existed among the Australian natives from the "earliest times."[222] Thus, "in its original form," the author concludes, "the division into classes is a striking confirmation of Morgan's theory, that the first step in the development of systems of relationship consists in the prevention of sexual union between parents and children in the wide sense."[223] For the same reason it follows that intermarriage between the nearest collateral relatives may not have been excluded.
Much more radical are the conclusions reached by Kohler in the monograph in which, by a minute examination of Morgan's tables and other materials, he seeks to establish the genetic relations subsisting between "totemism, group-marriage, and mother-right," as they appear among the Dravidians, the Australians, and the American aborigines. The researches of Curr and Westermarck are criticised as being too general and as lacking in rigid scientific method.[224] Abundant "survivals," such as the levirate, wife-lending, and, above all, the class-system, seem to demonstrate the former existence of group-marriage in Australia; and in the same way the same result is reached for the other peoples considered. The "key" to the problem is found in totemism, one of the most "formative and vitalizing impulses of mankind. In totemism lies the germ of the future family and state."[225] It is the characteristic feature of the social and religious life of the American Indians. From its very nature totemism favors the rise of mother-right and group-marriage. No person can belong to more than one totem, which is therefore of necessity exogamous. Choice must be made between the maternal and the paternal line—for totemism implies the blood-bond. This choice naturally leads directly to mother-right; since the relation of mother and child is the central fact in the genesis of social experience. The maternal system precedes the paternal, and no trustworthy examples of the opposite evolution have been discovered.[226] Furthermore, totemism leads straight to group-marriage. For if two totem-groups may intermarry, it follows that the "men of one totem may marry women of the other and vice versa." With kinship counted in the maternal line, this fact implies that a man may mate with his own daughter; while the union of mother and son or brother and sister would be excluded because of the identity of their totem.[227] Totemism is thus a means of differentiating matrimonial classes. "The whole history of group-marriage," the author concludes, "is a history of the restriction of marriage from totem to totem by the separation of under-totems through which marriage is subjected to definite conditions." Totemistic group-marriage appears to be the "starting-point" of social culture for all the races of mankind. Whether a more primitive stage of promiscuity may have preceded it, the author in this paper does not undertake to establish.[228]
Similar results are reached by Spencer and Gillen, who have given a remarkably clear and minute account of matrimonial, tribal, and totemistic institutions in central Australia. So far as the Australians are concerned, the theory that the classificatory nomenclatures are merely terms of address is positively rejected.[229] "When, in various tribes, we find series of terms of relationship all dependent upon classificatory systems such as those now to be described, and referring entirely to a mutual relationship such as would be brought about by their existence, we cannot do otherwise than come to the conclusion that the terms do actually indicate various degrees of relationship based primarily upon the existence of inter-marrying groups.... Whatever else they may be, the relationship terms are certainly not terms of address, the object of which is to prevent the native having to employ a personal name. In the Arunta tribe, for example, every man and woman has a personal name by which he or she is freely addressed by others—that is, by any, except a member of the opposite sex who stands in the relationship of 'Mura' to them, for such may only on very rare occasions speak to one another."[230] The fundamental idea of the Australian classes is "that the women of certain groups may marry the men of others." It is a device for limiting and defining the inter-marriage of persons supposed to be of common blood.[231]
Crawley likewise holds that the terms of the classificatory systems "are terms of kinship and not terms of address;" although being "in origin terms of relation," so far they are "terms of address also." For "all of the terms can be used as terms of address, just as our terms of relationship can be so used." The classificatory system, in some cases, appears clearly as a device to assist nature in confining marriage within the same generation.[232]
The results of the most recent research, therefore, seem to have advanced our knowledge of the early social condition of mankind; but not to have definitively settled the problem of the former existence of communistic marriage. One rises from an examination of the literature relating to totemism and to the classificatory systems of relationship with a feeling that much more material must be gathered and exploited before we shall escape entirely from the domain of speculation as to their full meaning.[233]
III. McLENNAN'S CONSTRUCTIVE THEORY
McLennan's theory[234] starts also with man in a condition, as he conceives it, resembling that of other gregarious animals. The unions of the sexes are "probably, in the earliest times, loose, transitory, and in some degree promiscuous."[235] There is no idea of consanguinity, though men may always have been held together by that "feeling of kindred" which arises in "filial and fraternal affection."[236] Everywhere when society emerges from this condition kinship is traced in the female line. Originally paternity is uncertain,[237] hence the recognition of relationship through the mother must of necessity have preceded the parental and the agnatic systems;[238] and this order of development is never reversed.
Though the "filial and fraternal affections may be instinctive," they are "obviously independent of any theory of kinship, its origin or consequences; they are distinct from the perception of the unity of blood upon which kinship depends; and they may have existed long before kinship became an object of thought." Such a group may have been held together chiefly "by the feeling of kindred;" but the "apparent bond of fellowship ... would be that they and theirs had always been companions in war or the chase—joint-tenants of the same cave or grove."[239] Slowly the idea of blood-relationship arose; and eventually observation led to a recognition of the system of kinship through the mother. Further than this, so long as paternity remained uncertain, primitive men could not go. For the theory in question, therefore, the maternal system of kinship existing in the homogeneous "group-stock" is the social fact of fundamental importance. But primitive man was rude, ignorant, relatively helpless. "Before the invention of the arts, and the formation of provident habits, the struggle for existence must often have become very serious. The instincts of self-preservation, therefore, must have frequently predominated." Society would tend toward one common type in which there was little place for the "unselfish affections." In this "struggle for food and security" the balance of the sexes would be disturbed. "As braves and hunters were required and valued, it would be to the interest of every horde to rear, when possible, its healthy male children." The weaker sex must obey the cruel law requiring the "survival of the fittest." Hence