PERHAPS the severest criticism with which the view I am maintaining has met is that which is passed on it by Mr. W. A. Pickard-Cambridge. I quote a passage which shows the general nature of his criticism.
‘It is evident at once that a theory which hides the process which it professes to defend, at the most critical point, in a cloud of negations, is not easy to attack convincingly. Anyone who maintains that we decide what is our duty by comparing together prima facie obligations and picking out one of them as the more stringent by inspection only, and on no principle at all, puts himself at a double advantage.
‘It is very difficult to subvert or to prove inappropriate any method which rejoices in working on no constant principle—as difficult as it is to confute a man who does not mind how much he contradicts himself. Such a man retains a power to strike at his opponent who does believe in some constant, universal principle of judgment, if he can only find a single case where the latter is inconsistent in his principles. On the other hand, he robs the opponent of any chance to strike at him with a like charge of inconsistency, because inconsistency of principle is, in his view, the normal and natural feature of all moral judgment. It is impossible to correct him at any point (as it is often easy to correct anyone who admits the existence of some one universal ground or character of our duties) by reference to his own judgment or to the principles he has employed elsewhere. He can always get out of the net by saying “I know that principle B on which I act to-day not only does not support but actually collides with principle A on which I acted yesterday: but that is as it should be, because an unprincipled intuition makes me of opinion” (or, in shorter American phrase, “I guess”) “that to-day B has the stronger claim”. Or again, if such order of priority between certain kinds of prima facie duty as the theory claims to be self-evident is ever reversed on any particular occasion (e.g. if ever it is believed to be right to give a present rather than pay a debt, when both cannot be done) the defender of the theory can always here again say that his intuition leads him in this case to this conclusion, and that that for him is sufficient. By thus professing to discard any consistent principle of judgment save to abide by his intuition at the moment, he puts himself outside the reach of any appeal to maintain any consistent principle of conduct—a form of appeal which is usually accepted and predominant in moral as in all other discussions. Argument, then, on these lines is impossible. The need for consistency in one form, indeed, he readily admits, viz. for consistency in judgment between different people. If he judges truly at any point, then anyone else ought to judge the same. If they disagree, one must be in error, and he will admit that the error may be his: his opinions on all such matters are “highly fallible”, never more than “probable”. This candid admission saves his judgments from the fate, to which the total disclaimer of consistency in any form would expose him, of being the expressions of pure caprice and so totally insignificant. But it still affords no possible basis for argument, nor opens any way to an agreement between the parties (except, of course, by a spontaneous and equally unprincipled change of opinion on one side or the other) upon the question which of them is in error and which not.’1
I find it difficult to seize what is the precise point of my theory which is here being attacked. But perhaps the most definite statement made is that in my view inconsistency of principle is the normal and natural feature of moral judgement. It is true that what distinguishes my view from Mr. Pickard-Cambridge's is that I believe in a plurality of moral principles, while he believes in but one, that we should always do the act which seems likely, and do it because it seems likely, to produce most good. This appears to me to be clearly not the only principle which we recognize. But why should recognition of a plurality of principles be thought to involve inconsistency of principle? That will depend on the nature of the principles. If my principles were, for instance, ‘I should always keep a promise’ and ‘I should always do that which seems likely to produce most good’, these principles, while not formally inconsistent in themselves, would sometimes involve inconsistent consequences; for sometimes the keeping of a promise does not seem likely to produce most good, and then the one principle would lead to the consequence ‘I ought to keep the promise’ and the other to the consequence ‘I ought to break it’. This is precisely the difficulty into which a view like Kant's falls. But I have tried to avoid this objection, by stating the two principles quite differently—the one in the form ‘an act, in so far as it is the fulfilling of a promise, tends to be right’, the other in the form ‘an act, in so far as it is the act which seems likely to produce most good, tends to be right’. And neither these principles, nor any consequences to which they lead, are inconsistent. What, then, is the justification for representing me as saying ‘I know that principle B on which I act to-day not only does not support but actually collides with principle A on which I acted yesterday: but that is as it should be, because an unprincipled intuition makes me of opinion… that to-day B has the stronger claim’? Surely Mr. Pickard-Cambridge himself must admit on his own principles the existence of conflicting tendencies in the same act. An act which seems likely to do good to A and harm to B tends to be right in the first respect and wrong in the second. And there would be no inconsistency in his saying that though one act which does good to A and harm to B is on balance right, another which, while doing an equal good to A, did more harm to B may be wrong. And similarly there is no inconsistency in saying that though an act which keeps a promise to A and does harm to B is right, an act which keeps a promise to A and does more harm to B may be wrong.
Next, I must examine Mr. Pickard-Cambridge's claim2 that the ideal utilitarian method is, compared with the intuitionistic, one which it is easy to apply in order to ascertain our duty. To this I would make two answers. (1) The claim seems to me to be justified as regards some moral situations, and unjustified as regards others. (a) Consider first a case in which we are really in doubt whether we ought to fulfil a particular promise, or to do an act which seems on the face of it likely to bring more good into the world. In such a case I believe the utilitarian method is one which it is easier to apply. For my view involves me in saying that in such a case we have to take account, just as a utilitarian would say we must take account, of the goodness and badness of the various consequences of either act. We are alike involved in this difficult evaluation. And, in addition, I am involved in the further problem of evaluating the prima facie obligatoriness of the one act qua fulfilment of promise with that of the other act qua productive of good; and this is in border-line cases a very real further difficulty. But (b) life is not full of such border-line cases. When we ask ourselves whether we ought in a given case to keep a promise, then if we hold as I do that there is a strong prima facie obligatoriness attaching to the fulfilling of a promise, which therefore no slight preponderance of good to be effected by an alternative act can prevent from being my duty, I am saved from the task of narrowly scrutinizing the situation to see whether there might not be some slight preponderance of good effected by breaking my promise. Accordingly, in perhaps nine out of ten cases in which I have given a promise, I have no difficulty at all in seeing that I ought to fulfil it; while the opposite view would involve me in anxious scrutiny of every case in which I have made a promise, before I could know that I ought to fulfil it. Putting it briefly, the utilitarian method is the easier to apply in difficult cases; but in return, utilitarianism would make practically all cases in which a promise is to be kept or broken seem difficult cases—even those which our natural moral consciousness informs us not to be so.
My main answer, however, would be a different one, viz. (2) that any appeal to the ease of applying the one view or the other is beside the mark. It is not the business of moral philosophy to provide us with a theory which is easy to apply. Its business, or the part of its business with which we are at present concerned, is to say on what the rightness or wrongness of actions in fact depends. The fact that it would be easier to recognize our duty if it depended on factor a only than it would be if it depended on factors a, b, and c has no tendency whatever to prove that in fact it depends on a alone. The kind of argument which Mr. Pickard-Cambridge turns against me could equally well be turned against him by a hedonist. A hedonist could fairly say that his own theory, which recognizes in pleasure the only intrinsic good, is easier to apply than one which recognizes, as Mr. Pickard-Cambridge's does, other types of good as well, and requires us to evaluate goods of different types against one another. To him Mr. Pickard-Cambridge would rightly reply: ‘That may be so, but the fact is that we think other things than pleasure good, and it is our business to take account of the facts and not to adopt a relatively easily applied theory because it is relatively easily applied.’ And similarly I would reply to Mr. Pickard-Cambridge: ‘The fact is that we recognize other grounds of rightness than productivity of good results, and the theory which points this out is the true theory, even if it is in many cases harder than the rival theory is to apply to the question what I should do here and now.’
We must, however, see whether his system will square with the facts of the moral consciousness. In particular, we must consider further the question of promise-keeping. Mr. Pickard-Cambridge of course admits that, in general, promises should be kept. He gives his reason for this by pointing out various goods which are produced, and evils that are avoided, by the keeping of promises. His analysis is as follows:—We may start with the evils produced by the breach of promise, viz.:
- The loss by the promisee of the promised benefit.
- The disappointment caused to him by the breach of promise.
- The damage to the promiser's reputation, and therefore to his power to co-operate with others, and therefore to his power to do good.
- The damage to society caused by his weakening the general confidence that promises will be kept.
If the promise is kept, on the other hand, the following goods will be produced:
- The gaining by the promisee of the promised benefit.
- The pleasure caused to him by the fulfilment of his expectation.
- The heightening of the promiser's reputation for trustworthiness, with the greater power to do good which this will bring in its train.
- The benefit to society caused by his increasing the general confidence that promises will be kept.
It is quite clear that the preservation of general mutual confidence between man and man is a source of very great good to a society. If men could not generally rely on others’ keeping their promises, commercial credit, for instance, would break down, and society would be conducted at a great disadvantage. Utilitarians, therefore, have naturally laid great stress on this, and have put forward, as perhaps the main reason why an individual promise should be kept, the fact that every breach tends to weaken this general confidence. I ventured to say elsewhere3 that ‘it may be suspected… that the effect of a single keeping or breaking of a promise in strengthening or weakening the fabric of mutual confidence is greatly exaggerated by the theory we are examining’ (the utilitarian theory), and Mr. Pickard-Cambridge takes me to task for this. He argues that the situation immediately suggested by my words (‘a single… breaking of a promise’) is not the situation we have to think of. My words, he says,
‘suggest a single breach of promise occurring as an exception to the general practice, and in face of a general opinion that such breaches are wrong. What we have to ask is what would happen if this general opinion were scrapped, and if a breach of promise could appear in public with no halter round its neck, admired as a shrewd device, or at least approved as quite unobjectionable and lawful. Imitators would pour in as thickly as applicants for the dole.… In six months or less credit would be as dead as national solvency with a universal dole’.4
His reasoning here appears to me to be at fault. We must distinguish two questions. One is the question whether condemnation of promise-breaking by public opinion is justifiable on utilitarian grounds? I am willing to admit that it is. Undoubtedly general condonation of promise-breaking would lead to general breaking of promises, and undoubtedly this would lead to results which a utilitarian would deplore. But the question I was asking was a different question, viz. whether it is by its results that a particular act of promise-breaking is made to be wrong, or by its intrinsic nature as a breach of promise? If I were tempted to break a particular promise, I, if I were a utilitarian, should not be entitled to ask ‘what would happen if promise-breaking were generally condoned’, except in so far as I thought this particular breach of promise likely to lead to the general condonation of promise-breaking; for particular acts must by a utilitarian be judged only in the light of their probable consequences. Now I do not think that generally speaking a particular breach of promise is likely to have much influence in promoting the general condonation of promise-breaking. Is it not rather the case that it will be condoned only in so far as public opinion is already weak or perverted on the matter of promise-breaking? And is it not the case that a flagrant breach of promise often tends to strengthen the general opinion against promise-breaking, so that so far as this effect goes a utilitarian should welcome and approve it? I should say, for instance, that the breach of Belgian neutrality by Germany in 1914 strengthened rather than weakened in all impartial observers the sense that treaties ought to be observed. ‘Recent history in Ireland or Chicago’, says Mr. Pickard-Cambridge, ‘sufficiently illustrates the situation which arises if murder be unsuppressed.’5 But that is an argument not against the doing of a particular murder, but against the general condonation of murder by the state.
So far Mr. Pickard-Cambridge seems to have fallen into an ignoratio elenchi. He has confused the question, What are the probable effects if I break this promise?, with the question, What are the probable effects if I do this and my act is generally condoned?; and since the bad effects in the latter case will be very great, and enough to condemn the act on utilitarian grounds, he assumes that in the former case also they will be so.
I pass to a part of his case which is more difficult to meet. He has shown great ingenuity in putting cases which do present some difficulty to one who believes in a prima facie duty to keep promises. (1) He puts first6 the case of two sick musicians, A and B, of whom A has promised B that before 5 p.m. to-day he will put a new E-string on B's violin. If at 4.45 B is evidently dying, is A still bound by his promise? I agree, as I suppose every one would, that he is not. But my comment would be that Mr. Pickard-Cambridge in using this case as an argument against my view is pressing on the letter of the promise and not on its spirit. I ask for some common sense in the interpretation of the promise. A's verbal promise is: ‘I will put a new E-string on your fiddle.’ But the underlying assumption is that B will be well enough to play on the fiddle. In spirit the promise was: ‘If you are alive and well enough to play on the fiddle, I will put it, in respect of its E-string, in condition for you to play on it.’ The admission of such a qualifying clause would be dangerous if the qualification were one made secretly by the promiser to himself at the time of the promise, or extemporized by him afterwards in his own interests; but in this case it is one which, if it had occurred to A to make it openly at the time, would have been accepted by B, and which implicitly determines the nature of the understanding between the two men. Suppose the assumption were different: suppose that A had broken the string of B's violin, and his promise to B were implicitly a promise to make good this slight financial loss, he would be still bound to prevent this loss to B's heirs.
(2) Secondly,7 he takes the case in which it is A's illness that takes a turn for the worse. If A is in pain and can hardly bear to move, then, says Mr. Pickard-Cambridge, he is under no obligation to fulfil his promise. Again, I would ask that the promise be interpreted sensibly. The expectation of both men is that when the time comes to fulfil the promise the circumstances will not be very different from what they are. It is on this assumption that the promise is made, and that it is received. If, when the time comes to fulfil it, A is as ill as Mr. Pickard-Cambridge supposes, he could fairly say to B, ‘Here is a change of circumstances which neither of us contemplated. The assumption on which the promise was made being falsified, do you consider my promise still binding?’ Suppose, on the other hand, A's promise had been: ‘However ill and weak I am, I will do this for you if I can’, then we should consider him still prima facie bound to fulfil the promise if he can; whereas here also Mr. Pickard-Cambridge would apparently say that, although A's being very ill and weak was a possible contingency explicitly named in the promise, yet since, when the time comes, he is ill and weak, so that the pain caused to him by the fulfilment of the promise would outweigh the good it will do to B, there is no obligation to fulfil the promise. Now this is a case in which, since the promise is probably known only to the two persons concerned, the sapping of general mutual confidence by general condonation is not a consequence that need be considered. The only consequences that a utilitarian need consider are the pain caused to A by his fulfilling the promise, and the pain caused to B8 by its not being fulfilled; and Mr. Pickard-Cambridge's view evidently is that, where other consequences are negligible, and A's pain would be greater than B's, A has no obligation to fulfil the promise. On this view, then, whenever A makes a promise to B, he is to be understood as saying, ‘I will do so and so if and only if my doing so and so will cause me less pain or inconvenience than my not doing so will cause to you.’ Mr. Pickard-Cambridge claims constantly that his view agrees better than the opposing view with the ordinary consciousness, but I think that this is not how the ordinary consciousness understands promises. I think we should want to know, when some one makes us a promise, whether it is an ordinary promise or a Pickard-Cambridge promise that he is making.
(3)9 I have promised (a) to play at a concert, (b) to send my score, (c) to send my partner's score. I fall ill, and it is clear that I shall not be able to play. Promise (a) therefore ‘must be broken, because its fulfilment is impossible’. But, according to Mr. Pickard-Cambridge, I (on my theory) ought to maintain that it is still my duty to fulfil promises (b) and (c), which I still can fulfil. Again I ask for common sense in the interpretation of the promise. My verbal promises were as stated, and if I am to be held to the letter of my promises I am guilty of breach of promise if I break any one of them. But it is clear that the promise to play was subject to the tacit condition ‘if I am able to play’, and that the other two promises are subject to the same condition, and cease to be binding when the assumed condition is not fulfilled.
(4)10 I promise (a) to call at X's house at 2 p.m. to-morrow and (b) to go for a walk with him. He falls ill, so that promise (b) cannot be fulfilled. Mr. Pickard-Cambridge says that on my theory I ought to maintain that promise (a) is still binding. But it is surely perfectly clear that the assumed condition of both promises is that both X and I are well enough to go for a walk, and that, this condition being unfulfilled, the prima facie obligation to fulfil either disappears. If, on the other hand, the promise to call at his house were independent of the assumed condition that he is well enough to go for a walk, it would still hold good. Once more Mr. Pickard-Cambridge is insisting on the letter of the promise and forgetting the spirit.
(5)11 The vicar of my parish asks me to send my piano to the parish hall next Thursday afternoon. I promise categorically. I afterwards learn that he is arranging for a concert, and later that the concert is off. Mr. Pickard-Cambridge says I ought (on my theory) to think myself still bound to send the piano. But it seems to me clear that though categorical in words, my promise was made and accepted on the tacit understanding that my piano would serve some purpose of the vicar's, and that when this condition is known to be unfulfilled the prima facie obligation disappears.
(6)12 A rich miser visits me in forma pauperis, and extracts a promise to pay him £100 in six months. ‘If I discover the fraud in time, will anyone hold that my promise binds me?’ The case seems to me a difficult one. I would deal with it tentatively as follows. If he had promised to do me a service, and I had promised to pay £100 on his doing so, it will be agreed that my promise would cease to be binding if its condition were not fulfilled by him. Now, in the actual case supposed, my promise to pay arose out of conversation with the miser, which was conducted under the implied contract to tell each other the truth. In breaking this contract, he has destroyed the basis on which my promise was made. It is as if I had said, ‘If you are telling me the truth, I will pay you £100’, and since he is not telling me the truth, the condition on which the promise was made is unfulfilled, and I shall not in spirit be breaking my promise in refusing to pay.
Suppose, however, that I do pay, is it not clear that he is under a very strong prima facie obligation to return the money—an obligation which good men would regard as in most cases overriding the question whether he or I or another man would make the best use of the money? Whatever we may think, in this difficult case, of the prima facie obligation to keep promises, does it not form a very clear case of another prima facie duty, that of restitution for injury done?
(7)13 A poor man extracts from me a like promise, and comes into a fortune before the six months have elapsed. ‘Am I bound to pay? Very few would think me bound to do so.’ The case does not seem to me so easy. I think that a man with a very delicate sense of honour would consider that he ought to pay for his own carelessness in making the promise unconditionally. But alternatively it might be urged that the promise was in spirit a promise to pay £100 to a poor man, and ceased to be binding when the poor man has become rich. In either case the prima facie obligation to keep promises does not seem to be placed in question. And surely, whatever we may think that the promiser ought to do, we should agree that if he does pay, the other ought to return the money. The case throws into high relief the prima facie duty of making a return for benefits received. So obvious is this that we should not think it necessary for the recipient of the £100 to consider anxiously whether he, or his benefactor, or some other person would make the best use of the money. He should send it straight back to his benefactor.
It would be tedious to go through all the cases put forward by Mr. Pickard-Cambridge; I have, I believe, said enough to indicate the lines on which they should be dealt with by one who believes in a prima facie obligation to fulfil promises, distinct from the obligation to seek the general good. But it may well be thought that I have been rather vague, in appealing to the spirit of a promise against its letter. ‘A promise’, it may be contended, ‘is a statement that one will do something, and is not fulfilled unless this thing is done.’ My reference to the spirit of the promise is simply a reference to the obvious fact that most promises, like most statements of any kind in ordinary life, are made without any attempt, and without any necessity, to state in full all their implied conditions and qualifications, since conversation would be very tedious if all of these were insisted upon. Between two men of good faith there will usually be agreement as to the unexpressed conditions to which a promise is subject; the spirit of the promise is perfectly well understood. To argue that one who takes the non-utilitarian view of promises is bound to interpret them according to the letter and not to the spirit is really to misinterpret the nature of ordinary speech.
Of course the presence of unexpressed conditions to which a promise is subject has its dangers; and in particular the presence of the very vague unexpressed condition ‘if circumstances have not become very different’. Those who wish to break a treaty, for instance, are usually able to say with truth that conditions have become very different since the treaty was made. There are various ways in which this difficulty can be dealt with. One, of course, is to make every effort to state the conditions explicitly instead of leaving them to be tacitly understood. But however much care be taken, it is impossible to foresee all the changes of circumstances which might make the maintenance of a treaty unreasonable. No treaty, therefore, should be made binding for ever; every treaty should either be for a definite period, or should contain provisions for its own possible revision under suitable arrangements. The difficulty is not so serious in the case of most promises between individuals. As a rule, these are promises to do something within a comparatively short time, in which there is less danger of a change of conditions which will make the fulfilment of the promise unreasonably onerous. But it is plain that even in such private promises it would often be well to express the conditions more exactly than we usually do.
Another argument to which Mr. Pickard-Cambridge attaches weight is to this effect: If the making of a promise in itself constituted a prima facie obligation to fulfil the promise, the obligation, springing as it does from a single source, ought to be always equally great; whereas every one in fact regards a promise to attend an At Home, for instance, as much less binding than one to attend a dinner-party.14 An attempt might be made to answer this objection by saying: ‘This would be a fair argument to use if I thought the making of a promise was the only source of a prima facie obligation. But I do not; I think there is also a prima facie obligation to bring into being as great a balance of good over evil as one can. I am therefore bound to consider the consequences of my act, and in the one case the obligation to fulfil the promise is reinforced by the obligation to save my host the very considerable inconvenience of an empty place at his dinner-table, while in the other, where the inconvenience to him will be much less, the reinforcement is much less, and a less strong countervailing obligation will make me think it right to break my promise and do something else instead—e.g. spend the afternoon in the company of a sick friend.’
This, however, would not be in my opinion the correct answer. For it divides my responsibility to my host into a responsibility to fulfil a promise+a responsibility to produce good, the first responsibility being of uniform obligatoriness and the second being more obligatory according as the good to be produced is greater; and that is not how we really think about the matter. What we really think is that we have a single responsibility to our friend, to confer on him the promised benefit (our presence at the party being assumed to be a benefit to him).
There appears to be no reason why one who does not take the utilitarian view of promises should consider the bindingness of all promises to be equal. In our natural thought about it, I believe we think of it as being, as it were, a product of two factors. One of these is the value of the promised service in the eyes of the promisee; we clearly think ourselves more bound not to fail another person in an important matter, than not to fail him in an unimportant one; and if any one doubts this and thinks I am being dangerously lax in allowing degrees of bindingness here, let him suppose himself (a) to have promised to visit a sick friend whom he knows to be longing for his company, or alternatively (b) to have promised to go to the theatre with a party which he thinks will not miss him much if he does not go. In the absence of any other responsibility competing with either of these, each of them will be binding; but the second responsibility will be much more readily overridden by any competing responsibility, such as that of turning aside to help the victims of an accident.
The other factor tending to increase the obligation to fulfil a promise depends on the way in which and the time at which the promise has been made. Any one would feel that a promise made casually in a moment of half-attention is less binding than one made explicitly and repeatedly, and perhaps reinforced by oath. The recency of the promise seems also to add something to its bindingness; we say, ‘Why, it was only yesterday that you promised to do it.’ Such considerations make the promise more binding, because they intensify the promiser's awareness of its existence and the promisee's expectation of its fulfilment.
We may, then, if we like to put the matter so, think of the responsibility for conferring a promised benefit as being n times as binding as the responsibility for conferring an exactly similar unpromised benefit, where n is always greater than 1, and, when the promise is very explicit, is much greater than 1. It will follow that it is always our duty to fulfil a promise, except when the uncovenanted benefit to be conferred is more than n times greater than the covenanted benefit. We are not able to assign a very definite value to n in any case, but I believe there is pretty general agreement that n is usually great enough to secure that when the alternative advantage to be conferred is not very different in amount, the promised advantage ought to be conferred.
From the duty of promise-keeping Mr. Pickard-Cambridge turns to consider briefly the duty of returning good for good, and that of punishment, and to argue that these also can be explained on utilitarian grounds. (1) ‘To return a service, especially where it requires a sacrifice, is to show most unmistakably that the service has been appreciated, and to keep alive the benefactor's good-will. Not to return it is to choke the life of friendship in its infancy, by discouraging the effective good-will of the benefactor; or, if his good-will still continues to live and to express itself, the relation is inevitably, if unintentionally, diverted downhill from friendship into patronage.’15 Is this the way in which any one not intent on defending a theory really thinks about the return of benefits? Is it really the wish to keep alive our benefactor's good-will, or to save him from becoming patronizing, that makes us think we ought to return benefits? Is not our actual thought more truly expressed in the simple phrase ‘one good turn deserves another’? And is not this the expression of the sense of a prima facie obligation to return good for good, quite distinct from the duty of caring for the character of our benefactor?
(2) With regard to punishment also, Mr. Pickard-Cambridge seems to give up any notion of desert. ‘The sufficient reason why prima facie it is a duty to punish only the guilty, is that they are the people who most need to be brought to reconsider their ways, and there is nothing so provocative of reflection as the inability to sit down in comfort.’16 It follows that if we thought that in any particular situation equally good consequences would be produced by punishing a guilty person and by inflicting pain on an innocent one, we should think the one act no more right than the other. But this is not what we really think. We think of infliction of pain on the innocent only as something to be taken to in the last resort, when some overwhelming reason of public policy demands it. And the very reformative effect of punishment of which he speaks will take place only if the punished person thinks of his punishment not as an administrative device for furthering the general benefit, but as something that he has deserved by his misdoing.
In his third article17 Mr. Pickard-Cambridge returns to the topic of promise-keeping. His argument is an elaborate one, but the only novel feature in it, I think, is a distinction which I had not brought out, and which is well worth bringing out. He distinguishes between what he calls the objective good brought about by the fulfilment of a promise, arising out of the nature of the service done to the promisee, and the subjective good, consisting in his gratification at getting what he expected to get; and similarly between the objective loss involved in the breaking of a promise, arising from the nature of what the promiser fails to give to the promisee, and the subjective evil consisting in his disappointment. And he argues that this enables him to justify on utilitarian grounds many acts of promise-keeping which it would be hard to justify on utilitarian grounds if we forgot the subjective good involved. He argues that if I should by keeping the promise confer 1,000 units of objective good on the promisee, and by doing an alternative act 1,002 units of objective good on some one else, I am on the strictest utilitarian grounds justified in keeping the promise if I think that more than one unit of subjective good will be conferred on the promisee in the one case, and more than one unit of subjective evil inflicted on him in the other.
He is surely not justified in laying the stress he does on the subjective good of a fulfilled promise to the promisee; for in general there is more subjective good involved in getting an unpromised and unexpected benefit than in getting a promised and expected one. But he is justified in attaching a good deal of importance to the subjective evil of not getting a promised and expected benefit; and this might, as he suggests, turn the scale, on utilitarian principles, in favour of keeping a promise which on other grounds a utilitarian would not think himself bound to keep. But I cannot accept the suggestion that, apart from the effect of my action in strengthening or weakening general mutual confidence, the obligation to keep a promise is entirely due to the sum of good objective and subjective to be conferred on the promisee. And I invite any one who accepts Mr. Pickard-Cambridge's view to consider the following cases.
(1) Where the objective goods to be conferred by keeping and by breaking the promise are equal, and the effects on general mutual confidence are negligible, he would make the reason for its being right to fulfil the promise, when it is right, lie in the gratification to be caused to the promisee. But now suppose that some one has through no fault of mine misunderstood me as promising to do a certain act—suppose, for instance, that he is a foreigner who has claimed to understand English perfectly but has in fact quite misunderstood some word or idiom I used, and therefore taken me to be promising to do act A when I was promising something different, or not promising anything at all; his gratification if I did or disappointment if I did not do what he was relying on me to do would be just as great as if I had promised to do it, but my thought about my duty would be entirely different. I should very likely as a matter of benevolence save him the disappointment, if I could do so without sacrificing some more stringent duty, but I should have none of the distinctive thought which I have when I have made a promise, that I am, simply for that reason, prima facie bound to fulfil it. Mr. Pickard-Cambridge's view leads to a conclusion completely contrary to what we actually think in such a case.
(2) Suppose that A, a dying man, has entrusted his property, or some part of it, to B on the strength of B's promise to hand it over to C, who knows nothing of A's wishes or of B's promise. Suppose that B does not believe in immortality, or believes that at any rate the dead know nothing of the fortunes of the living. Then there is for him no question of either subjective or objective good to be enjoyed by A, and since C knows nothing of the affair there is no question of subjective good or evil for him through the gratification or disappointment of his hopes. If the transaction has been private, B's act will have no effect on general mutual confidence, unless he divulges the transaction, which, if he breaks the promise, he is not likely to do. There is no question, for one who disbelieves in immortality, of strengthening or weakening A's friendship for him; for A is dead. Then according to Mr. Pickard-Cambridge B would be justified in simply considering what was the best use that could be made of the property. If he thought a fourth person D could make the best use of it, or that he himself could, he ought to bestow it on D or keep it for himself. There is, of course, the question of the effect on his own character. If he breaks the promise, he will probably weaken his own feeling for the sanctity of promises, and make it more likely that he will break other promises which even on utilitarian grounds he ought to keep. But, for Mr. Pickard-Cambridge, the duty of keeping promises is always subject to the more general duty of producing as much good as possible, so that the promiser will be doing a worse thing, and doing more harm to his character, if by keeping the promise he loses the opportunity of doing the maximum good, than if he breaks the promise.
Is it not clear that this utilitarian way of considering such a case is not the way in which honest men actually would consider it? We should, in fact, regard the breaking of this promise as an outrageous breach of trust, and if we fear the effect on our character, it is because we consider the act itself detestable.
Some one might ask why I take such an artificial case, in which many of the goods and evils that usually result from the keeping or breaking of promises are eliminated. My answer would be that I have no need of such cases to convince me, or (I believe) most thinking people, that there is a prima facie duty to fulfil promises, distinct from the prima facie duty to produce what is good. This seems to me to stand out as a salient fact in the moral situation, even when the moral situation bristles with good utilitarian reasons for fulfilling promises, or for breaking them. But when we have to deal with theorists who do not admit this, perhaps the method of isolating the issue by eliminating other considerations affords the best hope of convincing them.
The view that the obligation to fulfil promises is distinct from that of producing the greatest good has also been criticized from another point of view, viz. that of the school of Brentano. Mr. G. Katkov, in his Untersuchungen zur Werttheorie und Theodizee,18 has attempted to bring what we think about the sacredness of promises under the principle of the maximization of good, in the following way. He points out (1) that the receiver of a promise, on the strength of his belief that the promise will be fulfilled, undertakes certain sacrifices and risks which he would not otherwise have undertaken, so that the maker of the promise runs in turn the risk of inflicting great injury on the receiver if he breaks the promise.
This explanation plainly does not meet the case. The special characteristic of the situation created by a promise is that even when we do not think the promisee likely to lose more by the breaking of the promise than some one else will gain, we think ourselves prima facie obliged to keep the promise.
Mr. Katkov falls back (2) on a second line of defence, which is much more interesting. He holds that we are not justified in breaking a promise in order to confer even the greatest good on another than the promisee, if there is the slightest probability that through our action the promisee will suffer an injury which makes his total ‘balance of value’ predominantly bad. This is very different from Mr. Pickard-Cambridge's light-hearted treatment of the interests of the promisee as standing on exactly the same level as any one else's; it is, in fact, the admission of a prima facie obligation independent of the obligation to maximize value. It is true that Mr. Katkov rests this prima facie obligation on a different principle from that of fidelity to promise; he makes our special obligation in this case to be an instance of the general law that we are not justified in inflicting harm on one person in order to confer a greater good on another. To say this is at once to give up the crude utilitarian principle that an evil conferred on any one can always (in the determination of our duty) be set off against an equal good conferred on some one else; and I have already19 remarked that at this point Utilitarianism does not correspond to what we really think. But Mr. Katkov's admission fails to deal completely with the peculiarity of the situation created by a promise. This will be seen if we compare two situations. In one, A has made no promise to any one, and a certain act would confer a great good on B and a smaller injury on C. In the other, he has made a promise to C and none to B, and a breach of promise would confer a great good on B and a smaller injury on C. Any right-thinking person will, I believe, consider that the fact that this act would be a breach of promise is (whether decisive or not) a further argument against doing the act, over and above any objection which arises from its being an injuring of one person in order to benefit another.
(3) Finally, Mr. Katkov denies the possibility of any prima facie obligation which can conflict with the duty of producing the best state of affairs, on the general ground that ‘better’ simply means that which it is right to prefer. This is simply a cutting of the knot. The position is this. Those who think as I do, think (and claim that the moral consciousness when not sophisticated by a particular theory agrees with us) that there is a prima facie duty to fulfil promises even when no greater good can be foreseen as likely to come into being by the promise's being kept than by its being broken; and we think that this shows that the rule ‘produce the greatest good’ is not the only rule of conduct. The followers of Brentano agree with us in thinking that when we have made a promise we are under a special obligation to the promisee, but differ from us in thinking that it can be brought under the general rule ‘choose the greatest good’. But instead of trying to point out wherein the specific good to be produced by keeping promises consists, they content themselves with saying ‘it must be the greatest good, because it is what we ought to produce’. The utilitarian says ‘you ought to do so-and-so because by doing so you will produce the greatest possible good’: the follower of Brentano says ‘do what you ought to do, and you may be sure that in doing so you will be choosing the greatest good, since “better” means nothing but that to prefer which is right’.20
Mr. Katkov sees21 that if this is all that ‘better’ means, his principle ‘you ought to choose the best of what is attainable’ is in danger of being a mere tautology—‘it is right to prefer what it is right to prefer’. He can hardly be held to have escaped from this conclusion by the remark22 that ‘in choice not a simple preference, but the preference of an existence to a non-existence, or the opposite, is included, and so the utilitarian highest practical principle comes pretty much to this, that whenever one is faced by the choice between the existence of several goods, he should always prefer the existence of the more excellent’. If ‘better’ only means ‘such that one ought to prefer it’, nothing can save the principle ‘one ought to prefer the existence of the better’ from being a tautology. And unless we can see a goodness in the state of affairs produced by a fulfilment of promise—a goodness not resting on or consisting in the fact that we ought to fulfil the promise—we cannot say that the duty of fulfilling the promise rests on the general duty of producing what is good.23
Although I think it is quite clear that a promise creates a prima facie obligation, or a responsibility, quite different from that which alone the utilitarian system recognizes, we must guard against stating the obligation in a way which would lead to consequences which our reflective moral consciousness would reject. What then is it, exactly, that a promise creates an obligation to do? We must first note that, if the promise is phrased in the way in which promises usually are phrased, we are not under an obligation to do that which we have promised. For promises usually take the form of saying ‘I will do so-and-so’, where ‘doing so-and-so’ is the effecting of some change in the state of affairs, paying a debt, returning a borrowed book, travelling to some place to meet some person, &c.; but if we are right in the contention put forward in another context,24 we are under no obligation to effect changes, but only to set ourselves to do so. If, therefore, we want to limit ourselves to making promises which it will be our duty to fulfil, we should cast our promises in the form ‘I will set myself, or do my best, to do so-and-so’. But it would be pedantic to insist upon this alteration, for in fact our promises are understood so by both parties. No one thinks that he has failed to do his duty if he has done his best, without success, to fulfil his promise; and no one thinks another has failed in his duty, though we may think he has failed to discharge a legal obligation, if he has done his best to bring about the change that was promised.
But even the obligation to set oneself to do that which was promised does not necessarily remain binding until the self-exertion in question has been performed. For in the first place, if it has become clearly impossible to effect the change in question, there remains no duty to set oneself to effect it. If I have promised to return a certain book to a friend, and if the book is meantime destroyed, it ceases to be my duty to set myself to return it; for my duty, when the time comes at which I had promised to return the book, is determined not solely by my promise but also by the later developments of the situation; and here the situation has developed in such a way as to abolish the duty—though it has created another instead.
Suppose that the fulfilment of the promise has become not impossible, but much more difficult than it appeared likely to be when the promise was made. Mr. Pickard-Cambridge evidently thinks that in that case too the duty is usually abolished. But I do not think that our reflective moral sense would support him there. Ease of performance is no necessary or even usual characteristic of that which is our duty. No doubt, if the effort to fulfil some difficult promise were likely to be such as to incapacitate us for many other useful services, we might consider ourselves absolved from the duty; but then the prima facie obligation would not have been abolished, but would merely be over-ridden, as any prima facie obligation may be over-ridden by another that is more stringent. No doubt, too, we should regard the fulfilment of a promise which had become much more difficult to fulfil since it was given, as better evidence of moral goodness than the performance of one which had remained relatively easy, and the failure to perform it as less evidence of moral badness than the failure to perform one that had remained relatively easy. But to say there is less moral badness in failing to do A than in failing to do B is quite different from saying that A is less our duty than B, still more from saying that it is not our duty to do A at all.
Apart from impossibility of fulfilment, there is one other condition that abolishes the prima facie obligation. The duty of fulfilment of promise is not a duty of maintaining consistency between one's words and one's deeds. The sense of it is distinct from any sense that one may have that, to be self-consistent, one ought to do something that one has said one will do, just because one has said one will do it; and much stronger than any such sense. It is essentially the sense of a duty not to fail, not to ‘let down’, some one who has, on the ground of one's promise, trusted one to do something that he wishes one to do. If he convinces us that he has ceased to want it, we feel ourselves no longer under even any prima facie obligation to fulfil the promise, because we have no longer the thought that otherwise we shall be failing our promisee. Here again, the present duty depends not merely on the situation as it was when we made the promise, but on its subsequent development; and the subsequent development has been such as to make that cease to be a duty which at first was one. Indeed, if we feel convinced, even on other grounds than his saying so, that the promisee no longer wishes the promise to be fulfilled, the prima facie obligation has disappeared. At first sight, this seems to lend itself to the utilitarian explanation that it is the thought of the promisee's future gratification that makes the fulfilment of the promise a duty, so long as it is a duty. But that this is not the true explanation is shown by the fact that we all (I believe) think ourselves to be obliged on quite a different ground to fulfil a promise to another person, from that on which we think ourselves obliged to afford him an unpromised gratification. It is further shown by the fact that we think ourselves bound by a promise to a man who has died since the promise was made. Here, unless we think not only that he survives in another mode of being, but that he is, in that mode of being, aware of what happens on earth, and has the same desires that he had on earth, there is no thought of his future gratification; yet we feel the promise to be binding, whether or not we hold these views about a future life. In his case, too, we think the prima facie obligation is terminated if we are convinced that he, if he were alive and aware of the circumstances, would no longer wish us to do what we have promised. But that is a consideration which no utilitarian can consistently permit himself. He must concern himself only with consequences that he thinks likely to happen, not with what might happen if things were not as they are, if the promisee were alive instead of dead.
It might be suggested that though we do not seriously think the dead man will be gratified by our fulfilment of our promise, we are yet obscurely under the influence of a superstitious tendency to think of him as likely to be gratified. This, no doubt, is sometimes so; but that it is not a complete account of our attitude is shown by the fact that we should think just in the same way in the case of a promise to a man who by reason of distance, or of some other cause distinct from death, is not likely to know whether the promise has been fulfilled, and to be gratified or disappointed accordingly.
Again, if we have promised to do several related actions, some of which we believe to be desired by the promisee only if the others also are done, then if the latter become impossible, we naturally are under no obligation to do the former. On the other hand, if we have promised to do a number of related actions, the fact that some have become impossible and some are no longer desired by the promisee, while some perhaps are neither possible nor desired by him, does not abolish the duty to do the others which are both possible and desired by him.
There is another change of circumstances, regarding which we must ask ourselves whether it abrogates the prima facie duty of fulfilling a promise. Suppose that it is still possible to fulfil it, and that the promisee still wishes for its fulfilment, but that we have become convinced that its fulfilment will do him harm, or less good than something else that we might do instead. Here we think of the promise as still binding upon us; we think of ourselves as still under the prima facie obligation. But this, like all prima facie obligations, is open to the competition of others, and it is only to be expected that when the advantage to be given to the promisee by some alternative act is very great, the sense of obligation to do that act should outweigh the sense of obligation to fulfil the promise. And in principle the same is true when it is not the promisee but some one else on whom the great advantage will be conferred by the alternative act.
The various considerations I have pointed out as having the effect of abolishing or over-riding the original duty of fulfilling a promise seem to me to account completely for the whole variety of cases cited by Mr. Pickard-Cambridge in which promises are not felt to be binding, without in the least involving the conclusion which he draws, that when they are binding they are binding only on the ground that their probable consequences are better than those which could be achieved by any alternative action.
While we are concerned with the duty of fulfilment of promise, it is worth while to consider the relation of this duty to that of telling the truth. These are apt to be thought of as distinct and complementary duties, the one a duty to say what is true, the other a duty to make true what one has said. But the relation between the two is more complex than this; the two are connected in the following way. You can break a promise without telling a lie (for it is only when the promise is made with the intention of breaking it, that a lie is told); but you cannot tell a lie without breaking a promise. This arises from the nature of language. If words had a natural affinity with the things they usually stand for, you could deceive without breaking a promise, by simply using words in an unnatural meaning. But, apart from a few onomatopoeic words, there is not, between words and the things they stand for, any natural affinity which makes the one the natural symbol for the other. There is no more affinity between the noise ‘John is dead’ and the fact which we usually express by saying that ‘John is dead’, than between the noise ‘John is alive’ and the fact which we usually express by saying that ‘John is dead’. If some one asks me whether John is alive or dead, and I think that he is alive, it can be a duty for me to say ‘John is alive’ only in virtue of a pre-established expectation that when I make this noise I shall be thinking him to be in the state which we usually express by saying that he is alive. It is not to be supposed that there is a separate convention as to the meaning with which each word is to be used. But it must be supposed that there is a general convention that we shall not without notice use words in a meaning other than that in which they are generally used. If it were not for this, I should be at liberty to make the noise ‘John is dead’ when I think he is alive. But, there being this convention or understood promise, I am guilty of a breach of promise if I say he is dead when I think he is alive. Thus the telling of a lie is always a breach of promise.
The space I have given to discussing the duty of fulfilling promises might lay me open to the suspicion that I attach an undue importance to this duty. I do not think that that is the case. I have discussed it at length, partly because it is a very clear case of a duty which cannot be reduced to that of producing a maximum of good, and partly because the discussion has been forced on me by a particular critic. I do not suppose that I take, in fact, a more rigorous view of this duty than a utilitarian is likely to do; for on the one hand I am very conscious that there are other responsibilities which often outweigh that of fulfilling promises, and on the other hand a utilitarian is apt, in order to bring the rigour of his view about promises up to that of the ordinary moral consciousness, to make more of the utilitarian reasons for keeping promises than the facts warrant.25 My object is not to suggest that the duty is more binding than a utilitarian thinks it is, but to suggest that it is binding on quite a different ground. The other principles of duty which, I have suggested,26 fall outside the utilitarian scheme could be defended by arguments similar to those by which I have defended the independence of the principle of promise-keeping; but it would be tedious to develop such a defence. If I have convinced any one that there is one principle that falls outside the utilitarian scheme, he will probably be ready to admit that there are others also; and if I have not convinced him in this case, I should not be likely to do so in others.
Mind, xli (1932), 150–1.
The Right and the Good, 39.
Mind, xli (1932), 154.
Mind, xli (1932), 158.
I use this phase, for brevity, as including his loss of pleasure.
Mind, xli (1932), 163–4.
Mind, xli (1932), 159.
Mind, xli (1932), 168.
Mind, xli (1932), 169.
Op. cit. 140.
I return to this subject later, when the meaning of ‘good’ comes to be discussed, p. 289.
pp. 153–4, 160–1.
Cf. pp. 92–4.
Cf. pp. 69–77.