Tuesday, December 27, 2011

Essays on concepts and truth conditions

To: Gordon F.
From: Geoffrey Klempner
Subject: Essays on concepts and truth conditions
Date: 20 July 2006 13:57

Dear Gordon,

Thank you for your email of 11 July, with your essay for the Philosophy of Language program, in response to the question, 'What are concepts? How does analysing the concept of a 'concept' help to illuminate the way language works?', and your email of 20 July, with your essay in response to the question, 'Assess the philosophical significance of the claim that to understand a proposition is to know its truth conditions'.


Concepts classify, and apply to domains. An object x may be said to fall, or not fall under a concept F. Either way, to state that this is so is to convey information, either that x is F, or that x is not F.

So what is there to explain? What role is there for philosophy?

The question concerns, not so much the 'what' of concepts - that they classify, apply to domains etc. - but the 'how'. How do they do this? How is it possible to grasp a concept, and what is it that one grasps in grasping a concept? How is it possible for two or more persons to agree on a concept, on its 'meaning' or 'use'?

In your essay, you have given a very full description of the 'what', but not said too much about the 'how'. You do say, however, that concepts are associated with 'following rules', and this is the key move that takes us from the what to the how.

Consider the statement, 'A concept has an extension.' The extension of the concept 'dog' is all dogs, past, present and future, anywhere in the universe. No-one 'knows' this set, in the sense of being able to identify it as such (except for God, if God exists). Isn't it a bit extravagant to claim that, despite this, the concept 'dog' HAS an extension? And what is it, to us, if it has?

What this shows is that in any philosophical account of the nature of concepts, there has to be something which relates concepts to us, to language users. There is such a thing as 'grasping' a concept, 'using' it correctly or incorrectly. It is concepts 'in use' that are of interest to a philosopher.

Our best lead on this is that it has something to do with 'rules'. The question, now, is how these rules work and what they achieve.

One view, argued by Michael Dummett, is that the task for a 'theory of meaning' is to explain these rules, to make them explicit, in non-circular terms. It is not clear, how this can be achieved, since the only language we have to formulate our theory of meaning is the very same language that our theory of meaning is for. You might try to make a start by reducing 'dog' to some set of characteristics, but then each of these characteristics ('barks', 'has four legs' etc.) is a concept with rules that need to be articulated in a non-circular way.

An alternative approach would be to explain in general what the rules do without giving an explicit 'theory of meaning' in this sense. For example, some concepts are associated with a recognitional capacity. Arguably, there could not be a language where none of the concepts had this feature. Yet, equally, any language which was restricted to concepts which were associated with a recognitional capacity would be stuck at the ground level of experience, its users unable to formulate theories or explanations.

Hence the idea, expressed in Quine's image of the network which comes into contact with experience at the edges, of concepts performing a role which is not restricted to the description of the immediate contents of experience. Arguably, the role of concepts in embodying a 'theory' about the world is their most interesting aspect. And so on.

The question is - the point of the essay question - is how interesting is this? what is there of philosophical substance in the inquiry into the concept of a 'concept'? That is what the Language program is, in large part, about.

Truth conditions

You have chosen a good example for raising the question about truth conditions, both in the explicit example of Gertrude Stein's pigeons, and more generally in the question of the 'meaning' of a work of art.

The statement, 'Pigeons are on the grass, alas' involves more than just the statement that pigeons are on the grass. It is the expression of a propositional attitude. More explicitly, one might say, e.g. 'Gertrude regrets that pigeons are on the grass'.

As discussed in unit 7, Davidson, in his paper, 'On Saying That', offers an analysis of indirect discourse which avoids the 'Frege problem' arising when one construes terms within the context of indirect discourse (a 'that' clause) as referring to their normal sense rather than to their reference. One question that one might consider is how this analysis might be extended from 'X says that P' to 'X regrets that P.'

Why it important to do this? Language users know how to formulate indefinitely many sentences in indirect discourse, on the basis of their grasp of the individual words, or concepts. It follows that there must be rules which enable them to do this, rules which cover statements like, 'Gertrude regrets that pigeons are on the grass' or 'Gertrude said that pigeons are on the grass'. The lesson from Frege and Davidson is that this is not quite so easy as it looks.

But what about the more general question, the 'meaning' a statement like Gertrude Stein's statement about the pigeons, in the context of a work of art?

Frege based his account of language on a fundamental distinction between 'sense' and 'tone'. The first task is to explain how information is conveyed by using words with a sense. After that, comes consideration of the 'tone' of the statement, what further things are conveyed to a hearer or reader beyond what is strictly said.

Here is a more homely example. David Bowie, in one of his songs, laments that he is 'always crashing in the same car.' The proper Fregean or Davidsonian, truth-conditional way of understanding this is that Bowie owns a car, which he has crashed several times. However, Bowie may well have intended to mean something quite different, for example, 'I never learn from my mistakes'. There is no 'theory' which explains how we are able to understand the original statement in these terms.

Does that disprove the truth-conditions theory? On the face of it, no. That theory was intended to explain the basic phenomenon of factual meaning, not the other things that can be done with language. However, there is a counter-argument that far more of language than the Fregean or Davidsonian is ready to admit is 'metaphorical' in character, and that a far more fruitful area of inquiry would be into how 'metaphor' operates, what its 'rules' are.

All the best,