The Evolution of Modern Metaphysics: Making Sense of Things (57 page)

In order to be able to draw a limit to thought [understood as a
limitation
of thought], we should have to find both sides of the limit thinkable (i.e. we should have to be able to think what cannot be thought). (p. 3)

Provided that thought is understood in a suitably thin way (cf.
Ch 5
, §8), this argument is surely unassailable.

But does Wittgenstein understand thought in a suitably thin way? After all, the very starting point of this discussion was his insistence that logical propositions do not express thoughts. Is there perhaps scope, in Wittgenstein’s own terms, for resisting his version of the Limit Argument? Consider the expanded form of that argument, as set out in
Chapter 5
, §8. In producing propositions about how things must be, do we not falsify the first premise, the Limit-Drawing Principle? For do we not draw a limit to what we can make sense of without expressing any thoughts about that limit, and hence without making sense of it?

We do not do this in any sense that disrupts the Limit Argument. The point is exactly the same as it was before. We draw a limit to what we can make sense of
only
in the sense that we indicate essential features of what we can make sense of.
³⁵
We do not indicate any limitations of it. There is no inside and outside here, no immanent and transcendent. Logical truth does not
transcend
non-logical truth; it pervades it.
³⁶
It is revealing that the adjective that Wittgenstein uses to describe logic, in what is a clear echo of Kant – almost as if to emphasize that he does not take it to be transcendent – is ‘transcendental’ (6.13).

In other respects, however, this conception is radically un-Kantian. Wittgenstein is utterly hostile to the idea of the synthetic
a priori
. The synthetic admits of alternatives (cf. 6.11). But the
a priori
does not. It attaches only to absolute, logical necessity; only to the truth of that whose truth can be ascertained without reference to reality (see e.g. 2.225, 3.05, 5.4731, 5.634, 6.31, and 6.3211; and cf. 4.0412). To acknowledge the synthetic
a priori
is, in effect, to acknowledge limits that are limits in both senses of the term – essential features
and
limitations – precisely what Wittgenstein abjures. And, as Peter Sullivan well notes, the 6s in the
Tractatus
consist largely of a case-by-case rejection of Kant’s examples of synthetic
a priori
truth.
³⁷
It is thus that Wittgenstein answers what he calls at one point in his
Notebooks
‘the great problem round which everything that I write turns’ (p. 53). The question is: ‘Is there an order of the world
a priori
, and if so what does it consist in?’ (ibid.). The answer, which he gives at 5.634, is: ‘There is no
a priori
order of things.’
³⁸

5. ‘Anyone who understands me eventually recognizes my propositions as nonsensical’

Wittgenstein recoils from both a Fregean conception of the
a priori
and a Kantian conception of the
a priori
then. But this does not prevent him from inheriting some of the problems that afflict each of these. In this section we shall consider problems that he inherits from Frege. In §7 we shall consider problems that he inherits from Kant.

In §7 of the previous
chapter I
mentioned two problems that befell Frege. For reasons that need not detain us Wittgenstein takes himself to be immune to the first of these, the problem about the set of sets that do not belong to themselves (3.333; see also 6.031). But he certainly confronts an analogue of the second problem, the problem about the
Bedeutungen
of predicates.
³⁹
The relation between his semantics and Frege’s is too oblique for us to be able to say that he confronts exactly the same problem. But he does hold:

• that there is something in an elementary proposition that enables it to be more than a mere list of names

• that this corresponds to something in reality that enables objects to be combined together in states
of affairs

•
that the something in each case is the same, an abstract unity that he calls ‘logical form’ (this is the rationale for his holding that elementary propositions are facts)

and

• that logical form can never itself be the subject matter of any proposition, since propositions only ever say how
objects
are

or again,

• that logical form can never be the subject matter of any proposition since a proposition would need to transcend logical form, which is to say it would need to transcend reality, in order to say how, among the ways logical form might be, it is (see the 2.1s, 3.14–3.221, and the 4.12s).

Wittgenstein further holds that propositions ‘show’ logical form, where what can be shown cannot be said (4.121 and 4.1212).
⁴⁰
But this appeal to what propositions show cannot hide what in fact it serves only to emphasize, namely that Wittgenstein’s own numerous remarks about logical form, like Frege’s frequent references to properties, cannot be interpreted in a way that is consonant with his own views and must therefore, by his own lights, be regarded as nonsense.

Nor is this aporia the incidental aporia for Wittgenstein that it was for Frege. There is scarcely a sentence in the
Tractatus
that
can
be interpreted in a way that is consonant with his own views, or at least there is scarcely a sentence in the
Tractatus
that can clearly be so interpreted. Thus consider all the sortal noun phrases in the work that appear to have as one of their uses to indicate ‘internal properties’, where an internal property is a property such that whatever has it has it of necessity (see 4.122ff.). Examples are ‘object’ – nothing just
happens
to have the property of being an object – ‘state of affairs’, ‘sense’, ‘logical form’, and indeed ‘internal property’. The 4.12s foreclose any such use. (This is related to the principle that the only necessity is logical necessity.) There are related problems for any sortal noun phrase that is defined in terms of any of these, such as ‘fact’, ‘world’, ‘proposition’, or ‘thought’. Each occurrence of each of these noun phrases must therefore give pause. I put it no more strongly than that because of how much, in the interpretation of a term, depends on context.
⁴¹
Thus when Wittgenstein writes, ‘The whole modern conception of the world is founded on the illusion that the so-called laws of nature are the explanations of natural phenomena’ (6.371), it is not at all clear that what he writes is vitiated by its mere inclusion of the sortal noun ‘world’. (Here it is
worth remembering that, insofar as Wittgenstein specifies conditions that a combination of signs must meet in order to be a meaningful proposition, and insofar as we accept what he says, we are at just as much liberty to take what we believe to be a meaningful proposition and conclude by
modus ponens
that it must be a combination of signs meeting those conditions as we are to take what we believe to be a combination of signs failing to meet those conditions and conclude by
modus tollens
that it cannot be a meaningful proposition.
⁴²
) The fact remains that the 4.12s must make us wary of all of these noun phrases. And there is the further complication that each of them, with the exception of ‘object’, may suffer from a direct analogue of the problem afflicting the Fregean term ‘property’, namely that when it
is
being used to indicate an internal property, or is supposedly being used to do that, syntax demands that it is a property of objects and semantics demands that it is not.
⁴³
,
⁴⁴
Wittgenstein, like Hume and Kant before him, appears to be under serious threat from the challenge of applying his own principles to his own work.

However, unlike either of them, but like Frege, he is pre-emptive. He acknowledges that his book consists mainly of nonsense. In the famous penultimate remark of the book he writes:

My propositions serve as elucidations in the following way: anyone who understands me eventually recognizes them as nonsensical, when he has used them – as steps – to climb up beyond them. (He must, so to speak, throw away the ladder after he has climbed it.) …

He must transcend these propositions, and then he will see the world aright. (6.54)

Frege asked his reader not to begrudge him a pinch of salt (Frege (
1997e
), p. 192/p. 204 in the original German). Even a reader ready to grant Frege his wish might think twice about showing Wittgenstein the same indulgence. It is one thing to stumble into nonsense in trying to articulate one component
of a larger, otherwise coherent theory. It is another thing to produce nonsense at almost every turn throughout an entire book.
⁴⁵

To be sure, the fact that a book consists mostly of nonsense is not itself a ground for indictment. Books are written for all sorts of purposes. Some of these might well be served by nonsense. An obvious example is entertainment. Another is parody. Another is to ward the reader off engaging in a certain intellectual activity, a purpose that could be served by exhibiting the deleterious effects of doing so. But none of these seems to hold out much hope as far as exonerating the
Tractatus
is concerned.
Its
propositions appear to be attempts to convey genuine insights, insights into what it is to make propositional sense of things and into the ways in which bad philosophy hinders us in our efforts to do so. It looks as though we are supposed to share these insights by, as one would like to say, ‘understanding’ the nonsense. What is going on here?

6. Two Approaches to the
Tractatus
. A Rapprochement?

Recent exegesis of the
Tractatus
has involved two broad approaches. I shall adopt the simplifying assumption that there is enough unity in these approaches to warrant talk of two readings: what I shall call ‘the traditional reading’ and ‘the new reading’.

Advocates of the traditional reading take the
Tractatus
at face value, as an attempt to convey something. But they also take seriously the claim at 6.522 that ‘there are … things that cannot be put into words’, things that, in the terminology introduced in the previous section, can be shown though they cannot be said. They hold that what the
Tractatus
is an attempt to convey are things of just this sort.

The Traditional Reading
: The
Tractatus
is an attempt to convey things that are ineffable.
⁴⁶
The means that Wittgenstein uses to convey these things is language that he himself recognizes as nonsense. But it is nonsense of a special kind, what might be called ‘illuminating’ nonsense.
⁴⁷
It
enables us to grasp what he is trying to convey. And this in turn enables us, like him, to recognize it (the nonsense) as nonsense.
⁴⁸

Advocates of the new reading are prepared
not
to take the
Tractatus
at face value. They adopt a version of the suggestion that Wittgenstein is exhibiting the deleterious effects of engaging in a certain malpractice, namely bad philosophy. This brings the
Tractatus
more into line with what I said earlier we might have expected of it: a catalogue of examples of what not to do, presented in such a way as to enable the reader to see how and why not to do it.
⁴⁹

The New Reading
: There is nothing that cannot be put into words. There is only the temptation to see sense where it is lacking. Wittgenstein’s aim in the
Tractatus
is therapeutic. The reason why the book consists mainly of nonsense is that he is trying, by indulging the temptation, to eliminate it; by producing nonsense that appears to make sense, and then testing the appearance, to get the reader to acknowledge the illusion, so that the temptation disappears, and the reader is left realizing that the nonsense is precisely that: sheer lack of sense. It is an assemblage of signs to which, as used here, no meanings have been assigned. It conveys nothing whatsoever.
⁵⁰

These readings look as if they could scarcely be further apart. But two points should be made straight away. First, it is quite compatible with the traditional reading that the nonsense is ‘an assemblage of signs to which, as used here, no meanings have been assigned’. And second, conversely, it is quite compatible with the new reading that the nonsense is edifying nonsense. Indeed, advocates of the new reading would be among the first to insist that the nonsense is indeed edifying nonsense, in its own unusual and indirect way. But even apart from these two points there is a crucial and much deeper reason for thinking that the two readings may not be all that far apart.
⁵¹

Consider: what does someone mean who claims that there are things that cannot be put into words, or who claims that, on the contrary, there are no such things? What does the term ‘things’ range over here? If it ranges too widely, there is no interest in either claim. For it is plain that most things, on even a moderately broad construal, cannot be put into words. A brick, for example, cannot. It can be described, certainly. But it cannot be put into words. Nor, come to that, can the opening bar of Beethoven’s Fifth Symphony, which gives the (literalistic) lie to the famous quip that F.P. Ramsey directed at the
Tractatus
, that ‘what we can’t say we can’t say, and we can’t whistle it either’ (Ramsey (1931), p. 238).
⁵²