From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

1. Tractatus and Prototractatus

I suggest considering the so-called “Prototractatus” notebook (MS104) not as “an early version of the Tractatus Logico-Philosophicus”¹, but as the effective manuscript of Wittgenstein’s book. We know that the ultimate typescript was dictated in August 1918; it’s considered a final writing, despite a dozen of later inserted propositions². Well, the MS104 notebook (if we look at the whole of it and not only at its published part) contains in its turn all the material of that typescript, including title, dedication, motto and Preface, with the exception of only five propositions.³ In particular, the first fifty remarks of the manuscript, the backbone of the work, passed almost unaltered into the final book, and 41 of them maintained also the same decimal number: so we can consider the date of composition of these first pages as the real starting date of the Tractatus itself. Unfortunately, there isn’t any agreement on Prototractatus’ composition date.

The content similarity between Prototractatus and Tractatus was just what led von Wright in error when he advanced “the conjecture […] that work on the ‘Prototractatus’ immediately preceded the final composition of the book in summer of 1918” (Wittgenstein 1971, p.9). This may perhaps be true for the last part of the notebook, i.e. pp. 103-120, not edited and not considered “Prototractatus” by von Wright; but most of MS104 was written a long time before. The decisive philological proof was found by McGuinness in 1989, when he published a list taken from the correspondence of Hermine Wittgenstein and dated January 1917. It mentions some of Wittgenstein’s manuscripts; the fifth entry of the list (“a large chancery volume, containing the revision of [the first three notebooks] for publication”) seems to refer precisely to the Prototractatus notebook. McGuinness argues that at the end of 1916 the notebook was filled at least until page 71, in correspondence with proposition 7 insertion, or perhaps until the end of the successive layer of text at page 78 (McGuinness 2002).

2. Prototractatus first 70 pages

While we can accept his conclusion, it’s not so much clear in which circumstances these pages were composed. McGuinness expects that from Hermine’s list we can deduce the non-existence of an eventually lost diary connecting the three we have (MS101, 102 and 103 of von Wright’s catalog); the period between MS102 and MS103, from June 1915 to March 1916, would instead be dedicated to the Prototractatus compilation, until the line traced at page 70. This hypothesis is very uncertain. The work around the Prototractatus is utterly unlike the work on the diaries. Compared with the structured and formal aim of the Abhandlung, their date-ordered arrangement (which is identical before and after the interleaving period: left pages with encoded personal notes, right pages with philosophical free entries) answers to very different needs. Besides, it’s likely that an intermediate lost diary existed, as Geschkowski arguments in his book⁴. Finally, it’s probable that the third entry of the Hermine’s list just refers to this (now lost) notebook, and not to the successive MS103, as McGuinness thought.⁵ But from his objections to McGuinness, without any cogent reason, Geschkowski concludes that the first 70 pages of the Prototractatus were filled only in the autumn of 1916, on the basis of a gathering of material on loose sheets.

3. Prototractatus first 28 pages

On the contrary, as I elsewhere discussed (Bazzocchi 2007a and 2007b), I think that the method of composition of the notebook’s first layer, until p. 28, reveals a typical first writing, where the decimal numbers play the role of heuristic guide. It seems improbable that the proposition numbers were added later, as instead Geschkowski is forced to assert (if the decimals were present at the moment of the supposed copy from the loose sheets, the notebook wouldn’t be in such disorder as it is). We can prove, indeed, that the decimal numbers were in use from the beginning. In fact, the proposition 2.23 in second page was deleted by pencil and transferred to the fourth page under the new decimal 2.181: at the moment of deletion, it already had its number, perfectly coherent with all the others.

In short, I think that the first 28 pages were filled before the letter to Russell of October 22^nd 1915, because from the letter we can deduce that some time before a copy of the notebook first stratum into a “last summary on loose sheets” was made (see Bazzocchi 2006). So we can date this first layer compatibly with McGuinness’ dating (although by different reasons), i.e. around 1915 summer.

4. Prototractatus first 12 pages

Now I want to introduce a new argument, that until now critics haven’t noticed.

In the diary entry of June 18^th 1915, one can find a very baffling passage. In the middle of a long discussion on generality and particularity, illustrated by an example of a picture and its dots, there is an incongruous reflection: “Not: a propositions follows from another one, but: the truth of a proposition follows from the truth of the other”. Then the text continues about pictures and dots. What is the sense of this mention? Why such sudden inspiration? One idea is that Wittgenstein, incidentally remembering some other remark, decided on a new way of expression and hurried to fix it on the page. The remark to be modified is not in the diary. But if we look at page 12 of the Prototractatus notebook (a page that as in McGuinness’ as in my hypothesis takes place around that period), we find exactly the contested expression: “5.041 In particular a proposition follows from another one if all the truth-grounds of the first are truth-grounds of the second”⁶. Well, the remark is emended with the precise insertion of “the truth of”: “the truth of a proposition follows from the truth of another one”.⁷

Here we have two indubitable facts: on June 18^th 1915 Wittgenstein fixed a correction, and at Prototractatus page 12 the same correction took place. There are only two possibilities: or first Wittgenstein stated the amendment in abstract, and then the case took place and he corrected it exactly as stated some time before, or first he wrote the previous form on the Prototractatus, and then reviewed it and remarked the adjustment on the diary. The first case is very unlikely. It’s hard to believe that Wittgenstein decided in abstract such a particular (and indeed not so clear) correction of his thought; that then (a few weeks later, in McGuinness’ hypothesis) twice⁸ he made just the “mistake” he had already criticized; and that finally he corrected it following a previous such foresighted purely theoretical amendment. The only effective possibility is that the compilation of Prototractatus page 12 precedes the discovery of the inaccuracy and its record on the diary. Note that the question does not concern only the wording of propositions 5.04 and 5.041 – that at the time, one may think, could have been recorded on some other slip of paper – but properly Prototractatus page 12, because the correction is unquestionably on it.

Hence we can conclude that the Prototractatus notebook started before (and not after) the end of the MS102 diary, that in fact contains a reference to its page 12. McGuinness’ hypothesis seems to fall off anyway, but onto the opposite side compared to what Geschkowski argued.

5. Prototractatus first page

The Prototractatus compilation was indeed a very slow process, at an average speed of three or four pages a month: the total 120 pages of August 1918 were already 71 as the end of 1916, at least 28 in October 1915, and 12 in June⁹. So we can presume that the starting point was in April or May 1915. In this case, the letter to Russell of May 22^nd 1915 assumes a definite sense. In the previous communication to Russell, in November 1914, Wittgenstein said: “If I should not survive the present war, the manuscript of mine that I showed to Moore at the time will be sent to you, along with another one which I have written now, during the war” [Wittgenstein 1974, p. 62]. The second manuscript is evidently the 1914 diary, whose first notebook was completed in October 30^th. But in May the reference is quite different: “I’m extremely sorry that you weren’t able to understand Moore’s note – Wittgenstein writes – Now, what I’ve written recently will be, I’m afraid, still more incomprehensible. […] If I don’t live to see the end of this war, […] you must get my manuscript printed whether anyone understand it or not”.

Here Wittgenstein refers to only one coherent manuscript [“mein Manuskript”], started in the last period [“in der letzten Zeit”], very different and more incomprehensible than the one showed to Moore. This recent writing can hardly be identified with the two wartime notebooks MS101 and MS102, already cited in the previous letter and presented as similar to the pre-war notebook. Besides, this is the first time, despite Russell’s frequent solicitations, that Wittgenstein speaks about printing some work of his – or rather, insists it “must” be printed. After his reluctance to publish anything that is less than perfect, his diaries seem the less indicated works for publication.¹⁰ But the most puzzling reference is the final clause: “The problems are becoming more and more lapidary and general and the method has changed drastically. –”¹¹. Wittgenstein wasn’t in the habit of telling something without a good reason. Such a relevant change of method is not detectable in the diary entries, nor in the passage from MS101 to MS102. The method here remains discursive and dubitative, without any increasing “lapidarity”. On the contrary, everyone would say that with the first pages of the Prototractatus “problems become more and more lapidary and general”. Here Wittgenstein cannot refer to the diaries, but to new records (may be also in other sheets or notebooks) which in brief will converge (or are in the process of converging) into the Prototractatus notebook. No doubt that starting from its first page the method does “change drastically”, adopting Tractatus’ top-down numerical structure. So we aren’t far from the truth if we think that the first page of the notebook, the proper Abhandlung starting point, was filled between April and May 1915.¹²

This conclusion is not without consequences. If in general the 1915-16 notebooks do not precede the definition of the Abhandlung propositions on the Prototractatus register, nor are they independent and alternative, but accompany it, as a counter-song that discusses its apodictic statements, it’s useful to read the two documents in parallel. It’s essential to hypothesize a definite date scansion of the Prototractatus notebook, and above all to follow the sequence of its itinerary, which – it’s convenient to repeat here – doesn’t have anything in common with Tractatus’ arrangement in sequential order of decimal number. The notebook privileges a top-down process, from high-level sequences to ever deeper reflections; all the skeleton of the arguments is stated before the successive waves of specific comments.¹³ In particular, the first twenty-eight pages of the Prototractatus do not correspond to recorded propositions on the diaries, but are in general their structural ancestors. So, it becomes clear how could the 1915-16 notebooks contain so many propositions of detail which will find place, without corrections, in the final work, since at the moment of their first conceiving, the entire structure of reference was already fixed on the contemporary Abhandlung. The Prototractatus stratification sets a series of nuclear prototypes, in some way discussed and commented in the diary, whose inspection can improve the comprehension of the whole enterprise.