Can a theory of truth serve as a theory of meaning? If not, why not?

Résumé du document

Recent developments in the theory of truth by Tarski1 have brought new material for inquiries in the theory of meaning – this is what indeed thinkers such as Davidson believe; but are they right believing this ? More exactly, in what sense can we legitimately discuss the idea that a theory of truth can ‘serve as' a theory of meaning? As we understand it, this question commits us to elucidate two things: (i) what is required by (and for) a ‘good' theory of meaning; and (ii) to what extent a theory of truth can satisfy those requirements.

As Davidson puts it (and as most authors second), the main and crucial requirement for a good theory of meaning is that it fulfills the task of explaining how we come to master natural languages – that is, how it is that we have the a prior ability to understand and utter an infinite number of different sentences.

Besides, what one could logically expect from a good theory of meaning, is that it eventually produces unequivocal theorems of the form‘s means that p' – to put it another way : that it produces theorems that unequivocally link every possible sentence with some translation of its, in such a way that all of them are made directly intelligible. Given this, Davidson's core intuition is that Tarski's theory of truth could provide a convenient basis, for it contains the wanted ‘finite-to-infinite' connection, and an unequivocal form for its theorems – the famous ‘convention T'.

Extraits

[...] Thereby we measure how far from the ‘naïve' concept of truth is the sort of theory of truth Davidson's theory of meaning relies on : truth in ‘meaning that p for A' defined as ‘being true for A iff p' is at best a (one among many) conceptual (not textual) approximation of what A really had in mind. Thus we may conclude : if a theory of truth is to serve as a basis for a theory of meaning, it should not be a surprise that it has to be a semantic theory of truth (the first of which was coined by Tarski), for this sort of theory deals with sentences essentially, and (modestly) refers to truth as truth-in-a-given-object-language. Moreover, it has to come with further requirements, of the kind precited. [...]

[...] A hint is given with the very straightforward response to the third point : it is recommended that we ‘take truth to be a property, not of sentences, but of utterances, or speech acts' ; this means we should first rework Tarski's theory, adding data about the utterer, and the time and place of utterance. But more essentially, this also means that a theory of meaning for any language is a pragmatic program : as Segal notices, Davidson's intuition follows Quine's idea of a ‘radical interpreter', suggesting that a theory of meaning for any language L should be considered satisfactory if it ‘systematically yields correct interpretations of what an L-speaker says'. [...]

[...] Can a theory of truth serve as a theory of meaning? If not, why not? Recent developments in the theory of truth by Tarski1 have brought new material for inquiries in the theory of meaning this is what indeed thinkers such as Davidson believe2 ; but are they right believing this ? More exactly, in what sense can we legitimately discuss the idea that a theory of truth can ‘serve as' a theory of meaning? As we understand it, this question commits us to elucidate two things: what is required by (and for) a ‘good' theory of meaning; and to what extent a theory of truth can satisfy those requirements. [...]

[...] So Tarski's theory of truth is formally convenient ; but what should we think is the right connection between (precisely) truth and meaning ? How is the move made from is true if and only if p' to means that p' ? Davidson (1967) explains quite straightforwardly : It is this : the definition works by giving necessary and sufficient conditions for the truth of every sentence, and to give truth conditions is a way of giving the meaning of a sentence. [...]

[...] Unless such an account could be supplied for a particular language, it is argued, there would be no explaining the fact that we can learn the language: no explaining the fact that, on mastering a finite vocabulary and a finitely stated set of rules, we are prepared to produce and to understand any of a potential infinitude of sentences.4 Besides, what one could logically expect from a good theory of meaning, is that it eventually produces unequivocal theorems of the form‘s means that p' to put it another way : that it produces theorems that unequivocally link every possible sentence with some translation of its, in such a way that all of them are made directly intelligible. Given this, Davidson's core intuition is that Tarski's theory of truth could provide a convenient basis, for it contains the wanted ‘finite-to-infinite' connection, and an unequivocal form for its theorems the famous ‘convention T'. Tarski intended to define truth in any language L as the ‘logical conjunction' of all true sentences in L (considered ‘partial definitions'5), which eventually led him to accept, in the case of infinite languages (i.e. [...]