Ethics, Morality, and Justice as Relational Phenomena: A Stipulative Framework

Abstract

This paper stipulates a vocabulary for treating ethics, morality, and justice as relational rather than absolute: good/evil as a relation stipulated to hold directly between a phenomenon and a living being, right/wrong as a relation stipulated to hold between an action and a rule system, and just/unjust as a relation stipulated to hold between how rules are applied and the code a system sets for itself. None of these three is discovered; each is a criterion fixed by stipulation, differing from the others in what it is stipulated to apply to, not in whether it was stipulated at all. From this vocabulary, the paper develops three tools. Grounding Analysis traces a normative claim to the criterion it actually rests on, distinguishing claims resting on a rule system from claims resting on a direct relation of benefit and harm. The Three-Level Model separates a normative system into its declared goal, its concrete materialization, and its observable consequences — three things ordinary discourse tends to collapse into one, on a division this paper treats as the relevant one rather than derives from the system itself. The Consistency Test compares the three, identifying where a system's materialization or consequences fail to align with what it declares, independent of whether the goal declared is one the analyst shares — though not independent of the analyst accepting this division as the right one to apply. Each tool is developed with attention to what it can and cannot establish. The vocabulary's authority rests on internal consistency and analytical usefulness, not on correspondence to a fact about the world existing independent of the stipulation. None of the three tools instructs a reader toward any conclusion, but none is neutral either: each is built on a choice about what is worth distinguishing, a choice this paper makes and names rather than treats as given. A case examined in detail, early-twentieth-century eugenics, shows the tools' limits directly: they can identify that a system's stated purpose was not served by what it built, and separately, that severe harm resulted for those it affected — but they cannot weigh one finding against the other, or say what should have been done differently. The paper is descriptive rather than prescriptive throughout, offering a vocabulary and a comparison rather than a verdict on which normative systems ought to be adopted.