The conjecture, therefore, that a traditional marriage relation, differing from that based on parentage, grew up out of a previous totem friendship, is based primarily on the importance which totemic cult alliances in general possessed within the totemic tribal organization. Other causes, of course, may also have co-operated. Two further points must be noticed. In the first place, it is not at all likely that the transition from the parental exogamous relation to the traditional form occurred at the same time in all the totem groups. This is not only highly improbable in itself, but is also absolutely irreconcilable with the fact, shown by the example of the Dieri, that the earlier transition from unlimited to limited exogamy was gradual. Moreover, one must bear in mind that the transition from parental to traditional exogamy, represented by diagram III (p. 148), not only underwent several repeated transformations, but that, due to the power which tradition always exerts, a traditional exogamous union of two totems, after it once arose, may have persisted throughout several changing cult friendships. An existing marriage relation may not at all have corresponded to the cult friendship that immediately preceded it; it may have been based on any earlier friendship whatsoever that had been favoured by conditions and that had received a firm place in tradition. These facts show that the hypothetical 'wise ancestors' of the present-day Australians—sages who are said to have invented this complicated organization in the immemorial past for the purpose of avoiding endogamy—are just as superfluous as they are improbable. The phenomena arose in the course of a long period of time, out of conditions immanent in the life and in the cult of these tribes. The various forms of exogamy appearing in the course of this period were not the causes but the effects of the phenomena in question.
[7. THE FORMS OF POLYGAMY.]
Unless external influences have changed his mode of life, primitive man, as we have seen, is both monogamous and endogamous, the latter term being used in a relative sense as denoting a condition in which marriages are permitted between blood relations as well as between non-relations. As a result of the external conditions of life, however, particularly the common habitation of the same protective cave and the use of adjacent hunting-grounds, unions within a wider joint family generally predominate. Following upon the rise of exogamy, polygamy also regularly appears. These two practices give to the marriage and family relations of totemic society an essentially different character from that which they possess under primitive conditions. Even in the totemic era, indeed, polygamy is not universal; monogamy continues to survive. Monogamy, however, ceases to be a norm of custom. It is everywhere set aside, to a greater or less extent, in favour of the two forms of polygamy—polygyny and polyandry.
Now it is apparent that precisely the same conditions that underlie the development of the various forms of exogamy also generate polygyny and polyandry. From the standpoint of the general human impulses determining the relations of the sexes, both sorts of polygamy are manifestly connected very closely with the origin of exogamy. Here, again, the fact that exogamy originated in marriage by capture from within the tribe is of decisive importance. It is precisely this friendly form of the capture of brides, as we may learn from the example of the Australians and of others, that is never carried out by the individual alone, whether the custom be still seriously practised or exists only in playful survivals. The companions of the captor aid him, and he, in turn, reciprocates in similar undertakings. Thereby the companion, according to a view that long continued to be held, gains a joint right to the captured woman. Hence the original form of polygamy was probably not polygyny—the only form, practically, that later occurs—but polyandry. At first this polyandry, which originates in capture, was probably only temporary in character. Nevertheless it inevitably led to a loosening of the marriage bond, the result of which might easily be the introduction of polygyny. The man who has gained a wife for his permanent possession seeks to indemnify himself, so far as possible, for the partial loss which he suffers through his companions. Here, then, two motives co-operate to introduce the so-called 'group-marriage'—the dearth of women, which may also act as a secondary motive in the claim of the companions to the captured woman, and the impulse for sexual satisfaction, which is, in turn, intensified by the lack of women. Similarly, the right to the possession of a woman, even though only temporarily, also has two sources. In the first place, the helper demands a reward for his assistance. This reward, according to the primitive views of barter and exchange, can consist only in a partial right to the spoils, which, in this case, means the temporary joint possession of the woman. In the second place, however, the individual is a member of the clan, and what he gains is therefore regarded as belonging also to the others. Thus the right of the closest companions may broaden into a right of the clan. Indeed, where strict monogamy does not prevent, phenomena similar to marriage by capture persist far beyond this period into a later civilization. Thus, in France and Scotland, down to the seventeenth century, the lord possessed the right of jus primæ noctis in the case of all his newly married vassals. In place of the clan of an earlier period we here find the lord; to him has been transmitted the right of the clan. At the time when these phenomena were in their early beginnings, the temporary relation might very easily have become permanent. It is thus that group-marriage originates—an institution of an enduring character which not only survives the early marriage by capture but which is reinforced and probably first made permanent by its substitute, namely, marriage by purchase. In this instance again, Australian custom offers the clearest evidence. In the so-called 'Pirrauru marriage' of Australia, a man, M, possesses a chief wife, C1, called 'Tippamalku.' Another man, N, likewise has a chief wife, C2. This wife, C2, is, however, at the same time a secondary wife, S1, or 'Pirrauru' of M. In like manner the chief wife, C1, may, in turn, be a secondary wife, S2, of N. This is the simplest form of group-marriage. Two men have two wives, of whom one is the chief wife of M and the secondary wife of N, and the other is the chief wife of N and the secondary wife of M. Into such a group yet a third man, O, may occasionally enter with a chief wife, C3, whom he gives to M as a secondary wife, S3, and eventually to N as a secondary wife, S4, without himself participating further in the group. In this way there may well be innumerable different relations. But the marriage is a 'Pirrauru marriage' whenever a man possesses not only a chief wife but also one or more secondary wives who are at the same time the wives of other men. 'Pirrauru marriage' is a
form of group-marriage, for it involves an exchange of women between the men of a group according to the reciprocal relation of chief and secondary wives. The very manner in which 'Pirrauru marriage' originates, however, indicates that in all probability its basis is monogamy, and not, as is supposed by many ethnologists and sociologists, 'promiscuity,' or the total absence of all marriage. In harmony with this interpretation is the fact that in numerous regions of Australia, especially in the northern districts, it is not group-marriage but monogamy that prevails. There is also, of course, a form of group-marriage that differs from 'Pirrauru marriage,' and is apparently simpler. In it, the differences between chief and secondary wives disappear; several men simply possess several wives in common. Because this form of group-marriage is the simpler, it is also usually regarded as the earlier. This view, however, is not susceptible of proof. The supposition rests simply and alone upon the consideration that, if a state of absolutely promiscuous sexual intercourse originally prevailed, the transition to an undifferentiated group-marriage without distinction of chief and secondary wives would be the next stage of development. The reverse, however, would obtain were monogamy the original custom. For the group-marriage with chief and secondary wives is, of course, more similar to monogamy than is undifferentiated group-marriage. Moreover, this order of succession is also in greater consonance with the general laws underlying social changes of this sort. As a matter of fact, it would scarcely be possible to find grounds for a transition from undifferentiated group-marriage to the 'Pirrauru system.' If we assume that there was a growing inclination for single marriage, it would be difficult to understand why the circuitous path of 'Pirrauru marriage' should have been chosen. On the other hand, it is very easy to see that the distinction between chief and secondary wives might gradually disappear. Indeed, this is what has almost universally happened wherever pure polygyny prevails. Wherever polygyny may be traced back to its beginnings, it always seems to have its origin in the combination of a chief wife with several secondary wives. Later, however, when the wife comes to be regarded as property, we find a formal co-ordination of the wives. Or, there may be a distinction that arises from the accidental preference of the husband, as in the case of the Sultan's favourite wife, though in modern times such choice has again been displaced by a law of more ancient tradition. The latter change, however, was the result of the external influence of the culture of Western Europe. Such a retrogressive movement, in the sense of a reapproach to monogamy, is foreign to the motives immanent in the development itself. Furthermore, 'Pirrauru marriage' is very easily explicable by reference to the same condition that best explains the origin of exogamy, namely, the custom of marriage by capture as practised between groups enjoying a tribal or cult relationship. The captured wife is the Tippamalku, or chief wife, of the captor; to the companions who assist the latter she becomes a Pirrauru, or secondary wife. This latter relation is at first only temporary, though it later becomes permanent, probably as a result, in part, of a dearth of women. By rendering his companions a similar service, the original captor in turn gains the chief wives of the former as his secondary wives. As frequently happens, the custom which thus arises outlives the conditions of its origin. This is all the more likely to happen in this case, because the general motives to polyandry and polygyny persist and exercise a constant influence.
Proof that this is the forgotten origin of group-marriage may perhaps be found in a remarkable feature of the customs of these tribes—one that is for the most part regarded as an inexplicable paradox. Marriage with the chief wife is not celebrated by ceremonies or festivals, as is the union with the secondary wife. Thus, the celebration occurs, not in connection with that marriage which is of primary importance even to the Australian, but, on the contrary, on the occasion of the union which is in itself of less importance. The solution of this riddle can lie only in the origin of the two forms of marriage. And, in fact, the two result from radically different causes, if it be true that capture from a friendly clan is the origin of the Tippamalku marriage and that assistance rendered to an allied companion underlies Pirrauru marriage. Capture is an act which precludes all ceremony; alliance with a companion is a contract, perhaps the very first marriage contract that was ever concluded—one that was made, not with the woman or with her parents, but with her husband. The consummation of such a contract, however, is an act which in early times was always accompanied by ceremonial performances. These accompanying phenomena may also, of course, persist long after their source has been lost to memory. Thus, the difference between the two forms of primitive group-marriage also indirectly confirms the supposition that monogamy lies at the basis of group-marriage in general.
After a man has won one or more secondary wives in addition to his chief wife, in Pirrauru marriage, there will doubtless be a tendency for him to seek additional chief wives. This will be particularly apt to occur where, on the one hand, marriage by capture gives way to marriage by barter and later to marriage by purchase, and where, on the other hand, group-marriage is on the wane. Custom may then either recur to monogamy, or it may advance to a polygyny which is pure and not, as in the case of group-marriage, combined with polyandry. Whether the former or the latter will occur, will depend, now that marriage by purchase has become predominant, upon might and property. Since these are also the factors which insure man's supremacy within the family, the older forms of combined polyandry and polygyny almost universally (with few exceptions, conditioned by the dearth of women) give way, with the advance of culture, to simple polygyny, which is then practised alongside of monogamy. This polygyny, in turn, also finally recedes in favour of monogamy. The circle of development, accordingly, may be represented by the following diagram:—
