The KINSHIP system

Abstract

The KINSHIP program is a general computational tool for componential analysis of kinship terminology, supporting approaches from different theoretical persuasions. It can adequately handle the problems outlined in Chapter 2. The program accepts as input the kin terms of a given language with their attendant kin types and produces a componential analysis of the kinship vocabulary. The produced analysis is ‘consistent’, in that all kin terms are assigned definitions with necessary (nonredundant) and sufficient conditions, but the system also allows of producing various degrees of redundancy (‘partially’ or ‘fully’ redundant models) in case the user is for some analytical reason interested in such models. In addition, the system can generate all alternative solutions to kinship systems. The problem of the multiplicity of alternative solutions to kinship systems is approached by introducing three simplicity criteria, two choosing a minimum number of overall features (dimensions) and minimum components in kin term definitions, and a third one minimizing the use of features in definitions and ensuring coherence among kin term definitions within the whole data set. In this chapter, I describe the basic functional capabilities of the program (or what the program does) and the essential algorithm of its operation (or how the program does it).