Beschreibung:
<jats:p>This paper argues that both the traditional normative models and the more refined (`continuity') models of public communication of science fail to account adequately for cases of `deviation', i.e. those cases when scientists address the public directly by skipping the usual stages of scientific communication. It is hypothesized that most of such cases are related to crisis situations and to the definition of scientific boundaries. Therefore, at least two modalities of public communication of science should be distinguished: one is the routine, generally unproblematic itinerary of a scientific idea through the different levels of communication as presented by the continuity models; and the other is the dramatic (re)assessment of boundaries and professional competences in the public arena that is required by marginal situations. Continuity models need to be supplemented by a multilevel, multivariate model which enables us to explain this second modality and to understand in more detail the role that the level of public communication plays when such a modality is activated.</jats:p>