The ideas of motive and intention have played an important part, in recent years especially, in the discussion of moral evaluation. This is not surprising, for intention is usually thought to include the whole or some important part of what an agent anticipates as the result of his action, what he sets out to accomplish; and it seems evident that what a person supposes will come about in the course of his conduct is in some way a main consideration in the evaluation of what he is doing. The intention, thought of in this way, is sometimes thought to include only the aim the agent has especially in mind and the means directly relevant to it. The remainder will be consequences of our conduct in which we are not so expressly interested but which we must recognise and accept as unavoidable side effects of what we are more explicitly anxious to achieve.

On the view which I have been defending, intention in the present sense, in either its limited or more comprehensive form, is only indirectly relevant to estimation of moral worth, namely as establishing what the agent thinks is the course of his duty and the extent to which this deviates from what he most wants to do on the whole at the time. The proper moral worth will turn on the way he exercises his freedom of undetermined choice on the occasions when there is such deviation. If, for example, I believe my action will cause someone pain, that tells heavily against its being my duty and will need substantial compensation in other ways if I am still to consider it my duty to do it. In most cases it will also much affect what I want.

In this context it seems evident that our reckoning must be with the intention in the broader sense. For it would normally be thought to be unreasonable to overlook major consequences of one’s action, in estimating its conformity with our obligations, simply because we are not explicitly concerned with them. If what we hope to achieve is likely to have a disastrous accompaniment, that would tell heavily against its being in the line of duty. If, for example. I thought my view would be much improved by cutting down a large tree at the foot of my garden, there might seem to be nothing wrong in my setting out to do so. But this would not be the case if there was a likelihood that the tree would fall across the road in the way of traffic, and no precautions were taken. My aim in building a dam to create a reservoir to drain away water for a factory may be quite in order in itself, but not if it seriously diminishes the water supply of farms lower down the valley. It may not be part of my own concern that I should damage the farms, but I am not entitled for that reason to leave their fate out of my reckoning. In most cases, the damage which my action may incidentally cause, incidental, that is, to what appeals to me in the end I pursue and the means of achieving it, will also create a strong aversion to taking such a course.

If we are not to estimate moral worth in terms of our response to such conflict as there may be between duty and inclination, we are left with evaluation of our wishes and our character as a whole as these are reflected in what we set out to achieve. Intention will be important in this way, but here again the intention will need to be taken in its wholeness and not confined to what immediately appeals to us and the means directly pertinent to that. How good or bad our wishes and dispositions are will be much affected by what we are prepared to tolerate in the fulfilment of our aims.

A motive is sometimes thought to be that part of our intention as a whole which expressly appeals to us and for the sake of which we pursue some particular course. My motive, in the example a moment ago, is to get water for my factory. This is what prompts my action. Building the dam is a means, and I have no interest in it apart from the purpose it serves. But if ‘motive’ is used in this way, we have obviously to think of a great deal besides motives in determining the propriety of a course of action, and also in our assessment of what sort of person someone may be. We have to reckon with the means adopted to achieve what we want and the consequences we are prepared to tolerate in the course of this.

In the course of these considerations we have also to heed not only what seems to us obligatory in the light of our own assessment of the situation, of what will come about as a result of our actions, but also what in fact is our duty. My own sincere impression of where the course of my duty lies may be mistaken or inadequate, partly through wrong assessment of what the results of my action will be and partly through wrong assessment of the relevance or place of various results as ingredients in what is normally required, ignorance of fact and failure of moral judgement. There is thus something which is materially or outwardly required of me, what in actual fact is my duty. This is not directly relevant, on the views I have advanced, to the estimation of my moral worth. That will turn on how I respond to what seems to me required. But the recognition of there being obligation in this objective way, which is distinct from my own understanding of what is my duty, is nonetheless of great importance in many ways. One of these is that there is little point in having an understanding of what my duty is without the presumption that there is in actual fact something which I ought to do whether I appreciate this or not. Our fallible opinions claim to be about what is the case, in respect to moral requirement as to anything else. Without this presumption it is hard to see how there could be any ethical convictions — they would have no point apart from a claim of truth. This is what the relativist often overlooks.

In addition, it is plainly one of our major duties to discover what is one’s duty, to form as sound an opinion as we can about what a situation requires. We are not likely to do this effectively and with the persistence needed if the limit of our serious concern is set by what we happen to think from time to time. As in other contexts, the consciousness of our own fallibility, far from making us complacent, ought to spur us to more resolute efforts to correct our own judgments and bring our thoughts into as close a conformity as we can with what is in fact the case. At any particular time we have to be guided by such opinions as we have, in consultation with others, formed by that time. But that is no bar to our passing our opinions under constant review to guard against error and persist in the pursuit of truth. In this enterprise it is always an advantage to bear in mind that there is an objective truth to be sought after in our opinions, which also we have often good reason to believe are in line with the truth.

I shall not discuss this further here, as I have already considered it at length in my paper (first published in Mind 1945) on ‘Obedience to Conscience’ and related papers assembled in my book Morals and Revelation.

A further advantage of the insistence on objective truth is that it makes for tolerance, in part by reminding us of our own fallibility and also through the better realization that others who disagree with us may in fact be right and also holding their opinions, whether sound or not, with the same sincerity and the same devotion to what they understand to be their duty. The pacifist does not, for this reason, have to impugn the moral integrity of the soldier, or the soldier that of the pacifist. In one sense the one will have to tell the other that he has not done his duty, or that what he does is wrong; one or the other will have failed to meet what the situation requires. But that is no bar to their recognising one another’s integrity and scruples. In highly complicated matters there is much to be said on both sides; and, as we have no monopoly of wisdom, so we have no monopoly of the conscientious devotion to duty as we see it on which properly moral worth depends. To pursue the wrong course is not ipso facto morally evil.

It is also possible to pursue the proper course, to do what we ought, for the wrong reasons, which, if not bad, are morally worthless. This was much stressed in debates earlier this century about ‘the subjectively right act’ and ‘the objectively right act’. We may do the latter without being properly concerned about its rightness. It may be to our advantage to do right, honesty may pay, and it may be that this is what counts with us even when we are convinced that what we do is right. This is why it was often said, in the debates mentioned, that it was the motive which mattered, or mattered mainly, in the estimation of moral worth. But what was understood by ‘the motive’ in such discussions was not always clear.

The motive, on some views, would be understood in these contexts as that part of the intention, understood as the whole of what we envisage as the likely result of our action, which especially appeals to us and for the sake of which we adopt a particular course. A person may exhaust himself canvassing during an election, and make many speeches, but his motive is to get himself (or someone else) elected. Another may commit a felony but his motive may be to feed his starving family, as it may also be to enrich himself. The motive in this sense is plainly important, but it is not always clear just in what way it is important.

Our assessment would clearly be affected in some sense by whether a man robbed a bank to succour his destitute family or did so to live it up on an extended holiday or to be free of embarrassing debts. The latter thought, and quite plainly the first, would much affect our judgement of the situation. We would not view the first as harshly as the second. But what sort of judgement is involved here? Are we judging the material rightness of a course of conduct? If so, the question of what appealed to the agent, what ‘motivated’ him, would not be directly relevant. It would count only as part of the course of events to which his action leads and the judgement whether, on the whole, there being much else involved, his action was objectively right. But if we were assessing the moral worth of the agent we would have to regard what especially appealed to him, together with any reluctance to encompass other things. We might, therefore, conclude in this context that heed should be had to both the motive and the intention as a whole. A good motive, in the present sense, would not justify anything and everything. My own aims may be innocent enough, or even praiseworthy, but the means, or the consequences that I tolerate, may be bad. Many concluded in this way that both motive and intention affected the morality of the action, though not the material rightness of the act.

The debate was however much complicated in two ways. Firstly, the meaning of ‘motive’ was apt to be extended to include feelings such as love or hate. These may be said in some ways to move us to action, and much would seem therefore to turn on which it would be. To steal from love of an ailing child would be very different from stealing ‘to live it up’. It was also pointed out, as a second complication, that we could not summon up feelings such as love or compassion immediately on some occasion, any more than we could so expel a feeling of hate. We felt that way or we did not. We might cultivate the appropriate feelings, but we could not expressly command them.

These considerations were in due course extended to the desires with which our emotions would be associated and in this way to our desiring nature as a whole as that affected our conduct. Did it make sense in that case to maintain, as impressive ethical thinkers like H. W. B. Joseph and G. C. Field did, that it was our moral obligation to act from a certain motive, the motive counting in an impressive way in the determination of what we actually ought to do, the right act depending on the motive from which it was done? This seemed, however, to reverse the proper order of things. If the motive is what moves us to act how can we choose to act from that motive? Ross took refuge here in the sharpness of the distinction between the ‘morally good action’ and ‘the right act’. We set ourselves to accomplish the latter. He is, in one respect as we have seen, on very safe ground here, for we do have to draw a clear distinction between the objective rightness of what our action involves in virtue of what it brings about and the moral worth of the agent depending on his understanding of the situation and his state of mind in other respects. But for Ross there is a further obstacle in the path.

This is because he continued to think of motive, both in the way of feelings and of desires, as what moves us to act. If we cannot summon up feelings and desires at will, have we any better control of the acts which we are moved by such motives to perform? It is the act, the setting ourselves, that we choose. But if the motive and the act are inextricably linked can we be said to choose the one in any significant sense in which we do not choose the other? Ross was undoubtedly worried about this and, while adhering firmly to determinism, in the form of ‘self-determination’, confesses to being very uneasy about this, hoping that a philosophical genius will some day reconcile our thought about freedom and responsibility with acceptance of the law of causality.1

Some ingenious convolutions, and many cross-purposes, in the spate of ethical thinking to which I am alluding, could have been avoided if closer heed had been paid to the exclusiveness of the idea of obligation, and the special freedom and responsibility associated with it, in the determination of properly moral worth. It was a great mistake to extend this complex of ideas to other modes of evaluation where they have no place. Ross is quite right in asserting that we cannot summon up sorrow and love at will and that all that we can do, in the way obligation requires, is set ourselves to act in ways which will include the cultivation of appropriate attitudes. But he gave serious hostages to fortune when he included such attitudes of mind as love or hate directly in the determination of moral worth. This opened the way to thoughts of choosing to act from a certain motive etc. If we choose in that context it is only in a very attenuated way, for the motive itself is here the determining factor. We have to break away from the context in which our actions flow from motives, in the sense of emotions and desires, to an area of choice which is not so determined if we are to separate the right performance, as Ross seems concerned to do, from what we cannot modify at will on any occasion.

It does not follow that what we actually accomplish, and even more what we set out to do, is without relevance to estimations of our worth in non-moral ways. Directly, the actual effects of our conduct do not affect our quality as agents, for much may come about which we could not be expected to anticipate. But indirectly the course of the events we set in train provides major clues to what we intend to happen or to tolerate when not in itself to one’s purpose. That is a major determinant of what sort of persons we are, although there will be other ingredients. What we are, in various ways in this context, will be good or bad (or indifferent) in various measures just in virtue of what in fact we are, complicated and variable as it is, without our being able to modify what it is directly at any time, as in the obvious cases of intellectual, aesthetic or physical attainment. It is important therefore not to complicate our understanding of such matters, and set ourselves wheeling from one implausible extreme to the other, by importation into them of suppositions which can only be properly meaningful in the context of the sort of choice, in confrontation with a claim of duty, which is not determined by one’s character, or by anything other than the agent’s making it, except in the sense not directly relevant here of helping, in association with our understanding of our duties, to set the occasion for choices of that sort.

The sort of debates which held the centre of the stage at the time the deontologists, including the so-called Oxford intuitionists, were at the height of their influence have, to my mind, a lasting place in ethical thinking. They are much underestimated at present, and we neglect them to our cost. The writers in question had major insights, and their concern to set these out effectively is a source of profound illumination, as much in their mistakes as in their undoubted achievement. It is with a major, but understandable, error that I have been concerned just now. I have been so because I find it profoundly revealing. I have not, however, tried to bring this out in close detail here because I have on earlier occasions2 set out to do just that. It must suffice here to indicate in a general way the confirmation of the main view put forward in this work hitherto in the peculiar difficulties and complications to be found in the body of the philosophical literature to which I have just been alluding.

There is one further meaning of ‘motive’ which has played an important part in recent discussions of freedom and responsibility. It has been implicit in some of the things mentioned already, but it will be well to highlight it further now. This is the idea of motive as the disposition or trait of character of which an action on a particular occasion is the expression. It can be said, in this vein, that someone acted from vanity, fear, respect, affection etc. But, when we speak in this fashion, we are not, it is argued, thinking of causal determination in the normal way. A disposition is not an event which precedes another event and brings it about. It is not a happening, and therefore, in as much as our actions flow in this way from various attributes of character, they are not subject to causal determination and are therefore free in the way required by responsibility.

This has always seemed to me a specious argument. If the invocation of motives, in the present sense, is to have any significance it must be because such motives enter, directly or indirectly, into what we actually do. They are after all tendencies and, while they may be counteracted in various ways, normally by a stronger tendency, it is hard to see what ‘motive’, in the present form, could be or mean apart from some influence which it has. If it is invoked, in explanation of our conduct, it must be in respect of some way in which it is a determinant of conduct. The mode of such determination, however it is analysed, does not take away the sting of the insistence that, to the extent that some particular motive is involved, the action comes about because of that and would therefore have to be what it is when induced in that way.

This leaves it open for us to hold various views about the way a motive, in the present dispositional sense, affects the actual course of conduct, and that will be much affected in turn by our views of the sort of necessity we may ascribe generally to causal relations. If, for example, someone holds generally that causal relations are simply a systematic account of what in fact we find to be the case, with the assumption that things will continue to come about as we find that they do, he will accord his own kind of invariability to our conduct as affected by our motives. If he holds that more is involved, or that there is some more potent and unavoidable linkage in matters where our own understanding comes into play, he will include that in his account of the determination of conduct by motives. But it will be some form of determination in both cases. To the extent that our conduct falls under it, even if only in the way of generalisations, there must be the presumption that a motive which explains our conduct in this way is a factor which, granted that motive, makes our conduct what it is. There is not, in this frame of reference, a motive which falls outside the sort of determination which is thought to apply.

A proper exception would only be found if we were to go, as the libertarian does, altogether beyond the sphere where an action is explained by the invocation of a disposition or trait of character. Otherwise, we have to conclude that, granted the disposition as the explanation of our conduct and, in that way, the exclusion of some counter-acting stronger disposition, the action is made what it is by the disposition manifested in it. To that extent it still remains true that the agent could not help it. To violate that, we have to think of conduct which is not subject to any dispositional determination, the open choice to be found in the circumstances where the libertarian detects it.

There has been much discussion of late of the way a disposition comes to be activated in a particular case. On one view which seems, for the most part at least, to be the view powerfully commended in familiar writings by Professor Gilbert Ryle, there is nothing we can properly note in between the account of the observable physical course of our behaviour and the dispositions which regulate this. The alleged on-going course of mental processes, as the dualist understands them, is dissolved into the attitudes or dispositions which govern the observable conduct and explain it. Doing things on purpose reflects a general readiness to do certain things. The exceptions seem to be momentary aches or tickles and like sensations, and even of these there are desperate attempts to tell the story exhaustively in terms such as ‘inhibited disposition to scratch’ etc. But even if we were to go along with this bold reductionist programme, we would be invoking dispositions, if they are to count at all, as factors in some way regulating or determining what in point of observable fact we do outwardly. In that respect the determinism involved would be peculiarly rigid and depressing.

There are not, however, many who go along with the complete evacuation of the space between dispositions and visible behaviour. The relation of attributes of external things to their actual state or behaviour at any time presents a special problem which need not be considered here — it would be much bound up with the way we think of substance in this context. But in the case of persons, and I would add of other sentient existence and awareness, the disposition is activated in particular mental states in the course of which our own behaviour is maintained. If I am moved to act by love of someone I must have some thoughts at the time of ways in which I may please or serve that person, or some other thoughts and desires coloured by my concern for my friend. My love is not a blind force operating independently of what the course of my experience and understanding is like. This may not on all occasions require explicit thought of the person I love, but my thoughts at the time must owe something to my regard for that person.

Likewise, if someone acts out of fear, there will be some explicit thought at that time of some particular menace or, in the absence of some object or event we particularly dread, a vague thought of some unpleasant fate we may suffer. In the absence of this we would have to invoke some unconscious attitude or motivation. It is not easy to settle how precisely these should be envisaged, if we allow them. There is no particular problem about the possibility that past events of which we have no sort of recollection have shaped our attitudes in ways of which we can give no proper account at present. It is in this way that a Freudian unconscious motivation would usually be understood; and we may also invoke some unconscious feature of our experience in respect of the way a great deal of our passing experience drops rapidly into oblivion — this is how it is sometimes strangely suggested that we may drive a car, for example, in a state of deep abstraction with little awareness of what in fact we are doing. There is no strictly unconscious behaviour in this sense; and normally there would be much more explicit instantiations in our thoughts and conduct of the dispositions which they reflect.

To pursue this further into the wealth of detail which it offers on its own account is not however required for our purpose. It strengthens the position I have been presenting if it is insisted, in reply to those who observe that dispositions are not themselves events capable of functioning as causes in the determination of conduct, that, normally at least, dispositions become effective when they instantiate themselves in actual mental states. The latter can function as causes preceding their effects. If there is some further way in which dispositions affect what we do, then that must still be some sort of determination. That is all that we now require. The insistence on dispositions being determinants of conduct makes no difference, for our purpose, to the unavoidable determinism of the way conduct is thought to flow from one’s character or the sort of person one is at a particular time. Given the disposition, not counteracted by some other trait or peculiarity of our nature, the result within the sphere of this sort of determination is unavoidable.

It is worth noting that some while ago, and at the height of Professor Ryle’s influence, many thinkers who could hardly be thought unmindful of his views were concerned to question the finality of the distinction between the dispositional determination of our conduct and the efficient causality of one state bringing about another. A searching symposium on the subject as far back as 19523 is a fine example. In the course of his contribution, endorsing much in the criticisms of Ryle by the first symposiast. Dr. D. J. McCracken, Professor R. S. Peters leads us to his central theme with the insistence that ‘in the main. Professor Ryle is right in his insistence on the dispositional character of motives but wrong in making a sharp contrast between this sort of explanation and one in terms of efficient causes’.4 He himself supplements the mere law-like correlations of conduct with the notion of ‘a felt organic condition or state of mind’.5 There are goal-directed drives which require ‘certain typical commerces with the environment’ and ‘initiating conditions of drive or tension’.6 ‘Thus a motive-statement is a complicated causal statement’.7 This theme itself calls for further elaboration, and more requires to be said, as it has been by Professor Peters and others, about goal-directed drives and the way, in a full account, our purposes are shaped. But what matters at the moment is the idea of the supplementation of the notion of dispositional determination by a fuller account of various ways in which dispositions are instantiated.

Even Professor Elisabeth Anscombe, in making the distinction between a mental cause and a motive, as a reason for acting, a central theme of her book, Intentions, is most emphatic that there are mental causes. She firmly rejects the idea of motive as a mere ‘law-like proposition’8 and she also assures us that ‘it is worth noticing that what is so commonly said, that reason and cause are everywhere sharply distinct notions, is not true.’9 The main feature, in her own account of motive, is that there should be ‘a response to something as having a significance that is dwelt on by the agent in his own account’,10 and she also speaks of the agent having ‘a reason for acting’, which is his motive ‘if in treating it as a reason he conceives it as something good or bad, and his own action as doing good or harm’.11 This is not easily reconcilable with her equally firm declaration that she is ‘very glad not to be writing either ethics or literary criticism.’12 But even so she seems to be giving us an incomplete story.

This is because there are two ways in which my conceiving something as good or bad may be offered as an explanation for my acting. One of these would be as a justification of what I am doing, a reason for thinking it right. In that case we have to consider it in the full context of a claim of duty which may be opposed to my own wishes, and, in that event, we do not have an explanation of my response in the normal way — the agent makes his choice. But outside this sphere, my deeming something good or significant affects my practice only by affecting my inclinations which, in the absence of counteracting factors, in the case of a moral choice, determine my conduct. Both the way I come to deem something good and the way this in turn enters into my own wishes (and it does not inevitably do so unless goodness is itself conceived in terms of one’s wishes) are themselves shaped by various factors, heredity and my physical and social environment in particular. My evaluations at a particular moment cannot be altered by me at will, any more than my inclinations. We cannot in any final way vary what comes about here.

This is why it seems to me inconclusive, and often misleading, to say, as Professor Anscombe does, that ‘Motives may explain action to us; but that is not to say that they ‘determine’, in the sense of causing, actions’.^12a We need not quarrel seriously about words. There may be a case for a more restricted use of ‘cause’. But, however that may be, the idea of a motive as a reason for action, in explaining or, as Professor Anscombe also has it, ‘interpreting’ an action, can only mean that having the reason brings it about that we do what we do. What she offers us, in line with many others, including Hegelian idealists and those who say similar things in another idiom today, is simply a superior form of determination. But it is determination all the same.

The idea of intention in the sense of what the agent anticipates and seeks to bring about in the course of his conduct, including what expressly appeals to him and both the means to this and what the agent is prepared to tolerate, in distinction from what is in fact brought about in this way, is of great importance in ethics. It is on the former alone, in the way noted already, that moral estimation turns. That is all that we strictly control, it is all that we strictly do. The rest comes about in consequence of this and we intend as we do in the expectation, usually (but not always) high and confident, that what we seek to bring about will be accomplished. The most that we strictly do, one’s action in the proper sense, is the intending. It is for this that we are morally accountable and that also when we freely intend irrespective of our character at the time.

Many writers, however, especially of late, are reluctant to draw a sharp distinction, as I have done, between what we intend and what in point of fact we accomplish. The action, they hold, is or includes the outward behaviour as intended. There is no wedge we can drive between the two. There is just one whole which, in Brian O’Shaugnessy’s term, ‘encompasses’13 the physical change, there is no walking without the movement of my legs. I have criticised this elsewhere,14 and distinguished what we say for rough and ready purposes and what we should say in careful analysis. All I wish to do here is to note what needless complications some notable and careful writers bring on themselves by thinking of an action in this encompassing way, as being the outward performance in certain conditions.

There are intriguing examples in the recent work of Professor Donald Davidson. He has no qualms about accepting some kind of determinism. A number of writers, from Hobbes to Stevenson, have done all that is needed, in his view, ‘to remove the confusions that can make determinism seem to frustrate freedom’.15 This makes it important for him to indicate how our actions, though causally determined, are nevertheless free. This is interlocked for him with the problem of defining intentional action.16 He arrives in due course at the view that ‘to intend to perform an action is, on my account, to hold that it is desirable to perform an action of a certain sort in the light of what one believes is and will be the case’.17 We dispense in this way with ‘mysterious acts of the will, or kinds of causation foreign to science’.18 Indeed, in strictness, we do not need causation either. ‘Unavoidable mention of causality is a cloak for ignorance; we must appeal to the notion of cause when we lack detailed and accurate laws. In the analysis of action, mention of causality takes up some of the slack between analysis and science’.19 We may thus conclude that ‘to say when an agent is free to perform an action intentionally (i.e. with a certain intention) is to state conditions under which he would perform the action; to explain the performance of an action with a certain intention is to say that the conditions are satisfied’.20

This paves the way for Davidson to endorse J. L. Austin’s rejection of G. E. Moore’s analysis of freedom to act in terms of ‘would have’ or ‘could have’ ‘if I had chosen’, for there is no proper sense of choosing which is distinct from the action itself, and if there were we would have to ask about the freedom of that and so on ad infinitum. ‘In order to be eligible as a cause, the event mentioned must be separate from the action; but if it is separate from the action, there is, it seems, always the possibility of asking about it whether the agent is free to do it’.21 But this ‘doesn’t show that a causal “if” isn’t in the offing’,22 this being located in the desires and beliefs that rationalize the action. ‘The antecedent condition does not mention something that is an action, so the question whether the agent can do it does not arise’. Austin is vindicated, but not in the claim that he has exorcised the ghost of determinism as ‘causal theory of action’. For we have still to reckon with the causal conditions that rationalize the action.

I can go along with a great deal of this. There is no willing or trying distinct from the action itself. The search for such an antecendent invites the lampooning to which Ryle was so prone. But the action proper is the intending or setting ourselves to achieve our aim. In conditions in which there is no conflict of duty and interest, in the way noted earlier, this comes about through the way our desires sort themselves out to be what they are at the time. But it does not follow that there is no ‘setting’ or willing distinct from our inclinations and their being as they are. Having the desire is not intending. That supervenes upon the desire, a setting ourselves to achieve what we want; and it is not eliminated, as the core of conduct, in the circumstances where it is inevitable that it supervenes as it does. It is still detectable as the crucial feature of action, and it is hard to see how we could think of action except as it supervenes upon a desire that affects it, even when it does so inevitably upon the having of our desires as such. That is what makes it active.

Willing or intending in this way is not episodic, some irruption that requires the suspension of other activity and which may be noted and considered apart; it is our on-going activity all the time in all that we do. That is why some find it hard to detect; they have wrong expectations. But there is surely something in our conduct over and above our desires and beliefs, the attitudinal conditions which Davidson properly specifies.

Davidson would no doubt agree. But what he seems prepared to recognise beyond the attitudinal conditions is just the outward performance. The intention is the performance subject to the conditions which make it intentional. ‘Thus’, he observes, ‘if a man tries to hit a home run and succeeds, his try is his success, and cannot be its cause’. But, in that case, some curious problems arise. An example from Daniel Bennett is cited. A killer tries to shoot someone, he misses but the sound causes a herd of wild pigs to stampede and trample the victim to death. I do not know what issue of substance turns on this case. Morally, all that matters is what the killer anticipated and sought to bring about. It is doubtful whether anything else would matter even in law. The killer may be content with the outcome, but morally everything turns on what he set out to do — to kill in a special way. Whether, as things went, it could be said that he killed his victim is only important, if at all, as an account of a causal sequence that goes beyond the action proper.

Davidson outlines the proper solution to his problem — ‘Do we want to say that the man killed his victim intentionally?’ — in terms of the causal chain following ‘the right sort of route’. But what is the right route — and the corresponding ‘wayward’ one — apart from conformity with what the agent had in mind and set out to accomplish? Likewise in the example of a man wanting to be rid of the danger of holding another on a rope and being so unnerved by this as to cause him to loose his hold, we would not say that he chose to loosen his hold and did it intentionally. Davidson finds a serious problem here and hints at a solution in terms of a completer statement of the causal conditions of intentional action. But is there not a more straightforward solution in terms of what a person sets himself to do in complete distinction from what in fact he brings about?

Similarly, in the case of ‘intending’ ahead, deciding on Monday to sew a button on Tuesday. The fact is that we cannot strictly do this. I can do many things on Monday pertaining to sewing on the button on Tuesday, including the directing of my thoughts about it. But that is all. I can only set myself to sew the button at the time, although for rough and ready purposes we can talk of intending ahead in this way. Strictly we can only set ourselves to bring things about at the time when it seems to us appropriate, in the ongoing process of maintaining our purposes as occasion seems to require, to set ourselves actually to do so — preparation ahead is another matter.

This is the answer also to a further problem which much engages Professor Davidson, that of ‘pure intending’ — ‘intending without conscious deliberation or overt consequence’.23 It is not easy to determine what this sort of intending could be. It does not seem to be blind impulsive action, or some kind of unconscious activity or some process we suppress. The one example is quite specific — building a squirrel house. This we may intend, in the sense indicated, of intending ahead without any move to implement this. But if there is a genuine intending ahead, there must be some things that we do — if only arranging the set of our thoughts. Otherwise we would be just dreaming or, as we say, thinking about it. There is also a suggestion that ‘pure intending’ is a moment in a fuller process of explicit intending, as in the one other example of intending to write a letter as part of writing a proper word or sentence. My own hunch is that Davidson is aware that there must be more to intending than he has allowed, in noting our desires and beliefs as attitudinal conditions of action. In my own view what would fill the gap is a genuine on-going process of intending. But Davidson will have none of this. The last thing he wants to allow is that there may be ‘a mysterious act of will or special attitude or episode of willing’24 as ‘an embarrassing entity to be added to the world’s furniture’.25

What is provided, to account for intention in terms of judgement and desire as conditions of it, is the supplementation of partial ‘prima facie judgments’ about desirability in some limited respect by a more comprehensive ‘all-out judgment’. But if this is intended to subdue a lurking misgiving that all is not complete in the account of intentional action on the basis of conditioning belief and desires, it seems to be far off its target. For what leaves us uneasy about Davidson’s account is not that the initial story is not complete, but that something radically different has been left out in which our intending essentially consists.

It might be thought, however, that in practice there is not all that difference between Davidson’s view and my own. For I am only interpolating, at the peril of the famous razor, something which makes no difference to the outcome as determined by belief and desire. I reply, firstly that it is important, for proper understanding, that an account should be complete and take due heed of all that is vitally involved. But, secondly, once we move from the area of the confluence of desire and duty and have specifically moral situations and a properly open choice, it is of particular importance to allow the sort of willing or intending which can proceed, in extension to these situations, without determination by belief and desire beyond the way they set the situation for willing in this fully undetermined way.

What is remarkable, about the entire discussion of these questions by Davidson, is the almost total disregard of the way ethical considerations may make a radical difference to the issue of freedom and determinism. At one point he does take up the claim, as presented by Harry Frankfurt26, that ‘an action one is free to perform is an action that one is also free not to perform’. This is easily disposed of by bringing it fully within the main thesis. No one is free to do anything other than is required by the attitudinal conditions which rationalize it. But one could do something different if the action were not intentional.27 And this, it seems, is all that is involved.

There is, however, much more involved if we pass beyond the sphere where our conduct is in accord with our desires. In a properly moral situation, confronted with an obligation to do what in no final way appeals to us, there is a very strict requirement that we should, in a wholly unqualified way, be able to do something other than we do in fact. It is just here that we have the core of the free-will problem, and it is in respect to questions of moral obligation, of properly moral evaluation, that it has been felt that the agent must be the author of his actions, not in a limited way through special conditions, but totally and without any conditions other than the agent making the choice. It is over this requirement that so many have been uneasy down the ages about any form of determinism. But this hardly enters into the recent discussions of which the work of Professor Davidson is typical.

In such a context one cannot but repeat the sad but abidingly pertinent complaint of the late Professor W. G. Maclagan, in his eminently sensible and acute contribution to the symposium on ‘Freedom of the Will’ at the meeting of the Mind Association and the Aristotelian Society at Edinburgh in 195128. He said ‘It astonishes me that there should be discussions of this subject in which reference to the moral “ought” is conspicuous by its absence, and yet there are distinguished examples of this’.29 This is at least as relevant today as when it was written.

We owe to Maclagan also very helpful illumination of the sense in which it can be said that we only try to do what we do. Normally this would be misleading, for what we normally do is decide and maintain our decision; but we may speak of trying in such contexts if there is positive action and also an element of strain. None of this, however, affects the fact that, in actual accomplishment of one’s purpose, the most we can ever do is try or will. This has been challenged in recent discussion by reference to the occasions in which we would say that someone was trying.

When there is no obvious obstacle to what we are doing we would not normally say that someone was trying.30 I walk to my garage, enter the car and turn on the switch to start it and drive back to the door. No one would say in such cases that I had been trying to fetch the car or to start it. We would speak of trying if we thought there was some fault with the car, or with my skill in managing it, which would make it uncertain that I would succeed. It would be some reflection on myself or the car to say that I was trying to start it. Likewise, if I lifted a small stone by the sea to throw in the water, no one would normally say that I was trying to lift it. I was obviously doing so. But if I was struggling with a large boulder which it was not clear that I could lift, and yet not obviously out of the question, one would say that I was trying.

This is plain sailing so far as ordinary usage is concerned. It is misleading to use the word ‘try’ in the contexts where it is not appropriate. But this has no relevance to the claim that all that we do on all occasions is try to accomplish our aims, in the sense that all we can strictly do is will, though normally with the confidence that what we seek to bring about can be brought about in this way. Our ordinary use of ‘trying’ is not made in the context where we are seeking to set out what is involved in our conduct in an over-all way. Some willing, not episodic but sustained, is involved in doing all we do all the time, unless we are to defy the plain facts of experience and reduce our conduct to outward performance or such performance directed by dispositions which are not embodied in the ongoing mental processes we have all the time. To dismiss acts of will in this sense as hidden or preposterous mysteries is to belie what is not a mystery in any such sense but a plain fact of all our experience.

It does not follow that our acts of will have all to be free in the same sense. In much of our conduct the course we take does depend on the way our beliefs and desires shape themselves, the one outstanding and vital exception being the situation created by the deviation of what it is right for us to do from what we want at that time on the whole. This, as I have stressed already, is where we are ourselves supremely the arbiters and masters of what we are and become. But a great deal besides this is further deeply affected by this vibrant and uniquely important moral dimension of our existence.