ABSTRACT
Herbert Blumer has exercised a considerable influence on symbolic interactionist qualitative research in the United States. His contrast between definitive and sensitising concepts is frequently quoted to justify the use of sensitising concepts in qualitative work. The alternative solution to the problem of concept-formation is therefore to develop concepts which are both more adequate theoretically and also more faithful to the empirical social world than definitive concepts. There is constant interplay between the observation of realities and the formation of concepts, between research and theorising, between perception and explanation. An interesting parallel approach to the problem of concept-formation is that of Paul Lazarsfeld, who was intensely interested in problems in systematically analysing qualitative data. Sensitising concepts, analytic induction and grounded theory all do considerably less than justice to the process of double-fitting which goes on in the process of developing concepts out of case-study material.
