øøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøø ø ø ø File: 07-2-94.TXT - 51 KB ø ø ø øøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøø ø ø ø Michel ter Hark, Groningen - Netherlands ø ø ø øøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøø ø ø ø *Wittgenstein and Russell on psychology and other minds* ø ø ø øøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøø ø ø ø Hark, ter Michel (1994) Wittgenstein and Russel on psychology ø ø and other mindes; in: Wittgenstein Studies 2/94, ø ø File: 07-2-94; hrsg. von K.-O. Apel, F. Börncke, N. Garver, ø ø P. Hacker, R. Haller, G. Meggle, K. Puhl, Th. Rentsch, ø ø A. Roser, J.G.F. Rothhaupt, J. Schulte, U. Steinvorth, ø ø P. Stekeler-Weithofer, W. Vossenkuhl (3 1/2'' Diskette), ø ø ISBN 3-211-82655-6, ISSN 0943-5727 ø ø ø øøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøøø Michel ter Hark, Groningen - Netherlands *Wittgenstein and Russell on psychology and other minds* This chapter focuses on sections iv and v of part II of Wittgenstein's Philosophical Investigations. In these sections Wittgenstein deals with two closely knit problems: our knowledge of other minds and the subject matter of psychology. The interpretation of Wittgenstein's treatment of these problems cannot remain confined to these sections, however, as equally important references to these problems occur elsewhere in the Investigations as well as in the Remarks on the Philosophy of Psychology *1*. Moreover, Wittgenstein's very treatment of the two problems is intertwined with his treatment of some other problems in the philosophy of psychology. For instance, in the context of his discussion of the subject matter of psychology, he points out analogies with problems pertaining to theories of perception. I take these analogies seriously and, therefore, a discussion of perception is interwoven as appropriately. Yet another preliminary remark. Although references to other philosophers and psychologists do not occur in sections iv and v, it does not follow that they were not written in response to others. On the contrary, a substantial part of these (and other) sections is the product of reflecting upon theories produced by Russell, James and Köhler *2*. In my view, the best way to understand Wittgenstein is to involve the standpoints of his adversaries explicitly into the discussion. Otherwise seemingly trivial questions and condensed remarks, gain power and depth when explicitly seen as undermining and reversing dominant theories in philosophy and psychology. 1. Wittgenstein opens section iv with the following question: '"Ich glaube, dass er leidet." - Glaube ich auch, dass er kein Automat ist?'. His answer is that the question, asked under normal circumstances, makes no sense. For instance, to say of a friend 'He is not an automaton' is senseless. This is not to say that the proposition is false. But the proposition cannot be said to be true either. Truth and falsity belong to the logic of propositions that inform someone of something ('Mitteilung'). Although the proposition 'He is not an automaton' has the grammatical form of an informative proposition, like 'He is not ill', one must not be fooled by appearances. A communication typically conveys information, which is new, old but forgotten, founded on observations or other kinds of evidence. Moreover, it must be information, the denial of which is conceivable. Understanding what it would be like for a proposition to be true, implies understanding that, if it were not like that, the proposition would be false. But do we understand what it would be like if the proposition 'He is not an automaton' would be false? If not, affirming the truth of 'He is not an automaton' makes no sense either. The proposition, therefore, does not convey any information. As Wittgenstein puts it: 'Was wird hier mitgeteilt, und für wen wäre es eine Mitteilung? Für einen Menschen, der den Andern unter gewöhnlichen Umständen trifft? Was könnte es ihm mitteilen?' (PU II, iv, p. 178b). The first question that I would like to ask is not to whom the sentence would communicate anything but who wishes to convey this 'information' in the first place? Then we find ourselves in the epistemological problem of other minds. The question as to how each of us can know the nature of the mental states of others, or even that there are any other minds to have the mental states, which are not ours, has been widely discussed among analytic philosophers, such as J.S. Mill, Bertrand Russell and Rudolf Carnap. The question has been discussed too among German philosophers and psychologists, such as Köhler. Let us first have a look at Mill's and Russell's statement of the problem *3*. Mill puts the problem thus: 'I am conscious in myself of a series of facts connected by an uniform sequence, of which the beginning is modifications of my body, the middle is feelings, the end is outward demeanor. In the case of other human beings I have the evidence of my senses for the first and last links of the series, but not for the intermediate link' (Mill, 1867, pp. 237-238). From the same premises Russell draws the same conclusion: 'there seems no reason to believe that we are ever acquainted with other people's minds, seeing that these are not directly perceived' (Russell 1905, p.42). Not only can the minds of others not be perceived, the illusion of mental life can be produced where in fact it is absent. In this respect Russell refers to Descartes's argument that we may be dreaming but also to the possibility of ingenious automata, such as calculating machines and gramophone records that remember exactly what someone has said. Russell's sceptical conclusion is: 'How do you know that the gramophone does not think?' *4*. This threat of scepticism forces Russell to seek a postulate which establishes our belief in other minds rationally. He is after a postulate which establishes a rational connection between what is not perceived by the senses, for instance the belief that mother is angry, and what is perceived by the senses, such as the hearing of a loud voice and other events on the surface of her body. The postulate is provided for by an inductive generalization, i.e. the argument by analogy: if I have found, in my own case, that there is an association between certain mental states and behaviour and subsequently notice that there are other bodies similar to mine, manifesting similar behaviour, I am justified in concluding by analogy that mental states like the ones I experience are associated with those other bodies. 2. Wittgenstein's criticism is not so much against the argument by analogy, as against the other minds problem. His insight is that both scepticism and the argument by analogy share a common conception of mental states, which creates the other minds problem. This common assumption is, that the concept of a mental state is a state, the meaning of which can be defined independently of behaviour and occasions of behaviour. Scepticism about other minds suggests that mental states and behaviour are merely contigently related, and that it is quite possible for there to be the one without the other. The argument by analogy concedes this reasoning, but refuses to draw the conclusion, that no knowledge is possible. Wittgenstein's strategy is to show that this shared conception of the meaning of mental states is wrong. In section iv this dissolution of the other minds problem converges especially on one remark. After having said that the sceptical judgement 'I believe that he is not an automaton', just like that, makes no sense, Wittgenstein goes on: Meine Einstellung zu ihm ist eine Einstellung zur Seele. Ich habe nicht die Meinung, dass er eine Seele hat (PU II, iv, p. 178d). Especially two points ask for explanation. The first is that according to Wittgenstein 'zu ihm' can be substituted by 'zur Seele'. The second is that an 'Einstellung' is not a belief ('Meinung'). Although these two points are closely knit I will, for explanatory purposes, discuss them seperately. I will concentrate first on the latter point. In another remark the relation between 'Meinung' and 'Einstellung' becomes more explicit: Was aber ist der Unterschied zwischen einer Einstellung und einer Meinung? Ich möchte sagen: Die Einstellung kommt vor der Meinung. Eine Meinung kann sich irren. Aber wie sähe hier ein Irrtum aus? (LW vol. 2, §) The reference to error in this context points to an analogy between Wittgenstein's treatment of epistemic concepts in On Certainty and his treatment of the other minds problem. In On Certainty Wittgenstein opposes 'Irrtum' to 'Geistesstörung' and points out that 'Damit der Mensch sich irre, muss er schon mit der Menschheit konform urteilen' (OC, § 156). On the basis of this agreement certain errors are inconceivable. For instance, what would we think of somebody who claims that he cannot remember whether he has always had two hands? We no longer call this an error, but a mental derangement. The difference between the two is that an error has a reason and, therefore, can still be classified in the knowledge and opinions of the person who errs. An error can be explained and pointed out. The same conceptual distinction between error and mental derangement is made in the context of other minds: 'Es gibt doch Fälle, wo nur ein Wahnsinniger den Ausdruck des Schmerzes (z.B.) für unecht halten könnte'(LW vol. 2, § ) Evidently, Wittgenstein's conceptual analysis of the concept of error alludes to the epistemological argument from error, as exploited by Russell and others. This argument is to the effect that I cannot ever claim to know that p, unless I can show a relevant difference between two cases: the case in which I claim to know that p is true and the case in which I was wrong in claiming to know that p. Russell's sceptical question 'How do you know that the gramophone does not think?' alludes to this argument. Russell's argument is that, if nothing in your experience could count as evidence that the gramophone does not think, your belief that the gramophone does not think is unjustified. The belief is unjustified, because nothing you can point to, suggests that the gramophone is rather than is not thinking *5*. On the face of it, Wittgenstein seems to beg the question against Russell, for where the latter asks for a distinction between true and false cases the former simply says that there is such a distinction. Wittgenstein's point, however, is that the two cases, that is, the case in which he speaks of attitudes and the case in which he speaks of opinion, imply different uses of the concept of pain. Speaking in both cases of knowledge, evidence and error, Russell ignores this difference in use. If the concept of pain is used in the context of an 'Einstellung', the epistemological question for conclusive evidence makes no sense. For in this context our relation to the other is not one of knowing at all but of immediately reacting to the other, treating her or showing sympathy for her. This is not meant as an empirical observation but as constituting our concepts. As Wittgenstein puts it: 'My relation to the appearances is here part of my concept' (Z § 543). The content of an attitude is not based on epistemological grounds and inference, but is given by my very reactions to the other. Therefore, Russell's argument by analogy is rebutted by Wittgenstein as 'putting the cart before the horse' ((Z, § 542). Instead of saying, as Russell does, that we tend someone else, because by analogy with our own case we believe that he is experiencing pain too, we ought to say that it is our primitive reaction to tend the part that hurts, when someone else is in pain. Russell's mistake is to think that, unless such a reaction is given rational support, it is irrational and unsupported. As Wittgenstein remarks: 'Besteht eine Voraussetzung nicht, wo ein Zweifel besteht? Und der Zweifel kann gänzlich fehlen. Das Zweifeln hat ein Ende' (PU II, v, p. 180b). Our request for justification comes to an end and our assurance shows itself in our very attitude to other humans. An 'Einstellung', then, comes before an opinion and, hence, a mistaken opinion, in the sense that to be able to be mistaken about other minds already presupposes cases in which errors are inconceivable *6*. Immediately reacting to other humans as humans, is a response, which cannot be appreciated without the context of interaction, to which it belongs. But the response is not related to this context as a conclusion is related to its premisses, or a hypothesis to its evidence. In that sense, there is an anlogy between other minds and religious belief, of which Wittgenstein speaks too in section iv (cf. PU II, iv, p. 178e). And after the quoted alternative version of 'Einstellung zur Seele', he writes: '(Ist aber nicht der Glaube am Gott eben eine Einstellung?)' 'Wie wäre dies: Nur der glaubt es, der es als Mitteilung aussprechen kann' (LW vol.2, § ). This analogy with religion may reinforce the mistaken view of Wittgenstein, according to which the attitudinal approach amounts to a dogmatic protection of our views of others from criticism. But nothing Wittgenstein says, either in the context of religion or other minds, rules out criticism. He merely describes what religious belief and belief in other minds comes to. To say that it does not come to a belief, which can be communicated, like a true or false opinion, does not imply that this belief cannot be criticized. The point is that criticism of attitudes is different from criticism of opinions. In the former case criticism amounts to combatting ('bekämpfen') and persuading someone, not to point out contradictions *7*. Confusing the concepts of 'opinion' and 'belief' (in the sense of 'Einstellung') and, afortiori of the concepts 'combat' and 'criticism', leads to bad philosophy. Wittgenstein's diagnosis of the other minds problem rules out bad philosophy, without implying dogmatism. 3. Wittgenstein's 'attitudinal approach' to other minds, as I will call my reading of him, is in important respects different from the standard criteriological approach attributed to him *8*. According to the standard criteriological approach there exists a non-inductive evidential relation between mental states and behaviour. Because of this quasi-deductive, or quasi-logical relation holding between behaviour and mental state, behaviour is called a criterion on the basis of which the mental state is ascribed to another person. More precisely: painbehaviour is a criterion of pain, if and only if, the occurrence of painbehaviour is necessarily good but defeasible evidence for the truth of pain. And in the absence of contrary indications it is sufficient evidence. Hence in favourable cases the occurrence of painbehaviour perfectly justifies the belief that the other is in pain. There is no need for an analogical inference to other minds, for anyone who understands the concept of pain, knows that certain sorts of behaviour are criteria for ascription of pain. Therefore, it is not possible for there to be beings, who behave just as we do, but lack mental states as pain. In normal circumstances such behaviour counts as conclusive evidence that the person is in pain. Especially two features of Wittgenstein's remarks about other minds make the ascription of a criteriological approach in this sense to him unplausible. The first is that, whereas the criteriological approach sees knowledge of other minds as a matter of belief, the core of the attitudinal approach is precisely the absence of the notion of belief. The second is that the notion of a non-inductive connection between something called 'behaviour' and something else called 'mental state,' seems to appeal precisely to the dichotomy of the inner and the outer, the dissolution of which is at the heart of Wittgenstein's philosophy of language and psychology. As to the first point, he role of criteria is to provide justification for belief. Indeed, the role of criteria is to establish the existence of other minds with (deductive) certainty. This interpretation, however, assumes that Wittgenstein's goal was to refute scepticism, because he took the sceptical premise that we can never know with certainty other minds seriously. There is no evidence that Wittgenstein ever took scepticism seriously. On the contrary, according to Wittgenstein scepticism is not so much false as senseless. Of course, we apply psychological concepts to others on the basis of what we see and hear, i.e. their behaviour and expressions, but what we see and hear is no one's reason (criterion) for believing in other minds. We show in our reactions to others that we take this 'belief' for granted *9*. The same criticism applies to the successor of the criteriological approach to other minds, the theory approach. According to this approach we rely on a theory when attempting to explain other people's behaviour, just as in physics one relies on a theory that enables us to infer the presence of electrons as the best explanation of cloud-chamber tracks *10*. The mind is a theoretical postulation just as electrons and elasticity are postulated in order to explain certain phenomena. The analogy between inferences made in scientific context and knowledge of other minds is besides the point. Although it is true that in the language game *11* of science we postulate hypothetical entities, because we believe that they provide the best explanation of observed facts, no one believes in other minds for the reason that this provides the best explanation of observed behaviour. Reasons in the sense of criteria and grounds, rationally supporting our beliefs, are not part of attitudinal language games. Although the theory approach is different from both the analogical approach and the criteriological approach, the demand for good reasons for believing in other minds is a common and tacit presupposition. The attitudinal approach questions precisely this presupposition. 4. The second point of difference between the attitudinal approach and the criteriological approach concerned the inner-outer dichotomy. To think of criteria as a special kind of evidence connecting something called behaviour and something called mental state ivokes the inner-outer dichotomy Wittgenstein wishes to go beyond. Indeed, to say 'Meine Einstellung zu ihm ist eine Einstellung zur Seele' is not to say 'to him and to a soul'; the attitude is not towards two things but to one 'thing'. And in the phrase 'Der menschliche Körper ist das beste Bild der menschlichen Seele' (PU II, iv, S. 178g) the suggestion is that the soul can be seen *12*. Leaving telepathy aside, Wittgenstein seems to say that the notion of the soul is tied to perception of the human body. Whatever he may mean, the link between the problem of other minds and perception is rather different from what philosophers have understood by a criterial connection between two different sorts of facts. Wittgenstein's use of 'seeing' approaches more a psychological than an epistemological view of other minds. In that respect, a striking resemblance exists between Wittgenstein and Köhler. Köhler equally rejects the argument by analogy and the dichotomy of the inner and the outer upon which it is based. Surely, the outer conceived as a physical fact has nothing in common with mental processes, but in the context of the other minds problem we refer to the body as a perceptual fact and not to the organism as studied by anatomists and physiologists. Somehow subjective experiences tend to express themselves in the behaviour of people as we perceive them, and to this fact Köhler adds the observation, that the expression resembles what is being expressed. If that is the case, the main reason for indirect interpretations of other minds would be removed. For instance, if somebody is thinking angrily about a rival, he will start to walk faster, he will raise his voice. The increased tempo of the inner agitation is matched by a crescendo and accelerando in behaviour, expression and diction. In this way the outer is a direct picture of the inner: not indirect but direct evidence. The inner, the soul, mental processes are just as visible as form and colour. Wittgenstein's remark 'Der menschliche Körper ist das beste Bild der menschlichen Seele' approaches more Köhler's psychological observations than Russell's epistemological argument to the effect that we never directly perceive other minds. Indeed, like Köhler *13*, Wittgenstein intends to reverse the Cartesian conviction, according to which the body hides the soul. The soul is hidden within the body or by the body and the other person has to guess from the surface of the body what goes on inside. Wittgenstein's comment upon this traditional conviction is: Es ist nicht so, als hätte ich in mir direkte, er für mein Seelisches aber nur indirekte. Sondern er hat dafür Evidenz, ich (aber) nicht. (LW vol. 2, §...) The traditional conviction is that the adjectives 'inner' and 'outer' refer to properties of inner and outer processes much like the adjective 'fast' can refer to the inner process of digestion. But whereas 'fast' is a normal physical property, ascribing the property 'inner' to an inner process or substance is tantamount to ascribing a mysterious epistemological property of direct inner evidence. Wittgenstein attempts to undermine and replace this property as well as its inseperable opposite, i.e. indirect outer evidence. Wittgenstein insists upon the asymmetry between the first person and the third person *14*. The traditional view that each of us has direct knowledge in our own case and indirect knowledge in the case of others is replaced by an asymmetry according to which we use observation in the case of others, but not in our own. The conviction that we have direct knowledge of our own mental states is jettisoned by Wittgenstein earlier in the Philosophical Investigations, especially in § 244, where he remarks that the meaning of the verbal expression 'I am in pain' is not based on direct inner evidence. If we pay attention to the use of this sentence - and to Wittgenstein meaning is use - its meaning is to replace a prelinguistic expression of pain. 'I am in pain' corresponds to a cry of pain. Like a cry of pain, the role and purpose of the sentence 'I am in pain' is to express a feeling, to ask for help, etc. 'I am in pain' is not a report of my inner state, but neither is it a description of my behaviour. Wittgenstein is not a behaviourist. 'I am in pain' corresponds to a cry of pain, not to the statement that I am crying. His analysis of the first person, then, walks the narrow path between introspectionism and behaviourism. But what about his analysis of the third person? Is he not sliding off into the abyss of behaviourism, when saying that the body is the best picture of the human soul? In any case, he anticipates this 'accusation' in section v: 'So handelt die Psychologie vom Benehmen, nicht von der Seele?'(PU II, v, p. 179b). In order to see the difference between Wittgenstein and logical behaviourism one has to attend to the phrase 'das beste Bild'. First of all, when Wittgenstein speaks of 'Körper' he does not mean the anatomical or physiological body, but the behaviour ('das Verhalten') of the body , especially facial expressions, gestures and the most subtile form of behaviour, i.e. language *15*. The human body is the best picture of the human soul in the sense that looking at a face and reporting 'He was out of humor' is not a metaphorical description of what one sees. As metaphors can be paraphrased and translated into other sentences, or even reduced to other sentences, Wittgenstein opposes translation or reduction of psychological language into physicalistic language, which is the hallmark of logical behaviourism. For instance, according to Carnap every psychological sentence may be formulated in physical language, i.e. language describing 'the physical behaviour of humans and other animals' (1959, p. 165). Wittgenstein's comment on Carnap's physicalism and behaviourism is: 'Mann kann auch sagen "Er machte dieses Gesicht", oder "sein Gesicht veränderte sich so", indem man es nachmacht, - und ist wieder nicht im Stande, die Veränderung anders zu beschreiben. (Es gibt eben viel mehr Sprachspiele, als Carnap und Andere sich träumen lassen.)' (RPP, I, § 920). To describe someone's pain by looking at his facial expression, or by looking at the geometrical pattern on the face, are completely different language games. How different, is shown by the fact that we 'describe' someone's pain by pulling ourselves a painful face. We can do this instinctively, without needing to look in a mirror. And without being able to say what spatial alterations in the other person's face our imitation is based on. Similarly, to say of someone 'He was out of humour' is the best description of what is seen, as another description, in terms of what spatial alterations in the face made this impression, cannot be given (Cf. RPP vol. 1, § 287). The difference in language games is as great as the capacity of a painter to paint an eye so, that it stares, and his incapacity to describe the distribution of colour on the surface (Cf. RPP, I, § 1077). Attempts at translation are guided by the mistaken idea of a better, in the sense of an ideal and absolutely accurate, description. According to Wittgenstein this ideal is a myth. Logical behaviourism, in all its versions, is such a myth. Emphasizing that the human body is the best picture of the human soul, is to resist the temptation to translate mental propositions into behavioural propositions. Hence, Wittgenstein is not a logical behaviourist in the strong sense of the term. But he is not a logical behaviourist in the weaker sense either, for he does not argue that a translation is impossible 'for such superficial reasons as the greater ambiguity of mind talk, as compared with the relatively greater specifity of overt behaviour talk', as Putnam seems to attribute this view to Wittgenstein (Putnam 1965, p. 152). Again, the suggestion is that 'mind talk' is an ambiguous metaphor, ultimately to be reduced to a literal description of behaviour or physiology. Wittgenstein rejects these attempts at translation and reduction, not because he is biased against the science of psychology or physiology, but because these attempts are the product of bad philosophy. The mistake of behaviourism is, to think that sentences about other minds are not about experiences, but about something else, i.e. behaviour. In that respect, behaviourism is the mirror image of introspectionism, according to which sentences about other minds are not about behaviour, but about something else, i.e. private experiences. Wittgenstein resists both temptations. 5. This comes out clearly in two important remarks in section v: "Ich merke, er war verstimmt." Ist das ein Bericht über das Benehmen oder den Seelenzustand? ("Der Himmel sieht drohend aus": handelt das von der Gegenwart, oder von der Zukunft?) Beides; aber nicht im Nebeneinander; sondern vom einen durch das andere (PU II, v, p. 179c; RPP vol.1, § 288). Es ist hier ganz wie mit dem Reden über physikalische Gegenstände und Sinneseindrücke. Wir haben hier zwei Sprachspiele, und ihre Beziehungen zueinander sind kompliziert. Will man diese Beziehungen in einfacher Weise beschreiben, so geht man fehl' (RPP, vol. 1, § 289; PU II, v, p.180c). Let us first take up Wittgenstein's suggestion and see what the analogies are between the discussion about the subject matter of psychology and theories of perception. Although no mention is made of Russell, it seems clear that Wittgenstein has Russell's theory of sense data in mind *16*. Russell's theory of perception is a form of indirect realism. According to indirect realism we are indirectly aware of the material objects around us in virtue of a direct awareness of inner and non-material objects. Russell's version of indirect realism could best be called inferential realism. According to this theory we are never aware of material objects, but infer their presence from inner objects, with which we are immediately acquainted. Russell, then, discards the notion of an indirect, external material object, but leaves intact the notion of a direct internal object. The term 'sense datum' is coined by Russell to refer to these direct internal objects of perception. Examples are: color patches, shapes, sounds, smells and tactile feelings. These are the only objects of perceptual awareness, and we are not aware of external physical objects in any sense analogous to that, in which we are aware of sense data. Instead our knowledge of physical objects is inferential; we infer the presence and nature of material objects from sense data, which are caused by these objects. As Russell puts it in the context of an analysis of the method of introspection in psychology: 'When we "see a table", as common sense would say, the table as a physical object is not the "object" (in the psychological sense) of our perception. Our perception is made up of sensations, images and beliefs, but the supposed "object" is something inferential, externally related, not logically bound up with what is occurring in us'( 1921, p. 111-112). Russell's definition of psychology is in line with this inferential realism, for psychology is not the science of behaviour, but of something else: 'the science which deals with private data, and with the private aspects of data which common sense regards as public' (Russell 1948, p.45). The difference between mental states and behaviour, or between sense data and material objects, comes to a difference in number of entities: two instead of one, as behaviourism in psychology or direct realism in theories of perception would claim. Qua objects, the same logic applies to sense data material objects, mental states and behaviour. In Wittgenstein's terms: the same language game is played. Wittgenstein's point of view, however, is exactly the opposite of Russell and behaviourism. Wittgenstein wishes not to deny that there is a difference involved here, but emphasizes that it is a difference in language games; there are two language games, not two types of object. This is the meaning of the rather cryptic remark in the quote above: 'Beides; aber nicht im Nebeneinander'. Let us first see how this remark applies to the case of perception. Wittgenstein's point is that the concepts of visual impression and material object belong to different logical categories. Obliterating the distinction of categories, as indirect realism does, is committing a categorial mistake. However: 'Der Eindruck - möchte ich sagen - sei kein Gegenstand' (RPP I, § 1081). And to this we should add: not an object of perception. That is, the impressions we get when perceiving things are not themselves anything that we perceive. Indeed, this is the main reason for denying visual impressions the grammatical or logical status of objects: 'Den Eindruck kann man nicht betrachten, darum ist er kein Gegenstand. (Grammatisch) Denn man betrachtet den Gegenstand nicht, um ihn zu ändern. (Das ist eigentlich, was Leute damit meinen: die Gegenstände existieren "unabhängig von uns".)' (RPP I, § 1085). The suggestion is that, to the extent that introspection could be seen as a form of observation of inner objects, the objects are changed by the observation and therefore the observation-object model has to be discarded in the case of visual experiences. For, observation does not produce what is observed (Cf. PU II, ix, p. 187a). An observation must be applicable to the cases in which no observation takes place. But if introspective observation produces what is observed, the 'object' is dependent upon the act of 'observing' and cannot be referred to or spoken about in cases in which the 'observation' does not take place. Then the absurd consequence would follow that someone could say: no, I did not really mean 'auditory impression' for I have not listened attentively enough. If impressions are not objects because they cannot be observed, neither can they be said to resemble material objects. But this is precisely what indirect realism needs to say. Indirect realism owes us an account of the relation between the sense data which we do see and the physical objects we do not see. In order to say that sense data resemble physical objects, the two types of object ought to be compared with each other. But in respect of what? If we see an old tree it would be nonsense to say that we see an old sense datum of a tree. And from the fact that a grave looks purple, it cannot be inferred that the grave causes in us an impression which is purple. Indirect realism has been mislead by a false analogy. For instance, the analogy between a mental image and a physical object on the one hand, and a painting and a physical object on the other. In the latter case one can speak of an indirect and a direct description, but not in the former *17*. A painting of the queen is not the queen itself, but something else, something hanging on the wall. You can put the queen next to the painting ('im Nebeneinander') in order to convince a philosopher that the description 'This is the queen', while pointing to her, is more direct than pointing to the painting and saying 'This the queen'. As to the relation between a mental image and a physical object this procedure is not available. I cannot shift my attention from the physical object to something else I can show. The picture of two different objects has been wrongly applied here. The situation is unlike the one in which one can point to two different objects, one after the other, and say 'I meant this object, not that object'. The very temptation to construct a more direct account of visual impressions already assumes that to speak about material objects is to speak about another type of objects than about visual impressions. But visual impressions are not a type of object at all and speaking about material objects is the way to speak about visual impressions. 'Wir lernen Gegenstände beschreiben, und dadurch, in anderm Sinne, unsere Empfindungen' (RPP vol. 1, § 1082). We do not speak about them side-by-side, but 'vom einen durch das andere'(PU II, v, p. 179c). 6. The psychologist too speaks 'vom einen durch das andere', that is, about mental states via behaviour. Conceived as a sort of definition of the subject matter of psychology, this is different from both introspectionism and behaviourism. For according to the former, psychology is not the science of behaviour, but of something else, i.e. private objects, whereas to the latter psychology is not the science of private objects, but of public objects. Wittgenstein suggests a way out of this dilemma of introspectionism and behaviourism, by drawing attention to a shared, but mistaken, assumption as to the nature of psychological phenomena: both camps assume that psychological phenomena are objects. Indeed, behaviourism denies the existence of psychological phenomena just because it assumes that they are a kind of objects; objects the evidence for which is lacking. To a great extent, speaking about psychological objects is inspired by a comparison with physics. Wittgenstein, instead, emphasizes the disanalogy with physics: 'Sehen, Hören, Denken, Fühlen, Wollen sind nicht im gleichen Sinne die Gegenstände der Psychologie, wie die Bewegungen der Körper, die elektrischen Erscheinungen, etc., Gegenstände der Physik. Das siehst du daraus, dass der Physiker diese Erscheinungen sieht, hört, über sie nachdenkt, sie uns mitteilt, und der Psychologe die Äusserungen (das Benehmen) des Subjekts beobachtet' (PU § 571). Wittgenstein's point is not that psychology can only deal with behaviour and physiology, but that in one sense he reports or speaks about behaviour and, precisely through that, in another sense about mental states. In the manuscript version of PU, § 571, Wittgenstein vividly illustrates his idea of two senses (rather than two objects): 'Dass man in anderem Sinne eine Frau und ihr Geld heiratet, siehst du daraus, dass ihr Geld heiraten heisst, sie ihres Geld wegen heiraten' (MS 116, p. 333). The point of this analogy is that just as to the question 'How do I marry money?' the only conceivable answer is 'By marrying a rich woman', and not 'Marrying a woman plus marrying money', to the question of the psychologist 'How do I observe mental states?' the answer is 'Via observation of behaviour and in particular verbal utterances', and not 'By observing behaviour plus introspecting and inferring mental states'. Unless he is a behaviourist, the psychologist's reason for observing the behaviour of his testees is that this behaviour is expressive of mental states. As Wittgenstein says about these utterances: 'Aber diese handeln nicht vom Benehmen' (PU II, v, p. 179b). Put otherwise: the psychologist deals with utterances, and hence, with behaviour, but these utterances do not deal with behaviour. And precisely because they do not deal with behaviour, they are of interest to the psychologist. Wittgenstein's definition of the subject matter of psychology reflects more the experimental practice than Russell's definition does. For instance, in psychophysics experiments are conducted in order to measure the absolute threshold for hearing. The problem consists in determining the relation between an auditory impression and the magnitude of a physical stimulus. The experimenter presents a very faint, undetectable tone of a particular and constant frequency and increases its intensity in small steps until the testee reports 'I hear it'. Why does the experimenter pay attention to the moment his testee utters this report? Because the use of the utterance 'I hear it' is to be an utterance of an auditory impression. The semantical tie between the utterance and the auditory impression is such that to the experimenter the utterance is the impression. Just as to give a promise is not a forecast of my own behaviour, but is a forecast, the utterance 'I hear it' is an expression of an auditory impression, in the sense that no wedge can be driven between the expression and what it is an expression of. The semantical tie between the impression and the utterance is testified by the fact that the experimenter draws inferences from the utterances - and not from something else behind it - as to the threshold for hearing. In the context of this experiment to determine the threshold for auditory impressions is to determine when the testee says 'I hear it'. The difference between the testee and the experimenter, then, is not that the former knows or possesses something hidden from the latter. The difference, according to Wittgenstein, is that the same mental state is manifest in a different way: to the testee as an utterance, to the experimenter as an observation. Of course, the testee may be dishonest and report to hear a tone when in fact none is heard. But this is no reason for drawing an epistemological distinction between physical stimuli or physical objects and impressions or sense data and arguing, as Russell does, that, not until we have reached the rock bottom of impressions or sense data, we can be certain as to our knowledge of other minds or the external world. Wittgenstein's answer to Russell again testifies to an analogous treatment of other minds (whether in daily life or psychology) and perception. For his answer to Russell's sceptical question 'Is there really a table over there?' would be: 'Of course, there is a real table over there, as opposed to an imitated exemplar'. Equally, to the objection to his definition of the subject matter of psychology '"Freilich berichtet der Psychologe die Worte, das Benehmen des Subjekts, aber doch nur als Zeichen seelischer Vorgänge"' (RPP I, § 292) his answer is: 'Das ist richtig. Wenn die Worte und das Benehmen, z.B., eingelernt sind so interessieren sie den Psychologen nicht' (ibid.). The psychologist is interested in the sincere report of experiences, for these are the primary semantical expressions of experience. They are not the epistemological rock bottom of psychological phenomena, as Russell would have it, but the semantical rock bottom. The concepts of pretence and perceptual error - which belong equally to language games - logically presuppose language games of sincere expressions and reports of mental states. 7. Compared with Russell's inferential realism in the case of perception, Wittgenstein approaches more the position of naïve realism. We all are naïve realists in daily life in the sense that no one regards things around us as constructions out of sense data. But although sense data are not objects of any kind, Wittgenstein, in contrast to direct realism, refuses to draw the conclusion that we have no visual or auditory experiences. On the contrary, we often speak about experiences. The important thing is that speaking about experiences is a different language game than speaking about material objects. Similarly, compared with Russell's introspectionism and analogical inference, one might say that Wittgenstein's approach to other minds and psychology comes to a form of naïve behaviourism. Naïve behaviourism, however, is completely different from the philosophical and psychological forms of behaviourism. According to naïve behaviourism we make no sharp distinction between the observation of behaviour and experiences expressed by behaviour and, hence, no semantical gap needs to be bridged between the outer and the inner. But, then, we are all naïve behaviourists, even in the psychological lab *18*. Again, the important thing - totally overlooked by philosophical and psychological behaviourism - is that speaking about mental states is a different kind of language game than speaking about behaviour. This difference, as we have seen, is as great as can be, and there is no need of two different kinds of objects to ground this difference. *1* Remarks on the Philosophy of Psychology vol. 1 also contains the preparatory studies for sections iv and v, especially § 265-292. The Last Writings on the Philosophy of Psychology vol. 1 and 2 equally contain important material. *2* See ter Hark 1990 chapter VII for explicit references to especially James and Köhler in Wittgenstein's published and unpublished works. *3* Especially Russell is no arbitrary choice, for Wittgenstein was familiar with the former's view of other minds in various articles and books (Russell, 1905, 1912, 1921, 1926, 1948). Köhler (1929) was used by Wittgenstein in his lectures on the philosophy of psychology. Explicit references to Köhler's view of other minds can be found in RPP II, § 334 and in 1988, p. 42 and p. 167. *4* Russell (1948), p. 484 *5* The same reasoning is applied to other minds: 'We naturally suppose that other people's bodies are inhabited by minds more or less like our owns...yet it would seem also that there is no conclusive reason for believing it not mistaken'. (Russell 1984, p. 13). *6* Scepticism about other minds has point only in certain situations, such as the Turing Test, in which an interrogator has to guess from output on a screen, which answer comes from a human and which from a computer. The theory approach to other minds, to be discussed below, takes the Turing test as its epistemological point of departure and, hence, obliterates the important distinction between normal situations and exceptions. *7* See OC, § 609-612. *8* Notably Malcolm (1963) and Kenny (1967). See for an excellent survey of criteriologism Lycan (1971). This approach has had an unfavourable effect on the reception of Wittgenstein in contemporary philosophy of mind. See especially Chihara & Fodor 1965 and Putnam 1965. *9* See Strawson 1985 for a similar view as Wittgenstein. *10* The example is taken from Chihara & Fodor 1965. *11* Wittgenstein distinguishes between the language games of 'making scientific hypotheses and theories' and 'describing an immediate experience' (cf. BLB, p.68). Fodor&Chihara (1965) mention this distinction only to blurr it. See 1965, p.147. *12* Indeed, this is what Wittgenstein explicitly says in an unpublished remark which comes after PU § 285: 'Man kann wirklich sagen: der beseelte Körper hat Schmerzen. Und ob ein Körper beseelt ist, das nimmt man durch die Sinne wahr'(MS 124, pp. 244- 245). *13* According to Wittgenstein, however, Köhler is not entirely successful. See note 18. *14* The epistemological account of the first and third person is symmetric and can beexpressed in this way: I do not observe his pain, but he does and I observe myown mental states but he does not. The analogical approach speaks of introspection versus inference, the theory approach in both cases of inference and behaviourism in both cases of observation. *15* 'Welches bessere Bild des Glaubens könnte es geben, als der Mensch, der mit dem Ausdruck des Glaubens sagt "Ich glaube..."?' (RPP I, § 280). *16* Indeed, the analogy between theories of perception and theories about other minds is present in Russell's own work, for he is modelling his discussion about psychology on his treatment of the relation between sense data and physical objects. *17* Rejecting the distinction between a direct description and a indirect description of experience, Wittgenstein not only distances himself from Russell but also from Köhler. Although Köhler does not espouse introspectionism, he equally attempts togive a more direct description of other minds. Pointing to the resemblance between subjective experiences and perceptual phenomena Köhler wishes to say that, for instance, the concept of friendliness is as optical a feature as color and shape. Describing a facial expression as friendly is especially not a matter of interpreting what one sees. Wittgenstein makes two comments on this. First of all, to speak ofa resemblance again invokes a dichotomy of the inner and the outer, which has absurd consequences: 'Köhler talks of resemblance between a feeling and its expression - as though in rage the soul trembles as well as the body?'(Wittgenstein 1988, p. 42). Secondly, although Köhler is correct in rejecting the introspectionistic view that, seeing a friendly face, is an indirect affair, a form of thinking rather than seeing, he has erred in attempting to give a more direct account by reducing 'seeing-as' to seeing. According to Wittgenstein 'seeing-as' has logical relations with both thinking and seeing. However, I will not elaborate here upon Wittgenstein's discussion of seeing-as. See ter Hark 1990 chapter 6. *18* The example of psychophysics discussed above supports this contention. Another difference between naïve behaviourism and philosophical behaviourism is that theformer is not construed in order to pave the way for particular methodologies in psychology, for instance, psychological behaviourism. Naïve behaviourism is asemantical view and compatible with various methodologies. It is, for instance,compatible with Gestaltpsychology, which is neither introspectionistic, nor behaviouristic or reductionistic in any other sense. LITERATURE Armstrong, D.M. & Malcolm, N., Consciousness and Causality. Oxford: Basil Blackwell 1984. Carnap, R., 'Psychology in Physical Language' in A.J. Ayer (ed.), Logical Positivism. New York: The Free Press, 1959, pp. 165-197. Chihara, C.S. & Fodor, J.A., 'Operationalism and Ordinary Language: A Critique of Wittgenstein' American Philosophical Quarterly II, 4, 1965, 281-295. Quoted from: Rosenthal, D.M., The Nature of Mind. Oxford University Press: Oxford 1991, pp. 137-150. Chisholm, R.M., Theory of Knowledge. Prentice Hall, Inc., Englewood Cliffs: New York 1966. Hark ter, M.R.M., Beyond the Inner and the Outer: Wittgenstein's Philosophy of Psychology. Dordrecht: Kluwer Academic Publishers 1990. Kenny, A., 'Criterion' in P. Edwards (ed.) The Encyclopedia of Philosophy. New York: Macmillan 1967, pp. 258-261. Köhler, W., Gestaltpsychology. New York: Liveright Publishing Corporation 1929. Lycan, W.G., 'Noninductive evidence: Recent Work on Wittgenstein's "Criteria"' American Philosophical Quartely, 8, 2, 1971, pp. 109- 123. Malcolm, N., Knowledge and Certainty. Englewood Cliffs: Prentice Hall 1963. Mill, J.S., An Examination of Sir William Hamilton's Philosophy. London: Longmans, 1867. Putnam, H., 'Brains and Behaviour' in Butler, R.J. (ed.), Analytical Philosophy. Oxford: Basil Blackwell 1965. Quoted from Rosenthal, pp. 151-159. Russell, B., The Problems of Philosophy. London: Williams and Norgate 1912. Russell, B., 'On Denoting' Mind, 14, 1905, pp. 479-493. Quoted from Russell, B., Logic and Knowledge: Essays 1901-1950. London: George Allen and Unwin 1950, pp. 41-56. Russell, B., The Analysis of Mind. London: George Allen and Unwin 1921. Russell, B., Our Knowledge of the External World as a Field for Scientific Method in Philosophy. London: George Allen and Unwin 1926. Russell, B., Human Knowledge: its Scope and Limits. New York: Simon and Chuster 1948. Russell, B., Theory of Knowledge. London: George Allen and Unwin 1984. Strawson, P.F., Skepticism and Naturalism: Some Varieties. Methuen And Co. Ltd., 1985. Wittgenstein, L., Philosophical Investigations. Oxford: Basil Blackwell 1953. Wittgenstein, L., The Blue and the Brown Books. New York: Harper and Row 1958. Wittgenstein, L. Zettel. Oxford: Basil Blackwell Wittgenstein, L., On Certainty. Oxford: Basil Blackwell 1979 Wittgenstein, L., Remarks on the Philosophy of Psychology 2 vols. Oxford: Basil Blackwell 1980. Wittgenstein, L., Last Writings on the Philosophy of Psychology 2 vols. Oxford: Basil Blackwell 1982, 1992. Wittgenstein, L., Lectures on Philosophical Psychology 1946-1947. Notes by P.T. Geach, K.J. Shah and A.C. Jackson. New York: Harvester 1988.