In defense of beliefs as stably high credences: how stability theory of belief avoids the problem of conviction based on purely statistical evidence



Journal Title

Journal ISSN

Volume Title


Tartu Ülikool


In the thesis, I defended STB from Staffel’s criticism where she argues that STB provides neither a necessary nor a sufficient condition on rational belief. The reason for this is that STB is argued to allow for a belief in guilt based on purely statistical evidence, which is counterintuitive. I showed that this inference fails: conviction based on a narrative for guilt ensures that the conviction is not issued based on purely statistical evidence. Since in the cases that involve narratives for and against guilt STB (complemented with pragmatic constraints on accusing narrative) performs fine, the theory avoids the problem of purely statistical evidence. In the end, I dismissed a concern that STB when complemented with pragmatic constraints fails to provide a necessary condition on rational belief. A corollary of my defense of STB from Staffel’s criticism is an account of how STB can be fruitfully deployed in a court context. This makes room for further research on whether the combination of STB and the pragmatic constraints on accusing narratives provide a promising explication of the beyond reasonable doubt standard and, hence, a novel candidate to modeling legal standards of proof in both probabilistic and non-probabilistic ways.