Thursday, 26 June 2008

Making vs. Having


One thing that I keep noticing and feeling as very significant is a certain vacillation in Wittgenstein's terminology.  Sometimes he talks about a proposition making sense  and other times about a proposition having a sense.  Sometimes he changes between the two in the course of a sentence as in the following quote from the Notebooks:

[I]f a proposition is to make sense then the syntactical employment of each of its parts must be settled in advance.. must be completely settled before that proposition can have a sense!

In his later work, he seems to overcome the view that a proposition has a definite sense that the proposition connects to.  Yet he never seems to diagnose or directly address this slip in terminology as a source of the problems in his earlier work.  In fact, the later Wittgenstein is also sometimes inconsistent with how he talks about 'sense'.  Yet when he is trying to highlight mistakes in this area, he is (and perhaps without realising) does stick to talking about making sense:

“You understand this expression, don’t you?  Well then- I am using it in the sense you are familiar with.”- As if the sense were an atmosphere accompanying the word, which it carried into every kind of application.// …he should ask himself in what special circumstances this sentence is actually used.  There it does make sense.

But why the temptation to slip from one to the other?  Why not only talk about a proposition making sense?  Even in the quote from PI above he seems to talk about 'the sense' of a sentence (or at least a sentence as used in that context).  And if we do need to talk about the sense a sentence makes,or the sense that the sentence 'has', wherein lies the false conception earlier?

In PG he writes:

On hearing the assertion “This sentence makes sense” you cannot really ask “what sense?”

Yet, of course this is a question he asks both early and late.  A word doesn't by right mean this, that or the other.  It is only because it used in this sense or that.  The precise philosophical task (which can be done in the right way or otherwise) depends on explicating in what sense the word is used, and that determines the meaning of the word.

I will write more on this and try to work out where the arguments meet.  I think it is exceptionally important as it gives a way into diagnosing where the earlier Wittgenstein went wrong, without simply assuming he was being dogmatic.  It is one that if he was wrong, it was at a specific point he did so and was a subtle and understandable mistake in the context of a shared goal.  Both were trying to elucidate the sense of a proposition, but something led the earlier Wittgenstein to look in the wrong place for such an elucidation.

Tuesday, 3 June 2008

General Form of the Proposition Part 1: Where the criticism is not


In the Philosophical Investigations, Wittgenstein argues against his earlier claim that there is a ‘general form of the proposition’. Instead he argues that language has a great variety of different uses and there is no one way that a proposition can be used to mean something, or one form of proposition (that of being a truth function of elementary propositions, which are truth functions of themselves) that can have sense. However, Roger White is of the impression that no solid case has been made against the general form of the proposition:

What he does in the Investigations is simply parade before us the extraordinary diversity of uses of language that there are, and invites us to ask whether it is credible that they should conform to a simple underlying pattern, such as he had envisaged in the Tractatus. Here it seems to me a strong case can be made out for saying that his first thought may be nearer the truth than his later one. This is not the place to argue this in full, but I will indicate two considerations to think about. Firstly, Wittgenstein makes no distinction between what a sentence means, and the use to which we put it, and much of the diversity he illustrates in §23 is diversity, in use not meaning. Secondly, it is more than arguable that unless there was a simple underlying system to the language, it would lack the flexibility necessary for it to be put to such diverse uses.
It might seem that the later Wittgenstein believed that in the Tractatus it was simply taken for granted that there was a general form of the proposition. This was assumed, so the thought goes, because there was one use of language and this could be dispelled by: “Don’t say ‘there must…’ but look and see whether there is anything common to all.” This clearly would not do. Firstly, the author of the Tractatus is most certainly aware that there are different uses of language. Secondly, as White argues, it ignores the argument (and not, assumption) of 4.5 that there couldn’t be a proposition whose form could not be foreseen. We can understand propositions that have never been heard before and produce entirely new ones. As such, the most general form of the proposition must tell us ‘This is how things stand’. White argues that translated in this way, it sounds rather banal. It would be better to say ‘This is how things are arranged’: this is what the world would be like if the proposition were true.

In 4.002 he says, “the outward form of the clothing is not designed to reveal the form of the body, but for entirely different purposes. // The tacit conventions on which the understanding of everyday language depends are enormously complicated”. In other words, of course there are different uses and different purposes to which our language is put. More than this, these uses affect our understanding of which thought is expressed. However, it is not this outer clothing that is important (not the way in which a thought is expressed) but the thought beneath it (what the thought is that is expressed). Moreover given that “all the propositions of our everyday language, just as they stand, are in perfect logical order”, there is something that underpins these diverse uses and allows them all to have sense. Whilst there are many forms of proposition, there is only one general form. As such it does no good to simply point out that there many uses of language, because whilst that is true, it doesn’t bear on his arguments for a general form of proposition. As Wittgenstein says in 4.5 “It is clear that only what is essential to the most general propositional form may be included in its description- for otherwise it would not be the most general form”.


First of all, it should be noted that, in a certain sense, Wittgenstein is happy to sanction that there is a ‘general form of a proposition’. That is, there are certain features of a sentence without which we wouldn’t say that it is a proposition. In Philosophical Grammar he asks “Why not?” to giving a general propositional form. He says, “of course, you can’t draw a boundary if you have decided in advance not to recognise one. But of course the question remains: how do you use the word ‘proposition’? In contrast to what?” In answer to this he tells us that “A general propositional form determines a proposition as part of a calculus.” At this stage, it is being part of a calculus that allows a sentence to connect with an extra-linguistic reality.

Even in the PI there are certain features of a proposition that Wittgenstein holds fast to. Indeed, I would go so far as to say, they are the very features that he believes to be part of a proposition in the Tractatus. One could say that “This is how things stand” is co-extensive with the totality of propositions; and part of the reason for doing so is that that is what, in English, it is to ‘sound like a proposition’. It, itself, is a proposition in English that may be alternatively expressed (see §134) “such and such is the case” or “this is the situation”. Any genuine proposition, once we know what it means, will tell us ‘what is the case’. A second feature that he maintains is that a proposition is that which we, in our language, can apply a calculus of truth-functions to i.e. those sentences that can be considered true or false. That is to say that the proposition says something about reality and if that obtains, the proposition is true; if not, it is false.

However, the point being made in the PI is that this is yet to say anything at all. In §134 he says, “To say [about ‘This is how things are’] proposition agrees (or does not agree) with reality would be obvious nonsense”. In other words, it is not itself a proposition, nor does it say what it is for a proposition to have a sense (just that if it is a proposition, it will have one). As for a proposition being something we can predicate true or false of, this is something that “belongs to our concept ‘proposition’ but does not ‘fit’ it.” (§136) We can predicate ‘true’ [in English] of any sentence that says something that agrees with reality but is not a property that engages with a sentence to make it a proposition. To say that a proposition is true, is nothing other than to say what the sentence says (‘p’ is true = p // ‘p’ is false = not-p). He is using a deflationary view of truth. To know what truth means in a particular case we have to find out what it is for that proposition to agree with reality:

The use of ‘true or false’ has something misleading about it, because it is like saying, ‘It agrees with the facts or it doesn’t’, and the very thing in question is what ‘agreement’ is in here. (OC 199)

However, these aren’t immediate criticisms of the Tractatus. The Tractatus never saw ‘This is how things stand’ as a bona fide proposition and in the Notebooks, he grapples with the same worries “But how am I to explain the general nature of the proposition now? We can indeed say: everything that is (or is not) the case can be pictured by means of a proposition. But here I have the expression ‘to be the case’… It is just as problematic.” Nor did he see ‘truth’ as a mechanism that gives a proposition sense. In fact the positions from the PI mentioned above are of the very essence for the early Wittgenstein (and as such, form part of the motivation for there being a general form of proposition). Consider proposition 4.06 and its commentary:

4.06 A proposition can be true or false only in virtue of being a picture of reality.

4.061 It must not be overlooked that a proposition has a sense that is independent of the facts: otherwise one can easily suppose that true or false are relations of equal status between signs and what they signify

4.063 [A proposition] does not designate a thing (a truth-value) which might have properties called ‘false’ or ‘true’.

4.064 Every proposition must already have a sense: it cannot be given a sense by affirmation. Indeed its sense is just what is affirmed. And the same applies to negation, etc.

Here, truth and falsity do not add anything to the proposition; otherwise a statement would have a different sense depending on whether it was true or false. Instead, ‘p’ has the one and the same sense whether it is affirmed or denied (does or does not obtain). Of course, the important question, which the quote from On Certainty stresses, is for us to understand what ‘agreement’ consists in, in the proposition. It is precisely because a proposition must ‘already have a sense’ that we cannot say that the proposition tells us “This is how things stand”. Instead, we need to analyse the proposition in order to find out what it says about reality (what the agreement with reality consists in; what situation is shown by the proposition).

2.22 What a picture represents it represents independently of its truth and falsity, by means of its pictorial form

If the pictorial form shows us the sense of the proposition, independently of its truth and falsity, we must be able to foresee the form of all propositions (i.e. all sentences with sense). As such nothing Wittgenstein says in §134 or 136 undermines the Tractatus argument directly and even helps us see the motivation for it.


It might very well be denied that we have thus far found nothing to argue with about the Tractatus. How about that our propositions must be analysable into elementary propositions? How come the proposition must show its sense by means of its pictorial form or by its form at all? Such requirements were derided in the Philosophical Grammar as a form of chemical analysis trying to reveal what is hidden, where our normal propositions are fine as they are. However, we have not yet found reason for such complaint. It cannot be that we should ‘analyse’ (construed broadly) our language in order to find which propositions make sense. Consider the following:

4.023 A proposition constructs a world with the help of a logical scaffolding, so that one can see from the proposition how everything stands logically if it is true. One can draw inferences from a false proposition

Is this not the point of language-games in his later philosophy? Do we not want to see the internal relations between concepts? Do we not want to see the inferences that can be drawn from various propositions with sense? Cannot this be done even when the proposition is not true?

Secondly, we cannot say that the propositions in such an analysis are not meant to show us what is the case if true, that there are certain forms that can do this and certain ones that are used illicitly and lead to nonsense. Of course, the form is not decided independently of how we use our propositions. Despite this, it must be noted that Tractarian Wittgenstein isn’t saying that the sense of a proposition can be decided without any knowledge of how we use propositions to describe ‘reality’. For example he says, “The application of logic decides what elementary propositions there are. // What belongs to its application, logic cannot anticipate” (5.557). The general form of the proposition tells us that with a perspicuous sign-language we would be able to see what sense is being expressed from the form alone (without knowledge of truth/falsity), and that our propositions can be analysed as expressing one of these senses. However, we would not necessarily be able to construct such a sign language without knowing that certain situations have existed in reality (have been ‘true’).

This is why in theory the ‘general form of the proposition’ can survive his later criticism of the logical independence of elementary propositions. He says in Notes on Logical Form: “It is, of course; a deficiency of our notation that it does not prevent the formation of such nonsensical constructions, and a perfect notation will have to exclude such structures by definite rules of syntax… Such rules, however, cannot be laid down until we have actually achieved the ultimate analysis of the phenomena in question. This, as we all know, has not yet been achieved”. Analysis is not something magical and form does not do anything by itself. The final product should help us see how we actually do use our propositions. The point of analysis is just to tease out the way in which the proposition relates to reality, as this is sometimes obscured by our words.