‘I don't know whether I need a further level of disaster’: shifting media of sociology in the sandbox

If the media of research figure in the constitution of the sociological phenomenon, how is it possible to find out how they do so?1 Drawing on Garfinkel's idea of breaching experiments, we propose such an experiment to unearth the role of the media of sociology. The breaching experiment consists in an experimental setup to produce new disaster scenarios and accompanying forms of emergency provision. In the experiment, research subjects are asked to play in a sandbox with animal figures and other props to simulate disasters. The subjects are first asked to ‘build a world’, then to ‘turn the world upside down’, and finally to find an ‘emergency provision that would change the course of the disaster’. These plays are recorded with a purpose-built computer program and photographed and then transformed into fables and emergency provisions. The experiment breaks with three assumptions of media-use in sociology: First, sociologists use media exclusively for description, not creation of worlds. Second, sociologis...