Introduction to Part I

1.

Although accounts of action have been central to most philosophical systems from Plato to Kant, it is only in recent years (following the writings of Wittgenstein and Anscombe, Chapters 1 and 11) that philosophy of action has come to be seen as a subject in its own right. We begin this volume with enquiries into what we might call the most basic question in this area of study: what is action?

One obvious suggestion is that action is bodily motion. But not all bodily motion is action; when you jog my arm, the motion of my arm is not an action of mine - I haven't moved my arm - and it isn't an action of yours, either. So what is the difference between those bodily motions that are actions and those that are not? The most popular strategy is to adopt a causal theory, whereby the distinction between actions and other forms of behavior lies in their causal origins; a sneeze, for instance, is typically not going to count as an action, because it has the wrong sort of cause. So which causes are of the right sort? Davidson's influential answer to this question identifies the causes of action with (the onset of) beliefs and pro-attitudes (such as desires, preferences, and values) that rationalize the action, that is, show how the action that is their effect made sense to the agent, and so can be thought of as the agent's reasons for doing what he did (see Chapter 19). Most sneezes are not actions, because they are not caused by rationalizing beliefs and desires, but by such things as tickles. Davidson saw this account as an improvement on earlier views which identified the causes in question with inner acts of will. His view is a form of event-causalism (since the action is an event and its causes are events, too), and due to its prominence in the literature is frequently also referred to as 'the standard view'.

Event-causalism faces two general challenges. The first, recognized by Davidson himself, is that the right sort of cause (viz. a 'rationalizing' one) can bring about an action in the wrong sort of way (i.e. not in virtue of its rationalizing power). So we don't just need things of the right sort to do the causing, we need them to do their causing in the right sort of way. Davidson (Chapter 2) gives the now famous example of a climber who wants to rid himself of the weight and danger of holding another man on a rope, and who knows that the way to do this is to let go of the rope; but if this belief and desire together so unnerve him that his grip relaxes and the rope slips through his fingers, the loosening of the grip is something that happens to him rather than something that he does; so it is not an action of his even though it is caused by a rationalizing belief-desire combination (Davidson 1973). This has come to be known as the problem of deviant causes (addressed by Smith in Chapter 28).

The second challenge to event-causalism relates to the lack of any causal role played by agents themselves in all this. If actions are events caused by (the onset of) prior mental states and/or neural processes, we arguably lose sight of what, if any, role we play in all this. If we are not ourselves actively involved, are we really the agents of our own actions or are we mere vehicles for them? It seems insufficient for agency that the causes in question occur inside us. Our digestive processes, for example, are alien to our agency in a way in which our actions had better not be. This worry has come to be known as the problem of 'the disappearing agent'; it affects any account that, like Davidson's, understands actions as a species of events, viz. ones with a cause that is not identified with the agent. This problem is the focus of Hornsby's contribution in this part (Chapter 6). (There are other challenges to the details of Davidson's view, which are discussed in Parts IV and V.)

So an alternative strategy that is not event-causalist - and is sometimes even misleadingly described as non-causalist - identifies the cause with the agent himself (Chisholm 1964; Reid 1969; O'Connor 2000) rather than with some event. This idea, known as agent-causation, is thought to avoid the two problems discussed above. Agent-causalists disagree over whether the agent causes her action or whether the action consists in her causing a certain result (the latter is argued by Alvarez and Hyman in Chapter 5). But either way, there is the further question of whether an agent's causing something should itself be understood as an event, and if so, what, if anything, brings about that event. (Ruben 2003 denies that there are such events as the causing of things by agents; O'Connor 2000 denies that they need further causes.)

Not everybody agrees that action is bodily motion with a particular kind of cause. For instance Frankfurt (Chapter 4) defends the non-causalist view that what makes a bodily motion of yours an action is that you are embracing it as your own and that it occurs under your guidance. On this account there can be actions that do not involve the causation of bodily motion at all, so long they are embraced by the agent in the relevant way. Examples of such actions might be pressing one's hand against a door to keep it closed, refraining from apologizing, and omitting to send a card. In addition, some 'volitionist' philosophers identify actions not with bodily motions, however caused, but with the inner causes of those motions, which they take to be acts of will or volitions. Other volitionists take actions to be complex events composed of volitions followed (causally or otherwise) by bodily movements; on this view neither the volition nor the bodily motion is itself an action. These and other related views will be considered in more detail in the introduction to Part II.

Whatever the causes of action may be, most of the above views seem to identify actions themselves with events of some sort. But some thinkers identify actions with processes rather than events. The precise difference between the two characterizations is contentious, but it is generally agreed that - unlike events - processes need not occur throughout or across a temporal stretch (Mourelatos 1978). Dretske (1988) argues that an action is the causal process of a mental/neural event causing a bodily event. More recent process-theorists inspired by Aristotle (e.g. Stout 1997) prefer to think of actions as non-causal processes. These are teleological processes defined by an end or goal that need not be achieved in order for it to be true that the process has taken place. One may, for example, be in the process of baking a cake without ever succeeding in baking one, or crossing the road without ever making it to the other side. So understood, there can be cake-baking or road-crossing processes without there having been a cake-baking or road-crossing event.

Whether actions are events or processes, it may seem that they are at least occurrences or happenings. In Anscombe's terms, "I do what happens . there is no distinction between my doing and the things happening" (1957: §29). On this outlook, the problem of action we have been dealing with is that of offering a way of distinguishing the doings of an agent from what 'merely' happens to him (see the chapter by Frankfurt in this part). But even this framework can be, and has been, rejected. Some philosophers take actions to be instances of relations (e.g. Hyman 2001). Others remind us that to act is to do something (e.g. bring about x) and then proceed to distinguish between the thing done (the deed?) and the event of one's doing it (Macmurray 1938; Hornsby 1980; Ricour 1986). This distinction is often compared to that between the thing thought and one's thinking it, or between the thing said and one's saying it.

2.

The term 'basic action' was first introduced by Danto, in his 1963 paper "What We Can Do." Danto's goal was to identify the point at which agency begins (and arguably freedom and moral responsibility with it, but see the discussion of these issues in our introduction to Part VI). Danto's governing thought is that no matter how complex the action I am doing, there must always be a basic element to it, viz. something by doing which I do everything else that I am doing. But the notion of the basic needs careful handing everywhere in philosophy, not least in the case of basic action. Baier (1971) has raised the worry that there are at least eight kinds of basicness, some of which are a matter of degree rather than kind: causally basic, instrumentally basic, conventionally basic, ontologically basic, logically basic, genetically basic, ease basic, and isolation basic. If so, we need to be sure which one of these we are talking about. Danto's own example of a paradigmatic basic action is that of moving an arm "without having to do anything to cause it to move" (so pushing it with the other arm won't count). Volitionists, by contrast, maintain that such an action as moving one's arm is the effect of a volition; this volition is the basic action and its effect, the moving of the arm, is another action (done by means of the basic action of willing).

Chisholm has offered an alternative, teleological, definition of basic action intended to be neutral on these issues of causality: "'A is performed by the agent as...