<p>Judgement Architecture is a formal operating discipline for making human Judgement explicit, assigned, and enforceable at the points in AI-enabled workflows where decisions become consequential. This glossary establishes its canonical vocabulary.</p> <p>The terms defined here are the working language of Judgement Architecture in practice. They cover the conditions that require human Judgement, the mechanisms for enforcing it, the failure modes that arise when it is absent, and the organisational-level condition, Responsibility Erosion, that results when automation outpaces accountability.</p> <p>Every entry includes a Test: a single diagnostic question a practitioner can apply in real time to determine whether the concept is in play.</p> <p>This glossary is a companion to the Judgement Architecture Standard, which defines the requirements for implementing Judgement Architecture in AI-enabled work. The two documents are designed to be used together.</p> <p>The vocabulary introduced here is offered as a contribution to a field that urgently needs shared language, one precise enough to govern AI systems and plain enough to use in the room where decisions actually happen.</p>