The Participation
Lexicon
Creating a new category requires creating new language. These are the precise definitions of the terms, distinctions, and concepts that constitute the AI participation category.
This page installs the vocabulary. The methodology shows the mechanics →
The Sapir-Whorf hypothesis holds that the language available to describe a domain shapes how that domain can be thought about. Categories that lack precise vocabulary collapse into adjacent categories — losing their distinct character. The vocabulary on this page is not nomenclature for its own sake. It is the conceptual grammar of a new category. Fluency in it changes what you notice, what you measure, and what questions you ask.
Core Primitives
Four foundational concepts. Every other term in this lexicon derives from these. Misunderstanding any one of them produces systematic misinterpretation of everything downstream.
The condition of a network domain with respect to its structural accessibility, extractability, and attribution-readiness for autonomous AI systems. Participation is not a behaviour — it is a structural state that exists independently of whether it has been measured. A domain participates in the AI information layer when automated systems can reach, extract from, and cite it. Participation is derived; it is not assigned.
The term is deliberately chosen for its structural neutrality. Participation is not an achievement. It is a condition. It may be present without the domain operator's knowledge.
The computational process by which a participation state is determined from observable structural signals. Derivation is deterministic — the same domain, evaluated under the same protocol version, produces the same participation state every time. Derivation does not involve human judgement, editorial assessment, or weighted scoring. It is a pipeline with defined inputs and a defined output.
The distinction between derivation and scoring is not cosmetic. It is architectural. A score can be gamed by optimising individual signals in isolation. A derived state requires all preceding gates to pass — it cannot be manipulated by improving a single downstream signal.
The institutional frame within which aater.ai operates. An observatory observes pre-existing conditions — it does not create them, optimise them, or recommend interventions based on commercial relationships with the systems being observed. The observatory frame implies independence, reproducibility, and a commitment to describing what exists rather than prescribing what should exist.
The act of measuring a structural condition that exists prior to and independently of the measurement. In aater.ai's architecture, observation is distinguished from tracking (which implies following a human subject through time) and from monitoring (which implies active surveillance). Observation is the derivation of a state from structural signals at a moment in time.
The Three Gates
The gates are evaluated in strict sequence. Failure at any gate terminates evaluation — no gate can be compensated for by performance at another. This architecture is not a design choice. It reflects how AI systems actually process web content.
The first sequential evaluation layer in the PSS v2.1.0 derivation pipeline. The Reachability Gate assesses whether the domain's server-delivered content is accessible to non-JavaScript-executing autonomous agents — specifically: whether the server returns a valid HTTP response, whether server-delivered content meets a minimum completeness threshold, and whether the domain's machine-readable access policy (robots.txt) permits known AI agent classes. Failure at this gate terminates the derivation. Downstream gates are not evaluated.
The second sequential evaluation layer, activated only on passing of the Reachability Gate. The Legibility Gate assesses whether server-delivered content is structurally suitable for automated information extraction — specifically: heading hierarchy enabling content navigation and segmentation; named entity density indicating the presence of specific, attributable claims; and the proportion of content accessible without JavaScript execution (the render gap estimate).
The third sequential evaluation layer, activated only on passing of both preceding gates. The Authority Gate assesses whether the domain's content carries the structural signals that AI systems use to evaluate source credibility and enable attribution — specifically: authorship metadata (byline or schema.org author declarations), publication and modification timestamps, schema.org structured data markup, and topical coherence. The Authority Gate produces a graduated outcome: weak, moderate, or strong authority band.
The gate sequence above is the conceptual definition. The methodology shows exactly what signals are measured at each gate and how outcomes are determined.
See the full pipeline →The Five Participation States
Every public domain on the web occupies exactly one of these states at any given moment. States are not rankings. They are structural conditions derived from gate evaluation outcomes.
The participation state in which a domain's content cannot be accessed by autonomous AI systems under current structural conditions. The domain may be technically available to human browsers — it may load correctly, contain valuable content, and rank in search results. None of this is observable to AI systems if reachability conditions fail. Absence is not a penalty. It is a structural condition.
Absence is the most common state. Approximately 35% of surveyed domains resolve to Absent — most due to content requiring client-side JavaScript execution, robots.txt misconfiguration, or server response failures specific to non-browser clients.
The participation state in which a domain is structurally accessible to AI systems but content extraction yields insufficient usable information. The domain may be large, well-written, and frequently visited by human users. If the content is unstructured — lacking heading hierarchy, specific named entities, or verifiable claims — AI systems pass through without extracting usable material.
Approximately 28% of surveyed domains resolve to Marginal. The most common cause is content structure optimised for human reading experience rather than machine extraction — design choices that are invisible to human visitors but decisive for AI systems.
The participation state in which a domain's content is both accessible and extractable by AI systems, but structural authority signals are insufficient to support reliable attribution. AI systems may extract and synthesise content from Capturable domains, but without authorship, publication timing, or structured data signals, the content is unlikely to be cited as a named source.
The distinction between Capturable and Absent is not visible to the site operator in web analytics — AI systems leave no traceable signal when they pass through without extracting or citing. The distinction is only observable through structural classification or live telemetry.
The participation state in which all three gates pass and authority signals are present but not fully consistent. Authorship may be present on some pages but not others. Schema markup may be partial. Publication dates may be present but not structured. In the Emerging state, attribution is structurally possible but not reliably triggered — AI systems may cite the domain in some contexts but not others, depending on the completeness of the specific page's authority signals.
The participation state in which all three gates pass at full strength. The domain's content is accessible, extractable, and structurally positioned for attribution and citation by AI systems. All authority signals — authorship, schema markup, publication timestamps, topical coherence — are present and consistent. Authoritative is a structural condition, not a guarantee. External AI systems make independent decisions. Structural readiness is necessary but not sufficient for actual citation.
Approximately 6% of surveyed domains resolve to Authoritative. The scarcity reflects not the difficulty of the technical requirements but the absence of any prior standard defining them.
Observatory Instruments
The systems, protocols, and constructs that constitute the observatory infrastructure.
The current version of the derivation protocol used by aater.ai to classify network domains. PSS v2.1.0 defines the three-gate sequential evaluation architecture, the five-state participation taxonomy, and the derivation confidence meta-signal. Every classification produced by aater.ai is tagged with the PSS version that produced it, enabling historical comparability as the standard evolves.
See the full classification pipeline →The live telemetry instrument of the aater.ai observatory. Pulse records autonomous AI agent access events for monitored domains through a server-side detection mechanism that does not require client-side JavaScript execution. Pulse data provides the behavioural counterpart to the structural classification produced by the PSS derivation pipeline — enabling detection of agreement and contradiction between structural state and observed agent behaviour.
A second-order output of the PSS derivation pipeline that quantifies the epistemic reliability of the primary participation state. Derivation confidence is computed from four observable inputs: render accessibility state, render gap estimate, reachability outcome, and robots policy status. It is expressed as a discrete level — High, Moderate, or Low — and returned alongside the participation state as an independently interpretable signal. Absent structural signals are treated as neutral, not as negative indicators.
The public-read registry of derived participation states maintained by the aater.ai observatory. The Ledger contains the most recently derived participation state, the determining gate, the derivation timestamp, and the PSS version for all processed domains. The Ledger enables third-party verification of any domain's derived state and provides longitudinal state history enabling tracking of structural improvement over time.
The condition in which a domain's structural participation state has been derived but its live agent behaviour remains unobserved. Structural derivation reveals what conditions exist; Pulse telemetry reveals what agents are actually doing. The observation gap is the space between these two signal classes — closed only when both structural and behavioural data are available and can be compared.
An estimate of the proportion of a domain's content that is inaccessible to non-JavaScript-executing autonomous agents. Derived from detection of JavaScript framework construction patterns in server-delivered HTML and comparison with expected content depth. A high render gap indicates that significant content is invisible to AI systems regardless of reachability or authority conditions.
What We Don't Say
Cognitive priming works in both directions. Knowing what terms to avoid is as important as knowing which to use. Each substitution below reflects a deliberate architectural distinction — not a branding preference.
Reference: George Lakoff, Don't Think of an Elephant — framing shapes the terms of cognition before any argument is made.
Observatory Doctrine
Four statements that appear throughout the observatory. Each encodes a structural principle about how the system works — not a marketing claim.
“Participation conditions resolve whether they are observed or not.”
The structural conditions governing AI participation exist independently of measurement. Running a classification does not change the state — it reveals a state that already existed.
“Classification is not assigned — it is derived.”
No human judgement enters the derivation. The same domain, the same signals, the same gate sequence produces the same state deterministically.
“Structural readiness is necessary but not sufficient for AI citation or use.”
An Authoritative state does not guarantee that AI systems will cite the domain. It guarantees that no structural condition prevents them from doing so.
“Observed → Resolved.”
The observatory's operational ontology. A domain is observed (its structural signals are acquired) and a participation state is resolved from those signals. The process is directional and complete.
Now derive your participation state.
Classification is already occurring. The structural conditions that determine your domain's participation state exist whether or not you have observed them.