Abstract

It is common to believe that a speaker refers to x in virtue of intending to do so. I have four objections. The view is too intellectualized. It is theoretically incomplete: In virtue of what did the speaker intend to refer to x? Once completed it is redundant. It is misleading. A central idea of Gricean “intention-based semantics” is that speaker meaning is constituted by the speaker’s intention to communicate a certain content to an audience. I argue that the meaning is constituted by the content of the thought expressed. The requirement that the speaker intends to communicate that thought is theoretically unmotivated. It is standard among Griceans to believe that there is some constitutive constraint on what a speaker can intend by an utterance, a belief arising from one about a constraint on intentions in general. That alleged constraint ranges from the astonishingly strong “positive” one that X cannot intend to A unless X believes that she will A, to the much weaker “negative” one that X cannot intend to A unless she lacks the belief that she cannot A. I argue that there are no such constitutive constraints on intentions.