Domain-specific capacity rejects the global incapacity judgment — the finding that a person is simply “incompetent,” from which authority over their whole life then follows. Capacity is not one switch. Someone may be unable to manage complex medication and perfectly able to choose their relationships; unable to handle finances and entirely able to refuse a program they object to. So the only legitimate capacity question is narrow: can this person make this particular decision, at this time, under these conditions?
Judged that way, a real incapacity constrains only its own domain and leaves personal jurisdiction intact everywhere else. Where jurisdiction must narrow, it narrows by domain and on a clock — the bounded-override discipline of supported decision-making, periodic review, and a presumption of restored authority — rather than converting a single deficit into custody over the person. It is the capacity-side guard against custodial capture.
The same narrow question governs what may be asked of a participant, not only what they may decide. Reciprocal expectations track relevant capacity — can they reliably attend appointments, manage medication, meet shared-space responsibilities — rather than a moral judgment of whether they are trying hard enough (the commensurate logic, the reciprocity-side form of diagnostic differentiation). But such a reading is legitimate only as a decision-scoped assessment, never a standing classification: it is re-taken and expires rather than hardening into a persistent band or profile, and it does not travel to another domain without new justification — a reading of one capacity is not a portable score of the person. That is the discipline that keeps measurement an instrument, not an identity.