I PROPOSE next to attempt some discussion of the psychology of moral action, i.e. of any action to which either the epithet good or bad, or the epithet right or wrong, is applicable. For a reason which will appear later,1 I postpone for the present actions done from a sense of duty. It will, I think, be admitted that, with the possible exception of action from a sense of duty, all moral action is motivated by the desire of an end. Every student is familiar with Aristotle's account of action as either being, or (as I think he would rather have said) being immediately preceded by, choice of certain means, which by deliberation have come to be thought of as the means most likely to achieve a pre-desired end;2 and his account is pretty generally accepted. Nevertheless it needs reconsideration. To begin with, it is evident that not every desire of an end sets up a process of deliberation which in turn leads to the choice of means to that end. We have many idle desires which arise in us, and may remain as part of the colouring of our mental life, but never lead to deliberation, still less to action. There are careers, for instance, which one is attracted by, but nothing more; one realizes the attractions, in certain respects, of being Prime Minister or Lord Chancellor, or an England cricketer or a champion golfer, but of all those who are attracted, momentarily or even permanently, by such objects, few do anything about it, even to the extent of deliberating how they could be attained. It would seem that, to be effective as a determinant of deliberate action, an end must not merely be desired, but be chosen.

The difference between deliberate and impulsive action seems to be this, that impulsive action follows directly upon a desire, or upon the strongest of two or more conflicting desires, whereas in the case of deliberate action a choice intervenes between desire and action. This choice or decision is itself a conation, an activity not of intellect but of will; but it is often preceded by a decision in another sense, an intellectual activity of judging that something is the case. With regard to this intellectual activity two questions arise: (1) whether it is a necessary preliminary to choice, and (2) what it is that is judged to be the case. With regard to the first question, I can neither see a priori that such an intellectual decision is necessary, nor be sure by introspection that it always happens. It seems to me possible that there is a semi-impulsive form of action, which is preceded by a choice but not by an intellectual decision. With regard to the second question, two possibilities suggest themselves—(a) that we judge one desire to be the strongest, and (b) that we judge one end to be the most attractive. So far as I can make out by introspection, the latter is the judgement that we actually form. Our attention seems to be directed not to the comparative strength of our desires, but to the nature of their objects. It may be because the desire for A is stronger than the desire for B that we judge A to be more attractive than B, but what we actually judge seems to be that A is the more attractive.

It seems to me further that the attractiveness which we ascribe to ends may be ascribed to them on different grounds. One imagined state of affairs may be thought attractive because it will be a state of pleasure for ourselves; another because it will be a state of pleasure for some one else in whom we are interested; another because it will be a state that we think good in itself. There seems to be no foundation in fact for the view that what attracts us, and what is judged to be attractive, must be an imagined state of oneself, still less for the view that it must be a state of oneself qua pleasant.

If the one decision is an intellectual and the other a conative act, we must not for a moment suppose that there are two entities within us, an intelligence and a will, functioning side by side and independently. We should not choose or decide on the end if we had not made previous judgements about its. nature. We should not decide that it is more worth pursuing than any alternative if we did not start with certain conative tendencies, a tendency in one man, for instance, to a life of effort and in another to a life of ease, in one man to selfish and in another to unselfish activity. Intelligence and will act and react on one another, or rather, since even to say this is to reify them too much, the whole man, in virtue of certain judgements which he has made and certain desiderative tendencies which he has, both judges that a certain end is the most worth pursuing, and decides to pursue it.

We have already seen that to desire is not to choose; we must further realize that to desire an end more than its alternatives is not to choose it. A man may in fact be desiring an end more than any alternative, but he cannot be said to choose it until he performs a perfectly specific new kind of mental act, which can by introspection be distinguished from any sort of desire.

This act may equally well be called an act of decision. Any decision is in fact, even if the word ‘decision’ does not bring out this aspect of it as distinctly as the word ‘choice’, a choice of one thing in preference to all others. The others may be clearly conceived, or they may be very vaguely conceived. In either case preferential choice—of a rather than b or c, or of a rather than anything else—is involved.

It is clear that the mere desire of an end does not necessarily set in train a process of deliberation that leads up to action. Yet it would be wrong to say that an end must be chosen before the process of deliberation on means can be begun. To set up such a process, a strong attraction towards an end, which yet does not lead immediately to the choice of the end against all alternatives, is enough. Any fairly strong attraction, provided there is not some counter-attraction which is both stronger and recognized to be incompatible with it, may set up two processes of thought—(1) a process of thinking out more in detail what the end in question involves as part of its essence or of its necessary or probable consequences, and (2) a process of thinking out the steps necessary for its attainment. If one feels attracted by the idea of being Prime Minister, one naturally sets oneself to think on the one hand of the attractive and the unattractive features of the position itself, the enormous power and prestige balanced by the incessant toil and the crushing responsibility, and on the other hand of the long, toilsome, and often tedious course of action necessary for the attainment of the position. And either of these trains of thought may lead to the desired end not becoming a chosen end; we say ‘that object is attractive when considered abstractly but not when considered in detail’, or again ‘it is highly attractive in itself but it+the steps necessary for the attaining of it form a whole which is unattractive, or less attractive than some more modest career’. Let us suppose, however, that the desired end survives these objections, and becomes our chosen end. Then we continue no further the one train of thought, the thought about the detailed nature and consequences of the end. We know that by further reflection we could learn more about them, but we think this unnecessary; we think we know enough about them to know that this end is more worth pursuing than any alternative—though of course features of the end that come to our notice may later make us change our minds. For the present we have made a definite choice of the end as our end.

So far, it looks as if we should have to give up Aristotle's doctrine, that choice is not of ends but only of means. It looks, so far, as if we ought to recognize the existence of choice of an end (to be followed usually by another choice, which is the choice of certain means to it). But we must consider more closely what it is that we choose. To the word ‘choose’ we may append as object either a noun, or an infinitive preceded by ‘to’; and we see more clearly what it is that we choose if we concentrate on the latter form. That this is justified becomes obvious if we reflect that we have admitted choice to be identical with decision; for with ‘decide’ we can only use the ‘to’ form or what is in principle identical with it, the form ‘to decide on’ a certain course of action.

What then is it that we choose or decide to do? It is not to desire an end. That we must already be doing, for choice to take place. Not, again, to go on desiring an end which we desire already. Not, again, to have or possess a certain end, since that depends on circumstances beyond our control. What we choose or decide is to seek a certain end, i.e. to take whatever steps are expedient to the attaining of the end. To choose is to choose to take means to an end. And the distinction between this choice and the choice which follows is not that the one is a choice of end and the other a choice of means, but that the one is a choice to take whatever means are expedient to the attainment of an end, and the other is a choice to take certain means which we have come to think of as the means expedient to the end. It is only if we think of choice as a certain kind of desire that we shall feel tempted to relate the first choice to the end rather than to the means. But choice is by introspection seen to be a completely different activity from desire of any kind.

A certain amount of deliberation as to particular means has probably taken place before the choice ‘to take the means to a certain end’. For a wise man, at any rate, will not choose so unless he has satisfied himself that the unpleasantness or the unworthiness of the means does not make the end an end not worth pursuing at the price. But it is enough at this stage to have assured oneself that there are means such that the end is worth aiming at by the use of them. The agent may have assured himself of so much, without having assured himself that he has thought of the best means to the end. Thus after the first choice, the next step is a further deliberation, in which the problem he is trying to solve is not, as before, what end he is to take the means to, but what means he is to take to the end he has already decided to pursue.

Aristotle, perhaps only for purposes of exposition, describes this process very simply. He speaks as if we worked, in thought, steadily back from the distant end to be achieved to the means to be taken here and now, step by step, never retracing our steps. What he suggests is this: that we see that the means on which the end A would immediately follow is B; that the means on which B would immediately follow is C; and so on till we see that means Y can be achieved by means Z, which is immediately in our power. Really the process is much more complex. There may be several means which would directly produce end A. If we proceeded in the purely linear fashion, we should decide upon one of these means, presumably in virtue of two considerations—which of them would most probably secure end A, and which of them, if it secured it at all, would secure it in the fullest measure. Means B₁ would be preferred to means B₂, B₃, &c, as the result of a joint consideration of these two points. If, for instance, the probability of B₁'s securing A were ½ and that of B₂'s securing A were 1⁄3, but B₂ if it secured A at all would secure it in a certain degree d, while B₁ if it secured A would secure it in a certain degree d⁄2, then the choiceworthiness of B₁ would be d⁄4 and that of B₂ d⁄3. If we thought this, we should probably forthwith decide on B₂ as the means to A to be adopted in preference to B₁, and our sole concern henceforth would be with the means to be adopted as means to B₂.

These two considerations do undoubtedly come into our choice of means. But other considerations come in to complicate the choice, and to prevent us from proceeding in the purely linear way, from distant end to distant means, to less distant means… to immediate means. In the first place, though B₂ is preferable to all its alternatives in respect of the two considerations named above, there may be features about B₂, e.g. its painfulness to us, or its painfulness to other people, which may make us decide that means B₁ or B₃ is rather to be chosen, even if it be a less effective means to A. Secondly, even if B₂ survived this consideration, a wise man would not choose B₂ unless he had satisfied himself that B₂ can really be reached by means which it is in his power to effect here and now. Thus he must perform the whole process of deliberation in a tentative way, right down to the means to be adopted here and now, before he can properly decide on any of the distant means. And finally, even if he has satisfied himself that he could bring B₂ into being, he will not decide on B₂ unless he has assured himself that there are means by which it could be brought into being, which are not in themselves so repugnant as to make it better to adopt means to B₁ or B₃. Thus, instead of proceeding in linear fashion from distant to near means, we have to run over the whole series (and indeed many times), judging of means at each stage (1) in view of the likelihood of their leading to A, (2) in view of the degree of A which they will lead to, if they lead to it at all, (3) in view of their own attractiveness, (4) in view of the probability of their realization, and (5) in view of the attractiveness of the means by which they can be realized. And what we choose in the end is a line of action which in view of all these considerations applied to each of its stages is judged to be the best.

On this choice there supervenes a mental activity of quite a different type—the activity of setting oneself to bring about the change which is the chosen immediate means to the attainment of the chosen end. The distinctness of the two activities is often obscured by the fact that the second follows immediately on the first. When little time elapses between the arising of the situation and the time at which action must be taken if it is to be taken effectively, exertion follows on choice so rapidly that the difference may escape a hasty introspection. And the difference is also obscured by the fact that we tend to speak of both kinds of activity under the colourless name ‘acts of will’. Let us drop the colourless and indefinite phrase and ask ourselves whether we cannot by introspection distinguish the act of setting oneself to bring about a change, as entirely different from the choosing or deciding to bring it about. Not only does introspection show them to be quite different in character, but a considerable time may elapse between the one and the other; for a decision may be a decision to do something a minute, an hour, a day, a year, indeed any length of time later, or may be a decision to do it ‘sometime’, while a ‘setting oneself is a setting oneself to bring about a change forthwith.

The train of means which we choose as the best way of attaining the chosen end may be of either of two kinds. In the simpler kind, it is only in the bringing into being of the first link in the chain that our own activity is involved. We launch a single action, as it were, into the world, and trust to external circumstances to do the rest. Of this type of case psychology has, of course, nothing more to say. But more often further action on the part of the agent will be needed if his chosen means are in fact to lead to the chosen end. In the normal case, there will be a series of self-exertions which will reproduce in reverse order the series of choices which worked back from the end to the means to be adopted here and now. But in fact the normal series of actions contemplated by Aristotle often fails to take place. We have already seen that the process of deliberation does not proceed smoothly in one line from end to means, but that a choice of end is only provisional until there has been some consideration of means as well, and that the scale of end and means may be run up and down many times before either end or means is finally chosen. There is, however, another fact which makes our actual deliberation often depart still more widely from the Aristotelian model. Aristotle speaks as if desire for an end were the only possible starting-point of deliberation. But we are perfectly familiar with cases in which deliberation starts at the other end, by some action being suggested to us, and by our going on to consider what effects it is likely to produce. There are countless cases in which we are advised, requested, begged, or commanded to do some action, or it is hinted to us that we might, or we are asked whether we are going to do it. In all these cases the thought of the immediate action is in our minds before any thought of a distant end; the stimulus to deliberation comes from the opposite end to that contemplated by Aristotle. And apart from these very obvious cases of suggestion by another person, it constantly happens that something in our environment suggests to us the doing of some action here and now, and deliberation takes, not the Aristotelian form of asking what means will produce such-and-such an end, but that of asking what effects such-and-such an action will have.

Human beings might be divided into two types—what may perhaps be called the planning type and the suggestible type—according as the train of thought leading up to their choice of action habitually starts from the thought of an end and works back to means, or starts with the thought of an action suggested by the circumstances, and goes on to consider with more or less thoroughness its probable consequences. I do not suggest that these types are cut off from one another with a hatchet. Every one sometimes reasons in the one and sometimes in the other way; but some people tend to reason in the one and some in the other way. We have, on the one hand, the people with more or less settled purposes in life, who scrutinize possible lines of action, in all important matters at least, as lines of action tending to conduce to or to be unfavourable to the attainment of their purposes. We have, on the other hand, those who live more from day to day, acting more or less on the suggestion of the moment and as a rule not thinking out to the end the probable bearing of their actions on any ultimate purpose. The distinction is, of course, by no means the same as that between good men and bad men. The purposes of the first type may be mean and selfish, no less than in other cases they are high and unselfish; and the impulses of the second type may be generous or they may be low and narrow. It is rather the distinction between strength and weakness of character than between goodness and badness. It may safely be said that the men who make history, for good or for evil—great statesmen, for instance, and great soldiers—are men who by fixity of purpose correspond to the first type, and that the great mass of mankind corresponds to the second.

As a result of this to-and-fro movement of thought, partly working from desired ends to the means most likely to secure them, partly working from suggested actions to their probable effects, a decision is usually arrived at. Not always; for an indecisive mind, or any mind when faced with a situation of special difficulty, may prolong the time of deliberation till some more urgent problem supersedes the original problem. But suppose that a decision is taken. The circumstances may be such that the corresponding action, if it is to be done at all, must be done at once; and then the decision is immediately followed by the act decided on. Or they may be such that an interval should be allowed to elapse before the action is initiated. Or again, they may be such that, so far as we can see, it makes no difference whether the first practical step be taken at once or after some time. It is an interesting question whether, in the two latter cases, a renewed decision has to be taken before the act of ‘setting oneself can be done, or whether the momentum of the original decision is enough, as it were, to carry us on to the first practical step. This is a question to be decided by introspection, and I am not certain what the answer is. But so far as I can judge from observing the working of my own mind, I am inclined to think that the momentum of an original decision will only carry over a very short interval. Suppose, for instance, that I waken at seven, and decide to get up in half an hour. I find that at the end of the half hour I do not automatically set myself to get up, but have to renew my decision.

Whether a renewed decision is always necessary or not, it seems that it is usually desirable; for between the original decision and the time for action the circumstances may have so changed as to make the action decided on no longer the rational one to take, with a view to gaining one's end. Not that the original deliberation need all be performed over again; it may be enough to ask whether the circumstances have so changed. If we think they have not, the original decision should simply be renewed. If we think they have, the hard mental work of deliberation will not have, as a rule, to be done all over again; for usually the situation will not have been revolutionized but only changed in some details. Some probable consequences of any of the alternative actions will remain unchanged, and it will only be necessary to consider those that are changed.

Special interest attaches to the case in which it does not appear to matter when, within certain limits of time, the action decided on is to be taken. To reach the conclusion that it does not matter, we must satisfy ourselves of two things: (1) that, if the relevant circumstances do not change unexpectedly, an action done at any time within the limits is as likely as an action done at any other time within the same limits to produce a certain set of results, and (2) that the relevant circumstances are very unlikely to change unexpectedly, or that if they do change, they are not likely to change in such a way as to make any other action more effective. It is, of course, very hard to assure oneself on the latter point, and if we make this assumption, it must be in the main because we have to admit our almost total ignorance of whether the relevant circumstances are likely to change, and if so, how. When we are in such a state of ignorance, the assumption is a reasonable one to make. Why, then, does common sense prefer the man who ‘does it now’? By not doing it now, we run the risk that the circumstances may change so that the action decided on will become impossible or less effective, and we shall be left only with a choice between actions less effective than it would have been; but it is equally possible that the circumstances may change so as to make possible some other action more effective than it. This possibility also is recognized by common sense in the maxim ‘don't cross your bridges till you come to them’. If common sense on the whole prefers the man who carries out his resolutions speedily, that is to be justified on two grounds: (a) that if you put off action till near the end of the time available, unforeseen accidents may make it impossible for you within the time limit to carry out your policy of action, or to carry it out with the proper amount of care at each stage, and (b) that action which is resolved upon with some enthusiasm, when its linkage with the ends you desire is clearly recognized, is apt to become less attractive when that linkage has receded into the background of memory and the disagreeable features of the action have come into the foreground of attention; so that there is a danger that it may not be done at all if it is put off too long.

I have tried to give an account of the process leading up to and including a fully deliberate act, the kind of act which we are usually thinking of when we speak either of the rightness or of the moral goodness of acts. We have seen that in this process there is involved not only the activity of setting oneself to bring about some change, but the previous activity of resolving so to set oneself. Moral philosophy usually speaks as if it were the former alone that is the subject of moral judgements; but it is surely clear that resolutions also are manifestations of moral character, that they, no less than acts, can be morally good or bad. It seems to me also that they, no less than acts, can be right or wrong, obligatory or the reverse. But this is much less clear than it is that they can be morally good or bad. The view that they are right or wrong might be attacked on either or both of two grounds. (1) It might be said ‘what matters is that we should set ourselves to do certain things. Now, we can set ourselves to do things without having previously resolved to do so, and if so we shall have done our whole duty. The previous resolution, even if it is a manifestation of good character, is superfluous and no part of our duty.’ We have, then, to examine whether it is possible to ‘set oneself’ without previous resolution, and even without previous deliberation, without which I take resolution to be impossible. There is no doubt that this can happen. If some one asks us a question, for instance, we usually reply truly, without deliberating whether we shall, or resolving to do so. The momentum, as it were, of our character and habits carries us straight into the right act. Yet there remain two classes of cases in which previous deliberation and resolution seem to be included in our duty. The one is that class of case in which various prima facie obligations conflict, so that it is not at all clear what our duty is, and only careful deliberation will reveal this. He would be a foolish man who trusted to the inspiration of the moment to yield the right action in such cases as these. The other is the class of cases in which, while our duty is clear enough, we have a strong inclination to do something else. A man may, for instance, have a strong ingrained habit of telling the truth when his own interests are not involved; but this habit cannot be trusted to carry him straight to the telling of the truth when he stands to lose personally by doing so. In both these types of case the right action is in fact usually preceded by deliberation and resolution, and there can be no guarantee that it would be done if they were dispensed with.

(2) The other objection that can be made is this: ‘Even if resolution is in such cases necessary as a condition of the right action, why resolve now to do my duty to-morrow, or next week, or next year? Why not leave the resolution till just before the time comes for the action?’ The answer to this is that if we do not now put ourselves into the state of mind called intention (and resolving is just putting ourselves into this state), we shall be very likely to do meantime, on the suggestion of the moment, or in pursuit of other aims, things which will prevent us from doing, when the time comes, what it would be our duty to do if we could. Take, for instance, a debt which I ought to pay, but am under no obligation to pay till next year. If I do not now resolve to pay, and thus keep the duty of payment before my mind, I may easily be led into expenditure which will make it impossible to pay when the time comes. Or again, it is a man's duty to choose a career some time before the occasion arises for taking the first practical step in pursuit of it, because if he does not, he will very likely be led into actions which will make it impossible to pursue the career which but for them it would have been his duty to pursue.

The duty of resolving, then, seems to arise when either what is our duty is not perfectly clear, or there are desires which militate against the doing of it; and in either case it is a duty arising from the primary duty of acting in a certain way.

When I said, at the beginning of this chapter, that all moral action is motived by desire, I admitted one possible exception, viz. action from a sense of duty. Kant said that action from a sense of duty is not motived by desire at all—that its only preconditions are the knowledge that a certain act is one's duty, and the emotion of respect which that thought arouses in us. I do not think we can say a priori that this is impossible. It would seem a priori possible that an emotion may serve as the precondition of action, no less than a desire. But whether Kant was right or wrong in his view, it seems clear that his reasons for holding it were insufficient. He held it because he assumed that desire belongs altogether to a lower stratum of our nature than reason, a purely animal stratum to which no worth can attach. This complete degradation of desire is not justified. It is clear that, quite apart from a desire to do our duty, we have many desires which we could not have if we were not rational beings. Take, for instance, the desire to understand. It would be absurd to say that this belongs to the purely animal part of our nature. On the contrary, it springs directly from our possession of reason; and if it has developed continuously out of animal curiosity, I would rather say that animals, if they have curiosity, must have some spark of reason, than that curiosity must be irrational because animals can have it. And there are many other desires—the desire, for instance, to follow a certain career or occupy a certain position—which we should never feel unless reason had been at work, apprehending the nature of human relationships and the consequent desirability of such a career or such a position. We are, in fact, not limited in our choice to Hume's view of reason as the slave of the desires and Kant's view of it as their inveterate foe; many of our desires owe their very being to our possession of reason.

Kant's distrust of desire leads him to hold that all actions springing from desire are quite lacking in moral value—that an action done from kindness or love, unaccompanied by the sense of duty, is worth no more than the most selfish or the most cruel action. We can agree with him in thinking that the sense of duty is the highest motive, without following him in putting all other motives on the same dead level. Kant simplifies the moral life too much in making it a contest between one element which alone has worth and a multitude of others which have none; the truth rather is that it is a struggle between a multiplicity of desires having various degrees of worth.

If it be granted that we have desires that spring from our possession of reason, it is only natural that there should arise a desire, itself springing from our rational apprehension of principles of duty, not to be the slave of lower desires but to regulate our life by these principles. To say that conscientious action springs from such a desire, the desire to do one's duty, is in no way to degrade conscientious action, as Kant thought it would be. And in fact, when I ask myself why I do my duty (when I do it, and do it conscientiously), the truest answer I can find is that I do it because, then at least, I desire to do my duty more than I desire anything else.

The question now arises, whether conscientious action can, like other deliberate action, be described as the adoption of means to an end. An answer which naturally suggests itself is the following: ‘From one point of view, conscientious action is not the adoption of means to an end; for in so far as one's object is to do one's duty, this is something that is already achieved in the immediate act of self-exertion. The doing of one's duty may, indeed, be described as an object of desire, but not as a desired end, in so far as speaking of an end implies that the end is something to be achieved by the use of means. Yet this is not the whole truth, as we may see by considering some particular conscientious act. Suppose that someone conscientiously sets himself to relieve another person's pain; his desire to do his duty will be satisfied by his exerting himself in the way that seems to him best for the purpose, but he will not be completely satisfied unless the means adopted actually secures the relief of the other's suffering. He has two desires, the desire to do his duty, which is, of course, satisfied by his doing his duty (i.e. by his exerting himself in what seems the best way of helping the other man), and the desire to achieve the relief of a suffering person, which will be satisfied only if the means adopted turn out effective.’

This is, however, not the true account of a purely conscientious act, i.e. of one in which the only effective motive is the desire to do one's duty; it is an account of an act in which there co-operates with this desire an independent desire, springing from natural kindliness, to relieve the other man's distress. In a purely conscientious act (and in using the word ‘purely’ I do not mean to suggest that such an act is superior to one springing from the combination of the two motives, but only to call attention to the singleness of the motive3), there is no wish to promote the relief of another man's suffering, independent of the thought that it is one's duty to do so. Suppose the agent has decided that the greatest claim on him in a given situation is the claim of a sick man to his help. Then he has but one primary desire, the desire to do an act of self-exertion which will result in relief to the sick man, as being the doing of his duty. And this desire will not be satisfied unless his action has the effect in question. What he desires primarily is to do his objective duty. But, knowing the difficulty of knowing what act will be the doing of his objective duty, i.e. will produce the effect in question, he has also a secondary desire, the desire to do his subjective duty, i.e. the act which he thinks to be his objective duty. And he can derive some satisfaction from the doing of this even if his primary desire fails to be satisfied.

Thus conscientious action is the adopting of means to an end—not, however, because the end is desired, but because the self-exertion which will bring about the end is desired, as being the doing of one's duty.