Science Is Taking Fewer Big Leaps, 56-Million-Paper Study Shows

A new study analyzed 56 million scientific papers published between 1960 and 2025 and found something surprising: science is becoming more conservative. When scientists borrow ideas from other fields, they now tend to stick to closely related areas rather than reaching into completely different disciplines. Think of it as staying in your neighborhood instead of traveling to a foreign country for inspiration.

The authors call this a shift in how knowledge is recombined — how different ideas are mixed together to create new ones. They measured the 'semantic distance' between papers that cite each other (how far apart their topics are) and found that this distance has been steadily shrinking. In the past, a physics paper might inspire a breakthrough in biology. Today, that kind of crossover is less common. Instead, researchers prefer to build on overlapping knowledge bases where the ground is already familiar.

This trend matters because many of history's biggest breakthroughs — from the internet to vaccines — came from combining ideas from distant fields. The study also found that research is becoming more dependent on older literature and that the role of 'broker' (someone who connects different fields) is more spread out, but less powerful. No single lab or researcher is making those big leaps anymore.

The researchers warn that this could have real consequences for science policy. Funding agencies that want to encourage fresh ideas may need to design grants specifically for high-risk, cross-disciplinary projects. Otherwise, they say, science may keep playing it safe — and we could miss the next game-changing discovery.