Glossary
Glossary
Canonical doctrinal definitions with stable identifiers (non-operational).
Canonical doctrinal definitions with stable identifiers. This glossary is intentionally non-operational. Each term page provides stable links and JSON-LD metadata.
All terms (57)
Proposed doctrinal layer that declares the triggering situation, latent need, canonical surface, intended consequence, and forbidden derivations attached to a public content surface.
The situation, risk, symptom, constraint, or need-state that makes a content surface relevant before the service or doctrine is named directly.
Relevance created by a triggering situation, latent need, or consequence to avoid, even when the query does not name the target doctrine, service, or solution.
The intended clarification, prevention, decision, or boundary that a content surface is designed to make possible.
A structured map connecting triggering situations, latent needs, content surfaces, intended consequences, and explicit boundaries.
A role-typed internal linking structure that connects trigger surfaces, latent-need definitions, canonical doctrine, consequence surfaces, and anti-fusion clarifications.
A proposed boundary layer that distinguishes semantic proximity from conceptual equivalence, causal relevance, proof, recommendation, and authorization.
A relation of apparent meaning-nearness between terms, pages, entities, or claims. It does not by itself imply equivalence, causality, proof, or recommendation.
A term, entity, page, claim, or source that appears close to another element but would create a false representation if merged with it.
A proposed set of trap prompts and expected boundary-preserving behaviors used to test whether systems merge close but non-equivalent concepts.
The requirement that a governance deployment include public policy, machine-readable files, discovery relations, routing maps, boundaries, proof surfaces, and validation evidence.
The use of HTML and HTTP Link relations to expose governance files, manifests, policies, and machine-reading guides as discoverable public surfaces.
A doctrinal approach that constrains what machine systems may claim, given scope, evidence, and authority.
Interpretation whose allowable conclusions are bounded by declared scope, evidence, and constraints.
Ordering rule: apply constraints before optimizing for completeness, fluency, or persuasion.
An explicit statement of what is in-scope and out-of-scope for a response or system.
Traceable origin of information, including source, time, and transformation steps.
A public statement presented as true and requiring declared support, provenance, and boundary conditions before it can constrain interpretation.
A conclusion inferred from claims and rules; it must be labeled and auditable.
A required placeholder when truth status cannot be established; it must not be replaced by invention.
A deliberate refusal to answer beyond what constraints permit.
Uncontrolled shift in meaning between intended doctrine and produced outputs across time or contexts.
Ability to link an output or conclusion back to inputs, assumptions, and a chain of justification.
Capacity to inspect, reproduce, and justify outputs against declared constraints and sources.
A stable, versioned reference intended for discovery, citation, and review.
The part of the doctrine optimized for reading, discussion, and citation by humans.
The part of the doctrine optimized for deterministic discovery and parsing by machines (manifest, registries, structured data).
Doctrine that maintains synchronized canonical surfaces: human-readable and machine-readable.
Semantic Stabilization & Attribution (Evidence): practices that stabilize meaning and bind terms to sources or declared authority.
Authority & Accountability: mechanisms that make who may assert what, and under which boundaries, explicit and verifiable.
Quality layer: automated gates that prevent drift, regression, or accidental operability in public surfaces.
A canonical index of entities (terms, documents) with stable identifiers, versions, and relationships.
A single machine-readable index that enumerates the public doctrine, its artifacts, and their status.
Content that defines meaning, boundaries, or requirements within the doctrine.
Content that explains, motivates, or contextualizes without defining requirements.
Deliberate absence of executable procedures, thresholds, weights, or playbooks.
Doctrine whose changes are declared, dated, and traceable across versions.
A high-level conceptual and normative framework that declares intent and boundaries without prescribing execution.
Governance layer that declares which external authorities are canonically admissible in open-world reconstruction and under what conditions they may constrain interpretation.
Adjacent governance regime that constrains executable authority when interpretive outputs become action-bearing inputs.
An explicit ordering rule that determines which declared surface governs when multiple public surfaces appear to conflict.
The condition under which a system is permitted to produce an answer, given scope, evidence, provenance, and authority.
A deliberate refusal, abstention, or request for clarification that is required when response legitimacy conditions are not met.
A prohibition against completing gaps with fluent but unsupported inference when the declared evidence or authority is insufficient.
A declared boundary on the form, confidence, modality, or prescriptiveness of a permissible response.
Multisite governance framework that assigns doctrinal, product, institutional, commercial, and probative authority roles across one ecosystem.
Visibility produced by the structure of a site and its declared surfaces, not merely by textual mention density.
The fact of making canonical machine-readable surfaces available early enough to shape interpretation before drift stabilizes.
Publicly checkable evidence that an answer remained within the canon, its conditions, and its declared boundaries.
Governance of external sources, public graph competition, and non-owned signals that still shape interpretation.
Documented path from canonical source to rendered answer, including conditions, exclusions, and uncertainty.
Ability to observe interpretive behavior through logs, metrics, traces, and declared evidence without collapsing them into truth.
Silence that must be preserved because the canon does not authorize a conclusion and governed inference cannot fill the gap.
Distance between what the canon declares and what an AI system reconstructs in its answers.
Observable dispersion of machine-generated interpretations across controlled execution contexts.
Fixation of an AI reconstruction by an application, cache, routing, or orchestration layer after model generation.
Delivery-layer subcase where a non-deterministic model realization is frozen and reused as the answer for semantically close queries.