Obsession with novelty sidelines deeper learning
Too much focus on generating new ideas in science is driving the replication crisis.
22 November 2017
An overemphasis on novelty has meant that funders and journal editors are neglecting the equally important work of revisiting old problems, says molecular biologist, Barak Cohen, at Washington University School of Medicine in St. Louis. “If we always have to be finding something new to get funding or credit, it’s harder to pursue something in depth.”
Cohen set out his case in an opinion article in the journal eLife this year, calling for a renewed emphasis on research that validates existing ideas, deepens understanding, and improves predictive power.
Over the past two decades, the proportion of papers in the PubMed database containing the word ‘novel’ has increased from around 2% in 1995 to 7% in 2015.
Some journals stress the need for novelty in their guidelines to authors — Science looks for papers that “present novel and broadly important data, syntheses, or concepts”. Nature asks for “original scientific research...of outstanding scientific importance”.
Funders also highlight a desire for novelty. The National Institutes of Health’s instructions on preparing grant proposals suggest that those using novel concepts or approaches stand a better chance of success.
“Too much pressure on originality will create deviant behaviour,” says Yves Gingras, a historian of science at the University of Quebec in Montreal, who sees the overemphasis on novelty as partially responsible for the rise in retractions and findings of scientific fraud, and the current “reproducibility crisis” in many fields of science. “It’s worth nothing to reproduce someone else’s work. There is no reward for that.”
In his own field, Cohen says the emphasis on novelty has downgraded the valuable work of building models of biological systems that can accurately predict how they will react to perturbations. In his eLife paper, Cohen quotes an unnamed journal, which turned down his work on predicting drug sensitivity for being insufficiently novel. “There is immense value in going back to ancient problems and mathematically modelling them,” he says.
Mark Johnston, a molecular biologist at the University of Colorado Denver, says that the vast progress made in the field over the past 50 years has complicated the search for novel discoveries. Biology, says Johnston, is in a similar place to where physics was in the 1920s and 30s, when it transitioned from what he calls the “age of discovery” to the “age of application” — from Isaac Newton to Albert Einstein.
Biologists should now be doing the same, he suggests, focusing more on how to use the knowledge they have gained. “We know enough that we should be able to translate more of it.”