Again, Dr Lowie's error may be due to a vagueness of statement on my part. The words "objects and symbols which are originally of emotional value for individuals become through their totemic association transformed into social factors, referring to social units which are clearly defined," may be misleading. However, Dr Lowie could not possibly have misunderstood my statements on page 271. "The intimacy of the above associations could never become so absolute if not for the fact that the various elements, religious, aesthetic, ceremonial, and what not, become linked with definite social units (say, the clans), of which they henceforth become the prerogatives and symbols. This association with social units is what constitutes the peculiarity of totemic combinations. Elements which are per se indifferent or vague in their social bearings,2 such as dances, songs, carvings, rituals, names, etc., become associated with clearly defined social groups, and, by virtue of such association, themselves become transformed into social values, not merely intensified in degree, but definite and specific in character." The process is somewhat further elaborated in the following paragraph. Now this transformation into definite social values is what I call specific socialization. I also say: "The one obvious and important means by which the association with definite social groups is accomplished is descent." And my conception of the function of descent in this connection appears from the following sentence: "In clan totemism we start with a social group which in some way has acquired a totem, whether it be a worshipped or tabooed animal or plant, or merely a name [cf. Dr Lowie's own hypothetical instance on page 204 of his review]. Descent becomes henceforth a factor which tends to perpetuate the totemic clan as a social unit, as well as to consolidate it with those other elements which may from time to time become associated with it." And again: "In clan totemism, then, the social group is, for totemic purposes, the starting point."

115

1 Totemism, p. 275.

2 In the original these words are not italicized.

3 Totemism, p. 271.

4 Ibid., p. 271.

Ibid., p. 272.

It is very curious that Dr Lowie represents me as holding that my definition of totemism is based on assumptions such as that the clan names of the Iroquois or Australians designated individuals before they were applied to social groups, etc. This, of course, is the theory of the origin of totemism held by Hill-Tout, only that he starts, not with an individual taboo, or name, but with an individual guardian spirit. Dr Lowie has read my refutation of this theory, of which Frazer's conceptional totemism is a variant. I admit the possibility of such development, although there can be little doubt that, if origins were laid bare, the social group would in a great majority of cases be found at the very inception of the totemic process. One road to totemism may lead over the individual taboo, name, or guardian spirit, but it has not been an oft trodden road.

All this by the way, however, for, as I stated before, my definition of totemism does not involve any theory as to the origin of the institution. The connotations of the term "socialization," as used in the definition, are primarily psychological, not genetic. Dr Lowie seems to agree with the first definition in which the process is described from the point of view of the social units. Now, in the second and third definitions, I merely attempt to express the process in psychological terms, using the "emotional values" as the starting point. No new elements, or concepts, or hypotheses, are added. Social units become associated with objects of emotional value, or the objects become associated with social units, become socialized. As the social units are sharply defined, the socialization is specific. To take Dr Lowie's schematic example. Group A and group B have each certain taboos. The groups combine. Have we totemism? Not necessarily. For the result may be simply a larger group C, some of the members of which observe the taboos of former group A, others the taboos of group B. But A and B may combine while preserving their identity. They may thus become definite social units (say, clans) and the taboos, if practiced by the clans as social units, would then be socialized within the clans, forming the nucleus of a totemic community. 1 Totemism, pp. 268-9.

But there is really no need of such hypothetical constructions, for the term "socialization" is nothing but a description in psychological terms of what we actually find in totemic communities. The totem, as well as the concomitant beliefs, ceremonies, artistic representations, etc., are in totemic groups, always socialized within the social units to which they refer; they are their prerogatives, or symbols. Such a condition can not be regarded as primary; the specific socialization of a belief or practice is, of course, a psychological process in the minds of the individuals constituting the social unit. In the formative period of a totemic complex, this process must proceed for some time (say, several generations) before the new psycho-sociological relation becomes a fixed factor in the social consciousness of the group, although in a developed totemic community the time necessary for the socialization of a new totemic feature may be very brief indeed. In so far, then, as the connection between the socialized object and the social unit, while "in the making," must be conceived as a process, but only in so far, the term "socialization" is not merely psychologically descriptive but also genetic.

I should like to add a few words as to the application of the concept of convergent evolution to totemic phenomena. I think I have shown, as Dr Lowie insists, that totemic complexes must be regarded as the product of convergent evolution. On the other hand, all totemic complexes are genetically determined and psychologically constituted by the fact that the component social units of the complexes become associated with the various totemic features, or that the totemic features become socialized within the limits of the social units. This functional factor in all totemic complexes, whether we call it totemism or not, seems to be a constant. Moreover, it can not itself be conceived as a product of convergent evolution, but seems to be a primary socio-psychological fact.

This interpretation does not militate against the conception of totemic complexes as products of convergent developments. On the contrary, it brings the conception into relief by suggesting that the tendency to specific socialization reduces to a common denomi

nator the heterogeneous ethnic factors that go to the making of a totemic complex, by bringing them into that intimate relation with social units which is so characteristic of totemic communities.

In closing I want to join Dr Lowie in his final estimate of my work. My study was "not definitive, but programmatic." I have merely, "given a statement of first principles . . . The next step must be a more extensive ethnographic investigation of the field."

COLUMBIA UNIVERSITY,

NEW YORK CITY.

1 Totemism, p. 206.

AM. ANTH., N. S, 13-39

BOOK REVIEWS

Primitive Paternity, the Myth of Supernatural Birth in relation to the History of the Family. BY EDWIN SIDNEY HARTLAND. Vol. I (1909), pp. viii+325; vol. II (1910), pp. 328.

British anthropologists write so much and so well, that we, their American colleagues, are kept busy reviewing them. When the author of The Legend of Perseus, one who as folklorist and totemist ranks among the first, writes a book on Primitive Paternity, our attention is aroused and our anticipation kindled.

The author opens his argument by presenting a well-selected set of myths of supernatural birth, i. e., of "birth without sexual intercourse, and as the result of impregnation by means which we now know to be impossible" (I, p. 2). We read stories of impregnation by eating and drinking; of conception through stones or the consumption of a portion of a corpse; of children born from the wind, the rain, or the rays of the sun, etc. Having tasted of legend and myth, we follow the author through a maze of picturesque customs and beliefs which indicate that mythological fancy became reality in the innumerable devices for artificial impregnation which have been used in antiquity and continue to be used by modern savages and peasants (I, pp. 30–155).

The beliefs in supernatural birth or asexual conception are so widespread that they must evidently have existed from the remotest antiquity and must have sprung from some basic characteristic of primitive mentality. This the author finds in the primitive view of nature in which no sharp line was drawn between man, animal, plant, and stone, and transformations from one natural kingdom to another were of every day occurrence. Ninety-seven pages are devoted to a discussion of such beliefs in transformation, where, through death and rebirth or in some other way, man becomes an animal, stone, or plant, or vice versa.

Here the argument takes an unexpected turn, and we find ourselves confronted with the problem of mother-right. "During many ages" says Hartland, "the social organization of mankind would not have necessitated the concentration of thought on the problem of paternity" (I, p. 256). Such a type of social organization is found in mother-right, a state of society which must once have been universal for "the result of anthropological investigations during the past half century has been