This afternoon I attended a talk by Cornell computer scientist Jon Kleinberg at Stanford University, in the computer science department’s Distinguished Lectures Series. His research is fascinating, but the talk was also interesting at a different level. Among other things, he discussed his research on prestige in science, for which he developed a game-theoretic model. Now this is funny: here’s a computer scientist studying a pure sociology-of-science problem, using the economists’ toolbox. Of course, Kleinberg has also done quite some research on online networks, which is clearly related to sociology, but with this topic he really moves into social science territory.
I think this might be symptomatic of a relatively new trend of computer science starting to colonize social science, just like the economists have in the past (e.g., “network economics” or even freakonomics). In a sense, it’s not surprising: the rise of online social networks provided a natural bridge between the two fields, and computer science has a well-developed toolbox available to collect and analyze the data from these networks, in ways that the average sociologist can only dream of (e.g., web crawlers, efficient algorithms for analysis, huge hardware infrastructure and the skills to handle it). Why not use these tools for other social science topics as well? In addition, they seem to suffer less them economics form theoretical dogma; computer scientists are not committed to the rational choice approach and will very pragmatically come up with simple models or algorithms that seem to explain the data.
Sociologists, in the meantime, have to be careful not to be left behind. Before you know it, all the really cool work in our field is going to be done by others.