published together with the Tractatus, for then his results “could be seen in the right light only by contrast with and against the background of my old way of thinking”.3 It is evident that in speaking of his “old way of thinking” Wittgenstein doesn’t have in mind different ways of reasoning, deducing, etc., but to his old way of practicing philosophy. What has to be done is to contrast that “old way of thinking”, that is, the carrying out of logical analysis, with the new one, i.e., the practicing of grammatical analysis, embodied in the Investigations as well as in subsequent writings. But it would beabsolutely fantastic if it turned out that Wittgenstein had invented a “third way of thinking”. Really, the suggestion is almost grotesque. If this is so, then the idea of a “third” Wittgenstein begins to look as something rather artificial and incredible. Thus we do acknowledge two great periods of Wittgensteinian philosophizing which are, on the one hand, radically different from each other but which, on the other hand, contain some common elements which 2
It was G. H. von Wright who first used this term in relation with Wittgenstein’s philosophy. As he makes it clear himself, he borrows the term from the great Polish philosopher, T. KotarbiĔski. 3 Wittgenstein, PI, 1974, viii. (Emphasis mine. ATB).
Wittgenstein and the Myth of Hinge Propositions
89
in turn do indicate a certain continuity in Wittgenstein’s thinking. Having emphasized the second ones, I would like now to state some points concerning the drastic differences which prevail between both periods. My observations will be general in character and so I’ll limit myself to consider the relevant differences between the above mentioned periods from the point of view of global perspectives, philosophical lexicon and methods and strategies adopted in each of them. What I want now to show is that although there certainly are dramatic changes between the Wittgenstein of the Tractatus and the Wittgenstein of the Philosophical Investigations, there are no such changes between the latter and the Wittgenstein of On Certainty. a) General perspective. One of the most notorious changes between the two great periods of Wittgenstein’s philosophical activity concerns his global perspectives and their respective priorities. The philosophy of the Tractatus is marked by logic (we could even say, to be more precise, by Russell’s logic). The book is a study of the logical features of the world, of language, of experience, of knowledge and so forth. Everything is subordinated to logic, although never separated from it but always in intimate connection with it. Logic is always the logic of reality, of language, of scientific theories, etc. It is this perspective which gives the book its purely formal character. Accordingly, there is a sense in which everything is subordinated to it, language in particular. The Picture Theory is nothing but the logical theory of language, a theory from which absolutely any empirical consideration was expelled. All this explains why regarding the Tractatus we speak of a formal ontology, of a formal conception of scientific theories, of the formal theory of truth (by the way, the antecessor of Tarski’s famous “semantic theory of truth”), and so on. In the second Wittgensteinian period roles are inverted and what acquires priority above anything else (for strictly philosophical goals, of course) is natural language. From the new perspective, logic is seen as nothing more than a bunch of calculi, all of them illuminating from different points of view (propositions, adverbs, etc.) the functioning of our language. The foundation of everything which formerly was constituted by logic is now replaced by praxis, by human reactions and activities. Individual action, coordinated with others’ actions thanks to the brute fact of nature consisting in our agreeing in reactions, is the ultimate basis of language and thought beyond which it makes no sense to look for explanations. So it would have been difficult indeed to find greater differences than those which hold between the two periods’ perspectives.
90
Wittgenstein: Issues and Debates
b) Philosophical vocabulary. Although quite originally, during his first period Wittgenstein in general expresses himself by means of categories and expressions which were already in circulation. He even borrows from Russell and Moore some typical expressions of theirs, like ‘internal relations’. His originality is nonetheless present, as the very name ‘Picture Theory’ makes it clear.4 Other Wittgensteinian expressions in the Tractatus are, for instance, ‘nets’, ‘logical space’, ‘formal concept’ and ‘tautology’, to mention just a few. However, his coining some new terms doesn’t cancel the fact that he also used traditional philosophical vocabulary. When we turn to his second period we contemplate a completely different phenomenon, for what we find is a totally new philosophical lexicon, that is, a set of expressions that have no antecedent in the history of philosophy. Indeed, for several years Wittgenstein went on building his own linguistic tools, that is, the tools which would enable him to carry out his grammatical analyses. Probably the most prominent elements of this lexicon are ‘language-game’, ‘perspicuous representation’, ‘form of life’, ‘seeing as’, ‘grammar’, ‘grammatical proposition’, ‘criterion’ and ‘family resemblances’. This is the second Wittgenstein’s basic technical vocabulary. Now, he also employed many other expressions, which nonetheless it is not difficult to perceive that strictly speaking do not belong to his technical vocabulary. ‘Hinge proposition’ is a good example of this. ‘Bedrock propositions’, ‘hinge propositions’, etc., are useful metaphors which Wittgenstein occasionally employs in order to complete, to colour, to illustrate his descriptions and his analyses but which do not constitute technical terms, unavoidable expressions for the task of analysis. To transform ‘hinge’ into a technical term just because Wittgenstein uses it on a couple of occasions is like holding that ‘fly’ too is a technical term since he affirms that philosophers are like flies that can’t get out of the bottle. Equally absurd is the thesis that because Wittgenstein employed in several occasions (three, to be precise, in On Certainty) the expression ‘hinge proposition’, then it has been established beyond any doubt that he was elaborating a new theory, a third way of thinking. This is just absurd, but I’ll reserve the expounding of my arguments for the second part of the essay. 4
This could seem debatable, for it would be easy to point out that the expression comes from H. Hertz, who employs it in his The Principles of Mechanics. However, the way Wittgenstein uses the expression ‘Bild’ is certainly quite different from the way Hertz does. Hertz, 1956.
Wittgenstein and the Myth of Hinge Propositions
91
c) Strategies and methods. From what has been said up to now it can be inferred that the first and the second Wittgenstein’s philosophical approaches and tactics had to be completely different. For the Wittgenstein of the Tractatus, once the fundamental ideas had been outlined philosophical work just boiled down to the searching of references for names to determine whether or not a picture was meaningful and eventually to ascribe to it a truth-value. The method in this case was the method of logical-semantic analysis. For the second Wittgenstein, given his rejection of the Augustinian conception of language, nothing could be more futile and sterile than that. It is true that in both cases the ultimate goal was to separate off sense from nonsense, meaningful assertions from those which are so only apparently, but the method to achieve that had to be completely different. Actually, in the case of the second Wittgenstein we cannot speak of a single method, but rather of a whole variety of them. In fact the list is an open one, since new methods can be devised at any time and added to the list. Regarding this, section 133 of the Investigations is particularly relevant. I quote it in extenso: “It is not our aim to refine or complete the system of rules for the use of our words in unheard-of ways. For the clarity we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear. The real discovery is the one that makes me capable of stopping doing philosophy when I want to.– The one that gives philosophy peace, so that it is no longer tormented by questions which bring itself in question.– Instead, we now demonstrate a method, by examples; and the series of examples can be broken off.– Problems are solved (difficulties eliminated), not a single problem. There is not a philosophical method, though there are indeed methods, like different therapies”.5 It would be difficult to be more explicit than that: the general goal is just to exhibit the hidden nonsense of philosophical assertions, regardless of their content or of the branch of philosophy to which they belong, while the methods Wittgenstein employs are most diverse. Nowhere did he intend to give an exhaustive list of methods. It is our task to get them from his work, discerning them out of philosophical exercises. Here are some of them: 1) to replace questions like ‘what is x?’ by questions of the form ‘under what circumstances do we say that someone knows what “x” 5
Wittgenstein, PI, 1974, sec. 133.
92
Wittgenstein: Issues and Debates
means?, ‘what should happen in order for us to be able to say that something is an x?’, etc. 2) To ask: how is this word used in its original language-game? 3) Why is it absurd or grotesque to say of a that it is an x? Wittgenstein has no perfectly well ordered list of questions (in fact, I’m convinced that he frequently improvises, that is, as he advances he offers new suggestions about how to approach in a fruitful way a given philosophical problem), but when we study his writings we become aware of the fact that certain questions are recurrent, that certain ways of approaching issues repeat themselves, which on the other hand is perfectly understandable: whether we speak of the moon or of natural numbers, in both cases what we deal with is the same thing, that is, concepts, (“moon” and “natural number”). So the treatment is prima facie always the same, although obviously there will be questions which are more appropriate for certain subjects than for others. Hence for the Wittgensteinian kind of grammatical analysis dialogue and discussion are crucial for, as it happens with Socratic maieutics, the goal is to contribute to the birth of elucidating thoughts and to untangle conceptual knots. To achieve that, Wittgenstein allows himself to use all sorts of examples, similes, images, comparisons, etc.; he speaks of rivers, of myths and so on and, among other things, of hinges. Now what is undeniable is that for Wittgenstein such exercises are not carried out for the sake of themselves, but because they are indispensable for the dismantling of the philosophical puzzle one is dealing with. As can easily be confirmed, On Certainty does contain a number of questions and reminders concerning the use of expressions that exemplifies what I have been saying. In that book what Wittgenstein does is, as I have already stated, to carry out grammatical analyses of central epistemological notions by means of his own conceptual apparatus. There is not a single terminological change to be pointed out. What happens is that as he makes progress in his grammatical investigation, he gets and accumulates important conclusions about our propositional and belief systems. In fact in On Certainty he does reach the Tractatus’ ideal: to have the right view of his subject without putting it into words. 6 For the time being I’ll limit myself to point out that all of Wittgenstein’s remarks derive one way or another from the kind of philosophical exercises he actually invented and 6
Wittgenstein, TLP, 1978, 6.54 (b).
Wittgenstein and the Myth of Hinge Propositions
93
could not possibly have been formulated independently of them. Therefore, they could hardly constitute an independent philosophy from what could be called the ‘philosophy of grammatical analysis’, which is just the second Wittgenstein’s philosophy. If I have not deviated too much from the truth, it is clear that the content of On Certainty is a paradigmatic case of what can be referred to as the ‘second Wittgenstein’s thought’. If this is really so, then the idea that we can only speak of two Wittgensteins is doubly reinforced: there is on the one hand the Wittgenstein who exalts logic and the Wittgenstein who degrades it and puts natural language in its place. Practically everything he wrote after 1929 belongs to one and the same block of ideas, without denying of course that we can also speak of evolution, refinement, transitions and so on. And if this is so, then we can no longer doubt that to speak of a third Wittgenstein does represent a dangerous historical distortion. Without even trying to consider the issue in all its details, it surely is possible to get an overall picture of the second Wittgenstein’s philosophical contribution. To have it is important, for how shall we later read what he does in On Certainty depends upon the general picture we have of his work. From my point of view, the situation is the following: after his return to the world of professional philosophy, in 1929, Wittgenstein, already vexed by a series of doubts and worries concerning his former point of view and feeling invaded by a multitude of new ideas, ideas he would elaborate little by little, initiated his work of cutting off links with the philosophy of Tractatus. Thus starting from the criticism of the central notion of elementary proposition, Wittgenstein went on forging a wholly new conception of language, for the sake of which he had to get rid of all the notions usually employed in philosophy (“a priori”, “possible world”, “substance”, etc.). He obviously felt the need to operate with a set of fresh concepts in order to articulate his new thoughts. What was being submitted to a fierce attack and was about to be threatened as a whole was no more and no less than Western philosophy, that is, the whole set of trends of thought, problems, theses, prejudices, etc., and whose more brilliant presentation was to be found precisely in the Tractatus. Now once this first task was finished, once again Wittgenstein could have stopped doing philosophy. He was not forced to develop his new attained philosophical views: he had formulated a new conception of language which was a real alternative to all other theories in circulation. The
94
Wittgenstein: Issues and Debates
Tractatus had been refuted. However, for Wittgenstein it was obvious that his task just couldn’t finish at that stage: the rejection of the Augustinian conception of language and the construction of his new point of view in terms of language-games and forms of life was purely propaedeutic. From then onwards the goal was to demonstrate the superiority of the new approach showing that, when properly applied, it enables us to dissolve any philosophical puzzles one would analyze. In order to do that, Wittgenstein went on designing, as I already said, a variety of methods and tactics for philosophical debate thanks to which he was in a position to dismantle, one after the other, the philosophical theses and doctrines he decided to tackle. Classical philosophical enigmas like the mind-body issue, the problem of universals, the puzzles concerning meaning, all the philosophical difficulties around logical and mathematical truth, the question of the existence of God, and many others, were dismantled (some of us would have thought that once and for all, but this turned out to be an illusion). Now for purely contingent reasons, the last problem Wittgenstein came to face was a problem belonging to the theory of knowledge, that is, the problem of scepticism. In other words, once his basic conception and methods of work had been introduced, Wittgenstein passed to the next phase, that is, the phase of application of what had been established in, say, the first 135 or 140 sections of the Investigations, to concrete philosophical problems, regardless of the area or branch of philosophy. From the moment in which Wittgenstein approaches the issue of the nature of understanding, to pass to those of reading, rule-following, private languages, mental states and so on, what he does is to demonstrate in practice the superiority of his new way of thinking. As a matter of fact, all the problems he approached just vanished. So far as I am concerned, I wholeheartedly acknowledge that I am unable to point to a theme or a subject Wittgenstein dealt with which was not sufficiently elucidated by him, at least to the extent that we don’t have to take care of it again. If what I’ve so far stated is basically right, it follows that the most absurd thing that could be done would be to suggest a proliferation of Wittgensteins, specially if the only grounds to do that is the fact the on a couple of occasions he employed some expression not previously introduced. This is the case of the so called ‘hinge propositions’. Let’s review this a bit more in detail. Probably the first question to be asked is: Where does the idea of a “hinge proposition” first appear, where does Wittgenstein actually use that
Wittgenstein and the Myth of Hinge Propositions
95
expression? The answer is: in On Certainty, where he uses it three times. Let us then quote what in fact is our source: 1) “That is to say, the questions that we raise and our doubts depend on the fact that some propositions are exempt from doubt, are as it were like hinges on which those turn”7. 2) “But it isn’t that the situation is like this: We just can’t investigate everything, and for that reason we are forced to rest content with assumption. If I want the door to turn, the things must stay put”.8 3) “The mathematical proposition has, as it were officially, been given the stamp of incontestability. I.e.: ‘Dispute about other thing: this is immovable – it is a hinge on which your dispute can turn’”.9 From a purely textual point of view, this is everything there is. Of course, there is also the whole discussion developed throughout On Certainty concerning the propositional and belief background that each assertion we make brings with it, but the truth remains that as far as textual support of the idea of a third Wittgenstein is advocated the stuff is rather meagre. This hasn’t prevented scholars like D. Moyal-Sharrock, G. Baker and R. Harré from ascribing on that basis speculative aims to a supposed “third Wittgenstein”. I think that, regardless of how fine exegesis and analyses in favour of such an idea could be, the project itself is from the start utterly misguided and makes Wittgenstein incoherent. To show that let us now examine the issue in dispute itself. II.
Philosophical Considerations
1.
Background: propositions and moves in the language-game
It is slightly worrying to be forced at this stage to recognize that, after half a century of obsessive exegeses and intense debates, it is still the case that the most important lessons of the Philosophical Investigations, in particular those concerning Wittgenstein’s discussion and rejection of the Augustinian conception of language, have not yet been duly assimilated. It was reasonable to think that practically all dogmas related to language and its functioning had once and for all been eliminated: the idea that the meaning of a word is an object, the thesis that the fundamental function of 7
Wittgenstein, OC, 1969, sec. 341. Wittgenstein, OC, 1969, sec. 343. 9 Wittgenstein, OC, 1969, sec. 655. 8
96
Wittgenstein: Issues and Debates
language is to describe reality, that by means of predicates we point to essential features of things, and so on. Now in connection with this there is a topic, which is particularly relevant for our theme, namely, the nature of propositions. For several reasons, Wittgenstein felt forced to acknowledge two notions of propositions: a) Firstly, the notion of a proposition which emanates from the Augustinian conception of language, that is, a purely formal one, which we can describe simply as the sense of a sentence. The Augustinian conception doesn’t recognize differences between kinds of sentences, so that any sequence of signs or well formed formula from the point of view of surface grammar expresses a proposition. The fundamental feature of propositions is that they are systematically connected to the notions of truth and falsehood. In the early Wittgenstein’s terminology, their defining feature is bipolarity. Now this notion of proposition is not totally discarded by Wittgenstein for as a matter of fact it is a useful one, both in terms of analysis and in terms of exposition. b) Secondly, Wittgenstein replaces the old notion of a proposition by the notion of “move in the language-game”. From this new point of view, the determination of the sense of a sentence is given not in terms of truth and falsehood, of possible situations in virtue of which it is true or false, but in terms of the utility it lends, what it enables us to say or do. Contrary to the uniform character of traditional propositions, what we have here is an open class of elements, that is, linguistic moves, since we have an open set of language-games and practices (forms of life) associated with them. Keeping this distinction in mind, the first thing we have to ask now is: What sort of relation links propositions and moves in the languagegames? The answer seems to me rather obvious: not all propositions allow for a genuine move in the language-game. For instance, if in the middle of a conversation on irrational numbers someone abruptly asserts: “And of course bears are vegetarians too”, the speaker did construct or manipulated a proposition (the proposition “bears are vegetarians too”), but he certainly made no legitimate move in the language, that is, he didn’t contribute with anything to the conversation, he didn’t say anything relevant, etc. 10 10
It’s interesting to notice that this point of view which I’m ascribing to Wittgenstein somehow corresponds to P. Grice’s so-called “Clause (maxim) of Relation”. See: Grice, 1989.
Wittgenstein and the Myth of Hinge Propositions
97
Therefore, there are more propositions than genuine movements in the language-games, for the only thing needed in order to speak of propositions is to have syntactically and semantically well formed sentences. Within language-games other criteria are relevant. With respect to what I have just said there is something which is quite important to perceive, to highlight and to have in mind and it is that language is simultaneously a collective tool and a tool to be used individually, that is, to be used by each speaker in particular. Obviously, the individual use of the collective tool is viable just because it involves no spurious or philosophical notion of privacity. On the other hand, when we speak we make different sorts of assertions and, accordingly, we express different kinds of belief. For example, if I say ‘I am hungry’, I make a move in language which allows me to inform others about it, to get food, to announce that I’m about to eat something, etc. At the same time, however, I construct a proposition, for the sentence does belong to a natural language. What matters to me as a speaker is the linguistic move itself, not the proposition, since the sentence I use is nothing but a linguistic tool which has an immediate, prosaic, momentary utility; I use it for a concrete purpose and for nothing else. Now this linguistic tool which I use, that is, the sentence, is at my disposal as well as at other speakers’, for any time any of them can also say ‘I am hungry’. In traditional philosophy, in all variants of the Augustinian conception of language, the individual or particular character of the assertion immediately becomes problematic and is in general “explained” in terms of complicated theories about indexicals, logically proper names, temporal operators and so forth. Plainly we are never given a coherent and convincing explanation of why we all can say the same thing and understand what others say when what each of us speak about is what happens to the speaker in turn. From the Wittgensteinian perspective this phenomenon doesn’t represent a problem. The distinction “proposition-move in the language-game” is, therefore, decisive to understand Wittgenstein’s general stance and, in particular and above all, what he maintains in On Certainty. To begin with, it has to be understood that sentences automatically give rise to propositions and something important about the latter is that they form systems. When such systems are constituted by merely circumstantial propositions, by moves in the language-games like ‘I am hungry’, the propositions in question are disposable linguistic tools: nobody is interested in keeping in memory, either personal or collective, all the propositions which could possibly be built concerning passing situations.
98
Wittgenstein: Issues and Debates
Now there are truths which lack this ephemeral or fleeting character, for they describe or point to situations, phenomena, facts, etc., which simply just don’t vanish but remain or endure. For instance, I can assert that Paris is the capital of France and in saying this I both make a legitimate move in some language-game (e.g., to answer a question during an exam, to inform someone about it, etc.) and I also build a proposition which can serve me as well as others in a variety of circumstances. It is crucially important, therefore, to distinguish between propositional roles: there are sentences for the moment and sentences which possess a recurrent utility. ‘I am hungry’ exemplifies the former, ‘The Earth existed before I was born’ the latter. On the other hand, given the vital character of the Wittgensteinian conception of language in terms of language-games, it had to be expected that we should be given a dynamic view of assertions. Wittgenstein calls our attention to this in an interesting remark in the Philosophical Investigations where, after discussing Russell’s Theory of Descriptions he says: “The fluctuation of scientific definitions: what to-day counts as an observed concomitant of a phenomenon will to-morrow be used to define it”. 11 In other words: we witness every day what could be called ‘propositional transitions’, that is, changes in propositions’ different statu. Disposable propositions belong to the surface or, to employ Quine’s very similar metaphor, to the shore, whereas there are propositions which are rather fixed, the famous “bedrock propositions”, propositions that in general experience leaves untouched. But what should not be passed unnoticed is that there is nothing fixed once and for all: in principle, propositions from the bottom may disappear and others, which at a certain moment were on the surface may become more important, become indispensable and locate themselves in the bed of the propositional river. This is the way language functions, not to say ‘lives’. Thus, sentences are tools, which may give rise to propositions of a completely different nature. Needles to say that non well formed, elliptic, etc., sentences may nevertheless give rise to genuine moves in the language-games. So strictly speaking it could be the case that communication among speakers does take place without our being forced to talk about propositions at all, for sentences would be ill-formed. On the other hand, there are circumstances in which a sentence like ‘I’ve never traveled to Mars’ would allow me to make a legitimate move in the 11
Wittgenstein, PI, 1974, sec. 79.
Wittgenstein and the Myth of Hinge Propositions
99
language-game, but given its content it is highly probable that, although true, it would generate a proposition which more often wouldn’t be usable, that it would accomplish no purpose, it would convey nothing. In such cases what we have is a proposition that because of its not being false simply hides in the propositional background in such a way that it doesn’t interfere with the speakers’ moves and that is just there, until one discovers it and states it, if one finds it suitable. 2.
Language, beliefs, propositions and truth
What we have been saying is relevant to our subject for, regardless of how what we say is interpreted, that is, either as propositions or as moves in the language-games, what we have are beliefs. Naturally, the first criterion for the detection of a belief is what the speaker himself says. The Wittgensteinian approach immediately makes us realize that we have to divide our beliefs into two main groups: a) those associated with propositions b) those associated with moves in the language-games What has just been said has important and easily predictable implications, since the notion of a belief is linked at least to four other crucial epistemological notions, namely, the notions of truth, knowledge, doubt and certainty. Let us consider first the concept of knowledge. If what we assert when we make a move in some language-game is right, what we give expression to is a true belief. In accordance with our classification of propositions, what we now have to say is that the belief in question may be either an immediate or a mediate one, a useable or (to put it another way) a rather ornamental one. Differences between propositions are obviously a matter of degree. If I say ‘Yesterday it rained copiously’ my belief is circumstantial, it concerns a particular, short-lasting event, etc. We can hold that practically my belief disappeared together with the event in question: when I asserted it I believed it, I stopped being interested in it, it endured no more and in a sense my belief just vanished with it. These are beliefs which have, so to speak, a biography, birth and death certificates and their lives are rather short but, once again, what we have here is a gradation. We can, if we insist, keep it and say (with Frege, for instance) that it is eternally true that on such and a day such and such event took place. This of course is nothing but a philosophical extravaganza which we don’t have to accept, but it is worth noticing that it reintroduces the
100
Wittgenstein: Issues and Debates
Augustinian conception of a proposition and as a consequence the conception of a proposition associated to language-games stops being functional. Since we don’t want this, we have to draw other distinctions and classifications. Now in contrast to ephemeral, immediate or circumstantial beliefs we have others which are not like that, which resemble more classical propositions, that is, they are endurable, stable, etc., and which I’d describe as belonging to the background. As far as daily verbal intercourse is concerned they are not particularly required, but they are nevertheless there and we can have recourse to them any time. These beliefs are often expressed by propositions that we accept, although they can also often be expressed by legitimate moves in the language-games. Thus if I say that ‘The solar system existed long before I was born’, I am expressing a proposition and therefore a belief which although only occasionally may be of practical utility and accordingly only occasionally gives rise to a genuine move in the language-game, in the majority of cases in which it is used it shows itself as useless, as an unemployable proposition, which keeps engraved in the background of my assertions of a more immediate character; it is a proposition that I only slowly discover or retrieve, since it is not false, but I practically never use it. The Moore-type of propositions, which Wittgenstein examines in On Certainty are in general of this kind. A particularly important point which I wish to make is that what is at stake with this classification of beliefs is a matter of classes of knowledge, of truths and falsehoods, of objective doubt and certainty, but not of meaning. Contrary to what has been held, for instance, by D. MoyalSharrock,12 bedrock propositions like ‘I’ve never traveled beyond the solar system’ or ‘I had parents too’ do not determine the meaning of any sentence whatever. The issue has nothing to do with meaning, with the bounds of sense, with definitions, etc., but with classes of beliefs, with kinds of knowledge. The whole discussion is epistemological, not semantic. In On Certainty, Wittgenstein labors upon the platform conformed by secure meanings. The problems he deals with belong to the theory of knowledge, for in the Investigations he had already settled those belonging to the philosophy of language and they just don’t reappear. Naturally, if this is true then the advocates’ interpretation of “hinge propositions” turns out to be completely misguided, an absurd position 12
See, for instance, his Moyal-Sharrock, 2003, where she argues in extenso in this sense.
Wittgenstein and the Myth of Hinge Propositions
101
resulting from a profound misconception of Wittgensteinian thinking. We need now to reinforce this view. III. Language, beliefs, logic and knowledge The elements we have introduced so far enable us to infer that what follows from the Wittgensteinian treatment of epistemological concepts is that we have at least two concepts of knowledge, linked with each other in a rather complex way. On the one hand, there is collective knowledge, connected in a straightforward way with propositional systems and, on the other hand, individual knowledge, expressed through moves in the language-games and the speaker’s beliefs. Unfortunately, the question of classes of knowledge and beliefs and the relations which hold between them are neither as simple nor as neat as we would like them to be. Let’s review this in some detail. As we already pointed out, the speaker’s linguistic goals are first and foremost of a practical and immediate character. For them to be begot, the linguistic moves he makes have to take shape in well formed sentences of our language. While being used true sentences generate knowledge and what this kind of knowledge comprehends is everything which is related to the speaker’s direct or immediate experience: his goals, his plans, his actions, his feelings, his achievements, his memories, etc. I know for instance that yesterday it rained because I got wet and now I’m sick, I know that I’m going to enjoy a dish because I was invited to dinner, and so on. However, it is obvious that this personal, immediate “knowledge” is of very little interest. Practically, it concerns only its possessor, a concrete event, the environment, etc. Nevertheless, moves in the language-games are produced and therefore beliefs are expressed, which are true or false and, accordingly, knowledge or error arises. Naturally, in all these cases propositions are also involved but, given their nature, they are disposable: as soon as they are used they are thrown away. We, however, often make assertions which have a slightly more stable character, but that doesn’t prevent them from being useful. For example, I can say that I am Mexican. This is not something I constantly say here and there, to my neighbours, parents, pupils, colleagues, etc., but it is clear that there are countless occasions in which it may prove useful to say it, and even unavoidable. For instance, I would say ‘I am Mexican’ to a police agent of another country if I am questioned about my nationality. The truth involved, however, is not as recurrent as ‘I am hungry’. That is, I
102
Wittgenstein: Issues and Debates
don’t need to say so many times ‘I am Mexican’ as I may be induced to say ‘I am hungry’, but that makes it neither improper nor ill-formed nor absurd. It simply is less employed. To put it simply, it has more the status of a proposition than the status of a “move in the language-game”, but it certainly may serve to convey information and be useful. Something we should not lose sight of when we consider our sentences as generating propositions is the fact that a characteristic feature of propositions is that they form systems. Once we are dealing with propositions, what matters is logic, in the widest sense of the word. So if I say ‘I am Mexican’, somehow I logically imply or presuppose or assume or whatever the logical relation it is that holds between them that I am a person, that I had parents, that I was born somewhere on this planet, that I don’t breath by means of branchiae, that I don’t eat trees, etc., etc., that is, I imply or presuppose (or both things) an infinity of sentences which can’t be rejected for they are not false but are such that, even if I never use them, they are nonetheless there, as a matter of “logic”. For a variety of reasons, the truth is that users of language little by little become aware of the fact that, in saying something, automatically propositional systems are built, systems which as a matter of fact are simply irremovable. Once again, the panorama here is rather complex, since there are two senses of ‘propositional system’. On the one hand, there is the system of propositions created by logic as soon as something is said; on the other hand there is the collective or shared propositional system which is presupposed or implied by any assertion whatsoever. And what is fundamental to understand is that there are no clear cut, stable, fixed borders between some propositional groups and others, but rather unperceivable transitions, constant subtle changes, etc. Let’s give an example. Let’s suppose that I affirm that Warsaw is the capital of Poland. In certain contexts, to say such thing would be even ridiculous, as for example if I said it during a congress of history or geography. In such a case, it could even be understood as if I were making fun of someone. However, if the very same linguistic tool (that is, the English sentence ‘Warsaw is the capital of Poland’) is used during a lecture of geography for children, it is the appropriate instrument and what I say is relevant and true. Thus, the same sentence may both give rise to a genuine move in the language-game as generate a proposition which is incrusted in the more or less stable set of propositions with which it is logically connected and is left there. It is evident that however stable it might seem, it doesn’t reach
Wittgenstein and the Myth of Hinge Propositions
103
the status of a “truth” of arithmetic, for it is perfectly possible to imagine that tomorrow Warsaw will no longer be the capital of Poland whereas it is impossible for us to visualize a situation in which 2 plus 2 doesn’t equal 4. At any rate, propositions do form a structure, a sort of net which the less it is constituted by immediate propositions the more it tends to stabilize. Our propositional net, therefore, comprehends both propositions which so to speak either can be more or less quickly consumed (moves in the languagegames) or more or less stable propositions which in general are not used to make linguistic moves of any kind. Naturally, there is a wide variety of propositions maintaining between them no fixed or precise limits. We can now start joining together our results and getting some conclusions. Whenever we speak of sentences we have to emphasize the utility they lend, on the one hand, and their truth and falsity, on the other. Several combinations may be formed between these propositional qualities and any sentence ‘p’, but those that really matter are basically the following: 1) the sentence ‘p’ is true and enables one to make a move in the language-game, that is, it is useful, and 2) the sentence ‘p’ gives rise to a truth but in fact it doesn’t enable one to say something particularly useful, for it is only presupposed or implied by what is being said, it has a purely ornamental character, etc. Now the fact that it is not usable doesn’t imply that it is rejectable. In order to be rejectable it has to be false. ‘The Earth existed long before I was born’ may be useless, but it is not false. Related to this double possibility, we can speak of knowledge in two senses: 1) genuine knowledge, reachable through valid mechanisms acknowledged as such, as when I say in a history class that Napoleon was Corsican, and 2) knowledge in a broad or elastic and almost pejorative sense, therefore almost or totally useless, as when I affirm that I have never left the solar system, for what sort of knowledge do I convey if I say that to someone? Once again, we have to dwell upon the fact that we are not dealing here with two neatly distinguished groups. We are just pointing to the ends between which there is a whole propositional gradation and therefore, a cognitive one too.
104
Wittgenstein: Issues and Debates
IV. Doubt and certainty As is well known, in On Certainty Wittgenstein recognizes at least two concepts of certainty, namely, subjective and objective certainty. The former is simply a high or intense degree of conviction, but it doesn’t guarantee the truth of what is believed. I may be totally convinced, i.e., I may have subjective certainty about being the son of the couple who say they are my parents, but it is perfectly imaginable that I would be mistaken. In fact, this is something that has actually happened to lots of people: they had always believed that they were the genetic offspring of those who presented themselves as their parents and they later discovered that it was not the case. Subjective certainty, therefore, is not infallible. Naturally, a discovery such as the above mentioned carries with it drastic changes in our basic systems of beliefs (and probably not only in them). Psychologically, the person whose subjective certainty is put into question just breaks down. One has to establish a completely new order to find a place for the new belief, for the new truth. On the other hand, however, there is something like objective certainty. Contrary to what happens with subjective certainty, to try to put into question objective certainty is something that only someone mentally deranged might try to do, for it is tantamount to rejecting the propositional system that his own moves in the language-games would have created. This is precisely what the skeptic tries to do. Now it is peculiar of Wittgenstein’s philosophy that the classification of certainties and the corresponding classification of beliefs and pieces of knowledge are not drawn in formal terms, that is, for instance, in terms of propositional categories (“synthetic”, “analytical”, “contingent”, etc.). What is determinant is the role actually played by sentences. This role is not defined by means of fixed or ultimate categories. Roles change. The same sentence, as we saw, one day expresses a contingent proposition and next day a necessary one. It is because of their role within the propositional system that certain propositions turn out to be irremovable. These propositions which are taken for granted are not put into question any more. What they give is objective certainty. Obviously, they are not like those other propositions which are more directly linked to experience, but their difference doesn’t arise from differences of logical or grammatical form, but from their location within the system and therefore from their respective roles. Once more, the differences between propositions are a matter of degree and we pass without noticing from propositions which generate certainty but that enable us to make no
Wittgenstein and the Myth of Hinge Propositions
105
legitimate move in the language-game (“I am a living creature”) to propositions which do enable us to say something, but which no longer give us that class of certainty. Something similar to what I’ve just described happens with doubts. Doubt can legitimately fall upon propositions which can be eliminated. However, the more we get down towards the bed of the propositional river the more doubts become suspicious, ridiculous, absurd. To put it briefly, we simply can’t put into question propositions which by logic underlie propositions which we actually state and which, so to speak, automatically appear as soon as we say something. If I say to someone ‘Please, take a chair and have a sit’, both the person in question and I assume that the chair is a physical object, although we just don’t even dream of stating it. To make clarifications in this sense would be something very odd and under certain circumstances it would most probably have negative consequences.13 On the other hand, it is a fact that if we are ever asked whether or not we believe that the chair is a physical object we’ll unanimously say ‘yes’, even if the question seems trivial or childish or silly or vacuous to us. But it is very important to realize that the answer is not false. We should dwell upon the fact that what Wittgenstein does in On Certainty is something like outlining the map of knowledge, not a semantic map. The latter is plainly not at stake, it’s not something discussed in that book. To sum up: I can’t doubt that the chair is a material object in case I affirm of something that it is a chair. The chair’s being a material object is part of my linguistic background. I refer to such background when I need it, but most of the times it is just there, giving support to my moves in the language-games. Obviously, what happens in this example happens with any other sentence. This, however, has nothing to do with foundationalism. As was to be expected, our concept of knowledge is correlative to the concepts of doubt and certainty. In On Certainty, Wittgenstein made it clear that it is only when we deal with genuine knowledge that we can speak of doubt and that if what we have is objective certainty, then we speak of knowledge in a rather loose, not strict sense. Better: it makes no sense to speak of knowledge at all. The trouble, obviously, is that it just 13
Just imagine, for instance, that someone asks Caesar or Napoleon or Stalin to have a sit and that immediately after that he points out the any of them that the object he will sit on is a physical object. Such a joke could have rather unpleasant consequences. Mutatis mutandis, the same would happen in any other linguistic context (father-son, teacher-pupil, doctor-patient, police man-citizen, etc.).
106
Wittgenstein: Issues and Debates
can’t be maintained that ‘I’ve never traveled to the center of the Earth’ doesn’t give any kind of knowledge, for to do that we should have to deny that the proposition is true and we just can’t do that. What has to be done is to employ the Wittgensteinian terminology and point out that although the proposition is true, in fact with it we make no genuine move in the language-game and therefore we convey no knowledge at all. Thus, the concept of knowledge enables us to allude to quite different things. I’ll try to make this a bit clearer. First of all, we’ve got the paradigmatic, prototypical individual sort of knowledge, that is, the knowledge I report: what I see, I remember, I imagine, etc. This is immediate individual or personal knowledge. Now this knowledge carries with it all those propositions which automatically the speaker generates as soon as he says something. This is a derivative sense of the first sense of ‘knowledge’. For instance, if someone says that the lion is about to catch a zebra, he automatically generated the proposition that both the lion and the zebra are animals, that the lion eats meat, etc. The point is that these propositions, in the linguistic context in which they are located, are completely useless for the speaker. They are simply implied or presupposed or whatever is the epistemological relation that holds between them and the actual linguistic move. It is important to observe, on the other hand, that the same linguistic move made by another speaker generates a different propositional system, although it is probable that their respective systems will contain lots of common propositions. It is highly plausible that from all the propositional systems that speakers generate when they say something we could eventually form one which would be “complete”, that is, it would contain all the propositions which are common to all speakers’ propositional systems. That propositional system would give all speakers objective certainty and in relation to it doubt would be simply absurd. More or less the same holds for each propositional system generated by each linguistic move by each particular speaker. Secondly, we have the knowledge that required observation, measurements, calculations, specific methods of research, etc. This is collective (scientific) knowledge. In this case we’ve got a variety of propositions. Some of them may be put into question, but not all. Scientific knowledge tends to form propositional systems which aim at total rigidity, even if that ideal is never attained. Now what we assert is epistemologically legitimate as long as it doesn’t conflict with the body of scientific propositions. The problem, naturally, is that normal speakers are
Wittgenstein and the Myth of Hinge Propositions
107
not in contact with scientific propositional systems and that’s why very often people make assertions which strictly speaking they don’t have the epistemological right to make. For instance, someone can feel entitled to assert that beings from other galaxies have visited us for he just ignores that astrophysics has already discarded that possibility (because of the number of years that they would have to travel, the speed concerned, for biological reasons, etc.). But I must insist that the problems that might arise are epistemological or cognitive, but not semantic. What has so far been stated gives us, I think, a more or less acceptable apercu of the panorama that, patiently through his grammatical analyses, Wittgenstein outlines in On Certainty. Let’s quickly see now what consequences or implications this panorama has for some classical philosophical discussions. V.
Scepticism and Moore-type propositions
In my opinion, scepticism shares certain features with common-sense. Neither of them constitutes any “school” or philosophical trend, they are not sets of theses, etc. Rather they are something like mechanisms of reasoning one can appeal to at any time. They are obviously opposed to each other, that is, they cancel each other out: he who advocates a point of view on behalf of common-sense just can’t advocate a sceptical position, and the other way around. Scepticism is, therefore, the opposite mechanism to common-sense. In this sense, G. E. Moore was certainly right in opposing scepticism by taking sides with common-sense. In fact, the best way to understand each of these views is by contrast with the other one. Wittgenstein, on the other hand, showed that there is a sense in which both Moore and the skeptic are wrong. Scepticism means a challenge for all those who consider that it is obvious that there is something like “human knowledge”. The sceptic’s strategy is founded on two premises: 1) the purely formal and exclusive notion of a proposition, 2) the possibility of systematically denying any sentence whatever, based on the principle that we do not pass from a meaningful sentence to a senseless one just by adding or eliminating the negation sign.
108
Wittgenstein: Issues and Debates
Briefly, what the skeptic argues is that since it is logically possible to deny any sentence whatever and that all meaningful sentences express propositions, then we can question absolutely anything we are told. It follows that even the most elementary or natural claim to knowledge gets cancelled since it can be shown that any proposition, that is, any bearer of truth, can be denied. And naturally the skeptic in turn affirms that nobody can refute him. The sceptical stance can be presented in different ways. Perhaps the most ingenious of all forms so far suggested is that of a malign god, that is, an all-powerful, omniscient, eternal, etc., being but one who is also bad and is amused by fooling me making me believe that I know lots of things when in fact I know nothing. There is also the dream hypothesis. However, I will not go into the details of the skeptic’s strategies, for it would take me too far from what are my present purposes. G. E. Moore thought that he had refuted the skeptic by enumerating a series of beliefs that satisfy all the conditions which can be imposed upon them in order to speak legitimately of knowledge. The class in question is an open one, that is, new beliefs can be added at any time. Examples of true and indubitable beliefs are: the universe existed before I was born, I’ve always lived on the surface of this planet, I have a body, I have had a variety of experiences since I was born, and so forth. Now it is through or by means of propositions of this sort that Moore thinks that he can formally refute the skeptic, for since these propositions are certainly true, then he knows something with certainty. His line of argument is much more subtle and complex than what a crude presentation of his general stance could reveal, but for our purposes what we have said suffices. Wittgenstein takes part in the discussion and shows that although Moore is in some sense right, anyway there is another, philosophically important sense in which what he maintains is wrong, the reason being that he simply can’t be said to know what he affirms he knows. On the face of it, Wittgenstein’s point of view is strange, not to say ‘paradoxical’, but I’m convinced that if we have recourse to the categories we introduced above we shall be in a position to state his view more clearly and to show why he is right. Wittgenstein’s stance with respect to the “sceptic-Moore” controversy is more or less the following: vis à vis the sceptical thesis to the effect that knowledge is either impossible or unreal, Moore answers pointing to propositions which are true and which generate certainty; more than that: they are certain in the sense that it is plainly absurd (i.e., contrary
Wittgenstein and the Myth of Hinge Propositions
109
to common-sense) to deny them. The problem is that those propositions are such that in normal linguistic contexts they do not serve to make genuine moves in the language-games. But Moore, almost without noticing it, achieves something quite interesting, namely, he shows that any assertion we make automatically turns out to be like the peak of a propositional iceberg: there are a large and indefinite number of propositions underling the linguistic movement and which are indispensable to affirm or negate what is being asserted or negated. The speaker, so to say, doesn’t see those propositions, he may even ignore their existence, but they nevertheless are automatically produced by relations of cognitive presupposition, even if they are neither made explicit nor used in any sense whatever. Thus if I inform someone that I am Mexican, I am automatically committed to propositions like ‘I am a person’, ‘I was born in the country called “Mexico”’, ‘I come from North America’, etc., etc. In other words, my movement in the language-game brings with itself the truth of everything that has to be given for me to say what can be true. Again, it is particularly important here to insist that the relations between what is expressed by a linguistic move and its underlying propositions are not semantic in character. We are not saying that in order for what I say to be meaningful the corresponding propositional system has to be true. It’s not about meanings that we are talking, but about truth and falsehood. Now since we are dealing with true and not with false propositions, somehow these convey some kind of knowledge, but since it is so to speak useless knowledge, what they convey is not strictly speaking knowledge. Therefore Moore just can’t say that he knows what he says he knows. Likewise, since we, however vaguely, can speak in relation to those propositions of truth and knowledge, we can also speak of beliefs, but since we don’t have in mind beliefs we arrive at by the usual ways of acquiring beliefs, the beliefs in question are so only in a derivative sense; we could almost speak of a “belief simulacrum”. In other words, if I am seriously and explicitly asked whether I believe that I have a body I’ll reply that I do, but it’ll seem to me such a platitudinous or vacuous belief that I don’t even care to formulate it. The reason is that it is a belief which emanates from an assumed proposition, not from a genuine move in some language-game. Wittgenstein does offer different independent lines of argument against the sceptical position, but I won’t consider them in this essay. What I was interested in was in offering a plausible reconstruction and present his general outlook with respect to propositions, beliefs and pieces of
110
Wittgenstein: Issues and Debates
knowledge. From his perspective, both Moore and the skeptic make the same mistake: one says that he doesn’t know anything while the other affirms that he knows a lot of things while both of them misuse the verb ‘to know’. Their discussion is a typical pseudo-philosophical discussion. Once we have seen this, we can finally pass to consider “hinge propositions” themselves which, according to some, unambiguously point to a radical change in the general outlook of the second Wittgenstein. VI. Hinge propositions: a diagnostic. Before ascribing Wittgenstein any theory about hinge propositions we should perhaps ask: what is a hinge? What does it serve? It should be easy to provide answers to such simple questions because it is evident that Wittgenstein is simply using here a metaphor to indicate something, but he’s not introducing a new technical term, one more element of his conceptual apparatus. What matters is what the metaphor indicates, what by its means he wished to convey, not the metaphor itself. So then what is it that the metaphor helps to say? A hinge is a small tool which serves to keep two pieces together, in general made of wood but not necessarily (they could be made of steel, plastic, etc.), viz., the door and the framework. As Wittgenstein says, in order for the door to open or to close the hinge must be fixed. Here the question is: how does this metaphor apply to propositions? The answer doesn’t seem to be terribly difficult to find: there are propositions which have to be fixed in order for others to be applied, to be used. ‘To be fixed’ in this case means simply that they cannot be put into question, that they are not subject to doubt, for if they were they would no longer function as hinges. A mobile hinge is not a hinge. Now what propositions are like that? What propositions play in the realm of propositions a role similar to the role played by hinges in the realm of doors? First, there are “propositional systems”, like those of mathematics, which are such that once they are accepted by the community they remain untouched, that is, nobody even tries to modify or reject them. Indeed, only someone totally crazy would intend to refute Pythagoras’ theorem. Secondly, we have the propositional systems which are automatically generated as soon as the speaker says something. For instance, if I affirm that ‘John is my neighbour’ I am implicitly but automatically committed to propositions like ‘John is a human being’. It just couldn’t possibly be the case that John is my neighbour and that he’s not a human being (that is,
Wittgenstein and the Myth of Hinge Propositions
111
speaking seriously, knowing the meaning of words and so on). It is fixed propositions, propositions by which we stand, which stand fast for us, which “justify” our moves in the language-games. Perhaps a simple diagram could help to make this clear: Speaker (s)
Language-games (linguistic exchanges)
Epistemologically involved propositions
|
|
| | | || |
|
| | |
| | | |
|
|
| . .............. |
| | | ||
............ ...........
Each ‘|’ represents an epistemologically “presupposed” or “assumed” or “implied” proposition by the linguistic move carried out by the speaker. Let’s call ‘ǻ’ the totality (finite or infinite) of involved propositions. It takes us some effort, time, opportunities to discover the elements of ǻ, which support the linguistic moves. Indeed, the situation is very similar to what P. Strawson held in his polemics with B. Russell: if there are no kings in France, the issue of whether or not the King of France is bald simply doesn’t arise. In Strawson’s case, however, the point was to exhibit the propositions which are both necessary and sufficient for the meaningfulness of a sentence. In the case of Wittgenstein’s discussion, the issue concerns knowledge, our “general picture of the world”, our conception of reality, not the meaningfulness or the nonsensicality of our expressions and sentences. In fact Wittgenstein doesn’t criticize the skeptic for making semantic mistakes, something he certainly accuses Moore of doing for misusing the term ‘to know’. Wittgenstein’s arguments against the skeptic are strictly epistemological in character, they tend to point to holes in his lines of argument, argumentative vicious like circularity, etc. For example, against the argument from dreaming Wittgenstein points out that if I am dreaming, then also the remark that I am dreaming is dreamt and the meaning of what I say is dreamt too. The skeptic’s argument is thereby demolished, but the argument itself is not semantic or linguistic, because it is not the theory of meaning what is being examined by Wittgenstein. It is important to notice that ǻ will change from user to user, since their respective assertions or moves will be different. Nevertheless, the more we get down in the hierarchy of involved propositions the more universal the latter will be and therefore the more shared among speakers. When we’ve got at such propositions what we arrived at are basic elements of our conception of reality, of our implicit picture of the world. The fact
112
Wittgenstein: Issues and Debates
remains however that normally we, speakers, lack a clear idea of how we actually see the world, of what we in fact hold about it, for in general language has for us a purely practical utility. Thus the propositions which stand fast for us, which are never modified, which practically never change their status seem so obvious to us that we just never deign to consider them. With respect to this I’d like to say a few words. Since Wittgenstein is not concerned with language from a formal point of view, the classifications of propositions that he draws are classifications of roles. Therefore, traditional categories as applied to propositions, like “analytic” and “synthetic” become in this framework simply worthless. But then, apart from their roles: what features do propositions like ‘space is real’ or ‘there are animals’ or ‘I’ve have a wide range of experiences’ or ‘my life has a temporal dimension’ possess? To begin with, from the point of view of their internal constitution they certainly are synthetic. It is evident that they had to be so, for an analytic proposition would make no contribution to my picture of reality, since it says nothing, lacks content, is vacuous and results from a stipulation. On the other hand, given the role they play and the way we have access to them, there is a sense in which the most fundamental of bedrock propositions are neither a priori nor not a priori. They are a priori in the sense they are neither confirmed nor refuted by experience and they are not a priori for in a sense they were discovered empirically. Thus such categories simply don’t have a clear-cut application here. This is understandable: in a sense, those propositions are not established by us. As we saw, in so far as they simply underlie all our moves in the languagegames, it is language itself which automatically establishes them.14 What we can do is trace them, although ‘trace’ here doesn’t mean just ‘deducing’ them. It is not an axiomatic system that we build when we get the propositions from the bed of the propositional river. In fact we learn such propositions, we get in touch with them, after we learnt how to speak, much later in our development as speakers. That is, even though there is a sense in which we can say that they are assumed or presupposed, in the everyday practice of language they are not required at all. They simply serve to complete a picture. It is in this sense that what we say epistemologically depends on them. It is not the case that for what we say to be meaningful they have to be true; it is rather that for what we say to be understandable and not be a simple brushstroke on a canvass that they 14
About this I say below something else which I consider is important.
Wittgenstein and the Myth of Hinge Propositions
113
have to be there. The propositions of our linguistic background are indispensable in so far as they make our conception of reality coherent, they enlarge it and in this sense it is just impossible to question them. It’s only when we do philosophy that we may have doubts about them, which shows that Wittgenstein was perfectly right when he spoke about the utterly otiose and incongruous character of conventional philosophical theories. For normal speakers, such propositions do provide subjective certainty simply because they are bearers of objective certainties. They are indispensable to such extent that they look like tautologies, propositions which the speaker immediately grasps they just couldn’t possibly be false, but at the same time propositions he doesn’t know exactly what to do with. That is precisely why they do not contribute to the making of moves in the language-games. This doesn’t mean, however, that they never allow us to say something, that is, to make some real move in some language-game. After all, they are propositions. However, their role is not an empirical one, but rather systemic or structural. Apart from the fact that they at least in principle can move and be replaced, their number is indeterminate. It would be impossible (or rather absurd) to try to say something in this sense. One feature which makes the bottom propositions of our language look somewhat strange is simply that they are hidden and that only from time to time, so to speak, their heads come out. In this sense, Wittgenstein could even be seen as the Freud of language: he makes us aware of the fact that language has its own Unconscious, that is, a large group of elements (propositions) which in one way or another make themselves felt but that only rarely we are prepared to bring them to light. Just as in psychoanalytic therapy, where it’s only through a slow process that we obtain the data for the overcoming of our neurosis, so too in the Wittgensteinian kind of therapy we pick up the propositions which lie at the very bottom of our linguistic background only through slow grammatical analyses. In view of the discussions it has given rise to, I dare to think that, once we have a grasp of the overall panorama of On Certainty, the hinge metaphor probably turns out to be the least fortunate of all metaphors Wittgenstein ever coined (and there are lots of them). Through it the idea of foundation was reintroduced, a totally alien idea to the Wittgensteinian way of thinking, to account for the very last phase of Wittgenstein’s production. This amounts to a grave distortion of his thought. So far I’ve tried to discredit the idea of a “third Wittgenstein”, a Wittgenstein who would no longer be interested in carrying out grammatical analyses, but who would rather be engaged in some new sort
114
Wittgenstein: Issues and Debates
of foundationalist work. However, I should perhaps make an exception, although it concerns not On Certainty but texts like Remarks on Colour. It is worth observing that when the nature of his philosophical investigation changes, Wittgenstein is the first to warn us about it and he actually points out explicitly that he’s doing something different from what he usually does. So for example, in his Remarks on Colour he says that “there is no such thing as phenomenology, but there are indeed phenomenological problems”15 and what he goes on to do makes it clear that he’s not carrying out a purely logical or grammatical investigation. Now that doesn’t mean that his philosophical aim changed into a completely different one. Let us ask: what is a phenomenological problem? It’s a problem connected with or derived from “immediate experience”. It would seem at first sight as if Wittgenstein were interested in carrying out some sort of quasi-factual analysis of some sort. But to think that would be a sheer mistake, for Wittgenstein understands the phenomenological task as an analysis of the descriptions we make of our visual experience with respect to colours. So the study of the phenomenological problems that Wittgenstein deals with boils down to an analysis of colour concepts. Nevertheless, it is undeniable that we reach here the limits between a purely grammatical research and what would be an empirical study of visual experience. Wittgenstein, however, never crosses that limit. His phenomenological research lacks completely causal content. It is certainly not causal explanations what Wittgenstein provides when he examines, for instance, white colour. Now it is important to keep in mind that there are indeed texts in which Wittgenstein works philosophically in a way slightly different from the way he proceeds in other places, but the reason is simply that the difference comes from the subjects themselves, for it is evident that in the case of colours both introspective and behavioural analysis of visual experience become unavoidable. Now such a deviation, to call it such, by Wittgenstein with respect to his own standard writings, although understandable in the case of colours, is impossible to detect in On Certainty. In this book, what Wittgenstein carries out is a standard, grammatical analysis of epistemological notions (“knowledge”, “belief”, “doubt”, “certainty” and so on). That is, in this collection of remarks Wittgenstein’s goal went on being the same he always had, namely, the dismantling of diverse philosophical enigmas; more precisely, the dismantling of the venerable philosophical puzzle of scepticism, a subject that already in the Tractatus had been declared absurd. It’s difficult to see, 15
Wittgenstein, 1994, I, 53.
Wittgenstein and the Myth of Hinge Propositions
115
therefore, that in relation to the issue of the nature of philosophy an essential change in the very last phase of Wittgenstein’s production took place. Before ending, I’d like to highlight one more connection between the first and the second Wittgenstein’s thoughts. In my view, the idea of a “picture of the world” which emerges from the remarks of On Certainty has an important antecedent in the Tractatus. The Tractarian idea which seems to have anticipated what would come many years later is the idea of formal concept. Concepts like object, person, number, etc., are formal concepts. That’s why to say that John is a person is to say nothing: if I know that ‘John’ is a proper name for persons, to say that is simply to repeat what we already know. On the other had, as soon as I say ‘John is nice’, I automatically generate the proposition ‘John is a person’. ‘To be nice’ is a proper or genuine concept, it can be ascribed truthfully or falsely of someone, etc., but it just makes no sense to say either that John is a person or that he is not. Here we find a liaison between the two great phases in Wittgenstein’s thought, a connection which deserves to be carefully examined. As in many other cases, Wittgenstein’s insight seems to be the same in both periods and what changes is the approach which moved from logic to grammar. About this, however, I’ll say nothing here. VII. Conclusions I have argued against the idea of a “third Wittgenstein”, with all it implies. I hope to have shown that nothing would be more damaging as to invent a cut in the second Wittgenstein’s philosophical production. I must admit that I left out lots of interesting subjects. For instance, I didn’t consider the relations that hold between rules of grammar and “hinge” propositions. There are those who have maintained that they are the same. In my view that is a serious mistake. The subject is certainly worth investigating and debating, for lots of questions would be solved depending upon how we see the issue. Regardless of this, the reward we get from our discussion is the idea of a new diagnostic for at least some philosophical puzzles: we are now in a position to understand that they are not the result of a violation of rules of grammar (in the Wittgensteinian sense, of course), but rather of the confrontation between what is being said and the fundamental propositions which constitute our conception of the world, i.e., our Weltanscahuung, which underlies our discourse in all spheres of life.
116
Wittgenstein: Issues and Debates
Bibliography Baker, G. P., 2004: Wittgenstein’s Method: Neglected Aspects. Essays on Wittgenstein. Blackwell, Oxford. Cook, John W., 1994: Wittgenstein’s Metaphysics, Cambridge University Press, Cambridge. Grice, P., 1989: Studies in the Way of Words, Harvard University Press, Cambridge, Mas. Harré, R. 2008: Grammatical Therapy and the Third Wittgenstein, Metaphilosophy, 39, 4-5, pp. 484 – 491. ---
2009: Wittgenstein’s Therapies: From Rules to Hinges, New Ideas in Psychology, 27, 2, pp 118-132.
Hertz, H., 1956: The Principles of Mechanics, Dover Phoenix Editions, Dover. Moyal-Sharrock, D. 2003: Logic in Action: Wittgenstein’s Logical Pragmatism and the Impotence of Scepticism, Philosophical Investigations, 26, pp. 125-148. ---
2004: Understanding Macmillan, New York.
Wittgenstein’s
On
Certainty,
Palgrave
Moyal-Sharrock, D. (Ed.), 2004: The Third Wittgenstein. The PostInvestigations Works, Ashgate Wittgensteinian Studies, Farnham, Surrey. Stroll, A., 2002: Wittgenstein, Oneworld, Oxford. Wittgenstein, L., 1969: On Certainty, Basil Blackwell, Oxford. ---
1974: Philosophical Investigations, Basil Blackwell, Oxford.
---
1978: Tractatus Logico-Philosophicus, Routledge and Kegan Paul, London.
---
1994: Observaciones sobre los Colores. Translated into Spanish by Alejandro Tomasini Bassols, Instituto de Investigaciones Filosóficas/Paidós, Barcelona.
Hacker on Wittgenstein’s Ethnological Approach Lars HERTZBERG P. M. S. Hacker, in his essay ‘Wittgenstein’s Anthropological and Ethnological Approach’1, comments on Wittgenstein’s remark: If we use the ethnological approach, does that mean we are saying that philosophy is ethnology? No, it only means that we are taking up our position far outside, in order to see things more objectively.2
Hacker links this remark to what he considers to be a change in Wittgenstein’s philosophical outlook around the beginning of the 1930’s, from the point of view of the Tractatus, according to which philosophy was an investigation into the essence of the world, to that of the later philosophy, according to which “the task of philosophy is to investigate the uses of words that are the source of conceptual problems and confusions”3. I shall not address the question whether this construal of Wittgenstein’s development is adequate – according to some writers, the transition from the period of the Tractatus to that of the Philosophical Investigations is not so sharp as Hacker makes it out to be. What I wish to discuss, however, are some points concerning Hacker’s reading of the idea of an ethnological approach. I shall start by discussing, and agreeing with, Hacker’s claim that in philosophy language is described from an internal point of view. After that, I shall make some critical comments about Hacker’s use of the word “concept”, especially in connection with language learning. Finally, I address what appears to me to be an obscure point about Hacker’s idea of philosophical clarification. 1.
An internal vs. an ethnological point of view
Hacker’s presentation is lucid, and there is much in his essay that I agree with. Thus, a remark I find important is the following: 1
Hacker, 2010. Wittgenstein, 1998, 45e. 3 Hacker, 2010, 17. 2
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ), Ontos Verlag, Frankfurt a. M., 2010, 117-126.
118
Wittgenstein: Issues and Debates … it is the normative practices of the speech community that fix and hold firm the internal relations between a word and its application, between explanation of meaning and what counts, in the practice of using the word, as correct use, as well as what is determined as following from its use in an utterance.4
This point, as I understand it, might also be made by saying that it is through speakers’ responses to utterances that distinctions such as that between correct and incorrect enter into language. The misunderstanding we need to guard against is that we could base a description of correct use on neutral observations of the linguistic behaviour of the members of a speech community. That such an attempt could not achieve its purpose is clear from the fact that someone who does not herself have command of the language would not be able to tell what would be the relevant linguistic features, indeed, would have no way of distinguishing speech from other forms of behaviour. When Quine argues that the analytic/synthetic distinction cannot be upheld 5 , the reason for this seems to be that he considers observation of behaviour the sole source of judgments about meaning. Observation gives no basis for distinguishing between assent based on conceptual relations and assent based on agreement concerning empirical fact. However, if neutral observation were all that is allowed, we could not even get as far as Quine assumes. We need what Hacker calls an internal point of view; if that is granted, however, Quine’s reason for questioning the analytic/synthetic distinction dissolves.6 On the other hand, as will be seen, I have some disagreements with the way Hacker construes that point of view. 2.
Historicism without history
Hacker’s point about Wittgenstein’s approach involving historicism without history is also illuminating, while at the same time giving rise to some questions. Hacker writes: The concepts employed by different linguistic and social groups are the product of social interaction, responses to shared needs … common interests called forth by the varying circumstances of social life.7
Attending to the social circumstances of linguistic interaction is an important corrective to more traditional forms of philosophical analysis, in 4
Hacker, 2010, 19. Quine, 1963. 6 There are other reasons for questioning the distinction, but I shall not go into them here. 7 Hacker, 2010, 20. 5
Hacker on Wittgenstein’s Ethnological Approach
119
which one tends to focus narrowly on words or sentences considered, as it were, by themselves. Wittgenstein exhorts us to step back and attend to the situations in which those sentences have a role in conversation. This, however, should not be confused with a factual inquiry, aimed at finding an account of how culture has given rise to those forms of expression. What one is attempting to do, rather, is to get a clearer understanding of the senses of words in their actual use, so as to counteract the philosophical tendency to misconstrue their sense in accordance with certain preconceived ideas. This, I take it, is why Hacker speaks about a historicism without history. However, when Hacker speaks about concepts being the product of social interaction, this way of putting matters is liable to mislead. It suggests that one is speaking about a relation between two separate terms, the forms of interaction on the one hand and the concepts on the other hand. This misses the point that the relation in question is internal: it is only in the context of social interaction that concepts are what they are. Only when they have a role in human intercourse can words uttered be said to express a sense. Or better put: to speak of the sense of words is to speak of what speakers do in uttering them. In short, the important dependence here is logical, not historical. What there is to be done in uttering words is bound up with the forms of human interaction that exist in the speakers’ society. Reflecting on possible or actual cultural variations may be an important aid in trying to overcome philosophical confusion, since it helps free us from the idea that, as Wittgenstein puts it, in a passage quoted by Hacker, “certain concepts are absolutely the correct ones”8, and thus helps us turn the focus of our attention from the world we speak about to the things we do in speaking about it. 3.
Concepts and learning to speak
Hacker, following Wittgenstein, uses the word “concept” in speaking about these questions. It seems to me that this way of speaking is fraught with certain risks, and I shall try to point out ways in which those risks become manifest in Hacker’s essay. Consider, again, the notion that concepts are 8
Wittgenstein, PPF, § 366. References to Wittgenstein’s “Philosophy of Psychology – A Fragment” (previously known as Part II of Philosophical Investigations) will be given with PPF and section number.
120
Wittgenstein: Issues and Debates
the product of social interaction. What are we to suppose that product to consist in? If we take seriously, as I suggest we should, the notion that our concepts – our uses of words – are constituted by forms of social interaction, then what we get is the idea that forms of social interaction are produced by forms of social interaction. This in itself, while it may sound odd, could be a sensible claim to make, if it is taken to mean that the forms of interaction in a society tend to be the historical product of earlier forms of interaction. But this can hardly be what Hacker wants to say. For one thing, this is no longer a remark about language. What he does want to say, however, remains obscure. This obscurity, I would contend, is bound up with the use of the word “concept”. When we think of learning to speak as the acquisition of concepts, we may get the picture that the process is mediated through the formation of certain entities (“meaning kernels” as it were) that are then applied in the use and understanding of words. Hacker, it is true, explicitly rejects the suggestion that concepts are to be thought of as entities: Wittgenstein treats concepts not as entities to be discovered, but as techniques of using words. To have mastered a certain concept is to have mastered the technique of the use of a certain word in some language or other. To possess a concept is to be able to use a word or phrase correctly, to explain what one means by it in a given context, and to respond with understanding to its use.9
However, to speak of techniques in this connection is to retain the notion of a meaning kernel though in modified form. It has a strongly instrumental ring, and this is underscored by Hacker himself: Concepts … are comparable to instruments made for human purposes, and their acquisition is comparable to the mastery of the technique of using an instrument. They are rule-governed techniques of word use. … their techniques of application are exhibited in the use of words in practice.10
To regard a use of words as instrumental is to think of it as employed in the service of some purpose that exists independently of one’s mastery of those words. Certainly, such uses of words exist. An obvious case is the builders’ game in Philosophical Investigations § 2. Here the technique of calling for building blocks is simply added to the activity of constructing a building; it is a limited technique within a larger whole. But it is certainly a simplification to regard this as a paradigm of the kind of learning that goes on as we learn to speak. Wittgenstein warns us against the tendency to treat 9
Hacker, 2010, 18. Hacker, 2010, 18.
10
Hacker on Wittgenstein’s Ethnological Approach
121
various uses of words in accordance with a single pattern. Thus, after comparing the uses of words to tools in a tool-box, he writes: Suppose someone said: “All tools serve to modify something. So, a hammer modifies the position of a nail, a saw the shape of a board, and so on.” – And what is modified by a rule, a glue-pot, and nails? – “Our knowledge of a thing’s length, the temperature of the glue, and the solidity of a box.” – Would anything be gained by this assimilation of expressions? –”11
Consider, say, a child learning to ask for a drink of water. This is hardly to be understood in terms of the child’s recognizing that she is thirsty, then developing a technique for setting in motion a process that will ultimately lead to her having her thirst quenched; rather in learning to ask for a drink the child develops an understanding of what it means to be thirsty. This is part of the story: there is of course an element of reciprocity in learning to understand about thirst: I do not know what it means to be thirsty unless I realize (whether I act on it or not) that someone else’s expression of thirst may involve a call on me to give him something to drink. I believe this to show that it is not illuminating to speak of learning to express thirst or of learning to understand expressions of thirst in terms of the acquisition of a technique for achieving certain ends. A more natural way of describing what goes on is to say that the word “thirst” comes to be incorporated into the child’s life, into her relations to the people around her. There is no meaning kernel that can be considered in separation from the forms of social interaction in which the word “thirst” has its use. Similar observations can be made about the learning of many other types of expression, say, learning to express one’s intentions (as well as to understand other people’s expressions of intention). Here is what Hacker has to say about expressions of intention: Here we do not graft a piece of linguistic behaviour onto natural expressive behaviour, rather we introduce a piece of linguistic behaviour that heralds an action. We say ‘I’m going to V (throw the ball, give you the ball)’ and immediately go on to V. The child’s initial use of ‘I’m going to’ is to herald an action. And from this primitive beginning, long term intentions and their expression grow, and the nexus with immediate performance weakens.12
This is a somewhat simple story. The difference between declaring an intention as a way of heralding an action and using expressions of intention in the context of human interaction is not just a difference in time lapse. Learning to express intentions and to understand expressions of 11 12
Wittgenstein, PI, § 14. Hacker, 2010, 26.
122
Wittgenstein: Issues and Debates
intention, to an important degree, is coming to understand the ways in which human activities may interlock or clash, to understand notions like commitment, compliance, defiance, negotiation, threat, etc. A similar limitation of perspective gets expressed in Hacker’s claim that we are … inducted into a human community by being trained to imitate, drilled to repeat, and later: learning and being taught how to do things with words, how to engage in innumerable language-games in the human community of family and friends, and later strangers too. The words with which we learn to do things are, of course, rule-governed. Their rule-governed employment is manifest in a regularity that presupposes recognition of a uniformity.13
In fact, this passage expresses several different conceptions of what goes on in language learning, none of which, in my view, is able to accommodate the way in which speaking becomes an extension of the speaker’s life. Neither being trained to imitate, being drilled to repeat, nor learning to recognize a uniformity makes room for the notion that learning to speak means learning to express oneself by means of words.14 The same is probably true of what Hacker refers to by the Austinian phrase “learning to do things with words” – whether it means mastering their use as a tool for achieving various purposes along the lines described above, or learning to obey conventions for word use (later on, p. 29, Hacker speaks about conventions determining the limits of language).15 This is not to deny that elements like those listed by Hacker play a part in many of the processes involved in learning to speak. Imitation and repetition are surely important, for instance, at the early stages at which a child learns to produce the sounds of her language or begins to interact with her elders, as well as in learning to do things like greeting people, saying thanks, etc.; recognizing uniformities, on the other hand, is an important part of acquiring, say, a colour vocabulary (though it plays little
13
Hacker, 2010, 19. We should consider too that a normal speaker does not in a robot like fashion repeat words or phrases she has heard, but utters them with intonations of pleasure, distress, effort, concentration, etc, or accompanies them with the appropriate expressions. 15 As for the idea that our life with language is governed by rules or conventions (a frequent misunderstanding of Wittgenstein), consider Wittgenstein, PI, § 83. What Wittgenstein is suggesting there is that, in many situations, rather than rules determining how we act, we may pretend to follow rules, make up rules for the occasion, etc. Letting oneself be bound by rules or conventions is just one way of relating to them. 14
Hacker on Wittgenstein’s Ethnological Approach
123
role in many other kinds of language learning).16 This said, it is important to recognize how much gets left out of all these accounts. Simply having been taught the knack of naming the colours of objects on sight is a long distance away from the mastery of colour words. Even if recognition games may be a preparation for speaking about colour, we will not say that a child knows what it is to speak of colours until remarks about colour come to constitute an intelligible part of her life. Connected with this is the fact that coming to relate to someone as a speaker is not a matter of noting that she lives up to some standard or conforms to some pattern, rather, it is a matter of our finding ourselves interacting with her in ways involving words. 17 Elsewhere, Hacker hints at yet another account of language learning, when he says that concepts “are given by explanations of word meaning”18. Yet surely, being able to understand and apply explanations of word meaning requires a fairly high degree of linguistic sophistication. It cannot occur at an early stage of language learning. Right after this, Hacker says that “[t]he use of words is integrated into the activities of human beings in the stream of life”, but he gives no hint as to how we are to conceive of the process of integration. Is it external to the learning itself? Could a child be said to have learnt the use of certain words before the use was integrated into her activities, and if so, what is the role of the integration for how we think of language use: suppose there were speakers whose words were never integrated into activities; could they still be said to be using words? It is not clear how Hacker pictures the relation of the mastery of words to activities in the stream of life. For my part, I find it hard to see how that mastery could manifest itself in isolation from any activities. (Here, again, the idea of a meaning kernel seems to assert itself in Hacker’s thinking.). If Hacker’s account of learning to speak, thus, seems to leave out important elements, he himself seems to be aware that there are aspects of the process that his account is unable to capture. Thus, he acknowledges that [w]hat children learn is not how to translate their thoughts and wishes into words, but how to request, demand, beg, nag, ask and answer questions, call people and to respond to calls, tell people things and to listen to what others 16
At the end of the passage, just quoted Hacker refers to Wittgenstein’s, 1978, 348, where Wittgenstein speaks of the role of uniformity in learning to speak. But there is no suggestion that Wittgenstein intends this as a general account of what it means to learn to speak. 17 This is a central point of Segerdahl et al. 2005. 18 Hacker, 2010, 26.
124
Wittgenstein: Issues and Debates tell… As the linguistic behavioural repertoire of the child grows, so too the horizon of possible thought, feeling and volition expands. The child becomes able to think things he could not conceivably have thought, to feel things he could not possibly have felt, and to want things that no non-language using animal could intelligibly be said to want.19
Nevertheless, in representing the learning of concepts as the learning of a technique, in citing imitation, repetition and recognition as central to what it means to become a speaker, Hacker conveys the impression that language is a surface phenomenon, a mere set of conventions, something that could be skimmed off life like a cream. Surely, this is to misrepresent the place of language in our lives. This view of things may be an effect of regarding language learning under the aspect of concept formation, a perspective, which tempts us to regard the life we live with language in too abstract terms. (It is true that we find a similar tendency in Wittgenstein’s own work.). Next, I wish to argue that the emphasis on concepts has consequences for the way Hacker thinks about the role of philosophical clarification. 4.
On the nature of philosophical clarification
According to Hacker, philosophy “sketches the logical geography of those parts of the conceptual landscape in which we are prone to lose our way”20. The sketching of the landscape is not an end in itself, he adds, but is carried out in order to help us find our way. Now we may ask: what does it mean to lose our way in the conceptual landscape? This might be taken to be something that will occur in our day-to-day commerce with words. Or it might occur when we reflect on use. This contrast is important, but Hacker does not make explicit mention of it. In fact, his thinking on this score is somewhat obscure. Hacker says that philosophy, among other things, “invites us to bring to mind features of usage in order to get us to realise the way in which we are inadvertently misusing words”21, which seems to suggest that he thinks of philosophical confusions as arising in the actual use of language.22 But ordinarily philosophical confusion is taken to arise 19
Hacker, 2010, 19. Hacker, 2010, 17. 21 Hacker, 2010, 17. My italics. 22 Hacker also contrasts use with what he, somewhat confusingly, calls “comparative use”. He gives the example of the differences of use between “nearly” and “almost”, suggesting that while few “competent English speakers could, off the cuff, spell out the differences in use between ‘nearly’ and ‘almost’”, yet they would never say, 20
Hacker on Wittgenstein’s Ethnological Approach
125
when we reflect on the use of words, not when we use them. This, anyway, is Wittgenstein’s view. He quotes Augustine’s remark about time as an expression of the predicament typical of someone in the grips of a philosophical confusion: “quid est ergo tempus? si nemo ex me quaerat scio; si quaerenti explicare velim, nescio”.23 Why does Hacker ignore the all-important distinction between use and reflection? Here, of course, one can only speculate, but I would suggest that his thinking on this score is a natural consequence of the central role he gives to concepts. On the view I am attributing to him, the type of insight the philosopher needs in order to map the conceptual landscape is of a piece with the knowledge the child acquires in learning to speak. The philosopher makes explicit the child’s implicit knowledge. With this goes the idea that there might be a complete account of all the concepts of our language provided through what Hacker calls explanations of word meaning. Such an account, if we had it, would forestall the arising of philosophical puzzlement in advance. Against this, I would argue that the idea of a complete account of all our concepts is a chimaera; not because providing such an account would require an inordinate amount of time, but because the question of what the account would include is indeterminate. When we explain a word to someone, the form of our explanation will vary greatly depending on what the learner already knows or is able to do. Similarly, the type of clarification needed to resolve a philosophical puzzle will depend on the nature of our interlocutor’s bewilderment. Thus, we may have to discover what false analogies lead her thinking astray. As Wittgenstein puts it in PI: … One might say: an explanation serves to remove or to prevent a misunderstanding — one, that is, that would arise if not for the explanation, but not every misunderstanding that I can imagine.24 “There is not almost enough sugar in the pudding” as opposed to “There isn’t nearly enough sugar in the pudding”, Hacker, 2010, 18. But the inability Hacker is describing here clearly belongs to the context of reflection, not of use. Besides, it is hard to imagine the word pair “nearly” and “almost” giving rise to philosophical confusion. In all, it is hard to see what Hacker’s example is supposed to illustrate. 23 “What then is time? If nobody asks me, I know; but if I want to explain it to someone who asks me, I don’t know.” The quotation is in Wittgenstein, PI § 89. Consider also Wittgenstein’s oft-quoted remarks about philosophical confusions arising when language goes on holiday (Wittgenstein, PI § 38), or “when language is, as it were, idling, not when it is doing work” (Wittgenstein, PI § 132). 24 Wittgenstein, PI, § 87.
126
Wittgenstein: Issues and Debates
Accordingly, the type of clarification called for in philosophy is dependent on the actual confusions that arise. Where there are no confusions, there is nothing to be clarified, hence no task for the philosopher to carry out. If this point is correct, it is connected with what might be said about language learning. The child’s learning of new forms of expression is not to be thought of along the lines of the acquisition of knowledge, whether implicit or explicit. Rather, the child simply acquires new ways of acting and responding. Conclusion To give a substantial account of language learning is not, of course, a task for philosophy. Rather, philosophy’s concern with language learning is a matter of forestalling misunderstandings of what it means to become a speaker. This requires steering clear of both the Scylla of intellectualism (treating mastery of words as constituted by knowing how or knowing that) and the Kharybdis of mechanical conditioning. Both views fail to leave room for the way in which speaking develops organically within the life of the child. It is my sense that Hacker has not managed to avoid these risks. Bibliography Hacker, P.M.S., 2010: Wittgenstein’s Anthropological and Ethnological Approach, in: Philosophical Anthropology. Wittgenstein’s Perspective, (Ed. Jesús Padilla Gálvez), Ontos Verlag, Frankfurt a. M., 2010, 15-32. Quine, W. V. O., 1963: “Two Dogmas of Empiricism”, reprinted in From a Logical Point of View. Harper & Row, New York. Segerdahl, Pär et al., 2005: Kanzi’s Primal Language: The Cultural Initiation of Primates into Language, Palgrave Macmillan, Basingstoke. Wittgenstein, Ludwig, 2009: Philosophical Investigations, PI, revised 4th ed., Wiley-Blackwell, Chichester. ---
1978; Remarks on the Foundations of Mathematics, revised 3rd ed., Basil Blackwell, Oxford.
---
1998: Culture and Value: A Selection from the Posthumous Remains, German-English Edition, Ed. By G.H. von Wright. Blackwell, Oxford.
Wittgenstein’s Criticism against Gödel’s Project of Metalogic Jesús PADILLA GÁLVEZ Introduction Within the first decades of the last century, metamathematics had developed rapidly. This advance took place during a time when there existed a whole range of competing mathematical programs. Within this period of ongoing scientific progress, mathematicians used opposing philosophical positions within metamathematics. However, this provoked relatively few discussions. For instance, some mathematicians used an axiomatic approach, others applied an intuitionistic view, and certain questions could be solved by a Platonist view. G. Cantor’s metaphysic results were taken as an assumption and elements of intuitionism were used for argumentation. During the 1930s, the diverse prerequisites of the different approaches to metamathematics were rarely discussed. However, A. Tarski1, R. Carnap2, K. Gödel 3 and others appreciated the importance of L. Wittgenstein’s views. According to A. Tarski, R. Carnap and K. Gödel believed that L. Wittgenstein’s proposals were fundamental for syntax, metalogic and metamathematics. R. Carnap expresses this view in the following statement: My “syntax” has two historic roots: 1. Wittgenstein, 2. Metamathematics (Tarski, Gödel)4
In 1953, K. Gödel wrote a paper about mathematics in which he summarized Wittgenstein’s position: Around 1930 R. Carnap, H. Hahn, and M. Schlick, [...] largely under the influence of L. Wittgenstein, developed a conception of the nature of 1
Tarski 1936, 11. Carnap, Letter to Neurath of 23. 12. 1933. Hilman-Library, RC 29-03-06 A. PadillaGálvez 1999, 169, footnote 8. 3 Gödel 1995, 171. 4 “Meine ,”Syntax” hat historisch zwei Wurzeln: 1. Wittgenstein, 2. Metamathematik (Tarski, Gödel)”. Letter from R. Carnap to O. Neurath from 23 December 1933 in: Archives of Scientific Philosophy, Hilman-Library, RC, Nr: 029-03-06, SI. PadillaGálvez, 1998, 26. 2
Wittgenstein: Issues and Debates, (Eds. Eric LEMAIRE and Jesús PADILLA GÁLVEZ), Ontos Verlag, Frankfurt a. M., 2010, 127-148.
128
Wittgenstein: Issues and Debates
mathematics [...], which can be characterized as being a combination of nominalism and conventionalism.5
Inspite of Tarski’s and Gödel’s appreciation of Wittgenstein’s point of view, the latter did not accept his colleagues’ views. In the year 1933, L. Wittgenstein wrote an article with the provocative title: “Metamathematics does not exist”. 6 This article was published in Paragraph 109 of the Big Typescript 7 . However, this article was never really noticed by other mathematicians and did not generate any discussion in the secondary literature. The aim of this paper is to find out why his article did not generate any response within the scientific community. As we can see from the above statement, L. Wittgenstein was opposed to metamathematics. K. Gödel presented some interesting arguments at a time when metamathematics was a highly controversial issue. Nevertheless, why was L. Wittgenstein actually opposed to metamathematics? Does L. Wittgenstein opposition have anything to do with G. Frege and B. Russell’s logic rejection of the metatheoretic approach? Until now, these questions have not been touched upon in the secondary literature and they shall be brought to light in this paper. In order to get an insight into the argumentation we will first present K. Gödel’s arguments and then examine the assumptions on which his arguments are grounded. Secondly, we will present G. Cantor’s diagonal argument and Wittgenstein’s refutation of it. The contradiction between G. Cantor’s and L.
5
Gödel 1986 [1995], III, 334 and Gödel 1995, 171. “Es gibt keine Metamathematik” Wittgenstein PG, Teil II. Über Logik und Mathematik, 12 and Wittgenstein TS213, § 109, 539-541. 7 Many of the comments in his notebooks were summarized in 1932 in the form of a typescript, which became an important work of reference (Wittgenstein, Typescript based on 109-113 and the beginning of 114 (771 pp.), c 1932, Trinity College, 211). This was a big collection of type-written material containing all his thoughts and ideas which were prepared for publication. These were to be published later in his Philosophical Investigations. Wittgenstein seemed to handle his scripts in a rather unusual way by taking his typescript, cutting it into pieces according to topics of interest and having these topics retyped. These were to form The Big Typescript? (Wittgenstein, Typescript consisting of cut part from 208, 210 and 211 (3 boxes), c 1932-3, Trinity College, 212). In the course of this procedure he took his topics of interests, put them in several folders which he indexed by titles. One of these folders was labelled “Metamathematics does not exist”. Both, the original typescript and The Big Typescript contain parts of texts which were taken out of their context (Wittgenstein, The “Big Typescript” (776 pp), c 1933, Trinity College, 213). 6
Wittgenstein’s Criticism against Gödel’s Project of Metalogic
129
Wittgenstein’s view of the diagonal argument will be analysed. Finally, we will reconstruct Wittgenstein’s critical view of metamathematics. 1.
Gödel’s Result
D. Hilbert underlines that strict formalization of a theory requires a full abstraction from the meaning.8 The result is a formal system or formalism. Metamathematics involves the view of approaching the formal system as a whole. Metamathematics includes the description or definition of formal systems as well as the investigation of the properties of such formal systems. One of the most important aims of metamathematics is to provide a justification for the predicate calculus. It is rather based on mathematics and logic rather than on ordinary language. Consequently, its task is to find a proof for the internal consistency of a formal system. Any theory of calculus should avoid contradictions. Therefore a theory of a formula A cannot simultaneously prove A and non-A. The internal consistency of logic corresponds to that of mathematics. 9 D. Hilbert expressed this thought as 8
Hilbert, 1923, 151-165. and Hilbert / Ackermann, 1972. Wittgenstein summarized his discussions of 28 December 1930 as follows: The problem of internal consistency of mathematics stems from two different sources: 1. From the ideas of non-Euclidean geometry, which was dealing with the problem of proving the parallel axiom after the principle of reductio ad absurdum. 2. From the antinoms of Burali-Forti and from Russell. Also: “Das Problem der Widerspruchsfreiheit der Mathematik stammt aus zwei Quellen: 1. Aus der Ideen der nicht-euklidischen Geometrie, wo es sich darum gehandelt hat, nach dem gegebenen Vorbild einer reductio ad absurdum das Parallelaxiom zu beweisen. 2. Aus den Antinomien von Burali-Forti und von Russell.” (Wittgenstein, WWK, 121). The arguments can be outlined as follows: The problem of internal consistency is derived from the antinomy. But Wittgenstein thinks that these two elements don’t have anything in common, because the antinomy never occurs in the calculus but it appears in the ordinary language. The reason is that words are often used in an ambiguous way. This problem can be solved by using words precisely and distinctly, which would lead to the disappearance of antinomy, “but by analysis rather than by proof.” (Wittgenstein, WWK, 122) Then the types of different proofs applied in mathematics were listed: 1. The proof, that gives evidence for a certain formula. This formula occurs in the proof as its last element. 2. The inference proof. What is striking is that the proposition that is supposed to be proved does not appear in the proof. This means, that the inference is not a method that leads to a proposition, but it shows us infinite possibilities, which is a relevant characteristic of the inference proof (Wittgenstein, WWK, 135). Mathematical induction is a paradigm of a method for proving generality propositions about the natural numbers. (When Poincaré defends mathematical induction as an irreducible tool of mathematical reasoning, is also a forerunner of the constructivist point of view (See: Poincaré 1902). A proof by 9
130
Wittgenstein: Issues and Debates
follows: “The absolute truths are those insights that were given by my theory of proof and the internal consistency of those formula-systems.”10 In an article written in 1930, K. Gödel obtained his results by metamathematic reasoning that has a formal structure and is considered a system of objects. What is new about this view? He starts from the assumption that objects of a formal system are formal symbols, formal expressions and finite sequences of formal expressions. There is an enumerable infinity of formal symbols given at the outset. Formal objects form an enumerable class. By specifying a particular enumeration of them, and letting, our metamathematical statements refer to the indices in the enumeration instead of the objects enumerated. Metamathematics becomes a branch of number theory. In this result, the use of G. Cantor’s diagonal
induction of the proposition for all n, P(n) shows that any given n would have to have the property P, by reasoning which uses only the numbers from 0 up to n. Certainly, for a particular proof by induction to be constructive, also the reasoning’s used within its basis and induction step must be intuitionistic. An existence statement there exists a natural number n having the property P, or briefly there exists an n such that P(n), has its constructivist meaning as a partial communication of a statement giving a particular example of a natural number n which has the property P, or at least giving a method by which in principle one could find such an example. Therefore a constructive proof of the proposition there exists an n such that P(n) must be constructive in the following sense. The proof actually exhibits an example of an n such that P(n), or at least indicates a method by which one could in principle find such an example. L. Wittgenstein believes that the antinomy has encouraged him to deal with the problem of internal consistency. His line of argument seems to be wrong in this aspect because the antinomy does not relate to the internal consistency of mathematics. The antinomy normally appears in the ordinary language rather man in the calculus, because words are often used in an ambiguous way. The solution to this problem would be to replace ambiguous meanings with precise ones. The antinomy disappears through analysis rather than by proof. Thus, if ambiguity leads to contradictions in mathematics, it can never be solved by a proof. And as a result of this, he states, that there can never be a proof for internal consistency (comparing the contradictions in mathematics to those in predicate calculus). On the basis of the discussion with Hilbert’s ‘New Foundation of Mathematics’, Wittgenstein points out, that a proof for internal consistency must evolve to understand the rules. These rules are usually described by the inference proof. The inference suggests infinite possibilities. A contradiction cannot be considered as rule because the grammar of the word “rule” is of a kind, that a contradiction is not a rule (Wittgenstein, WWK, 194). 10 “Als die absoluten Wahrheiten sind vielmehr die Einsichten einzusehen, die durch meine Beweistheorie hinsichtlich der Beweisbarkeit und der Widerspruchsfreiheit jener Formelsysteme geliefert werden.” Hilbert, 1923, 153.
Wittgenstein’s Criticism against Gödel’s Project of Metalogic
131
method can be found a closed formula, which asserts its own unprobability. We will not present the general argument. K. Gödel’s propositions includes that every system of proof is incomplete for the set of arithmetic true formula. Every formal theory that contains cardinal numbers is not provable within this theory. If a first order predicate logic is to fulfil the conditions for incompleteness it is necessary that, firstly, a complement to each expression G(x), representing a set, can be described. Secondly, each expression G(x) of a described as set M is characterised by the set M*={x|d(x)M}. The symbol d(x) represents the diagonalisation of x. And thirdly, the set of provable expressions of the theory must be describable by an expression of a form of G(x). According to the theorem of Löwenheim-Skolem a model with the cardinality of signature can be found.11 For normal theories, a countable model exists, such as the natural numbers. K. Gödel proposed to put the formulas of the theory in the position of being their own object. Therefore, there is a Gödel number of the formula by bijectively transforming them into natural numbers. This can be achieved, for instance, by relating each symbol of a signature to a number, which is then concatenated. If ‘0’ is related to ‘1’ and identity (“=”) is related to “2” then the Gödel number of the formula “0=0” is “121”. The operation of concatenation can easily be achieved through exponent. Both, the syntactically well-formed formula and the provable formula can be described as arithmetic expressions. The diagonalisation in Gödel’s proof is the application of an expression P(x) on the Gödel number. If the Gödel number of an expression, and therefore of the series, P(x) is for example 121, then the diagonalisation of this number 11
L. Löwenheim proved that for any set of propositions of standard predicate logic, if there is an interpretation in which they are true in some domain, there is also an interpretation that makes them true in a countable subset of the original domain (Löwenheim, 1915, 447-470 and Skolem, 1929, 1 ff.). The result was called a paradox since it was believed that certain sets of axioms characterised the real numbers, and now Löwenheim’s result showed that the same axioms must hold in a countable subset of the real numbers. It also seemed to attempt to axiomatic set theory somewhat of a problem. Although it seems to contradict common sense (as do other results which depend on the Axiom of Choice), there is no paradox. The result implies that no uncountable mathematical system, such as those involved in analysis, geometry, and set theory, can be characterised up to isomorphism using only first-order propositions. If one examines the case of the real numbers more closely, then the axioms for an ordered field are all first-order propositions. Löwenheim’s result then shows that the real numbers contain a countable ordered field, which then cannot satisfy the leastupper-bound axiom, which is a second-order proposition.
132
Wittgenstein: Issues and Debates
is the Gödel number of P(121). If expression B(x) says that x is provable and the Gödel number of B(x) is 121, then its negation ¬B(121) is a nonprovable expression. What are the steps to come to this implication? The Gödel number 121 is provable. The Gödel number of B(x) is 121. Therefore, the expression ¬B(121) is non-provable. If these assumptions are correct then they are true but not provable.12 As indicated at the beginning, L. Wittgenstein underlined that the definitions proposed by Gödel were not exclusively arithmetic. Therefore, he made the following suggestion:13 (1)
Proposition An(n) is improvable in S.
Definitions contain syntactic expressions such as “proposition”, and for instance, provable, booth of which are expressions that cannot be formalised.14 But Gödel has not reflected upon this incongruence in his results. The expressions belong to the syntax of system S and are no number predicates. In order to avoid this problem proposition (1) must be transformed into the following proposition (2) that contains free variables: (2)
¬ [ f ( n , n ) T ]
In (2) all syntactic expressions such as “proposition” and provable are replaced by arithmetic expressions. As it denotes a quality of numbers it is transformed into an arithmetic proposition. There must be a proposition that expresses the quality of proposition (2) within S that may be achieved by way of explicit description. We assume the statement Ar of the infinite sequence A1, A2, A3, ... denoting the quality (2) within S. If Cantor’s diagonal method is applied within Ar the free variable is substituted by the n-tuple figure. However, this results in a contradiction that has the form of equivalence: A r (r)Ł¬A r (r). L. Wittgenstein had always been very critical towards Gödel’s project of metalogic. Accordingly, we have tried to show that the project has left several questions unsettled. Firstly, expressions used such as “proposition” or provable lack definiteness and L. Wittgenstein objected that Gödel’s definitions were not exclusively arithmetic. They contain syntactic expressions such as proposition and provable that cannot be formalised. K. Gödel has not taken this deficiency into account in his results. Secondly, the introduction and application of Cantor’s diagonal method leads to several problems. L. 12
Padilla Gálvez, 1999, 167ff. See: L. Wittgenstein’s original statement: (1) Der Satz An(n) ist unbeweisbar in S. 14 Padilla Gálvez, 1994, 43ff.
13
Wittgenstein’s Criticism against Gödel’s Project of Metalogic
133
Wittgenstein was aware of the fact that many metamathematic results had used Cantor’s diagonal method. Consequently, he was interested in the tacit presuppositions on which Cantor’s proof was based. In this paper, we will analyse this second objection. 2.
Cantor’s Diagonal Argument
As mentioned above, a relevant problem within Gödel’s results is the tacit presuppositions on which his proofs are based. One of these presuppositions has to do with Cantor’s diagonal argument. L. Wittgenstein made the following conceptual conjunction: if it were possible to set up an opposing model to the diagonal argument, Gödel’s results would loose their overall importance. In other words, a relevant prerequisite becomes doubtful and metamathematics would loose its significance. We will show that L. Wittgenstein had actually set up an alternative diagonal method that put Cantor’s results into question. G. Cantor’s diagonal argument is a proof that there are infinite sets, which cannot be put into a one-to-one correspondence with the infinite set of natural numbers.15 These are commonly known as uncountable sets, the 15
Cantor, 1890-1, 75-78 and Cantor, 1980, 278-281. L. Wittgenstein challenged the belief according to which the rules of classic logic have absolute validity, irrespective of the content to which it is applied. His discussion of two obvious paradigms reveals that principles that are valid for thinking about finite sets cannot be transferred to infinite sets. One is the principle that the whole is greater than any proper part, when applied to a 1-1 correspondence among sets. Another is that a set of natural numbers contains a greatest number. A principle of classical logic, valid in reasoning about finite sets, which Wittgenstein does not accept for infinite sets, is the law of the excluded middle. In its general form the law assumes for every proposition A, either A or non-A. If A is the proposition there exists a member of the set S that have the property P. Then non-A is equivalent to every member of S that does not have the property P, or in other words every member of S has the property non-P. The law, applied to this A, hence gives either there exists a member of S having the property P, or every member of S has the property non-P. For definiteness, let us specify P to be a property such that, for any given member of S, we can determine whether that member has the property P or does not. Now suppose S is a finite set (Now suppose the set S is finite). Then we could examine every member of (the set) S in turn, and thus either find a member having the property P, or verify that all members have the property not-P. There might be practical difficulties, e.g. when S is a very large set having say a billion members, or even for a small (set) S when the determination whether or not a given member has the property P may be tedious. But the possibility of completing the search is possible. It is this possibility (choice) which for the constructive point of view makes the law of the excluded middle a
134
Wittgenstein: Issues and Debates
size of infinite sets being treated by the theory of cardinal numbers initiated by Cantor. The diagonal argument is a general technique that had been applied in a great number of proofs, the most well-known being Gödel’s incompleteness theorem. The simple version of Cantor’s diagonal method is employed in the classic proof of uncountability of the real numbers between 0 and 1 that finds its expression in the form of decimal fraction. An uncountable set is an infinite set that is too big to be counted. A set is called countable if it is either finite or else there exists a bijection to the set of whole natural numbers. A set is considered uncountable if its cardinality (quantity/number of elements) is bigger than that of the set of natural numbers. Or rather, a set is considered uncountable if each list x1, x2, x3, ... of the elements of the set is incomplete. In fact, G. Cantor’s second diagonal argument is a proof for inconsistency which he used in order to verify the uncountability of real numbers. Each list of real numbers is obviously incomplete. This implies the possibility of each list being extended to a new list that remains countable and incomplete.16 valid principle for reasoning with finite sets S and properties P of the kind specified. For an infinite set, the situation is basically different. It is no longer possible in principle to search through the entire set. Moreover, in a situation the law is not saved for the constructivist by substituting, for the impossible search through all the members of the infinite set, a mathematical solution of the problem posed. We may in some cases, i.e. for some sets S and properties P, succeed in finding a member of S having the property P; and in other cases, succeed in showing by mathematical reasoning that every member of S has the property not-P, e.g. by deducing a contradiction from the assumption that an arbitrary member of S has the property P. However, we have no ground for affirming the possibility of obtaining either one or the other of these kinds of solutions in every case. L. Wittgenstein differs essentially in their view of the infinite from the theories of Weierstrass, Dedekind and Cantor. From the constructive point of view, the infinite is treated as potential or constructive. In the classical position, the infinite is treated as actual or completed. An infinite set is regarded as existing as a completed totality. L. Wittgenstein’s criticism of the Gödel’s metalogic as applied to an infinite set S arises from this perspective respecting infinity. 16 In mathematics, an uncountable set is an infinite set, which is too big to be countable. The uncountability of a set is closely related to its cardinal number; a set is uncountable if its cardinal number is larger than that of the natural numbers. The related term nondenumerable some authors use set as a synonym, for “uncountable set” while other authors define a set to be nondenumerable if it is not an infinite countable set. There are many equivalent characterizations of uncountability. A set X is uncountable if and only if any of the following conditions holds: 1. there is no injective function from X to the set of natural numbers. 2. X is nonempty and any Ȧsequence of elements of X fails to include at least one element of X. That is, X is nonempty and there is no surjective function from the natural numbers to X. 3. the
Wittgenstein’s Criticism against Gödel’s Project of Metalogic
135
The result of the diagonal method is to reveal with simple means that no list or sequence of real numbers can ever include all real numbers. Therefore, the totality of real numbers is uncountable. We assume that there is an infinite sequence within arithmetic language. Within this language, a property of numbers is expressed through a proposition. This description results in a contradiction. G. Cantor asked if the set of real numbers ( ) is uncountable. He analysed the open interval (0,1) and started by claiming that it is not countable. In order to prove this contradiction we assume that (0,1) be countable. This leads to the following relation:
f
x
¦x
j
2
j
, whereby
j 1
xj{0,1} and xj0. If j then is xj=1 and jj0 is not allowed. We then denote (xj)j as a sequence with a range of {0,1}, which means that x: ĺ{0,1} is an arbitrary mapping. If the transformation from A to B through mapping as Map(A,B) is (sufficiently) defined it results in the following injective mapping: \:(0,1) x(x(j))j Map( {0,1}), with Map( {0,1})\\((0,1)) being countable. If (0,1) is countable the same is true for \((0,1)), and consequently the Map(( {0,1}) is also countable. That produces the following bijective mapping: Ȝ: kx(k)Map( {0,1}). In that case, we can apply G. Cantor’s diagonal method in the following form: 1
2
3
4
5
1 x(1)(1) x(1)(2) x(1)(3) x(1)(4) x(1)(5) 2 x(2)(1) x(2)(2) x(2)(3) x(2)(4) 3 x(3)(1) x(3)(2) x(3)(3) 4 x(4)(1) x(4)(2)
5 x(5)(1)
cardinality of X is neither finite nor equal to ʠ0. 4. the set X has cardinality strictly greater than ʠ0. The best known example of an uncountable set is the set of all real numbers; Cantor's diagonal argument shows that this set is uncountable. The diagonalization proof technique can also be used to show that several other sets are uncountable, such as the set of all infinite sequences of natural numbers and the set of all subsets of the set of natural numbers. The Cantor set is an uncountable subset of .
136
Wittgenstein: Issues and Debates
Interestingly enough a constellation where y: ĺ{0,1} is combined with y(l)=1-x(l)(l), provided that y is not part of Map( {0,1}), results in a contradiction. The comments in L. Wittgenstein’s Bemerkungen über die Grundlagen der Mathematik reveal that he must have intensively dealt with Cantor’s argument. He realized the relevance of such argument and tried to falsify it by constructing a counter argument. It was many years later that he clearly described the method utilized and made the following comments on Hilbert’s statement about Cantor’s paradise: “Nicht einmal im Traum würde ich versuchen, jemandem aus diesem Paradies zu vertreiben. Ich würde etwas ganz anderes versuchen, nämlich zu zeigen, daß es gar kein Paradies ist – so daß man es aus eigenem Antrieb verläßt.”17 3.
Wittgenstein’s Alternative
L. Wittgenstein analysed the assumption according to which the diagonal sequence d defines a b-adic fraction and therefore denotes a real number. In this respect he doubted whether Cantor’s diagonal argument could actually prove the uncountability of real numbers.18 Wittgenstein presupposed that the diagonal method would not create an infinite amount of real numbers countable by means of the finite method. He pointed out that the sequence of numbers produced by the diagonal method directly depended on the axiomatic system and not on the diagonal method. He said: Wendet man meine Betrachtung auf das Cantorsche Diagonalverfahren an so ergibt sich: Eine unendliche Menge von Dezimalbrüchen kann nur ein Gesetz bedeuten nach dem Gesetze gebildet 0 a11a12 a13a14 .... werden und das heißt eigentlich eine Funktion von 0 a12 a22 a23a24 .... zwei veränderlischen. F(x,y) ist die funktion von 0 a31a32 a33a34 zwei Veränderlichen. F(x,n) ist der n-te von ihnen und F(m,n) seine m-te Stelle. Der Dezimalbruch nach der Diagonale genommen ist F(x,x) und verändert lautet er etwa F(x,x)+1 (dazu müßte festgesetzt werden, daß 0+1=1, 1+1=2, ...., q+1=0 etc ist) Und nun zeigt ein Inductionbeweis daß F(x,x)+1 eine andere Entwicklung hat als jedes beliebige F(x,y). Wo aber ist hier das höhere Unendliche? (oder gar das “eigentlich Unendliche”).19 17
Wittgenstein 1978, 121. Wittgenstein 1974, BGM, II, §§ 1-8, 125 ff. 19 Wittgenstein 1999, WA, 2, 268. 18
Wittgenstein’s Criticism against Gödel’s Project of Metalogic
137
If we take this statement into account, it seems that L. Wittgenstein did not consider diagonal numbers as numbers. Moreover, he asked whether it is justified that the diagonal numbers are different from the sequence f of enumeration.20 Presumably, he must have mistrusted the diagonal method as he points to Cantor’s assumption that the vertical axis is mapped onto the expansion of diagonal numbers. 21 According to Cantor’s definition enumeration is not an element of enumeration. The diagonal number D=limnĺDn has an infinite number of n-tuple sets. As each n-tuple expansion of diagonal number exactly avoids n numbers of the enumeration, each n-ary expansion of diagonal number covers the n-ary part of the expansion of the vertical sequence (described as Vn ). Therefore, Cantor’s diagonal method delimits the length of the diagonal sequence by the amount 1, producing the following result:22 (3)
Df Vf
lim nof
( Di )i d n (Vi )i d n
lim nof
n n
1
In (3) it is assumed that the finite expansion of the vertical sequence is covered by a finite expansion of the diagonal sequence. According to this view, every value of a finite enumeration is defined by the fact that this range is different from the diagonal sequence. L. Wittgenstein argued, however, that part of the enumeration is finite. He tried to show that the part of the range, which the diagonal number successfully manages to avoid, is very small in contrast to the length of the total list. Therefore, the limit of finite diagonal numbers must result in cero. He underlined that there exist infinite values for counting infinite sets, “...für die nicht bestimmt ist, ob sie von der Diagonalreihe verschieden [sind] oder nicht.” 23 This alternative version may be described by the following formula: (4)
lim
n of
Df Vf
lim n of
( Di )i d n (Vk ) k d 10
n
n n n of 10
lim
0
Formula (4) shows some relevant aspects, such as that an existing limit (D ) does not automatically require a diagonal number that functions as i i d n
(Vk ) k d 10 n
a limit of the sequence of its finite expansion. Neither does it require a range
20
Wittgenstein 1974, BGM, II, §§ 9-15, 127-129. Cantor, 1890-1, 75-78 and Cantor, 1980, 278-281. 22 Redecker, 2006, 67. 23 Wittgenstein 1974, BGM, II, § 9, 127. 21
138
Wittgenstein: Issues and Debates
of all infinite decimal numbers. The last part such as lim nof
n 10 n
0
indicates that
for every value İ>0 there is a natural number, so that for all n the formula n