Scientists disrupt less but more persistently with career aging

Disruption describes the extent to which a scientific work overturns the prior knowledge on which it builds and leads to a new line of research. A scientist's tendency to produce disruptive work has been found recently to decline over the career. However, the disruption index captures only how strongly a paper overturns prior work and ignores whether that disruption is later accepted or overturned by subsequent research. The persistent disruption (PD) framework adds this second perspective, identifying work that disrupts prior research and is then consolidated rather than displaced by later work. We examine how PD evolves over the career and find the opposite pattern: The probability of producing a persistent disruptive paper increases with career age. Senior scientists disrupt prior work less often compared to young, but once they do, the disruption tends to persist rather than being quickly superseded. Our results suggest that this reversal feature is due to the experience and we corroborate it across several dimensions of recognition. We also show that, at a fixed level of disruptiveness, the more senior the scientist is, it attracts more citations, social-media attention, and research grants. Nobel laureates also show delayed persistent disruptions. They exhibit high disruptive ability early in their careers, yet their prize-winning work appears years later. This lag further indicates that experience is needed to turn disruptive ability into the field's most celebrated contributions.