Interdisciplinarity

It is difficult these days to go through calls for research proposals from government agencies, and not see at least some emphasis on “interdisciplinary research” (IR), i.e., research centered on a project or theme involving a collaboration among scientists of different backgrounds. IR is the focus of a growing number of scientific journals (see, here, for instance), including some with a clearly stated intellectual affiliation to one specific discipline (see, for instance, this one).
Interdisciplinarity is also the darling of university administrations, eager to establish new curricula of studies, aimed at imparting broad (if perhaps less in-depth) knowledge of science, spread across several of the traditional fields as opposed to focused on one of them.

In some respects, it is easy too see why the “interdisciplinary” paradigm may appeal to some researchers and educators. The case is plausible that too rigid a division among disciplines, may prove detrimental to the investigation of novel problems, which may not fall squarely within one of the existing, generally accepted areas. After all, there are examples of progress in some fields of science being instigated and underlain by advances in other fields. Such valuable cross-fertilization may be rendered more difficult, if not prevented altogether, by the somewhat parochial mentality fostered by the traditional division of science into “turfs” — the idea that problems in biology are best tackled by biologists, in physics by physicists and so on.
From the educational standpoint, the argument is often put forth that the type of employment that most science graduates find (not just with a Bachelor’s but even with a doctoral degree), even when research-based, seldom has them focus on the same subject, or theme, for extended period of times. Especially in industry, researchers are expected to switch easily from one subject to another, typically never going to significant depth in any of them, but rather assessing reliably the practical feasibility and potential impact of a line of work on the bottom line of their employers, for example. This seems to be seriously at odds with the philosophy underlying most science graduate programs, where a student will spend years literally becoming one of a handful of worldwide experts in a very narrowly defined subject, while trying all possible approaches in order to answer a single outstanding question.

My experience of researcher and educator, however, has made me aware of what I regard as potential pitfalls of the notion of “interdisciplinarity”, especially when elevated to the role of guiding criterion for research and education.
I have seen many research and educational initiatives in my view scarcely innovative and of little breadth, predicated based on the interdisciplinary mantra, where promoting collaboration among physicists, chemists, biologists, engineers etc. seemed to be an end to itself, rather than the key to making fundamental advances.

Which problems are truly interdisciplinary ?
IR does not come from putting a physicist, a chemist and a biologist in a room and telling them “you people do something interdisciplinary” (although some funding program directors and university administrators seem to think that way). Attempts to promote new research in this way, e.g., through the creation of “interdisciplinary research institutes”, are wasteful and misguided, useful perhaps to talk gullible politicians into funding expensive new infrastructure, but not likely to yield any useful scientific result.
The starting point for IR is usually a well-defined, concrete outstanding problem in a specific area of inquiry, solidly situated within one of the traditional fields.
In other words, while the idea may sound plausible that there exist scientific problems containing elements drawn from several of the known disciplines, but not really falling into any of them, my observation is that in practice most research branded as “interdisciplinary” is fairly easily labeled according to the traditional classification — biology, physics, chemistry and so on. What makes it IR is largely the fact that practitioners in the field where it belongs, are technically ill-equipped to tackle the problem at hand, and benefit from input from scientists in other fields who happen to possess the knowledge needed at that juncture.
A case in point is econophysics, a field of research that became fashionable in the 90s, as theoretical physicists began to apply methods of statistical physics to solve problems in economics. I am no expert of this subject, and I am sure that significant results were achieved, but it seems excessive to have coined such a word for research work that can be properly seen as economics ad not physics, into which an exceedingly small fraction of physicists are engaged [0], which has not resulted into any advances in physics, and that economists will soon be perfectly capable of doing themselves, once economics curricula are updated to include the few notions drawn from statistical physics that are useful or needed.
Biophysics is another example of application of basic physics notions, ideas and methods (for instance, molecular dynamics simulations) to problems of relevance to biology. It seems difficult to imagine that novel ideas or fundamental advances in physics will come from biophysics [1], and one may expect that biologists will soon be able to do without a “resident physicist”, as soon as they start taking their physics courses more seriously (just kidding…).

Now, do not get me wrong: there is absolutely nothing bad about, for instance, physicists devoting part of their research effort to problems in, say, oceanography; they will probably have fun, do something useful for a change, learn new stuff and at the end of the day go home with a publication in an exotic, cool field. More generally, scientists should talk to colleagues in different fields and yes, progress can come from collaborations across disciplines. However, I still think that these fortunate cases arise largely as a result of the individual interest and effort of the scientists involved; I am very skeptical of any attempt to catalyze this type of process by means of financial incentives or other practices.
Moreover, it should be clear that this is not about exploring new areas of research, it is about pushing the envelope in existing ones. It is about putting the knowledge built in one field at the service of another. The type of IR described above crucially hinges on the existence and preservation of well-defined, separate disciplines.

Jack of all trades, master of none…
Based on the above, it seems that attempts to modify science curricula in the direction of blurring the intellectual distinction among the various disciplines, as well as dispersing a student’s education across several fields as opposed to promoting some degree of concentration, are only likely to undermine the very foundation that makes most IR possible.
It is a simple fact that the overwhelming majority of scientific research work continues to be disciplinary, and requires therefore the type of strong foundation in a given field that only targeted course work and years of focused research can provide. I am worried about educating science students, especially at the graduate level, in some “interdisciplinary” fashion, where they would familiarize themselves with ideas from different fields without really going to any depth — these students seriously risk to end up short changed on the job market. Their background is likely to come across as shallow and unfocused, and doubts may well arise about their capability of tackling difficult problems independently. They are not likely to be strong competitors for most research jobs, both in academia as well as in national laboratories.
The notion that science graduates are too narrowly trained, and that their excessive specialization renders them somewhat ill-suited to perform adequately in a setting that requires instead flexibility and constant retraining, may again sound plausible but does not seem to be supported by any quantitative measure (I have written about it here and here). And I do think that there is something valuable about becoming an expert, albeit in a narrow field of inquiry. It endows one with a sense of appreciation for scientific progress, the difficulty and hurdles that must be overcome in order to achieve significant new results, as well as the importance of comparing different approaches to the same problems.

Notes

[0] I am personally unaware of any physics department currently trying to build a sizable econophysics research group, or even having hired an “econophysicist” at any time.

[1] See, for instance, this article.

Tags: Academia, Collective Fixations, Graduate Studies, Physics, Research, Science, Science Education, Science Policy