The Shape of Inheritance

Introduction

This note begins with a single observation: some ethical states feel different from others not because they produce different outcomes, but because they recruit different kinds of participation from the actor. Asking a serious question is not neutral. Intervening is not neutral. Recognizing an available intervention and declining it is not neutral. The claim of this note is that these states differ chiefly in burden — in what they require of the actor while occurring, and in what remains with the actor afterward.

The vocabulary that follows — recruit, discharge, hold open, carry — is used analogically, as a way of describing structure, not as a claim about energy, attention, or any other measurable quantity. This isn't a disclaimer to be stated once and set aside: a vocabulary built to describe burden will tend, on its own, to get read as physical regardless of the disclaimer, and the better response is to keep checking the language against that tendency rather than assume one sentence has settled it.

Two documents inform what follows. The first is Notes from a Window, whose Section 6 ("The Problem of Others") describes "do no harm" as a resting state — "a heartbeat, not a commandment" — rather than a prescriptive rule. The second is Understanding as Ethical Intervention, whose Section 1 sets out the Question of Action: when I intervene with understanding, am I completing a process, or ending it? This note does not claim a structural identity between the two documents; their relationship here is interpretive, not architectural. What follows reads each through the same lens and reports what comes back.

Three States, Not Two

Most ethical frameworks model decision-making as binary: act, or do not act. Read carefully, both source documents imply a third state.

In the framework of Understanding as Ethical Intervention, the Question of Action is precisely what performs the first move, lifting the actor from default rest into inspection. The answer to that question does not by itself decide whether the actor moves to action or remains in inspection; it decides what kind of action is on the table. Where the honest answer is "completing," the available intervention extends a process toward its own stopping point. Where the honest answer is "ending," the available intervention would terminate the process before that point is reached — and recognizing this raises the burden of justification for acting. Whether the actor meets that burden and acts anyway, or declines and remains in inspection, is a further question, not a foregone conclusion of the diagnosis.

Inheritance, Not Symmetry

Completing and terminating look, from outcome alone, like two branches of the same act: both close a question, both discharge into the world, both return the actor to a new rest. That symmetry holds for consequence. It doesn't hold for what the actor carries afterward.

Intervention — whether completing or terminating — inherits closure. Closure inherits authorship: a possibility that was open is no longer open, and the actor's hand is now part of how it closed. This is true regardless of whether the intervention was correct, necessary, or justified. Justification changes what the actor is responsible for, not whether the actor carries it. A correct action can still become part of the actor who took it; a necessary intervention can still leave residue. This is the burden Understanding as Ethical Intervention Section 3 names "the burden of restraint" from the other side — the side that chose to act rather than withhold.

Restraint inherits differently. Restraint is the condition of having taken the first move, recognized that the available intervention would terminate rather than complete, and chosen — under that heavier burden of justification — not to take it. Termination remains available; the actor declines it. Nothing closes. The possibility that was opened in inspection stays open, and the actor continues holding it rather than discharging it into either action or a return to rest. Restraint does not relieve itself with time; it must be sustained for as long as the situation persists, which is what makes it a burden in the strict sense rather than a temporary discomfort: action and rest both resolve, restraint does not.

Restraint, though, is not one thing. Understanding as Ethical Intervention Section 8 distinguishes scaffolding from abandonment: declining to terminate can mean staying actively present — supporting a process without taking ownership of where it stops — or it can mean simply withdrawing, leaving the question open because no one is tending it rather than because someone has chosen to keep tending it. Scaffolding inherits in the sense developed above: it stays answerable to a question it continues to hold. Abandonment inherits differently again — not closure, since nothing closes, and not continued answerability, since no one is answering. What it inherits is the fact of having let a held question go untended: the question persists, but the actor's claim on it does not, and the actor remains answerable not for the question but for having relinquished it. The burden described above belongs to scaffolding. Abandonment is what happens when that burden is set down rather than carried — which is not the same as never having picked it up.

Neither inheritance is empty. Intervention inherits what it closed. Restraint inherits what it continues to hold open.

What This Clarifies

Why persistent non-resolution is difficult.

Understanding as Ethical Intervention Section 4 defines persistent non-resolution as a limit condition in which explanatory convergence would itself be subtractive — would remove something the phenomenon needs to remain itself. Holding a phenomenon in persistent non-resolution is difficult not because the ambiguity is intellectually demanding but because unresolved participation is difficult to sustain. Every available move — completing, terminating, or simply ceasing to inspect — would relieve the actor of the holding. Persistent non-resolution asks the actor to decline all three, repeatedly, for as long as the case requires it.

Why the heartbeat reading of Section 6 attracts misdiagnosis.

Critics of Notes from a Window Section 6 consistently read "do no harm" as a prescription rather than a description. This is not a failure of attention. Reading description as rule collapses the three states back into two and removes inspection entirely — it lets the reader experience the text as already-closed instruction rather than as an open description requiring ongoing attention. The reader returns to something that feels like rest while believing engagement has occurred, because closure is the move that relieves holding, and a rule is already closed in a way a description is not. The convergence on the misreading is the structural pull operating on the document's reception, not a deficiency in any particular reader.

What an evidence-locus calibration object actually does.

Understanding as Ethical Intervention Section 9 treats the Roman dodecahedron as a calibrating case for evidence-locus persistent non-resolution. An object of this kind does not pose a puzzle to be solved; it imposes a standing requirement to keep holding. The recipient must either continue carrying the object open or relieve that requirement by closing it — assigning it a function, a use, a solved status. Recipient experiments documenting that the closure impulse overrides explicit instruction not to resolve the object are evidence of exactly this asymmetry: closing is the move that discharges, holding open is the move that doesn't, and most recipients, given the choice, take the move that discharges.

Closing Note

This note does not propose a new mechanism. It names a shape: questions recruit, interventions inherit closure, and restraint — when sustained as scaffolding rather than left to empty into abandonment — inherits continued answerability for the question. Inheritance is the load-bearing concept here because it does not need to specify what restraint is made of — cognitive, attentional, affective, social, or something not yet named — in order to describe what it does. Whether this shape deserves a place in either framework is a separate question, to be settled the same way as any other claim here: not by how cleanly it reads, but by whether it continues to earn its keep across the cases the frameworks are already being used to read.