Abstract

Arguments, suggested by readings of Durkheim and Kroeber, for the integrity and autonomy of social theory are examined. These arguments may be construed as closure arguments on domains of social events and of social facts. Causal closure, ontic closure, and several kinds of nomic and explanatory closure are distinguished. Discussion of the relations of various kinds of closure, integrity, autonomy, etc. under plausible assumptions concerning causation and explanation leads to the conclusion that one main strand of the integrity arguments is defensible; special ontological assumptions are not necessary and are dubiously sufficient for autonomy. This general conclusion accords with the positions of the later Kroeber and of D. Kaplan, that integrity-autonomy is best considered a methodological, not an ontological issue--a matter of distinct levels of description and explanation, not distinct levels of reality