Meaning and Truth

Robert Cummins

Donald Davidson, in a seminal paper, porposed the bold conjecture that a Tarski-style truth definition for a natural language would constitute a theory of meaning for that language. This proposal has two evident virtues. (1) It opens the way to an explanation of how infinite expressive power can be acquired and deployed with finite means; (2) It sets an agenda of apparantly tractable problems for the theory of meaning and understanding.

These virtues were enough to essentially capture the philosophy of language, and a good part of the philosophy of mind and linguistics. But there are reasons to suspect that Davidson's conjecture is mistaken. It does not really open the way to explaining understanding; it has dubious implications for mental representation, and the Tarskian combinatorics it relies on are inappropriate for understanding how concepts can be combined.