So, then, the central notion in intentional action is that of an agent intentionally bringing about an event. Bringing about is causing. In all other cases, causation can be expressed as a relation between events. When a substance (the brick) is said to cause an event (the shattering of the window), this can be expressed as an event which
61 There are a few actions which, if I try to do them, I must succeed. Trying is obviously one example; if I try to try to φ, necessarily I try to φ. Thinking that p in the sense of
bringing about the thought that p (where p is a certain proposition) is another example. If I try to have the thought that 2+ 2= 4, necessarily I succeed—because the very endeavour to produce that effect involves having in mind the effect, being aware of what affect I am trying to produce. But to do that is to have that thought. However, as I claimed above, the same does not apply to images. I can try to form an image of something and fail.
is a state of that substance causing the second event—the motion of the brick caused the shattering of the window. And the wind which causes the leaves to blow about just is the motion of air. Again, when the sun causes the ice to melt, it is the sun being hot which causes the ice to melt. It is in virtue of being in this state (i.e. characterized by this property), rather than that state, that a substance causes some event, normally because it is a consequence of laws of nature that events of that kind cause events of the subsequent kind. But when an agent intentionally brings about an event, what is the state of the agent which brings it about? When I move my hand, meaning so to do, what is the state of me which brings about its motion?
There are answers of two possible kinds normally given to this question. The answers of the first kind are false; the answers of the second kind are true but misleading. Answers of the first kind consist in citing passive states of the agent which cause the events in question–desires (alias wishes or wants) plus beliefs. On such an account, when I move my hand for its own sake (not in order to achieve something thereby), what happens is that my desire to move my hand causes the movement of my hand. When I move it in order to achieve something further thereby, e.g. get a book from the shelf, what happens is that my desire to have a book from the shelf and my belief that by moving my hand I shall get it, cause the motion of my hand. We may express such an account more formally as follows. When an agent performs an intentional action, what happens is that his desire for the performance of that act under its ultimate act- description, and his belief that this will be achieved by the performance of the act under its basic act-description (via a certain series of events), cause its basic result, and thereby, via the expected series of events, achieve the performance of the act.62My intentionally killing the man consists in my desire that I kill him and my belief that I would kill him if I
were to squeeze my finger and thus cause the gun to fire and put a bullet in him, causing my finger to be squeezed and thereby the gun to fire and a bullet to enter the man causing his death. My insulting you (or however else that act is described) consists in my desire that I insult you and my belief that I will do so by saying
62 Except, that is, in the case where there is no basic result of the action but simply an unsuccessful attempt to achieve one; and in such a case there is simply the desire and the
‘You are a fool’, causing the words ‘You are a fool’ to come out of my mouth, and thereby you to be insulted. Such analyses are false in more than one way. First, they do not give sufficient conditions for action. Desires or wants (with appropriate beliefs) can cause the events desired without any intentional action taking place. A number of examples of this have been given in the literature. Here is one from Davidson:
A climber might want to rid himself of the weight and danger of holding another man on a rope, and he might know that by loosening his hold on the rope he could rid himself of the weight and the danger. This belief and want might so unnerve him as to cause him to loosen his hold, and it might be the case that he never chose to loosen his hold, nor did he do it intentionally.63
As I shall explain more fully in the next chapter, desires, like beliefs, are passive states. They are natural inclinations to action, which an agent finds himself having. And not merely that—they are involuntary; a man cannot immediately control his desires; they come to him unsought, and sometimes unwanted. All men find themselves from time to time with desires for food, drink, rest, sex, etc.; and with more complicated desires—for power or fame, say, of this or that kind, which vary with the individual. Yet, although a man cannot help his desires, he can hold them in check, refuse to ‘go along with’ them and let them have sway; and so long as he so refuses, they will give rise to no intentional action of bringing about or even trying to bring about the thing desired. But the desires may still have effects which the agent does not intend or welcome—including the occurrence of the thing desired, as in Davidson's example. There is more to intentional action than desire having its effects (directly or via a route believed efficacious).
Nor is desire a necessary condition of action. An agent may perform an intentional action without in any way desiring its performance under its ultimate act-description, or any other, in the normal sense of ‘desire’. I may repay my debt, not because I desire to do so nor because I have any desire (in the sense of natural inclination) to do my duty. I just do my duty because I choose to do so. (This point will be developed more fully in the next chapter.) Davidson attempts to make the general approach of intentional action as involving something like desire, plausible by
talking, instead of ‘desires’, of ‘pro-attitudes’, which term covers ‘desires, wantings, urges, promptings, and a great variety of moral views’, etc., etc.64 The claim, then, put in a more formal way, is that when an agent performs an
intentional action, his pro-attitude towards the performance of that action under its ultimate act-description and his belief that this will be achieved by the performance of the act under its basic act-description (via a certain series of events), cause its basic result and thereby, via the expected series of events, achieve the performance of the act. Now, it is certainly true that in a very wide sense of pro-attitude, an agent must have a pro-attitude towards the performance of any action of his under its ultimate act-description, if he is to perform that action. But, now, this account, although giving a necessary condition for intentional action, on other grounds fails to give a sufficient condition. For we have pro-attitudes of some sort towards most things (it is in some way a good thing that I go upstairs, that I go downstairs, that I stay put), and those attitudes may make differences to what happens without any intentional action being performed. If the cited claim is to be made in any way interesting and illuminating, it must be tightened up to read not just ‘a pro-attitude’ towards the performance of an action, but something like ‘an attitude which holds the performance of that action to be on balance preferable to the performance of any incompatible action’. Not any pro-attitude but an overall pro-attitude must cause effects if they are to be one's actions. But now the analysis will not work because of a different problem, the problem of akrasia or weakness of will. As Davidson himself draws to our attention in another context,65an agent may do the action towards which he has a lesser pro-attitude in preference to that which he believes
most worth doing. A man may believe that he ought to use his money to repay his debts, that it is overall better, that he has overriding reason to repay his debts—and yet
64 D. Davidson, ‘Actions, Reasons and Causes’ in his Essays on Actions and Events, p. 4. I do not wish to suggest that Davidson would endorse the detailed use of this notion
below. His paper was making a claim only about the explanation and not the analysis of action.
65 ‘How is Weakness of Will Possible’ in op. cit. Davidson regards the view of this later essay (‘that a reason that is causally the strongest need not be a reason deemed by the
actor to provide the strongest (best) grounds for acting’) as a correction to the view of the earlier essay (‘that the propositional expressions of the reasons for an action are deductively related to the proposition that corresponds to the action as explained by those reasons’) by denying that ‘the strongest reasons are the strongest causes’. See
spend the money on drink or gambling instead. So he does an action, towards the performance of which he has no overall pro-attitude. For these reasons I conclude that analyses of intentional actions in terms of actions (or events) brought about by passive states fail.
The question to which we are seeking an answer was—when an agent performs an intentional action, what is the state of the agent which brings about the basic result of the action? The alternative to the answer that it is some passive state or states such as desire or want (normally plus a belief) is that it is an action (normally plus a belief), only not a bodily action but an act of will. Such acts of will have often in the past been called ‘volitions’. Volition theory runs somewhat as follows. When an agent does an action such as moving his arm for its own sake, what happens is that his volition to move his arm causes the motion of his arm. When he performs an action in order to achieve something further thereby, e.g. turns a handle by twisting his hand, what happens is that his volition to turn the handle and his belief that twisting his hand will cause the handle to turn causes his hand to twist and thereby the handle to turn. Intentional actions are volitions (plus beliefs) causing bodily movements, or other effects such as mental events—except in the case where the only action is trying, and then the action consists just in the volition without any effect. The volition is basically a trying, but it is only appropriate to talk of ‘trying’ where effort or failure is involved. When I move my hand easily and spontaneously, it seems odd to say that I tried to move my hand. However, the very same event which is called ‘trying’ when failure occurs is clearly involved in, and is the crucial active part of, any ordinary easily performed basic action. For suppose that I try to move my hand, yet fail, because suddenly I am paralysed. What I have contributed in the way of endeavour would clearly have been sufficient for my moving my hand intentionally to take place if I had not been paralysed. So volition theory gives to this element the technical name of ‘volition’. It is an element which we usually notice only when failure or effort are involved, and then we call it ‘trying’. But this thought experiment (‘if I had been suddenly paralysed’) reveals that it is there in every intentional action.
Volition theory seems to me basically on the right lines in the crucial respect that it does not analyse away the active element in intentional action in terms of some passive component, but acknowledges that it is there unanalysably—in the form of an act
of will. However, there seem to me two highly misleading aspects of the theory as I have expounded it so far and as it is expounded by its advocates.
The first misleading aspect is that volition theory represents intentional action as a more active thing than it often is. A ‘volition’ is supposed to be ‘an act of will’, and that suggests something like a decision or a choice. This suggests that every action is a quick decision subsequently efficacious. But there are three different things wrong with that suggestion. First, intentional action is often a matter of the agent allowing certain sequences of movement to take place rather than actively bringing them about. When I fidget, tap my foot, or scratch my hand, my body is, as it were, dictating the pace, as in a reflex movement. But what makes these things intentional actions is that I allow them to happen and could at any instant, if I so chose, stop them. The volition is, in these cases, no decision: it is a giving of permission; but not the active permission of signing a permission form, but the passive permission of doing nothing to stop something. So much action is a matter of the agent not stopping his body from doing what it does naturally under its own steam. But what is allowed to happen is the performance of an action if it is something which the agent, by choosing, can readily bring about or stop occurring. Further, actions often take a long time. My walking from Madeley to Keele takes an hour, yet walking to Keele is certainly an intentional action of mine. What makes it intentional, however, is not that at the start of the walk, I decided to walk; but rather that purpose or intention guided my movements throughout the walk. The ‘volition’ is to be regarded as a continuing permission rather than merely an original permission. And, finally, when actions do take a long time, they are performed by performing many separate actions. I write this chapter by writing first this sentence, then that sentence, and so on. Yet there is not, as volition theory sometimes seems to suggest, a separate volition corresponding to each of the latter intentional actions. If there was, my mental life would be overcrowded with volitions. What is more plausible, surely, is that when I perform an intentional action, either there is a separate ‘volition’ for that action; or there is a volition to perform a longer action, one way of performing which is, I believe, by bringing about a certain series of events. In the latter case the bringing about of each event of the series constitutes an intentional action.
The volition has the general character—‘to walk from Madeley to Keele’. The body selects (without any conscious choice of the agent's) a particular bodily routine to execute the volition—to put this foot here and that foot there, and so on. However, the occurrence of each part of the routine, so long as the agent believes that it forwards what he is seeking to achieve, does count as his performing an intentional action. My putting this foot there was an intentional action of mine, because, although there was nothing like a decision that the foot be there, there was a purpose of mine (‘to walk to Keele’) which my body was executing through a routine which involved the foot going there, and I had the belief (which, as we have already noted and as we shall see further in Chapter 7, is no conscious occurrence) that that routine (among other possible routines, perhaps) would achieve that purpose.
It is to avoid these misleading implications of volition theory, making intentional action too active a thing—implications suggested in part, I believe, by the word ‘volition’ itself, that I wish to replace the word ‘volition’ by a different word, ‘purposing’, and now state volition theory formally in a way which will cope with the difficulties. Purposing to bring about an event is trying to bring it about (minus any implications of effort or failure), or allowing an event to occur of a kind which the agent, by trying, can bring about, or by trying, can easily stop. When an agent performs an intentional action A, what happens is that his purposing to perform A (or a larger act of which A is, he believes, part) under its ultimate act-description and his belief that this will be achieved (in part or whole) by the performance of A under its basic act-description (via a certain series of events), cause the basic result of A, and thereby, via the expected series of events, achieve the performance of A. (The exception to this is where the intentional action does not have a basic result, viz. is simply a trying, and then it consists simply in a purposing to bring about something plus a belief as to how that is to be done.) When the basic result is a long event (e.g. the bodily movements involved in a one-hour walk), the purposing and belief last as long as, and cause, the whole event.
Note that on this theory there is not a separate purposing for the act under each description—the agent does not purpose to pull the trigger, and also purpose to shoot, also purpose to kill; simply a purposing to perform the act under its ultimate act-description.
Nor is there a separate purposing to perform each small constituent act which makes up a larger act; simply a purposing to perform the larger act which the body executes in a way which the agent believes will achieve his aim.