66 SOCIAL SCIENTIST mere objects of study, continuing to be unaware of how their responses to questionnaires can be used to know and control them better. With the information extractive tendency of the academia becoming more pronounced, it is the oppressed themselves, rather than the problems of the oppressed which become the focus of study. The idea of a discussion on AR arose out of the dissatisfaction with existing theories and methods, not only of research but of understanding atjieast some aspects of the rather complex social structure that we have to encounter everyday. While it was recognized that AR could be looked at from the point of view of either action or research, it was decided to confine the discussion primarily to problems of the latter variety. To this end a wide spectrum of individuals and institutions were requested to participate in an initial seminar on the problem. For purposes of convenience the issues under discussion were clubbed together under four categories: Conceptual Problem: This involved not only trying to define the basic idiom, as well as the cut-off points from just action or just research, but tackling issues like the validity of such a method, the ideological premises under which it operates, the problems posed by the 'insider informant method5, and so on. Strategies of intervention: This related to questions such as: What are the various techniques of intervention sought/attempted?-How does one decide the nature of the entry point? What is the methodology of appraisal? How does one decide when (and if) to withdraw? Implications of intervention: The problems under this category get posed at three levels—with the researcher, with her/his institution, with the group. The problems are not merely technical, or legal, or political. Very often they stray into ethical grounds. The learning mechanism: The questions dealt with here were of the following kind: What is the feedback from such a programme? How is the social data generated through such a process used? How much of it goes back into refining teaching/research/action, institutional reform, reforms of professional training and so on? The papers presented were basically of two kinds: a) Those dealing with conceptual problems; and b) Reports on experiments. In what follows we deal with the substantial issues raised during the process of discussions. On Questions of Method Discussions of AR inevitably tend to start from the primal