Answerability and Institutional Commitment: Why Missing Bearership Is Not Defective Ownership

Institutional systems increasingly present claims, refusals, recommendations, explanations, and decisions through automated interfaces. The central normative question is not exhausted by asking whether the output is intelligent, useful, accurate, or legally attributable. It is whether the presented commitment has a bearer who remains answerable for it. This paper isolates and defends a four-part criterion of Answerability: attribution continuity, challenge-responsive reconsideration, repair competence, and standing-consequence. Each relation is defended by a necessity argument and a contrast case. The paper then develops a distinction that is essential to the criterion’s use: missing bearership is not the same as defective ownership. A commitment may be clearly owned even when its owner reasons badly, obstructs review, refuses correction, or adopts a machinemediated result without adequate warrant. Conversely, an apparent interlocutor may lack bearership even where an institution elsewhere remains legally or normatively responsible. The distinction prevents a perverse inference from institutional failure to institutional innocence. It also supplies a substrate-neutral test: human, corporate, collective, and artificial candidates are assessed by the same public relations of continuity, challenge, repair, and consequence. A broader overview of the framework and its application to institutional AI is given in Cott (2026); the present paper develops only the criterion and the missing/defective distinction.