Over the past few decades, the number of science and technology research papers published has soared, rising at a rate of nearly 10% each year. In the biomedical field alone, there are more than a million papers pouring into the PubMed database each year, or around two studies per minute.

At face value, that sounds like this is good for science, but despite this tremendous rate of publishing, the rate of ground-breaking discoveries and technological innovations is slowing down significantly, according to a new analysis of millions of papers and patents published in Nature.

The new study revealed that the “disruptiveness” of contemporary science has decreased, rendering ever diminishing returns. In this particular context, authors define disruptiveness as the degree to which a study departs from previous literature and renders it obsolete. In other words, a highly disruptive study is one that completely changes the way we think about a particular topic and renders previous research on the subject obsolete.

Measuring Disruptiveness

Researchers at the University of Minnesota’s Carlson School of Management used citation data from 45 million scientific papers and 3.9 US-based patents to calculate a “CD index”, a measure of disruption which ranged from -1 for the least disruptive work to 1 for the most disruptive. The CD index is basically a proxy for innovation and is computed by looking at the number and quality of citations a paper or patent received five years after publication, assuming that the more disruptive a study is, the less its predecessors would be cited because it is, by then, considered outdated knowledge.

The authors were stunned to learn that the average CD index declined by more than 90% between 1945 and 2010 for research manuscripts, and by more than 78% from 1980 to 2010 for patents. This decline in disruptiveness was observed in all fields and patent types, even when taking into account potential differences in citation practices.

What these figures mean is that research papers and patents have become increasingly conservative, consolidating or building upon previous knowledge, rather than breaking new ground.

Even our language seems to reflect a dwindling spirit of innovation. The authors performed an analysis of the most common verbs in the manuscripts included in their study, finding that whereas research in the 1950s was more likely to use words like ‘produce’ or ‘determine’ that evoke discovery, research in the 2010s was more likely to use terms such as ‘improve’ or ‘enhance’.

Imagine the introduction of the combustion engine, the first commercial aircraft, the polio vaccine, the discovery of DNA and nuclear fission, or the first Apollo-era rockets to reach the moon. We now have better and faster cars, airplanes, and rockets than ever before, and scientists responded to the pandemic with an incredible reaction speed, bringing novel mRNA-based vaccines to the market with unprecedented speed, but these aren’t exactly new technologies, just incremental improvements to previous iterations. Sure, there’s CRISPR and quantum computers which appeared in the 2000s, but compared to the rate of groundbreaking discoveries in the 20th century, we seem to be lagging behind innovation-wise.

The study is the first to “emphatically, convincingly document this decline of disruptiveness across all major fields of science and technology,” according to lead author Michael Park, a doctoral student at the University of Minnesota.

The decline in disruptive research was most pronounced in the physical sciences, such as physics and chemistry. According to senior study author Russell Funk, “the nature of research is shifting” as incremental innovations become more common. But why is this happening?

Why is Disruptiveness Declining?

One theory, known as “the burden of research,” suggests that there is now so much that scientists must learn in order to master a particular field that they have little time left to push boundaries and make truly disruptive discoveries.

As Park explains, this can lead scientists and inventors to “focus on a narrow slice of the existing knowledge, leading them to just come up with something more consolidating rather than disruptive.” Another possible reason for the decline in disruptiveness is the increasing pressure on academics to publish, publish, publish. With the number of published papers often used as a metric for success in academia, researchers may be more inclined to focus on producing a high volume of work rather than taking the time to pursue more innovative, risky ideas.

These findings have raised concerns about the future of science and its ability to continue driving progress and improving our lives. However, it’s important to note that the institution of science is still capable of making major breakthroughs, as evidenced by recent developments such as the use of mRNA technology in COVID-19 vaccines or the measurement of gravity waves in 2015.

So, what can be done to help restore the disruptive edge to science? The researchers involved in the study have called on universities and funding agencies to focus more on quality rather than quantity, and to consider full subsidies for year-long sabbaticals to allow academics the time and space to read and think more deeply. They have also highlighted the importance of encouraging diversity and inclusivity within the scientific community, as diverse perspectives and experiences can lead to new and innovative ideas.