What is guilt beyond a reasonable doubt (BARD) for a Bayesian? Is thinking of BARD in terms of probabilities a nonstarter? I propose an account of BARD compatible with Subjective Bayesianism that rejects the view that BARD is met by a threshold probability. BARD is a judgment, not merely about the credal state the factfinder endorses as her own (i.e. not merely as one’s own credence in guilt), but as about alternative possible credences, specifically those the factfinder does not endorse, but finds reasonable. To this end, I employ a Bayesian framework, expounded by Lange (1999), that permits revision of past prior probability assignments. Such a framework presupposes a point of view free from one’s prior from which a prior is judged. A trier-of-fact asks whether doubt persists among any reasonable starting point one might take; if it does, acquittal is warranted.
