Published on
Updated
Reading 3 mins.
Scientific publications and technological innovations are growing exponentially, but major discoveries that dramatically advance science are declining over time, according to a Nature study published Wednesday.
This paradox, that of a science in slow motion despite an acceleration in the production of knowledge, was already documented for certain research fields such as semiconductors or pharmacy, note the authors of the study, researchers at the Carlson School of Management from the University of Minnesota (USA).
They wanted to measure it for the first time across all disciplines. Based on 45 million scientific articles over six decades (from 1945 to 2010), and 3.9 million patents (1976 to 2010), extracted from global databases including the Web of Science information platform .
Over the years, their work dissects the most cited papers, the way scientists cite each other, to develop a “disruptivity index“. Which makes it possible to estimate whether an article has changed the situation or whether a technological innovation has created a breakthrough.
They distinguish two types of discoveries: on the one hand, the contributions which come to consolidate a “status quo”, like the Nobel prize winners Khon and Sham (1965) who based themselves on existing theorems to develop an original method of calculating the electronic structure of atoms.
continuous fall
On the other hand, the “disruptive” discoveries that shake up theories and propel science in another direction, like Nobel Watsons and Cricks, who revolutionized biology by demonstrating the double helix structure of DNA.
Their index ranges from a negative score of -1 (consolidation) to a positive score of +1 (disruption). According to their calculations, the percentage of this index decreased considerably and continuously between 1945 and 2010 for publications (91.9%), affecting the medical, physical and social sciences as well. Ditto for patents, with a fall of 80%.
The decline is felt in the vocabulary: during the first decades studied, verbs evoking creation and discovery predominate; in recent decades, they have been replaced by words evoking improvements or applications.
“This does not mean that there are no more advances“, nuance Michael Park, the main author. Despite a slow pace, “there is always a fairly consistent number of disruptive works in recent discoveries, such as gravitational waves in 2015 or the messenger RNA vaccine against Covid-19“, he explains to AFP.
But “a healthy scientific ecosystem is a system where disruptive and consolidating discoveries intertwine. The nature of science is changing“, emphasizes Professor Russell Funk, who led the work.
“Publish or Perish”
In question, the growing weight of knowledge that researchers and inventors must acquire, who spend more time training than “pushing the limits of science”.
This “burden of research” leads to relying “on increasingly narrow portions of existing knowledge“, regrets Professor Funk. The study notes, for example, a tendency to cite more of the same previous work, therefore a lack of renewal.
Another factor: the pressure for “publish, publish, publish“, a guarantee of academic success as the expression says”publish or perish“, he comments.
“The race for publication has accelerated since the early 2000s“, abounds Jérôme Lamy, historian and sociologist of science, who did not take part in the work.
“If the notion of +disruptive+ discovery is a little difficult to define, this study precisely puts the finger on a global trend pushing researchers to +split up+ their papers to improve their performance indicator and find funding.“, continues the CNRS researcher, interviewed by AFP.
That “atomization“work ends”weaken research, which takes fewer and fewer risks“, analyzes the expert.
To address this, the study calls on national research agencies to prioritize quality over quantity, and to free up more time for scholars to think deeply about big topics.