We are grateful to Michael Lynch for his very helpful comments and suggestions.

University 0/Biele/eld I_ges are objects ofwork in the laboratory. On its face, tltis work is achieved through talA:. Yet the talle attaclled to diese images _kes reference to other images, whieh are drawn from various environments. In dlis aniere, four such environments are identifred: the domain of",boNItory practice; the COlltut ofinvisible pIIysieal reactiOllS; the future ilfUlge as it will appear in publication; and the domain ofcase precetlems alld reference scenariosfrom lhe {ield. The work ofimage analysis brill8s the outside ofthe image WO it alld takes the inside olll. Thus it can be seen that images are not just laÜn, tIrey are desigMd IiUId made. Philosophers, historians, and soeiologists of science have long considered writing to be a central part of scientific activities. From an outside observer's perspective on science, this conception of scientific writing is plausible enough. Written articles are the ostensible goal and the raison d'etre of much scientific activity. They are the topic of seientific arguments, the materials that integrate scientific communities, and the carriers of scientific information and "progress" in a field. They are the tangible result of transient and undocumented research activities, the palpable "traces" left for analysts of seience to study. Scientific inquiry, from this perspective, is centrally concerned with producing and quaJifying texts. 1 Yet from within scientific inquiry, the focus of many laboratory activities is not texts, but images and displays. The difference is crucial. Images and texts are differently occasioned and differently handled in the process of inquiry. On occasion, images and texts are assembled together to ereate one another's referents-in the process of writing up data for the purpose of publication, for example. Yet the very necessity of their assemblage, at the stage of publication, points to their singular relevance in other contexts, sueh as in inquiry. In a sense, natural AUTHOR'S NOTE: The research for Ihis article hlS been made pouible by a grant from the Deutsche Fon.chungsgemeinschaft for the study of "Complex Knowledge Processes." We are

No abstract

No abstract

No abstract