independent data for the construction and verification of theory, is in fact a very com plicated compound of local realities and the contingencies of metropolitan theory (5:360).A third front has recently been opened in the assault on the edifice of ethnography. Having deconstructed ethnographic form and historicized the ethnographic subject, some have now turned to regionalizing its conceptual claims. Their presumption is that all ethnography is regional, a local transposition of general disciplinary concerns. It must be read critically for the problems it highlights through the mutual adaptation of anthropological dis course and locally prominent features and issues: prestige economy in Melanesia, marriage rules in Australia, lineage in Africa, caste in South Asia. "Localizing strategies" is Fardon's felicitous phrase (81) for the complicated dialectic of region and problematic, which was illustrated so effectively in Abu-Lughod's review (2) of "zones of theory" in Arab world anthropology.Some of the work I consider in this essay may be so analyzed. I Yet I argue IThis review is limited to social-cultural anthropology (for archaeology, see 77, 109; for linguistic anthropology, see 244; for primatology, see 9, 97:244--58). I deal only with English language works. The disciplinary surveys in the periodic but misnamed Introductory Bibliogra phy for Japanese Studies provide helpful English-language reviews of anthropology in Japan (114,197, 261,262,288,289; see also 17,124, 183, and 157). Long (157) has surveyed family sociology.
395