Tyler Miniati
University of Florida
Whether abductive reasoning (more commonly known as “inference to the best explanation”) can be synthesized with Bayesianism is a central question in contemporary philosophy of science and epistemology. A common proposal is that abduction is a (perhaps heuristic) method for estimating or constraining the Bayesian probabilities of explanations. In “Can there be a Bayesian explanationism? On the prospects of a productive partnership” (2017), Frank Cabrera casts doubt on this proposal by arguing that the proposed explanatory virtues of mechanism and precision, which measure different aspects of an explanation’s informativeness, are not positively relevant to explanations’ probabilities. Contrary to Cabrera, I will argue that explanatory completeness, a measure of informativeness which combines elements of precision and mechanism, is positively relevant to the comparative prior probabilities of explanations. On this account, a hypothesis accepted as an explanation carries an implicit commitment to the existence of a complete explanation containing the hypothesis. This commitment should affect the prior probability assigned to the hypothesis; I will argue that explanatory completeness reflects the degree to which the complete explanation can be expected to differ from the hypothesis, and thus the epistemic risk of the additional commitment. Thus, other things being equal, a more complete explanation should receive a higher prior probability than another. As Cabrera’s arguments can be adapted to apply to completeness, I will directly respond to them on two counts: that they neglect the additional commitments involved in accepting a hypothesis as an explanation, and that they apply only to special cases which will not occur in actual instances of scientific reasoning. Thus, through the virtue of explanatory completeness, abduction can favor informative explanations in a manner compatible with Bayesianism.
Chair: tba
Time:
Location:
SOPhiA 2026 