Componential Analysis of Kin Terms - Some Problems and their Solutions

Componential analysis of kinship vocabulary has a long tradition, but a look at the practicing of the method reveals two basic problems. The first pertains to ensuring “consistent” kin term definitions (i.e., definitions with necessary and sufficient components), and the second to the discovery and handling of multiple solutions to kinship systems. The paper introduces a computer program implemented to handle these two problems, and in general, designed to study kinship vocabulary. The program guarantees consistent componential models and the discovery of all alternative models, as well as introduces simplicity constraints to reduce the usually huge number of alternatives to a unique or just a couple of models. We illustrate the problems with examples from the literature and apply the program to a complex kinship vocabulary (Serbo-Croatian) to show its operation.