NADIR, a studio in the RISCON family, publishes question–judgment pairs about decline, filed under a research frame that runs from a decade of erosion through a collapse in 2026 to a turning point in 2027. Every pair has the status of a hypothesis; the frame's ending is not an observed fact, and the studio calls its judgments and options a model's conditional proposals. This paper reads the sixteen hypothetical decision exercises that the studio files under fairness and exchange after the erosion phase. They are unreviewed and set in two invented situations: a final settlement in which extra labour and personal costs that no official record kept are raised, and a restart in which those who joined first and those who joined late drift apart in authority and reward. No pair names a real person, organisation or event, and none is treated as a finding about the world. The paper asks how the set frames what is owed, and to whom, at an organisation's end and at its new start.
Four findings follow. The exercises choose no principle of distribution and turn instead to the procedure by which a claim is heard. They treat the record as both the measure of a claim and a source of unfairness. Their recurring move is to keep apart accounts that a settlement tends to merge. And the condition for revising a settlement is placed in reasons kept and in a procedure for objection, while the options beside the questions almost never speak in the settlement's own terms.