,14=/,15=#,23=%,24==> 9Tractatus 1921{-22 T~H~E book we treat separately because it had a career of its own. In its content we shall see that it mirrored its author's career to that time. Its mode of publication also had about it something characteristic of him: he left his work for others to decide on and to do as they would with. It was to be the same with his later writings. All was to be as God willed, as he said earlier when he had his manuscripts sent to him through the field post. The Martha-like task of arranging publication was left in the hands of Dorothy Wrinch, a young woman already mentioned as a prote$0ge$0e of Russell's. Russell himself left for China in the autumn of 1920 and did not return for a year. She set about the task with competence.1 Cambridge University Press (probably the first to be approached) sent its negative reply on 17 January: no publishing house is free of vain regrets, but Cambridge, parvi mater amoris in this respect, was to reject Wittgenstein's later work also. Miss Wrinch turned to Germany, apparently first sending no more than a description of the work, which was short enough to constitute a longish article, to a selection of periodicals interested in a scientific approach to philosophy, or in the view that philosophy contained the essence of the sciences. Schumann of the Zeitschrift fu$3r Psychologie und Physiologie der Sinnesorgane stipulated, however, that the work should be psychological rather than philo- sophical, thus ruling it out; Ludwig Stein of the Archiv fu$3r systematische Philosophie was prepared to take an article by a pupil of Russell's, but there would be some delay in printing, could Miss Wrinch write again in May? These letters were both dated 12 February. Wilhelm Ostwald of the Annalen der Naturphilosophie dated his letter 52.21, one of his many quirks (21 February was doubtless meant and he apologizes for his delay in replying): In any other case I should have declined to accept the article. But I have such an extremely high regard for Mr. Bertrand Russell, both for his researches and for his personality, that I will gladly publish Mr. Wittgenstein's article in my Annalen der Naturphilosophie: Mr. Bertrand Russell's Introduction will be particularly welcome. 1.Details in the Russell Archive under W.51, "Wrinch'. See also von Wright's "The origin of the Tractatus' already quoted. Miss Wrinch was to become a distinguished spectroscopist. 297 9Tractatus 1921{-22 It would be published in a couple of months: offprints would be available, perhaps earlier. Closing with this offer Miss Wrinch sent the typescript (Ostwald acknowledges it on a card postmarked 10 March 1921) along with a copy of Russell's introduction, which was once again translated, rather roughly, into German.2 The work duly appeared and (conformably to its message) this turned out to be the last number of the periodical. "I consider this a pirated edition: it is full of errors,' said Wittgenstein _himself, as so often happens, committing a pair of errors in his complaint.3 But there is justice in that complaint: the proofs were read only in Germany with little understanding of the subject matter and the printer, pardonably, reproduced rather than interpreted various type- written devices that stood proxy for Russell's and Sheffer's logical notation. In one or two places instructions to the typist, or even handwritten queries by Russell, were incorporated in the text.4 Still the work was intelligible enough to a first generation of continental readers, and anyone who picks up a copy of this fragile fascicule, poorly printed on yellowing paper like so many post-war publications from Germany, will feel in touch with the time and the circumstances of the work itself. If the comparison be permitted, it is like handling a Shakespeare quarto_and of as little use for establishing a better text. These were, all the same, strange auspices and defenders for Wittgenstein: Russell, whose introduction he disapproved of, and Ostwald, who seemed to unite many of Wittgenstein's greatest aversions. In the middle of a distinguished career establishing physical chemistry as a fundamental subject and publishing an indispensable series of scientific classics, he had retired to a large house (actually called "Landhaus Energie' in honour of his King Charles's head) to promote a whole series of causes: pacifism, a universal language, scientific monism (on which he published so called Sunday sermons) and others still. Wittgenstein styled him "an utter charlatan' and feared that he might put the book into "his idiotic spelling', another fad.5 Still, such are the fates of books, such is, almost, the "normal way'6 in which a work is accepted by the world. The English edition which followed also had a contingent origin: Miss Wrinch was a friend of C.K. Ogden, founder in 1912 of the Cambridge Magazine, highly regarded 2.The earlier translation presumably remained in Austria: it is now lost. For variations between the version of the introduction published by Ostwald and the subsequent English publication, see M. Teresa Iglesias in Russell 25{-28 (1977). 3.LW to PE 5.8.22: "voller Fehlern' confuses two German constructions; and of course the publication was in no sense a "Raubdruck'. 4.At 5.535, for example, where Wittgenstein claims to have solved all the problems that Russell's axiom of infinity brings with it, a pencilled question-mark after the word "Alle' (probably a demurrer by Russell) was duly printed. 5.LW to BR 28.11.21 (written, though not knowingly, after the publication in the Annalen). 6.Wittgenstein's phrase in his letter to Ficker, mid-October 1919, already quoted. 298 Wittgenstein: A Life intellectually before the war and an enormous success during and after it. Wittgenstein was known to Ogden from the Russell circle and the Moral Science Club of pre-war years. Recently Ogden had put the results of his studies of language teaching to use in The Meaning of Meaning, written together with I.A. Richards, for its time an unusually concrete psychological and sociological account of language. At this point he was launched on a new publishing career: as editor both of Psyche, an international journal with a central interest in language, and of The International Library of Psychology, Philosophy, and Scientific Method published by Kegan Paul, whose very title conveys its manifesto. Jung, Vaihinger, Russell, Moore, and Hardy's Mathematics for Philosophers figured among the authors and books projected. The new sciences were to be brought to the aid of British culture, Cambridge voices were to be heard alongside the last word from Vienna or Trieste. The aftermath of the war seemed to promise a new science, a science of man. The interest above all in language was its hallmark. These ambitions of Ogden and his friends may fairly be said to have been realized. This was the man to whom Dorothy Wrinch showed Wittgenstein's treatise shortly after the rejection from the Cambridge Press arrived.7 He took no immediate action, so she went the rounds of Germany, with what result we have seen. Still, Ogden was more perceptive, showed more Spu$3rsinn, than Ficker and in due course he put to Russell the idea of publishing the work. Russell scrupulously sought a second authority to publish from Wittgenstein. This was of course forthcoming and Wittgen- stein in fact collaborated, marking up corrections in a copy of the Annalen publication, reading the proofs supplied by Ogden and also commenting on the translation as it was sent him. The translation seems principally to have been the work of Frank Ramsey, who in a short life was to make considerable contributions both to economics and the philosophy of mathematics. Not yet twenty years of age, but already recognized as a genius, son of an academic family, an Apostle, of course, and a prote$0ge$0 of Keynes's, he was on the point of leaving mathematics for the moral sciences, to give them their Cambridge name, and Trinity for King's. It was perhaps known that he had won at Winchester a German prize intended for non-specialists. Braithwaite describes him as sailing into Miss Pate's typing office in Cambridge and dictating the translation with Ogden at his side.8 Russell's own version was used for those passages originally cited at length by him: Wittgenstein insisted that these translations had equal authority with the German text. There was also much consultation with Wittgenstein himself, conducted by Ogden. The idea of publishing the original with the translation en face was Russell's 7.So she told me, and it explains how Ogden was able to approach Russell about the book in the autumn following. 8.Personal communication. 299 9Tractatus 1921{-22 but Ogden readily embraced it as an intelligible solution to a difficult problem. It was almost impossible, he said in his editor's note, to convey the literary and epigrammatic flavour of the original in any other way. It is a nice question whether in their attempt to convey this flavour he and Ramsey overdid it or not. A whole generation of English-speaking philosophers came to know the work through a translation which seems to have been not so much liberated as shackled by the presence of the German on the opposite page. It reads as if made from a dead language, though that may be thought to assist the stylistic effect of a revelation which is certainly one of the aims of the original. A smoother, and perhaps more philosophical version has been attempted by the present writer with a colleague.9 Almost as influential as the translation was the title chosen, which even in German sometimes supplants Logisch-philosophische Abhandlung. That, the original title, seems to me deliberately chaste: "a treatise_as it might be one of many_that goes about philosophy in a logical way'. It is understated in a way that the book itself is not. The first suggestion for a translation was Philosophical Logic, which indeed appeared in Kegan Paul's advertisements. This, of course, gets things just the wrong way round, and Wittgenstein poured scorn on it: I don't know what it means! There is no such thing as philosophic logic. (Unless one says that as the whole book is nonsense the title might as well be nonsense too.) (LW to CKO 23.4.22.) So instead Moore's suggestion of Tractatus Logico-Philosophicus was accepted: "not ideal,' Wittgenstein said, "still it has something like the right meaning'. The allusion to Tractatus Theologico-Politicus, a work of Spinoza's more familiar for its title than for its content, was perhaps not altogether a happy idea: it is a work censured by Schopenhauer for its optimism and Wittgenstein will hardly have known it. He does indeed echo a phrase of Spinoza's (not from this work) when he equates the mystical view of the world as a limited whole with the view of the world "sub specie aeterni',10 but the context in his Notebooks suggest that, as with many of his allusions, it is Schopenhauer's quotation of Spinoza that he is echoing here rather than the original.11 The new title reinforced the tendency of the translation to underline the dogmatic and quasi-biblical tone of Wittgenstein's work. There is, as has been suggested above, some justification for that: the work begins with a sort of creation-myth: 9.With D.F.Pears, 1961. In 1971 a second edition was published taking account of Wittgenstein's own wishes as expressed in the correspondence with Ogden (also published 1971). 10.TLP 6.45. 11.Cf. Notebooks 7.10.16 with Die Welt als Wille und Vorstellung vol.II (Frauensta$3dt) p.211. 300 Wittgenstein: A Life The world is everything that is the case. The world is the totality of facts, not of things. (1{-1.1) and ends with a mystical adjuration to silence in the face of the ineffable, as it were with a form of negative theology: Whereof one cannot speak, thereof one must be silent. (7) (Here, to give the effect, I use the first translation.) Also there is the impression of something revealed, as if some sage, some Zarathustra, were speaking. There are, to be sure, explicit arguments (here, as normally, I use the newer translation): Objects make up the substance of the world. That is why they cannot be composite. If the world had no substance, then whether a proposition had sense would depend on whether another proposition was true. In that case we could not sketch any picture of the world (true or false). (2.021{-2.0211) But the framework of assumptions within which these arguments are effective is not clearly set out, and their importance in the whole scheme is left implicit in the system of numbering, which no one has yet completely elucidated. That is not to say that the work is without structure. On the contrary it has a carefully organized one, which will be described shortly; but, in detail, it retains many of the features of the Zibaldone, the series of notebooks of aphoristic paragraphs from which it is drawn. In them, as in the collections of aphorisms by Schopenhauer and Lichtenberg, which must to some extent have served as a model, there is only a very loose connexion between the topics treated in the course of a day or a section. The writer may feel that there is some inner unity: the reader, at all events, is left to guess at it_nay more, the reader is encouraged to think that there is here a whole world-view, a way of thinking and feeling, from which all that is said will seem obvious. The aphorism that follows our last excerpt illustrates this: It is obvious that an imagined world, however different it may be from the real one, must have something_a form_in common with it. (2.022) It is a challenge to the reader to say that this is obvious. The whole idea that different possible worlds vary in form not in content is one that Wittgenstein is trying to impose on us: it is precisely not something obvious in itself. We, to be sure, in the course of this book have seen the 301 9Tractatus 1921{-22 sort of reasons that led Moore to think that the furniture of the world consisted of concepts and Wittgenstein to amend this into the view that it consisted of asserted propositions or facts. Perhaps there is an initial plausibility in the notion that any alternative to the actual world must differ in some describable way and hence would render true a different set of propositions; but the thoughts that at the same time there will be a substance, as it were the substance of all possible worlds, which remains untouched, and further that this substance is a kind of form or pattern for all possible worlds, are by no means self-evident. By telling us, none the less, that they are so, Wittgenstein is in part insisting that his reader must think through everything with him, so that it becomes obvious. He gives us warning of this when he says in the first sentence of his preface: Perhaps this book will be understood only by someone who has himself already had the thoughts that are expressed in it_or at least similar thoughts._So it is not a textbook. It is a further and very characteristic piece of irony that the book is in fact written and arranged like a textbook. The numbering of the propositions mimics, as we saw when discussing the last stages of the genesis of the book, the logical ordering of Principia Mathematica (and in general that of any treatise arranged on mathematical or Euclidean lines). But the principle of arrangement actually being followed is by no means clear. We are being told, by the literary form and its contrast with the content, that things are, and are not, as simple as they seem in this presentation. But in these propositions about the form of the world, paradoxically said to be obvious, Wittgenstein is also indicating that there is something odd about this very ontology of his (as perhaps about any ontology). And indeed the objects that make up the substance of the world must be very different from the solid and complex objects, with their properties and relations, that make up the world of our acquaintance. His objects are, rather, simply designed to be possible constituents of states of affairs. In another ironic, and to some extent self-refuting, statement early in the book, Wittgenstein says, Nothing in the province of logic can be merely possible. Logic deals with every possibility and all possibilities are its facts. (2.0121, in part) Again we see that Wittgenstein's whole approach, his literary style, has as part of its aim to tell us that he may be only seeming to be uttering straightforward propositions about palpable objects_trattando l'ombre 302 Wittgenstein: A Life come cosa salda.12 The status of his own propositions is always in his mind. He was right to say to Ficker in a letter we have quoted that the work was strictly philosophical and at the same time literary. One aspect of its literary character is that, like a poem, it is not an indifferent vehicle for something expressible in other ways but shows or conveys something unique by its own form of expression. In Wittgenstein's own terms, it shows or manifests something_sie zeigt etwas, as we find him saying in the last sentence of the preface. The interesting thing, and the real riddle of the Tractatus, is that at the same time it does give expression to thoughts, claimed to be incontestably true. So the work is philosophical as well as literary, but it is in the nature of philosophy (according to the Tractatus) that its thoughts are ones whose point is to turn us away from philosophy. Thus, while Wittgenstein is convinced that his work contains, in all essentials, the final solution of the problems treated13 and that its value in the first place consists in how well this is expressed, the very fact that he has the final solution enables him to say in the last sentence of his preface: And if I am not mistaken in this belief, then the second thing in which the value of this work consists is that it shows how little is achieved when these problems are solved. A strange book, half of whose value consists in showing its own unimportance! The reader will, I believe, find it invaluable to keep these literary characteristics of the work in mind as he reads it. The aphoristic style and the sometimes puzzling transitions of thought are part of its commentary on itself. They show something about the nature of philosophy_and that after all is the subject of the work. There are plenty of internal indications that its propositions are not to be taken quite au pied de la lettre. It is, of course, not the only philosophical work with such an aim. The self-doubt of the philosopher is to be found in many sceptical writings. Sextus Empiricus is echoed here, who tells us that, according to the Pyrrhonian, scepticism is not a set of dogmas but a way of life, and Plato devotes the last part of the Phaedrus to a demonstration that written works (and hence the very dialogue which teaches this) are not the way to convey the lessons of true philosophy, which is found only in the living speech and written in the soul of the participant. It is in the same spirit that Wittgenstein says: 12."Treating these shadows as material things', as Statius did Virgil's shade in Dante (Purg. 21.136). 13.Wittgenstein here echoes Frege's preface to The Foundations of Arithmetic p. v: "I hope to settle the question finally, at least in essentials'. 303 9Tractatus 1921{-22 Philosophy is not a body of doctrine but an activity. Philosophy does not result in philosophical propositions but rather in the clarification of propositions. It will signify what cannot be said by presenting clearly what can be said. (4.112, in part, and 4.115) The idea of "signifying' is important here. To be sure it is only one possible translation of bedeuten_Ogden and Ramsey had "mean'_but at all events the idea is that nothing can be said. Philosophy is a sort of knowledge which can be presented only obliquely, as Heraclitus says: The Lord whose oracle is in Delphi neither speaks out nor conceals, but gives a sign. (DK 22 B 93) And to make the parallel with Plato even more striking, Wittgenstein seems to condemn the method of his own book: The correct method in philosophy would really be the following: to say nothing except what can be said, i.e. propositions of natural science_i.e. something that has nothing to do with philosophy_and then, whenever someone else wanted to say something metaphysical, to demonstrate to him that he has failed to give a meaning to certain signs in his propositions. Although it would not be satisfying to the other person_he would not have the feeling that we were teaching him philosophy_this method would be the only strictly correct one. (6.53) It is not unfair to see his whole philosophy as a kind of mystic revelation, remembering that mystic means what cannot or should not be spoken. He himself says so just before this last proposition: There are indeed things that cannot be put into words. They make themselves manifest. They are what is mystical. (6.53) The reading of the book, then, has a purpose: it is like an initiation into the mysteries, and when they are reached it can be forgotten. These are, of course, only prolegomena; and while the reader of the present volume will not expect a substitute for reading the Tractatus, he may be surprised at an introduction that lays so much emphasis on its hidden content. The fact is, however, that a direct description of its doctrines (which many of my predecessors have attempted_vestigia terrent)14 often leads to surprise at the sudden turn in the argument by which those doctrines are themselves invalidated. The edifice erected in 14.Norman Malcolm in Nothing is Hidden, 1986, (especially chapter 2) at least discusses views like mine, though he reaches an opposite conclusion. Earlier commentators often end up trying desperately to put the bathwater back with the baby. 304 Wittgenstein: A Life the work must be examined, but with the sort of awareness of its insubstantiality which I hope to have given here. The student will do well, I think, to look first at what purport by their numbering to be the main theses, those numbered with whole numbers or at most one decimal figure. These, as we have already noted, were mostly written at the beginning of the Prototractatus noteboook: the decimal parts of 5 and 6 and the final proposition, 7, being added in the course of composition. The aim should be to note how the work is structured round them, and the student must not, of course, suppose that what comes first is what is most basic, still less most intelligible. It is possible, in fact, to defend the view that the work is a kind of systole and diastole around proposition 4, where the most basic propositions of its argument are to be found. Wittgenstein opens then (we have quoted the first two capital propositions already) with the idea that the world is all that is the case. It emerges in comment (hence the number, 1.1) that this means that the world consists of facts, not of things, and further (1.2) that these facts divide the world up, i.e. that they are independent. These comments on 1 lead up to, and in some measure are seen as arguments for, proposition 2. This is a common feature and perhaps part of the point of the system of numbering. Proposition 2 runs: What is the case_a fact_is the existence of states of affairs. The thought is that a fact (by definition actual) presupposes one or more potentialities that are realized. This in turn is only conceivable if there are ways of grasping these potentialities independently of their realiz- ation. In other words, in order to be able to affirm any fact we must have access to a magazine of possibilities, which we grasp mentally, and that indeed is proposition 2.1: We picture facts to ourselves. Note that the argument constantly turns on the human ability to affirm (and hence to understand) propositions. We have already quoted a brief subsidiary argument (2.021{-2.0212) where, in similar fashion, the necessary existence of simple objects (which is roughly equivalent to the existence of that determinate magazine of potentialities I have spoken of) is proved by our power to imagine a world at random, i.e. to be sure that we are describing a possibility before we know any facts. The features that enable pictures to represent can be generalized (and are throughout the comments on 2.1) into the demand that any picture should have formal properties in common with what it depicts, and these properties are (in proposition 2.2) called, really by prescriptive 305 9Tractatus 1921{-22 definition, logico-pictorial form. This means that every picture must display a possibility regarding the world (and indeed a contingency_a possibility whose opposite is also possible). Conversely something will be a possibility in the world if and only if it can be represented by a picture. It follows (after the 2.2's) that: A logical picture of facts is a thought. (3) Only that can be thought which satisfies the minimal conditions that we have just laid down. The next step perhaps needs no argument: it is 3.1 In a proposition a thought finds an expression that can be perceived by the senses.15 That is, it is only in language that we can identify a thought, and thus (as we shall ultimately see) all that is logically true of language will also define for us the limits of what can be thought and what can exist. For the present Wittgenstein is concerned with the articulation of the proposition as such. Not a set of signs, but only the fact that signs are disposed in a certain way can express a sense. This has the important consequence that: In a proposition a thought can be expressed in such a way that elements of the propositional sign correspond to the objects of the thought. (3.2)16 The importance of this for Wittgenstein is that it brings out the possibility in principle of analysing every proposition so that it consists of primitive signs which stand for objects. These_names, as he calls them_are correlative in status with propositions: each requires the other: Only propositions have sense; only in the nexus of a proposition does a name have meaning. (3.3) Wittgenstein here, in the second clause, echoes a sentiment of Frege's and uses it with his accustomed irony to controvert a favourite doctrine of Frege's, that both names and sentences have both sense and reference. The new thought is, that implicit in our simplest linguistic abilities is the existence of a level of language at which there could (theoretically) be seen the two interdependent moments or factors, one being the fact of 15.The Prototractatus at this point has: "The perceptible expression of a thought is a propositional sign.' The final work is less careful to indicate when (as here) the sign is meant rather than the "proposition' expressed by it. 16.The single decimals under 3 are slightly different in Prototractatus, though essentially the same line of thought is followed. 306 Wittgenstein: A Life combination, the other being that which is so combined_function and argument these are called in Wittgenstein's semi-technical vocabulary.17 With them, equally present in every judgement, is the key to all that can be done with language. We are here speaking, it must be understood, of the theoretical limit of analysis. It is not pretended, it need not be pretended, that we can reach that limit. What is being said, through the remainder of the 3's, which we shall not rehearse in detail, is that the notions of representation, of sense, of possibility, of necessary con- sequence can only be anchored in our grasp of what is implicit in a proposition (we remember how central the problem of the proposition was in Wittgenstein's wartime notebooks). Thus when he says as the culmination of this section: A thought is a proposition with a sense. (4) he is drawing the conclusion that the laws of the sense of propositions are the laws of thought generally. No wonder the comment immediately follows that philosophy is nothing but critique of language.18 We are here at the heart of the Tractatus. The essence of its account of propositions is contained in the propositions number 4.01{-4.0641, which lead off with the picture theory: A proposition is a picture of reality. A proposition is a model of reality as we imagine it. (4.01) It is an interesting circumstance that nearly all the arguments in this section were developed before the idea came to Wittgenstein that propositions might be seen as pictures. One might say that nothing hinges on this way of seeing them. The picture paradigm is a convenient way of summing up what must logically be true of a proposition. As so often, what looks like a premiss is a conclusion. The real premisses are, for example, that we can understand the sense of a propositional sign without its having been explained to us (4.02). The proposition can do this only by having the correct multiplicity, only by representing, implicitly or explicitly, all the elements of reality that are relevant. It must, as it were, (this is an extension of one of Wittgenstein's own metaphors) build a model with bricks in the certainty that sufficiently similar bricks are available in reality. If it does this, the proposition will itself be enough to guarantee the possibility of 17.See 5.47, later in the book, for a very clear statement of this. 18.4.0031. "Not in Mauthner's sense,' says Wittgenstein, because Mauthner, in his large work with the title Critique of Language, set out to show various ways in which language misrepresented thought and reality, whereas, according to Wittgenstein, their general nature is identical with the possibilities afforded by language. 307 9Tractatus 1921{-22 what it says, because (i.e. this is equivalent to the fact that) it has sense. And this means that there is a further stage, after the understanding of the proposition, which is the comparison of that proposition with reality (4.5). All this talk about sense and possibility, where we might wonder how Wittgenstein knew with such certainty what were the correct relations between these concepts, turns out to be a way of saying (or exploring the meaning of the fact that) a proposition is essentially something capable of being true or false. This is precisely what Wittgenstein told Whitehead in 1913: A proposition has two poles, "a' and "b' and when asked, mildly_for Whitehead was not without his satirical side_what "a' and "b' might be, replied "in a voice of thunder': "a' and "b' are indefinable.19 In the Tractatus all this is put by saying: A proposition can be true or false only in virtue of being a picture of reality. (4.06) where Wittgenstein is telling us that this minimum condition for being a proposition, this bipolarity, entails the view of possibilities and logical space that is set out in the sections we have been summarizing. It could be argued that the 4.06's, which give an account of what is required for truth (i.e. for truth-or-falsity) are the pivot of the Tractatus, the logical centre of the work. Wittgenstein himself picks out a slightly earlier proposition in our present context: My fundamental idea is that the "logical constants' are not representative; that there can be no representation of the logic of facts. (4.0312 in part)20 We remember that this echoes an early problem of Russell's and Wittgenstein's, and, perhaps for that reason it here serves as a turning-point, leading us into the negative side, the downward slope, as it were (though also much the longer part) of the book. We have seen that the principle of the representation of objects by signs makes possible the true-or-false proposition: in other words, our magazine of signs that can be combined without first inquiring as to the facts assures us of sense 19.This is the only anecdote about Wittgenstein in V.Lowe's life of Whitehead. It is taken from Russell's Autobiography vol. 2, p. 101. 20.Translation slightly modified. I have discussed this passage at, at any rate, greater length in "The Grundgedanke of the Tractatus' in G. Vesey (ed.): Understanding Wittgenstein (1974). 308 Wittgenstein: A Life for our utterances (though not of course of truth). Now we find that these conditions, which make propositions possible, are not themselves expres- sible in propositions, not themselves states of affairs that can be constructed experimentally thanks to language. This has consequences on the level of the constituents of propositions. There is something capable of being affirmed when, within a logical framework, words are combined (objects are thought of as being combined) in a new way. When there is nothing but the logical framework, when no risk of error is taken as regards any objects mentioned, there is also no proposition true or false. In Wittgenstein's own example, I know nothing about the weather when I know that it is either raining or not raining:21 do I, then, know something about negation and disjunction and the truth and falsity of propositions in general? No, says Wittgenstein, what I can know I can put into a proposition, but here we have something that must be given me before I can frame any proposition at all. If there were such a question as what truth and falsity are, how could it be answered? So in this enunciation I am not giving a general truth about the meaning of terms like "not' and "either... or...' (which are the only terms that matter in the example we are considering, the only terms, in a later jargon, that "occur essentially' in the relevant sentence). There are no truths in this area, and hence there is no subject-matter here either, and hence the logical constants (such as the terms just mentioned) do not stand for objects, are not representative. (Once again the thought said to be fundamental turns out to be something encapsulating a whole complex way of thinking.) What we are none the less doing in uttering such sentences, what, by implication, Wittgenstein himself is doing in his Tractatus, is revealed shortly. If in the 4.0's we have the chief premisses of the work, in the 4.1's we have a summary view of its conclusions. Hence the presence there of the propositions about philosophy that we quoted above: philosophy is precisely contrasted with the proposition proper, now identified (4.1 itself) as the representation of the existence and non-existence of states of affairs. Propositions satisfying this description are (when true) the content of natural science, and philosophy is not a natural science, or indeed a science or doctrine of any kind. Distinct from the utterance of truths, there is another activity of understanding and making clear. It is to this realm, the realm of showing as opposed to saying, that philosophy belongs; and it is not surprising that Wittgenstein's own book excels in presenting in a new and surpising form what is in some sense familiar to us. This is true, for example, of his use of the truth-tables (a discovery of 1912, as far as he was concerned) to show the true nature of a proposition (essentially a selection from truth-possibilities), and to show 21.4.461: he had been out of England for some time when he wrote this. 309 9Tractatus 1921{-22 thereby that a so-called logical truth (and a contradiction likewise) was not a proposition at all, since it said nothing. Its truth-table was (Miss Anscombe has well said) like a map that remains an absolute blank. Thus all discussion in the formal sciences becomes a sort of error, though at the same time the importance of the understanding or the showing is not in the least diminished. It only does not fit a certain preconception (which was certainly that of Russell, the scientist). Naturally all of this requires demonstration in detail, and to this the remaning sections are devoted_4.5, 5, and 6 tell us in different terms what the general form of a proposition is, the first colloquially and impressionistically, the second in terms of the vocabulary we have acquired in the 4's, the third in symbols. Thus: ...The general form of a proposition is: This is how things stand. (4.5, in part) A proposition is a truth-function of elementary propositions. (An elementary proposition is a truth-function of itself.) (5) The general form of a truth-function is [p$H, x$H, N(x$H)]. This is the general form of a proposition. (6) The programme of the sections so headed (though there are incidental illuminations of other topics) is to work out these insights (or this insight). It is shown, for example, in the course of the 5's that all propositions and all our operations with propositions can be described (and indeed illuminatingly described) using the model already given. Thus inference is said to be simply a reading off of the truth-grounds common to two propositions: a most ingenious column displays how this would be applied to a simple case, namely all possible combinations of two independent propositions. This table also allows the calculation of the probability that the various combinations give to one another. From 5.2 on Wittgenstein uses the notion of an operation to show in a general way how truth-functions differ from material functions: this will have as a consequence that it is never an hypothesis, and hence also never a truth properly so called, that two propositions have a certain logical relation. The general point is that when we utter a complex proposition (e.g., that it is not the case that it is not the case that it is raining) we are not saying of some object that it has a certain property, and then of the object so qualified that it has the same property. (In Wittgenstein's terms, we have not got a material function here.) Rather we are applying to the actual subject matter (rain, say, and today and now) one of the possible schemata of truth-grounds, a possible truth- table. There would be no proposition unless we applied some schema, 310 Wittgenstein: A Life but of those we apply none are really more complex than others.22 We can certainly display the available truth-tables in a systematic way, and in other ways show how an expression for one possible assertion about our subject matter can be constructed out of that for another; but in doing so we shall be merely exhibiting features of our chosen system of signs, features which will vary in interest according as we ourselves, or those to whom they are exhibited, have failed to notice them. Not many will have had difficulty seeing, "It is not not raining,' as another way of saying "It is raining', but some more complicated correspondences or inclusions will at first escape all of us. The notion of an operation thus introduced is characteristic of Wittgenstein. It has been comparatively little studied and, leading as it does to the apparent trivialization of logic and mathematics, it seems almost too ambitious. The difficulty is that it asks us to depart from a favourite model_that of the formal sciences as a set of truths. Wittgenstein is not so much contributing to the solution of a set of well-recognized problems as demanding a change in our whole ap- proach. It may be put to the debit side that he does not then develop a whole system of logic and mathematics on a new basis. He is content with the insight itself_and this is in some ways intelligible, for it is a difficult insight. The idea that there are at least two quite different uses of language, one to put forward truths, another to exhibit features of language itself, is on the one hand attractive, but also something that we have to be forced to accept. Just as Wittgenstein says that everyday language has features that are designed to do anything rather than reveal its true logical form, so a great deal of our activity as regards the formal sciences_the writing of logic books, the examination of children in arithmetic_seems to presuppose that here too we can conjecture and get things wrong, while other parts of our activity_our reliance on self-evidence, the extraordinary freedom we have in our starting- points_seem to place us in a different world. The difference from factual enquiries is one that we sometimes seem to see clearly and then suddenly lose again. An operation powerful enough to generate all truth-functions is defined at 5.5, and in the course of the the 5.5's it is shown that it will indeed permit the expression not only of any degree of complexity in the selection of truth-possibilities, but also, and by the same token, of general propositions. Given that propositions consist in the application of a truth-schema to a suitable material, or as Wittgenstein says in function and argument (which two cannot come together save by the application 22.This might be put by saying that in Wittgenstein's table of possible truth-functions of two propositions (if we ignore tautology and contradiction, which are not really truth-functions of the two propositions in question) all lines serve to define a proposition in exactly the same way: none is the preferred possibility. 311 9Tractatus 1921{-22 of the logical constants expressed by truth-schemata), we do not need a further logical constant of generalization, as Frege and Whitehead and Russell supposed. By a suitable use of Wittgenstein's one logical operation we can apply truth-schemata to a propositional variable (to the sort of function that is implicit in every proposition) and obtain a generalized proposition of any desired degree of complexity. Thus, in a simple example, if we simultaneously negate the range of propositions indicated by, "x is both human and immortal' we obtain the proposition that all men are mortal. The essential move is made by supposing that a range of propositions can be specified by a function, i.e. by the use of a logical prototype (Urbild) such as "x has F': there is no need for the fictitious enumeration of all possible objects, a, b, c, ..., in order to construct the equally fictitious proposition, Fa v Fb v Fc ... (where F stands for "is both human and immortal'), whose negation might be used to define the generalization in question. In succeeding brief sections Wittgenstein attempts to show (and his attempt is at least interesting) that identity is not an extra logical constant requiring separate treatment, and that propositions of the form, "A judges that p' (which, as we have seen, caused so much trouble to Russell) can also be dealt with. This is not a relation of some independent subject A to a proposition, which would leave it open that A might inadvertently have the same relation to, might entertain, say, something not a proposition at all. Rather, A is nothing but a succession of thoughts, and when we specify one of them, there is no further question how it manages to be the thought of just the state of affairs it envisages: "p', as Wittgenstein puts it, says_i.e. it cannot but say_that p. The occurrence of a thought is not established by taking the mental temperature, but by observing or understanding a proposition. It follows easily that a nonsense cannot be thought (technically that "p or not-p' follows from p's being thought).23 If the subject of judgement is, in this way, a complex (and hence not really a soul at all_5.5421), it is perhaps natural that Wittgenstein should turn next to the topic of solipsism:24 The limits of my language mean the limits of my world. I am my world. (The microcosm.) (5.6 and 5.63) At some point certainly Wittgenstein needed to make clear that there was not a metaphysical subject, outside both world and language and set over against them, an idea to be found in some form in Schopenhauer, 23.This interpretation is given at slightly greater length in my "Language and Reality in the Tractatus', Teoria 5 (1985) pp. 140ff. It derives some plausibility from Russell's adopting it in the second edition of Principia Mathematica. 24.In Prototractatus, however, the topic is treated earlier, and that of judgement later. 312 Wittgenstein: A Life and perhaps in the history of philosophy generally. Otherwise his demonstration (if we may allow it to be one) that natural science consisted in the normal use of language to affirm what was true but might be false, and formal science in the exhibiting of inevitable features of that language, would after all leave room for a metaphysical science of necessary features of the world: the subject would be one such feature or would incorporate all such. His answer is that everything describable, and therefore everything psychological also, belongs in the world and is mere fact. When it is further said that this world is still described from a point of view, the answer is that the point of view is given by language and its limits: there is no other restriction, nor is this restriction describable_its removal or lifting is literally inconceivable and hence it too is inexpressible. The 6's draw the conclusions we have been pointing to all along. Logic (conceived of as a system of propositions like Principia Mathematica) is found to consist of tautologies, of pseudo-propositions which, when properly viewed, immediately manifest their own truth_or rather their own lack of truth-or-falsity, their own status as pseudo-propositions (the 6.1's). Mathematics is found to consist likewise of pseudo-propositions, but this time they do not even appear to have a truth-value: they are equations, revealing, in each case, that two apparently distinct applic- ations of operations are really identical applications. The 6.3's give an account of science, showing that apparently necessary or non- hypothetical principles in science are really arbitrary choices of a framework in which to set the laws (which of course may themselves simply turn out to be true or false). The final two sections are parts of the work that we have constantly anticipated, but their content might surprise an innocent reader (unless he had been alerted by the hieratic tone in which some of the purely logical propositions were uttered). The 6.4's (and a few of the 6.3's) deal with ethics_and indeed with God, freedom, and immortality. The 6.5's with philosophy and the problems of life. Three pages, approximately, to give the real message of the book. It can, of course, only be done by indirection. There is no fact that makes an action or a life or the world good or bad: yet they can be seen or experienced as such. Wittgenstein says of the world of the happy or unhappy man: It can wax or wane as a whole_expand or contract (6.43). Literally, of course, this would be nonsense_if everything expands, nothing does; yet the terms have often been used by mystics to describe hope and despair.25 The world can be seen in a certain way, and accepted (or rejected). Essentially it is precisely the acceptance of the lack of value of the facts in the world that sets the mind at rest. Thus ethics is a matter of seeing what is fact in the 25.Some examples are given in my "The Mysticism of the Tractatus', Philosophical Review (1966) pp.324{-5. 313 9Tractatus 1921{-22 world and what not: it is for this reason that it is said to be transcendental, just as logic is. And, to speak very summarily, it is the same with philosophy and the problems of life. Only by seeing (as Tractatus enables you to) that there is no scientific (still less philo- sophical) answer to the problems of life, the problems of Goethe and Beethoven, as Wittgenstein was later to say,26 do you come near to answering them. In this way the book serves to produce a kind of docta ignorantia and who reads it will be able to say in the right spirit: What we cannot speak about we must pass over in silence. (7) On any showing it was a tour de force to combine in the book all the problems of his philosophic life_and to say much also that reflected and bore on the helpless and hopeless situation of an Austrian officer in a war that was bound to be lost. Russell's problems about the logical constants were Wittgenstein's starting-point. He met them with an account of language which gave for the first time an exact characterization of what was true of logical propositions and of no others. No matter that it later turned out that not all propositions of logic can be proved by means akin to truth-tables; that there can be no decision procedure for the whole of logic. Partly for that reason it is no longer supposed that there is a logic of language, a logic of all thought, which we can try to formulate.27 Instead our modern logicians examine different logics, different calculi and supposed possible interpretations. Still this move itself was made with Wittgenstein's help: it is one possible reaction to the difficulties about the ineffability of logic that he raised. What is more, it is still necessary to read his book in order to conduct, or to restrain oneself from, any general discussion of the relations between language and reality such as even the modern logician must be prepared for.28 The work's approach to mathematics has found fewer followers. The general importance, and the difficulty, of trying to hold apart in one's mind different forms of discourse (even the word "discourse' might make one shudder) must be a valuable lesson, but the detail of his doctrine has yet to be worked out, if anyone but the author can. Some anticipation of his later interest in intuitionism can be detected, and the philosopher of mathematics who wants material for reflection will certainly not discard the book. Its insistence that mathematics can be conducted without a theory of classes (6.031) is as outmoded, perhaps, as its idea of the logic of language, but here too it has contributed to its own desuetude. The 26.In 1931: see Culture and Value (a slightly unfortunate title in view of the considerations I am advancing), p.9. 27.See Jean van He{ienoort: "Logic as language and logic as calculus', Synthese 17 (1967) pp.324{-30. 28.See Michael Dummett: Frege's Philosophy of Language (2nd ed. 1981) for an example of the frequent reference that must be made to Tractatus. 314 Wittgenstein: A Life apparently more technical problem how much of set theory is needed for how much of mathematics is still a central one and its investigation leads ever and again to surprising simplifications. And the basic issue between intuitionism and Platonism is still alive in a form which might lead philosophers to seek a fresh vantage point such as the Tractatus, in the infant days of art, aspired to give. The answer to Russell's "Problem of Matter' is as decisive in intention as that to the problems of the foundations of logic and mathematics. The Tractatus renders meaningless any attempt to describe the elements of reality. They are bound to be logical elements, incapable of any description. At a higher level, indeed, it is possible to paraphrase any statement of fact in physical terms, or equally to paraphrase it in terms of the sense data of an individual. These are merely arbitrary choices of a framework within which description can take place: neither is the correct one, though either may be illuminating for a particular purpose.29 The claim that either represented the real furniture of the world would be metaphysical.30 On the question of the status of science, of that cognitive or intellectual value that had led Russell to aspire to a scientific method in philosophy, the Tractatus position is twofold. On the one hand, science (in the sense of natural science) alone has cognitive value, since it is the totality of true propositions. On the other hand, it provides us with the explanation of nothing. Darwinism gives the essence of the world as little as phenomenalism. Like Newtonian physics, or Hertz's, it is simply a choice, which might be made a priori, of the form in which we are going to express our general descriptions of the world. Various considerations, of convenience, of convention, and of tolerance of error or of complic- ation in description, may decide us for one such possibility in preference to others, but the possibilities themselves are all clearly constructions implicit in language. Here, perhaps, Wittgenstein fails to do justice to the fertility of the scientist in the formation of new concepts and to the essential role in such formation of a corpus of experience and of interrogation of nature, concerted or consilient, by generations of scientists without which the conditions for the formation of such concepts would not exist.31 Such misgivings do not affect the main point: the 29.Thus at 6.3751 a physicalist translation in terms of movements of particles is used to show why (i.e.how) it is contradictory to ascribe, say, redness and greenness to a single place in the visual field. When Carnap and Neurath made great play with the notion, Wittgenstein told Schlick that the Tractatus had already dealt with physicalism though not under that rebarbative (scheusslichen) name and with its usual brevity (LW to MS 8.8.32). 30.Thus, as Wittgenstein reminded Schlick in the letter just quoted, the Tractatus is directed as much against the metaphysics of the physicists (physicalism) as against that of the philosophers (of which phenomenalism would be one form). 31.Mach's Theory of Heat describes this well for one domain, but it is also evident, I think, in the work of Boltzmann and Hertz, so much admired by Wittgenstein. I have developed this theme in an address given at Kirchberg in 1986. 315 9Tractatus 1921{-22 rejection of any claim by science to explain phenomena, if (that is) explanation is taken to be anything other than presenting the phenomena in some clear and easily grasped form. Of course, Mach had already effected this rejection: but Wittgenstein's position is of special interest in that it proceeds from purely logical considerations not from an empiricist prejudice.32 This conclusion is besides important for the fundamental lesson of Wittgenstein's book. It is no accident that the account of science in the 6.3's culminates in 6.4: All propositions are of equal value. Just as factual propositions cannot explain in an intellectual context, so also they cannot give sense to the world. All the value of the world lies in the inexpressible, as we have already seen. For the third range of questions that Russell and Wittgenstein had to deal with, those concerning the self, we have given the essence of the answer, or rather the new question, in the course of our exposition. It seemed that the self was bound either to be dissolved by phenomenalism into a series of conscious states or, if an idealist starting-point was taken, to absorb everything into itself, and result in solipsism. Wittgenstein adopts both positions, but draws the sting from both. The self of acquaintance is not an object known to me alone_rather as Russell thought that the real reference of the phrase "The Iron Chancellor' was an object_call it O_known only to Bismarck, but is instead a series of thoughts which I of course can describe.33 Yet there is another attitude to the world, which is that of identifying with the metaphysical subject, of seeing the world, one might say, with the objectivity and neutrality of language itself. This, the refusal to identify with the interests of the individual human being, or to be exercised by hope and fear, is no doubt what, quite early in the War, he described in his diary as the strict solipsistic position,34 and in his notebooks in the latter part of 1916 he returns again and again to this theme. It is less prominent in the Tractatus itself, yet there too he sketches an attitude to the world which renounces any special position for the empirical self and attributes no value to the holding, or the bringing about, of any particular states of affairs. The lesson is not itself so strange as the way it is taught. He has indeed found his way through the technical problems of logic in which he did not want Russell to find value: yet precisely in them he has also found his answer to the problems 32.This is why Schlick and some others regarded Wittgenstein as one of the founders of logical positivism (though the latter of the two words does not fit him very well). 33.In a book The World as I Found it there would be no mention of the subject, according to Wittgenstein (5.631). 34.In his code diary for 8.12.14 Wittgenstein asks whether the idea of leading a life without sense is reconcilable with the strict solipsistic position, clearly thought of as being the one he aspires to. 316 Wittgenstein: A Life of life. Philosophy requires logic and logic shows that there can be no philosophy, yet the fact that there can be none turns out to be the greatest liberation. Frege convinced him that his youthful adherence to the idealism of Schopenhauer was mistaken. Now, thinking out Frege's position he finds that after all idealism and realism coincide,35 that he can master the world only by turning his back on it. How he fared when he came to do so will be our next topic. 35.Notebooks 2.9.16, echoed in Tractatus 5.64.