In 1985, Oxford University diabetes researcher David Matthews published a study in the journal Diabetologia about a computerized mathematical model that could determine blood glucose and insulin levels. For more than two decades, the paper attracted little attention. But something changed in 2008, and in the 11 years since, the paper has been cited more than a thousand times each year, adding up to just under 20,000 citations. That makes Matthews’ study one of the most highly-cited pieces of research published in 1985.
Matthews’ study is what is known as a “sleeping beauty,” a paper that awakens from its dormant state of attracting little to no attention to suddenly become a major body of work. Though the number of such papers was on the increase in medical literature until 1998, they have since remained at a constant level, according to a recent analysis of 230 million citations from seven million papers published between 1980 and 2017. The trend holds true when the authors account for the fact that scholarly literature has exploded in the last few decades, with the global scientific output roughly doubling every nine years.
This is likely due to the fact that scholarly literature has become easier to find due to the rise of open access policies, which require journals to make papers free to read, according to Ton van Raan, a bibliometrician at Leiden University in the Netherlands who co-authored the analysis. But it could also be that researchers are conducting less of the sort of risky, out-of-the-box science that can create such sleeping beauty papers, due to increasing academic pressures to publish more and more research.
“The huge pressure to publish means many researchers often cut corners and fail to do a proper systematic review before starting their research,” says Adrian Barnett, a statistician and metascience researcher at Queensland University of Technology in Brisbane, Australia, who was not involved with the new analysis.
“A new scientific truth does not triumph by convincing its opponents, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.”
Academia’s publish-or-perish culture is often thought to drive researchers to cut corners to produce as many journal papers as possible, which, in turn, can be used to win grant funding and professional advancement. This means scientists often lack the time and incentives to pursue unusual research that may not directly result in journal papers and contribute towards future grant applications.
It’s still unclear why some studies that eventually become so prominent initially go unnoticed. Barnett’s hunch is that many of these studies offer radical ideas which are highly controversial when they are first proposed, but eventually become more widely accepted as old influencers in the field die. As the Nobel Prize-winning physicist Max Planck once said: “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.”
Barnett thinks there are now fewer “princes” in literature, which are new studies that start citing older sleeping beauties to wake up them up from their slumber. “The explosion in papers over time has also made systematic reviews harder to do because, in general, researchers have to do far more work today than 20 years ago in terms of reading and summarizing the previous work,” he says.
The amount of time sleeping beauty papers remained under the radar was considerably longer in the 1980s than it is now, however. The longer a study remains in the literature without being cited, the more unlikely a paper will be frequently cited. In other words, the longer a potential sleeping beauty is asleep, the less likely it will ever be awakened.
In recent years, it has also become more likely that sleeping beauties will be cited in patents, even though they aren’t yet frequently cited by other studies, according to the analysis. “This phenomenon is most probably related to two important developments: the increasing number of patents and the increasing number of patent citations to scientific publications,” van Raan explains.
For Barnett, the analysis highlights the importance of sheer good fortune — beyond quality or originality — in whether a scientist’s work will eventually earn prominence. “Some great papers need that lucky break of being read by someone who really understands their importance and is able to influence the field by commenting on the paper or using it to further their own research,” he says.
Although access to research is becoming easier worldwide, sleeping beauties aren’t going away, argues Qing Ke, who studies complex computer networks at Northeastern University in Boston. “Scientific discoveries can be made before their implications are realized by the scientific community, and such lag could be decades long,” he says. Ke and his colleagues previously sifted through 22 million scientific papers to find sleeping beauties, and in the process spotted a now-famous 1935 quantum mechanics paper by Albert Einstein, Boris Podolsky, and Nathan Rosen that was hiding in plain sight for several years.
Barnett says analyzing sleeping beauties is necessary, since “missing or ignoring important papers can create serious research waste.” Perhaps most importantly, scientists need to move away from judging researchers solely on how many citations their work accumulates. “The more we measure and reward the careers of scientists based on citations,” says Barnett, “the more we introduce luck into our reward systems.”