ABSTRACT
A central problem in the sociology of knowledge has been to show that sane people can intelligibly have quite different alternative understandings of the same problem, such as a kind of risk, without abandoning the idea that there is a real problem about which to disagree, and to show the social basis of both plurality and viability. In recent decades, attempts to make this problem tractable have focused on the idea of a 'frame'. Theories of frames offer accounts of the range of content, as distinguished from theories of processes of diffusion, of which risk amplification theory is the best known example. In this article, several theories of frames – those of Goffman, D'Andrade, Moscovici, Gamson, Schön and Rein, and of prospect theory – found are to be inadequate, because of their lack of clarity and plausibility in their answers to four key questions: 'what is the relationship between sense-making and bias?', 'how are frames to be individuated?', 'where do frames come from?', and 'how far and how can people move between frames?'. The article makes the case for a neo-Durkheimian institutional theory developed by Douglas and others. This approach derives frames as concrete applications to specific contexts from thought styles, which are in turn the product of solidarities or institutional styles of social organization, because it can offer clear, testable, parsimonious hypotheses with which to answer these four questions. The theory therefore provides an account of the institutional logic of framing, and presents reasons for preferring this to non-institutional approaches such as the various kinds of cognitivism. The article offers three conceptual innovations with which to develop the neo-Durkheimian theory, in order better to deal with the crucial fourth question about the scope for mobility between frames. These innovations and some specific hypotheses about the scope for mobility between frames are supported by consideration of some exploratory qualitative empirical research on privacy risk perception. The theory provides a more satisfactory strategy for tackling the core problem than most others, by showing plurality to be limited, by showing clear and specific social bases for plurality of frames, by neither wholly endorsing nor wholly rejecting any basic bias, and by showing that their conflictual and systemic interdependence is what makes for viability.
