The papers in this volume present a wide range and impressive quantity of new empirical data on how people talk and on the rules which govern that talk and its interpretation. Authors have used their data in several different ways, variously (1) organizing descriptive frames from which testable propositions can be derived; (2) creating theoretical frames which facilitate the search for new data sets; (3) actually testing theories. In this connection I would like to elaborate somewhat on the theme of the interaction of data and theory – or more specifically, the uses of societally or culturally specific social behaviors in moving toward identification of behaviors (and structural features constraining those behaviors) characteristic of men in all societies. I will do this by remarking on: (1) sources of data (and their dangers); (2) some modes of organizing data in our search for descriptive adequacy, and (3) reasons for formalization and possibilities for theory which has explanatory adequacy. While I will not always make it explicit, all of what I say should be seen against a comparative backdrop – the ‘theys’ of specific ethnographic reports and the ‘us’ of our own society combined into a global ‘us’ – social man wherever he is found.

Data

Concern about the validity of informant reports is central to the ethnographic enterprise. Labov (1972b) recently chastised a group of linguists about the dangers of circularity and reflexivity in the theory-informed intuition-theory confirmation cycle.