A Conundrum in Bayesian Epistemology of Disagreement.” Unpublished manuscript

The proportional weight view in epistemology of disagreement generalizes the equal weight view and proposes that we assign to the judgments of different people weights that are proportional to their epistemic qualifications. It is known that (under the plausible Context-Free Assumption) if the resulting aggregate degrees of confidence are to constitute a probability function, they must be the weighted arithmetic means of individual degrees of confidence, but aggregation by the weighted arithmetic means violates the Bayesian rule of conditionalization. The double bind entails that the proportional weight view is inconsistent with Bayesianism. The paper explores various ways to respond to this challenge to the proportional weight view