This text examines the problems of concept formation in the social sciences. Beginning with a discussion of positivistic, hermeneutic, rationalist and realistic philosophies of science, it argues that realism is best able to furnish rational criteria for the choice and specification of social scientific concepts.