2. Action and Generality

Abstract

The philosophy of action defines itself by reference to a pair of canonical divisions. First, among events, a distinction is drawn between that which is a "mere event" and that which is an "action." Then, a second distinction is drawn among actions, between that which is action in some qualified way---because it is unintentional, or unconscious, or unfree, or what have you---and that which is action unqualifiedly. "The standard approach," as Anscombe called it, is to take for granted the genus EVENT, and to hunt for the differentia of action; or to take for granted the genus ACTION, and to hunt for that of unqualified action. The negative aim of the dissertation is to argue against the standard approach; the positive aim is to develop an alternative. I first distinguish three different forms of generality---forms that are associated with the traditional ideas of an accident, a category and an essence. I then ask: What kind of generality is exemplified by each of the two canonical divisions? The standard approach is viable only if both divisions exemplify what I call "accidental generality." In fact, neither does. The division of action into qualified and unqualified action is an example of what I call "essential generality." I argue that, as in all such cases, the question, "What is unqualified action?" reduces into the question, "What is action?" The other division is an example of what I call "categorial generality." The concept "action" refers to a category of a distinctively practical kind: an agent must think that what she is doing falls under this category, if, in fact, it does fall under it. Then any attempt to describe a differentia must be circular: sooner or later it must refer the agent's thought; and the agent's thought must in turn make reference to that which it needed to explain. On the positive account defended here, an action is a certain sort of temporally-ordered system of ends and means. The claim is that the agent herself must think of what she is doing as being such a system---if, indeed, it is one.