CLASSIFICATION OF KINSHIP TERMINOLOGIES The foundation of the scientific study of this subject was laid by Lewis H. Morgan in his Systems of Consanguinity and Affinity of the Human Family (Smithsonian Contributions to Knowledge, XVII., Washington, 1871). In this great work were assembled comprehensive schedules for every major area of the globe except Africa and Australia, for which data were largely or wholly inac cessible to the author. Morgan grouped all terminologies under two main heads—the descriptive and the classificatory. The de scriptive system, ascribed to the Aryan, Semitic and Uralian families, "describes collateral consanguinei, for the most part, by an augmentation or combination of the primary terms of rela tionship" (pp. 12, 468). The classificatory system never describes consanguinei by a combination of primary terms but ranges them into great classes or categories : "All the individuals of the same class are admitted into one and the same relationship, and the same special term is applied indiscriminately to each and all of them" (p. 143). Two varieties of the classificatory type were recog nized—the Malayan, which merged all kindred of one generation irrespective of proximity in one category; and the Turanian Ganowanian, in which only some of the collateral kin of a genera tion were merged with the lineal, the remainder being separated by distinctive terms.
This dichotomy is unacceptable. As Kroeber' and Rivers' re mark, our words for uncle, aunt and cousin likewise range con sanguinei into large classes, so that the absoluteness of the dis tinction breaks down. Furthermore, Rivers notes' that "descrip tive" does not strictly apply to the English and related terminol ogies, which rarely augment or combine the primary terms. That epithet may be properly reserved for such Norwegian words as farbror and morbror for father's and mother's brother, respec tively; and hence for certain nomenclatures, e.g., the Arabic, which largely employ expressions of this order. Thus, Rivers supplanted
Morgan's dualism with a tripartite scheme, recognizing denotative, descriptive and classificatory types. Elsewhere he suggests the L. Kroeber, "Classificatory Systems of Relationship," Journal Royal Anthrop. lnstit., 5909, 39: 'W. H. R. Rivers, Social Organization, 1924, article "Kin, Kinship" in Hastings' Encyclopaedia of Religion and Ethics.
appellations family, kindred and clan systems, the "kindred" being a major bilateral unit.' These latter terms are, however, objec tionable since they inject the inferential basis of a system into its descriptive definition. There are clanless tribes with clan systems in Rivers' sense, and the result would be confusion. Apart from this terminological difficulty, the scheme fails to obviate a logical fault in Morgan's earlier effort. "Descriptive" and "classifica tory" do not relate to the same logical universe, hence are not complementary terms. "Classificatory" relates to the number of individuals defined, "descriptive" to the technique of defining a relationship. It is possible to augment or combine primary terms and then apply the resultant to an indefinitely large class of persons. The antithesis to "classificatory" is the concept "individu alizing," as Morgan felt, though he failed to express it.
But the purging of the traditional terminology does not carry us very far. How shall we characterize our simple English system if it embodies both individualizing and classificatory principles? The answer is provided in Kroeber's above-cited essay : a kinship terminology is not a logically coherent whole but must be resolved into the several categories recognized. Those enumerated by Kroeber are : the difference of generations; the difference between lineal and collateral kin; the difference of age within one genera tion; sex; the speaker's sex; the connecting relative's sex; the difference between consanguinity and affinity; and the life or death of the connecting relative.