Communications of the ACM

Bertrand Meyer asks why too many research agencies seem obsessed with funding only groundbreaking projects.

Paul Valckenaers

Bertrand Meyer is right and wrong.

He rightly states that ground-breaking research needs incremental research. Indeed, many first results are hard-to-digest. Incremental research brings these results to society. Stronger, much of the contribution to society by researchers is communication, often as a side-effect of incremental research.

He is right stating that this funding of ground-breaking research is unlikely to achieve its goal. Indeed, many ground-breaking achievements even surprised their inventors. How can external experts judge which research proposal to fund? What are effective evaluation criteria? How fast will these evaluation criteria become obsolete when researchers adapt (analogous to insects with black and yellow stripes without being able to sting)? To what extent will insider knowledge on these criteria and their relative importance falsify the competition? To what extent does even subtle lobbying become decisive in the funding selection?

He is wrong when suggesting that "quality, originality, methodology, and track record" are sufficient grounds to decide on funding research, incremental or otherwise. For society, supporting a scientific community only makes sense because of ground-breaking results. As an analogy, funding a community of Mensa-members inventing evermore complicated manners to perform calculus in roman numbers is a poor investment for our society. In contrast, the invention of Arab numbers makes funding research a very profitable business for society. Here, a key issue is that incremental research communities are failing to recognize ground-breaking results in an embryonic stage and, when presented with ground-breaking research, often are reluctant to switch.

The solution unfortunately goes against the nature of current funding practices. To facilitate ground-breaking research, it suffices to give researchers their academic freedom. Indeed, 99.99% of researchers want to make a difference to society and contribute something extraordinary. There is no need to (micro)manage something that we get for free. It is not necessary to allocate premium funding for researchers on a trail to ground-breaking results. They know that they are on a trail that makes sense and that has extraordinary potential. The only special requirements for such research is access to resources necessary to go where none has gone before. For the most, this involves the ability to go where no evaluation board would allow it. In a fictitious example, experts in the stone age would not approve experiments attempting to insert colored stones in ceramic pots (because thermal stress will make this fail). However, some of those stones are copper ore. Current funding practices prevent transitions of this nature (from the stone age into the bronze age) and thus close the main path along which research is able to contribute where it counts.

Moreover, by emphasizing accountability over academic freedom, funding agencies put constant pressure on researchers. It is precisely such pressures that deny incremental researchers from welcoming ground-breaking but immature and poorly formulated research results. They are denied the time and resources to re-orient themselves. They are not rewarded because accountability is based on whatever happens to be measurable, which are very rough activity level indicators (paper counting, citations, impact factors). The accuracy of current practices is analogous to measuring the climate quality in buildings by looking at the electricity bills (note: accurate measurement of the electricity consumption does not solve the problem). There are no rewards for early detection of ground-breaking results. To the contrary, incremental researchers are induced to smother ground-breaking research to ensure their own success and even survival as a researcher. The unavoidable immaturity of ground-breaking research results makes them easy victims.

The core issue of the above is that allocating the available funding to the applications for funding (asking for much more than available) needs innovation. The base against which novel approaches have to compete shall not be current practice. The base solution is a structured and weighted lottery in which random selection is applied as soon as no information is available to steer the choices. Structure ensures that funding is spread across domains, short/medium/long term research... Weights account for track records... The prices ensure that talent is attracted (e.g. tenure). Once the zero-sum-game is resolved, the policies, evaluations, reviews... will be able to facilitate rather than deny researchers to do what they deem to be most beneficial to society. Indeed, the latter is something we get for free and does not require management.