Misdisquotation and substitutivity: when not to infer belief from assent

Abstract

In A Puzzle about Belief Saul Kripke appeals to a principle of disquotation that allows us to infer a person's beliefs from the sentences to which she assents (in certain conditions). Kripke relies on this principle in constructing some famous puzzle cases, which he uses to defend the Millian view that the sole semantic function of a proper name is to refer to its bearer. The examples are meant to undermine the anti-Millian objection, grounded in traditional Frege-cases, that truth-value is not always maintained when coreferential names are intersubstituted in belief reports. I argue here that our disquotational practice is sensitive to certain shifts in conversational context, and it is only if we overlook these shifts-if we misdisquote-that we can draw the conclusions Kripke wants to draw from his examples. In the wake of this conclusion, I provide a contextualist treatment of Kripke's puzzle cases. I show how this treatment is motivated by certain nonns of rationality, and I defend these norms against an intriguing anti-Cartesian theory of mind. Throughout the paper, I develop the larger implications that my treatment of Kripke's argument has for the semantic theory of names and belief reports, and, more generally, for our picture of the relation between linguistic behaviour and our states of mind.