Survey finds scientists disagree about definition of science more than public
A survey of 91 scientists and 99 members of the public found that scientists were less aligned on what science means, often resisting single definitions even within the same subfield. The public more consistently saw science as a reliable path to objective truth, while scientists emphasized uncertainty, creativity, and judgment. The gap may affect how scientific findings are communicated, trusted, and used in policy and AI-driven research.
Researchers at the Santa Fe Institute and Princeton University surveyed 91 scientists and 99 members of the public, asking them to evaluate 18 historical descriptions of science, including science as detective work, a pursuit of truth, and a mirror reflecting reality. The study, led by Marina Dubova, found scientists rarely accepted any single description without reservations, even within the same subfield. The public, by contrast, expressed more consistent views, tending to see science as a dependable route to objective truth.
The evidence came from ratings of 18 definitions drawn from different scientific traditions. Both groups agreed on broad themes: curiosity, exploration, and mystery-solving. The split appeared when participants considered whether science is mainly objective discovery, creative judgment, or rule-following. Some scientists emphasized improvisation and personal judgment; others stressed established procedures. The public more often valued certainty and objectivity, while scientists highlighted uncertainty and creativity. The methods note: the study used a cross-sectional survey of 190 participants, so it captures reported attitudes, not a causal account of why views differ.
The finding fits a longer debate about science as both a method and a social practice. Textbook accounts often present science as a neat ladder from hypothesis to fact, but working researchers routinely navigate incomplete data, competing models, and judgment calls. The study also found the public's views more closely resembled physicists and biologists than psychologists and social scientists. Dubova suggests this may reflect which fields dominate science education, though she calls that explanation tentative and says more work is needed.
These differences can shape how scientific claims are communicated and received. If the public expects firm answers while scientists see uncertainty as normal, changing findings may look like failure rather than correction. The stakes extend to funding decisions, policy debates, and trust in research institutions. The paper also raises a practical question for AI tools used in discovery: if there is no shared definition of science, systems meant to automate research may embed one contested view over another.
The findings affect how researchers communicate uncertainty and how the public judges changing scientific conclusions, with implications for trust, funding, and the design of AI tools used in discovery.
Sources (1)
- [1]Even scientists can't agree on what science actually is(https://www.sciencedaily.com/releases/2026/11/261119194541.htm)