The Strength in Numbers
Don't be mistaken by the title The Strenght in Numbers. This book is not about number theory or mathematics. The numbers in the title refer to the number of people cooperating in a research team. So the subtitle The New Science of Team Science says it more clearly. This book is about the management of research teams. In particular research teams in topics that are nowadays referred to as STEM (Science, Technology, Engineering, Mathematics). So this is a book by two social scientists researching the dynamics of STEM research teams in order to advise the participating scientists on how they should organize their team research, improve the effectiveness, and avoid bad experiences. The methods, style, and culture of research conducted by these authors cannot be more different from the way their subjects of study conduct and report their research. They are well aware of this difference since at some point they write "However the social sciences do have one, perhaps only one, clear advantage over the natural and physical sciences: our phenomenon talks back to us in relatively plain language that we sometimes understand". And this is what they did: they collected and analysed what STEM researchers told them about their cooperative experience. They conducted a Web survey, had personal interviews, and collected anonymous Web posts to get their raw material. These results are used to classify collaboration experiences that can range from a dream collaboration over a generally good or generally bad collaboration to a nightmare collaboration. The authors identify the problems that can occur and give some advise on how these can be prevented or remedied. This book is a detailed report on their data, their method, and their conclusions.
Especially in mathematics the idea of a lonely genius scribbling and working in an isolated ivory tower has long been around, but that is rapidly melting away in the real world, although a bit slower than in the other STEM disciplines. On the other hand, a paper of 33 pages in Physcal Review Letters about measuring the mass of a Higgs Bose particle of which the first 24 pages are just listing the 5154 authors will hopefully remain an anomaly.
A first question one might ask is where this trend towards an increase in collaboration in smaller or larger teams comes from. This can be explained by the simple fact that knowledge has become so vast that it is impossible for one person to master it all. The authors of the book tag a phenomenon of drastic change in the science landscape as "revolution" and point to some other revolutions as possible causes for an increase in cooperating teams: academic capitalism, gender diversity, and multiculturalism. From a lifelong mathematics research career I experienced that at an institutional level research management has drastically increased and not only in engineering but also in mathematics and other fundamental disciplines, results and PhD's are considered as "products" and management is focussing on increasing the productivity. Thus I agree that academic capitalism has become a fact. Gender and multicultural issues are in my opinion more buzz words than influential in team building.
On the other hand, there is, for mathematics at least, an increased interest in using more and more mathematics in other disciplines. New mathematical problems do arise that are specific for the application which often forces a collaboration between for example engineers and mathematicians or computer scientists. Wavelets were born out of theoretical physics, Fourier analysis, and signal and image processing; the Internet applications enforced a renewed mathematical interest in analysis of graphs, required studying dynamics and optimisation of networks, and a new science of preference systems posed new challenges to numerical analysis. One can come up with many other disciplines now collaborating intensively that developed previously in isolation independent from each other. This explosive scientific activity has attracted more researchers and it increased the global mass of research potential and individuals are plainly forced to collaborate in order to survive. The individuals digging their own private niche, away from mainstream research that eventually become successful are very rare mutations in the fabric of scientific evolution. Nevertheless, they can be the ones that cause the paradigm shifts that are necessary for the advance of scientific knowledge according to Thomas Kuhn.
I believe that other important factors facilitating team work are the increased mobility and the democratization of long-distance flights so that more people meet at international conferences, and of course the World Wide Web which doesn't even require travelling allows collaboration that would have been impossible before its existence. These trends were of course noticed by the institutions and the funding instances that stimulate and even enforce mobility, interdisciplinarity, and collaboration between different institutions nationally or internationally. Cooperation is then just a requirement to get funding. Some of the teams set up in this way are artificial, including an industrial partner where production prevails over research and development, or including a group of a famous "Rock Scientist" who doesn't have the time to invest in yet another collaboration.
The gender and multicultural issues may be a problem when measuring the effectiveness of the collaboration. It is in fact not straightforward to measure the output that is supposedly reflecting the effectiveness of a research team. Most often, its output is measured by counting publications and impact. Then the number and the order of the authors listed on the papers becomes important. However, collaboration can have many forms, and it need not always be materialized in a paper or a patent. Who should be listed as an author? This raises issues of ghost authors and honorary authors who did not really collaborated on the particular paper, and where on the list should they be inserted? This can often be the source of frictions and uneasiness in the team. What is usual and what is not depends on the scientific discipline. For example in mathematics, economics, and computer science alphabetical order is more common although also there this is a decreasing habit. But the authors of this book recognise the Tolstoy principle, meaning that they pay disproportionately attention to a minority of the cases that indeed gave problems. The extreme cases of a dream or nightmare collaboration than can make or destroy a career are exceptions. The majority of cases investigated work out fine and were reported as generally good.
So the authors start by classifying the collaborations according to the four qualitative experiences that were mentioned above. They discuss how to measure the effectiveness, and identify the issues that gave rise to problems. The organization and management of current teams come also in several gradations from tyrannical to a fully democratic consultative hierarchy. With many examples and citations from the results of their survey, they amply illustrate the types of collaboration, the different ways of managing the teams, and most of all, they give examples the different kinds of problems that occurred. For the social scientists (like themselves) who want to work on the same topic, they also provide a survey of the literature that is relevant to the object of their research.
However the main result is of course an analytical model they call an Aggregate Model of Research Collaboration Effectiveness (AMRCE). The AMRCE puts an emphasis on the management of the team and how this influences the effectiveness of the collaboration. This leads to the eventual aggregate of recommendations that are contained in the AMRCE model. They propose a consultative collaboration with clear rules that can be institutional or of special purpose for the team. There should however be a possibility to file a complaint or openly discuss any problem that can arise. Communication and trust are the most important components for a successful cooperation. Not really shocking conclusions, but beyond these, there is not a unique set of guidelines that will solve all problems for all possible types of collaboration. More detailed recommendations are found in the book. Fortunately most collaborations are positively experienced. In my experience, an important decisive factor allowing success or failure of collaboration often depends on a micro scale on the compatibility of the characters of the individual members that need to interact directly with each other. Communication and mutual trust and respect are indeed essential.
The purpose of the book is thus twofold. Clearly the STEM researcher can bring the recommendations of the book to good use to improve the effectiveness of their collaboration. On the other hand, the book is the result of research in the social sciences about a topic the authors consider to be a new emerging discipline: the management of research teams (in the realm of STEM disciplines) with the purpose to improve effectiveness of the collaboration. In extensive appendices the interested reader may consult all the numbers regarding how the data were collected and what they reveal, and a systematic annotated list of references used in the book and a long plain list of references.