ABSTRACT
Is modern social anthropology engaged in a retreat from empirical reality? We are concerned with 'deep structure' rather than with content; with models rather than with behaviour; with symbols rather than with customs. Leaving aside the question of how far this is a reaction to fundamental changes in the degree to which men feel they can control their social environment, I think at least three elements can be identified in this anthropological focus. There is a search for abstraction as a heuristic device, to eliminate idiosyncratic variation and perceive general principles more easily. There is the recognition of observer-effect, by which what purports to be a record of fact is admitted to be a record of opinion, or at least to incorporate some quantum of personal interpretation. There is also a challenge to positivism, whereby from a relatively neutral position of criticism or from some specific commitment the autonomy, even priority, of the non-empirical is insisted upon. In all this the language of symbolism, and inquiry into the nature of symbolism, have come to the forefront.
