{"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__A","search_lanes":{"direct_problem":{"queries":["digital humanities scholarly editions versioning superseded versions confusion citation","research data superseded versions discoverability stale version citation problem","historical research corpus version control annotations versions provenance","digital archive catalog description outdated superseded metadata version"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"closest_prior_art":{"queries":["research data repository withdrawal tombstone superseded dataset preservation citation policy","digital repository retention policy dependency check citations before deletion dataset","digital archive restore testing preservation copies audit recoverability policy","research repository deaccession legal hold privacy withdrawal versions"],"source_ids":["SRC2","SRC4","SRC5","SRC6","SRC7","SRC8"],"no_result_note":null},"historical_terminology":{"queries":["archival reappraisal deaccessioning digital records retention schedule superseded copies review provenance","records continuum digital archives superseded versions disposition schedule historical research data"],"source_ids":["SRC5","SRC8"],"no_result_note":null},"products_practices_standards":{"queries":["site:w3.org/TR/dwbp data versioning best practice versions discoverable superseded","site:datacite.org versioning DOI previous version superseded official","site:zenodo.org versioning DOI specific version official","digital preservation retention review disposal quarantine restore testing standard archival repositories"],"source_ids":["SRC3","SRC4","SRC5","SRC6","SRC7"],"no_result_note":null},"non_english_regional":{"queries":["Versionierung digitale Editionen veraltete Versionen Archivierung Zitierbarkeit","gestion versions données recherche version obsolète archivage citation"],"source_ids":["SRC2"],"no_result_note":null},"composition_subproblems":{"queries":["research data repository withdrawal tombstone superseded dataset preservation citation policy","digital repository retention policy dependency check citations before deletion dataset","digital archive restore testing preservation copies audit recoverability policy","research repository deaccession legal hold privacy withdrawal versions"],"source_ids":["SRC4","SRC5","SRC6","SRC7","SRC8"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Digital Editions and Version Numbering","url":"https://dhq.digitalhumanities.org/vol/14/2/000455/000455.html","publisher":"Digital Humanities Quarterly","date_or_year":"2020","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["Mutable digital scholarly editions create citation and scholarly-record problems when a cited state cannot be identified.","Defined, citable version numbers were rare among surveyed editions, and no consensus practice existed.","Version identity and history are proposed as practical means to preserve intelligibility, citation, and reuse."]},{"source_id":"SRC2","title":"Versionierung von digitalen Editionen in der Praxis","url":"https://kups.ub.uni-koeln.de/70452/1/09_Buergermeister.pdf","publisher":"Institut für Dokumentologie und Editorik / Kölner UniversitätsPublikationsServer","date_or_year":"2023","source_type":"PRIMARY_RESEARCH","language":"German","claims_supported":["A quantitative inventory found versioning in 65 of 257 accessible digital-edition projects, showing that it was present but not routine.","Four practices were observed: change descriptions, revision descriptions, version control, and whole-system versioning; only eight examined projects exposed archived whole-system versions.","Archived predecessors, current versions, change logs, and recoverable historical states are implemented in some historical-edition projects.","The study reports user-navigation difficulty and substantial maintenance cost for recoverable whole-system versions, and says digital humanities lacked best-practice models for this task."]},{"source_id":"SRC3","title":"Data on the Web Best Practices","url":"https://www.w3.org/TR/dwbp/","publisher":"World Wide Web Consortium","date_or_year":"2017","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["The W3C recommends version indicators, version histories, provenance, persistent identifiers for individual versions and series, and a stable latest-version URI.","The recommendation explicitly connects versioning with preventing confusion, repeatability, comparison, trust, and discovery of newer versions.","It supplies testable implementation checks for distinguishing specific versions from the latest version."]},{"source_id":"SRC4","title":"Preservation policy","url":"https://library.ed.ac.uk/research-support/research-data-service/after/data-repository/service-policies/preservation-policy","publisher":"University of Edinburgh Research Data Service","date_or_year":"2024-08-21","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Updated research-data versions may be linked while the most recent version is clearly identified and an earlier version is withdrawn from public view.","Withdrawn items are preserved rather than deleted, identifiers remain, and tombstones record reasons and replacement links.","Original bitstreams, checksums, backups, legal withdrawal grounds, and succession arrangements provide preservation and authority safeguards."]},{"source_id":"SRC5","title":"Digital Preservation Policy","url":"https://www.naa.gov.au/about-us/who-we-are/accountability-and-reporting/archival-policy-and-planning/digital-preservation-policy","publisher":"National Archives of Australia","date_or_year":"2020-06-30","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["Archival metadata models must represent multiple versions, relationships, and complete preservation-action audit trails.","The Archives retains original records, uses fixity checks, and periodically tests backup viability and restoration ability.","The policy assigns responsibilities, requires authorization and validation, recognizes access restrictions and intellectual-property rights, and uses risk-based strategies by record type."]},{"source_id":"SRC6","title":"Deposit Policy","url":"https://dataservices.library.jhu.edu/data-sharing/deposit-policy/","publisher":"Johns Hopkins Sheridan Libraries Data Services","date_or_year":"undated; accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["The repository uses professional curators and identifies repository administrators as decision makers.","Published datasets receive five-year retention reviews considering citations, downloads, hosting cost, transfer, and continued retention.","Removed datasets receive citation-bearing tombstones, while privacy, consent, legal, and ethical requirements constrain deposit and access."]},{"source_id":"SRC7","title":"ICPSR Deaccession Policy","url":"https://www.icpsr.umich.edu/sites/icpsr/about/policies/deaccession","publisher":"Inter-university Consortium for Political and Social Research, University of Michigan","date_or_year":"undated; accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["ICPSR evaluates holdings on an ongoing basis using use, research value, duplication, errors, disclosure, and custodial criteria.","Deaccessioned data remain preserved but cease to be publicly accessible and searchable; their persistent description remains viewable.","The policy demonstrates access demotion distinct from destruction and allows stewardship transfer to preserve citability."]},{"source_id":"SRC8","title":"Guidelines for Reappraisal and Deaccessioning","url":"https://www2.archivists.org/sites/all/files/GuidelinesForReappraisalAndDeaccessioning-May2012.pdf","publisher":"Society of American Archivists","date_or_year":"2012","source_type":"TRADE_PROFESSIONAL","language":"English","claims_supported":["Reappraisal, retention review, disposition schedules, and documented deaccessioning are established archival terminology and practices.","The guidelines require systematic, transparent, documented review, administrative approval, ownership resolution, donor-agreement checks, and consideration of research use and value.","They distinguish reappraisal from automatic removal, allow deferment and later review, and identify transfer as preferable to destruction where continued access and preservation matter."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The core problem exists: mutable digital editions and datasets can lack clearly citable states, confuse the scholarly record, and make older versions difficult for users to interpret. Domain-specific evidence found versioning in only 65 of 257 examined digital editions and whole-system archived versions in only eight. However, the retained sources do not quantify wrong-version citation rates, reconciliation time, or active-index growth across the full proposed mixture of transcriptions, annotations, catalog descriptions, datasets, and drafts.","source_ids":["SRC1","SRC2","SRC3"],"uncertainty":"The evidence establishes version ambiguity and uneven practice more strongly than the proposed downstream error and labor magnitudes. Repository policies show concern and remedies but not measured incidence in historical-research corpora."},"adopter_evidence":{"status":"SUPPORTED","finding":"Identifiable adopters and authorizers include digital-edition project leads, repository curators and administrators, archive collection-management staff, principal investigators or depositors, and institutional administrative authorities. Existing policies assign review and preservation duties to these roles and require administrative, ownership, donor, legal, privacy, or records-schedule approval as applicable.","source_ids":["SRC4","SRC5","SRC6","SRC7","SRC8"],"uncertainty":"The exact authority split would depend on whether a pilot corpus is project working material, deposited research data, accessioned archival material, or an institutional record."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"Nearly every proposed mechanism has an operational analogue: explicit version identifiers and histories; latest-version pointers; archived predecessors; periodic retention review; access demotion without destruction; tombstones; documented appraisal; legal and ownership checks; preservation audit trails; and restore testing. These mechanisms are implemented across standards and repository policies, but no retained source demonstrates the complete composition operating on a closed historical-project corpus or reports its effect on wrong-version selection and retrieval time.","source_ids":["SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"],"uncertainty":"Published policies do not prove consistent execution, and the search did not provide implementation logs, dependency-graph completeness measurements, or comparative user-task results."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"Versioned digital scholarly editions","source_ids":["SRC1","SRC2","SRC3"],"same_problem":true,"same_causal_lever":true,"overlap":"They identify version states, expose histories, distinguish current from archived editions, preserve addressable predecessors, and support reproducible citation.","remaining_difference":"They generally do not combine scheduled artifact-class review, evidence-based staleness flags, dependency-gated search demotion, reversible quarantine, tombstones, and sampled restore tests, nor test retrieval outcomes against a baseline."},{"name":"University research-data preservation and retention policies","source_ids":["SRC4","SRC6"],"same_problem":true,"same_causal_lever":true,"overlap":"They identify newer versions, withdraw earlier or noncompliant items from public access, preserve identifiers or bytes, create tombstones, apply retention review, and name repository decision makers.","remaining_difference":"They govern deposited datasets rather than the heterogeneous internal derivative layers of historical research and do not require citation-dependency tracing or blinded wrong-version retrieval trials."},{"name":"ICPSR ongoing deaccession and access demotion","source_ids":["SRC7"],"same_problem":false,"same_causal_lever":true,"overlap":"Ongoing review uses research value, use, duplication, error, disclosure, and custodial criteria; deaccessioned material leaves search and public access while remaining preserved and persistently described.","remaining_difference":"Its stated trigger is collection deaccession rather than superseded-version confusion, and it does not describe reversible quarantine, dependency checks, or restore-sampled user testing."},{"name":"Archival reappraisal and digital-preservation controls","source_ids":["SRC5","SRC8"],"same_problem":false,"same_causal_lever":true,"overlap":"Systematic lifecycle review, documented decisions, preservation exceptions, authority checks, relationship metadata, audit trails, and restore testing closely match the proposal's safeguards.","remaining_difference":"The guidance addresses holdings appraisal and technical preservation, not an integrated active-discovery lifecycle for successive historical interpretations and derivatives."}],"contrastive_claim_remaining":"In a heterogeneous historical-project corpus that already has ordinary versioning or provenance, adding explicit discovery lifecycle states plus dependency-gated demotion, reversible quarantine, tombstones, and sampled restoration will reduce wrong-version selections and retrieval time without hiding task-relevant current artifacts, breaking dependencies, or losing recoverability.","contrastive_claim_falsifier":"The claim is falsified if a stratified baseline already yields unambiguous intended-version selection, or if the intervention produces no improvement in blinded selection accuracy or retrieval time, or causes any additional missed dependency, hidden task-relevant artifact, or failed restoration.","confidence":"MODERATE","search_limitations":"The bounded search retained exactly eight web-accessible sources and covered English, German, and French search terminology, but the final evidence is English and German. It did not inspect proprietary repository configurations, internal project folders, usage logs, unpublished policies, patent literature, or every regional vocabulary. Policies establish prescribed practice, not routine compliance or realized outcomes."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"Independent domain research and a W3C recommendation support the existence and consequence pathway of mutable, insufficiently identified versions, while the German inventory shows incomplete adoption in digital editions. The exact error and labor rates remain unmeasured but are suitable pilot outcomes rather than prerequisites.","source_ids":["SRC1","SRC2","SRC3"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Repository curators and administrators, edition project leads, archival staff, and institutional administrative authorities are identifiable; existing policies specify responsibility and approval constraints.","source_ids":["SRC4","SRC5","SRC6","SRC7","SRC8"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Prior art covers versioning, preservation, retention review, and deaccession separately, but leaves a distinct comparative claim for the integrated discovery-state, dependency-check, quarantine, and restore-tested workflow in historical project derivatives.","source_ids":["SRC1","SRC2","SRC4","SRC5","SRC7","SRC8"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A shadow pilot on at most 200 artifacts, with no deletion or public-index change, can compare blinded intended-version accuracy and retrieval time, trace dependencies, and restore ten samples. The constituent controls are operationally grounded in existing repository practice.","source_ids":["SRC2","SRC4","SRC5","SRC6","SRC7"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first step is labels-only and reversible, preserves all bytes, and halts on unresolved dependencies or failed restores. Existing guidance supplies administrative, ownership, donor, legal, privacy, records-schedule, audit, and restoration controls. Authority must be confirmed locally before the pilot, but no inherent stop remains for a closed, non-mutating shadow test.","source_ids":["SRC4","SRC5","SRC6","SRC7","SRC8"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially, including archival historical terms, standards and deployed repository policies, German and French terminology, and combined subproblems. Eight retained sources span seven publisher organizations and include primary research, a formal W3C recommendation, government policy, university repository policies, and professional archival guidance.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":true,"screen_survival":true,"remaining_research_value":"MODERATE","recommended_next_step":"Run the authorized six-week shadow pilot on one closed corpus, stratifying artifacts by type and age. Pre-register baseline versus proposed-state comparisons for wrong-version selection and retrieval time; require complete dependency review and zero failed restores among ten sampled archived copies; make no deletion or public discovery changes.","world_novelty_boundary":"This bounded public-web review supports only an adjacent-prior-art finding and a testable incremental claim. It cannot establish world novelty, patentability, freedom to operate, market size, routine institutional adoption, realized impact, or absence of undiscovered prior art."}