A Bayesian simulation model of group deliberation and polarization.

Abstract

(from the chapter) This chapter describes a simulation environment for epistemic interaction based on a Bayesian model called Laputa. An interpretation of the model is proposed under which the exchanges taking place between inquirers are argumentative. The model, under this interpretation, is seen to survive the polarization test: If initially disposed to judge along the same lines, inquirers in Laputa will adopt a more extreme position in the same direction as the effect of group deliberation, just like members of real argumentative bodies. Our model allows us to study what happens to mutual trust in the polarization process. We observe that inquirers become increasingly trusting which creates a snowball effect. We also study conditions under which inquirers will diverge and adopt contrary positions. To the extent that Bayesian reasoning is normatively correct, the bottom line is that polarization and divergence are not necessarily the result of mere irrational “group think” but that even ideally rational inquirers will predictably polarize or diverge under realistic conditions. The concluding section comments on the relation between the present model and the influential and empirically robust Persuasive Argument Theory (PAT), and it is argued that the former is essentially subsumable under the latter. (PsycINFO Database Record (c) 2014 APA, all rights reserved)