Seminar: Epistemic Social Dilemmas: Strategic Interactions in Debate and Peer Review

The pursuit of truth in academic discourse often requires cooperation, yet is plagued by tensions between individual and collective interests. This talk examines how strategic behaviour can undermine epistemic goals in two critical contexts: rational debate and academic publishing. Drawing on game-theoretic models, I explore how debaters face competing motivations between discovering truth and minimising the cognitive costs of changing their beliefs. This creates cooperative dilemmas where individually rational strategies systematically undermine collective truth-seeking. In the second part, I examine how different incentive structures affect journal behaviour in peer review, showing that journals incentivised by rejection rates have perverse incentives to maintain lower-quality peer review than those focused solely on publication quality. Both cases illuminate a broader theme: individual rational choices in epistemic contexts can systematically undermine the collective pursuit of knowledge. These findings have practical implications for designing better institutions for academic discourse, journal policies, and peer review systems. I conclude by reflecting on how AI and automated systems may reshape these strategic dynamics and create new opportunities for improving our collective epistemic practices.

Julian García is an Associate Professor in the Department of Artificial Intelligence and Data Science at Monash University in Australia. He uses game theory and computational models to study how individuals come together to achieve collective goals, with applications including cooperation, multi-agent systems and social behaviour dynamics. Julian completed his PhD at Vrije Universiteit Amsterdam in 2011, focusing on evolutionary game theory, and held a postdoctoral position at the Max Planck Institute for Evolutionary Biology before joining Monash in 2014. His research features in Nature, PNAS, Journal of Economic Theory, PLOS CB amongst others.

Alternatively, join online:

Microsoft Teams
Meeting ID: 314 744 371 653
Passcode: FD7uW2ca