Narrow Specialization in Modern Science

Lawrence J. Henderson summarized the problem of increasingly narrow research specialization in his lecture Natural Science: General Introduction (part of the Harvard Classics series).

The number of working scientists, if not their quality, has enormously increased. An army has been organized and disciplined, and an amount of work which can scarcely be imagined has been produced. Scientific literature has now become a flood that has to be canalized with the help of special journals of various descriptions devoted solely to its review, description, and orderly classification, in order that it may be utilized at all.

Descartes was philosopher, scientist, and mathematician; some of the great men of the eighteenth century were hardly less so. Even through a large part of the nineteenth century many of the greater men ranged widely over the field of science and mathematics. To-day the force of circumstances has largely changed all that. The chemist is likely to look upon the physicist, or even the physical chemist, with suspicion on account of his mathematical interests. On the other hand, the mathematician, unlike Newton, Euler, and Gauss, is commonly no longer a physicist at all. There are to-day very few men who possess even a superficial acquaintance with all the principal departments of science, and between the work of the astronomer, on the one hand, and that of the anatomist, on the other, there is perhaps no closer relationship than the fact that both employ optical instruments in their researches.

Amazingly enough, Prof. Henderson wrote this in 1914! If he complained then about the flood of literature and increasingly narrow specialization, what would he say about the state of science today?

Interdisciplinary Communications

Each of the traditional scientific disciplines comes with its own research philosophy, which for the purpose of this discussion can be thought of as a system of common assumptions and postulates shared by most of the practitioners. For example, a physicist is likely consider the following principles to be important and even self-evident: reductionism, use of model systems, and quantitative analysis.

Michael W. Friedlander provides a concise description of the differences between the life and physical sciences in his book At the Fringes of Science. He points out that the physical sciences have been successful "largely because it has been possible to reduce their complexity to the scale of controllable and inanimate systems, with the focus of attention on a very few quantities to the exclusion of all others". The same approach is not common in life science research, mainly because "it is not as easy to identify much less exclude unwanted variables in the study of living organisms".

Research in new fields like nanoscience or biointerface science, however, is inherently multidisciplinary, and thus requires a more flexible research philosophy. In particular, interactions between physicists, chemists, and biologists often call for mutual awareness of the differences in underlying assumptions that each of the disciplines brings to the table. Accordingly, uncovering and then resolving any apparent conflicts between research philosophies becomes the first important step to finding a compromise and thus moving from a multidisciplinary to a interdisciplinary collaboration.

Basic Concepts and Vocabulary

The following correction from MRS Bulletin illustrates that even the simplest and most basic concepts do not easily cross disciplinary boundaries.

In the February 2010 issue of MRS Bulletin 35 (2) p. 154, the statement, "One such surprise is the presence of an amino acid, glycine, which is one of the four building blocks found in DNA" should read "One such surprise is glycine, one of the amino acids commonly found in proteins."

Model Example of a Realistic Problem

A molecule that exhibits an interesting property contains two functional groups separated by an "inert" spacer. A reductionist approach, typically favored by a physicist, suggests isolating the functional groups in order to test which one of them, if either, produces the effect. After all, why spend time studying the "unimportant" parts of a molecule? An organic chemist, however, just as instinctively, may see the solution in testing a series (or even a matrix) of control molecules, where each of the three parameters (two functional groups and the length of the spacer) is systematically varied. The idea is that testing such a comprehensive matrix will provide sufficient number of positive and negative controls for determining which part of the molecule produces the effect. A physical chemist, in turn, may suggest to keep the molecule the same, but rather perform a series of experiments under systematically varying conditions (e.g., run a pH titration curve), to study how the effect responds to such changes and thereby gather clues to the underlying mechanism.

This is a simplified example of a real situation—a reductionist model of a problem, as it were. As an anecdotal composite, it is meant to illustrate how many respective scientists perceive each other's "favorite" methods, rather than to reinforce those stereotypical views. But specific views arguably are not important in this example. As long as the approaches are conceptually different, the essential question remains:"Is there a right (or a wrong) choice of a research philosophy?"

Obviously, the answer is not choosing the single "right" philosophy, but rather developing an integrated one. Each discipline has been successful in its own right and can provide helpful insight; furthermore, neither discipline is inherently superior. But until the three archetypal scientists from the above example realize why they are arguing about their methods, they will simply talk through each other, because superficially each of the possibilities essentially negates and opposes the other two: "Reducing the complexity, how absurd! Complexity is the most important characteristic of this system!" Or the perennial favorite: "But you just don't understand!"

There are no "Dumb" Questions!

The paramount requirements in any interdisciplinary project are then appreciating the strengths of different collaborators and working out a way to combine these strengths for a productive and mutually beneficial collaboration. Many people do not realize how much effort this approach sometimes involves ("haven't I explained this to you five times already?!") but the benefits are invariably worth the trouble. A physicist may always advocate reductionism and look for model systems that can be measured directly, but he can also learn to think about controls (positive and negative controls) from colleagues in life sciences, and to appreciate the value of a well-designed experimental series (albeit, perhaps not always the same series that a chemist would have chosen). A scientist cannot become expert in every field, but the ability and willingness to learn from anyone who is willing to walk their half of the way to a middle-ground compromise is crucial for successful interdisciplinary research.

Not surprisingly, willingness to learn and sometimes to eat the proverbial "humble pie" is as important in multi- and interdisciplinary research as it is in one's own field. Whether learning the basics of Chemistry, Biology, Physics, or English grammar, asking naive questions simply cannot be avoided. Fortunately, only a minority of scientists will send someone to the library before bothering to even attempt an explanation. In fact, because most scientists love to talk about what they do, people who ask questions are actually more often perceived as inquisitive, rather than sophomoric.