Page:Folk-lore - A Quarterly Review. Volume 14, 1903.djvu/470

 428 Reviews.

relations, to co-ordinate them or to subordinate them one to another. All classification implies a hierarchic order, of which neither the sensible world nor our consciousness offers us the model. The very expressions we make use of to characterise it authorise us in presuming that all these logical notions are of extra-logical origin. We speak of the " relationship " between the species of a given genus. We call certain classes "families." Nay, the word genus itself designated first of all a family group (yeVov). These facts tend to the conjecture that the scheme of classification is not a spontaneous product of the abstract understanding, but the result of an elaboration into which all sorts of strange elements have entered.

The object of the paper is not to treat in its entirety the ques- tion, what has led men to dispose their ideas in this form and whence they could have found the plan of this remarkable dis- position, but simply to bring together a certain number of indica- tions which may throw some light on it. The only way to answer the question fully is to seek the most rudimentary classifications adopted by mankind, in order to see out of what elements they have been constructed. All it is proposed to do is to bring together in the pages following those I have thus summarised a certain number of classifications which are certainly very primitive, and the signification of which does not appear doubtful.

The systems of classification analysed and compared in pur- suance of this intention are those of the Australians, the Zuni, and the Chinese. It is not possible here to follow the authors into the details of their lengthy exposition ; but an outline may be given of their argument.

Among the Australians each tribe is divided into two great fundamental sections, which we call phratries. Each phratry in its turn includes a certain number of clans, that is, groups of individuals who bear the same totem. Save in rare cases, the totems of one phratry are not found in the other. Beside this division into clans each phratry is divided into two matrimonial classes, or classes having the right of connubium with one another, each composed of a certain number of clans. The point insisted on by the authors is that this classification is not confined to the members of the tribe, but extends to the universe at large. All the objects of the universe are divided between these phratries and clans; all belong either to one or another. The classification