Claims & corrections
This ledger separates what QAEDA verifies, what the evidence merely supports, what remains attributed or unverified, and what is explicitly editorial inference.
Method
A broad subject can be well documented while a dramatic quotation, causal claim, or attribution inside it remains uncertain. The ledger therefore evaluates narrow propositions rather than granting an entire report one credibility score.
Public ledger
Reviewed 2026-09-02. This is a source-aware publication layer, not a claim of omniscience or a live-news feed.
Basis: The provenance-focused source set identifies dated public statements, petitions, protest documentation, graffiti, and later prosecutions involving the wording.
What would change this: Revise only if primary-source chronology materially changes the claim’s dating or prominence.
Basis: Several supplied interpretive reports assert movement-level use, but later provenance-focused reports did not locate a securely dated anti-war artifact or canonical Russian wording. Better-documented Russian slogans exist.
What would change this: A dated first-generation artifact—such as an original post, protest photograph, manifesto, police record, or contemporaneous report—showing movement-level anti-war use would justify reassessment.
Basis: Ben Macintyre’s 2020 biography attributes the shorter sentence to Smedley, and accessible 1931 Smedley writing strongly corroborates the context, but the underlying letter/manuscript or archival shelf mark has not yet been located in the reviewed material.
What would change this: Promote to verified quotation if the primary Smedley artifact behind the Macintyre attribution is recovered and matches the wording.
Basis: The reviewed research locates the full two-clause wording in a 2023 Daniel Tutt essay, while Macintyre’s earlier Smedley attribution stops after the shorter sentence.
What would change this: A pre-2023 Smedley primary source or securely sourced archival quotation containing the added sentence would justify reassessment.
Basis: The source corpus separates street protest, emigration, administrative/legal avoidance, conscientious objection, family activism, unauthorized absence/refusal by serving personnel, and a smaller destructive fringe with different constituencies and goals.
What would change this: Revise if stronger systematic research demonstrates a level of centralized organization or ideological unity not reflected in the current corpus.
Basis: The repression report synthesizes human-rights documentation, detention/prosecution data, and later cases, while also warning that destructive-incident totals are difficult to interpret because political motives, fraud, coercion, and proxy recruitment can be conflated.
What would change this: Revise if a more comprehensive comparable dataset materially changes the relative scale or classification of anti-war-related coercion and violence.
Basis: The reconstruction follows directly from the surviving character sequence and requires no spelling substitution.
What would change this: Revise only if the original surrounding source context shows that the interruption was not an inserted UI/token artifact.
Basis: The source-critical report finds multiple independent referents and search collisions. A second report intentionally uses the phrase as an interdisciplinary metaphor, which is interpretive rather than evidence of historical transmission.
What would change this: A documented transmission history or explicit source lineage linking the domains would be required to establish a shared historical concept.
Basis: The site’s own cognitive-liberty and anti-Nanny-AI materials explicitly distinguish QAEDA terminology from the external standards reference.
What would change this: Revise if QAEDA changes its terminology or the external standard later adopts and defines the same term.
Basis: This is QAEDA’s editorial inference from its evidence-first and cognitive-liberty commitments, reinforced by the provenance conflicts found in the supplied research corpus.
What would change this: This is a normative editorial principle rather than an empirical proposition; changes require an explicit editorial decision.
Basis: The new source set converges on query logs as unusually intimate records and describes modern inference systems that combine query content with account, location, session, and behavioral context.
What would change this: Narrow or revise if empirical evidence shows that the relevant query data are not retained, linked, or inferentially useful in the system being discussed; this claim is about capability and risk, not every provider implementation.
Basis: The supplied surveillance reports synthesize post-Snowden observational research and related studies showing changes in reading/search behavior while also noting causal and generalization limits.
What would change this: Revise the strength or scope if stronger causal studies contradict the effect, or if a specific surveillance context shows no meaningful behavioral change.
Basis: The source set documents many legitimate research, journalistic, educational, defensive, health, and critical reasons to investigate dangerous subjects and emphasizes the ambiguity of query-level intent inference.
What would change this: The principle should be narrowed only where additional context supplies independent evidence of concrete harmful conduct, a true threat, unlawful transaction, or directly dangerous operational request.
Basis: The architecture reports identify on-device processing, privacy-preserving relays, private retrieval, distributed aggregation, differential privacy, and anonymous-credential patterns that separate particular safety functions from general-purpose raw-history retention.
What would change this: Revise for any safety objective where credible evaluation shows that the privacy-preserving alternative cannot meet the required effectiveness, latency, integrity, or accountability threshold.
Basis: The historical and adversarial reports identify recurring mechanisms such as broad definitions, sunk infrastructure, secrecy, secondary use, weak sunsets, and institutional incentives while also documenting examples of rollback or termination.
What would change this: Revise if comparative historical evidence materially changes either the documented expansion mechanisms or the evidence that meaningful rollback can occur.
Basis: The adversarial safety report rejects both safety absolutism and cognitive-liberty absolutism, then converges on proportionality, least-restrictive alternatives, transparency, review, deletion, and protection of lawful inquiry.
What would change this: Revise the implementation, not the core principle, when evidence shows that a narrower measure fails to address a specific high-severity harm and a more restrictive intervention has demonstrably better marginal efficacy.
Basis: The minors report combines developmental evidence with a policy distinction between algorithmic amplification and deliberate inquiry, arguing that protecting vulnerable users need not require broad subject censorship.
What would change this: Revise by age group, risk domain, or intervention if developmental evidence or outcome studies show a materially different balance between autonomy and protection.
Basis: This is the published QAEDA methodology derived from the supplied audit research and encoded in /audit and /audit-framework.json.
What would change this: Revise when QAEDA formally changes the methodology; historical versions should remain identifiable by build/version.
Basis: The minors, privacy-architecture, and adversarial safety reports describe anonymous/tokenized age assertions and attribute-proof approaches as alternatives to universal identity-linked browsing.
What would change this: Revise for jurisdictions or implementations that legally or technically require stronger identity proof, or where the verifier architecture creates equivalent persistent linkage through other means.
Basis: The audit, search-profiling, and website-evaluation reports converge on notice, reasons, appeal, correction, and independent review as procedural checks on opaque automated gatekeeping.
What would change this: Implementation can vary with consequence and abuse risk; revise if a specific disclosure or appeal mechanism would itself create a demonstrable security or privacy harm that cannot be mitigated.
Basis: The censorship corpus distinguishes direct prohibition from visibility-layer controls and documents ranking, moderation, recommendation, hosting, and distribution systems that can change reach without removing the underlying artifact.
What would change this: Narrow any provider-specific application unless evidence establishes that a particular visibility change was intentional, sustained, and consequential; capability does not prove censorship motive.
Basis: The comparative speech-governance reports distinguish constitutional/state-action constraints, private editorial discretion, intermediary pressure, and situations where government coercion or delegation can alter the analysis.
What would change this: Apply the distinction jurisdiction by jurisdiction and revise where law treats a private actor as state action, common carriage, delegated enforcement, or another legally distinct category.
Basis: The reader-privacy, surveillance, and censorship-psychology reports converge on intellectual privacy as upstream of expression and on observation as a mechanism that can alter lawful inquiry.
What would change this: The principle can be implemented differently by context; narrow specific privacy expectations where a transaction necessarily requires identity or where targeted lawful process is justified by independent evidence.
Basis: The democratic-drift, early-warning, speech-regulation, and wartime reports identify recurring expansion mechanisms while also distinguishing legitimate targeted restrictions, judicial review, and reversible controls.
What would change this: Revise the strength of the warning when comparative outcome data show that particular safeguards reliably prevent expansion or that an alleged drift mechanism did not produce material restriction.
Basis: The wartime report distinguishes tactical operational-security needs from political embarrassment, broad sedition controls, and suppression of accountability; QAEDA applies its existing harm-specificity and least-restrictive tests to that distinction.
What would change this: Implementation must vary with conflict conditions and genuine operational sensitivity; revise where disclosure risk cannot safely be described publicly but can be independently reviewed.
Basis: The archival, diaspora, and historical-memory reports converge on geographic, institutional, and format diversity as resilience against confiscation, deletion, institutional collapse, and narrative capture.
What would change this: Revise implementation advice as preservation technology changes; the core redundancy claim should narrow only if independent custody creates greater integrity risk than single-custodian preservation in a defined context.
Basis: The financial-pressure and journalism reports document financial exclusion, advertiser pressure, banking restrictions, and payment chokepoints as mechanisms that can affect whether a speaker or publisher can continue operating.
What would change this: Do not infer censorship from every commercial termination; evaluate contractual reason, state involvement, market alternatives, notice, discrimination, and the materiality of the financial dependency.
Basis: The language and translation reports synthesize historical and contemporary examples of lexical control, selective translation, omission, reframing, and state-mediated foreign-content presentation.
What would change this: Treat any specific translation as an empirical artifact requiring source-text comparison; do not infer coordinated censorship from ordinary translation variation, ambiguity, or error alone.
Basis: The personalized-censorship and algorithmic-suppression reports describe user-specific ranking, behavioral profiles, recommendation systems, and differentiated visibility as existing technical capabilities.
What would change this: Do not claim intentional personalized censorship by a provider without controlled evidence showing systematic differential treatment tied to a protected or viewpoint-relevant variable.
Basis: The AI training-data and regulation reports distinguish corpus construction and filtering from later fine-tuning, preference optimization, system instructions, retrieval, and runtime moderation.
What would change this: Revise model-specific claims as architectures change; this claim concerns distinct intervention layers rather than asserting that every filter is ideological censorship.
Basis: The digital-identity report and earlier privacy-preserving safety corpus discuss selective disclosure, anonymous credentials, and age/eligibility assertions as designs that separate an attribute from a full identity dossier.
What would change this: Revise where law, fraud risk, accountability, or technical limitations genuinely require stronger identification, and evaluate whether supposedly anonymous systems remain linkable in practice.
Basis: The censorship-resistance, uncensored-network, archival, and diaspora reports repeatedly treat resilience as a multi-layer property and describe the limits and trade-offs of individual technologies.
What would change this: Revise the specific architecture as threat models change; retain the layered-system principle unless a demonstrably universal single mechanism emerges.
Basis: The Russian censorship-system, information-access, and internet-sovereignty reports independently describe a layered legal and technical architecture including Roskomnadzor/TSPU controls, blocking, throttling, platform restrictions, and punitive speech laws.
What would change this: Current implementation details require fresh verification; revise any component as law, infrastructure, provider behavior, or enforcement changes.
Basis: The Russian information-access, resistance-history, diaspora, and psychology reports describe continued information-seeking, preference falsification, alternative channels, and adaptive circumvention under restrictive environments.
What would change this: Do not generalize the amount of access or dissent across populations; revise with stronger behavioral data for a specific country, period, or technology regime.
Basis: The historical-record, educational-censorship, diaspora, and global-history reports describe multiple mechanisms by which access, preservation, curriculum, and official narrative affect which records remain socially visible.
What would change this: Specific historical cases require source-level verification; revise where evidence shows preservation and public access were not materially affected by the alleged control mechanism.
Correction log
These entries record material narrowing or correction of claims in the supplied research corpus or QAEDA synthesis. Preserved source reports are not rewritten to hide the disagreement.
Earlier/source position: Several supplied interpretive reports described the phrase as an underground or defining Russian anti-war/partisan slogan.
Current position: QAEDA does not verify movement-level provenance and now labels the claim unverified unless a dated primary artifact emerges.
Why: Later provenance-focused research found a much stronger documentary record for ‘Нет войне’, anti-mobilization wording, coded dissent, and survival/refusal language than for ‘Rebel or Die’.
Earlier/source position: Some later writing presents the full two-clause ‘rebel or die / rebel and die’ wording as Smedley’s quotation.
Current position: The shorter sentence is a qualified Smedley attribution through Ben Macintyre; the added sentence remains unauthenticated to Smedley in the reviewed evidence.
Why: The primary Smedley artifact behind the shorter attribution remains unidentified, and the complete two-clause wording was located later in the source trail.
Earlier/source position: One supplied report deliberately universalizes the phrase as an interdisciplinary archetype.
Current position: QAEDA preserves that as metaphor while rejecting it as proof of one historical referent or transmission chain.
Why: The source-critical investigation finds multiple unrelated exact and near-exact uses, plus search/OCR false friends.
Earlier/source position: The strongest rhetoric around the thesis can imply that safety controls are equivalent to surveillance and that exceptional restrictions inevitably become permanent.
Current position: QAEDA treats function creep, opacity, weak efficacy review, and generalized inquiry surveillance as serious risks while explicitly rejecting the claims that every safeguard is surveillance or that expansion is inevitable.
Why: The supplied full adversarial investigation found real current safety harms, meaningful narrow interventions, and historical examples of rollback, while preserving the case for least-restrictive design and strong review of broad powers.
AEO quick answers
No. It means the narrow proposition is strongly established by the evidence QAEDA reviewed. New primary evidence can still require revision.
Because uncertainty is part of the record. Hiding disputed claims makes it easier for the same attribution error to return later.
Not necessarily. It is a different kind of statement: an interpretation or normative judgment that should be argued for rather than disguised as a source fact.