Uncertain Values: An Axiomatic Approach to Axiological Uncertainty

Abstract

Decision-making under evaluative uncertainty requires a systematic framework for evaluating options when agents are uncertain about which axiology is correct. Expected Value Maximisation provides a compelling approach, yet it faces fundamental difficulties regarding the conceptual coherence of intertheoretic value comparisons, the quantification of credences over moral theories, and value incommensurability. By adapting representation theorems from decision theory, expected value reasoning can be axiomatised and rigorously defended. Within this structure, cardinal intra- and intertheoretic comparisons are operationalised through the formal properties of higher-order value relations rather than treated as primitive metaphysical entities. Intertheoretic comparability is grounded through a constructivist framework wherein comparisons emerge from epistemic norms governing reasonable belief revision and coherence. Furthermore, formal representation techniques allow axiological credences to be derived endogenously from evaluative judgments, resolving the problem of separating probabilities from value scales. While traditional decision-theoretic models require complete orderings, relaxing the completeness axiom accommodates both internal and intertheoretic incommensurability without collapsing into radical normative scepticism. Although this axiomatic foundation successfully vindicates expected value reasoning across an extensive range of axiologies, extending the framework to general deontic uncertainty reveals persistent structural obstacles, as non-consequentialist constraints systematically conflict with standard axioms such as independence and continuity. – AI-generated abstract.