It would appear, then, that there must have been a time when several of the same totems by descent occurred in both phratries, or, at least, in both the local groups that became phratries. In that case, by phratry law, a Snipe in Eagle Hawk phratry might marry, out of his own phratry, in Crow phratry, a Snipe. By totem law, however, he may not do this. There was thus a clash of laws, as soon as totem law was fully developed, and the totems were therefore deliberately arranged so that one totem never appeared in both phratries. This law made it necessary, when Snipes occurred in both phratries, that some Snipes, say, in Eagle Hawk phratry, must cross over and join the other Snipes in Crow phratry, or vice versa. They obviously could not change their totems, and, of two evils, preferred to change their phratry, the representative of their old local group. Totems were beginning to override and flourish at the expense of phratries, a process in the course of which many phratry names are now of unknown meaning, many phratry names have even ceased to exist (the later matrimonial class names doing all that is needed), and outside of Australia, America, and parts of Melanesia, phratries seem not to be found at all among totemists—(the Melanesians have only rags of totemism left).
But where totems, under male kinship (as among the Arunta), have decayed, phratries, named or nameless (and, where nameless, indicated by the opposed matrimonial classes in Australia), do regulate exogamy still.
Thus the possibility of members of a totem kin changing phratries, as we suppose Eagle Hawk and Crow kins to have done, seems to have been demonstrated by actual fact, by that redistribution of totem kins in the phratries—never the same totem in both phratries—which cannot be due to accident, and is universal, except in the Arunta nation. In that nation the absence of the universal practice has been explained. (Cf. Chapter IV.)
It is clear that the first great change in evolution was the addition to the rule, "No marriage in the local group of animal name," of the rule, "No marriage in the animal name of descent," or totem, the totem being nearer and dearer to a man than his local group name, when that became a phratry name, including several totem kins.
Now that this feeling—to which the totem of the kin was far nearer and dearer than the old local group animal whence the phratry took its name—is a genuine sentiment, can be proved by the evidence of Mr. Howitt, who certainly is not biassed by affection for my theory—his own being contrary. He says: "The class name" (that is, in our terminology, the phratry name) "is general, the totem name is in one sense individual, for it is certainly nearer to the individual than the name of the moiety" (phratry) "of the community to which he belongs."[3] Again, "It is interesting to note that the totems seem to be much nearer to the aborigines, if I may use that expression, than the" (animals of?) "the primary classes," that is, phratries.[4]
As soon as this sentiment prevailed, wherever a clash of laws arose men would change their phratries, rather than change their totems, and we have seen that, to effect the present distribution of totems (never the same totem in each phratry), many persons must have changed their phratries, as did the two whole totem kins of the phratriac names, on my hypothesis. I reached these conclusions before Mr. Howitt informed us of the various dodges by which several tribes now facilitate marriages that are counter to the strict letter of the law.
It seems needless to dwell on the objection that my system "does not account for the fact that phratriac names—say Eagle Hawk, Crow—are commonly found over wide areas, and are not distributed in a way that Mr. Lang's 'casual' origin would explain."[5]
We have seen, though we knew it not when the objection was raised, that the institutions were perhaps in some cases diffused by borrowing, from a centre where Kilpara meant Crow, and Mukwara meant Eagle Hawk; and that these names, and the phratriac institution, reached regions very remote, and tribes in whose language Kilpara and Mukwara have no everyday meaning. If borrowing be rejected, then the names spread with the spread of migration from a given Mukwara-Kilpara centre, and other names for Eagle Hawk and Crow were evolved in everyday life.
Except as regards late "abnormalities," we have now surveyed the whole course of totemic evolution. May it not be said that my theory involves but a small element of conjecture? Man, however he began, was driven, by obvious economic causes, into life in small groups. Being man, he had individual likes and dislikes, involving discrimination of persons and some practical restraints. A sense of female kin and blood kin and milk kin was forced on him by the visible facts of birth, of nursing, of association. His groups undeniably did receive names; mainly animal names, which I show to be usual as group sobriquets in ancient Israel and in later rural societies. These names were peculiarly suitable for silent signalling by gesture language; no others could so easily be signalled silently; none could so easily be represented in pictographs, whether naturalistic or schematised into "geometrical" marks. It is no conjecture that the names exist, and exist in the diffused manner naturally caused by women handing on their names to their offspring, as, under a system of reckoning in the female line, they do to this day. It is no conjecture that the origin of the totem names has long been forgotten.
It is no conjecture that names are believed, by savages, to indicate a mystical rapport, and transcendental connection, between the name and all bearers of the name. It is no conjecture that this rapport is exploited for magical and other purposes. It is no conjecture that myths have been invented to explain the rapport which must, it is held, exist between Emu bird and Emu man, and so in all such cases. It is no conjecture that the myths explain the rapport, usually, as one of blood connection, involving duties and privileges. It is no conjecture that blood is held sacred, especially kindred blood, and that this belief involves exogamy, "No marriage within the blood of the man and the totem." We give reasons for everything, whereas, if a reformatory bisection of a promiscuous horde were made, by an inspired wizard, why did he do it, and why should each moiety take an animal name? Again, if there were no recognised pre-existing connection between human groups and animals, why should one group do magic for one animal, rather than for another, in cases where they do this magic?