In any attempt to explain the origin of exogamy there are, in my opinion, three parallel groups of facts of general occurrence which necessarily must be taken into consideration:—Firstly, the prohibitions of incest and rules of exogamy themselves; secondly, the aversion to sexual intercourse between persons living together from early youth; thirdly, the injurious consequences of inbreeding. As for the facts of the first group, Frazer and I agree that they all have the same root, exogamy being in some way or other derived from an aversion to the marriages of near kin. As for the facts of the second group, Frazer at all events admits that “there seems to be some ground” for believing in them. As for the facts of the third group, there is complete agreement between us. I ask: Is it reasonable to think that there is no causal connection between these three groups of facts? Is it right to ignore the second group altogether, as does Frazer, and to look upon the coincidence of the first and the third as accidental? I gratefully acknowledge that Frazer’s chapter on the Origin of Exogamy has only strengthened my belief in my own theory.
Other objections to my theory have recently been made by Messrs. Hose and McDougall in their work on The Pagan Tribes of Borneo, vol. ii., p. 197, note. They observe that intercourse between a youth and his sister-by-adoption is not regarded as incest in these tribes, and that they know at least one instance of marriage between two young Kenyahs brought up together as adopted brother and sister. “This occurrence of incest between couples brought up in the same household,” they say, “is, of course, difficult to reconcile with Professor Westermarck’s well-known theory of the ground of the almost universal feeling against incest, namely, that it depends upon sexual aversion or indifference engendered by close proximity during childhood.” They moreover maintain that “the occurrence of incest between brothers and sisters, and the strong feeling of the Sea Dyaks against incest between nephew and aunt (who often are members of distinct communities),” are facts which are fatal to this theory.
In my attempt to explain the rules against incest I certainly did not overlook the fact that these rules very frequently have reference to persons who are, or may be, members of different communities, and I found no difficulty in accounting for it (see supra, [ii. 369]; History of Human Marriage, p. 330 sq.). Curiously enough Messrs. Hose and McDougall’s own attempt to solve the problem is, if I understand them rightly, based on the supposition that the prohibitions of intermarriage originally referred to persons who belonged to the same community. They write:—“If we accept some such view of the constitution of primitive society as has been suggested by Messrs. Atkinson and Lang (Primal Law), namely, that the social group consisted of a single patriarch and a group of wives and daughters, over all of whom he exercised unrestricted power or rights; we shall see that the first step towards the constitution of a higher form of society must have been the strict limitation of his rights over certain of the women, in order that younger males might be incorporated in the society and enjoy the undisputed possession of them. The patriarch, having accepted this limitation of his rights over his daughters for the sake of the greater security and strength of the band given by the inclusion of a certain number of young males, would enforce all the more strictly upon them his prohibition against any tampering with the females of the senior generation. Thus very strict prohibitions and severe penalties against the consorting of the patriarch with the younger generation of females, i.e. his daughters, and against intercourse between the young males admitted to membership of the group and the wives of the patriarch, would be the essential conditions of advance of social organisation. The enforcement of these penalties would engender a traditional sentiment against such unions, and these would be the unions primitively regarded as incestuous. The persistency of the tendency of the patriarch’s jealousy to drive his sons out of the family group as they attained puberty would render the extension of this sentiment to brother-and-sister unions easy and almost inevitable. For the young male admitted to the group would be one who came with a price in his hand to offer in return for the bride he sought. Such a price could only be exacted by the patriarch on the condition that he maintained an absolute prohibition on sexual relations between his offspring so long as the young sons remained under his roof.” I should like to know how Messrs. Hose and McDougall, on the basis of this theory, would explain “the strong feeling of the Sea Dyaks against incest between nephew and aunt (who often are members of distinct communities),” and, generally speaking, the rules prohibiting the intermarriage of persons belonging to different local groups. For the rest, I must confess that the assumptions on which their whole theory rests seem to me extremely arbitrary. Brothers are prohibited from marrying their sisters because the old patriarch drove away his grown-up sons out of jealousy; but his jealousy was not strong enough to prevent other young males from joining the band. On the contrary, he allowed them to be incorporated in it, because they added to its strength; nay, he gave them his own daughters in marriage, and refrained henceforth himself from intercourse with these young women so rigorously that ever since a father has been prohibited from marrying his daughter. But the young men had to pay a price for their wives. It may be asked: Why did not the old patriarch accept a price from his own sons or let them work for him, instead of mercilessly turning them out of their old home, although they would have been just as good protectors of it as anybody else? And why did he give the young men his daughters? He might have kept the young women for himself and let the young men have the old ones. This is what is done by the old men in Australia, where the young girls are, as a rule, allotted to old men, and the boys, whenever they are allowed to marry, get old lubras as wives (Malinowski, The Family among the Australian Aborigines, p. 259 sqq.). Yet, in spite of this custom, there is no country where incest has been more strictly prohibited than in Australia.
Messrs. Hose and McDougall maintain that the occurrence of incest between brothers and sisters and the feeling of the Sea Dyaks against incest between nephew and aunt are facts which seem “to point strongly to the view that the sentiment has a purely conventional or customary source.” I ask: Is it reasonable to suppose that, if this were the case, the feeling against sexual intercourse between the nearest relatives could have so long survived the conditions from which it sprang without showing any signs of decay? As I have pointed out above, the prohibited degrees are very differently defined in the customs or laws of different peoples, generally being more numerous among peoples unaffected by modern civilisation than they are in more advanced communities; and it appears that the extent to which relatives are prohibited from intermarrying is closely connected with the intimacy of their social relations. Whilst among ourselves cousins are allowed to intermarry, there is still a strong sentiment against intercourse between parents and children and between brothers and sisters, who in normal cases belong to the same family circle. Why should the feeling against incest have survived in this case but not in others, if it had a purely conventional origin? And how could any law based on convention alone account for the normal absence of erotic feelings in the relation between parents and children and brothers and sisters? It is true that cases of intercourse between the nearest relatives do occur, but they are certainly quite exceptional. Messrs. Hose and McDougall say themselves (p. 198) that “incest of any form is very infrequent” among the tribes of Borneo, and they seem to know of only one instance of marriage between young Kenyahs brought up together as adopted brother and sister, although such marriages are allowed. To maintain that cases of this kind are fatal to my theory seems to me as illogical as it would be to assume that the occurrence of a horror feminæ in many men disproves the general prevalence of a feeling of love between the sexes.
[P. 396, n. 1.]—In his recent work, The Family among the Australian Aborigines, Dr. Malinowski has come to the same conclusion. He observes that the individual family plays a foremost part in the social life of those aborigines; it has a very firm basis in their customs and ideas, and “by no means bears the features of anything like recent innovation, or a subordinate form subservient to the idea of group marriage.” The Australian husband had generally a definite sexual “over-right” over his wife, which secured to him the privilege of disposing of her, or at least of exercising a certain control over her conduct in sexual matters, even though this “over-right” did not, as a rule, amount to an exclusive right. There were customs like wife-lending, exchange of wives, ceremonial defloration of girls by old men, the different forms of licence practised at large tribal gatherings, and especially the Pirrauru relationship found in several of the southern central tribes. But all this does not constitute group marriage, the complete content of which does not consist in sexual relations alone. Dr. Malinowski emphasises the fact that marriage cannot be detached from family life; “it is defined in all its aspects by the problems of the economic unity of the family, of the bonds created by common life in one wurley, through the common rearing of, and affection towards, the offspring.” In nearly all these respects even the Pirrauru relationship essentially differs from marriage, and cannot, therefore, seriously encroach upon the individual family. Nor can we regard this relationship as a survival of previous group marriage. Dr. Malinowski also points out (p. 89 sq.) how highly objectionable it is that “our best informants (especially Howitt and Spencer and Gillen) describe the facts of sexual life of to-day in terms of their hypothetical assumptions.”
[P. 419, n. 5.]—For Moorish beliefs relating to contact between sexual uncleanness and holiness see my essays, The Moorish Conception of Holiness (Baraka), p. 123 sqq., and Ceremonies and Beliefs connected with Agriculture, certain Dates of the Solar Year, and the Weather in Morocco, pp. 17, 22, 23, 28, 46, 47, 54.
[P. 463, n. 8.]—During the years that have passed since the first edition of this book was issued, the study of homosexuality has been carried on with remarkable activity. The following books are exclusively devoted to this subject:—Das gleichgeschlechtliche Leben der Naturvölker, by F. Karsch-Haack (1911), Intermediate Types among Primitive Folk, by Edward Carpenter (1914), and Die Homosexualität des Mannes und des Weibes, by Magnus Hirschfeld (1914). Carpenter’s book chiefly deals with the invert in early religion and in warfare. Hirschfeld’s work is a veritable encyclopædia of homosexuality—according to Dr. Havelock Ellis, “not only the largest but the most precise, detailed, and comprehensive—even the most condensed—work which has yet appeared on the subject.” In 1915 Dr. Havelock Ellis issued a third, revised and enlarged, edition of his Sexual Inversion.
[P. 485, n. 1.]—This passage and, generally, the suggestion that there is a certain relationship between the social reaction against homosexuality and against infanticide, have been excluded from the last edition of Dr. Havelock Ellis’s book.
[P. 584, n. 1.]—There is hardly any subject which during the last four or five years has been more eagerly discussed by students of social anthropology than the relation between religion and magic. It has been dealt with, e.g., by Sir J. G. Frazer in The Magic Art, by Professor Durkheim in Les formes élémentaires de la vie religieuse, by Dr. Marett in The Threshold of Religion and other writings, by Dr. Irving King in The Development of Religion, by Professor Leuba in A Psychological Study of Religion, by Mr. Sidney Hartland in Ritual and Belief, and by the present Archbishop of Sweden, Nathan Söderblom, in his book Gudstrons uppkomst. According to the French school of sociologists, religion is social in its aims and magic antisocial; and this distinction has lately been accepted by Dr. Marett, who writes (Anthropology, p. 209 sq.): “Magic I take to include all bad ways, and religion all good ways, of dealing with the supernormal—bad and good, of course, not as we may happen to judge them, but as the society concerned judges them.” But this use of the terms is neither in agreement with traditional usage nor, in my opinion, suitable for the purpose of scientific classification. Besides black magic, or witchcraft, there is also white magic; even a medieval theologian like Albertus Magnus asserts that “magical science is not evil, since through knowledge of it evil can be avoided and good attained.” The French distinction between magic and religion implies that a prayer to a god for the destruction of an enemy must be classified as religion if it is offered in a cause which is considered just by the community, but as magic if it is disapproved of. If a man makes a girl drink a love-potion in order to gain her favour, it is religion if their union is desirable from the society’s point of view, but if he gives the same drink to another man’s wife it is magic. The best part of what has been hitherto called imitative or homœopathic magic no longer remains magic at all; if water is poured out for the purpose of producing rain it is homœopathic magic only in case rain is not wanted by the community, but if it is done during a drought it is religion. Thus the very same practices are qualified as religious or magical according as they have social or antisocial ends; and, as Mr. Hartland rightly asks (Ritual and Belief, p. 76): “How shall we define these ends?”
It should be added, however, that the definition of religion which I have given in the text has reference only to religion in the abstract, not to the various religions. In the popular sense of the word, a religion may include many practices which are what I have called magical. As I have said above ([p. 649]), “both Christianity in its earlier phases and Muhammedanism are full of magical practices expressly sanctioned by their theology.” Although the magical and the strictly religious attitude differ from each other, they are not irreconcilable, and may therefore very well form parts of one and the same religion; there is no such thing as a magic being opposed to a religion. By a religion is generally understood a system of beliefs and rules of behaviour which have reference to men’s relations to one or several supernatural beings whom they call their god or gods, that is, supernatural beings who are the objects of a regular cult and between whom and their worshippers there are established and permanent relationships. If it be admitted that the word religion may be thus legitimately used in two different senses, I think there is little ground left for further controversy on the subject. After all, sociologists may more profitably occupy their time than by continuous quarrelling about the meaning of terms.