In two recent papers, John W Cook presents a radical interpretation of some of the underlying ‘sub-texts' of key arguments in Wittgenstein's later philosophy. In both papers Wittgenstein's conception of the ‘person' is the point of concern. Cook makes some other revisionary remarks as well, for example, suggesting that logical necessity is the only modality Wittgenstein allows, even in the later works. My comments will be directed mostly to the question of the concept of a person.
In the first of these papers he introduces his take on Wittgenstein's view of the concept of a ‘person' in the course of a discussion of Wittgenstein's analysis of the alleged ‘problem of other minds' as an exemplar of method in philosophy.Footnote 1 The argument of the second paper turns on the interpretation of Wittgenstein's rather sketchy treatment of the uses of the first person pronoun, ‘I'.Footnote 2 I am not satisfied with either of Cook's discussions, though they do help us to get a clearer understanding of how Wittgenstein understood the notion of a person.
Persons in the Context of the Private Language Argument
In commenting on the way the notion of ‘person' appears in the Private Language Argument Cook's remarks turn on two main ways that he believes Wittgenstein's practice deviates from his declared method:
1. He failed to ditch the underlying distinction which, according to his lights, ought to have been seen as the source of the ‘other minds' impasse as an intractable philosophical problem. By implicitly sticking to the Material Body/Immaterial Mind dichotomy Wittgenstein's demolition of the latter leaves him with only the former as the basis of an account of the meaning of psychological concepts. ‘So [says Cook] philosophers find themselves having to choose [concepts with which to categorise a human being] between a brick and a spook, and this should strike us as peculiar'.Footnote 3 According to Cook this is the underlying source of the implicit behaviourism which many have claimed to detect in Wittgenstein's writings on psychological topics.
2. Cook offers pretty convincing evidence for the view that Wittgenstein's claims about which statements using the words ‘know' and ‘believe' make sense and which are meaningless are not grounded in an examination of how these concepts are actually used by lay persons.Footnote 4 The therapeutic move of assembling reminders of such uses to highlight grammatical illusions is not invoked in that section of the Philosophical Investigations. Instead, Cook argues, Wittgenstein seems to have picked up for comment the way some of the philosophers of his acquaintance used these words, notably Moore and Broad.
However, I believe that the first part of Cook's argument is not convincing. It seems to me that Wittgenstein worked up a detailed taxonomy of material beings, distinguishing persons from various other kinds of things, such as stones and dolls. He did ditch both poles of the apparent dichotomy, as I shall argue.
In the second part of the argument Cook makes a good case for tracing the origin of Wittgenstein's claims about the meaninglessness of certain uses of the words ‘know' and ‘believe' to the discursive practices of philosophers. Be that as it may, I want to try to show that in his discussion of the ‘other minds' problem Wittgenstein remained true to some of his other and deeper methodological commitments. His diagnosis of the sources of scepticism concerning what someone can know of the subjectivity of other people as rooted in implicit adherence to false grammatical analogies goes deeper than misunderstandings of the grammar of first and third person uses of words like ‘know' and ‘believe'. He makes use of the descriptive/expressive distinction as the main diagnostic tool for picking up the relevant misleading grammatical analogies that had underlain centuries of futile debate. Wittgenstein's points about the uses of ‘know' and ‘believe' derive from the distinction between descriptive and expressive grammars. What is more Wittgenstein's account of first person discourse depends on another deep distinction that Cook overlooks, namely that between denotational and indexical reference.
Abandoning a Tempting Distinction
Contrary to Cook's argument I want to show how Wittgenstein made use of the concept of a person as an animated body, neither just a brick nor merely a spook, to probe the way that publicly meaningful ascriptions of psychological states to something are possible. The discussion in the Philosophical Investigations immediately following the exposition of the Private Language ArgumentFootnote 5 strongly suggests that Wittgenstein was quite clear that the traditional Cartesian distinction between what is ‘material' and what is ‘mental' should be abandoned in favour of a spectrum of material exemplars, from stones to people.Footnote 6
From birth we are presented with material beings. Do we make something like a scientific discovery that the evidence is overwhelming that some of these beings have thoughts and feelings while others do not? Wittgenstein seems to me to be quite consistent in holding that taking some things to be people and so blessed with subjectivity and others to be stones is a matter of the attitude we take to them. It is an attitude we cannot, as human beings, fail to take. We do not discover the subjectivity of other people. Our human lives hinge on it. I have no doubt that Cook thinks so too. My argument with him concerns whether Wittgenstein was one of us.
Most people, I dare say, manage their worlds with a variety of cases along a continuous transition from an embodied person, say as exemplified by the living Elvis Presley, through a video of him performing at a concert, a life like but material representation of a personFootnote 7 [just behaviour], to the lunar formation (‘Elvis found on the Moon') well known to readers of the National Examiner, to a stone found on the beach without the slightest human characteristics. It seems to me that something very like such a continuous dimension between radically disjoint poles is implicit in several remarks in this section of the Investigations.
Here are some of the remarks that support this suggestion.
PI § 281.Footnote 8 ‘But doesn't what you say come to this: that there is no pain, for example, without pain-behaviour?–It comes to this: only of a living human being and what resembles (behaves like) a living human being, can one say: it sees; is blind; hears; is deaf; is conscious or unconscious.'
This is not because of any ingress into the ‘soul' of that living being. Nor does it imply that it is only by its behaviour that I know someone is a human being. I credit chimps with some measure of subjectivity because they behave like we do. In the human case there are natural expressions and we are endowed with the ability to produce them and to read them. Autistic children lack this ability. Autism is now treated as a natural, that is genetically determined, defect. This of course shows that the intact ability to share our emotional lives with others is also natural. It is not that autistic people fail to make inferences that the rest of us make. Their attitude to others is different.
PI § 302. ‘Pain-behaviour can point to a painful place–but the subject of pain is the person who gives it expression. A stone cannot ‘give pain an expression'.'
Wittgenstein does not treat all material beings as brick-like. There are stones and there are flies and at the end of that slide, there are people. We don't need to make hazardous inferences as to their inner lives. Understanding how others feel is a natural endowment; an aspect of the human form of life, a hinge on which our uses of psychological concepts turn.
PI § 284. ‘Look at a stone and imagine it having sensations … And now look at a wriggling fly and at once these difficulties vanish and pain seems to get a footing here, where before everything was, so to speak, too smooth for it.
And so, too, a corpse seems to us quite inaccessible to pain. – Our attitude to what is alive and to what is dead is not the same. All our reactions are different.'
Whatever a corpse does, say emitting a sound like a sigh, is not readable by another human being as a natural expression of anything. Whatever it is, it is not a sigh. It seems to me, contra Cook, that Wittgenstein works with exactly the right distinction between stones and people.
The paragraph about ascribing feelings to dolls shows clearly that Wittgenstein took our capacity to understand how others are feeling to be a matter of the attitude we take to them rather than to any ‘irresponsible' inferences from behavioural evidence to the existence of the unobservable mental states of others.
PI § 282. ‘We do indeed say of an inanimate thing that it is in pain: when playing with dolls for example. But this use of the concept of pain is secondary one. [To show how pain talk is established in living human contexts] [i]magine a case in which people ascribed pain only to inanimate things; pitied only dolls. [Then it would not be the same language game].’
Truth and Sincerity
In commenting on Wittgenstein's remarks about the meaningfulness of assertions using the words ‘belief' and ‘knowledge' Cook claims, no doubt rightly, that Wittgenstein is not reporting the linguistic practices of real or imaginary tribes, but drawing on what Moore and Broad had said to him, qua philosophers. However, the remarks about the meaningfulness or not of claims to know and/or believe should be seen in the light of a deeper distinction that runs through the whole of Wittgenstein's writings on psychological concepts. This is the distinction between uses of words to describe some state of affairs and their uses to express an attitude, a feeling and so on. The former are assessed by means of epistemological concepts and criteria, such as truth and falsity, while the latter fall under moral concepts and criteria, such as sincerity and insincerity. Typically Wittgenstein takes our first person uses of psychological words to be expressive, that is avowals, sincere or insincere, honest or prevaricating, while third person uses of the same words are putative descriptions and so true, accurate, vague and so on. While the question of evidence arises for descriptions no such question could arise for avowals. Only if we take the grammar of first person avowals to be similar to that of third person descriptions does the possibility of meaningful scepticism about the thoughts and feelings of other people get any purchase. Distinguishing the grammar of first and third person uses of psychological words is the move that allows the Private Language Argument to get a grip on the ‘other minds problem'. The intelligibility of first person avowals stems from their origin as substitutes for natural expressions. It would not be pain if one had a tendency to smile and laugh when suffering it. Thus the groaning and the personal experience of being in pain are inseparable parts of the one psychological complex. The utterance of the words ‘I'm in pain' inherits the grammar of groaning.
In PI § 317 Wittgenstein says apropos of the distinction between expressive and descriptive or propositional uses of language:
‘Misleading parallel: the expression of pain is a cry – the expression of a thought is a proposition. As if the purpose of the proposition were to convey to a person how it is with another; occurring, so to speak, in his thinking part and not in his stomach.'
Closely allied to this point is Wittgenstein's treatment of the apparent paradox of my knowing the meaning of a sensation word but not being sure whether the sensation I am having is this one. If one fails to pay attention to the grammar of pain-talk one might slip into thinking it was descriptive, opening up the possibility that one might mistake one's own feelings. That one does not routinely question the correctness of an expression of a feeling is another step in undermining the grammar of description as a model for the use of psychological words.
PI §288 ‘[However] the expression of doubt has no place in the language-game; but if we cut out human behaviour, which is the expression of sensation,Footnote 9 it looks as of it I might legitimately begin to doubt afresh.'
This temptation arises only of ‘I assume the abrogation of the normal language game with the expression of a sensation'. So – don't abrogate it!!
To disrupt the psychological complex by pretending to be in pain is a moral fault. Groaning to gain sympathy is exploitation, not an odd way of uttering a false proposition.
It is a mortal sin in philosophy to mistake the grammar of expression for that of description. The distinction cannot be by-passed since it is the expressive use of psychological words that make it possible for sensation words to have a place in the language. Taking that use to be descriptive is what leads directly to the impasse of the ‘problem of other minds'. Sensation words could not be learned nor could their uses be tested empirically as if they stood outside the experiences they express.
This is what underpins the variety of uses of ‘believe' that Cook comments on. Thus ‘I believe in God' is an avowal and the question is only of my sincerity in holding that there is a God. ‘I believe that I have a broken tooth' is an expression of an opinion. The related proposition ‘My tooth is broken' typically comes up for judgement as to its truth or falsity. The role of the Pope apropos of God's existence is quite different from the role of my dentist apropos of the state of my tooth. Since ‘know' demands the truth of what one is said to know, the word is doing no work at all in avowals, except as Wittgenstein points out, as a mark of emphasis. ‘Believe' and ‘know' are not appropriate tools for commenting on expressive uses of words, nor as determining grammatical frames in which to insert them. Hence the meaninglessness of knowledge or belief utterances meant descriptively in contexts in which the grammar is expressive.
The consequences of trying to impose the grammar of descriptive talk on to contexts in which the uses of words can only be expressive are even more deeply subversive of meaning. In PI § 293 Wittgenstein points out that ‘ … if we construe the grammar of the expression of sensation on the model of “object and designation” the object drops out of consideration as irrelevant'. So construed it could play no role in the establishment of meanings nor in assessing an expressive act for sincerity or insincerity. But its relevance is of supreme importance in such commonplace matters as distinguishing between malingering and genuinely being ill. That the ‘object drops out of consideration' if we adopt the descriptive grammar is the best possible reason for refusing to construe sensation talk descriptively.
Indexicality of the First Person Singular
In his second paper John Cook argues that Wittgenstein was committed to a simple materialist account of the human being as person, and that this was the core of the sub-text on which the intelligibility and force of such clarificatory discussions as the Private Language Argument depends. Just as Cook's radical interpretations of the Wittgenstein texts in his first paper overlook the distinction between descriptive and expressive uses of language, so in the second paper they turn on a failure to see that the sub-text of Wittgenstein's discussions is the distinction between indexical and anaphoric uses of the first person, particularly pronouns. Cook says, correctly, that the word ‘I' does not designate a person.Footnote 10 It does not follow that it has no referential function at all. But that function must not be construed on the model of name and object. There is another equally important way in which such expressions as ‘I', ‘here' and ‘now' have a referential function, that is as indexicals. In PI § 410 Wittgenstein makes the point as follows:
‘“I” is not the name of person, nor “here” of a place, and “this” is not a name. But they are connected to names. Names are explained by means of them. It is also true that it is characteristic of physics not to use these words.’
The last sentence in this quotation would also be an interesting topic for reflection.
In those languages in which explicit pronoun use is mandatory, such as English and French, the first person singular is surely not the name of anything, neither of the whole person nor of the Cartesian soul. Its function is not referential like that of a proper name. It is indexical.Footnote 11 ‘I' indexes the content of the utterance it introduces with the spatiotemporal location of the speaker. The spatiotemporal indexicality of the first person in ‘I heard a nightingale' displays the speaker's claim to have been then and there in the forest. In other words an indexical completes the sense of the utterance by the listener's extra linguistic knowledge as to who is speaking. ‘He', typically anaphoric, in ‘Gordon went into the forest and he heard the nightingale' can be understood without local knowledge of who is the speaker. The reference is determined intralinguistically via a proper name or a definite description. In so far as the speaker's body is the spatiotemporal locus of his or her speaking, the use of ‘I' is tied to embodied beings. In many languages, though not in all, the first person singular also indexes the speaker's utterance with his or her personal commitment thus elaborating the performative force of the speech act. For example ‘I will do the dishes' commits only the speaker. Hence to know who is committed by that speech act requires knowledge of the actual occasion of the utterance.
Wittgenstein's comparison between the uses of the first person and exemplary indexicals such as ‘here' and ‘this' leaves no doubt as to his understating of the role of ‘I'. However indexicality depends on the beings in question being speakers. Cook's claim for the thing-like nature of those beings we, in our quotidian way, took to be persons, extends to a claim that talking should be interpreted as making sounds which influence the behaviour of other people, presumably causally. Cook takes this to be a consequence of the ‘use' conception of meaning. However, the efficacy of speech acts depends on the interlocutor, the builder's assistant, understanding what the builder wants him to do, and that the builder's cry of ‘Slab!' is an order. Once, when I was explaining the Block/Slab game to a person of an extreme radical persuasion, she became indignant at the way the builder was treating the assistant and declared that he should have gone on strike. Even in this world in which the four words are the only vocabulary a complex form of life is possible.
Cook also claims that Wittgenstein's remarks about training youngsters as if they were circus animals shows that Wittgenstein's attitude to human life was strongly behaviourist. This too misses another point of consequence. On what sort of bedrock was Wittgenstein's spade famously turned? It was the bedrock of those practices which determined the local form of life. These are the hinges on which the door of language turns.Footnote 12 Of course, these bedrock practices ought to be inculcated as habits. Only thus could the river of life flow smoothly. The discussions of the deep working bases of life in On Certainty bring out the way that doubt cannot be allowed to infect a hinge.Footnote 13 Trained to keep left on the highway I am protected from dangerous vacillations that would occur if I began to doubt whether that was the rule of the road. What is the logical status of the propositions in which we might try to express the content of hinges? Since the hinges on which our lives turn might have been different from what they are, such propositions must be, in some large sense, contingent. However, the practices they describe could not be abandoned without our form of life disappearing, so they have a tinge of necessity about them. Cook's claim that the only necessity Wittgenstein admits is logical necessity misses an important qualification to the seeming disjunction between those propositions that are necessary and those that are contingent.
Conclusion
Cook is surely right to point to the ways that Wittgenstein slips away from the methodological technique of relying on the everyday talk of ordinary people and their customary linguistic practices. However, there is a good deal more than this to Wittgenstein's methodological engagement with linguistic matters. The key to his arguments lies in the powerful methodological device of picking up the subtle ways that inadvertently giving authority to misleading grammatical models can lead to typical philosophical impasses.
In the case of the problem of ‘other minds' the result of carefully distinguishing expressive from descriptive grammars is put very clearly in PI § 304. If we adopt the descriptive grammar to try to understand ‘I am in pain' we are forced to imagine a ‘something' independent of the utterance. Our route to knowing about such a being would have to be by a quasi-scientific inference from behaviour taken as evidence for the existence of an unobservable. But that procedure is hopeless. Indeed a nothing would be just as a good as something since neither could be reached this way. Drop the descriptive grammar in favour of the expressive grammar and the paradox disappears.
In reminding us that when we express our personal feelings we are not ascribing properties to a singular substance at all, that is not describing ourselves at all, Wittgenstein depends on the indexical analysis of the uses of the first person. If we try to understand the first person on the model of ‘name and object' we fall into philosophical error yet again. The spatiotemporal aspect of first person indexicality depends on the material embodiment of persons but it leaves the continuum from bricks to people intact. Cook's phraseology in commenting on the unfortunate Hal, the murderous computer of 2001 fame, as an occasion ‘on which Hal had spoken of himself' reveals the slippage I am drawing attention to.Footnote 14 Hal is speaking as himself, indexing the content of his utterance with his mode of personal being. Cook remarks that Wittgenstein ‘in his discussions of the word “I” … typically said that when we use the word in speaking of immediate experience – of pain, for example, we do not identify a subject of experience.'Footnote 15 No doubt, but we index the content of the utterance with the spatiotemporal location of the embodied speaker. We get to know where and when the pain occurred. As Wittgenstein remarks in PI § 404 there are no criteria of identity involved when someone says ‘I am in pain'. The pronoun is indexical and the utterance expressive.
Cook's suggestions as to the sub-texts of Wittgenstein's arguments seem to me to point in the right direction, but to be far too simplistic. People are necessarily embodied but not very brick-like. The first person is not referential, but it does serve to focus the interlocutor on attributes of the speaker. Natural necessity is not logical necessity, but hinges do determine the roots of a form of life.
