We have now considered a number of primitive ways of life. Each of them we regarded as being, in ultimate analysis, the result of the attempts of a particular people to give concrete embodiment to their idea of the good life. I have tried to show that the peoples who developed these ways of life and on the whole find them good are people like ourselves, with the same endowment of reason and passions, the same nature and needs, so that any differences between their ways of life, or between them and our own, cannot be explained as due to differences between the natures or the minds of the peoples who entertain them. We have also seen that each of them has a secular morality, that their moral ideas and ideals, the virtues they approve and the vices they condemn, do not derive their authority from religion, and that most of their moral rules are not regarded as obligatory merely or even mainly because of fear of supernatural powers.

We may assume, therefore, that the primitive moral consciousness proceeds on the same principle or principles as our own, and that its judgements are as much entitled to be regarded as deliverances of the moral consciousness as ours. What light, then, do they throw on the nature of morality, the principle or principles of moral judgement, the nature of the moral ideal, the authority of moral rules, the conditions of moral progress? How far are they consistent with, or capable of being explained by, or give support to recent and contemporary ethical theories which are based mainly on the study of the ways of life of white men and the moral ideas and ideals of Western civilisation? How far do they confirm or disprove the provisional theory of the structure of the moral life which I outlined in an earlier lecture? If the primitive moral consciousness proceeds on the same principle as our own, how are we to account for the great difference between many of its moral judgements and ours? These are some of the questions to which the consideration of primitive morality gives rise. I have time to consider only a few of the most central of them.

It might be thought more appropriate that I should devote all my time to the task of making explicit and illustrating the ethical theory which primitive moralities seem to suggest and support, rather than to showing that they are inconsistent with certain other ethical theories—always a thankless task. There are, however, certain advantages in following the latter, more indirect, course. For we are concerned not merely with primitive moral judgements but with the moral judgements of all men everywhere. We have, therefore, to consider how far any theory suggested by an analysis of primitive morality can account for civilised moral judgements; just as we have to examine whether theories suggested by civilised morality can account for primitive moral judgements. And in considering civilised moral judgements, it would be unwise to dispense with the help of the careful and critical analyses of them given by recent ethical writers.

Moreover, the attempt to compare the anthropological evidence regarding primitive morality with some widely held contemporary ethical theories will help to bring into clearer relief than I have hitherto done certain aspects of primitive morality itself. It will also help to bring to light the differences between the presuppositions of the anthropologists, and of the methods which they have developed in their attempts to understand primitive ways of life, and some of the assumptions which not only underlie the methods of many contemporary ethical writers, but are also widespread in our ways of thinking about moral phenomena. It is desirable, therefore, that we should see the full range of the facts for which a satisfactory ethical theory must account—the moral judgements of all men everywhere—and that we should consider the presuppositions and methods of procedure of different ethical writers in the light of them. Moreover, if it is found that the anthropological evidence regarding primitive moral judgements is, in important respects, inconsistent with certain widely held contemporary theories, and with some common assumptions about moral facts which they presuppose, this may dispose the reader to consider more favourably the claims of the alternative theory which I am suggesting, a theory which, though not entirely without adherents among contemporary philosophers, is certainly very much out of fashion at present.

I think the most fundamental difference between the views of anthropologists dealing with primitives and most recent and contemporary ethical theorists is to be found in the difference between the answers which they respectively give to the question: What is the simplest unit of conduct which has to be taken into consideration in passing a considered moral judgement, or in trying to understand the moral judgement of another? As I pointed out earlier, a moral judgement is a final or ultimate judgement, one from which there is no appeal to a higher court or a wider context. What, then, is the simplest entity, which is sufficiently self-contained to warrant a final judgement of rightness or goodness—a judgement which is not liable to require alteration or modification in the light of any further facts or factors?

Recent ethical theorists tend to regard the units in question as relatively simple. They tend to use what I have called the method of isolation. They regard isolated, or at least isolable, elements in the moral life or moral ideal as capable of sustaining such a judgement. Thus it is held that the obligatoriness of certain moral rules can be recognised as self-evident, when the rules are considered by themselves apart from any context. The judgement that certain sorts of acts are right is regarded as final or ultimate, in the sense that it does not need any reason or justification beyond itself. It is true that the obligatoriness of such acts is no longer regarded as absolute but as prima facie, i.e. they are actually obligatory unless they conflict with more urgent obligations; but the judgement that they have this obligatoriness is infallible or self-justifying without reference to any context. Again, certain ends or states of affairs or experiences are regarded as intrinsically good, by which is meant that they would be good even if they existed quite alone, even if there were nothing else in the universe. We find this view in a variety of forms—not only in the Ideal Utilitarianism of Moore, or combined with Intuitionism as by Ross, but also in some theories of the self-realisation type. Urban,1 e.g., gives a list of values which, when considered by themselves, can, he thinks, be arranged on a scale which has universal validity.

We are not here concerned with the details of these views but only with the method of isolation by which they try to get rules, ends or values which have by themselves an absolute or self-evident claim that they should be obeyed, pursued or realised. It is worth pointing out, however, that, even if we could get such atomic absolutes the task of the moral agent would not be so easy as it might at first sight appear. For an act is not normally just an instance of one rule or just the pursuit of one end or just the realisation of one value. It may also be the violation of another rule, the neglect of another end or the denial of another value. And the real problem of the moral agent, in trying to discover what he ought to do, is to discover which rule is the more urgent, which intrinsic good or sum of intrinsic goods is the greater, which act will realise the higher value. The theories in question suggest no principle by which the moral agent can make his choice. They counsel him merely to do his best and trust his fallible judgement. But until some principle is forthcoming on which he can make this, his most crucial choice, the principles ‘obey the most urgent rule’, ‘produce the greatest amount of intrinsic good’ or ‘realise the higher value’ are of little practical use to the moral agent.

What concerns us, however, is whether such atomic absolutes, which contain the grounds of their rightness or goodness within themselves, can in fact be found. The unanimous and oft-repeated view of recent and contemporary social anthropologists is that, among primitives, at least, they cannot. They hold in effect that the results of any attempt to apply the method of isolation to the moral judgements of primitives are not only valueless but misleading. Indeed, as we have seen, their main criticism of the earlier anthropologists is just that they used the method of isolation, and tried to understand primitive actions and customs and value-judgements without taking account of the cultural context which alone renders them intelligible. According to their view, what we find among primitives is not rules whose obligatoriness is self-evident when they are considered in isolation, but rules which are the conditions of the working of certain interrelated institutions; not acts which have intrinsic goods as consequences, but acts which are good in their context; not a scale of values whose relative order can be decided in abstraction, but a system of values embodied in a way of life which determines their order of preference at any given moment.

I have already repeatedly quoted the views of anthropologists that we cannot understand any item in a primitive culture without taking account of its concrete setting in the context of the way of life of which it forms part; and, in my description of the ways of life of particular primitive peoples, I illustrated in some detail the intimate interconnections between the elements of their ways of life. We see it in the items on their scales of value, in the provision which they make for different human needs, in the ends they pursue as good and in the duties which they consider binding. None of these can be understood by itself.

This of course does not mean that it is not legitimate, and even necessary, to distinguish the different elements in a way of life and to consider them separately. What it does mean is that we cannot appreciate the way in which they appear to the person who lives the life and passes judgements of value on them, unless we look at them in their context. Nor does it mean that the ways of life of most, or indeed perhaps of any, peoples are the consistent expression of one principle. Their degree of coherence varies from people to people, and the unity of the resulting patterns generally takes the form of functional interdependence between the parts, rather than the expression of a rationally coherent plan of life. Similar variations are to be found in the extent to which different individuals, even in the same community, enter into the spirit of their way of life and appreciate its pattern as a whole. But whatever be the degree of integration of the pattern and to whatever extent individuals appreciate it as a whole, the anthropological evidence suggests that their conception of it provides the operative ideal which determines their duties and obligations, that what appears to them right and good is conditioned by it, and that, therefore, however immediate and intuitive their judgements may seem, they are not unconditional or self-justifying, but at least partly dependent on, and conditioned by, the background of the way of life of the people concerned. In the discussion of moral questions this background is normally taken for granted; and as long as it is common to all the individuals concerned in the discussion, the neglect of it, and of the way in which it conditions their judgements, need occasion no difficulty. But when we consider the differences between the value-judgements of different peoples, and especially of primitives and ourselves, the relevance of the different ways of life which they presuppose becomes clear. We find that they are all conditioned by the context in relation to which they are made. The only judgement which is strictly unconditional and self-justifying is that passed on a way of life as a whole.2

Now if this view is sound and if it applies to civilised as well as to primitive morality, however useful the analyses of the contemporary ethical theorists to whom I referred may be for certain purposes, their prima facie obligations, their intrinsic goods, and their scales of value are the results of undue simplification and abstraction, and the judgements on acts and ends to which their method of isolation leads must be reconsidered, and it may be corrected, when the acts and ends are looked at in the context of the way of life in which they actually occur. For what the view of the anthropologists really amounts to is that the simplest self-contained unit of conduct, which can justify or render intelligible a final moral judgement, is a way of life as a whole, or at least a very substantial part of such a way of life. I want, therefore, to consider in some detail how far theories which proceed by the method of isolation are consistent with the facts of primitive morality; and how far the views of the anthropologists about primitive morality apply also to civilised morality; and, if they do, what implications this has for ethical theory. In this lecture I propose to discuss these questions with special reference to moral rules. In the next I shall consider them with reference to goods and scales of value.

Those who proceed by the method of isolation contend that certain moral rules are self-evident, unmediated, containing the grounds of their rightness within themselves. Now rules which have this character should be recognised as such by all who have the capacity to grasp the concepts involved in them and have sufficiently attended to them. And if such rules are recognised at all, they should be recognised as being strictly universal. For their rightness is independent of any context and, therefore, holds in every context. Their rightness is independent of, and additional to, the goodness of any state of affairs to which their observance leads, or of which it may be a condition. That certain moral rules have this self-evidence and strict universality and independent rightness is the claim of the intuitionists. What we have to consider is how far this claim is consistent with the facts about primitive morality.

In considering this question we have to ask: (1) How far do primitives recognise as binding the same rules as we do, in particular those which intuitionists claim to be self-evident? And (2) so far as they do recognise them, how far do they regard them as self-evident or self-authenticating? There is a way of interpreting these rules according to which it may be said that all primitives recognise some, and some primitives recognise all of them, but only within definitely prescribed limits. And this seems to me to be sufficient to show that they understand the concepts used in them and have sufficiently attended to them to recognise them as self-evident, if they are in fact self-evident. They have satisfied the two conditions usually regarded as necessary and sufficient for such recognition.

What, then, is the reservation subject to which they recognise them? Primitives do not recognise the rules which they regard as right as universal in the sense that they apply to their relations with all men. There is always, or almost always, a group of men beyond which they do not regard themselves as having any duty to the outsider. This in-group may be the family, the clan, the tribe, the group of tribes, the inhabitants of a certain locality, and so on; but there is always an in-group and an out-group; and the members of the latter are not regarded as subjects of rights or objects of duties.

Is this state of affairs consistent with the view that the rightness of the rules in question is self-evident, and that the rules are therefore strictly universal and applicable to the relations between all moral beings? It seems to me that there are two, and only two, possible ways in which the intuitionist might try to show that it is; but I do not think either of them will account for the known facts.

One way is to hold that, in so far as the primitive recognises the outsider as a fellow-man, he does recognise prima facie obligations to him; but that fear of the outsider gives rise to a stronger obligation to promote his own welfare and that of his group, and that this overrides the obligations which he recognises towards the outsider as a man. Now it is quite true that many primitive people entertain very strange ideas about outsiders, even about those who live beyond a mountain range, an arm of the sea or a belt of forest, let alone about those who live in more remote parts. They are unknown and the unknown is apt to be a source of danger and dread. It is also true that many primitives have good reasons to fear the hostility of outsiders. But there is no evidence whatsoever that they do recognise obligations to the outsider, and that they are just overborne by stronger obligations to the in-group. All the evidence points to the opposite conclusion, that they recognise no obligations to the genuine outsider and that they feel no compunction in treating him in ways which involve breaking every rule in the moral code of the in-group.

The other possible way of trying to reconcile the facts with the intuitionist theory is to contend that primitives do not recognise outsiders as human or at least as moral beings like themselves. It is, however, difficult to accept this contention. For as Goldenweiser has pointed out,3 it is difficult to believe that one dog will recognise another as his fellow whatever the difference in size or shape or colour or breed, and that a man will not do the same. But even if we accept this contention, while it might explain the failure of primitives to recognise any obligations towards persons from remote parts of the earth whose language and habits they do not understand, and about whose purposes and powers they are ignorant and often entertain very strange notions, it would not explain their failure to recognise such obligations—if they are really self-evident—to members of neighbouring peoples with whom their relations are relatively intimate and often not unfriendly. It is not true that all primitives regard their neighbours as dangerous enemies. Some of them are quite friendly and unwarlike, and yet most of them do not recognise any obligations to those outside the group.

Moreover, if this contention were true, members of one group would recognise either all or none of the moral rules of the in-group as binding in their relations to members of another group. But this is not what we in fact seem to find. The line of demarcation between the in-group and the out-group is not so clearly defined as it is often represented. There are sometimes intermediate groups to whom, or to some of whom, either all or some of the members of a particular community recognise some obligations and not others. For example, the Trobrianders recognise the inhabitants of neighbouring islands such as the Amphletts and Dobu as human beings like themselves, and they engage in some forms of exchange with them as in the institution of the Kula. The partners in these exchanges recognise duties to one another, duties of promise-keeping, protection, hospitality, friendship and so on, but outside such partnerships there seems to be no recognition of any moral rules as applicable to the relations between the members of the communities concerned. They will cheat and deceive and, until the white man prohibited it, they would kill one another without any sort of compunction. Thus a Trobriander will recognise moral rules as applicable to his relations to a native of Dobu who is his partner in the institution of the Kula, but not in his relations to him simply as a man or a Dobuan. In other words, he finds some forms of co-operation with the members of another group good and he recognises as obligatory, in his relations to them, the rules required to make such co-operation possible; and the close association with them which such co-operation involves creates a situation in which man's natural interest in his fellow-men and their welfare has an opportunity to assert itself; and as a result other moral rules than those which are involved in the particular form of co-operation are extended to his relations with them. Again, many primitive people recognise a duty of hospitality to the stranger and of respect for the person of messengers and heralds, while they may recognise no other duties either to the individuals concerned or to the other members of the groups from which they come. We find a similar state of affairs in the relations of some natives to the white men who live among them and with whom they are on friendly terms. They seem to regard some but not others of their moral rules as applicable to their relations with them. Thus the Trobrianders feel gratitude for benefit received from white men and an obligation to make a return for them; but while they condemn stealing among themselves, Malinowski tells us that they do not regard stealing from the white man as “a breach of law, morality, or gentlemanly manners”.4

I could quote many other examples, both from the peoples whose ways of life I have already described and from other primitive peoples, to show that all the moral rules which they regard as right they do not regard as applicable to their relations with all whom they recognise as human beings like themselves. But I must be content with one further illustration. I take it from one of the very few peoples whose attitude to breaches of their moral and social rules, both among the members of a group and between the members of different groups, has been the subject of a special investigation.5 They are the inhabitants of Wogeo, a small island off the coast of New Guinea. The island is about fifteen miles in circumference and has about 900 inhabitants; and all, or almost all, the adults among them know one another. They are relatively untouched by Western civilisation. When Hogbin went to carry out his investigation in 1934, he was the first white man who had ever lived on the island. The island is divided into five districts, each of which consists of several villages, and in each village there are two and in some cases three clans. For some purposes the clan is the unit, and the relations between clansmen are very close and intimate; but for many other purposes the unit is the village, and the natives recognise the implications of their common citizenship and the conditions necessary for co-operation. Members of different districts, while recognising one another as human beings, tend to regard each other as strangers if not foreigners; but the same culture and customs prevail throughout the island.

Hogbin illustrates the state of affairs which he found by the attitude of the natives to theft and adultery, which he tells us is typical of their attitude to breaches of moral rules in general. In the abstract all the natives regard theft and adultery as wrong. They have been so taught from their youth up, They strongly disapprove of them between the members of a clan or village, whether their own or another makes no difference. Their disapproval is less marked when they are committed against a member of another village within the same district; and when they are committed against someone from another district they are not considered wrong at all. For example, they will openly boast of adultery with the wife of a man from another district; but they feel embarrassed and ashamed in the presence of a member of their own district, and especially of their own clan, whom they have wronged in this way; even though he is not aware, and they know he is not aware, that they have done so.

The same principle applies to their attitude to theft. For instance, some of them killed and ate a pig belonging to a man from another district which had strayed into their village. In discussing this incident with Hogbin the natives did not try to find extenuating circumstances for their action; they just denied that it was theft, though they would have regarded it as theft if the pig belonged to a man from their own district. Hogbin succeeded in getting the more thoughtful of them to admit that it really was theft, and in recognising the inconsistency of saying that theft is wrong and yet that killing and eating a pig belonging to a man from another district is not. He says the natives found it difficult to explain away this inconsistency; but no doubt they regarded the act as theft in the sense of taking another man's property, but not as theft in the sense of the wrongful taking of another man's property, i.e. not as theft in the sense in which theft is morally wrong. Be this as it may, we are told that they still continued to denounce action of this kind within the village (any village and not just their own), but failed to recognise it as equally or indeed at all wrong when committed against a member of another district. And they were prepared to give reasons for their view. They pointed out that theft was inconsistent with the relations between members of a clan or village, that it would involve a disturbance of the life of the group, and that if it became general the group would disintegrate; whereas the relations between members of different districts were less intimate and involved an element of suspicion, if not even of hostility.

Whether we regard this attitude as due to a lack of imaginative capacity to put themselves in the owner's place, or, as seems to me more likely, to the lack of any form of co-operation between the different districts and, therefore, to the absence of any need to comply with the conditions necessary to make co-operation effective, and the absence of social sentiments which result from co-operation, what is significant from our present point of view is that they do not regard the moral rules, which they recognise as right in the relations between the members of the smaller group, as applicable in their relations to all the inhabitants of the island, though they undoubtedly regard them all as human like themselves, and though their relations with them are on the whole not unfriendly. This clearly shows that they do not recognise their moral rules as rules of prima facie obligation, whose claims are overborne by more urgent claims. In other words, it is not the sort of act or forbearance as such or in its own right which they regard as obligatory. For the same sort of act is regarded as right in one context and wrong in another.

Hogbin produces evidence that the judgements of these natives are genuinely moral judgements, that they are not deterred from wrongdoing merely by fear of punishment, that they have their ideals of conduct and that these relate to the inner as well as the outer aspect of conduct, that they are no strangers to moral conflict and successful resistance to temptation, and that they not only experience but value the satisfaction of having a good conscience. One native, e.g., is reported as saying that when an “opportunity to do evil presented itself and he declined to take it, ‘his inside felt good’”;6 and he seemed to regard that as a sufficient reward for following the path of duty.

The evidence regarding the attitude of primitives towards the rules which they recognise as binding between the members of the in-group, the group which is always regarded as fully human and moral beings, seems to confirm the conclusion that such rules are not regarded as self-authenticating or carrying the grounds of their rightness within themselves. It is true, as I have said, that there is a way of interpreting moral rules according to which we may say that, as between the members of the in-group, many primitives recognise most, and some recognise all, the rules which we recognise as binding. And this way of interpreting moral rules seems to me to have misled many people into minimising the differences between primitive moral rules and ours. But according to this interpretation moral rules are mere tautologies, not significant or synthetic statements. Moreover, the terms used in them have different meanings in different communities; and so the rules do not really tell us what conduct is considered right in different communities. If, e.g., by a lie is meant not all deliberate telling of what is known to be untrue, but only such forms of untruth as are disapproved by the community to which the individual making the judgement belongs, if by homicide or adultery is meant only such killing of another human being or such intercourse between a married man and a woman not his wife as is disapproved by the particular society, then the propositions, that lies, homicide and adultery are wrong, are universal and self-evident among all peoples. But in that case the terms ‘lie’, ‘homicide’, and ‘adultery’ have different meanings in different communities, and in each community the propositions are merely verbal, indicating how the terms are being used. If, on the other hand, by a lie is meant any deliberate mis-statement of fact with intent to deceive, by homicide the killing of any man, by adultery any intercourse between a married man and a woman not his wife, the proposition that all lies, homicide and adultery are wrong, does not accurately describe the beliefs of many, or perhaps of most, primitive peoples.

Similar considerations apply to all other rules. Among all primitives some, and among most primitives all, moral rules admit of exceptions which are not only permissible but obligatory. It is true that most of these exceptions are not recognised as exceptions, because, as we illustrated in the case of the Australians7 and the Trobrianders,8 they are not thought of in relation to the rules in question but as the requirements of other institutions. Instead, therefore, of being thought of as e.g. lies, homicide or adultery, they are thought of as e.g. loyalty to a friend, regard for the welfare of the group, taking vengeance on an enemy, consecrating a new village or generosity to a friend.

It is, of course, true that we too admit exceptions to many of our moral rules, and that this is admitted by, and quite consistent with the contention of those who hold that the prima facie rightness of some of them is apprehended as self-evident. The point I want to make is rather that the interpretations which different peoples put on the terms used, even in the rules which they seem to have in common, are so different, and the exceptions which they admit to the rules vary so widely that it is difficult to say that the rules which they recognise are the same rules in any precise sense. These differences tend to be concealed by the use of general terms, and especially general terms like lies and murder, which not only describe facts but also convey moral judgements. For example, to say that a statement is a mis-statement of fact is merely to describe it, but to say that it is a lie is not merely to describe it but to pass a moral judgement on it. Now the factual part of the meaning of a term may vary, while the moral judgement involved in its use may remain the same, and vice versa. It is failure to recognise this which is responsible for the confusion to which I am calling attention, the confusion, that is, between the sense in which moral rules are universal but only verbal tautologies and the sense in which they are significant or synthetic, but generalisations true in the main and yet having many exceptions.

Even so acute an observer as Boas seems to be guilty of this confusion; and as a result he regards the moral rules of primitives as more akin to our own than the evidence warrants. “There is no evolution of moral ideas”, he writes. “All the vices that we know, lying, theft, murder, rape, are discountenanced in the life of equals in the closed society. There is progress in ethical conduct, based on the recognition of larger groups who participate in the rights enjoyed by the members of the closed society.”9 By the closed society Boas means what I have called the in-group; and his contention seems to be that the only difference in moral ideas between the most primitive and the most advanced peoples is the greater extent of the group to whom moral considerations apply. In the relations between the members of the in-group, those who are regarded as moral beings, all peoples recognise the same moral rules and the same virtues and vices. Now if this only meant that the fundamental nature of the moral consciousness is the same everywhere, I should entirely agree. But in saying that the same moral rules, the same virtues and vices, are recognised by the primitives and ourselves, Boas seems to have been misled by the ambiguity of such words as lying, murder, theft and so on—terms which not only describe forms of conduct but convey moral judgements.

It is true that among all peoples there are forms of conduct regarded as right and others regarded as wrong; and these relate, among other things, to life, property, sex relations and intercommunication between individuals. Now if by murder, theft, adultery and lying, we mean the forms of conduct disapproved by particular peoples in relation to life, property, sex relations and intercommunication, we can say that they all discountenance them. But this does not mean that they all discountenance the same acts or sorts of acts. For the forms of conduct in relation to life, property, etc. of which different peoples disapprove and which they, therefore, describe by the terms, murder, theft, etc., vary from people to people. Take, e.g., murder. When, as Boas himself tells us,10 a Chuckchee son kills his father before he loses his vigour and vitality, because he believes that a man will continue for ever in the state in which he is at the time of his death, the Chuckchee do not regard this as murder, but as a filial act, the discharge of a difficult and painful duty. Or when a Bantu parent kills one of his twin offspring to prevent, as he believes, disaster to himself and his people, his act is not regarded as murder. Indeed Boas himself provides the corrective to the passage which I have quoted above, when he tells us in an earlier work:11 “The person who slays an enemy in revenge for wrongs done, a youth who kills his father before he gets decrepit in order to enable him to continue a vigorous life in the world to come, a father who kills his child as a sacrifice for the welfare of his people, act from such entirely different motives, that psychologically a comparison of their activities does not seem permissible. It would seem much more proper to compare the murder of an enemy in revenge with destruction of his property for the same purpose, or to compare the sacrifice of a child on behalf of the tribe with any other action performed on account of strong altruistic motives, than to base our comparison on the common concept of murder.”

Whether we call such acts murder because we consider them wrong, or refuse to call them murder because the people concerned consider them right, it is clear that they are forms of killing which they approve and we disapprove. Similarly, both we and the Australian Aborigines may be said to condemn adultery, but, as we have seen,12 there are acts such as wife-lending which we consider adultery and they do not, but regard as obligatory. It is, therefore, misleading to say that the primitives and ourselves disapprove of the same vices; and nothing but confusion can result from a failure to distinguish the forms of killing and intercourse of which they disapprove from those of which we disapprove, despite the use of the common terms murder and adultery, in the sense of wrongful killing and intercourse, applied to both.

Now these considerations have a direct bearing on the intuitionist view that the rightness of certain moral rules is self-evident and apprehended by an act of direct insight. Ross seems to recognise that this is so; and while he does not, so far as I know, discuss the evidence in detail, he assumes that the differences between the moral rules of different peoples are less than the facts which we have been considering suggest. He does not go as far as Boas and hold that all men recognise the same moral rules; nor is it necessary for him to do so in order to reconcile the facts with his theory. But in several places he implies, and at least once he explicitly states, that the differences between the rules accepted by different peoples apply only to the more specific “rules such as those which prescribe monogamous marriage and forbid unchastity”, and not to “the very general rules”13 such as truth-telling and promise-keeping, which he claims to be self-evident. The former he regards as empirical generalisations, “the crystallised product of the experience and reflection of many generations”;14 the latter he regards as necessary truths, apprehended by direct insight.15 But if we have correctly interpreted the anthropological evidence, it would seem to show that the differences apply to all rules, the more general as well as the more specific. For if a proposition is to be apprehended as self-evident, the concepts involved in it must be conceived and stated in a precise and unambiguous way as is the case with mathematical propositions to which the intuitionists often compare moral rules. And in view of the variations of the meanings of the terms used in moral rules from people to people, it is impossible to state even the more general rules with the necessary precision and yet claim universal recognition for them.

It has often been pointed out that even among ourselves, where acts and sorts of acts are interpreted against the background of a common way of life, few, if any, of the terms which we use in moral rules have the precision and freedom from ambiguity which would entitle us to claim self-evidence for the propositions in which they occur. When, e.g., we say that the obligatoriness of promise-keeping is self-evident, what precisely are we to understand by a promise? Is a ‘promise’ extracted under physical torture or threat of blackmail a promise in the required sense? And even if we say that the promise must be freely given, is a promise, however freely given, to assist in a burglary, or to show favouritism to a candidate for an appointment, or to give financial assistance to another man, when the promise was made as the result of misrepresentation by him as to his means, a promise within the meaning of the rule? Again, in considering whether an act of promise-making comes within the scope of the rule, is any account to be taken, and if so how much, of unforeseen changes in the situation contemplated when the promise was made? When we ask what promises ought to be made, and if made ought to be kept, it is by no means easy to get an answer that will command universal assent even among ourselves, let alone among all peoples. Similar considerations apply to other rules, most of which are even less precise. We should tell the truth, but how much, to whom and on what occasions? We should not steal, but what taking of property is theft? We should make a return for services rendered, but what and how much should be given in return for a particular service, and indeed what is a ‘service’ in the required sense? If we reply: ‘What is considered right by our own society’ the rule becomes a tautology, and in addition it is not precise; if we give any other answer the rule fails to command anything like universal assent.16

Thus while the normal application of many of the terms used in moral rules is clear enough, there is a considerable margin where this is not so, and we have to consider the context of acts in order to discover how they should be characterised. Intuitionists themselves seem to me at times to admit this, implicitly if not explicitly. For example, in answering criticisms of the view that the obligatoriness of promise-keeping is self-evident, Ross17 insists, and it seems to me rightly insists, on the need to take account of the ‘spirit’ of a promise; and he calls attention to the fact that there are “implicit conditions and qualifications”, “underlying assumptions” and “tacit conditions” subject to which the promise has to be interpreted. Similar considerations apply to other prima facie obligations, most obviously to reparation for injuries and return for services rendered. The implication of this seems to me to be that we must think of such obligations subject to the conditions of a way of life, as interpreted in the spirit of that way of life by those who actually live it. And if this is so, we are far from considering promises or other sorts of acts which are regarded as obligatory in isolation, apart from any context or conditions.

Thus, even among one people we find a degree of ambiguity and lack of precision in the terms used in moral rules, which seems to be incompatible with the claim that their rightness is self-evident. And when we compare the meaning of such terms as used by different peoples, and especially by primitives and ourselves, the ambiguity is much greater and the precision much less. Until these rules are made precise, it is difficult to see how their claim to self-evidence can be granted; and in the degree to which they are made even relatively precise, it is impossible to claim that they are universally recognised as right, let alone as self-evidently right, even in their application to the in-group among primitives or to all men among ourselves.

It is, of course, true that any tolerable form of social life requires that there should be rules governing the relations between persons in regard to such matters as intercommunication, return for services rendered, sex relations, respect for life and property, etc., and that they should be generally obeyed. And the rules contained in lists of prima facie obligations are in a general way such obvious conditions of individual and social well-being that most of them are included in the moral codes of most peoples. Their value, and indeed their fundamental importance, in this sense is not in question here. What is in question is the ground of their rightness—whether they contain it within themselves and are therefore self-evident, or whether they derive it from the form of life whose conditions they are. My contention is that the ambiguity of the terms used, the great variety of the ways in which the rules are construed by different peoples and of the exceptions which they admit to the rules which they recognise, show that the rules lack the definiteness and precision necessary to self-evident intuitions; and that the rightness of the rules themselves, in those cases in which they are recognised as right, the ways in which they are construed, the exceptions to them which are regarded as justified, and the relative order of urgency which is assigned to them, are explicable by, and derive their authority from, the way of life of the people concerned. In other words, they have not a rightness which is independent of the goodness of the state of affairs whose conditions they are.

This conclusion is further supported by the fact that some primitive people do not recognise some of the so-called self-evident rules as in any way binding, unless conformity to them is regarded as being for the good either of particular individuals or of the group as a whole. It is not just that they regard exceptions to them as right when their requirements clash with those of other rules: they recognise no obligation at all to obey them as such, i.e. unless obedience to them promotes a state of affairs which they independently recognise as good. Take, e.g., the rule of truth-telling. Junod, as we have seen, reports that the Tonga do not regard a breach of this rule as wrong unless it harms someone. This is not because they do not understand what lying is. For they condemn and punish both perjury in court and defamation of character. They do so, however, not because they regard lying as such as wrong but because such lies harm others, either in their reputation or through interfering with the administration of justice.

The Tonga are by no means unique in this respect. According to Smith and Dale, a similar state of affairs exists among the Ba-Ila, another Bantu tribe. This highly intelligent people have a detailed code of morals, and they teach it to their youths during initiation. “Their ideas about family duties and privileges”, we are told,18 “scarcely differ in any way from our own”; and they have a system of justice and judicial administration not unlike that of the Tonga. Yet they neither regard truth as a virtue nor lying as a vice. “Among themselves”, Smith and Dale write,19 “they lie in the most barefaced and strenuous manner . . . without the least shame. They lie often when it is to their advantage to tell the truth.” And in case we should conclude from this that they are amoral, our authorities state explicitly that this is not so. “Immoral they may be,” they write, “they are not unmoral”.20 “The principles of conduct are there and they are well known”, they continue, but they do not include truth-telling. Other rules they sometimes break; the rule to tell the truth they do not recognise. It is not among the virtues taught to their youths during initiation; and in their folk-lore, which, we are told, faithfully reflects the character of the people, success comes “by lying and cheating”, “by downright chicane”, “by treachery”, “by promises made and not fulfilled”, and so on.21

Smith and Dale offer a possible explanation of this attitude to lying. “Much lying”, they say,22 “may be attributed to their sense of politeness; they do not wish to hurt one's feelings.” But they recognise that such an explanation can at best be only partial. Nor can we attribute this attitude to incapacity to understand what lying is or to not having attended to it. Like the Tonga, they condemn lying by a witness in court; and Smith and Dale called their attention to what they regarded as the nature and evils of lying, but the Ba-Ila still held that they saw nothing wrong with lying as such.

Contrast with this the attitude of the Manus people of the Admiralty Islands, a people with a much less keen sense of social justice and, in other ways, less advanced morally than the Bantu. The Manus moral code, as we have seen, includes strict rules of truth-telling and promise-keeping; and they regard these rules as having a powerful supernatural sanction. For them one of the two supreme virtues is a high level of commercial morality, and this requires honesty and mutual trust. In case we should imagine that the supernatural sanction will explain the difference between the Manus and the Bantu attitude to truth-telling, we may note that the Murray Islanders of the Torres Straits, a still more primitive people whose morality is entirely secular, have an equally strict rule of truth-telling.23

What, then, is the explanation of this difference of attitude to truth-telling? It might be suggested that differences in the economic conditions and the natural environments of the peoples concerned provide an explanation. The Tonga and the Ba-Ila are agriculturists who till the soil and rear cattle; and the land which they occupy is relatively infertile and sparsely populated. Their kraals or small villages are often separated by considerable distances. Each is relatively self-sufficient. There is little trade with or dependence on the members of other kraals. Mutual trust and confidence between the members of different kraals is, therefore, relatively unimportant for them. In the Manus way of life, on the other hand, commercial transactions—some of them based on a credit system—play a very important part. Such a form of life could not succeed without a considerable measure of mutual confidence and the virtues which make confidence possible. Therefore, truth-telling and promise-keeping rank high in the Manus code. But I do not think that this can be accepted as a satisfactory explanation. For one thing it is difficult to say whether the Manus attitude to truth-telling and promise-keeping is the cause or the effect of their particular form of economic life; and no doubt the one tends to support and strengthen the other. In any case, this explanation could not account for the attitude of the Murray Islanders, among whom commerce, though not unknown, plays a relatively unimportant part. In addition, the Bantu kraals are only relatively independent. Members of one kraal have to get their wives from another; and the members of different neighbouring kraals all owe allegiance to the same headman or chief, who administers justice between them in cases of dispute. Moreover, the Bantu attitude to lying applies not merely to relations between members of different kraals but also between members of one kraal.

The real explanation of the difference of attitude to truth-telling is, I think, to be found in the important part which the belief in black magic or sorcery plays in the life of the Bantu and its relative insignificance in the lives of the Manus and the Murray Islanders. If a man believes, as the Bantu does, that any of his neighbours may be, and that some of them in fact are, endowed with supernatural powers which may be used to do him or his neighbours harm; and if he further believes that giving them information provides them with the means of using these powers, we have a state of affairs in which not mutual trust and confidence but mutual suspicion and fear are likely to flourish. In such circumstances, truth-telling is not likely to be regarded as a virtue except in situations like giving evidence in court, where truth-telling is likely to result in good to others and lying in harm to them.

When we think of truth-telling and promise-keeping as right, we think of them in relation to conditions in which the possession of accurate knowledge and the performance of what is promised are on the whole good, independently of any rightness of telling the truth and keeping promises as such. Under such conditions it is obvious that truth-telling and promise-keeping have a value, whether or not they have, in addition to that value, an independent rightness. But in a way of life riddled with the belief in black magic, the free interchange of information is likely to do more harm than good. This is the state of affairs among both the Tonga and the Ba-Ila, whereas among the Manus and the Murray Islanders black magic plays a very insignificant role. What we find, then, is that wherever the possession or interchange of information is not regarded as good, truth-telling is not regarded as obligatory; and this suggests that its rightness, where it is regarded as obligatory, does not belong to it as such, but is dependent on the goodness of the state of affairs of which it is a condition.

In further confirmation of this interpretation of the facts, and in case it might be thought that the difference of attitude to truth-telling is a geographical difference between the Bantu and the natives of the Pacific islands, we may note that there are many Pacific islands in which the belief in black magic is rife, and that wherever this is the case the rule of truth-telling is not to be found. If, e.g., we consider the moral code of the natives of Dobu—the home of sorcery par excellence—we find no rule of truth-telling.24 Even perjury in court (since courts have been established by the white man) is not regarded by the natives as wrong. In the atmosphere of fear and suspicion and hatred which exists in Dobu, there are even fewer than among the Bantu of the conditions which would make the free interchange of information contribute to individual and social well-being. The traveller from one part of the island to another has to go round the villages on the way in case he should use magic against person or property, or gather the information which would enable him to do so. Children are used to spy on their parents; and the surviving partner to a marriage is liable to be accused of having killed his or her spouse by magic. In such circumstances it is natural that truth-telling should not be regarded as a virtue nor lying as a vice. Nevertheless, even in Dobu, a certain minimum of co-operation and of the mutual trust required to make it possible is essential to any form of social life. But the amount of it is very strictly limited, and it is significant that it tends to be confined to circumstances in which, and to occur between persons between whom, the powers of black magic are believed to be either non-existent or less effective.

Now all this suggests that the rule of truth-telling is regarded as obligatory only in the societies and within the limits in which it is believed, in the light of the beliefs which are entertained about human beings and their natural and supernatural environment, that a free interchange of information is a condition of or contributes towards individual and social well-being. I believe the differing attitudes of different peoples to other rules could be explained on the same principle. I find no evidence of any primitive people who recognise any rule as right unless general compliance with it in the spirit of their way of life seems to them, in the light of their view of the human situation, likely to result in a state of affairs or to contribute to a form of life which they regard as on the whole good.

This conclusion is also supported by the testimony of field workers about the attitude which natives take to the rules which they do recognise as right. Junod, as we have seen, tells us again and again that the Tonga regard their rules of conduct as right not, as he puts it, because breaches of them are regarded as immoral,25 but because such breaches would be disruptive of their way of life, or detrimental to the good either of the group as a whole or of some of its members. We get similar testimony from Haddon about the Murray Islanders, and from many others.

Apart from the evidence which I submitted to show that the primitive mind is not markedly inferior in powers to the civilised, the examples of their moral rules which I have quoted should by now have made it clear that, if primitives do not recognise particular rules as binding or do not regard those which they do recognise as self-evident, it is neither because they lack the capacity to grasp the concepts used in them nor because they have failed to attend to them. It may well be that there are some primitive people who do not formulate such rules in abstract terms, though there are others who do. We find such formulation of them, e.g., in the instruction given to the youth of some tribes during initiation and in the discussions which take place in the courts of justice of the Bantu. But even those who do not explicitly formulate their rules in abstract terms should still be able to appreciate their rightness and their self-evidence, if they are in fact self-evident. For, as the intuitionists themselves admit,26 the way in which people come to recognise the self-evidence of such propositions as are really self-evident is by seeing their truth in particular cases, and, in so doing, seeing the universal in the particular. And, of course, all peoples have seen plenty of concrete examples of the sorts of act whose prima facie rightness is claimed by intuitionists to be self-evident. If, then, as the evidence suggests, primitives do not recognise their self-evidence, and if the moral judgements of primitives are genuine pronouncements of the moral consciousness, it is difficult to accept the view that the rules in question are really self-evident in the required sense, i.e. evident to all who have the capacity to grasp them and have had the opportunity of attending to them.

Moreover, many primitive people regard as obviously right rules which we do not recognise, and which even seem to us abhorrent, e.g. rules about wife-lending, scalp-hunting, infanticide of twins or the punishment of one man for the sins of another. I believe a detailed examination of these rules, in the context of the ways of life of the peoples who think them right, would show that they, too, are regarded as obligatory, because, in view of the beliefs these people entertain, the rules seem to them to be conditions of a way of life which they regard as good.

But, even if the moral rules recognised by primitives as right thus appear to be relative to the ways of life of the peoples who recognise them, there is an important sense in which they are not relative. They are not relative to the individual moral agent or to his inclinations or desires or likings. As far as the members of the in-group to whom moral considerations are regarded as applicable are concerned, moral rules are universal in the sense that if it is right for one individual to act in a particular way in certain circumstances, it would be right for anyone else similarly situated to behave in the same way. In that sense primitive people claim objective validity for their moral rules. They do not admit that there is anything arbitrary or capricious or strictly individual about them.

Let me try to sum up the results of this lengthy survey of the attitude of primitive peoples to their moral rules: (1) Few, if any, primitive people regard their moral rules as universal in the sense that they apply to all men. Some of them do not regard all their moral rules as applicable to all even of those to whom some of them are applicable. (2) While many moral rules abstractly stated may appear to be common to the codes of many peoples, the interpretations which different peoples put on the terms used in them, and the exceptions to them which they regard as right, are so different that the rules in question, in any sense in which they can be said to be common to most peoples, lack the precision and freedom from ambiguity necessary to any rules which are to be apprehended as self-evident. (3) Many people fail to recognise as right at all some of the rules regarded by contemporary intuitionists as prima facie obligations. (4) The rules which primitives do recognise as binding they do not regard as self-authenticating or unmediated, but as deriving their authority from the form of life whose conditions they are. And (5) some primitives regard as obviously right some rules which we do not recognise as right at all and which are indeed inconsistent with some of those which we do regard as binding obligations.

If these conclusions are well founded, they seem to show that most of the moral rules, for which self-evidence has been claimed, are not really self-evident in the sense that they are recognised as such by all who understand them and have attended to them. And if we accept the view, as I think we must, that a satisfactory ethical theory must be consistent with the moral judgements of all men everywhere, this means that intuitionism cannot in any of its forms be regarded as a satisfactory ethical theory. This, however, does not mean that moral rules are less important or obedience to them less obligatory than if they were self-evident intuitions. It means only that the grounds of their obligatoriness have to be sought elsewhere than in themselves. It is to be found in their being the conditions of a way of life which is an attempt by the people concerned to embody what they take to be the good for man. It is their relation to such embodied or operative ideals which justifies both the rules and the exceptions to them which are regarded as right, and which also determines the relative urgency of different rules when their requirements clash.

This is not to say that, when even the more thoughtful, let alone the ordinary unreflective, member of a primitive community regards an act or a rule as right, he always consciously thinks of it as a condition of a way of life which he recognises as on the whole good. To the suggestion that the interpretation of the facts which I have given need involve such conscious reference to the good of the society as a whole, Ross has, in another connection, given what he himself calls “a perfectly proper answer”. I cannot put this answer more clearly than Ross has done; but in quoting it I substitute the present theory for Ideal Utilitarianism which he was discussing when he used it. “Certain types of acts have in practice been found [to fit into a way of life, or to be conditions of its smooth working], and have in consequence been judged to be right; and so, for plain men, the character of rightness has come to seem to belong to such acts directly, in virtue of their being, e.g. fulfilments of promises, and the middle term which established their rightness has come to be forgotten. Media axiomata such as ‘men should keep their promises’ have come to be accepted as if they were self-evidently true, and people habitually judge acts to be right on the strength of the media axiomata, forgetting the method by which the media axiomata have themselves been established.”27 This is, I think, as Ross says, “a fair answer”. But I should add that in a small, closely knit primitive community it is much easier to see that the observance of a rule is necessary for the good of the group as a whole than it is in a large-scale modern society; though even in primitive societies the rule is more often thought of as the requirement of a particular institution than in reference to the way of life of the people as a whole.

There are, however, two rules which are often included in lists of prima facie obligations which seem to be in a different position from the others. They are the rule to produce as much good as we can and the rule of justice or equity. These are not so much particular rules among others as general statements of what every way of life is trying to do, and, for that reason, whatever other rules are or are not recognised by particular peoples, these rules are in some sense recognised by all. There could not be any way of life without them. But these rules, or, as I should prefer to call them, principles, remain theoretically vague and practically valueless till they are embodied in a way of life whose structure determines the particular rights and duties in which their requirements find expression; and within that structure the one cannot be understood except in relation to the other.

The rule to produce as much good as possible is just a rule to realise the moral ideal regarded as the pursuit of ends or the realisation of interests. And it remains vague and indefinite till the ends and interests are systematised into a structural pattern, which embodies the conception of the people concerned of what is required to satisfy their nature as a whole. When this is done, the principle of justice or equity is found to be the structural principle of the pattern—the principle which determines the interrelation of ends and interests, and especially those of different persons, to one another. This principle takes a large variety of forms, but in one form or another we find it expressed, sometimes more and sometimes less adequately, in the way of life of every people. It is a kind of moral symmetry which underlies every system of rights and duties. Its ultimate basis is the recognition, however dimly, by individuals of other people as moral persons and, therefore, entitled, in virtue of their common humanity, to a certain respect and consideration of their interests. But it, too, remains vague and general until it is embodied in a system of interrelated institutions constituting a way of life in which the ends and interests of individuals are dovetailed into a relatively coherent pattern.

With what is involved in the rule to produce as much good as we can I shall deal in my next lecture, and I shall try to explain the nature of justice in the last lecture. Meantime I want to state in general terms what I take to be the relations of the different prima facie rules to one another in the moral ideal, and in the ways of life which are attempts to embody it. The rule to produce as much good as possible expresses the requirement to realise the moral ideal from the point of view of ends or goods. The principle of justice is the principle of interrelation between ends, and between the individuals whose they are, which is necessary to enable the individuals to realise their ends in such a way that they will be on the whole satisfied; while the more specific and detailed rules are the conditions of the co-operation and mutual trust and respect between persons on which the realisation of their ends, in accordance with the principle of justice, depends; and the concrete systems of rights and duties of particular peoples, and the institutions in which they are embodied, are the flesh and blood of the body of which the more abstract rules, and especially the principle of justice, is the skeleton.

There are other questions about moral rules which would have to be answered, and objections which would have to be considered, before the view of their nature and authority for which I am contending could be regarded as established. The answers to most of them seem to me to be contained in what I have said already. But even at the risk of some repetition, I want to consider one or two of them further. Others will be dealt with in later lectures.

If moral rules are not binding in their own right but have only a borrowed authority, why, it may be asked, do they occupy so important a place in the life and thought of every people, primitive and civilised? Why do we, as we often seem to do, recognise the rightness of moral rules, and react to breaches of them more directly and immediately and spontaneously than we recognise the goodness of a way of life, or react to what hinders or promotes it as such? Why do we feel compunction, as we undoubtedly do, when we feel obliged to break a moral rule, such as that of truth-telling or promise-keeping, even when we believe that it is our duty to do so, and that more good is likely to follow from the action in which we do so than from any alternative action open to us? Why do we feel compunction about breaking the rule of truth-telling and the Tonga do not, or about certain forms of adultery and the Australian Aborigines do not?

There is no doubt that, in the ordinary business of living, morality presents itself to most people most of the time, and perhaps to some people all the time, not so much as the promotion of a way of life or the production of good, but rather as obedience to rules which are accepted as right. I do not want to deny or even to minimise the importance of rules in the moral life. As I have repeatedly pointed out, any form of ordered social life requires that there should be rules and that they should be generally obeyed. These rules, as Ross reminds us in the passages which I have already quoted, embody the accumulated wisdom of our ancestors and their value is attested by each generation in its own experience. Some of them are such obvious conditions of individual and social welfare that even the least imaginative can recognise the fact. Accordingly, to break one of them is to run a grave moral risk. The moral agent is usually much more certain of the value of their general observance than he is of the problematic good which he expects to follow an exception to one of them.

Moreover, to break a rule which is generally recognised as right is to be guilty of discourtesy or disrespect towards the person or persons in relation to whom we do it, and who have a right to expect us to observe it; without the general observance of such rules people would not know what to expect of one another. In this, and in other ways, breaches of rules tend to undermine mutual confidence. For example, other people may see the breach and may not be aware of the good motive for it. And such breaches are in danger of producing in the individual who is guilty of them a habit of ignoring the rules concerned in other connections, and this may also affect his attitude to other rules—as we have seen in the recent great increase in disrespect for both laws and moral rules which was the result, in large part at least, of people living in war-time under conditions in which it was often regarded as a merit to break them, or at least to make exceptions to them. Again, the individual may quite rightly hesitate to trust his own moral insight as to the good which he expects to follow from the breach of a rule against the deeper insight and the greater wisdom of those whose experience is embodied in the rule. The fact that rules are generally recognised as right and embodied in a moral code is not what makes them right; but it may be a good ground for believing them to be right. It suggests that they have been tested by experience and found good.

For the greater part of the time, the moral thinking of most people is imitative or repetitive, not original, and there is therefore no need for them to go behind or beyond the accepted moral rules. They sufficiently indicate their duties; if they obey them they will have done their duty. And if there is any situation in which it is more unwise than another to question the authority, or to ask for the credentials of a moral rule, it is when an individual is faced with a practical decision in which it would be to his interest or convenience to make an exception to it. Rules act as a warning against making exceptions in favour of oneself.

And there is another reason why rules are much in evidence in our ordinary moral thinking, and occupy more than their fair share of our attention. They are often in conflict with our natural inclinations and desires. This is so not only in childhood, when they are presented to us mainly in negative form as restraints on what we want to do, but throughout life; and the highest moral conduct, action done from a sense of duty, always involves such conflict with our immediate desires. Now anything which conflicts with our desires attracts our attention and is apt to occupy the focus of consciousness. Just as the citizen does not notice the laws of the land except when they prevent him from doing something which he would like to do, or compel him to do something which he does not very much want to do, so the moral agent is naturally more conscious of the rules which conflict with his desires than he is of the purposes for which the rules are imposed or of the ends which their observance promotes. Moral rules are, therefore, apt to occupy the foreground of consciousness, while the form of life whose conditions they are tends to occupy the background. But in our more reflective moments when we engage in original moral thinking, the form of life comes into the foreground. This is specially apt to happen when an accepted rule is challenged and requires justification or modification. When we say, e.g., that a particular rule is not right, or that the institution or form of life of which it forms part should be changed, we tend to give as our reason that those who live under it have no chance of living a reasonable human life or of developing their powers or personalities. This is the sort of argument which we find used, and it seems to me rightly used, in recent discussions on such matters as divorce laws, education, the conditions in slum areas revealed by the air raids, the conditions in certain industries, and so on.28 In the last resort the appeal is to a way of life as the ultimate criterion by which rules are to be tested.

It is also worth noting that in the smaller and more intimate societies, like the family, the team, the college, in which relations are personal and friendly, and in which the individual feels at home and at ease in co-operating with others to carry out the common plan—in such groups, while there are rules and patterns, they are not much in evidence, and they are not usually felt as irksome restraints. It is in the larger groupings, where relations are more impersonal and less intimate, that rules are more in evidence; and even there they are, at most, the framework and not the whole of the moral life.29 Indeed as I said earlier, if many or even most men in most societies were not better than the rules required of them, if they were not at times prepared to do more and exact less than the strict letter of the rules lays down, it is doubtful if any form of social life would really prosper.

At present, however, we are concerned not with the limitations of a life based on rules alone, but to reconcile the prominent place which rules occupy in the moral life with the theory of the nature and authority of moral rules which the anthropological evidence, as I have interpreted it, suggests. And the considerations which we have been discussing seem sufficient to explain the importance which we attach to moral rules, the necessity for general compliance with their requirements, and the compunction which we feel when we consider ourselves obliged to break one of them, even if they have not an independent rightness in addition to the goodness of the form of life whose conditions they are. Breaches of them require justification; to comply with them does not. And this is all that we seem to get on the intuitionist theory; for intuitionists admit that particular moral rules ought to be broken, both when the requirements of different rules clash and when much greater good would result from breaking them than from complying with them. On any theory the practical problem of deciding whether in particular circumstances it is right to break a moral rule presents considerable difficulties. The only advantage of the present theory is that it claims to bring to light the principle which guides the reflective moral agent in his attempts to solve this problem, however difficult it may be to apply the principle in detail; whereas the intuitionist theory leaves the moral agent without any principle to guide his efforts.

But why, it may be asked, do we feel compunction when we break certain moral rules and the primitive does not? The answer, I think, varies from people to people and from rule to rule. We may distinguish three sets of circumstances in which primitives break, without compunction, rules which we recognise as binding: (1) when they do not recognise the rule in question as right at all; (2) when a rule which is recognised in the relations between members of the in-group is not regarded as applicable in their relations to outsiders; and (3) when a rule which is accepted as right in some relations between the members of the in-group is broken without compunction in others. The explanation of cases of the first kind is to be found, I think, in the principle which we discovered in our analysis of the attitude of certain tribes to the rule of truth-telling. In their circumstances, as they conceive them, general compliance with the rule in question would not promote their welfare; indeed, it would do more harm than good. Therefore, they do not recognise the rule as binding. The same principle, I think, supplies at least part of the explanation of cases of the second kind. The way in which the principle operates in these cases and the other considerations which enter will be discussed in my last lecture. Many factors seem to contribute to produce cases of the third kind. For example, the individuals towards whom the exceptions show disrespect may not be recognised as persons in the full sense, as seems to be true of women and children among the Australian Aborigines. But the main explanation seems to be that the exceptions are not thought of in relation to the rule but as the requirements of particular institutions and, therefore, not as exceptions. What the anthropological evidence suggests is that most or even all rules are first recognised as the requirements of particular institutions and in the relations between small groups, where the forms of conduct of which the sorts of act in question form parts are found to be good. Once a rule is thus recognised as right in one connection, there is a tendency for it to be carried over to other institutions and relations, till it comes to be regarded as right in all relations between those who constitute the group of those who are recognised as persons in the full sense. The same principle seems to apply among ourselves, as Russell has illustrated in detail in his treatment of professional ethics.30 The smaller and more intimate the group, the higher the ideal and the more exacting the standard demanded in the relations between its members; and progress in ideas of rightness or righteousness largely consists in extending these standards, and the rules which they imply, to the relations between individuals in larger groups.