{"actors":["Library director and assessment office commissioning the explanation","Head of metadata services who is publicly credited for the change","Catalogers, metadata librarians, paraprofessionals, and temporary grant-funded staff who performed remediation","Systems librarians and repository vendor staff who implemented indexing and workflow changes","Special-collections archivists who supplied collection knowledge and revised descriptions","Community representatives whose terminology feedback altered access points","Grant administrator and governing board that supplied resources and deadlines","Researchers and community users affected by discovery delays and descriptive choices"],"affected_objective":"Produce a defensible explanation of why a digitized special-collections backlog became discoverable more quickly during one fiscal year, so future staffing and workflow decisions target mechanisms that carried the change rather than the most visible participant.","arm":"COMMON_P1","authority_safety":{"authorized_first_step":"The assessment librarian may assemble and analyze a de-identified sample of 30 completed collections using existing workflow logs, change records, meeting minutes, and credit statements; interviews, personnel judgments, and public corrections require later authorization.","decision_authority":"The library director owns any resource or staffing decision; the head of special collections and community advisory body separately control changes to public collection descriptions. The analyst may classify causal evidence but may not assign employment responsibility, honors, blame, or compensation.","excluded_actions":["Do not infer causal weight from title, name frequency, seniority, or public visibility.","Do not publish individual performance rankings or use the analysis for discipline.","Do not alter finding aids, annual reports, staffing, or budgets during the evidence step.","Do not treat counterfactual substitutions as observed facts.","Do not equate causal contribution with moral credit, liability, or authority over community description."],"halt_rollback":"Stop the review if personnel-sensitive records cannot be de-identified, community-supplied evidence would be exposed, or the outcome definition changes mid-analysis. Quarantine the working matrix, retain the unmodified source records, and resume only after the relevant records or community authority approves a revised scope."},"baseline":"The annual report and internal retrospective center the head of metadata services because that person announced the initiative and appears in executive records. Explanations alternately credit that leader or call the improvement an automatic result of the repository migration, without separating initiation, resource authorization, standards work, distributed remediation, system execution, or user terminology feedback.","candidate_id":"agency_structure_attribution_balance__library_information_science__COMMON_P1","causal_chain":["The library fixes the explanatory target as the change in median interval from digitization completion to successful discovery through the public catalog for the selected backlog, compared with the preceding fiscal year.","A focality audit records why the department head dominates the account, including formal authority, announcement authorship, meeting attendance, and source abundance, without treating those signals as causal magnitude.","A collective inventory identifies cataloging, systems, archival, vendor, grant, and community work, while a structural map records the grant deadline, staffing allocations, metadata standard, repository capabilities, authority rules, and inherited backlog condition.","Dated evidence links specific decisions and omissions to workflow changes: prioritization and coalition work may initiate or coordinate, grants and positions enable, staff remediate records, systems changes execute and amplify, and community feedback changes access terminology.","A paired workflow-and-outcome timeline tests whether those actions preceded the relevant handoffs and discovery-state changes at collection level rather than merely coinciding with the annual result.","Actor-substitution probes compare bounded alternatives such as delayed leadership approval, another qualified role occupant, or absence of temporary staff, while structural probes vary the grant, batch-editing capability, or approval rule and explicitly record assumptions.","The synthesis states which contributions were conditional on position, resources, standards, distributed execution, and timing; preserves competing explanations; and separates this causal account from praise, blame, or personnel authority.","Future staffing and workflow decisions can then be tied to evidenced mechanisms, with review triggered by contradictory logs, additional participant evidence, a changed outcome definition, or consequential reuse."],"cell_id":"agency_structure_attribution_balance__library_information_science","consequence":"A person-centered account can direct future investment toward leadership symbolism while overlooking temporary capacity, standards, system affordances, and distributed description work; a structure-only account can erase consequential choices about sequencing, coalition formation, and terminology. Either error can distort organizational learning and credit without establishing responsibility.","diversity_from_prior_proposals":"No prior proposals were inspected under runtime isolation. This candidate is framed as a bounded organizational attribution review of a library discovery-workflow outcome, not as a general provenance audit, collection-description redesign, founder-legacy decision, or legal accountability process.","experiment_id":"eoa_inverse_innovation_exp13_second_slot_policy60_20260806","intervention":"Before using the backlog-clearance story for staffing, budgeting, or recognition, create a versioned Collection Discovery Attribution Dossier. It fixes one outcome and comparison period; inventories focal and background actors; maps positions, resources, standards, technologies, and constraints; traces dated actions across collection-level workflow states; classifies initiation, enabling, coordination, execution, amplification, and legitimation; audits narrative visibility; runs bounded actor-substitution and structural-relaxation probes; and publishes conditional conclusions with alternatives, uncertainty, and responsibility limits.","mechanism_mapping":[{"counterfactual_removal":"Without the matrix, visible announcements and repository migration can be listed beside staff work without showing which action or condition changed which collection-level state.","mechanism_slug":"actor_structure_evidence_matrix","role":"Cross-tabulates dated actor actions and structural conditions with evidence provenance, causal role, workflow stage, and observed discovery-state transition."},{"counterfactual_removal":"Without cross-level tracing, executive decisions may be mistaken for execution, or record-level changes may be disconnected from the catalog-wide outcome.","mechanism_slug":"process_tracing_across_levels","role":"Connects approval and resource decisions through team workflows and metadata transformations to indexing and user-visible discovery."},{"counterfactual_removal":"Without substitution probes, the review cannot distinguish effects of the named occupant from authority or capabilities attached to the position.","mechanism_slug":"counterfactual_actor_substitution_probe","role":"Tests absence, delay, and a qualified alternative occupant under the same bounded information, authority, resources, and timing."},{"counterfactual_removal":"Without structural probes, the leader's apparent contribution may absorb effects of grant labor, batch-editing technology, standards, or approval rules.","mechanism_slug":"structural_constraint_relaxation_probe","role":"Tests whether traced actor effects remain plausible when one resource, technical affordance, institutional rule, or deadline is changed."},{"counterfactual_removal":"Without focality auditing, source abundance and annual-report prominence can continue to function as unstated evidence of causal magnitude.","mechanism_slug":"narrative_focal_actor_audit","role":"Compares public and managerial visibility with workflow evidence and records contributors or affected groups missing from the dominant account."},{"counterfactual_removal":"Without plural review, one convenient causal story can close the inquiry despite conflicting timing, missing records, or sensitivity to assumptions.","mechanism_slug":"plural_causal_synthesis_review","role":"Retains actor-led, capacity-led, technology-led, and interaction models with their supporting evidence, contradictions, uncertainty, and review triggers."}],"nearest_rivals":["A metadata provenance audit can show who edited which records but does not by itself attribute the cross-level outcome, test positional replaceability, or vary structural conditions.","A workflow root-cause analysis can identify bottlenecks and technical dependencies but may omit narrative focality, distributed credit, actor substitution, and the boundary between causation and responsibility.","A narrative construction audit can reveal why the department head became the protagonist but does not determine whether that person's actions initiated, enabled, accelerated, or merely symbolized the operational change.","A personnel performance review evaluates employees against role standards and belongs to authorized management; this dossier explains a bounded organizational outcome and cannot supply disciplinary or reward judgments."],"negative_tests":{"intervention_falsifier":"Reject the intervention as decision-relevant if the sampled collection timelines cannot connect distinct actor actions or mapped structural conditions to observable workflow transitions, or if removing the focal actor's name leaves the same explanatory conclusions without any loss of mechanism or timing information.","problem_falsifier":"The proposed attribution problem is absent if existing records already define the same bounded outcome, distinguish person from position, include distributed and affected actors, trace dated mechanisms across scales, test bounded actor and structural counterfactuals, preserve alternatives, and separate causal contribution from responsibility.","risks":["Incomplete logs may make invisible labor appear causally absent.","Managerial records may overrepresent formal decisions, while informal repair and community input remain underdocumented.","Interviews conducted under hierarchy could create retaliation concerns or retrospective conformity.","Counterfactuals may smuggle in unsupported assumptions about substitute competence, legitimacy, or available time.","A simplified synthesis may be reused as a personnel judgment despite its stated boundary.","Collection-level sampling may miss heterogeneous workflows and produce aggregation error."],"strongest_counterevidence":"The strongest counterevidence would be collection-level timestamps showing that the improvement began before the focal head's relevant decisions, persisted across collections untouched by the attributed workflow, and is fully accounted for by an automated repository change applied uniformly without discretionary sequencing, coalition work, or distributed remediation."},"next_evidence_step":"For a stratified sample of 30 completed collections—10 processed before, 10 during, and 10 after the reported turning point—the assessment librarian will build a read-only matrix from existing timestamps, metadata diffs, tickets, approval records, meeting minutes, and public credit statements. The output is limited to a chronology, actor/position inventory, condition map, missing-evidence log, and two preregistered probes: delayed focal approval and absence of batch-editing capability. No interviews or causal verdict will occur at this step.","observable_state":"For each sampled collection, existing systems expose dated transitions from digitization complete to description ready, authority review, repository ingest, index publication, and a documented successful catalog retrieval. The same evidence set shows who initiated, approved, edited, implemented, or advised each transition; which resources and rules were active; and whose contribution appears or is absent in the annual report.","prior_art_status":"UNSEARCHED","problem":"A library plans to use a celebrated account of one metadata-services head 'clearing' a digitized special-collections backlog to guide the next budget and staffing cycle. The outcome is underspecified, the head is unusually visible in managerial sources, repository migration is sometimes treated as automatic causation, and the actions of catalogers, systems staff, archivists, temporary workers, vendors, and community advisers are compressed or omitted. The library therefore cannot tell which stage depended on individual choice, formal position, distributed execution, technical capacity, grant resources, inherited standards, or timing.","proposal_index":1,"remaining_contrastive_claim":"The dossier is warranted only if it changes the explanation by distinguishing the focal occupant's evidenced choices from authority attached to the role and from distributed, technical, and institutional mechanisms; merely adding contributor names or contextual variables would not instantiate the intervention.","revision_record":{"claim_changes":[],"conceptual_changes":[],"evidence_changes":[],"operational_changes":[],"parent_version":null,"progress_targets_addressed":["Instantiate the archetype in a concrete library discovery and metadata workflow.","Preserve actor–structure interaction, causal-role classification, cross-scale tracing, narrative audit, and bounded counterfactuals.","Define authority, safeguards, falsifiers, risks, and a limited first evidence step without asserting an effect."]},"schema_version":1,"structural_mapping":[{"archetype_element":"Explanatory Target, Scope, and Use","domain_realization":"Median interval from digitization completion to documented successful public-catalog discovery for the specified backlog, compared across two fiscal years and used only for organizational learning and future resource design."},{"archetype_element":"Focal Actor and Collective Inventory","domain_realization":"The credited metadata-services head is separated from catalogers, systems staff, archivists, temporary workers, vendors, grant administrators, community advisers, governance, and users."},{"archetype_element":"Structural Condition and Path Map","domain_realization":"Grant deadlines, staffing allocations, repository functions, metadata and authority standards, approval rights, procurement arrangements, inherited records, and prior backlog sequencing define unequal options and constraints."},{"archetype_element":"Agency Contribution Trace","domain_realization":"Dated approvals, prioritization choices, coalition work, metadata edits, system configurations, terminology advice, and omissions are connected to particular workflow-state transitions."},{"archetype_element":"Opportunity, Constraint, and Selection Map","domain_realization":"Formal position grants agenda and approval access; technical roles grant configuration access; funding creates temporary capacity; standards constrain descriptions; community relationships affect the legitimacy and usability of terms."},{"archetype_element":"Causal Role and Timing Classification","domain_realization":"Contributions are classified by collection and stage as initiation, enabling, coordination, execution, acceleration, blocking, amplification, legitimation, or symbolic focality."},{"archetype_element":"Evidence Provenance and Narrative Weighting Record","domain_realization":"Workflow logs, record diffs, tickets, minutes, interviews if later authorized, and public reports are distinguished by origin and coverage; managerial visibility is recorded separately from mechanism evidence."},{"archetype_element":"Actor and Structural Counterfactual Tests","domain_realization":"Bounded probes vary approval timing, qualified role occupants, temporary labor, batch editing, standards, and governance rules while retaining explicit assumptions and uncertainty."},{"archetype_element":"Cross-Scale Actor–Structure Interaction Map","domain_realization":"Individual and team actions are linked through departmental approvals and repository behavior to collection-level discovery states and the aggregated annual indicator."},{"archetype_element":"Attribution Uncertainty, Responsibility Boundary, and Review Rule","domain_realization":"Competing causal models, missing evidence, sampling sensitivity, and review triggers remain attached to the synthesis, which expressly cannot decide praise, blame, discipline, compensation, or descriptive authority."}],"title":"Collection Discovery Attribution Dossier","version":0}