A January 2026 Royal Society Open Science paper offers the first game-theoretic explanation for a puzzle that discipline-by-discipline surveys of authorship convention have never fully resolved: why do some fields settle on author-order rules that ignore who actually did the most work — and why does that choice measurably discourage collaboration? “The evolution of scientific credit: when authorship norms impede collaboration”, by Toby Handfield (Monash University) and Kevin Zollman (Carnegie Mellon University), was published in Royal Society Open Science, volume 13, issue 1, article 251301, on 7 January 2026 (DOI 10.1098/rsos.251301), open access under CC BY 4.0.
CASRAI already documents what authorship-order conventions look like field by field — see our guide to authorship-order conventions by discipline. This paper is not more of that catalog. It supplies a mechanism: an evolutionary game-theoretic account of why contribution-insensitive conventions emerge in the first place, and a second model showing how they create friction that a purely descriptive comparison of norms cannot capture.
Two authorship regimes, one open question
The paper starts from a distinction familiar to anyone who has compared authorship practice across fields: contribution-sensitive systems, where first-author position tracks who contributed most and the rest of the byline reflects relative input (common in the biomedical and natural sciences), versus contribution-insensitive conventions such as strict alphabetical ordering or an automatic senior-author-last slot (common in economics, mathematics, and much of the humanities). What had not been formally modeled is why a discipline ends up with one regime rather than the other, or what that choice does to researchers’ willingness to collaborate once the convention is in place.
The mechanism: a “Red King” dynamic
The authors’ first model borrows a concept from evolutionary biology — the Red King effect, in which the slower-adapting party in a mutualistic relationship ends up capturing more of the benefit, the inverse of the better-known Red Queen dynamic where faster adaptation wins. Applied to authorship, the model shows that contribution-insensitive norms tend to evolve when the researchers who stand to lose the most positional advantage — typically senior authors managing larger labs, bigger grant portfolios, and heavier reputational stakes — face the strongest pressure to adapt their strategy. Because they have more to lose from a failed or contested collaboration, their behavior shifts fastest toward the more accommodating, credit-sharing convention. Junior researchers, facing lower per-collaboration stakes, adapt more slowly and hold their preference longer, which pulls the overall population toward the junior-favorable equilibrium (alphabetical order, or a fixed non-first slot for the senior author) rather than the reverse.
That is the paper’s central, counterintuitive finding: contribution-insensitive norms are not evidence that a field discounts individual credit as a matter of culture or fairness ideology. They can emerge specifically because senior, high-status researchers are the ones under the most adaptive pressure to accommodate — the exact opposite of the assumption that senior researchers set self-favoring conventions and junior researchers simply inherit them.
The cost: why the convention then impedes collaboration
The second model asks what these conventions do once established. Here the paper’s finding is more directly practical for research-integrity offices and PIs: contribution-sensitive norms outperform contribution-insensitive ones at sustaining valuable collaborations, because a norm that ignores actual contribution generates two related sources of friction. A researcher who did disproportionately more work has no byline mechanism to have that recognized, which discourages the kind of asymmetric, high-effort collaboration that produces the best science; and where two contributors made genuinely comparable efforts, a norm that assigns credit by an arbitrary rule (alphabet, seniority) rather than by contribution can still register as unequal treatment between near-peers. Both effects push some potential collaborations toward not happening at all, or toward being restructured to avoid the norm’s blind spots — which the authors frame as contribution-insensitive conventions functioning as a real institutional friction on scientific productivity, not a neutral local custom.
What this adds to CASRAI’s existing coverage
CASRAI’s discipline-by-discipline guides describe the conventions themselves: alphabetical-order economics and mathematics, the AEA’s certified random author order system built specifically to work around alphabetical bias, law review single-authorship norms, and computer science’s conference-vs-journal split. None of those pages previously had a formal account of why a field lands on one convention rather than another, or a quantified argument for what the convention costs in foregone collaboration. This study is the first to supply both, and it reframes contribution-insensitive conventions as a design choice with a measurable downside rather than a values-neutral disciplinary habit.
It is also a useful complement to the ICMJE’s four-part authorship criteria and the CRediT taxonomy, both of which are explicitly contribution-sensitive standards in the paper’s terms — they tie authorship/credit to what was actually done rather than to a fixed positional rule. The paper’s finding is consistent with the general rationale behind both frameworks: making credit legible to actual contribution is not just a fairness argument, it is now also backed by a formal account of why it should produce more (and more willing) collaboration.
A note on scope
This is a theoretical, model-based paper — evolutionary game theory, not an empirical survey of researchers’ behavior or attitudes. Its claims are about what the models predict and formalize, not about measured outcomes in a specific field or institution. Readers who want the underlying assumptions, payoff structures, and full derivations should consult the paper directly; this page summarizes its argument and does not attempt to reproduce or extend its formal results.
Frequently asked questions
What is the “Red King” dynamic in this paper?
It is an evolutionary-game-theory concept, borrowed from biology, describing how the party with the most to lose from a failed interaction (here, senior researchers with larger collaboration portfolios) faces the strongest pressure to adapt, and so ends up conceding positional advantage — the opposite of the more familiar Red Queen dynamic where faster adaptation wins.
Does the study say alphabetical or senior-last ordering is unethical?
No. It argues these conventions are a predictable evolutionary outcome of the incentives senior researchers face, and separately models them as less effective than contribution-sensitive norms at sustaining collaboration — a productivity argument, not primarily an ethics claim, though it has ethics-adjacent implications for how institutions design authorship policy.
Who wrote the study and where was it published?
Toby Handfield (Monash University) and Kevin Zollman (Carnegie Mellon University), published in Royal Society Open Science, vol. 13, issue 1, article 251301, 7 January 2026, DOI 10.1098/rsos.251301, open access.
How does this relate to CASRAI’s discipline-by-discipline authorship guides?
Those guides catalog what conventions exist per field; this paper explains a mechanism for why contribution-insensitive conventions emerge and models their downstream cost to collaboration — see our guide to authorship-order conventions by discipline for the descriptive picture this paper complements.
Editorial note: this page summarizes the paper’s stated models and findings as published in the open-access article and verified against the publisher’s Crossref metadata. It does not reproduce the paper’s formal derivations; readers needing the full mathematical treatment should consult the primary source.
