Jan. 26, 2004
Six (or fewer) degrees of science
ANN ARBOR, Mich.— Kevin Bacon has nothing on Sir Martin Rees. Like the actor who can be connected to virtually any other living movie star through a six-link-or-shorter chain of mutual movie appearances, the Astronomer Royal of Great Britain is at least as well-connected to his scientific peers.
That's just one of the intriguing details University of Michigan physicist Mark Newman uncovered when he analyzed collaboration patterns among scientists. Newman's findings—along with those of other researchers doing similar work—are summarized in a paper to be published online this week in the Proceedings of the National Academy of Sciences (PNAS).
Working with authorship data from three fields of science—biomedicine, physics and mathematics—Newman used a set of techniques he developed to reveal patterns of connection in all sorts of networks, from social connections among friends to food webs in nature. In the current work, he constructed networks based on coauthorship of scientific papers over a five-year period. Two scientists were considered connected if they had written one or more papers together. Here are some of the findings summarized in the PNAS paper:
• A small number of extremely well-connected scientists serve as "brokers" for communications between others, with most connections among collaborators passing through them.
• Most of a particular scientist's connections to others in the field are through only one or two collaborators. "Those few people alone give them most of their contact with the rest of the community," Newman said.
• A few scientists produce "enormous" numbers of papers, while most have more modest levels of productivity. Similarly, a few scientists have hundreds or even thousands of collaborators, while most work with only a few peers.
• The number of authors per paper varies widely among disciplines, probably reflecting the way research is conducted in each field—by large teams conducting experiments (biomedicine) or by individuals doing theoretical work (math). At one extreme is the sub-discipline of high- energy physics, in which authors had an average of 173 collaborators over five years.
• The number of collaborators a scientist has does not necessarily reflect the quality of his or her research; many who toil alone or with only a few other people turn out important work. However, researchers identified as extremely well connected in the analysis tended to be well known in their fields. For example, Rees, whose connections make him "the Kevin Bacon of astrophysics" in Newman's analysis, is a very influential scientist.
• The world of science is a small one. All three disciplines turned out to be highly interconnected. Mathematicians were separated from one another by 7.6 links, while the 1.6 million biomedical researchers in the analysis were separated by only 4 links.
Results such as these may offer new insights into the way science is practiced, Newman said. "My interest is in the mathematics of networks, and I'm not a sociologist, but the sociology of science is fascinating. From the human point of view, looking at differences in collaboration patterns in different sciences and speculating about why the differences exist can have implications for the way science is conducted, the way it's published, and the way it's funded."
Contact: Nancy Ross-Flanigan