Search for: All records

Abstract The Disruption Index (D-index) provides the first quantitative framework for identifying breakthroughs in science and technology. As its use expands, questions have emerged about its meaning, strengths, and limitations. Because the D-index measures how a focal paper competes with its references for citation attention, some worry that it is distorted by historical changes in citation practices. For example, if papers cite more references over time—a trend known as “citation inflation”—then newer papers might appear less disruptive even when equally inventive. We show that this concern is unfounded. Citation counts follow a long-tailed distribution, meaning competition is overwhelmingly shaped by the focal paper and its most-cited reference, while other references are negligible. Thus, the D-index captures whether a paper overturns a dominant idea in its field. The metric is fundamentally relational: It measures competition with predecessors rather than innovation in a vacuum. From this perspective, breakthroughs arise not only from generating novel ideas but also from replacing established ones—much like light bulbs replacing candles. We support this interpretation with mathematical analysis and large-scale bibliometric evidence.