{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"layer_decay_and_expiration_management__history_historiography","opaque_id":"layer_decay_and_expiration_management__history_historiography__B","search_lanes":{"direct_problem":{"queries":["museum catalog outdated descriptions harmful language review policy historical records","research users trust museum labels authority outdated interpretation visitor study","museum labels perceived authoritative visitor interpretation study historical exhibits","archives users outdated description confusion current language study"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"closest_prior_art":{"queries":["site:wikidata.org deprecated rank statements reason for deprecated preferred rank help","digital historical encyclopedia versioning superseded interpretations archive editions","Stanford Encyclopedia of Philosophy archives substantive revision current version citation fixed edition policy","living history encyclopedia archived versions current article revision superseded"],"source_ids":["SRC4","SRC5","SRC6"],"no_result_note":null},"historical_terminology":{"queries":["\"legacy descriptions\" review archive preserve previous descriptions","\"reparative description\" version history preserve original description","\"retrospective cataloguing\" outdated descriptions museum","\"superseded\" \"historical interpretation\" museum"],"source_ids":["SRC2","SRC3","SRC5"],"no_result_note":"Older terms such as legacy description, reparative description, retrospective cataloguing, dynamic reference work, and retirement from the active portion surfaced adjacent practices, but no older source located the complete claim-level experimental package."},"products_practices_standards":{"queries":["knowledge graph deprecated statements lifecycle provenance historical interpretation","W3C PROV invalidation revision entity deprecated claim standard","museum collection management standard description review version history outdated","site:icom.museum museum catalogue outdated terminology contextualize records guideline"],"source_ids":["SRC4","SRC5","SRC6","SRC7"],"no_result_note":null},"non_english_regional":{"queries":["museum veraltete Begriffe Sammlungsdatenbank historische Beschreibungen aktualisieren Kontextualisierung","musée termes obsolètes notices catalogue contextualiser anciennes descriptions","museo lenguaje obsoleto catálogo descripciones históricas actualizar conservar contexto","archivo descripción lenguaje obsoleto revisión metadatos patrimonio"],"source_ids":["SRC8"],"no_result_note":"German-language Austrian practice produced a direct regional analogue; the French and Spanish searches produced standards and metadata material but no closer retained source."},"composition_subproblems":{"queries":["historical knowledge base claim lifecycle superseded status label","knowledge graph deprecated statements lifecycle provenance historical interpretation","Stanford Encyclopedia of Philosophy archives substantive revision current version citation fixed edition policy","\"reparative description\" version history preserve original description"],"source_ids":["SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Learning in the history museum: designing exhibit labels that engage visitors","url":"https://unbscholar.lib.unb.ca/items/d879b56f-4c02-45b2-a16b-5cf11348a652","publisher":"University of New Brunswick","date_or_year":"2015","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["A qualitative history-museum study using redesigned labels, a post-visit survey, and interviews found that labels can substantially influence visitor engagement and connections with displayed objects.","Interpretive display text is consequential enough that user-facing status and framing interventions are empirically testable."]},{"source_id":"SRC2","title":"Statement on potentially harmful content","url":"https://collections.reading.ac.uk/about-us/statement-on-potentially-harmful-content/","publisher":"University of Reading Museums and Collections","date_or_year":"February 2026","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["An operating cultural-heritage institution reports potentially harmful language or imagery in online catalogue entries.","Its practice includes surveying collections, contextualizing content, making some terms less immediately visible, flagging content at first engagement, preserving former catalogue descriptions, soliciting reports, and annual policy review.","This establishes an identifiable institutional adopter and a preservation-versus-current-presentation problem closely resembling the proposal."]},{"source_id":"SRC3","title":"Problematic Language in Archival Description","url":"https://www.lib.uwo.ca/archives/problematic_language.html","publisher":"Western University Libraries","date_or_year":"Undated; accessed 2026-08-04","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Western Archives states that archival titles and descriptions can contain outdated language and that descriptions may embody decades-old archivist or creator biases.","It commits to continuous updating and contextualization while preserving previous descriptions for context and research.","The institution provides a feedback route for harmful or inaccurate descriptions, identifying both an adopter and a review trigger."]},{"source_id":"SRC4","title":"Help:Ranking","url":"https://www.wikidata.org/wiki/Help%3ARanking/en-gb","publisher":"Wikidata, Wikimedia Foundation","date_or_year":"Continuously maintained; accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["Wikidata implements statement-level normal, preferred, and deprecated lifecycle-like ranks.","Preferred ranks can identify current or consensus statements and affect default query results; deprecated statements represent errors or outdated knowledge and are excluded by default unless specifically requested.","Deprecated statements remain verifiable and preserved, can carry reasons for deprecation, and support reconstruction of the evolution of ideas rather than deletion.","The system distinguishes outdated knowledge from correct historical values, which instead receive temporal qualifiers, directly addressing the proposal's analogy-break concern."]},{"source_id":"SRC5","title":"Guidelines and Policies for Entry Content","url":"https://plato.stanford.edu/archives/fall2025/guidelines.html","publisher":"Stanford Encyclopedia of Philosophy, Stanford University","date_or_year":"Fall 2025 edition","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["SEP requires expert authors to maintain and periodically update dynamic scholarly entries, typically every four to five years, in response to research and valid criticism.","Editors and the editorial board determine the validity and timing of requested revisions.","Entries not revised within an agreed timetable may be retired from the active encyclopedia and retained in archives until successfully revised.","The policy preserves citable archived versions while allowing discussions no longer central to be retired from current text."]},{"source_id":"SRC6","title":"About the Stanford Encyclopedia of Philosophy","url":"https://seop.illc.uva.nl/about.html","publisher":"Stanford Encyclopedia of Philosophy, official ILLC mirror","date_or_year":"Current page describing the publishing model; accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["SEP operates a dynamic scholarly reference work whose entries and substantive updates are reviewed by an editorial board.","Its infrastructure assigns editors and authors, tracks deadlines and entry states, logs interface actions, sends reminders, supports side-by-side revision comparison, and creates quarterly fixed archives.","The practice demonstrates that scheduled revalidation, accountable human authorization, current publication, stable citation, and recoverable archival history can coexist in a curated interpretive knowledge system."]},{"source_id":"SRC7","title":"The PROV Namespace","url":"https://www.w3.org/ns/prov","publisher":"World Wide Web Consortium","date_or_year":"2013; continuously hosted","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["W3C PROV supplies standard concepts for revision, derivation, attribution, responsibility, usage, and invalidation or expiry.","These concepts provide established technical primitives for provenance, lineage, accountable lifecycle events, and revised entities, although they do not prescribe editorial status or discovery ranking."]},{"source_id":"SRC8","title":"„Problematische Objekte Online“ – Digitalisierung mit Kontextualisierung im Salzburg Museum","url":"https://www.noemuseen.at/news-list/blog-detail/news/detail/News/problematische-objekte-online-digitalisierung-mit-kontextualisierung-im-salzburg-museum/","publisher":"Museumsmanagement Niederösterreich","date_or_year":"2024-03-18","source_type":"TRADE_PROFESSIONAL","language":"German","claims_supported":["The Salzburg Museum found outdated classifications and language inherited from card catalogues and earlier databases persisting as legacy deposits in its current database.","Its working group inventories affected records, reviews them with collection leads, decides case by case whether historically necessary terms should remain, adds automated distancing notices and contextual glossaries, and preserves broad online visibility.","The regional example confirms both the practical problem and the safety tension between contextualization and suppressing difficult historical evidence."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"Public institutions explicitly report outdated, biased, or harmful legacy descriptions persisting in catalogues and databases, and a history-museum study shows that labels shape visitor engagement. The exact asserted outcome—readers mistaking superseded interpretive claims for the institution's current conclusion and then making citation or teaching errors—was not directly measured by the retained sources.","source_ids":["SRC1","SRC2","SRC3","SRC8"],"uncertainty":"The evidence is strongest for legacy descriptive language and interpretive framing, not for a quantified rate of mistaken-current judgments about individually citable historiographic claims."},"adopter_evidence":{"status":"SUPPORTED","finding":"Museums, archives, university collections, collaborative knowledge bases, and editorial boards are identifiable adopters and authorizers. The retained sources show operational responsibility assigned to collection staff, working groups, expert authors, editors, and editorial boards.","source_ids":["SRC2","SRC3","SRC5","SRC6","SRC8"],"uncertainty":"A particular pilot institution would still need to designate its board, community-consultation process, staff owner, and research-ethics pathway."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"Major components are already implemented separately or in adjacent combinations: statement ranks and default retrieval filtering in Wikidata; scheduled scholarly revision, retirement, archives, deadline tracking, and audit logs in SEP; inventory, contextualization, reduced prominence, preservation of former descriptions, and human review in museum and archival practice; and provenance primitives in W3C PROV. No retained source demonstrated the whole claim-level historical-interpretation registry or the proposed randomized comparison against label-only treatment.","source_ids":["SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"],"uncertainty":"Feasibility is supported by composable precedents, but metadata cost, inter-rater reliability, downstream citation effects, representation effects, and treatment efficacy remain unmeasured."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"Wikidata statement ranks and query filtering","source_ids":["SRC4"],"same_problem":true,"same_causal_lever":true,"overlap":"Individually represented claims can be preferred, normal, or deprecated; outdated knowledge is retained with sources and reasons while deprecated values are excluded from default query use. Historical values and disputed claims receive separate treatment.","remaining_difference":"Wikidata does not supply the proposal's age-based review cadence, historical-value and representation-risk assessment, institution-specific editorial authority, appeal safeguards, or controlled test in curated historical interpretation."},{"name":"Stanford Encyclopedia of Philosophy dynamic publishing and retirement model","source_ids":["SRC5","SRC6"],"same_problem":true,"same_causal_lever":true,"overlap":"A curated interpretive reference work schedules updates, tracks responsibility and deadlines, logs work, permits editorial retirement from the active resource, and preserves fixed citable archives with restoration possible after revision.","remaining_difference":"Lifecycle management occurs mainly at entry and passage level rather than through an exposed registry of individually citable interpretive claims; no retained evidence compares status-label-plus-demotion with label-only discovery."},{"name":"Reparative archival and museum-description workflows","source_ids":["SRC2","SRC3","SRC8"],"same_problem":false,"same_causal_lever":true,"overlap":"Institutions inventory legacy descriptions, review and contextualize them, reduce immediate visibility in some cases, preserve earlier descriptions, use human and community input, and maintain feedback routes.","remaining_difference":"These workflows focus chiefly on harmful terminology, cataloguing description, and problematic objects rather than scholarly supersession status across granular historiographic claims; they do not establish a general current/contextual/archived claim lifecycle."},{"name":"W3C PROV revision and invalidation model","source_ids":["SRC7"],"same_problem":false,"same_causal_lever":false,"overlap":"The standard supplies reusable provenance, responsibility, derivation, revision, and invalidation semantics for an implementation.","remaining_difference":"It does not decide epistemic status, review urgency, preservation policy, display rank, community authority, or intervention effectiveness."}],"contrastive_claim_remaining":"Within one bounded curated historical resource, a human-governed claim-level registry that preserves provenance and prior states while visibly marking and demoting eligible superseded interpretations will reduce mistaken-current selections by at least 20% relative to equally prominent label-only treatment, without materially worsening retrieval, citation resolution, viewpoint representation, or curator workload.","contrastive_claim_falsifier":"The incremental claim is falsified if a representative baseline finds fewer than 5% mistaken-current selections, if an already operating history-specific system is found to use substantially the same claim-level lifecycle and discovery treatment routinely, or if the proposed pilot fails its 20% relative reduction threshold or breaches any retrieval, citation, representation, workload, appeal, or suppression guardrail.","confidence":"MODERATE","search_limitations":"This was a bounded public-web search retaining exactly eight direct sources. It covered six lanes and English, German, French, and Spanish terminology, but retained regional evidence only in German. Paywalled literature, internal museum systems, unindexed local policies, and patent or commercial-product databases were not comprehensively examined. The search cannot establish world novelty, patentability, freedom to operate, prevalence, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"Multiple operating cultural institutions document legacy descriptions that require updating or contextualization, and primary visitor research shows that history-museum labels affect engagement. That is sufficient external support for testing the narrower mistaken-current hypothesis, though it does not establish its prevalence.","source_ids":["SRC1","SRC2","SRC3","SRC8"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Museums and archives have collection staff and working groups, while SEP demonstrates expert authors and an editorial board with explicit revision and retirement authority. The proposal also names an institutional editorial or collections board and affected-community consultation.","source_ids":["SRC2","SRC3","SRC5","SRC6","SRC8"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Despite close analogues, the retained evidence does not show a history-specific, claim-level lifecycle registry tested through randomized discovery demotion against label-only treatment. The remaining claim has quantitative efficacy and non-inferiority guardrails.","source_ids":["SRC1","SRC4","SRC5","SRC6"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A reversible audit and eight-week randomized pilot on 100 claims in one digital exhibit is bounded in objects, institution, duration, treatments, metrics, and rollback conditions. Existing systems show that status, ranking, revision, and archival mechanics are implementable.","source_ids":["SRC2","SRC4","SRC5","SRC6","SRC7"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"The pilot can avoid destroying sources or changing stable citations, require human authorization and community consultation, retain audit history, and restore prior display states. The main risks—orthodoxy, presentism, unequal review, and suppression—are serious but are monitorable stop conditions rather than reasons to prohibit a reversible pilot.","source_ids":["SRC2","SRC3","SRC4","SRC5","SRC8"]},"adequate_search_evidence":{"status":"PASS","rationale":"The bounded search covered all six required lanes, retained exactly eight opened direct sources from independent publishers, and included primary research, official institutional guidance, first-party operational systems, and an official standard, plus non-English regional evidence.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"MODERATE","recommended_next_step":"Preregister a two-stage reversible study in one digital history exhibit. First, run the blinded representative baseline to estimate mistaken-current selections and stop if the rate is below 5%. If it is not, have authorized reviewers classify 100 claims with documented provenance, dependencies, preservation exceptions, and representation checks; randomize only eligible stale claims to label-plus-demotion versus label-only for eight weeks; measure mistaken-current selections, retrieval, citation resolution, viewpoint representation, appeals, and curator time; halt and restore displays on any stated guardrail breach.","world_novelty_boundary":"The search found close component-level and package-level analogues but no retained source for the exact history-specific claim-level randomized package. This supports only a bounded contrastive research claim. It does not establish world novelty, patentability, freedom to operate, market size, routine feasibility across institutions, or realized impact."}