I’ve had one too many conversations with Singaporeans in which the distinction between “pure” and “applied” research is conflated with that between research that aims to produce tangible results in the long term and research that aims for quicker visible returns.

“Pure” research and “applied” research differ from one another on the how easily we can imagine them leading directly to technological applications (I include “directly” to exclude spin-offs from “pure” research projects that could have been directly funded by other means, like advances in superconductor magnets spun off from constructing particle accelerators). This distinction is thus a function of the intentions of the researchers and the conceivable results of the research. Relatively uncontroversial examples of “pure” research include high energy physics, cosmology, and palaeontology. It is hard to conceive of the main goals of projects in those fields leading directly to advances in technology.

The distinction between long-term-oriented research and short-term-oriented research, in contrast, is not based on our powers of imagination. They differ on the length of time over which they are expected to “succeed”. Of course whether one takes a project to be short-term or long-term then depends on one’s standards of “success” for the project. So this distinction is a function of one’s definition of success and the expected time-frame of the research projects. If one takes the standards of “success” for quantum computing to be that of producing a quantum computer that outperforms classical computers, then quantum computing will “succeed” only in the long term, if ever. But at the same time, it is very clearly “applied” research: its possible applications are very direct and easily conceivable (even if practical difficulties get in the way of realising our vision of its applications). In contrast, some projects in observational astronomy or experimental particle physics are expected to produce “success” in a shorter time frame than quantum computing is, but we have a hard time envisioning how their success directly leads to technological applications. Thus, “pure” research can be either “short-term” or “long-term”.

A recent article on Biopolis suggested that unlike scientists working in academia, scientists at A*Star are held to standards that demand results over a shorter time period. It is often declared that it is impractical for the Singapore government to support “blue-sky” (a.k.a. “pure”) research. This is often taken as a justification for policies that demand short-term success. But if I am right that “pure” versus “applied” is a distinction that is orthogonal to that of “short-term” versus “long-term”, then short-termism cannot be justified by an aversion to “pure” research. One would also have to justify penalizing those who pursue “applied” research that can be reasonably expected to be “successful” only in the long term.

There are some other assumptions made in the above justification that deserve to be highlighted. Suppose we grant, despite what I’ve said, that all “applied” research is short-term-oriented. One might think it then follows that the best strategy to maximise short-term returns from research is to demand that all research is short-term. But that conclusion is contingent on one assuming that the success of short-term research is independent of what the situation for long-term research is like. It is assuming that within a single organisation or region, the success of some research groups is independent of the quality and nature of research going on in other research groups. Anyone who has worked in research knows that this is patently untrue. If I take any given physics department and fire all the faculty who work on “pure” projects, keeping only those who work on “applied” projects, the decreased flow of ideas and collaboration opportunities will have a tangible effect on the quality of research of the “applied” people. One might counter that the money saved could be used to fund more “applied” research, and that the gains from this would exceed the losses from reducing the flow of ideas, but it is far from clear that this would be the case, given that the positive influence of the “pure” people on the “applied” people is rather difficult to measure (while being difficult to deny, as well). At the very least, one expects some kind of principled argument and evidence for why it would be so. None, so far, has been forthcoming.

Note that a similar assumption is operating when one asserts that the best way to have short-term gains in research is to allow only short-term-oriented research. That assumes that the quality of short-term-oriented research is independent of whether that research is done in an environment that contains more long-term-oriented research.

The lesson is that policies that value only certain quantities do not necessarily lead to a maximization of those quantities. Such policies have myriad unintended consequences that deserve wider attention than they get in the current naive discourse that surrounds Singapore’s R&D strategies.