In Knowledge in a Social World, Alvin Goldman presents a framework to quantify the epistemic effects that various policies, procedures, and behaviors can have on a group of agents. In this essay, I show that the framework requires some modifications when applied to agents with credences. The required modifications carry with them an interesting consequence, namely, that any group whose members disagree can become more accurate by forming a consensus through averaging their credences. I sketch a way that this result can be used to show that individual norms of rationality and group norms of rationality can dictate conflicting behaviors for the members of some groups. I conclude by discussing how some of the assumptions used to generate the consensus result might be loosened.
MATTHEW KOPEC is a doctoral candidate in the Department of Philosophy at the University of Wisconsin–Madison. His research interests include group rationality, probability theory, epistemic normativity, theory testing in the social sciences, and the evidential and ethical issues surrounding the use of race in the biological sciences.