Abstract

This paper examines the apparent tension in Michael Smith’s The Moral Problem between his commitment to convergence in ideal desires and his acceptance of agentrelative reasons, particularly those grounded in first-personal perspectives like the parent-child relationship. While Smith maintains that ideal desires are agent-invariant and converge on what is universally desirable, he also endorses agent-relative reasons that imply agent-centered normative commitments. I argue that resolving this tension requires rethinking convergence. Specifically, I propose extending the first-personal („de se“) nature of agent-relative reasons to the objects of convergence, which I term „de se aims.“ By recognizing these aims as value bearers, we can reconcile agentrelativity with the universality of desirability, preserving Smith’s broader metaethical commitments. The proposal avoids the pitfalls of agent-relative value theories and illuminates the role of perspective-dependent aims in systematic justification.