The problem of action is to explicate the contrast between what an agent does and what merely happens to him, or between the bodily movements that he makes and those that occur without his making them. According to causal theories of the nature of action, which currently represent the most widely followed approach to the understanding of this contrast, the essential difference between events of the two types is to be found in their prior causal histories: a bodily movement is an action if and only if it results from antecedents of a certain kind. Different versions of the causal approach may provide differing accounts of the sorts of events or states which must figure causally in the production of actions. The tenet they characteristically share is that it is both necessary and sufficient, in order to determine whether an event is an action, to consider how it was brought about.

Despite its popularity, I believe that the causal approach is inherently implausible and that it cannot provide a satisfactory analysis of the nature of action. I do not mean to suggest that actions have no causes; they are as likely to have causes, I suppose, as other events are. My claim is rather that it is no part of the nature of an action to have a prior causal history of any particular kind.