Structures of Scientific Collaboration

  • Wesley Shrum,
  • Joel Genuth &
  • Ivan Chompalov
MIT Press: 2007. 296 pp. $35, £22.95 0262195593 9780262195591 | ISBN: 0-262-19559-3

As scientists across the world work on ever more ambitious projects, they are collaborating more and on larger scales. How does that collaboration emerge and develop? And how does the nature of the collaboration affect the success of a project?

Scientists often have a narrow view of collaborations, based on their own experiences. Most studies of collaborations have also focused on individual cases. Structures of Scientific Collaboration by Wesley Shrum, Joel Genuth and Ivan Chompalov surveys 53 collaborative projects and reveals remarkable diversity. The authors analysed oral histories collected in the 1990s for the American Institute of Physics and covering a range of fields including geophysics, particle physics and space science, primarily in the United States.

They developed a quantitative approach using statistical and organizational concepts. They shy away from extended case studies, as it can be difficult to generalize from them, but their studies retain some individual character, to give far more than a dry bibliometric analysis of who did what with whom. This approach helps them provide an enlightening, easy to read and sometimes surprising view of how collaborations work.

Credit: A. MARTIN

Trust, conflict and performance are crucial factors. One counter-intuitive finding is that trust does not correlate with performance. It is widely assumed that trust is required to achieve success, but this may stem from the fact that lower levels of trust tend to result in higher levels of conflict. Conflict is unpleasant for individual scientists but is not necessarily a barrier to progress. In fact, many collaborations inherently tend to protect against damage from conflict between members.

There are several types of organizational structure, characterized mostly by the breadth of activities and manner of governance. Most types occur in a range of scientific fields. One sort — the 'quasi-Athenian democracy', typified by bottom-up consensus building — is almost unique to particle physics. The book's authors argue that this model would not benefit other science communities, and indeed, as particle-physics collaborations grow larger, they might need to change from this traditional form.

One weakness of this characterization of particle physics is that collaborations have already grown in size well beyond the cases studied, all of which occurred in the 1970s and 1980s. It would be interesting to compare two competing but nearly identical projects, such as the BaBar experiment at the Stanford Linear Accelerator Center in California and the Belle experiment at the High Energy Accelerator Research Organization (KEK) in Tsukuba, Japan. These have different organizational and management styles, with different strengths and weaknesses.

The study's applicability is limited by the US-centric and physics-dominated data set. It would have been interesting to read about international collaborations, such as the ITER fusion-reactor project in Cadarache, France, the Human Genome Project and the Large Hadron Collider at CERN, the European particle-physics laboratory near Geneva. Do those projects fit the authors' models even though international diplomatic and political factors have played a more prominent role?

Under what conditions is collaboration desirable? This, too, needs further exploration, especially as more projects are located and managed in large laboratories. A recent report on university participation in US particle physics by the Department of Energy's High Energy Physics Advisory Panel suggests that the concentration of collaborative projects in large laboratories could be weakening the physics effort in universities and contributing to structural deficiencies in the science programme.