prichard, action, and volition

BRUCE AUNE

P R I C H A R D , A C T I O N , A N D V O L I T I O N

(Received 31 October, 1972)

H. A. Prichard's posthumously published essay 'Acting, Willing, and Desiring' is rarely read with sympathy nowadays. 1 It is included in recent anthologies, but it is generally taken as an example of how far wrong even a shrewd philosopher can go in dealing with the topic of action. I cannot deny that his essay contains some very startling claims - for example, that a human action is an activity of willing some change, and that when a man fails to do what he wills to do, his activity of willing is nevertheless "o f the same sort as what we ordinarily call and think of as an action." Yet it seems to me that, properly understood, Prichard's claims are really very close to the truth - much closer to it than the claims of his recent critics. My aim in this paper is to present a critical analysis of Prichard's basic reasoning about action and to relate his conclusions to problems of current interest and debate.

Prichard begins his essay by considering the suggestion, still very much alive, 2 that "to do something is to originate or to bring into existence, Le.,

really, to cause, some not yet existing state either of ourselves or of someone else, or, again, of some body" (p. 59). He acknowledges that this suggestion may be qualified in various ways - for example, by distinguish- ing the direct from the indirect production of a change - but he insists that it is nevertheless untenable. An action, he says, is an activity; and "though we think that some man in moving his hand, or that the sun in attracting the earth, causes a certain movement, we do not think that the man's or the sun's activity is or consists in causing the movement." (p. 60) To this he adds, " i f we ask ourselves: 'Is there such an activity as originating or causing a change in something else?', we have to answer that there is not." (ibid.)

The main thrust of Prichard's objection is clear: An action is an activity; there is no activity of causing or originating something; therefore, an action cannot be identified with the causing or originating of some change.

Philosophical Studies 25 (1974) 97-116. All Rights Reserved Copyright �9 1974 by D. Reidel Publishing Company, Dordrecht-Holland

98 B R U C E A U N E

What is not immediately dear is what an activity is supposed to be and why Prichard is so confident that there is no such thing as an activity of causing or originating something. He grants that we may speak of a subject causing a change, and he acknowledges that "there is such a thing" as causing something. Yet he insists that, though an episode of causing may occur and may even require an activity, the causing itself is not an activity and therefore not an action. What is his basis for these claims?

I cannot say exactly what Prichard meant by the word 'activity', though he no doubt restricted its application to events and occurrences that, in contrast to what might be called 'passivifies', are regarded as causes rather than effects. The principle behind his claim that causing is not an activity is easier to identify, however. It appears to be this: Statements of the form 'Thing or person S causes event E' are to be understood as elliptical for statements of the form '(3A) (.4 is an activity by S and A causes E)'. Thus, we may speak of a thing or person causing a change, but if we have our philosophical wits about us we shall mean that the thing or person 'does something' (that is, acts in a certain way) as the result of which the change occurs.

Prichard remarks that we often speak of a thing causing some result without knowing what change or activity of it is actually responsible for that result. When we say that a body caused a change "by a force of at- traction or repulsion, we are only expressing our knowledge that there is some activity at work, while being ignorant of what the kind of activity is." (p. 61) Statements of the form 'Thing S causes E' thus allude to activities only indirectly; they affirm that there is some activity or other of the subject that causes the relevant change to occur. To say, as Prichard does, that "there is such a thing" (but not such an activity) as causing something is only to say, therefore, that there are, or can be, complex states of affairs that involve activities and what results from them.

After rejecting the claim that an action, being an activity, can be identified with the causing or originating of something, Prichard proceeds to develop his own view. As I have indicated, he contends that a human action is an activity of willing some change. His contention is not meant to deny that 'John moves his hand' attributes to John a certain physical movement. His point is simply that John's activity in moving his hand - the action he thus performs - is accurately understood as an activity of willing his hand to move.

PRICHARD, ACTION, AND VOLITION 99

To understand the rationale for Prichard's peculiar view, consider the action statement 'X moves Y's hand'. This statement affirms that the subject X brings about (that is, causes) a certain movement - namely, the movement of Y's hand. According to Prichard's thesis about causation, this claim about X is equivalent to '(3A) (A is an activity by X and A causes the movement of Y's hand)'. Observe that, on this interpretation, the 'moves' in 'X moves Y's hand' does not itself connote a physical movement. Of course, if X moves Y's hand by pushing it with some part of his body, then X will have caused it to move by making a physical movement. But X need not, logically speaking, have proceeded this way; as Pilchard points out, it is at least conceivable that X moves it by some kind of psychokinetic action. Although such a 'moving' may be no more than a bare logical possibility, 8 it allows us to see that the verb 'moves' has fundamentally different meanings in its intransitive and transitive uses. In 'Y's hand moves' it connotes a motion of Y's hand, but in 'X moves Y's hand' it alludes to an event or activity that causes Y's hand to move. Prichard's contention is that when 'moves' connotes an action, it is used in the latter sense: it means 'does something that causes a movement'.

Prichard's odd sounding view of action follows naturally from these considerations. The statement 'John moves his hand' is equivalent to 'John moves John's hand'. Given Prichard's principles, we must analyze 'John moves John's hand' as '(3A) (A is an activity by John and A causes the motion of John's hand)'. The relevant activity here may, of course, involve a movement, since John might move his hand by pushing it with his foot. But if this action is voluntary, he will probably (though not necessarily) move his foot without making some other movement that causes his foot to move. 4 In this case the moving of his foot (in the action sense) will not itself be a movement of his body. What activity, then, will cause his foot to move? Prichard says "it almost goes without saying... [that it is] a mental activity of a certain kind, an activity of whose nature we are dimly aware in doing the action and of which we can become more clearly aware by reflecting on it .... If we ask 'What is the word for this special kind of [mental] activity?' the answer, it seems, has to be 'willing'. (p. 61)

Although willing is an activity, it cannot, for Prichards, be directly brought about by further willing. What we will, he says, is always "some change of state of some thing or person" (p. 64), and the activity of will-

100 BRUCE AUNE

ing is not, in his view, a change of this kind (see p. 69). He grants that willing can result from a desire to will something, but the denies that it can result from a further act of willing. To speak of willing as he does is not, therefore, to invite the kind of vicious regress Ryle speaks of when criticizing 'The Myth of Volitions' in The Concept of Mind. 5

As is well known, Prichard maintains that the action a man performs in moving his hand could also be performed i f , owing to paralysis, his hand failed to move. This suggests that the action statement 'John moved his hand' asserts the same thing as 'John willed his hand to move'. This suggestion cannot be correct, however, for 'John moved his hand' implies that his hand moved whereas 'John willed his hand to move' does not. Prichard was aware of this difficulty and said that John's activity of willing his hand to move would not 'ordinarily' be called an action if his hand did not move. He insisted, nevertheless, that John's activity of willing is still 'of the same sort' as what we ordinarily call an action (see p. 65). This claim requires elaboration.

According to Prichard, the statement 'John moved his hand' asserts that John performed a certain action - namely, that of moving his hand in the activity sense of the word. Assuming that the resulting movement of his hand was not brought about by some other movement he made, we may say that the action he performed in moving his hand was that of willing his hand to move. But in addition to saying that John performed such an action, the statement 'John moved his hand' also implies that his hand moved as the result of his action. I f we take 'John moved his hand' as a typical action statement, we must therefore acknowledge that such statements do more than merely report the performance of Prichardian actions; they also imply that such actions have certain results or upshots - in particular, the results that the agent willed to occur.

As I have expounded it, Prichard's theory of action is open to attack on a number of counts. Some philosophers would immediately reject his view that statements affirming that a subject S brings about an event E are reducible to statements of the form '(3A) (A is an activity by S and A causes E)'. In other words, they would reject his view that so-called agent causation is reducible to what we may call 'event causation' n. They may well grant that, when the agent in question is an inanimate object, the agent causation is reducible to event causation; but they would insist that the same kind of reduction is not possible when the agent is a person.


It seems to me, however, that Prichard is clearly correct in this matter. I f we say that John voluntarily moved his hand at time t, we may indeed say that he causes his hand to move at th~tt time. But to say this is not to say that he was the irreducible cause of his hand's motion; if he were, then, since he existed yesterday, his hand should have moved the same way yesterday too. After all, if A is the irreducible cause of B, then whenever we have A, we should have B as well. Clearly, if John did move his hand at a certain time, then there must have been something about him at the time that accounts for the occurrence, then, of his hand's movement. In this respect the case of John does not differ from that in which we say that an empty car knocked over a lamppost: the car, like the man, must 'do' something (in this case, strike the post) that causes the result to occur.

When we think of a man's action, we do not, of course, think of a mere response to some physiological change. We think of a man as a conscious agent, and we regard the movements he voluntarily makes as due to his mental activity. Traditionally, the kind of mental activity considered the cause of a man's voluntary movements is called 'willing'. This use of the verb ' to will' is not just a philosopher's technicality, for it has been familiar in English since before the time of Shakespeare. Unless the traditional notion of willing can be shown to be confused or objectionable in some fundamental way, the idea that our voluntary movements are willed into being should be regarded as an obvious presupposition of the ordinary concept of action.

It is well known, of course, that philosophers in our time have offered many reasons for rejecting the notion of willing as confused, mythical, or otherwise unacceptable. 7 Yet it seems to me, after surveying the literature on the topic, that these reasons are, without exception, untenable. I shall comment briefly on some of them toward the end of this paper; for the moment I shall merely say that, when the concept of willing is appro- priately clarified, it is easily seen to be far less objectionable than is commonly supposed.

As I have developed it thus far, Prichard's theory applies most directly to actions like moving one's hand. But not all cases in which we say that a man moves his hand concern actions. When a sleeping man 'moves' his hand, he does not act in the sense described by Prichard, for he does not will anything. Although Prichard sometimes gives the impression that we can bring about a movement (in the sense of acting) only by willing that

102 BRUCE AUNE

movement to occur, he makes it clear in at least one passage that we may will one movement but actually bring about another (see p. 64). This admission is important, because there are hosts of cases in which, though we are awake and alert, we bring about changes that we do not will to occur. Such cases are different from that of the sleeping man, for the changes brought about are somehow due to our activity as conscious agents. In a recent article Donald Davidson has described these cases by saying that the person bringing about the changes is 'the agent' of their occurrence. 8 Although Prichard has little to say about such cases, his theory seems to commit him to the following definition:

(1) S brings about E in the sense of being the agent of E's occurrence just in case E's occurrence is the causal consequence of S's willing something.

Consider the example of John who, as we should say, knocks over a teacup unintentionally. According to Prichard, John can be said to have acted only if he exercised his volition in some way. Assume that it was a movement of John's body that caused the teacup to be knocked over. I f this movement was willed by him, we can say:

(2) (3V) (3M) (Vis an activity of willing by John & M i s a moev- ment of John's body & Vcaused M & M caused the teacup to

be knocked over).

Assuming that the causing here is transitive, we may also affirm:

(3) (311) (V is an activity of willing by John & V caused the teacup to be knocked over).

But (3), according to Prichard's theory, is equivalent to

(4) John brought about the knocking over of the teacup.

Given that it was a movement of John's body that caused the teacup to be knocked over, we can use (4) to infer that John was the agent of the teacup's being knocked over (that is, he knocked it over) whether he did so intentionally or not.9

Anyone familiar with Davidson's theory of agency will see that it is similar, in fundamental respects, to the view I am attributing to Prichard. According to Davidson, S is the agent of an event E just in case there is a


description D of E such that r- brought about D intentionally "1 is true of S. 1~ To see the similarity of this view to Prichard's we may note that, for Davidson, the following descriptions might apply to the same event: 11

(5) (6) (7)

John's knocking over the teacup John's lifting the platter of cookies John's surprising his wife.

Davidson would regard the events described here as movements by John; and if only one movement were involved in all three cases, he would say that (5), (6), and (7) have the same reference, respectively, as the following:

(8)

(9)

(10)

(~M) (M is a movement by John & M causes the knocking over of the teacup) (~M) (Mis a movement by John & M causes the raising of the platter of cookies) (TM) (Mis a movement by John & M causes John's wife to be surprised).

On the assumption that just one movement by John is involved in (8), (9), and (10), we may infer that the events denoted by (5), (6), and (7) are identical - that is, one and the same event. Given that

(11) John's lifting the platter of cookies was intentional,

Davidson's theory allows us to infer that the events denoted by 'Johns' knocking over the teacup' and 'John's surprising his wife' are actions John performed.

The similarity between Davidson's and Prichard's views appears even closer if Davidson's remarks on intentional action are taken into account, According to his 'Actions, Reasons, and Causes', statements of the form 'S did A intentionally' appear to be equivalent to corresponding statements of the f o r m ' S did A as the result of a primary reason R'. 12 Since a primary reason, for Davidson, is a complex of belief and pro attitude, we should be able to say that, for him, S is the agent of an event E if and only if E is caused by the joint occurrence in S of some belief and pro attitude concerning E. The events immediately produced by such reasons - i.e., the events explicitly alluded to by the primary reasons that cause them - are, for Davidson, 'mere movements' of the agent's body; and these, Davidson

104 BRUCE ALINE

says, "are all the actions there are. We never do more than move our bodies: the rest is up to nature. ''18 As we have seen, Prichard's view is more austere than this. For him we merely will to move our bodies: even our physical movements are up to nature.

To avoid hasty objections to these theories we should emphasize that such strong claims by Davidson and Prichard are hyperbolic. Obviously, both men would agree that we can do a lot more than merely move or bodies or will them to move: we can knock over teacups, start wars, and proceed to sink battleships. Yet our actions in doing such ambitious things are analyzable, they will insist, into doing more elementary things that have the falling over of teacups, the onset of wars, and the sinking of battleships as their more or less distant consequences. The claim that actions described in causal terms are analyzable into more elementary activities (or events) that have certain consequences is not, in my opinion, bizarre; and neither theory is obviously mistaken when its acutal com- mitments are taken into account.

One might expect that Davidson's theory, being the more recent of the two, is bound to be preferable. But the advantages seem to lie with Prichard. Davidson identifies and action with a movement, but this maneuver is questionable. As Prichard insisted, although we commonly do bring about a movement when we act, the movement brought about does not seem to be identical with our action; it seems to be its result or upshot. This view is supported by examples of actions where no movement is even intended. If a man deliberately blocks a doorway by not moving away from where he is standing, there need be no movement - not even an unintended one, such as a tensing of certain muscles - with which his action could plausibly be identified. His activity, his doing, seems to be that of willing to remain where he is. It is true that the notion of willing is not well understood, but then Davidson's view has its own obscurities. In explicating the concept of agency Davidson employs the familiar but really obscure notion of believing that an event has a property 'under a description'. Prichard's view lacks this feature; and if the notion of willing can be clarified satisfactorily, then his theory might be preferred, other things being equal, on the grounds of simplicity. 14

The last part of Prichard's paper is concerned with the relation between action and desire. Prichard maintains that willing a movement results (at least in part) from, and could not occur without, a certain desire (p. 65).

]PRICHARD, ACTION, AND VOLITION 105

Concerning the nature of this desire, two general views are (he says) possible: (a) that it is a desire for what is willed, namely a certain movement; or (b) that it is a desire for the willing of the movement. Prichard accepts the last alternative on the ground, among others, that when a man desires to sink a putt, he desires to be the agent of a certain result (pp. 66f). The mere belief that the ball goes into the cup, or that the putt is sunk (by someone), need not satisfy his desire; it is certain to be satisfied only if he believes that he himself was the agent of the ball's going into the cup (in an appropriate way).

It seems to me that Prichard is clearly wrong in adopting this second alternative. I f I desire to sink a putt, I desire to be the agent of a certain result; but surely the object of this desire includes the physical result just as much as it includes my willing that result. What I desire, in other words, is that a certain state of affairs obtain - namely, my willing a certain result and that result coming to pass as a consequence of my willing. Prichard apparently did not notice this shortcoming of his view, because he identified doing A with willing A. As we have seen, however, although 'S moved his hand' and 'S willed the movement of his hand' may both attribute an activity of willing to S, the former (like all action statements) also implies that the agent's activity had a certain upshot. When, therefore, I desire to do a thing, typically I desire to succeed in doing it; if I desire to move my hand, I desire that my hand move as the result of my moving it.

If we amend Prichard's view along these lines, a possible interpretation of'bringing about E intentionally' comes to mind:

(12) S brings about E intentionally iff S brings about E because of his desire to do so.

This interpretation is not suggested by anything Prichard says, but it is close to the view propounded by Davidson in 'Actions, Reasons, and Causes'. According to him, "the expression 'the intention with which James went to church' has the outward form of a description, but in fact it is syncategorematic and cannot be taken to refer to an entity, state, disposition, or event." Its function, he says, is "to generate new descriptions of actions in terms of their reasons." (p. 84) He also says, "To know a primary reason why someone acted as he did is to know the intention with which his action was done." (p. 83) These remarks suggest t h a t ' S brought about E intentionally' can be interpreted as 'S brought about E

106 B R U C E A U N E

because of some primary reason for doing so." If we recall that, for Davidson, a primary reason is a complex of belief and pro attitude (desire being a pro attitude) then the similarity of this view to (12) is un- mistakable. In fact, since anyone defending (12) would no doubt allow that a certain desire is likely to lead to action only when accompanied by some be l i e f - for example, that the conditions are appropriate for satis- fying the desire - we can amend (12) so that it comes even closer to the view suggested by Davidson's remarks:

(13) S brings about E intentionally iff S beings about E because of his desire to do so together with some belief.

The question we must now consider is whether this interpretation of 'S brings about E intentionally' is really satisfactory. In my view it is not; it seems fundamentally defective as an analysis or explication of the locution is question. There are a number of reasons for this.

For one thing, taken as an analysis or explication, (13) seems to imply that the expression 'S's intention in bringing about E ' is, in Davidson's words, syncategorematic and does not refer to any state, event, or disposition. But this seems false. On the face of it, at least, there is as little reason to deny that intentions are genuine mental states as there is to deny that beliefs or desires are genuine mental states. Davidson, as far as I know, has offered no explicit defense for his reading of 'S's intention in doing such and such' and in the absence of such a defense his reading should be regarded as prima facie implausible. This implausibility gains support from Austin's observation that the word 'intention' is closely related to 'choice' and 'decision', 15 both of which clearly apply to mental episodes invoved in deliberation. In deciding what to do one is faced with alternatives between which one must choose. If, on reflection, one chooses to do A rather than B, one has made a decision. But to decide to do A is, among other things, to form the intention of doing A, at least in typical cases. Any theory that denies, or even casts doubt on, the existence of such familiar mental states or episodes as choosing, deciding, or intending is, to my mind, extremely dubious.

A more direct difficulty with (13) as an analysis or explication is that a belief-desire (or belief-pro attitude) complex may elicit responses that are not intentional under any description. Intuitively, to do something intentionally is to mean to do it - to carry out, as it were, a plan, idea, objective,


or even resolution. 16 But a desire, passion, or other pro attitude may conceivably be so strong that a man loses control of himself (as people used to say) and 'does' things that he does not mean to do. Again, one could conceivably give in to a passing desire, whim, craving, or yen - also without really meaning to do so. In such cases one need not have lost control; ane may simply not have exercised it. If desires and Davidson's pro atti- tudes are not restricted to intentions, then the fact that they, together with a belief or group of beliefs, may prompt a man to respond in a certain way will not be sufficient to insure that his response has the deliberate, meant character of an intentional response - that his response is something he really intended to produce. For this reason, it seems to me that (13) and what I take to be Davidson's analysis are inadequate treatments of intentional behavior.

If I am right in what I have been arguing, the most plausible analysis of 'S brings about E intentionally' is no doubt the more natural 'S brings about E because he intends to do so'. To clarify this natural analysis, one must, of course, clarify the concept of an intention. Exactly what is distinctive of such a mental state? My answer, in brief, is that intending is very much like believing: it is a dispositional state. Unlike a belief, however, an intention has a particularly close relation to action. In general, if one really intends to do A in circumstances C and believes truly that one is in C, then, given that one has no conflcting intentions and that one can do what one intends to do, one will do A in C.

I have argued elsewhere that believing that snow is white is essentially a disposition to use the premises 'Snow is white' in one's reasoning. 17 Although I cannot argue the point here, I think that intending to do A in C can be understood, in part, as a disposition to use the premises 'I will do A in C' in one's reasoning. Such a premiss provides the tie between intention and action, for it is the thought 'I will do A in C' that leads the agent to do, or try to do, A when he believes he is in C. This transition from intention and belief to action may be mediated, in the case of conditional intentions, by a practical conclusion. I f one has the intention 'I will do A if p ' and also believes that p, then one may conclude 'I will do A now'. This latter thought, which expresses a categorical intention, will generally result (if the agent can do what he intends to do) in the action A. Much more can, of course, be said about intention - in particular, its complicated role in practical reasoning requires careful discussion TM

108 BRUCE AUNE

- but the few remarks I have just made will be sufficient for the purposes of this paper.

This brings me to the topic of volition. I f the intention 'I will now do A' leads one to act, and if Prichard is right that acting consists in willing some movement, then 'I will now do A' must be very closely related to volition. How is this relation to be understood ?

Prichard himself offers no help in the matter, for his account of volition is extremely unsatisfactory. After remarking that willing is "a mental activity.., o f whose nature we... [are] dimly aware in doing... [an] action and of which we can become more clearly aware by reflecting on i t" (p. 61), he has to admit that its "nature is sui generis and so incapable o f being defined, i.e., of having its nature expressed in terms of the nature of other things." Yet if the nature of a volition cannot be clarified by some kind of verbal explanation, there is little hope that the relation between intention and volition can be clarified either.

Fortunately, an account of volition has been developed recently that seems tailor-made for our purposes. It is due to Wilfrid Sellars. 19 As I interpret him, a volition is a practical thought of the form 'I will do A here and now'. Such a thought might be called an occurrent intention, for it is the expression, in thought, of a restricted kind of categorical intention. Not all categorical intentions concern occurrences one can immediately bring about, but the target of a volition is something one believes is in one's power here and now. Sellars calls the behavioral upshot of a volition a 'minimal act'. Such acts need not be simple; in fact, we can be trained to perform highly complicated minimal acts. If a skilled musician can perform a cadenza without willing to play its individual notes, then playing that cadenza is, for him, a minimal act.

The fundamental relation of willing to minimal action can be further clarified as follows. It is a truism that we must learn to think in terms of a conceptual scheme: we must learn to think of ourselves as doing this or that, as having done this or that, or as being about to do this or that. To think that we will do something minimal requires some tendency to do what we think we will do; it is only if we have this tendency that we shall be thinking in a volitional way. The words 'I will now do A' may run through our minds, but if they do not carry with them the appropriate behavioral tendency, they will not constitute a volitional thought, which is to say an 'act' of willing. The relation between willing and appropriate


behavior is thus causal, but the concept of willing has an intimate logical relation to the concept of such behavior. This logical relation can be exhibited by the conceptual (or analytic) truth that if an agent has the ability and the opportunity to do something A, then if he wills to do A (and does not immediately will to cancel this volition), he will do A.

It is not difficult to see that Sellar's conception of volition avoids the standard objections to acts of willing that have been developed in the literature since Ryle's The Concept of Mind. Like Prichard, he denies that willing is a voluntary activity in the sense that it can be directly brought about by further willing. 20 As an act of practical thinking, it is, moreover, nothing like the 'inner thrusts' Ryle made fun of; in fact, it is no more mysterious than the thoughts one has in playing chess, many of which are expressions of conditional intentions like 'If he attacks with his Queen, I will castle'. Since we conceive of thoughts in terms of their formal structure rather than in terms of qualitative features they invariably possess (to think that snow is white is not necessarily to experience verbal or other imagery), it is no surprise that philosophers have failed to detect distinctive 'impressions' whenever they do something voluntarily. 21

To avoid misunderstanding two facts about thoughts should be em- phasized. First, the formal structure of a thought concerns its semantic features: the logical powers, the 'sense and reference', it possesses. Such features are not qualitative or empirical; they might be called 'functional'. Consequently, it is entirely possible that the empirical features of the episodes possessing the formal structure of both indicative thoughts and volitions may be describable in the language of neurophysiology. Second, our ability to 'introspect' our thoughts - to have noninferential knowledge of what we are thinking - should not be understood on the analogy of sense-perception. We do not know what we are thinking because we look inwardly and 'see' what thoughts are floating through our minds. Our ability to avow our thoughts with a high degree of accuracy may be difficult to explain in common sense terms, but we should not expects to provide illumination by postulating some peculiar eye of the soul. The fact that we cannot form images of our thoughts, or discern the material features they possess, does not warrant the conclusion that we do not think at all. The same holds true for those volitional thoughts called 'acts' of willing.

110 BRUCE AUNE

Although Sellars' conception of willing is, at least in my view, extremely attractive, it is not entirely clear that it can be brought into line with Prichard's view of action. Prichard insists that 'what we will' are physical movements; in fact, he takes special pains to attack the view that what we will are actions (see pp. 62ff). But consider a volitional thought like 'I will now raise my arm', which Sellars regards as the willing of something. Such a thought certainly seems to concern an action rather than a mere bodily movement. If we must interpret it as the willing of an action then Pilchard would have to deny that it could be a proper volition. Is there any way of resolving this apparent conflict between Prichard and Sellars?

As a first step in answering this question, we may observe that the meaning of 'what we will' is not entirely clear. On the face of it, at least, it could mean either of two things: it could mean (a) what we directly bring about by willing or (b) what we aim or intends to realize in willing. It is clear that, if Prichard had sense (a) in mind, his thesis would not conflict with what seems to be implied by Sellar's account, for Sellars could perfectly well agree that what we bring about by willing is, at least among other things, a physical movement. Unfortunately, Prichard makes it clear that he does not have this sense in mind. In criticizing James and Stout he says that they seem to treat the distinction between willing and what is willed as the distinction between willing and its effect, "though they are totally different." (p. 63)

In saying that what we will is a movement Prichard thus seems to mean that our aim or intention in willing is always that some movement or change occur. But why does he hold this view? The answer is that the alternative and more natural view that our aim is to act in a certain way involves, in his opinion, a vicious regress. Thus, if an action is, as he claims, the willing of something, then the willing of an action "must in turn really be the willing of the willing of something else, and so on." (p. 64) This argument and the related arguments he gives are invalid, however, for 'the willing of something' is an opaque context given interpretation (b) of 'what we will'. Therefore, the premises 'we will an action' and 'an action is the willing of something' do not warrant the conclusion 'we will the willing of something'.

Although his argument is invalid, its spirit can be saved by an amend- ment. If we accept Prichard's thesis about action, we shall use the ex-

PRICHARD, ACTION, AND VOLITION l l l

pression 'I will an action' to mean 'I will the willing of something'. If, further, we accept the view that what we will is always an action, we shall have to acknowledge that the 'something' in 'I will the willing of something' refers, when we use it, to an action. Consequently, when we say 'I will the willing of something', we shall be committed to saying 'I will the willing of an action'. But this latter statement will mean, for us, 'I will the willing of the willing of something'. This last claim can in turn be expanded, and we shall be on our way to an infinite regress. Since the regress is generated by the thesis that we always will some action, it is thus reasonable to reject that thesis.

In spite of this argument it is difficult to deny that we generally will to do this or that. Since the thesis that we will an action seems to be a technical way of saying that, when we will, we will to do something, it is natural to seek some means of avoiding Prichard's conclusion. A tempting strategy is to identify the action of doing A with the 'action' of bringing about some result (some movement or change) M. Given this identifica- tion, we could then argue t h a t ' S wills to do A' simply means 'S wills to bring about M'. I f we interpret Prichard's claim that what we will is a movement as meaning 'what we will to bring about is a movement', his thesis will be equivalent to the natural view that, when we will, we will to do something.

Tempting as this line of thought may initially be, it cannot be acceptable to Prichard. I f M is a minimal movement - that is, one under our imme- diate voluntary control - then the activity by us that results in M is an activity of willing. For Prichard, there is, as we may recall, no activity of bringing something about: the locution 'S brings about M ' is simply short for 'S does something that results in M' . Given this, the claim that S wills to bring about M can only mean that S wills to do something that results in M. But if M is minimal, the only activity that can have M as a result is a willing. Therefore, the claim that we might will to do something that has M as a result can only mean that we might will to perform or execute a willing that has M as a result. But Prichard declares that we cannot will to will (see above, p. 5).

Since acting, for Prichard, is willing, there is no escaping the conclusion that he must disallow the possibility of our willing to act in the full sense of the word 'act'. I f we deliberately move our finger we will our finger's movement, or we will that it move - but we do not will to bring about its

112 BRUCE AUNE

movement. We may say, it is true, that we bring about its movement, but saying this is simply short for saying that we 'do something' (in this case, will something) as the result of which our finger moves. To will to bring something about is to will to act, and Prichard cannot allow that this is possible.

Following Sellars, I have interpreted an act of willing as a volitional thought of the form 'I will now do such and such'. In view of Prichard's denial that we can will to act it might appear that there is no way of reconciling his theory with my interpretation of willing. This is not so, however, for there is a sense in which we can will to do something without thereby willing to act - at least in Prichard's sense of 'act'. If such a non-action sense of 'do' is appropriate in the con tex t 'S wills to do such and such', then my conception of willing is perfectly consistent with Prichard's theory.

As we saw in introducing the topic of agency, action verbs like 'move' are not always used to ascribe actions to agents. I f a sleeping man's arm moves spontaneously, it is permissible (even natural) to say that he moves his arm - even though it is clear that the movement is not voluntary. Similarly, when a man is the victim of a coughing spasm, it is natural to say that he coughs or even that he does something (though not voluntarily). In the case of such nonvoluntary 'doings', we are using typical action language to connote movements, changes, or mere responses. The same is true, Prichard would have to contend, of the world 'do' in the con t ex t 'S wills to do A'" S's aim here is that certain movements occur - those movements, namely, whose occurrence is implied by the statement that he has done A. (Recall t h a t ' S moves his hand' implies S"s hand moves'.)

Sellars, who claims that we can and do will to act, uses 'action' in something like this weak sense. He says "A voluntary action in the primary sense is one that is caused by an act of will. ''2~ If we ask what is caused by S's willing his arm to move, the answer must be 'The movement of his arm' - and this would seem to be an action in the relevant sense. Where Sellars and (we may recall) Davidson use 'action', Coiling- wood and others use 'deed'. Collingwood used 'deed' because the word 'action' is, as he claimed, "often by ancient and respectable usage em- ployed for the will as opposed to the deed." As examples of this usage, he cited the acts of parliament, the acta of any committee or deliberative


body, the use of 'act' in devotional literature, and the use of 'action' in well-known passages from Shakespeare. 93

Although the decision to speak of the upshots of volitions as actions, deeds, or mere movements may seem to be a mere matter of nomenclature having little if any philosophical interest, Prichard held strong views to the contrary. In fact, he made a special point of criticizing a Professor Macmurray, who claimed that the word 'action' is really ambiguous and may properly refer either to what is done, which he called 'the deed', or to the doing of it, which he called 'the action'. In opposition to this, Prichard flatly asserted, "Obviously... there is no ambiguity whatever." He continued:

When I move my hand, the movement of my hand, though an effect of my action, is not itself an action, and no one who considered the matter would say it was, any more than he would say that the death of Caesar, as distinct from his murder, was an action or even part of an action (p. 63).

For my part, the question whether to use the word 'action' for the willing or the result of the willing is not very important in the context of a philosophical theory. What is important are the distinctions drawn and the illumination provided. Prichard's use of 'action', which is to a degree technical, is based on an analysis of cases in which a man's volitional causality (as it might be called) is exercised. Such causality is implied by certain uses of transitive verbs like 'moves', 'hits', or 'murders'. I f a man moves his arm in the way that is distinctive of a rational agent, then he brings about a certain movement: the movement results from his volitional activity. For reasons we have discussed, Prichard contends that there is no activity of bringing something about: to say that S brings about E is only to say that E results from some activity of S's. This provides a reasonably clear, straightforward analysis of what most philosophers would regard as action-ascribing uses of 'S moves his arm', 'S kills Jones', and the like.

There are, of course, as most philosophers would acknowledge, non- action uses of 'S moves his arm' - for example, 'S moved his arm while asleep'. How are we to understand such locutions? Clearly, no conscious agency is implied. We seem to be presented merely with the movement of a body - a bodily movement. In saying 'He moved his arm' in such cases we are certainly not ascribing responsibility or whatnot to the agent, for there is no question of his being responsible (generally speaking) for

114 B R U C E AUNE

the movements of his body while he is asleep. A Prichardian would no doubt contend that we are using action language here largely as a result of habit. We are presented with the kind of movement that occurs when a man voluntarily moves his arm; and although we believe that the man is not voluntarily making that movement, we describe him as if he were. The import of what we are saying, when we give such a description, is that the man's body moved in the way it would move if he voluntarily moved it. This non-action use of locutions like 'He moved his arm' is thus logically secondary; it makes sense only in relation to talk about actions proper.

Although it is possible to call the movements or changes resulting from our volitional activity 'actions', this practice, according to the Pri- chardian, is unnecessarily confusing. An action is normally understood as a doing, which is technically an activity. Yet the movement of an arm clearly seems to be the result of an activity, the activity of willing. Why call such results 'actions'? Why adopt a view having the consequence that the movement of Jones' arm might be Smith's action, as in a case of psychokinesis? Indeed, if Brutus could kill Caesar merely by willing him to die, which is certainly 'logically' possible, then Brutus' act of killing Caesar might be identical with Caesar's dying. But surely an act of killing could never be identified with an 'act' of dying. Prichard's way of speaking is certainly preferable to this!

As I said, however, the important thing about a theory of action is the distinctions it draws and the philosophical illumination it provides. I f anyone objects strenuously to Prichard's technical use of 'action', he can substitute 'volitional activity' for it, and speak of actions where Prichard would speak of the results of actions. This would involve no significant change in Prichard's theory, though I think it would make it less intuitive. If, however, we employ Prichard's terminology, we must realize that, strictly speaking, there can be no voluntary or involuntary actions. What are properly voluntary are movements, or doings in the secondary sense. For a Prichardian, a movement is voluntarily brought about if it is caused by an act of willing; a kind of action is voluntary, or under our voluntary control, if it can be so brought about. It seems to me that this terminology is not really out of line with ordinary usage; it is likely to sound odd only to a philosopher.

My aim in this essay has been to provide a critical analysis of Prichard's


views on act ion and to relate his views to current th inking on the subject.

I have contended that, in spite of his controversial and possibly mis-

leading language, Prichard 's major claims abou t action were essentially

correct. The chief defect of his posi t ion was a totally inadequate concep-

t ion of willing, bu t I have indicated at length how this can be repaired.

He did no t offer an analysis of the concept of in tent ional movemen t and

his views on what we desire do not lend themselves to a sound analysis

of tha t concept. Yet the general structure of his th inking makes it relatively

easy to in t roduce a plausible analysis that rounds out a very penetrat ing

concept ion of h u m a n action. ~4

University o f Massachusetts

at Amherst

NOTES

1 H. A. Prichard, 'Acting, Willing ,and Desiring', in Moral Obligation, Oxford 1945, pp. 89-98. Reprinted in A. R. White, The Philosophy of Action, Oxford 1968, pp. 59-69. My references are to this reprint.

See White, p. 2. 3 According to a recent article, 'Parapsychology in the U.S.S.R.', Saturday Review (March 18, 1972), feats of psychokinesis have been carried out experimentally at the Institute for Technical Parapsychology in Moscow. The article includes a photograph of"the noted Russian 'sensitive' Nina Kulagina... [causing] a plastic sphere seemingly to float in the air...." We are told that "Mrs. Kulagina's other psychokinetic ac- complishments supposedly include separating the yolk from the white of an egg, moving a pitcher filled with water across a table, and stopping the heartbeat of a frog" (p. 36). 4 Here I am ignoring the various physiological 'movements' that must occur if the man's foot is to move. 5 See Gilbert Ryle, The Concept of Mind, London 1949, pp. 62-69. 6 See e.g., R. M. Chisholm, 'Freedom and Action', in Freedom and Determinism (ed. by Keith Lehrer), New York 1966, pp. 28-44.

Wilfrid Sellars has given clear, brief replies to a number of standard objections to volitions in 'Metaphysics and the Concept of a Person', in The Logical Way of Doing Things (ed. by Karel Lamber0, New Haven 1969, pp. 243-246. 8 See Donald Davidson, 'Agency', in Agent, Action, and Reason (ed. by Robert Binkley et al.), Toronto, 1971, pp. 2-25. 9 If the movement of John's arm startles Harry so that he strikes the teacup, John will still have knocked it over in the sense of being the agent of its being knocked over; Harry will have knocked it over in the non-agency sense in which a stick swung by a man could be said to knock something over. If, however, John's movement leads Harry to will some movement that results in its being knocked over, then John will not knock it over, for he will not be the agent of the result: Harry will be that. We might say that the relation of agency is not transitive through persons.

116 BRUCE AUNE

lo Ibid., p. 7. The formulation of Davidson's view is mine; it differs from his more informal statement in what I hope is an unimportant respect. 11 Ibid., pp. 21-25. 12 See Davidson, 'Actions, Reasons, and Causes', Journal of Philosophy 60 (1963), 685-700; reprinted in White, pp. 72-92. My references are to White. 18 See 'Agency', p. 23. 14 A kind of difficulty recently brought against Davidson's theory is illustrated by the following: Jones kills Smith by shooting him. To shoot him Jones pulls the trigger of a gun. Since Jones makes just one movement that (a) causes the trigger to be pulled and (b) causes Smith to be killed, his action of pulling the trigger ought to be identical, on Davidson's theory, with his act of killing Smith. But although Jones' act of pulling the trigger causes the gun to go off, his act of killing Smith did not (it is argued) cause this. Hence, it is a mistake for Davidson to identify these two acts of Jones.

The correct response to this objection is that Jones' particular act, which may be called a killing of Smith because of a relational property it possesses, may perfectly well be the cause of the gun's going off. It might sound odd to say such a thing, but this is because we tend to confuse a causal relation between events with an explanatory relation between propositions. The fact that the gun went off is not causally explained, obviously, by the fact that Jones killed Smith. Yet the action Jones performed in bringing about Smith's death, which we might call 'His killing of Smith' may still have caused the gun to go off.

Other objections to Davidson's theory, which I believe can also be met, are given by Alvin Goldman in A Theory of Haman Action, Englewood Cliffs 1970, Ch. 1. 15 See J. L. Austin, 'Three Ways of Spilling Ink', The Philosophical Review 75 (1966), 437. 16 See ibid. 17 It is commonly held that a belief is a complex behavioral disposition, but such a view is fundamentally mistaken. There is no overt behavior distinctive of a given belief - not even the tendency to avow it. Belief does affect behavior, but only via practical reasoning in which the proposition believed occurs as a premiss. For a thorough discussion of belief, see my Knowledge, Mind, and Nature, New York 1967, pp. 213-218. is I have discussed this subject in connection with a critical examination of Wilfrid Sellar's views on practical reasoning in 'Sellars on Practical Reason', forthcoming in a volume of essays on Sellars, edited by H. N. Castafieda. SeUars has expounded his views on the subject in Science and Metaphysics, London 1968, Ch. VII. 19 See esp. his 'Thought and Action', in Lehrer, pp. 105-129. s0 See 'Metaphysics and the Concept of a Person', pp. 243f. 31 For a thorough discussion of the claims in this and in the following paragraph, see my Knowledge, Mind, and Nature, Ch. 7. 22 See 'Fatalism and Determinism', in Lehrer, p. 159. z3 See R. G. Collingwood, The New Leviathan, Oxford 1942, pp. 97f. ~4 For helpful comments on this paper, I am indebted to Gary Matthews, Fred Feldman, Jeff Sicha, and Wilfrid Sellars.

prichard, action, and volition

Documents