Title:
Analyticity Without Conventionalism: A Game-Theoretic Account
Abstract:
To philosophers who want to explain the apparent necessity of mathematics and logic without positing anything metaphysically suspect, the idea of conceptual or analytic truths has long held appeal. However, standard accounts of the notion (truth in virtue of meaning, truth by convention, truth knowable through meaning alone) face powerful objections, e.g. the objection that our meaning something by a sentence can only explain what it means, not make it so that it is true, and Quine’s objections in Two Dogmas that analyticity circularly depends on synonymy, which in turns itself depends on analyticity.
This talk develops an account of analyticity that meets these objections, while preserving the central insight of the traditional view: that denying an analytic truth involves a kind of unintelligibility or incoherence. The account extends a game-theoretic theory of semantic content, on which the meaning of a term is fixed by the equilibrium of a basic constitutive practice of using it. By extending the account, first to sentences and then to propositions, a proposition is defined to be analytic just in case asserting it is correct at every possible occasion of use.
Because a proposition’s correctness conditions partly constitute its identity, to suppose that an analytic truth is false is to suppose that a different proposition, involving different concepts, is true, not the analytic proposition itself. This explains why denying an analytic truth has always seemed unintelligible. The account also separates the analytic from the a priori, since some analytic truths turn out to be knowable only a posteriori.