When people take part in joint activities – business transactions, chess games, piano duets – they perform a variety of joint actions. They say things to each other, hand things to each other, nod at each other, gaze at each other, and through these advance their joint activities. Many of these joint actions, or their parts, are communicative acts through which they get others to understand what they mean. What sort of acts are these, and how do they work?

The traditional view is that communicative acts are performed by a speaker autonomously. In the drugstore, when Stone said “I'll be right there,” she was making a promise on her own. Although she directed it at me, I had no real part in it. A promise expresses a commitment to do something in the future, and speakers express such commitments on their own. In that tradition, the focus is on speakers. There is no mention, no hint, that addressees have any role.

Paradoxically, the traditional view carries the seeds of its own destruction. The very notion of meaning – speaker's meaning – requires addressees to join speakers in a special way, and so do other notions of speech acts. We will discover, on closer examination, that communicative acts are inherently joint acts, and that they are just one level of an entire ladder of joint actions.