American Scientist

Structures of Scientific Collaboration. Wesley Shrum, Joel Genuth and Ivan Chompalov. xiv + 280 pp. The MIT Press, 2007. $35.

The continuing growth of "big science" since the 1970s has presented historians and archivists of science with new challenges. The ever-increasing scale of the work has demanded greater numbers of people, larger budgets, longer periods of time, and equipment that grows ever more gargantuan and sophisticated. In writing about science and technology, historians have thus had to expand their scope to include not just scientists working alone or in small teams but the vastly more complex activities of collectives. In this endeavor they have received enormous assistance from anthropologists and sociologists of science—people such as Karin Knorr Cetina, Bruno Latour, Andrew Pickering and Sharon Traweek.

For science archivists, those vastly underrated professionals responsible for preserving and organizing the materials that allow scholars to make sense of science, it has always been a daunting task to identify, preserve and catalog the right materials, even when documenting just the scientific work of individuals or small teams. The archival problems have multiplied since large research teams emerged at institutions such as Brookhaven National Laboratory, Lawrence Berkeley National Laboratory, the Stanford Linear Accelerator Center (SLAC), the European Organization for Nuclear Research (CERN) and the Fermi National Accelerator Laboratory (Fermilab). The problems are greatest at labs that host projects in which outside investigators, often from a number of universities, collaborate with the "in-house" groups, because once the collaborations have dispersed, the question of which documents to archive is complicated by that of where they are to be found.

To help develop ways to address these thorny problems, the late Joan Warnow-Blewett, the original archivist and subsequently the associate director of the Center for History of Physics at the American Institute of Physics, organized and conducted an imaginative research project. Working with colleagues, including the center's director, Spencer Weart, and the archivist R. Joseph Anderson, she conceived a unique project aimed at studying the historical and sociological patterns of multi-institutional or computer-mediated collaborations in high-energy physics, geophysics, heavy-ion and nuclear physics, medical physics, space science, astronomy and materials science.

Warnow-Blewett assembled a multidisciplinary research team of her own, one that included sociologists and anthropologists of science as well as historians and archivists. Starting in 1989 and working for almost a decade, she and her colleagues employed an ethnographic methodology, conducting interviews with members of the various scientific teams and compiling an enormous database of information about roughly 60 collaborations. A huge effort then went into coding the interviews by categories that could be compared. The information thus organized was diverse, covering such topics as how these collaborations were started and how they are led and governed; trust and conflicts along the way; invention and practice; and issues of specialization, decision-making, values, subcontracting, instrumentation, funding and, of course, credit allocation.

Wesley Shrum, Joel Genuth and Ivan Chompalov worked with Warnow-Blewett on this effort and subsequently wrote Structures of Scientific Collaboration, a new book that attempts to summarize some of what they learned and make it accessible to scholars of science.

The authors, observing in their preface that "It is good to write a small book on a large project," draw on a substantial section of their vast array of data based on 53 of the collaborations (some were omitted because they hadn't lasted long enough to actually receive resources) to address such questions as these: How do far-flung collaborators find one another and organize? Does the degree of bureaucracy increase with the size of the project? Do the organizational forms relate to technology? To what extent do the people involved depend on trust and similar factors?

Most historians might prefer to know other things instead—for instance, how the results of the collaboration led to specific scientific findings or how the timescales of the research performed relate to academic timescales that rule the careers of scientists. But the research effort instead looked at problems that were more general and more amenable to being addressed by performing sociological analysis on the generated data.

The level of detail provided in chapters 1 through 5—which contain many bar graphs, dendrograms and charts, as well as much statistical analysis—is likely to appeal more to social scientists than to humanists. In addition, in these chapters the authors make organizational generalizations that will be of greater interest to administrators than to scholars of science. In chapter 5, for example, they note that "The primary lines of conflict in research collaborations are between major structural groups." And in chapter 4 they observe that

Collaborations that do not allocate problems but leave management to their structural components are more likely to possess certain bureaucratic characteristics than projects in which the analysis is managed by the collaboration as a whole. Similarly, collaborations that do not examine or cross-check results from multiple teams and instruments are more likely to exhibit those bureaucratic features than those that do so.

Nevertheless, many humanists interested in science will find a wealth of interesting and useful information in this volume. The immense body of data brought together here to characterize both the structure of large collaborations and their archival problems is far more extensive than in typical historical studies.

This book is, to my knowledge, the first to bring together the approaches of history, sociology and anthropology in the analysis of multi-institutional research. The methodology produces a number of interesting conflicts, perhaps worth an analysis of their own. Historians aim to study the particular reasons why things happened while recognizing the full complexity of the situation, whereas this work attempts to simplify—the authors try to strip away the massive detail of events and find the general outlines, using the more telescopic and statistical view of science that is typical of social scientists.

To end with a clear and convincing set of conclusions may be too much to ask of such a book, but the general reflections in the final chapter will certainly interest many readers. Among these is that the functioning and ultimate success of collaborations has much to do with how they formed; indeed, the authors say, "the beginning of a project may have more to do with its success than its end." They think that although bureaucracy is often seen as an evil force, it typically helps keep the group on task. They find that collaborations that avoid sharing data seem, at least from the point of view of participants, to have more likelihood of success. And, surprisingly, trust within the collaborations does not seem to be generally important, nor does conflict appear to be a negative thing. Among the most influential factors, they say, are the decisions made regarding instrumentation.

Most historians would prefer more colorful and detailed conclusions, perhaps reflecting particular scientific goals of collaborations or of the people involved, or their excitement for the scientific work done—the kinds of elements that have attracted countless readers to case studies, a genre the authors severely critique in the introduction.

Case studies are inevitably narrower, but they are much richer in detail than any ethnography, and they may well help us to understand better how science really works. It might be possible to bring together the best of the two conflicting approaches by starting with case studies to reveal patterns in a complex of messy and detailed human activity and then employing the database used in writing this book to help confirm and generalize findings.

Despite the let-down at the end of the book and its lack of human spark, it does provide an unusual, interesting and thought-provoking analysis of structures of research in modern big science. No one concerned with the life of science as it is lived in multi-institutional collaborations can afford to miss reading it.