December 8, 1916] 



SCIENCE 



827 



the peculiar — one is tempted to say reckless- 

 manner in which he applies the principle of 

 the diffusion of culture in the second part of 

 his theoretical argument. It is true, the 

 author is not guilty of that mechanical hand- 

 ling of cultural features, like units of a phys- 

 ical mixture, which is so characteristic of 

 Graebner's procedure. Dr. Rivers gives due 

 weight to the psychological aspects of culture 

 contact; he emphasizes, for instance, the ob- 

 servation that the very circumstances of the 

 contact of two cultures may give rise to 

 features foreign to both cultures before con- 

 tact. He also devotes an entire chapter, 

 perhaps the most valuable part of the volume, 

 to an ordered consideration of the mechan- 

 isms and conditions, physical as well as psy- 

 ' chical, of the diffusion of culture. But for 

 all that the glaring unreality of the author's 

 method remains the most striking feature of 

 his book. Deliberately evading any attempt 

 to furnish proof of diffusion in specific in- 

 stances, the author erects a purely hypothet- 

 ical structure, based on a bewildering maze 

 of assumptions invariably favoring interpre- 

 tations through diffusion while disregarding 

 alternative interpretations. In the discussion 

 of the secret societies of Mota, for instance, 

 the author ascribes the secrecy of the socie- 

 ties, their multiplicity, as well as their grad- 

 ing in rank, to the fact that the societies were 

 introduced by an immigrant people; they were 

 secret because an open ritual in the presence 

 of a hostile indigenous population (at an- 

 other stage in the argument the population is 

 assumed to be friendly to the newcomers) was 

 dangerous; they were numerous because a 

 constant stream of applicants for membership 

 from the natives led to the formation of new 

 societies; they were graded as to rank because 

 a line had to be drawn between a society 

 wholly of immigrant origin and one into 

 which natives had already been admitted, and 

 so on. Now, it is a well-known fact that re- 

 ligious societies such as those of Mota, 

 whether they belong to other parts of Mela- 

 nesia, to West Africa or to North America, 

 are very commonly secret, multiple and 

 ranked. No ground is found, in other places, 



to ascribe them to an immigrant people. 

 Why, then, in Mota? The author is, indeed, 

 aware of this circumstance. He admits the 

 possibility of an alternative interpretation, 

 but he rejects it in favor of his own, and 

 proceeds with his argument (II., 213). Simi- 

 larly, when discussing decorative art the 

 author chooses to neglect the psychologic- 

 ally plausible and experientially verified tend- 

 ency of designs to pass progressively from 

 realistic forms to geometrical ones or of 

 geometrical designs to become elaborated and 

 often transformed through the addition of 

 realistic appendages. For Dr. Rivers con- 

 ventionalization is a factor " depending on the 

 blending of peoples and of their cultures." 

 By conventionalization he means " essentially 

 a process by which a form of artistic ex- 

 pression introduced into a new home becomes 

 modified through the influence of the conven- 

 tions and long-established technique of the 

 people among whom the new notions are in- 

 troduced" (II., 383). The most striking in- 

 stance of such procedure is perhaps the case 

 of language, where the author ascribes the 

 presence of the inclusive and exclusive plural 

 to the necessity of differentiating between two 

 classes of society. An inclusive and exclusive 

 plural as well as dual occur, for instance, in 

 quite a number of American Indian languages. 

 In these instances Dr. Rivers himself would 

 probably not find a sociological interpretation 

 necessary. Why then so radical an assump- 

 tion in the case of Melanesia, unless indeed it 

 can be made something more than a mere 

 assumption? An examination of several of 

 the features used by the author as tests of 

 his theory shows with great clearness how 

 easy as well as futile it is to advance an in- 

 terpretation of the facts through diffusion, 

 unless proof can be furnished. We note a set 

 of dual features : the sacred and the profane ; 

 higher and lower grades; chiefs and com- 

 moners; geometric and realistic designs; two 

 communities with products to exchange; in- 

 clusive and exclusive plural; maternal and 

 paternal descent; religion and magic. Now, 

 it occurs at once that numerous instances 

 could be cited where one or more of the 



