{"actors":["Seminar participants interpreting the passage","Instructor or discussion facilitator","Editor or research assistant maintaining the passage record"],"affected_objective":"Develop a text-grounded interpretation of whose perspective governs the passage without converting a plausible attribution into settled fact.","arm":"COMMON_P1","authority_safety":{"authorized_first_step":"The facilitator may create the comparison record and request a bounded check of the passage, its immediate context, and designated editions; participants may make only reversible annotations and inquiry choices from the result.","decision_authority":"The interpretive group retains authority to accept a provisional ranking or preserve a tie; an editor controls any change to published annotation, and the protocol itself cannot determine authorial intent.","excluded_actions":["Changing a published edition or canonical attribution solely from the working explanation","Grading, excluding, or discrediting a participant for advancing a live rival interpretation","Representing the selected explanation as proof of authorial intention","Silently removing rivals or disconfirming evidence from the record"],"halt_rollback":"Reopen the ranking and withdraw any dependent annotation if the discriminating check contradicts the selected explanation, reveals an edition-dependent wording difference, or leaves the leading candidates tied."},"baseline":"Unstructured discussion in which participants move directly from striking diction to an attribution such as narrator, character, or author, with evidence, inference, rivals, and confidence mixed together in notes or conversation.","candidate_id":"abductive_explanation_selection__literature_literary_theory__COMMON_P1","causal_chain":["A passage presents an anomalous cluster of deictic, evaluative, and tense shifts without an explicit speech marker.","Readers separately record the exact textual marks and the bounded question of whose perspective organizes them.","They generate a candidate set containing free indirect discourse, overt narratorial irony, a focalization transition, and edition or transcription corruption.","A passage docket compares each candidate for coverage of the local marks, continuity with adjacent passages, narratological mechanism, timing, required ad hoc exceptions, and predicted traces elsewhere.","If one candidate accounts for the cluster more coherently than its nearest rival, it becomes an action-limited working reading; otherwise the docket records an unresolved tie.","The working reading guides a reversible annotation and a bounded probe of nearby attribution cues and edition variants.","Contradictory cues, variant wording, or failure of predicted recurrence triggers reopening, reranking, or splitting the explanation.","Interpretive claims remain linked to cited marks, visible rivals, and an explicit confidence boundary."],"cell_id":"abductive_explanation_selection__literature_literary_theory","consequence":"A vivid first interpretation can harden into an unsupported speaker or perspective attribution, causing later claims about irony, character, and theme to inherit an unexamined premise and making revision difficult when conflicting textual evidence appears.","diversity_from_prior_proposals":"No prior proposals or experiment cells were inspected; this candidate is differentiated only against the named domain rivals by centering a passage-level, defeasible perspective-attribution docket.","experiment_id":"eoa_inverse_innovation_exp13_second_slot_policy60_20260806","intervention":"Introduce a Passage Attribution Docket for anomalous perspective shifts. It fixes the explanandum, stores transcription and location separately from interpretation, compares a bounded set of explanations in an explicit fit matrix, records either a best-so-far reading or unresolved tie, labels its confidence and permitted uses, and specifies one discriminating textual or edition-based probe plus revision triggers.","mechanism_mapping":[{"counterfactual_removal":"Without the matrix, preferences about narrators or favored theories can remain hidden and no reviewable reason explains why one attribution outranks another.","mechanism_slug":"inference_to_best_explanation_matrix","role":"Compares perspective-attribution candidates against shared criteria: textual coverage, contextual continuity, mechanism plausibility, ad hoc burden, and discriminating predictions."},{"counterfactual_removal":"Without the memo, later participants may inherit the selected interpretation while losing the original observation, surviving rivals, and confidence limits.","mechanism_slug":"explanatory_case_memo","role":"Preserves the passage, observed marks, explanandum, candidate ranking, selection rationale, defeasibility label, and action boundary for handoff."},{"counterfactual_removal":"Without a disconfirming probe, the favored reading can absorb every later detail and become effectively unrevisable.","mechanism_slug":"disconfirming_probe_plan","role":"Identifies nearby cues or edition variants expected under the leading account and specifies observations that would weaken it or favor a rival."},{"counterfactual_removal":"Without an evolving log, changes in interpretation can be presented as if they were always obvious, obscuring which new evidence caused revision.","mechanism_slug":"abduction_log","role":"Tracks reranking when contextual, bibliographic, or comparative textual evidence is added."}],"nearest_rivals":["Ordinary close reading that binds claims to local textual detail but does not require an explicit candidate set, ranked rival comparison, or revision trigger","A narratological classification checklist that assigns a stable category from recognized textual cues without treating the attribution as a puzzling, defeasible explanation","Interpretive pluralism that preserves several readings without requiring a best-so-far ranking when one materially outperforms the others","Authorial-intent research that seeks external historical evidence before forming a passage-level working explanation"],"negative_tests":{"intervention_falsifier":"Across blinded passage reviews, the docket fails if reviewers using it cannot state the observation separately from interpretation, omit a credible rival, or produce probes that would change their ranking more often than reviewers using the baseline notes.","problem_falsifier":"The proposed problem is absent for the focal passage if independent readers already agree on the bounded textual observation and attribution through an explicit, stable rule, and no plausible rival would alter the permitted interpretive action.","risks":["Formal scoring may create false precision or privilege criteria favored by the facilitator.","A bounded candidate set may exclude an explanation grounded in historical language, genre convention, translation, or material bibliography.","The working label may still be quoted downstream without its confidence and action limits.","Checking only immediate context may overvalue local coherence and miss book-scale patterns.","Edition comparison may introduce new ambiguity rather than resolve the attribution."],"strongest_counterevidence":"If passages with ambiguous perspective are handled just as transparently and revisably through ordinary close reading—while preserving observations, rivals, confidence limits, and discriminating checks—the added docket may be unnecessary administrative structure."},"next_evidence_step":"Select one public-domain passage containing an unmarked perspective shift. Give the same passage and immediate context to two small reader groups: one uses ordinary notes and one uses a blank Passage Attribution Docket. Without judging which interpretation is correct, compare whether each record separates observation from inference, includes the four prespecified candidate types, names its nearest rival, states an action limit, and supplies a probe whose possible outcomes would rerank candidates. Stop after this single passage and retain all records for review.","observable_state":"In a specified passage, evaluative diction, temporal reference, or deictic language shifts abruptly without quotation marks or an explicit reporting clause; the record contains the exact words, punctuation, line or page location, and immediate context, while claims such as free indirect discourse or narratorial irony are stored only as candidate interpretations.","prior_art_status":"UNSEARCHED","problem":"When a literary passage shifts diction, deixis, or tense without an explicit speaker marker, readers may need a provisional account of whether the language belongs to a character's represented thought, an ironic narrator, a focalization transition, or a textual variant. Discussion often needs to proceed before the attribution is demonstrable, but selecting the first compelling reading can conceal rivals and make downstream interpretation brittle.","proposal_index":1,"remaining_contrastive_claim":"Unlike close reading alone, the docket's defining claim is that an anomalous textual pattern should guide interpretation only after explicit rival explanations are compared and the selected attribution carries a disconfirming probe, revision trigger, and action limit.","revision_record":{"claim_changes":["Initial candidate; no prior claim was revised."],"conceptual_changes":["Initial mapping of abductive explanation selection to ambiguous perspective attribution in literary passages."],"evidence_changes":["No external evidence or prior-art material was consulted."],"operational_changes":["Initial specification of a one-passage comparison using a blank docket and ordinary-note baseline."],"parent_version":null,"progress_targets_addressed":["Concrete explanandum and actors","Observable state separated from interpretation","Bounded rivals and explanatory-fit criteria","Defeasibility, authority limits, and rollback triggers","Problem and intervention falsifiers","Bounded first evidence step"]},"schema_version":1,"structural_mapping":[{"archetype_element":"Surprising observation and explanandum boundary","domain_realization":"An unmarked shift in evaluative diction, deixis, or tense in one located passage, bounded to the question of perspective attribution."},{"archetype_element":"Observation–interpretation separation","domain_realization":"Verbatim words, punctuation, location, and context occupy an evidence field; labels such as free indirect discourse or irony occupy separate hypothesis fields."},{"archetype_element":"Candidate explanation set","domain_realization":"Free indirect discourse, overt narratorial irony, focalization transition, and edition or transcription corruption remain visible candidates."},{"archetype_element":"Explanatory fit comparison","domain_realization":"Candidates are compared for local textual coverage, continuity with surrounding passages, narratological mechanism, ad hoc burden, and predicted additional cues."},{"archetype_element":"Best-so-far explanation or plurality state","domain_realization":"The docket records either a materially better perspective attribution or an unresolved ranked tie rather than forcing closure."},{"archetype_element":"Defeasibility status","domain_realization":"The selected reading is labeled working and action-limited, permitting discussion and reversible annotation but not a definitive claim of authorial intention."},{"archetype_element":"Discriminating evidence plan","domain_realization":"Readers inspect a bounded surrounding context and designated edition variants for attribution cues predicted differently by the live explanations."},{"archetype_element":"Revision trigger and action–confidence boundary","domain_realization":"Contradictory cues, wording variants, or a surviving tie reopen the docket and withdraw dependent annotation while leaving exploratory discussion permitted."}],"title":"Passage Attribution Docket for Ambiguous Narrative Perspective","version":0}