To: Graham C.
From: Geoffrey Klempner
Subject: Deflationist view of the predicate 'true'
Date: 10 April 2008 12:05

Dear Graham,

Thank you for your email of 30 March, with your University of London Logic essay in response to the question, 'If the predicate 'true' does not pick out any property, what role does it play?'

There is a lot about this essay that I liked. I approve of its sceptical attitude, although I would describe myself as mostly a minimalist about truth. You lay out the issues clearly, offering sound criticism of some of the less than convincing arguments for deflationism, and also make your own contribution to the debate.

I also approve, in principle, of the stance that 'the predicate 'true' plays a variety of roles which any theorist ought to be concerned to capture', which has resonances of Wittgenstein's famous point in 'Philosophical Investigations' about the 'family resemblance' between different things we call 'games'. It is a view which can be developed to good effect, for example, in analysing the concept of 'knowledge'.

I wonder, however, whether you have seen that this is ideology, no less than the other ideological views listed by McGinn. If one were to give it a name, along the lines of McGinn's DIME classification which you cite, it would be 'Complexify'. Complexifying is different from Domesticating (because the complexity may resist unravelling to some extent), yet it is not a claim about Irreducibility (because there remains analytical work to do). Those ideologically opposed to Complexifying might accuse its proponents of Mystifying, but that is just an unkind slur. It is not the same as Elimination, although in principle there is no reason why a Complexifying account need not also be Eliminative, at least, in its ambitions.

I appreciate that you have used the question to write an essay which gives a bit more than was asked for. In an exam, it would be within the remit of the question to briefly give your reasons why you reject the claim that 'true' does not pick out any property, before going on to examine the possible ways (from your point of view, in 'counternecessaryfactual mode') in which 'true' would have a role were it not the case that it picked out a property -- and perhaps then going on to indicate how the shortcomings of this explanation support your sceptical view.

My memory may be a little hazy, but I am not sure that Strawson does offer a performative analysis of 'true'. (If I'm wrong, please let me have the reference.) In his Aristotelian Society debate with Austin, Strawson took a deflationary line against Austin's attempt to resurrect a version of the correspondence For Strawson, talk of facts is otiose, since every expressible true statement has an associated 'fact'.

Indeed, it is not necessary to attribute any special performative function to the predicate 'is true' (i.e. in addition to any performative aspect of assertion as such) in order to view the predicate 'is true' as having an endorsing role. You consider one aspect of this, in the case of 'Catholicism is true'. If I say, 'What Graham says about truth is true', I am saving myself the trouble of listing the claims you make about truth. In real life, admittedly, we rarely agree with everything a person says, so one is more likely to come across a statement like, 'What Graham says about truth is largely true' (or 'mostly true'), meaning that if one were to list all the things Graham says about truth, one would be prepared to assert most of them.

(The semantics for the quantifier 'most' are a matter of debate, but we can bracket that for present purposes.)

'Catholicism is true' is an interesting case because it seems to imply much more than agreeing with a given set of propositions. It is closer to, 'My world is a Catholic world' which is arguably not a truth claim in the literal sense at all. On the other hand, I accept that this is could be seen as a legitimate example of how the term 'true' as used in our language has more than one role. The notion of a 'true believer' is one that is difficult to capture without the use of the term 'true', yet it seems a far cry from the 'predicate of disquotation'.

What you really need to say, however (and what the question is asking for) is what follows from the view that 'true' serves a generalization function. Here, you would need to go much further into the difference between the simple redundancy claim, and the claim that the predicate 'is true' adds greatly to the representational powers of a language, and is therefore far from being redundant: I mean, the view that the use of this predicate is equivalent, in some sense, to propositional quantification.

For example, 'Graham said something true' would be represented as:

(EP)(Graham said that P and P).

(My keyboard can't do a backward 'E'.)

The problem is that propositional quantification is itself a questionable notion: whether viewed 'substitutionally' or 'objectivally'. This would definitely be within the remit of the question.

Another significant line would be Michael Dummett's critique of the redundancy theory, his argument that it is indeed (as you claim) equivalent to anti-realism, and that therefore the proper realist response is not an old-fashioned correspondence theory of truth but rather an account of the form of a theory of meaning where truth is the central concept.

Dummett strongly endorses Frege's vicious regress argument, in Frege's essay 'The Thought' for the view that 'truth' cannot be defined. The only alternative to the discredited correspondence theory for the would-be realist, argues Dummett, is to give a theory of meaning in terms of truth conditions.

Opponents of Dummett argue that you can give a Davidsonian style theory of meaning in terms of truth conditions without being committed to any metaphysical view about truth. Hence, the idea of minimalism. The point here, however, is that truth does indeed turn out to play a very important role despite the admission that there is no 'real property' which it picks out.

All the best,

Geoffrey