Wittgenstein's Second Half in Forty Five Minutes
by
Denis Paul
BARRA HEAD
A friend has lent me Paul Strathern's Wittgenstein in 90 minutes, Constable, 1996. Contrary to what its title might suggest to people who have spent fifty years studying Wittgenstein, this deals with Tractatus Logico Philosophicus, its origins and the years in Austria following it in a fair and helpful way, with only minor inaccuracies which I shall not quibble about. Strathern also has the honesty to call the Tractatus "philosophical poetry of the highest order", a point which I, as a logical and textual nit-picker, have not gone out of my way to make, and here gladly acknowledge.
On the subject of Wittgenstein's deciding, at the end of 1928, to return to writing philosophy, however, and on the twenty two years that followed (February 1929 to April 1951), Strathern is wildly adrift, as most commentators have been until the last ten years or so, when a new movement began, devoted to a painstaking study of what Wittgenstein actually wrote in those twenty two years.
Since the achievements of this movement have not yet become available to the general public, I should like to take the opportunity to offer a summary of them on a scale that fits Strathern's book. He himself deals with these final twenty two years in a mere nine pages, after using forty four for his earlier story. So I feel entitled to let myself spread well beyond nine pages of Constable's small format, but I do promise to stay short of forty four, and my compromise aim will be forty five minutes of comfortable reading time.
I take up the story near the foot of page 50, where Wittgenstein has realised that "perhaps he hadn't quite succeeded in killing off philosophy after all".
The first thing to say is that his renewed interest in philosophy began at least in 1927, when he met Schlick and other members of the Vienna Circle. Stories about an occasion when he refused to talk philosophy with them do not mean that he never did. His interest in a 1928 lecture by Brouwer on the nature of mathematics was only one small detail in his reapplication to philosophy. In a series of draft prefaces for a proposed pre-war publication of Philosophical Investigations he distinguished between his return to writing philosophy in 1929 and an earlier return to thinking about it.
His deciding step was to arrange a place at Trinity as a postgraduate student (his degree being provided by the University's willingness to accept the Tractatus as a PhD thesis). He did not believe the Tractatus to be completely wrong, only to be in need of some tinkering. From his preparatory notes for the Tractatus (Notebooks 1914-1916, written from 1914 to early in 1917) one can infer that from the beginning he had been sitting on the fence between two interpretations of its atomic objects --- actual physical particles or components on the one hand and sense data on the other (see the entries from 17.6.15 to 22.6.15). In Cambridge, dating his first notes 2.2.29, he proceeded to write in a new set of large notebooks, investigating a 'phenomenological language' in which the second of these assumptions could be expressed. After eight months of hard work, and roughly half way into the third of these new notebooks, he abandoned the attempt.
Unfortunately the first two of them had been written in a most idiosyncratic fashion: writing his philosophy only on the right hand pages of Volume I, he proceeded to the right hand pages of Volume II; then he wrote (not working from its end but starting again at the beginning) in the left hand pages of Volume II, and finally in the left hand pages of Volume I. As Volume II had roughly twice as many pages as Volume I, a passage early in left-hand I would actually have been written in the last sixth of this 'Siamese Twin' enterprise (as Rush Rhees has called it), but could give a careless reader the impression of being written at its beginning. This very mistake is made by Monk, who quotes a remark about a mathematical idea of Brouwer's as if it was an early one. To the best of my knowledge, until the last dozen years or so only three people had cracked this page system: Rhush Rhees, who had access to the originals as a trustee, Micheal Nedo, who began making new microfilms for Trinity in 1976, and myself, working in 1976 from inferior microfilms published by Cornell University. From at least 1991 onwards all sorts of people made the discovery (among whom I owe an especial debt of gratitude to Dr Josef Rothhaupt of Munich). It is now common knowledge among Wittgenstein scholars but, judging by Strathern's book, it has still not penetrated to non-specialists, let alone non-philosophers.
There are two interesting passages which show that Wittgenstein was feeling uneasy about his quest for a phenomenological language before he expressly abandoned it. The first comes near the end of left-hand Volume I (MS 105 if you want to look for it in the new Bergen CD Rom edition), and so it isn't surprising that it is little known. It comes (undated) on page 116.
With the phenomenological language it is as if I came into a swamp with a magic spell on it where everything that can be grasped vanishes.
This is useful to me personally because it gives me a title for my book on Wittgenstein: Climbing out of the Swamp. The second comes on the first page of the next volume (III, MS 107) and is also undated.
The kind of consideration that leads us down as if into a Devil's Punchbowl is considering the present as the only real thing. This present, conceived as being in constant flow or rather in constant change, will not let itself be grasped. It vanishes before we can even think of grasping it. We stay bewitched in this punchbowl in a swirl of thoughts.
With one exception all the paragraphs of the first three pages of this third volume escaped Wittgenstein's paste pot (described ahead), but they constitute one of his best purple passages. They will not only be read easily on CD Rom but are already in print in the Springer Verlag Wiener Ausgabe, as are all the first nine and a half of Wittgenstein's new large manuscript volumes, MSS 105-114. (The second half of Volume X is separated from the first by a long pause for dictating the so-called 'Big Typescript'.)
As I can attest from my own microfilm work of 1976, it is only too easy to jump to the conclusion that at least from October 1929 onwards Wittgenstein began to demolish the Tractatus. He did not. Declaring his abandonment of the phenomenological language on 22.10.29 and again on 25.11.29, he was still able to write, on 1.12.29, at the foot of page 223 of MS 107,
Phenomena are not symptoms for something else but are reality.
and
Phenomena are not symptoms for something else that is still needed to make a sentence true or false but are themselves what verifies it.
If the point he is making does not seem obvious, one has to remember that for Wittgenstein in this period phenomena were simply sense data, related (though not precisely corresponding) to the Greek meaning of the word. Later he used the German "Erscheinungen" when he wanted to make clear that he was talking about appearances. In a 1946 passage Elizabeth Anscombe, a Greek scholar, translates this word as "phenomena", but by that time Wittgenstein had come to mean by phenomena what we mean when we talk about the phenomena of rust, or of metal fatigue, or of stage fright, or of garden gnomes.
The first of December may seem so close to November the twenty fifth that nothing much can be inferred, but Wittgenstein had a very precise mind. He is letting us know that renouncing his phenomenological language did not mean abandoning what we should call phenomenalism or idealism. To find evidence of a move towards common sense, we have to jump to Vienna on 13.12.29, when Wittgenstein had left his third notebook (unfinished) in Cambridge and started on a fourth. On page five of that we find
"Realism", "idealism", etc are from the very beginning metaphysical names. That is, they indicate that their devotees believe they can say something definite about the nature of the world.
Denying that language could achieve anything of the kind had been the subject of his very first note on returning to Vienna. Then, coming back to page five,
Anyone who wants to contest the proposition that only present experience is real (which is just as false as to assert it) will perhaps ask whether a sentence like "Julius Caesar crossed the Alps" describes only my present state of mind in considering the matter.
We have here only one of many cracks that began to appear in the structure of the Tractatus, both during this Vienna vacation and back in Cambridge before Easter 1930, but the structure was far from collapsing. In Vienna for Easter, Wittgenstein dictated a typescript from what by then were three and a half manuscript volumes. Altogether the process of freeing himself took three years, from 2.2.29 to 1.7.31 in his notebooks and then some months more in his lectures.
Back in England in April 1930, Wittgenstein left his typescript with Russell and took the carbon copy back to Cambridge, where he cut it up into paragraphs and pasted a selection from them into the blank pages of a further manuscript book, cunningly reordering them to give the impression that the phenomenological language quest had been given up before the resulting text was written, instead of, merely, before it was pasted together. The trustees' nickname for this quite substantial collection of pastings was the Moore Volume. Wittgenstein's name, on the identical title pages of both, was Philosophische Bemerkungen. This still causes confusion among commentators, and it really would be easier if they were called the Russell Typescript and the Moore Volume to distinguish them. Fortunately photographs were made of the latter, enabling Rhees to publish it under Wittgenstein's title in 1964, having left the original in a telephone box.
Even though this printed volume, later translated as Philosophical Remarks, disguised the part played by the phenomenological language, it did reflect the fact that through his Christmas visit to Vienna and into January 1930 Wittgenstein had been as firmly in favour of the verification principle (that no unverifiable assertion could mean anything) as any member of the Vienna Circle. What one might well take as a softening of the principle is written on 21.1.30 on page 252 of Volume III (which he filled up on returning to Cambridge), namely
An hypothesis stands to reality, as it were, in a looser connection than that of verification.
but one must beware. This comes in a very dogmatic context in which an hypothesis like "this is a chair that I am sitting on" (unverifiable) is contrasted with my sensations as I sit on it (which I do not need to verify). A real relaxation of the principle does follow eventually, heralded perhaps on 30.6.31 (on page 238 of Volume VI) by
A description of the verification of a proposition is a contribution to its grammar.
Even here we are not out of the verification wood. There is evidence from lectures delivered in the autumn of that year that Wittgenstein was still taking a very dogmatic line about the hypothetical nature of propositions which in common sense we take to be in no need of more than cursory verification --- such as, let us say, that as I write this I am wearing two black shoes.
Wittgenstein's adherance to the verification principle is also evident from the volume Wittgenstein und der Wiener Kreis, published in German around the same time as Philosophische Bemerkungen. In this, Waismann's notes on conversations between himself, Wittgenstein and Schlick, starting at the end of 1929, are printed. They are extremely readable. Naturally, they do not expound the phenomenological language, abandoned before they began, but their 1929-30 pages do give evidence of a strong Tractatus bias. Some 1931 passages, however, record Wittgenstein arriving in Vienna with the opening pages of a new typescript (TS 211), which do at least presage a break with the Tractatus. These pages were dictated from the seventh of Wittgenstein's new manuscript volumes and include the origin of the 'Augustine language-teaching' idea with which Investigations opens (dated 15.7.31 in manuscript). Beware again, however: this does not mean that the full ideas of Investigations began on that date.
A very good way to appreciate the rigour of Wittgenstein's first three years back in Cambridge in contrast to the later, softer views for which he became famous is to read the lectures of the early thirties as noted by Lee and compare them with those of the mid-thirties as noted by Ambrose. (His Olympian years, as she called them, because of their iconoclasm --- a beautiful view of Zeus!) These later lectures are full of the spirit of The Blue Book, dictated to Ambrose and Skinner for the benefit of their fellow students and intended to be kept secret from everyone else. It was an electrifying piece, elaborating on Wittgenstein's mid-thirties war-cry, "the meaning of a word is its use". Contraband copies of it were circulated in Oxford and Cambridge after the war and formed the basis of our knowledge of post-Tractatus Wittgenstein until Investigations was published in 1953. Another dictation circulated in samizdat typescript was The Brown Book, which was less inspiring and gave rise to Investigations (begun in 1936). The Brown Book and Investigations were essentially didactic works, intended for learners to work through and think about step by step.
Having made clear that the easing of Tractatus rigour did not come all in one go, I must mention two more easings and one that, perhaps, was never made at all. On 1.1.30 there is a Vienna note on page 52 of Volume IV:
The concept of an "elementary proposition" now loses its meaning entirely.
However, Vienna Circle conversations of 2.1.30 and 5.1.30 (pages 73-91 in print) show that things were not quite so clear cut. What Wittgenstein had abandoned was the mutual independance of elementary (atomic) propositions, while he still believed that they were there, waiting to be discovered by analysis, although their form was unpredictable.
On the first of July 1931 Wittgenstein wrote a note indicating that he was giving up what I believe to have been the worst of the superstitions of the Tractatus. This was his belief (disguised in the Tractatus itself to sound much more reasonable) that once a sentence was rewritten in its logical fundamentals it had to have the complexity of the facts it described. One bit of sentence had to correspond to one bit of fact-constituent --- the real meaning of the picture theory of language which Wittgenstein derived from a report on a Paris street accident (see Strathern, page 37). The note itself only says "Complex # Tatsache", and for an explanation of what that is shorthand for you must read an essay called "Komplex und Tatsache", which Rhees thought important enough to include in the appendices of both Philosophical Remarks and Philosophical Grammar. You must also be sceptical, because you will find on page 85 of the Lee lecture volume, for the Michaelmas Term of 1931,
A proposition is a fact which is a picture of another fact.
The 'superstition' which may never have been given up only dawned on me while I was writing this. On page 279 of Volume IV, 31.7.30, there is
The truth functions belong to the nature of sentences, and that is why they cannot be explained by means of sentences.
to which I said "At the blackboard they can", remembering the many times I had done so when teaching Wittgenstein's truth functions and truth tables as a form of Boolean algebra. The opening Vienna note which I mentioned said that language cannot express the nature of the world. This suggests that we can only dismiss philosophical talk about the nature of the world as metaphysical nonsense by appealing to what language can do, and a theory about how it does it. On page 6 of MS 151, which must be either 1935 or 1936, there is this parenthesis:
[A scientist says that he is only describing empirical science, or a mathematician only mathematics, and not [doing] philosophy, --- but he is subject to the temptations of language like everyone else, he is in the same danger and must be on the look-out for it.]
After a recent spate of scientists and mathematicians giving way to the temptations of language, I cheered when I found that, but was Wittgenstein giving way to the temptation to explain matters by appealing to language? Investigations is full of warnings against harbouring models of how language ought to work, but after he had abandoned every typical model I fear that there was still, at the back of his mind, an idea of language as something that, while it could never be typified, still influenced our thought. As in particular cases it clearly does, but when one meets an example one can usually say which particular quirk of which particular language is at fault. And when one cannot, then instead of constructing an abstract background to language in order to make it the culprit, one should have the honesty to say simply that some things are metaphysical nonsense up with which one will not put. Wittgenstein might well have agreed with this at the end of his life, but I cannot be sure.
The 1936 manuscript and the 1938 typescript of Investigations, both now available at last, or shortly to be so, are easier to read than the 1953 printed version --- the latter, as well as containing a lot of extra material, also condenses the earlier material to a degree that makes it bewildering and rebarbative. Condensation was a vice that Wittgenstein only gave up shortly before the end, when he realised he had no time for it. His last, best and most flowing work, published in the form in which he wrote it in his final notebooks, can be read as §300 onwards of the volume entitled On Certainty.
The most extreme example of condensation (apart from the Tractatus itself) was published by the trustees as Part II of Investigations, although it was indubitably intended as a separate work. Wittgenstein dictated its typescript (both top and carbon copies of which were lost soon after it was published) after he had already made a preliminary 'selection towards a condensation' in two long typescripts which have been in print for some time, dictated from the first eight of nine large notebooks. Then he wrote a very condensed manuscript (MS 144) before dictating the final typescript (which the trustees shuffled around a little to make it end impressively with some remarks on the confusions of psychology). The two printed volumes of the first selection (they are Parts I and II of Remarks on the Philosophy of Psychology) have some fascinating things in them, but their basis, MSS 130-138, are what people ought really to be queueing up to read in the Bergen edition. This is not to say that everybody will enjoy these lengthy, penultimate notebooks (1945-1949). Far from it. They do not read like philosophy at all. The fact remains that they were, in so far as they were preliminary notes for an analysis of concepts related to psychology. Wittgenstein regarded them as a tidying up of the work of the academic psychologist William James (the brother of the novelist). To me they are the philosophical equivalent of Johnson stubbing his foot on a stone to refute Berkeley, except that the pain went on for much longer (roughly four and a half years).
These notes are certainly not nonsense, like the attempted propositions of the Tractatus, which one was expected to think of as a ladder to be climbed up and then thrown away, but they do share the character of being best thrown away when they have been read. There is hearsay evidence that Wittgenstein was very dubious about whether anybody would be able to continue his work. In the sense of continuing his effort, his Klärungszweck as he called it, I am in entire agreement with that. Whether one need fear not being able to continue doing philosophy after reading Wittgenstein is quite another matter. There is nothing to stop one accomplishing one's own Klärungszweck, or any other kind of philosophy --- prompted by the world's activities around one, by remarks made by contemporaries (in print or otherwise), or by passages in philosophers of the past. The recipe is simple. Read the entirety of Wittgenstein's corpus in the Bergen CD Rom edition (preferably in the order: Volume 2, Volume 1, Volume 3, Volume 4), and then, remembering the ladder, throw it away, or preferably give it to an impoverished philosophy student who cannot afford it, provided he or she promises to pass it on in turn. This task will take you at least two years, possibly as many as five, depending on your stamina, your powers of concentration and your eyesight.
Then, take a deep breath and start doing your own philosophy. I say "doing" because you may not want to write it, only to speak it aloud as you walk the hills, or argue it with your friends. There are many possibilities. What I am certain of is that provided (which you will discover within a month or two) you can undertake this task at all, you will never regret carrying it through, and always be grateful that you paid no attention to people who told you that Wittgenstein's work was a load of rubbish.
I have said nothing (except to recommend my favourite part) about the work that followed the 'philosophy of psychology', on colour concepts, the human 'interior' and the difference between certainty and knowledge, which elsewhere I have called his group of last quartets. It is well, if not perfectly, published in book form, and if you like it you can go on to read it in Bergen form, and see how it was written. So I shall end by giving a few quotations that will give a flavour of the 'philosophy of psychology' work. Incidentally, if you still think it perverse to call this philosophy, simply call it the conceptual analysis of words that describe the normal psychology of everyday life.
This interest did not begin in 1946. To set the scene for it, here is a short remark written on page 202 of Volume III on 20.11.29.
But could one say "I seem to be sad, I am letting my head hang so"?
Indeed, 1929 and its struggles with the phenomenological language continued to reverberate with Wittgenstein until the end of his life. This was written on page 51 of MS 136 on 3.1.48:
In philosophy one must climb down into the old chaos and feel good there [sich dort wohlfühlen].
In a 1938 notebook (MS 121) he had called the phenomenological language "the Fatamorgana of a language", in a poetic remark which it will be well worth the cost of the Bergen edition to find. (Unpoetically, "Fatamorgana" just means "mirage".)
A remark whose meaning is easy to see, though people who have worried about the state of their teeth may disagree with it, can be read in print as §959 of Volume I of Remarks on the Philosophy of Psychology:
Love is not a feeling. Love is tested, pains aren't.
This was written on 28.4.47 on page 159 of MS 134. Quibbles about what one is testing when one probes a tooth with one's tongue are left as an exercise for the reader. What is more interesting to mention here is that in the Cornell microfilm of this notebook seven lines were blocked out by the trustees lest we should discover what was on Wittgenstein's mind. This, of course, will be revealed in the Bergen edition, and if the seven lines were written in Wittgenstein's code they will be translated.
§7 of the second Remarks volume includes (written in November or December 1947 on page 153 of MS 135)
It is not true that thinking is a kind of speaking, as I once said. 'Thinking' is categorically distinct from the concept 'speaking'.
At Oxford we used to translate Wittgenstein's usage there as "categorially" but I do not think he would have approved. His German was "kategorisch". His earlier view was expressed at length and by implication in arguments about whether a statement of intention or expectation leaves anything unsaid (as it were as a wordless thought) about exactly what is intended or expected. Although he had said (on 28.6.30, on page 195 of his fourth volume, MS 108)
It's in language that it's all done.
which one might think was intended to settle the matter, his arguments continued, and a month and two days later, on page 270, he wrote
A step is needed here similar to that of the theory of relativity.
and he seems to take this step the next day, on page 277, after declaring that thought is what can be expressed in a language and that he wants to say that all thinking must take place in symbols, with the beautiful argument
But if someone said "how can I know what he means, I can only see his symbols", my answer is: "how is he supposed to know what he means, he has only got his symbols".
I am very glad that in my book I called this relativity-step 'the self-sufficiency of language' because (to return to our muttons of the penultimate notebooks) one could express Wittgenstein's later view by saying that language is sufficient for thought but not necessary to it. A reference to a wordless thought can be found on page 49 of MS 134, 19.3.47:
"I thought 'how unkind he looks!' " As I thought it I did not say these words to myself. Perhaps I looked thoughtfully at him and shook my head. Did I thereupon translate that shaking of the head into those words? No. And hadn't I really thought it? To be sure! --- That is how we use the past tense "I thought".
This ought to be compared with an Investigations paragraph, §330, in which Wittgenstein agrees that thought is not identical to speech but gives an illustration that is rather airy fairy compared to that. He is measuring objects and behaves in a way that could lead someone watching him to say that he had, without words, thought: if two measurements are equal to a third they are equal to each other. No doubt he had applied that principle but few people would say he had thought that thought, whereas it is the most natural thing in the world to agree with him (if he tells us) that shaking his head had constituted thinking that chap unkind, whether or not he had translated this into words before telling us.
A similar example of thinking more naturally after (at least the preliminary revision of) Investigations can be found in an early but undated passage in MS 130, almost certainly written in 1945 when finishing touches were still being put to its pre-war core. One of these is its §125, which ends with the exasperating
The bourgeois position of a contradiction, or its position in bourgeois life: that is the philosophical problem.
This (made to sound much more reasonable by the translation "the civil status") is one of the remarks one cannot possibly be expected to understand without a context, and this is given by the rest of §125, culled from pages 11-15 of MS 130. The real context, however, is an argument with Turing, mentioned by Strathern on his pages 53-4, which can be read in lectures XXI, XXII and XXIII of Wittgenstein's Lectures on the Foundations of Mathematics, Cambridge 1939, pages 207-230. §125 is a kind of acknowledgement of Turing's points, but pages 11-15 include more. With a remark about Russell's Theory of Types starting at the foot of page 11, and some further remarks which (the exasperating one omitted) can be read as §§ 685-690 of Zettel, Wittgenstein comes even closer to admitting that Turing was right.
(Zettel --- the word means "slips" --- was a kind of Moore Volume without paste, consisting of cuttings from various typescripts collected by Wittgenstein in a box file --- but, I later discovered, pasted by Peter Geach into two unused notebooks.)
There is one sequence in Part II of Investigations which I have always found bewildering and still cannot expect other people to enthuse about. This is the first two thirds of section xi, on aspect or 'seeing as' --- for example, seeing a diagram as convex or concave. Wittgenstein goes on to say (on page 214 in print) that the philosophical importance of this problem is its connection with how the meaning of a word is experienced, but I am convinced that it had another significance for him too. This is its reverberations with the 1929 problems of the phenomenological language. We do not see any difference in the diagram and yet we see it differently, and this seems to contradict the concept of a sense datum, that it is something given and thus unquestionable. On page 40 of MS 132, dated 19.9.46 (just before Wittgenstein's last academic year as a Cambridge professor), seeing a capital K as I< is cited. It may be only my reaction that changes, but it is almost as if my image changes, and I want to say that my visual impression, my sense datum, does actually change, even though I know it doesn't. One would think that there is no more to be said, but he goes on to say a great deal more, and I cannot believe that the problem would have nagged at him if it had not been for his 1929 self-indoctrination.
On page 12 of MS 133, dated 24.10.46 (the last academic year under way), there is a moving interjection:
Oh, a key can lie there for ever, where the master [locksmith] put it down, and never be put to use to unlock the lock the master made it to open.
MSS 136-7, begun in December 1947 after retirement, bring us to the last of the passages that Wittgenstein dictated (written on 25.8.48, page 76 of 137), and the rest of MSS 137-8 can be read, except for censorings, in Last Writings Volume I. They start, after a pause for the dictation, on 19.10.48. These three volumes, which end in May 1949, present the philosophy of psychology in a way that perhaps requires less apology than I have been giving the first six of the 'penultimate' volumes. I therefore end with just one illustrative quotation, in print as §714 of the second Remarks volume, dated 9.7.48 and on page 71 of MS 137.
For our considerations it is important that there are people of whom someone feels he will never know what is going on inside them. He will never understand them. (Englishwomen for Europeans.)
The Wiener Ausgabe will continue with the complete TS 211 (the opening of which is mentioned above), with an important group of small preparatory notebooks and with the Big Typescript, TS 213. One must hope that eventually it will also include MSS 130-138, which certainly deserve our having the pleasure of reading them in print as well as on screen.
I must acknowledge the help of Dr Michael Nedo in correcting errors in this squib (as it was originally intended to be) and suggesting additions. I hope these leave it still deserving that term. It will (with apologies to Strathern) be put on the Internet, subject to the stipulation that any reproduction of it must acknowledge my copyright and give my name and address (below), together with my promise that I will do my best to answer queries that are sent to me by ordinary post.
Copyright 1999
Denis Paul
Min yr Afon
Aberath
Wales
SA46 0LL