On Not Being Able to do Otherwise

Citation preview

On Not Being Able to do Otherwise Winston Nesbitt; Stewart Candlish Mind, New Series, Vol. 82, No. 327. (Jul., 1973), pp. 321-330. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28197307%292%3A82%3A327%3C321%3AONBATD%3E2.0.CO%3B2-U Mind is currently published by Oxford University Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri Jun 22 07:33:21 2007

VOL. LXXXII NO. 3271

[July, 1973

MIND

A QUARTERLY REVIEW

PSYCHOLOGY A N D PHILOSOPHY I.-ON NOT BEING ABLE

T O DO OTHERWISE

BY WINSTONNESBITTand STEWARTCANDLISH WE shall call the following argument, Premiss 1

"

argument A " :

If a man could not have done otherwise than he in fact did, then he is not responsible for his action. (That is, ' He could not have done otherwise ' is a recognised excuse.) Premiss 2 If determinism is true, it is true of every action that the agent could not have done otherwise. Conclusion If determinism is true, no one is ever responsible for his actions. Argument A sets out, in broad outline, the reasoning by which those who think that determinism rules out the possibility of moral responsibility have often arrived a t their view. The attempted proof contains no obvious fallacy. We shall try to show that it does, however, fail to establish its conclusion; but also that this failure is not for the reasons usually presented by critics of the argument. The first premiss seems difficult to deny, and it has even been suggested that it states a logical truth.l At any rate, philosophers who wish to avoid the conclusion of argument A usually accept the first premiss and concentrate their attacks on the By Roderick Chisholm, in his contribution to Determinism and Freedom, ed. Sidney Hook (Collier Books, 1961), p. 157. 11

321

322

WINSTON NESBITT AND STEWART CANDLISH :

second. I t is assumed by these philosophers that the second premiss will have been shown to be false if it is shown that, whether or not determinism is true, we may still have the ability, capacity, or power to do otherwise than we do. Two examples will illustrate the point. First, P. H. Nowell-Smith, in the course of an attempt to reconcile the truth of determinism with our practice of holding people responsible for their actions, notes that the truth of the statement ' He could have done otherwise ' is a necessary condition of the justifiable application of the judgment ' He deserves censure ',I and then goes on to analyse ' He could have . . .', as it occurs here, in terms of ' He would have . . . if . . .'. This is the same analysis that he provides for the phrase when it is used to assert the presence of abilities llke the ability to run a mile or to play the Appassionuta2; and although he later accepts certain of Austin's criticisms of his analysis of statemelrts about abilities, he still remains convinced that ' the concept of being able to do something . . . lies at the heart of the free-will p r ~ b l e m . ' ~ Austin himself did not question this conviction, and seems even to endorse it by suggesting that his criticisms of Nowell-Smith's analysis go some way towards showing that determinism is not compatible with ' the things we ordinarily say about what we can do and could have done '.4 Secondly, 11. R. Ayers, also intent on proving the second premiss to be false, notes that ' can ' may be used to imply that both capacity and opportunity are present, and claims that " This, of course, is how we must take the phrase so dear to the hearts of the freewill controversialists. ' He could have acted otherwise ' ".5 Once more it is assumed that what is at issue in the consideration of determinism is a man's capacity or power to do otherwise in a given situation. However, these philosophers fail to show that there is a mistake in argument A. They fail because they misunderstand the nature of the statement ' He could not have done otherwise.' In demonstrating the misunderstanding, we shall begin by P. H. ?\Ton-ell-Smith,Etllics (Penguin Boolis, 1964), p. 273. "bid. pp. 275-278.

Op. cit., 1961 reprinting, p. 290.

J. L. Austin, ' Ifs and Cans ', Philosophical Papers (Oxford University

Press, 1961), p. 179.

"1. R. Ayers, Tile Refutation of Deternzinism (Methuen, 1968), p. 104.

The title of Ayers' book msy seem to cast doubt on our clsim that he

wishes t o prove the second premiss of Argument A t o be false. Such

doubt is illusion, caused by his use of the word ' determinism ' (see his

p. 4) in s sense which is not perhaps the most usual. We have chosen

his second sense, which we consider more usual. It is made clear in

Section I1 of this paper.

ON NOT BEING ABLE TO DO OTHERWISE

323

noting what Ayers has to say about some cases in which people are inclined to agree that the agent could not have done otherwise because of the presence of factors of a certain sort which, according to Ayers, nevertheless logically cannot limit a man's capacities. (His reasons for claiming that they cannot are irrelevant to the present purpose.) The cases he considers are those of a man who said while drunk things which he later regretted; of a man who, on being insulted, threw a bottle (presumably a t the person who insulted him); and of a one-time patriot ~vhospied on his own country after being captured and brainwashed by the enemy. He deals with these ' apparent counter-examples to [his] thesis ' by pointing out that " ' I could not help it ' and ' He could not do otherwise ' are used very loosely as what we rnight call ' excusing expressions ', even when the excuse offered is not really that the agent was literally incapable of doing anything else."l The suggestion here is clearly that these are nonstandard instances of the use of these expressions, and that in the normal cases, the excuse offered i s that the agent was literally incapable of doing otherwise. But this is precisely the reverse of the truth, which is that the remark ' He could not have done otherwise ' is inappropriate unless the agent did have the capacity to do otherwise. This is made clear by the following considerations. It has already been seen that what Ayers means by ' literally incapable of doing otherwise ' is ' lacking the capacity and /or the opportunity to do otherwise '. A nlan is literally incapable of doing otherwise, then, when what he does is something like descending vertically to the ground after being pushed from a high bridge, or remaining motionless while in a certain physiological condition which causes paralysis-when, in short, his doing other%-iseis empirically impossible. Now it is peculiar, to say the least, to refer to the fact that a man lacked both capacity and opportunity to do otherwise than fall vertically from a bridge as his excuse for landing on and killing a child. For talk of excusing to be appropriate, landing on the child must have been a course selected by him in preference to others which were within his pourer. (Perhaps he thought that he alone had in his head the formula which would save the human race from extinction, and so deemed his life more important than that of the child.) If n o other course was within his power, he did not do anything, and a man can be excused only for what he does. A corollary of this is that it is nonsense to say that a man is not responsible for his action if he lacked the power to do otherwise, for if he lacked this power then what he Ibid., pp. 163 and 164.

324

WINSTON NESBITT AND STEWART CANDLISH :

" did " was not an action1; and the question of his responsibility, in any sense stronger than that of mere causal instrumentality, for what got done, can arise only on the mistaken assumption that it was a result of his action. It is, of course, true that if one is literally incapable of doing other than one does, then one cannot be morally responsible for what gets done; but all this amounts merely to the fact that a man is directly responsible only for his actions as opposed to mere movements of his body. In trying to show that determinism is compatible with our having the power to do otherwise than we do, then, Ayers and Nowell-Smith do not show that it is compatible with the fulfilment of the condition of moral responsibility which is implicitly stated in the first premiss of argument A. However, it might be objected, why should they be criticised for failing to show that determinism is consistent with the fulfilment of this condition? They do show that it is consistent with the fulfilment of a condition of moral responsibility, namely an agent's capacity to do otherwise than he in fact did; and presumably this is all that they were trying to do. Two points can be made in reply to this objection. In the first place, Nowell-Smith and Ayers do not distinguish clearly, as we have tried to do, between these two conditions. While discussing the analysis of uses of ' could have ' in ' moral contexts ' . 2 Nowell-Smith considers ' excuses ' ranging from u u

physical compulsion to a bad upbringing, apparently assuming 1 There are many intermediate cases which might seem t o cast doubt upon this contention. For example, if I stop running because of a bad stitch, I have performed a n action, but one may want to say that I lack the power to go on running. Whether I do or not, in fact, is unclear, but what is clear is that if I undeniably lack the capacity to go on (e.q. I collapse from exhaustion), my not going on is not a n action of mine in the sense, e.q., elucidated by Charles Taylor on p. 29 of The Explanation of Behaviour (Routledge, 1964), where ' it is not only necessary that it end in the result or meet the criterion by which actions of this kind are characterized, but it must also be the case that the agent's intention or purpose was to achieve this result or criterion.' The addition of a further necessary condition for action, namely that the intention be productive of the result (i.e. it is not action if the intended result is accidentally achieved), makes the point even clearer. Nowell-Smith, op. cit., ch. 20. His implicit conflation of ' He could not have acted otherwise ' with ' He lacked the c a ~ a c i"t vto do otherwise ' is shown also in ch. 19, in the following passage: Capacities are a sub-class of dispositions. To say that a man ' can ' do something is not t o say that he ever has or ever will; there may be special reasons why the capacity is never exercised, (p. 277). The point is rendered conclusive when Nowell-Smith (1961 edn., p. 290) accepts Austin's contention that ' in the relevant cases ', ' could have ' is the past indicative of ' can '. &

...

ON NOT BEING ABLE TO DO OTHERWISE

325

that the form of the excuse in all these cases is essentially the same. While Ayers, though he stresses that the ' excusing expression ', ' He could not have acted otherwise ', is sometimes used even when it is not meant that the agent was literally incapable of doing otherwise, suggests by this that the sentence is normally used with this meaning. Again, to say ' He could have acted otherwise ' is not perhaps the most natural way of saying that a man had the capacity to do other than he did (if we should ever in fact mant to say such a thing), especially as a man who literally lacked this capacity could not be said to have acted. Yet both Ayers and Nowell-Smith choose this particular formula to indicate the type of judgment with which they are concerned in their discussion of capacities, and this is surely because they have in mind that to say ' He could have acted differently ' is the most natural way to reject the applicability of the common ' excusing expression ' ' He could not have acted otherwise '. I n the second place, there is more than one condition of moral responsibility: for example, it is a condition of an agent's being fully responsible for his actions that they correspond to his intentions. But if one's aim is to reconcile determinism with the possibility of moral responsibility, proofs that determinism does not entail that actions do not correspond to intentions would be relevant only if there were some reason to doubt that this is so. Now why should it be doubted that determinism allows for the possibility of someone's being able to do other than he in fact does? Determinism certainly rules out the possibility of anyone's ever exercising his capacity, on any particular occasion, to do other than he in fact does, but it is not inzmediately clear why it should be thought that one's having this capacity is similarly ruled out. An obvious retort to this leads on to the next question which we mant to consider. It may be said that it is relevant to demonstrate that determinism is compatible with our having the capacity to do other than we do, because determinists often deny this fact.

I t is true that many determinists have claimed that men can never do anything other than just 1171iat they do. But one should by nowr be wary of assuming that this must be interpreted as the claim that they always lack the capacity to do otherwise than they in fact do. How else, though, might it be interpreted? Richard Taylor, who agrees with those who claim that deter-

326

WINSTON NESBITT AND STEWART CANDLISH :

minism is inconipatible with moral responsibility, reconstructs1 the argument which he thinks has led them to this conclusion. The argument, which he believes to be sound, is essentially our argument A. However, he interprets the second premiss, not as the claim that determinism excludes one's having the capacity to do other than one does, but as the tautology that, if determinism is true all actions are necessary. Against those who assert that causes do not necessitate their effects., he ~ o i n t sout that there is ' a clear and common sense in which, for example, a man who has been decapitated necessarily dies, or c a n not go on living '. We suggest that Taylor interprets the determinist's claim correctly here. Pirst, to say ' He could not have acted differently' could be a legitimate way of saying ' His behaviour was causally necessary '. Secondly, as we have already remarked, it is not clear why determinism should be claimed to entail that one always lacks the capacity to do anything but just what one does. Finally, neither Ayers nor Nomell-Smith produces an example of a philosopher who claims this, though the former thinks that Hobbes was committed to the view, on the grounds, significantly, that he held that ' everyone acts necessarily, and could not act differently '.2 Ayers also quotes3 a remark from Hunie's T r e a t i s e which implies that people never have the power to do anything but just what they do, but Hume does not suggest that one's having such power is ruled out by determinism; his suggestion stems, rather, from his scepticism regarding unobservable entities. Nevertheless, it may be that some philosophers have meant to claim that determinism is incompatible with one's having" the power to do other than one does. But if they have, the appropriate response is not to set about an elaborate refutation but rather to ask why on earth they expect such a claim to be taken seriously. In view of this, we shall turn to an examination of the reasoning which, according to Taylor, establishes that if determinism is true: no one is ever responsible for what he does. He argues that an agent is not responsible for what he does if he could not have done otherwise, and that since determinism is the view that all events are causally necessary this condition of moral responsibility is never fulfilled if determinism is true. Now as was remarked earlier, even those who deny that determinism is inconlpatible with moral responsibility accept this condition of such responsibility. But they do not accept it if it means that an action for 11-hich one is responsible is contingent, i.e. not A

0

1 In

his contribution t o Hook (ed.), op. cit., p. 225. Ibid., p. 16. Ayers, op. cit., p. 5

ON NOT BEING ABLE TO DO OTHERWISE

327

causally necessary. Yet according to Taylor the condition means precisely that! For after stating that it is a condition of a man's being held responsible for what he has done ' that he could have done, or could do, something else ', he adds ' that is, that the occurrence for which he is responsible . . . is contingent '.l Now if this be a condition of moral responsibility it is very simple indeed to show that such responsibility is incompatible with determinism, for the latter is just the thesis that this condition is never fulfilled. Only those who have already accepted that determinism is incompatible with moral responsibility are likely to accept that a nlan is not responsible for what he does if his action is causally necessary, because for such people these are different ways of saying the same thing. Taylor introduces as if self-evident a premiss which presupposes the conclusion for which he is arguing, and there is little doubt that he is prepared to beg the question in this way because he takes himself to be referring to the indisputable fact that the statement ' He could not have done other\\-ise ' is always a sound defence. Like Ayers and Nowell-Smith, however, he has misinterpreted this fact, with dire consequences for his argument.

We have argued so far that the defence ' He could not have done otherwise ' is to be interpreted neither as ' He was incapable of doing otherwise ', nor as ' His action was causally necessary '. Won-, then, is it to be interpreted? The question has been partially answered in part I, but a further example will clarify the issue. Suppose that we are defending the action of a bank teller who has handed money over to a bandit a t gunpoint, and we say ' He couldn't have done anything else; he had a gun pointed a t him '. The person to whom we say this replies, ' K h a t do you mean, couldn't have done anything else? He was quite capable of hanging on to the money, wasn't he?' Assuming that this is not intended as a joke in poor taste, a natural reply would be ' We don't mean that he \\-as i9zcapable of doing anything else. What we're getting a t is that it would have been unreasonable to have expected him to do anything else.' If we were then pressed to explain why it would have been unreasonable to have expected the teller to do anything else. we n~ouldprobably point out that the presence of the gun provided him with overwhelming reasons against doing anything else, that is, for doing what he did. And this, we suggest, is what must be the case for the " excuqing Taylor, op. cit., p. 225.

328

WINSTON NESBITT AND STEWART CANDLISH

:

expression ", ' He couldn't have done otherwise ' to be appropriate: there must have been circumstances which provided the agent with overwhelming, or very good, reasons for doing what he did. For first, in all cases in which such an expression is an appropriate defence, the factor which makes it a defence is in fact such as to provide the agent with good or overwhelming reasons for doing what he did. I t is difficult to imagine how this contention might be proved; we have tried to show that it is plausible, and can only ask anyone who thinks differently to produce counter-examples. Secondly, to say ' He had overwhelming (or good) reasons for doing what he did ' is itself a sound defence, and it would be implausible to claim that although this defence is a h a v s available whenever ' He could have done otherwise ' is appropriate, some other defence is in fact invoked. It may be thought that we have already had occasion to mention some cases which constitute counter-examnles to what has just been said. It will be remembered that ~ y & makes s his remark, that the " excusing expressions " ' He could not help it ' and ' He could not have done otherwise ' are used even when it is not meant that the agent was literallv inca~ableof doingw otherwise, in connection with some examples in which the excusing factors are, respectively, the agent's having been drunk, the agent's having been insulted, and the agent's having been brainwashed. Now being drunk. being insulted. and being brainwashed do not, on: would' h a v c thought, ' provide &e with overwhelming or even good reasons, respectively, for saying things one later regrets, throwing bottles and spying on one's country. However, Ayers is simply wrong in suggesting that in any of these cases it is appropriate to defend the agent by saying ' He could not have done otherwise '.l To say ' He couldn't have done anything else-he was drunk ' might be in order if what he did wis reffuvse to drive home, a course ?or which his being drunk provided good reasons, but not if what he did was something he later regretted, thereby acknowledging that it was not a course supported by good reasons. The strangeness of defending an act of bottle-throwing by saying ' He couldn't have done otherwise-he'd been insulted ' and of defending an act of spying with ' He couldn't have done otherwise-he'd been brainwashed ' is sufficiently obvious not to need stressing. We will remark only that neither would be strange if insults and brainwashing were the sorts of things which provided people with good reasons for the actions in question. For example, the former would not be strange in a country where failure to ' He could not help it ' seems to be used, confusingly, in both kinds of w

W8e.

ON NOT BEING ABLE TO DO OTHERWISE

329

avenge an insult was considered cowardice and bottle-throwing a recognised way of avenging insults. Now the consequences of all this for argument A are disastrous. For one thing, if the argument is not to involve equivocation, its second premiss will have to be interpreted as the bizarre claim that if determinism is true, we always have good reason for doing what we do. But this is by no means the only difficulty. We have so far allowed, though not explicitly condoned, the references of Ayers and Nowell-Smith to ' He could not have done otherwise ' as an excuse. But our account of the nature of this defence shows that, in that strict sense of the word in which excuses proper are distinguished from justifications, it is not an excuse but a justification; for to claim that one had good or overwhelming reasons for one's action is clearly to try to justify rather than excuse it. Now as Austin has pointed out,l to justify one's action is to accept responsibility for it but claim that it was a good, right or sensible thing to have done; whereas to offer an excuse is to concede that what one did was not good, right, or sensible but claim that one was not responsible, or at least not wholly responsible, for it. But this means that the first premiss, though unquestioned by either side in the dispute over argument A, is false. For if the defence that a man could not have done otherwise is accepted, it does not follow that he was not res~onsiblefor what he did-what does follow is that although " he was responsible for what he did, it was in the circumstances a good, right, or sensible thing to have done. I t is tempting to dismiss this as mere linguistic hair-splitting, and to say that it remains true that when we accept this defence of a man's action, we accept at least that he is not to blame for it, which is the crucial point. But this is confused. To accept that an action was justified is not to accept that the agent was not to blame for it; it is to accept that, since the action turns out to have been a good, right, or sensible one, the question of blame for it simply does not arise. We conclude that the first premiss is false aid-argument A a non-starter.

To summarise the main points for which we have argued : Argument A is often accepted as demonstrating the incompatibility of determinism and moral responsibility. Those who wish to deny this incompatibility have in consequence frequently advanced purported refutations of the argument. 1

J. L. Austin, ' A Plea for Excuses ', q.cit., p. 124.

Since the first premiss has seemed undeniable, they have concentrated on trying to show that the second premiss is false, interpreting the key phrase ' He could not have done otherwise ' as ' He was incapable of doing otherwise '. But if this was how those who accept argument A understood the second premiss, proofs of its falsity would be superfluous, as the argument would then involve a fallacy of equivocation; for ' He could not have done otherwise ', as it occurs in the argument's other premiss, does not mean ' He was incapable of doing otherwise '.I However, what those who accept the second premiss have meant by the phrase ' He could not have done otherwise ' is ' His action was causally necessary '; that is, they have meant this premiss to be understood as the tautology that if determinism is true, all actions are causally necessary. But this also is a different sense of ' He could not have done otherwise ' from that in which the phrase is used in the first premiss, so that on this interpretation the argument still involves an equivocation, though a different one. 'IThat is meant bv ' could not have done otherwise ' in the first premiss is ' had good or overwhelming reasons to do what he did '. This means that, if the argument is to avoid equivocation, the second premiss must be interpreted as the claim that if determinism is true we alwavs have " good reason for what we do, a claim which, presumably, no one would wish to make. [For the consequences of, alternatively, re-writing the first premiss, see footnote below.1 Worse still. it means that the first premiss is false; for ' He could not have done otherwise ' turns out to be not an excuse but a justification, so that to offer this defence of an action is not to deny the agent's responsibility for it but to claim that it was not a bad thing to have done. There is, therefore, nothing but sophistry in argument A in its usually discussed forms, and this means that the appeal to the fact that ' He could not have acted otherwise ' is a sound defence, is a broken reed in the dispute over the compatibility or otherwise of determinism and moral responsibility.

Unicersity of Adelaide Unicersity of Westeryz Australia 1 One may wish to ask here whether it could mean this. The short answer is that it could; that if it did, no action, and thus no responsibility for action, is in question; hence that argument A thus rewritten demonstrates the incompatibility of determinism with the notioil of huinan action, which entails far more sweeping revisions than the mere closing of prisons; and that the question of the truth of the seco~ldpremiss is worth reopening. Yor qualifications, cf. note 1, p. 324.