ABSTRACT
This chapter begins by evaluating Herbert Blumer's own account of naturalistic research, before assessing analytic induction, grounded theorizing, and the pattern model of explanation. It emphasizes that Blumer's methodological writings were very much a reaction against growing trends in US sociology away from case-study research. The chapter argues that the essence of analytic induction is the hypothetico-deductive method, the logic of the experiment in the broad sense used by Blumer and Znaniecki. It explores the possibility of resolving Blumer's dilemma by re-defining science or by abandoning our commitment to it. At first sight the argument about the distinctiveness of social phenomena is even less convincing than the argument about Verstehen, since all phenomena, including physical objects and events, are unique. Symbolic interactionism portrays the social world as generated by social interaction among people; interaction that itself produces, and is shaped by, participants' interpretations of the world.
