Content area
Full Text
THE MOTHER-IN-LAW TABOO1
In this article a hypothesis is developed and tested that asserts that avoidance between son-in-law and mother-in-law is associated with culturally expected economic interaction between these relatives. On the grounds of both the correlation that was found and an analysis of the participants' views on son-in-law/mother-in-law avoidance, the custom is interpreted as a device for distinguishing the son-in-law/mother-in-law relationship from the husband-wife relationship in societies where these relationships tend to be similar as far as their economic aspect is concerned. The final part of the article seeks to identify the conditions that give rise to economic interaction between son-in-law and mother-in-law. (Avoidance, mother-in-law, sexual division of labor, extended households, matrilocality)
Various hypotheses have been propounded to explain avoidance behavior between mother-in-law and son-in-law (hereafter M-S avoidance). Tylor (1889:24648) claims that the custom is associated with matrilocal residence; Stephens and D'Andrade (1962:124-50, 213-26) assert that the custom is significantly correlated with a long postpartum sex taboo; and according to Sweetser (1966) the custom tends to occur in societies where a strong emphasis on unilineal affiliation goes together with the occurrence of fragmented and impermanent residential family groups. Subsequent research has failed to support any of these hypotheses. The association found by Tylor did not appear in tests performed by Stephens and D'Andrade (1962:147-48), Jorgensen (1966:165-66), Sweetser (1966:309-10), and Witkowski (1972:244-49).2 Stephens and D'Andrade's (1962) hypothesis failed to be supported by Sweetser's (1966:310-13) replications of their test, and Sweetser's assertion that the rate of parent-in-law avoidance significantly decreases with extended-family organization was not corroborated by Witkowski's (1972:249-50) retesting. Only Sweetser's view that parent-in-law avoidance goes together with a strong emphasis on unilineal affiliation is to a certain extent supported by Stephens and D'Andrade's (1962:147) finding that there is a positive correlation between kin avoidances and unilineal descent, and Jorgensen's (1966:165-66) finding that M-S avoidance is significantly correlated with matrilineal descent.
The remaining explanations fail to account for the numerous cases where avoidance is absent. Cases in point are Eggan's (1955:75-77) and RadcliffeBrown's (1950:54-60) view that kin avoidances serve to minimize the possibility of conflict between kinsmen, and Frazer's (1910:i, 503, iii, 112-13, iv, 108-09; 1913:75, 84-85) view that kin avoidances serve to prevent sexual contact between certain categories of relatives. Murdock (1949:273-74,...