Wittgenstein and the Philosophical Investigation by Marie McGinn

(Routledge: ISBN 0-415-11191-9)

These are excerpts from the above book and some guidance to read the book. This page is made for Yanase's students' convenience. In some parts the original sentences of the book are slightly altered for editing, although the meaning of is it stays the same to the best of Yanase's editing skills.


>Wittgenstein's life briefly described in Introduction is worth reading. Although one's philosophy is not a direct reflection of one's life, both are, I believe, undeniably related.

>Although there are undoubtedly continuities between the early and the later work ---- in particular, the conviction that 'most of the propositions and questions of philosophers arise from our failure to understand the logic of our language' (TLP 4.003) ---- Wittgenstein's manner of approaching the task of resolving philosphilcal confusion gradually undergoes a fundamental shift of focus. Where the Tractatus attempts to set a limit to what can be thought by means of an abstract theory that sets precise and exact limits to what can e said, the later Wittgenstein moves more and more towards a concern with the detailed workings of the concrete, complex, mutifarious and indeterminate phenomenon of language-in-use. Instead of trying to resolve the confusions about the logic of our language ---- which he believes to be the root of philosophical problems ---- by means of a clear theory of what the the essence of language consists in, Wittgenstein gradually develops a number of techniques for conutering these misunderstandings by means of a clarified view of how language actually functions within the everyday lives of speakers. (pp.4-5)


>Wittgenstein's Philosophical Investigations is concerned with two principal topics: the philosophy of language and philosophical psychology. (p.9)

>W emphasizes over and over again that it is a method or a style of thought, rather than doctrines, that characterizes his later philosophy. It is, moreover, his insistence that his philosophical aims do not involve him in putting forward 'any kind of theory' (PI 109) that makes the question of method, and of how to read his remarks, such a difficult one, for it suggests that we cannnot approach the book in the usual way, with a view to finding and extracting the views which are expressed in it. (p. 10)

>Think of the following remarks by W: "we may not advance any kind of theory (PI 109)," "philosophy neither explains nor deduces anything(PI 126)," "we must do away with all explanation and description alone must take its place (PI 109)."


>Read the very important paragraph on Page 12 and the three quotations tht follow. W sees the difficulty of his work 'not as the intellectural difficulty of the science, but the difficulty of a change of attitude.

>The idea of 'a grammatical investigation' is central to W's later philosophy, and it is the key to understanding his work. The Investigations can be seen as a large collection of particular grammatical investigations, each one of which examines the detailed workings of an area of our language that has become a focal point for philosophical myth and confusion. (p. 13)

>W's use of the concept of 'grammar' relates, not to language considered as a system of signs, but to our use of words, to the structure of our practice of using language. The concept of 'our practice of using language' is here intended to invoke the idea of language, not as 'some non-spatial, non-temporal phantasm', but as a 'spatial and temporal phenomenon', i.e. as the phenomenon of language-in-use. W's grammatical method is one in which 'we remind ourselves' of the details of the distinctive patterns of use that constitute what he calls 'the grammar of our concepts'. (p.14)

>When W speaks of our need to 'command a clear view ofthe use of our words', he is thinking both of our need to uncover the conflict between our philosophical notions and the way our concepts actually function, and of our need to become aware of the grammatical differences in how the concepts in the different regions of our language are used. (p.15)

>Read quotations of PI 340 and PI 118 on page 15.

>You might quite disagree with W: Surely, we feel, language and mental states are phenomena that cry out for explanation. There must, for example, be some explanation of what language's ability to represent the world consists in, of what our understanding of our language consists in, of what thinking is, of what an intention, or a sensation, is, and so on. how could it possiblly be wrong or inappropriate to try to elucidate these phenomena, to say what they consist in, or to offer some sort of explanatory account of them? (p.16)

>We feel that it is only by means of elucications which offer us some sort of account of these phenomena that our urge to understand them more clearly could possibly be met. (p.16)

>In order to understand unambiguously what W says in the first place, let us be clear about what W opposes. Theorizing or thoeretical attitude is the attitude that W means to characterize when he says that 'we feel as if we had to penetrate phenomena' (PI 90).


>Read the first paragraph and the three quotations on page 17.

>W does not intend here to express any general opposition to sicence. It is rather that the methods of science, or more particularly the way of asking and answering questions in science, are misleading and inappropriate when applied to questions like 'What is meaning?', 'What is thought?', and so on. (p. 17)

>Read the two quotations on page 18.

>What we are concerened with when we ask questions of the form 'What is time?', 'What is meaning?', 'What is thought?' is the nature of the phenomena which constitute our world. These phenomena constitute the form of the world which we inhabit, and in asking these questions we express a desire to understand them more clearly. (p.18)

>When we try to answer these questions by uncovering or explaning something, we are led into a state of frustration and philosophical confusion. We think that the fault lies in our explanations and that we need to construct ever more subtle and surprising accouts.

>The real fault, W believes, is not in our explanations, but in the very idea that the puzzlement we feel can be removed by means of a discovery. (p.19)

>The method we really need is one that 'simply puts everything before us, and neither explains nor deduces anything. ----Since everything lies open to view there is nothing to explain'. (PI 126)

>Read the last paragraph of page 19.

>Read the middle paragraph of page 21.

>The confusions arising from the theoretical attitude are not, therefore, mere mistakes. They are misunderstandings which, when we become reflective about it, language itself has the power to draw us into.

>Read the last quotation on page 22.


>W's preferred description of the above process is 'therapy' (PI 133) or the 'treatment of an illness' (PI 254). This description is apt for a number of reasons.

>First of all, it conveys the idea that our concern with the construction of elucidations or models is in some way itself an obstacle to our progress, something that holds or arrests us and prevents us from moving on. (p.23)

>Secondly, it captures the fact that W's method is aimed, not at producing new, stateable conclusions, but at working on us in such a way as to change our whole style of thinking or way of approaching problems. (p.23)

>Finally, the concept of therapy recognizes that this process is in its nature protracted. Therapy is essentially a slow process. (p.23)

>Read the first paragraph on page 24 from the 6th line of the paragraph.

>W adopts two different approaches in response to our sense that these pictures somehow give us the essence of a phenomenon. (p.24)

>The first response that he adopts is to attempt, by a number of characteristic techniques, to reveal the emptiness of the explanatory accounts that we're inclined to offer, or of the pictures and models that we construct. We think that we are putting forward clear, explanatory models of what a given phenomenon consists in that will satisfy our urge to understand. However, when W calls on us to examine these models more carefully, or to connect them with what actually goes on in our practice of employing language, we find that they crumble to dust in our hands. We come to see that the pictures and models that initially seem to us so straightforward and explanatory acutually make no connection with the phenomena they are designed to illuminate. (pp.24-25)

>The second response that W makes is directed at evoking an appreciation of the neglected details of concrete examples of the 'spatial and temporal phenomenon of language'. It is through the appreciation of the differences in the way the different regions of our language function that he believes we can escape from philosophical confusion, and find that 'the satisfaction we are trying to get from an explanation comes of itself' (RFGB, p. 30). For our puzzlement concerned the nature or essence of particular phenomena ---- meaning, thought, sensations ---- and this puzzlement is removed, 'not by giving new information, but by arranging what we have always known' (PI 109)

>W's grammatical enquiry aims to produce a kind of understanding which consists in seeing a pattern or form in what is there before our eyes, but which we had previously neglected or overlooked. It is through an emerging sense of this form that the essence of language, meaning, understanding, etc. is gradually revealed and understood. We gradually come to see that 7nothing out of the ordinary is involved' (PI 94), that no further (deeper) explanation is needed, that the essence 'lies open to view' (PI 126). (pp. 26-27)


>One might, however, wonder why W is so against the idea of a systematic description of the grammar of our language.

>The answer to this question lies, at least in part, in the essential responsiveness of W's philosophical method. Thus, the anti-systematic nature of W's philosophy is linked with the idea that a grammatical investigation is one that 'gets its light, that is to say its purpose, from the philosophical problems' (PI 109). The self-conscious awareness of the forms of our human practice of employing language that a grammatical investigation results in does not represent an increase in our knowledge of the kind we associate with science; it merely 'reminds us' of something which, insofar as we are masters of the practice of using language, we already know. Bringing this knowledge to self-consciousness does not increase or improve our mastery of this practice, but it gives us a kind of understanding which 'consits in "seeing connexions"', and which frees us from both false pictures and the inappropriate urge to explain the phenomena that puzzle us. (pp. 27-28)

>Read the first paragraph on page 28.

>Thus we will miss the whole point of W's philosophical method if we attempt to extract from his remarks a series of philosophical claims about what constitutes meaning, understanding, sensations, and so on. (p. 28)

>We must, therefore, resist the attempt to sum up, or to state philosophically exciting conclusions, and allow instead for a series of clarifications to take place in which the philosophical problem . . . completely disappears' (PI 133).

>Read the last paragraph of Chapter 1 on pages 30-31.


>Read the first paragraph of this chapter, which is also a brief summary of Tractatus.

>The appeal of the theory in Tractatus lies, not only in its apparent explanation of language's representational capacities, but in its providing a solution to a number of philosophical problems about the nature of the proposition that had been thrown up by the work of Russell and Frege. (p. 34)

>Read the paragraph which begins at the bottom of page 34. (cf. name/object - propositon/state of affairs)

>In the picture theory of the proposition, which he puts forward in the Tractatus, we see an example of the style of thought, and of the temptations associated with it, that are the real target of his remarks.


>Read the quotation from Augustine's Confessions on page 36.

>By using a passage from Augustine's Confessions, written in the fourth century AD, W. brings out the universality of the temptations he is concerned with.

>Read the quotation from W.'s on 37.

>We can see Augustine's tendency to think of the4 human subject in terms of a private essence or mind ---- in which there are wishes, thoughts, desires, etc. ---- and a physical interface with the outside world. The private essence is conceived as somehow already fully human, but as lacking the capacity to communicate with others. (p.38)

>Read W.'s language game (PI 1) on page 39.

>Although this example presents a simple language, or a simple use of language, it does not involve the sort of oversimplification that we find in Augustine. (p.39)

>First, it presents this simple language in its natural environment, when it is acutually functioning, and not as a system of words or sentences abstracted from use. (p.39)

>Secondly, W. does not use this particular concrete case as the basis for deriving any claim about the essence of language, but simply as a means to draw our attention to the natural embedding of language withing the active lives of speakers, and to a richness and complexity in the phenomenon of language. (p.40)

>W. asks: But what is the meaning of the word 'five'? ---- No such thing was in question here, only how the word 'five' is used. (PI 1)

>Thus, W.'s concrete example begins to work against Augustine's temptations both to think about language in abstraction from its use and to look for the essence of meaning. (p.40)

>When language is functioning, it is clear that different expressions play quite different roles. W. invents this exapmple of a simple language, in which the individual words have clear, distinct techniques associated with them, in order to bring out, on the one hand, how language is interwoven with non-linguisitic activity, and on the other, how it is in use that the different functions of expressions become apparent. In this way, a concrete example is made to work against our philosophical temptations, without itself being used as a source of philosophical doctrines. (p. 40)


>We are asked to imagine a language consisting of the words 'block', 'pillar', 'slab', and 'beam', whose function, like that of 'table', 'chair', 'bread', etc., is to pick out a particular sort of object.

>W. now uses this example to explore the picture of language acquisition that Augustine presents us with, by asking us to imagine in detail how a child would be taught this language, and how we would judge that he had succeeded in mastering it.

>The teacher's initial acts of pointing at a shape and saying the appropriate sound cannot be taken as an ostensive definition of a word, since the child is not yet in a position to understand what the adult does as defining a name, for the child is not yet master of the technique of naming, he 'cannnot as yet ask what the name is'. Thus W. calls the process we'&ve just described 'the ostensive teaching of words' (PI 6)

>Does the fact that word 'slab' prompts the pupil to form an image of a slab mean that he has understood the word, or mastered the language? To answer this question, W. suggests, we need to ask what the purpose of the word 'slab' is in the language of this tribe.

>Again, we find that W's response to Augustine here is concessive rather than confrontational. He does not claim that it couldn't be the case that the purpose of a language is to evoke images in the mind of a hearer.

>However, this is not the purpose of the use of the language of the imaginary tribe of builders.

>We have been thinking of what is accomplished by the ostensive teaching of words purely in terms of what goes on in the child's mind, and have neglected the way in which the training with the language is embedded in an overall training in the tribve's practice of building. When we focus on the embedding of linguisitic training in the wider context of learing to build, we see how this neglected aspect of the process is actually paramout to our ordinary idea of what constitutes understanding: 'Don't you understand the call "Slab!" if you act upon it in such-and-such a way?' (PI 6)

>W's aim is to show us that in this act of abstraction we turn our backs oneverything that is essential to the actual functioning of language; it is our act of abstracting language from its employment within our ordinary lives that turns it into something dead, whose ability to represent now cries out for explanation.

>W. hopes gradually to bring us to see that 'nothing out of the ordinary is involved' (PI 94), that everything that we need to understand the essence of language 'already lies open to view' (PI 126)

>Can the words 'there' and 'this' be taught ostensively? If the child takes the words 'there' and 'this7 be taught ostensively? If the child takes 'there' and 'this' to describe tthe place or the thing pointed to, then it will not have understood the function of these words. The oversimplified picture of meaning that Augustine adopts invites the idea that ostensive teaching provides a model for the process of language acquisition. yet once we become aware of the variety of linguisitc techniques that exist even in this very simple language-game, then we begin to see that Augustine has taken one striking case as a model for all, and that in fact training with the language is as varied as the techineques that constitute it.

>When we ask the question 'What do the words of this language signify?', we have a particular idea of the form of answer that is required, one that is connected with the idea of pointing to what is signified by a word. But we have seen that it is the use of a word that shows its significance, and not an object that can be pointed to. Yet our urge for generalization may sitll prompt us to look for some canonical form for specifying the meaning of expressions; the canonical form, just by being generally applicable, may sitll seem to us to capture something essential about meaning.

>W. does not suggest that no such form could be found.

>W. shows how little is thereby accom-plished.

>Firstly, although the sentence '"slab" signifies this object' may distinguish the shape of building block that 'slab', as opposed to 'block', refers to, it does not indicate anything about how the word 'slab' is actually used. The technique for using the word is simply presupposed in this canonical description. Thus, the canonical description does not bring us any closer to the essence of meaning, for the entire framework of employing names within a language is not described but presupposed.

>Secondly, the construction of a canonical form for specifiying the meaning of expressions only serves to make fundamentally diffeent types of expression look more similar than they acutally are.

>What makes a tool a tool is simply that it is used as a tool, each kind of tool in its own specific way. The aspects of language that are brought to the fore by the tool analogy are ones that escape our notice when we do philosophy precisely because of our tendency to think of words, not as they are applied and used, but as they are written or spoken, aside form their actual use or application.

>We might, W. says, group words into kinds for some particular purpose, with some specific aim in mind, but, by implication, none of these classifications should be thought of as uncovering the intrinsic structure of languge.

>W.'s interest in these particular, concrete cases is exhausted by their role in overcoming specific philosophical confusions and in countering our urge to generalize.

>We might feel that the language-game of PI 2, and the extended version of it, cannot possibly be complete language because they consist only of orders. W. responds to this by asking whether our language is complete.

>W. has long ago abandoned both the Tractatus theory of representation and the theoretical style of thought that it exemplifies. He has offered us instead a picture of language as a set of instruments or techniques that are employed by speakers in the course of their everyday lives. This picture is not presented in the form of a new theory of representation, but as a way of looking at language ---- 'an object of comparison' (PI 131) ---- which counters our temptation to think of language in abstraction from its use.

>There is no essential structure or function against which the notion of completeness can be defined; it makes no more sense ot speak of a complete language than to speak of a complete tool-kit.

>It makes no sense to speak of language as either complete or incomplete, for language represents a form of limit to those who speak it; it represents the point from which we judge. Our language is not superior to the language-games W. describes in the sense of being closer to some ideal, or complete, symbolism; it is simply richer and more complex. The concept of incompleteness, like the concept of completeness, belongs with the false idea of an absolutely correct or essential system of representation.

>It is hrere that W. introduces the notion of a form of life: 'To imagine a language means to imagine a form of life' (PI 19). The idea of languge as a form of life, like the idea of a language-game, is to be set over against the idea of language as an abstract system of signs; it again serves to bring into prominence the fact that language is embedded within a horizon of significant, non-linguisitic behaviour.

>The idea of a form of life applies rather to historical groups of individuals who are bound together into a community by a shared set of complex, language-involving practices. These practices are grounded in biological needs and capacities, but insofar as these are mediated and transformed by a set of intricate, historically-specific language-games, our human form of life is fundamentally-specific language-games, our human form of life is fundamentally cultural (rather than bilogical) in nature.

>W.'s idea that the concept of language is internally linked with that of a form of life suggests that the adult human subject emerges slowly, as its life becomes structured through the acquisition of new and more complex language-games. In acquiring a language, the child comes to inhabit a social world of practices, the structure of which is grounded, not merely in concepts, but in ways of acting and responding that essentially involve the use of language.


>The temptation to appeal to an intrinsic, inner structure ---- to the thought that is really expressed by 'Slab' ---- is very strong. For we want to say that what the speaker of the language of PI 2 really means when he calls 'Slab!' is: 'Bring me a slab'.

>This thought certainly contains echoes of the Tractatus, where the logical form of a proposition is something that must be revealed through analysis, and where analysis is thought of as uncovering the intrinsic logical form of the propositon (thought) that a sentence expresses. But more generally, the thought above reveals a temptation to think of the mind as possessing an inner structure, which may be only loosely connected with the manifest forms of human behaviour, and which constitutes the essence of psychological states like meaning and intending.

>The interlocutor responds: 'But when I call "Slab!", then what I want is, that he should bring me a slab!' W. replies: 'Certainly, but does "wanting this" consist in thinking in some form or other a different sentence from the one you utter?' (PI 19)

>This question touches on our temptation to picture psychological phenomena (such as understanding or mastery) in terms of a determinate, inner state that exists, as it were complete, inside the subjects mind (or brain). Thus we imagine the linguistic mastery that grounds my meaning 'Bring me a slab' as four words in terms of an inner structure that lies behind my uttrance of this sentence. For it must, we feel, be the case that my meaning the sentence as four words rather than one consists in something that obtains at the time of my saying these words. This picture of the mind as an internal mechanism, whose inner structure underlies and explains the structure of outward behaviour, is immensely compelling, but it stands in fundemental opposition to the idea that everything we need to understand psychological phenomena 'already lies open to view' (PI 92). It is this picture of the mind as an internal mechanism, or repository of psychological states, that forms one of the major targets of the Investigations, and W. introduces his oppositon to it as follows: I have admitted that the foreigner [who thinks that 'Bring me a slab' is one word rather than four] will probably pronounce a sentence differently if conceives it differently; but what we call his wrong conception need not lie in anything that accompanies the utterance of the command. (PI 20)

>W. is here directing us to look for what grounds the structure of an utterance of the sentence 'Bring me a slab', not in what accompanies the saying of the sentence, but in what, as it were, surronds it. Thus he introduces the idea that it is nothing that occurs at the time, or in the speaker's mind, that determines that he means it as four words rather than one. We can imagine, for example, that exactly the same thing goes on in the mind of the foreigner who means the sentence as one world and the native speaker who means it as foru. To understand the difference between thenative speaker and the foreigner we should look at the background to the utterance of these words, that is, at something which does not accompany their utterance, but rather forms its context or horizon.

>The structure and complexity of psychological states does not lie in the strucute and complexity of some inner mechanism (mental or physical), but in the complex (temporally extended) form of life that is apparent in an individual's behaviour and accessible from a third-person point of view.

>The interlocutor then raises the following question: ' "You grant that the shortend and unshortened sentence have the same sense. ---- What is this sense then? Isn't there a verbal expression for this sense?"' (PI 20). W. responds by directing us away from a concern with an ideal expression of this sense towards the shared use that forms the background to a particular utterance of each of these sentences. Thus it is the fact that these sentences share a use, or play the same role, within a wider languge-game that constitutes their having the same sense. Instead of looking for an idealized expression of this sense, we are directed towards the function that these sentences have within the practice of using language which forms the framework for the use of these sentences on particular occasions.

>The emphasis on formal, grammatical differences between sentences that our way of thinking creates has led philosophers to suppose that there are three basic types of sentence ---- assertions, questions, commands ---- corresponding to the three distinct grammatical forms.

>When we look at how sentences are used ---- at the distinct language-games we play with the sentences of our language ---- then we are faced not with three types, but with a countless number: There are countless kinds [of sentences]: countless different kinds of use of what we call "symbols", "words", "sentences". And this multiplicity is not something fixed, given once and for all; but new types of language, new language-games, as we may say, come into existence, and others bocome obsolete and get forgotten. (PI 23)

>Learning our language, or coming to participate in our form of life, is essentially connected with acquiring mastery of countless kinds of language-game.

>Learning our language means becoming accuturated, that is, coming to participate in a vast network of structured activities that essentially employ language. This rich conception of what is involved in the acquisition of languge contrasts sharply with Augustine's impovrished idea of learning language, which his conception of language as a system of meaning signs makes almost inevitable. Approaching language in abstraction from its use leads us to neglect, or misunderstand, the rich diversity of language-games that we come to participate in as we acquire a mastery of our language. This diversity in how we use language is regarded by the philosopher as something incidental to its essence; there is no acknowledgement that the structure and function of language are inextricably linked with the structure and function of the complex activities in which its use is embedded.

>The philosoper's concern with the construction of an accunt of the essence of meaning, of assertion, of querstions, etc. does not only 'send us in pursuit of chimeras' (PI 94), but it ignores the real distinctions and complexities that are revealed only when we look at language when it is functionning, within our day-to-day practice of using it.


>In PI 27, W. W. writes: '"We name things and then we can talk about them: can refer to them in talk."' This sentence expresses our inclination to think of naming in very simple, unproblematic terms. It's easy: we just name things, and then we can use the name to talk about them.

>If we think of defining the words 'NN', 'three', 'table', 'red', 'square', and so on, by pointing to an appropriate object and repeating the name, then it is clear that what counts as 'going on to talk about' these things ---- the technique of using the name ---- will be very different in each case.

>Read the quotation from PI 28 on page 62.

>It is gradually becoming clear that we have no very clear idea of what is involved in teaching someone something's name.

>But how do we know whether an ostensive definition has succeeded or not? It depends, W. suggests, 'on whether without it the other poeron takes the definition otherwise than I wish. And that will depend on the circumstances under which it is given, and on the person I give it to' (PI 29).

>But how do I know whether someone has taken a difinition 'as I wish'? W. goes on: 'And how he "takes" the definition is seen in the use that he makes of the word defined' (PI 29). It is not what happens when he hears the definition, but what happens afterwards ---- the use that he goes on to make of the word defined ---- that shows how the pupil takes the definition.

>It is when someone is already master of the practice of employing the techniques that constitute our language-games that he is in a position to understand an ostensive definition, or equally, to ask for somehing's name.

>Given this practical mastery of linguistic techniques, we might go into a foreing country and learn the language of its inhabitants by means of ostensive definitions.

>W. suggests that Augustine describes the child's leaning its first language as if it was a foreigner coming into a strange country.

>W. does not dispute that there is such a thing as concentrating on the colour rather than on the shape of an object. Someone might, for example, point to a vase and say, '"Look at that marvellous blue ---- the shape isn't the point."' (PI 33). Equally, someone might say, '"Look at that marvellous shape ---- the colour doesn't mater."' (PI 33). However, does what we do in each case constitute what it is to concentrate on (or point to, or mean) the colour rather than the shape? Read the bottom quotation on page 66.

>Here W. introduces an analogy with chess. We don't make a move in chess simply by moving a piece in such-and-such a way on the board. No more do we make a move simply because we have certain thoughts and feelings as we move the piece. Rather, moving a piece on a board constitutes a move in chess only 'in the circumstances that we call "playing a game of chess", "solving a problem in chess", and so on' (PI 33). Again W. is directing our attention away from what accompanies the move to the context or circumstances in which the act occurs; it is what happens before and afterwards, rather than anything that accompanies my act, that constitutes what I do as a move in chess. By analogy, it is what happens before and afterwards, rather than what happens while I attend to the colour of an object, that constitutes what I do as attending to (pointing at, meaning) the coloour rather than the shape.

>Read from the last paragraph of page 67 to the end of page 68.


>What is fundamentally at issue is the question of how we approach the problem of understanding the structure and function of language. Augustine approaches language as a system of signs whose capacity to represent requires elucidation in a theory that tells us what meaning consists in. It is not merely that W. believes that the theories which Augustine presents are wrong in their details, but that in the very first step of abstracting language from its application Augustine situates himself towards it in a way that makes it impossible for him to achieve the understanding he seek.

>In thinking about language as a system of signs, abstracted fromits use within the ordinary lives of speakers, Augustine turns away from 'the spatial and temporal phenomenon of languae' (PI 108) and towards an abstract phantasm. In this act of turning away from the practice of using language, Augustine inevitably loses sight of the way language actually functions, and is led into picturing language's aboilty to represent as 'some remarkable act of mind' (PI 38). Thus we are led to look for explanations of the phenomena of understanding, meaning, meaning one thing rather than another, and so on in the realm of the mind, or in what accompanies the actural use of words.

>The particular examples that W. uses to counter these temptations to peel language off from our form of life, and to connect it with what is occuring in the mind of speakers, are not intended to form the basis of an alternative account of what meaning (or naming, or understanding) is. Rather, these examples are used as a means to recognizing, on the one hand, how false the pictures of language that we've constructed are, and on the other, that the concepts of language and linguisitc mastery are not to be explicated by reference to hidden accompaniments to the use of words, but are essentially tied up with the idea of a distinctive pattern of behaviour or form of life.

>Any attempt to derive thoeretical elucidations from W.'s remarks is clearly at odds with the central idea that the value of the particular cases he describes lies in their ability to persuade us that 'everything lies open to view' (PI 126), and to overcome our sense that there is something wich needs to be explained.

>The whole point of W.'s opposition to Augustine's approach to language lies in his using particular cases to show that there is no need tospeculate about the hidden accompaniments to the use of language in order to understand how language functions; we have simply to look and see how it functions.


>We sometimes speak of the meaning of a word being grasped 'in a flash', or of understanding a word when we hear it. This might nturally lead us to picture meaning as something that can be grasped by the mind in an instant. The use that a speaker makes of a word is, by contrast, something extended in time, something that happens after he has heard the word and understood it.

>How does this criterion connect with the idea that meaning is something that can be grasped in an instant, or in a flash? For 'what we grasp [in a flash] is surely something different from the "use" which is extended in time!' (PI 138)

>Behind W.'s wide-ranging and elusive remarks on meaning, understanding and rules, Kripke detects an entirely novel form of sceptical argument that allegedly establishes that there is no fact, either in my mind or in my external behaviour, that constitutes my meaning something by the words I utter, or that fixes what will cont as a correct application of a rule that I grasp.


>Let us begin by presenting the sceptical argument that Kripke finds in Wittgenstein.

>It is very natural to suppose that in learning to add I grasped the rule for addition in such a way that my intention to follow this rule in the future will determine a unique answer to subsequent addition in such a way that myintention to follow this rule in the future will determine a unique answer to subsequent addition problems in indefinittely many new cases.

>In particular, it is natural to suppose that when I respond '125' to the question '68+57=?', I am acting in conformity with my previous intention to use '+' in accordance with the rule for addition. My previous intention to mean addition by '+' determines '125' as the correct answer, as the answer I should give, even though I have never explicityly considered this particular case before, and even though I have never explicityly considered this particular case before, and even though I have never before added any number greater than 56.

>What is there about my previous intention that rules out the possibility that I actually intended to use '+' in such a way that the correct answer to '68+57=?' is actually '5', so that in answering '125' I have actually changed what I mean by '+'?

>The sceptic can point out that I have only ever given myself a finite number of examples manifesting this fuction, which have all involved humbers less than 57, and that this finite nuber of examples is compatible with my meaning any one of an infinite nuber of functions by '+'. There is, for example, nothing to rule out that I intended to use 'plus' and '+' to denote a function which Kripke calls 'quus' and symbllizes by '(+)', and which is defined as follows:

x (+) Y = x + y, if x, y < 57

= 5 otherwise

If the sceptic is right, then there is no fact about my past intention, or about my past performance, that establishes, or constitutes, my meaning one function rather than another by '+'.

>The concept of meaning or intending one function rather than another has been shown to make no sense. Every possible answer is compatible with some possible function, and thus the idea of any answer being the correct answer becomes completely idle or empty.

>Kripke quickly scotches any idea that we can escape this sceptical problem by constructing an instruction that specifies, in terms of other rules, how the rule for addition is to be applied in new cases, e.g. by using the rule for counting to give an algorithm for addition. The problem is that whatever further rule I employ in giving myself an instruction about the rule for addition, there will always be a question as to how this further rule is itself to be applied.

>Nor is this problem restricted to the mathematical case. For any word in my language, we can come up with alternative interpretations of what I mean by it that are compatible with both my past usage and any explicit instruction that I might have given myself.

>Kripke calls what he believes to be W.'s way out of this paradox a 'sceptical solution' to it, and his point in so calling it is that he believes that W.'s account of rule-following begins by conceding to the sceptic that there is no fact about me that constitutes my meaning addition by 'plus', and which determines in advance what I should do to accord with this meaning. The intolerable paradox that this appears to create arise3s, Kripke suggests, only because we mistakenly insist upon construing the meaning of 'I mean addition by "+"' on the model of a truth-conditions conception of meaning, i.e. one that assumes that the meaning of a sentence is given by a condition that specifies what must be the case in order for it to be true. For if we assume that the meaning of a sentence is given by its truth-conditions, then it would follow from the sceptic's discovery that there is no fact about me which distinguishe4s my meaning something specific by a word from my meaning nothing at all, and that any sentence of the form 'A means . . . by "-" ' is at best false, and at worst meaningless. W.'s claim, according to Kripke, is that we can avoid this paradox, while accepting the sceptic's conclusion that there is no fact, inside or out, that constitutes my meaning something by a word, provided we adopt an assetability-condition model of meaning, i.e. one that assumes that the meaning of a sentence is given by the conditions under which it can be asserted.

>In the Investigations, Kripke claims, W holds that there are two aspects to a sentence's meaning. First of all, there must be conditions under which the sentence is appropriately asserted (or denied); secondly, this practice of asserting (and denying) the sentence must play a significant role in our everyday lives.

>What is needed, on this account, is that 'there be roughly specifiable circumstances under which they are legitimately assetable, and that the game of asserting them under such conditions has a role in our lives' (Kripke, 1982, p.78). The paradoxical, self-defeating consequences of the sceptic's discovery are avoided, on this account of meaning, insofar as 'no supposition that "facts correspond" to those assertions is needed' (p.78).

>The normative element ---- i.e. the distincition between a correct and incorrect use of a wrd or application of a rule ---- that is an essential part of our ordinary concept of meaning or rule-following only enters in when we consider the individual in relation to the wider community of sepakers.

>Any individual speaker will be deemed to have mastered the concept of addition, or to mean addition by '+', whenever his responses to particular addition problems agree with those of the members of a wider community in a sufficient number of cases.

>Kripke's interpretation of W.'s remarks on rule-following has been subject to a vast and wide-ranging critical response.

>Some commentators, however, have taken a more radical line against Kripke, by raising the question whether he is warranted in transforming W.'s wide-ranging remarks into a more or less systematic account of how meaning, or (on Kripke's interpretations) talk about meaning, is possible. In transforming W.'s remarks into the articulation of a clear philosophical problem and its solution, Kripke inevitably erases all trace of the idea of a grammatical investigation, which I have suggested is the key to how to read W.'s work.


>Let us return, then, to the issue that W. raises in PI 138: what is the connection between the act of grasping the meaning of a word (of hearing a word and understanding it) and the use that is subsequently made of it? How can what I grasp in an instant connect with a use that is extended in time?

>W. begins the investigation of the phenomenon of understanding by asking us to imagine the particular case of hearing and understanding the word 'cube' when someone says it to us. He asks us to reflect on what does actually come before my mind when I hear and understand the word 'cube' in this way. perhaps, he suggests, what comes before my mind is a schematic drawing of a cube. Now the question is how does the schematic drawing that comes before my mind connect with the use that I go on to make of the word 'cube'. The interlocutor replies: '"It's quite simple; ---- if that picuture occurs to me and I point to a triangular prism for instance, and say it is a cube, then this use doesn't fit the picture"' (PI 139). W. responds: 'Doesn't it fit? I have purposely so chosen the example that it is quite easy to imagine a method of projection accoding to which the picturedoes fit after all' (PI 139).

>I am inclined to think that the picture of a schematic cube that comes before my mind when I hear and understand the word 'cube' is such that it cannot be applied to anything but a cube. W. has purposely chosen the example, so that when we look carefully at this particular case we can quite easily see that this initial sense that the picture itself somehow imposes a particular use on us is quite empty. For when I reflect on matters, I see that it is quite easy to imagine another method of projecting the picture, e.g. one by which it fits a triangular prism.

>What is at issue here is not a sceptical argument about whether there is anything that constitutes the meaning of my words, but whether a particular picture I am tempted to form of the connection between what comes beofe my mind and the use I make of a word ---- the picture of what comes before my mind 'forcing' a particular use on me ---- has any real content.

>The same picture might come before our minds and the application still be different: the picture doesn't, after all, force a particular use on us. In the case where the applications are different, it is the use that is made of the word that shows what a speaker means by it, and not what comes before his mind when he hears the word and understands it.

>W.'s interlocutor does not, however, immediately abandon the idea of a picuture that forces a particular use on us. Couldn't it be the case, he asks, that, not merely the picture of the cube, but also the method of projecting it comes before my mind when I hear thw word and understand it?

>W. now asks: 'can there be a collision between a picture and application?' (PI 141). Clearly there can. But this is not because the picture forces a particular application on us; not because it cannot be applied differently from the way we apply it. The possibility of a collision arises simply inso far as 'the picture makes us expect a different use, because people in general apply this picture like this' (PI 141). W. goes on: 'I want to say: we have here a normal case, and abnormal cases.'

>The connection between a picture and its application, which gives the idea of a clash between picture and applicationg its content, is grounded in the practice of using the picture, in the fact of its having a use, and not in a mysterious power of the picture to force a particular use on us.

>The temptations to misunderstand arise because we are too hasty in constructing what initially seems to us to be a clear picture of hwo a part of our language functions. What W. shows us is that when we come to try to apply our picture to a particular concrete example, it proves unusable. In coming, through a grammatical investigation of particular cases, to a clearer perception of what is actually involved, we begin to recognize, not only the emptiness of our initial idea, but also that everything that we need to understand how this part of our language functions is already there in fromt of our eyes, observable in the surface structure of our ordinary practice. By comparing our philosophical picture with reality, we begin to see that our urge to explain the link between the act of grasping the meaning of a word and the use that is made of it has produced only an empty construction.


>The false picture of the connection between what occurs in a speaker's mind when he hears and understands a word and the use that he subsequently makes of it is linked with the following idea: that a speaker's understanding of his language is a form of mental state which is the source of his ability to go on and use the words of his language correctly. The thought that what comes before a speaker's mind when he hears and understands a word must somehow force a particular application on him is clearly linked with ---- is one way of giving expression to ---- this picture of what the state of understanding must consist in. Even if this particulr picture of meaning has proved empty, we may still feel that the concept of understanding requires that something like this must be the case. To deny it would, it seems, be to claim that our infinite ability to use language cannot be explained in terms of a finite, generative base; it would be to suppose that the difference, between someone who understands a language and someone who does not, does not lie in the presence or absence, respectively, of a finite internal state which is the source of the infinite capacity.

>Suppose we are teaching a pupil how to construct different series of numbers according to particular formation rules. When will we say that he has mastered a prticular series, say, the series of natural numbers? Clearly, he must e able to produce this series correctly: 'that is, as we do it' (PI 145). W. points here to a certain vagueness in our criteria for judging that he has mastered the system, in respect of both how often he must get it right and how far he must develop it. This vagueness is something that W. sees as a distinctive characteristic of our psychological language-game, one that distinguishes it from the language-game in which we describe mechanical systems.

>It may well be that, for exapmle, a formula's coming to mind is a criterion of a pupil's having understood a series, but only because this formula is used in a particular way, and not because it is somehow 'the source' of a given series. We have only to reflect that it is quite possible for the appropriate formula to come before a pupil's mind and yet for him still not to understand the series to see once again that our picture of 'logical compulsion' is empty.

>It is part of the distinctive grammar of concepts that describe conscious mental states ---- e.g. the concept of being in pain, of being depressed, of hearing a buzzing sound, etc. ---- that temporal concepts like duration, interruption and continuity, as well as concepts of intensity or degree, all make clear sense when applied to them. The concept of understanding, by contrast, is not linked with the idea of anything's 'occurring in our minds', but introduces an idea of a disposition or ability.

>What sort of state of understanding? The danger in calling it a mental state is that, instead of trying to describe its grammar by observing how the concept actually functions, we just form a vague picture of understanding as like a conscious mental state only unconscious. This does not make the grammar of the state of understanding any clearer (in fact, it could be said to obscure it), but at the same time it allows us to neglect the real work involved in making the grammar clear.

>In the absence of a clear view of the grammar of the concept of understanding, there is also a temptation to picture it in terms of the internal strcuture of a mental apparatus (the brain, say) which explains a speaker's ability to use the language. Again, Wittgenstein suggests that this picture does not fit the grammar of the concept, for our use of the concept shows that we are not describing the determinate state of an internal mechanism. Thus there is, for example, no criterion for a speaker's understanding a word, or knowing how to add, that invokes 'a knowledge of the apparatus, quite apart from what it does' (PI 149). Rather, our use of the concept of understanding is linked in complex ways with a speaker's participation in a characteristic form of life.

>It misrepresents Wittgenstein's thought here to say that the criteria for understanding are purely behavioural. This rendering of his thought does not capture the particular form of complexity that he detects in the way our concept of understanding functions. Zhe does not claim that the sense of the concept of understanding consists in the behavioural criteria for its application. It is rather that the field within which this concepts gets its sense (that is to say, its use) is the complex form of life that is revealed in the way speakers live and act, both in their past history and intheir current and their future ways of acting and responding.

>(On a series of numbers) What justifies me, on a particular occasion, in saying that I can go on? Wittgenstein writes: 'If there has to be anyghing 'behind the utterance of the formula' it is particular circumstances, which justify me in saying I can go on ---- When the formula occurs to me (PI 154). Thus I use the words 'Now I understand' with a particular background or history.

>Wittgenstein's directive ---- 'Try not to think of understanding as a mental process at all' (PI 154) ---- is not, therefore, to be understood as an assertion of an uncompromising behaviurism, or as an expression of anti-realism concerning mental states. It is rather to be construed as a remark about the grammar of the concept of understanding. Wittgestein intends his directive to warn us against adopting an oversimplified and inappropriate picture of the grammar of this concept, and to get us to open our eyes to the way in which this concept reverberates with the characteristic structures of our formm of life.

>Teaching a child to use the words 'Now I understand' will not turn on drawing his attention either to anything 'occuring in his mind' or to an empirical connection between his thinking of a formula and his going on correctly. Rather, the teacher responds with encouragement to the emerging patterns in the child's behaviour that signify increasing mastery. The child respons in turn with a growing sense of confidence and facitlty that prompts him to take the vital step of continuing independently. The use of the words 'Now I understand', 'Now I can go on', and so on are learnt in the context of the development of autonomous, confident responses that are in harmony with a particdular practice, and not in connection with either the introspection of internal menalstates or the hypothesizing of empirical regularities.

>W's remarks are purely descriptive of how our words are used, how they are taught, and so on. The whole purpose of his description, like the purpose of his grammatical enquiries in general, is to reveal that the picture we're inclined to construct in response to the question 'What is understanding?' is empty, and that everthing that we need to remove our puzzlement lies already open to view in our practice of using language. Thus the idea of an inner state, either conscious or unconscious, plays no role in the language-game that is played with the words 'Now I understand'.


>Suppose we get a pupil to continue a series (say +2) beyond 1000 ---- and he writes 1000, 1004, 1008, 1012.

>The issue here is not, as Kripke suggests, scepticism about meaning. Rather, W. is using the abnormal case as an object of comparison, in order to throw light on aspect of the normal case that we are tempted to misunderstand. In particular, the abnormal case is used to shed light on what is involved in my meaning or understanding the order '+2' in one way rather than another, response to the order the one that is correct. In bothe cases we are involved in a grammatical investigation, and not in providing a philosophical gloss on, or a justification of, our ordinary practice.

>We certainly didn't explicityly think of the steps 1000, 1002, 1004, 1006 when we gave the order; or if we did, there are clearly other steps that we didn't think of. And here we are pushed back on to the temptation to think of meaning or understanding as a state that already anticipates all the applications that can be made of a word. For we want to say; '"What I meant was, that he should write the next but one number after every number that he wrote; and from this all those propositions follow in turn"' (PI 186). It is, it seems, only by appeal to some such picture of an infinite number of propositions having already been determined that we can make sense of the idea that '1000, 1002, 1004, 1006 . . . ' is the correct, and '1000, 1004, 1008, 1012 . . . ' an incorrect, response to the order ' Add 2'.

>W. puts as follows: "But I already knew, at the time when I gave the order, that he ought to write 1002 after 1000." ----Certainly; and you can also say you meant it then; only you should not let yourself be misled by the grammar of the words "know" and "mean". (PI 187)

>We can say 'I already knew at the time ...', but these words connect, not with an act of meaning that mysteriously anticipates the future, but with our practice of using a rule and our trained responses to it. Thus my training in the practice of using the rule has brought it about that I now act, without hesitation, at each new point.

> 'I already knew at the time ...' does not mean that all my responses must in some mysterious way be already present or anticipated. Rather, these words receive their significance from the form of life that constitutes 'the scene of the language-game'; it is not what accompanies the saying of these words that gives them their sense, but the practice of using language in which their use is embedded; it is my practice that is invoked by these words, and not a mysterious mental act.

>In PI 201, W. writes: This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule. The answer was: if everything can be made out to accord with the rule, then it can also be made out to conflict with it. And so there would be neither accord nor conflict here.

>The paradox that W. here describes is not to be understood as a sceptical crisis that he has generated for the concept of a rule as such. Rather, the paradox refers to our initial sense of perplexity at the discovery that the idea of 'logical compulsion', which we want to use to describe the connection between a rule and its application, cannot be given any content.

>What our practice of teaching and folowing rules reveals is 'that there is a way of grasping a rule which is not an interpretation' (PI 201), that is, which does not depend upon the formulation and selection of hypotheses about how the rule is to be applied. This way of grasping the rule consisits simply in our responding to the rule in the way we have be3en trained, in accordance with the practice of using it. This unreflective, practical grasp of a rule 'is exhibited in what we call "obeying the rule" and "going against it" in actual cases' (PI 201), that is, in the ways of acting or responding to the rule that fall within our practice.

>In PI 202, W. strikingly links these grammataical observations concerning the concept of a rule with the question of whether there can be a private rule, or whether it is possible 'to obey a rule "privately"': And hence also 'obeying a rule' is a practice. And to think one is obeying a rule is not to obey a rule. Hence it is not possible to obey a rule 'privately': otherwise thinking one was obeying a rule would be the same thing as obeying it.

>It is not the psychological accompaniments to following a rule (thinking one is obeying a rule) that constitute one's act as an act of obeying a rule or going against it, but what surrounds one's act. Without this surrounding the concept of obeying a rule or going against it is completely empty; for it is nothing that occurs at the time, or 'in the mind', that gives the concept of a rule its content, but the existence of a particular form of practice. Yet to make sense of the notion of a private rule (of a rule that is not linked with a practice of using it), we would have only the psychological accompaniments to rule-follwing to appeal to, and these do not connect with the grammar of the concept of a rule.


>W. sums up his treatment of the idea of 'logical compulsion' in PI 218-21. He begins by asking, 'Whence comes the idea that the beginning of a series is a visible section of rails invisibly laid to infinity?' (PI 218). The image of rails is clearly one way of expressing our sense that all the steps have already been taken before we physically arrive at this or that point. Thus, 'infinitely long rails correspond to the unlimited application of a rule' (PI 218).

>W. asks: But if something of this sort really were the case, how would it help? (PI 219)

>Let us suppose that there really are rails laid down in a Platonic heaven corresponding to all the future applications of a rule. What difference would this make? Could we, for instance, consult these rails in order to determine how the rule is to be applied in a new case? Clearly not. These objectively existing rails can make no contact with our human practice of following rules, or of judging, in particular cases, whether a rule has been applied correctly.

>W. now goes on: No; my description only made sense if it was understood symbolically. ---- I should have said: This is how it strikes me. When I obey a rule, I do not choose. I obey it blindly. (PI 219); But what is the purpose of that symbolical proposition? It was supposed to bring into prominence the difference between being causally determined and being logically determined. (PI 220); My symbolical expression was really a myghological description of the use of a rule. (PI 221)

>Thus I use the picture of rails invisibly laid to infinity in order to give expression to my own sense that I no longer have a choice in how a word that I understand, or a rule that I've mastered, is to be applied in a new case. My response to the training that I have received is such that at each new step, e.g. in the development of a series, no doubt arrises; I simply react without hesitation, in accordance with the practice of employing the rule, and no other way of responding presents itself as a possibility. It is this characteristic experience of rule-follwing ---- this characteristic sense of inexorability in how the rule is to be applied ---- that we express 'symbolically' (i.e. metphorically) in the picture of infinitely long rails.

>Problems arise, however, when we're tempted to interpret this picture, not as a metaphor, but literally, so that the image of Platonic (i.e. real but non-material) rails stretching to infinity, marking out all the correct applications of a rule, comes to represent the essential difference between being causally determined and being logically determined.

>What makes this a case of 'logical' , rather than causal, compulsion (i.e. a case of following a rule) lies, not in the way in which I'm compelled ---- not in the rule itself somehow compelling me ---- but in the fact that my response to the rule is part of a particular language-game (or practice) of rule-following. The idea that there really are Platonic rails laid to infinity proves tobe no more than a philosophical illusion, thrown up by a combination of our ignorance of the grammar of the concept of 'a rule' and a temptation to put a false construction of the things that someone in the grip of a rule is inclined to say.

>Our concept of a language describes, not an abstract system of signs with meaning, but a particular form of life, namely, one that displays the characteristic regularities or patterns that constitute the following of rules. Central to the idea of the form of life that our concept of language picks out is that there exists a pattern or structure in the activity of using words which fixes what counts as applying the words of the language correctly or incorrectly.

>The agreement or harmony that W. suggests is essential to our concept of language is the agreement that constitutes the characteristic form of life that speaking a language (giving orders, making reports, and the rest) consists in. he characterizes this harmony or agreement as foloows: 'It is what human being say that is true or false; and they agree in the language they use. That is not agreementin opinions but in form of life' (PI 241).

>If language is to be a means of communication there must be agreement no only in definitions but also (queer as this may sound) in judgements'; that is to say, in our use of words.

>This brings us to the end of the interpretation of W.'s remarks on meaning and rule-following. It is, I hope, clear that it would be a mistake to suppose that the force of these remarks is to be summarized by means of the following claims: Meaning is use. Speaking a language is a practice. Understanding is not a mental state that is the source of correct use. And so on. The philosophical singnificance of W.'s remarks does not lie in these 'conclusions', which are on their own thin and obscure, and which are clearly not intended to serve as the basis for the construction of a more elaborate theoretical machinery. The real purpose of the journey that has been made lies, on the one hand, in the overcoming of certain deep-seated philosophical myths and inclinations, and on the other, in the reorientation of our style of thought towards what lies open to view in our practice of using language.

>It is this reorientation of our thought away from abstractions and generalizations towards a creful attention to what lies before our eyes, in the forms of our concrete practice of using language, that is the real lesson of W.'s remarks, and this is something that cannot be expressed in a generalization.


>In this picture we shall begin to look at how W. applies the philosophical approach we've seen him adopt in the philosophy of language to the philosophy of psychology.

>He does not address the familiar doctorines of traditional philosophy directly, but goes back to their roots in the first temptation to form false pictures of language or psychological phenomena, which the grammar of our concepts presents. Thus he wants us to step back from our concern with the construction of theorectical accounts of the nature of consciousness, or of the relation betweeen consciousness and the brain, and examine the steps by which we are led to approach the problem of understanding the nature of psychological processes in the way that we do. Why do we go on trying to explain the relationship between consciousness and brain-process? Why do we feel a need to prove that other people have minds? How do we come to feel these problems as problems, when these things do not normally come into consideration within our ordinary lives?

>We have already seen this technique at work in the investigation of the concept of understanding. The underlying force of W.'s investigation of this concept is to direct us away from a concern with speculating about what is occurring, as it were, inside the subject (either in his mind or in his brain) towards a concern with what the grammar of the concept (how the concept functions) reveals about the nature of the state it describes.

>When we looked carefully at the grammar of this concept ---- at how we use the words 'I understand', 'He understands', and so on ---- we find that it simply doesn't function in the way we are inclined to suppose, namely as a name of a process occurring inside the speaker when he hears a word and understands it.

>In the current chapter, I will concentrate on W.'s remarks on privacy and private language, which occur between PI243 and PI275. In the next chapter, on the inner and the outer, I will focus much more widely on W.'s extended discussion of hte relation between psychological concepts and behaviour.


>W.'s discussion of philosophical psychology begins with the remarks on the idea of a private language that have become familiar as 'W.'s private language argument'.

>The first thing we need to get clear about in interpreting W.'s remarks on private language is just what he has in mind when he speaks of a 'private language'. He defines such a language, at PI 243, as one in which 'the individual words . . . are to refer to what can only be known to the person speaking; to his immediate private sensations. So another person cannot understand the language.' The idea of a pviate language is introduced in explicit contrast ot our ordinary psychological language, and the question W. raises conerning it is whether we can imagine such a language.

>But what might make us think that we could imagine such a language? To help us answer this question, and thus to focus on the issue raised by W.'s remarks, I wan to look at a discussion of psychological language that occurs in William Jame's book The Principles of Psychology, a text which W. was well acquainted and to which he often refers.

>In the context of a discussion of sources of error in psychology, Willam James introduces the idea of an ideal psychological language that would constitite 'a special vocabulary for subjective facts'. James imagines this pure or ideal psychological language as one which would have no link with the objective world, but which would simply record or name the distinct subjective states that are revealed to a subject by means of pure acts of introspection. This idea of an ideal or pure language of subjectivity is connected in James, not with a commitment to Cartesian dualism, but with what he takes ot be the undeniable sense of 'the bare phenomenal fact' of consciousness, and thus with the thought that introspection is the method by which we discover the essence of the states and processes to which our psychological expressions refer. Thus, he believes that 'introspective observation is what we have to rely on first and foremost and always' in psychology, because it is by means of introspection that we 'discover states of consciousness' (1981, p.185)

>James does not explicitly raise the question whether the pure language of subjectivity, whose absence he laments, would be one that could be understood only by the person speaking. However, I believe that we can still use James's discussion to focus on the central issue which is raised by W.'s remarks on private language.

>This sense that we grasp what a sensation, a thought, an image, etc. is on the basis of an introspective knowledge of our own case is a central theme of W.'s remarks on the philosophy of psychology, and his remarks on private language can be seen as the beginning of his exploration of the way this picture influence our idea of how psychological concepts are defined.

>To reject the idea that introspection is essential to a grasp of what, for example, a sensation is seems like a rejection of the distinctively subjective nature of sensations, and a consequent blurring of the distincition between the psychological and the physical. It is not dualism as such, but simply the thought that, as Nagel puts it, ' for an organism to have conscious experience at all means . . . that there is something it is like to be that organism' (1979, p.166), which underlies our sense that it is thruough introspection, or looking inwards, that we achieve an understanding of what our psychological concepts refer to or describe.

>It is not that W. wishes to deny that introspection is possible, or that its results may be of interest to us, but only to show that introspection is not a means by which we discover what sensations, thoughts, images, and so on are; it is not a means to defining a psychological term: 'Introspection can never lead to a definition. It can only lead to a spychological statement about the introspector' (RPP 1, 212)


>It is a striking feature of W.'s discussion of the idea a private language that, having introduced it in PI 243, he immediately drops it and takes up a grammatical investigation of our ordinary sensation language. In PI 244, our initially vague and over-simplified picture of what is involved in giving names to sensations ---- 'There doesn't seem to be any problem here: don't we talk about sensations every day, and give them names?' ---- is replaced by a realistic account of how we use contexts in which the child has hurt himself to teach him, first of all, exclamations, and later, words, with which to express his pain: 'A child has hurt himself and he cries; and then adults talk to him and teach him exclamations and, later, sentences. They teach the child new pain-bahaviour.' In looking at how we teach a child the word 'pain', W. is already drawing our attention to the fact that we teach the use of the word without ever attempting to direct the child's attention inwards.

>The way that these words connect with what the child feels is through his learning to use them as a technique for giving expression to how he feels. No act of inner ostensive definition is required for the words 'I'm in pain' to connect with what is felt; the connection is secured by the grammar of the concept, by virture of the fact that it is used as a new means to express what is felt. The connection between 'pain' and what a subject feels is achieved, not by where the child looks when he learns the word, but by the grammar of the concept that he is being taught.

>If we reflect on our ordinary use of these words, then we shall see that I am never said to know of my pain ---- I just have pain ----and that people very often do know of another's pain. The attempt to do justice to the distinction between pain and crying in terms of a subject's unique access to his sensations leads to our saying something which, at any other time, we would regard as nonsense.

>The distinction that we have just tried to understand as a qualitative difference between kinds of object ---- private (accessible only to the subject) versus public (accessible3 to everyone) ---- W. now presents as a grammatical difference between two kinds of concept. The distinction that we want to make when we say that the pain is private is one which is to be understood in terms of the grammatical differences between the concept pain and for example, the concept crying. One important element in this grammatical difference lies in a distinctive asymmetry that characterizes the use of the former concept. Thus the first-person use of the concept pain diverges from the third-person use of the concept in a variety of ways. For example, the words 'I'm in pain' give expression to pain, whereas the words 'She's in pain' do not; I cannnot be said to learn of my pain, whereas others can; other people may doubt whether I'm really in pain or only pretending, whereas it makes no sense for me to doubt it; and so on.

>The distinction that we attempt to mark with the words 'pain is private' is not a perceived, qualitative difference between kinds of object, but a grammatical difference between sensation concepts and concepts of behaviour, which itself reveals that we have to do with two quite distinct kinds of thing. It is through the grammatical differences between kinds of concept that we distinguish the kinds of thing that our concepts describe: 'Essence is expressed by grammar' (PI 371)

>The words 'Sensations are private', or 'Only you can know if you had that intention', are, therefore, to be understood as grammatical remarks: 'The proposition "Sensations are private" is comparable to "One plays patience by oneself"' (PI 248)

>The expressions 'private', 'inner', 'hidden' are all attempts to capture this distincitive grammar, and might therefore be regarded as marking a boundary between our psychological language-game and the language-game of physical description. W. does not wish to deny the aptness of these pictures, but his overall aim is to get us to recognize that the distinction which they are intended to capture is, at bottom, a grammatical one.

>W.'s interlocutor says in PI 253: '"Another person can't have my pains." ' The interlocutor here presents a picture of pain as something that is owned and accessible only to the person who has it, as something that is owned and accessible only to the person who has it, as something that he identifies or knows simply by turning his attention inwards to what only he can perceive.

>'Another person can't have my pains' tries to draw a distincition between pains that are qualitatively similar (my pains are like his) but numerically distinct (my pains are distinct from his pains); in this way, it treats the grammar of the word 'pain' as on a par with the grammar of the word 'chair', for which we do understand the distinction between 'a similar chair' and 'the same chair'. When we look at how these concepts are actually used, however, we see that they function grammatically in quite different ways, for while we talk of people feeling pain, and of their feeling the same (i.e. similar) pains, we don't identify or count pains in the way that we identify and count chairs.

>We talk about pain, about my having the same pain as I had yesterday and about my having the same pain as you, but all of this talk depends for its sense upon a language-game which serves to fix the grammar of the word 'pain'; and this grammar simply lacks the structure that is needed for the interlocutor's words ---- '"Another can't have my pains"' ---- to make sense. There is nothing independent of this language game ---- nothing that is fixed simply by looking inwards ---- that can determine a criterion of identity for pains.


(updated on 13 April, 1999: to be continuted)

Go to the index page