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

I

The Hacker-Diamond controversy over how to read the Tractatus is likely to stir much debate. The issues involved call for rigorous scholarship (and therefore also a length of exposition not allowed here). But why bother? - Stanly Cavell has recently claimed that there is still work to do in receiving "the event of the Investigations in Western philosophical culture" (2001, 92). The import of the Hacker-Diamond controversy seems to me lie in how it contributes to the task Cavell defines: It gives new life to the old idea that in order to see what is new and important in the Investigations we must see to what extent conventional conceptions of philosophy are challenged already in the Tractatus.¹

A complication here is the hot air surrounding the controversy. P.M.S. Hacker and Cora Diamond present themselves as authors having a case. So, it might seem that their task, and ours, is to find out who is right? But that might lead into the dark. My suggestion will be that on crucial points Hacker and Diamond are both right in their critique of each other and, hence, that neither has achieved a conclusive position. I also wish to indicate that both efforts to get the Tractatus right create a risk of inhibiting our communication with Wittgenstein's later thinking.

II

Hacker reads the Tractatus as advancing the idea that there are truths which cannot be stated, but which can be shown. (2000, 353, 381.) Diamond disagrees. Put technically, the gist of Diamond's critique of Hacker is that he is mistaken in maintaining the following: The Tractatus promotes "the idea of there being such a thing as violating the principles of logical syntax by using a term in what, given its syntax, goes against what can be said with it" (Diamond 1991a, 195). The critique has been outlined by Diamond and some of it has been worked out in greater detail by James Conant.² Responding to the criticism Hacker suggests that we compare the rules of logical syntax with contract law. Failure to follow the rules for making contracts does not result in illegal contracts but, rather, in no contracts at all. Likewise, failure to comply with the rules of logical syntax does not result in "forms of words that describe a logical impossibility" (as Diamond et.al. think that Hacker thinks) but "in nonsense". (Hacker 2000, 365-367.) Hacker adds: "If one fails to comply with ... the rules of logical syntax one transgresses the bounds of sense, which are given by logical syntax, and to transgress the bounds of sense is to talk nonsense." (ibid., 367.)

There is a question of where the mistake lies and I do not think Diamond and Conant are always helpful. We might agree with Hacker that "to transgress the bounds of sense /or, the rules of logical syntax/ is to talk nonsense". That might be a stipulation about the term "the bounds of sense" or an announcement that from now on we will use "transgression of the bounds of sense" as a synonym for "nonsense". This need not lead us wrong. (Compare: there might be no harm in speaking about "the activity of butter when it rises in prise". Wittgenstein 1953, part 1, no. 693.) The entire idea that it clearly is or is not helpful to speak about the limits of knowledge or of sense in philosophy and of their transgression is problematic.³ Therefore, Diamond's notion that there is an outright mistake on Hacker's part appears suspect. Nevertheless, there may be problems in what he says.

Here is one diagnosis of the problem. As noted, Hacker speaks about something which is "given by logical syntax". And he attributes to the author of the Tractatus the notion that we can infer from our understanding of logical syntax whether something which appears under that pretension is or is not a proposition. He claims that "logical syntax allows us, for example, to substitute certain symbols for certain other symbols" and that "to disobey the rules of logical syntax is to string together words in manner that is excluded ... by logical syntax". Apparently, if we see a case of disobedience, a substitution which is not allowed by logical syntax, we are entitled to the claim that what is proposed is not a proposition. Hacker's idea, then, is that on the Tractarian view knowledge of certain rules will entitle us to claims about whether something is or is not a proposition. As if there was a something, a set of rules, logic, the logical syntax of our language, or even, of any possible language, that decides about what belongs to language and what does not. (This reading opens the doors to a curiously flat reading of Wittgenstein as one who works out theories: The early Wittgenstein can then be classified as a conventional transcendentalist. Furthermore, the later Wittgenstein will easily be taken as producing an interesting new variety of transcendental argument. As if the point of the so called private language argument were to prove something. Cf. Diamond 1991b.) I would agree with Diamond and Conant that Hacker's position is difficult to reconcile with the Tractatus 5.473 - 5.4733. Hacker will have some difficulty with the passage in 5.473 saying that "the proposition is nonsensical because we have failed to make an arbitrary determination, and not because the symbol, in itself, would be illegitimate." His difficulties culminate when we turn to the occurrence of the word nur (only) in 5.4733. It is difficult to understand why Wittgenstein would have used that word, unless to mark his disagreement with the Fregean line Hacker is taking.⁴ An interpretation improving on Hacker's and in harmony with some of what he says and with all of the 5.473's would be that our determination of what is and what is not a proposition is decisive for what we are prepared to understand as the rules of logical syntax. But then, contrary to what Hacker (and his comparison with contract law) suggests, "logical syntax" (or "logical form") could not be referred to and relied on when (if) we wish to legislate about sense and nonsense.

III

At the core of Hacker's criticism of Diamond is the thesis that her attempt to throw way the ladder resolutely is a "case of trying to have one's cake and eat it". (Hacker 2000, p. 361.) Hacker's decisive argument comes early in his long discussion. He notes that Diamond et. al. are "methodologically inconsistent" (p. 360). While they claim to follow the instruction of Tractatus 6.54 and to throw away the work as nonsense they make exceptions. They read some remarks in the book, in particular parts of the Preface, the 5.473's and 6.54 itself, as meaningful. These meaningful remarks they call the "frame" of the Tractatus. But on their terms, how can they account for this conception of a meaningful frame or for the frame/body -distinction?

Hacker does not develop the point much and he fails to address in detail how Diamond and her allies could respond and have, partly, responded to it. These responses include three main strategies. One is to say that whenever Diamond et. al. create the impression that they read sense into a passage in the Tractatus they are engaged in transitional discourse aiming at freeing us from the illusion that the passage invoked makes sense. The second is to say that there is no objective way of distinguishing between sense and nonsense in the Tractatus; that the distribution of sense and nonsense, the division between frame and body of the work, is a question to be left to the judgment of each individual reader. The third is to say that the task of separating sense (frame) from nonsense (body) in the Tractatus is trivial and involves no philosophical effort and, therefore, no illegitimate reliance on the nonsense in the body of the Tractatus.⁵

The crux of all these lines of defence against Hacker's paradox charge is this. They all (overtly or covertly) presuppose that the author of the Tractatus or its "resolute interpreters" have competence to distinguish frame from body, sense from nonsense, while they simultaneously give up all possible resources for sustaining this authority. The most complex case from the point of view of my criticism may be that of Conant. For him the paradox ultimately arises from the tension between claiming that Hacker gets the Tractatus wrong and claiming that how we understand Sinn and Unsinn to be distributed in the Tractatus "depends on us" (2000, 216). The question Conant is unable to answer on the terms he sets is: If it is up to me to perceive or not to perceive sense in the Tractatus with what right does Conant speak about Hacker's, or anyone's, getting the Tractatus wrong?

The upshot of Diamond's "methodological inconsistency" is a tendency to read the Tractatus not as a failed book but as a successful practice of philosophy as therapy and, hence, to read too much continuity into Wittgenstein's development. Hacker helps us see this.

IV

If Hacker and Diamond both get the Tractatus wrong, what can we do? Here is a sketch. The dispute between Hacker and Diamond is a dispute about the exact meaning of "nonsense" in Tractatus 6.54. But does it have an exact meaning? My suggestion: the term is ambiguous and the ambiguity derives from historically and motivationally different layers in Wittgenstein's work on the book. The 6.54 can be taken as saying that the sentences of the Tractatus, and perhaps all sentences presenting themselves as philosophy, are nonsensical on the terms of the theory of logic and meaning of the Tractatus. This is Hacker's "whistling Wittgenstein", energetically sawing off the branch on which he is sitting. But 6.54. can also be taken as saying that the Tractatus is nonsense in an ethical sense only: it solves (by way of argument or therapy) the problems of philosophy but leaves all problems of life unsolved.

Hacker writes that Diamond et.al. suggest that Wittgenstein practised Kierkegaardian irony designed to reveal that all substantial philosophical claims are plain nonsense. Hacker stresses that "there is not a single trace of any such strategy" in any of Wittgenstein's conversations with his pupils. (2000, 381.) Perhaps, as Hacker says, there was no trace. But now there is. At a seminar in Helsinki in February this year a paper on the Hacker-Diamond controversy was presented. In the discussion Georg Henrik von Wright remarked that Wittgenstein "wanted to finish philosophy, all of it". If this pupil's report is taken at face-value we have reason to think that "Kierkegaardian irony" might have suited Wittgenstein's purposes. Likewise, the ambiguity of 6.54 explained above would have suited him because the logical and the ethical reading of it both spell the end of philosophy in the double sense of solving all problems of philosophy and of showing how little is thereby achieved.

The decisive point of rupture between the early and the late Wittgenstein could now be outlined as follows. The early Wittgenstein thought that there is such a things as a grand distinction between sense and nonsense and that the task of philosophy is to make it clear where sense is to be found and where nonsense reigns. The later Wittgenstein gives up the idea of such a grand distinction and of any lofty ladder to take us to a position from which to confer ultimate judgment about sense. Instead he works out a (Socratic?) idea of philosophy as "work on oneself ... on how one sees things" (1998, 24); a work which requires that we keep to the rough ground of conversation between earthly creatures. That also involves admitting, as Wittgenstein later did, that perhaps we, or some of us, cannot stop doing philosophy. (Rhees 2001, 158.)