Academic questions: Interdisciplinarity

No Discipline Is an Island

Watching younger friends struggling for grant money, I see that the granting agencies, and thus university administrations eager to rake in funding, are still pushing pretty hard to foster interdisciplinary research. You jack up your chances of getting a project approved if you can haul people from other departments into it.

There's a certain amount of tosh in this push for interdisciplinarity, surely??  ...  The tosh arises from some interesting (and related) misconceptions, I think: (a) Interdisciplinary research is desirable as such, because it is likely to be less narrow and more creative than within-discipline research. (b) Interdisciplinary research means research conducted by a team of individuals from different departments. (c) A principal obstacle to interdisciplinary research is the administrative separateness of university departments and colleges (or in non-US institutions, institutes and faculties).

The toshness of (a) comes from its assumption that intellectual distances are greater between than within disciplines—the assumption that a discipline is an island of intellectual uniformity, separated by some sea-miles from islands of different intellectual uniformities.  This is not the case. To be sure, particle physics and art history are miles apart. But disciplines are not islands, they are more like countries on a mainland where borders are historically unstable and there's a lot of internal conflict. A discipline often covers a fairly diverse intellectual territory—diverse enough to prompt uncivil wars within the discipline—and scholars on one side of the discipline may have more intellectually in common with scholars just over the border in the next discipline than with colleagues over on the far side of their own discipline. 

Thus staying within your own discipline doesn’t guarantee narrowness, nor does linking outside it guarantee breadth.  

Moreover, if you offer big rewards for research that involves people from a couple of disciplines, in the hope of getting novel research that doesn't just color within the lines of a single disciplinary tradition, there’s a fair likelihood that you’re going to get research that just colors inside the intersection of multiple sets of lines. The intersection of A, B, and C—that is, what people from the various disciplines can agree on—cannot be larger than A or B or C alone.  Hence the kind of interdisciplinary research that is performed just in order to be interdisciplinary will fairly often (not always) be narrow not broad and will have a vapid committee-report quality.

The toshness of (b) comes from the fact that having people from three different departments working together isn't the only--and maybe isn't even the best--way of prompting the liberating-destructive influence that one discipline can exert on another. The people from the three different departments may think alike already. Or they may stick to the limited intersection of their three disciplines. 

People who are all within a discipline can bring in radical change, inspired by another discipline, and this kind of inspiration isn't necessarily identical with formal relations with another department (joint appointments, joint grant projects and so on). When Kahneman and Tversky got started on the projects that ended up bringing psychology into economics and economic questions into psychology in a big way, they were both psychologists, and they didn't proceed by setting up a joint project with a bunch of economists. Eventually there was more cross-department interaction, but they did a lot of groundbreaking work as psychologists doing psychology. Similarly in accounting-land: Ball and Brown were both accountants, so their 1968 paper, which remade accounting research in the image of empirical finance, would not have counted as “interdisciplinary” in the sense of multiple-department research.  

The toshness of (c) ...  Well, for example: conflict a few years ago in a German university. A high-level administrator was bent on breaking up and recombining colleges, partly to increase economic efficiency (sensible: there were probably too many little colleges), partly to appear to be doing something (sigh), and partly to foster interdisciplinary research. His view: Administrative units are what matter. Recombining administrative units will lead to recombining ideas. And of course it's true that meeting different people by the coffee machine can lead to interesting conversations when departments and thus office space have been reconfigured.  But then there's also the opposing view taken by the professors in the relevant colleges: We have research projects with other colleges already. But if we break up and recombine the administrative structures, we'll have to spend the next two years rewriting all our bylaws and degree requirements, and we'll hardly get any research done at all.