CHAPTER XI

MR. FRAZER'S THEORY OF TOTEMISM

Mr. Frazer's latest theory—Closely akin to that of Professor Spencer—Arunta totemism the most archaic—Proof of Arunta primitiveness—Their ignorance of the facts of procreation—But the more primitive south-eastern tribes are not ignorant of the facts—Proof from Mr. Howitt—Yet south-eastern tribes are subject to Mr. Frazer's supposed causes of ignorance—Mr. Frazer's new theory cited—No account taken of primitive tribes of the southern interior—Similar oversight by Mr. Howitt as regards religion—Examples of this oversight—Social advance does not explain the religion of tribes which have not made the social advance—Theory of borrowing needed by Mr. Howitt—Mr. Frazer's suggestion as to the origin of exogamy—Objections to the suggestion.

Throughout these chapters, when there was occasion to mention the totemic theories of Mr. J. G. Frazer, we have spoken of them with reserve, as the theory of this or that date. Fortunately his article, "The Beginnings of Religion and Totemism among the Australian Aborigines," in the Fortnightly Review (September 1905), enables us to report Mr. Frazer's latest, perhaps final, hypothesis. "After years of sounding," he says, "our plummets seem to touch bottom at last."

In essence Mr. Frazer's latest hypothesis is that of Professor Baldwin Spencer. He accepts Pirrauru as "group marriage," and holds that the Arunta retain the most archaic form of totemism now known to exist. In Chapter III. we believe ourselves to have proved that Pirrauru is not "group marriage"; and that the "classificatory names for relationships "do not demonstrate the existence of "group marriage" in the relatively near, or of promiscuity in the very distant past.

In Chapter IV. we show that, by Professor Spencer's statement, the Arunta are in a highly advanced social state for Australians. Inheritance of local office (Alatunjaship) and of the paternal totemic ritual goes in the male, not in the female line of descent, which is confessedly the more archaic. (Mr. Frazer, however, now thinks this point open to doubt.) The institutions are of a local character; and the ceremonials are of what Professor Spencer considers the later and much more complex type. Arunta totemism, Mr. Spencer shows, depends on the idea of ancestral spirits attached to stone churinga nanja, amulets of various forms usually inscribed with archaic patterns, and these churinga nanja, with this belief about them, are not found outside of the Arunta region. Without them, the Arunta system of totemism does not, and apparently cannot exist On this head Mr. Frazer says nothing. For these and many other reasons, most of which have been urged by Dr. Durkheim, Mr. Hartland, Mr. Marett, and other students, we have explained the Arunta system as a late, isolated, and apparently unique institution. As the Arunta ceremonials and institutions, with inheritance in the male line and local magistracies hereditable in the male line, are at the opposite pole from the primitive, while the Arunta totemic system reposes on an isolated superstition connected with manufactured stone objects, and not elsewhere found in Australia, it has seemed vain to regard Arunta totemism as the most archaic.

This, however, is the present hypothesis of Mr. Frazer, as of Mr. Spencer, and he adduces a proof of Arunta primitiveness concerning which too little was said in our Chapter IV. The Arunta system "ignores altogether the intercourse of the sexes as the cause of offspring; and further, it ignores the tie of blood on the maternal as well as the paternal side."[1] The theory "denies implicitly, and the natives themselves deny explicitly, that children are the fruit of the commerce of the sexes. So astounding an ignorance of natural causation cannot but date from a past immeasurably remote."[2]

Now when the Arunta "ignore the tie of blood on the maternal side," they prove too much. They ignore that of which they are not ignorant. Not being idiots, they are well aware of the maternal tie of blood; but they do not permit it to affect the descent of the totem, which is regulated by their isolated superstition, the doctrine of reincarnation combined with the churinga nanja belief. Nor do they ignore fatherhood, as we saw, in affairs of inheritance of local office and totemic rites.

But they do deny that the intercourse of the sexes is the cause of birth of children. Here the interesting point is that tribes much more primitive, the south-eastern tribes, with female reckoning of descent, inheritance in the female line, and no hereditary local moderatorships, are perfectly well aware of all that the more advanced Arunta do not know. Yet they, quite as much as the Arunta, are subject to the causes which, according to Mr. Frazer, produce the Arunta nescience of the facts of procreation. That nescience, says Mr. Frazer, "may be explained easily enough from the habits and modes of thought of savage men." Thus, "first, the sexual act precedes the first symptoms of pregnancy by a considerable interval." Je n'en vois pas la nécessité. Secondly, savage tribes "allow unrestricted licence of intercourse between the sexes under puberty," and thus "familiarise him" (the savage) "with sexual unions that are necessarily sterile; from which he may not unnaturally conclude that the intercourse of the sexes has nothing to do with the birth of offspring." The savage, therefore, explains the arrival of children (at least the Arunta does) by the entrance of a discarnate ancestral spirit into the woman.

The conspicuous and closing objection to this theory is, that savages who are at least as familiar as the Arunta with (1) the alleged remoteness in time of the sexual act from the appearance of the first symptoms of pregnancy (among them, such an act and the symptoms may be synchronous), and (2) with licence before puberty, are not in the Arunta state of ignorance. They are under no illusions on these interesting points.

The tribes of social organisation much more primitive than that of the Arunta, the south-eastern tribes, as a rule, know all about the matter. Mr. Howitt says, "these" (south-eastern) "aborigines, even while counting descent—that is, counting the class names—through the mother, never for a moment feel any doubt, according to my experience, that the children originate solely from the male parent, and only owe their infantine nurture to their mother."[3] Mr. Howitt also quotes "the remark made to me in several cases, that a woman is only a nurse who takes care of a man's children for him."[4]

Here, then, we have very low savages among whom the causes of savage ignorance of procreation, as explained by Mr. Frazer, are present, but who, far from being ignorant, take the line of Athene in the Eumenides of Æschylus. I give Mr. Raley's translation of the passage:—

"The parent of that which is called her child is not really the mother of it, she is but the nurse of the newly conceived fœtus. It is the male who is the author of its being, while she, as a stranger for a stranger (i.e. no blood relation), preserves the young plant...."—Eumenides, 628-631.

These south-eastern tribes, far more primitive than the Arunta in their ceremonials, and in their social organisation, do not entertain that dominant factor in Aruntadom, the belief in the perpetual reincarnation of the souls of the mythical ancestors of the Alcheringa. That belief is a philosophy far from primitive. As each child is, in Arunta opinion, a being who has existed from the beginning of things, he is not, he cannot be, a creature of man's begetting. Sexual acts, say Messrs. Spencer and Gillen, only, at most, "prepare" a woman for the reception of a child—who is as old as the world! If the Arunta were experimental philosophers, and locked a girl up in Danae's tower, so that she was never "prepared," they would, perhaps, be surprised if she gave birth to a child.

However that may be, the Arunta nescience about reproduction is not caused by the facts which, according to Mr. Frazer, are common to them with other savages. These facts produce no nescience among the more primitive tribes with female descent, simply because these primitive tribes do not share the far from primitive Arunta philosophy of eternal reincarnation. If the Arunta deny the fact of procreation among the lower animals, that is because "the man and his totem are practically indistinguishable," as Mr. Frazer says. What is sauce for the goose is sauce for the gander.

The proof of Arunta primitiveness, the only proof, has been their nescience of the facts of generation. But we have demonstrated that, where Mr. Frazer's alleged causes of that nescience are present, among the south-eastern tribes, they do not produce it; while among the Arunta, it is caused by their system of philosophy, which the south-eastern tribes do not possess.

Mr. Frazer next applies his idea to the evolution of a new theory of the Origin of Totemism. Among the Arunta, as we know, each region has its local centre of totemic spirits awaiting reincarnation, one totem for each region. These centres, Oknanikilla, are, in myth, and for all that I know, in fact, burial-places of the primal ancestors, and in each is one, or there may be more, Nanja trees or rocks, permanently haunted by ancestral spirits, all of the same totem, whose stone amulets, churinga nanja, are lying in or on the ground. When a woman feels a living child's part in her being, she knows that it is a spirit of an ancestor of the local totem, haunting the Nanja, and that totem is allotted to the child when born.

Mr. Frazer from these known facts, deduces thus his new theory of the Origin of Totemism. It is best to give it in his own words:[5]—

"Naturally enough, when she is first aware of the mysterious movement within her, the mother fancies that something has that very moment passed into her body, and it is equally natural that in her attempt to ascertain what the thing is she should fix upon some object that happened to be near her and to engage her attention at the critical moment. Thus if she chanced at the time to be watching a kangaroo, or collecting grass-seed for food, or bathing in water, or sitting under a gum-tree, she might imagine that the spirit of a kangaroo, of grass-seed, of water, or of a gum-tree, had passed into her, and accordingly, that when her child was born, it was really a kangaroo, a grass-seed, water, or a gum-tree, though to the bodily eye it presented the outward form of a human being. Amongst the objects on which her fancy might pitch as the cause of her pregnancy we may suppose that the last food she had eaten would often be one. If she had recently partaken of emu flesh or yams she might suppose that the emu or yam, which she had unquestionably taken into her body, had, so to say, struck root and grown up in her. This last, as perhaps the most natural, might be the commonest explanation of pregnancy; and if that was so, we can understand why, among the Central Australian tribes, if not among totemic tribes all over the world, the great majority of totems are edible objects, whether animals or plants.[6] Now, too, we can fully comprehend why people should identify themselves, as totemic tribes commonly do, with their totems, to such an extent as to regard the man and his totem as practically indistinguishable. A man of the emu totem, for example, might say, 'An emu entered into my mother at such and such a place and time; it grew up in her, and came forth from her. I am that emu, therefore I am an emu man. I am practically the same as the bird, though to you, perhaps, I may not look like it.' And so with all the other totems. On such a view it is perfectly natural that a man, deeming himself one of his totem species, should regard it with respect and affection, and that he should imagine himself possessed of a power, such as men of other totems do not possess, to increase or diminish it, according to circumstances, for the good of himself and his fellows. Thus the practice of Intichiuma, that is, magical ceremonies, performed by men of a totem for its increase or diminution, would be a natural development of the original germ or stock of totemism.[7] That germ or stock, if my conjecture is right, is, in its essence, nothing more or less than an early theory of conception, which presented itself to savage man at a time when he was still ignorant of the true cause of the propagation of the species. This theory of conception is, on the principles of savage thought, so simple and obvious that it may well have occurred to men independently in many parts of the world. Thus we could understand the wide prevalence of totemism among distant races without being forced to suppose that they had borrowed it from each other. Further, the hypothesis accounts for one of the most characteristic features of totemism, namely, the intermingling in the same community of men and women of many different totem stocks. For each person's totem would be determined by what may be called an accident, that is, by the place where his mother happened to be, the occupation in which she was engaged, or the last food she had eaten at the time when she first felt the child in her womb; and such accidents (and with them the totems) would vary considerably in individual cases, though the range of variation would necessarily be limited by the number of objects open to the observation, or conceivable by the imagination, of the tribe. These objects would be chiefly the natural features of the district, and the kinds of food on which the community subsisted; but they might quite well include artificial and even imaginary objects, such as boomerangs and mythical beasts. Even a totem like Laughing Boys, which we find among the Arunta, is perfectly intelligible on the present theory. In fact, of all the things which the savage perceives or imagines, there is none which he might not thus convert into a totem, since there is none which might not chance to impress itself on the mind of the mother, waking or dreaming, at the critical season.

"If we may hypothetically assume, as the first stage in the evolution of totemism, a system like the foregoing, based on a primitive theory of conception, the whole history of totemism becomes intelligible. For in the first place, the existing system of totemism among the Arunta and Kaitish, which combines the principle of conception with that of locality, could be derived from this hypothetical system in the simplest and easiest manner, as I shall point out immediately. And in the second place, the existing system of the Arunta and Kaitish could, in its turn, readily pass into hereditary totemism of the ordinary type, as in fact it appears to be doing in the Umbaia and Nani tribes of Central Australia at present. Thus what may be called conceptional totemism pure and simple furnishes an intelligible starting-point for the evolution of totemism in general. In it, after years of sounding, our plummets seem to touch bottom at last."

How the totemic spirits became localised, is, Mr. Frazer says, "matter of conjecture," and he guesses that, after several women had felt the first recognised signs of maternity, "in the same place, and under the same circumstances "—for example, at the moment of seeing a Witchetty Grub, or a Laughing Boy—the site would become an Oknanikilla haunted by spirits of the Laughing Boy or Grub totem.[8] The Arunta view is different; these places are burial-grounds of men all of this or that totem, who have left their churinga nanja there. About these essential parts of the system, Mr. Frazer, as has been observed, says nothing. His theory I do not criticise, as I have already stated my objection to his premises. "The ultimate origin of exogamy ..." he says, "remains a problem nearly as dark as ever," but is a matter of deliberate institution. The tribes, already totemic, but not exogamous, were divided into the two exogamous phratries, and still later into the matrimonial classes, which the most pristine tribes do not possess, though they do know about procreation, while the more advanced Arunta, with classes and loss of phratry names, do not know. In the primitive tribes, with no churinga nanja, the totems became hereditary. Among the advanced Arunta, with churinga nanja, the totems did not (like all other things, including the right to work the paternal totemic ritual), become hereditary, though their rites did, which is curious. Consequently, Mr. Frazer suggests, the Arunta did not redistribute the totems so that one totem never occurs in both exogamous phratries; and totems in the region of churinga nanja alone are not exogamous.

Finally the tribes of Central Australia, which we prove to have the more advanced ceremonial, system of inheritance, local magistracies hereditary in the male line, and the matrimonial classes which Mr. Frazer proclaims to be later than the mere phratries of many south-eastern tribes—"are the more backward, and the coastal tribes the more progressive."[9]

This is a very hard saying!

It seems to rest either on Mr. Frazer's opinion that the south tribes of Queensland, and many on the Upper Murray, Paroo, and Barwan rivers are "coastal" ("which is absurd"), or on a failure to take them into account. For these tribes, the Barkinji, Ta-Ta-Thai, Barinji, and the rest, are the least progressive, and "coastal," of course, they are not.

This apparent failure to take into account the most primitive of all the tribes, those on the Murray, Paroo, Darling, Barwan, and other rivers, and to overlook even the more advanced Kamilaroi, is exhibited by Mr. Howitt, whose example Mr. Frazer copies, in the question of Australian religious beliefs.

I quote a passage from Mr. Howitt, which Mr. Frazer re-states in his own words. He defines "the part of Australia in which a belief exists in an anthropomorphic supernatural being, who lives in the sky, and who is supposed to have some kind of influence on the morals of the natives ... That part of Australia which I have indicated as the habitat of tribes having that belief" (namely, 'certainly the whole of Victoria and of New South Wales up to the eastern boundaries of the tribes of the Darling River') "is also the area where there has been the advance from group marriage to individual marriage, from descent in the female line to that in the male line; where the primitive organisation under the class system has been more or less replaced by an organisation based on locality—in fact, where those advances have been made to which I have more than once drawn attention in this work."[10]

This is an unexpected remark!

Mr. Howitt, in fact, has produced all his examples of tribes with descent in the female line, except the Dieri and Urabunna "nations," from the district which he calls "the habitat of tribes in which there has been advance ... from descent in the female to that in the male line." Apparently all, and certainly most of the south-eastern tribes described by him who have not made that advance, cherish the belief in the sky-dwelling All Father.

I give examples:—

About other tribes Mr. Howitt's information is rather vague, but, thanks to Mrs. Langloh Parker, we can add:—

Euahlayi — Female descent — All Father.

Here, then, we have eight tribes with female descent and the All Father, against five tribes with male descent and the All Father, in the area to which Mr. Howitt assigns "the advance from descent in the female line to that in the male line." The tribes with female descent occupy much the greater part of the southern interior, not of the coastal line, of South-East Australia.

Mr. Frazer puts the case thus, "it can hardly be an accidental coincidence that, as Dr. Howitt has well pointed out, the same regions in which the germs of religion begin to appear have also made some progress towards a higher form of social and family life."[11]

But though Dr. Howitt has certainly "pointed it out," his statement seems in collision with his own evidence as to the facts. The tribes with female descent and the "germs of religion" occupy the greater part of the area in which he finds "the advance from descent in the female line to that in the male line." He does find that advance, with belief in the All Father, in some tribes, mainly coastal, of his area, but he also finds the belief in the All Father among "nations" and tribes which have not made the "advance"—in the interior. As the northern tribes who have made the "advance" are mainly credited with no All Father, it is clear that the "advance" in social and family life has no connection with the All Father belief. Mr. Howitt, in saying so, overlooks his own collection of evidence. Large tribes and nations, in the region described by him, are in that social organisation which he justly regards as the least advanced of all, yet they have the "germs of religion," which he explains as the results of a social progress which they have not made.

In these circumstances Mr. Howitt might perhaps adopt a large theory of borrowing. The primitive south-east tribes have not borrowed from the remote coastal tribes the usage of male descent; they have not borrowed matrimonial classes from the Kamilaroi. But, nevertheless, they have borrowed, it may be said, their religion from remote coastal tribes. Of course, it is just as easy to guess that the coastal tribes have borrowed their Bunjil All Father from the Kamilaroi Baiame, or the Mulkari of Queensland.

I have not commented on Mr. Frazer's suggestion as to the origin of exogamy. It was the result, he thinks, of a deliberate reformation, and its earliest form was the division of the tribe into the two phratries. "Exogamy was introduced ... at first to prevent the marriage of brothers with sisters, and afterwards" (in the matrimonial classes) "to prevent the marriage of parents with children."[12] The motive was probably a superstitious fear that such close unions would be harmful, in some way, "to the persons immediately concerned," according to "a savage superstition to which we have lost the clue." I made the same suggestion in Custom and Myth (1884). I added, however, that totemic exogamy might be only one aspect of the general totem tabu on eating, killing, or touching, &c., an object of the totem name. We seem to have found the clue to that superstition, including the blood tabu, emphasised by Dr. Durkheim. But, on this showing, the animal patrons of phratries and totem kins, with their "religion," are among the causes of exogamy, while some unknown superstition, in Mr. Frazer's system, may have been the cause. As we have a known superstition, of origin already explained, it seems unnecessary to suppose an unknown superstition.

Again, if the reformers knew who were brothers and sisters, how can they have been promiscuous? Further, the phratriac prohibition includes vast numbers of persons who are not brothers and sisters, except in the phratry. Sires could prohibit unions of brothers and sisters, each in his own hearth circle; the phratriac prohibition is much more sweeping, so is the matrimonial class prohibition. Once more, parent with child unions do not occur among primitive tribes which have no matrimonial classes at all.

For these reasons Mr. Frazer's system does not recommend itself at least to persons who cherish a different theory.

He may, perhaps, explain the Kaitish usage, in which totems, though not hereditary but acquired in the Arunta manner, remain practically exogamous, by suggesting that the Kaitish are imitating the totemic exogamy of the rest of the savage world. But this hardly accounts for the fact that, among the Arunta, certain totems greatly preponderate in one, and another set of totems in the other exogamous moiety of the tribe. These facts indicate that the Arunta system is relatively recent, and has not yet overcome among the Kaitish the old rule of totemic exogamy. Mr. Frazer, too, as has been said, does not touch on the concomitance of stone churinga nanja with the Arunta system of acquiring totems.

APPENDIX

SOME AMERICAN THEORIES OF TOTEMISM

With some American theories of the origin of totemism, I find it extremely difficult to deal. They ought not to be neglected, that were disrespectful to the valued labours of the school of the American "Bureau of Ethnology." But the expositions are scattered in numerous Reports, and are scarcely focussed with distinctness. Again, the terminology of American inquirers, the technical words which they use, differ from those which we employ. That fact would be unimportant if they employed their technical terms consistently. Unluckily this is not their practice. The terms "clan," "gens," and "phratry" are by them used with bewildering inconsistency, and are often interchangeable. When "clan" or gens, means, now (i) a collection of gentes, or (2) of families, or (3) of phratries, and again (4) "clan" means a totem kin with female descent; and again (5) a village community; while a phratry may be (1) an exogamous moiety of a tribe, or (2) a "family," or (3) a magical society; and a gens may be (1) a clan, or (2) a "family," or (3) an aggregate of families, or (4) a totem kin with male descent, or (5) a magical society, while "tribal" and "sub-tribal divisions" are vaguely spoken of—the European student is apt to be puzzled! All these varieties of terminology occur too frequently in the otherwise most praiseworthy works of some of the American School of Anthropologists. I had collected the examples, but to give them at length would occupy considerable space, and the facts are only too apparent to every reader.[1]

Once more, and this point is of essential importance, the recent writers on totemism in America dwell mainly on the institution as found among the tribes of the north-west coast of the States and of British Columbia. These tribes are so advanced in material civilisation that they dwell in village settlements. They have a system of credit which looks like a satirical parody of the credit system of the civilised world. In some tribes there is a regular organisation by ranks, noblesse depending on ancestral wealth.

It seems sanguine to look for the origins of totemism among tribes so advanced in material culture. The origin of totemism lies far behind the lowest savagery of Australia. It is found in a more primitive form among the southern and eastern than in most of the north-western American tribes, but the north-western are chiefly studied, for example, by Mr. Hill-Tout, and by Dr. Boas. A new difficulty is caused by the alleged intermixture of tribes in very different states of social organisation. That intermixture, if I understand Mr. Hill-Tout, causes some borrowing of institutions among tribes of different languages, and different degrees of culture, in the west of British Columbia and the adjacent territories. We find, in the north, the primitive Australian type of organisation (Thlinket tribe), with phratries, totems, and descent in the female line. South of these are the Kwakiutl, with descent wavering in a curious fashion between the male and female systems. Further south are the Salish tribes, who have evolved something like the modern family, reckoning on both sides of the house. I, with Mr. McGee of the United States Bureau of Ethnology, suppose the Kwakiutl to be moving from the female to the male line of descent. In the opinions of Mr. Hill-Tout and Dr. Boas, they are moving from the advanced Salish to the primitive Thlinket system, under the influence of their primitive neighbours. It is not for me to decide this question. But it is unprecedented to find tribes with male reverting to female reckoning of descent

Next, Mr. Hill-Tout employs "totem" in various senses. As totems he reckons (1) the sacred animals of the tribe; (2) of the religious or magical societies (containing persons of many totems of descent); (3) of the individual and (4) the hereditary totems of the kin. All these, our author says, are, by their original concept, Guardian Spirits. All such protective animals, plants, or other objects, which patronise and give names to individuals, or kins, or tribes, or societies, are "totems," in the opinion of the late Major Powell, and the "American School," and are essentially "guardian spirits." All are derived by the American theory[2] from the manitu, or guardian, of some individual to whom the animal or other object has been revealed in an inspired dream or otherwise. The object became hereditary in the family of that man, descended to his offspring, or, in early societies with reckoning in the female line, to the offspring of his sisters (this is Mr. Hill-Tout's theory), and so became the hereditary totem of a kin, while men of various totem kins unite in religious societies with society "totems" suggested by dreams. These communities may or may not be exogamous, they may even be endogamous. By the friends of this theory the association of exogamy with hereditary kin-totemism is regarded as "accidental," rather than essential.

Using the word "totem" in this wide sense, or in these many senses, which are not ours, it is plain that a man and woman who chance to have the same "personal totem," (i) or belong to the same religious society with its "totem," (i) or to the same local tribe with its "totem," (3) may marry, and, by this way of looking at the matter, "totems" do permit marriage within the totem, and are not exogamous. But we, for our part (like Mr. E. B. Tylor, and M. Van Gennep[3]), call none of these personal, tribal, or society sacred animals "totems." That term we reserve for the hereditary totem of the exogamous kin. Thus it is not easy, it is almost impossible, for us to argue with Mr. Hill-Tout, as we and he use the term "totem" in utterly different senses.

On his theory there are all sorts of "totems," belonging to individuals and to various kinds of associations. The totems hereditary in the kins when they are exogamous, are exogamous (on Mr. Hill-Tout's theory) because the kins, in certain cases, made a treaty of alliance and intermarriage with other kins for purely political purposes. They might have made such treaties, and become exogamous, though they had no totems, no name-giving animals; and they might have had name-giving animals, and yet not made such treaties involving exogamy. Thus totemic exogamy is, on this theory, a mere accident: the totem has nothing to do with the exogamous rule.

Mr. Hill-Tout writes to me, "The totem groups are exogamous not because of their common totem, but because of blood relationship. It is the blood-tie[4] that bans marriage within the totem group, not the common totem. That exogamy and the totem group with female descent go together is accidental, and follows from the fact that the totem group is always, in Indian theory at least, blood related. Where I believe you err is in regarding exogamy as the essential feature of totemism. I cannot so regard it. To me it is secondary, and becomes the bar to marriage only because it marks kinship by blood, which is the real bar, however it may have arisen, and from whatever causes."

Here I am obliged to differ from Mr. Hill-Tout. I know no instance in which a tribe with female kin (the most primitive confessedly), and with hereditary totems, is not exogamous. Exogamy, then, if an accident, must be called an inseparable accident of totemism, with female descent, till cases to the contrary are proved to exist. Mr. Hill-Tout cites the Arunta case: totems among the Arunta are not exogamous. But of that argument we have disposed (see Chapter IV.), and it need no longer trouble us.

Again, it is not possible to agree with Mr. Hill-Tout when he writes, "It is the blood-tie that bars marriage within the totem group, not the common totem." The totem does not by its law prevent marriages of blood kin. A man, as far as totem law goes, may marry his daughter by blood, a brother may marry his sister on the father's side (with female descent), and a man may not marry a woman from a thousand miles away if she is of his totem, though she is not of his blood. It is not the real blood-tie itself, but the blood-tie as defined and sanctioned by the totem, that is not to be violated by marriage within it.

To return to the theory that totems are tutelary spirits in animal or other natural forms. A man may have a spirit guardian in animal form, that is his "totem," on the theory. He may transmit it to his descendants, and then it is their "totem"; or his sisters may adopt it, and hand it down in the female line, and then it is the totem of his nephews and nieces for ever; or the man may not transmit it at all. Usually, it is manifest, he did not transmit it; for there must have been countless species of animal protectors of individuals, but tribes in America have very few totems. If a man does transmit his animal protector, his descendants, lineal or collateral, may become exogamous, on the theory, by making other kins treaties of intermarriage to secure political alliances; or they may not, just as taste or chance direct. All the while, every "totem" of every sort, hereditary or not, is, on this theory, a guardian spirit. That spiritual entity is the essence of totemism, exogamy is an accident—according to Mr. Hill-Tout.

Such is his theory. It is, perhaps, the result of studying the North-West American Sulia, or "personal totem" answering to the nyarongs of Borneo, the naguals of the Southern American tribes, the yunbeai of the Euahlayi of New South Wales, and the "Bush Souls" of West Africa. All of these are, as the Ibans of Borneo imply in the term nyarong, "spirit helpers," in animal or material form. Some tribes call genuine totems by one name, but call animal familiars of an individual by another name. Budjan, among the Wiradjuri, stands both for a man's totem, and for the animal familiar which, rduring apparently hypnotic suggestion," he receives on being initiated.[5] Among the Ibans (but not among the few Australian tribes which have yunbeai), the spirit helper may befriend the great-grandchildren of its original protégé.[6]

But in no case recorded does this nyarong become the hereditary totem of an exogamous kin.

The "spirit helper" does not do that, nor am I aware, on the other hand, that the hereditary totem of an exogamous kin is ever, or anywhere, regarded as a "tutelary spirit." No such idea has ever been found in Australia. Again, if I understand Dr. Boas, among his north-western tribes, such as the Thlinket, who have female descent and hereditary exogamous totems, the totem is no more regarded as a tutelary spirit than it is among the Australians. Of the Kwakiutl he says, "The manitu" (that is, the individual's tutelary spirit) "was acquired by a mythical ancestor, and the connection has become so slight, in many cases, that the tutelary genius of the clan has degenerated into a crest."

That the "crest" or totem mark was originally a "tutelary genius" among the Thlinket, seems to be merely the hypothesis of Dr. Boas. Even among the Kwakiutl, in their transitional state, the totem mark now is "in many cases a crest." "This degeneration" (from spirit to crest), our author writes, "I take to be due to the influence of the northern totemism," such as that of the Thlinket.[7] Thus the Thlinket, totemic on Australian primitive lines, do not regard their hereditary exogamous totems as "tutelary spirits."[8] No more do the Australians, nor the many American totemists who claim descent from the animal which is their totem.[9]

The tutelary spirit and the true totem, in my opinion, are utterly different things. The American theory that all things (their name is legion) called "totems" by the American School are, in origin and essence, tutelary spirits, is thus countered by the fact that the Australian tribes do not regard their hereditary totems as such; nor do many American tribes, even when they are familiar with the idea of the tutelary spirits of individuals. The Euahlayi, in Australia for instance, call tutelary spirits yunbeai; hereditary totems they call by a separate name, Dhe.[10]

The theory that the hereditary totem of the exogamous kin is the "spirit helper" or "tutelary genius," acquired by and transmitted by an actual ancestor, cannot be proved, for many reasons. We know plenty of tribes in which the individual has a "spirit helper," we know none in which he bequeaths it as the totem of an exogamous kin.

Again we find, (1) in Australia, tribes with hereditary totems, but with no "personal totems," as far as our knowledge goes. Whence, then, came Australian hereditary totems? Next, (2) we find tribes with both hereditary and "personal totems," but the "personal totems" are never hereditable. The "spirit helpers," where they do occur in Australia, are either the familiars of wizards (like the witch's cat or hare), or are given by wizards to others.[11] Next, (3) we find, in Africa and elsewhere, tribes with "personal totems," but with no hereditary totems. Why not? For these reasons, the theory that hereditary kin-totems are personal tutelary spirits become hereditary, seems a highly improbable conjecture. If it were right, genuine totemism, with exogamy, might arise in any savage society where "personal totems" flourish. But we never find totemism, with exogamy, just coming into existence.

To sum up the discussion as far as it has gone, Mr. Hill-Tout had maintained (1) that the concept of a ghostly helper is the basis of all his varieties of so-called "totems." I have replied that the idea of a tutelary spirit makes no part of the Australian, or usually of the American "concepts" about the hereditary totems. This is matter of certainty.

Mr. Hill-Tout next argues that hereditary totems are only "personal totems" become hereditary, which may happen, he says, in almost any stage of savage society. I have replied, "not plus the totemic law of exogamy," and he has answered (3) that the law is casual, and may or may not accompany a system of totemic kindred, instancing the Arunta, as a negative example. In answer, I have shown that the Arunta case is not to the point, that it is an isolated "sport."

I have also remarked frequently, in previous works, that under the primitive method of reckoning descent in the female line, an individual male cannot bequeath his personal protective animal as a kin-name to his descendants, so that the hereditary totem of the kin cannot have originated in that way. Mr. Hill-Tout answers that it can, and does, originate in that way—a male founder of a family can, and does, found it by bequeathing his personal protective animal to the descendants of his sisters, so that it henceforth passes in the female line. I quote his reply to my contention that this is not found to occur.[12]

"The main objection brought against this view of the matter by Mr. Andrew Lang and others is that the personal totem is not transmissible or hereditable. But is not this objection contrary to the facts of the case? We have abundant evidence to show that the personal totem is transmissible and hereditable. Even among tribes like the Thompson, where it was the custom for every one of both sexes to acquire a guardian spirit at the period of puberty, we find the totem is in some instances hereditable. Teit says, in his detailed account of the guardian spirits of the Thompson Indians, that 'the totems of the shamans[13] are sometimes inherited directly from the parents'; and among those tribes where individual totemism is not so prevalent, as, for instance, among the coast tribes of British Columbia, the personal totem of a chief or other prominent individual, more particularly if that totem has been acquired by means other than the usual dream or vision, such as a personal encounter with the object in the forest or in the mountains, is commonly inherited and owned by his or her posterity. It is but a few weeks ago that I made a special inquiry into this subject among some of the Halkomelem tribes of the Lower Fraser. 'Dr. George,' a noted shaman[14] of the Tcil'Qe'Ek, related to me the manner in which his grandfather had acquired their family totem,[15] the Bear; and made it perfectly clear that the Bear had been ever since the totem of all his grandfather's descendants. The important totem of the Sqoiàqî[16] which has members in a dozen different tribes of the coast and Lower Fraser Salish, is another case in point. It matters little to us how the first possessor of the totem acquired it. We may utterly disregard the account of its origin as given by the Indians themselves, the main fact for us is, that between a certain object or being and a body of people, certain mysterious relations have been established, identical with those existing between the individual and his personal totem; and that these people trace their descent from and are the lineal descendants of the man or woman who first acquired the totem. Here is evidence direct and ample of the hereditability of the individual totem, and American data abound in it."

All these things occur under the system of male kinship. Even if the "personal totem" of a chief or shaman is adopted by his offspring, it does not affect my argument, nor are the bearers of the badge thus inherited said to constitute an exogamous kin.[17] If they do not, the affair is not, in my sense, "totemic" at all. We should be dealing not with totemism but with heraldry, as when a man of the name of Lion obtains a lion as his crest, and transmits it to his family. Meanwhile I do not see "evidence direct and ample," or a shred of evidence, that a man's familiar animal is borrowed by his sisters, and handed on to their children.

Next, as to that point, Mr. Hill-Tout writes:[18]—

"To return to Mr. Lang's primary objection, that the evolution of the group totem cannot proceed from the personal, individual totem because in the more primitive forms of society where totemism originated "male ancestors do not found houses or clan names," descent being on the female side. As Mr. Lang has laid so much stress upon this argument, and is able apart from it to appreciate the force of the evidence for the American point of view, if it can be clearly shown that his objection has no basis in fact, that his conception of the laws of inheritance under matriarchy is faulty, consistency must needs make him a convert to the American view. The singular error into which Mr. Lang has fallen is in overlooking the fact that male property and rights are as hereditable under mother-right as under father-right, the only difference being that in the latter case the transmission is directly from the father to his offspring, and in the former indirectly from the maternal uncle to his sister's children. What is there to prevent a man of ability under matriarchy from 'founding a family,' that is, acquiring an individual totem which by his personal success and prosperity is looked upon as a powerful helper, and therefore worthy of regard and reverence? Under mother-right the head of the clan is invariably a man, the elder male relative on the maternal side; and the clan name is not so much the property of the woman as of her elder brother or her conventional 'father,' that is, her maternal uncle. The 'fathers' of the group, that is, the maternal uncles, are just as much the heads and I founders of houses' and clans in the matriarchal state as under the more advanced state of patriarchal rule. And that they do found family and group totems the evidence from our northern coast tribes makes clear beyond the shadow of a doubt.

"The oft-quoted case of the Bear totem among the Tsimshians is a case in point, and this is but one of scores that could be cited. The origin of this totem came about in the following manner: 'A man was out hunting and met a black bear who took him to his home and taught him many useful things. After a lengthy stay with the bear the man returned home. All the people became afraid of him, he looked and acted so like a bear. Some one took him in hand and rubbed him with magic herbs and he became a man again. Thereafter whenever he went hunting his friend the bear helped him. He built a house and painted the bear on the front of it, and his sister made a dancing blanket, the design of which represented a bear. Thereafter the descendants of his sister used the bear for their crest, and were known as the Bear clan.'[19]

"Who was the 'founder of the family' here, and the source of the clan totem? Clearly and indubitably the many and so it invariably was, as the study of the myths accounting for the clan totems plainly shows.[20] It matters not, I may point out, that these myths may have been created since the formation of the clans to account for their origin, the point for us is that the man was regarded by the natives as the 'founder' of the family and clan. The founders of families and totem-crests are as invariably men under matriarchy as under patriarchy, the essential difference only between the two states in this regard being that under one the descent is through the 'conventional father,' under the other through the 'real or ostensible father.' Such being the case, Mr. Lang's chief argument falls to the ground, and the position taken by American students as to the origin of group-totems is as sound as before."

Now where, outside the region of myth, is there proof that Mr. Hill-Tout's processes ever do occur?

Mr. Hill-Tout argues that the founder of the totem kin is "invariably the man, as the study of the myths accounting for the clan totems plainly shows." But myths have no historical authority, and many of these myths show the very opposite: in them a beast or other creature begets the "clan."[21] To be sure, Mr. Hill-Tout says nothing about these myths, or about scores of familiar American myths[22] to the very same effect.

Again, as mythical evidence is worthless, Mr. Hill-Tout argues that "the man was regarded by the natives themselves as the 'founder' of the family or clan." Yes, in some myths, but not in those which Mr. Hill-Tout overlooks.

That the natives in some myths regard the man as founder of a totem kin under female descent proves nothing at all. Does the Tsimshian Bear myth prove that the natives themselves turn into Bears, and become men again? Does it even prove that such an occurrence, to-day, would now seem normal to them? Nothing is proved, except that in myth-making the natives think that this metamorphosis may have occurred in the past. In the same way—when myth-making—they think that a man might convey his badge to his sisters, to be hereditary in the female line. To prove his case, Mr. Hill-Tout must show that men actually do thus convey their personal protective animals and badges into the female line. To that evidence I shall bow.

If I reasoned like our author, I might argue, "The South African tribes say that their totems (siboko) arose in nicknames given to them on account of known historical incidents, therefore my conjecture that totems thus arose, in group names given from without, is corroborated by the natives themselves, who testify thus to the actuality of that mode of getting tribal names and siboko."[23]

But I, at least, cannot argue thus! The process (my process) does not and cannot occur in South African conditions, where tribes of an advanced culture have sacred protective animals. The natives have merely hit on my own conjecture, as to the remote germ of totemic names, and applied it where the process never occurs. The Tsimshians, in the same way, are familiar with the adoption of protective animals by male individuals. They are also familiar with the descent of the kin-totem through females. Like the famous writer on Chinese Metaphysics, the Tsimshians "combine their information." A man, they say, became a bear, and became a man again. He took the Bear for his badge; and to account for the transmission of the badge through women, the Tsimshians add that his sister also took and transmitted the Bear cognisance, as a hereditary totem. They think this could be done, exactly as the Bakwena think that their tribal protective animal, the Crocodile, the Baboon, or another, could arise in a nickname, given recently. It could not do so, the process is no longer possible, the explanation in this case is false, and does not help my theory of the origin of totemism. In the same way the Bear myth does not help Mr. Hill-Tout's theory, unless he can prove that sisters do actually take and transmit to their descendants, as exogamous totems, the sulia or individual protective animal of their brothers. Of this process I do not observe that Mr. Hill-Tout gives a single verifiable example.

As to this argument, Mr. Hill-Tout writes to me, "I cannot accept your criticism on the poor evidence of the Tsimshian accounts of the origin of their totem kins. You could not take such a view, I think, if you had personal, first-hand knowledge of the Indian mind. Your objections apply to 'classic myths,' but not to the accounts of tribes who are still in the totemic stage."

I fail to understand the distinction. It is now universally recognised that most myths, "classic" or savage (the classic being survivals of savage myths), are mere fanciful hypotheses framed to account for unexplained facts. Moreover, I am discussing and comparing the myths of various savage races, I am not speaking of "classic myths." Savages have anticipated us in every one of our hypotheses as to the origin of totemism, but, of course, they state their hypotheses in the shape of myths, of stories told to account for the facts. Some Australian myths favour Mr. Howitt's hypothesis, others favour that of Mr. Spencer, one flatters that of Dr. Haddon, one African myth is the fore-runner of my theory, and a myth of the Tsimshians anticipates the idea of Mr. Hill-Tout. But all these myths are equally valueless as historical evidence.

As to heritage under female kin, which I am said not to understand, no man reckoning by female kin has hitherto been said to inherit his totem from his maternal uncle! A man inherits his totem from his mother only, and inherits it if he has no maternal uncles, and never had. If a man has a manitu, a nagual, a yunbeai, a nyarong, or "personal totem," his sister does not take it from him and hand it to her children, or, if this ever occurs, I say once more, we need proof of it. A man may inherit "property and rights" from his maternal uncles under female kin. But I speak of the totem name, which a man undeniably does not inherit from his maternal uncle, while there is no proof offered that a woman ever takes such a name from her brother, and hands it on to her children. So I repeat that, under the system of reckoning in the female line, "male ancestors do not found houses or clan names," or are not proved to do so.

It is apparent, probably, that a theory of totemism derived in great part from the myths and customs of a few advanced tribes, dwelling in village communities, and sometimes in possession of the modern family, with male kin, is based on facts which are not germane to the matter. The origin of totemism must be sought in tribes of much more backward culture, and of the confessedly "more primitive" type of organisation with female descent To disprove Mr. Hill-Tout's theory is of course impossible. There may have been a time when "personal totems" were as common among the Australians as they are now rare. There may have been a time when an Australian man's sisters adopted, and transmitted, his "personal totem," though that is no longer done to our knowledge. It may have chanced that stocks, being provided, on Mr. Hill-Tout's plan, with tutelary spirits of animal names descending in the female line, made marriage treaties, and so became exogamous. Then we should have explained totemism, perhaps, but a considerable number of missing facts must be discovered and reported before this explanation can be accepted.

Mr. Hill-Tout's scheme, I presume, would work out thus: there are sets of human beings, A, B, C, D, E, F. In all of these every man acquires an animal, plant, or other friendly object. Their sisters adopt it as a name, and hand it on to their children. The stocks are now named after the familiar animals, as Grouse, Trout, Deer, Turtle, Buffalo, Salmon, and hundreds more. They have hitherto, I presume, married as they please, anyhow. But stocks Grouse and Deer think, "We shall be stronger if we give our women to each other, and never let a Grouse marry a Grouse, or a Deer a Deer." They make this pact, the other stocks, Salmon, Turtle, Buffalo, &c., come into it, ranging themselves under Deer or Grouse, and now Deer and Grouse are phratries in a tribe with the other animals as heads of totem kins in the phratries. The animals themselves go on being tutelary spirits, and are highly respected.

This scheme (whether Mr. Hill-Tout would arrange it just thus or not) works perfectly well. It explains the origin of exogamy—not by an inexplicable moral reform, and bisection of the horde, but as the result of a political alliance. It explains the origin of totemism by a theory of animal-shaped tutelary spirits taken on by sisters from brothers, and bequeathed by the sisters when they become mothers to their children. It explains the origin of phratries, and of totem kins in the phratries. It works out all along the line—if only one knew that very low savages deliberately made political alliances; and if all low savages had animal-shaped tutelary spirits; and if these were known to be adopted from brothers by sisters, and by sisters bequeathed, for an eternal possession, to their children; and if these transactions, once achieved, were never repeated in each line of female descent—no sister in the next generation taking on her brother's personal tutelary animal, and bequeathing it to her children for ever. Finally, if savages in general did regard their hereditary totems as tutelary spirits, the sketch which I make on Mr. Hill-Tout's lines would leave nothing to be desired. But we do not know any of these desirable facts.

If I have stated Mr. Hill-Tout's ideas correctly, he agrees with me in regarding the tribe as formed by aggregation of many more primitive groups. He does not regard the phratries and totem kins as the result of the segmentation of a primordial indiscriminate mass or horde, split up at the injunction of an inspired medicine man, or by a tribal decree. Against our opinion, Mr. Howitt argues that only one writer who "has or had a personal acquaintance with the Australian blacks" accepts it, the Rev. John Matthew. It is accepted, however, as far as "sub-phratries" go (as an alternative hypothesis), by Mr. Hewitt's friend, Dr. Fison.[24] But I have given my reasons for not accepting Mr. Howitt's doctrine, and I await some reason for his rejection of mine. Even authors who have "a personal acquaintance with the Australian blacks" should, I venture to think, give their reasons for rejecting one and persisting in another theory of "the probabilities of the case."[25] I have shown why I think it improbable that a postulated prehistoric tribe split itself up, for no alleged reason, at the suggestion of a medicine man. Now I am anxious to know why my postulated groups should not make marriage alliance for the reason of securing peace—a very sufficient motive for betrothals.