an article forthcoming in the British Journal of Sociology (pdf). (table)
Sociologists have increasingly come to recognize that the discipline has unduly privileged textual representations, but efforts to incorporate visual and other media are still only in their beginning. This paper develops an analysis of the ways objects of knowledge are translated in research into other media, in order to understand the visual practices of sociology and to point out unused possibilities. I argue that the discourse on visual sociology, by assuming that photographs are less objective than text, is based on an asymmetric media-determinism and on a misleading notion of objectivity. Instead, I suggest to analyze media with the concept of translations. I introduce several kinds of translations, most centrally the one between tight and loose ones. I show that many sciences, such as biology, focus on tight translations, using a variety of media and manipulating both research objects and representations. Sociology, in contrast, uses both tight and loose translations, but uses the latter only for texts. For visuals, sociology restricts itself to what I call “the documentary”: focussing on mechanical recording technologies without manipulating either the object of research or the representation. I conclude by discussing three rare examples of what is largely excluded in sociology: visual loose translations, visual tight translations based on non-mechanical recording technologies, and visual tight translations based on mechanical recording technologies that include the manipulation of both object and representation.