Studies in Hegelian Cosmology. John McTaggart Ellis McTaggart 1901
97. What may we conclude, on Hegelian principles, about the Supreme Good? The Logic has given us the Absolute Idea, which stands to knowledge in the same relation as the idea of the Supreme Good, if there is one, stands to action. In examining the Absolute Idea, we find it involves the existence of a unity of individuals, each of whom, perfectly individual through his perfect unity with all the rest, places before himself an end and finds the whole of the universe in complete harmony with that end.
If we have been justified in taking the Absolute Idea as only expressible in a unity of individuals, the rest of this description clearly follows. The individuals must be in harmony, and how can a conscious individual be in harmony with another, except by proposing an end to which that other is a means, though not, of course, a mere means? Besides this, if we look at the final stages of the Logic, we shall find that the idea of End, once introduced at the close of the Objective Notion, is never again lost. It is identical with the category of Life: in Cognition (which includes Volition) the only change is that the End has become conscious; while the transition to the Absolute Idea only alters the manner in which the harmony is held to be produced.
This is the supreme reality – the only reality sub specie aeternitatis, the goal of the process of the universe sub specie temporis. It will be very desirable if we can identify the supreme reality with the supreme good.
It is not the supreme good simply because it is the supreme reality.
This is scarcely more than a truism. But it always wants repetition, and never more than at present. It is often asserted that ideals are real because they are good, and from this it follows by formal logic that, if they were not real, they would not be good. Against this we must protest for the sake both of truth and of goodness. The idea of the good comes from that paradoxical power which is possessed by every conscious member of the universe – the power to judge and condemn part or all of that very system of reality of which he himself is a part. If the whole constitution of the whole universe led, by the clearest development of its essential nature, to our universal damnation or our resolution into aggregates of material atoms, the complete and inevitable reality of these results would not give even the first step towards proving them good.
98. But although the supremely real, as such, is not the supremely good, we may admit, I think, that if the supreme reality be such as Hegel has described it to be, then it will coincide with the supreme good. For, in the reality so defined, every conscious being – and there are no other beings – will express all his individuality in one end, which will truly and adequately express it. The fulfilment of such an end as this would give satisfaction, not partial and temporary, but complete and eternal.
And since each individual finds the whole universe in harmony with his end, it will necessarily follow that the end is fulfilled. Here is a supreme good ready to our hands.
99. Hegel has thus helped us to the conception of the supreme good, firstly by suggesting it, and secondly by proving that it contains no contradictory elements. Such a supreme good, we notice, is not purely hedonistic. It contains pleasure, no doubt, for the fulfilment of the ends of conscious beings must always involve that. But the pleasure is only one element of the perfect state. The supreme good is not pleasure as such, but this particular pleasant state.
100. It does not follow, however, that, because we have determined the supreme good, we have therefore determined the criterion of morality.
They can be identical, no doubt, but they need not be so. The object of a criterion is merely practical – to guide our actions towards good.
For this purpose we require something which shall be a sure sign of the good. But a thing may have many marks besides its essence, and one of these may often be the more convenient test. A stock is not made safe by a stock-broker’s belief in it. But an ordinary investor will find the opinion of a good stock-broker a much surer test of the safety of a stock than could lie furnished by his own efforts to estimate the forces, which will be the real causes of safety or danger.
We must remember, also, that for a satisfactory criterion of morality we do not require a sure test of all good, but only a sure test of such good as can possibly be secured by our voluntary efforts to secure it. If we find a criterion which will tell us this, it will be unnecessary to reject it because it is not also a satisfactory test of some other element of good, which we may enjoy when we get it, but cannot get by our own action.
101. But is a moral criterion wanted at all? It might be maintained that it was not. It would only be wanted, it might be said, if we decided our actions by general rules, which we do not. Our moral action depends on particular judgments that A is better than B, which we recognize with comparative immediacy, in the same way that we recognize that one plate is hotter than another, or one picture more beautiful than another. It is on these particular intuitive judgments of value, and not on general rules, that our moral action is based.
This seems to me to be a dangerous exaggeration of an important truth. It is quite true that, if we did not begin with such judgments, we should have neither morality nor ethics. But it is equally true that we should have neither morality nor ethics if we stopped, where we must begin, with these judgments, and treated them. as decisive and closing discussion. For our moral judgments are hopelessly contradictory of one another. Of two intelligent and conscientious men, A often judges to be right what B judges to be wrong. Or A, at forty, judges that to be wrong, which at twenty he judged to be right. Now these judgments are contradictory. For every moral judgment claims to be objective, and demands assent from alt men. If A asserts that he likes sugar in his tea, and B asserts that, for his part, he does not, both statements may be true. But if A asserts that to be right which B asserts to be wrong one of them must be in error, since they are making contrary statements about the same thing.
It is therefore impossible to treat all particular judgments of value as valid. We must do with them as we do with the particular judgments of existence – that is to say, treat them as the materials in which truth may be discovered, but not as themselves all true. We must reject some, and accept others. Now I do not see how this is to be done except by discovering some common quality which the valid judgments, and they alone, possess. And, if we test the particular judgments by means of this quality we have a moral criterion. Even if we confine ourselves to saying that the judgments of the best, or of the wisest, men are to be followed, there will be a criterion. For we cannot recognize the best, or the wisest in ethical matters, without a general idea of the good; To make the recognition itself depend on one of the particular intuitive judgments to be tested would be a vicious circle.
102. A criterion is therefore necessary. Before considering its nature, we must consider an ambiguity as to the matter which it is to judge. The ethical significance of the content of any moment of time is double. It may be considered in itself. In that case its moral significance will depend on the closeness with which it resembles the content which would realise the supreme good. Or it may be considered as a means towards a future end. In that case its moral significance will depend on the degree in which it tends to advance or to retard the eventual complete realisation of the supreme good.
It would be desirable, no doubt, if these two standards always coincided – if every action which was immediately good hastened the coming of the supreme good, and every action which was immediately bad retarded it. But we have no reason to believe that this is so in any particular case, and we have many reasons to believe that it is not so always.
We know that good often comes out of evil, and evil out of good.
This is a matter of every-day observation. And Hegel has shown us that good never comes, except out of conquered evil, and that evil must arise before it can be conquered. To bring our conduct to-day as close as possible to the supreme good may be to help or to hinder the coming of the supreme good in all its perfection.
This does not, however, introduce any conflict into our moral life.
For of the two possible standards ‘by which omniscience might judge a proposed action only one is practicable for us. We can see, to some extent, what conduct embodies the supreme good least imperfectly. But we have no material whatever for deciding what conduct will tend to bring about the complete realisation of the supreme good. That lies so far in the future, and involves so much of which we are completely ignorant, that we are quite unable to predict the road which will lead to it. What we do know, if we follow Hegel, is this – that the road we do take will lead to it, because the supreme good is also the supreme reality, and is therefore the inevitable goal of all temporal process.
It follows that the criterion of moral action which we require is not one which will determine what actions will most conduce to the eventual establishment of absolute perfection. It is one which will tell us what actions will bring about, immediately, or in the comparatively near future which we can predict with reasonable certainty, the state which conforms as closely as possible to that perfection.
The points I wish to prove in this chapter are (1) that the idea of perfection cannot give us any criterion of moral action; (2) that the hedonic computation of pleasures and pains does give us a definite criterion, right or wrong; (3) that the use of this latter criterion is not incompatible with the recognition of perfection as the supreme good, and would give us, if not unerring guidance, still guidance less erroneous than would be afforded by any other applicable criterion.
103. Let us consider the first point. When two courses are presented to a man who wishes to act rightly, and he is in doubt which of them he shall adopt, will he be assisted by reflecting on the nature of the supreme reality, which we have decided to be also the supreme good? It is clear, to begin with, that if either of the courses would result in the immediate realisation of the supreme good, it would be the course to take. But it is equally clear that this cannot ever take place, in the present state of ourselves and our experience. The reality contemplated by Hegel in his Absolute Idea is absolutely spiritual, absolutely timeless, absolutely perfect. Now none of us ever get a chance of performing an action the result of which would satisfy these three conditions. The result of any actions possible to us now would be a state in which spirit was still encompassed with matter, in which change still took place, and in which perfection, if rather nearer than before, was still obviously not attained.
It is useless then to test our actions by enquiring if they will realise the supreme good. None of them will do that, and we are reduced to considering which of them will enable us to reach rather nearer to supreme good than we were before.
104. This question has not, I think, been faced quite fairly by the school who assert the idea of perfection to be an adequate criterion.
They generally take a case in which some form of the desire for good as good – some form of specifically moral feeling – is opposed to something desired regardless of, or in opposition to, morality. They have then comparatively little difficulty in asserting, with some probability, that the idea of perfection would be a sufficient guide to direct us to the first rather than to the second. For perfection clearly includes a positing of the supreme good by each person as his end; and this positing would only differ from desire in excluding all thought of the possibility of non-fulfilment.
Surely, then, the good will must raise any state, of which it is a moment, above all other states which do not participate in it. But even if this criterion is true, it is almost always useless. It is of some use if there is a question of another will besides that of the agent. For then there would be some meaning in saying that A’s duty to B was to endeavour to make B do that which B himself thought morally right. Here the will to be made good is not the agent’s own will, and so there is no tautology.
But we have other duties besides our duty to influence the wills of our neighbours. And the attempt to use the criterion more generally, by applying it to the agent’s will, breaks down. If A demands which of two courses the ideal of perfection prescribes for him here and now, all the reply that can be made is that it will be best for him to take the course which he takes believing it to be the best. Now he certainly will take the course which he believes to be morally the best. For, if not, he would not have sought guidance in an ethical criterion.. Such a criterion can never give a reason why the morally good should be desired. All it can do is to tell us what things are morally good.
If A has not decided to act morally, the criterion will be ineffective, for, if he has not decided to act rightly, why should he refrain from an action because it is wrong? If, on the other hand, he has decided to act morally, and appeals to the criterion to tell him what course he should take, it is clear that each course claims to be the morally right one, and he is undecided. In that case to tell him that he will be right, if he pursues the course which he judges to be right, is to tell him nothing, for what he wants is to be helped in judging which of them is right.
105. The practical use of ethics – and it is this we are considering – can only occur, then, when a man has resolved to act in conformity with duty, and is not certain what course duty prescribes. Two courses of action may each be in itself morally desirable, and may be incompatible, so that we are in doubt which to pursue.
Two courses of conduct, let us suppose, are presented to us. By taking a we shall further the end a, by taking b the end ▀. Both a and ▀ are good, but a and b are incompatibles. Can the principle of perfection tell us whether a or ▀ is under the circumstances, to be preferred? It seems to me that it is impossible in most cases, if not in all. It is clear that neither a nor ▀ can be expected to be realised unchanged in the supreme good. For any end which can be attained by an action in our, present state would still be an element in an imperfect and incomplete world, would still be tainted with imperfection, and could not therefore, as such, form part of absolute perfection. On the other hand every end which a man could represent to himself as a moral ideal has some real good in it. It would therefore form an element, however transcended and altered, of the supreme good. Thus we should only be able to say, of both a and ▀, that they were imperfect goods. Which is the least imperfect? That could only be settled by comparing each of them with the supreme good – a comparison which it is scarcely ever possible to carry out so as to assign a preference to either alternative.
106. Let us take an example. Most people think that the institution of marriage, as it at present exists in civilized countries, is on the whole a good thing. But others think it a mistake. They hold that all unions between man and woman should be terminable at any moment by the simple desire of the parties, who should then each of them be free to form a fresh union. And this they put forward as a moral advance. Can the contemplation of the supreme good help us to decide this question? It is clear, at any rate, that we cannot solve the difficulty by simply copying the pattern which the supreme good lays down for us. For there the difficulty could not arise at all. In a world of pure spirit there could be no sexual desire, and in a world which was timeless there could be no propagation of children – two elements which have considerable importance when we are dealing with marriage. And in such a state all relations would be permanent together, so that the question could never be raised whether outside relations ought to change in harmony with internal changes.
Whichever course we take, therefore, we shall not be able to model ourselves completely on the supreme good. Which course will lead us to the result least removed from the supreme good We find ourselves in a hopeless antinomy – hopeless, not from the actual want of a solution, but because the solution requires a knowledge of detail far beyond our power, The conservative side may assert, and with perfect truth, that in perfection all relations are absolutely constant. But if they infer from this that a minimum of change in the relationships between particular men and particular women is most consonant to the supreme good, the innovators may reply, with equal truth, that in perfection all relations will be the free expression of the inner nature of the individual, and draw from this – with equal right – the contrary conclusion that every relation between a man and a woman should cease with the cessation of the feelings which led them into it. If it is answered to this – as it certainly may be – that true freedom, as we find it in perfection, is not capricious but manifests itself in objective uniformities, it may with equal force be retorted that true constancy does not lie in clinging to external arrangements which have become unfit expressions of the internal nature of the persons concerned, but in the continuous readjustment of the external to the developing nature of the internal. It there is a rejoinder that true development does not consist in yielding to caprice, there may be a rebutter that true order does not lie in blindly accepting experience, but in moulding it. And so on, and so on, until the stock of edifying truths runs out, if it ever does. We can never get forward. One side can always prove that there is some good in a, and some imperfection in ▀. The other side can prove the converse propositions. But to know which is best, we should have to discover whether we should he nearer to perfection if at the present moment we emphasised freedom, even at the price of caprice, or emphasised order, even at the price of constraint. And how are we to discover this? And yet the particular problem we have been discussing is one on which most people in the world, and most of the independent thinkers of the world, have come to the same conclusion. But that, I fancy, is because they take a more practical criterion. If we estimate the gain or loss of happiness which would follow from the abolition of marriage, we may perhaps find excellent reasons for declining to make the change.
But we shall not have been helped in our decision by the idea of the supreme good.
Innumerable similar cases could be found. Public schools knock a great deal of pretence out of boys, and knock a certain amount of Philistinism into them. In heaven we shall be neither shams nor Philistines.
But are we nearer to heaven, if at this moment we buy genuineness with Philistinism, or buy culture with Schwńrmerei? The man who answers that question would need to be deep in the secrets of the universe.
107. But although the supreme good is useless as a help in a real investigation of an ethical question, it is a dangerously efficient ally in a barren and unfair polemic. For a is always partly good, ▀ never quite good. Ignore the corresponding propositions, that ▀ is also good, and a also imperfect, and we have an admirable argument for anything. For this purpose the words “true” and “higher” are useful. Thus the opponent of marriage, if confronted with the goodness of order, may reply that the true, or the higher, order is freedom. But then the supporter of marriage may enter on the same sophistry, by representing that the true, or the higher, freedom is order. Both propositions are quite true. In the supreme good, order and freedom are so transcended, that they are compatible – indeed, identical. It is true that the perfect forms of each are identical, and that the perfect form of either would always include and surpass the other’s imperfect form. The sophistry lies in making this the ground for preferring the imperfect form of the one to the imperfect form of the other. When we consider how short and simple such a device is, as compared with a laborious empirical calculation of consequences, and that it can be applied on any side of any dispute, we may expect that it will in the future furnish as convenient a shelter for prejudices and indolence as innate moral ideas, or the healthy instincts of the human mind.
108. Another class of difficulties occurs in which the ends are not in themselves incompatible, but in which the inadequacy of the means renders it necessary to sacrifice one – at any rate partially. We have continually to divide our energy, our time, and our money, between several objects, each of which has admittedly a claim to some, and which could absorb between them, with good results, more than the total amount we have to divide. Ethical problems arise here to which the answers must be quantitative, and I fail to see what hope there is of settling them by means of the idea of the supreme good.
A man with some leisure may admit – and will generally be wise if he does – that he should devote some of it to work of public utility, and some to direct self-improvement. But how much to each? He could very probably use all his leisure for either purpose with good results. At any rate, he will – in the great majority of cases – often find an hour which he could use for either. Which shall he sacrifice? Shall he attend a committee meeting, or spend the evening studying metaphysics? These difficulties come to all of us. The contemplation of the supreme good will tell us, it may be granted, that both metaphysics and social work have an element of good in them. But our contemplation cannot tell us which to prefer to the other, for the supreme good chooses neither, but, transcending both, enjoys both in their full perfection simultaneously, which is just what, in the present imperfect state of things, we cannot do. And it is no good telling us to neglect neither, or to make a division of our time. For a division cannot be made in the abstract. We must make it at a particular point, and assign the marginal hour of which we have been speaking either to philanthropy or to metaphysics.
The distribution of wealth presents us continually with similar questions. A man with a thousand a year would probably feel that he ought to give something to relieve distress, and also to give his children a better education than the average child gets at present. But this abstract conviction will not divide his income for him. Shall he send his sons to a second-rate school, and pension his old nurse, or shall he send them to a first-rate school, and let her go to the workhouse? Problems like these are the real ethical difficulties of life, and they are not to be solved by generalities – nor even by contemplating the idea of the supreme good, in which there are neither school-bills nor workhouses, and whose perfections are in consequence irrelevant to the situation.
109. It may be said that it is not within the province of ethics to deal with individual cases such as this. And in one sense this is true. A system of ethics is not bound to lay down beforehand the precise action a man ought to take in every conceivable contingency. This would, to begin with, be impossible, owing to the number of possible contingencies.
And, even if possible, it would be undesirable. In applying rules to a given set of circumstances we require not so much philosophical insight as common sense and special knowledge of those circumstances.
The philosopher is not likely, perhaps, to have more common sense than the man whose action is being considered. And the latter is much more likely to understand his own circumstances for himself than the philosopher is to understand them for him. The particular problems of conduct, therefore, are best solved at the place and time when they actually occur.
But it is, none the less, the duty of ethics to provide the general principles upon which any doubtful point of conduct ought to be settled.
It would plainly be absurd to assert that any one distribution of our time and wealth among good objects would be as good as any other distribution.
It would he still more absurd to assert that a man who desired to act rightly would not care whether he made the best possible distribution.
Surely the only alternative is to look to ethics for the principle on which we must make the distribution. And it is just this in which the idea of the supreme good fails to help us.
110. It has been suggested that a suitable formula for ethics may be found in “my station and its duties.” Each of us finds himself in a particular place in the world. The particular characteristics of the situation suggest certain duties. Do these, and in this way the supreme good will be most advantaged.
As an analysis of morality this theory has many recommendations, and it was not, if I understand rightly, originally put forward as a moral criterion. But, for the sake of completeness, it will be well to point out that it is not available as a criterion. In the first place, it fails to tell us how we are to judge those persons who have endeavoured to advance the good by going beyond, or contrary to, the duties of the station in which they then were, and so transforming their society and their own station in it. The number of these may be comparatively small. But the effect of their action is so important for everyone that it is essential for a moral criterion to be able to determine when such innovations should be accepted and when rejected.
These cases can be brought within the scope of the formula, if it is only taken as an analysis of morality. For there is no contradiction in saying that my duty in a certain station – e.g., that of a slave-holder, or of a slave – may be to destroy that station. But such cases are clearly fatal to any attempt to use the formula as a criterion. Some fresh criterion would be wanted to tell me whether my duty in my station did or did not involve an attempt to fundamentally alter its nature.
Again, even in the ordinary routine of life, such a principle would give but little real guidance. It lays down, indeed, the wide boundaries within which I must act, but it does not say precisely how I shall act within these boundaries, and so leaves a vast mass of true ethical questions unsettled. My station may include among its duties that I should seek a seat in Parliament. If I get one, my station will demand that I should vote for some bills and against others. But which? Shall I vote for or against a Sunday Closing Bill, for example? Such questions can in the long run only be answered by reference to an ethical ideal. And the ideal of my station and its duties will not help us. For while the ideal M.P. will certainly vote for the bills be thinks ought to pass, and against those he thinks ought not to pass, there is nothing in the conception of a perfect member of parliament which can tell us in which of these classes he will place a Sunday Closing Bill.
Or my station may be that of a schoolmaster. This defines my duties within certain limits. But it cannot tell me whether in a particular case it is worth while to make a boy obedient at the cost of making him sulky.
Thus the principle, if taken as a criterion, is not only inadequate, but it proclaims its own inadequacy. For the duty of an M.P. or a schoolmaster is not only to vote on bills, or to act on boys, regardless of the manner, but to vote rightly on bills, or to act rightly on boys. And, since the right way in each particular case can never be got out of the mere idea of the station, the formula itself shows that some other criterion is needed for the adequate guidance of our action.
111. I now proceed to the next branch of my subject – namely that the calculation of pleasures and pains does give a definite criterion of action. (Calculation is, I think, a better word than calculus, which, as a technical term of mathematics, seems to imply a precision unattainable, on any theory, in ethics.) I am not now maintaining that it is a correct criterion that it will enable us to distinguish right from wrong, but merely that it is sufficiently definite to be applied to actions in an intelligible way. The question of its correctness from an ethical point of view must be postponed for the present.
112. The elements at any rate of such a calculation are clear. We do know what a pleasure is, and what a pain is, and we can distinguish a greater pleasure or pain from a lesser one. I do not mean, of course, that the distinction is always easy to make in practice. There are some states of consciousness of which we can hardly tell whether they give us pleasure or pain. And there are many cases in which we should find it impossible to decide which of two pleasures, or of two pains, was the greater.
This, however, while it no doubt introduces some uncertainty into our calculations, does not entirely vitiate them. For when we can see no difference, as to amount of pleasure or pain between two mental states, we may safely conclude that the difference existing is smaller than any perceptible one. And, in the same way, if we are unable to tell whether a particular state is more pleasurable than painful, we may safely conclude that the excess of one feeling over the other must be small. Thus the margin of vagueness which is left is itself limited. This is quite different from the far more dangerous vagueness which we found in considering perfection. When we were unable to tell whether the maintenance or the abolition of marriage would bring us nearer to the supreme good, this uncertainty by no means gave us the right to infer that it made little difference which happened. The choice might make a very great difference. The uncertainty came from our ignorance, and not from the close equality of the two alternatives. But if we are doubtful whether a plate of turtle or a bunch of asparagus would give us most pleasure, or whether the pleasure of a long walk outweighs the pain of it, we may at least be certain that we shall not lose very much pleasure, whichever alternative we finally select.
113. It has been objected to hedonistic systems that pleasure is a mere abstraction, that no one could experience pleasure as such, but only this or that species of pleasure, and that therefore pleasure is an impossible criterion. It is true that we experience only particular pleasant states which are partially heterogeneous with one another. But this is no reason why we should be unable to classify them by the amount of a particular abstract element which is in all of them. No ship contains abstract wealth as a cargo. Some have tea, some have butter, some have machinery. But we are quite justified in arranging those ships, should we find it convenient, in an order determined by the extent to which their concrete cargoes possess the abstract attribute of being exchangeable for a number of sovereigns.
114. Another objection which is often made to hedonism lies in the fact that pleasures vanish in the act of enjoyment, and that to keep up any good that might be based on pleasure, there must be a continuous series of fresh pleasures. This is directed against the possibility of a sum of pleasures being the supreme good. As we are here only looking for a criterion, we might pass it by. But it may be well to remark in passing that it seems unfounded. For so long as we exist in time, the supreme good, whatever it is – perfection, self-realisation, the good will – will have to manifest itself in a series of states of consciousness. It will never be fulfilled at any one moment. If it be said that all these states have the common element of perfection or the good will running through them, the hedonist might reply that in his ideal condition all the states of consciousness will have the common element of pleasure running through them. Pleasure, it may be objected, is a mere abstraction. Certainly it is, and the element of a pure identity which runs through a differentiated whole must always he to some extent an abstraction, because it abstracts from the differentiation. In the same way, perfection or good will, if conceived as timeless elements of a consciousness existing in time, are just as much abstract, since abstraction is thus made of the circumstances under which alone they can be conceived as real and concrete.
So long, therefore, as our consciousness is in time, it can be no reason of special reproach to pleasure that it can only be realised in a continuous succession. And if our consciousness should ever free itself of the form of succession, there is no reason why pleasure should not be realised, like all the other elements of consciousness, in an eternal form.
Indeed pleasure seems better adapted for the transition than the other elements of consciousness. A timeless feeling is no doubt an obscure conception. But we can, I think, form a better idea of what is meant by it than we can of the meaning of timeless cognition or of timeless volition.
115. We now come, however, to a more serious difficulty. Hedonic calculations require, not only that we should compare the magnitudes of pleasures, but that we should add and subtract them. The actions which we propose to ourselves will not each result in a single pleasure or pain.
Each will have a variety of results, and, as a rule, some of them will be pleasures, and some pains. To compare two projected actions, therefore, it will be necessary in each case to take the sum of the pleasures, subtract from it the sum of the pains, and then enquire which of the two remainders is the larger positive, or the smaller negative, quantity.
Now pleasures and pains are intensive, not extensive, quantities. And it is sometimes argued that this renders it impossible for them to be added or subtracted. The difference between two pleasures or two temperatures is not itself, it is said, pleasant or hot. The possibility of adding or subtracting intensive quantities depends, it is maintained, on the fact that the difference between two of them is a third quantity of the same kind – that the difference between two lengths is itself a length, and the difference between two durations is itself a duration. And, since this characteristic is wanting in intensive quantities, it is concluded that it is impossible to deal with them arithmetically.
The question is one of great importance, and the answer affects more than the hedonic criterion of moral action. It will, I believe, be found on further consideration that, reasonably or unreasonably, we are continually making calculations of pleasures and pains, that they have an indispensable place in every system of morality, and that any system which substitutes perfection for pleasure as a criterion of moral action also involves the addition and subtraction of other intensive quantities.
If such a process is unjustifiable, it is not hedonism only, but all ethics, which will become unmeaning.
116. Introspection. I think, will convince us that we are continually adding and subtracting pleasures and pains, or imagining that we do so, and acting on what we suppose to be the result of our calculations.
Whether we do it as a moral criterion or not, we are continually doing it in cases in which we do not bring morality into the matter. Suppose a man to be presented with two bills of fare for two equally expensive and equally wholesome dinners, and to be invited to choose which he shall take. Few of us, I fancy, would either find ourselves unable to decide the question, or admit that our answer was purely capricious and unmeaning. Yet how can such an answer be given except by adding pleasures? Even the most artistic composition can scarcely give such unity to a dinner as to admit of the pleasures we derive from it being regarded as anything but a succession of pleasures from each dish – not to say each mouthful. And, if we still prefer one dinner to the other, does not this involve the addition of pleasures? Such cases make up a great part of our lives. For even when distinctively moral considerations come in, they very often leave us a choice of equivalent means, which can be settled only by our own pleasure. My duty may demand that. I shall be at my office at a certain hour, but it is only my pleasure which can give me a motive for walking there on one side of the street rather than on the other. My duty may demand that I shall read a certain book, but there may be no motive but pleasure to settle whether I shall use a light copy with bad print, or a heavy copy with good print. And almost all such decisions, if made with any meaning at all, require that pleasures and pains should be added and subtracted.
It is only in this way that we can decide, whenever several pleasures and pains of each course have to be taken into consideration, and.
whenever a pleasure has to be balanced against a pain. Moreover, even if a single pleasure or pain from one has to be balanced against a single pleasure or pain from another, we still require addition if each of these feelings is to be looked on as an aggregate of several smaller ones. And they must be looked at in that way, at any rate, in the very common case in which the greater keenness of one feeling is balanced against the greater length of the other.
117. This calculation of pleasures is not only requisite for life, but it fills an indispensable, though subordinate, place in even non-hedonist morality. If, with two courses a and b before me, I can find no perceptible difference either to the welfare of others, or to my own perfection, while at the same time a is pleasanter than b, is it morally indifferent which course I shall take? Surely it cannot be held to be indifferent, unless we deny that pleasure is better than pain – an outrage on common sense of which the great majority of non-hedonist moralists cannot be accused. If pleasure is better than pain, then, caeteris paribus, it is our duty to choose it – a duty which may not require very constant preaching, but the neglect of which is none the less morally evil.
But, even if we leave this out, it can scarcely be denied that there are cases when it is our duty to give pleasure, simply as pleasure, to others.
Even Kant admits this. And if we have to do this we must either confess our actions to be utterly absurd, or else base them on a calculation of pleasures. Whenever either course produces a succession of pleasures or pains, whenever pleasures and pains have to be balanced against one another, whenever the intensity of one feeling has to be balanced against the length of another, or the intensity of one man’s feeling against a plurality of weaker feelings in many men – in all. these cases we must either add pleasures and pains, or work absolutely in the dark.
118. I have, I think, said enough to show that the rejection of all calculations about pleasure is not a simple question, and that it would necessarily lead to a good deal of doubt – almost amounting to positive denial – of the possibility of our acting rationally at all. But we may carry this line of argument further. The only reason which we have found for doubting the legitimacy of such calculations is that they involved the addition of intensive quantities. Now if it should be the case that the opposed theory of ethics, which would have us take perfection as a criterion, also requires the addition of intensive quantities, we should have got, at the least, an effective argument ad hominem against our chief opponents.
We should, however, have got more than this. For every ethical theory accepts either perfection or pleasure as a criterion, except the theory which holds that the good is shown us by immediate intuitive judgments, which, as we have seen above, rejects all criteria whatever.
Even that other form of Intuitionism, which maintains that we are immediately conscious of the validity of certain general moral laws, requires one or both of these criteria. For some of the moral laws are always represented as laws of imperfect obligation.. We are to be as good as possible, or to do as much good as possible. And such laws always involve either perfection or pleasure as a standard.
The only criteria offered are perfection and pleasure. Pleasure as a criterion admittedly involves the addition of intensive quantities. If perfection as a criterion does the same, we shall be reduced to a dilemma.
Either we must find room within ethics for the addition of intensive quantities, or we must surrender all hope of directing our conduct by an ethical principle.
119. Is it then the case that the criterion of perfection does require the addition of intensive quantities? I do not see how this can be avoided.
Absolute perfection – the supreme good – is not quantitative. But we shall not reach absolute perfection by any action which we shall have a chance of taking to-day or to-morrow. And of the degrees of perfection it is impossible to speak except quantitatively. If we can say – and we must be able to say something of the sort, if perfection is to be our criterion – that a man who stays away from the poll acts more perfectly than a man who votes against his conscience for a bribe, and that a man who votes according to his conscience acts more perfectly than one who stays away – then we are either talking about quantities or about nothing.
And these quantities are clearly intensive. The difference between one perfection and another cannot be a third perfection.
The incomplete stages of perfection, which, on this theory, must be the immediate ends between which we have to choose, are quantities then, and intensive quantities. Does the regulation of our conduct require that they should be added and subtracted? Again I do not see how this can be denied.
120. A boy is to be sent to one of two schools. At A he will get better manners, and a purer Latin style, than he would at B. But at B he will acquire habits of greater industry, and greater bodily vigour, than he would at A. How is the question to be decided, with perfection as the criterion? I have already tried to show in the preceding part of this chapter that it cannot be decided at all on such principles, since we have absolutely no data to enable us to guess whether a particular English boy, in 1901, will be nearer to the supreme good with industry and bad manners, or with good manners and indolence. But supposing this obstacle got over, the success of the method would then depend entirely on our being able to add intensive quantities. For here you have two elements of perfection – manners and Latin style – on the one hand, and two more elements – industry and bodily vigour – on the other. And unless the perfections attained at A have a sum which can be compared with the sum of the perfections attained at B, your action will be absolutely unreasonable, on whichever school you may decide.
Nor would it be fair to attempt to evade this by saying that perfections of character cannot be taken as units which can be aggregated or opposed, but should be considered as forming a unity. No doubt this is true of absolute perfection. All moments which form part of the supreme good are not only compatible, but essentially and indissolubly connected in the supreme good. In the supreme good whatever elements correspond to those imperfect goods which we call manners, and Latin style, and industry, and bodily vigour, will imply and determine one another. But not even a public school can land us straight in heaven.
And in this imperfect world these four qualities must be considered as four separable goods, for every one of the sixteen combinations which their presence and absence could produce is notoriously possible. We must consider the problem before us as one in which two separate goods are gained at the expense of two others. And how we are to come to any opinion on this point, unless we add the goodness of the goods together, I fail to conceive.
Or again, with a limited sum to spend on education, shall we educate a few people thoroughly, or many less thoroughly? Let us assume – and it seems at least as reasonable as any other view – that a slightly educated person is nearer to perfection than one completely uneducated, and that a thoroughly educated person is still nearer to perfection. How are we to decide between the greater improvement in each one of a few people, and the smaller improvement in each one of many people, except by estimating the sum of the perfections gained by each course? Or the difficulty may arise about oneself. Two foreign tours may each offer several quite heterogeneous goods. If I go to Italy, I may study pictures and improve my knowledge of Roman antiquities. If I go to Germany, I may hear Wagner and investigate German socialism. If we are to use perfection as a criterion here, must we not begin by summing the good which would result from each course? 121. And thus it would seem that ethical criteria in general must share the fate of the hedonic criterion. For the only serious charge that has been brought against the latter is that it involves the addition and subtraction of intensive quantities. And we have now seen that the only other criterion which has been suggested is equally impotent to act, in most cases, except by the addition and subtraction of intensive quantities.
This would destroy all ethical systems except those which made our particular moral judgments immediate and ultimate. And this position, as I have endeavoured to show above, is as destructive to ethics in another way, since it destroys all possibility of saying that any moral judgment is wrong.
And not only ethics, but all regulation of conduct with regard to consequences, seems equally involved. For what consequence of action, which we can regard as valuable, has not intensive quantity? And how can we act rationally with regard to consequences, unless the different intensive quantities in different sets of consequences can be compared?
122. Let us now consider whether the arguments which lead to such a negative result are really valid. I do not think that they are. If we have two pleasures of different intensities, it is true, no doubt, that the excess of A over B is not a pleasure. For we cannot imagine that part of the intensity of A existing by itself. Its meaning depends on its being in combination with the rest of A’s intensity. It would be meaningless to ask what the heat of an average June day would be like after the heat of an average December day had been subtracted from it. The remainder would cease to be what it had been as soon as it was separated from the other part.
But although the excess of A’s intensity over B is not a pleasure, I submit that it is, nevertheless, pleasure. Whatever has quantity must be homogeneous in respect of some quality, and is only quantitative in respect of that homogeneous quality. If therefore pleasure has an intensive quantity, then each part of that quantity must be pleasure, including that part by which it is greater than another.
If then the excess of intensity of A over B is pleasure, and a quantity, it must be capable of being brought into numerical relations with other quantities of pleasure. And thus, while it is true that we cannot imagine that excess as a separate pleasure, we can imagine a separate pleasure which shall be equal to that excess. If this is called C, then we shall be able to say that the pleasure in A is equal to the pleasure in B and C. And this is all that is wanted for the hedonic criterion.
I must confess that I find no difficulty in making such judgments, and that they seem to me to have a perfectly definite meaning. I feel no hesitation in affirming that the pleasure I get from a plate of turtle-soup is more than twice the pleasure I get from a plate of pea-soup, or that the pleasure I get from reading a new novel, together with the pain of a hot walk to get it, leaves a balance of pleasure greater than the pleasure from reading an old novel off my shelves. Of course I may make mistakes over these judgments. But mistakes can be made about extensive quantities also. I may judge A to be six feet high, when he is really an inch less. But this does not prevent his height from having a real and definite relation to the length of a yard-measure.
123. The certainty of any particular judgment as to an intensive quantity, and the minuteness to which such judgments can be carried, is far less, certainly, than is the case with judgments as to space, or as to anything which can be measured by means of spatial standards. It would be impossible to say with any confidence that one pleasure was 3.77 times as great as another, or even exactly twice as great. This has sometimes been taken as a proof of the impossibility of the hedonic criterion.
But it is unfair to argue from the impossibility of absolute certainty or exactitude in any class of judgments that the judgments are without any meaning, and that there is no objective truth to which the judgments approximate. This would render all judgments of quantity invalid. When we pronounce a yard-measure to be equal to the standard yard at Westminster, that is only an approximation, dependent on the accuracy of our instruments, which may be great but is never perfect. The approximation in the measurement of pleasure is no doubt much rougher, but there is only a difference of degree, and if the uncertainty does not completely invalidate the judgment in one case, it cannot do so in the other.
It may be objected that the uncertainty of this criterion, while not destroying its theoretical validity, deprives it of all practical use. Even if this were the case, it would be no worse off than any other criterion.
For, as was pointed out above, the value of an action cannot be judged by the standard of perfection without the addition and subtraction of intensive quantities. The only difference is one which is to the advantage of hedonism, for no one ever mistakes intense pain for intense pleasure, while ideals of perfection have been so different and incompatible that, whoever is right, many people must have mistaken great defects for great excellencies.
But there seems no reason for supposing that our estimates of pleasures and pains are so inaccurate as to be useless. We all make these estimates many times daily – even those of us who do not accept them as moral criteria. Can it be asserted that they have no worth whatever, and that everyone would on the whole be just as happy if he always took the course which seemed to him in anticipation to be less pleasant? Supposing that, on the next Bank Holiday, every person who should think that he would enjoy Epping Forest more than Hampstead Heath, should nevertheless go to Hampstead, is there any doubt that there would be a net loss of pleasure? Much uncertainty and error there certainly is in our estimates. But the only fair consequence to draw from this is that the conduct of human life is often a doubtful and difficult matter. And this conclusion is neither novel nor absurd.
124. We now pass on to the third division of the subject. Even if pleasure gives us a criterion which is applicable, does it give us one which is correct? The supreme good, as defined at the beginning of this chapter, may be analysed into two moments. On the one hand each individual has a nature, whose satisfaction he postulates. On the other hand, the relation of each individual with others is such that it satisfies the natures of all of them. This analysis of the supreme reality, which is also the supreme good, is not the only one which is possible. Indeed it may be said that it is not a perfectly adequate analysis, since it gives a primacy to the nature of the individual over the nature of the whole which misrepresents the perfectly equal and reciprocal relation indicated in the Absolute Idea.
But it is, I think, the most adequate analysis of absolute reality which is possible for Ethics. Ethics is based on the idea of Volition – an idea which the Logic shows us is transcended by the Absolute Idea – and cannot rise above the view of reality under the category of Volition, the peculiarity of which is exactly this over-emphasis on the nature of the individual as compared with the nature of the whole. The imperfection by which we fall short of the supreme good is two-fold. On the one hand the ideals of which we postulate the fulfilment are not absolutely the same ideals which would be found in a state of perfection. On the other hand the ideals which we have are not completely satisfied. The two sides are closely connected. Nothing but perfect ideals could ever be perfectly satisfied, nor could an unsatisfied ideal be quite perfect. For all things react on one another, and the perfection of any part of the universe is only possible on condition that the rest is perfect too. At the same time, the two sides are sufficiently distinct for progress in the one to co-exist, for a time at least, with retrogression in the other. A man may become less in harmony with his surroundings as his ideal rises, and may become more in harmony with them by lowering his ideal.
125. Other things being equal, a man is happier in proportion as he is more in harmony with his environment. In so far, therefore, as our efforts are devoted to the increase of happiness, they will tend to produce a greater amount of harmony between individuals and their environment, and so will be directed to the increase of one moment of the supreme good.
So far, then, the hedonic criterion would be a trustworthy guide.
But there is the other element in the supreme good to be considered. Our ideals must be developed more fully as well as more completely satisfied.
And to this element the criterion of happiness has no necessary or uniform relation.
Very often, indeed, a man is led by desire for his own happiness to actions which develop his ideals towards perfection. A man with a certain taste for music, for example, may be desirous of the intense happiness which music gives to those whose taste is more developed, and may consequently give such time and attention to it, as will make his taste purer and more subtle than before. Or, again, without any desire for a higher musical ideal, he may give his attention to music simply to satisfy the desire which he already has for it, and may, through the knowledge and experience thus gained, find that his appreciation of music has become more discriminating and more intense.
Very often, again, a man develops his own ideals by his desire for the happiness of others. If he educates himself in order that he may support his parents, or serve his country, he will probably find that one effect of his education has been to develop his ideals of knowledge and beauty. Again, benevolence is a disposition which increases by being indulged, and one result of acting for the happiness of others is often to desire that happiness more keenly than before.
There are also the cases where the agent’s action is directed to improving the ideals of another person on the ground that this will conduce to the happiness either of the person improved, or of a third person.
Much of the moral education of children falls under this head. In some cases, no doubt, a quality is inculcated because it is thought desirable per se, but very frequently the reason is to be found in a consideration of the future happiness of the child, or of the people with whom it will associate in after life.
126. But there are cases in which the hedonic criterion would by no means lead us to the development of what we should regard as a higher ideal. It is true that, if we accept Hegel’s principles, and if we see reason to include among them the immortality of the individual, we should be bound to hold that every heightening of the ideal would eventually mean increased happiness. For happiness depends for its amount, not merely on the completeness with which the environment answers to our ideals, but also on the vividness and completeness of those ideals. The more numerous and the more earnest are our wishes, the happier we shall be in their satisfaction, if they are satisfied. The more completely we are self-conscious individuals, the greater will be the happiness and the misery of which we are capable. Since the end of the time-process will be absolute harmony, we may safely assert that anything which makes our ideals more perfect will in the long run be the cause of greater happiness, since it will increase the intensity of our demands, and so of their eventual satisfaction.
But although the complete development of our ideals might be known a priori to involve the greatest happiness, it does not follow that the hedonic criterion would lead us in the direction of the complete development of our ideals. For this coincidence of development and happiness is only known to be certain in the indefinitely remote future, a future far too remote to be known by any empirical calculations. We may be certain that complete development will mean complete happiness. But it by no means follows that, if we aim at the greatest happiness which we can perceive to be attainable by our present action, we shall be aiming in the direction of complete development.
127. And there are many cases in which we should judge that the development of our ideals would indicate a course which would rather diminish than increase happiness. A man is generally admitted to be nearer to perfection in proportion as his love of truth, or his concern for the happiness of others, increases. And yet the love of truth may force us to change very comforting beliefs for very depressing ones. And in so imperfect a state as the present increased sympathy for the happiness or misery of others often produces more misery than happiness for the sympathiser.
Of course the hedonic criterion does not take account of the pleasure of the agent only, but of all people who are affected by the action.
This makes a considerable difference, for it not infrequently happens that a development which makes a person more miserable makes him also more useful. But there are cases where the opposite happens. To lose a false, but inspiriting, belief may diminish a man’s utility as well as his happiness. And, if my chances of helping others are few, an increase of benevolence on my part may deprive mc of much more happiness than it enables me to bestow upon others.
There are circumstances in which an exclusive regard for happiness would lead us not only to shrink from development, but actually to endeavour to fall back in the scale. It would be generally admitted that a man who was chronically under the influence of drugs had fallen, so far as his ideals went, below the level of a man who kept his intellect and will unclouded. And there are men whose physical and mental sufferings are so great that they would be happier – or at least less unhappy – if they were kept continually drugged with opiates. This might increase not only their own happiness, but happiness in general, for a man who is in great and constant pain is not likely to cause much pleasure to anyone, while his condition will certainly cause pain to his friends.
There are thus cases in which the hedonic criterion would direct us to a goal which, as far as we can see, is, in respect of the other moment of the supreme good, something lower, and not higher than the starting-point.
Under such circumstances ought we to follow the hedonic criterion, or to reject it?
128. The question is not put fairly if it is represented as a choice between happiness and perfection. For the happiness is also an element of perfection. The supreme good consists in a complete development of our ideals, and a complete satisfaction of them when developed. We are more perfect in proportion as either of these takes place, and less perfect in proportion as it is wanting. Happiness is not by itself the supreme good, but any happiness, so far as it goes, is good, and any absence of happiness bad.
This comes out more clearly if we take examples in which the happiness at stake is not that of the agent. For so much sin comes from attaching excessive weight to the happiness of the sinner, and morality has to check self-interest so much oftener than to encourage it, that we are apt to fall into the delusion that happiness should not be measured against development. But if we ask whether I ought always to choose to slightly elevate another person’s ideals, at the cost of great suffering to him, or if I ought always to choose to slightly elevate my own ideals, at the cost of great suffering to some one else, it becomes clear that happiness and development are ethically commensurable, and that we have no right to treat a loss of either as ethically indifferent.
Thus the conflict is between two elements of the good. Now we saw above that it was impossible to compare such elements with any hope of discovering which was the most desirable. And in this case the difficulty is greater than in any other, because we are comparing the two primary elements, which exhibit the greatest heterogeneity to be found in the content of the good. How miserable would civilized men have to be, before it would be better for them to change their state for that of happy savages? How much more misery would make it worth their while to accept the passivity of oysters?
129. Common Sense generally deals with this class of questions by judging that a great change for the good in one element will counterbalance a moderate change for the bad in the other. It would approve of a man who sought refuge from extreme physical pain in drugs which left his mind slightly less clear, but not of one who paid this price to avoid a slight discomfort. It would count a keen insight into fallacies as good, although life was thereby made somewhat more dreary, but not if the result was to destroy entirely the happiness of the thinker, and to injure the happiness of his friends.
130. But such a position as this is theoretically indefensible. It implies that we have some means of knowing, within very broad limits; how much happiness will be more worth having than a given degree of development. And it is impossible to settle this. On the other hand the position is so vague that it has very little practical value. For, in most of the cases which present themselves, the gains and losses are not so extreme in proportion to one another as to allow Common Sense to give an opinion at all.
The matter can often be settled, no doubt, by adhering strictly to the hedonic criterion on the ground that we are much more certain of the happiness or the misery than we are of the advance towards, or the retreat from, the goal of a perfectly developed ideal. But this is not always true. It sometimes happens that the retrogression in development, which accompanies the increased happiness, seems beyond all doubt.
131. To sum up – we have seen that a moral criterion is necessary, if any sincere ethical judgment is to be pronounced either right or wrong – that is, if morality is to have any objectivity at all. We have seen that the possible criteria appear to be confined to pleasure and perfection. We have seen that perfection breaks down, if we attempt to use it in this way. Pleasure, on the other hand, does seem to be a possible criterion – difficult, indeed, to apply, but offering no greater difficulties than those which appear to be inherent in ethics. But when we enquire if it is a correct criterion of the good, we find that it only measures one of the two elements into which the good may be analysed.
There are four possible cases. In the first, the action to which the hedonic criterion would guide us, involves in our judgment a greater development of ideals. In this case it is clear that we should take this course, since both elements of the good are increased.
In the second case, our action, whichever way we act, will as far as we can see make no difference to the development of ideals. Here too we can safely abide by the hedonic criterion, since that measures the only element of the good which our decision can be seen to affect.
In the third case, our action may make a considerable difference to the development of our ideals, but we are unable to tell whether the difference will be for good or for evil. Once more we shall do well to follow the hedonic criterion. For then, at any rate, we shall gain in respect of one element of the good. We may indeed lose much more in respect of development. But then we may gain in respect of that element also. Since the effect on development is unknown, the only rational course, if we must act, is to be guided by the effect on happiness, which is known.
But in the fourth case the course to which the hedonic criterion would guide us has in our judgment an unfavourable effect on the development of ideals, as compared with the alternative course. In this case there seems no reasonable solution. For we cannot estimate the quantity of loss to development, and, if we could, we are ignorant of the common standard by which this could be compared with the gain in pleasure.
132. In considering how much uncertainty this brings into ethics, we must remember once more that the question is not limited to the pleasure and the development of the agent but includes the consideration of the pleasure and development of all people affected by the action.
This diminishes the number of cases of the fourth class, for the happiness a man gives is generally more closely proportioned to the development of his ideals than is the happiness he enjoys.
And, again, we must remember that the object of a moral criterion is strictly practical. Its object is to guide our action. It follows from this that it is comparatively unimportant if it fails to indicate which of two events would be the better, in those cases in which our action cannot bring about or hinder either alternative. It is no doubt convenient to know what would be gain and what loss, but the real need to know arrives only when our knowledge can help us to bring about the gain or avoid the loss.
Now the development of our ideals is, in many cases, entirely out of our power, to help or hinder. It is possible that a man might get more pleasure if he could retain his childish taste for sweetmeats, and avoid the growth of a taste for claret. At any rate he could satisfy himself at less expense. But no efforts, on his own part or that of his teachers, will prevent the relative places of sweetmeats and claret in the scale of pleasures being different for the average man from what they were for the average boy.
It is possible, again, that the general religious attitude of the twelfth century gave a greater balance of pleasure than was given by the general religious attitude of the nineteenth century. But if the majority had known this beforehand, and had acted on the most rigidly Utilitarian principles, could their united efforts have averted the Renaissance, the Reformation, or the Illumination? Our desires have a dialectic of their own, and no finite ideal can satisfy us indefinitely. Some we transcend as soon as we have attained them. For others a period of enjoyment is necessary before they pall. In other cases, again, the mere desire for an unattained ideal seems to be sufficient to demonstrate, after a time, its inadequacy. Our volition has, no doubt, a certain influence on this process. But there are many cases in which it would proceed in spite of all our efforts to restrain it. And even if in these cases, the process should diminish happiness, we should do but little harm if we directed our action by the hedonic criterion. For, while such action would be mistaken, it would be also ineffective.
133. But after all these deductions it remains true that there are cases of the fourth class in which our decisions will have a decisive effect on the result, and that ethics offers us no principle upon which to make the decisions. There is thus no possibility of moral action in deciding them.
This is a less revolutionary conclusion than it appears at first sight.
It does not deny that one of the two alternatives is always objectively better than the other. One of the two finite and incompatible goods – the particular gain in pleasure, or the particular gain in development of ideals – would raise us nearer to the supreme good than the other. This is the one to be accepted. But, since they have no common standard but the supreme good, we could only compare them if we knew the exact relation of each of them to the supreme good, and this we do not know.
134. The impossibility of decision arises, then, not from the facts of the case, but from our ignorance about them. Now every system of ethics, with the exception of those which believe in an immediate and unerring intuition for every particular choice, must hold that there are some cases where it is impossible to see what the best course is. If we take the hedonic criterion, there are cases in which the alternative actions seem to present such equal balances of pleasure that it is impossible to see which is the greater. If we take perfection, two incompatible goods may seem so equally good that no reason can be found for choice.
Indeed an ethical system which denied that the best and wisest men were sometimes compelled to act utterly in the dark would he in glaring contradiction to the facts of life.
There is only one difference between the difficulties I have described above as arising on my theory and these others which exist on any theory.
The latter are merely quantitative. They arise from the complexity, or the equality, of data whose nature is not incompatible with a reasoned choice, and which admit of such choice when the instance is simpler or less evenly balanced. In the cases of the fourth class, which I described in Section 131, on the other hand, the problem is one to which the only methods of decision possible to us, in our present imperfect state, do not apply at all.
My theory does thus involve rather more ethical scepticism than the others. But this is of no importance in practice. For in practice the important point is not to know the reason why some moral problems are insoluble. Practice is only concerned to enquire how many, and how serious, are the insoluble problems.
135. And, fortunately, the attainment of the good does not ultimately depend upon action. If it did, it might be rather alarming to think that there were certain cases in which we did not know how to act. But, after all, if it did depend on action, things would be so bad on any theory of ethics that minor differences would be unimportant. If the nature of reality was hostile or indifferent to the good, nothing but the most meagre and transitory gains could ever be made by creatures so weak and insignificant as we should be in such a universe. But if, as Hegel teaches us, that which we recognize as the supreme good is also the supreme reality, then it must inevitably realise itself in the temporal process, and no mistakes of ours can hinder the advance and the eventual attainment.
136. There is therefore nothing in this occasional failure of the only available criterion which should make us think more meanly of reality, or more hopelessly of the good. And we should count it a gain, and not a loss, if it emphasises the inadequacy both of the practice of morality, and of the science of ethics. For this is one of the most profound and important consequences of all metaphysical idealism. Virtue, and the science which deals with it, imply the possibility of sin, they imply action, and they imply time. And they share, therefore, the inadequacy of matter and of the physical sciences. The conception of virtue is, indeed, more adequate than such conceptions as matter and notion. But, like them, it reveals its own imperfection, and, like them, it must be transcended and absorbed before we can reach either the absolutely real or the absolutely good.