Colloquium Abstract

Friday, August 01, 2008 • 3:30 PM • Robert N. Noyce Conference Room, SFI

Herb Gintis Emeritus Professor, University of Massachusetts; Professor, Central European University, Budapest; Visiting Professor, University of Siena; and External Professor, Santa Fe Institute

Bayesian Rationality and Social Norms

Classical game theory holds that social behavior can be modeled as the interaction of Bayesian rational agents, and that social norms are Nash equilibria of appropriately chosen games.

I show, by contrast, that social norms are emergent properties of human society, conceived of as a complex, adaptive, nonlinear system governed dynamically by evolutionary processes.

Specifically, social norms are not simply descriptions of equilibrium states, but rather are active causal agents in coordinating social interaction, and humans have evolved psychological machinery for interpreting, and in general conforming to, the social norms that regulate their social life.

The central analytical tool that I use is epistemic game theory, which combines classical game theory with the modal logic of knowledge developed by Saul Kripke and first applied to game theory by Robert Aumann.

The goal of this research is to provide an analytical foundation for a model of human strategic interaction that includes insights from several behavioral disciplines, including economics, sociology, psychology, anthropology, and sociobiology.

Host: J. Doyne Farmer