By Sarah Steimer
As scientists age, they become more creative in the ways they find links between previously unconnected work; however, their innovation is limited by their desire to tap familiar references. Younger scientists, on the other hand, are more likely to instigate disruption in their field, replacing established ideas with new ones. While both novelty and disruption are key to innovation, science has become increasingly dependent on its aging core — threatening the ability to disrupt.
A new policy article published in the journal Science uses large-scale data and novel deep learning measurements to measure the impact of scientific aging. The data analyzed show how younger scientists tend toward disruptive contributions while older scientists engage in combinatorial innovation with aging works. According to the authors, these findings highlight the benefits of policies that reimagine funding and encourage independent support for young scientists, unexpected collaborations, among other interventions.
“As scientists grow older, their science ages, too — the work to which they anchor, and draw inspiration and expectation,” says study author James A. Evans, the Max Palevsky Professor in Sociology and Data Science, and director of the Knowledge Lab. “We call it intellectual aging because it’s a social as well as a biological process.” He gives the example of fans preferring to see their favorite, aging musician perform the hits that made them famous, rather than their new music. “Well-known scholars are pushed to return to the work that made them well-known, which accelerates their aging and orientation to older work.”
Evans and his coauthors — Haochuan Cui, Yiling Lin, and Lingfei Wu of the University of Pittsburgh — began with a core question centered on how academic age influences creativity. They began by distinguishing between two measures of creativity: novelty and disruption — both of which drive progress, but in different manners. Novelty extends established paradigms by revealing hidden complementarities between ideas; disruption, on the other hand, overturns ideas by exposing hidden substitutions of old ideas with new ones. They identified name-disambiguated scientists, landing on a sample of 12.5 million scholars from various fields who have published between 1960 and 2020, each with three or more publications. They defined academic age as the number of years since a scientist’s first publication, and followed how their primary knowledge sources, captured by average reference age, changed over time. To assess novelty, the researchers looked at how papers draw upon previously unconnected domains, diverging from conventional citation patterns. To assess disruption, they considered how subsequent papers cite a focal paper but not its references, illustrating when new ideas displace older ones. They corroborate these patterns with scientific text, reviews, and the analysis of many natural experiments.
The researchers note in the paper that their analyses allowed them to examine the life cycle of scientific aging: As scientists move from being trainees to principal investigators, their roles and constraints also notably change. Spending time on leadership, administrative, and reviewing responsibilities can limit the time they have to stay up-to-date on research, which in turn affects how frequently they renew their research references. This points to how individual aging also has collective consequences, as senior scientists’ preferences can shape their collaborators’ citation practices. Older peer reviewers may also direct authors toward citing familiar or preferred work. To consider this potential group effect, the team explored both team dynamics and peer-review dynamics.
Additionally, the team analyzed the impact of the 1994 U.S. Supreme Court ruling on mandatory retirement, which lifted a rule that allowed colleges and universities to legally require faculty to retire at age 70 and any related requirement, and led to an accelerated aging in American science. They also explored how unexpected collaborations — which occur when scientists move to a new university, as a result of their new proximities to scientists — affect citation and attention patterns. They find an aggregate effect at the level of teams and fields, where as the average age grows, a superlinear aging of ideas follows, with young people pressured to adopt older ideas.
The results of the studies are the first to offer careful measurements of innovation over time, versus prior speculative and small-scale work, and they contribute to three areas of innovation study. First, it clarifies why previous research on age and creativity has been mixed, while showing that aging enhances combinatorial innovation but limits disruptive breakthroughs. Second, the research connects these cognitive dynamics to team structure, providing insight into why larger, hierarchical teams tend to innovate less, illustrating the social foundations of intellectual aging. Lastly, the researchers identified the Nostalgia Effect, suggesting that, on average, the age of a scientist’s references increases by about one month per career year — shaping how scientists remember, connect, and sometimes struggle to move past ideas over their careers.
Evans and his coauthors offer several policy considerations as well. To begin with, they suggest that funding and promotion systems should promote diverse pathways that enable both continuity and renewal, each of which contribute to scientific progress. Next, as it relates to the lifting of mandatory retirement, they recommend science policy should assess how funding mechanisms, tenure structures, and retirement norms affect the age distribution of fields along with patterns of innovation. Third, they note that age matters for global competition: Younger scientific workforces such as China and India tend to produce more disruptive work, while older workforces such as in the U.S. and Japan are more likely to incrementally integrate and recombine knowledge. In the U.S., immigrant scientists (younger on average) often help offset the aging workforce here.
“Our scientific workforce, like that of the UK and Japan, is growing ever older, and increasingly focused on old ideas and work, leading to a lower and slower churn of new ideas,” Evans says, noting that he hopes the paper helps shape discussion around immigration. “This has measurable consequences for unfolding science.”
Lastly, the paper emphasizes that team-level actions are an immediate solution to concerns about scientific aging. Institutions are capable of expanding early-career principal investigator and co–corresponding author roles; encouraging intergenerational, flat collaborations; lowering barriers to mobility and cross-institutional work; and recognizing disruptive contributions alongside recombinational synthesis.
“Yes, older scientists more creatively combine the old elements of their field, but they are largely immune to new ideas and actively fight against them,” Evans says. “Memory is important, but it resists change and transformative discovery.”
That said, the paper shows the key role that aging scientists and their creativity play in innovation — but they can be a predictable liability as the aggregate age continues to rise.
This story was reposted from the Chicago Center for Computational Social Science
People
