At the Intersection of Language, Logic, and Information

This paper offers an initial investigation into how the number of choices available to individual agents may influence choice at the group level by formalizing and simulating a social version of the No Alternatives Argument (NAA). The Social NAA assumption predicts that strength of belief in the most strongly held hypothesis in a group of agents will increase when the number of available hypotheses decreases. Social network simulations using connected Bayesian networks show that this assumption can be violated, but infrequently. Implications of the Social NAA assumption and when it holds in social networks are discussed, and future work is outlined.