Beyond the binary: Nuanced categories of good, neutral and bad support lasting human cooperation

Gaby Clark
scientific editor

Robert Egan
associate editor

A research team led by Hitoshi Yamamoto has unveiled new insights into how humans build and update reputations in cooperative social interactions. Human societies have achieved remarkable levels of cooperation, facilitated mainly by mechanisms of indirect reciprocity, where reputation and social norms play crucial roles.
While theoretical models have proposed complex, multi-layered systems for how reputation information sustains cooperation, experimental studies often rely on oversimplified binary categorizations. This research aimed to bridge this gap by investigating the type of information and level of granularity required to define and maintain reputation-based cooperation in real-world contexts. The study's appeared in PLOS One on August 8, 2025.
Through scenario-based experiments combined with mathematical modeling, the study revealed that reputations are best understood not as binary (good/bad) but as ternary (good/neutral/bad) categories that change gradually based on cooperative or defective behaviors. Notably, defecting against individuals with bad reputations did not degrade the defector's reputation, suggesting that justified defection is viewed neutrally rather than positively.
Theoretical analyses demonstrated that under certain conditions, a tolerant action rule—cooperating with those of good or neutral reputation and defecting only against those with bad reputations—becomes dominant and stabilizes cooperation.
These findings challenge traditional views of abrupt, binary shifts in reputation and highlight the importance of gradual, nuanced updates that mirror real-world social dynamics. The research offers a more realistic model of how social norms evolve and sustain cooperation, providing empirical evidence to support the theoretical claim that multi-valued reputation systems are crucial for maintaining stable cooperation in large-scale human societies.
Lead author Yamamoto notes that future research will expand on these findings by examining cross-cultural differences in reputation evaluations and exploring how introducing greater reputation granularity may further influence cooperative behavior. Understanding these dynamics could inform the design of effective mechanisms for sustaining cooperation in digital platforms and globalized communities.
More information: Hitoshi Yamamoto et al, Gradual reputation dynamics evolve and sustain cooperation in indirect reciprocity, PLOS One (2025).
Journal information: PLoS ONE
Provided by Rissho University