Graham McFee

Main page Curriculum Vitae
Baker and Hacker without Baker? Wittgenstein: Text and Context
"On first looking into ...": The placing of the 'slips of paper' ... Wittgenstein, Systematicity and the Use of Philosophy
A Nasty Accident with One's Flies [Inaugural Address] Everything Goes with Beer

Wittgenstein: Text and Context

My aims, in this work-in-progress, are to exemplify the importance that I would give to a proper historicising of claims of philosophy by showing how their historical position bears on problems in reading the texts that have come down to us as "the later Wittgenstein" -- and, in particular, the published texts. For, with any philosopher, a key line of objection concerns the consistency of his claims: and, to assess consistency, we must know what words to assign to the philosopher, how to read or interpret those words, and (faced with differences) how to weigh some as against others. This task is complicated in the case of Wittgenstein: since all those published works are Nachlass (that is, a legacy of material unpublished at his death), they may appear to all have the same status. And that impression is strengthened by the details typically given in the editorial forwards, translator's introductions, and the like. In this vein, I try to offer a commentary which will allow a useful classification of the works. (Although I have read versions of this material at philosophy societies, I have no plans to publish it.)


This is an odd paper (in addition to being a 'work in progress') because, if its principal thesis is rejected, it probably follows that this paper is not after all a (potential) contribution to philosophy ... but at best a contribution to (say) the history of philosophy. Yet/so a part of my purpose is to contest that contrast, at least in that version.

Suppose you were a teacher of philosophy, faced with teaching (say) Descartes: you could, of course, begin with the French and Latin texts of his works -- not perhaps them alone, but certainly including them. Yet, for many of us, that is not a helpful starting place -- our mastery of French and/or Latin is not adequate to the task. So you might decide -- for most of the class at least -- to simply use an agreed translation: I quite like that by John Cottingham and his pals, published by Cambridge University Press. Then, there may be some issues where I would comment on the translation (for instance, the way conscientia is translated as "consciousness" conceals its relation to conscience; and hence its moral character (DD p. 102ff.[1]) -- students might benefit from seeing that). But, in general, we would read together the text (in English) in front of us.

My discussion of texts of Wittgenstein -- if you like, of the twenty-plus books on the shelf in my study with "Wittgenstein" on the spine, in the author's spot -- is explained partly because the attitude I have just described for Descartes will not work for those texts, for reasons I will come to. But my thought is that -- once we have learned from Wittgenstein's case -- we will revise our mode of approach to all (or, anyway, most) philosophers: of course, that revision may not make much difference in practice ...

 The key question in the background is how does philosophy relate to its history? And, of course, I will not even be trying to resolve that question today.

 So the centre of the talk is a discussion of Wittgenstein's works (a scholarly discussion[2]) with a top/tail of general issues and implications ...

 §1       Preface and problem

Two general points, by way of preface:

1.        The first concerns Anglo-American analytical philosophy's problematic relation to its past: that is, its ahistorical conception of the project of philosophy -- for many figures from philosophy's past, the conclusion is "genius, but ...". The most extreme case here is probably Descartes, where -- despite Descartes' widely-acknowledged greatness, as a founding father of modern philosophy -- almost all the books on him tell the reader, on the first page or so, what a great genius Descartes was; but (by page three) he is accused of making all the silly mistakes one (or the author) can think of. Were this right, the admiration for Descartes would be odd, and unwarranted.

           But many theorists studied are taken to be sufficiently far into the genius category for us to keep reading them after all these years, and sufficiently far from it that we will safely conclude that most of their claims are misguided, misplaced, or some such ... And this treatment derives from regarding them as working on our problems. But (for me) it won't do to treat key figures from our past as colleagues on sabbatical, with no cell-phones. Or so I will urge in a moment.

2.        The second issue concerns the importance of determining, for any philosopher under discussion, what X said, since accusations of inconsistency are at the centre of our repertoire (as they almost always are in philosophy): we need to know what words to take, and what place to award them, in considering what problems or issues particular philosophers were facing -- and what they could (and could not) take for granted. We can only look for inconsistency once this is achieved.

Importantly, we cannot simply rely on "going with the words" (as some postmodernists [among others] urge), since the words (chosen) may be contentious or difficult to determine, often in four ways (which I'll mention here, with some exemplification later), as follows:

 (a)     determining the correct words to consider (and what weight to give to them) can be problematic, especially with unpublished or posthumously published works;

(b)      we often are considering translations, where what word we select may be crucial, in lots of different ways [[Three examples: [i] PI §116 original translation has Sprache translated "language-game" -- that is just plain bad translation; [ii] PI §402 includes French [faute de mieux] as way to translate German [zur Not] into English; [iii] Ms 137/ Z §165 on Partie, where what is literally "I have won the party" becomes (correctly) "I have won the match": but that does not help us to see what analogy Wittgenstein was drawing -- I'll return to this one[3]]];

(c)       the words may amount to something in the specific context (perhaps, the specific temporal context) which we do not take into account;

(d)      (even if/when the other three problems are fixed) apparent inconsistencies may have other explanations: it might be that X changed his mind (as, say, Collingwood arguably did) or addressed an inconsistent-sounding remark to a different context.

More importantly, the question, "What did so-and-so mean?" is connected to the question, "What are so-and-so's issues?" -- and we cannot determine that (at least readily) simply by scrutinising so-and-so's text alone [[I won't discuss this in detail, but ...]]

 In part, then, the problem here lies in how we make sense of the context within which the philosophers we consider make their interventions; and what weight we give to that context. And I take this to be crucial: so I'd claim, as Gordon Baker (1986 p. 45[4]) urged, that:

[t]here is no such thing as grasping the significance of a philosophical thesis independently of understanding the archetectonic of the thought of its author.

And this can be problematic: philosophers do change their minds, so that later works might be given preference for that reason. Equally, some very late work might not find a publisher until after the philosopher's death -- if then. So, at least sometimes, we might find reason to prefer unpublished works (or posthumous works) to published ones. But that would not be our typical assumption about priority, especially for philosophers who had ample opportunity to publish. Moreover, we expect philosophers to own up to changes of mind -- as, say, Hilary Putnam has done.

           My plan has me both exemplifying and discussing these points in the later writings of Wittgenstein. To introduce this discussion, I offer (as a third general point) a kind of slogan:

3.        For Wittgenstein, (all?) later works are Nachlass (legacy): that is, works left unpublished at the author's death. And this is important because publication of works from the Nachlass will inevitably depend for their authenticity on editorial practice, principles and decisions. In illustration consider one case: in a discussion of the editing of a text published as "Cause and Effect: Intuitive Awareness" (PO p. 368), we are told that its editor [Rush Rhees]:

... noticed ... that though Ms 119 had been mined at least twice by Wittgenstein for apparently different purposes, there still remained a number of interesting passages ...

As I read it, this description (which accurately reflects Rhees's own comments):

(a)       presents the remarks as (typically) 'used up' by having found their way in some version into some typescript -- that is why Rhees chose from what had not been so "mined ... by Wittgenstein";

(b)       formalises or instantiates a view of what is good or important in Wittgenstein -- hence creates a kind of 'self-fulfilling prophesy' for Rhees's conception of Wittgenstein's project. And, of course, this should be contrasted with other editorial decisions across the whole range of publications from the Nachlass. (I'll give one example shortly.)

           So the thought that Wittgenstein's (later) works are all Nachlass might make it seem that they all have the same footing, but this cannot be right -- in ways I will try to make clear.


§2       Wittgenstein's role

I chose Wittgenstein's work partly because I am especially interested in it, partly because I have done some work on it that I'd like to share with you, and partly because its problems -- although not all shared with many philosophical oeuvres -- are revealing in ways that do apply quite widely. It particular, it is both puzzling and essential for a proper appreciation of Wittgenstein's work to ascertain exactly what he was saying. And that will certainly involve us in weighing various writing since (as I said earlier) all this later stuff is Nachlass.

           We might see my thinking here as especially timely. For it is pertinent to think again about Wittgenstein's texts in the light of the publication of the so-called "Big Typescript" [BT -- Ts 213], since the standard view urged that the publication of individual works of Wittgenstein (as distinct from the whole Nachlass) was complete "... save for the Big Typescript, perhaps" (PO p. 504). And, interestingly, this work has appeared in a "German-English Scholars' Edition"[5] -- unlike any of the previous texts published by Blackwell in including textual variants, and such like. So its publication is seen more directly in terms of a scholarly project in respect of Wittgenstein.

           With some other philosophers -- for example, Descartes -- a part of the problem of coming to an acceptable reading lies in not jumping too soon to take his projects in terms of ours. Suppose, for instance, Descartes tells us (as though it were commonsense) that we humans have seven senses, including two internal senses, where this was  "... an integral part of the framework of Aristotelian psychology which he [Descartes] inherited" (DD p. 126). We should not treat Descartes's comment as automatically just a misleading or confusing way of referring to the five senses plus introspection (DD pp. 48-50;124-138[6]). So, there, the danger lies in running together the different 'world-pictures' (or, better, mind-styles) of different periods. This was a topic to which Collingwood was especially alive. In a discussion of the state, Collingwood [Autobiography (Oxford: Clarendon, 1939) p. 62] comments:

... the history of political theory is not the history of different answers given to one and the same question, butthe history of a problem more or less constantly changing, whose solution changes with it.

The central thesis, then, is that Plato's problems, or Hobbes' problems, for example, are radically different from those of a modern political theorist. There is not a single issue to which all are directing their attention. Dismissing the past is basically ignoring this sort of point, since to recognise it would suggest that the problems of the past are not obviously ours.

           As I said, that version of the 'problem of past philosophical works' is not primarily our problem with Wittgenstein, although I will argue that the problem of making sense of Wittgenstein is still one of giving due weight to -- and hence not running together -- what I am following Gordon Baker in calling mind-styles.

           For instance, it is a standard among some classic Wittgensteinians -- Anthony Kenny and Normal Malcolm most prominently, although Peter Hacker can also write in this vein --that Wittgenstein's target (especially in the so-called "Private Language Argument" or PLA) is Descartes. Since Descartes is part of the conceptual landscape of most of us with philosophical training, it seems so likely that Wittgenstein would have needed (somehow) to make his peace with Descartes. Yet, if we do think that, we forget that Wittgenstein's training was very different from ours. Descartes does not loom large in the Nachlass -- indeed, there is no evidence that Wittgenstein had any direct acquaintance with his writings, as Gordon Baker (2004 p. 117) has pointed out (see also Baker, 2004 p. 3). But we know that Wittgenstein had studied in detail both Frege and Russell: and that both of these did see their project as contesting (what they took to be) Descartes's concerns and conclusions. As a route to conclusions about Descartes -- say, as part of an appreciation of his philosophy -- tracking Wittgenstein's concerns through Russell's (or Frege's) looks improbable: did Russell really grasp Descartes's mind-style well enough to begin contesting Descartes's answers to Descartes's questions (as opposed to what Russell took to be Descartes's answers to Russell's question)? [Those tempted to answer "yes" should read, in particular, Russell's account of Kant, in any of his works but most vividly in History of Western Philosophy: Russell does not seem to know what is going on in Kant.]

           My point is just that reference to Descartes is a red-herring here: we have to treat the topics Wittgenstein wrote about (and his manner of treating them) to see whether he was at odds with Descartes. And one might, say, insert Ryle's Concept of Mind, as taking Wittgenstein's points into Descartes's territory -- but this work too is not marked by its scholarly concern with Descartes. Moreover this whole strategy is likely to be unproductive for both our understanding of Wittgenstein and our understanding of Descartes: there seems no reason to imagine they were 'working the same street'. Moreover, our grasp of Wittgenstein's biography explains to us at least one reason why this might be so. But my suggestion -- the moral here -- is that we should always be approaching a philosopher by aiming (as best we can) to approach what I am calling his/her mind-style. Of course, this will always be a kind of dialectical relation with the texts he/she left us -- when this happened (compare Aristotle). For Wittgenstein, our real problem is that we have a lot of texts, but lack a way to treat them judiciously.

           But perhaps that does not quite catch my problem: I suppose I really think that the net effect of the publication of Wittgenstein's work in its present form is to make all (or anyway, most) of the published texts appear 'on a par' -- and to prioritise those texts over the unpublished work. And, if asked, I would begin explaining this fact partly by pointing to the force (if one is not paying close attention) of the beautiful 'black and white' of the printed word -- especially when contrasted with the scribble of some manuscript -- and partly through the inappropriate comparison with philosophers who do publish their mature thoughts: we have good reason with (say) Collingwood to prefer his published thoughts on certain topics to unpublished ones -- and that reason amounts to saying that he could have published them if he had wanted to. And that, in turn, finds a place for works where this was not true -- in Collingwood's case, what is now published (posthumously) as Principles of History (Oxford: Clarendon, 1999), for instance, which he wrote very late in life.

           {{As a digression: compare Nietzsche -- on my understanding, we know he could have published whatever he wanted [he was doing a kind of vanity publishing at the end of his most productive period], but weight is still given in the secondary literature (a) to a manuscript he scrupulously did not publish ["Truth and Lies in the Moral Sense" -- De Man, etc.]; and (b) to a collection cobbled together by his crazy, Nazi sister [The Will to Power -- of which, incidentally, she sent a copy to Wittgenstein: we know he thanked her, but I know of no evidence that he read it. Given Wittgenstein's political sympathies, it is hard to imagine he would not have commented on it if he had read it. But, then, we know that Wittgenstein had an aversion to reading philosophy: as he put it, in a letter to Von Wright (of 22 December, 1947), "real reading is always bad for me" (PO p. 466).]}}

           But neither stressing what is published (over the rest) nor stressing late (and therefore) unpublished works explains our problem with Wittgenstein's later work, since it all started off as Nachlass; nor does this offer good reasons in Wittgenstein's case for prioritising this work over that one. So a judicious weighing of the published texts we have and of the unpublished work is called for. Tonight I will both offer some of that, and some comments on it.

 §3       Some principles

Let us turn to some ideas to be deployed in making sense of philosophical texts. In granting that it might be relevant that Wittgenstein was concerned about, and pursued, the problems raised by Russell and Frege, we are acknowledging one basis for deciding how to approach Wittgenstein's works: do they make sense if we think of them as addressing some assertion (or some problem) in Frege or Russell -- for instance, Russell's advocacy of a causal theory of meaning? The general connection between, say, question and answer supports such a strategy for trying to make sense of Wittgenstein's ideas, given (only?) his Nachlass -- together with what may be inferred from other sources: for example, from his biography, from who he read, and who he did not read ... and at what issues or questions his work was directed.

           For example, asked by John Wisdom in 1944 for content for a proposed dictionary entry, Wittgenstein offered just one sentence: "He has concerned himself principally with questions about the foundations of mathematics" (Nedo, 1993 p. 57[7]), which shows something of his view of himself and his work at that time --but then (to give that substance) we might add the connection, explicit through his dealings with Waismann, to the specifics of these topics; especially on the issue of exceptionlessness (where Wittgenstein thought about the infinite, about [mathematical] induction, and about the scope of "all".) Note too Rhees's down-playing of Wittgenstein's work on mathematics [see also Nedo, 1993 p. 58]: if we do not see Wittgenstein's work being as centrally related to mathematics as his own comment to Wisdom suggests, one explanation might rest with the editing of the Nachlass -- equally, Wittgenstein was surely over-stating the case, perhaps reflecting what he was presently working on.

           A contrary basis for reading the Nachlass is offered in the Vienna Edition, which identifies a 'logical centre' to the Nachlass in one of its constituents:

"The backbone of the complete works is formed by the manuscript volumes" (Nedo, 1993 p. 81)

{{(a) Von Wright's catalogue: PO p. 482 ff; (b) Nedo (parallel) catalogue p. 81.}}

Were this right, we should be publishing the manuscript volumes as finished works (as the Vienna Edition intends to).

           But it seems odd to think of the manuscript volumes as so crucial, when (in fact) Wittgenstein seems to have worked with the typescripts -- cutting and pasting them, etc. (and, occasionally, re-inserting material from manuscripts). As he wrote to Von Wright on 6th November, 1947 (PO p. 464), he wished to have his current work "... in a handy form, i.e. typewritten".

           I am not of course trying to resolve this dispute here. Rather, and at the very least, this shows that we must come to the words in the Nachlass with ways of deciding between competing bases for editorial intervention or structure: we have seen how biographical considerations might play a role here.

           So that recognition of the force of biography, with its connection to the problems approached or the perplexity addressed, is our first general insight here. To make the other matters more concrete, I will highlight three further, related kinds of issue, which might directly offer principles for making sense of Wittgenstein's texts, with some examples (for the first two):

 ·          Knowing the linguistic context: my example is the translation of the word Partie in Ms 137 p. 20 ff. (15th February, 1948), which is published as Z §165: "I won the match"

The context for the analogy Wittgenstein suggests is given by his discussing a form of words that, despite the fact that we readily say it, can be misleading: if we say, "I experienced that [musical] passage quite differently" as a way to describe our understanding a musical passage differently, it is easy to give too much weight to the term "experience". But, Wittgenstein insists, this form of words -- with its apparent reference to experience -- is only misleading to those not "at home in the special conceptual world that belongs to these situations" (see Z §165: see McFee, 2001 p. 96; McFee, 2004 p. 112[8]): that is, to the world of music. Then Wittgenstein offers his informative analogy, obscured in the translation: the German text literally says, "Analogy: I have won the party". It is a useful analogy here because this way of describing victory is used only for matches at chess (and other serious board games)[9]. And so is accurately rendered into English (as the translation has it), "I won the match". Someone who understood the German words but knew nothing of how serious board games such as chess are discussed (in German) would/could take this as some oblique remark about parties. So the analogy is that, in the cases of both music and this remark about (say) chess, one needs more than just an understanding of individual elements. Then those not part of "the special conceptual world" of chess etc. will typically fail to make sense of that remark -- taking it as about parties or failing to make sense of it at all. And we (readers) will miss this point to the extent that we fail to understand the analogy: that is, fail to grasp its linguistic context!

·          Understanding the archeology of the works: PI §109 speaks of "... battle against the bewitchment of our intelligence by means of language"

           Ossie Hanfling[10] thinks that the battle is by means of language: hence, Wittgenstein is a philosopher concerned (centrally) with language and, when we add the idea that we "... bring words back ... to their everyday use" (PI §116), Wittgenstein turns out to be an 'ordinary language' philosopher.

           But the ancestor of PI §109 is Ms 153a [Nedo, 1993 p. 90]: "We are engaged in a fight with language" -- and then we see that the bewitchment is by means of language, so the battle is with language (with language as opponent), rather than (as on Hanfling's view) conducted via it. And a consideration of Ts 213 (PO p. 167: see Baker, 2004 pp. 101-102; McFee, 2002 pp. 99) reinforces for us the importance Wittgenstein placed on the passage Hanfling omitted: that we bring words back from their metaphysical to their everyday use -- so that the metaphysical use was the one Wittgenstein hoped to discuss, the other being simply a default position. Indeed, before settling on "everyday" as the word here, Wittgenstein expressly wondered about "correct" or "normal" -- either of which would have a very different relation to, say, recognising that Wittgenstein was not suggesting an empirical enquiry..

           This respects Baker's emphasis on archeology (Baker, 2004 p. 3), highlighting that it must be a judicious archeology, making choices ... But it offers at least one way to address a genuine problem.[11]

 ·          Taking Wittgenstein at his word (for which, of course, we must have his words, etc -- not those of students): As Baker (2004 p. 66) urges:

... we should proceed on the basis that the texts which Wittgenstein constructed for himself consist of carefully thought out arrangements of remarks whose precise wording was of paramount importance.

That is, we should recognise Wittgenstein as a meticulous writer. But, as Baker continues ruefully, this assumption:

... does not apply either to the texts compiled by editors in various more or less systematic ways from his manuscripts or to texts compiled from lecture notes taken down by others.

Now I do not want to give the impression that this would be easy in any case -- and certainly not in Wittgenstein's. Everyone (well, almost everyone) knows about "Wittgenstein's frustrating search for adequate formulations of his ideas" (PO p. ix).

           So he regularly struggled to find the right way of making his point (tout court) rather than, as his ideas should have predicted, the right way in this context or that one. Thus we know that -- perhaps contrary to some ideas in his philosophy[12] -- Wittgenstein the person believed in a kind of mot juste, the exact expression of his ideas; and this was connected (although not as directly as might be hoped) with a philosophically crucial idea: what he called the erlösende Wort (the magic or redeeming word [PO pp. 164-5], the word that sets us free [Nedo, 1993 p. 91 {top}]) -- which, if only we could arrive at it, would deal with our perplexities. But, unlike the mot juste, this was person-specific and occasion-specific -- it dealt (at best) with your problems, in this context (for instance, Baker, 2004 p. 159). That this is not a promising basis for a philosophical text is visible in the best moment in Derek Jarman's film Wittgenstein (1993), where we see the erlösende Wort in action: the 'bright young things' say that the reason 'people' believe that the sun goes round the earth is that this is how it looks -- and Wittgenstein asks how it would look if the earth went round the sun! It takes a moment but the scales fall from their eyes -- it must look like this, since this is what does happen.

           But notice first how specific this is: it deals only with the particular perplexities of these particular persons. And perhaps a book should find itself in other hands. Again, Wittgenstein stresses the way that philosophical perplexities appear as disquiets, etc.: "... the focus is on a distinct range of intellectual emotions or disturbances"(Baker, 2004 p. 182[13]) {NB PO p. 165 "(psychoanalysis)"}:

...unrest (PLP p. 7; F 93), torment, disquiet (PI §111; LSM 33), discomfort (BB p. 26; PLP p. 4), drives (PI §109), obsessions (HISP p. 18), craving (BB p. 17; LPM 58), revulsion (BB p. 15, p. 57), Angst (BB p. 27; F 94), irritation (PLP p. 7; F 62), profound uneasiness of mind (HISP p. 3), profound mental discomfort (HISP p. 6), obsessional doubt (HISP p. 8), shock (PLP p. 7), troubles (BB p. 46), compulsions to say things (BB p. 47), irresistable temptations (BB p. 18), alarm (HISP p. 4), etc.

Here too the ideas developed are ill-suited to the kinds of general presentation one might imagine for a book.

[But Wittgenstein had a response on the final point, although it is one we typically do not recognise: he imagines a debate with the kind of intelligent concerned person one might find in a Cambridge seminar (Baker, 2004 p. 113): so he has a set of perplexities (or perplexed persons) in mind -- and, again, this is poorly served by seeing the works as a debate with "the interlocutor" if he/she is taken as a cantankerous philosopher of a hostile persuasion or even worse when the debate is with "the Private Linguist".]


§4       Texts and categories

I will begin the specifics here by mentioning just some of the complexities we must deal with; and some of the categories we might need -- and then (in the next section) I'll expand these a bit in terms of four problematic cases (or less, given time constraints), before concluding. (There are a bunch of fairly uncontentious categories and a bunch of problem cases.)

[Explain Ms/Ts contrast]

Let is begin with some uncontentious categories:

1.        The only candidate for a (nearly) finished work -- PI Part One: originally Ts 227, not worked on after 1946. Von Wright, in his catalogue, makes no comment, but "[t]he original typescript of the final version of Wittgenstein's ... [Philosophical Investigations] can no longer be found" [Nedo, 1993 p. 43]; "the carbon copy has the number Ts 227" [Nedo, 1993 p. 43].

This is undoubtedly Wittgenstein's masterpiece, a work that he left in a more or less finished state: but there are a few problems, of which I will mention four:

(a)       The carbon copy has hand-written remarks by Peter Geach, which he claims are 'transcriptions' of remarks by Wittgenstein, but we can check the accuracy of neither the remarks themselves nor their location;

(b)       There were  slips of paper which "... Wittgenstein had cut from other writings and inserted in these pages, withoutfurther indication as to where they were to come in" (PI "Editors' Note"): the remarks are printed at the bottom of pages -- having no record, we cannot be sure where these remarks were intended to be: at best, we might assume that Wittgenstein had a consistent policy here, so that the 'insertion' always related (say) to the page behind it, but we have no real basis for that assumption -- and the revision of the pagination of the text of PI (for the "50th Anniversary" edition) further complicates matters [REF to my notes?].

(c)       Is Part One a finished work? Especially, what is its relation to Part Two? Von Wright [Nedo, 1993 p. 55], one of Wittgenstein's literary editors, comments:

I lean, myself, towards the opinion that Part I of the Investigations is a complete work and that Wittgenstein's writings from 1946 onwards represent in certain ways departures in a new direction.

In a similar vein, Baker and Hacker (Commentary p. 7) accurately note:

There is no published evidence, nor any indication in the Nachlass, that Part II was conceived either as a continuation of Part I or as material to be worked into it (as suggested by the editors).

So there is (at least) a question about the unity of the text as we have it (and the question is easy to miss).

(d)       Problems with the translation: some of these have been fixed (for instance, the translation of Sprache as language-game in PI §116 -- although one might still query whether "language" is the best translation: compare Baker, 2004 pp. 60-62). As an aid to commenting on the published translation here, we have Ts 226, a translation of the early parts of PI, corrected by Wittgenstein -- and some of his (harsh) comments suggest the translation is still not to his liking [compare PI §32, where Wittgenstein crossed out the version now in the text].

           [[We might here comment on people reading "a book by Miss Anscombe" (Hanfling).]]

           The net effect is to make us hesitant in committing ourselves wholeheartedly to PI as we have it -- although the significance and the broad direction of Part One are not really contentious.


2.        The most finished of the (very) late texts -- PI Part Two: originally Ts 234 "... probably dictated in 1949" (PO p. 491) [[The rest of the very late work is problematic -- see §5 (c).]]

           "This typescript was unfortunately lost ... during the publication in 1953" [Nedo, 1993 p. 46]; it "... can unfortunately no longer be located" [Nedo, 1993 p. 55] -- both uses of "unfortunately" here betray a lack of understanding of the publishing process of the time, whereby the typescript used to prepare the pages was heavily marked-up by the printers, and discarded after use (a similar process was still in place in 1992, for a typescript of mine!) -- compare the fate of Collingwood's typescript used for the publication of Idea of History (1946; revised edition, 1993[14]) -- which is not, of course, to deny that the loss of these typescripts is not unfortunate!

           'Luckily', we have Ms 144, a "fair manuscript copy ..." (PO p. 487), from 1949, of roughly Part Two of PI. But:

(a)       there remain issues for the dating of the Ms -- Wittgenstein certainly thought of it as a kind of summary of the work he had been doing -- did he make it to take with him to the USA, to show to Malcolm (as I believe), or was it dictated on his return (see Monk, 1990 p. 542[15]; also Nedo, 1993 p. 46)?

(b)       there are differences with the organisation of Ms 144, beyond the one noted (concerning the placing of the final section). In particular, since there are no section divisions in the Ms except those given by 'starting on a new page', section xi might profitably be sub-divided (see McFee 1999 pp. 267) -- it is by far the longest. And such a sub-division might alter how it should be understood. Of course, the Ts might have had such sub-divisions ...

(c)       Its relation to Part One remains problematic.

Nevertheless, the work as a whole, Philosophical Investigations, certainly is Wittgenstein's masterpiece, the nearest thing he left to a finished work: Part One was not worked on after 1946, and Part Two was at least a summary of some of his work after that, for a trusted colleague. Nothing else from the Nachlass has anything approaching a comparable claim to authenticity and authority as "later Wittgenstein". But, even here, we should recognise the text as at least probably a false unity!

 3.        Writings on mathematics -- the crucial text is RFM, initially as occupying:

... a nearly unique, and not altogether happy, position among the posthumous publications. (PO p. 502)

           Even in the later (1978) edition, the text is a collection of passages from Wittgenstein's writing over a variety of periods, but without that being clear from the text. Most sections consist simply of "selections from" (PO p. 508) the various manuscripts. But Wittgenstein clearly regards this as a central area of his thinking[16]. (We should at least consider that the present policy of separate publication obscures that fact. Equally, he did lecture on this as a separate topic.)

           Wittgenstein's views here have been criticised [see Nedo, 1993 p. 58]; but at least part of this criticism is explained by the fragmentary character of the text as we have it -- in particular, by the separation of the passages dealing with mathematics from others. (for example, in Ms 119 which supplies many remarks for RFM [see PO p. 368], the "Cause and Effect: Intuitive Awareness" passages are much less clearly differentiated from the rest than Rhees implies). We know (from WWK, "Diktat für Schlick", PLP) that Wittgenstein regularly used mathematical examples to make other points: this is far less clear in the other texts, because the mathematical examples have (largely) been collected here.

           A better policy might see the material here as more strongly integrated with the rest of Wittgenstein's work.

 4.        Works presenting a moment in Wittgenstein's thought. A. example of this fourth category: PR -- a text, from a particular moment, 'replicating' one of Wittgenstein's (Ts 209 -- 1930) [[similarly, RPP 1 is Ts 229 (and TS 213 is ... Ts 213!).]]

 5.        Dictations: most famously, BB as dictations (BlB 1933 {Nedo, 1993 p. 32}; BrB 1934 {Nedo, 1993 p. 33}); but see also "Diktat fur Schlick" (Ds 302: see VW pp. 2-83); Wi: MS {see below, but also PO p. 500: "essentially a typescript version of [Ms] 140"} -- these dictations offer Wittgenstein's words, more or less, and at a particular moment. So, as with category 4, these should be taken seriously -- especially when, as with BB, Wittgenstein revised them. But we should watch for editorial intervention, and be wary of seeing all in the published texts as in this category: here, for instance, the thought that BB is "Preliminary Studies for PI" (the sub-title on the text) is very misleading.

 6.        In a separate, special category belongs the text published as Zettel -- As Von Wright explains it:

Among Wittgenstein's papers there was ... a box containing a huge collection of cuttings from various typescripts but mainly from the time after 1945. The fragments were partly loose, partly clipped in bundles. An arrangement of the cuttings by Peter Geach was published in 1967 under the title Zettel. (Von Wright, 1982 p. 59; PO pp. 502-503)

           I should explain that method which generated the Zettel (Ts 223) was not unique, in that Wittgenstein often worked by cutting up typescripts and then re-glueing the slips of paper into a new order (an old version of 'cut-and-paste' on a word-processor). And these were clearly slips which Wittgenstein thought worth saving[17].

           However, the description I quoted from Von Wright is inadequate in two respects. First, it doesn't make clear that the published arrangement is wholly Geach's: that no record exists of what remarks were and were not clipped into bundles. So any organisation Wittgenstein gave to these remarks is lost -- or at least cannot be checked against Geach's. Second, Von Wright's account gives a misleading impression as to the variety of the sources. While Von Wright is correct in saying that most of the remarks in the Zettel originate in 1945 or later (approximately two thirds do so), this does no justice to the variety of sources -- that some go back to manuscripts from 1929 (source: Maury REF).

           The Zettel, then, clearly represents material Wittgenstein thought worthy of preservation -- more generally, important. But the text as we have it is at best an uncertain guide to Wittgenstein's view of how these remarks fitted together. The present organisation is Geach's, "... Wittgenstein's [partial] arrangement of the cuttings was destroyed" (Nedo, 1993 p. 85) -- certainly it has an organisation: Von Wright thought it the middle panel of a triptych.

 7.        students notes (I will comment on the kind of errors which we both might expect, and do find)  Of course, there is a general problem with using students notes: I certainly wouldn't want my thought represented by the notes students have taken. But even when we can be fairly confident that the lecture-notes reflect some of what Wittgensein said (as we are in these cases), one huge problem is that -- unlike Wittgenstein's own writings -- these notes lack the modifications, the "here one might say ..." or "I would like to say ...", which pepper Witgenstein's own writings. It makes the ideas seem definite, non-contextual, in ways his own writings never are. As a result, we can never look to more than gist from such such notes -- in particular, they can never be used to resolve issues of consistency.

 These categories must be deployed for any account of the relation between Wittgenstein's published works and his thought, given that all the publications are posthumous. There are, however, a few cases which fit poorly into the categorisation -- and which (in some cases) perhaps suggest further categories to be added. These will be addressed in the next section.


§5.      Problem cases.

My cases of kinds of problems (which I'll designate with letters):

A.       A clearly problematic text is RPP 2 (= Ts 232): given Wittgenstein's 'cut-and-paste' (literally) procedures, I suggest that this typescript was just a 'cut-and-paste' object, not a book-draft. Here, although it is published, it really should not have been.

           The Cornell microfilm describes Ts 232 as "A typescript rearrangement of Bande Q & R" [that is, Mss 136 and 137]. The editors' decision to publish indicates that they took Ts 232 to be a relatively finished object. Hence, for them, this passage has some role which relates to its being in this place in the 'argument' of the typescript.

           In contrast, I conjecture that Wittgenstein intended Ts 232 as raw material for the (literal) 'cut-and-paste' he used to compose later typescripts. My evidence? Well, it simply types-up remarks from the manuscripts, following almost exactly the ordering of the original notes (except for a few minor changes [for instance, the transposition of §561 and §562]), as though merely copy-typed from the notebook. In that sense, it is not "a rearrangement" of that manuscript material at all. If Wittgenstein were composing a book here, it seems likely that he would have done at least some re-ordering at this stage, in line with his usual policy. And he did not. First, issues Wittgenstein would have been able to resolve remain: one word from the last remark, difficult to read in the manuscript, is queried in the typescript (and ultimately changed: I do not know how, or by whom -- the editors' "Preface" to RPP 2 gives no indication of such changes). Had Wittgenstein meant to change the word (say, during dictation), he would have done so (or indicated alternatives -- a common practice of his). Second, the last passage is enclosed within double-lines, as though meant to be moved: this simply replicates the manuscript's double-lines -- Wittgenstein could have moved the passage. Although not more than suggestive, together these features suggest that Wittgenstein did not select this ordering for the typescript: hence this location is not revealing. (As I've said, the editors of RPP 2 implicitly took a different view.) At the least, this is much less obviously a moment in Wittgenstein's thought.

           So, to repeat, in my view, this work should never have been published: its publication suggests that it has a unity which (as the comments above suggest) it lacks.


B.        A different case is reflected by PG, since this could have been a text reflecting a moment in Wittgenstein's thought -- as say, PR is: see §4.4. (That would have require the publication of BT = Ts 213 -- which has now happened!) The story, in outline is this: Wittgenstein redrafted the first two-thirds of Ts 213, into Ms 114 [PO p. 502]; then he re-redrafted the first third (into Ms 140). And this is (almost) what we have in the published version -- the first third double revised, the second third once revised, the last third unrevised.

           This clearly does not represent a moment in Wittgenstein's thought: there is no reason to think he was happy about the bit he did not revise, but only that he didn't get round to it.

           However, the published text is in reality a yet bigger mess; especially in respect of:

           (a) omissions -- there were chapters in the texts as I've descriobed it 'missing' in the published version. But since, since there were unchanged in final revision, they shopuld have been included {so Rhees didn't even follow his own policy}; and he did not mention these chapters (in his editorial comments). Moreover, since one of them -- subsequently published -- is on the nature bof philosophy, it might well be thought a central concern in a Wittgenstein text;

           (b) in relation to MS 116 -- this was a yet later revision that the ones Rhees used (as more recent scholarship -- especially by Stephen Hilmy -- has shown). So, first, (by Rhees self-avowed principles) this should have been the version included for publication; and, secondly, Wittgenstein clearly returned to this with a plan to use (re-use?) it.

           And there is a further complexity, that it is difficult to place:

           (c) Wi: MS, a version of Ms 140 (the Grosses Format, offering the 'Zweite Umarbeitung' of Ts 213: see VW p. xlvi note). [[suggests Wittgenstein wanted to use this bit -- but what about the rest?]]


Both A and B above are specific cases of problematic texts: a different problem arises when we turn to a collection of texts which put together elements from the Nachlass.


C.       Some important late works (OC, ROC, LW 2): clearly, these were from notes towartds the end of Wittgenstein's life, which he did not revise. It is tempting (and probably correct) to think them important.

           For these, though, the issue is more strictly editorial. For the problem is that -- although interesting in their own right -- the published texts suggest both that Wittgenstein was working on three disparate topics towards the end of his life  (I'd say that the three were more firmly integrated) and that the publication was just a matter of dividing the material where Wittgenstein had done: That is, that there was little or no editorial intervention[18]. Moreover, the status of both OC and, to a lesser degree, ROC as collections of material brought together by the editors is obscured in the published texts -- indeed, the account of its origin in OC reads as though it was basically all (and only) the content of a cetain notebook. But, as the account of these texts in PO (p. 509[19]) makes plain, there was a fair amount of decision here.

           Morover, the texts are not all treated the same: LW 2 is presented, in published version, as Nachlass. But, although this suggests (correctly) that the text is a bunch of fragments, it obscures their relation to the other pieces of writing from the same time and the same note books. (see the account in PO p. 509, referreed to above.)

           So my suggestion for making sense of Wittgenstein's last writings must lie in unscrambling them. Or, since that is unlikely to be done by the publishers, in treating them in one's own work as a set of overlapping texts. [In justification of the unity and provenance of OC, Malcolm tells us the Wittgenstein was 'worked up' about Moore's proof -- but Wittgenstein was always worked up by what he was reading; and that included Goethe's book on colour!]

           For instance, consider Ms 172, which is described as "... dealing with the topics of colour and certainty" (PO p. 498) -- but are these the real topics? See Nedo, 1993 p. 56:

... 1. The question of discernibility (Unterscheidbarkeit), investigated by means of the phenomena of colour discrimination, and... 2. the determination of the point of observation, which Wittgenstein called the Ort der Weltanschauung, that is, of the collection of all those propositions which the observer does not doubt. Both of these questions are entirely in keeping with the tradition of philosophy in the German-speaking countries.

           Moreover, for Nedo, 1993 [p. 56], this investigation is into "the foundations of mathematic proof".

           Certainly, first, Wittgenstein's discussion in these manuscripts is couched in terms of texts Wittgenstein was reading (Moore on certainty; Goethe on colour) -- but Wittgenstein often did this; for example, re. Kohler's work;  Thus, for me at least, his uses of Kohler's ideas and images[20] or his discussion of Goethe's views of colour both surface at times when he was actively reading these works -- he had not, say, read them previously and flicked through for a relevant reference. We have biographical information for the (brute) fact that he was doing this reading at these times, but that leaves unclear whether he was doing so because he was (already) thinking along certain lines or whether his thinking about them resulted from the reading. And, of course, this is probably a distinction without difference -- there was doubtless an interesting interplay between what philosophical thoughts Wittgenstein was having and the way he read texts such as these (what he took from them), as well as more traditional relations between ideas and readings.

           Second, as we now have it, Rush Rhees (2003 p. 13[21]) shows the connection of On Certainty to Wittgenstein's other projects, and (especially) to his concern with the logic of concepts, by showing that he was struck by the peculiar character of certain 'propositions': that, appearances to the contrary not withstanding, what can appear to be empirical propositions are (often? sometimes?) not. Interestingly, one worked example Rhees deploys is "Blue is darker than yellow". A summary here might be that the connections between such claims (in some uses?) and the claims Wittgenstein discusses in detail in On Certainty (where the discussion of often appropriated by epistemology) are that both reflect this sense of "'the peculiar role' played by certain empirical propositions" (Phillips, in Rhees, 2003 p. ix).

           What should be noted, of course, is that actually Wittgenstein's writing at the time (the notes from which for On Certainty was excerpted) include material now published as Remarks on Colour: and reflect Wittgenstein's reading of Goethe. So this is not a case of Rhees bringing together in an illuminating way two disparate ideas in Wittgenstein -- or, if it is, it is because Rhees and his cronies in the editing of Wittgenstein had first put these two ideas asunder. And, mirabile dictu, Rhees (2003 p. 14) goes on to say that "... the clearest examples of concept-formation are in mathematics" (which, recall, Nedo took to be he topic in these passages). But as the editing of "Cause and Effect -- Intuitive Awareness" [mentioned above] highlight most clearly, a net effect -- almost a guiding principle -- of Rhees's editing of Wittgenstein has been the separation of the discussions of mathematics from other considerations of concepts.

           Given earlier comments on the idea of Wittgenstein as being (or, better, not being) a 'linguistic philosopher', it is worth spending a minute contrasting Rhees's own formulations here with the one I quoted earlier (from D. Z. Phillips' introduction to Rhees's work): for Rhees, these were propositions which -- despite appearances -- were not empirical, while for Phillips they remain "certain empirical propositions" (Rhees, 2003 p. ix). But what seems held firm by both formulations is that this proposition will (or will not) be empirical. Yet this claim too is rendered complex by unclarity as to what exactly is (or is not) a proposition. Suppose we think that a particular form of words, on some occasion, states a particular proposition -- and part of the test here might be that it will be truth-bearing: that is, it can be true or false. Well, if that is what we mean by the term "proposition", the fluctuation between criteria and symptoms (PI § 354) will not affect propositions. Rather, (roughly) the sentence used to state a symptom will be one proposition, that used to state a criterion will be another (and different) one. And then we will not be able to identify propositions by quoting sentences, since the same sentence might pick out criterion and symptom. Rather, propositions will be a bit more like utterances than they are like sentences: roughly, sentences in certain uses on certain (kinds of) occasions.

           But Rhees and Phillips seem to endorse the opposed view: they write as though the proposition were "Blue is darker than yellow": that is, as though the proposition were a kind of sentence. And as we know (PI §50), the very same sentence (say, "This bar is one metre long") sometimes functions as an empirical claim or description (telling us the length of the bar) and sometimes as an exemplar, telling us (roughly) what "metre long" amounts to. And this second use is not bi-polar. So, if this view of propositions were adopted, the very same proposition would sometimes state the symptoms, or the empirical description, and sometimes state the criterion, or function to pick out an exemplar. Then, roughly, it would be empirical on some occasions and logical on others.

           There may be no profound contrast here, viewed one way, but just a terminological one. Still, it has a direct influence on how the theses Wittgenstein deploys (if "thesis" is the right term -- as opposed to, say, "slogan" [REF?]) are regarded: in particular, the degree to which it makes any sense to regard them as theses about language.

[Interestingly, Rhees's summary has it that:

Wittgenstein's earliest and last concern was: what does it mean to say something? (Rhees, 2003 p. 6).

I cannot find Wittgenstein's philosophy in this, nor this in Wittgenstein's philosophy.]


D.       Importance of sometimes focusing on the unpublished (for example, in respect of art): it might be interesting to get together a version (paralleling CV) of those of Wittgenstein's remarks in the Nachlass which might plausibly be thought a contribution to aesthetics -- this was a topic on which he did lecture, but where we lack his own notes. Careful attention to such a document -- and especially to its chronology -- might allow us to see what a contribution ot aesthetics by Wittgenstein might have looked like. But this would need great care -- on most occasions when Wittgenstein used examples from, say, music or painting, he took them as the clear cases, to shed light on the mind or action.


{TS 235 as "table of contents": but I've made so many of these ...}


§6       Summarising the catologue

The thought, then, is (first) that one should always recognise -- for any of Wittgenstein's published texts -- its place in a catalogue of this sort; (second) that treating these texts in this way will involve granting a differential status (and also an historically-inflected status) to some as against others -- and hence to any views ascribed to Wittgenstein on the basis of them; (third) a sense of which of the published texts should be thought of as more purely or directly representing Wittgenstein's (pre-edited) words or arguments: and hence how we might look for what Wittgenstein meant. There is no single axis for any taxonomy (and hence any catalogue) here. But this one is most likely to weigh differently texts with a different status.

           In essence, then, I suggest that we see a taxonomy which has ten or eleven different categories: the seven outlined in section 4 and the four from section 5 (or perhaps exclude D above as unjustified from Wittgenstein's texts) -- but we might decide to treat PI Part Two as actually part of the 'late writings': that is, in the problem case of §5 C. In this way, we deal with all the published later writings, except CV; and we do so by valuing differentially different texts (where only for these 'late writings' does that amount to treating together texts published apart). For CV, the solution is simply to combine its English and German titles, calling it Miscellaneous Remarks on Culture and Value -- that makes its arbitrary character quite clear, as well as identifying its principle of selection!

           If the argument for a taxonomy of texts is at all plausible -- and if the taxonomy suggested here is then adopted -- one has a rationale for focusing one's consideration of Wittgenstein's work primarily on Philosophical Investigations (and especially PI Part One) as the nearest thing to not-Nachlass in the later work. Moreover, one can begin to prioritise among the other published texts. For one would then recognise that they were not all equal! [[NB the categorisation of works in Hacker (2005) is misleading -- although less than some -- precisely in ways our taxonomy can identify.]]


§7       Conclusion

It is fairly clear, I take it, why -- if what I have said is true -- one should prioritise PI (and especially Part One): this is the nearest Wittgenstein left us to a book on his later philosophy -- and there is the maximum of Wittgenstein's meticulous writing, with little editorial intervention. By contrast, it is fairly clear why both the students' notes and (for a different reason) RPP 2 should be given little credibility.

           Giving PI (and especially Part One) this importance will invite us to take it seriously as a structure, rather than regarding it as a collection of aperçus. Further, it will encourage us to take seriously what Wittgenstein says he is/was doing; and then -- faced with (say) an apparent tension between what Wittgenstein says his method is and what he seems to be doing in his discussion of "private language" (the so-called PLA: see Baker, 2004 pp. 119-140) -- to give serious thought to hunting for a reading of one or other set of passages (or both) that makes them consistent. For this would employ a 'principle of charity' commonly employed for writers we take seriously. Thus, for instance, suppose we thought (a) that there was a prima facie tension if the PLA were read as a reductio of 'Cartesianism', but (b) that Gordon Baker's account of Wittgenstein's project (from his 2004 book Wittgenstein's Method: Neglected Aspects [Oxford: Blackwell, 2004]) removed this tension -- well, that would speak in favour of a Baker-style reading. And all the more so if this 'reading' respected other of our considerations: for instance, if it aimed at likely targets of Wittgenstein (Russell, Frege) rather than likely targets of academic philosophy (Descartes, Locke, Hume). (Of course, this is just to apply here ideas we familiarly apply to, say, readings of Aristotle or Plato.) And we would not expect any such 'resolution' here to be absolutely conclusive -- rather, we would look for 'best fit' -- for "... interpretations, like ways of seeing line drawings [such as the duck-rabbit], can ... be more or less natural (as opposed to 'strained') ... To show (to the extent that  such a thing can be shown) that ... [so-and-so] is a very unnatural reading is not to disprove it, but it ought to make it less attractive, especially if there is an available alternative" (DD pp. 5-6).

           Again, the late work is an interesting example here: LW 2 has some passages which simply reprint Wittgenstein's texts (from the Nachlass) -- but it is also involved in the construction of the two (apparently) independent late texts, OC and ROC, since it comprises what is left when they are removed. And the credentials to independence of those two texts is at least open to doubt. Deciding to doubt them would lead to a pretty different reading of Wittgenstein's corpus -- and, of course, a revision of how one took all those secondary texts which differed from one on this.

           Our basis for choosing here included a judgement as to how Wittgenstein's working procedures impinged on the texts: or, if this is different, whether editorial decisions in respect of them were well-supported.

           Suppose that, following the lines of argument I've offered, we come to some decisions about Wittgenstein's works: what have we learned? In particular, have we learned anything which might be relevant to our 'reading' -- our treatment -- of philosophers of the past? [To get us back to my central question]

           Two or three areas seem to me at least strongly suggested (and perhaps more): first, we cannot be learning directly from the study of Wittgenstein in respect of problems Wittgenstein did not address. So that even if there were features of what Wittgenstein called our method (VW p. 69) -- not philosophy's method, but just the one we use -- that can be applied elsewhere than where he applied them, that is work that still remains to be done. And there may be more constraints on that method than we know/realise/acknowledge: after all, it too is found in part of one of these texts. But we have no chance of finding an insight in Wittgenstein if we do not pay attention to his project -- to proceed otherwise is to treat the text simply as a kind of Rorschach test.

           The second area concerns the worth of Wittgenstein's ideas (as examples of the ideas of a past philosopher), and operates somehow dialectically. If Wittgenstein's ideas, understood in such-and-such a way, seem coherent and plausible -- given his assumptions -- that may be a prima facie good reason to read his work that way: and hence a prima facie reason to read his work! Yet, if we deny its coherence and plausibility, that gives us a reason to read it differently. (But how do we decide who to consider, given this policy of charitable reading? That too will have to be argued, case-by-case and time-by-time. For instance I would insist on John Wisdom's importance: if other disagree -- and he seems to have disappeared from reading lists in universities I know -- we begin to look at what granting his importance would offer.)

           This brings we to the third area: that what a philosopher of the past can offer us is just that -- a consistent picture, given its assumptions, of a set of conceptual relations. In seeing another mind-style, we learn (in part) that there are other ways (than ours) of organising conceptual relations. For example, in their writing on Descartes, Baker and Morris highlight (a) how knowledge of past philosophy "... it might lead us to a see an aspect of concepts to which we have been blind" (DD p. 218) and (b) how "... making a sustained effort to grasp Descartes' vision might lead us to a greater self-awareness ..." (DD p. 219): it might lead to our coming to "... read his texts with more humility" (DD p. 219) -- surely appropriate for a major thinker -- where the outcome might be "... the possibility of recognising conflicts of frameworks and responding to them intelligently" (DD p. 219). In general, doing so might amount to changes in "... nothing less than ways of seeing things or norms for describing them" (DD p. 219). Recognising that a plausible way to conceptualise a certain issue[22] is not the only way, and that gifted thinkers have conceptualised it differently, is not only a sobering thought -- faced with the pretensions of contemporary views -- but a basis for a re-thinking of our own position(s).

 If this is right, then, we have been exploring at least some framework conditions, and possibly some details, of Wittgenstein's position -- what, earlier, I called his mind-style. And doing that is doing philosophy. Or so it seems to me.



[1]              Gordon Baker & Katherine Morris Descartes' Dualism London: Routledge, 1996 - cited as "DD".

[2]              Since I began writing (and discussing) the first versions of this paper,. I have come across David Stern's "The Availability of Wittgenstein's Philosophy" in Hans Sluga & David Stern (eds) The Cambridge Companion to Wittgenstein (Cambridge: Cambridge University Press, 1996 pp. 442-476), which overlaps with some things I urge here. Like me, Stern laments some of the difficulties that follow from the fact that the later philosophy of Wittgenstein is composed entirely of Nachlass; that is, of material unpublished at Wittgenstein's death. And, of course, one part of the complexity is that we now have a number of texts published from this material, with Wittgenstein's name (as author) on the spines.

                              Stern claims to be offering a much-needed " outline of the relationship between Wittgenstein's published and unpublished writings" for those " for whom the published Wittgenstein is Wittgenstein enough" (p. 469).

                              As I hope this text makes plain, I accept that the confusing character of the present situation is exacerbated by that fact that a (roughly) uniform edition makes all the published material appear to have the same weight. So one needs a clarification here. Unlike Stern, I am not optimistic about the prospects of such clarity being provided by the availability - now achieved - of the whole Nachlass on CD-ROM [REF?]. Indeed, that might cloud the picture, by contrasting those 'published works of Wittgenstein' available in one format (books), with those now available in another (through the Nachlass).

                              What is needed (and what I begin to provide below) is a catalogue which emphasises the differences within the published (in book-form) works - so that we more readily understand the status of each: hence distinguish (a) those most nearly 'books/texts' by Wittgenstein (the masterpieces); (b) those whose primary interest lies in reflecting a moment in Wittgenstein's thought; (c) those that - despite their value - are flawed (in many different ways, but most relating to the editing process); (d) those that really ought not to be published in this form (if I had my way - but 'always a god-mother, never a God').

[3]              Of course, sometimes these thoughts about translation can obscure the issue - for instance, Descartes's use of corpore: translating it "body" gives it the same confusing character in English, where the term "body" can sometimes mean human body (or my body) and sometimes thing (as in "heavenly bodies").

[4]              "Philosophia ": See Baker, 2004 p. 12

[5]              Ludwig Wittgenstein The Big Typescript: TS 213, (ed. and trans. C. Grant Luckhardt & Maximilian Aue), Oxford: Blackwell, 2005.

[6]              Med VI [CSM II p. 53]:

 Sometimes towers which had looked round from a distance appeared square from close up In these and countless other cases, I found that the judgements of the external senses were mistaken. And this applied not just to the external senses but to the internal senses as well. For what can be more internal than pain? And yet I had heard that those who had had a lefg or an arm amputated still seemed to feel pain intermittantly in the missing part of the body.

               Principles IV [CSM I p. 280] gives "the full-dress account of sense-perception" (DD p. 125).

[7]              Michael Nedo Ludwig Wittgenstein: Wiener Ausgabe/Vienna Edition - Introduction. Vienna: Springer-Verlag, 1993.

[8]              Graham McFee "Wittgenstein, Performing Art and Action" in R. Allen & M. Turvey (eds) Wittgenstein, Theory and the Arts. London: Routledge, 2001 pp. 92-116; Graham McFee "Wittgenstein and the Arts: Understanding and Performing" in P. Lewis (ed.) Wittgenstein, Aesthetics and Philosophy. Aldershot: Ashgate, 2004 pp. 109-136. (These paper are based on the same research project, reflecting two aspects of it.)

[9]              For help with this, and with the translation generally, my thanks to my (native German-speaking) colleague Udo Merkel.

[10]            Ossie Hanfling Philosophy and Ordinary Language. London: Routledge, 2000 p. 38, p. 40.

[11]            [[·          Combined case: translation inflected through reading. Here, we might follow Wittgenstein (PI§445):

It is in what we say that expectation and fulfilment make contact. (my emphasis: translating due to Baker, 2004 p. 64)

               This makes clear that it is our activities, or practices, of language-using that are at issue, rather than some mysterious properties of 'language' (or, worse of the German/English language) [[Baker, 2004 p. 72 note is also clear that one does not add the practice to give life to as sign: compare SRV pp. 77-83.]]. As a simple example to explain, consider predicting that it will rain and taking an umbrella: these two go together, as part of what is involved in meaning that (say) it will rain tomorrow. So if I did not take the umbrella, you would with justice doubt that I was really expecting it to rain, given my well-known antipathy to getting wet; if I put on the wellingtons and pick up the umbrella now, you will with justice doubt that I expected it to rain tomorrow; and so on. To repeat, the point is just that what we say, in characterising the event, and what counts as an occurrence of that event (and not some other), come together.]]

[12]            Wittgenstein (notoriously?) connected the substance of philosophy to its style - as though the manner of delivery had a direct bearing on what was said.[[see Nedo, 1993 p. 53, quoting from Alice Ambrose Wittgenstein's Lectures: Cambridge 1932-1935, Oxford: Blackwell, 1979 p. 43]]:

One difficulty with philosophy is that we lack a synoptic view. we encounter the kind of difficulty we should have with the geography of a country for which we had no map, or else a map of isolated bits. We can walk about the country quite well, but when we are forced to make a map we go wrong. A map will show different roads through the same country, any one of which we can take, though not two, just as in philosophy we must take up problems one by one though in fact each problem leads to a multitude of others. We must wait until; we come round to the starting point before we can proceed to another section, that is, before we can either treat of the problem we first attacked or proceed to another. In philosophy matters are not simple enough for us to say 'Let's get a rough idea', for we do not know the country except by knowing the connections between the roads. So I suggest repetition as a means of surveying the connections.

               If this point were granted, it would be another way to raise again a range of ideas about how XXXX

[13]            "The focus is on individuals' states of mind: unrest (PLP p. 7; F 93), torment, disquiet (PI §111; LSM 33), discomfort (BB p. 26; PLP p. 4), drives (PI §109), obsessions (HISP p. 18), craving (BB p. 17; LPM 58), revulsion (BB p. 15, p. 57), Angst (BB p. 27; F 94), irritation (PLP p. 7; F 62), profound uneasiness of mind (HISP p. 3), profound mental discomfort (HISP p. 6), obsessional doubt (HISP p. 8), shock (PLP p. 7), troubles (BB p. 46), compulsioons to say things (BB p. 47), irresistable temptations (BB p. 18), alarm (HISP p. 4), etc. That is to say, the focus is on a duistinct range of intellectual emotions or disturbances." (NB sustaining all of these comparisons - as opposed to some - will require adopting Baker's view of the relation of Wittgenstein to Waismann.)

[14]            Some of the detail here is recorded in the introduction to Collingwood's Principles of History (cited earlier).

[15]            Re Part Two of PI: Monk, 1990 p. 542 comments:

in Dublin it was presumably during this time [i.e. May 1949: beforing going to the USA] that he prepared the fair manuscript copy of what is now Philosophical Investigations, Part II.

               This fits with my view: he had it ready to show to Malcolm. But Monk (1990 p. 542) takes it that:

before he left for America he wanted to spend a few weeks in Cambridge preparing a final, polished typescript of the work with which he had been occupied since 1946.

               Again for a time in Cambridge prior to going to the USA:

His chief concern during these few weeks in Cambridge was to dictate to a typist the manuscript containing his final selection of the remarks written over the last three years, which now forms Part II of Philosophical Investigations. (Monk, 1990 p. 544)

               [[NB Von Wright letters??]]

               {{This means that that work as a whole was more-or-less finished by 12 July 1949: Monk, 1990 p. 550.}}

               By contrast, see Von Wright Wittgenstein (Blackwell, 1982) p. 135:

It is not known when the typescript [that became Part Two] was made nor even whether it was dictatated by Wittgenstein himself.

               [And lots more.]

               However, in respect of a letter from Wittgenstein  of 3rd July 1949, Von Wright comments:

It was presumably then [during his stay in Cambridge] that he dictated the typescript for the second part of the Investigations. (PO p. 473)

               NB Nedo, 1993 p. 46: who implies he worked on the Ts after his return from the USA. QUOTE?

[16]            Stern's solution ("Availability" pp. 468-469) is the regard RFM Part One as one of the 'finished works of Wittgenstein', drawing on criteria proposed by Schulte: but to continue to regard the situation as unsatisfactory.

[17]            Indeed Von Wright (1982, p.136) conjectures that this material might 'bridge the gap' between the present Parts I and II of the Investigations. Certainly (and at least) these are an important set of remarks on the philosophy of psychology.

[18]            Here I agree exactly with Stern ("Availability" p. 447) that " there is no indication that Wittgenstein conceived of [this body of work, including On Certainty] as separate pieces of work": hence On Certainty - widely regarded as one of Wittgenstein's triumphs - does not fare well under detailed scrutiny.

[19]            And even more once we correrct errors in that presentation: thus, to go by the work's pagination, the correct account of of Ms 176 should read that part (pp. 1-22) was published as Part I of ROC [this is presently correct]; that part (pp. 22-47, and pp. 51-81) were published as §§ 426-523, 524-637 respectively of OC; and that the remainder (pp. 47-51) was published as the final part of LW 2. [Notice that my version make the continuity, especially of OC, less strong than the original version.]

[20]            See John Cook Wittgenstein's Metaphysics (Cambridge: Cambridge University Press, 1994 Ch 10; Graham McFee "Wittgenstein on Art and Aspects", Philosophical Investigations, Vol. 22 No 3, 1999 pp. 262-284, esp. pp. 266-268.

[21]            Rush Rhees Wittgenstein's On Certainty, Oxford: Blackwell, 2003

[22]            Note here the dubious assumption that there is one issue, once we agree that there are two accounts.


April 2006

Return to top of page