Event

Abstract: We live in a society. Human societies consist of unrelated individuals with little or no history of past interaction, coexisting in relative harmony. We help each other, even at personal cost; we forgo opportunities to cheat each other for personal gain. Why? Traditional mechanisms for explaining the proliferation of cooperative behavior, such as kin selection, group selection, or tit-for-tat, struggle to answer this question. But reputation can. We care about each other's moral standing, we seek the approval of our peers, and we price the risk of being assigned a bad reputation into our actions. Reputation is even more salient in a polarized, digital world; our behavior is subject to constant scrutiny, and information about us spreads very quickly.

How does reputation shape the evolution of cooperative behavior via social learning? What role does gossip play in incentivizing good behavior? And what kinds of moral rules are likely to evolve? I use evolutionary game-theoretic models to answer these questions. I also weaponize the machinery of adaptive dynamics, a well-worn set of tools in evolutionary and ecological theory.

My results are markedly at odds with existing research on the subject. I find that gossip, which effectively synchronizes players' views of each other's reputations, is a necessary precondition for cooperation; without gossip, it's pointless to think about reputation at all, and you might as well just defect all the time. Conversely, moral rules that are too tolerant of disagreement about reputations cannot sustain any level of gossip. Fortunately (depending on your point of view), the only rule that can evolve is a "cancel culture"-type rule that punishes those who reward bad behavior––and thus is intolerant of disagreement about reputations. I comment on some implications for legal systems, moral philosophy, and social justice.

Twitter/X: @nashintasapaino