Critics of attempts to explain meaning in terms of truth-conditions have tended to charge their opponents with misconceptions regarding truth. I shall argue that the 'naïve' version of the truth-conditional theory which best accounts for its resilience fails for a different and more basic reason, namely, circularity arising from the contingency of meaning. One reason why this problem has been overlooked is a tendency (noted by Dummett in a different connection) to assimilate the naïve truth-conditional theory to an idealized verificationism.

Richard Heck has contested my argument that the equation of the meaning of a sentence with its truth-condition implies deflationism, on the ground that the argument does not go through if truth-conditions are understood, in Davidson's style, to be stated by T-sentences. My reply is that Davidsonian theories of meaning do not equate the meaning of a sentence with its truth-condition, and thus that Heck's point does not actually obstruct my argument.

A response to certain parts of Rumfitt : I defend Davidson's project in semantics, suggest that Rumfitt's use of sentential quantification renders his definition of truth needlessly elaborate, and pose a question for Rumfitt's handling of the strengthened Liar.