To: Kathleen C.
From: Geoffrey Klempner
Subject: Truth conditions and metaphysics
Date: 3 August 2006 14:51

Dear Kay,

Thank you for your email of 24 July, with your third essay for the Metaphysics program, in response to the question, 'When you grasp the meaning of a proposition, you know the conditions under which it would be true.' -- Is this statement correct? What light does it shed on the notion of truth?

How does this question relate to the realist/ anti-realist debate?

One question - which we can come to in a moment, is in what way the realist or anti-realist is entitled to talk about 'truth', or, what each 'understands' by that notion.

However, as I have argued in the program, there are reasons why one would want to reject the identification of an account of the meaning of a proposition in terms of truth conditions with 'realism'. You can be an anti-realist and yet still agree that 'When you grasp the meaning of a proposition, you know the conditions under which it would be true.'

The first point to make is that a Tarski-style 'theory of truth' - which derives 'theorems' like 'Snow is white' is true iff snow is white from 'axioms', 'Snow' refers to snow, 'white' applies to x iff x is white - is the basis for a compositional theory of meaning which explains how a speaker is able to understand potentially infinitely many sentences on the basis of a knowledge of the meanings of a finite number of words. This is argued in Donald Davidson's famous paper 'Meaning and Truth'.

This is something we ideally want. In order to achieve the goal of a Davidsonian theory of meaning it would first be necessary to reduce ALL English idioms to first-order predicate logic - something which to date has not yet been fully achieved.

Anti-realists like Dummett, however, would argue that the entire project is misconceived because a truth conditions theory can never be sufficient to explain understanding. The best case, for Dummett's point, as I argue in the program, is the that of unrestricted infinite generalisations which you quote.

The best way to defend a truth conditions theory, on the other hand, is by resisting the temptation to construct imaginary scenarios, or talk about beings with increased perceptual or physical powers. We understand the proposition, 'Every planetary atmosphere contains nitrogen' because we understand the meanings of the words and the import of putting them together in a sentence. If you ask what it is that my understanding 'consists in' there is nothing informative to say.

So what has this got to do with realism and anti-realism?

The realist does not have to give a 'correspondence theory of truth'. Although Tarski called his account of the T-predicate a 'correspondence theory' this is so only in a weak sense which does not involve any heavy metaphysical commitment to 'facts' or suchlike. Once you have explained how the truth-predicate works, you have said all there is to say on the subject of truth. Or, in other words, truth is indefinable.

The anti-realist does not have to give an account of truth which equates truth with verification, however construed. It is possible to make the points which the anti-realist makes without going against the basic truism that 'truth transcends verification' - however sure you may be that you have 'verified' a proposition, there remains the logical possibility that you are wrong. Or, in other words, truth is indefinable.

So either way, you have a truth conditions theory of meaning. either way, truth is indefinable. Yet a huge metaphysical gulf separates the realist and the anti-realist.

All the best,

Geoffrey