Against the Idea of a “Third” Wittgenstein
Against the Idea of a “Third” Wittgenstein


In a recently published book, The Third Wittgenstein, the editor, Danièle Moyal-Sharrock, writes that it “seeks to correct the traditional bipartite conception of Wittgenstein’s thought into his Tractatus and Philosophical Investigations by focusing on his neglected last masterpiece, On Certainty, and works contemporaneous with it”. This paper aims to show that such a view is mistaken. It wrongly takes for granted, first of all, that namely On Certainty constitutes a work by Wittgenstein. But the idea of a “third” Wittgenstein is also mistaken because it presupposes that the Investigations, such as posthumously published in 1953, form another work by Wittgenstein, something which is in need of a re-evaluation. I focus then on some issues concerning the origins of the Investigations and try to make clear that the material edited in On Certainty or Last Writings still belongs to (Part I of) the only book Wittgenstein worked at after the Tractatus.

Table of contents


    Much has been written about Wittgenstein’s posthumous publications but, with a few notable exceptions, little attention has been paid to their origins. An example of such lack of consideration is a book edited by Danièle Moyal-Sharrock, The Third Wittgenstein, recently published. As stated in the editor’s introduction, it “stems from the conviction that there is a third Wittgenstein, a Wittgenstein who went beyond what he had achieved in the Investigations”, aiming, then, to “supersede the traditional bipartite division of Wittgenstein’s philosophy crowned by the Tractatus and Philosophical Investigations, and indicate not only a new phase in Wittgenstein’s thinking, but also that Wittgenstein was the author of three, not two, philosophical masterpieces” (2004, 1). This alleged third philosophical masterpiece is On Certainty, something which, as Moyal-Sharrock stresses, was first recognized by Avrum Stroll (cf. 1994, 5). But she goes further, taking “the third Wittgenstein corpus as essentially consisting of all of his writings from approximately 1946”, and “[t]his includes On Certainty, Remarks on Colour, Zettel, and all the writings on philosophical psychology, including Part II of Philosophical Investigations” (2004, 2).

    Moyal-Sharrock is here following G.H. von Wright’s view, shared by P.M.S. Hacker, “that Part I of the Investigations is a complete work and that Wittgenstein’s writings from 1946 onwards represent in certain ways departures in new directions” (1982, 136; cf. also Hacker 1996, xvi). Von Wright, whose name erroneously appears in various editions of theInvestigations as one of the editors, was actually the first to raise doubts about the publication of the two parts together (see 1982, 135-136; 1992, esp. 186-188). However, maintaining that Part I is “a complete work”, he failed to see why Part II really does not fit into the other. I shall thus begin by considering the bipartition of the Investigations.


    The typescript from which Part I of the Investigations was printed is lost but a copy has survived, corresponding to item 227 in von Wright’s catalogue of Wittgenstein’s Nachlass (see von Wright 1993).1 The typescript of Part II, numbered 234, is also lost and, in this case, no copy has been preserved. We know that Wittgenstein worked intensively on theInvestigations, inclusively submitting early versions of it to Cambridge University Press, in 1938 and in 1943 – the latter including the Tractatus. The so-called “Early Version” was based, as the printed Investigations, on two typescripts, items 220 and 221, to which Wittgenstein attached a preface (TS225), where he speaks of two distinct parts of the work. Moreover, in a letter to von Wright, dated 13 September 1939, Wittgenstein refers to “what would be the first volume of [his] book” (LvW, 461), a reference already made in two other letters, to J.M. Keynes, of 1 February 1939, and to G.E. Moore, of the next day (cf. CL, 304-305). Given the reworking carried out in TS(S)222(-224), composed of cuttings from a copy of TS221, it is not likely that the second submission had included only TS239, a revised version of TS220, consisting, rather, also of two parts.2 And, finally, in a letter to Rush Rhees - who edited with G.E.M. Anscombe the Investigations -, dated 13 June 1945, Wittgenstein informs that “[he had] been working fairly well since Easter” and that “[he was] dictating some stuff, remarks, some of which [he wanted] to embody in [his] first volume” (cited in von Wright 1982, 127).

    Now, if TSS220 and 239 correspond to §§1-189a of the Investigations published in 1953, the same does not apply to Part II of the book, since TS234 has nothing to do with TSS221-222, which deal with the philosophy of mathematics and not with the philosophy of psychology. In a recent study, Brian McGuinness actually reports that “the package containing a surviving copy of typescript 227 […] is labeled ‘Philosophie der Psychologie’”, emphasizing that “[t]he title ‘Philosophical Investigations’ was always meant to cover the mathematical material as well” (2002, 286). Wittgenstein himself makes it clear in the published preface to the Investigations. He writes:

    The thoughts which I publish in what follows are the precipitate of philosophical investigations which have occupied me for the last sixteen years. They concern many subjects: the concepts of meaning, of understanding, of a proposition, of logic, the foundations of mathematics, states of consciousness, and other things. (PI, ixe)

    Nevertheless, the editors of the Investigations did not understand the matter that way, editing, three years after – this time together with von Wright –, TSS222 and 223, as well as a wide selection from later manuscripts (sc. 117, 121-122 and 124-127), under the title Remarks on the Foundations of Mathematics.3 This, then, opened the door to all the “post-Investigations works” we know. I shall now turn to them.


    Edited by Anscombe and von Wright, the Zettel were the first “post-Investigations work” that came to light. Both the original edition of 1967 and the revised (English) one of 1981 are based on an arrangement made by P.T. Geach from a large quantity of cuttings found in a box, some of them clipped together, but others lying loose in it, which resulted in two collections of cuttings, TSS233a and b. The problem of deciding, in several cases, where the material should be assigned and the need of completing it in other ones - making Geach use of copies of the cut-up typescripts (mainly 228-229 and 232) or, in certain cases, of his own inspiration - gave rise to a work which may be at odds with Wittgenstein’s intentions.

    Two years after the publication of Zettel, Anscombe and von Wright would take, nonetheless, a further step in the Wittgenstein editing. Selecting remarks from MSS172 and 174-177, they published On Certainty, from which a revised edition appeared in 1974. In the editors’ preface it is argued that “[i]t seemed appropriate to publish this work by itself” because “[i]t is not a selection”, insofar as “Wittgenstein marked it off in his notebooks as a separate topic, which he apparently took up at four separate periods during [the last] eighteen months [of his life]”, constituting it, thus, “a single sustained treatment of the topic” (OC). That Wittgenstein had taken up such topic only then is denied, as Kim van Gennip (2003) rightly pointed out, by a number of related remarks in the undated MSS169-171, which were edited by von Wright and Heikki Nyman in, say, chapters 1-3 of Volume II of Last Writings on the Philosophy of Psychology, published in 1992, as well as in the - according to van Gennip – preceding MSS137(II)-138 (the sources for more than a half of MS144, from which TS234 was dictated), also edited by von Wright and Nyman, in 1982, as Volume I of Last Writings. As a matter of fact, Wittgenstein had explicitly dealt with the topic of certainty already in MS119, which dates from 1937, an item partly edited, with lecture notes, by Rhees as “Cause and Effect: Intuitive Awareness”, in 1976.4 And if this clearly indicates that On Certainty is far from representing a “single sustained treatment of the topic”, any remaining doubts concerning its status of “masterpiece” will be completely removed when one verifies, as van Gennip nicely put it, that “not only are Wittgenstein’s ‘marks’ ambiguous, but the editors applied their own demarcations […] as well” (2003, 129).

    The same holds obviously true for Remarks on Colour, solely edited by Anscombe, from MS173 and, again, from MSS172 and 176, in 1977. And it immediately follows from all this that the edition of Volume II of Last Writings, whose chapters 4-6 derive, once again, from MSS173-174 and 176, is problematic too.

    Yet, these arguments do not seem powerful enough to meet Moyal-Sharrock’s most general claim, that “the third Wittgenstein corpus […] essentially [consists] of all of his writings from approximately 1946”. This brings me back to the Investigations.


    There are plenty of reasons to suppose that, contrarily to what is commonly assumed, Wittgenstein was still working on (Part I of) the Investigations in the final years of his life. In their editorial note, Anscombe and Rhees point to this very fact for, after having written that “[w]hat appears as Part I […] was complete by 1945”, they concede that “[i]f Wittgenstein had published his work himself, he would have suppressed a good deal of what is in the last thirty pages or so of Part I and worked what is in Part II, with further material, into his place” (PI). A much similar, somewhat contradictory, view is held by Geach in his preface to Wittgenstein’s Lectures on Philosophical Psychology 1946-47. He says, on the one hand, that “Part I of the Investigations was complete when Wittgenstein died, and [that he and Anscombe] had already seen the MS of what is now printed as Part II”, but, on the other, that “Wittgenstein intended to have revised the final pages of Part I to incorporate the new material, but he died before he could do this” (xiii). I shall not discuss here all the pieces of evidence for the truly unfinished character of the Investigations I can think of.5 But I shall look at some textual facts which are particularly illustrative of that.

    In a parenthetical remark, written down in MS137 (92b), on 9 November 1948, Wittgenstein observes that “[i]t’s no accident that [he]’s using so many interrogative sentences in this book” (LW I, §150), meaning him with “this book” nothing but the Investigations. This is clear from a remark written down just a few days later, more specifically on 28 November, in the same notebook (112a), where it is said, without no further reference, that “[i]f the language-game, the activity, for instance, building a house (as in No 2), fixes the use of a word, then the concept of use is flexible, and varies along with the concept of activity” (LW I, §340), referring that “No 2”, undoubtedly - and the editors of Last Writings were the first to recognize it -, to §2 of the Investigations.

    In fact, there are various allusions to that same section in previous items6 and two more in MSS175 and 176, the first in a remark from 18 March 1951 (67v) and the second in one from 19 April (62v), which constitute §§396 and 566 of On Certainty.

    If we add then to this another allusion, without no further reference as well, this one to §8 of the Investigations, which is to be found in a remark written down on 7 February 1949 in MS138 (16a), §833 of Volume I of Last Writings, it becomes manifest that Wittgenstein had been occupied, until his death, with Part I of the Investigations, whose final pages, as Geach reports, he “intended to have revised”, in order “to incorporate the new material”. And this, I am convinced, not only shows the inaccuracy of a “third” Wittgenstein, but also, and fundamentally, that if we want to make sense of the Investigations, we have to read them in all their extent.


    1. Conant, James 1998 “Wittgenstein on Meaning and Use”, Philosophical Investigations 21, 222-250.
    2. Geach, P.T. (ed.) 1988 Wittgenstein’s Lectures on Philosophical Psychology 1946-47. Notes by P.T. Geach, K J. Shah and A.C. Jackson, New York: Harvester.
    3. Hacker, P.M.S. 1996 Wittgenstein: Mind and Will. Volume 4 of an Analytical Commentary on the Philosophical Investigations, Oxford: Blackwell.
    4. McGuinness, Brian 2002 “Manuscripts and Works in the 1930s”, in: Approaches to Wittgenstein. Collected Papers, London: Routledge, 270-286.
    5. McGuinness, Brian and von Wright, G.H. (eds.) 1995Ludwig Wittgenstein. Cambridge Letters, Oxford: Blackwell. (CL)
    6. Moyal-Sharrock, Danièle (ed.) 2004 The ThirdWittgenstein. The Post-InvestigationsWorks, Aldershot: Ashgate.
    7. Stern, David G. 1996 “Toward a Critical Edition of the Philosophical Investigations”, in: Kjell S. Johannessen and Tore Nordenstam (eds.), Wittgenstein and the Philosophy of Culture. Proceedings of the 18th International Wittgenstein Symposium, Vienna: Hölder-Pichler-Tempsky, 298-309.
    8. Stroll, Avrum 1994 Moore and Wittgenstein on Certainty, Oxford: OUP.
    9. van Gennip, Kim 2003 “Connections and Divisions in On Certainty”, in: Winfried Löffler and Paul Weingartner (eds.), Knowledge and Belief. Papers of the 26th International Wittgenstein Symposium, Kirchberg am Wechsel: ALWS, 129-131.
    10. von Wright, G.H. 1982 “The Origin and Composition of the Philosophical Investigations”, in: Wittgenstein, Oxford: Blackwell, 111-136.
    11. von Wright, G.H. 1992 “The Troubled History of Part II of the Investigations”, Grazer Philosophische Studien 42, 181-192.
    12. von Wright, G.H. 1993 “The Wittgenstein Papers”, in: James C. Klagge and Alfred Nordmann (eds.), Ludwig Wittgenstein. Philosophical Occasions 1912-1951, Indianapolis: Hackett, 480-506.
    13. von Wright, G.H. (ed.) 1993 “Letters from Ludwig Wittgenstein to Georg Henrik von Wright”, in: James C. Klagge and Alfred Nordmann (eds.), Ludwig Wittgenstein. Philosophical Occasions 1912-1951, Indianapolis: Hackett, 459-479. (LvW)
    14. Wittgenstein, Ludwig 21974 On Certainty, ed. by G.E.M. Anscombe and G.H. von Wright, tr. by Denis Paul and G.E.M. Anscombe, Oxford: Blackwell. (OC)
    15. Wittgenstein, Ludwig 1982 Last Writings on the Philosophy of Psychology, Vol. I, ed. by G.H. von Wright and Heikki Nyman, tr. by C.G. Luckhardt and M.A.E. Aue, Oxford: Blackwell. (LW I)
    16. Wittgenstein, Ludwig 2000 Wittgenstein’s Nachlass. The Bergen Electronic Edition, Bergen/Oxford: The Wittgenstein Archives at the University of Bergen/OUP. (MSS & TSS)
    17. Wittgenstein, Ludwig 32001Philosophical Investigations, ed. by G.E.M. Anscombe and Rush Rhees, tr. by G.E.M. Anscombe, Oxford: Blackwell. (PI)
    A second copy was found in 1993, containing extensive corrections, in different hands. These corrections differ from those of the first copy and from the printed version. The two typescripts are now known as items 227a and b. For a detailed discussion of this issue see Stern 1996.
    It is worth noting that in MS124 (150-151: 18.3.1944) we still find additions to TS222 (87-88).
    A new English edition of this text appeared in 1978, incorporating, apart from other selections from the same manuscripts, TS224 and material from MS164.
    I say that Wittgenstein had explicitly dealt with the topic of certainty in MS119 because, as James Conant (1998, 238ff.) convincingly argued, such topic (or what is at stake in it) is nothing but what underlies Part I of the Investigations.
    That would notably imply an analysis (and dating) of MS182 and TSS227-232, which I cannot undertake here. Note that TSS229 and 232 were the sources for Volumes I and II of Remarks on the Philosophy of Psychology, the former edited by Anscombe and von Wright and the latter edited by von Wright and Nyman, both in 1980.
    Cf. MS165, 94-95 (c. 1941-44), MS124, 192 (13.4.1944), MS132, 203 (21.10.1946), MS136, 53a (3.1.1948), as well as TS 233a, 20-21.
    The writing of this paper was supported by a Postdoctoral Fellowship from the Portuguese Foundation for Science and Technology.
    Nuno Venturinha. 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.