Is science and technology becoming a bit meh?
Discovery and technological innovation appear to be slowing over time.
Happy new year to all my readers!
As far as you lot are concerned, my new year resolution is to grow the audience of this blog, and do paying readers the service they deserve by publishing far more on it. If you like what you read here, please consider promoting this Substack via your own websites and social media channels.
To begin 2023, I give you a report that, according to an analysis1 of 45 million research papers and 3.9 million patents published over the past 60 years, it appears that discovery and invention are slowing down and becoming less ‘disruptive’. One image brought to mind is of a golden age of science and engineering slowly atrophying, but this is a ridiculously simplistic perspective.
Progress in science and technology is the result of theoretical breakthroughs and engineering advances. In the modern age we have grown accustomed to an exponential growth in scientific and technical knowledge, and for a tangible example one only has to look at the rapid rise in computing power available to the masses. But even here you may have thought that in recent years nothing much has changed, and each year’s smartphone upgrade is a bit underwhelming.
Anecdotal evidence of techno-atrophy is all well and good, but is there something deeper going on? According to Minnesota sociologist Russell Funk and his colleagues Michael Park and Erin Leahey, research paper and patent citations are becoming less disruptive across all disciplines. Their analysis of the way in which papers and patents change networks of citations draws on theories of scientific and technological changes which distinguish between two types of breakthrough: contributions that incrementally advance existing knowledge; and those that disrupt existing knowledge and propel science and engineering in new directions.
Taking the period between 1945 and 2010, a tool known as the CD index shows that the decline in disruptive advances ranges from 92-100% for research papers, and 79-92% for patents. Since 1980, the decline has been less steep in the life sciences, biomedicine and physics, and most obvious in social sciences and technology. These results are reinforced by analysing alternative indicators such as the diversity of words and use of disruptive language in research papers.
Funk and his coworkers say that the observed decline is unlikely to be due to research quality or changes in citation policies. Rather, they suggest that scientists and technologists are relying increasingly on a narrower set of existing knowledge.
My own conjecture is that research culture and practice has in recent decades become more collective and labour-intensive, which can leave less room for individual creativity. This may account for at least some of the innovation decline.
There has been some attempt to address this through programmes such as the European Research Council that award research grants on individual merit, but to date these have been relatively small in scale. As for the private sector, it rarely innovates in itself, but instead builds on and realises discoveries made in academic and public sector laboratories. It has always been thus, and I suspect always will.
Park et al., “Papers and patents are becoming less disruptive over time”, Nature 613, 138–144 (2023); doi:10.1038/s41586-022-05543-x.