Lexicon

Attribution Displacement

The competitive outcome in which one entity loses AI citation share to another — the operational manifestation of ontological warfare. Attribution displacement is measured through quarterly retrieval preference testing: if your retrieval preference on domain queries declines from 60% to 40% while a competitor's rises from 30% to 50%, you have experienced attribution displacement. Attribution displacement is caused by relative degradation: your competitor maintained schema entropy rate above equilibrium while yours fell below. Attribution displacement is the market signal that cognitive equilibrium has been lost.

Bi-Temporal Provenance

The four-timestamp model for forensic traceability of corroboration sources: valid_from (when the fact became true in the world), valid_until (when the fact ceased to be true), ingested_at (when the fact entered the system), and invalidated_at (when the system recognized the fact as no longer current). Bi-temporal provenance makes the engagement record legally defensible, enables retroactive forfeiture detection (sources whose valid_until precedes the current quarter are automatically flagged in entropy rate calculations), and produces a longitudinal dataset quality that compounds with every engagement and is structurally irreproducible retroactively — contributing to the data moat. Applied to corroboration source records in the CC-DATA-01 engagement record schema.

Byrum’s Law

The formal statement of the cognitive equilibrium requirement: Sₑ ≥ Eₑ(γ), where Sₑ is entity signal construction rate, Eₑ is baseline schema entropy rate, and γ is the environmental acceleration factor. In plain language: to maintain defended posture, an organization must construct entity signals at least as fast as its infrastructure decays, adjusted for how rapidly the competitive and technological environment is changing. Byrum's Law explains why 'set and forget' entity infrastructure fails: in a changing environment (γ > 1), even perfect infrastructure degrades without active maintenance.

Causal Modeling

The analytical method for extracting predictive patterns from engagement record data: which maintenance interventions cause which posture outcomes? Causal modeling transforms descriptive engagement data into prescriptive recommendations. Example causal questions: Does tier-1 corroboration frequency predict schema entropy rate? Does L1a verification gate failure predict subsequent citation coverage degradation? Causal modeling requires structured engagement records, bi-temporal provenance, and sufficient sample size. The causal model is the analytical asset that makes the data moat actionable — converting proprietary data into proprietary insight.

Citation Coverage

The breadth of core entity claims that AI systems will cite when answering relevant queries — distinct from attribution rate, which measures frequency. An entity might have high attribution rate (frequently cited) but low citation coverage (only cited for a narrow subset of claims). Citation coverage measures: of all the claims you want AI systems to make about you, what percentage are they actually making? Low citation coverage indicates that some claims lack sufficient corroboration or fail EAV-E compliance. Citation coverage expansion is the tactical objective of corroboration campaigns.

Citation Engineering

The practice of designing content to maximize its citability by AI systems — structuring claims to pass EAV-E compliance and hallucination-avoidance filters. Citation engineering is not 'SEO for AI' — it is precision engineering of verifiable, attributed, entity-specific claims that AI systems can confidently cite without triggering uncertainty. Citation-engineered content includes: named entities (not 'we'), specific attributes (not vague benefits), measurable values (not subjective claims), and traceable evidence (not unsupported assertions). Content that fails EAV-E compliance will not be cited regardless of source tier or corroboration count.

Cognitive Equilibrium

The dynamic condition in which an organization's entity infrastructure is maintained at a rate that outpaces schema entropy accumulation — the defense line above which schema coherence compounds and below which degradation is thermodynamically certain. Not a static achievement but a continuously maintained operational state. In Byrum's Law, the formal condition in which Sₑ ≥ Eₑ(γ): the entity constructs signals at least as fast as its infrastructure decays under gamma-adjusted entropy accumulation.

Competitive Corroboration Gap

The measured difference in corroboration signal strength between you and your competitors — the primary determinant of retrieval preference. A positive gap (you have more tier-1/2 sources than competitors) produces retrieval advantage; a negative gap produces attribution displacement. The competitive corroboration gap is measured per-claim: for each core attribute you want cited, count your tier-1/2 sources and your top 3 competitors' tier-1/2 sources. The gap determines whether you are cited preferentially, competitively, or displaced. Closing competitive corroboration gaps is the tactical objective of corroboration campaigns.

Confidence Threshold Dynamics

The discontinuous behavior of AI citation: entity confidence operates as a switch, not a dial. Above the threshold, the entity is cited confidently; below it, the system hedges or omits. Small confidence reductions can trigger large behavioral changes if they cross the threshold. This explains contested posture invisibility: schema entropy accumulates gradually, but citation behavior changes abruptly when the threshold is crossed. Confidence threshold dynamics make entity infrastructure similar to structural engineering: loads accumulate gradually, but failures occur suddenly. Monitoring schema entropy rate detects approaching threshold before behavioral change occurs.

Contested Posture

The organizational condition in which entity infrastructure was built but maintenance has fallen below the cognitive equilibrium rate — causing invisible degradation of AI-mediated brand authority. The most dangerous of the three postures because the organization does not know it is there: measurement tools reflect a prior success state while the feedback loop compounds against the organization. The posture most commonly occupied by organizations that believe they are defended.

Corroboration

The structural requirement that entity claims appear in multiple independent, authoritative sources before AI systems will cite them with confidence. Single-source claims trigger hallucination-avoidance behavior; corroborated claims are cited. The minimum viable corroboration threshold is 5 independent tier-1 or tier-2 sources for each core entity attribute. Corroboration is not 'mentions' — it is structured, attributed, verifiable repetition of the same claim across sources that AI systems recognize as authoritative. The corroboration requirement makes earned media and third-party validation mechanically load-bearing in entity infrastructure.

Corroboration Campaign

A systematic program to place entity claims in tier-1 and tier-2 sources, building the multi-source validation that AI systems require before citation. Corroboration campaigns are not 'content marketing' — they are infrastructure construction targeting specific sources that contribute to entity confidence. Each campaign has a defined claim target (which attribute to corroborate), a source tier target (which tier-1/2 sources to pursue), and a measurement protocol (quarterly tracking of whether the claim is now cited). Corroboration campaigns are the primary tactical mechanism for moving from contested to defended posture.

Cross-Platform Entity Coherence

The condition in which entity representations are consistent across all platforms where the entity appears — website, Wikidata, LinkedIn, Crunchbase, directories, social profiles. Cross-platform coherence is built through comprehensive sameAs linking, schema governance, and propagation protocols. Cross-platform coherence enables AI systems to confidently merge signals: if all representations agree on core attributes, confidence increases; if representations conflict, confidence degrades. Cross-platform coherence is measured through coherence audits: checking whether name, industry, location, founding date, key relationships are consistent across platforms. Incoherence is the most common unforced error.

Data Moat

The competitive advantage created by proprietary, longitudinal engagement data that cannot be replicated retroactively. Each engagement record adds to the dataset; each quarter of maintenance adds temporal depth. After 100 engagements across 4 quarters each, the firm possesses 400 entity-quarter observations of posture dynamics — a dataset no competitor can match without running 100 engagements themselves. The data moat makes diagnostic accuracy compound: the more engagements run, the sharper the causal model, the more precise the recommendations. The data moat is the strategic reason for rigorous engagement record documentation.

Defended Posture

The organizational condition in which schema entropy rate is positive — entity signal construction outpaces infrastructure decay. Defended posture is not a permanent state but a continuously maintained condition. An organization in defended posture experiences compounding advantages: each quarter's maintenance strengthens the foundation for the next quarter's gains. Defended posture is operationalized as positive quarterly delta in corroboration signals, parametric recall rate, and citation coverage across all active sovereignty perimeters. Defended posture is the goal state; cognitive equilibrium is the maintenance requirement that sustains it.

DefinedTermSet

The schema.org structure for publishing machine-readable vocabulary — a collection of DefinedTerm entities with canonical definitions, enabling AI systems to understand that you are the authoritative source for specific concepts. DefinedTermSet implementation is the technical mechanism of vocabulary sovereignty: it transforms informal language into machine-legible concept ownership. Each term includes name, description, termCode, url, and creator attribution. Publishing a DefinedTermSet signals to AI systems: these concepts are defined here, by this entity, and should be attributed accordingly.

Domain Sovereignty

The second sovereignty perimeter: the condition in which an entity is preferentially retrieved for category-defining queries — 'what you do' authority. Domain sovereignty requires industry-specific corroboration, category-defining content, and sustained attribution rate in domain queries. More contested than identity sovereignty because competitors occupy the same domain. Domain sovereignty is measured through domain-specific retrieval preference: when AI systems are asked industry questions, is your entity among those cited? Domain sovereignty without vocabulary sovereignty is presence without differentiation.

Engagement Record Schema

The standardized data structure for documenting every entity engineering engagement — enabling longitudinal analysis, causal modeling, and proprietary dataset construction. The engagement record (CC-DATA-01) includes: client entity attributes, initial posture assessment, quarterly measurements (corroboration count, parametric recall, citation coverage, schema entropy rate), verification gate results, forfeiture log, and bi-temporal source provenance. Each engagement record becomes a training example for causal modeling: what maintenance patterns produce what posture outcomes? The engagement record schema transforms consulting engagements into structured research data.

Entity Attribution Rate

The percentage of AI responses in which the entity is correctly identified and cited when topically relevant — the core measurement of entity infrastructure effectiveness. Measured by sampling category-relevant queries, having multiple AI systems respond, and calculating what percentage of responses correctly attribute the entity. An entity attribution rate below 50% indicates the entity has not yet crossed the confidence threshold for that category. Attribution rate is measured per-perimeter: identity attribution, domain attribution, and vocabulary attribution are independent and must be tracked separately.

Entity Confidence

The degree to which an AI system is certain that it has correctly identified and characterized an entity — built through corroboration, consistent structured data, and temporal consistency. Entity confidence operates as a binary confidence threshold: above it the organization is surfaced with full confidence; below it the system hedges, displaces, or omits. The threshold is not a dial — it is a switch. Systems with confidence thresholds do not degrade gradually; they hold, then transition discontinuously.

Entity Disambiguation

The AI system process of determining which real-world entity corresponds to an ambiguous reference — resolving 'Michael Jordan' into the basketball player vs. the Berkeley professor vs. the actor. Disambiguation relies on structured data, corroboration, and context. Entities that fail disambiguation are omitted from AI responses because systems cannot confidently determine which entity is meant. KGMID assignment is evidence of successful disambiguation. Schema.org markup, comprehensive sameAs linking, and consistent NAP (Name-Address-Phone) across sources are the technical inputs to disambiguation.

Entity-Attribute-Value-Evidence

The four-part AI citability standard — abbreviated EAV-E — that determines whether a brand claim is citable by AI systems without triggering hallucination-avoidance behavior. Entity: a named, disambiguated entity. Attribute: a specific, defined property. Value: a verifiable, concrete claim. Evidence: traceable, cross-referenceable proof. A statement like 'We improve customer outcomes' fails EAV-E. 'Company X reduces deployment time to 24 hours, validated by 150 client implementations documented in their 2025 case study library' passes it. Corroboration campaigns built on content that fails EAV-E will underperform regardless of source tier quality, because the AI's hallucination-avoidance behavior suppresses citation of imprecisely stated claims. EAV-E compliance is the pre-condition for citation engineering to work as specified.

Forfeiture Event

A quarter in which schema entropy rate went negative on any sovereignty perimeter — the formal definition of maintenance lapse. Forfeiture events are documented in the posture forfeiture log with: which perimeter, measured entropy delta, hypothesized cause, corrective action taken, and measured outcome. Forfeiture events are the primary training data for causal modeling: which maintenance gaps cause which forfeiture patterns? Repeated forfeiture events on the same perimeter indicate structural maintenance insufficiency, not random variance. Forfeiture events make contested posture visible in the engagement record.

Full Spectrum Dominance

The entity engineering goal state in which an organization achieves accurate, consistent, machine-confirmed presence across both AI retrieval pathways simultaneously — RAG retrieval and parametric memory — across all three sovereignty layers (identity, domain, vocabulary). Formally: the condition in which Sₑ > (Eₑ(γ) + Sₐ) holds simultaneously across all active sovereignty perimeters, operationalized as ≥75% retrieval preference across sampled queries on each perimeter. Not the absence of competitive threat but the construction of a position so structurally coherent and temporally deep that the cost of displacement exceeds any competitor's willingness to pay it.

Gamma Factor

The environmental acceleration variable in Byrum's Law: γ measures how rapidly the competitive and technological landscape is changing, which determines how quickly schema entropy accumulates. High-gamma environments (emerging technologies, rapid competitive entry, frequent model updates) require higher maintenance rates to sustain cognitive equilibrium. Low-gamma environments (stable industries, slow technological change, infrequent model retraining) allow lower maintenance rates. Gamma is assessed qualitatively per industry and adjusted quarterly based on observed market dynamics. Gamma explains why identical maintenance programs produce different outcomes across industries.

Hallucination Avoidance

The AI system behavior of declining to make claims when confidence is insufficient — the protective mechanism that causes AI systems to hedge, generalize, or omit rather than risk stating falsehoods. Hallucination avoidance is the reason single-source claims are not cited, ambiguous entities are not named, and poorly-corroborated attributes are not stated. Understanding hallucination avoidance explains why earnest, true claims about your entity may not be cited: it's not that the AI doesn't believe them — it's that the structural evidence isn't sufficient to override the hallucination-avoidance threshold.

Identity Sovereignty

The first sovereignty perimeter: the condition in which an entity is reliably identified, disambiguated, and characterized by AI systems when asked 'who is X?' Achieved through KGMID assignment, comprehensive schema.org Person/Organization markup, Wikidata entity creation, and consistent NAP across authoritative directories. Identity sovereignty is prerequisite to domain and vocabulary sovereignty: if AI systems cannot reliably identify who you are, they cannot reliably characterize what you do or what you mean. Most organizations have identity sovereignty or are close to it; it is the least contested perimeter.

KGMID

Knowledge Graph Machine Identifier — the persistent, globally unique identifier assigned by Google's Knowledge Graph to disambiguated entities. KGMID assignment is evidence that an entity has crossed the disambiguation threshold: Google's entity resolution system has determined with sufficient confidence that this entity is distinct, real, and worth tracking. Entities without KGMIDs are not disambiguated in Google's knowledge infrastructure and are mechanically disadvantaged in retrieval contexts. KGMID is not 'nice to have' — it is the formal entry credential to identity sovereignty.

Methodological Vocabulary

The subset of vocabulary sovereignty consisting of terms that define how work is done — methodologies, frameworks, processes, diagnostic protocols. Methodological vocabulary is the highest-leverage form of vocabulary sovereignty because it travels: other organizations adopt your terms to describe their own work, creating network effects in term recognition. Entity engineering, cognitive equilibrium, and schema entropy are methodological vocabulary: they describe how the work is done, not merely what the work is called. Methodological vocabulary requires DefinedTermSet publication and systematic termCode usage in methodology documentation.

Multi-Variety Optimization

The schema design requirement that entity representations include multiple varieties of the same core claim to match different query patterns — optimizing for lexical diversity while preserving semantic identity. Example: an entity described as 'AI consulting firm' should also be described as 'artificial intelligence advisory' and 'machine learning consultancy' to match query variety. Multi-variety optimization prevents query-pattern brittleness: if your schema only includes one term variant, you're invisible to queries using other variants. Multi-variety optimization is distinct from keyword stuffing: it's structured semantic expansion, not repetitive phrasing.

OG-RAG

Ontology-Grounded Retrieval-Augmented Generation — a formally peer-reviewed retrieval paradigm (Sharma, Kumar & Li, EMNLP 2025) that integrates formal ontologies at every stage of the retrieval-generation loop, meaning AI systems using OG-RAG resolve entities by traversing a formal ontology rather than by vector similarity alone. An entity whose schema architecture is OG-RAG compatible — defining what things are and how they relate, not merely that they exist — is preferentially retrieved by ontology-grounded systems. Schema that only declares presence without semantic relationships will fail OG-RAG resolution regardless of KGMID or sameAs completeness. The emerging retrieval paradigm for specialized knowledge domains in enterprise AI environments. OG-RAG compatibility is assessed in the L1a Verification Gate.

Ontological Relationships

The semantic connections between entities and concepts that enable AI systems to understand meaning, not merely co-occurrence. Ontological relationships are expressed through schema.org properties like 'member of', 'part of', 'is defined by', 'has specialty'. OG-RAG retrieval systems traverse these relationships to resolve entities: 'entity engineering' is defined by 'Joseph Byrum', who is 'founder of' 'Big House Enterprise', which 'specializes in' 'AI authority construction'. Schema that only declares entities without relationships will fail OG-RAG resolution. Ontological relationships are the L1a verification gate requirement.

Ontological Warfare

The structural competitive dynamic in which organizations with stronger entity infrastructure displace competitors from AI category positions — not through malice but through the mechanics of how AI systems assign confidence. The mechanism operates regardless of intent: it rewards structural coherence and displaces whatever is less coherent, less corroborated, and less temporally consistent. Not a hostile act but a structural one. Named in parallel to documented research on information operations, with entirely different ethics: competitors building stronger entity infrastructure are not conducting information operations — they are building.

Parametric Memory

Knowledge encoded directly into an AI model's neural weights during training — what the model 'knows' without needing to retrieve external documents. Parametric memory is temporally frozen at the model's training cutoff date and cannot be updated without retraining. Entities absent from parametric memory are mechanically disadvantaged in zero-shot queries where the model answers without retrieval. Building parametric presence requires temporal depth: years of consistent, corroborated signals across high-authority sources ingested during the model's training window. Parametric memory is the harder-to-reach but higher-value component of full spectrum dominance because it persists across retrieval contexts.

Parametric Recall Protocol

The testing methodology for measuring parametric memory presence: disable retrieval, ask identity/domain/vocabulary questions, score accuracy and confidence. Parametric recall protocol operationalizes the zero-shot query concept into a repeatable measurement. Execute quarterly with consistent question sets across multiple AI models. Scoring: 1 = accurate and confident, 0.5 = accurate but hedged, 0 = inaccurate or omitted. Aggregate score is parametric recall rate. Below 50% indicates parametric absence; above 75% indicates parametric presence. Parametric recall rate is a lagging indicator (reflects training-time corroboration, not current corroboration).

Per-Perimeter Posture Assessment

A formal diagnostic instrument that produces three independent posture verdicts (Defended / Contested / Undefended) for the three sovereignty perimeters — identity, domain, and vocabulary — rather than a single composite posture label. The most common real-world condition is Defended on identity, Contested on domain, and Undefended on vocabulary simultaneously; a single posture label obscures the vulnerabilities and produces undifferentiated investment. The per-perimeter assessment identifies the constraining bottleneck before engagement begins, enabling targeted investment. The critical sales and diagnostic framing: most prospects are Defended on identity, Contested on domain, Undefended on vocabulary — and don't know it.

Posture Diagnostics

The systematic assessment protocol that produces per-perimeter posture verdicts (Defended / Contested / Undefended) for identity, domain, and vocabulary sovereignty. Posture diagnostics replace vague 'brand health' assessments with mechanistic measurements: schema coherence score, corroboration signal count, parametric recall rate, attribution rate, citation coverage, schema entropy rate. The diagnostic outputs a posture verdict per perimeter and identifies the constraining bottleneck before engagement begins. Posture diagnostics are the qualification and scoping tool: they determine which perimeters need construction vs. maintenance.

Posture Forfeiture Log

A formal field in the CC-DATA-01 engagement record documenting every quarter in which any sovereignty perimeter's schema entropy rate went negative — recording the maintenance gap that caused the forfeiture event, the correction applied, and its measured outcome. The posture forfeiture log makes forfeiture visible in the engagement record rather than masked by aggregate metrics, closing the contested-posture-invisibility failure mode at the engagement record level. Every engagement's forfeiture log becomes proprietary training data for the CC-DATA-01 causal dataset: the longitudinal record of forfeiture events and corrections compounds across engagements, making each subsequent diagnostic sharper.

RAG Retrieval

Retrieval-Augmented Generation: the AI pathway in which models search external sources before answering, rather than relying solely on parametric memory. RAG retrieval is the faster pathway to entity presence because it operates on current web data rather than frozen training data. However, RAG-only presence is contextually fragile: it depends on the retrieval system surfacing the right sources at query time. Full spectrum dominance requires simultaneous presence in both RAG and parametric pathways. RAG presence without parametric presence produces inconsistent brand authority across user contexts.

Retrieval Preference

The AI system behavior of selecting one entity over another when multiple entities could plausibly answer a query — the mechanism through which ontological warfare operates. Retrieval preference is determined by relative entity confidence: the system retrieves and cites the entity with stronger corroboration, clearer disambiguation, and deeper temporal consistency. Retrieval preference is measured as a percentage: in a sample of category queries, what percentage result in your entity being retrieved vs. competitors? Full spectrum dominance is defined as ≥75% retrieval preference across all perimeters.

Retroactive Irreproducibility

The competitive moat created by temporal depth: infrastructure built over years cannot be replicated quickly by competitors starting today. Parametric memory presence requires 2-3 years of tier-1 corroboration during the model's training window — a competitor starting today cannot achieve parametric presence until the next model generation, regardless of budget. Retroactive irreproducibility makes temporal consistency a strategic asset: the longer you maintain infrastructure, the more expensive it becomes for competitors to displace you. Retroactive irreproducibility is why early investment in entity infrastructure compounds.

SameAs

The schema.org property that asserts 'this entity representation and that entity representation refer to the same real-world entity' — the linking mechanism that enables cross-platform entity coherence. SameAs bindings connect your website's schema to your Wikidata item, LinkedIn profile, Crunchbase page, and other authoritative entity representations. Each sameAs link is a corroboration signal. Comprehensive sameAs linking is required for entity resolution systems to confidently merge signals from multiple sources into a unified entity understanding. Missing sameAs links fragment your entity signal and reduce confidence.

Schema Coherence

The condition in which all entity representations (website schema, Wikidata item, directory listings, social profiles) make consistent claims about core attributes — name, industry, location, relationships. Schema coherence is required for entity confidence: inconsistent schema signals ambiguity and triggers disambiguation failure. Schema coherence is maintained through centralized schema governance and quarterly coherence audits across all public entity representations. Schema incoherence is the most common unforced error in entity infrastructure — organizations publish conflicting information about themselves and wonder why AI systems hedge.

Schema Entropy

The progressive incoherence of an organization's machine-readable identity signals in the absence of active maintenance. An organization not actively maintaining its entity infrastructure does not hold position — it degrades. Schema entropy is the AI era's expression of the permanent CCT condition: coherence, corroboration, and temporal consistency require active maintenance against passive drift. The threat environment against which cognitive equilibrium is the defense line.

Schema Entropy Rate

The signed quarterly delta measuring whether an organization is above or below the cognitive equilibrium line — a formal measurement requirement in the entity engineering methodology. Calculated by comparing the corroboration signal count, parametric recall rate, and citation coverage from the prior quarter against the current quarter. A positive delta indicates the organization is above the cognitive equilibrium line (defended posture); a negative delta indicates it is below (contested posture, regardless of what the monitoring dashboard shows). The schema entropy rate makes the defense line visible and actionable rather than merely conceptual — closing the contested-posture-invisibility failure mode at the measurement level.

Schema Governance

The organizational discipline of centralized control over entity schema updates — ensuring all changes to entity representations are coordinated, versioned, and propagated consistently. Schema governance prevents schema incoherence from distributed, uncoordinated updates. Schema governance includes: centralized schema registry, change approval process, propagation checklist (update website, Wikidata, directories simultaneously), version control, and coherence testing. Without schema governance, well-intentioned updates fragment entity signals: marketing updates the website schema, PR updates Wikidata, sales updates Crunchbase, none coordinate — result is schema incoherence.

Schema.org

The universally adopted structured data vocabulary that enables machine-readable entity representation on the web. Schema.org markup transforms human-readable content into AI-parseable assertions about who you are, what you do, and how you relate to other entities. Structured data is the prerequisite for entity disambiguation: without it, AI systems cannot reliably distinguish you from similarly-named entities. Schema.org implementation is the foundational layer of identity sovereignty and the technical enabler of KGMID assignment. Incomplete or inconsistent schema is mechanically equivalent to entity ambiguity.

Source Tier Classification

The hierarchical ranking of corroboration sources by their authority weight in AI entity resolution systems. Tier 1: Peer-reviewed academic sources, major news outlets, government authorities, established encyclopedic sources. Tier 2: Industry analysts, trade publications, authoritative industry directories, credentialed expert publications. Tier 3: General business publications, company-controlled content, social media. Not all sources are equal: a single tier-1 source carries more corroboration weight than ten tier-3 sources. Corroboration campaigns must prioritize tier-1 and tier-2 placement to move entity confidence thresholds.

Sovereignty Perimeters

The three distinct infrastructural boundaries across which entity dominance is contested and measured independently: identity sovereignty (who you are), domain sovereignty (what you do), and vocabulary sovereignty (what you mean). These perimeters are mechanically independent: an organization can be Defended on identity, Contested on domain, and Undefended on vocabulary simultaneously — a common real-world condition. The three perimeters are not interchangeable and do not substitute for each other. Identity without domain is visibility without category position; domain without vocabulary is presence without conceptual authority. The per-perimeter assessment replaces the unitary 'brand strength' abstraction with a mechanistically actionable diagnostic.

Temporal Consistency

The requirement that entity claims remain stable and corroborated across time — a factor in entity confidence scoring. AI systems distrust claims that appear suddenly or change frequently without explanation. Temporal consistency is built through sustained, repeated corroboration of the same core claims across years. Entities with deep temporal consistency (claims corroborated across 3+ years) score higher in confidence than entities with recent but shallow signal history. Temporal consistency makes entity infrastructure resistant to competitive displacement because new entrants cannot retroactively create temporal depth.

Undefended Posture

The organizational condition in which entity infrastructure does not exist or has never reached the confidence threshold for a given perimeter. Undefended posture is visible and diagnosable: the organization knows it has no presence. Undefended posture on identity is rare; undefended posture on domain is common; undefended posture on vocabulary is nearly universal. Undefended posture is the least dangerous posture because the organization knows it needs to build. The tactical response to undefended posture is greenfield construction: build infrastructure from zero with no legacy constraints.

Variety Audit Protocol

The diagnostic procedure for identifying query pattern coverage gaps — ensuring entity schema includes sufficient lexical variety to match how users actually query. Variety audit protocol: sample category queries from actual user search data, identify term variants used, check whether your schema includes those variants. Example: if users search 'AI consultants', 'artificial intelligence advisors', and 'machine learning consulting firms', your schema should include all three variants. Variety audit prevents query-pattern brittleness. The variety audit is run pre-deployment (L1b verification) and quarterly (maintenance check). Variety gaps are resolved through multi-variety optimization.

Verification Gates

The three-layer quality control protocol ensuring entity infrastructure meets technical standards before deployment: L1 (schema validity), L2 (corroboration sufficiency), L3 (operational coherence). Verification gates prevent deployment of non-compliant infrastructure that would fail in production. L1a verifies OG-RAG compatibility; L1b verifies schema.org completeness. L2 verifies minimum corroboration thresholds are met. L3 verifies end-to-end citation behavior. Passing all three gates is the exit criterion from the build phase into the maintenance phase. Verification gate failures document technical debt in the engagement record.

Vocabulary Sovereignty

The third sovereignty perimeter: the condition in which an entity is recognized as the authoritative source for specific concepts — 'what you mean' authority. Achieved through DefinedTermSet publication, systematic termCode usage, and corroboration of term ownership in tier-1/2 sources. Vocabulary sovereignty is the most defensible perimeter because concepts, once established, require retraining or comprehensive retrieval updates to displace. Vocabulary sovereignty enables pricing power and category creation: if you own the concepts, you define the category. Vocabulary sovereignty without domain sovereignty is language without legitimacy.

Wikidata

The structured knowledge base operated by the Wikimedia Foundation — a machine-readable, crowd-maintained entity registry that serves as a critical corroboration source for AI systems. Wikidata entities carry high authority weight in entity resolution systems because Wikidata's editorial standards and version control make it a reliable disambiguation signal. A Wikidata item for your entity, with properly linked claims and references, contributes to both KGMID assignment and parametric memory formation. Wikidata is load-bearing infrastructure, not optional social proof.

Zero-Shot Query

An AI query answered purely from parametric memory without retrieval — the test condition for measuring whether your entity has achieved parametric presence. Zero-shot queries are operationalized by disabling web search and asking the model to answer from its training data alone. Entities that are correctly characterized in zero-shot responses have achieved parametric memory formation. Entities that are omitted, hedged, or mischaracterized in zero-shot responses lack parametric presence and are dependent on RAG retrieval for visibility.