The topic of this project is social beliefs and their informational dynamics, with a particular focus on their truth-tracking potential. The plan is, first, to look at some of the known studies and examples in Social Epistemology, and more concretely the known truth-tracking successes ("wisdom of the crowds") and failures (informational cascades, pluralistic ignorance, group polarization) of standard belief-aggregation mechanisms. Second, we will look at the attempts in the literature to provide explanations for these phenomena (mainly based on the Bayesian theory of belief update). Third, (time-permitting) we plan to investigate possible formal-logical treatments of these phenomena, using ideas and techniques from (probabilistilic) Dynamic Epistemic Logic and Formal Learning Theory.