Anthropologists are interested in reconstructing the kinship system as it might have existed at the time of a proto-language. A kinship system can be regarded as composed of two correlated systems: a system of kinship terminology and a set of behaviors that are patterned in relation to the terminological system. We know of no way in which we can rigorously infer the kinds of behavior directly, but it is generally regarded as possible to reconstruct the terminology at least in part. If rigor can be introduced in the procedure of reconstructing kinship terminology, then a generalization of that rigorous procedure is lexical reconstruction.

Isidore Dyen, David F. Aberle (1974), Lexical Reconstruction: The Case of the Proto-Athapaskan Kinship System. p. 7