Erik Stenius, Wittgenstein's Tractatus: A Critical Exposition of Its Main Lines of Thought, 1960.



In the preceding chapters we have mainly directed our attention to the author of the Tractatus as a logician and a semanticist. In this final chapter we will give a short analysis of Wittgenstein's philosophical system.

In so far as Wittgenstein adhered to 'logical atomism' he could be characterized as a metaphysician of a rather Anglo-Saxon type. But I believe this sort of metaphysics to be in fact of only secondary importance in his philosophical system, which is, on the whole, more related to German metaphysics, and in particular to the metaphysics of Kant.

Wittgenstein rejected the possibility of any synthetic propositions a priori. Since the basic task of Kant's Critique of Pure Reason was to show how synthetic propositions a priori are possible, we might infer that Wittgenstein -- like the logical positivists -- was strongly anti-Kantian. But this is not so -- at least it is not undisputably so. Unlike the logical positivists Wittgenstein was in essential respects a Kantian philosopher; his anti-Kantianism meant only that he -- like other Kantians -- transformed the system of Kant and thus created a Kantianism of a peculiar kind.

This is not to be understood to mean that Wittgenstein had been directly influenced by Kant's writings.1 One did not need to have read Kant to be influenced by a more or less clearly stated Kantianism; it belonged to the intellectual atmosphere in the German speaking world. Moreover we know that Wittgenstein read and appreciated Schopenhauer,2 who was in his way a Kantian of a peculiar kind. In what sense Wittgenstein was so I shall now try to explain.


I shall begin with emphasizing one particular feature in Kantian philosophy. Kant's question 'How are synthetic judgements a priori possible?' was, to be sure, a basic question in the Critique of Pure Reason, but his interest in this question was not only an interest in an epistemological problem as such. His investigation of the a priori had a purpose which was still more fundamental to his philosophical outlook. This was to make a dichotomy between what belongs to the genuine province of 'theoretical reason' and what does not belong to it. The error in Leibniz-Wolffian metaphysics, which Kant called 'dogmatic', was that it applied the forms of theoretical reason to questions which do not belong to the province of theoretical knowledge -- as questions about God, the immortal soul, the universe as a whole, the free will, morals, and so on. All those questions belong to the province of 'practical reason' and can be answered only by methods proper to it. We can thus state as a fundamental line in Kantian thought the dichotomy between questions belonging to the provinces of theoretical and practical reason.3 The same distinction -- though certainly interpreted in a different, more mystical, way -- is found in the very title of Schopenhauer's main work The World as Will and Idea, where the 'World as Will' corresponds to the province of practical reason and the 'World as Idea' to the province of theoretical reason.

Starting from this distinction we can indicate as the main purpose of Kant's basic question that it serves as a basis for fixing the limits of the province of theoretical reason. 'How are synthetic judgements a priori possible?' Kant's famous 'Copernican revolution', which gave him the answer to this question, consisted in the 'discovery' that there are two components in human experience qua human experience. One component emanates from our sensations, and this is empirical and a posteriori, but there is also a component emanating from our theoretical reason, and this is a priori. The latter component consists of two parts. Our 'sensations' (Empfindung) ate not yet 'perceptions' (Wahrnehmung); in order to give rise to a perception a sensation must be interpreted within the framework of space and time; thus space and time constitute the a priori form of perception.4 And our perceptions are still not 'experiences'; in order to be so they must be subjected to the conceptual framework of experience, which -- together with the form of perception -- constitutes the a priori form of experience. The fact that all our experiences must have the a priori form of experience is the source of a number of synthetic judgements a priori: the form of perception is the source of the mathematical propositions, the conceptual framework of experience is the source of the propositions of 'pure science' -- as, for instance, the law of causality. Such statements are not analytic, but they are nevertheless true a priori, because they are prior to all experience qua experience: they are prescribed to nature in so far as nature is capable of being experienced at all by theoretical reason.

An investigation of our theoretical reason shows the limits of all possible experience and thus also what kind of questions lies outside this limit. Such an investigation Kant calls a transcendental deduction -- it shows the limits of our theoretical reason. And what is 'transcendental' in this sense must be distinguished from what is transcendent? that is, what transcends this limit. To the domain of the transcendent belongs also the Ding an sich, i.e., a 'thing' which exists independently of the form of experience. Schopenhauer assumed that the mind has access to the Ding an sich through the intuition of the will, which rends the 'Veil of Maya' by which theoretical reason covers reality. But this was a definitely non-Kantian turn of Kantianism. To Kant himself any kind of 'knowledge' was a matter of theoretical reason; principles of practical reason cannot be known by any special faculty of intuition but only postulated as necessary conditions for the existence of a moral world order.

The method used by Kant in his transcendental deductions, and in particular in his deduction of the form of perception is of interest if we wish to compare his views with the philosophy of Wittgenstein. Anything that can be perceived by a human mind must be submitted to the laws of (Euclidean) geometry, Kant argues. But how can he know this to be so? The method he actually uses for arriving at this knowledge is an analysis of what he calls our Anschauung. The word Anschauung is usually translated as 'intuition', but from our present point of view this is a very bad translation. Rather one ought to render the word as 'imagination', because the a priori character of the geometrical theorems is guaranteed by the fact that we can prove them by means of reine Anschaunng, i.e. 'pure imagination' without any reference to sensations. This idea could, however, be expressed thus: we cannot even imagine a world in which the theorems of geometry were not valid, and since the real world, in so far as it is capable of being perceived, must be an imaginable world, the theorems of geometry must necessarily hold true in the real world. A world which is 'possible' in terms of our theoretical reason must be an imaginable world. Thus an investigation of what is imaginable to us shows what is a priori true with regard to perception. In the same way an investigation of what is intelligible to theoretical reason in other respects gives us additional a priori knowledge of the world of experience.

What I call Kantian philosophy could thus be summed up in the following theses:

  1. The task of theoretical philosophy is to make transcendental deductions concerning the limits of theoretical discourse, not to speculate over what transcends this limit and thus cannot be theoretically known.
  2. A world is a possible world of experience only if it is 'possible' to theoretical reason, i.e. if it is imaginable and intelligible.
  3. Our experience has a 'form' which is founded in theoretical reason and a 'content' which is based on our sensations.
  4. True synthetical propositions are a priori if they refer only to the form of experience, a posteriori if they refer also to the content.
  5. Thus there exist synthetic propositions a priori (e.g. mathematical statements, the law of causality).
  6. There are also 'transcendent' propositions (e.g. propositions on God, the immortal soul, the universe as a whole, free will, morals, etc.). Such propositions cannot be known to be true by theoretical knowledge, but only 'postulated' by practical reason (Kant) or intuited by the will (Schopenhauer). The establishment of such propositions belongs to the task of practical philosophy.
  7. The Ding an sich is transcendent.

2. Wittgenstein's transcendental deductions

If formulated in this, to be sure rather Wittgensteinian, way our theses of Kantianism need only to be modified in one respect in order to be transformed into the main points of Wittgenstein's philosophical system.

I start from thesis (b). As it stands there it could be accepted by Wittgenstein. But we have to realize that what is 'imaginable' and 'intelligible' is what is 'thinkable' and that 'thought' is the 'logical picture of reality', which means that what is 'thinkable' is that which we can present by a logical picture, or in other words, that which can be described in a depicting language. And since Wittgenstein assumes all language to be depicting language, it follows that what is in (b) characterized as 'imaginable' and 'intelligible' is identical with what can be described in meaningful language.

Thus to be possible to theoretical reason corresponds in Wittgenstein's philosophy to possibility in terms of what is describable in meaningful language. This is the essential modification of the Kantian view which gives rise to all differences between Wittgenstein and Kant. The task of (theoretical) philosophy is for Wittgenstein as for Kant to indicate the limits of theoretical discourse. But since what belongs to theoretical discourse is what can be 'said' at all in language, the investigation of this limit is the investigation of the 'logic' of language, which shows the 'logic of the world'. 'Logic is not a theory but a reflexion of the world,' Wittgenstein says in 6.13, and adds: 'Logic is transcendental', which can be interpreted in this way: What Kant's transcendental deductions are intended to perform: this is performed by the logical analysis of language.

And this is the way in which thesis (a) is transformed in Wittgenstein's system. 'Philosophy limits the domain of scientific discourse' (4.113). 'It has to limit the thinkable and thereby the unthinkable. It has to limit the unthinkable from within through the thinkable. It will indicate the unsayable by clearly presenting the sayable' (4.114-115).

Wittgenstein moves the limits of theoretical reason to the limits of language. Whereas Kant thought 'possible to theoretical reason' to be a more narrow concept than 'logically possible', these two concepts are identical according to Wittgenstein. Therefore, what Kant calls the 'form of experience' is the common form of all logically possible worlds, that is, the 'logical form of substance' or 'internal structure of substance', which -- if we take logical atomism for granted -- is shown by the internal structure of language revealed by logical analysis. Thus thesis (c) holds true in this sense in Wittgenstein's system -- though we must include in the 'content' spoken of here also the external structure of reality (cf. Ch. VI, § 10).

Since the logical form of substance is prior to all experience it could be said to be a priori (cf. above, Ch. V, § 10). Thus Wittgenstein like Kant could regard statements on the 'form' as a priori, if there only were any such statements. But since the 'logical form' is the form of language there cannot be any meaningful sentences on the 'form'. Therefore (d) must be replaced by the thesis that the 'a priori' form of reality can only be exhibited by language but not expressed by sentences. It follows that thesis (e) is false. That a sentence is 'synthetical' means that its negation is logically possible, and thus a world in which the negation holds true is thinkable.

But what about mathematical statements? 'Mathematics', Wittgenstein argues, 'is a logical method' (6.2); its sentences 'are equations, and therefore pseudo-statements' (6.2) (cf. Ch. VIII, §5), i.e. they 'express no thoughts' (6.21). Perhaps we could interpret this as meaning that the mathematical statements are 'tautologies' in the sense that their descriptive content is empty, though they are not tautologies in the truth-functional sense,6 and thus they 'show' the logic of the world. And to the question whether we need Anschauung for the solution of mathematical problems it must be answered that language itself here supplies the necessary Anschauung (6.233).

As for the law of causality there is only one kind of laws a priori, and these are the logical laws (6.3, 6.31). The 'law of causality' is not a 'law' but the form of a law (6.32) -- that is, an empirical law can be called a 'law of causality' if it is of a certain type (6.321). That there are laws of this kind is a priori true only in so far as we can form hypotheses of this form in language (6.3211 f.). What a 'law of causality' ought to say if it is to be true a priori is that 'there are natural laws' (6.36) in the sense that we can give our hypotheses of 'connections' in nature the form of general laws -- and Kantianism is right in so far as this logical form is the only form under which connections in nature are thinkable (6.361) -- but this can in effect only be 'shown' but not 'said' (6.36). And since what is 'shown' here by language is 'shown' by all description of 'natural connections' what the 'law of causality' is intended to exclude cannot be described (6.362)7.

To sum up: it is essential to Wittgenstein's outlook that logical analysis of language as he conceives of it is a kind of 'transcendental deduction' in Kant's sense, the aim of which is to indicate the a priori form of experience which is 'shown' by all meaningful language and therefore cannot be 'said'. From this point of view the Tractatus could be called a 'Critique of Pure Language'8: 'The book will . . . draw a limit to thinking, or rather -- not to thinking, but to the expression of thoughts; for, in order to draw a limit to thinking we should have to be able to think both sides of this limit (we should therefore have to be able to think what cannot be thought). -- The limit can, therefore, only be drawn in language and what lies on the other side of the limit will be simply nonsense' (Preface).


Kantianism has been called 'Critical Idealism' or 'Transcendental Idealism'. This terminology emphasizes the fact that Kantianism is an 'idealistic' philosophy in the restricted sense that the form of experience, but only the form is imposed on the world of experience by the structure of the human mind. In a way the form of experience can therefore be said to be 'subjective', but 'subjective' only in a 'metaphysical' or 'transcendental' sense, since the 'subject', the 'ego', as we ordinarily conceive of it, is not this 'metaphysical' or 'transcendental' subject but an object in the world of experience, that is, an object in a world which already has the 'form' prescribed by the mind of the metaphysical subject. With regard to this terminology Wittgenstein's philosophical system could be called 'Critical Lingualism' or 'Transcendental Lingualism' or even 'Lingualistic Idealism'. For Wittgenstein, too, the form of experience is 'subjective' in the transcendental sense, the metaphysical subject

being the 'subject' which uses and understands language, and which must be distinguished from the empirical self, which is part of the world describable in language. Therefore, 'the limits of my language mean the limits of my world' (5.6). The limits of the world of the metaphysical subject, or rather, the limits of the metaphysical subject's 'logical space' of possible worlds, is determined by the limits of his language.

The real world of science is the world as describable in meaningful language, and this is, in a sense, 'my world:'

5.62 ........
Was der Solipsismus nämlich meint, ist ganz richtig, nur lässt es sich nicht sagen, sondern es zeigt sich.

Dass die Welt meine Welt ist, das zeigt sich darin, dass die Grenzen der Sprache (der Sprache, die allein ich verstehe) die Grenzen meiner Welt bedeuten.

Like Russell9 Itake the word allein in the second paragraph as referring to the relative pronoun die,10 and think the passage could be translated as follows: 'That the world is my world, shows itself in the fact that the limits of Language (which is the only language I understand) mean the limits of my world.' The parenthesis can be interpreted in this way: since the limits of language are the transcendental limits of the world and since Language is my language (the only language which I understand) the limits of Language are the limits of my world. The Ego to which the word 'my' refers here is the metaphysical subject, which in a sense, does not exist (5.631), because it is transcendental, it 'does not belong to the world but it is a limit of the world' (5.63 2) -- it is like the eye in relation to the field of sight; the eye cannot see itself (5.633-6.331). And the metaphysical subject is to be distinguished from the empirical ego: in the book 'The world as I found it' there is an ego which must be described, but of the metaphysical ego there could be no mention in this book (5.631). The world is my world if the word 'my' refers to the metaphysical ego, and that this is so implies that 'solipsism'11 can be considered true in a sense; that is, what solipsism intends to say is quite correct, but -- since we cannot speak of the metaphysical ego -- this cannot be said. That the metaphysical ego is transcendental implies also 'that solipsism strictly carried out coincides with pure realism. The Ego in solipsism shrinks to an extensionless point and there remains the reality co-ordinated with it' (5.64).


The limit to thinking drawn by Wittgenstein's transcendental deductions can, as we have seen, properly be drawn only in language. We cannot think what is unthinkable, but we can form linguistic expressions which do not express thoughts, since they are simply nonsensical. Thus the limit between what in Kantian terms belongs, and belongs not to theoretical reason is shown by the logical distinction between sense and nonsense.

It follows that problems which according to Kant are unsolvable by theoretical reason cannot even be raised; 'the deepest problems are properly speaking not problems' (4.003, italics mine). 'For an answer which cannot be expressed the question too cannot be expressed. -- The riddle does not exist' (6.5).

There are no 'transcendent' statements, because what they try to say cannot be said. But this does not mean that our Kantian thesis (f) is wholly wrong. 'There is, however, the inexpressible', Wittgenstein admits in 6.522. 'This shows itself; it is the mystical.' Thus there is indeed a domain of 'practical reason', but this lies outside what is expressible in language. God belongs to this domain: 'God does not reveal himself in the world' (6.432). Ethics lies also beyond language. One cannot (as Kant did) speak of the 'will' as the subject of ethics, because the 'will' of which one can speak 'is only of interest to psychology'(6.423).12 'It is clear that ethics cannot be expressed. Ethics is transcendental' (6.421) -- I think Wittgenstein would rather have said 'Ethics is transcendent' if he had adopted the above distinction between 'transcendental' and 'transcendent'.13 As for the immortality of the soul, 'the solution of the riddle of life in space and time lies outside space and time' (6.4312).

In a logical empiricist's vocabulary 'nonsensicality' means something purely negative. Wittgenstein's identification of the inexpressible with the mystical seems to show that to him 'nonsensicality' has a rather positive ring. The German word Unaussprechlich means not only 'inexpressible' but also 'ineffable'. Thus the identification in 6.522 of das Unaussprechliche with the mystical could be interpreted as a statement of the old thought that the mystical is ineffable.

The inexpressible in this sense 'shows itself'. In Chapter X, section 8, we made a distinction between two kinds of things that cannot be said: that which can be shown in language but not said, and that which can be neither shown nor said in language. The mystical belongs, I think, to the latter kind of inexpressible things. It does not reflect itself in language. How does it then express itself? I believe this is to be understood as a matter of feeling. One 'experiences' the mystical as a form of emotional experience which in German would be called Erlebnis in contradistinction to ordinary fact-stating 'experience' that is called Erfahrung. 'Die Anschauung der Welt sub specie aeterni ist ihre Anschauung als -- begrenztes -- Ganzes' (6.45). I think this ought to be translated in this way: 'The contemplation of the world sub specie aeterni is the intuition of it as a -- limited -- whole'. And Wittgenstein adds: "The. feeling of the world as a limited whole is the mystical feeling' (italics mine). One could perhaps characterize this feeling as an emotional experience of the world from what one feels as God's point of view. This has nothing to do with 'facts' describable in language: 'The facts belong all to the task (Aufgabe) and not to its performance (Lösung)' -- that is, to the inexpressible problem, not to its solution (6.4321).

This argument leads us to the idea of the transcendent Ding an sich (thesis (g)). The 'things' of Wittgenstein's logical atomism belong to the framework of a world description, and could therefore be called 'transcendental' -- and this would actually conform with one aspect of Kant's own view. But the Ding an sich which is thought of as existing independently of the form of experience is not this transcendental thing, but appears in Kantianism, and in particular in Schopenhauer's philosophy, as a symbol for the unreachable transcendent. And I think there is a reminiscence of this idea in the Tractatus.

In the second chapter I asked myself why Wittgenstein gives his first thesis in the Tractatus the form of an answer to the question what the world is. I believe this form is connected with the old philosophical feeling that a what, if it could be answered, would bring us into a mystical contact with the world whereas an answer to a how only takes the form of an impassive description. 'Logic precedes every Erfahrung -- that something is so', and therefore 'it is before the How', but it is 'not before the What' (5.552). 'How the world is, is completely indifferent for what is higher' (6.432). 'Not how the world is, is the mystical, but that it is' (6.44).

We have criticized Wittgenstein's logical atomism in many respects. Does this criticism essentially affect his Kantian metaphysics? Even if we admit that logical atomism is wrong, even if we admit that there are many ways of saying things, and that the idea of a 'depicting' language can be extended in many directions, cannot the distinction between what, on the one hand, can be said or shown in language, and what, on the other hand, lies entirely outside language always be reconstructed, though the basis for this distinction must be more complicated than that assumed by logical atomism?

It is difficult to give a definite answer to this question. But I share the opinion of many others that a distinction of this kind does not in any case answer to the demands made on it in the Tractatus. So, for instance, the difference between ethical judgements and scientific statements is not a difference between a transcendent reality and a reality of science, but a difference between two kinds of problems corresponding to different moods of language. What is of lasting value in the Tractatus is not the philosophical system which is its alleged result, but the views proposed in the different steps of the argument 'leading' to it, that is, of the ladder which according to 6.54 is to be thrown away after one had climbed up on it.

In fact the inconsistencies in the Tractatus are in the main inconsistencies between the philosophical system and the investigations performed in the different steps. And in view of this fact the following point in Wittgenstein's attitude to nonsense is of interest. Since Wittgenstein believed our deepest problems to lie beyond language one could expect him to extend his conception of 'philosophy as an activity' in such a way as to comprise practices of an Indian kind for the attainment of a contact with the mystical Ineffable. That his philosophy did not take this turn is significant. If my second-hand impression of Wittgenstein's personality is correct he did not in fact have any strong inclination to mysticism; at least it was counterbalanced by an opposite tendency. Logical positivists did not simply invert Wittgenstein's attitude to nonsensicality, because it is in fact ambiguous in the Tractatus. On the one hand the 'inexpressible', as we have seen, has a positive ring, but on the other hand Wittgenstein seems to share the positivistic tendency to regard it as nonsense which does not deserve our attention.

In 6.5 2 we read: 'We feel that even if all possible scientific questions be answered, the problems of life have still not been touched at all.' This conforms with the positive evaluation of the inexpressible. But the positive evaluation becomes dubious when Wittgenstein adds in 6.52 and 6.521: 'To be sure, there is then no question left, and just this is the answer. -- The solution of the problem of life is seen in the vanishing of the problem.' And when he states in 6.53 that the only correct method in philosophy is to demonstrate to anyone who 'wishes to say something metaphysical that he had not given any meaning to certain of the signs in his sentences', then we have a definite feeling that what is inexpressible is just nonsense and nothing else. Wittgenstein says not merely 'The riddle is inexpressible', but 'The riddle does not exist'.

'Whereof one cannot speak, thereon one must be silent.' As a matter of fact this is not a reverence for the ineffable. It could rather be characterized as the expression of a way of escape. When Wittgenstein determined to be silent he turned away from philosophy and tried to enter an active life. The philosophical activity neither meant to him an amusing occupation of the mind, nor was it a means for establishing some independent doctrine of life or reality; it was a passion from which he could never free himself. He did not feel himself as a captive of Language which bars the contact with what is ineffable; but he felt himself indeed as a captive -- in philosophy. And therefore the essential aim of philosophy was to find a way out from itself.

I have stated earlier that the 'nonsensicality' of all philosophical statements according to the Tractatus did not imply that philosophy as an activity is aimless. But I think one ought to add to this that an essential aim of the philosophical activity in the Tractatus actually was to make philosophy aimless. The 'definitiveness' of the truth of the thoughts expressed in the book thus meant that Wittgenstein considered this aim to be reached -- at least for his own part.

Later he changed his mind in this respect. But he made no essential change in his attitude toward the aim of philosophy. The main error in the Tractatus in this was the endeavour to find a once-for-all remedy for the philosophical disease. Wittgenstein discovered that one could find means of interrupting the philosophical activity when one wants to, without finding a method for stopping it altogether. Or to quote his own words in the Philosophical Investigations, § 133:

. . . For the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear.

The real discovery is the one that makes me capable of interrupting doing philosophy when I want to. -- The one that gives philosophy peace, so that it is no longer tormented by questions which bring itself in question. -- We demonstrate a method, by examples; and the series of examples can be broken off. Problems are solved (difficulties eliminated), not a single problem.

There is not one philosophical method; but there are indeed methods, like different therapies.

This passage, in its formulation strangely intermediate between the statement of a personal predicament and the establishment of a rule for general acceptance, can be regarded as a pathetic expression of Wittgenstein's struggle with the problems of philosophy. And it gives us an idea of the form in which the belief in the nonsensicality of philosophical statements is retained in the thought of the later Wittgenstein.

'All philosophy is "Critique of language"' (4.0031). For my part I do not believe this thesis to be an exhaustive characterization even of what ought to be the aim of philosophical analysis. But in so far as philosophy is critique of language it is an investigation which must be carried out step by step like investigations in science. This does not mean that philosophical investigations are 'empirical' -- their result is indeed 'clarity' rather than 'knowledge' in a scientific sense: 'The word "philosophy" must mean something which stands above or below, but not beside the natural sciences' (4.111). But we cannot content ourselves with considering the results of philosophical analysis as inexpressible. We have to find means of expressing them, and expressing them with increasing clarity, unless the philosophical activity is to remain an eternal vicious circle.


1 According to von Wright ('Wittgenstein', p. 543) Wittgenstein said that he could get only 'occasional glimpses of understanding' from Spinoza, Hume, and Kant.

2 See von Wright, /.a, p. 530. Concerning the influence of Kantianism on Wittgenstein, cf. also above, p. 87 n.

1 This dichotomy was in fact already made in the Dreams of a Visionary Explained by Dreams of Metaphysics, which has been regarded as the first of the works from Kant's 'critical' period, and can thus be said to form a starting-point of his critical philosophy.

11 use the term 'Form of Perception' (Form der Wahrnehmung) for what Kant calls either Form der Anschauung or Form der Sinnlichkeit, in order to indicate the analogy between this concept and the Kantian concept 'Form of Experience' (Form der Erfabruttg) mentioned below.

a The distinction between the terms 'transcendental', which refers to the limits of theoretical knowledge, and 'transcendent', which refers to what falls outside this limit, is hinted at in the Critique of Pure 'Reason, but not consistently upheld. It is, however, clarifying to use the terms in this way.

1 I.e., equations may be regarded as tautologies in sense (a) of p. 76, though they are not so in sense (b). According to Ramsey (foundations of Mathematics, p. 17) this was Wittgenstein's view, whereas Ramsey himself tried to show that all mathematical propositions are tautologies also in sense (b). Cf. also Waismann: 'Identitac', p. 63.

11 think this argument is confused in a way which answers to a corresponding confusion in Kant's thinking.

21 owe this expression to Mr. P. Geach,

1 See his introduction to the Tractatus, p. 18.

8 Cf. Hintikka: 'On Wittgenstein's "Solipsism",' p. 88. 'Wittgenstein's formulation is, to be sure, ambiguous, but I suppose that Wittgenstein, if he had meant 'the language, which I alone understand' (Urmson: Philosophical Analysis, p. 135), would have written die nur ich verstehe. Moreover a thought of this kind is unparallelled in the Tractatus and would not fit into the context. On the interpretation of what Wittgenstein wants "to say here, cf. also Rhees' argument in his review of Cornforth, pp. 388 f.

s According to Hintikka Wittgenstein does not mean by 'solipsism' what is ordinarily meant by it. I think he is right at least in so far as Wittgenstein ought to have used the word 'idealism' rather than 'solipsism', because what he calls 'solipsism' is exactly his linguistic turn of Kantian idealism. Urmson seems to understand 'solipsism' as the doctrine that there are no other minds instead of the doctrine that there is no external world, but this peculiar meaning of 'solipsism' can hardly have been known to wTttv genstein when he wrote the Tractatus; it is in fact unknown to all philosophical dictionaries I have consulted.

1 This implies also that the Problem of Free Will cannot be treated in the spirit of Kant. Wittgenstein hints at his attitude to this problem in 5.1362.

2  In the world describable in language there are no (objective) values, since all sentences are 'of equal value' (6.4).