Moral Uncertainty, Pure Justifiers, and Agent-Centred Options

Abstract

Standard decision procedures for resolving moral uncertainty, most notably Maximize Expected Choiceworthiness (MEC), treat moral latitude as an improbable coincidence of numerical ties. By measuring choiceworthiness along a single dimension, MEC generally mandates a single uniquely appropriate action, failing to accommodate agent-centred prerogatives and supererogation when an agent assigns credence to non-consequentialist theories. This deficiency is resolved by distinguishing between two independent dimensions of normative strength: requiring strength, which supports obligations, and justifying strength, which supports permissions. Integrating this distinction into decision-making under normative uncertainty yields the Expected Balance of Reasons (EBR) framework. Under EBR, an option is appropriate if its expected justifying strength is at least as great as the expected requiring strength of any alternative, and uniquely appropriate when its expected requiring strength strictly exceeds the expected justifying strength of alternatives. EBR effectively generalizes MEC to multi-dimensional reason structures, outperforming rival decision rules such as Maximizing Expected Requiring Strength and Minimizing Expected Wrongfulness. This model preserves genuine agent-centred options under moral uncertainty, with direct implications for practical dilemmas such as the moral status of charitable donations. – AI-generated abstract.