Moral Decision-Making Under Uncertainty

Abstract

Agents facing moral uncertainty can reasonably manage normative risks in a manner analogous to decision-making under empirical uncertainty. This approach, known as moral uncertaintism, relies on standard decision-theoretic principles but faces three central theoretical challenges: the assignment of intermediate probabilities to moral propositions, the commensurability of values across rival moral frameworks, and the threat of an infinite regress generated by higher-order normative uncertainty. Intermediate probabilities for moral propositions are defensible when interpreted as epistemic probabilities, which function as mind-independent expressions of credence rather than objective chances or mere psychological reports. Meanwhile, inter-theoretic comparisons of value differences can be grounded endogenously, as the specific credal objects agents evaluate often consist of paired theoretical structures that inherently specify relative value intervals. Finally, the possibility of higher-order uncertainty concerning decision-rules does not undermine action-guidance or normative coherence. From the deliberating agent’s first-person perspective, decision-rules at adjacent orders of uncertainty do not generate practical conflicts; rather, higher-order evaluations represent iterative attempts at subjective rightness that tend to converge on specific actions or successfully delimit permissible subsets of choices. Moral uncertaintism therefore retains the theoretical resources necessary to guide rational deliberation despite the complexities of normative doubt. – AI-generated abstract.