25535419

7/30/2019 25535419

1/31

Aristotle on Action

Ursula Coope and Christopher Shields

IURSULA COOPE

ARISTOTLEON ACTION

When I raise my arm, what makes it the case that my arms going up is aninstance of my raising my arm? In this paper, I discuss Aristotles answerto this question. His view, I argue, is that my arms going up counts as myraising my arm just in case it is an exercise of a certain kind of causal

power of mine. I show that this view differs in an interesting way bothfrom the Davidsonian standard causal account of action and from ac-counts put forward by recent critics of Davidson, such as Hornsby, Alva-rez and Hyman.

I

Aristotle and Modern Philosophy of Action. Aristotle holds thatwhen I raise my arm, my action of raising my arm is the same eventas my arms going up.1 We can, then, ask of Aristotle a question thatis often posed in modern philosophy: when I raise my arm, whatmakes it the case that my arms going up is an instance of my raisingmy arm? His answer, I shall claim, is that my arms going up countsas my raising my arm just in case it is an exercise of a certain kind ofcausal power of mine.2

This combination of views distinguishes Aristotles position fromeach of two opposed lines of thought in modern philosophy of ac-tion. On the one hand, Aristotle would disagree with what is oftencalled the standard causal account of action, the view that myarms going up counts as an action of mine in virtue of being causedin some appropriate way by my beliefs and desires.3 On the otherhand, he would also disagree with those critics of the standard caus-al account who insist that the action of raising my arm cannot be

1Aristotle, Physics III.3. I defend this interpretation of this chapter in Coope (2004).2 It is an action in virtue of being an exercise of a certain kind of causal power. It is an action

7/30/2019 25535419

2/31

IURSULACOOPE110

the same event as my arms rising.4 In this essay, I shall explain whyAristotle would object to both of these modern positions and shalldiscuss to what extent his own view presents an attractive alterna-

tive.

Aristotles view might, at first sight, seem to have much in com-mon with the standard causal account. The standard causal ac-count, like Aristotles, identifies my action of moving my body withmy bodys movement. So according to both accounts, when I raisemy arm, my arms going up is same as the event of my raising myarm.5 Moreover, Aristotle, like the proponents of the standard ac-count, holds that human and animal actions must be caused, insome sense, by a combination of the agents desire and the agentsbeing in some cognitive relation to the object of this desire.6 Butthere the similarity ends. For according to the standard causal ac-count, what makes the movement the agent undergoes an action ofthe agents is the fact that it is caused in some appropriate way bythe agents beliefs and desires: my arms rising counts as my raisingmy arm just because it is caused by my beliefs and desires. In con-

trast, though Aristotle thinks that human actions are explained bydesires and beliefs, it is not because they are explicable in this waythat they count as actions. On Aristotles view, what makes a partic-ular change an action is the fact that it is the exercise of a certainkind of causal power: it is a causingof something. My arms goingup counts as an action of mine (that is, it counts as my raising myarm) only because it is the exercise of one of my causal powers.

This feature of Aristotles view allows him to escape a certain

powerful criticism that has been levelled against the standard causalaccount. The criticism is that the standard account fails to be an ac-count of action or agency at all. On the standard account, when Iraise my arm, all that happens is that my beliefs and desires causemy arm to go up. But in that case, so the criticism goes, I dont real-ly do anything: I am merely an arena in which some events (my be-

4 The objections I shall consider here are those raised by Alvarez and Hyman (1998) and byHornsby (2004a; 2004b)

5 Or at least, on Davidsons account, this is typically the case when I raise my arm. Davidsononly defends this view in relation to primitive actions (those actions we do not do by doing

thi l ) H h diff t t b t h t h h f l I i

7/30/2019 25535419

3/31

ARISTOTLEONACTION 111

liefs and desires) cause others (the movements of my body). Whatthe standard account fails to recognize (on this view) is that an ac-tion is an agents bringing something about. This is not somethingthat can be reduced to one events causing another. The standard ac-count assumes that the only sorts of things that can be causes areevents and states. Because of this, it cannot make sense of a personsdoing something. As Hornsby (2004a, p. 22) complains, on thestandard account of action, the fact that the person exercises a ca-pacity to bring something about is suppressed.

It should be obvious why Aristotle is not vulnerable to this criti-cism. He does not share the assumption that the only things that doany causal work are events and states. On the contrary, for Aristo-tle, what it is to be an action is to be an exercise of one of the agentscausal powers. However, from a modern perspective, it can seemvery puzzling that anyone could hold this view, while maintainingthat the action of moving Xis the very same event as the movementthat Xundergoes. How can my action of raising my arm be the ex-ercise of a causal power, if it is itself a movement my arm under-

goes?There are three different reasons why this is puzzling. The firstreason is that the views I have attributed to Aristotle seem, whentaken together, to have an implausible consequence. Alvarez andHyman have argued that if an action is the exercise of a causal pow-er of the agent it cannot be the same event as the movement that theagents body undergoes. Their argument is very simple. The event ofcausing a movement could not, they say, be the same event as the

movement caused:My raising my arm is my causing my arm to rise. Hence, if my raisingmy arm is an event, it is the same event as my causing my arm to rise.And hence, if my raising my arm and my arms rising are one and thesame event, then my causing my arm to rise and my arms rising areone and the same event. But it cannot be plausible that causing anevent to occur is not merely an event itself, but the very same event asthe event caused. (Alvarez and Hyman, 1998, p. 229)

This argument poses a challenge to Aristotle (at least as I have inter-d hi ) I h h i I h ib d hi

7/30/2019 25535419

4/31

IURSULACOOPE112

There is also a second reason why the position I have attributedto Aristotle might seem flawed. I have claimed that on Aristotlesview what makes an event an action is that it is an exercise of a cer-

tain kind of causal power: it is a causingof something. However,some modern philosophers maintain that causing is not itself anevent. This, for instance, was von Wrights (1963, pp. 356) reasonfor denying that actions are changes: An act, he said, is not achange in the world. But many acts may quite appropriately be de-scribed as the bringing about or effecting (at will) of a change.Alvarez and Hyman endorse this view that actions are not events.They suggest that the doctrine that actions are events should lose itsallure once one realizes that an action is the causing of an event: Itis, after all, far from obvious that a causing, like the event caused, isitself an event (Alvarez and Hyman, 1998, p. 229). Given that Aris-totle, like these modern philosophers, holds that an action is a caus-ing, what is his reason for maintaining that it is nevertheless a kindof event (a change)?

The third source of puzzlement is over what reason anyone could

have for identifying the action of moving ones body with the bodysmotion, if the reason is not a commitment to something like thestandard account. Hornsby (2004a, p. 20), who holds that my ac-tion of moving my body is an event that causes my bodys move-ment, asks rhetorically, Would anyone be inclined to think ofsomeones moving her foot as a foots movement unless they imag-ined that a persons activity could be dissolved into the goings-on ofstates and events? As we have seen, the answer to this is Yes: Aris-

totle would. But Hornsbys question helps to bring out how oddAristotles position seems from a modern perspective. We need toexplain why Aristotle thinks that the agents action of moving (orchanging) Xis the same event as Xs movement (or change), giventhat this view is not based on a commitment to the picture of causa-tion that Hornsby is criticizing.

As I shall explain, Aristotles discussion of agency provides uswith the materials to answer each of these points. To the charge that

he is committed to identifying a change with the action of causingthat change, he can reply that an action is the causing of a state,h h h i f h O hi i i f i

7/30/2019 25535419

5/31


objection that a causing is not a change, Aristotle can reply bypointing to certain characteristic features of change that are sharedby actions (or at least, are shared by those actions, such as building

a house and raising ones arm, that he identifies with changes). Fi-nally, we can extract, from Aristotles discussion of agency in Phys-ics III.3, an argument that the action of moving (or, in general,changing) Xis one and the same event as the movement (or change)that Xundergoes.

II

Causing and What Is Caused. The first of these objections is thatAristotle seems to be committed to identifying a change with the ac-tion of causing that change. To answer this on Aristotles behalf, weneed to know more about what it is, on his view, that we are caus-ing when we engage in an action of moving or changing something.A difficulty here is that the word causing has no obvious equiva-

lent in Greek. When I say that, on Aristotles view, an action is acausing of something, what I mean is that it is an exercise of a cer-tain kind of power (or potential): an active power rather than a pas-sive power. In exercising an active power, one brings somethingabout (building a house and moving somethingare exercises of ac-tive powers); in exercising a passive power one undergoes some-thing (getting hot and being moved are exercises of passivepowers).7

Aristotles view, I shall argue, is that the power that is exercised inan action ofmoving Xis a power to produce the endofXs move-ment: a power to produce a state, rather than a movement. In thissense, what I am causing when I move Xis the state that Xs move-

7 When I move myself (e.g. when I walk from one place to another), I exercise both an activepower (a power to bring something about) and a passive power (a power to be affected in acertain way). (As we shall see, Aristotles view is that one part of me exercises the activepower and another exercises the passive power.) The fact that actions like walking fromplace to place involve the exercise of an active power as well as a passive power explains, I

think, why Aristotle describes them as self-changes. Waterlow (1982, p. 216) objects thatordinary discourse applies the language of self-change to movements and actions that

t ll t t t i li ti t t th hi h t h th

7/30/2019 25535419

6/31

IURSULACOOPE114

ment is directed towards.8 For example, when I raise my arm, whatI am causing is my arms being up, rather than my arms going up.More generally, the action of changing something towards being Fis, for Aristotle, a particular kind of causing of the state being F.

Aristotle never endorses this view of action explicitly. However, Ishall claim that if we attribute this view to him, we can explainwhat would otherwise be a puzzling argument in Physics VIII.5. Theargument concerns self-movement. Aristotle claims that somethingthat moves itself must have distinct agent and patient parts: the onlyway it is possible for something to change itself is for one part of itto be the agent of the change and another part of it to undergo thechange. As we shall see, his argument depends crucially on the viewthat when an agent changes something, what it is producing orcausing is the state the change is directed towards, rather than thechange itself.

Here is the argument:

Movement is an incomplete actuality of the moveable. But the moveris already in actuality, e.g. it is the hot thing that produces heat, and in

general, that which produces the form possesses it. So that the samething in respect of the same thing will be at the same time both hotand not hot Therefore, when a thing moves itself, one part of it isthe mover and another part is moved.(Physics VIII.5.257b813)

Aristotle seems to be reasoning as follows:

(i) If something is changing (or moving) towards being F, thenit is not (yet) F.

(ii) What changes (or moves) something towards being Fmustitself be F(e.g. something that heats must be hot).9

So: what changes (or moves) something towards being F mustbe distinct from what undergoes the change (or movement) to-wards being F (otherwise the same thing would at the sametime both be Fand not be F).

8 The state the movement/change is directed towards rather than the state in which themovement/change results: there are circumstances in which, because of interference, a

7/30/2019 25535419

7/31


The claim that if something is changing towards being F, then it can-not yet be F is uncontroversial. The premiss in this argument thatcries out for some support is premiss (ii). Why does Aristotle think

that what changes something towards being Fmust itself be F?

One obvious thing to say is that Aristotle is appealing to someprinciple oflike causes like, or to (what in this context amounts tothe same thing) agivingmodel of causation. I cannot give you some-thing I dont myself possess. Similarly (so the thought goes) an agentcannot make something else Funless it is (in some sense) already Fitself.10

But this cannot be the whole story. The principle that like causeslike does not, in itself, provide us with any reason for adopting Aris-totles premiss:

(ii) What changes something towards being Fmust itself be F(since it is causing something to be F),

9 This claim needs to be qualified in two ways. (i) In order to accommodate obvious coun-terexamples, Aristotle needs to say that the agent of a change to Fis something that is Fin a

certain special sense. The housebuilder changes the bricks and mortar into a house, but thehousebuilder is not himself a house. Aristotles view is that the housebuilders soul containsthe form of a house. In this very special sense, the housebuilders soul can be said to havethe form that the bricks and mortar are acquiring. (A more difficult question is how theclaim can be true of the agent of a spatial movement. The end state of a movement from Ato B is: being at B. But surely Aristotle cannot hold that anything that is the agent of such amovement must already be at B. Perhaps in reply he would appeal to his view that, at leastin the case ofself-movement, there must be some object of desire that is the goal of themovement. A movement from A to B to reach an apple at B could be thought of as havingan apple, as its end. Aristotle might claim, then, that it is because the self-mover has a rep-resentation of the apple as something desired that he is able to be an agent of this move-

ment. To have a representation of the apple is (according to Aristotle) to have a part of onessoul that is (in a sense) an apple. In that sense, the self-mover has the form that his move-ment is towards.) (ii) The view that what changes something towards being Fmust itself beFdoes not, presumably, apply to those moved movers that are mere instruments. Aristotlesays that such movers change together with, and in the same respect as, the thing that ismoved (Physics VIII.5.256b1718). This suggests that an instrumental mover that is chang-ing something towards being F is itself (in a certain sense) changing towards being F. Suchan instrumental mover cannot, then, already be F. If this is right, then Aristotles viewdepends on a distinction between instrumental movers, which merely transmit agency, andtrue agents, which change something towards being Fby themselves being F. (For more oninstrumental movers, see n.31 below).10

Aristotle defends this view of causation in Generation and Corruption I.7. He explainsthere that for an agent to change a patient the two must initially be unlike each other. Theagent, in acting, assimilates the patient to itself. The agent (which is already F) makes the

7/30/2019 25535419

8/31

IURSULACOOPE116

as opposed to the alternative:

(Cii) what changes something towards being Fmust itself bechanging (since it is causing a change).

I shall argue that in our passage, Aristotles primary concern is tooppose a Platonic view of self-movement. The Platonic view he op-poses is a view that assumes principle (Cii): it assumes that an agentthat changes something must itself be undergoing change. HenceAristotle needs some justification for holding his premiss (ii) as op-posed to the alternative (Cii). In order to provide this justification, Ishall claim, it is necessary to invoke the view that an action ofchanging something towards being Fis a causing of the state beingF, not a causing of a change.

Aristotle claims that within a self-mover there must be someagent part (a part that is the mover) and that this part does not itselfundergo the motion of which it is the agent. In saying this, he is re-jecting Platos account of self-movement. This becomes clear if we

look at the discussion of self-movement in Platos Laws. There arestriking similarities between the line of thought sketched out by theAthenian in the Laws and Aristotles account of self-movement inthe Physics. Platos Athenian, like Aristotle, distinguishes betweenthings that move themselves and moved movers (things that moveother things while, in their turn, being moved by something else).And in both Platos Laws and Aristotles Physics, we get an argu-ment that it is impossible for a chain of moved movers to go on for-

ever; any such chain must terminate in a self-mover:When we find one thing producing a change in another, and that inturn affecting something else, and so forth, will there ever be, in such asequence, an original cause of change? How could anything whosemotion is transmitted to it from something else be the first thing to ef-fect an alteration? It is impossible. In fact, when something which hasset itself moving brings about an alteration in something, and that inturn brings about something else, so that the motion is transmitted tothousands upon thousands of things one after another, the entire se-quence of their movements must surely spring from some initial prin-ciple (Laws 894e95a)11

7/30/2019 25535419

9/31


But the Athenians view about the nature of this initial principle isquite different from Aristotles. The initial principle, we are told,can hardly be anything except the change brought about by self-

generated motion (895a). It is, in other words, a change that causesitself.

What we need to understand, then, is the reason for this disagree-ment about the initial principle. The assumption in Platos Lawsseems to be that a change must always be produced by a change.This is, at least, a plausible reason for supposing that a change thatis not produced by another change must be a change that is causingitself. Aristotle agrees that it is impossible to have infinite chains ofchanges, each produced by the one before, but he rejects the as-sumption that a change can only occur if it is produced by a change.Indeed, he claims that it is impossible for something to be changingitself in Platos sense. It is impossible for one thing, in one and thesame part, to be undergoing a change and to be causing the verychange that it is undergoing. Our question, then, is why Aristotle re-jects this part of Platos view. Why does he think that, in a self-mov-

er, there must be some agent part that is responsible for the self-movers motion but does not itself undergo this motion?We cannot answer this by appealing to a principle of like causes

like, or to a giving model of causation. The reason is that the Athe-nian of Platos Laws seems to base his argument on an appeal to aprinciple, or model, of the very same sort. As we have seen, theAthenians argument is based on the assumption that a change mustbe produced by something that is changing. That is why he holds

that a change that is not produced by another change must be pro-duced by itself. But this Platonic assumption can itself be seen as aninstance of the principle that like causes like (or of the giving modelof causation): change can only be produced by change; an agentmust itself be changing, if it is togive change to something else. Ar-istotle and his Platonist opponent agree about endorsing some prin-ciple of like causes like, or some version of the giving model ofcausation.12 What we need to ask, then, is why Aristotle thinks that

this principle should be applied in the way he applies it (why, that is,he concludes that an agent that changes something towards being F

7/30/2019 25535419

10/31

IURSULACOOPE118

must itself be F), rather than in the way that the Platonist does (tosupport the conclusion that a change must be produced by some-thing that is changing). In other words, we need to ask why Aristo-

tle endorses what I earlier called premiss (ii), as opposed to thealternative (Cii).

My suggestion is that the reason why Aristotle claims that whatchanges something towards being F must itself be F (instead ofclaiming that the agent must be something that is changing) is thathe thinks of the action of changing something towards being Fas akind of causing of the state being F (rather than a causing of achange). If what is caused is the state being F, and if the agent mustbe like what it causes, then the agent must itself be F.

I have argued that, on Aristotles view, the action of changingsomething towards F is a causing of the state being F (and not acausing of a change). Attributing this view to him enables us to un-derstand why he thinks that an agent that changes something to-wards being Fmust itself be F. But this, of course, raises a furtherquestion. If he holds this view, is it simply an arbitrary stipulation?

What would he say to a Platonist (or indeed, to anyone else) who in-sisted that an action of changing something was an action of pro-ducing a change?

Aristotles answer, I think, would be to appeal to his account ofchange, and in particular to his view that change is an incompleteactuality.13 He describes change in this way at Physics III.12:Change seems to be a kind of actuality, but an incomplete one. Thisis because the potential, of which it is the actuality, is incomplete

(201b3133). He returns to this point about the incompleteness ofchange in the lines leading up to the argument that we have beendiscussing:

13 It is worth noting here that Aristotle may think that the Platonist himself is committed notonly to the giving model of causation but also to the view that change is in some senseincomplete. After all, when Aristotle introduces his view about the incompleteness ofchange in Physics III.12, he boasts that this captures the truth his predecessors were ges-turing at when they described change as difference, inequality and that which is not(201b2033). It is likely that one of the predecessors he has in mind is Plato. (The reference

to inequality recalls Timaeus57e58c; the description of change as that which is not recallsSophist256de.) If so, then Aristotle is arguing that the Platonist, by his own lights, oughtto accept that what changes something towards being Fmust be something that is itself F.

7/30/2019 25535419

11/31


What changes [or is changed, kineitai] is the changeable. But thischanges in virtue of potentiality, not in virtue of actuality.14 Thatwhich is potentially such and such progresses to actuality, and the

change is the incomplete actuality of the changeable. (PhysicsVIII.5.257b69)

What, then, does he mean when he says that change is an actualitythat is incomplete? In Physics III.1, he defines change in terms of po-tentiality and actuality: change is the actuality of what is potentially,qua such (Physics III.1.201a1011). How to interpret this definitionis a matter for debate. The debate centres on how to understand thephrase what is potentially. Two alternatives have been defended inthe literature. Aristotle could mean: (i) change is the actuality ofwhat is potentially changing, qua such, or he could mean (ii) changeis the actuality of what is potentially in some end state, qua such.15

The second of these alternatives makes best sense of the claim thatchange is an incomplete actuality. The thought would be this. Achange towards being Fis a kind of fulfilment (or actuality) of thechanging things potential to be F. The changing things potential to

be Fis completely fulfilled (or actual) when it is F(and the change isover). But its potential to be Fis also fulfilled in a way, though in-completely, when it is in the process of becoming F. Its change is theincomplete fulfilment (that is, the incomplete actuality) of its capac-ity to be in this new state. Consider, for instance, an acorn that isgrowing into an oak tree. The acorn is potentially an oak tree. Be-fore the acorn has begun to grow, this potential is not, in any sense,fulfilled. When the acorn has finally become an oak tree, its poten-

tial to be an oak tree is completely fulfilled, and it can no longer bedescribed as changing into an oak. The acorn is becoming an oakjust when its potential to be an oak is incompletely fulfilled. A

14For a defence of this translation see Waterlow (1982, p. 244, n.27). The most obvioustranslation would be This is potentially, but not actually, changing. But this cannot beright. When it is changing, the changing thing is actually changing. Moreover, the next sen-tence refers to the progression to actuality. If the actuality in question here were the state ofchanging, then the relevant progression would be a progression to the state of changing. Buton Aristotles view, there is no progression to a state of changing: there cannot be a change

that is towards another change.15 For a defence of interpretation (i), see Heinaman (1994). For a defence of (ii), see Kosman(1969) and Waterlow (1982) According to a third interpretation Aristotle is defining

7/30/2019 25535419

12/31

IURSULACOOPE120

change, on Aristotles view, is essentially a going towards somestate. In defining change as the incomplete actuality of a potential tobe in some end state, he is saying that changing to Fis an incompleteway of fulfilling a potential to be F.

For our purposes, the important point here is that change is onlyan incomplete actuality. There is no potential that is a potential forchangingtowards F. In other words, a change towards Fis not thecomplete fulfilment of any potential. It is, instead, the incompletefulfilment of a potential for being F.

Consider now the agent that changes something towards being F.Since a change is not a complete actuality, the only complete actual-ity that is produced in the patient16 is the state being F. If the agentmust have, in actuality, whatever it is bringing about in the patient,this suggests that the agent must itself be F, not that the agent must(like the patient) be changing.

Moreover, just as the patient does not have a potential for chang-ingtowards being F, so also (on Aristotles view) the agent does nothave a potential for producing a change towards being F. The poten-

tial that an agent exercises in changing something towards being Fisa potential that is completely fulfilled when the patient is in the statebeing F. It is a potential for producing this state, not a potential forproducing the change towards this state.17 The teacher is exercisinga potential for producing understanding, not for producing theprocess of learning. According to Aristotle, the process of learningthat occurs is not something that the teachers action is causing;rather, it is itself one and the same as the teachers action of produc-

ing understanding. Similarly, in moving my hand towards P I amcausing my hand to be at P. My hands motion towards P is notwhat I am causing; it is, rather, one and the same event as my actionof causing my hand to be at P.

We can now return to the challenge with which we began. Thecharge was that Aristotles account had an absurd consequence. Al-varez and Hyman argue that it is not plausible to claim both (i) thatthe action of moving Xis the same change as the movement Xun-

dergoes and (ii) that the action of moving Xis an exercise of a caus-al power of the agent. For taken together, these views seem to imply

7/30/2019 25535419

13/31


that the causing of an event and the event caused are one and thesame. And that, surely, is absurd.

We are now in a position to see that, though Aristotle holds both(i) and (ii), he is not, in fact, committed to the view that the causingof an event and the event caused are one and the same. This is be-cause, as his argument about self-change shows, he holds that whatI am causing, when I change Xtowards being F, is Xs being F. It is amistake to call this action an action of producing (or causing) thechange to F. It is, rather, an action of producing the state, being F.His claim, then, is that a certain kind of causing of being F, the kindthat is an action of changing something towards being F, is one andthe same event as that things coming to be F. To return to our earli-er example, he holds that my raising my arm is my causing my armto be up, and my causing my arm to be up is one and the same eventas my arms going up.

This explains why it ispossible for Aristotle to hold both (i) and(ii), but it does not yet explain why he holds these views. I have al-ready said something about the attractiveness of (ii). An action is a

doing: it is the exercise of one of the agents causal powers. But thisleaves the question of why Aristotle is committed to (i), why, that is,he thinks that my action of moving X is the same change as themovement that X undergoes. To understand this, we need to seewhy he rejects alternative possible views. Among modern philoso-phers who emphasize that an action is an exercise of a causal power,there are at least two alternative views about the relation betweenan action and a change. Some philosophers hold that the action of

moving Xis not a change (or even an event) at all. Others think thatthe action of moving Xis an event that causes the change that Xun-dergoes.18 In the next two sections, I shall explain why Aristotlewould reject these modern positions.

III

Aristotles View that the Action of Moving X is Itself a Movement.In modern philosophy, those who hold that an action is a bodilyi d f b i i l d b i k f h

7/30/2019 25535419

14/31

IURSULACOOPE122

English language. The word movement can be used either in atransitive or in an intransitive sense. Clearly, my raising my arm is amovement in the transitive sense: it is my movement of my arm. But

this does not imply that my raising my arm is a movement in the in-transitive sense: that it is a movement that I undergo.19 Aristotlegives no argument that an action is a change or a movement. Wemight wonder, then, whether he is subject to a similar confusion. InGreek, the transitive and the intransitive uses of the verb move (ki-nein, kineisthai) are distinct, the transitive being active and the in-transitive middle. But there is no corresponding difference in thenoun, kinsis (change/movement). Moreover, Greek words for ac-tions of changing something often have distinctive sis endings, likethe ending of the word for change itself, kinsis. The word forhousebuilding, for instance, is oikodomsis; curing is iatreusis. Hasthis similarity of ending simply misled Aristotle into assuming thatactions like housebuilding and curing are themselves changes?

There is, I think, no need to suppose that Aristotle is confused inthis way. Though he does not give an argument that actions, such as

housebuilding and arm-raising, are changes, there are two points wecan make on his behalf. Both points also provide reason to reject themodern view that an action of changing something is not an event atall.

The first is that one can build a house, raise ones arm and, ingeneral, move or change something slowly or quickly. Aristotle else-where uses this as a test of whether or not a certain candidate is achange. He claims that taking pleasure in something cannot be a

change because it cannot be slow or quick. Being slow or quick is,he says, thought to be proper (oikeion) to every change.20 A change,as we have seen, is a progression towards some definite end. It is theincomplete fulfilment of a potential to be in some particular endstate. A sign that something is a progression of this sort (and henceis a change) is that we can ask how quick or slow it is: we can askhow much time was needed to reach the end in question. This is a

19 The point is made, for instance, in Hornsby (1980, ch. 1).20Nicomachean Ethics X.3.1173a321173b4. To say that being slow or quick is oikeion toevery change is to say, not merely that being slow or quick is a necessary feature of change,

7/30/2019 25535419

15/31


question that can be asked not only about changes that we undergo,but also about our actions. Because building a house is a progres-sion towards an end, we can ask how long it took to reach the end

in question; that is, we can ask how quick or slow the action ofbuilding was. Similarly, I can raise my arm more or less quickly.This suggests that the actions of building a house and raising myarm are both changes.

The other sign that such actions are changes is that they can beinterrupted before they are over. For instance, it makes sense to say,He was building a house, but got interrupted before he had fin-ished. Again, this is an indication that an action is a progression to-wards an end. Something that is not a progression of this sortcannot be cut off before it is over. If you disturb me while I amgazing at the sky, you might prevent me from gazing for as long as Iwanted to, but you do not cause me to leave my gazing unfinished.If you restrain my arm when it is only half way up, you have actedin such a way as to prevent my action of arm raising from beingcompleted. Arm-raising (and in general, any action of changing or

moving something) isin a way that gazing at the sky is notthesort of thing that can be unfinished or incomplete. According to Ar-istotle, this shows that actions like arm-raising and housebuilding(actions of changing or moving something) must themselves bechanges.

IV

Aristotles View that the Action of Raising My Arm is the SameChange as My Arms Going Up. Given that my action of raising myarm is itself a change, why does Aristotle think that it must be iden-tical with the change that is my arms going up? Why, for instance,would be reject Hornsbys view that my action of raising my arm is

an event that causes my arm to go up (or, more generally, that myaction of changing Xis an event that causes the change that Xun-d )?

7/30/2019 25535419

16/31

IURSULACOOPE124

use of his discussion of agency and patiency in Physics III.3.21 In thischapter, he discusses a puzzle about agency. The puzzle arises whenwe ask whether an agents action and the change that occurs in the

patient are, or are not, one and the same change. Absurd conse-quences seem to follow both from the assumption that they are twodifferent changes and, equally, from the assumption that they areone and the same. Aristotle resolves the puzzle by attacking the ar-guments that purport to derive absurd consequences from the as-sumption that the action and the patients change are one and thesame.

We can, then, piece together his reasons for holding this view thatthe agents action and the patients change must be one and the samechange, by looking at the half of the puzzle that contains an argu-ment that they cannot be two different changes. The argument thatemerges is as follows:

(i) There cannot be two distinct changes, both in one and thesame thing and both directed towards the same end.

(ii) The action of changing something towards being Fmust it-self be a change towards being F.

(iii) The agents action must be in the thing acted upon (that is,in the patient).

Hence: the agents action of changing the patient towards being

Fand the patients change towards being Fare both in the samething (the patient) and directed at the same end (being F).

Hence (by (i)) they are one and the same change.

It will be helpful to look at each of these premisses in turn.

Premiss (i):

Aristotle makes this claim explicitly, in laying out the puzzle. Heasks: How can there be two changes in one subject towards onef ? d li I i i ibl ( 6) H j ifi

7/30/2019 25535419

17/31


this claim and it might seem open to an obvious objection: couldntsomething be progressing, in two different ways, towards one andthe same end? For instance, couldnt a sapling be simultaneously un-

dergoing two changes towards being an oak: putting out longerroots and getting thicker about the trunk? Aristotle must hold thatin a case like this the sapling would be undergoing one, albeit com-plex, change. Perhaps he thinks this follows from his account ofchange as the incomplete fulfilment of a potential to be in some endstate. A thing either is, or is not, incompletely fulfilling a potentialto be in a certain end state: it cannot, somehow, be incompletely ful-filling such a potential twice over.

Premiss (ii):Aristotle does not explicitly say that the action of changing some-thing towards being Fis itself a change towards being F. However,this point is presupposed by the argument he presents in PhysicsIII.3. He assumes that, if he can show that the agents action is achange in the patient, he can use premiss (i)the claim that there

cannot be two distinct changes towards the same form in the samethingto draw the conclusion that the agents action and thechange in the patient must be one and the same change. He assumes,that is, that the agents action and the patients change are bothchanges towards the same form. Why does he think this?

We have already seen that, on Aristotles view, the action ofchanging something is itself a change. According to his account,then, an action of changing something must be the incomplete fulfil-

ment of a potential for some end state. But what, in the case of suchan action, is the relevant end state? What potential, for instance, isincompletely fulfilled in the act of teaching or in the act of house-building?

The potential in question cannot itself be a potential for chang-ing. Housebuilding cannot, for instance, be the incomplete actualityof a potential for the coming to be of a house. Aristotle insists that itis impossible for a change to have, as its end, some further change.22

By defining change as an incomplete actuality, Aristotle is contrast-ing it with the complete actuality that is its end. The potential in

7/30/2019 25535419

18/31

IURSULACOOPE126

terms of which a change is defined, is incompletely fulfilled in thechange, but completely fulfilled in the end-state towards which thechange is progressing. Hence, the end must itself be a complete actu-

ality. It cannot be a further change. A change that had as its end achange towards being F would have to be a change towards achange towards being F and that can be nothing other than achange towards being F.

But nor can the potential in question be a potential for the agentto be in some new state. There is no new state that an agent is typi-cally in, as a result of successfully changing something. We can say,of course, that the housebuilder, when he has completed his task, isin the state ofhaving builtand that the teacher, at the end of theday, is in a state ofhaving taught. But it is misleading to think ofthese as new states of the housebuilder or of the teacher. Of course,if a housebuilder is still learning his craft, then when we describehim as having built somethingwe may be attributing to him newproperties (such as the possession of a certain amount of expertiseand experience). But the fully qualified housebuilder does not ac-

quire any new causal powers, when it becomes true of him that hehas built another house. This suggests that having built such andsuch a house is not a new property that he acquires, when he finish-es building.

Even if having built were a state the housebuilder acquired, itwould be very odd to think that building was an action directed atthis state, or that the action of teaching was a change towards beingin a state of having taught. The point of housebuilding is to produce

a house. The point of teaching is to produce understanding. An ac-tion of housebuilding is over when there is a house; an action ofteaching, when the pupil understands the thing that is being taught.The new states towards which these actions are directed are, in oth-er words, states in the things acted upon: the bricks in the one case,the pupil in the other.23

According to Aristotle, the action of building a house is (like thechange that is a houses coming to be) a change towards there being

a house; the action of teaching is (like the change that is the pupilsacquiring understanding) a change towards the pupils having un-

7/30/2019 25535419

19/31


derstanding. The action of changing something towards being F isthe incomplete actuality of a potential that the agent has; this poten-tial of the agent is a potential for something else (the thing acted up-

on) to be in the state F. Hence, the agents action of changingsomething towards being Fis (like the change in the thing that theagent is acting upon) a change towards being F.

Premiss (iii):Aristotle might have argued that premiss (iii) follows directly frompremiss (ii). If an action is a change towards being F, then it must bea change in the thing that (barring interference) ends up being F(that is, it must be a change in the thing acted upon). Interestingly,he does not argue in this way. Instead, in presenting the puzzle, hetries to establish that the action of changing something cannot be achange that is in the agent. His argument is that, if the agents actionwere a change in the agent, then either (a) every agent would, whenit acted to change something, itself be changed or (b) there would bean agent that had change in it but was not changed (202a301). He

takes it as obvious that neither of these consequences is possible. AsAristotle understands it, (b) is clearly absurd: what would be thesense of saying that a certain change was in a thing, if that thingwere not changed?24 Why, though, does Aristotle take it for grantedthat (a) should be rejected? Why is it so obviously a mistake to sup-pose that every agent, when it changes something, is itself changed?The answer must be that Aristotle is assuming here that there are (orat least, can be) unmoved movers: things that change other things

without themselves being changed.To understand why Aristotle thinks he can assume this, we mustrecall our earlier conclusions about self-change. As we saw, Aristotlethinks chains of changes must end in something that moves itself.And he thinks that a self-mover must have a part that is an un-moved mover: a part that is a mover but does not itself undergomovement.

24 I assume that kinsetai, here, is a future middle, not a passive, form. Thus, Aristotlesobjection is to the claim that something could have change in it but not undergo thatchange. Ifkinsetai were passive, then he would be objecting to the claim that something

ld h h b b h d b h ld f b h

7/30/2019 25535419

20/31

IURSULACOOPE128

In fact, Aristotle could have provided a more direct argument forthis conclusion. He could, for instance, have appealed to his viewthat an agent must have the form that it is producing in the patient.

An agent that is changing something towards being Fmust, in somesense, be F. As we have seen, this claim plays an important role inhis argument about self-movement. He uses it there to show that theagent of a change to F must, in some sense, be distinct from thething that undergoes the change to F. However, if (as premiss iiclaims) the action of changing something towards being Fis itself achange directed towards being F, the very same consideration willshow that the action cannot be in the agent either: the agent is al-ready F, but the agents action (being a change towards being F) can-not be in something that is already F.

This shows how Aristotle could defend the claim that the agentsaction is not in the agent. It is worth noting, though, that this is notyet premiss (iii). If he is to arrive at the conclusion that the agentsaction and the patients change are one and the same, he needs toshow that the agents action is in the patient. In the puzzle he

presents in Physics III.3, he simply assumes that if the action is notin the agent, it must be in the patient. However, given that he him-self allows, elsewhere, that it is possible for an agent to move a pa-tient using an instrument,25 it is not at all clear that this assumptionis justified. If the agent moves the patient by means of an instru-ment, why couldnt the agents change be a change in the instru-ment, and hence a change that was distinct from the change thatoccurred in the patient?26 Aristotle never considered this. If he had

done, it might have led him to modify his claim that an agents ac-tion is the same change as the change that occurs in the patient. Per-

25 See, for example, Physics VIII.5.256b1415.26Since the action of changing something towards being Fis itself a change towards being F,this would imply that the instrument must itself undergo a change towards being F. Thereare obvious difficulties with this view, since typically the instruments by which we changesomething towards being Fdo not themselves become F. The housebuilders tools do notbecome a house (note, though, that a pan transmits the stoves heat to the water by becom-ing hot itself). However, there is reason to believe that Aristotle must, in any case, accept

something like this view about instrumental changes, on pain of being committed to thepossibility of a certain kind of action at a distance. For if the instrument by which some-thing changes towards being Fis itself something that is neither Fnor becoming F, how can

7/30/2019 25535419

21/31


haps he should have made this claim only with regard to cases inwhich an agent changes something directly, without making use ofinstruments.

V

Aristotles Account as an Alternative to Modern Accounts: SomeAdvantages and Disadvantages. I have explained why Aristotlewould disagree with certain modern accounts of action. Many of hisarguments, though, depend upon assumptions that modern philoso-phers would reject: the assumption that like must be caused by like,for instance, or the view that change is, by definition, an incompletefulfilment of a potential. Does Aristotles account have anything torecommend it, independently of these assumptions?

As we have seen, Aristotle has a distinctive combination of viewsabout the nature of action. On the one hand, he holds that the ac-tion of changing X is the same event as the change that Xunder-

goes. Thus, on his view, my raising my arm and my arms rising areone and the same event. But on the other hand, he also holds thatthis event counts as an action in virtue of being an exercise of a po-tential of mine to bring something about. My arms going up is theevent of my raising my arm because it is an exercise of a potential Ihave for making it the case that my arm is up. My question, then, iswhether this combination of views constitutes an attractive alterna-tive to modern accounts of action.

I have already touched on one reason for preferring Aristotlesview to a view like Davidsons. This is the central role that Aristotleassigns to an agent in his account of action. On Davidsons view,what makes my arms rising an action of arm raising is just that it iscaused by certain other states or events. My arms rising counts asmy raising my arm in virtue of being caused, in an appropriate way,by my beliefs and desires. But this leaves out of account the role Iplay in raising my arm. My beliefs and desires could cause my arm

to move without my doing anything at all. Accounts of this sortmiss a crucial feature of actions: the fact that an action must be ani f t A i l i l bl hi i i

7/30/2019 25535419

22/31

IURSULACOOPE130

counts as an action because it itself is the exercise of an agents po-tential for bringing something about.

However, by claiming that the action of raising my arm is thesame event as my arms going up, Aristotle retains something that isattractive about Davidsons view. For one thing, Aristotle is notcommitted to denying that an action is an event. Though he thinksthat the action of changing something is an exercise of the agentscausal power, he claims that this exercise is itself a change. Becauseof this, he has no difficulty accounting for the fact that an action ofthis sort can (like other changes) occur more or less quickly. Similar-ly, he can accommodate the fact that it is possible to interrupt an ac-tion of changing something before it is over: this, again, is a featurethat the action of changing (or moving) something shares with otherkinds of change.

In addition, because he identifies the action of moving my bodywith my bodys motion, Aristotle avoids some of the drawbacks ofviews, like Hornsbys, on which these are two separate events. OnHornsbys view, my action of moving my body is an event that caus-

es my body to move. But this makes it hard to avoid the conclusionthat actions (such as raising my arm) are invisible: we see their ef-fects (since we see my arms rising), but we do not see the actionsthemselves (the events that cause the arm rising or other bodilymovement).27 Moreover, if an action is a distinct event that causesmy bodys motion, then (at least according to a common view ofcausation), it must occur before my body moves. But do we reallywant to say that I raise my arm before my arm goes up? Aristotle is

not vulnerable to either of these objections. On his view, to see myarm rising is to see my action of raising it (since these are one andthe same event), and for the same reason, my action of raising myarm does not precede my arms going up.

However, there is one objection sometimes leveled at Hornsbysview that can be brought, with at least equal force, against Aristotle.Hornsby is sometimes accused of alienating the agent from herbody. The charge is that the view on which an agents action is an

event that causes her body to move represents the agents relation toher body as being too like her relation to other objects. Thus AdrianH dd k ( 6 ) l i h

7/30/2019 25535419

23/31


[On Hornsbys view,] our bodies are pictured as entities whose powersare wholly distinct from our powers of agency, as entities that we can(at best) only cause to moveand in this respect they are the same as

any other worldly object. Jane moves her body just as she moves (say)a glass of water.28

If it is a mistake to think our actions are no more closely related toour bodies than to other worldly objects, then this is a mistake thatAristotle makes too. In a sense, Aristotle is committed to this viewfor reasons opposite to Hornsbys. Though Aristotle (unlike Horns-by) identifies my movement of my body with my bodys motion, thisis not (for him) a sign that there is anything special about my rela-tion to my body. As we have seen, he holds that, quite generally, anaction of changing Xis the same event as the change that Xunder-goes. Thus, he would also identify my action of moving a glass ofwater with the motion of the glass of water, and he would identifymy action of building a house with the change that the bricks andmortar undergo in becoming a house. For Aristotle, my action isonly the same event as my bodys motion in those cases in which I

am merely moving my body (rather than using my body to movesomething else). If Hornsby seems to alienate us from our bodies byclaiming that our actions are internal events that cause our bodies tomove, Aristotle alienates us from our bodies by claiming that someof our actions are external to our bodies.

One way to bring out the Aristotelian version of this problem isto consider what Aristotle would have to say about the timing of anaction. I have already argued that it is an advantage of his view that

he can deny that my raising my arm is an event that precedes myarms going up. However, he too is committed to implausible con-clusions about the timing of certain actions. Suppose, for instance,that Fred puts a slow-acting poison in Bills drink. If Bill dies as a re-sult, then Fred has killed Bill. On Aristotles view, Freds action ofkilling Bill and Bills dying are one and the same process, a processthat ends in Bills being dead. The worry here is that Bills deathmight occur some time after we would intuitively want to say that

Fred has stopped doing anything. Indeed, Fred himself might die in28 Of H b i t itt d t th i th t J h b d i j t th

7/30/2019 25535419

24/31

IURSULACOOPE132

the meantime. Can Fred really be killing Bill after Fred himself hasceased to exist?29

Finally, Aristotles account faces a further difficulty that is, Ithink, unique to it.30 This difficulty stems from his view that the ac-tion of changing something is the causing of the end state of thechange: raising ones arm is causing ones arm to be up; walking tothe pier is causing oneself to be at the pier. This raises an obviousquestion. There are, surely, different types of action that result inones arm being up, and also different types of action that result inones being at the pier. How is Aristotle to distinguish betweenthese? How, for instance, is he to distinguish between walking to thepier and swimming to the pier? Each of these actions would, afterall, be an incomplete fulfilment of the potential for being at the pier.

It is interesting that Aristotle himself never considers this prob-lem. Perhaps the reason is that, in his discussion of animal move-ment, he tends to think of different types of movement as beingappropriate to different animals. For instance, he begins The Move-ment of Animals with the remark that some animals move by fly-

ing, some by swimming, some by walking, some in othercomparable ways (698a57). He may be assuming that, in the typeof animal that walks, the incomplete fulfilment of an active poten-tial for being at P will be an act of walking to P, whereas in an ani-mal that swims, the incomplete fulfilment of such a potential will bean act of swimming. The thought would be that the manner inwhich an animal incompletely fulfilled this potential would dependupon the animals bodily make-up (e.g. whether it had feet or fins).

Of course, this by itself is not a full solution to the problem. Afterall, some animals (humans, for instance) are capable of swimmingand of walking. Perhaps, though, we can get some hint of a solutionfrom the idea that swimming to P and walking to P are differentways of incompletely fulfilling one and the same potential. Aristotle

29 This problem with Aristotles view is pointed out by Charles (1984, pp. 7981) and byHeinaman (1985). Davidson also discusses a problem about the timing of an action. Heidentifies Freds killing Bill with Freds administering the poison. His problem, then, is to

explain how Freds action of killing Bill can be over before Bill is dead. See Davidson (1971,pp. 579; 1969, pp. 1778; and 1985, pp. 299300).30 At least, none of the modern accounts I have considered here face this problem. Davidson

7/30/2019 25535419

25/31


could say that walking to P is incompletely fulfilling (in a walkingmanner) the potential to be at P, whereas swimming to P is incom-pletely fulfilling (in a swimming manner) the potential to be at P,and so on.

VI

The Relevance of Aristotles Account: His Questions and Ours. Ihave suggested that there are certain ways in which the account Ihave extracted from Aristotle is superior to those found in modernphilosophy, but I have also drawn attention to problems that wouldneed to be resolved in any defence of this account. In this final sec-tion, I shall discuss an objection to the whole project of comparingAristotle, in this way, to modern philosophers of action. It might beclaimed that the kind of comparison I have been drawing is basedon a false assumption: the assumption that Aristotles discussion ofagency is designed to answer the very questions that are posed in

modern philosophy of action, by philosophers such as Davidsonand Hornsby. If it is a mistake to think that Aristotles questions arethe same as ours, can his account nevertheless be relevant to thesemodern discussions?

One important difference between Aristotle and Davidson is inthe scope of their accounts. Aristotles account has a scope that is, inone respect, narrower and in another broader, than Davidsonian ac-counts of action. The scope of Aristotles account is narrower be-

cause it is concerned only with actions of changing (or moving)something. Although Davidson (1971, p. 49) says that all actionsare bodily movements, he makes it clear that he means his accountto apply to actions that are not, in any ordinary sense, movements.The term movement must, he says, be interpreted generouslyenough to include such actions as standing fast. Aristotle wouldnot count standing fast as a movement or change. And it is not justan accident that his account is only about actions of changing, or

moving, something. The arguments I have presented in its defencedepend crucially on this fact. This, for instance, is what allows Aris-l l i h i i i f d f h

7/30/2019 25535419

26/31

IURSULACOOPE134

As I said above, there is also a way in which Aristotles accounthas a broader scope than Davidsons. When Davidson asks whatmakes it the case that my arms going up is an instance of my raising

my arm, he is asking what makes this event an intentionalaction.But the Physics III.3 account of agency is not, primarily, an accountof intentional action. For this account applies also to the actions ofinanimate things. The fires action of heating the stone is the incom-plete fulfilment of the fires potential for the stone to be hot, andthis action of the fires is one and the same change as the stones be-coming hot.

However, these differences of scope do not show that Aristotlesaccount is irrelevant to the questions with which Davidson is con-cerned. Instead, they pose an interesting challenge to certain modernways of answering these questions. Is it right, for instance, to sup-pose (with most modern philosophers) that there must be one ac-count of action that applies both to the action of changingsomething and to actions, such as standing fast, that do not involvechanging or moving anything? It might be considered a merit of Ar-

istotles discussion that he does not lump together changes and otherkinds of activity under the general heading events. Once it is recog-nized, though, that there are significant differences between changesand other kinds of activity, it becomes less obvious there must besome single account that is an account both of what it is to changesomething and of what it is to engage in an activity.

As for Aristotles assumption that there can be a single account ofwhat it is to change something, an accountthat applies both to hu-

man and to inaminate action, I want to suggest that this assumptioncaptures an important insight that is missed by Davidsonian ac-counts. Davidson asks what makes it the case that my arms goingup is (in certain cases) an instance of my raising my arm. On Aristo-tles view, human agency and inanimate agency have something im-portant in common: they are both exercises of a certain kind ofcausal power. For Aristotle, we cannot say what makes a certainevent an intentional action, unless we can say first what makes it an

action. On this view, giving a general account of agency is a neces-sary precursor to explaining what it is to be an intentional action.31

7/30/2019 25535419

27/31


I shall turn now to one final way in which Aristotle differs signif-icantly from all the modern philosophers of action I have discussed.This difference arises not from the scope of Aristotles account, but

rather from his conception of change. The question What makesmy arms going up a case of my raising my arm? cannot mean quitethe same for Aristotle as it does for modern philosophers. Whenmodern philosophers ask this question, they assume that the eventof my arms going up is an event that could have occurred eventhough there was no action of arm-raising: my arm could have justhappened to go up, without me or anyone else raising it. For in-stance, someones arm might go up against her will because she hasanarchic hand syndrome. The question, then, is how to distinguishcases like this from cases in which her arm goes up because she rais-es it. Davidsons answer is that when someones arm goes up be-cause she suffers from anarchic hand syndrome, the event of herarms going up is not caused, in the appropriate way, by her beliefsand desires. Hornsbys answer is that the sufferer from anarchichand syndrome, when her arm rises because of this condition, is notexercising her capacity to raise her arm at will. On Hornsbys view(2004b, 5, p. 179), one must be exercising such a capacity if one isto count as intentionally raising ones arm.

Aristotle does not face quite the same question. The reason is thathis conception of the event ofmy arms going up is different fromthat presupposed in these modern accounts. On Aristotles view, achange is the incomplete fulfilment of a things potential to be insome end state. For something to count as a change, then, it must al-

ready exhibit a kind of directedness: it must be a going towardssome end. This account of change is sometimes criticized on thegrounds that there are certain things that we would, intuitively,want to count as changes that are not, in this way, directed towardsan end. One example of such a change might be the movement of astone that rolls out of the way when it is accidentally knocked by awalker.32 Another might be the arm movements to which a suffererof anarchic hand syndrome is subject. A movement of this kind is

not the fulfilment of a potential anything has to be in some particu-lar end state. If it is right that such accidental movements do not

7/30/2019 25535419

28/31

IURSULACOOPE136

qualify as Aristotelian changes, then this has an interesting conse-quence for his views about action. On Aristotles view, there is notone type of event, my arms going up, which could occur because Iraise my arm, but could equally occur because I suffer from anar-chic hand syndrome. The event which occurs when I raise my arm isof quite a different kind from the event that would occur if my armmoved about because I suffered from this condition. The one eventis a change and hence a progression towards some end; the other isnot. According to Aristotle, if my arms rising is a change, it musthave some agent (though the agent could be someone other thanme, and it could, even, be something inanimate).33 On the otherhand, if it is nota change, then it is not even a candidate for beingidentical with an action of the sort I have discussed in this paper: anaction of changing something.34

Because of this, if Aristotle were asked What makes my arms go-ing up a case of my raising my arm?, he could not understand thisquestion in the way that it is generally understood in modern philos-ophy of action. On his view, when my arms rising is a change, it is

an event that simply in virtue of being a change has a kind of direct-edness towards an end: it is the incomplete fulfilment of a potentialof my arm to be up. Thus, in asking what makes this event an ac-tion, one is not asking what gives it this end-directedness (since thisend-directedness is presupposed by its being a change at all).

Again, this feature of Aristotles account raises important ques-tions for modern philosophers. It challenges us to ask whether therereally is one and the same type of event, arm-rising, which occurs

both when I raise my arm and also when my arm moves upwardswithout my (or anyone else) raising it (perhaps because of a condi-tion like anarchic hand syndrome). Even if one is not convinced that

33 Aristotle makes it clear that he thinks a change must have an agent in PhysicsIII.3.202b2628, when he redefines change in terms of agency and patiency.34 Elsewhere, Hornsby (1997, pp. 10210) puts forward a disjunctive conception of bodilymovements that has something in common with the view that I am attributing to Aristotle.She suggests that the concept of bodily movements subsumes events of significantly differ-ent kinds: there is no unitary category to which both the effects of actions and mere B-move-

ments [bodily movements that are not the effects of actions] both belong (p. 104). Ofcourse, as we have already seen, Aristotle holds (unlike Hornsby) that my action of movingmy body is itself a bodily movement (not something that causes a bodily movement) His

7/30/2019 25535419

29/31


Aristotles definition of change succeeds in capturing the notion ofchange in general, there remains a question whether this account (orsomething like it) applies to certain changes: changes that are the

exercise of an agents power to bring something about. More gener-ally, once we recognize the central role Aristotles account of changeplays in his philosophy of action, we are faced again with a questionabout the modern notion of an event: does the general notion of anevent obscure distinctions that are needed for a satisfactory accountof agency?

Though Aristotles own concerns are often different from ours,this is not a reason to suppose that his account has no bearing onthe questions asked by modern philosophers of action. What I havetried to show in this paper is that these very differences suggest in-teresting alternatives to the ways in which modern philosopherstend to approach such questions. It would be a much larger projectto assess whether any of these alternative approaches should, infact, be adopted. My aim here has simply been to argue that such aproject is worth pursuing.35

REFERENCES

Alvarez, M. and J. Hyman 1998: Agents and their Actions, Philosophy,73, pp. 21945.

Bach, K. 1980: Actions Are Not Events, Mind, 89, pp. 11420.Charles, D. 1984: Aristotles Philosophy of Action (London: Duckworth).Coope, U. 2004: Aristotles Account of Agency in Physics III.3, Proceed-

ings of the Boston Area Colloquium in Ancient Philosophy.forthcoming: Change and its Relation to Potentiality and Actuality, in

G. Anagnostopoulos (ed.), Blackwell Companion to Aristotle (Oxford:Blackwell).

Cooper, J. M. 1997: Plato: Complete Works (Indianapolis: Hackett).Davidson, D. 1963: Actions, Reasons and Causes, in Davidson (2001).1967: The Logical Form of Action Sentences, in Davidson (2001).1969: The Individuation of Events, in Davidson (2001).1971: Agency, in Davidson (2001).

1985: Adverbs of Action, Appendix A in Davidson (2001).2001: Essays on Actions and Events (Oxford: Clarendon Press).

7/30/2019 25535419

30/31

IURSULACOOPE138

Frankfurt, H. 1988: The Problem of Action, in The Importance of WhatWe Care About(Cambridge: Cambridge University Press).

Haddock, A. 2005: At One With Our Actions, But At Two With Our Bod-

ies: Hornsbys Account of Action, Philosophical Explorations, 8 (2),pp. 15772.

Heinaman R. 1985: Aristotle on Housebuilding, History of PhilosophyQuarterly, 2, pp. 14562.

1994: Is Aristotles Definition of Change Circular?, Apeiron.Hornsby, J. 1980: Actions (London: Routledge & Kegan Paul).1997: Simple Mindedness (Cambridge, MA: Harvard University Press).2004a: Agency and Actions, in H. Steward and J. Hyman (eds.),

Agency and Action (Cambridge: Cambridge University Press), pp. 123.2004b: Alienated Agents, in M. De Caro and D. Macarthur (eds.),

Naturalism in Question (Cambridge, MA: Harvard University Press),pp. 17387.

Hussey, E. 1983: Aristotles Physics. Books III and IV(Oxford: ClarendonPress).

Kenny, A. J. P 1963: Action, Emotion and Will(London: Routledge & KeganPaul).

Kosman, L. A. 1969: Aristotles Definition of Motion, Phronesis, 14,

pp. 4062.Kostman, J. 1987: Aristotles Definition of Change, History of PhilosophyQuarterly.

Makin, S. 1991: An Ancient Principle about Causation, Proceedings of theAristotelian Society.

Sedley, D. 1998: Platonic Causes, Phronesis, 43 (2), pp. 11432.von Wright, G. H. 1963: Norm and Action (London: Routledge & Kegan

Paul).Waterlow, S. 1982: Nature, Change and Agency in Aristotles Physics (Ox-

ford: Clarendon Press).

7/30/2019 25535419

31/31

25535419

Documents