This paper suggests that lessons from the field of environmental ethics and sociological perspectives on knowledge are important tools for rethinking what type of ethical analysis is needed for building up further the field of development ethics and, more generally, for addressing some of the most fundamental ethical problems related to global poverty and development. The paper argues for a methodologically pragmatist approach to development ethics that focuses on the interplay between facts, values, concepts and practices. It views development ethics (...) as a hybrid between a public moral?political philosophy and a public conception of social science. Ethical analyses of poverty and development must lead to fundamental changes in the ways knowledge is produced and justified and must challenge the dominance of global institutions and orthodox economics as the single sources of expert knowledge for development. Two of the main tasks of ethical analysis is to provide tools for the formulation of alternative knowledge for development centred on the equal moral worth of all human beings, and to influence global policy making as well as multilateral institutions' goals and policies. The last section of the paper argues that Amartya Sen's version of the Capability Approach is already methodologically pragmatist and points to some convergence between Sen's work and John Dewey's. Further sociological and methodologically pragmatist analysis of the approach is needed to assess the ways in which it is influencing debates on human development and leading to policy changes, and the possible distortions it suffers when adopted by multilateral agencies and policy makers. An earlier version of this paper was presented at the international conference on Ethics and Development at Michigan State University, East Landing (USA) 8?15 April 2005. The section on methodological pragmatism draws from an unpublished paper written with Andrew Light called ?A Pragmatist Methodology for Development Ethics,? presented at the American Philosophical Association (APA) meeting in New York, December 1999. I have updated, transformed and used some parts and insights developed with Light in a way he may not recognize. (shrink)