Morality and religion are both complex subjects and the relationship between them, even among primitives, is neither single nor simple. Generalisations about it are therefore apt to be ambiguous and misleading, the results of abstraction and over-simplification. Indeed there is no other aspect of our subject about which there seems to be so much confusion and about which many of our authorities are so unhelpful and such unsafe guides; and, as I said in an earlier lecture, I believe this confusion to be largely responsible for the neglect of primitive morality by both social anthropologists and ethical theorists. This confusion is, no doubt, partly due to the complexity and difficulty of the subject1 and to differences of opinion as to the nature of morality and religion; but it is also partly due to the lack of clear definition of the sense in which the terms are being used, and especially to failure to distinguish carefully between the different aspects of morality and religion, and between the different questions which may be raised about their interrelations. This is specially true of morality, which even careful writers like Radcliffe-Brown and Malinowski, who are at pains to analyse and define most of their other terms, seem to use uncritically without distinguishing its different aspects or elements. As a result, they seem to use the term with different senses in different connections, meaning by it sometimes moral sentiments and motives, at other times moral rules and principles. This is not only apt to confuse their readers, who often find it difficult to discover in what sense the term is being used in a particular context, but, what is more important, it is apt, as I shall try to show, to confuse the writers themselves. They are apt to assume that, when they have shown that religion is related to morality in a particular way in one sense of the term ‘morality’, they have proved that it is so related in a different sense of the term ‘morality’.

The position is further complicated by the fact that few anthropologists have devoted any separate consideration to the relationship between morality and religion, and that their views about it have to be gathered from incidental references to it which are scattered throughout their treatment of other subjects. Accordingly, each such statement has to be considered in the light of its context and of the evidence submitted in support of it; and, when so considered, many of them do not mean what they appear to say. We may find, for example, that for the purpose of a particular writer in a certain connection one aspect of the complex interrelation of morality and religion may be more important, and that therefore it is the only one which is mentioned; but the other aspects may also be there, and though the writer does not mention them, he may not wish to deny their existence or their importance. But, however we interpret them, some of the statements about our subject made in different connections even by the same authority are not only ambiguous and misleading but actually inconsistent.

My procedure throughout these lectures has been based on the assumption that morality is autonomous in the sense that it is possible to discover the grounds of the goodness of moral ideals and of the rightness of moral rules by an analysis of the moral consciousness, without introducing the idea of the supernatural or the religious view of life, as the source of the authority of moral rules or the justification of moral ideals; and that it is therefore permissible to treat morality by itself without taking account of its relation to, or any support it gains from or gives to, religion. There are, however, several reasons which make it necessary to examine and defend this assumption.

(1) The belief is still widely held that the rules of conduct of primitive peoples are magico-religious taboos, not understood to be obligatory in their own right but accepted on authority and obeyed because of their magico-religious sanctions; that, in short, primitives obey their moral rules and customs, mainly if not merely, because they believe that failure to do so will bring misfortune or disaster on themselves or their people, either through the displeasure of personal supernatural agents or through the automatic working of impersonal supernatural forces. Now, as we have seen in part already in considering the beliefs of the representative primitive peoples whose ways of life I tried to describe, this view is not in accordance with the facts, and it is not shared by field workers among primitives; but it is still so common that it will be necessary to consider it further.2

(2) Many anthropologists who explicitly reject this view use language which suggests, and sometimes explicitly states, that morality is dependent on, or even a part of, religion, and that moral ideals and rules are prescribed and sanctioned by it and derive their authority from it. This seems to me the result of a confusion due to a loose use of terms and a failure to distinguish between different questions. The confusion seems to have come about in the following way. Whatever differences there may be among anthropologists as to the nature of religion, they are all agreed on two points: (a) that primitive peoples have a religion, that is, that they believe that they are surrounded by supersensible or supernatural powers; and (b) that the beliefs, the emotional attitudes and the ritual practices, which constitute their reaction to these powers, profoundly affect their characters, their attitude to their fellows and the cohesion of their society. It is from these psychological and social effects of religion that many anthropologists conclude that primitive morality is dependent on, or sanctioned by, or a part of religion. But the inference seems to me hasty, the result of a confusion between functional interdependence, or mutual influence and support, and one-sided or causal dependence of morality on religion, or of a confusion between religion producing moral effects and religion prescribing moral rules or ideals.

(3) There is yet another sense in which morality is sometimes held to be dependent on religion. It is contended that man, whether primitive or civilised, has not in himself the wisdom to recognise what is good and right nor the strength to do it; and that for both he is dependent on religion.

Now any views which assert, on whatever grounds, that religion is the source of moral authority, or that the relation of morality to religion is a relation of one-sided dependence, deny the autonomy which I claim for morality, and imply that we cannot treat the nature and foundation of morality satisfactorily without taking account of the religious view of life. And, owing to the confusion to which I have referred, such views can appeal for support to many statements by anthropologists of eminence. I shall therefore have to examine carefully the evidence on which such statements are based in order to discover whether it warrants the conclusions which have been drawn from it.

I am concerned with only some of the questions which may be asked about the relation between morality and religion; and it will be necessary to distinguish these questions from others with which they sometimes seem to me to be confused. All I want to establish is that primitive morality is independent of religion in the sense that the moral ideals of primitive peoples do not derive their goodness, or their moral rules their rightness, from a religious source; and that, even among primitives, fear of supernatural punishments or hope of supernatural rewards, here or hereafter, is not the only or even the main incentive to pursuing the one or obeying the other.

This conclusion is quite consistent with a very intimate connection in the way of mutual influence and support between morality and religion. I do not wish to belittle, much less to deny, the undoubtedly profound influence which religion exerts on the character and conduct of its devotees or the support which it has given to individual and social morality; but I want to suggest that the influence is mutual and not onesided and that it is not inconsistent with the autonomy which I claim for morality. Such mutual influence is what we should expect on the functional theory regarding primitive culture, for which I have already quoted so much evidence. Nor do I wish to deny that morality has theological implications or that we may legitimately argue from morality to religion. Even if we can discover our moral duties without any reference to religion, even if we can see their binding character by considering the nature of persons and the relations between them, it may still be that, when we consider the view of reality which the existence and nature of morality presuppose, we may find ourselves driven to a religious view of the world as the most reasonable interpretation of the facts. In that case, though morality would still be autonomous in the sense that our knowledge of moral rules and ideals would be independent of religion, the religious view of the world would be the ontological condition of morality. So that morality would be first in order of knowledge or learning, but religion would be first in order of reality. All I contend is that, whatever be the conditions of the possibility of morality and whatever inferences we may draw from it, in the order of knowledge it comes first; and that we are more certain of the value of moral goodness than we are of any inferences, religious or other, that we may draw from it. Nor do I suggest that the view which a man takes about the supernatural makes no difference to what he believes to be right or good. In order to understand a man's judgements of value, his views as to what ends he ought to pursue and what acts he ought to do, we must take account of his beliefs about his supernatural as well as his natural and social environment. But I suggest that a man's ideas about what is right and good may exercise as much influence on his conception of the supernatural as his ideas about the supernatural exercise on his moral conceptions. My contention is simply that, whatever its presuppositions and implications and whatever support it may receive from or give to religion, morality is autonomous, containing its authority, the grounds of its goodness and rightness, within itself. I am concerned with the relation between morality and religion only so far as is necessary to establish this conclusion. There are therefore many interesting and important and highly controversial questions about their relation which are not germane to my purpose and with which I do not propose to deal. And I want to make it plain that I am not attempting to answer these questions, nor prejudging what the answer to them should be.

In view of the complexity of the subject and the confusion regarding it which I seem to find in the utterances of even distinguished anthropologists, it seems desirable that we should try to disentangle the issues in order to expose and, if we can, to remove the confusion, before we try to draw the conclusions as to the relation between primitive morality and religion which the available evidence suggests. This seems to me all the more necessary because, if I am not mistaken, such confusion is not confined to anthropologists dealing with primitive people. It seems to be no less evident in the writings of publicists and theologians, and perhaps even of philosophers, concerned with our contemporary situation; and it is today a source of bewilderment to many men and women of goodwill in our midst. For one of the most significant and perplexing features of our time is a growing divorce between morality and religion, combined with frequent statements that morality needs the support of religion if it is to survive and flourish.

As evidence of this divorce I need mention only two facts. On the one hand, there is the decay of religious beliefs and practices and the increasing secularisation of morality which is today admitted by both those who deplore the fact and those who rejoice at the prospect. Nor is the decay confined to the careless and irresponsible. Many sincere and thoughtful people in all walks of life, people with a high standard of personal conduct and a keen sense of social justice, people who devote themselves in the spirit of unselfish service to the well-being of their fellows, find themselves unable to accept any creed and have given up the separate practice of religion. On the other hand, a considerable and apparently growing body of theologians, in revolt against the rationalistic idealism and the liberal theology of the early part of the century, preach a form of theology which not only rejects any support from the moral consciousness, but seems to be inconsistent with its main assumptions, and to regard any suggestion that upright and honest living has any religious value as evidence of an irreligious spirit, if not even of being “in the bondage of sin”.3 I have no time to develop this point; nor would it be relevant to my subject to do so; but I am profoundly convinced that the absence of mutual understanding and support between the saint and the moral and social reformer is a misfortune for both. Unless the vision of the saint is transformed into moral energy to sustain and strengthen the drive, not only to change the individual, but to transform society, not elsewhere and hereafter but here and now, it will be apt to grow dim and religion to become a private luxury or an emotional thrill. And unless the ethical and reforming spirit is reinforced and strengthened and sustained by the vision of the saint, and the life of moral goodness put in a cosmic setting and regarded as the service of God, the moral agent will be apt to lose heart in the struggle and perhaps also to lose his direction. The ethical religions at their best succeeded in combining the inner peace which religion gives with the moral dynamic of the reformer in such a way that each supported the other: they called on men to do justly as well as to walk humbly; while the moral consciousness has always claimed the right to criticise the teaching of religion in the light of its own insight. Today it seems to me they are in danger of parting company and ceasing to understand or support one another; organised religion being content to look backwards rather than forwards, to sit at the feet of ancient teachers rather than stand on their shoulders; while many in whom the moral spirit is strong find neither spiritual refreshment nor moral stimulus in the services of the church and are tending to give up the separate practice of religion. When side by side with this growing rift between morality and religion in both theory and practice he finds grave concern expressed in many quarters, that the decay of religious faith and practice is in danger of undermining the moral fibre of our people and the very foundations of our way of life, is it surprising that the ordinary man should be perplexed and bewildered?4

Whether or not a consideration of the relation between morality and religion among primitives will help to throw light on our contemporary situation, the present widespread bewilderment and uncertainty about the question, as well as the confusion which I seem to find in the writings of anthropologists, makes it desirable to try to analyse and define our terms, to distinguish the different aspects of or elements in morality and religion and to disentangle the different questions which may be raised about their interrelation, so that we may discover which of them are relevant to our enquiry. To this preparatory work of clarifying the issues and clearing the ground I propose to devote the remainder of the present lecture. In the next I shall consider the answers, which the facts seem to warrant, to those questions about the relation of primitive morality and religion which specially concern our subject.

Following the general practice of anthropologists, and in order to avoid difficult and controversial issues, such as the distinction between magic and religion, I am going to use the term ‘religion’ in a very wide sense, to include all experiences and beliefs, practices and emotional attitudes in relation to the supernatural, i.e. to cover what might be more accurately described as magico-religious beliefs, attitudes and practices. Religion so defined includes (1) an element of belief—beliefs regarding the existence and nature of the supernatural, whether it be thought of in personal or impersonal terms; (2) an emotional element, the sense of the sacred or the uncanny and the emotional response which it evokes, what some anthropologists call the religious thrill; and (3) practices in which the beliefs and emotions find expressions, actions and abstentions, rites and ceremonies, dances and sacrifices and prayers. Closely connected with these are the legends and myths, the sacred traditions, which contain what we might call the theology and cosmology of the primitives. These tend to confirm by precedent, i.e. by reference to events which are believed to have happened in the past, the present beliefs and practices. They provide the reasons why certain things should be believed and done. For so far as ritual practices at any rate are concerned, the primitive does not usually draw a very clear distinction between reasons which explain and reasons which justify. As Robertson Smith5 puts it, “the precedent once established is authoritative and does not appear to require any proof”.

In morality also, as we have already seen, we may distinguish three aspects or elements: (1) the ends or ideals which are regarded as good or worthy of pursuit; (2) the moral rules or principles according to which actions are judged as right or wrong; (3) the motives or sentiments or attitudes of mind which find expression in the pursuit of ends and the obedience to rules. The term moral is commonly applied to all three, that is, to motives, ends and rules. But it is important to distinguish them because what is true of one need not be true of another.

The reference to good and bad or right and wrong conduct, however, does not enable us to distinguish between morality and religion, for these terms may be used either in a moral or a religious sense. It is necessary, therefore, to try to distinguish specifically moral from specifically religious conduct. I shall call specifically moral, conduct in relation to one's fellow-men, individually and collectively. It is true that a man may have duties to himself and to supernatural beings, and such duties, if they exist, are in a quite legitimate sense moral duties. But if we are to distinguish at all between morality and religion or between moral and religious duties, it must be on the basis of a distinction between, on the one hand, duties to himself and his fellows which can be described and understood without introducing any reference to the supernatural and, on the other, those duties which involve a direct reference to the supernatural. It is true that duties to one's fellow-men, duties of truth-telling and promise-keeping, acting justly and respecting life and property, may be, in whole or in part, regarded by a particular people from a religious point of view as duties to God who is believed to require from them that they treat their fellow-men in certain ways. But they can be understood without any such reference. In other words, they are moral duties whether or not they are also religious duties. And even if they are also religious duties, these moral duties are to be distinguished from strictly religious duties which cannot be understood without introducing the idea of the supernatural. The latter are of two kinds: (1) duties to personal supernatural agents in the way of ritual or worship, actions and abstentions, such as sacrifices or refraining from certain foods, which are believed to be prescribed by such agents; and (2) duties, I should not say to, but in respect of impersonal supernatural powers or forces which are believed to affect the welfare of an individual or his people, and which can be influenced by magico-religious means. It is true that both of these classes of strictly religious duties may be, and the latter must be, also, either prudential duties to oneself or social duties to one's fellow-men, or both, because a breach of them is believed to bring misfortune on the individual or his group. Nevertheless, they are primarily religious duties, because they cannot be understood without introducing the idea of the supernatural.

When the terms morality and religion are used in this sense, there is no people of whom we have any record, whether contemporary primitives or the early ancestors of people who are now advanced, who have not both morality and religion. A complete account of the relation between the morality and the religion of any of these peoples would have to consider how the different elements in their religion, the beliefs, the emotional attitudes and the ritual practices, are related to one another, and how, if at all, each of them is related to each of the three elements in morality, the rules, the ends and the motives. My purpose, as I have explained, is much more limited, but one or two further remarks on the wider issues may not be irrelevant to it.

On the relation between, and the relative priority of, the different elements in religion there is considerable difference of opinion among anthropologists as well as among other students of religion. The earlier anthropologists, like Tylor and Frazer, regarded the element of belief as fundamental and the other elements as secondary. Similarly, they emphasised the cognitive element in morality, rules and codes rather than motives and sentiments. Recently the tendency has been to regard the emotional and ceremonial elements in religion as basic and to treat religious beliefs as rationalisations to explain and justify the other elements. Thus, Marett, Goldenweiser and Lowie tend to regard the emotional element—the religious thrill as they call it—as the distinguishing characteristic of religion; while Radcliffe-Brown, following the line taken earlier by Robertson Smith and Durkheim, treats the ritual or ceremonial element as fundamental. According to this view, the performance of the ceremonies gives rise to the emotional aspect, and the beliefs about the supernatural are rationalisations to justify the ritual behaviour. Those who take these views tend also to emphasise the emotional element in morality, motives and sentiments rather than codes and rules, ends and ideals.

I cannot enter into the merits of these views, but so far as they bear on our problem they seem to me to call for two comments: (1) Wherever we find religion, while one of the elements may be more explicit or prominent, all of them are present in some degree or at least they are liable to emerge at any moment.6 They seem to develop together and in their development they interact and mutually modify one another. (2) No doubt the earlier views were over-intellectual, but the later views seem in danger of erring in the opposite direction. They tend to regard the supernatural as unknowable if not non-existent and all beliefs about it as irrational; and so they treat one set of myths, or one form of wishful thinking about the supernatural, as being as good as any other as long as it produces desirable effects on the lives of those who believe it. They therefore find it difficult to explain why changes in religious beliefs come about.7 Nor can they account for the fact that religion is never content to regard itself as merely a matter of feeling or ritual practices. The element of ideas and beliefs about the supernatural seems to be an essential part of it, a part without which it could not fulfil the functions in the lives of individuals and societies to which the anthropologists attach so much importance. And so we have the paradoxical situation of professed sceptics and agnostics stressing the necessity of religion, including religious beliefs, if morality and civilisation are to survive. Let me give just one example. Malinowski begins his Riddell Lectures on The Foundations of Faith and Morals by declaring himself a rationalist and an agnostic, who cannot accept the dogmas of any religion and has no faith except a faith in humanity and its powers of improvement;8 but he confesses himself so alarmed by certain tendencies of our age and by the failure of science to provide a basis for ethics that he goes on not only to profess his belief in the value of religion, but to call on his readers to defend its “eternal truths” against all attack;9 and he concludes with words which remind us of Voltaire's dictum that, if God did not exist, men would have to invent him. “The rationalist and the agnostic”, he writes, “must admit that even if he himself cannot accept these truths, he must at least recognise them as indispensable pragmatic figments without which civilisation cannot exist.”10

I have quoted this personal confession of Malinowski's not just as an interesting item of biographical information, but partly because it will help us to understand some confusions in his thinking about the relation of morality and religion to which I want to call attention later, and partly because the idea of “pragmatic figments” which are to be defended as “eternal truths” seems to me a rather glaring example of an attitude, and a sort of confusion about our subject, which is not uncommon among our contemporaries.

Now the questions which I want specially to consider are: (1) How far and in what sense, if in any, do primitive people regard their moral duties to their fellow-men as also religious duties, in the sense that they are prescribed or sanctioned by their religion? And (2) how far and in what ways, if at all, do their beliefs about the supernatural, the emotions which it evokes and the performance of their ritual or strictly religious duties, directly or indirectly influence their attitude and behaviour to their fellows, that is, the nature and performance of their strictly moral duties? As I have said, these are not the only, or even the only important, questions about the relation between morality and religion, even among primitives. But I want to concentrate attention on them partly because they are specially relevant to my subject, and partly because it seems to me essential that we should distinguish clearly between them. I believe that some anthropologists have confused them, and that when they submitted evidence in support of an answer to the one they assumed that they had answered the other also. I want, therefore, to bring out the distinction between them and between the evidence which is relevant to an answer to them, in more detail, and then to illustrate the way in which they have been confused and the consequences of such confusion.

When we say that a moral rule is prescribed by religion or has a religious sanction, this normally means one or more of three things. (1) That the rule was promulgated or revealed as the will of a supernatural power. This may take the form of a direct revelation to the individual, as we saw among the Crow Indians, or of a pronouncement by an accredited representative of the supernatural power such as priest or shaman or medium, or of a revelation to some wise man of bygone days, handed down by tradition in myth or legend. Moreover, this promulgation of the rule may be interpreted either as what makes the rule right or as a reason for believing it to be right. (2) That a breach of the rule will be followed by disastrous consequences to the individual or his group, either automatically through the operation of some impersonal supernatural power or as the result of the intervention of a superhuman agent who disapproves of the breach. Or (3) that a man's destiny in an after life will be affected according as he observes or breaks the rule.11

Now some or all of the moral rules of a particular society may have supernatural sanctions in one of these senses and not in another; and the same rule may have a supernatural sanction in one sense in one society and in another sense or not at all in another society. Moreover, a rule which has a supernatural sanction in any or all of these senses may have moral and social and legal sanctions as well—just as among ourselves a particular sort of act such as murder may be regarded as morally wrong, legally a crime against society, and, from the religious point of view, a sin against God. Thus the rules which have, and the senses in which they have, supernatural sanctions vary from one primitive people to another. For what we find are not primitive morality and primitive religion but primitive moralities and primitive religions; and the rules which are regarded as right and the sanctions, supernatural and other, which are attached to them, vary from people to people.

I shall return to the question which of the moral rules of primitive peoples have supernatural sanctions and in what senses they have them. At present I am concerned not with the answer to this question but with the distinction between it and my second question, namely: Does primitive man's religion influence his attitude to his neighbours, either in the way of determining or helping to determine what he takes his duties to them to be, or in the likelihood that he will perform those duties to them which he independently regards as binding? For even if we find, either generally or in the case of a particular people, that few or none of their moral rules have supernatural sanctions in the senses I have indicated, it by no means follows that their religion does not exercise an important influence on their individual and social morality.

To understand the ways in which this may happen it would be necessary to go in greater detail than is here possible into primitive man's conception of the supernatural—a conception which varies within wide limits from one primitive people to another.12 One or two illustrations must therefore suffice. If, for example, a man believes, however mistakenly, that his neighbour is a wizard or a magician endowed with deadly supernatural powers, this is likely to influence his attitude to him and what he believes his duty towards him to be. Again, the beliefs which a man entertains regarding the situations to which he has to respond, and the consequences of the actions open to him, will make a difference to what he thinks he ought to do. If, for example, he believes that human heads are necessary for the fertility of the soil, or that to allow twins to live will cause drought and bring disaster and possible starvation on himself and his people, this will make a difference to what he regards as his duty. In this way what we should call beliefs about non-ethical matters of fact, whether about the situations in which he has to act, the consequences of actions, or the persons towards whom he has to act, may help to determine what he regards as his moral duty; and such beliefs may be, and often are influenced by his views about the supernatural.

But whether or not the religious beliefs of the primitive in this way help to determine what he takes to be his moral duties, there is another and more important way in which his religion influences both his character and his attitude to his neighbours. These influences are mainly due to the ritual and ceremonial aspects of religion and to the emotional attitudes to which cither they or the conception of the supernatural gives rise. Some of these ritual duties are difficult and arduous and their performance teaches endurance and self-restraint. Some are undertaken in the belief that the community as a whole would suffer if they were neglected, and their performance teaches unselfishness and devotion to the common good. Others, like religious dances and other ceremonial performances in which many or all the members of the community take part together, produce an intensified tribal consciousness which makes for social harmony and cohesion. Nothing binds men more closely together than the common sharing of deeply moving experiences; and the more deeply moving the experiences the stronger the tie they produce between those who share them. Now many of the situations which are matters of religious concern to the primitive are themselves such as to arouse deep emotion; for they concern his most vital interests and the crises of his fate—his means of livelihood, illness and death and war and so on; and the sense of the supernatural in these situations further intensifies the emotions to which they naturally give rise. For nothing moves a man so profoundly or to the depth of his being as the consciousness of the presence of the supernatural. Consequently the religious ceremonies in which they take part together unite a people to one another and separate them off from others. In this way they foster strong social sentiments, and a spirit of goodwill and mutual trust and helpfulness, among the members of the group. Moreover, the sense of the presence of the supernatural drives a man back upon himself, at once humbling and exalting him. It tends to restrain the cruder and more selfish passions and to produce an attitude of seriousness and a sense of responsibility which make it more likely that he will do what he believes to be his duty to his neighbours.

Now it is these psychological and social consequences of religion which have specially impressed field workers among primitive peoples, and it is on them that they dwell when they discuss the moral value of religion. But however important these effects of religion may be, it seems to me essential for our understanding of the relation between morality and religion that we should not confuse them with the ethical content of religion nor regard them as evidence that morality is a part of or dependent on religion, in the sense that it prescribes the duties to be fulfilled, the rights to be respected or the ideals to be pursued, or that it provides a sanction for moral rules either by way of supplying grounds for believing them to be right or direct incentives to comply with their requirements. Many people, anthropologists and others, seem to me to be guilty of this confusion. They, therefore, assume that by producing evidence to show that religion has desirable effects on the character of individuals and the cohesion of society—effects which may in one sense of the term be called moral—they prove that religion is the foundation and justification of morality.

In trying to expose this confusion and to show the way in which evidence in support of the one conclusion is treated as a proof of the other to which it is irrelevant, I think it is more profitable to examine the statements of one writer in some detail than to deal in a sketchy way with those of several. The writer whose views I am going to consider is Malinowski, but I should add that I regard his views, in this respect, as an example of a confusion which seems to be widespread. My reasons for choosing him as my illustration are: his writings illustrate the point I want to make very clearly; he is one of the few recent anthropologists who have devoted some attention to the relation between primitive morality and religion; I have already given some of his views about their relation and it is only just that I should also give the rest; the evidence which he submits in support of his view is taken mainly from two of the peoples whose ways of life I have described in detail; his deservedly high reputation as a field worker and as interpreter of primitive life and thought gives his utterances great weight; and, finally, he states the facts which he discovered in his field work so clearly and he distinguishes so carefully between the facts and the conclusions which he draws from them, that we can check his theories by facts which he has himself provided. This, he tells us,13 was the ideal which he set before himself in his descriptive work, and he has succeeded to an unusual degree in realising it.

Malinowski is so impressed with the undoubtedly important influence of primitive religion in promoting the mental integrity of the individual and the social harmony of the group that he often writes as if their morality were merely a part of their religion and as if their religion were the source and justification of all their moral rules and ideals and provided their incentive for obeying the one and pursuing the other.

This is what I now want to illustrate. I shall first quote statements from his writings which seem to show (1) that he thinks of morality as a part of, or at least as dependent on religion, and (2) that in doing so he is referring not only to the motives of moral actions or to the influence of religion on character, but also to moral rules and principles, to what is regarded as right. I shall then examine the evidence which he gives in support of these statements in order to show what it proves and what it does not, and in particular to show that evidence in support of a conclusion regarding the inner or subjective aspect of morality is treated as a proof of a conclusion regarding the outer or objective aspect, to which it is irrelevant. Having done this, we shall be in a better position to consider the empirical evidence which Malinowski and others have provided regarding the relationship between primitive morality and religion, and in particular, to see whether, and if so how far, their religion prescribes or sanctions their moral rules.

Let me begin then with some quotations from Malinowski. In his Riddell Lectures, which he describes as “an anthropological analysis of primitive beliefs and conduct with special reference to the fundamental problems of religion and ethics”, and which he claims “is documented with the most relevant and most telling facts”,14 he tells us that in all religions there are three elements which are inseparably connected, dogma, ritual and ethics.15 The first element, the dogma, consists of legends and myths about the nature and activities of supernatural beings. It constitutes the sacred tradition of the people concerned. Its function is to provide reasons for the other two elements. These reasons take the form of precedents and, according to Malinowski, they are reasons which justify rather than reasons which explain. He speaks of them as “validating”, “sanctioning”, “justifying”, “being the charter of” both ritual and moral rules. The second element, the ritual, consists of rites and ceremonies, whether magical or religious, the ways in which individuals should behave towards supernatural powers. The nature of the third element, the ethical or moral element, is far from clear. At times it seems to be an emotional attitude, an attitude of friendliness and loyalty and goodwill between the fellow-worshippers which is produced by belief in the dogma and, more especially, by the performance of the ritual. At other times, in his references to it, Malinowski explicitly includes moral rules and principles and codes of conduct, the kind of actions which are regarded as right or morally obligatory. For example, he writes: “Every religion, however humble, carries instructions for a good life; it invariably provides its followers with an ethical system”.16 In every religion “there must be an ethical code of rules which binds the faithful and determines their behaviour to one another”.17 And again: “Every religion carries its own morals”,18 “it implies some reward of virtue and punishment of sin”;19 and yet again: into the ritual practices of ancestor worship “ethics comes in because the spirits and their reactions are determined by moral principles . . . they expect . . . good behaviour".20 In another work, he writes of” the place of morals in early, primitive, religion”, and of “the problem of morals as an early religious function”.21 “Religious ritual, in the ceremonies of initiation, . . .” he tells us, “establishes the existence of some power or personality from which tribal law is derived and which is responsible for the moral rules imparted to the novice.”22 Again, “all the morality of primitives is derived from religious belief”;23 and even the public performance of religion is necessary for “the maintenance of morals”.24 Of the function of myths and their relation to conduct he writes: sacred myth “safeguards and enforces morality . . . and contains practical rules for the guidance of man”.25 We “find sacred stories wherever there is . . . some fundamental ethical process at stake”.26 “The sacred tradition, the myth . . . controls moral and social behaviour.”27 “Myth supplies a . . . pattern of moral values.”28 Again, he describes religion as “moral in its very essence”,29 “always the mainspring of moral values”,30 “the permanent source of moral control”.31 Finally, he writes of “the ethical element intrinsically inherent in all religious activities”, and he describes “taboos, vigils and religious exercises” as “essentially moral”.32

I admit that, torn from their context, these statements convey a very inadequate impression of Malinowski's position taken as a whole, but I submit that, in any context, the natural interpretation to put upon them is a one-sided dependence of morality on religion, and that this dependence applies to moral rules as well as to the motives of moral actions. Partly to confirm that this is the natural interpretation of such statements, and partly to show the difficulties and confusion to which, as so interpreted, they give rise, I want to refer to an article in which Godfrey Wilson, one of Malinowski's most distinguished pupils, so interprets them. Wilson tried to apply what he took to be his master's principles to the morality of an African tribe. He gives the result in an article33 which he tells us was inspired throughout not only by the writings, but also by the oral teaching of Malinowski. Wilson begins his account by defining morality as “those forms of right conduct which have a supernatural sanction”.34 He soon finds, however, that some of the rules and forms of conduct which are usually regarded as moral have no supernatural sanction in any ordinary sense of the term among the tribe with which he is concerned. And what does he do? He just points out that he is precluded by his definition from calling such rules and conduct moral. Instead he calls them rules of good manners.35 It does not seem to have occurred to him that the facts called rather for a rejection of his original definition—so strong and persistent is the assumption that morality must have a religious sanction.

Now, as far as I am aware, Malinowski has not anywhere in his writings committed himself to Wilson's definition of morality, but it is significant that his views should be so interpreted by one of his ablest pupils; and this interpretation is the natural one to put on many of the statements I have quoted from him. So interpreted, however, they are not warranted by Malinowski's evidence and, as we have already seen in considering his account of the Trobrianders and shall see further, there is much in his writings that is inconsistent with them.

What, then, is the evidence which he submits in support of these statements? I shall confine myself mainly to that which is given in the Foundations of Faith and Morals. The examples which he there gives of duties which have a supernatural sanction, in the sense that breaches of them are punished by supernatural means, are of the nature of ritual obligations and do not directly concern man's relation to his fellows—such, for example, as carrying out fishing magic and the observance of mortuary ritual.36 This is true of all his examples from the Trobrianders and the Australian Aborigines from whom most of his illustrative material is taken. It is equally true of his examples from the Bemba and the Toda. The only possible exception is from the Pueblo Indians. Among them, we are told, “virtue and morality in the ordinary conduct of life are enjoined under supernatural sanctions”.37 Even this, however, is supported by a quotation (from Coolidge, The Rainmakers, p. 204) which leaves something to be desired in the way of clearness. It is: “These virtues (i.e. goodness, unselfishness, truth-telling, respect for property, etc.) are [among the Pueblo] closely connected with religious belief and conduct, but not their principal object”.38

Similarly, the examples which he gives of forms of conduct, which are prescribed or justified by the sacred tradition, the legends and myths of these people, are of the same ritual and ceremonial nature.39 They concern strictly religious duties, not moral duties to their fellow-men. And none of the peoples to whom he refers believe that they will be called to account hereafter for the way in which they behave to their fellow-men; nor that their happiness in another world will in any way depend on such behaviour. Thus, Malinowski gives very few, if any, examples of moral rules or ideals being prescribed or sanctioned by primitive religion in any of the three senses which these terms normally bear.

The rest of the evidence in support of his thesis—and it contains the greater part of the argument of the Lectures—consists in pointing out the psychological and social consequences of religion, the effects produced on the character and attitude of mind of individuals, and on the cohesion and solidarity of society, by the performance of the ritual and ceremonial duties which their religion obliges them to undertake. Many of these duties, he points out, are irksome and exacting. Their performance, therefore, makes for self-control and discipline and submission to leadership. Some of them are believed to be necessary for the welfare of the community. Their performance, therefore, fosters the spirit of unselfish service, the subordination of one's own inclination to the good of others; and, as Malinowski points out, though the belief is mistaken and the benefit to others illusory, the unselfishness is real and valuable.40 Similarly, the joint participation by the members of a community in ritual ceremonies makes for social unity and harmony, and produces a sense of common responsibility and a spirit and attitude of mind in which they are more likely to perform their duties to one another. For, as Malinowski says in support of the contention that “ethics is an essential element in religion”, “men cannot worship in common without a common bond of mutual trust and assistance, that is, of charity and love”.41 In the same way, the conclusion that “taboos, vigils and religious exercises are essentially moral” is derived from the fact that they require the sacrifice of a man's personal comfort for what he believes to be for the common weal.

Malinowski's argument, therefore, seems to be: Religion prescribes and sanctions ritual and ceremonial duties. The performance of these duties produces in the individual certain qualities of character and a spirit of goodwill to his neighbours. These in turn stimulate him to do his duties to his neighbours. Therefore, religion prescribes and sanctions these duties; it tells a man what his duties are and provides the justification for doing them; it is, in fact, the foundation and justification of morality.

Now we may accept all Malinowski's premises; his evidence in support of them is conclusive.42 But they do not warrant the conclusion which he draws from them. All that they entitle us to conclude is that religion produces or promotes goodwill and moral motives. They do not entitle us even to conclude that religion alone produces or promotes them or that they would be impossible without it; much less that it prescribes or sanctions them. For it is one thing to influence a character for good or to produce moral motives or a spirit of goodwill; it is quite another to prescribe or sanction these things. And even if we were to admit that religion prescribes and sanctions—and not merely produces and promotes— goodwill and moral motives, it does not follow that it tells a man what his duties are, that it prescribes moral rules or attaches sanctions to them.

Moreover, despite the ambiguous and misleading language which Malinowski's enthusiasm for the important function which religion fulfils in the life of primitive people leads him to use, I do not believe that he really wants us to draw such a conclusion. For it is inconsistent with positions which he takes in other parts of his writings, and which constitute some of his most significant contributions to social anthropology. Indeed, no one has done more than Malinowski to show that most of the rules of conduct of primitive peoples have no supernatural sanction, and that they do not obey them, merely or even mainly, from fear of supernatural powers. But he did not work out the implications of these facts for morality or its relation to religion; and I believe he was prevented from doing so largely by the presupposition, which he never seems to have questioned, that morality must in some sense have a religious sanction. Far from belittling the importance of the contributions which he has made to the subject, I am merely trying to work out the implications of the facts to which he himself called attention in a field to which he did not devote much attention; and to reconcile these implications with the important function which, as he and other field workers realise, religion performs among primitives; I want also to point out the modification which a recognition of these implications requires in some of his statements, which, as they stand, seem to be misleading.

What, then, are some of the facts which he has himself recorded which are inconsistent with the above conclusion? There is, so far as I have been able to discover, only one rule governing the relations between individuals which he reports as having a direct supernatural sanction among the Trobrianders. It is the rule against incest, breaches of which are believed by them to be followed automatically by eruptions and boils and, it may be, by death.43 Respect for the chieftainship and rank is also believed by them to have at least a semi-supernatural sanction, due to the powerful magic which the chief is supposed to possess. He also mentions a vague feeling which is sometimes found among them that their ancestors will be angry at breaches of custom, and that disasters may follow as a result;44 but whether this applies to secular as well as to ritual customs is not clear. The examples which he gives are of ritual duties only.45 As regards all other rules governing the relations between persons as members of a society, i.e. the ordinary rules of individual and social morality, not only their principles of justice, distributive and corrective, but their moral code about respect for person and property, about truth-telling and promise-keeping—all these, he tells us explicitly and repeatedly in his detailed discussion of them in Crime and Custom in Savage Society, have neither a religious origin nor a supernatural sanction. “They have in no way the character of religious commandments”, he writes;46 “they are not endowed with any mystical character . . . not set forth in the name of God, not enforced by any supernatural sanction, but provided with a purely social binding force”.47

As we have seen, in that work Malinowski is concerned primarily with the incentives which the natives have to obey the rules which they regard as right and the sanctions or social machinery which encourage compliance with their requirements. They are sanctioned, he tells us, by the working of the principle of reciprocity, the mutuality of services, which underlies all their institutions; and among the motives which the natives have for obeying them, he mentions “sense of duty and recognition of the need for co-operation”,48 “regard for the rights of others”,49 and loyalty to the group, as well as self-interest and fear or hope of consequences, economic or other. In that work, he calls these rules of right conduct rules of primitive law. In a later work, in response to the criticism that it is misleading to speak of rules of law among people who have no ad hoc machinery for enforcing them, he calls them sanctioned customs, but he again emphasises that the sanctions for them are not supernatural rewards or punishments but the working of the principle of reciprocity and moral and social approval and disapproval.50 But, by whatever name they are called, they include practically the whole of what we normally call the rules of secular morality; and I suspect that nothing but the preconceived idea that morality is either a part of, or intimately connected with, religion prevents Malinowski from regarding them as moral as well as legal rules or sanctioned customs.

This interpretation of the situation is supported by another fact. Malinowski does sometimes speak of these rules and principles as moral or ethical. He does so when he considers them in relation to religion and thinks of ritual and ceremonial practices as strengthening the bonds between members of the community and promoting a spirit of goodwill among them which acts as a motive for carrying out their duties to one another. In other words, when he regards them as moral rules and principles, he tends to think of them as a part of religion or, at least, as having a supernatural sanction. Thus, for example, he writes of “the moral principle of give and take”, “the ethical principle of co-operative services” as applying not only to the relations between the living members of the community but also to those between the living and the dead, and so having a supernatural sanction.51 Again, he speaks of “the principle of mutuality of services on which the possibility of co-operation depends” as requiring for its maintenance the sanction of “the public performance of ritual” and even of religious or sacred myths.52

Thus we find the same rules and principles described sometimes as rules of primitive law or sanctioned customs, without any religious or supernatural sanction but enforced by the automatic working of the principle of reciprocity, and at other times as moral rules having a religious sanction and maintained only by the public performance of religious ritual. And the facts which these apparently inconsistent statements, when considered in the light of their contexts and the evidence offered in support of them, describe seem to be: (1) that the rules in question are moral rules which are recognised as right, because they are the conditions of effective co-operation on which individual and social well-being and the smooth working of institutions depend; (2) that the need for complying with their requirements is brought home to the individual through the working of the principle of reciprocity; and (3) that the performance of magico-religious duties and ceremonies promotes social sentiments and a spirit of goodwill, as a result of which individuals are more likely to comply with the rules which they regard as right.

There are certain considerations—such as failure to distinguish clearly between different aspects of morality, the relative importance of these aspects for the work on which anthropologists are engaged, and the somewhat unusual terminology which they use—which may help us to understand how anthropologists come to make statements such as I have quoted from Malinowski. I want to explain how these considerations seem to me to have led the anthropologists to make these statements; but however they came to make them, nothing will enable us to reconcile some of them with one another or with the facts, except on the assumption that some of them do not mean what on the surface they appear to say.

To every piece of conduct there is an inner or subjective side, the motives or attitudes of mind from which it is done; and an outer or objective side, the external actions performed. In passing moral judgement on an act we have to take account of both aspects. For no action is, in the full sense, morally good unless it is both done from a good motive and aims at producing what is believed to be objectively right. Now, anthropologists tend to think of the objective aspect, what is right or obligatory, as determined by the social structure; and, therefore, they deal with it under the heading of social organisation. For example, A. W. Hoernle writes: “by ‘social organisation’ or ‘social structure’. . . is meant the more or less permanent framework of relationships between members of the community which manifests itself in an ordered group-life, with reciprocal rights and duties, privileges and obligations, of members, determining behaviour-patterns for each individual member towards other members, and moulding the feelings, thoughts, and conduct of members according to these patterns, so that it is only in and through them that the individual can achieve his personal self-realisation and participate in the satisfactions offered by the life of his community”.53 In the same way, Radcliffe-Brown describes his chapter on ‘social organisation’ as concerned with “the customs and institutions by which the natives regulate the conduct of persons to one another”;54 and we find a similar view expressed by Malinowski in his article on culture in the Encyclopaedia of the Social Sciences.55

Now if the outer aspect of morality, the rules of right conduct, as determined by a person's station and its duties, are regarded as part of social organisation and thus taken for granted as fixed and known, it is natural, in considering moral conduct, to think mainly or even merely of the inner aspect, the motives and attitudes of mind of the actors. This is the variable aspect. The appropriate attitude of mind, the right sentiments, the spirit of goodwill may or may not be present; and even if it is present, it may be of varying degrees of intensity. But while this inner aspect is morally very important, motives and sentiments and goodwill are not the whole of morality. They may provide the drive and dynamic, but moral conduct also needs direction; and for this right rules and good ideals are necessary. Indeed, as Radcliffe-Brown points out in connection with religious feeling, feelings and sentiments tend to be vague and indefinite until they are crystallised into definite modes of behaviour.

One effect of their exclusive emphasis on the inner aspect of conduct as being alone moral is that anthropologists tend to think of the same rule of conduct as moral, when people comply with it from one motive, as legal, when they comply with it for another, and as economic, when they comply with it for a third, and so on. In addition, some of them write at times as if all rules of conduct, at least among primitives, required a non-moral or extra-moral sanction. It is significant in this connection that in classifications of rules of conduct, which are based on the sanctions attached to the rules, moral rules as such are seldom included.56 It is, however, difficult to reconcile such a position with frequent statements that natives sometimes comply with rules from a sense of duty or from regard for the rights of others. In any case, a rule does not derive its moral rightness from the motives from which people comply with its requirements; and if a rule is morally right and recognised as such, it does not lose this character when legal or economic or other sanctions are attached to it, or even when people comply with it from such non-moral motives as fear of punishment by society or hope of economic reward. Only when this happens the actions of those who obey it from such motives are not morally good.

Now once the nature of a rule is supposed to be determined by the motives from which people comply with it, it is not a long step to the view that rules have a religious sanction, if religion produces the motives which make people comply with them. This, at any rate, is the view which some anthropologists seem to take, and such is the only evidence which they give in support of it. But does the evidence justify the conclusion that all or most primitive moral rules have a religious sanction, even in this sense of the term? In other words, are we justified in concluding that, even if primitive religion does not generally prescribe moral duties and does not attach supernatural rewards or punishments to their performance or neglect, primitive morality is still dependent on religion in the sense that religion produces the motives without which moral duties would not be performed? That is what some of Malinowski's statements suggest. But such statements give us only one side of the picture; and there are others which give the other side, and show that, even in respect of its inner side, the side of moral and social sentiments, the relation between morality and religion is not one-sided dependence but mutual influence and support. In the latter passages he represents moral and social sentiments as spontaneous developments of the nature of men who live in close contact with one another; and he traces the origin of certain religious beliefs and ritual ceremonies to these independently existing sentiments. “Human relations”, he writes in one such passage, “do not rest merely or even mainly on constraint coming from without. Men can only work with and for one another by the moral force which grows out of personal attachments and loyalties. . . . Lifelong bonds of co-operation and mutual interest create sentiments, and sentiments rebel against death and dissolution.”57 Hence arise mortuary ritual, and the belief in immortality and ancestor worship; and once these come into existence, they strengthen and support the social sentiments in which they had their source. These moral and social sentiments he here regards as the source of the inner constraint which finds expression in moral motivation and which he describes variously as sense of duty or loyalty to the group, conscience or sense of responsibility; and the sentiments themselves he describes as products partly of “lifelong bonds of co-operation and mutual interests”, and partly of “gradual training . . . within a definite set of cultural conditions”.58

Elsewhere, he tells us that, especially among ancestor-worshipping peoples, certain forms of religious ritual and the beliefs connected with them are “the religious extension of the ethical rules of conduct as between the members of the family”.59 This extension in turn tends to give the rules a supernatural flavour and thus to strengthen and intensify the motives to obey them; but it is neither the source of the motives nor the justification of the rules. So that however important religion may be in promoting goodwill and intensifying moral motives, it is not the sole source of such motivation.

Thus, when we take Malinowski's different statements into consideration, we find that, from whatever point of view we look at the relation between primitive morality and religion, what we have is not one-sided dependence but a complex system of interactions in which each influences and supports and strengthens the other. This is what his detailed account of the way of life of the Trobrianders shows to be true in their case. But whatever support their morality gains from or gives to their religious beliefs and practices, none of their rules of conduct governing the relations of individuals to one another, except the rules about incest, is either prescribed by their religion or has a direct supernatural sanction.

Moreover, as I have pointed out already, Malinowski's account makes it clear that there is a significant difference between the attitude of the natives to the rules of their secular morality which have not, and their magical and ritual rules which have, supernatural sanctions. They do not understand why the ritual and magical rules should be as they are. They accept them merely on authority. They obey them partly out of respect for custom, partly because the custom is sacred and partly because of fear of the consequences to themselves and their people which they believe would follow any breach of them. To their rules of individual and social morality their attitude is quite different. While they accept them in the first instance because they are the customs of their people, they understand why they should be as they are. They see that they are necessary for the smooth working of their institutions which they understand and find good. They regard them as right and reasonable because they recognise that they are the conditions of effective co-operation on which their own well-being and that of their people depend.

Among the Trobrianders, then, we find a relatively clear distinction between strictly religious duties with a supernatural sanction and duties to their fellow-men with a merely moral and social sanction. We find a similar state of affairs among the Aborigines of Central Australia—the other primitive people to whom Malinowski specially refers in his Foundations of Faith and Morals. I have already quoted the statement of Spencer and Gillen on whom Malinowski mainly relies for his information about these tribes. They write: “The Central Australian natives . . . have not the vaguest idea of a personal individual other than an actual living member of the tribe, who approves or disapproves of their conduct, as far as anything like what we call morality is concerned”. “But”, they continue, “it must not, however, be imagined that the Central Australian native has nothing in the nature of a moral code. As a matter of fact, he has a very strict one . . . but he quite understands that any punishment for [its] infringement . . . will come from the older men.”60 On the other hand, as we have already seen, their totemic and other magico-religious beliefs and practices exercise a profound effect on their character and their attitude to their neighbours. While they do not tell them how to behave towards one another or prescribe or sanction moral rules, they strengthen the social bonds between individuals and promote goodwill and a spirit of co-operation. Some of them even provide incentives, direct and indirect, for the performance of moral duties.61 So that among the Australian aborigines, too, we find religion producing important effects on the character of individuals and the social harmony of groups, while the main content of their moral rules and ideals is of non-religious origin and has no direct supernatural sanction, personal or impersonal.