Omissions and Causalism *

Omissions are puzzling, so puzzling that people tend to say puzzling things about them and give up otherwise attractive philosophical theories in order to accommodate them.1 In this paper I suggest that omissions make trouble—serious trouble, and trouble of a new, sui generis kind—for “causalism,” the standard view or family of views of agency. In particular, I am interested in causalism as an attempt to explain what it is for an agent to behave intentionally. I will argue that causalism cannot accommodate intentional omissions—or, at least, it cannot account for them in the same way it accounts for (positive) actions. As a result, causalism is incomplete—or, at best, highly disjunctive—as a theory of what it is to behave intentionally. I will bypass the question whether omissions can be, properly speaking, actions—“negative actions” or “active nondoings,” as they have been called (see, e.g., Kleinig 1976). For some people (notably, Thomson 1977), actions are a subclass of events, where events are particulars with specific spatiotemporal locations, intrinsic properties, etc. On this kind of view, it’s hard to count omissions as actions, for omissions don’t appear to have specific spatiotemporal locations, intrinsic properties, etc. Nevertheless, even if omissions aren’t actions, it seems that agents can still fail to do things intentionally, and it makes sense to ask under what conditions an agent’s not doing something is intentional (see, e.g., Ginet 2004). Thus, even if omissions aren’t actions, a theory of what it is to behave intentionally should be able to accommodate omissions. (Note that, if omissions aren’t actions, a theory of what it is to behave intentionally is not the same thing as a theory of what it is to perform an intentional action, and it might not even be the same thing as a theory of what it is to act intentionally.)2