{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"computability_boundary_mapping__philosophy","opaque_id":"computability_boundary_mapping__philosophy__A","search_lanes":{"direct_problem":{"queries":["philosophical argument platform automated entailment verdict formal logic UNKNOWN timeout","argument mapping software formal logic validity automated evaluation philosophy platform","unrestricted first order logic validity undecidable Church theorem entailment","computational argumentation platform undecidable reasoning explicit unknown status argument evaluation"],"source_ids":["S1","S6","S7","S8"],"no_result_note":"The theoretical impossibility and relevant platforms were found, but no retained source documents a deployed philosophical platform that both admits an unrestricted formal language and promises a terminating binary entailment verdict for every input."},"closest_prior_art":{"queries":["theorem prover unknown status timeout satisfiability standard SZS ontology","site:smt-lib.org standard unknown reason-unknown solver response sat unsat unknown","site:w3.org OWL 2 Full undecidable OWL 2 DL decidable conformance unknown","formal argumentation systems complexity undecidable infinite frameworks decision problems"],"source_ids":["S2","S3","S4","S5","S6","S7"],"no_result_note":null},"historical_terminology":{"queries":["Entscheidungsproblem philosophical logic universal decision procedure Church Turing","site:plato.stanford.edu Entscheidungsproblem Church Turing first order logic undecidable","Entscheidbarkeit philosophische Argumente Plattform automatische Gültigkeitsprüfung unentscheidbar"],"source_ids":["S1","S8"],"no_result_note":null},"products_practices_standards":{"queries":["TPTP SZS ontology status theorem unknown gaveup timeout official","site:smt-lib.org language standard 2.7 responses unknown get-info reason-unknown pdf","site:w3.org/TR/owl2-conformance consistency checker Unknown OWL 2 Full undecidable complete terminating","Carneades argument evaluation proof standard software philosophical arguments"],"source_ids":["S2","S3","S4","S5","S6","S7"],"no_result_note":null},"non_english_regional":{"queries":["Entscheidungsproblem philosophische Argumente Plattform automatische Gültigkeitsprüfung unentscheidbar","plateforme argumentation philosophique vérification logique indécidable résultat inconnu"],"source_ids":["S8"],"no_result_note":"French and German terminology confirmed the established distinction among decidability, undecidability, and indeterminate status, but no regional philosophical platform matching the proposed universal-binary failure mode was located."},"composition_subproblems":{"queries":["\"undecidable\" entailment \"decidable fragment\" \"Unknown\" proof certificate","theorem prover timeout must not mean not entailed explicit unknown certificate","\"proof certificate\" \"reason-unknown\" decidable fragment entailment checker","site:w3.org/TR/owl2-profiles decidability restriction profiles official recommendation"],"source_ids":["S2","S3","S4","S5","S6"],"no_result_note":null}},"sources":[{"source_id":"S1","title":"The Church-Turing Thesis","url":"https://plato.stanford.edu/archives/fall2024/entries/church-turing/","publisher":"Stanford Encyclopedia of Philosophy, Stanford University","date_or_year":"2024","source_type":"SECONDARY_RESEARCH","language":"English","claims_supported":["The Entscheidungsproblem asks for an effective method deciding every formula in a logical calculus.","Church and Turing independently established that first-order predicate logic has no such decision method.","Decidable propositional and monadic fragments were known before the unrestricted negative result."]},{"source_id":"S2","title":"OWL 2 Web Ontology Language Conformance (Second Edition)","url":"https://www.w3.org/TR/owl2-test/","publisher":"World Wide Web Consortium (W3C)","date_or_year":"2012-12-11","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["A standardized entailment-checker interface distinguishes True, False, Unknown, and Error.","True and False are constrained by soundness conditions; Unknown means the checker cannot determine entailment.","The standard distinguishes termination from completeness and warns when Unknown can make query answers incomplete."]},{"source_id":"S3","title":"OWL 2 Web Ontology Language Profiles (Second Edition)","url":"https://www.w3.org/TR/owl2-profiles/","publisher":"World Wide Web Consortium (W3C)","date_or_year":"2012-12-11","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["OWL profiles are mechanically specified syntactic restrictions that trade expressiveness for reasoning guarantees.","The unrestricted RDF-based reasoning problems listed are undecidable, while restricted EL, QL, and RL profiles have stated complexity bounds.","The document separates decidability from computational complexity."]},{"source_id":"S4","title":"The SMT-LIB Standard, Version 2.7","url":"https://smt-lib.org/papers/smt-lib-reference-v2.7-r2025-07-07.pdf","publisher":"SMT-LIB Initiative","date_or_year":"2025-07-07","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["The standard check response is explicitly ternary: sat, unsat, or unknown.","Unknown means the search was inconclusive because of resource limits, incompleteness, or other reasons.","The optional reason-unknown response can identify memory exhaustion or incompleteness for the submitted formula class."]},{"source_id":"S5","title":"SZS Ontology","url":"https://tptp.org/UserDocs/SZSOntology/","publisher":"TPTP World, University of Miami","date_or_year":"undated (accessed 2026-08-04)","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["Automated theorem-proving practice distinguishes semantic success from no-success.","No-success statuses separately encode Unknown, Timeout, MemoryOut, Incomplete, Inappropriate, input errors, and unverified outputs.","Proofs, refutations, and models can be attached as standardized justification data."]},{"source_id":"S6","title":"Oak","url":"https://oakproof.org/","publisher":"Oak project","date_or_year":"undated (accessed 2026-08-04)","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["A first-party proof checker explicitly targets statements from philosophy as well as mathematics and theology.","Oak translates proof steps into first-order logic and delegates validity checks to an automated theorem prover.","If a step is invalid or the prover exceeds a work bound, Oak stops and reports a problem rather than claiming to construct a proof."]},{"source_id":"S7","title":"Carneades Argumentation System","url":"https://carneades.github.io/carneades/","publisher":"Carneades project","date_or_year":"undated (accessed 2026-08-04)","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["A concrete open-source argument-evaluation system and multi-user web application exists.","Its formal model covers argument graphs, practical reasoning, case-based reasoning, and multi-criteria decision analysis.","It supplies an identifiable class of platform maintainers and argument-system operators, although it does not document the proposed universal entailment guarantee."]},{"source_id":"S8","title":"Indécidabilité","url":"https://www.larousse.fr/encyclopedie/philosophie/ind%C3%A9cidabilit%C3%A9/191752","publisher":"Larousse, Dictionnaire de la philosophie","date_or_year":"undated (accessed 2026-08-04)","source_type":"OTHER","language":"French","claims_supported":["French philosophical terminology distinguishes theory-level undecidability from whether an individual proposition and its negation are derivable.","It identifies propositional logic as decidable and first-order predicate logic as undecidable following Church and Turing in 1936.","It cautions that decidability is relative to the selected theory and language."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The computability premise is well supported: unrestricted first-order entailment cannot have a total correct decision procedure, while real philosophy-facing and argument-evaluation tools invoke formal proof checking. However, the searched evidence does not establish the proposal's crucial operational premise that an actual philosophical platform simultaneously admits an unrestricted language and forces every timeout, failed search, or unsupported input into ENTAILED or NOT_ENTAILED. Oak instead stops when its prover exceeds a work bound, and the major reasoning standards preserve inconclusive statuses.","source_ids":["S1","S2","S4","S5","S6","S7","S8"],"uncertainty":"The exact grammar, semantics, output contract, timeout handling, and user interpretation of any intended target platform were not identified."},"adopter_evidence":{"status":"PARTLY_SUPPORTED","finding":"Concrete adopter classes exist: maintainers of Oak-like philosophy proof checkers and Carneades-like argument platforms, plus formal-logic tool operators implementing OWL, SMT-LIB, or TPTP conventions. Platform maintainers and content or logic leads are therefore identifiable roles, but no organization was found acknowledging the specific forced-binary defect or committing to the proposed shadow test.","source_ids":["S2","S4","S5","S6","S7"],"uncertainty":"A named target platform and its actual decision authority remain to be confirmed before intervention."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"Most technical components are established outside the proposed context: syntactically enforced profiles, sound True/False conditions, explicit Unknown, standardized timeout and incompleteness reasons, and proof/model artifacts. Oak also demonstrates bounded first-order proof checking for philosophy-related statements. What remains untested is their integrated routing, guarantee-label comprehension, coverage, and error-reduction effect on archived philosophical arguments.","source_ids":["S2","S3","S4","S5","S6"],"uncertainty":"No retained source evaluates the complete package on a philosophical argument platform or shows that natural-language formalization errors are subordinate to computability errors."},"prior_art":{"disposition":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"OWL 2 entailment-checker conformance plus syntactic profiles","source_ids":["S2","S3"],"same_problem":true,"same_causal_lever":true,"overlap":"This standard package already separates True, False, Unknown, and Error; imposes soundness conditions; distinguishes termination from completeness; and defines enforceable restricted profiles with known computational properties.","remaining_difference":"It concerns ontology entailment rather than philosophical argument adjudication and does not test whether platform users understand guarantee labels or whether routing preserves useful philosophical coverage."},{"name":"SMT-LIB and TPTP/SZS result contracts","source_ids":["S4","S5"],"same_problem":true,"same_causal_lever":true,"overlap":"Both operationalize non-binary outcomes for automated reasoning, distinguish inconclusive search from negative results, record reasons such as timeout or incompleteness, and support proof or model artifacts.","remaining_difference":"They are solver-facing interoperability standards, not a philosophy-platform workflow combining formalization review, fragment classification, authorizer governance, and user-facing empirical evaluation."},{"name":"Oak philosophy-facing first-order proof checker","source_ids":["S6"],"same_problem":true,"same_causal_lever":true,"overlap":"Oak applies automated first-order validity checking to statements that may be philosophical and stops when a resource/work bound is exceeded instead of constructing a universal proof procedure.","remaining_difference":"Its public description does not expose a computability-status lattice, prove supported-fragment deciders, distinguish invalidity from resource exhaustion with standardized guarantee labels, or report a comparative shadow test."},{"name":"Carneades Argumentation System","source_ids":["S7"],"same_problem":false,"same_causal_lever":false,"overlap":"It supplies the nearby substrate: a deployed formal model and software for structured argument evaluation, including practical reasoning and argument graphs.","remaining_difference":"Its evaluation semantics are not presented as unrestricted classical entailment, and the retained page does not document universal binary entailment, undecidability routing, proof certificates, or explicit timeout/unknown policy."}],"contrastive_claim_remaining":"For a named philosophical argument platform whose admitted language is first shown to contain an undecidable entailment class and whose baseline is shown to coerce unresolved cases into binary verdicts, mechanically enforced fragment routing plus certificate-backed positive results and explicit guarantee-labeled UNKNOWN will reduce unsound or forced-negative verdicts on a preregistered archived corpus without reducing usable definitive coverage below a declared threshold or causing users to confuse entailment with philosophical truth.","contrastive_claim_falsifier":"The claim is falsified if the platform's full admitted language has a verified total decider, if baseline logs show no unresolved-to-binary coercion, or if the shadow test yields any unsound definitive verdict, no reduction in forced-binary errors, unacceptable coverage loss, or persistent user confusion about UNKNOWN and philosophical truth.","confidence":"HIGH","search_limitations":"This was a bounded eight-source search across six lanes, not an exhaustive literature, code, patent, procurement, or market review. No target platform's internal logs, full source code, contractual guarantee, or user research were available. French and German terminology were searched, but only one non-English source was retained. The evidence cannot establish world novelty, patentability, freedom to operate, market size, or realized impact."},"researchability_gates":{"externally_supported_problem":{"status":"INDETERMINATE","rationale":"Undecidability of unrestricted entailment and the relevance of automated checking to philosophy are externally supported, but no source verifies the required conjunction of unrestricted admission, guaranteed termination, binary-only output, and actual coercion of unresolved cases on a target philosophical platform.","source_ids":["S1","S6","S7","S8"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Existing philosophy-facing proof and argument systems make platform maintainers and formal-logic operators identifiable adopter classes; a target can be selected and its content owner and logic lead named before testing.","source_ids":["S6","S7"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Although the core solver mechanisms are established, a distinct empirical transfer claim remains: whether the integrated router reduces forced-binary errors while preserving coverage and label comprehension in philosophical argument assessment.","source_ids":["S2","S3","S4","S5","S6"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A read-only audit and shadow test on 120 stratified archived arguments can verify the language boundary, baseline coercion rate, certificate soundness, routing behavior, coverage, and label comprehension without publishing verdicts.","source_ids":["S2","S3","S4","S5"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first step is read-only, keeps outputs out of grading and publication, preserves manual review, and has explicit halt conditions. Existing standards provide conservative status semantics suitable for the shadow test. Formalization mismatch remains a monitored risk, not a reason to prohibit the bounded test.","source_ids":["S2","S4","S5","S6"]},"adequate_search_evidence":{"status":"PASS","rationale":"All six required lanes were searched adversarially using historical, product, standards, non-English, and component-combination terminology. Exactly eight retained sources were opened; they span six publishers and include five official-standard or first-party sources.","source_ids":["S1","S2","S3","S4","S5","S6","S7","S8"]}},"strict_success":false,"screen_survival":false,"remaining_research_value":"MODERATE","recommended_next_step":"Select one named platform and first freeze its admitted grammar, semantics, and output contract; then run the authorized 120-case read-only shadow test only if logs confirm both an undecidable admitted class and unresolved-to-binary coercion, with preregistered soundness, error-reduction, coverage, and comprehension thresholds.","world_novelty_boundary":"The bounded search shows that undecidability-aware fragment restriction, explicit UNKNOWN, reason-coded inconclusive outcomes, and proof/model artifacts are established in automated-reasoning standards and products. It did not find the complete package evaluated on a philosophical argument platform, but that absence is not evidence of world novelty and makes no claim about patents, freedom to operate, markets, or impact."}