{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"computability_boundary_mapping__rhetoric","trajectory_id":"R","attempt_index":0,"archetype_slug":"computability_boundary_mapping","domain_slug":"rhetoric","decision":"CANDIDATE","problem_id":"universal_persuasive_dialogue_harm_reachability","causal_lever_id":"formal_rhetoric_language_scope_and_verdict_contract","proposal":{"problem":"A rhetoric platform promises an exact, terminating analyzer that decides whether every arbitrary executable persuasive-dialogue campaign can ever reach a formally prohibited rhetorical state, despite allowing unbounded scripts, audience-model updates, and interaction histories.","actors_substrate":["rhetoric-platform designers","campaign authors","moderation analysts","modeled and real audiences","executable dialogue scripts and audience-state models"],"observable_state":"The analyzer returns Boolean safe/unsafe verdicts for unrestricted campaigns; timeouts are treated as safe, tested campaigns substitute for class-wide proof, and the shipped claim does not distinguish the formal audience model from real persuasion.","consequence":"Unsafe campaigns may be cleared, benign campaigns blocked, and engineering effort repeatedly spent pursuing an impossible unrestricted decider.","affected_objective":"Provide auditable moderation guarantees without misrepresenting what automated rhetorical analysis can decide.","structural_mapping":[{"archetype_element":"open-ended problem class","domain_realization":"All executable persuasive-dialogue campaigns with unbounded state and interaction are admitted.","claim_kind":"HYPOTHESIS"},{"archetype_element":"semantic property","domain_realization":"Whether some execution reaches the platform's formally defined prohibited rhetorical state.","claim_kind":"HYPOTHESIS"},{"archetype_element":"universal exact terminating requirement","domain_realization":"The product promises safe/unsafe for every admitted campaign.","claim_kind":"HYPOTHESIS"},{"archetype_element":"impossibility boundary","domain_realization":"If campaign scripts can simulate arbitrary computation, prohibited-state reachability can encode halting.","claim_kind":"INFERENCE"},{"archetype_element":"decidable region","domain_realization":"Finite-state, bounded-history, mechanically enforceable campaign fragments admit exhaustive checking.","claim_kind":"INFERENCE"},{"archetype_element":"honest fallback","domain_realization":"Route campaigns to exact finite-state checking, bounded witness search, or UNKNOWN/out-of-scope review with guarantee labels.","claim_kind":"HYPOTHESIS"}],"component_map":[{"component":"Problem-Class Specification","status":"adapted","domain_realization":"Define campaigns, audience models, executions, and prohibited states."},{"component":"Instance Representation Contract","status":"adapted","domain_realization":"Version the campaign grammar, state encoding, and history bounds."},{"component":"Computation Model Contract","status":"adapted","domain_realization":"Declare script operations, memory, nondeterminism, and external inputs."},{"component":"Solvability Guarantee Profile","status":"direct","domain_realization":"Separate exact decision, violation recognition, bounded checking, and UNKNOWN."},{"component":"Quantifier and Scope Map","status":"direct","domain_realization":"Distinguish every campaign/every execution from one campaign/some execution."},{"component":"Computability Status Lattice","status":"direct","domain_realization":"Classify fragments as decidable, recognizable, relative, or unresolved."},{"component":"Constructive Procedure Witness","status":"direct","domain_realization":"Supply a terminating checker for the finite fragment."},{"component":"Reduction Preservation Contract","status":"direct","domain_realization":"Map machine halting to prohibited-state reachability in both answer directions."},{"component":"Computability Impossibility Certificate","status":"direct","domain_realization":"Retain the checked reduction for unrestricted scripts."},{"component":"Assumption Register","status":"direct","domain_realization":"Record expressiveness, formal semantics, and environment assumptions."},{"component":"Decidable Subclass Map","status":"direct","domain_realization":"List enforceable finite-state and bounded-history fragments."},{"component":"One-Sided Recognition Contract","status":"direct","domain_realization":"A witnessed harmful trace supports UNSAFE; silence never supports SAFE."},{"component":"Unknown and Nontermination Policy","status":"direct","domain_realization":"Keep UNKNOWN, timeout, out-of-scope, and SAFE distinct."},{"component":"Fallback Solution Contract","status":"direct","domain_realization":"Label each routed result with its actual scope and strength."},{"component":"Computability Guarantee Record","status":"direct","domain_realization":"Version-link grammar, model, proof, checker, and public claim."},{"component":"Recheck Trigger","status":"direct","domain_realization":"Reclassify after grammar, audience model, or external-service changes."},{"component":"Termination Condition","status":"direct","domain_realization":"Finite checking terminates; bounded search stops at a declared budget."},{"component":"Scope Boundary","status":"direct","domain_realization":"Mechanically reject or quarantine unrestricted constructs."},{"component":"Decision Record","status":"direct","domain_realization":"Record why each shipped mode and guarantee was selected."},{"component":"Uncertainty Residue","status":"direct","domain_realization":"Expose unproved obligations and model-to-reality mismatch."},{"component":"Independent Proof Review","status":"direct","domain_realization":"Have an independent reviewer check reduction and constructive proof."},{"component":"Complexity Follow-On Gate","status":"direct","domain_realization":"Estimate cost only after a fragment is shown decidable."}],"mechanism_dispositions":[{"slug":"abstract_interpretation_or_model_checking","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Exactly check a finite over-approximation; SAFE is conditional on sound abstraction.","counterfactual_removal":"No scalable sound exact mode for modeled finite behavior."},{"slug":"bounded_domain_exhaustive_search","disposition":"selected_supporting","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Exhaust campaigns and traces only within explicit bounds.","counterfactual_removal":"Pilot loses complete within-bound evidence."},{"slug":"computability_boundary_decision_record","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Version boundary, guarantee, and triggers.","counterfactual_removal":"Guarantees can drift silently."},{"slug":"computational_complexity_analysis","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Assess feasible bounds after decidability.","counterfactual_removal":"Correct fragments may still be impractical."},{"slug":"constructive_algorithm_and_correctness_proof","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Prove finite-fragment checker total and correct.","counterfactual_removal":"Decidability claim lacks a witness."},{"slug":"diagonalization_impossibility_proof","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"A reduction is clearer than a bespoke self-reference proof.","counterfactual_removal":"No material change."},{"slug":"enumeration_and_dovetailing","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Bounded trace search supplies a more operable recognizer.","counterfactual_removal":"No material change."},{"slug":"fallback_mode_router","disposition":"selected_load_bearing","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Route by enforceable class and label guarantees.","counterfactual_removal":"Weaker outputs can be mistaken for universal verdicts."},{"slug":"halting_problem_reduction","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Encode halting as reachability for unrestricted scripts.","counterfactual_removal":"No principled basis to reject the universal decider."},{"slug":"language_fragment_restriction","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Enforce a finite-state campaign grammar.","counterfactual_removal":"No decidable production region."},{"slug":"many_one_reduction_proof","disposition":"selected_supporting","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Formalize the halting-to-reachability map and preservation.","counterfactual_removal":"The impossibility transfer becomes rhetorical analogy."},{"slug":"promise_problem_restriction","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Unenforced promises permit authoritative behavior outside scope; syntax is enforceable.","counterfactual_removal":"No material change."},{"slug":"proof_by_counterexample","disposition":"selected_supporting","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Refute overbroad safety claims with an in-scope missed trace.","counterfactual_removal":"Cheap claim-breaking test is lost."},{"slug":"proof_checking","disposition":"selected_supporting","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Independently verify proofs and stated theorem.","counterfactual_removal":"Boundary rests on author authority."},{"slug":"reduction_direction_checklist","disposition":"selected_supporting","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Gate source-to-target direction and assumptions.","counterfactual_removal":"A reversed reduction may be accepted."},{"slug":"semi_decision_with_explicit_unknown","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Return witnessed UNSAFE or budgeted UNKNOWN, never timeout-as-SAFE.","counterfactual_removal":"Operational pressure recreates false Boolean verdicts."},{"slug":"theorem_prover_guided_search","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Optional tooling adds no required causal role in the first test.","counterfactual_removal":"No material change."},{"slug":"turing_reduction_analysis","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Relative oracle degrees do not answer the shipped boundary question.","counterfactual_removal":"No material change."}],"causal_chain":["Specify formal rhetorical reachability and computation model.","Prove unrestricted undecidability by a checked source-to-target reduction.","Enforce a finite campaign fragment admitting a total checker.","Route other inputs to bounded witness search or explicit UNKNOWN.","Attach scope and guarantee labels so downstream users cannot read UNKNOWN as SAFE.","Recheck when expressiveness or models change."],"baseline":"Continue unrestricted Boolean analysis, treat timeout as safe, and rely on campaign benchmarks.","nearest_rival":"Use a heuristic harmful-rhetoric classifier with confidence scores but no computability classification or enforceable scope.","authority_safety":{"affected_parties":["campaign audiences","campaign authors","moderation analysts","people represented in audience models"],"decision_authority":"Platform safety owner may authorize only offline validation; governance and affected-party representatives approve deployment claims.","authorized_first_step":"Offline pilot on a synthetic finite-state campaign language: prove checker obligations, exhaust small instances, seed known harmful traces, and verify SAFE/UNSAFE/UNKNOWN labels without exposing real audiences.","excluded_actions":["live persuasion experiments","automatic clearance of unrestricted campaigns","treating UNKNOWN as SAFE","claiming conclusions about real human response from formal-model reachability"],"halt_rollback":"Halt if any known harmful bounded trace is cleared, fragment membership is bypassable, the abstraction drops modeled behavior, or labels collapse; withdraw the guarantee and revert to manual quarantine."}},"negative_tests":{"strongest_counterevidence":"The deployed campaign language may already be finite-state and bounded; then exact reachability is decidable and the real issue is complexity, not computability.","analogy_break":"Formal reachability concerns a stipulated audience model, while actual persuasion is contextual and human; a theorem about the model cannot establish real-world rhetorical harm or safety.","failure_condition":"The restriction removes necessary rhetorical interactions, membership is unenforceable, or UNKNOWN rates make the fallback unusable.","problem_falsifier":"Show that all admitted campaigns and histories are effectively finite under enforced bounds and that the requested guarantee concerns only this class.","intervention_falsifier":"Within the declared fragment, find a well-formed campaign for which the checker fails to terminate or returns an incorrect verdict, or show the router mislabels an out-of-scope result as exact.","risks":["Formal-model safety may be mistaken for audience safety.","Restrictive syntax may displace authors to ungoverned channels.","False alarms may burden speakers or suppress legitimate advocacy.","A flawed reduction or abstraction may institutionalize a wrong boundary."]},"null_rationale":null,"classification":{"candidate_kind":"MECHANISM_COMPOSITION","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.84,"generator_notes":"Closed-book structural transfer. The candidate is conditional on the rhetoric platform using an executable language expressive enough to simulate arbitrary computation; that premise requires local verification."}