Abstract

Imperative sentences admit of many different uses, from imposing obligations to answering questions, granting permissions, and giving advice. Some declarative sentences, such as statements about what one should do or about what the speaker wants one to do, can also serve similar purposes. These utterances all share a single discourse role. They work as directives, whose defining effect, I argue here, is to propose that their addressee publicly commit to performing an action. Understanding directives this way can help explain four core features of directive speech acts. For one, directives are subject to norms of consistency that accord with rational coherence requirements for intentions. Second, directives can answer questions about what the questioner is to do. Third and fourth, imperatives can seldom be used in conjunction with prioritizing modal and preference-expressing declaratives. I propose a strictly pragmatic explanation for these phenomena, with minimal semantic commitments. Priority modals and expressions of preference can only convey the same type of information as imperatives insofar as those declaratives can be used to indirectly tell their addressees what to do. Imperatives, in turn, propose executive commitments directly and can only induce doxastic commitments to priorities or speaker-preferences indirectly.