{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","cell_id":"bounded_rivalry_governance__sociology_anthropology","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_id":"brg_sa_bounded_ethnographic_reanalysis_v0","proposal_index":4,"version":0,"title":"Bounded Rival Reanalysis of Ethnographic Explanations","problem":"A sociology journal and an authorized data steward invite research teams to explain why some neighborhood mutual-aid organizations persisted after emergency funding ended while others dissolved. Two commissioned synthesis articles and one follow-up research award are scarce. Teams examining the same permissioned longitudinal corpus can improve their prospects by selecting favorable episodes, ignoring negative cases, escalating analyst labor, seeking privileged contextual guidance, overstating the reach of one interpretation, lobbying reviewers, attacking rival competence, or coordinating nominally distinct submissions. Ordinary peer review compares finished manuscripts but does not govern these cross-team strategies or test whether the winning explanation survives evidence it was not tuned to fit.","actors":["Competing qualitative and mixed-method research teams","Journal editors and research-award sponsor","Authorized ethnographic data steward","Original field researchers","Independent qualitative-methods reviewers","Participant or source-community advisory panel","Research-ethics and privacy officer","Secure-analysis-environment administrator","Independent appeal and misconduct reviewer"],"observable_state":"The problem is present when submission records, secure-environment logs, team rosters, codebooks, case-selection decisions, reviewer disclosures, citation patterns, and complaints show selective use of supportive episodes, omitted contradictory cases, undisclosed analyst expansion, privileged curator contact, reciprocal reviewer nominations, coordinated submissions, personal disparagement, or claims that extend beyond the cases examined. The immediate trigger is that multiple eligible teams strategically pursue fewer publication and follow-up awards than there are defensible interpretations.","consequence":"Selection can reward rhetorical confidence, labor scale, privileged context, or evidentiary cherry-picking rather than traceable explanatory adequacy. A winning account may then be treated as the authoritative description of participants or organizations, narrowing subsequent inquiry and exporting reputational or interpretive harm to people whose records supplied the evidence.","affected_objective":"Use disciplined comparison to identify explanations that account for observed organizational persistence and failure while preserving contextual interpretation, participant protections, explicit scope limits, and the continued contestability of scholarly claims.","intervention":"Run a staged bounded reanalysis arena inside a secure environment. Before teams enter, editors, the data steward, methods reviewers, and the participant advisory panel freeze a rulebook defining the explanatory question, prizes, eligible evidence, permitted contextual consultation, prohibited conduct, disclosure duties, scoring, appeals, and noncontestable participant protections. Qualified teams receive the same corpus version, data dictionary, clarification archive, funded analyst-hour allowance, workspace period, and submission format. Each team preregisters a primary mechanism, diagnostic observations, scope conditions, case-selection rule, and treatment of disconfirming evidence before seeing a reserved set of permission-cleared cases. Teams then submit an auditable evidence map linking claims to field episodes, alternative explanations, negative cases, and uncertainty. Scoring combines evidentiary coverage, traceability, contextual coherence, treatment of contradiction, calibration of scope, reproducibility of coding operations where applicable, and performance of preregistered diagnostics on reserved cases; no single numerical proxy determines rank. The leading submissions are independently reproduced and audited before award. The portfolio may commission two individually qualified, genuinely distinct explanations when their scope conditions or unresolved anomalies are complementary, while the follow-up award targets a discriminating observation rather than declaring a cultural truth. Uniform cross-submission screens can refer shared undisclosed analysts, coordinated case omissions, reciprocal review arrangements, or cover submissions for inquiry but cannot establish guilt. Graduated sanctions cover data leakage, participant contact, reidentification attempts, fabricated evidence, concealed conflicts, retaliation, collusion, and deliberate misquotation. A predeclared portion of each honorarium is retained until correction, permitted quotation review, artifact deposit, and secure deletion duties are verified. Awards convey no exclusive claim to the corpus or automatic priority in later rounds. A post-contest review examines whether follow-up observations, corrections, participant concerns, and later reanalysis support revising the arena or abandoning competitive selection.","structural_mapping":[{"archetype_element":"Rivalry purpose statement","domain_realization":"Use rivalry to compare the evidentiary reach and limits of explanations for a bounded organizational outcome, not to select an authoritative account of a people or culture."},{"archetype_element":"Scarce prize or selection constraint","domain_realization":"Two commissioned synthesis slots and one funded follow-up investigation are available to more eligible teams than can receive them."},{"archetype_element":"Competitor eligibility boundary","domain_realization":"Teams must demonstrate relevant methodological competence, complete ethics and privacy requirements, disclose personnel and conflicts, accept secure access, and agree to participant-protection and artifact-deposit rules."},{"archetype_element":"Contest arena boundary","domain_realization":"Teams may analyze the common corpus, use the shared clarification channel, and challenge rival claims through evidence; they may not contact participants, seek private curator coaching, reidentify people, leak records, fabricate evidence, attack persons, conceal labor, or coordinate nominally independent entries."},{"archetype_element":"Performance metric and scoring basis","domain_realization":"A multi-part rubric evaluates evidence coverage, claim-to-record traceability, contextual coherence, negative cases, alternatives, scope calibration, auditable operations, and reserved-case diagnostics without reducing interpretive adequacy to one score."},{"archetype_element":"Fair process and due process layer","domain_realization":"Question, evidence boundaries, weights, reserved-case protocol, and tie-breaks are frozen in advance; reviewer conflicts are disclosed, written reasons are issued, and factual or procedural errors can be appealed independently."},{"archetype_element":"Anti-sabotage and anti-collusion guardrail","domain_realization":"Uniform screens and a graduated penalty schedule address data interference, personal retaliation, fabricated quotations, coordinated omissions, cover submissions, reciprocal reviewing, and undisclosed shared analysts."},{"archetype_element":"Externality and spillover boundary","domain_realization":"Secure access, quotation permissions, participant-advisory review, claim-scope limits, correction duties, and retained honoraria keep privacy, reputational, and interpretive harms inside the contest’s obligations."},{"archetype_element":"Escalation and arms-race damper","domain_realization":"Common corpus access, funded analyst-hour limits, roster disclosure, fixed workspace duration, and standardized submissions constrain escalation through team size, privileged access, and presentation expenditure."},{"archetype_element":"Winner power and lock-in review","domain_realization":"Awards are explicitly scope-bound, confer no corpus exclusivity or automatic future priority, preserve qualified rival interpretations, and are reopened when new evidence supplies a meaningful challenge."},{"archetype_element":"Learning and recalibration loop","domain_realization":"Follow-up observations, audit failures, corrections, participant concerns, and later reanalyses feed revisions to the question, corpus, metric, eligibility rules, or decision to discontinue the contest."}],"mechanism_mapping":[{"mechanism_slug":"contest_rulebook","role":"Freezes the explanatory target, evidence boundary, legal analytical moves, scoring, conflicts, sanctions, tie-breaks, and appeal procedures before teams adapt to them.","counterfactual_removal":"Without a binding rulebook, editors could alter standards after recognizing favored theories or teams, and private contextual access could become an unregulated winning strategy."},{"mechanism_slug":"prize_challenge","role":"Defines bounded, verifiable deliverables: an auditable explanation with preregistered diagnostics and a follow-up design capable of distinguishing among accounts.","counterfactual_removal":"Without outcome-defined awards, commissions could reward intellectual reputation or effort without requiring a traceable explanatory result."},{"mechanism_slug":"ranked_leaderboard_with_audit","role":"Produces comparable component standings while requiring independent reproduction, claim-to-evidence checks, and reserved-case validation before the leading submissions are commissioned.","counterfactual_removal":"Without auditing the leaders, teams could top the comparison through selective quotations, irreproducible coding, unsupported confidence, or tuning to the visible corpus."},{"mechanism_slug":"multiple_award_or_portfolio_selection","role":"Allows two qualified explanations to be commissioned when distinct mechanisms, scope conditions, or anomaly coverage make the set more informative than a single winner.","counterfactual_removal":"Without portfolio selection, a narrow first-place account could crowd out a differently scoped explanation needed to cover contradictory cases or heterogeneous processes."},{"mechanism_slug":"spending_cap_or_resource_cap","role":"Standardizes funded analyst hours, declared team size, secure-environment time, submission length, and access to contextual clarification.","counterfactual_removal":"Without resource constraints, large laboratories could escalate coding labor, specialist recruitment, and curator access merely to preserve rank, making bankroll and network position part of the result."},{"mechanism_slug":"anti_collusion_monitoring","role":"Applies fixed cross-submission screens to undisclosed shared labor, coordinated case omissions, reciprocal review relationships, cover entries, and suspiciously complementary claims, referring anomalies for inquiry only.","counterfactual_removal":"Without cross-entry observation, teams could divide explanatory territory, coordinate weak opposition, or present one collaboration as independent corroboration."},{"mechanism_slug":"sabotage_or_foul_penalty_schedule","role":"Attaches predeclared, graduated consequences to verified leakage, reidentification attempts, fabrication, deliberate misquotation, retaliation, collusion, and concealed conflicts.","counterfactual_removal":"Without enforceable consequences, harming rivals or compromising records could remain advantageous whenever misconduct is detected but does not affect the award."},{"mechanism_slug":"externality_bond_or_liability_rule","role":"Retains part of the honorarium until correction, authorized quotation review, artifact deposit, participant notification where required, and secure deletion are verified.","counterfactual_removal":"Without secured completion resources, a team could publish and depart while the steward, journal, or represented participants absorb correction and remediation work."},{"mechanism_slug":"challenger_access_window","role":"Permits a bounded reanalysis when specified new cases, audit failures, or follow-up observations materially challenge an awarded account and denies incumbents automatic priority.","counterfactual_removal":"Without a genuine challenge path, editorial recognition could harden a provisional explanation into a durable canon unrelated to its continuing evidentiary performance."},{"mechanism_slug":"post_contest_impact_review","role":"Compares awarded claims with follow-up evidence, corrections, participant concerns, subsequent use, and concentration of citations or corpus access before another round.","counterfactual_removal":"Without retrospective review, a rubric that selected overconfident, harmful, or poorly transportable explanations could be repeated because the original contest appeared procedurally successful."}],"causal_chain":["Scarce commissions and follow-up funding preserve a reason for eligible teams to develop and defend competing explanations.","A frozen purpose and protected domain prevent the contest from treating cultural identity, participant credibility, or lived meaning as the prize.","Equal evidence access, preregistration, roster disclosure, and resource caps reduce advantages from privileged context, post-hoc case selection, and labor escalation.","Common evidence maps make the relationship among claims, supporting episodes, contradictory cases, alternatives, and uncertainty observable to judges and rivals.","Reserved permission-cleared cases test predeclared diagnostic expectations after teams have committed to mechanisms and scope conditions.","Multi-part scoring and independent audits make evidentiary coverage, traceability, negative-case handling, and calibrated scope more reliable routes to an award than confidence or selective quotation.","Portfolio selection retains useful explanatory heterogeneity when accounts cover distinct cases or mechanisms instead of forcing a single totalizing winner.","Monitoring, inquiry, sanctions, and retained remediation resources constrain collusion, sabotage, privacy violations, and participant-facing spillovers.","Nonexclusive awards and challenger windows keep editorial victory from controlling future corpus access or permanently closing interpretation.","Follow-up evidence and post-contest review determine whether the claims, contest rules, or competitive format should be revised or retired."],"baseline":"Editors solicit or receive separate manuscripts and use ordinary peer review to choose limited publication and research opportunities. Teams often analyze different corpus subsets, receive unequal contextual guidance, disclose labor inconsistently, and encounter negative cases at different stages. Reviewers judge completed narratives without a shared preregistration point, common evidence map, reserved-case test, cross-submission abuse screen, resource boundary, or planned challenge to an awarded explanation.","nearest_rivals":["A noncompetitive plural interpretive symposium: it can preserve contextual and theoretical diversity without ranking accounts and may be preferable when interpretations are valuable but not commensurable.","Adversarial collaboration among proponents: it can produce shared tests and negotiated claims while replacing rivalry with cooperation, but agreement over questions and evidence may suppress approaches held by actors outside the collaboration.","Ordinary double-blind peer review: it is lighter-weight and protects exploratory scholarship, but it does not equalize evidence access, govern cross-team conduct, or test preregistered claims on reserved cases.","A registered-report process: it constrains post-hoc analysis and publication bias for one team or study, but does not govern rivalry, collusion, sabotage, comparative scoring, or winner lock-in among several explanations.","Independent qualitative synthesis by a neutral team: it can compare interpretations without prizes or campaigning, but concentrates framing and case-selection authority in the synthesizers.","Sequential replication or follow-up fieldwork without ranking: it can test claims directly and may be superior when the key uncertainty concerns new observations rather than strategic comparison of existing explanations."],"remaining_contrastive_claim":"The proposal’s contrast is conditional: when multiple teams strategically pursue scarce recognition and follow-up resources for comparable explanations of a bounded outcome, a governed reanalysis arena may couple winning to traceable evidence, negative-case handling, calibrated scope, and prospective diagnostics while preventing one award from closing interpretation. If the accounts are not meaningfully comparable, situated interpretation cannot survive common evidence rules, or collaboration, synthesis, registered review, or plural publication resolves the epistemic problem without rivalry, the arena is not warranted.","authority_safety":{"decision_authority":"The journal and sponsor may allocate their commissions, honoraria, and follow-up funds. The data steward may authorize only uses permitted by consent, governance agreements, and applicable rules. Editors and judges have no authority to alter participant permissions, declare a culture’s true meaning, transfer corpus ownership, or confer representational authority on a research team.","authorized_first_step":"Run the entire arena on a fictional corpus containing invented field notes, interviews, organizational timelines, contradictory episodes, ambiguous mechanisms, and planted privacy traps. Use fictional teams and permuted status cues; do not access or score any real participant record or active manuscript.","excluded_actions":["Contacting participants or organizations as part of competition without separate authorization","Using records outside their consent, governance, embargo, or cultural-access conditions","Treating people, communities, identities, credibility, or lived meanings as ranked competitors","Attempting reidentification or publishing identifying combinations","Equating a reserved-case score with cultural truth or complete ethnographic validity","Requiring disclosure of restricted records or interpretive knowledge to make an entry reproducible","Publishing misconduct allegations based only on anomaly screens","Allowing editors, original researchers, or advisory members with conflicts to judge affected entries","Granting winners exclusive corpus access, ownership of rival codebooks, or automatic future priority","Suppressing qualified dissenting interpretations because they did not receive a scarce award"],"halt_rollback":"Stop before real-data use if synthetic scoring cannot distinguish evidentiary weakness from legitimate interpretive difference, rankings change materially with team-status cues, reserved cases reward decontextualization, secure review cannot preserve participant permissions, or the advisory panel finds the question itself harmful or totalizing. Preserve ordinary submission rights, issue no finding about real teams, delete synthetic credentials and workspaces, and revert to plural peer review, registered reports, collaborative follow-up, or noncompetitive synthesis."},"negative_tests":{"strongest_counterevidence":"The strongest counterevidence would show that explanatory validity in the target inquiry depends on situated relationships and contextual knowledge that cannot be equalized or meaningfully assessed through a shared corpus; that competitive incentives reduce disclosure and interpretive generosity; or that independent plural review already surfaces negative cases without creating a canonical winner.","problem_falsifier":"The problem pattern is falsified if there is no scarce publication, funding, or status prize; teams do not strategically affect one another’s prospects; the candidate accounts answer materially different questions; consent or governance forbids comparative reanalysis; or the inquiry concerns noncontestable meaning or normative authority rather than a bounded explanatory claim.","intervention_falsifier":"Do not advance if identity-permuted entries receive unstable rankings, planted cherry-picking or fabrication survives audit, reserved-case performance conflicts systematically with contextual adequacy, portfolio scoring can rationalize any preferred pair, off-book labor defeats the resource boundary, or participant-protection review becomes a proxy for theoretical conformity. Suspend live use if awarded claims generate uncorrected misrepresentation, privacy incidents, premature canonization, reduced access for challengers, or worse treatment of contradictory evidence despite formal compliance.","risks":["Common scoring may impose false commensurability on distinct interpretive traditions.","Reserved cases may reward prediction-like simplification and penalize legitimate contextual revision.","A shared corpus may detach records from relationships and circumstances needed to interpret them.","Preregistration may discourage exploratory noticing or make early framing errors persistent.","Resource caps may favor teams with existing code, theory, personnel, or tacit familiarity.","Evidence maps and reproducibility requirements may expose restricted contextual information.","Participant-advisory review may become tokenistic or be mistaken for unanimous participant endorsement.","Multiple awards may manufacture superficial theoretical diversity while excluding unscored perspectives.","Published rankings may encourage citation concentration and premature canonization.","Collusion screens may mistake shared intellectual lineage or legitimate collaboration for coordination.","A retained honorarium may burden smaller teams or inadequately fund participant-facing remediation.","Editors may defend the contest’s winner after investing institutional prestige in the result.","Follow-up funding may privilege easily testable mechanisms over important but slower or less observable processes."]},"next_evidence_step":"Build a synthetic longitudinal corpus covering 18 fictional mutual-aid organizations, with planted support for two partially overlapping mechanisms, genuine ambiguity, negative cases, misleading vivid episodes, unequal documentation, and privacy traps. Create 8 fictional submissions exhibiting preregistered restraint, post-hoc cherry-picking, labor escalation, fabricated corroboration, legitimate contextual revision, coordinated coverage, and complementary scope conditions. Have separate methods, contextual, privacy, participant-advisory, audit, and appeal roles compare the bounded arena with ordinary blinded peer review and a noncompetitive plural symposium. Repeat after permuting team names and institutional status. Record planted-failure detection, false misconduct referrals, ranking stability, negative-case visibility, scope calibration, audit reproducibility, portfolio dependence, appeal reversals, reviewer time, and authenticity or cultural-truth judgments. Use the exercise only to decide whether a separately authorized trial on nonrestricted institutional material is warranted.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Proposal 1 governed research teams competing for field-site placements, liaison hours, and prospective participant access; its causal path bounded recruitment, gatekeeper capture, cumulative burden, and sampling distortion. This proposal begins only after a corpus exists and governs competition among explanations through common evidence, preregistration, negative cases, reserved-case diagnostics, and claim audits; it neither allocates field access nor authorizes participant contact. Proposal 2 governed repositories competing for custody, preservation funding, and control of an existing archive; its causal path tested recovery, cultural-access execution, exportability, and institutional lock-in. This proposal leaves custody with one authorized steward and allocates editorial and follow-up opportunities among analyst teams; portability and repository selection are outside its intervention. Proposal 3 governed civic associations competing for municipal liaison mandates and agenda access; its causal path addressed manufactured consensus, identity policing, organizational intimidation, and representative gatekeeper capture. This proposal confers no civic or cultural representation role and instead bounds an epistemic contest over scholarly explanations. It is independently adoptable by a journal, data steward, or research sponsor without changing fieldwork placements, archival custody, or municipal consultation channels.","revision_record":{"parent_version":null,"progress_targets_addressed":[],"conceptual_changes":[],"operational_changes":[],"evidence_changes":[],"claim_changes":[]}}