But this explanation cannot apply to the Tlingit, with female descent, for with female descent, unless the men go to the women's homes, no local clan of descent is possible. I have shown that I do not pretend to know precisely what are the facts of the Tlingit system, as accounts contradict each other. But in other American cases, as in those of the Apaches and Navahos, the tribes "are divided into a large number of exogamous clans with descent in the female line, but the names of the clans appear to be local, not totemic...."[59] Such names are Lone Tree, Red Flat, House of the Cliffs, Bend in a Canyon, and so forth. Are such names inherited? Is every child of a woman of Red Flat called "Red Flat"? Persons of the same clan or phratry (from eight to twelve phratries) may not intermarry. The phratries "have no formal names"; speaking of his phratry a man will often refer to it by the title of its oldest or most numerous clan—and that, it seems, is always a local name, "Dr. Washington Matthews," says Mr. Frazer, "who spoke with authority on the subject, was of opinion that the Navahos clans were originally and indeed till quite recently local exogamous groups and not true clans." What else can they be? But Dr. Washington Matthews found a legend which suggests that the Navahos were once totemic. If this be an explanatory myth its point is to explain why the clans have now local names, and why do the clans think that the fact needs explanation? " It is said that when they set out on their journey each clan was provided with a different pet, such as a bear, a puma, a deer, a snake, and a porcupine, and that when the clans received their local names these pets were set free."[60] That is, place-names ousted totem names.
It appears to me that when a tribe acquires settled habits and lives in villages, territorial names may oust totem names, and exogamy may become, as among the Navaho, local, just as it becomes local in several Australian tribes with male descent. But nothing in my theory compels me to suppose that every people has passed through totemic exogamy. Exogamy, in my view, was prior to totemism; totem names were a later way of designating local groups which were already exogamous.[61] "The rule would be, No marriage within the local group." The totemic names were a later addition, and I can think of no reason why all peoples should necessarily accept totemic names; only, as it chances, the enormous majority among the lower races have done so.
Perhaps the Navaho and Apaches never had totemic names for their exogamous local groups. They are not known to exhibit any sign or vestige of totemism beyond the legend or myth of the wild animal pets.
All such cases of exogamous units bearing non-totemic names, in tribes of female descent, where no vestige of totemism is found, are outside of the field of totemism. Why should we treat people as totemic who have no totems? If we held the opinion that totemism was the cause of exogamy, the position would be different. At one time I thought that the totem and the totem blood taboo, clinched, as it were, and sanctified a pre-existing exogamy. But as I never found that marriage within the totem was automatically punished by sickness or death; (as, in many tribes, the offence of eating the totem is supposed to be); I saw that marriage within the totem was a breach of secular law, punished capitally by "the State." There is no taboo in the case. But as we repudiate the opinion that totemism was the cause of exogamy, in studying totemism we have no concern with peoples who are exogamous but show no trace of having ever been totemic.
[59] T. and E., vol. iii, p. 243.
[60] T. and E. vol. iii. p. 245. note 5, citing Washington Matthews. J. A. F. iii., 1890, p. 105, and Navaho Legends, p. 31, 1897.
[61] Secret of the Totem, pp. 114, 115.
VIII.
The case of the Tlingit is quite different. Here the phratries have totemic names; the "clans" in the phratries are said, by early authorities, to have totemic names; the "crests" (mainly the same animals as those said to give names to the Tlingit "clans") are readily to be explained by totemism evolving into heraldry.