This paper considers the relevance of the sociology of science for the study of power. Though there is by no means complete agreement between sociologists of science, recent work in this area has suggested that (a) scientists negotiate not only about scientific but also social reality, and (b) the distinction between macro- and micro-sociology is an impediment rather than an aid to analysis. Thus, though there are indeed differences in scale, il is argued that these should be seen as the outcome of differentially effective attempts by scientists to impose versions of scientific and social reality.

The present paper extends this argument by considering the way in which a set of pharmacological experiments was undertaken in order to generate results and control aspects of the scientific and social environment. It is suggested that the experimentalist acted like an entrepreneur, combining a variety of potentially unruly resources with the aim of simplifying these and reducing them to docile figures on a sheet of paper. The strategies and materials used in this process of control are considered and three classes of potential resources are identified: natural objects or devices, people and inscriptions. It is argued that these have certain properties that render them relatively durable and transportable and hence convenient for the purpose of long distance social control.