Radical Pooling and Imprecise Probabilities

Abstract

This paper focuses on radical pooling, or the question of how to aggregate credences when there is a fundamental disagreement about which is the relevant logical space for inquiry. The solution advanced is based on the notion of consensus as common ground (Levi in Synthese 62:3–11, 1985), where agents can find it by suspending judgment on logical possibilities. This is exemplified with cases of scientific revolution. On a formal level, the proposal uses algebraic joins and imprecise probabilities; which is shown to be compatible with the principles of marginalization, rigidity, reverse bayesianism, and minimum divergence commonly endorsed in these contexts. Furthermore, I extend results from previous work by (Stewart & Ojea Quintana in J Philos Logic 47:17–45, 2016; Erkenntnis 83:369–389, 2018) to show that pooling sets of imprecise probabilities can satisfy important pooling axioms.