ABSTRACT
The practical privilege in which all scientific activity arises never more subtly governs that activity (insofar as science presupposes not only an epistemological break but also a social separation) than when, unrecognised as privilege, it leads to an implicit theory of practice which is the corollary of neglect of the social conditions in which science is possible. The social world may be the object of three modes of theoretical knowledge, each of which implies a set of (usually tacit) anthropological theses. The language of rules and models, which seems tolerable when applied to "alien" practices, ceases to convince as soon as one considers the practical mastery of the symbolism of social interaction-tact, dexterity, or savoir-faire-presupposed by the most everyday games of sociability and accompanied by the application of a spontaneous semiology, i.e. a mass of precepts, formulae, and Codified cues.
