The search function is not Maga

For once, we must thank the Trump administration, which has forced the American academic world to confront an issue that has long required attention: that of peer review. The proposed solution, however, risks being worse than the problem itself. Let’s start with the news story. A regulation currently under review by the US Office of Management and Budget stipulates that, before any federal research funding is awarded, a political representative must certify that the project ‘advances the President’s policy priorities’. Funding could also be withdrawn should those priorities change.

The modern American research system was founded on very different principles. In 1945, Vannevar Bush, an engineer from my own university, MIT, published the report ‘Science, the Endless Frontier’, proposing a three-way pact. The government would fund research; scientists, rather than politicians, would assess its merit; and society would benefit from it. That model forms the basis of the contemporary American university and of countless inventions in our daily lives – from neural networks to the internet.

Peer review certainly has its limitations. It can turn into ‘peer reassurance’: even we academics are not immune to herd behaviour. Furthermore, in recent years, other pressures have emerged. Various requirements relating to diversity, equity and inclusion introduced during the Biden administration have introduced non-scientific criteria into the evaluation of research – precisely the problem that the current proposals claim to want to resolve.

These are real problems that warrant a rethink. But the solution to flawed peer review is better peer review, not government oversight. Replacing scientific judgement with political alignment risks destroying the very engine of innovation we were discussing earlier. None of the great scientific discoveries of the past was inevitable; on the contrary, each was a long and uncertain gamble on ideas that might well have failed.

This is precisely the sort of research that a system of funding which can be withdrawn at any time risks eliminating: not by openly banning it, but by making it clear to every scientist that a multi-year project can be scrapped as soon as the political climate changes. At that point, the most daring discoveries are not even attempted.

So what should be done? In chats amongst colleagues at MIT over the past few weeks, several ideas have emerged. I proposed dividing the peer review process into two stages. The first would be to assess the methodological soundness of a piece of research, following traditional procedures, perhaps supplemented by artificial intelligence. The second would be to allow the relevance of the work to be discussed openly from the bottom up, enabling ideas to compete over time. Some agencies in New Zealand and Switzerland are already moving in this direction.

If enacted in its current form, the regulation in question would risk undermining American leadership in science. In "Urn Burial", Sir Thomas Browne captured the cruelty of written history. “Erostratus, who set fire to the Temple of Diana, lives on; he who built it is almost forgotten.” History remembers the arsonist more readily than the architect.

For much of the last century, America has been the architect of modern science. It would be a bitter epitaph if, on its 250th anniversary, it were to be remembered not for what it has built, but for what it has destroyed.