Maybe scientific progress isn’t slowing, after all

颠覆性下降的质疑

One notable contribution came in 2023, when Michael Park, then a PhD student at the University of Minnesota, and his colleagues published a paper in Nature. It analysed citation patterns in 45m scientific papers and 3.9m patents and concluded that the “disruptiveness” of both had fallen off a cliff since the 1950s. This was widely reported (including in T he Economist) . Its findings found their way into Congressional hearings and speeches by White House officials.

On August 12th came a twist. Nature published a follow-up paper arguing that Dr Park and his colleagues’ conclusions were mostly illusory, caused by problems with the data set they had analysed. Correct those, argues Vincent Holst, a PhD student at Vrije University, in Brussels, and his co-authors, and most of the decline in disruptiveness goes away (see chart).

CD指数与数据误分类

The original paper’s claim hinges on something called the consolidation-disruption index, or CD index, which aims to measure how much a given paper or patent shakes up its field. It compares a paper’s references with those of papers published later. A paper (or patent) is classed as “consolidating” if subsequent papers cite both the paper itself and the earlier research that it refers back to. It is deemed “disruptive” if future papers cite the paper itself but ignore its own references—the assumption being that a paper like that has shaken things up so much that older work has become irrelevant. The CD index runs from -1 (maximally consolidating) to 1 (maximally disruptive).