THE SECRET OF THE TOTEM

BY

ANDREW LANG

LONGMANS, GREEN, AND CO.

39 PATERNOSTER ROW, LONDON

NEW YORK AND BOMBAY

1905

CONTENTS

[INTRODUCTION]

[APPENDIX: AMERICAN THEORIES]

[INTRODUCTION]

This book is the natural sequel of Social Origins and Primal Law, published three years ago. In Primal Law, Mr. J. J. Atkinson sought for the origin of marriage prohibitions in the social conditions of early man, as conceived of by Mr. Darwin. Man, in the opinion of the great naturalist, was a jealous animal; the sire, in each group, kept all his female mates to himself, expelling his adolescent male offspring. From this earliest and very drastic restriction, Mr. Atkinson, using the evidence of "avoidances" between kinsfolk in savage society, deduced the various prohibitions on sexual unions. His ingenious theory has been received with some favour, where it has been understood.

Mr. Atkinson said little about totemism, and, in Social Origins, I offered a theory of the Origin of Totemism; an elaboration of the oldest of all scientific theories, that of Garcilasso de la Vega, an Inca on the maternal side, the author of the History of the Incas. Totems, he conceived, arose in the early efforts of human groups to differentiate each from the others. Mr. Max Müller and Dr. Pikler set forth the same notion, independently. The "clans," or, as I say, "groups," needed differentiation by names, such as are still used as personal names by savages, and by names easily expressed in pictographs, and easily signalled in gesture language. The origin of the group names, or sobriquets, once forgotten, the names, as usual, suggested a relation between the various name-giving objects and the groups which bore them. That relation was explained by the various myths which make the name-giving animals, plants, and other objects, mystic kinsmen, patrons, or ancestors of the groups named after them. From reflection on this mystic rapport between the objects and the human groups of the same names, arose the various superstitions and tabus, including that which prohibits unions between men and women of the same animal group-name, whether by locality or maternal descent.

Critics objected that such a "trivial accident" as a name could not be the germ, or one of the germs of a great social system. But "the name goes before everything," as the Scots used to say; and in this book I have set forth the great importance of names in early society, a fact universally acknowledged by anthropologists.

It was also objected that names given from without would never be accepted and gloried in, so I now prove that such names have often been accepted and gloried in, even when they are derisive; which, among savages, names derived from plants and animals are not; they are rather honourable appellations.

So far, I have only fortified my position. But some acute criticisms offered in Man by Mr. N. W. Thomas enabled me to detect a weak point in my system, as given in Social Origins, and so led on to what I venture to think not unimportant discoveries regarding the Australian social organisations. To Mr. Thomas's researches, which I trust he will publish in full, I am much indebted, and he kindly read part of this book in type-written MS.

I also owe much to Mrs. Langloh Parker, who generously permitted me to read, in her MS., her valuable account of the Euahlayi tribe of New South Wales, which is to be published by Messrs. Archibald Constable. No student has been so intimately acquainted as this lady with the women of an Australian tribe; while the men, in a place where they could be certain that they were free from tribal espionnage, were singularly communicative. Within its limits, Mrs. Langloh Parker's book, I think, may be reckoned almost as valuable as those of Messrs. Spencer and Gillen.

By the irony of fortune, I had no sooner seen my book in print, than Mr. J. G. Frazer's chapter on "The Beginnings of Religion and Totemism among the Australian Aborigines" (Fortnightly Review, September 1905) came into my hands. I then discovered that, just when I thought myself to have disentangled the ravelled thread of totemism, Mr. Frazer also thought, using another metaphor, that his own "plummets had found bottom"—a very different bottom. I then wrote Chapter XI., stating my objections to his theories. Many of these, mainly objections to the hypothesis of the relative primitiveness of the Arunta "nation," had often been urged before by others. I was unaware that they had been answered, but they have obviously been deemed inadequate. Meanwhile the question as between two entirely different solutions of the old mystery remains open.

Since critics of my Social Origins often missed my meaning, I am forced to suppose that I may in like manner have misconstrued some of the opinions of others, which, as I understand them, I am obliged to contest. I have done my best to understand, and shall deeply regret any failures of interpretation on my own part.

Necessarily I was unaware that in Mr. Frazer's opinion, as set forth in his essay of September 1905, "the common assumption that inheritance of the totem through the mother always preceded inheritance of it through the father need not hold good." I have throughout argued on that assumption, which I understood to be held by Mr. Frazer, as well as by Mr. Taylor, Mr. Howitt, and most authorities. If it be correct, as I still think it is, it cannot but be fatal to the Arunta claim to primitiveness. But Arunta society is, in many points, so obviously highly organised, and so confessedly advanced, that I am quite unable to accept this tribe as an example of the most archaic state of affairs extant. If I am wrong, much of my argument is shaken, and of this it is necessary to warn the reader. But a tribe really must be highly advanced in organisation, if it can afford to meet and devote four months to ceremonials, as it did, in a region said to be relatively deficient in natural supplies.

In this book I have been able to use the copious materials of Mr. Howitt and Messrs. Spencer and Gillen in their two recent works. It seems arrogant to differ from some of the speculative opinions of these distinguished observers, but "we must go where the logos leads us."

I end by thanking Mr. H. J. Ford for his design of Eagle Hawk and Crow, heading the totems in their phratries, and betrothing two interesting young human members of these divisions.

THE SECRET OF THE TOTEM

CHAPTER I

ORIGIN OF TOTEMISM

The making of the local tribe of savagery—Earliest known stage of society—Result of complex processes—Elaborate tribal rules—Laws altered deliberately: sometimes borrowed—Existing legislative methods of savages not primitive—The tribe a gradual conquest of culture—The tribe a combination of small pre-tribal kinships—History of progress towards the tribe traceable in surviving institutions—From passion to Law—Rudeness of native culture in Australia—Varieties of social organisation there—I. Tribes with two phratries, totems, female descent—Tribes of this organisation differ as to ceremonies and beliefs—Some beliefs tend to polytheism: others towards monotheism—Some tribes of pristine organisation have totemic magic and pirrauru: others have not—The more northern tribes of pristine organisation share the ceremonies and beliefs of central tribes: not so the south-eastern tribes—Second form (a) of social organisation has male descent—Second form (b) has female descent plus "matrimonial classes"—Account of these—Eight-class system—The Arunta nation—Their peculiar form of belief in reincarnation—Churinga nanja—Recapitulation—The Euahlayi tribe.

The question of the origin of totemism has more than the merely curious or antiquarian interest of an historic or prehistoric mystery. In the course of the inquiry we may be able to discern and discriminate the relative contributions of unreflecting passion, on one hand, and of deliberate reason, on the other, to the structure of the earliest extant form of human society. That form is the savage local tribe, as known to us in America and in Australia.

Men live in united local communities, relatively large, and carefully regimented, before they have learned to domesticate animals, or to obey chiefs, or to practise the rudest form of agriculture, or to fashion clay into pottery, or to build permanent hovels. Customary law is older than any of these things, and the most ancient law which we can observe unites a tribe by that system of marriages which expresses itself in totemism.

It is plain that the processes of evolution which have resulted in the most backward societies known to us, must have been very complex. If we reflect that the society of the Australian aborigines presents the institution of local tribes, each living peacefully, except for occasional internal squabbles, in a large definite tract of country; cultivating, on the whole, friendly relations with similar and similarly organised tribes; while obeying a most elaborate system of rules, it is obvious that these social conditions must be very remote from the absolutely primitive.[1] The rules of these tribes regulate every detail of private life with a minuteness and a rigour that remind us of what the Scottish Cavalier (1652) protested against as "the bloody and barbarous inconveniences of Presbyterial Government." Yet the tribes have neither presbyters, nor priests, nor kings. Their body of customary law, so copious and complex that, to the European, it seems as puzzling as algebra is to the savage, has been evolved, after a certain early point, by the slow secular action of "collective wisdom." We shall find that on this point, early deliberate modification of law, there can be no doubt.

The recent personal researches of Mr. Howitt and Messrs. Spencer and Gillen make it certain that tribal affairs, now, among many tribes at least, are discussed with the utmost deliberation, and that modifications of institutions may be canvassed, adopted, or rejected, on the initiative of seniors, local "Headmen," and medicine men.[2] It is also certain that tribe borrows from tribe, in the matter of songs, dances, and institutions, while members of one tribe are permitted to be present at the sacred ceremonials of others, especially when these tribes are on intermarrying terms.[3] In such cases, the ceremonials of one tribe may affect those of another, the Arunta may influence the Urabunna, who borrow their sacred objects or churinga for use in their own rites. We even hear of cases in which native religious ideas have been propagated by missionaries sent from tribe to tribe.[4]

Thus, conservative as is the savage by nature, he is distinctly capable of deliberate modification of his rites, ceremonies, and customary laws, and of interchanging ideas on these subjects with neighbouring tribes.

All this is true, to-day, and doubtless has long been true.

But at this point we must guard against what we consider a prevalent fallacy. The legislative action of the natives, the initiative of local Headmen, and Heads of Totems and of "Classes" (social divisions), and of medicine men inspired by "some supernatural being, such as Kutchi of the Dieri, Bunjil of the Wurunjerri, or Daramulun of the Coast Murring,"[5] is only rendered possible by the existence, to-day, of social conditions which cannot be primitive. To-day the Tribe, with its innumerable rules, and its common faith in Kutchi or Daramulun, with its recognised local or social Headmen, with its regulations for dealing with other tribes, and with its heralds or messengers, is an institution "in being." But, necessarily, this was not always so; the Tribe itself is a great "conquest of culture," and that conquest must have been made very slowly.

The prevalent fallacy, then, is to take unconsciously for granted that the people was, from the beginning, regimented into tribes, or existed in "hordes" already as capable as actual tribes of deliberative assemblies and legislative action, and that, in these hordes, a certain law, "the universal basis of their social system, was brought about by intention," as Mr. Howitt believes.[6]

The law in question, "the universal basis of their social system," was nothing less than a rule compelling people who had hitherto been promiscuous in their unions, to array themselves into a pair of tribal divisions, in which no member might marry another member of the same division, but must marry a member of the opposite division. The mere idea of such an act of legislation, for which no motive is assigned (and no motive is conceivable) postulates the pre-existence of a community like the Tribe of to-day, with powers to legislate, and to secure obedience for its legislative acts. This postulate cannot be granted, it refracts the institutions of to-day on a past state of society which, in all probability, could possess no such institutions. The "chaotic horde" of the hypothesis could not allot to various human groups the duty of working magic (to take an instance) for the good of various articles of the common food supply, nor could it establish a new and drastic rule, suddenly regulating sexual unions which had previously been utterly unregulated.

Human history does not show us a relatively large mass segregating itself into smaller communities. It shows us small communities aggregating into larger combinations, the village into the city, the European tribes into the kingdom, the kingdoms into the nation, the nation into the empire. The Tribe itself, in savage society, is a combination of small kins, or sets of persons of various degrees of status; these kins have not been legislatively segregated out of a pre-existing horde having powers of legislation. The idea of such a legislative primeval horde has been unconsciously borrowed from the actual Tribe of experience to-day.

That tribe is not primitive, far from it, but is very old.

Tribal collective wisdom, when once the tribe was evolved, has probably been at work, in unrecorded ages, over all the world, and in most places seems, up to a certain point, to have followed much the same strange course. The path does not march straight to any point predetermined by man, but loops, and zigzags, and retreats, and returns on itself, like the course of a river beset by rocks and shoals, and parcelled into wandering streams, and lagging in morasses. Yet the river reaches the sea, and the loops and links of the path, frayed by innumerable generations of early men, led at last to the haven of the civilised Family, and the Family Peace.

The history of the progress must necessarily be written in the strange characters of savage institutions, and in these odd and elaborate regulations which alarm the incurious mind under the names of "Phratries," "Totems," "Matrimonial Classes," "Pirrauru," and "Piraungaru." In these, as in some Maya or Easter Island inscription, graven in bizarre signs, lies the early social history of Man. We pore over the characters, turning them this way and that, deciphering a mark here and there, but unable to agree on any coherent rendering of the whole, so that some scholars deem the problems insoluble—and most are at odds among themselves.

Possibly we can at last present a coherent translation of the record which lies half concealed and half revealed in the savage institutions with their uncouth names, and can trace the course of an evolution which, beginning in natural passions, emotions, and superstitions, reached a rudimentary social law. That law, again, from a period far behind our historical knowledge, has been deliberately modified by men, much as a Bill in Parliament is modified by amendments and compromises into an Act. The industry of students who examine the customs of the remotest races has accumulated a body of evidence in which the various ways out of early totemic society towards the civilised conception of the family may be distinctly traced.

Meanwhile we are concerned rather with the way into totemism out of a prior non-totemic social condition, and with the development of the various stages of totemic society in Australia. The natives of that country, when unspoiled by European influences, are almost on one level as to material culture. Some tribes have rather better and more permanent shelters than others; some have less inadequate canoes than the rest; some drape themselves against cold weather in the skins of beasts, while others go bare; but all are non-agricultural hunting wanderers, without domesticated animals, without priests, and without chiefs on the level of those of the old Highland clans. They are ignorant of pottery, a fact which marks the very lowest culture; they know not the bow and arrow; their implements of stone vary from the polished "neolithic" to the rough-hewn "palæolithic" type: a man will use either sort as occasion serves.

While everyday life and its implements are thus rude, there are great varieties of social organisation, of ceremonial institutions, and of what, among Europeans, would be called speculative and religious ideas, expressing themselves in myths and rites.

Taking social organisation first, we begin with what all inquirers (except one or two who wrote before the recent great contributions to knowledge appeared) acknowledge to be the most pristine type extant Each tribe of this type is in two intermarrying divisions (which we call "exogamous moieties," or "phratries"), and each phratry bears a name which, when it can be translated, is, as a rule, that of an animal.[7] We shall show later why the meaning of the names has often been lost. Take the animal names of the phratries to be Emu and Kangaroo, no man of the Emu phratry may marry a woman of the same phratry, he must marry out of his phratry ("exogamy"); nor may a man of the Kangaroo phratry marry a woman of the same. Kangaroo phratry must marry into Emu, and Emu into Kangaroo. The phratry names in each case are, in the more primitive types of the organisation (which alone we are now considering) inherited from the mother.[8] A man of the Emu phratry marries a woman of the Kangaroo phratry, and to that phratry her children belong. Thus members of either phratry must be found in any casual knot or company of natives. Within each phratry there are, again, kinships also known by hereditary names of animals or plants. Thus, in Emu phratry, there may be kins called, say, Emu, Opossum, Wallaby, Grub, and others; in the Kangaroo phratry different names prevail, such as Kangaroos, Lizards, Dingoes, Cockatoos, and others. The name-giving animals, in this case, are called by us "totems," and the human kins which bear their names are called "totem kins." No man or woman may marry a person of his or her own totem. But this, in fact, as matters stand in Australia, puts no fresh bar on marriage, because (except in four or five tribes of the Centre) if a man marries out of his phratry he must necessarily marry out of his totem kin, since there are no members of his totem name in the phratry into which he must marry. In America, in cases where there are no phratries, and universally, where totems exist without phratries, marriage between persons of the same totem is forbidden.

The organisation of the more primitive tribes presents only the two exogamous moieties or phratries in each tribe and the totem kins in the phratries. We have Crow phratry and Eagle Hawk phratry, and, within Crow phratry, Crow totem kin,[9] with other totem kins; within Eagle Hawk phratry, Eagle Hawk totem kin, with other totem kins, which are never of the same names as those in Crow phratry.

This we call the primitive type, all the other organisations are the result of advances on and modifications of this organisation. It also occurs in America,[10] where, however, the phratry is seldom extant, though it does exist occasionally, and is known to have existed among the Iroquois and to have decayed.

On examining Mr. Howitt's map[11] it will be seen that this type of social organisation extends, or has extended, from Mount Gambier, by the sea, in the extreme south, past Lake Eyre, to some distance beyond Cooper's Creek or the Barcoo River, and even across the Diamantina River in Queensland. But it is far from being the case that all tribes with this pristine organisation possess identical ceremonies and ideas. On the other hand, from the southern borders of Lake Eyre, northwards, the tribes of this social organisation have peculiar ceremonies, unknown in the south and east, but usual further north and west. They initiate young men with the rites of circumcision or subincision (a cruel process unknown outside of Australia), or with both. In the south-east the knocking out of a front tooth takes the place of these bloody ordeals. The Lake Eyre tribes, again, do not, like those south and east of them, hold by, and inculcate at the rites, "the belief as to the existence of a great supernatural anthropomorphic Being, by whom the ceremonies were first instituted, and who still communicates with mankind through the medicine men, his servants."[12] Their myths rather repose on the idea of beings previous to man, "the prototypes of, but more powerful in magic than the native tribes. These beings, if they did not create man, at least perfected him from some unformed and scarcely human creatures."[13]

Thus, the more northern tribes of primitive tribal organisation (say the Dieri and their congeners) have beliefs which might ripen into the Greek mythology of gods and Titans, while the faith of the tribes of the same social organisation, further south by east, might develop into a rude form of Hebrew monotheism, and the two myths may co-exist, and often do. The northern tribes about Lake Eyre, and the central and north tribes, work co-operative magic for the behoof of their totem animals, as part of the common food supply, a rite unknown to the south and east. They also practise a custom (Pirrauru) of allotting men and women, married or unmarried, as paramours to each other, after a symbolic ceremony. This arrangement also is unknown in the south and east, and even north by west, though almost everywhere there is sexual licence at certain ceremonial meetings. It is thus plain that the more northern tribes of the primitive organisation described, differ from their southern and eastern neighbours (i.) in their most important initiatory rites, (ii.) in some of their myths or beliefs,[14] (iii.) in their totemic magic, and (iv.) in their allotment of permanent paramours. In the first three points these northern tribes of primitive type resemble, not the south-eastern tribes of the same social type, but the more socially advanced central, western, and northern "nations," with whom some of them are in touch and even intermarry. It is a dangerous fallacy to suppose that all tribes of the primitive tribal organisation are solidaires as to marriage, ceremonial rites, and beliefs.

It is difficult to say which is the second type of tribal organisation. We have in Victoria, in a triangle with its apex on the Murray River, the organisation already described (1), but here descent is reckoned in the male, not in the female line. This implies some social advance: social institutions, with male descent of the totem name, are certain to become local, rather than totemistic. The Kangaroos, deriving the totem name from the father, are a local clan, in some cases, like the MacIans in Glencoe. The Kangaroo name prevails in the locality. This cannot occur, obviously, when the names are derived from mothers, and the women go to the husband's district. We may call the organisation thus described (2a), and as (2b) we should reckon the organisation which prevails, as a rule, on the east of Southern Australia, in Queensland and New South Wales, from the northerly and southern coast-line (with a gap in the centre of the coast-line), to the eastern limits of (1). Here we find (2b) a great set of tribes having female descent, but each individual belongs not only to one of two phratries, and to a totem, but also to a "Matrimonial Class." In each phratry there are two such classes. Among the Kamilaroi, in phratry Dilbi, are "classes" named Muri (male) and Kubi (male). In phratry Kupathin are Ipai (male) and Kumbo (male), while the women bear the feminine forms of these names. Their meaning is usually unknown, but in two or three tribes, where the meaning of the class names is known with certainty, they denote animals.

The arrangement works thus, a man of phratry Dilbi, and of matrimonial class Muri, may not marry any woman that he chooses, in the other phratry, Kupathin. He can only marry a Kubatha, that is, a female of the class Kumbo. Their children, female descent prevailing, are of Kupathin phratry, and of the mother's totem, but do not belong to the class either of father (Muri) or of mother (Kumbo). They must belong to the other class within her phratry, namely Ipai. This rule applies throughout; thus, if a man of phratry Dilbi, and of Kubi class, marries a woman of Ipai class in phratry Kupathin, their children are neither of class Kubi nor of class Ipai, but of class Kumbo, the linked or sister class of Ipai, in Kupathin phratry.

Suppose for the sake of argument that the class names denote, or once denoted animals, so that, say—

In phratry

While in phratry

It is obvious that male Turtle would marry female Cat, and (with maternal descent) their children would, by class name, be Carpet Snakes. Bat would marry Carpet Snake, and their children would, by class name, be Cats. Persons of each generation would thus belong to classes of different animal names for ever, and no one might marry into either his or her own phratry, his or her own totem, or his or her own generation, that is, into his or her own class. It is exactly (where the classes bear animal names) as if two generations had totems. The mothers of Muri class in Dilbi would have Turtle, the mothers in Kupathin (Ipai) would have Carpet Snake. Their children, in Kupathin, would have Cat. Not only the phratries and the totem kins, but each successive generation, would thus be delimited by bearing an animal name, and marriage would be forbidden between all persons not of different animal-named phratries, different animal-named totem kins, and different animal-named generations. In many cases, we repeat, the names of the phratries and of the classes have not yet been translated, and the meanings are unknown to the natives themselves. That the class names were originally animal names is a mere hypothesis, based on few examples.

Say I am of phratry Crow, of totem Lizard, of generation and matrimonial class Turtle; then I must marry only a woman of phratry Eagle Hawk, of any totem in Eagle Hawk phratry,[15] and of generation and class name Cat. Our children, with female descent, will be of phratry Eagle Hawk, of totem the mother's, and of generation and class name Carpet Snake. Their children will be of phratry Crow, of totem the mother's, and of generation and class name Cat again; and so on for ever. Each generation in a phratry has its class name, and may not marry within that name. The next generation has the other class name, and may not marry within that. Assuming that phratry names, totem names, and generation names are always names of animals (or of other objects in nature), the laws would amount, we repeat, simply to this: No person may marry another person who, by phratry, or totem, or generation, owns the same hereditary animal name or other name as himself or herself. Moreover no one may marry a person (where matrimonial classes exist) who bears the same class or generation name as his mother or father.

In practice the rules are thus quite simple, mistake is impossible—complicated as the arrangements look on paper. Where totem and phratry names only exist, a man has merely to ask a woman, "What is your phratry name?" If it is his own, an amour is forbidden. Where phratry names are obsolete, and classes exist, he has only to ask, "What is your class name?" If it is that of either class in his own phratry of the tribe, to love is to break a sacred law. It is not necessary, as a rule, even to ask the totem name. What looks so perplexing is in essence, and in practical working, of extreme simplicity. But some tribes have deliberately modified the rules, to facilitate marriage.

The conspicuous practical result of the Class arrangement (not primitive), is that just as totem law makes it impossible for a person to marry a sister or brother uterine, so Class law makes a marriage between father and daughter, mother and son, impossible.[16] But such marriages never occur in Australian tribes of pristine organisation (1) which have no class names, no collective names for successive generations. The origin of these class or generation names is a problem which will be discussed later.

Such is the Class system where it exists in tribes with female descent. It has often led to the loss and disappearance of the phratry names, which are forgotten, since the two sets of opposed class names do the phratry work.

We have next (3) the same arrangements with descent reckoned in the male line. This prevails on the south-east coast, from Hervey River to Warwick. In Gippsland, and in a section round Melbourne, there were "anomalous" arrangements which need not now detain us; the archaic systems tended to die out altogether.

All these south central (Dieri), southern, and eastern tribes may be studied in Mr. Howitt's book, already cited, which contains the result of forty years' work, the information being collected partly by personal research and partly through many correspondents. Mr. Howitt has viewed the initiatory ceremonies of more than one tribe, and is familiar with their inmost secrets.

For the tribes of the centre and north we must consult two books, the fruits of the personal researches of Mr. Baldwin Spencer, M.A., F.R.S., Professor of Biology in the University of Melbourne, and of Mr. F. J. Gillen, Sub-Protector of Aborigines, South Australia.[17] For many years Mr. Gillen has been in the confidence of the tribes, and he and Mr. Spencer have passed many months in the wilds, being admitted to view the most secret ceremonies, and being initiated into the myths of the people. Their photographs of natives are numerous and excellent.

These observers begin in the south centre, where Mr. Howitt leaves off in his northerly researches, and go north. They start with the Urabunna tribe, north-east of Lake Eyre, congeners of Mr. Howitt's Dieri, and speaking a dialect akin to theirs, while the tribe intermarry marry with the Arunta (whose own dialect has points in common with theirs) of the centre of the continent These Urabunna are apparently in the form of social organisation which we style primitive (No. 1), but there are said, rather vaguely, to be more restrictions on marriage than is usual, people of one totem in Kiraru phratry being restricted to people of one totem in Matteri phratry.[18]

They have phratries, totem kins, apparently no matrimonial classes (some of their rules are imperfectly ascertained), and they reckon descent in the female line. But, like the Dieri (and unlike the tribes of the south and east), they practise subincision; they have, or are said to have, no belief in "a supernatural anthropomorphic great Being"; they believe in "old semi-human ancestors," who scattered about spirits, which are perpetually reincarnated in new members of the tribe; they practise totemic magic; and they cultivate the Dieri custom of allotting paramours. Thus, by social organisation, they attach themselves to the south-eastern tribes (1), but, like the Dieri, and even more so (for, unlike the Dieri, they believe in reincarnation), they agree in ceremonies, and in the general idea of their totemic magic, rites, and mythical ideas, with tribes who, as regards social organisation, are in state (4), reckon descent in the male line, and possess, not four, but eight matrimonial classes.

This institution of eight classes is developing in the Arunta "nation," the people of the precise centre of Australia, who march with, and intermarry with, the Urabunna; at least the names for the second set of four matrimonial classes, making eight in all, are reaching the Arunta from the northern tribes. All the way further north to the Gulf of Carpentaria, male descent and eight classes prevail, with subincision, prolonged and complex ceremonials, the belief in reincarnation of primal semi-human, semi-bestial ancestors, and the absence (except in the Kaitish tribe, next the Arunta) of any known belief in what Mr. Howitt calls the "All Father." Totemic magic also is prevalent, dwindling as you approach the north-east coast. In consequence of reckoning in the male line (which necessarily causes most of the dwellers in a group to be of the same totem), local organisation is more advanced in these tribes than in the south and east.

We next speak of social organisation (5), namely, that of the Arunta and Kaitish tribes, which is without example in any other known totemic society all over the world. The Arunta and Kaitish not only believe, like most northern and western tribes, in the perpetual reincarnation of ancestral spirits, but they, and they alone, hold that each such spirit, during discarnate intervals, resides in, or is mainly attached to, a decorated kind of stone amulet, called churinga nanja. These objects, with this myth, are not recorded as existing among other "nations." When a child is born, its friends hunt for its ancestral stone amulet in the place where its mother thinks that she conceived it, and around the nearest rendezvous of discarnate local totemic souls, all of one totem only. The amulet and the local totemic centre, with its haunted nanja rock or tree, determine the totem of the child. Thus, unlike all other totemists, the Arunta do not inherit their totems either from father or mother, or both. Totems are determined by local accident. Not being hereditary, they are not exogamous: here, and here alone, they do not regulate marriage. Men may, and do, marry women of their own totem, and their child's totem may neither be that of its father nor of its mother. The members of totem groups are really members of societies, which co-operatively work magic for the good of the totems. The question arises, Is this the primitive form of totemism? We shall later discuss that question (Chapter IV.).

Meanwhile we conceive the various types of social organisation to begin with the south-eastern phratries, totems, and female reckoning of descent (1) to advance to these plus male descent (2a), and to these with female descent and four matrimonial classes (2b). Next we place (3) that four-class system with male descent; next (4) the north-western system of male descent with eight matrimonial classes, and last (as anomalous in some respects), (5) the Arunta-Kaitish system of male descent, eight classes, and non-hereditary non-exogamous totems.

As regards ceremonial and belief, we place (1) the tribes south and east of the Dieri. (2) The Dieri. (3) The Urabunna, and north, central, and western tribes. (4) The Arunta. The Dieri and Urabunna we regard (at least the Dieri) as pristine in social organisation, with peculiarities all their own, but in ceremonial and belief more closely attached to the central, north, and west than to the south-eastern tribes. As concerns the bloody rites, Mr. Howitt inclines to the belief (corroborated by legends, whatever their value) that "a northern origin must ultimately be assigned to these ceremonies."[19] It is natural to assume that the more cruel initiatory rites are the more archaic, and that the tribes which practise them are the more pristine. But this is not our opinion nor that of Messrs. Spencer and Gillen. The older rite is the mere knocking out of front teeth (also used by the Masai of East Central Africa). This rite, in Central Australia, "has lost its old meaning, its place has been taken by other rites."[20] ... Increased cruelty accompanies social advance in this instance. In another matter innovation comes from the north. Messrs. Spencer and Gillen are of the opinion that "changes in totemic matters have been slowly passing down from north to south." The eight classes, in place of four classes, are known as a matter of fact to have actually "reached the Arunta from the north, and at the present moment are spreading south-wards."[21]

Again, a feebler form of the reincarnation belief, namely, that souls of the young who die uninitiated are reincarnated, occurs in the Euahlayi tribe of north-western New South Wales.[22] Whether the Euahlayi belief came from the north, in a limited way, or whether it is the germinal state of the northern belief, is uncertain. It is plain that if bloody rites and eight classes may come down from the north, totemic magic and the faith in reincarnation may also have done so, and thus modified the rites and "religious" opinions of the Dieri and Urabunna, who are said still to be, socially, in the most pristine state, that of phratries and female descent, without matrimonial classes.[23] It is also obvious that if the Kaitish faith in a sky-dweller (rare in northern tribes) be a "sport," and if the Arunta churinga nanja, plus non-hereditary and non-exogamous totems, be a "sport," the Dieri and Urabunna custom, too, of solemnly allotted permanent paramours may be a thing of isolated and special development, not a survival of an age of "group marriage."

[1] Howitt, Native Tribes of South-East Australia, p. 41. 1904.

[2] Cf. for example Spencer and Gillen, Northern Tribes of Central Australia, p. 26. Howitt, Native Tribes of South-East Australia, pp. 88, 89.

[3] Howitt, ut supra, pp. 511, 513.

[4] Hale, U.S. Exploring Expedition, p. 410. 1846.

[5] Howitt, ut supra, p. 89.

[6] Op. cit., p. 89.

[7] There are exceptions, or at least one exception is known to the rule of animal names for phratries, a point to which we shall return. Dr. Roth (N.W. Central Queensland Aborigines, p. 56) suggests that the phratry names Wutaru and Pakuta mean One and Two (cf. p. 26). For Wutaru and Yungaru, however, interpretations indicating names of animals are given, diversely, by Mr. Bridgman and Mr. Chatfield, Kamilaroi and Kurnai, pp. 40, 41.

[8] That reckoning descent in the female line, among totemists, is earlier than reckoning in the male line, Mr. Howitt, Mr. Tylor, Dr. Durkheim, and Messrs. Spencer and Gillen, with Mr. J. G. Frazer, till recently, are agreed. Starcke says "usually the female line only appears in connection with the Kobong (totem) groups," and he holds the eccentric opinion that totems are relatively late, and that the tribes with none are the more primitive! (The Primitive Family, p. 26, 1896.) This writer calls Mr. Howitt "a missionary."

[9] That this is the case will be proved later; the fact has hitherto escaped observation.

[10] Frazer, Totemism, p. 6l. Morgan, Ancient Society, pp. 90, 94 et seq.

[11] Native Tribes of South-East Australia. Macmillan, 1904.

[12] Native Tribes of South-East Australia, p. 640. For examples, pp. 528-535.

[13] Ibid., p. 487.

[14] That is, on our present information. It is very unusual for orthodox adhesion to one set of myths to prevail.

[15] Sometimes members of one totem are said to be restricted to marriage with members of only one other totem.

[16] Howitt, Native Tribes of South-East Australia, p. 284, citing Mr. J. G. Frazer.

[17] Native Tribes of Central Australia, 1899. Northern Tribes of Central Australia, 1904. Macmillan.

[18] Cf. Howitt, Native Tribes of South-East Australia, pp. 188-189. Native Tribes of Central Australia, p. 60.

[19] Howitt, op. cit., p. 676, N.T., p. 20.

[20] Native Tribes of Central Australia, p. 214. The same opinion is stated as very probable in Northern Tribes of Central Australia, p. 329.

[21] N. T., p. 20.

[22] Mrs. Langloh Parker's M.S.

[23] I am uncertain as to this point among the Urabunna, as will appear later.

[CHAPTER II]

METHOD OF INQUIRY

Method of inquiry—Errors to be avoided—Origin of totemism not to be looked for among the "sports" of socially advanced tribes—Nor among tribes of male reckoning of descent—Nor in the myths explanatory of origin of totemism—Myths of origin of heraldic bearings compared—Tribes in state of ancestor-worship: their totemic myths cannot be true—Case of Bantu myths (African)—Their myth implies ancestor-worship —Another African myth derives tribal totems from tribal nicknames—No totemic myths are of any historic value—The use of conjecture—Every theory must start from conjecture—Two possible conjectures as to earliest men gregarious (the horde), or lonely sire, female mates, and off-spring—Five possible conjectures as to the animal names of kinships in relation to early society and exogamy—Theory of the author; of Professor Spencer; of Dr. Durkheim; of Mr. Hill-Tout; of Mr. Howitt—Note on McLennan's theory of exogamy.

We have now given the essential facts in the problem of early society as it exists in various forms among the most isolated and pristine peoples extant. It has been shown that the sets of seniority (classes), the exogamous moieties (phratries), and the kinships in each tribe bear names which, when translated, are usually found to denote animals. Especially the names of the totem kindreds, and of the totems, are commonly names of animals or plants. If we can discover why this is so, we are near the discovery of the origin of totemism. Meanwhile we offer some remarks as to the method to be pursued in the search for a theory which will colligate all the facts in the case, and explain the origin of totemic society. In the first place certain needful warnings must be given, certain reefs which usually wreck efforts to construct a satisfactory hypothesis must be marked.

First, it will be vain to look for the origin of totemism either among advanced and therefore non-pristine Australian types of tribal organisation, or among peoples not Australian, who are infinitely more forward than the Australians in the arts of life, and in the possession of property. Such progressive peoples may present many interesting social phenomena, but, as regards pure primitive totemism, they dwell on "fragments of a broken world." The totemic fragments, among them, are twisted and shattered strata, with fantastic features which cannot be primordial, but are metamorphic. Accounts of these societies are often puzzling, and the strange confused terms used by the reporters, especially in America, frequently make them unintelligible.

The learned, who are curious in these matters, would have saved themselves much time and labour had they kept two conspicuous facts before their eyes.

(1) It is useless to look for the origins of totemism among the peculiarities and "sports" which always attend the decadence of totemism, consequent on the change from female to male lineage, as Mr. Howitt, our leader in these researches, has always insisted. To search for the beginnings among late and abnormal phenomena, things isolated, done in a corner, and not found among the tribal organisations of the earliest types, is to follow a trail sure to be misleading.

(2) The second warning is to be inferred from the first. It is waste of time to seek for the origin of totemism in anything—an animal name, a sacred animal, a paternal soul tenanting an animal—which is inherited from its first owner, he being an individual ancestor male. Such inheritance implies the existence of reckoning descent in the male line, and totemism conspicuously began in, and is least contaminated in, tribes who reckon descent in the female line.

Another stone of stumbling comes from the same logical formation. The error is, to look for origins in myths about origins, told among advanced or early societies. If a people has advanced far in material culture, if it is agricultural, breeds cattle, and works the metals, of course it cannot be primitive. However, it may retain vestiges of totemism, and, if it does, it will explain them by a story, a myth of its own, just as modern families, and even cities, have their myths to account for the origin, now forgotten, of their armorial bearings, or crests—the dagger in the city shield, the skene of the Skenes, the sawn tree of the Hamiltons, the lyon of the Stuarts.

Now an agricultural, metallurgic people, with male descent, in the middle barbarism, will explain its survivals of totemism by a myth natural in its intellectual and social condition; but not natural in the condition of the homeless nomad hunters, among whom totemism arose. For example, we have no reason to suspect that when totemism began men had a highly developed religion of ancestor-worship. Such a religion has not yet been evolved in Australia, where the names of the dead are usually tabooed, where there is hardly a trace of prayers, hardly a trace of offerings to the dead, and none of offerings to animals.[1] The more pristine Australians, therefore, do not explain their totems as containing the souls of ancestral spirits. On the other hand, when the Bantu tribes of Southern Africa—agricultural, with settled villages, with kings, and with many of the crafts, such as metallurgy—explain the origin of their tribal names derived from animals on the lines of their religion—ancestor-worship—their explanation may be neglected as far as our present purpose is concerned. It is only their theory, only the myth which, in their intellectual and religious condition, they are bound to tell, and it can throw no light on the origin of sacred animals.

The Bantu local tribes, according to Mr. M'Call Theal, have Siboko, that is, name-giving animals. The tribesmen will not kill, or eat, or touch, "or in any way come into contact with" their Siboko, if they can avoid doing so. A man, asked "What do you dance?" replies by giving the name of his Siboko, which is, or once was, honoured in mystic or magical dances.

"When a division of a tribe took place, each section retained the same ancestral animal," and men thus trace dispersed segments of their tribe, or they thus account for the existence of other tribes of the same Siboko as themselves.

Things being in this condition, an ancestor-worshipping people has to explain the circumstances by a myth. Being an ancestor-worshipping people, the Bantu explain the circumstance, as they were certain to do, by a myth of ancestral spirits. "Each tribe regarded some particular animal as the one selected by the ghosts of its kindred, and therefore looked upon it as sacred."

It should be superfluous to say that the Bantu myth cannot possibly throw any tight on the real origin of totemism. The Bantu, ancestor-worshippers of great piety, find themselves saddled with sacred tribal Siboko; why, they know not. So they naturally invent the fable that the Siboko, which are sacred, are sacred because they are the shrines of what to them are really sacred, namely, ancestral spirits.[2] But they also cherish another totally different myth to explain their Siboko.

We now give this South African myth, which explains tribal Siboko, and their origin, not on the lines of ancestor-worship, but, rather to my annoyance, on the lines of my own theory of the Origin of Totems!

On December 9, 1879, the Rev. Roger Price, of Mole-pole, in the northern Bakuena country, wrote as follows to Mr. W. G. Stow, Geological Survey, South Africa. He gives the myth which is told to account for the Siboko or tribal sacred and name-giving animal of the Bahurutshe—Baboons. (These animal names in this part of Africa denote local tribes, not totem kins within a local tribe.)

"Tradition says that about the time the separation took place between the Bahurutshe and the Bakuena, Baboons entered the gardens of the Bahurutshe and ate their pumpkins, before the proper time for commencing to eat the fruits of the new year. The Bahurutshe were unwilling that the pumpkins which the baboons had broken off and nibbled should be wasted, and ate them accordingly. This act is said to have led to the Bahurutshe being called Buchwene, Baboon people—which" (namely, the Baboon) "is their Siboko to this day—and their having the precedence ever afterwards in the matter of taking the first bite of the new year's fruits. If this be the true explanation," adds Mr. Price, "it is evident that what is now used as a term of honour was once a term of reproach. The Bakuena, too, are said to owe their Siboko (the Crocodile) to the fact that their people once ate an ox which had been killed by a crocodile."

Mr. Price, therefore, is strongly inclined to think "that the Siboko of all the tribes was originally a kind of nickname or term of reproach, but," he adds, "there is a good deal of mystery about the whole thing."

On this point Mr. Stow, to whom Mr. Price wrote the letter just cited, remarks in his MS.: "From the foregoing facts it would seem possible that the origin of the Siboko among these tribes arose from some sobriquet that had been given to them, and that, in course of time, as their superstitious and devotional feelings became more developed, these tribal symbols became objects of veneration and superstitious awe, whose favour was to be propitiated or malign influence averted...."[3]

Here it will be seen that these South African tribes account for their Siboko now by the myth deriving the sacredness of the tribal animal from ancestor-worship, as reported by Mr. Theal, and again by nicknames given to the tribes on account of certain undignified incidents.

This latter theory is very like my own as stated in Social Origins, and to be set forth and reinforced later in this work. But the theory, as held by the Bahurutsche and Bakuena, does not help to confirm mine in the slightest degree. Among these very advanced African tribes, the Siboko or tribal sacred animal, is the animal of the local tribe, not, as in pure totemism, of the scattered exogamous kin. It is probably a lingering remnant of totemism. The totem of the most powerful local group in a tribe having descent through males, appears to have become the Siboko of the whole tribe, while the other totems have died out. It is not probable that a nickname of remembered origin, given in recent times to a tribe of relatively advanced civilisation, should, as the myth asserts, not only have become a name of honour, but should have founded tribal animal-worship.

It was in a low state of culture no longer found on earth, that I conceive the animal names of groups not yet totemic, names of origin no longer remembered, to have arisen and become the germ of totemism.

Myths of the origin of totemism, in short, are of absolutely no historic value. Siboko no longer arise in the manner postulated by these African myths; these myths are not based on experience any more than is the Tsimshian myth of the Bear Totem, to be criticised later in a chapter on American Totemism. We are to be on our guard, then, against looking for the origins of totemism among the myths of peoples of relatively advanced culture, such as the village-dwelling Indians of the north-west coast of America. We must not look for origins among tribes, even if otherwise pristine, who reckon by male descent. We must look on all savage myths of origins merely as savage hypotheses, which, in fact, usually agree with one or other of our scientific modern hypotheses, but yield them no corroboration.

On the common fallacy of regarding the tribe of to-day, with its relative powers, as primitive, we have spoken in Chapter I.

By the nature of the case, as the origin of totemism lies far beyond our powers of historical examination or of experiment, we must have recourse as regards this matter to conjecture.

Here a word might be said as to the method of conjecture about institutions of which the origins are concealed "in the dark backward and abysm of time."

There are conjectures and conjectures! None is capable in every detail of historical demonstration, but one guess may explain all the known facts, and others may explain few or none. We are dealing with human affairs—they whose groups first answered to animal group-names were men as much as we are. They had reason; they had human language, spoken or by gesture, and human passions. That conjecture, therefore, which deals with the first totemists as men, men with plenty of human nature, is better than any rival guess which runs contrary to human nature as known in our experience of man, savage, barbaric, or civilised.

Once more, a set of guesses which are consistent with themselves is better than a set of guesses which can be shown to be even ludicrously self-contradictory. If any guess, again, colligates all the known facts, if any conjectural system will "march," will meet every known circumstance in the face, manifestly it is a better system than one which stumbles, breaks down, evades giving an answer to the problems, says that they are insoluble, is in contradiction with itself, and does not even try to colligate all the known facts. A consistent system, unmarred by self-contradictions; in accordance with known human nature; in accordance, too, with recognised rules of evolution, and of logic; and co-ordinating all known facts, if it is tried on them, cannot be dismissed with the remark that "there are plenty of other possible guesses."

Our method must be—having already stated the facts as they present themselves in the most primitive organisation of the most archaic society extant—to enumerate all the possible conjectures which have been logically (or even illogically) made as to the origin of the institutions before us.

All theories as to how these institutions arose, must rest, primarily, on a basis of conjecture as to the original social character of man. Nowhere do we see absolutely primitive man, and a totemic system in the making. The processes of evolution must have been very gradually developed in the course of distant ages, but our conjecture as to the nature, in each case, of the processes must be in accordance with what is known of human nature. Conjecture, too, has its logical limitations.

We must first make our choice, therefore, between the guess that the earliest human beings lived in very small groups (as, in everyday life, the natives of Australia are in many cases still compelled to do by the precarious nature of their food supplies), or the guess that earliest man was gregarious, and dwelt in a promiscuous horde with no sort of restraint. One or other view must be correct.

On the former guess (men originally lived in very small groups), the probable mutual hostility of group to rival group, the authority of the strongest male in each group, and the passions of jealousy, love, and hate, must inevitably have produced some rudimentary restrictions on absolute archaic freedom. Some people would be prevented from doing some things, they must have been checked by the hand of the stronger; and from the habit of restraint customary rules would arise. The advocates of the alternative conjecture—that man was gregarious, and utterly promiscuous—take it for granted (it seems to me) that the older and stronger males established no rudimentary restrictions on the freedom of the affections, but allowed the young males to share with them the females in the horde, and that they permitted both sexes to go entirely as they pleased, till, for some unknown reason and by some unknown authority, the horde was bisected into exogamous moieties (phratries), and after somehow developing totem kins (unless animal-named magical groups had been previously developed, on purpose to work magic), became a tribe with two phratries.

It is not even necessary for us to deny that the ancestors of man were originally communal and gregarious. What we deem to be impossible is that, till man had developed into something more like himself, as we know him, than an animal without jealousy, and ignorant of anything prejudicial to any one's interests in promiscuous unions, he could begin to evolve his actual tribal institutions. This is also the opinion of Mr. Howitt, as we shall see later.

Thus whoever tries to disengage the evolutionary processes which produced the existing society of Australia must commence by making his choice between the two conjectures—early man gregarious, promiscuous, and anarchist; or early man unsociable, fierce, bullying, and jealous. A via media is attempted, however, by Mr. Howitt, to which we shall return.

Next, it is clear and certain that some human beliefs about the animals which give their names, in known cases, to the two large exogamous divisions of the tribe (phratries), and about the other animals which give names to the totem kins, and, in one or two cases, to the matrimonial classes, must be, in some way, connected with the prohibitions to marry, first within the phratries, then, perhaps, within the totem kins, then within the Classes (or within the same generation).

Thus there are here five courses which conjecture can logically take.

(a) Members of certain recognised human groups already married habitually out of their group into other groups, before the animal names (now totem names) were given to the groups. The names came later and merely marked, at first, and then sanctioned, the limits within which marriage had already been forbidden while the groups were still nameless.

Or (b) the animal names of the phratries and totem kins existed (perhaps as denoting groups which worked magic for the behoof of each animal) before marriage was forbidden within their limits. Later, for some reason, prohibitions were enacted.

Or (c) at one time there were no marriage regulations at all, but these arose when, apparently for some religious reason, a hitherto undivided communal horde split into two sections, each of which revered a different name-giving animal as their "god" (totem), claimed descent from it, and, out of respect to their "god," did not marry any of those who professed its faith, and were called by its name, but always married persons of another name and "god."

Or (d) men were at first in groups, intermarrying within the group. These groups received names from animals and other objects, because individual men adopted animal "familiars," as Bear, Elk, Duck, Potato, Pine-tree. The sisters of the men next adopted these animal or vegetable "familiars," or protective creatures, from their brothers, and bequeathed them, by female descent, to their children. These children became groups bearing such names as Bear, Potato, Duck, and so on. These groups made treaties of marriage with each other, for political reasons of acquiring strength by union. The treaties declared that Duck should never marry Duck, but always Elk, and vice versa. This was exogamy, instituted for political purposes, to use the word "political" proleptically.

Or (e) men were at first in a promiscuous incestuous horde, but, perceiving the evils of this condition (whatever these evils might be taken to be), they divided it into two halves, of which one must never marry within itself, but always in the other. To these divisions animal names were given; they are the phratries. They threw off colonies, or accepted other groups, which took new animal names, and are now the totem kins.

Finally, in (f) conjectures (a) and (c) may be combined thus: groups of men, still nameless as groups, had for certain reasons the habit of not marrying within themselves; but, after receiving animal names, they developed an idea that the animal of each group was its kinsman, and that, for a certain superstitious reason, it was even more wrong than it had been before, to marry "within the blood" of the animal, as, for Emu to marry Emu. Or (f2) the small groups did marry within themselves till, after receiving animal names, they evolved the superstition that such marriage was a sin against the animals, and so became exogamous.

On the point of the original state of society conjecture seems to be limited to this field of possible choices. At least I am acquainted with no theory hitherto propounded, which does not set out from one or other of these conjectural bases. We must not attack each other's ideas merely because they start from conjectures: they can start in no other way. Our method must be to discover which conjecture, as it is developed, most consistently and successfully colligates all the ascertained facts and best endures the touchstone of logic.

Of the hypotheses enumerated above, the system to be advocated here is that marked (f 1 and 2). Men, whatever their brutal ancestors may have done, when they became men indeed, lived originally in small anonymous local groups, and had, for a reason to be given, the habit of selecting female mates from groups not their own. Or, if they had not this habit they developed the rule, after the previously anonymous local groups had received animal names, and after the name-giving animals came to receive the measure of respect at present given to them as totems.

The second hypothesis (b) (that the animal names of the groups were originally those of societies which worked magic, each for an animal, and that the prohibition on marriage was later introduced) has been suggested by Professor Baldwin Spencer and Mr. J. G. Frazer, and is accepted by Mr. Howitt.

The third conjecture (c) (man originally promiscuous, but ceasing to be so from religious respect for the totem, or "god") is that of Dr. Durkheim.

The fourth theory (d) is that of Mr. Hill-Tout.[4]

The fifth theory (e) was that of Mr. Howitt. He now adopts the similar theory of Mr. Spencer (b).

[1] The Dieri tribe do pray to the Mura-Mura, or mythical ancestors, but not, apparently, to the remembered dead.

[2] "Totemism, South Africa," J. G. Frazer, Man, 1901, No. III. Mr. Frazer does not, of course, adopt the Bantu myth as settling the question.

[3] Stow, MSS., 820. I owe the extract to Miss C. G. Burne.

[4] I have not included the theory of Dr. Westermarck, in the History of Human Marriage, because that work is written without any reference to totemism.

NOTE

I have not included the theory of Mr. J. F. McLennan, the founder of all research into totemism. In his opinion, totemism, that is, the possession by different stocks of different name-giving animals, "is older than exogamy in all cases." That is, as Mr. Robertson Smith explains, "it is easy to see that exogamy necessarily presupposes the existence of a system of kinship which took no account of degrees, but only of participation in a common stock. Such an idea as this could not be conceived by savages in an abstract form; it must necessarily have had a concrete expression, or rather must have been thought under a concrete and tangible form, and that form seems to have been always supplied by totemism." (Kinship and Marriage in Early Arabia, p. 189, 1885). This means that, before they were exogamous, men existed in groups of animal name, as Ravens, Wolves, Ants, and so on. When they became conscious of kinship, and resolved to marry out of the kin, or stock, they fixed the name, say Raven, Wolf, or what not, as the limit within which there must be no marriage. But Mr. McLennan's theory as to why they determined to take no wives within the stock and name, has never been accepted. (See Westermarck, History of Human Marriage, pp. 311-314.)

Mr. McLennan supposed that female infanticide made women scarce in each group, and that therefore they stole each other's girls, and, finally, abstained from their own. But the objections to this hypothesis are infinite and obvious. At one time Mr. McLennan thought that tattooing was the origin of totemism. Members of each group tattooed the semblance of an animal on their flesh—but, as far as I am aware, he did not ask why they adopted this practice. Manifestly a sense of some special connection between the animal and the group must have been prior to the marking of the members of the group with the effigy of the animal. What gave rise to this belief in the connection? (See Chapter VI., criticism of Dr. Pikler). Mr, McLennan merely mentioned to me, in conversation, this idea, which he later abandoned. It had previously occurred to Garcilasso de la Vega that the germ of totemism was to be found in the mere desire to differentiate group from group; which is the theory to be urged later, the names being the instruments of differentiation.

Mr. A. K. Keane, as in Mr. McLennan's abandoned conjecture, and as in the theory of Dr. Pikler, makes totemism arise in "heraldic badges," "a mere device for distinguishing one individual from another, one family or clan group from another ... the personal or family name precedes the totem, which grows out of it." (Ethnology, pp. 9, II).

[CHAPTER III]

THEORY OF PRIMAL PROMISCUITY

Why did man, if once promiscuous, regulate the relations of the sexes?—Theory of Professor Spencer—Animal-named magical societies were prior to regulation of marriage—Theory of Mr. Howitt—Regulations introduced by inspired medicine man—His motives unknown—The theory postulates the pristine existence of the organised tribe of to-day, and of belief in the All Father—Reasons for holding that men were originally promiscuous: (1) So-called survival of so-called "group marriage"; (2) Inclusive names of human relationships—Betrothals not denied—A form of marriage—Mitigated by Pirauru—Allotment of paramours at feasts—Is Pirauru a survival of group marriage?—Or a rare case of limitation of custom of feasts of license—Examples of such saturnalia—Fiji, Arunta, Urabunna, Dieri—Degrees of license—Argument against the author's opinion—Laws of incest older than marriage—Names of relationships—Indicate tribal status, not degrees of consanguinity—Fallacy exposed—Starcke versus Morgan's theory of primal promiscuity—Dr. Durkheim on Choctaw names of relationships—A man cannot regard his second cousin as his mother—Dr. Fison on anomalous terms of relationship—Grandfathers and grandsons call each other "brothers"—Noa denotes a man's wife and also all women whom he might legally wed—Proof that terms of relationship do not denote consanguinity—The Pirrauru custom implies previous marriage, and is not logically thinkable without it—Descriptions of Pirrauru—The Kandri ceremony merely modifies pre-existing marriage—Pirrauru is not "group marriage"—Is found only in tribes of the Matteri Kiraru phratries—Not found in south-eastern tribes—Mr. Howitt's "survivals" do not mean "group marriage."

In the theories which postulate that man began in a communal horde, with no idea of regulating sexual unions at all—because, having no notion of consanguinity, or of harm in consanguine marriages, he saw nothing to regulate—the initial difficulty is, how did he ever come to change his nature and to see that a rule must be made, as made it has been? Mr. Howitt endeavours (if I grasp his meaning) to show how man did at last see it, and therefore bisected the horde into intermarrying phratries. Mr. Spencer has only asserted that, while man saw nothing to regulate in marriages, he evolved an organisation, that of the phratries and classes, which did come, somehow, to regulate them. Dr. Durkheim takes it, that man if he was originally promiscuous, later regulated marriages out of respect to his totems, which were his gods. Mr. Hill-Tout supposes that the exogamous rules were made for "political" reasons.

The theories of Mr. Howitt and Mr. Spencer differed from each other, originally, only in so far as that Mr. Spencer supposes animal-named magical societies (now totemic) to have arisen before man regulated marriage in any way; whereas this conception of animal-named groups not bound by totemic restrictions on marriage had not occurred to Mr. Howitt or any other inquirer, except Mr. J. G. Frazer, who evolved it independently. Mr. Spencer's theory in this matter rests entirely on his discovery, among the Arunta, in Central Australia, of totems marking magical societies, but not regulating marriage, and on his inference that, in the beginning, animal-named groups were everywhere mere magical societies. To work co-operative magic was their primary function. To that opinion Mr. Howitt has now come in, and he adds that "the division of the tribe" (into the two primary exogamous moieties or phratries, or "classes") "was made with intent to regulate the relations of the sexes."[1] On one point, we repeat, namely, why division was made, Mr. Spencer utters no certain sound, nor does Mr. Howitt explicitly tell us for what reason sexual relations, hitherto unregulated, were supposed to need regulation. He conceives that there is "a widespread belief in the supernatural origin of the practice," but that explains nothing.[2]

Thus Mr. Howitt postulates the existence of a "tribe," divided into animal-named magical societies, and promiscuous. The tribe has "medicine men" who see visions. One of these men, conceiving, no one knows why, that it would be an excellent thing to regulate the relations of the sexes, announces to his fellow-men that he has received from a supernatural being a command to do so. If they approve, they declare the supernatural message "to the assembled headmen at one of the ceremonial meetings," the tribe obeys, and divides itself into the two primary exogamous moieties or phratries.[3] Mr. Howitt thus postulates the existence of the organised tribe, with its prophets, its "All Father" (such as Daramulun), its magical societies, its recognised headmen, and its public meetings for ceremonial and legislation, all in full swing, before the relations of the sexes are in any way regulated.

On reflection, Mr. Howitt may find difficulties in this postulate. Meanwhile, we ask what made the very original medicine man, the Moses of the tribe, think of the new and drastic command which he brought down from the local Sinai? Why did this thinker suppose that the relations of the sexes ought to be regulated? Perhaps the idea was the inspiration of a dream. Mr. Spencer, acquainted chiefly with tribes who have no All Father, has not advanced this theory.

The reasons given for supposing that the "tribe" was originally promiscuous are partly based (a) on the actual condition as regards individual marriage of some Australian tribes, mainly Dieri and Urabunna, with their congeners. These tribes, it is argued, are now no longer absolutely promiscuous, but men and women are divided into intermarriageable sets, so that all women of a certain status in Emu phratry are, or their predecessors have been, actual wives of all men of the corresponding status in Kangaroo phratry. The only bar to absolute promiscuity is that of the phratries (established by legislation on this theory), and of certain by-laws, of relatively recent institution. The names for human relationships (father, mother, son, daughter, brother, sister), again, (b) are, it is argued, such as "group marriage," and "group marriage" alone, would inevitably produce. All women of a certain status are my "mothers," all men of a certain status are my "fathers," all women of another status are my "sisters," all of another are my "wives," and so on. Thus Mr. Spencer is able to say that "individual marriage does not exist either in name or in practice in the Urabunna tribe" at the present day.[4]

This, however, does not mean that among many such tribes a man is not betrothed to a special woman, and does not marry that woman, with certain filthy initiatory "rites," contravening the usual rules of intercourse.[5] Nor is it denied that such man and wife habitually cohabit, and that the man, by hunting and fishing, provides for the wife and all her children, and recognises them as his own.

It is meant that each man has only a certain set of nubile women open to him (Nupa, or Noa, or Unawa), and that out of these, in addition to his allotted bride, an uncertain number of women are assigned to him and to others, mainly at tribal festivals, as paramours (Pirauru or Piraungaru), by their elder brothers, or the heads of totem kins, or the seniors of the Urabunna tribe. "This relationship is usually established at times when considerable numbers of the tribe are gathered together to perform important ceremonies."[6] One woman may, on different occasions, be allotted as Piraungaru to different men, one man to different women. Occasionally, though rarely, the regular husband (he who marries the wife by filthy "rites") resists the allotting of his wife to another man, and then "there is a fight."

The question is, does this Urabunna custom of Piraungaru (the existence of which in some tribes is not denied) represent a survival of a primary stage in which all men of a certain social and phratriac status were all alike husbands to all women of the corresponding status (group, or rather status, marriage); and was that, in turn, a survival of the anarchy of the horde, in which there were no grades at all, but anarchic promiscuity?

That is the opinion of believers in "the primary undivided horde," and in "group marriage," or rather "status marriage."

Or is this Piraungaru custom, as we think more probable, an organised and circumscribed and isolated legalisation, among a few tribes, of the utterly unbridled license practised by many savages on festive occasions corresponding to the Persian feast of the Sacaea, and to the Roman Saturnalia?[7]

The Piraungaru allotments are made, as a rule, at great licentious meetings, but among the Urabunna, though they break the rules of individual marriage, they do not break the tribal rules of incest. By these rules the Piraungaru men and women must be legal intermarriageable persons (Nupa); their regulated paramourship is not, by tribal law, what we, or the natives, deem "incestuous." On the other hand, at Fijian seasons of license, even the relationship of brother and sister—the most sacred of all to a savage—is purposely profaned. Brothers and sisters are "intentionally coupled" at the feast of license called Nanga. The object is to have "a regular burst," and deliberately violate every law. Men and women "publicly practised unmentionable abominations."[8]

The Fijians are infinitely above the Urabunna in civilisation, being an agricultural people. Their Nanga feast is also called Mbaki—"harvest" Yet the Fijians, though more civilised, far exceed the license of the Piraungaru custom of the Urabunna, not only permitting, but enjoining, the extremest form of incest.

The Arunta, again, neighbours of the Urabunna, though said to have more of "individual marriage" than they, in seasons of license go much beyond the Urabunna, though not so far as the Fijians. Women, at certain large meetings, "are told off ... and with the exception of men who stand in the relation of actual Uther, brother, or sons, they are, for the time being, common property to all the men present on the corroboree ground." Women are thus handed over to men "whom, under ordinary circumstances, they may not even speak to or go near."[9] Every known rule, except that which forbids the closest incest as understood by ourselves, is deliberately and purposely reversed[10] by the Arunta on certain occasions. Another example will be produced later, that of the Dieri, neighbours of the Urabunna.

We suggest, then, that these three grades of license—the Urabunna, adulterous, but more or less permanent, and limited by rules and by tribal and modern laws of incest; the Arunta, not permanent, adulterous, and tribally incestuous, limited only by our own ideas of the worst kinds of incest; and the Fijian, not permanent, adulterous, and of an incestuous character not only unlimited by laws, but rather limited by the desire to break the most sacred laws—are all of the same kind. They are not, we suggest, survivals of "group marriage," or of a period of perfect promiscuity in everyday life, though that they commemorate such a fancied period is the Arunta myth, just as the Roman myth averred that the Saturnalia commemorated the anarchy of the Golden Age.

"In Saturn's time
Such mixture was not held a crime."

The Golden Age of promiscuity is, of course, reported, not in an historical tradition recording a fact, but in a myth invented to explain the feasts of license. Men find that they have institutions, they argue that they must once have been without institutions, they make myths about ancestors or gods who introduced institutions, they invent the Golden Age, when there were none, and, on occasion, revert for a day or a week to that happy ideal. The periods of license cannot be true commemorative functions, continued in pious memory of a time of anarchy since institutions began.

But of the three types, Urabunna, Arunta, Fijian, the Urabunna, except in its degree of permanence, is the least licentious, least invades law, and it is a curious question why incest increases at these feasts as culture advances, up to a certain point. The law invaded by the Urabunna Piraungaru custom is not the tribal law of incest, nor the modern law of incest, but the law of the sanctity of individual marriage. It may therefore be argued (as against my own opinion) that the sanctity of individual marriage is still merely a nascent idea among the Urabunna, an idea which is recent, and so can be set aside easily; whereas the tribal laws of incest are strong with the strength of immemorial antiquity, and therefore must have already existed in a past age when there was no individual marriage at all. On this showing we have, first, the communal undivided horde; next, the horde bisected into groups which must not marry within each other (phratries), though why this arrangement was made and submitted to nobody can guess with any plausibility. By this time all females of phratry A might not only marry any man of phratry B, but were, according to the hypothesis, by theory and by practice, all wives of all men of phratry B. Next, as to-day, a man of B married a woman of A, with or without the existing offensive rites, but his tenure of her is still so insecure and recent that it is set aside, to a great extent, by the Piraungaru or Pirauru custom, itself a proof and survival of "group marriage," and of communal promiscuity in the past. Such is the argument for "group marriage," which may be advanced against my opinion, or thus, if I did not hold my opinion, I would state the argument.

This licentious custom, whether called Piraungaru or by other names, is, with the tribal names for human relationships, the only basis of the belief in the primal promiscuous horde. Now, as to these names of relationships, we may repeat the adverse arguments already advanced by us in Social Origins, pp. 99-103. "Whatever the original sense of the names, they all now denote seniority and customary legal status in the tribe, with the reciprocal duties, rights, and avoidances.... The friends of group and communal marriage keep unconsciously forgetting, at this point of their argument, that our ideas of sister, brother, father, mother, and so on, have nothing to do (as they tell us at certain other points of their argument) with the native terms, which include, indeed, but do not denote these relationships as understood by us.... We cannot say 'our word "son" must not be thought of when we try to understand the native term of relationship which includes sons—in our sense,' and next aver that 'sons, in our sense, are regarded [or spoken of] as real sons of the group, not of the individual, because of a past [or present] stage of promiscuity which made real paternity undiscoverable.'"

Manifestly there lurks a fallacy in alternately using "sons," for example, in our sense, and then in the tribal sense, which includes both fatherhood, or sonship, in our sense, and also tribal status and duties. "The terms, in addition to their usual and generally accepted signification of relationship by blood, express a class or group relation quite independent of it."[11]

Thus the tribal names may result from an expanded use of earlier names of blood relationship, or names of tribal status may now be applied to include persons who are within degrees of blood relationship. In the latter case, how do we know that a tribe with its degrees of status is primitive? Starcke thinks that Mr. Morgan's use of terms of relationship as proof of "communal marriage" is "a wild dream, if not the delirium of fever." "The nomenclature was in every respect the faithful reflection of the juridical relations which arose between the nearest kinsfolk of each tribe. Individuals who were, according to the legal point of view, on the same level with the speaker, received the same designation. The other categories of kinship were formally developed out of this standpoint." The system of names for relationships "affords no warrant" for Mr. Morgan's theory of primitive promiscuity.[12]

Similar arguments against inferring collective marriage in the past from existing tribal terms of relationship are urged by Dr. Durkheim.[13] He writes, taking an American case of names of relationship, as against Professor Kohler: "We see that the (Choctaw) word Inoha (mother) applies indifferently to all the women of my mother's group, from the oldest to the youngest. The term thus defines its own meaning: it applies to all the women of the family (or clan?) into which my father has married. Doubtless it is rather hard to understand how the same term can apply to so many different people. But certain it is, that the word cannot awake, in men's minds, any idea of descent, in the usual sense of the word. For a man cannot seriously regard his second cousin as his mother, even virtual. The vocabulary of relationships must therefore express something other than relations of consanguinity, properly so-called.... Relationship and consanguinity are very different things ... relationship being essentially constituted by certain legal and moral obligations, which society imposes on certain individuals."[14]

The whole passage should be read, but its sense is that which I have already tried to express; and Dr. Durkheim says, "The hypothesis of collective marriage has never been more than an ultima ratio" (a last resource), "intended to enable us to envisage these strange customs; but it is impossible to overlook all the difficulties which it raises."

An analogous explanation of the wide use of certain terms of relationship has been given by Dr. Fison, of whom Mr. Howitt writes, "Much of what I have done is equally his."[15] Dr. Fison says, "All men of the same generation who bear the same totem are tribally brothers, though they may belong to different and widely separated tribes. Here we find an explanation of certain apparently anomalous terms of relationship. Thus, in some tribes the paternal grandson and his grandfather call one another 'elder brother' and 'younger brother' respectively. These persons are of the same totem."[16] "Many other designations" in Mr. Morgan's Tables of Terms of Relationship "admit of a similar solution."[17] The terms do not denote degrees of blood relationship, but of brotherhood in the totem (or phratry, or matrimonial class). It is so, too, with the Choctaw term for Mother. Every one knows who his mother, in our sense, is: the Choctaw term denotes a tribal status.

If it be said that, because a man calls his wife his Noa, and also calls all women whom he might have married his Noa, therefore all these women, in past times, would have been his wives; it might as well be said that all the women whom he calls "mother" would, in times past, have collaborated in giving birth to him. As far as these terms indicate relationship, "a man is the younger brother of his maternal grandmother," and the maternal grandfather of his second cousin![18] The terms do not denote relationship in blood, clearly, but something quite different.

The custom of Piraungaru, or Pirrauru, and cases of license at festivals, and the names for tribal relations, are, we repeat, the only arguments in favour of the theory of the communal horde.[19] We have shown that the terms of relationship do not necessarily help the theory. That theory, again, is invalidated by its inability to account for the origin of the rules forbidding marriage between persons of the same phratry (for it does not tell us why the original medicine man conceived the idea of regulations), or even to account for the origin of the phratriac divisions.

But why, on our system, can the Piraungaru custom break the rule of individual marriage more easily than the law prohibiting incest? Why it can do so on the theory of pristine promiscuity we have explained (p. 41, supra).

We reply that individual marriage has not, among savages, any "religious" sanction; it is protected by no form of the phratry or totem tabu; by no god, such as Hymen; but rests, as from the first it rested, on the character and strength of the possessor of the woman or women, and falls into abeyance if he does not choose to exert it. If the males of the Urabunna have so far departed from the natural animal instincts as usually (with exceptions) to prefer to relax their tenure of women, being tempted by the bribe of a legalised change of partners all round, they exhibit, not a primitive, but a rather advanced type of human nature. The moral poet sings:—

"Of Whist or Cribbage mark the amusing Game,
The Partners changing, but the Sport the same,
Then see one Man with one unceasing Wife,
Play the long Rubber of connubial Life."[20]

This is the "platform" of the Urabunna and Dieri, as it is of the old Cicisbeism in Italy, and of a section of modern "smart society," especially at the end of the ancien régime in France. Man may fall into this way of thinking, just as, in Greece, he actually legalised unnatural passions by a ceremony of union. "That one practice, in many countries, became systematised," as Mr. J. F. McLennan wrote to Mr. Darwin.[21]

This is not the only example of a legalised aberration from nature, or from second nature. Abhorrence of incest has become a law of second nature, among savage as among civilised men. But Dr. Durkheim publishes a long list of legalised aberrations from the laws of incest among Hebrews, Arabs, Phœnicians, Greeks, Slavonic peoples, Medes, Persians, Egyptians, Cambodians, and Peruvians.[22] If these things, these monstrous aberrations, can be legalised "in the green tree," why should not jealousy fall into a kind of legalised abeyance among the Urabunna, under the law of partner-shifting? The Piraungaru custom does not prove that earliest man was not ferociously jealous; it merely shows that certain tribes have reached a stage in which jealousy is, at present, more or less suppressed in favour of legalised license.

We catch the Urabunna and Dieri at a moment of development in which the abandonment of strict possession of a wife is compensated for by a legalised system of changing partners, enduring after the feast of license is over. But even so, a man is responsible, as father, for the children of his actual wife, not for the children of his Piraungaru paramours. For these their actual husbands (Tippa Malku) are responsible.

Mr. Howitt says, in his earlier account of this institution, that among the Dieri, neighbours of the Urabunna, the men and women who are made Pirauru are not consulted. The heads of the tribe do not ask whether they fancy each other or not. "The time is one of festivity, feasting, and amusement," only too obviously! "Dancing is carried on." "A man can always exercise marital rights towards his Pirauru, if they meet when her Noa (real husband) is absent, but he cannot take her away from him unless by his consent," except at the feasts. But the husband usually consents. "In spite of all this arrangement, most of the quarrels among the Dieri arise out of this Pirauru practice.... "A son or daughter regards the real husband (Noa) of his mother as his Apiri Muria, or "real father"; his mother's Pirauru is only his Apiri Waka, or "little father." At certain feasts of license, such as intertribal marriages, "no jealous feeling is allowed under penalty of strangling, but it crops up afterwards, and occasions many bloody affrays."[23] Thus jealousy is not easily kept in abeyance by customary law.

The idea of such a change of partners is human, not animal, and the more of a brute the ancestor of man was the less could he dream, in times truly primitive, of Piraungaru as a permanent arrangement. Men, in a few tribes, declined into it, and are capable of passing out of it, like the Urabunna or Dieri man, who either retains so much of the animal, or is rising so far towards the Homeric standard, as to fight rather than let his wife be allotted to another man, or at least to thump that other man afterwards.

The Dieri case of the feast of license, just mentioned, is notable. "The various Piraurus (paramours) are allotted to each other by the great council of the tribe, after which their names are formally announced to the assembled people on the evening of the ceremony of circumcision, during which there is for a time a general license permitted between all those who have been thus allotted to each other." But persons of the same totem among the Dieri may not be Piraurus to each other, nor may near relations as we reckon kinship, including cousins on both sides.

In this arrangement Mr. Howitt sees "a form of group marriage," while I see tribe-regulated license, certainly much less lawless than that of the more advanced Fijians or the Arunta. Mr. Howitt did not state that the Pirauru or Piraungaru unions are preceded (as marriage is) by any ceremony, unless the reading the banns, so to speak, by public proclamation among the Dieri is a ceremony.[24] Now he has discovered a ceremony as symbolic as our wedding ring (1904).

Little light, if any, is thrown on these customs of legalised license by philology. Mr. Howitt thought that Pirauru may be derived from Pira, "the moon," and Uru, "circular." The tribal feasts of license are held at the full moon, but I am not aware that, by the natives, people are deemed peculiarly "moonstruck," or lunatic, at that season. If Urabunna Piraungaru is linguistically connected with Dieri Pirauru, then both Piraungaru and Pirauru may mean "Full Mooners." "Thy full moons and thy festivals are an abomination to me!"[25]

Among the Dieri, "a woman becomes the Noa of a man most frequently by being betrothed to him when she is a mere infant.... In certain cases she is given by the Great Council, as a reward for some meritorious act on his part." "None but the brave deserve the fair," and this is "individual marriage," though the woman who is wedded to one man may be legally allotted as Full Mooner, or Pirauru, to several. "The right of the Noa overrides that of the Pirauru. Thus a man cannot claim a woman who is Pirauru to him when her Noa is present in the camp, excepting by his consent." The husband generally yields, he shares equivalent privileges. "Such cases, however, are the frequent causes of jealousies and fights."[26]

This evidence does not seem, on the whole, to force upon us the conclusion that the Urabunna Piraungaru custom, or any of these customs, any more than the custom of polyandry, or of legalised incest in higher societies, is a survival of "group marriage"—all men of certain social grades being actual husbands of all women of the corresponding grades—while again that is a survival of gradeless promiscuity. We shall disprove that theory. Rather, the Piraungaru custom appears to be a limited concession to the taste, certainly a human taste, for partner-changing—which can only manifest itself where regular partnerships already exist. Jealousy among these tribes is in a state of modified abeyance: like nature herself, and second nature, where, among civilised peoples, things unnatural, or contrary to the horror of incest, have been systematically legalised.

I have so far given Mr. Howitt's account of Pirrauru (the name is now so written by him) among the Dieri, as it appeared in his works, prior to 1904. In that year he published his Native Tribes of South-East Australia, which contains additional details of essential importance (pp. 179-187). A woman becomes Tippa Malku,[27] or affianced,[28] to one man only, before she becomes Pirrauru, or what Mr. Howitt calls a "group wife." A "group wife," I think, no woman becomes. She is never the Pirrauru of all the men who are Noa to her, that is, intermarriageable with her. She is merely later allotted, after a symbolic ceremony, as a Pirrauru to one or more men, who are Noa to her. At first, while a child, or at least while a maiden, she is betrothed (there are varieties of modes) to one individual male. She may ask her husband to let her take on another man as Pirrauru; "should he refuse to do this she must put up with it." If he consents, other men make two adjacent ridges of sand, and level them into one larger ridge, while a man, usually the selected lover, pours sand from the ridge over the upper part of his thighs, "buries the Pirrauru in the sand." (The phrase does not suggest that Pirrauru means "Full Mooners.") This is the Kandri ceremony, it is performed when men swop wives (exchange their Noa as Pirraurus), and also when "the whole of the marriageable or married people, even those who are already Pirrauru, are reallotted," a term which suggests the temporary character of the unions.

I am ready to allow that the Kandri ceremony, a symbol of recognised union, like our wedding ring, or the exchanged garlands of the Indian Ghandarva rite, constitutes, in a sense, marriage, or a qualified union recognised by public opinion. But it is a form of union which is arranged subsequent to the Tippa Malku ceremony of permanent betrothal and wedlock. Moreover, it is, without a shadow of doubt, subsequent in time and in evolution to the "specialising" of one woman to one man in the Tippa Malku arrangement. That arrangement is demonstrably more primitive than Pirrauru, for Pirrauru is unthinkable, except as a later (and isolated) custom in modification of Tippa Malku.

This can easily be proved. On Mr. Howitt's theory, "group marriage" (I prefer to say "status marriage") came next after promiscuity. All persons legally intermarriageable (Noa), under phratry law, were originally, he holds, ipso facto, married. Consequently the Kandri custom could not make them more married than they then actually were. In no conceivable way could it widen the area of their matrimonial comforts, unless it enabled them to enjoy partners who were not Noa, not legally intermarriageable with them. But this the Kandri ceremony does not do. All that it does is to permit certain persons who are already Tippa Malku (wedded) to each other, to acquire legal paramours in certain other wedded or Tippa Malku women, and in men either married or bachelors. Thus, except as a legalised modification of individual Tippa Malku, Pirrauru is impossible, and its existence is unthinkable.[29]

Pirrauru is a modification of marriage (Tippa Malku), Tippa Malku is not a modification of "group marriage." If it were, a Tippa Malku husband, "specialising" (as Mr. Howitt says) a woman to himself, would need to ask the leave of his fellows, who are Noa to his intended fiancée.[30] The reverse is the case. A man cannot take his Pirrauru woman away from her Tippa Malku husband "unless by his consent, excepting at certain ceremonial times"—feasts, in fact, of license. Pirrauru secures the domestic peace, more or less, of the seniors, by providing the young men (who otherwise would be wifeless and desperate) with legalised lemans. By giving these Pirrauru "in commendation" to the young men, older men increase their property and social influence. What do the Tippa Malku husbands say to this arrangement?

As for "group" marriage, there is nothing of the kind; no group marries another group, the Pirrauru literally heap hot coals on each other if they suspect that their mate is taking another of the "group" as Pirrauru. The jealous, at feasts of license, are strangled (Nulina). The Rev. Otto Siebert, a missionary among the Dieri, praises Pirrauru for "its earnestness in regard to morality." One does not quite see that hiring out one's paramours, who are other men's wives, to a third set of men is earnestly moral, or that jealousy, checked by strangling in public, by hot coals in private, is edifying, but Pirrauru is not "group marriage." No pre-existing group is involved. Pirrauru may (if they like jealousy and hot coals) live together in a group, or the men and women may often live far remote from each other, and meet only at bean-feasts.

You may call Pirrauru a form of "marriage," if you like, but, as a later modification of a prior Tippa Malku wedlock, it cannot be cited as a proof of a yet more pristine status-marriage of all male to all female intermarriageable persons, which supposed state of affairs is called "group marriage."[31]

If Pirrauru were primitive, it might be looked for among these southern and eastern tribes which, with the pristine social organisation of the Urabunna and their congeners, lack the more recent institutions of circumcision, subincision, totemic magic, possess the All Father belief, but not the belief in prehuman predecessors, or, at least, in their constant reincarnation. (This last is not a Dieri belief.) But among these primitive south-east tribes, Pirrauru is no more found than subincision. Nor is it found among the Arunta and the northern tribes. It is an isolated "sport" among the Dieri, Urabunna, and their congeners. Being thus isolated, Pirrauru cannot claim to be a necessary step in evolution from "group marriage" to "individual marriage." It may, however, though the point is uncertain, prevail, or have prevailed, "among all the tribes between Port Lincoln and the Yerkla-mining at Eucla," that is, wherever the Dieri and Urabunna phratry names, Matteri and Kararu, exist.[32] Having identical phratry names (or one phratry name identical, as among the Kunandaburi), whether by borrowing or by original community of language and institutions: all these tribes southward to the sea from Lake Eyre may possess, or may have possessed, Pirrauru.

Among the most pristine of all tribes, in the south by east, however, Pirrauru is not found. When we reach the Wiimbaio, the Geawe-gal, the Kuinmarbura, the Wakelbura, and the Narrang-ga, we find no Pirrauru. But Mr. Howitt notes other practices which are taken by him to be mere rudimentary survivals of "group marriage." They are (i.) exchange of wives at feasts of marriage, or in view of impending misfortune, as when shipwrecked mariners break into the stores, and are "working at the rum and the gin." These are feasts of license, not survivals of "group marriage" nor of Pirrauru. (ii.) The jus primae noctis, enjoyed by men of the bridegroom's totem. This is not marriage at all, nor is it a survival of Pirrauru. (iii.) Very rare "saturnalia," "almost promiscuous." This is neither "group marriage" (being almost promiscuous and very rare) nor Pirrauru. (iv.) Seven brothers have one wife. This is adelphic polyandry, Mr. Howitt calls it "group marriage." (v.) "A man had the right to exchange his wife for the wife of another man, but the practice was not looked upon favourably by the clan." If this is "group marriage" (there is no "group" concerned) there was group marriage in ancient Rome.[33] This, I think, is all that Mr. Howitt has to show for "group marriage" and Pirrauru among the tribes most retentive of primitive usages.

The manner in which Tippa Malku betrothals are arranged deserves attention. They who "give this woman away," and they who give away her bride-groom also, are the brothers of the mothers of the pair, or the mothers themselves may arrange the matter.[34]

Mr. Howitt, on this point, observes that, if the past can be judged of by the present, "I should say that the practice of betrothal, which is universal in Australia, must have produced a feeling of individual proprietary right over the women so promised." Manifestly Mr. Howitt is putting the plough before the oxen. It is because certain kinsfolk have an acknowledged "proprietary right" over the woman that they can betroth her to a man: it is not because they can betroth her to a man that they have "a feeling of individual proprietary right over her." I give my coppers away to a crossing-sweeper, or exchange them for commodities, because I have an individual proprietary right over these coins. I have not acquired the feeling of individual proprietary right over the pence by dint of observing that I do give them away or buy things with them.

The proprietary rights of mothers, maternal uncles, or any other kinsfolk over girls must, of course, have been existing and generally acknowledged before these kinsfolk could exercise the said rights of giving away. But, in a promiscuous horde, before marriage existed, how could anybody know what persons had proprietary rights over what other persons?[35]

Mr. Howitt here adds that the "practice of betrothal ..." (or perhaps he means that "the feeling of individual proprietary right"?) "when accentuated by the Tippa Malku marriage, must also tend to overthrow the Pirrauru marriage." Of course we see, on the other hand, and have proved, that if there were no Tippa Malku marriage there could be no Pirrauru to overthrow.

As to the Pirrauru or Piraungaru custom, moreover, Mr. Howitt has himself candidly observed that, on his theory, it "ought rather to have been perpetuated than abandoned" (so it is abandoned) "under conditions of environment" (such as more abundant food) "which permitted the Pirrauru group to remain together on one spot, instead of being compelled by the exigencies of existence to separate into lesser groups having the Noa" (or regular) "marriage."[36] So Pirrauru don't live in "groups"!

As a fact, the more that supplies, in some regions, as on the south coast, permit relatively large groups to coexist, the less is their marital license; while, on the other hand, the less favourable the conditions of supply (as in the Barkinji region), the less do we hear of Pirrauru, or anything of the kind, except among tribes of the Kiraru and Matteri phratries. For these reasons, Pirrauru unions appear to mark an isolated moment in culture, not to be a survival of universal pristine promiscuity. They are almost always associated, in their inception, with seasons of frolic and lust, and with large assemblages, rather than with the usual course of everyday existence.

For the reasons here stated, it does not seem that Australian institutions yield any evidence for primitive promiscuity.

[1] Native Tribes of South-East Australia, p. 89.

[2] Native Tribes of South-East Australia, p. 90.

[3] Loc. cit. Mr. Howitt says "classes," but we adhere to the term "phratries."

[4] Natives of Central Australia, Spencer and Gillen, p. 63.

[5] Spencer end Gillen, pp. 92-98.

[6] Natives of Central Australia, Spencer and Gillen, p. 63.

[7] For a large account of these customs see The Golden Bough, second edition.

[8] Fison, J.A.I., xiv. p. 28.

[9] Natives of Central Australia, Spencer and Gillen, p. 97.

[10] Ibid., p. 111.

[11] Roth, N.W.C. Queensland Aborigines, p. 56.

[12] Starcke, The Primitive Family, p. 207.

[13] L'Année Sociologique, i. pp. 313-316.

[14] L'Année Sociologique, i. p. 315.

[15] Native Tribes of South-East Australia, xiv.

[16] Can Dr. Fison mean of the same matrimonial class?

[17] Kamilaroi and Kurnai, pp. 166, 167.

[18] Native Races of South-East Australia, p. 163. Pointed out by Mr. N. W. Thomas.

[19] The participation of many men in the jus primae noctis is open to various explanations.

[20] Poetry of the Antijacobin.

[21] Studies in Ancient History, ii. p. 52.

[22] L'Année Sociologique, i., pp.38, 39, 62.

[23] J. A. I., pp. 56-60, August 1890.

[24] Howitt, J. A. I., August 1890, pp. 55-58.

[25] What the Dieri call Pirauru (legalised paramour) the adjacent Kunan-daburi tribe call Dilpa Mali. In this tribe the individual husband or individual wife (that is, the real wife or husband) is styled Nubaia, in Dieri Noa, in Urabunna Nupa. Husband's brother, sister's husband, wife's sister, and brother's wife are all Nubaia Kodimali in Kunandabori, and are all Noa in Dieri. What Dilpa Mali (legalised paramour, or "accessory wife or husband") means in Kunandabori Mr. Howitt does not know. But he learns that Kodi Mali (applied to Pirauru) means "not Nubaia," that is, "not legal individual husband or wife." If we knew what Dilpa means in Dilpa Mali (legalised paramour of either sex), we should know more than we are apt to do in the present state of Australian philology.

At Port Lincoln a man calls his own wife Yung Ara, that of his brother Karteti (Trans. Phil. Soc. Vic., v. 180). What do these words mean?—Report of Regents of Smithsonian Institute, 1883, pp. 804-806.

[26] Report of Regents of Smithsonian Institute, 1883, p. 807.

[27] Tippa, in one tongue, Malku in another, denote the tassel which is a man's full dress suit.

[28] Mr. Howitt says that the pair are Tippa Malku "for the time being" (p. 179), though the association seems to be permanent. May girls Tippa Malku—"sealed" to a man—have relations with other men before their actual marriage, and with what men? We are not told, but a girl cannot be a Pirrauru before she is Tippa Malku. If Pirrauru "arises through the exchange by brothers of their wives" (p. 181), how can an unmarried man who has no wife become a Pirrauru? He does. When Pirrauru people are "re-allotted" (p. 182), does the old connection persist, or is it broken, or is it merely in being for the festive occasion? How does the jealousy of the Pirrauru, which is great, like the change? These questions, and many more, are asked by Mr. N. W. Thomas.

[29] Will any one say, originally all Noa people were actual husbands and wives to each other? Then the Kandri ceremony and Pirrauru were devised to limit Tom, Dick, and Harry, &c., to Jane, Mary, and Susan, &c., all these men being Pirrauru to all these women, and vice versa. Next, Tippa Malku was devised, limiting Jane to Tom, but Pirrauru was retained, to modify that limitation. Anybody is welcome to this mode of making Pirrauru logically thinkable, without prior Tippa Malku: if he thinks that the arrangement is logically thinkable, which I do not.

[30] Or his seniors would hare to ask it. But his kin could not possess the tight to betroth him before kinship was recognised, which, before marriage existed, it could not be.

[31] I have here had the advantage of using a MS. note by Mr. N. W. Thomas.

[32] Native Tribes of South-East Australia, p. 191.

[33] Native Tribes of South-East Australia, pp. 195, 217, 219, 224, 260.

[34] Native Tribes of South-East Australia, pp. 177, 178.

[35] Ibid., p. 283.

[36] J. A. I., xiii. p. 34.

[CHAPTER IV]

THE ARUNTA ANOMALY

How could man, if promiscuous, cease to be so?—Opinion of Mr. Howitt—Ethical training in groups very small, by reason of economic conditions—Likes and dislikes—Love and jealousy—Distinctions and restrictions—Origin of restrictions not explained by Professor Spencer—His account of the Arunta—Among them the totem does not regulate marriage, is not exogamous, denotes a magical society—Causes of this unique state of things—Male descent: doctrine of reincarnation, belief in spirit-haunted stone churinga nanja—Mr. Spencer thinks Arunta totemism pristine—This opinion contested—How Arunta totemism ceased to regulate marriage—Result of isolated belief in churinga nanja—Contradictory Arunta myths—Arunta totemism impossible in tribes with female descent—Case of the Urabunna—Origin of churinga nanja belief—Sacred stone objects in New South Wales—Present Arunta belief perhaps based on myths explanatory of stone amulets of unknown meaning—Proof that the more northern tribes never held the Arunta belief in churinga nanja—Traces of Arunta ideas among the Euahlayi—Possible traces of a belief in a sky-dwelling being among southern Arunta—Mr. Gillen's "great Ulthaana of the heavens"—How arose the magic-working animal-named Arunta societies?—Not found in the south-east—Mr. Spencer's theory that they do survive—Criticism of his evidence—Recapitulation—Arunta totemism not primitive but modified.

Next we have to ask how, granting the hypothesis of the promiscuous horde, man ceased to be promiscuous. It will be seen that, on a theory of Mr. Howitt's, man was, in fact, far on the way of ceasing to be promiscuous or a "horde's man," before he introduced the moral reform of bisecting his horde into phratries, for the purpose of preventing brother with sister marriages. Till unions were permanent, and kin recognised, things impossible in a state of promiscuity, nobody could dream of forbidding brother and sister marriage, because nobody could know who was brother or sister to whom. Now, Mr. Howitt does indicate a way in which man might cease to be promiscuous, before any sage invented the system of exogamous phratries.

He writes,[1] "I start ... from the assumption that there was once an undivided commune ... I do not desire to be understood as maintaining that it implies necessarily the assumption of complete communism between the sexes. Assuming that the former physical conditions of the Australian continent were much as they are now, complete communism always existing would, I think, be an impossibility. The character of the country, the necessity of hunting for food, and of removing from one spot to another in search of game and of vegetable food, would necessarily cause any undivided commune, when it assumed dimensions of more than that of a few members, to break up, under the necessities of existence, into two or more communes of similar constitution to itself. In addition to this it has become evident to me, after a long acquaintance with the Australian savage, that, in the past as now, individual likes and dislikes must have existed; so that, although there was the admitted common right between certain groups of the commune, in practice these rights would either not be exercised by reason of various causes, or would remain in abeyance, so far as the separated but allied undivided communes were concerned, until on great ceremonial occasions, or where certain periodical gatherings for food purposes reunited temporarily all the segments of the original community. In short, so far as the evidence goes at present, I am inclined to regard the probable condition of the undivided commune as being well represented now by what occurs when on certain occasions the modified divided communes reunite."[2]

What occurs in these festive assemblies among certain central and northern tribes, as we have seen, is a legalised and restricted change of wives all round, with disregard, in some cases, of some of the tribal rules against incest. On Mr. Howitt's theory the undivided communal horde must always have been, as I have urged, dividing itself, owing to lack of supplies. It would be a very small group, continually broken up, and intercourse of the sexes even in that group, must have been restrained by jealousy, based on the asserted existence of individual "likes" and "dislikes." These restrictions, again, must have led to some idea that the man usually associated with, and responsible for feeding, and protecting, and correcting the woman and her children, was just the man who "liked" her, the man whom she "liked," and the man who "disliked" other men if they wooed her.

But that state of things is not an undivided communal horde at all! It is much more akin to the state of things in which I take marriage rules to have arisen.

We may suppose, then, that early moral distinctions and restrictions grew up among the practically "family" groups of everyday life, as described by Mr. Howitt, and we need not discuss again the question whether, at this very early period, there existed a community exactly like the local tribe of to-day in every respect—except that marriage was utterly unregulated, till an inspired medicine man promulgated the law of exogamy, his own invention.

Mr. Howitt began his long and invaluable studies of these problems as a disciple of Mr. Lewis Morgan. That scholar was a warm partisan of the primeval horde, of group marriage, and (at times) of a reformatory movement. These ideas, first admitted to Mr. Howitt's mind, have remained with him, but he has seen clearly that the whole theory needed at least that essential modification which his practical knowledge of savage life has enabled him to make. He does not seem to me to hold that the promiscuous horde suddenly, for no reason, reformed itself: his reformers had previous ethical training in a state of daily life which is not that of the hypothetical horde. But he still clings to the horde, tiny as it must have been, as the source of a tradition of a brief-lived period of promiscuity. This faith is but the "after-image" left in his mental processes by the glow of Mr. Morgan's theory, but the faith is confirmed by his view of the terms of relationship, and of the Piraungaru, Pirrauru, and similar customs. We have shown, in the last chapter, that the terms and the customs are not necessarily proofs of promiscuity in the past, but may be otherwise interpreted with logical consistency, and in conformity with human nature.

The statement of Mr. Howitt shows how the communal horde of the hypothesis might come to see that it needed moral reformation. In daily life, by Mr. Howitt's theory, it had practically ceased to be a communal horde before the medicine man was inspired to reform it. The hypothesis of Professor Baldwin Spencer resembles that of Mr. Howitt, but, unlike his (as it used to stand), accounts for the existence of animal-named sets of people within the phratries. Mr. Spencer, starting from the present social condition of the Arunta "nation" or group of tribes (Arunta, Kaitish, Ilpirra, Unmatjera), supposes that these tribes retain pristine traits, once universal, but now confined to them. The peculiar pristine traits, by the theory, are (1) the existence of animal-named local societies for magical purposes. The members of each local group worked magic for their name-giving animal or plant, but any one might marry a woman of his own group name, Eagle Hawk, Cockatoo, and the like, while these names were not inherited, either from father or mother, and did not denote a bond of kinship. Mr. Spencer, then, supposes the horde to have been composed of such magical societies, at a very remote date, before sexual relations were regulated by any law. Later, in some fashion, and for some reason which Mr. Spencer does not profess to explain, "there was felt the need of some form of organisation, and this gradually resulted in the development of exogamous groups."[3] These "exogamous groups," among the Arunta, are now the four or eight "matrimonial classes," as among other tribes of northern Australia. These tribes, as a rule, have phratries, but the Arunta have lost even the phratry names.

Mr. Spencer's theory thus explains the existence of animal-named groups—as co-operative magical societies, for breeding the animals or plants—but does not explain how exogamy arose, or why, everywhere, except among the Arunta, all the animal or plant named sets of people are kinships, and are exogamous, while they are neither the one or the other among the Arunta. Either the Arunta groups have once been exogamous totem kinships, and have ceased to be so, becoming magical societies; or such animal-named sets of people have, everywhere, first been magical societies, and later become exogamous totem kinships. Mr. Spencer holds the latter view, we hold the former, believing that the Arunta have once been in the universal state of totemic exogamy, and that, by a perfectly intelligible process, their animal-named groups have become magical societies, no longer exogamous kinships. We can show how the old exogamous totem kinship, among the Arunta, became a magical society, not regulating sexual relations; but we cannot imagine how all totemic mankind, if they began with magical societies in an unregulated horde, should have everywhere, except among the Arunta, conspired to convert these magical societies into kinships with exogamy. If the social organisation of the Arunta were peculiarly primitive, if their beliefs and ceremonials were of the most archaic type, there might be some ground for Mr. Spencer's opinion. But Mr. Hartland justly says that all the beliefs and institutions of the Arunta "point in the same direction, namely, that the Arunta are the most advanced and not the most primitive of the Central Australian tribes."[4]

The Arunta, a tribe so advanced that it has forgotten its phratry names, has male kinship, eight matrimonial classes, and local totem groups, with Headmen hereditary in the male line, and so cannot possibly be called "primitive," as regards organisation. If, then, the tribe possesses a peculiar institution, contravening what is universally practised, the natural inference is that the Arunta institution, being absolutely isolated and unique, as far as its non-exogamy goes, in an advanced tribe, is a local freak or "sport," like many others which exist. This inference seems to be corroborated when we discover, as we do at a glance, the peculiar conditions without which the Arunta organisation is physically impossible. These essential and indispensable conditions are admitted by Messrs. Spencer and Gillen to be:—

1. Male reckoning of descent—which is found in very many tribes where totems are exogamous—as everywhere.

2. Local totem groups, which are a result of male reckoning of descent. These also are found in many other tribes where, as everywhere, totems are exogamous.

3. The belief that the spirits of the primal ancestors of the "Dream-Time" (Alcheringa)—creatures evolved out of various animal shapes into human form—are constantly reincarnated in new-born children. This belief is found in all the northern tribes with male descent; and among the Urabunna, who have female descent—but among all these tribes totems are exogamous, as everywhere.

4. The Arunta and Kaitish, with two or three minor neighbouring tribes, believe that spirits desiring incarnation, all of one totem in each case, reside "at certain definite spots." So do the Urabunna believe, but at each of these spots, in Urabunna land, there may be spirits of several different totems.[5] Among the Urabunna, as everywhere, totems are exogamous. None of these four conditions, nor all of them, can produce the Arunta totemic non-exogamy.

Finally (5) the Arunta and Kaitish, and they alone, believe not only that the spirits desiring reincarnation reside at certain definite spots, and not only that the spirits there are, in each case, all of one totem (which is essential), but also that these spirits are most closely associated with objects of stone, inscribed with archaic markings (churinga nanja), which the spirits have dropped in these places—the scenes where the ancestors died (Oknanikilla). These stone objects, and this belief in their connection with ancestral spirits, are found in the Arunta region alone, and are the determining cause, or inseparable accident at least, of the non-exogamy of Arunta totemism, as will be fully explained later.

Not one of these five conditions is reported by Mr. Howitt among the primitive south-eastern tribes, and the fifth is found only in Aruntadom. Yet Mr. Spencer regards as the earliest form of totemism extant that Arunta form, which requires four conditions, not found in the tribes of primitive organisation, and a fifth, which is peculiar to the Arunta "nation" alone.

That the Arunta tribe, whether shut off from all others or not (as a matter of fact it is not), should alone (while advanced in all respects, including marriage and ceremonials) have retained a belief which, though called primitive, is unknown among primitive tribes, seems a singularly paradoxical hypothesis. Meanwhile the cause of the Arunta peculiarity—non-exogamous totems—is recognised by Messrs. Spencer and Gillen, who also declare that the cause is isolated. They say "it is the idea of spirit individuals associated with churinga" (manufactured objects of stone), "and resident in certain definite spots, that lies at the root of the present totemic system of the Arunta tribe."[6]

Again, they inform us that the churinga belief, and the existence of stone churinga, are things isolated. "In the Worgaia tribe, which inhabits the country to the north-east of the Kaitish" (neighbours of the Arunta), "we meet, so far as we have been able to discover, with the last traces of the churinga—that is, of the churinga with its meaning and significance, as known to us in the true central tribes, as associated with the spirits of Alcheringa ancestors" (mythical beings, supposed to be constantly reincarnated).[7] Thus, "the present totemic system of the Arunta tribe,"—in which, contrary to universal rule, persons of the same totem may inter-marry—reposes on a belief associated with certain manufactured articles of stone, and neither the belief nor the stone objects are discovered beyond a certain limited region. It is proper to add that the regretted Mr. David Carnegie found, at Family Wells, in the desert of Central Australia, two stone objects, one plain, the other rudely marked with concentric circles, which resemble churinga nanja. He mentions two others found and thrown away by Colonel Warburton. The meaning or use of these objects was not ascertained.[8]

We differ from Messrs. Spencer and Gillen when they think that this peculiar and isolated belief, held by four or five tribes of confessedly advanced social organisation and ceremonials (a belief only possible under advanced social organisation), is the pristine form of totemism, out of which all totemists, however primitive, have found their way except the Arunta "nation" alone. Messrs. Spencer and Gillen write: "... the only conclusion which it seems possible to arrive at is that in the more northern tribes" (which have no churinga nanja, no stone churinga), "the churinga represent the surviving relics of a time when the beliefs among those tribes were similar to those which now exist among the Arunta. It is more easy to imagine a change which shall lead from the present Arunta or Kaitish belief to that which exists among the Warramunga, than it is to imagine one which shall lead from the Warramunga to the Arunta."[9] Now among the Warramunga, as everywhere, the division of the totems between the two (exogamous) moieties is complete, "and, with very few exceptions indeed, the children follow the father."[10] (These exceptions are not explained.) Among the Kaitish the same totems occur among both exogamous moieties, so persons of the same totem can intermarry, but "it is a very rare thing for a man to marry a woman of the same totem as himself."[11]

The obvious conclusion is the reverse of that which our authors think "alone possible." The Kaitish have adopted the Arunta churinga nanja usage which introduces the same totem into both exogamous moieties, but, unlike the Arunta, they have not yet discarded the old universal rule, "No marriage within the totem." It is not absolutely forbidden, but it scarcely ever occurs. The Kaitish, as regards exogamy and religion, are a link between the primitive south-eastern tribes and the Arunta.

We go on to show in detail how Arunta totems alone ceased to be exogamous, and to demonstrate that the more northern tribes have never been, and never can have been, in the present Arunta condition. Among the Arunta, in the classes, none of them his own, into which alone a man may marry, there are plenty of women of his own totem. Thus, in marrying a woman of his totem, but not of his set of classes, a man does not break the law of Arunta exogamy. Now how does it happen that a totem may be in both sets of exogamous classes among the Arunta alone of mankind? Was this always the case from the beginning?

It is, naturally, our opinion that among the Arunta, as everywhere else, matters were originally, or not much later, so arranged that the same totem never appeared in both phratries, or, afterwards, when phratries were lost, in both opposed sets of two or four exogamous matrimonial classes. The only objection to this theory is that the Arunta themselves believe it, and mention the circumstance in their myths. These myths cannot be historical reminiscences of the "Dream-Time," which never existed. But even a myth may deviate into truth, especially as the Arunta must know that in other tribes the same totem never occurs in both phratries, and are clever enough to see that their method needs explanation as being an exception to general rule; and that, even now, "the great majority of any one totem belong to one moiety of the tribe." So they say that originally all Witchetty Grubs, for instance, were in the Bulthara-Panunga moiety (as most Grubs still are to this day), while all Emus were in the opposite exogamous moiety (Purula-Kumura). But, say Messrs. Spencer and Gillen, "owing to the system according to which totem names are" (now) "acquired, it is always possible for a man to be, say, a Purula or a Kumura, and yet a Witchetty; or, on the other hand, a Bulthara or a Panunga, and yet an Emu."[12] The present system of acquiring totem names has transferred the totems into both exogamous moieties, and so has made it possible to marry within the totem name.

This suggests that, in native opinion or conjecture, Arunta totems, like all others, were once exogamous; no totem ever occurred originally in both exogamous moieties. It also indicates that, in the opinion of Messrs. Spencer and Gillen, they only ceased to be exogamous when the present method of acquiring totem names, an unique method, was introduced. Happily, to prove the historical worthlessness of Arunta legendary myth, the tribe has a contradictory legend. The same totem, according to this fable, occurred in both exogamous moieties, even in the mythic Dream-Time (Alcheringa); by this fable the natives explain (what needs explaining) how the same totem does occur in both exogamous moieties to-day, and so is not exogamous.[13]

This is nonsense, just as the other contradictory myth was conjecture. Messrs. Spencer and Gillen have themselves explained why the same totem may now occur in both moieties, and so be non-exogamous. The unique phenomenon is due to the actual and unique method of acquiring totem names.[14] Thus the modern method is not primitive. These passages are very instructive.

The Arunta have been so long in the relatively advanced state of local totemism that their myths do not look behind it. A group, whether stationary or migratory, in the myths of the Dream-Time (the Alcheringa) always consists of persons of the same totem, with occasional visitors of other totems. The myths, we repeat, reflect the present state of local totem groups back on the past.

The myths allege (here the isolated superstition comes in) that the mythical ancestors of the Alcheringa died, or "went into the ground" at certain now haunted spots, marked by rocks or trees, which may be called "mortuary local totem-centres"—in native speech, Oknanikilla[15] Trees or rocks arose to mark the spot where the ancestors, all of one totem in each case, went into the ground. These trees or rocks are called Nanja. Thereabouts the dying ancestors deposited possessions peculiar to Aruntadom, their stone amulets, or churinga nanja, with what are now read as totemic incised marks. Their spirits, all of one totem in each case, haunt the Nanja rock or tree, and are especially attached to these stone amulets,[16] called churinga nanja. The spirits discarnate await a chance of entering into women, and being reborn. When a child comes to the birth, the mother, whatever her own or her husband's totem may be, names the spot where she supposes that she conceived the child, and the child's Nanja tree or rock is that in the Oknanikilla, or mortuary local totem-centre nearest to the place where the child was conceived. Its male kin hunt for the churinga, or stone amulet, there deposited by the dying Alcheringa ancestor; if they find it, it becomes the child's churinga, for he is merely the ancestor spirit reborn. He (or she) "comes into his own"; his Nanja tree or rock, his churinga nanja, and his original totem, which may be, and often is, neither that of his father or mother.

Thus inheriting his own old Nanja tree and churinga, and totem, the child is not necessarily of his father's or mothers but is of his own old original totem, say Grub, or Hakea Flower, or Kangaroo, or Frog. His totem is thus not inherited, we repeat, as elsewhere, from either parent, but is derived, by the accident of his place of conception, from the local totem, from the totemic ghosts (all of one totem) haunting the particular mortuary totem centre, or Oknanikilla, where he was conceived. His totem may thus be in both of the exogamous moieties, and for that reason alone is not exogamous. To take an example. A woman, by totem Cat, has a husband by totem Iguana. She conceives a child, and believes that she conceived it in a certain district. The local totem of that district is the Grub, Grub ghosts haunt the region; the child, therefore, is a Grub. He inherits his exogamous class, say Bukhara, from his father, and he must marry a woman of Class Kumara. But she may also be a Grub, for her totem, like his, has been acquired (like his, not by inheritance, but) by the accident that her mother conceived her in a Grub district. Thus, and thus only, are totems not exogamous among the Arunta. They are not inherited from either parent.

It is probable that, after male descent came in, the Arunta and Kaitish at first inherited their totems from their fathers, as among all other tribes with male descent. This appears to be proved by the fact that they still do inherit, from their fathers, totemic rites, and the power of doing totemic mummeries for their fathers' totems, even when, by the accident of their places of conception, they do not inherit their fathers' totems. When they did inherit the paternal totem, they were, doubtless, totemically exogamous, like all other tribes with either male or female descent.

One simple argument upsets the claim of Arunta totems to be primitive. In no tribe with female descent can a district have its local totem, as among the Arunta. A district can only have a local totem if the majority of the living people, and of the haunting ghosts of the dead, are of one totem only. But this (setting aside the occasional results of an isolated Urabunna superstition) can only occur under male reckoning of descent, which confessedly is not primitive. In a region where reckoning in the female line exists a woman could not say, "I conceived my child in Grub district, the country of totem Grub"—for such a country there is not and cannot be. Consequently, among the Urabunna as everywhere with reckoning of descent in the female line, every child is of its mother's totem.

Let us examine other tribes who, like the Arunta, have the theory of reincarnation, but whose totems are, as elsewhere, exogamous, unlike those of the Arunta. The Urabunna have female descent, and their myth about the origin of totemic ancestors approximates to that of the Arunta, but, unlike the Arunta fable, does not produce, or account for, non-exogamy in totems. Things began, say the Urabunna, by the appearance of a few creatures half human, half bestial or vegetable. They had miraculous powers, and dropped spirits which tenanted lizards, snakes, and so on, all over the district. These spirits later became incarnated in human beings of the Lizard, Snake, or other totem, and are constantly being reincarnated. The two Urabunna phratries were originally a green and a brown snake: the Green Snake said to the Brown Snake, "I am Kirarawa, you are Matthurie"—the phratry names. It does not appear that these names now mean Green Snake and Brown Snake, though they may once have had these significations. The spirits left about by these snakes, like all the other such spirits (mai aurli) keep on being incarnated, and, when incarnated, the children bear the totem name of their mothers in each case. A Green Snake woman is entered by a spirit, which she bears as a Green Snake child. The accident of the locality in which the child was conceived does not affect his totem, so Urabunna totems remain in their own proper phratries, and therefore, by phratry law, are exogamous, as everywhere, except among the Arunta.[17]

This arrangement is merely the usual arrangement, with female descent A woman's child is of the woman's totem. Believing in reincarnation, the Urabunna merely adapt that belief to the facts. With female descent an Emu woman's child is Emu. If a tribe has male descent, an Emu father's child is Emu. With female descent, a spirit has entered an Emu woman and been born Emu: with male descent, a spirit has entered the wife of an Emu man, and, by inheritance from his father, is Emu. Yet Messrs. Spencer and Gillen think that the Arunta and Kaitish rule—demanding the non-primitive male descent, local groups, local ghosts all of one totem, and churinga stones of the mark of that totem (all of which are indispensable), "is probably the simplest and most primitive."[18]

Most primitive, by our author's own statement, the Arunta method cannot be, for, as they show, it demands male descent, local totemism, and the peculiar belief about manufactured stone churinga. But they think it "most simple," because the Urabunna have a complicated myth, which, however, in no way affects the result, namely, that each child takes its mother's totem. Each spirit, according to the myth, changes its phratry and sex, and, necessarily, its totem, at each reincarnation, but that does not affect the result. Each child, as in all tribes with female descent, is still of its mother's totem.[19] No churinga nanja cause an anomaly among the Urabunna, for the churinga nanja, and the belief about them, among the Urabunna do not exist.

The Urabunna myth, adapted to male descent, occurs in all the northern tribes, from the northern bounds of the Kaitish to the sea, which have no stone churinga nanja; and in all of them totems are exogamous, because they never occur in both phratries, being uninfluenced by the Arunta churinga belief. They cannot, for they are duly inherited from the father, and they are so inherited because the tribes have not the exceptional Churinga Nanja creed, attaching the spirit to the amulet of a local totem group, which fixes—by the accident of place of conception—the totem of each child.

The Arunta non-exogamous totems, in Australia, as we saw, are only found where stone churinga nanja are in use; these amulets being peculiarly the residence of the spirits of totemic ancestors.

The origin of that belief is obscure. It could not arise in the present condition of Arunta or Kaitish affairs, for, now, every stone churinga in the tribe has already its recognised legal owner, and, on the death of an owner, or the extinction of a local totem group, the churinga are not left lying about to be found on or in the earth, but pass by a definite rule of inheritance; and they are all carefully warded and frequently examined, in Ertnatu-lunga, or sacred storehouses.[20] Thus stone churinga nanja, to-day, are not left lying about on the surface, or buried in graves, like those which, on the birth of each Arunta child, are sought for, and sometimes found, at the local totem-centre, and near the Nanja tree or rock, where the child was conceived. There churinga nanja must have been buried, of old, if our authors correctly say that the mythical ancestors "went into the ground, each carrying his churinga with him."[21] Again we read, "Many of the churinga were placed in the ground, some natural object again marking the spot." The spot was always marked by some natural object, such as a tree or rock.[22]

Though our authors tell us that they know Arunta natives who, on the birth of a child, have sought for and found his churinga nanja near the Nanja rock or tree next to the place where he was conceived, they do not say that the churinga are found by digging.[23] If they are, or if the Oknanikilla really are ancient burying-places (about which we are told nothing), the association of the churinga nanja with the ghost of the man in whose grave it is buried would be easily explained. But the impression left is that the stone churinga nanja found after search are discovered on the surface, dropped there by the spirit when about to be reincarnated.[24]

Here a curious fact may be filed for reference. Stone amulets, fashioned and decorated by man, are not known to be in use south of the Arunta region. But a cousin of my own, Mr. William Lang, found a stone object not unlike one figured by Messrs. Spencer and Gillen, on his station near Cooma, New South Wales. The decoration was of the rectilineal type prevalent in that region. Mr. Lang knew nothing of the Arunta churinga till I drew his attention to the subject. He then visited the Sydney Museum, and found several stone objects, "banana-shaped," exactly like the specimen (wooden?), one out of five known to Messrs. Spencer and Gillen, and published by them in their first work (p. 150). The New South Wales ornament, however, was always rectilineal. The articles appear to be obsolete among the tribes of New South Wales. It is said that they were erected of old round graves of the dead. Whites call them "grave stones." Careful articles on these decorated stone objects of New South Wales have been written by Mr. W. R. Harper and Mr. Graham Officer.[25] As a rule, they are not banana-shaped or crescentine, but are in the form of enormous stone cigars. They used to be placed, twelve or thirteen of them, on graves, and their weight, averaging about 3 lbs. to 4 lbs., makes them less portable than most of the churinga of the Arunta. It does not seem at all probable that Arunta stone churinga were ever erected round graves, but excavations at Oknanikilla, if they could be executed without a shock to Arunta sentiment, might throw some light on the subject.

In my opinion, the churinga found at Oknanikilla by the Arunta may have had no such original significance as is now attached to them. The belief may be a mere myth, explaining the sense of objects found and not understood—relics, as the myth itself avers, of an earlier race, the Alcheringa folk. The only information about those New South Wales decorated cigar-shaped and banana-shaped stone objects which could be got out of a local black was: "All same as bloody brand." He meant, conceivably, that the incised markings were totem marks, I think, and in that sense the marks on Arunta stone churinga are now interpreted.

It would not be surprising if the Arunta—supposing that they possessed the belief in "spirit trees," and the belief in reincarnation, and then found, near the Nanja trees or rocks, the stone amulets or "grave stones" of some earlier occupants of the region—evolved the myth that ancestral souls, connected with the spirit trees, abode especially in these decorated stones, common enough in American and European neolithic sites.

This is, of course, a mere conjecture. But Messrs. Spencer and Gillen agree with us when they say: "It is this idea of spirit individuals associated with churinga, and resident in certain definite spots, that lies at the root of the present totemic system of the Arunta tribes."[26]