Key points are not available for this paper at this time.
Social studies of science have ascribed a central rôle to trust in the constitution of knowledge, yet there are few studies of its operation in the large, interorganizational structures increasingly required for scientific work. We begin an examination of 53 collaborations in physics and related sciences with two unexpected findings: (1) trust is no higher in projects formed through pre-existing relationships than those without such ties; and (2) there is no relationship between trust and performance. Why, then, is trust viewed as important? Because trust is inversely associated with conflict. In the second part of this paper, three axes of conflict are described, as well as their sources in the interdependencies of collaborative projects. More important than trust for an understanding of large scientific collaborations is the organization of interaction between structural components such as research teams. In the third part, we examine participant accounts of performance. Collaborations that experience uncertainties in resource acquisition are more likely to be viewed as successful than those formed under more routine conditions. In conclusion, we suggest that the segmentation of scientific collaborations through bureaucratic organization can impose a structure for interaction resembling work that is actually noncollaborative.
Shrum et al. (Mon,) studied this question.