Scientific progress—and scientific careers—depends on novel ideas. One might therefore expect the scientific community to welcome intellectual innovations as boons to advancement. Two recent studies, however, show just the opposite. One examines how the presence of a prominent, established leader in a research field discourages new people from entering it and presenting ideas considered unconventional. The other reveals a “bias against novelty” that often keeps innovative research from gaining the attention and recognition it deserves, at least initially.
Together, these results imply career risks for scientists who put forward new theories or interpretations—even though, in the right circumstances, doing so can also make for big advances and big professional successes. In deciding what type of research to pursue, the authors of both studies suggest, scientists must therefore weigh the relative costs and benefits of doing risky but potentially more scientifically significant work against taking a safer and more conventional path that can appear likelier to lead to funding and advancement. Other observers have also described how today’s extremely competitive funding environment can make the safe course appear by far the wiser one, at least from a career standpoint. But, the articles’ authors also note, changing certain policies could make science more receptive to the novelty it needs to move forward.
Death and progress
“Does Science Advance One Funeral at a Time?” empirically tests the widely quoted witticism of Nobel Prize-winning Physicist Max Planck. (Apparently he didn't exactly state this. He instead appears to have said, as translated from the original German and quoted in the paper, “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.”) Regardless of how Planck expressed his belief that senior scientific stars are often loath to accept ideas that threaten their views, the paper’s “results lend credence” to the notion, write Pierre Azoulay of the Sloan School of Management at the Massachusetts Institute of Technology in Cambridge; Christian Fons-Rosen of the Barcelona Graduate School of Economics in Spain; and Joshua S. Graff Zivin of the University of California, San Diego, in their National Bureau of Economic Research (NBER) working paper.
The authors explored how established stars influence the flow of ideas by examining what happens to scientific fields when a dominating figure unexpectedly dies. They identified 452 eminent academic life scientists whose deaths were premature—defined as happening before the scientist entered pre-retirement or took on a predominantly administrative role—and studied how these demises affected the “vitality (measured by publication rates and funding flows) of the [scientists’] subfields.”
“[P]recipitous declines” in the publication rates of the star’s collaborators was the most immediate effect, the authors write. “Remarkably, however, … increased publication rates” by scientists who had not previously collaborated with the star eventually “more than offset” the declines, they continue. These publications in fact “represent [the noncollaborators’] first foray into the extinct star’s subfield.”
Equally strikingly, the new contributors are “less likely to be part of the scientific elite” than the star was, though they are “not necessarily younger on average” than the star. But in the long run—over a period of 5 years—these contributions tend to attain high levels of citation impact. Moreover, the entrants’ contributions are “less likely to cite previous research in the field, and especially less likely to cite the deceased star's work at all.” Together, these findings suggest that these newcomers bring important new ideas. In fact, the authors write, they “appear to tackle the mainstream questions within the field … by leveraging newer ideas that arise in other domains.” While the stars are alive, however, they appear to “regulate access” to their fields and enjoy “outsized opportunities to shape the direction of scientific advance in that space,” the authors add.
Stars don’t use such obvious tactics as rejecting newcomers’ journal submissions or downgrading their grant proposals. At the time of their deaths, for example, only three of the 452 deceased stars were journal editors and just another three served on study sections. Clearly, so few individuals “could not possibly drive the robust effects we have uncovered,” the authors argue.
Instead, subtler barriers, such as what the authors call “Goliath’s shadow,” discourage outsiders from “challenging a [living] luminary.” Obstacles can also include “intellectual closure,” which the authors define as strong agreement among the field’s members “on the set of questions, approaches, and methodologies that propel the field forward,” as well as “social closure,” which develops when the field’s members function as “a tightly-knit clique, often collaborating with each other, and perhaps also reviewing each other's manuscripts.” But just as the felling of a large tree allows young plants to sprout in the newly sunny ground that it formerly shaded, the passing of a major scientist permits new ideas to spring up in the intellectual space she or he once dominated.
The new growth does not flourish immediately, however. In fact, understanding the dynamics of how novel ideas enter a scientific field requires a long view, reports another NBER working paper: “Bias against Novelty in Science: A Cautionary Tale for Users of Bibliometric Indicators” by Jian Wang and Reinhilde Veugelers of the University of Leuven in Belgium and Paula Stephan of Georgia State University. Research based on ideas new to a field has “a high potential for major impact but also carries a higher uncertainty of having impact” because it can be slow to gain attention, the authors write. In the long run, however, novel papers “are more likely to be a top 1% highly cited paper …, to inspire follow on highly cited research, and to be cited in a broader set of disciplines” than more conventional ones.
Along with the “strong evidence of delayed recognition” for papers that present new and original combinations of knowledge, these papers tend to appear in journals with lower impact factors than more conventional work. Beyond that, they are “significantly more highly cited in ‘foreign’ fields but not in [their] ‘home’ field.” In other words, they need substantially longer periods of time to attract the attention they ultimately deserve.
This delay in recognition can place authors of “high risk/high gain” papers at a disadvantage in the contest for funding and career advancement, because their work does relatively badly on the “classic bibliometric measures” of article impact that generally “use short citation windows” of only a few years, the authors note. The lower impact factors of the journals where novel work appears add to that effect. Other recent research confirms that “evaluators give lower scores to proposals that are highly novel,” they add.
“[T]he increased reliance funding agencies place on [such] classic bibliometric indicators” as short-term citation metrics and journal impact measures thus hampers the chances that innovative work will win support and constrains both the progress of science and the career prospects of innovative but lesser-known researchers, the authors observe. Agencies that rely on such measures tend to be “more and more risk-averse, choosing ‘safe’ projects over those” with greater potential for major gain.
To catch underrecognized work, funders should instead use citation windows considerably longer than the customary 2 or 3 years as well as “a wider portfolio of indicators,” the authors advise. Funders should also eschew “a monodisciplinary approach in peer review” and instead draw reviewers from a range of disciplines. And to help keep established stars from hampering innovative ideas by monopolizing funding, Azoulay, Fons-Rosen and Graff Zivin propose that agencies should use double-blind peer review and cap the amounts that any single laboratory can obtain. Funders, they write, should also award “bonus points” to new researchers. In addition, “emeritus awards to induce senior scientists to wind down their laboratory activities” could help clear the way for newcomers.
How much such steps could mitigate novelty’s disadvantage in today’s intense funding competition is unclear. Surely, however, anything that might help tip decisions toward riskier research would be good for science—and for the scientists whose novel ideas will help it progress.