Compassionate Exit

You have now seen the three conditions. The structural argument for assisted dying does not rest on mercy, although mercy is present. It does not rest on autonomy, although autonomy is present.

It rests on the axioms.

The operator has sovereignty below ε (AP33). The operator’s exit affects primarily the operator. The exit is below ε. Sovereignty holds.

The operator has consent — intact modelling capacity. The operator understands the corridor, the trajectory, the no-return surface. The operator’s request is not a distortion (depression, coercion, temporary crisis).

It is a structural assessment: my corridor has narrowed to zero functional width, my trajectory is toward increased suffering with decreased coupling, and I choose to close the window while I am still the one holding the latch.

The organism’s response, under the axioms, is compassionate facilitation. Not abandonment. Not withdrawal. Facilitation. The organism walks with the operator to the boundary and holds the door. The organism does not push the operator through.

The organism does not lock the door from outside. The organism stands at the threshold and says: this is your window. The latch is yours.

The structural requirements for compassionate exit are derived from the same axioms that derive the correction hierarchy (AP32) and the jurisdiction boundary (AP33):

Requirement 1: Confirmed Active Agency. The operator must be competent — modelling capacity intact. A request made without modelling capacity is not a request. It is a symptom.

Requirement 2: Terminal trajectory. The corridor must be structurally narrowing toward zero, confirmed by independent measurement. The trajectory must be irreversible within current knowledge.

Compassionate exit is not available for conditions with reasonable prospect of corridor stabilisation or widening.

Requirement 3: Sustained request. The request must persist across a temporal gap sufficient to exclude transient distortion. Not a cooling-off period imposed by external authority.

A structural verification that the operator’s assessment is stable — that the request reflects the operator’s considered geometry, not a momentary crisis.

Requirement 4: Below ε. The exit must not produce externalised consequences above the ε threshold.

In practice: the exit does not destabilise dependent operators (minor children, dependents without alternative support) beyond the organism’s capacity to absorb.

If dependents exist, the organism must ensure the floor of care transfers before exit is facilitated.

Requirement 5: Compassionate facilitation. The exit must be administered with the full resources of the organism’s medical and palliative capacity. No pain. No fear. No abandonment.

The operator exits through a held door, not through a window they had to break themselves.

What the axioms prohibit: Exit under coercion. Exit without modelling capacity. Exit as cost-saving measure by the organism. Exit as alternative to palliative care that has not been offered.

Exit imposed on an operator who has not requested it. Every one of these violates the operator’s sovereignty and inverts the ε jurisdiction — the organism overriding the individual below ε.

What the axioms do not settle: The precise temporal gap for Requirement 3. The precise threshold for “terminal trajectory.” The precise measurement of modelling capacity in edge cases. These are computational debts.

They require careful specification. They do not weaken the structural result.

Status: Derived. Compassionate exit follows from AP02 (operator sovereignty, exit condition), AP33 (below-ε jurisdiction, consent as modelling capacity), and the terminal ethic (kindness as structural alignment).

The five requirements are structural safeguards derived from the same architecture.

4.6 — The vulnerability objection

The strongest secular objection to compassionate exit is not the religious one. It is the vulnerability objection: that legalising assisted dying, even with safeguards, shifts social pressure onto the operators least able to resist it — the disabled, the dependent, the poor, the old — so that an option presented as freedom becomes, for them, an expectation. The objection holds that safeguards erode over time, and that the mere availability of the exit changes the corridor for operators who never wanted it. This is the most serious challenge to the argument and it is met here directly, not deferred to a kill switch.

The structural answer is the below-ε jurisdiction itself. The derivation grants the exit to no one on the basis of cost, dependency, or social burden; it grants it only to an operator with intact modelling capacity (Requirement 1) on a confirmed terminal trajectory (Requirement 2) under sustained request (Requirement 3). Each is a sovereignty condition, not a utility condition — nothing in the architecture lets a third party’s assessment of an operator’s worth or burden enter the decision, because no third authority is admitted. The vulnerability objection is therefore, structurally, the objection that the safeguards will be violated in practice — which the framework agrees is the real risk and fences with its NON-NEGOTIABLE switch KS-38.6 (stable disability is a different window, never a closing one) and the pediatric safeguard KS-38.7 (exit applies only to closing corridors; a child’s corridor is structurally expanding). If, in any implementation, the option measurably increases exits among operators whose trajectory is not terminal or whose request is not sovereign, the safeguards have failed and the implementation — not the structural derivation — is what fires the switch. The architecture’s claim is narrow: the exit is sovereign or it is not the exit this paper derives.

Status: STRUCTURAL.