AMAL SANYAL^ Choice, Welfare and Measurement WHERE does one place man in a model of the society—this so-called "problem of man59—is one of the most mishandled questions in "mainstream55 economic theory. The individual agent at the centre of the analysis makes his choice according to his preference and the totality of such choices comprises an alleged description of the economy. This format for the analysis of social behaviour describes the motivation of choice as arising within the individual, ignoring the dialectics of the interaction between the numerous agents through their choices, actions and the results thereof. The totality of these latter types of interactions and interrelationships however are of the essence of the choice of a social man. To be sure, the individual's conception of his welfare has been sometimes modelled as altruistic, where the individual's welfare is taken to depend on others' welfare as well, but the choice of the agent still remains shackled to a motivation arising from within himself, i e, his perception regarding his own welfare. This mechanical homocentrism is certainly one of the major weaknesses of the conceptualisation of the social man by "mainstream" economic theory. Amartya Sen's persuasive critique1 of the mistaken identification of preference with choice in economic theory is a most welcome development in this context, since his critique is based on the recognition of the important interaction between the choice of the agent and his economic environment. In his critique spread over a number of essays of this volume, Prof Sen first argues that the theory of Revealed Preference is based on the correspondence between choice and preference, despite the usual contention that it has banished the idea of preference from economic theory. Then he argues that while it is possible to construct a ranking for an individual over commodities from his observed choices, to take the former as his welfare ranking is an unjustified step. He shows, using situations like a Prisonners' Dilemma, that the preference revealed under such a situation will not correspond to individuals' ranking of their own welfare. Indeed, their welfare rankings would be correctly revealed only if they drastically deviate from behaviour based on their own welfare consideration. The consequence of this demonstration, that the study of choice in general will not lead to an unveiling * Centre for Economic Studies and Planning, Jawaharlal Nehru University, New Delhi.