Credal sensitivism: threshold vs. credence-one

Download Edit this record How to cite View on PhilPapers
Abstract
According to an increasingly popular view in epistemology and philosophy of mind, beliefs are sensitive to contextual factors such as practical factors and salient error possibilities. A prominent version of this view, called credal sensitivism, holds that the context-sensitivity of belief is due to the context-sensitivity of degrees of belief or credence. Credal sensitivism comes in two variants: while credence-one sensitivism (COS) holds that maximal confidence (credence one) is necessary for belief, threshold credal sensitivism (TCS) holds that belief consists in having credence above some threshold, where this threshold doesn’t require maximal confidence. In this paper, I argue that COS has difficulties in accounting for three important features about belief: i) the compatibility between believing p and assigning non-zero credence to certain error possibilities that one takes to entail not-p, ii) the fact that outright beliefs can occur in different strengths, and iii) beliefs held by unconscious subjects. I also argue that TCS can easily avoid these problems. Finally, I consider an alleged advantage of COS over TCS in terms of explaining beliefs about lotteries. I argue that lottery cases are rather more problematic for COS than TCS. In conclusion, TCS is the most plausible version of credal sensitivitism.
PhilPapers/Archive ID
GAOAP-2
Upload history
First archival date: 2020-08-04
Latest version: 3 (2020-11-22)
View other versions
Added to PP index
2020-08-04

Total views
137 ( #35,134 of 2,427,989 )

Recent downloads (6 months)
64 ( #11,344 of 2,427,989 )

How can I increase my downloads?

Downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.