R. M. Hare, The Language of Morals (1952)

4. DECISIONS OF PRINCIPLE

4. 1. There are two factors which may be involved in the making of any decision to do something. Of these, the first may at any raite theoretically be absent, the second is always present to some degree. They correspond to the major and minor premisses of the Aristotelian practical syllogism. The major premiss is a principle of conduct; the minor premiss is a statement, more or less full, of what we should in fact be doing if we did one or other of the alternatives open to us. Thus if I decide not to say something, because it is false, I am acting on a principle, 'Never (or never under certain conditions) say what is false', and I must know that this, which I am wondering whether to say, is false.

Let us take the minor premiss first, since it presents less difficulty. We plainly cannot decide what to do unless we know at least something about what we should be doing if we did this or that. For example, suppose that I am an employer, and am wondering whether or not to sack a clerk who habitually turns up at the office after the hour at which he has undertaken to turn up. If I sack him I shall be depriving his family of the money on which they live, perhaps giving my firm a reputation which will lead clerks to avoid it when other jobs are available, and so on; if I keep him, I shall be causing the other clerks to do work which otherwise would be done by this clerk; and the affairs of the office will not be transacted so quickly as they would if all the clerks were punctual. These would be the sorts of consideration that I should take into account in making my decision. They would be the effects on the total situation of the alternative actions, sacking him or not sacking him. It is the effects which determine what I should be doing; it is between the two sets of effects that I am deciding. The whole point about a decision is that it makes a difference to what happens; and this difference is the difference between the effects of deciding one way, and the effects of deciding the other.

It sometimes seems to be implied by writers on ethics that it is immoral, on certain sorts of occasion, to consider the effects of doing something. We ought, it is said, to do our duty no matter what the effects of doing it. As I am using the word 'effects', this cannot be maintained. I am not making a claim for 'expediency' (in the bad sense) as against 'duty'. Even to do our duty — in so far as it is doing something — is effecting certain changes in the total situation. It is quite true that, of tthe changes that it is possible to effect in the total situation, most people would agree that we ought to consider certain kinds more relevant than others (which than which, it is the purpose of moral principles to tell us). I do not think that the immediacy or remoteness of the effects makes any difference, though their certainty or uncertainty does. The reason why it is considered immoral to fail to right an injustice whose effects will maximize pleasure, is not that in such a choice the effects are considered when they should not have been; it is that certain of the effects — namely, the maximization of pleasure — are given a relevance which they should not have, in view of the prior claim of those other effects which would have consisted in the righting of the injustice.

For reasons which will become apparent when we have examined the logic of value-words, it is most important, in a verbal exposition of an argument about what to do, not to allow value-words in the minor premiss. In setting out the facts of the case, we should be as factual as we can. Those versed in the logic of these words, and therefore forewarned against its pitfalls, may in the interests of brevity neglect this precaution; but for the inexperienced it is very much better to keep value-expressions where they belong, in the major premiss. This will prevent the inadvertent admission of an ambiguous middle term, as in the example in 3.3 sub fine. I do not mean that in discussing the facts of the case we should not admit any words which could possibly have an evaluative meaning; for this, in view of the way in which evaluative meaning pervades our language, would be well-nigh impossible. I only mean that we must be sure that, as we are using the words in the minor premiss, there are definite tests (not themselves involving evaluation) for ascertaining its truth or falsity. In the last paragraph I was using the word 'pleasure' in such a sense, though it is not always so used.

4.2. The relation between the two premisses may perhaps be made clearer by considering an artificial example. Let us suppose that a man has a peculiar kind of clairvoyance such that he can know everything about the effects of all the alternative actions open to him. But let us suppose that he has so far formed for himself, or been taught, no principles of conduct. In deciding between alternative courses of action, such a man would know, fully and exactly, between what he was deciding. We have to ask to what extent, if any, such a man would be handicapped, in coming to a decision, by not having any formed principles. It would seem beyond doubt that he could choose between two courses; it would be strange, even, to call such a choice necessarily arbitrary or ungrounded ; for if a man knows to the last detail exactly what he is doing, and what he might otherwise have done, his choice is not arbitrary in the sense in which a choice would be arbitrary if made by the toss of a coin without any consideration of the effects. But suppose that we were to ask such a man 'Why did you choose this set of effects rather than that? Which of the many effects were they that led you to decide the way you did?' His answer to this question might be of two kinds. He might say 'I can't give any reasons; I just felt like deciding that way; another time, faced with the same choice, I might decide differently'. On the other hand, he might say 'It was this and this that made me decide; I was deliberately avoiding such and such effects, and seeking such and such'. If he gave the first of these two answers, we might in a certain sense of that word call his decision arbitrary (though even in that case he had some reason for his choice, namely, that he felt that way); but if he gave the second, we should not.

Let us see what is involved in this second type of answer. Although we have assumed that the man has no formed principles, he shows, if he gives the second answer, that he has started to form principles for himself; for to choose effects because they are such and such is to begin to act on a principle that such and such effects are to be chosen. We see in this example that in order to act on principle it is not necessary in some sense to have a principle already, before you act; it may be that the decision to act in a certain way, because of something about the effects of acting in that way, is to subscribe to a principle of action — though it is not necessarily to adopt it in any permanent sense.

Ordinary men are not so fortunate as the man in our artificial example. They start, indeed, without any knowledge of the future at all; and when they acquire knowledge it is not of this intuitive kind. The kind of knowledge that we have of the future — unless we are clairvoyant — is based upon principles of prediction which we are taught, or form for ourselves. Principles of prediction are one kind of principle of action; for to predict is to act in a certain way. Thus, although there is nothing logically to prevent someone doing entirely without principles, and making all his choices in the arbitrary manner exhibited in the first kind of answer, this never in fact occurs. Moreover, our knowledge of the future is fragmentary and only probable; and therefore in many cases the principles which we are taught or form for ourselves say, not 'Choose this kind of effect rather than that', but 'You do not know for certain what will be the effects; but do this rather than that, and the effects are most likely to be such as you would have chosen, if you had known them'. It is important to remember, in this connexion, that 'likely' and 'probable' are value-words: in many contexts 'It is probable (or likely) that P' is adequately rendered by 'There is good reason (or evidence) for holding that P'.

4.3. We may distinguish, so far, two reasons why we have principles. The first reason applies to anyone, even a man with complete insight into the future, who decides to choose something because it is of a certain character. The second reason applies to us because we do not in fact have complete knowledge of the future, and because such knowledge as we do have involves principles. To these reasons a third must now be added. Without principles, most kinds of teaching are impossible, for what is taught is in most cases a principle. In particular, when we learn to do something, what we learn is always a principle. Even to learn or be taught a fact (like the names of the five rivers of the Punjab) is to learn how to answer a question; it is to learn the principle 'When asked "What are the names of the five rivers of the Punjab?" answer "The Jhelum, the Chenab, &c.".' By this I do not of course mean, that to learn to do anything is to learn to recite by rote some universal imperative sentence. This would involve us in a vicious regress; for learning to recite is a kind of learning, and must have its principles; but in that case we should have to learn to recite the principles of reciting. The point is rather this, that to learn to do anything is never to learn to do an individual act; it is always to learn to do acts of a certain kind in a certain kind of situation; and this is to learn a principle. Thus, in learning to drive, I learn, not to change gear now, but to change gear when the engine makes a certain kind of noise. If this were not so, instruction would be of no use at all; for if all an instructor could do were to tell us to change gear now, he would have to sit beside us for the rest of our lives in order to tell us just when, on each occasion, to change gear.

Thus without principles we could not learn anything whatever from our elders. This would mean that every generation would have to start from scratch and teach itself. But even if each generation were able to teach itself, it could not do so without principles; for self-teaching, like all other teaching, is the teaching of principles. This may be seen by recurring to our artificial example. Let us suppose that our clairvoyant made all his choices on some principle, but always forgot, as soon as he had made the choice, what the principle had been. He would have, accordingly, each time he made a decision, to go over all the effects of the alternative actions. This would be so time-consuming that he would not have the leisure to make many decisions in the course of his life. He would spend his whole time deciding matters like whether to step off with the right or the left foot, and would never reach what we should call the more important decisions. But if he could remember the principles on which he acted, he would be in a much better position; he could learn how to act in certain kinds of circumstance; he could learn to single out quickly the relevant aspects of a situation, including the effects of the various possible actions, and so choose quickly, and in many cases habitually. Thus his powers of considered decision would be set free for more momentous decisions. When the cabinet-maker has learnt how to make a dovetail without thinking much about it, he will have time to think about such things as the proportions and aesthetic appearance of the finished product. And it is the same with our conduct in the moral sphere; when the performance of the lesser duties has become a matter of habit, we have time to think about the greater.

There is a limit in practice to the amount that can be taught to someone by someone else. Beyond this point, self-teaching is necessary. The limit is set by the variety of conditions which may be met with in doing whatever is being taught; and this variety is greater in some cases than in others. A sergeant can teach a recruit almost all there is to be known about fixing bayonets on parade, because one occasion of fixing bayonets on parade is much like another; but a driving instructor cannot do more than begin to teach his pupil the art of driving, because the conditions to be met with in driving are so various. In most cases, teaching cannot consist in getting the learner to perform faultlessly a fixed drill. One of the things that has to be included in any but the most elementary kinds of instruction is the opportunity for the learner to make decisions for himself, and in so doing to examine, and even modify to suit particular types of case, the principles which are being taught. The principles that are taught us initially are of a provisional kind (very like the principle 'Never say what is false' which I discussed in the last chapter). Our training, after the initial stages, consists in taking these principles, and making them less provisional; we do this by using them continually in our own decisions, and sometimes making exceptions to them; some of the exceptions are made because our instructor points out to us that certain cases are instances of classes of exceptions to the principle; and some of the exceptions we decide on for ourselves. This presents no more difficulty than our clairvoyant had in deciding between two sets of effects. If we learn from experiment that to follow a certain principle would have certain effects, whereas to modify it in a certain way would have certain other effects, we adopt whichever form of the principle leads to the effects which we choose to pursue.

We may illustrate this process of modifying principles from the example already used, that of learning to drive. I am told, for instance, always to draw into the side of the road when I stop the car; but later I am told that this does not apply when I stop before turning into a side-road to the offside — for then I must stop near the middle of the road until it is possible for me to turn. Still later I learn that in this manoeuvre it is not necessary to stop at all if it is an uncontrolled junction and I can see that there is no traffic which I should obstruct by turning. When I have picked up all these modifications to the rule, and the similar modifications to all the other rules, and practice them habitually as so modified, then I am said to be a good driver, because my car is always in the right place on the road, travelling at the right speed, and so on. The good driver is, among other things, one whose actions are so exactly governed by principles which have become a habit with him, that he normally does not have to think just what to do. But road conditions are exceedingly various, and therefore it is unwise to let all one's driving become a matter of habit. One can never be certain that one's principles of driving are perfect — indeed, one can be very sure that they are not; and therefore the good driver not only drives well from habit, but constantly attends to his driving habits, to see whether they might not be improved; he never stops learning.1

It is hardly necessary to point out that principles of driving, like other principles, are normally not inculcated by their verbal repetition, but by example, demonstration, and other practical means. We learn to drive, not by precept, but by being shown how to do particular bits of driving; the precepts are usually only explanatory or mnemonic of what we are being shown. Thereafter, we try to do the particular manoeuvres ourselves, and are criticized for failures, commended when we do them well, and so gradually get the hang of the various principles of good driving. For although our instruction is far from being purely verbal, nevertheless what we are being taught are principles. The fact that the derivation of particular acts (or commands to do them) from principles is normally done non-verbally does not show that it is not a logical process, any more than the inference:

The clock has just struck seven times
The clock strikes seven times at seven o'clock only
∴ It is just after seven o'clock
is shown to be non-logical because it is never made explicitly in words.
Drivers often know just what to do in a certain situation without being able to enunciate in words the principle on which they act. This is a very common state of affairs with all kinds of principles. Trappers know just where to set their traps, but often cannot explain just why they have put a trap in a particular place. We all know hnw to use words to convey our meaning; but if a logician presses us for the exact definition of a word we have used, or the exact rules for its use, we are often at a loss. This does not mean that the setting of traps or the use of words or the driving of cars does not proceed according to principles. One may know how, without being able to say how — though if a skill is to be taught, it is easier if we can say how.

We must not think that, if we can decide between one course and another without further thought (it seems self-evident to us, which we should do), this necessarily implies that we have some mysterious intuitive faculty which tells us what to do. A driver does not know when to change gear by intuition; he knows it because he has learnt and not forgotten; what he knows is a principle, though he cannot formulate the principle in words. The same is true of moral decisions which are sometimes called 'intuitive'. We have moral 'intuitions' because we have learnt how to behave, and have different ones according to how we have learnt to behave.

It would be a mistake to say that all that had to be done to a man to make him into a good driver was to tell him, or otherwise inculcate into him, a lot of general principles. This would be to leave out the factor of decision. Very soon after he begins to learn, he will be faced with situations to deal with which the provisional principles so far taught him require modification; and he will then have to decide what to do. He will very soon discover which decisions were right and which wrong, partly because his instructor tells him, and partly because having seen the effects of the decisions he determines in future not to bring about such effects. On no account must we commit the mistake of supposing that decisions and principles occupy two separate spheres and do not meet at any point. All decisions except those, if any, that are completely arbitrary are to some extent decisions of principle. We are always setting precedents for ourselves. It is not a case of the principle settling everything down to a certain point, and decision dealing with everything below that point. Rather, decision and principles interact throughout the whole field. Suppose that we have a principle to act in a certain way in certain circumstances. Suppose then that we find ourselves in circumstances which fall under the principle, but which have certain other peculiar features, not met before, which make us ask 'Is the principle really intended to cover cases like this, or is it incompletely specified — is there here a case belonging to a class which should be treated as exceptional?' Our answer to this question will be a decision, but a decision of principle, as is shown by the use of the value-word 'should'. If we decide that this should be an exception, we thereby modify the principle by laying down an exception to it.

Suppose, for example, that in learning to drive I have been taught always to signal before I slow down or stop, but have not yet been taught what to do when stopping in an emergency; if a child leaps in front of my car, I do not signal, but keep both hands on the steering-wheel; and thereafter I accept the former principle with this exception, that in cases of emergency it is better to steer than to signal. I have, even on the spur of the moment, made a decision of principle. To understand what happens in cases like this is to understand a great deal about the making of value-judgements.

4.4. I do not wish to seem to be pressing too far my comparison, in respect of the way in which they are learnt, between principles of driving and principles of conduct. It is necessary also to bear in mind some distinctions. In the first place, the expression 'good driver' is itself ambiguous in that it is not immediately clear what standard is being applied. It might be simply a standard of expertness; we might call a person a good driver if he were able to do just what he wanted with his car; we might say 'Although a very good driver, he is most inconsiderate to other road users'. On the other hand, we sometimes expect a good driver to have moral qualities as well; we do not, according to this criterion, call a man a good driver if he drives expertly, but without the slightest heed for the convenience or safety of other people. The line between these two standards of good driving is not easy to draw in practice. There is also a third standard, according to which a driver is said to be good if he conforms to the accepted principles of good driving as laid down, for example, in the Highway Code. Since the Highway Code is compiled with a definite purpose in view, this standard coincides to a great extent with the second.

Secondly, there are two ways of looking at driving instruction:

We establish at the beginning certain ends, for example the avoidance of collisions, and instruction consists in teaching what practices are conducive to those ends. According to this way of looking at them, the principles of good driving are hypothetical imperatives.
We teach at first simple rules of thumb, and the learner only gradually comes to see what the ends are, at which the instruction is aimed.
It must not be thought that either (1) or (2) by itself gives a complete account of our procedure. Which method we adopt depends to a great extent on the maturity and intelligence of the learner. In teaching African soldiers to drive, we might incline more to the second method; if I had to teach my two-year-old son to drive, I should have to adopt the same methods as I now adopt for teaching him to refrain from interfering with the controls when I am driving myself. With a highly intelligent learner, on the other hand, we may adopt a method which has more of (1) in it than of (2). It must not be thought, however, that method (2) is ever entirely without a place even in the case of the most rational of learners. It may be that the desirability of avoiding collisions is at once understood and accepted even by comparatively stupid learners; but there are a great many more ends than this which a good driver has to aim at. He has to avoid causing many kinds of avoidable inconvenience both to himself and to others; he has to learn not to do things which result in damage to his vehicle, and so on. It is of no use to establish at the beginning a general end, 'the avoidance of avoidable inconvenience'; for 'inconvenience' is a value-word, and until he has had experience of driving, the learner will not know what sorts of situation are to count as avoidable inconvenience. The general end or principle is vacuous until by our detailed instruction we have given it content. Therefore it is always necessary to start, to some extent, by teaching our learner what to do, and leaving it for him to find out later why. We may therefore say that although moral principles, which are normally taught us when we are immature, are taught largely by method (2), and principles of driving preponderantly by method (1), there is not an absolute division between the two sorts of principle in this respect. What I have just said about first learning what to do, and about the initial vacuity of the general end, is borrowed from Aristotle.2 The one fundamental distinction between principles of driving and principles of conduct is that the latter are, in Aristotle's term, 'architectonic' of the former; for the ends of good driving (safety, the avoidance of inconvenience to others, the preservation of property, and so on) are justified ultimately, if justification is sought, by appeal to moral considerations.3
It would be folly, however, to say that there is only one way of learning a skill or any other body of principles, or of justifying a particular decision made in the practice of it. There are many ways, and I have tried to make the above account sufficiently general to cover all of them. It is sometimes said by writers on morals that we have to justify an act by reference to its effects, and that we tell which effects are to be sought, which avoided, by reference to some principle. Such a theory is that of the utilitarians, who bid us look at the effects, and examine these in the light of the principle of utility, to see which effects would maximize pleasure. Sometimes, on the other hand, it is said (as by Mr. Toulmin)4 that an act is justified directly by reference to the principles which it observes, and these principles in their turn by reference to the effects of always observing them. Sometimes it is said that we should observe principles and ignore the effects — though for the reasons given above 'effects' cannot be here intended in the sense in which I have been using it. What is wrong with these theories is not what they say, but their assumption that they are telling us the only way to justify actions, or decide what actions to do. We do, indeed, justify and decide on actions in all these ways; for example, sometimes, if asked why we did A, we say, 'Because it was a case falling under principle P', and if asked to justify P in turn, we go into the effects of observing it and of not observing it. But sometimes, when asked the same question 'Why did you do A?' we say 'Because if I hadn't, E would have happened', and if asked what was wrong about E happening, we appeal to some principle.

The truth is that, if asked to justify as completely as possible any decision, we have to bring in both effects — to give content to the decision — and principles, and the effects in general of observing those principles, and so on, until we have satisfied our inquirer. Thus a complete justification of a decision would consist of a complete account of its effects, together with a complete account of the principles which it observed, and the effects of observing those principles — for, of course, it is the effects (what obeying them in fact consists in) which give content to the principles too. Thus, if pressed to justify a decision completely, we have to give a complete specification of the way of life of which it is a part. This complete specification it is impossible in practice to give; the nearest attempts are those given by the great religions, especially those which can point to historical persons who carried out the way of life in practice. Suppose, however, that we can give it. If the inquirer still goes on asking 'But why should I live like that?' then there is no further answer to give him, because we have already, ex hypothesi, said everything that could be included in this further answer. We can only ask him to make up his own mind which way he ought to live; for in the end everything rests upon such a decision of principle. He has to decide whether to accept that way of life or not; if he accepts it, then we can proceed to justify the decisions that are based upon it; if he does not accept it, then let him accept some other, and try to live by it. The sting is in the last clause. To describe such ultimate decisions as arbitrary, because ex hypothesi everything which could be used to justify them has already been included in the decision, would be like saying that a complete description of the universe was utterly unfounded, because no further fact could be called upon in corroboration of it. This is not how we use the words 'arbitrary' and 'unfounded'. Far from being arbitrary, such a decision would be the most well-founded of decisions, because it would be based upon a consideration of everything upon which it could possibly be founded.

It will be noticed how, in talking of decisions of principle, I have inevitably started talking value-language. Thus we decide that the principle should be modified, or that it is better to steer than to signal. This illustrates the very close relevance of what I have been saying in the first part of this book to the problems of the second part; for to make a value-judgement is to make a decision of principle. To ask whether I ought to do A in these circumstances is (to borrow Kantian language with a small though important modification) to ask whether or not I will that doing A in such circumstances should become a universal law.5 It may seem a far cry from Kant to Professor Stevenson; but the same question could be put in other words by asking 'What attitude shall I adopt and recommend towards doing A in such circumstances?'; for 'attitude', if it means anything, means a principle of action. Unfortunately Stevenson, unlike Kant, devotes very little space to the examination of this first-person question; had he paid due attention to it, and avoided the dangers of the word 'persuasive', he might have reached a position not unlike that of Kant.

4.5. As Kant points out in the important passage on the Autonomy of the Will, to which I referred earlier, we have to make our own decisions of principle.6 Other people cannot make them for us unless we have first decided to take their advice or obey their orders. There is an interesting analogy here with the position of the scientist, who also has to rely on his own observations. It might be said that there is a difference here between decisions and observations, to the detriment of the former, in that an observation, once made, is public property, whereas decisions have to be made by the agent himself on each occasion. But the difference is only apparent. A scientist would not have become a scientist unless he had convinced himself that the observations of other scientists were in general reliable. He did this by making some observations of his own. When we learnt elementary chemistry at school, we had some theoretical periods and some practical. In the theoretical periods we studied books; in the practical periods we made experiments, and found, if we were lucky, that the results tallied with what the books said. This showed us that what the books said was not all nonsense; so that even if, by reason of disturbing factors ignored by us, our experiments came out wrong, we were inclined to trust the books and acknowledge that we had made a mistake. We were confirmed in this assumption by the fact that we often discovered later what the mistake had been. If our observations, however carefully we did them, were always at variance with the textbooks, we should not be tempted to make science our profession. Thus the confidence of the scientist in other people's observations is ultimately based, among other things, on his own observations and his own judgements about what is reliable. He has in the end to rely on himself.

The case of the moral agent is not dissimilar. When in our early days we are given our elementary moral instruction, there are some things that we are told, and some things that we do. If, when we did as we were told, the total effects of our so doing, when they happened, were always such as we would not have chosen, had we known, then we should seek better advice, or, if prevented from so doing, either work out our own salvation or become moral defectives. If we are in general given what we subsequently come to see to have been good advice, we decide in general to follow the advice and adopt the principles of those who have given us this good advice in the past. This is what happens to any child who is well brought up. Just as the scientist does not try to rewrite all that is in the textbooks, but takes that for granted and sticks to his own particular researches, so this fortunate child will take over bodily the principles of his elders and adapt them in detail, by his own decisions, to suit his own circumstances from time to time. This is how in a well-ordered society morality remains stable, and at the same time gets adapted to changing circumstances.

4.6. There are, however, many ways in which this happy state of affairs can deteriorate. Let us consider a process that seems to occur quite often in history; it occurred in .Greece during the fifth and fourth centuries, and it has occurred in our own time. Suppose that the people of a certain generation — I will call it the first generation — have got very settled principles, inherited from their fathers. Suppose that they have become so settled as to be second nature, so that generally speaking people act on the principles without thinking, and their power of making considered decisions of principle becomes atrophied. They act always by the book, and come to no harm, because the state of the world in their time remains much the same as that for which the principles were thought out. But their sons, the second generation, as they grow up, find that conditions have changed (e.g. through a protracted war or an industrial revolution), and that the principles in which they have been brought up are no longer adequate. Since, in their education, much stress has been laid on observing principles, and very little on making the decisions on which these principles are ultimately based, their morality has no roots, and becomes completely unstable. Books on 'The Whole Duty of Man' are no longer written or read. Often, when they do what it says in such books, they subsequently find cause to regret their decisions; and there are too many cases of this kind for any confidence in the old principles, as a body, to remain. No doubt there are among these old principles certain very general ones, which will remain acceptable unless human, nature and the state of the world undergo a most fundamental change; but the second generation, not having been brought up to make decisions of principle, but to do what it says in the book, will not, most of them, be able to make those crucial decisions which would determine which principles to keep, which to modify, and which to abandon. Some people, the Polemarchuses of the second generation, will have been so steeped in the old principles that they just follow them come what may; and these will on the whole be more fortunate than the others, for it is better to have some principles, even if they sometimes lead to decisions which we regret, than to be morally adrift. The bulk of the second generation, and still more perhaps of the third, will not know which of the principles to keep and which to reject; and so they will come more and more to live from day to day — not a bad thing, because it trains their powers of decision, but it is an unpleasant and dangerous state to be in. A few among them, the rebels, will shout from the housetops that some or all of the old moral principles are worthless; some of these rebels will advocate new principles of their own; some will have nothing to offer. Though they increase the confusion, these rebels perform the useful function of making people decide between their rival principles; and if they not only advocate new principles, but sincerely try to live by them, they are conducting a moral experiment which may be of the utmost value to man (in which case they go down in history as great moral teachers), or may, on the other hand, prove disastrous both to them and to their disciples.

It may take several generations for this disease to play itself out. Morality regains its vigour when ordinary people have learnt afresh to decide for themselves what principles to live by, and more especially what principles to teach their children. Since the world, though subject to vast material changes, changes only very slowly in matters that are fundamental from the moral point of view, the principles which win the acceptance of the mass of people are not likely to differ enormously from those which their fathers came to distrust. The moral principles of Aristotle resemble those of Aeschylus more than they differ from them, and we ourselves shall perhaps come back to something recognizably like the morality of our grandfathers. But there will be some changes; some of the principles advocated by the rebels will have been adopted. That is how morality progresses — or retrogresses. The process is, as we shall see, reflected by very subtle changes in the uses of value-words; the impossibility of translating Aristotle's catalogue of virtues into modern English may serve as an example, and the disappearance without I trace of the word 'righteous' may serve as another.

4.7. The question 'How shall I bring up my children?' which we have mentioned, is one to the logic of which, since ancient times, few philosophers have given much attention. A child's moral upbringing has an effect upon him which will remain largely untouched by anything that happens to him thereafter. If he has had a stable upbringing, whether on good principles or on bad ones, it will be extremely difficult for him to abandon those principles in later life — difficult but not impossible. They will have for him the force of an objective moral law; and his behaviour will seem to give much evidence in support of intuitionist ethical theories, provided that it is not compared with the behaviour of those who stick just as firmly to quite different principles. But nevertheless, unless our education has been so thorough as to transform us into automata, we can come to doubt or even reject these principles; that is what makes human beings, whose moral systems change, different from ants, whose 'moral system' does not. Therefore, even if for me the question 'What shall I do in such and such a situation?' is almost invariably answered without ambiguity by the moral intuition which my upbringing has given me, I may, if I ask myself 'How shall I bring up my children?' pause before giving an answer. It is here that the most fundamental moral decisions of all arise; and it is here, if only moral philosophers would pay attention to them, that the most characteristic uses of moral words are to be found. Shall I bring up my children exactly as I was brought up, so that they have the same intuitions about morals as I have? Or have circumstances altered, so that the moral character of the father will not provide a suitable equipment for the children? Perhaps I shall try to bring them up like their father, and shall fail; perhaps their new environment will be too strong for me, and they will come to repudiate my principles. Or I may have become so bewildered by the strange new world that, although I still act from force of habit on the principles that I have learnt, I simply do not know what principles to impart to my children, if, indeed, one in my condition can impart any settled principles at all. On all these questions, I have to make up my mind; only the most hide-bound father will try to bring up his children, without thinking, in exactly the way that he himself was brought up; and even he will usually fail disastrously.

Many of the dark places of ethics become clearer when we consider this dilemma in which parents are liable to find themselves. We have already noticed that, although principles have in the end to rest upon decisions of principle, decisions as such cannot be taught; only principles can be taught. It is the powerlessness of the parent to make for his son those many decisions of principle which the son during his future career will make, that gives moral language its characteristic shape. The only instrument which the parent possesses is moral education — the teaching of principles by example and precept, backed up by chastisement and other more up-to-date psychological methods. Shall he use these means, and to what extent? Certain generations of parents have had no doubts about this question. They have used them to the full; and the result has been to turn their children into good intuitionists, able to cling to the rails, but bad at steering round corners. At other times parents — and who shall blame them? — suffer from lack of confidence; they are not sure enough what they themselves think, to be ready to impart to their children a stable way of life. The children of such a generation are likely to grow up opportunists, well able to make individual decisions, but without the settled body of principles which is the most priceless heritage that any generatioiican leave to its successors. For, though principles are in the end built upon decisions of principle, the building is the work of many generations, and the man who has to start from the beginning is to be pitied; he will not be likely, unless he is a genius, to achieve many conclusions of importance, any more than the average boy, turned loose without instruction upon a desert island, or even in a laboratory, would be likely to make any of the major scientific discoveries.

The dilemma between these two extreme courses in education is plainly a false one. Why it is a false one is apparent, if we recall what was said earlier about the dynamic relation between decisions and principles. It is very like learning to drive. It would be foolish, in teaching someone to drive, to try to inculcate into him such fixed and comprehensive principles that he would never have to make an independent decision. It would be equally foolish to go to the other extreme and leave it to him to find his own way of driving. What we do, if we are sensible, is to give him a solid basis of principles, but at the same time ample opportunity of making the decisions upon which these principles are based, and by which they are modified, improved, adapted to changed circumstances, or even abandoned if they become entirely unsuited to a new environment. To teach only the principles, without giving the opportunity of subjecting them to the learner's own decisions of principle, is like teaching science exclusively from textbooks without entering a laboratory. On the other hand, to abandon one's child or one's driving-pupil to his own self-expression is like putting a boy into a laboratory and saying 'Get on with it'. The boy may enjoy himself or kill himself, but will probably not learn much science.

The moral words, of which we may take 'ought' as an example, reflect in their logical behaviour this double nature of moral instruction — as well they may, for it is in moral instruction that they are most typically used. The sentences in which they appear are normally the expression of decisions of principle — and it is easy to let the decisions get separated, in our discussion of the subject, from the principles. This is the source of the controversy between the 'objectivists', as intuitionists sometimes call themselves, and the 'subjectivists', as they often call their opponents. The former lay stress on the fixed principles that are handed down by the father, the latter on the new decisions which have to be made by the son. The objectivist says 'Of course you know what you ought to do; look at what your conscience tells you, and if in doubt go by the consciences of the vast majority of men'. He is able to say this, because our consciences are the product of the principles which our early training has indelibly planted in us, and in one society these principles do not differ much from one person to another. The subjectivist, on the other hand, says 'But surely, when it comes to the point — when I have listened to what other people say, and given due weight to my own intuitions, the legacy of my upbringing — I have in the end to decide for myself what I ought to do. To deny this is to be a conventionalist; for both common moral notions and my own intuitions are the legacy of tradition, and — apart from the fact that there are so many different traditions in the world — traditions cannot be started without someone doing what I now feel called upon to do, decide. If I refuse to make my own decisions, I am, in merely copying my fathers, showing myself a lesser man than they; for whereas they must have initiated, I shall be merely accepting.' This plea of the subjectivist is quite justified. It is the plea of the adolescent who wants to be adult. To become morally adult is to reconcile these two apparently conflicting positions by learning to make decisions of principle; it is to learn to use 'ought'-sentences in the realization that they can only be verified by reference to a standard or set of principles which we have by our own decision accepted and made our own. This is what our present generation is so painfully trying to do.

Notes
1 Cf. Romans 221.

2 Nicomachean Ethics, i. 4.

3 Op. cit. i. 1, 2.

4 Reason in Ethics, pp. 144 ff.

5 Cf. Groundwork of the Metaphysics of Morals, tr. H. J. Paton, p. 99.

6 Op. cit., pp.108 ff.