Jeffery Paige is right on many points, but not on all, and not, I think, by reason of the data that appear in the tables of his paper, "Kinship and Polity in Stateless Societies." Paige argues that, in societies in which "group ties are overwhelmingly based on kinship," descent and residence rules determine whether the polity will be "factional" or "communal." He follows Schneider (1961) and Goodenough (1956) in noting that matrilineal descent brings together in a single family males from different kinship groups, thereby making it more difficult for a man to support the interests of his own kinsmen in opposition to those of others. (As Nicholas [1965] and Schlegel [1972] show, this is a somewhat idealized view of matriliny.) Paige then observes that a patrilineal rule of descent keeps males from the same kinship group together, thereby making it easy for them to unite in pursuing their joint interests in opposition to the interests of others. It is in this way that a matrilineal rule promotes a communal polity, whereas patriliny promotes factionalism. Tables 1 and 2 of Paige's paper follow similar tabulations by Aberle (1961) and verify certain of the points just made. Table 1 shows that societies having a patrilineal rule of descent are more likely than matrilineal societies to have local communities in which all males belong to the same kinship group in Paige's term, the same "lineage." Table 2 shows that, among societies in which males in given communities belong to different lineages, the societies having a patrilineal rule of descent are more likely than those having a matrilineal rule to have patrilineal polygynous families, a form of marriage that Paige believes can easily operate as an "interest group."2 There seems little doubt that the associations found in tables 1 and 2 are valid. We need, however, to understand that, in themselves, they simply verify that matrilineal and patrilineal patterns are as the ethnographers have defined them. The results are therefore more a demonstration of construct validity than evidence for or against any theory about the relations in these societies between kinship and polity. Paige regards table 3 as more important for the theory he advances. He writes: "It is clear that descent rules create distinct patterns of kinship organization. The relationship between descent and polity type, however, depends on the political behavior of these kinship groupings. Propo-