SOPhiA 2017

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

How to Revise Credences in Light of Disagreement
(Epistemology, English)

According to conciliationists, when individuals encounter disagreement, they ought to revise their credences. But how? Suppose Jack's credence in p is 0.9 and Jill's is 0.4. One possibility is to adopt, for each proposition, the (linear) average of the disagreeing individuals' credences (in this case, 0.65). Another possibility is to take their geometric mean (in this case, 0.6). However, I argue that both linear and geometric averaging, which many conciliationists endorse, are problematic.
More generally, a broad class of rules for determining how disagreeing individuals ought to revise their credences faces the following problem: they cannot distinguish cases involving genuine points of consensus among disagreeing individuals from certain pathological cases where no such consensus exists. In particular, all 'non-holistic' rules face this problem. These are rules according to which how disagreeing individuals ought to revise their credence in p just depends on what credence each individual assigns to p. That is, the credences assigned to propositions other than p are irrelevant.
The problems with non-holistic rules have important implications for the epistemology of disagreement. According to conciliationists, disagreeing individuals ought to revise their credences. This view is tenable only if there is a plausible rule for determining how individuals ought to revise their credences. What I argue is that prima facie plausible rules like linear and geometric averaging are problematic. So, unless there are viable alternative rules that circumvent this problem, we should be pessimistic about conciliationism about peer-disagreement.

Chair: Andrea Togni
Time: 12:25-12:55, 14 September 2017 (Thursday)
Location: SR 1.005

Weng Kin San 
(University of Oxford, United Kingdom)

I am currently a BPhil student at the University of Oxford. My interests are mainly in epistemology, especially formal epistemology.

Testability and Meaning deco