From the Prototractatus to the Tractatus
From the Prototractatus to the Tractatus


The final, unpublished section of Wittgenstein’s so-called "Prototractatus" manuscript allows us to rebuild the gradual course he took to arrive at the definitive structure of the Tractatus. So that it’s now possible to solve the remaining enigmas of the "Korrektur" cited by the author himself in the final pages of the Notebook. The text’s continuity with the preceding parts confirms the top-down technique coherently followed by Wittgenstein in developing the logical tree of his composition.

Table of contents

    Wittgenstein’s Ms104 Notebook contains a final, unpublished section with 80 propositions, which almost perfectly match the text of the definitive Tractatus. In the manuscript, the distinction in relation to the previous part is pointed out by the inclusion of a little cross near the decimals, commencing with the second proposition on page 103, and by a note that informs: "+n are the numbers of the Korrektur". What did in fact take place on page 103 of the notebook?

    In his Historical Introduction to the edition of firs part of Ms104, published as “Prototractatus”, on this point von Wright speculates: «Wittgenstein did to the manuscript the same thing we have done to it here, viz. he copied it out, or had it copied out, with the remarks arranged in their proper number-order. Probably either this copy was a typescript or a typescript copy of it was made. Then he worked with this text (the ‘Korrektur’) changing the formulation, grouping remarks which carried separate numbers under one single number, rearranging the order in places. Having done this (or when doing this) he also made some additions. These additional entries he noted down in handwriting in the same notebook as the ‘Prototractatus’» [Wittgenstein 1996, p. 8].

    On the one hand, von Wright underestimates the complexity of the operation; on the other he doesn't focus correctly on the core of the problem. No-one would be able to reorganize the Abhandlung – von Wright argues – if they started with the chaos of the Notebook; therefore, first of all it was vital to reorder its propositions: exactly as he had to do even only to succeed in reading it. He doesn't consider, however, that until this point Wittgenstein has been successful in overcoming all that structured disorder; better, he has constructed it without mistaking any number, here and there carefully tuning the decimals and changing some dependencies. All the propositions had found their exact place in the virtual structure of the comments, forming a coherent whole: to the point that, at the moment in which one wants to draw an orderly copy, the hierarchical cascade is recomposed and the resulting 800 decimals miraculously connected. It is obvious that Wittgenstein must have always possessed the necessary representative tools to do what he has done; he is certainly able to preserve a perspicuous representation of his construction. So, why should Wittgenstein write down a previous copy of the book, just at the exact point when he decides that the structure is no longer adequate and must be modified?

    More correctly, McGuinness believes that Wittgenstein already worked with an ordered copy, in parallel to the Notebook, and would not have needed a new, preliminary one. On the contrary, according to McGuinness “[in July 1918] Wittgenstein prepared a Korrektur – a version revised and renumbered which closely resemble our Tractatus. These corrections were not written into Ms104, but presumably onto a typescript, then, when new material was needed, it was written into Ms104.” [Wittgenstein 1996, p. xi] In other words, while for von Wright the typed copy was necessary before restructuring, according to McGuinness it was required principally later, to record the Korrektur itself.

    The point here is that still nobody has taken the job of looking and analysing how the "additions", found on pages 103 to page 118 were assembled. “There are additions to every part of the previous Abhandlung (i.e. to the ultimate ‘Prototractatus’),” McGuinness says, calling them also “additional remarks” [p. xii].

    “The second part [of the Notebook] – von Wright writes – has the character of additions to, and further elucidations of, the thoughts contained in the Prototractatus”.1 These statements would be perfect if referring to any preceding section of the Notebook: every proposition of Ms104, up to page 103, is indeed an addition to the structure, a further elucidation that doesn't alter the pre-existing propositions and also usually leaves their decimals intact. Referring to the last part of the Notebook, instead, such descriptions are quite misleading. In fact, most of the so-called final "additions" absorb, modify and integrate preceding propositions. Besides, their numeration doesn't generally put them after already established sequences, but rather at their beginning, or in elevated places of the hierarchy, and implies a rearrangement of surrounding material. Propositions and numbers don't imply any preliminary, consistent copy. If such copy should have existed, it would have been destroyed and rebuilt every time the new arranged propositions are inserted: on the contrary, it is the “additions” which suggest and determine the new correct version.

    Then, we can exclude that, when he reached page 103, Wittgenstein has suspended the work, has created a supplementary copy, has broadly restructured it (possibly writing down an integral copy, typed or not), and then has returned to the job of adding material, annotating it from page 103 onwards. If we had to speculate about such an operation, then we would need to imagine new restructurings and complex re-writings taking place for each "addition", or to suppose that Wittgenstein had left holes and gaps, suspended material and provisional versions. We would also have to imagine some intermediary phases, incompatible with the procedure of the Notebook. Perhaps McGuinness suspected just this when he noted: “It requires some ingenuity and perhaps a little too much imagination to work out what these additions replaced” [McGuinness 2002, p. 281].

    In truth, the final "additions" don't replace anything unknown other than, if at all, real propositions of the old "Prototractatus", propositions now taken up and re-written. McGuinness considers the Korrektur as "a version revised and renumbered so as to resemble closely our Tractatus", and then doubts that it can be ever reconstructed. But, a restructured version, propaedeutic to the last part of the Notebook is not reconstructable, simply because there has never been. The "additions" integrate and modify the compositional phase that precedes page 103, whose propositions contextually, on the parallel copy, are put together and moved up until they become, step-by-step, the Tractatus that we know. If we really want to rebuild the version of "Korrektur", we have it in its entirety by the end of the "additions": for it is none other than the typescript of the Abhandlung itself (Ts202).

    It can be hypothesized that the Korrektur starts as a normal intervention of maintenance on the decimals. In fact, the first four additions, which constitute all the new material of Group 2, are “additions” in the strictest sense and it would have easily been possible to integrate them into the preceding Abhandlung. It may be believed that they have been inserted into the twin structure (managed on loose sheets, in my opinion), requiring, at the first point, only the following simple interventions:

    • a) PT 2.013 is eliminated, and the old 2.014, along with comment 2.0141, takes its place (becoming 2.013, with comment 2.0131). So, now the new 2.014 and 2.0141 can be inserted;
    • b) new entry 2.033 takes the place of PT 2.033, which thus assumes the number 2.034;
    • c) the new record now numbered 2.151 is actually born as 2.1511, a new main comment to PT 2.151; the old PT 2.1511 is renumbered 2.15101 (this remark remains the first note to 2.151, but at secondary level).

    Now, as always, Wittgenstein has to transpose to the Notebook these modifications to the previous decimals. He starts with number 2.1511 on page 14, changing it into 2.15101: on the Notebook, this correction remains clearly readable. Here all should be fine, but, as Wittgenstein starts to appraise the general form of the whole page, he realizes that he ends up creating a structure with an aesthetical-formal meaning. Within the context of searching for formal equilibrium, he shifts the whole sequence and the new proposition assumes the number 2.151; and, as such, it is inserted on page 103. Likewise, the 2.014 graft, with the elimination of PT 2.013, suggests changes to branch 2.012, and finally to branch 2.06, with new attention to the relationships of decimal subordinals. The assemblage he arrives at is difficult to carry into the shattered panorama of the manuscript, so Wittgenstein postpones updating the book, and finally renounces it entirely. The first adjustment, i.e. the correction of number 2.1511 into 2.15101, remains the only one done to the Notebook and becomes one of the rare incongruities in the Prototractatus’ numeration (as the Editors note on p. 52, note 1). It marks the factual beginning of the Korrektur, ultimately constituting a point of no return in the procedure of composition.

    Without revising the book, Wittgenstein revises the whole structure of his Abhandlung on the parallel copy, recording on Ms104 only the more remarkable textual changes. On page 103, Wittgenstein reproduces propositions 3.1, 3.11 (that summarises PT 3.111, 3.12 and 3.13) and 3.12. Groups 3.1 and 3.2 include, in wide part, already-annotated steps, forcing the correction of the whole previous text. PT 3.14 has been modified and turned into the new 3.2, as recorded on p. 103; accordingly, section 3.2 suffered a drastic “cleaning up”, with two new additions inserted on page 104. Proposition 3.3 is really an exact reissue of PT 3.202; while 3.31 includes three previous remarks; 3.323 (p. 105) condenses four of them, also picking up from section 4 (pulling in PT 4.00151).

    Restructuring Group 4 involves only one further addition to the Notebook: proposition 4.24, which comes from rehashing PT 4.2211 and PT 4.2212. Then, there are five annotations belonging to Group 5.1, almost entirely original. Undoubtedly this phase required new numeration, on the parallel copy, to the other propositions of branches 5.0 and 5.1; Reassigning the decimals is equivalent to a complete redrawing of the form and structural dependences of these sections.

    At this point, an entirely different operation appears. Wittgenstein returns to proposition 3.31 of page 104, that defines the relationship between "proposition" and "expression": he adds a last paragraph ("An expression characterizes a form and a content", condensed from ex 3.253), and then decides to comment on it in greater detail. So, he builds four new observations (3.311-4), to deepen the particular use of the terms, "expression" [Ausdruck] and "propositional variable" [Satzvariable]; yet confirming, in 3.313: “In the limiting case the variable becomes constant, the expression a proposition”. These new comments aren’t highlighted by a cross because they don't belong to the offline elaboration (the Korrektur), but are evidently elaborated directly during adjustments to the Notebook. The section concludes with a long line of separation that traverses the width of the page.

    Now, Wittgenstein returns to aligning the text, continuing with Section 5.2. The whole section first appears in one form (on pp. 108-9) then, on page 110, in a different one. The sequence initially proposed (5.2, 5.21, 5.22, 5.232, 5.233, 5.24, 5.241) is partially marked as wrong and replaced by another series.

    Since the discarded propositions, unlike those preserved, aren’t highlighted by the small cross, it may be deduced that the mark sign (and the correspondent note saying of the Korrektur) had not yet been introduced at the moment of the deletion; therefore, up to this point there were no graphic distinctions between the propositions published today as "Prototractatus" and those that follow. At this point, perhaps, Wittgenstein wasn’t warred about being able to update the whole Notebook; but, by now the collateral restructurings are so huge that any possible attempt to do so would be rendered entirely fanciful. It can be inferred that the crosses which signal the Korrektur have been introduced after the elimination of remark 5.241 (not yet marked), but before reassigning 5.233. In fact when, as a result of the formulation of a different 5.233 on page111, the 5.233 of the preceding series was eliminated, it was already marked by a cross. Therefore, also the note on the Korrektur would be added only now, when the compilation already included page 110. The mark continues to be used up to page 114.

    In summary, the text drawn from the phase of Korrektur is transcribed in strict numerical order, touching upon the sections 2.0, 2.1, 3.1, 3.2, 3.3, 4, 5.1, 5.2, 5.3, 5.4 and 5.5.

    By now it is clear, I believe, that the Korrektur could not entail any full-scale writing-out, neither before nor after the restructurings: in no moment, in fact, does a rewriting appear reasonable. Not at the beginning, when nothing has yet changed; nor after each further restructuring, because what is finished no longer creates any problems, and what has still to be checked doesn’t yet create any problem either. The adjustments happen one by one on the parallel structure that, although changing some sheets and rebuilding some parts, always remains coherent and complete.

    It is evident, besides, that the work in the Korrektur involved a notable lapse of time, with changes of strategy and a progressive realisation of the depth of the transformation. A lot of corrections are the direct consequence of absorbing propositions coming from diverse sub-branches, some of which, accordingly, disappear entirely. The last proposition with a cross, 5.501, occupying the whole of page 114, absorbs eight propositions previously belonging to sections 5.003, 5.004 and 5.005.

    On page 115, Wittgenstein recovers some parts which, until that point, have been unused: with 4.1273, he picks up and integrates the last four propositions of branch 5.0053 (originally annotated on page 21). Likewise, he reforms as 5.2522 a proposition intrinsically "orphaned", the PT 5.005351 of page 102 noted by the Editors for its incongruous and duplicated numeration. By the end, material not recovered in the new organization remains rejected: this corresponds to about thirty propositions referred as "Non-correspondences ‘Prototractatus’-Tractatus" on Table 3 of the edited volume [Wittgenstein 1996, p. 253].

    It is probable that it is at this point that Wittgenstein strives to revise Proposition 6, modifying its symbolism: in place of α, he uses the character ξ. This forces him to correct the just-transcribed proposition 5.501 (the change is visible on page 114) and to re-phrase 5.5. In a similar way, he modifies propositions 5.502, which introduces and defines the term "N ()", and 5.503.

    The final 10 propositions belong to Group 6.2 on mathematics; and in contrast to those that precede them, they are entirely original. Insertion 6.21 determines the revision of the whole branch; while other sections of Group 6 receive only minor adjustments. Only at this point does the revision come to an end; there was no reason to prepare an ad hoc copy before completing the schema: for by now, the copy is none other than the typescript of August 1918.

    Surely this is the imminence of dictation; the last numbered proposition, 6.241, carries directions for the typist: "[6 free lines]" instead of a formula. It is also possible that in the intervening time the operation of typing had already begun, and that this page was compiled during some break in dictation. Perhaps this is the reason why the last reflection of the manuscript ("The concept of operation is equivalent to the concept ‘and so on’ ") doesn't find its correct position and remains without number2.

    On pages 119-121, we find the Preface to the Abhandlung, identical to the typed version (the elimination of its conclusive sentence happens on the typescript itself). It is likely that the typing of the preface took place after the typing of the propositions: in fact in Ts202, the pages of the Vorwort don't follow the numeration of the text, which starts on page 1, but are numbered "I" and "II". In version Ts204 (of which Ts202, in relation to the other parts, is a carbon copy), the preface is typed separately, on unnumbered sheets.

    As we have seen, the last part of the notebook is divided into distinct sections following the new organization that Wittgenstein was creating. It is a characteristic of the top-down procedure, which is often combined with backward, or "purpose-driven", techniques. This has given the illusion that the final structure had already been built and that it was now only a question of modifying it. On the contrary, this structure was only virtual, beginning with the high-level schemas, and was developed bit by bit during the process of revision. There was therefore a “live” process of control and change, maintaining as far as possible the functionality of the whole and the consistency of the various parts, both the consolidate ones as well as those yet to be examined. The coherence with all the preceding phases of the manuscript is therefore very strong.


    1. McGuinness, Brian 2002: Wittgenstein’s 1916 Abhandlung, inWittgenstein and the Future of philosophy, Haller, H., Puhl, K., Wien, pp.272-282.
    2. Wittgenstein, Ludwig 1971, 21996: Prototractatus, B.F. McGuinness, T. Nyberg and G.H. von Wright eds., Routledge & Kegan Paul, London.
    Ivi, p. 2. The tendency to consider them new additions leads to paradoxical affirmations. "With the exception of two propositions the new material does not draw on previous manuscripts that we know", McGuinness affirms [p. xii]. Yet, most of the "new material" is derived from the “Prototractatus” itself, i.e. from the preceding part of the manuscript; sometimes, from propositions that in their turn were derived from other manuscripts. This is the case with the two supposed exceptions that McGuinness specifies in his notes: in reality, 3.322 originates from PT 3.2012, which in turn was derived from the Notes on Logic; 5.503 originates from PT5.3003, and in turn originated from Ms103, 13.7.16. An analogous double passage also occurs in 5.2341, 5.25, 5.252, 5.254, 6.1201 etc. The only proper addition that is taken directly from other manuscripts we know of, is the last (unnumbered) remark of the book, which corresponds to the diary of November 21st 1916: "The concept of the operation is equivalent to the concept ‘and so on’ ” (see TLP 5.2523).
    Only later will it be inserted, by hand, into the typescript.
    Luciano Bazzocchi. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.