Using mis- and realignment cases in the development of interactional sequences in which prospective actions and events are being parleyed in everyday English conversation, this study shows that participants distinguish between the initiating actions of Proposal, Offer, Request, and Suggestion, if these labels are read as technical terms for different reply constellations to the questions (i) who will perform the future action? and (ii) who will profit from it?. These different action types are routinely linked to different sets of social action formats, and via these speakers can frame their turns as performing one action type in contrast to another and recipients can identify these actions as such and suitably reply to them. When misalignments and realignments happen, this is often due to the fact that ‘polysemous’ linguistic formats have been used to carry out the initiating action.