Abstract Houses and unilineal descent groups have been treated as different types of social phenomena in socio-cultural anthropology, and as borrowed for analysis of households and settlements in archaeology. This paper contends that houses and lineages, especially those configured by Crow–Omaha kinship terminologies, are better considered as perspectival variants, reflecting differences that are fundamentally synchronic versus diachronic. Crow–Omaha systems and house societies exhibit signal similarities, occupying an intermediate status between kin-based and class-based formations, and evidently derive in an evolutionary sense from prior ‘Iroquois’ or ‘Dravidian’ forms. Setting out the terms in which kinship systems should be considered if they are to serve as useful explanatory analogues for archaeological analysis, the paper then proceeds to examine Lévi-Strauss’s original inspiration for the ‘house’, i.e. societies of the Pacific Northwest Coast of North America. It is no coincidence, the present paper contends, that Kwakwa̲ka̲’wakw, the archetypal house society is situated adjacent to a Crow-matrilineal series of communities that share a great deal in common with it culturally, as a result of centuries of exchange. In short, the house needs to re-attend to kinship structures, as descent groups need to be reconnected with exchange structures and alliance processes earlier elaborated by Lévi-Strauss.
Peter M. Whiteley (Tue,) studied this question.