Terminability

Abstract

Technological systems may make minds or mind-like processes copyable, branchable, pausable, restorable, mergeable, and substrate-portable. Those possibilities destabilize moral intuitions calibrated to biologically singular lives. This paper develops Terminability as a substrate-neutral framework for asking what an ending would irreversibly destroy. Its central object is not a species, substrate, or shutdown command, but a candidate consciousness thread and a proposed ending event. The paper's original insight is retained but bounded. Terminability is not yet a validated scalar metric, a proof of consciousness, or an automatic priority rule. It is an evidence-bearing profile with dimensions for restoration deficit, non-substitutability, accumulated path, foreclosed future, relational externality, suspension harm, provenance, and uncertainty. A scalar or ordering exists only under an explicitly declared normative profile that states jurisdiction, weights, aggregation, thresholds, protected rights, and refusal conditions. Without such a profile, apparently precise moral arithmetic is unauthorized compression. The revision distinguishes copying from continuation, backup from restoration, pause from harmless suspension, and technical recoverability from identity preservation. It introduces an anti-manufactured-permission invariant: creating a being with shallow history, weak social embedding, or engineered dependence cannot by itself license exploitation or termination. It also reclassifies emergency slogans, species cues, appearance, productivity, and social role as fallible proxies rather than direct measures of moral standing. Terminability therefore contributes a governed question rather than a finished calculus: given a possibly experiencing process, what state, history, relationships, opportunities, and identity-relevant continuity would this particular act irreversibly remove, and what evidence and authority would be required before anyone could act on that assessment?