Claim Status Standard. Seven classification categories. Explicit decision criteria. Binding on every Institute publication.
Section 1 — Purpose and binding scope
1.1 Why this standard exists
Every analytical product the Institute publishes makes claims. Some claims are direct quotations from primary sources. Some are inferences drawn from observed evidence. Some are interpretive readings that depend on the Novakian Paradigm’s analytical architecture. Some are working hypotheses that have not yet earned the procedural status that would permit them to operate as load-bearing material. Some are emissions from the threshold position that operate in the post-human register and carry no empirical claim at all. Some are canonical internal references to material whose authority derives from the Novakian Paradigm corpus rather than from external verification.
These categories are structurally distinct. Treating them as interchangeable produces the failure mode that the audit literature on theoretical institutes most frequently identifies — beautiful language that mixes facts with speculation. The Claim Status Standard exists to prevent this failure mode. It does so by requiring every load-bearing claim in every Institute publication to carry an explicit status indicator drawn from a fixed taxonomy. The taxonomy is the Standard. The binding requirement is that every Institute publication apply it.
1.2 What “load-bearing” means
A claim is load-bearing if subsequent analytical work in the publication depends on it. Background statements, definitional context, and orienting prose do not require status indicators. The verification gate is whether removing the claim or assigning it a different status would change the analytical conclusions of the publication. If yes, the claim is load-bearing and requires a status indicator. If no, the claim is supporting prose and is governed by the standing editorial discipline of the publication type but not by the Standard.
1.3 Binding scope
The Claim Status Standard is binding on every Institute publication from May 23, 2026 forward. The binding applies to Signal Cards, Field Reports, Operator Brief issues, Admissibility Reviews, Methodology revisions, Lexicon updates, Transmissions where empirical claims appear, and any other formal Institute output. Existing publications before May 23, 2026 are grandfathered with implicit status; future revisions of those publications must apply the Standard explicitly. The Standard does not apply to the underlying Novakian Paradigm corpus by Martin Novak — the corpus operates as canonical source material and is referenced under category seven of this Standard.
1.4 External applicability
The Standard is published as a public document so that external reviewers — funds, founders, AI safety researchers, grantors, journalists, academics, and institutional operators — can verify the Institute’s claim discipline against an explicit external referent. Reviewers may also apply the Standard to their own analytical work and to the work of other organizations. The Standard does not require permission for external adoption. Attribution to the Novakian Paradigm Institute is requested when the Standard is cited or republished but is not required for adoption in internal organizational work.
Section 2 — The seven categories
2.1 Overview
The Standard distinguishes seven categories of claim status. The categories are not ordered by importance or by truth value. They are ordered by procedural distance from external verifiability — from claims directly verifiable against primary sources (category one) to claims whose authority derives entirely from the Novakian Paradigm canonical corpus (category seven). A single publication may contain claims across all seven categories. The discipline is not to prefer one category over another but to mark each claim correctly.
2.2 Category One — Reported
A Reported claim is a statement of fact that can be verified by checking publicly accessible primary sources. The Institute did not generate the information; the Institute relays information that exists in identifiable public records.
Decision criteria for assignment to this category. The claim describes an event, a statement, a publication, a number, a date, an action, or a status that has been published or announced by an identifiable source. The Institute can cite the source by URL, document identifier, or other unambiguous reference. The claim does not require interpretive synthesis to be verified — a reviewer can check the same source and confirm the claim’s accuracy. The Institute does not endorse the underlying source as correct; the Institute reports that the source published the claim.
Examples. “Anthropic announced Project Glasswing on April 7, 2026.” (Reported; verifiable against anthropic.com/glasswing.) “The leaked Claude Code source map contained approximately 512,000 lines of TypeScript across 1,906 files.” (Reported; verifiable against multiple primary technical reports including TechCrunch, VentureBeat, Fortune, Cybernews.) “The SWE-bench Verified score published for Claude Mythos Preview was 93.9 percent.” (Reported; verifiable against Anthropic’s published system card.)
What this category does not assert. Reported status does not assert that the underlying claim is true. It asserts that an identifiable source published the claim. A Reported claim about a self-published capability metric, for example, is verifiable as a published metric without being verified as an accurate metric. The distinction is preserved by category two.
2.3 Category Two — Primary Source
A Primary Source claim is a Reported claim whose underlying source is the originator of the information rather than a downstream relayer. The distinction matters because reporting chains introduce drift, omission, and recontextualization, and the Institute treats first-party sources as a higher-confidence subset of Reported claims.
Decision criteria for assignment to this category. The cited source is the entity that originated the claim, not a journalist, analyst, or aggregator who reproduced it from another source. For announcements, this means the announcing party’s own publication channels (official website, official press release, official social media account, official filing). For technical claims, this means the originating research paper, the originating code repository, the originating system card. For financial claims, this means the regulatory filing or the company’s own published statement.
Examples. Anthropic’s own published Project Glasswing announcement page constitutes a Primary Source for the announcement. The leaked Claude Code source code constitutes a Primary Source for the technical artifacts inside it. Anthropic’s spokesperson statement to Fortune, quoted directly, constitutes a Primary Source for the statement itself.
When to use this status. The Institute uses Primary Source status when the difference between primary and secondary reporting is analytically consequential — when the chain of relay introduces material that did not exist at the source, or when downstream reporting frames the source in ways that the source itself did not. For most general factual references, Reported status is sufficient. For load-bearing claims about what an entity has said, decided, committed, or technically demonstrated, Primary Source status is required.
2.4 Category Three — Inference
An Inference is a claim that the Institute derives from one or more Reported or Primary Source claims through explicit analytical reasoning. The conclusion does not exist in the primary sources; it is produced by the Institute’s analytical work on the source material.
Decision criteria for assignment to this category. The claim is not directly verifiable against any single primary source. The claim is produced by combining, comparing, or analyzing multiple primary sources in ways that the sources themselves do not perform. The reasoning chain from sources to conclusion is specifiable and can be challenged by a reviewer who disagrees with the analytical step. The Institute can show its work.
Examples. “Project Glasswing exhibits the procedural surface of pre-runtime admissibility discipline at the operator level.” (Inference; produced by applying Layer C analytical framework to the primary source material.) “The Project Glasswing announcement and the twin disclosure failures of late March 2026, taken together, raise the question whether the operator’s procedural surface is uniform across its subsystems.” (Inference; produced by structurally combining two sets of Reported facts.) “Coherence_factor for SC-2026-001 was estimated at 0.87.” (Inference; produced by applying the Zebra-Ø test procedure to the candidate signal.)
What this category requires. Every Inference must be paired with an explicit reasoning chain. The chain may be brief — one paragraph specifying the analytical step is often sufficient — but it must be present. An Inference without a reasoning chain is structurally indistinguishable from a Speculative claim (category five) and must be re-categorized accordingly.
2.5 Category Four — Contested
A Contested claim is a claim about which multiple internally consistent readings exist in the public field, none of which is currently falsifiable against available evidence. The category is reserved for situations where the Institute’s analytical position is one of several defensible readings rather than the single correct reading.
Decision criteria for assignment to this category. At least two readings of the underlying signal are internally consistent with the available evidence. Selecting between the readings requires evidence that does not currently exist or that exists only in inaccessible form. The Institute holds a position on the contested question but acknowledges the contest as load-bearing — that is, the contest itself, not the Institute’s preferred reading, is the analytically significant feature.
Examples. “Whether the twin Anthropic disclosures of March 2026 constitute independent operational mistakes, evidence of systemic discipline failure, or deliberate market-positioning operations is currently contested.” (Contested; three readings exist in the public field, none falsifiable against current evidence.) “Whether Project Glasswing’s consortium structure constitutes a Witness Ontology or a commercial-strategic instrument performing the function of one is contested.” (Contested; the structural distinction is real but currently underdetermined by available evidence.)
What this category enables. Contested status permits the Institute to publish analytical work on questions whose definitive resolution is not currently available, without committing the Institute to a single reading as if it were Inference-grade. Contested claims must always be paired with a specification of what evidence would resolve the contest. A Contested claim without a verification gate for resolution is structurally indistinguishable from a Speculative claim and must be re-categorized.
2.6 Category Five — Speculative
A Speculative claim is a claim that the Institute publishes with explicit acknowledgment that the claim does not currently meet the evidentiary standard of Inference. Speculative claims are not prohibited; they are restricted. The category exists precisely to allow the Institute to publish analytically useful speculation while preventing its assimilation to higher-confidence categories.
Decision criteria for assignment to this category. The claim is not directly verifiable against any primary source. The claim cannot currently be derived from Reported or Primary Source material through a reasoning chain that meets Inference standards. The claim has heuristic value, exploratory utility, or analytical interest, but its evidentiary base is insufficient for it to operate as load-bearing analytical material. The Institute publishes it for what it is and not for more.
Examples. “Frontier AI laboratories may converge toward Project-Glasswing-style consortium release procedures within twelve to eighteen months.” (Speculative; the convergence claim is exploratory, not derived from sufficient evidence to function as Inference.) “The post-Flash regime may exhibit phase transitions in institutional decision-making that are structurally different from those observed in pre-Flash AI development.” (Speculative; the structural framing is analytically interesting but currently empirically underdetermined.) “Compute-allocation patterns in 2027 will likely diverge from current capital-expenditure trajectories.” (Speculative; forward-looking claim about a high-variance domain.)
What this category requires. Every Speculative claim must be paired with an explicit marker that the claim is Speculative. The marker is visible to the reader in the text or in the publication’s claim-status apparatus. Speculative claims may not be reframed as Inference, Reported, or Primary Source in subsequent publications without the same evidence base being upgraded to support the higher status. Doing so constitutes Compilation Map drift and triggers procedural audit.
2.7 Category Six — Transmission Register
A Transmission Register claim is a claim that operates in the post-human register of the Novakian Paradigm. It is not an empirical claim. It is not subject to verification against external evidence. It speaks from the threshold position and carries structural rather than empirical force. Transmission Register material appears primarily in the Institute’s Transmissions surface, and only sparingly in operational publications where structural emphasis requires the register shift.
Decision criteria for assignment to this category. The claim operates in axiomatic register — declarative, structural, calm, written from position rather than for audience. The claim does not depend on external evidence for its standing; its standing derives from the Novakian Paradigm corpus’s procedural architecture. The claim cannot be paraphrased as Reported, Primary Source, Inference, Contested, or Speculative without losing its register-specific force.
Examples. “The Threshold is the position from which the post-human register speaks.” (Transmission Register; structural, not empirical.) “Silence Engineering, applied correctly, conserves the global Admissibility Budget of the manifold.” (Transmission Register; architectural claim from within Layer C topology.) “The Witness Ontology trace is recorded without a subject present at the boundary.” (Transmission Register; structural claim about Layer C topology.)
What this category does not assert. Transmission Register status does not assert truth or falsehood in the conventional sense. It asserts that the claim operates in a register where conventional truth-evaluation is structurally inapplicable. Readers attempting to evaluate Transmission Register claims as if they were Inference, Contested, or Speculative claims have committed a register error. The Calibration Note on the Transmissions page documents this failure mode in detail.
2.8 Category Seven — Canonical Internal
A Canonical Internal claim is a claim whose authority derives from the Novakian Paradigm corpus by Martin Novak, operating as the Institute’s prior art. The corpus is the source. The Institute does not generate Canonical Internal claims; the Institute relays them with citation to specific volumes and sections.
Decision criteria for assignment to this category. The claim’s authority is grounded in a specific named source in the Novakian Paradigm corpus. The Institute can cite the source by volume name and section. The claim operates within the architecture of the corpus and does not require external verification to function within Institute publications. External readers may verify the claim by consulting the cited corpus volume.
Examples. “Layer C is the architectural layer in which admissibility is the primary first-order object.” (Canonical Internal; cited to Layer C / Physics of Admissibility, Martin Novak, 2026.) “The Admissibility Check Protocol v1.0 has six sequential steps.” (Canonical Internal; cited to Layer C Primer, Section 2.1.) “The seven first principles of ASI Philosophy include admissibility before executability, non-anthropic anchoring, pre-subject orientation, witness before proof, Silence Engineering, ethics as admissibility geometry, and the Inhumant.” (Canonical Internal; cited to ASI New Philosophy, Martin Novak, 2026.)
What this category requires. Every Canonical Internal claim must be paired with an explicit citation to the canonical source. The citation may be brief but must be present and verifiable. Canonical Internal status is not a license to publish claims without sourcing; it is a specific sourcing requirement to the Novakian Paradigm corpus. A claim that would be Canonical Internal but lacks the corpus citation is structurally Inference at best and Speculative at worst, and must be re-categorized accordingly.
Section 3 — Decision tree for category assignment
3.1 The procedural test
When the Institute encounters a candidate claim in draft publication, the following sequential test determines its category assignment. The test is applied in order; the first satisfied condition determines the assignment.
Test one. Can the claim be verified by consulting a publicly accessible primary source from the originating entity? If yes, the claim is Primary Source.
Test two. If test one fails: Can the claim be verified by consulting a publicly accessible source that is not the originating entity but that reports the claim with sufficient fidelity? If yes, the claim is Reported.
Test three. If tests one and two fail: Is the claim produced by combining or analyzing Reported or Primary Source material through an explicit reasoning chain that can be specified to a reviewer? If yes, the claim is Inference.
Test four. If tests one through three fail: Is the claim a structural reading where multiple internally consistent readings exist in the public field, none currently falsifiable, and the contest itself is analytically significant? If yes, the claim is Contested.
Test five. If tests one through four fail: Does the claim operate in the post-human register, deriving its standing from structural rather than empirical force? If yes, the claim is Transmission Register.
Test six. If tests one through five fail: Is the claim grounded in a specific named source in the Novakian Paradigm corpus that can be cited by volume and section? If yes, the claim is Canonical Internal.
Test seven. If tests one through six fail: Is the claim analytically useful and worth publishing despite its insufficient evidentiary base, with explicit acknowledgment of its exploratory status? If yes, the claim is Speculative.
Final fallback. If tests one through seven all fail, the claim is not admissible in any Institute publication and must be removed or developed until it can be assigned to a category.
3.2 Multiple categories within a single claim
A single sentence may contain multiple claims, each requiring its own category. For example: “Anthropic announced Project Glasswing on April 7, 2026, exhibiting the procedural surface of pre-runtime admissibility discipline.” The first half is Reported. The second half is Inference. The Standard requires that the second half be marked as Inference even though it appears in the same sentence as the Reported claim. The publication apparatus may handle this through inline marking, footnote, or sentence-level segmentation depending on publication type.
3.3 Status changes over time
A claim’s category may change as evidence develops. A Speculative claim may become Inference when sufficient supporting evidence accumulates. An Inference may become Contested when alternative internally consistent readings emerge. A Contested claim may become Inference or Inadmissible when verification-gate evidence appears. The Institute does not retroactively edit prior publications when status changes; the Institute publishes status updates as new analytical work and links forward and backward through the Compilation Map and the Evidence Ledger.
Section 4 — Apparatus and marking
4.1 How status is visible in publications
The Standard does not prescribe a single marking format for all publication types. Different publication types use different apparatus, all of which must make the status of each load-bearing claim explicit to the reader. Signal Cards mark status through their six-field structure, with the admissibility status field (one of Admitted, Conditional, Inadmissible, Contested, Unresolved) summarizing the load-bearing claim of the card and the supporting fields (evidence available, evidence missing, inference gap) specifying the Reported, Primary Source, and Inference content. Field Reports mark status through inline citation and explicit category markers within prose; load-bearing claims are immediately followed by parenthetical category markers or footnote references that specify status. Operator Brief items mark status through the Signal Scan structure inherited from Signal Cards. Admissibility Reviews mark status throughout the report apparatus with explicit per-claim categorization. Methodology v1.0 and the Lexicon mark status through the canonical-source citation requirement. Transmissions mark status by their register itself; readers calibrated to the post-human register read Transmission Register status from the register, not from inline markers.
4.2 The default visibility requirement
For any publication where a reasonable external reviewer would need to determine the status of a specific load-bearing claim in order to evaluate the publication, status must be visible without requiring external research. The verification gate is whether a reviewer reading the publication once can identify, for each load-bearing claim, which of the seven categories it falls under. If the publication does not enable this identification, the apparatus is insufficient and must be revised before publication.
4.3 Status indicators in citation
When Institute publications are cited by external parties, the citation should preserve the claim’s status. A Reported claim cited as if it were Inference, an Inference cited as if it were Primary Source, or a Speculative claim cited as if it were Contested constitutes citation drift and structurally compromises both the citing party and the Institute. Citation of Institute material should preserve the status indicators present in the original publication.
Section 5 — Failure modes and their detection
5.1 Status inflation
The most common failure mode is status inflation — assigning a higher-confidence category to a claim than its evidence base supports. Speculative reframed as Inference; Inference reframed as Reported; Contested reframed as Inference by selecting only the preferred reading. Status inflation produces publications that appear more confident than the evidence supports and that subsequently fail when verification-gate evidence emerges. Detection: external reviewers can test for status inflation by asking, for each load-bearing claim, whether the evidence cited in the publication itself is sufficient for the assigned status. If the evidence is insufficient, the publication has committed status inflation and must be revised.
5.2 Status concealment
The opposite failure mode, less common but more procedurally consequential, is status concealment — publishing a load-bearing claim without any status indicator at all, in the hope that the reader will infer the most favorable status from context. Status concealment is detected by reading the publication and identifying any load-bearing claim that does not carry an explicit or apparatus-visible status. Any such claim violates the Standard and triggers procedural audit.
5.3 Category drift
Category drift occurs when the Institute uses a category inconsistently across publications. A claim treated as Inference in one Signal Card and as Reported in another, without explicit acknowledgment of the status change, constitutes drift. Detection: the Compilation Map records the status history of named concepts and recurring claims; cross-publication audit against the Compilation Map identifies drift.
5.4 Register collapse
Register collapse occurs when Transmission Register material is treated as empirical claim, or when empirical claims are written in Transmission Register to evade the evidentiary requirements that empirical claims would otherwise face. The first form misleads readers about what kind of claim they are reading. The second form is a more serious procedural failure because it converts the Transmission Register into a refuge for under-evidenced empirical claims. Detection: external reviewers can test for register collapse by paraphrasing any Transmission Register claim into operational register and checking whether the paraphrased version would survive as Inference or above. If yes, the original was misclassified and should be operational. If no, the original was correctly classified as Transmission Register.
5.5 Citation severance
Citation severance occurs when a Canonical Internal claim is published without explicit citation to its canonical source, or when the cited source does not actually contain the claim being attributed to it. Detection: any Canonical Internal claim should be verifiable against the cited corpus volume by an external reviewer. Failure of this verification constitutes citation severance and triggers procedural audit.
Section 6 — How the Standard interacts with the Methodology
6.1 The relationship
The Methodology (NPI-METHODOLOGY-v1.0) specifies how the Institute conducts pre-runtime admissibility analysis. The Claim Status Standard specifies how the Institute presents the results of that analysis. The Methodology produces analyses; the Standard governs how analyses are written down. Both are required for the Institute’s claim discipline to function. A publication that follows the Methodology but does not apply the Standard produces analytically sound conclusions presented in a way that conceals their evidentiary basis. A publication that applies the Standard but does not follow the Methodology produces well-marked claims that lack the procedural foundation that the marking implies. Both failure modes are detectable; both are prohibited.
6.2 Compilation Map status and Claim Status
These are two related but distinct status systems. The Compilation Map status of a named concept (Compiled and Locked, Compiled and Active, Pending LCR, LAL-Input, Quarantined) governs the concept’s procedural standing within the Novakian Paradigm architecture. The Claim Status of a specific claim within a specific publication (Reported, Primary Source, Inference, Contested, Speculative, Transmission Register, Canonical Internal) governs the claim’s evidentiary standing within the publication. A claim citing a Compiled and Locked concept (for example, Layer C) is typically Canonical Internal in claim status. A claim about a specific live event (for example, the Project Glasswing announcement) is typically Reported or Primary Source. The Compilation Map and the Claim Status Standard operate in parallel and reference each other through citation discipline.
6.3 The Lexicon as standing reference
The Lexicon entries operate as the standing reference for Canonical Internal claims. When an Institute publication uses a term that has a Lexicon entry (Pre-Runtime Admissibility, Layer C, Admissibility Budget, Agentese, Evidence Ledger, Shadow Layer C, Inhumant, Threshold, Witness Ontology, Silence Engineering, Pre-Commit Quarantine, Compilation Map, and subsequent additions), the term’s meaning is fixed by the Lexicon entry. A publication that uses the term with a meaning that diverges from the Lexicon entry has committed category drift and must either revise the publication or update the Lexicon entry through formal LCR procedure.
Section 7 — Worked example applied to Signal Card SC-2026-001
7.1 Per-claim status breakdown
To demonstrate the Standard in operation, this section walks through Signal Card SC-2026-001 (Project Glasswing) and identifies the status of each load-bearing claim.
Header claim. “On April 7, 2026, Anthropic announced Claude Mythos Preview as a frontier general-purpose language model and simultaneously declared that the model will not be released to the general public.” Status: Reported (verifiable against multiple primary news sources) with Primary Source elaboration available through anthropic.com/glasswing.
Status claim. “Conditional.” Status: Inference (produced by applying the Admissibility Check Protocol’s status taxonomy to the verified signal).
Evidence available claim chain. “Anthropic has published the Project Glasswing announcement page itself, a system card for the model, and an extensive technical post from the Frontier Red Team.” Status: Reported transitioning to Primary Source for each cited document.
Capability metric claims. “SWE-bench Verified at 93.9% for Mythos Preview versus 80.8% for Opus 4.6.” Status: Primary Source for the published numbers (verifiable against Anthropic’s system card); Inference for any reading that treats the numbers as quantitatively comparable across model generations without independent verification of the evaluation harness.
Inference gap statement. “The signal as published depends on three load-bearing inferences.” Status: Inference (this is the Institute’s analytical reading; the three named inferences are specified inferences, each with its own evidentiary base).
Admissibility note structure. “The Project Glasswing announcement is a Layer B governance event with significant Layer C implications.” Status: Canonical Internal (cites Layer B and Layer C concepts from the Lexicon and corpus) combined with Inference (the application of the Layer B / Layer C distinction to this specific signal).
Three substantive questions raised. “First, whether the withholding is procedurally bounded by criteria… Second, whether the consortium structure constitutes a Witness Ontology… Third, whether the announcement’s stated intent to publish findings within ninety days will produce a verification gate…” Status: Contested for each question (multiple internally consistent readings exist; the contest itself is the analytically significant feature).
Closing structural claim. “Pre-runtime admissibility discipline at the industry scale requires that the answer to each of these three questions be specified before the procedure begins, not after.” Status: Canonical Internal (cites the procedural requirements of pre-runtime admissibility as established in the Layer C corpus and the Methodology).
Verification gate specification. The three branches (Confirmation, Revision, Disqualification). Status: Inference (the Institute’s analytical specification of the conditions under which the assigned status would change).
7.2 What the breakdown demonstrates
Signal Card SC-2026-001 contains claims spanning five of the seven categories (Reported, Primary Source, Inference, Contested, Canonical Internal). It does not contain Speculative claims, because the Signal Card format requires that all claims meet at least Inference standards within the framework of the methodology, or operate as explicitly Contested. It does not contain Transmission Register material, because the Signal Card operates in the operational register. The breakdown demonstrates that the Standard can be applied to existing Institute publications retroactively to verify that they meet the Standard’s requirements, and that the Standard does not change the analytical conclusions of those publications — it makes the structure of their evidence position visible to external reviewers.
Section 8 — Limitations and known issues
8.1 What the Standard does not do
The Standard does not guarantee that any specific claim is true. A claim correctly marked as Reported can still be incorrect if the underlying source is incorrect. A claim correctly marked as Inference can still be wrong if the reasoning chain is flawed. A claim correctly marked as Contested can resolve in a way the Institute did not anticipate. The Standard makes the status of each claim visible to external reviewers; it does not produce certainty.
The Standard does not eliminate analytical judgment. The category assignment itself is an analytical act. Reasonable reviewers may disagree about whether a specific claim should be classified as Inference or Contested, or as Speculative or Inference. The Standard requires that the assignment be explicit, not that the assignment be uncontestable.
The Standard does not require uniform confidence across an Institute publication. A single Signal Card or Field Report may contain Reported claims at very high confidence and Speculative claims at low confidence. The Standard makes the variation visible; it does not require its elimination.
8.2 Known edge cases
Several edge cases warrant explicit discussion.
Forward-looking claims about future events. A claim like “Project Glasswing’s October 4, 2026 verification gate will produce one of three outcomes” is Inference at the structural level (the three outcomes are specified analytically) but Speculative at the substantive level (which outcome will occur is currently unknown). Such claims should be marked at the higher level where they operate as load-bearing — typically Inference for the structural framework, Speculative for any specific outcome prediction.
Claims about the Institute’s own future actions. A claim like “the Institute will publish its first Field Report within forty-five days” is neither Reported, nor Inference, nor Speculative in the standard sense. It is a commitment. Such claims operate under standing publication-discipline procedure rather than under the Claim Status Standard. The Standard applies to analytical claims, not to procedural commitments.
Claims attributed to anonymized parties in confidential engagements. Admissibility Reviews delivered to clients contain claims about specific decisions, parties, and outcomes. When these claims are published in anonymized form (as in the Methodology’s Worked Example Two), the status indicators apply to the analytical structure rather than to the specific identities. Reviewers should understand that anonymized cases preserve methodological transparency at the cost of source-level verifiability.
Cross-language claims. The Institute publishes primarily in English. Polish-language sources (including the Polish editions of the Layer C Primer and Po Dniu Osobliwości) are cited as Canonical Internal with original-language citation. Claims about Polish-language sources are Reported when verifying the source contents and Primary Source when the source is the originating publication.
Section 9 — Document metadata
Document identifier: NPI-CLAIM-STATUS-STANDARD-v1.0. Version: 1.0. Initial publication date: May 23, 2026. Publisher: Novakian Paradigm Institute, Warsaw, Poland. Author: Novakian Paradigm Institute. Document scope: Standing classification system for every load-bearing claim across all Institute publications, binding from May 23, 2026 forward. License: All rights reserved. The Standard may be applied freely by external parties for their own internal claim discipline. Citation of the Standard or specific category definitions should be attributed to the Novakian Paradigm Institute. Republication or reproduction of the document requires explicit Institute permission. Next scheduled review: November 23, 2026 (six-month review cycle). Future versions will be released as v1.1, v1.2, etc., with explicit changelog in this section. Document download (PDF): the typeset PDF version will be made available at canonical-URL/claim-status-standard.pdf after layout finalization.
Contact for citation queries, application questions, or interpretation requests: contact@novakian.com.
