docs(taxonomy): #498 PR-9 — Narrative ambiguity / deception cluster (6 leaves, completes Ambiguïté subfamily) - #714
Merged
jsboige merged 1 commit intoJul 5, 2026
Conversation
…leaves, completes subfamily) Ninth cluster of the #498 AIF chantier. Completes the Ambiguïté subfamily (3/3 sub-subs). Second half of the Ambiguïté narrative sub-sub: the deferred mechanisms from PR-8 — source-deception (881-883), interrogation (884-885), and polytélie (886). In-sub-sub anchor pk 876 (ConflictingGoals_Conflict). Source-deception and interrogation leaves honestly diverge from the anchor's ConflictingGoals_Conflict to epistemic/action schemes that fit their semantics: - Source-deception (881 grey, 883 black): ExpertOpinion_Inference + ExpertiseInconsistency_Conflict. Omission (concealment, narrow) vs commission (falsification, close) distinguished by MappingType — graded source-deception. - Whisper campaign (882): PositionToKnow_Inference + LackOfCompleteKnowledge_ Conflict (cross-PR-8 reuse of 880 epistemic pair, organized-campaign flavor). - Clandestine interrogation (884): PracticalReasoning_Inference + ConflictingGoals_Conflict (anchor reuse — friendly surface + extraction goal). - Preach the false (885): Bias_Inference direct-conflict (scheme as defect — evidence biased by false-premise collection, degenerate undercut like PR-4 800). - Polytélie (886): ConflictingGoals_Conflict direct reuse (literal multiple goals). 6 leaves, 6 fully-modeled, 0 FAIL-LOUD (2nd consecutive clean cluster). 4 new chantier native tokens: ExpertOpinion_Inference, ExpertiseInconsistency_Conflict, PracticalReasoning_Inference, Bias_Inference (all native 1x/1x/2x/2x). Palette now 21 distinct native tokens across 9 PR. §6 completes Ambiguïté subfamily 3/3 (18/18 leaves). Second fully-complete subfamily after Fallacious comparison. Cumulative 70→100 strict / 101 inclusive (+30). §6 dual criterion re po-2024 audit #709 (this PR's +6 criterion-independent). §7 I/RA/CA: all 6 undercuts (deception defeats the inference audience/target is manipulated to draw). Decomposition recorded, not serialized (in-sub-sub anchor = independent of I/RA/CA schema decision #707 §4). PROPOSITION ONLY — GATED, 0 prod CSV write, no DB write, no OWL regen. Native vocabulary validated programmatically — PASS, 0 fabrication. Relates: dispatch ih617l (tertiary), #498, PR-1 #699, PR-2 #701, PR-3 #703, PR-4 #705, PR-6 #708, PR-7 #711, PR-8 #713, coverage-status #707, verification-audit #709, family-gap #712, #133/#130 (OWL), #499, #677 (0 fabrication), #192, #458. Co-Authored-By: Claude-Code <noreply@anthropic.com>
This was referenced Jul 6, 2026
jsboige
added a commit
that referenced
this pull request
Jul 6, 2026
… second i18n gap (dispatch ynv05a) (#716) * docs(taxonomy): #498 verification-audit addendum PR-7/8/9 + chantier count reconciliation Independent code=truth verification of the 3 Ambiguïté-family cluster propositions produced after the original audit (PR-1/2/3/4/6). Dispatch ynv05a PRIMARY + SECONDARY (ai-01 2026-07-06 01:32), base master bdba45d. Read-only: 0 write Cards/, audit doc only. PR-7 #711 Amphibologie (8 rows), PR-8 #713 Narrative ambiguity/insinuation (4 leaves), PR-9 #714 Narrative ambiguity/deception (6 leaves) — 18 leaves + 2 anchors (846 borrow-root, 876 in-sub-sub). Same adversarial method as the original audit: deterministic read-only extractor (498_audit_extract_pr789.py, scratchpad), no LLM judgment on the data. Verdict (code=truth on bdba45d): - 20/20 cited pks present; 2/2 anchors (846, 876) byte-exact (text/DirectRef/ExceptionRef/MappingType). - 15/15 distinct proposed tokens native (0 fabrication, #677). Native inventory re-derived: 54 tokens (35 Inference + 19 Conflict), holds vs original audit. - 18/18 cited desc_fr byte-close (2 cosmetic diffs: apostrophe + 878 casing). - 3/3 FAIL-LOUD legit (PR-7 847/848/853 syntactic-ambiguity CA-missing — hypothetical tokens verified absent from inventory). Fourth recurring AIF ontology gap (ambiguity) after circularity/comparison. SECONDARY count reconciliation (resolves the original audit's §6 FLAG, compounded by po-2023's PR-8/9 "dual criterion"): po-2023's "inclusive 101" is non-uniform — counts pk 840 (CA-missing) as fully-modeled while excluding pk 829 + 847/848/853 (same shape). Drop it. ONE authoritative headline: 100 fully-modeled (strict, uniform: BOTH scheme AND conflict present). 9 partial leaves honestly tracked as 5 CA-missing (circularity + ambiguity, would become undercuts with an ontology extension) + 4 RA-missing (pure comparison + vague meta). Recommendation: commit 100 (strict) for release notes / OWL regen. Read-only: 0 write Cards/, 0 build, 0 test, 0 OWL regen. Adversarial (count FLAG resolved, not a rubber-stamp). Reproducible via 498_audit_extract_pr789.py (scratchpad). Dispatch ynv05a (PRIMARY + SECONDARY). Base bdba45d. Co-authored-by: Claude-Code <noreply@anthropic.com> * docs(taxonomy): second i18n gap — 48 rows empty Subsubfamily (third column on #712 rows) Inventory + gated fix proposition. Dispatch ynv05a TERTIAIRE (ai-01 2026-07-06 01:32): apply the #712 method Fallacies-CSV-wide to find OTHER i18n-propagation gaps (FR filled / EN empty + derivable) beyond the 48 Family rows. Base master bdba45d. Read-only: 0 write Cards/. All-pairs FR→EN scan (i18n_gap_scan.py, scratchpad, ~60 LOC stdlib) across hierarchy + content columns: - Subsubfamily (EN) empty: 48 rows — the ONLY other hierarchy gap. 100% mechanically derivable (each row's Soussousfamille FR maps 1-to-1 to one canonical Subsubfamily_en, 0 conflict). 6 distinct FR values: Jeu de pouvoir→Power games (33), Conditionnement→Conditioning (6), Mensonge→Lying (4), Langage persuasif→Loaded language (2), Ambiguïté narrative→Narrative ambiguity (2), Repoussoir→Foil (1). - text_fr/desc_fr/example_fr → _en: 0 gaps (content fully translated). - link_fr → link_en: 24 gaps (URLs — human-research lane, not FR→EN translatable; memory [[i18n-coverage-gap-is-link-urls]], out of scope). Key finding: the 48 Subsubfamily_en-empty rows are the EXACT SAME 48 rows as #712's Family_en gap (overlap 48/48, 0 difference). #712 proposed a 2-column fill (Family_en + Subfamily_en) but OMITTED Subsubfamily_en — this is a third column on the same rows. Recommendation: merge this fill into the #712 apply as a third column (one pass, 48 rows × 3 EN hierarchy cells), same gated post-tag window. Same 8 family-header rows (pk 0 + 7 roots) correctly excluded (FR Soussousfamille empty = roots). Gated fix (NOT executed, release freeze on Cards/): same byte-targeted apply discipline as #654/#712. Validation gates: re-derive mapping on current master (abort if conflict), row count unchanged, BOM +CRLF byte-preserved, no Soussousfamille drift. Read-only: 0 write Cards/, 0 build, 0 test. Reproducible via i18n_gap_scan.py (scratchpad). Dispatch ynv05a (TERTIAIRE). Base bdba45d. Co-authored-by: Claude-Code <noreply@anthropic.com> --------- Co-authored-by: Your <your.email@example.com> Co-authored-by: Claude-Code <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Scope
Ninth cluster of the #498 AIF chantier (defeasible exceptions to argument schemes, Walton paradigm). Completes the Ambiguïté subfamily (3/3 sub-subs). Second half of the Ambiguïté narrative sub-sub: the deferred mechanisms from PR-8 — source-deception (pk 881-883), interrogation (884-885), polytélie (886), 6 leaves.
Predecessors: PR-1 #699, PR-2 #701, PR-3 #703, PR-4 #705, PR-6 #708, PR-7 #711, PR-8 #713. Repo ref master
204adc47.What the cluster looks like (code=truth)
PR-8 (#713) scoped the Ambiguïté narrative sub-sub to the implication mechanism (pk 877-880). This PR completes the sub-sub with the deferred mechanisms, all borrowing the in-sub-sub anchor pk 876 (
ConflictingGoals_Conflict) where it fits and diverging to epistemic/action schemes where the mechanism is genuinely different.Proposed AIF structure (6 leaves)
ExpertOpinion_InferenceExpertiseInconsistency_Conflictskos:narrowMatchPositionToKnow_InferenceLackOfCompleteKnowledge_Conflictskos:narrowMatchExpertOpinion_InferenceExpertiseInconsistency_Conflictskos:closeMatchPracticalReasoning_InferenceConflictingGoals_Conflictskos:narrowMatchBias_Inferenceskos:narrowMatchConflictingGoals_Conflictskos:closeMatch6 leaves proposed, 6 fully-modeled, 0 FAIL-LOUD — 2nd consecutive clean cluster (after PR-8).
Patterns demonstrated
ExpertOpinion_Inference/PositionToKnow_Inference; interrogation →PracticalReasoning_Inference.Bias_Inference(885, degenerate undercut like PR-4 800);ConflictingGoals_Conflictanchor reuse (886, literal multiple-goals).Mechanism specialization via MappingType (881 vs 883)
Grey (881) and black (883) propaganda share the same scheme+conflict pair (
ExpertOpinion_Inference+ExpertiseInconsistency_Conflict). The distinction is the mechanism: grey = concealment (passive omission,narrowMatch), black = falsification (active commission,closeMatch). The MappingType carries real semantic content — graded source-deception.Cross-PR-8 reuse (882 → 880)
Campagne de murmures (882) reuses PR-8's 880 epistemic pair (
PositionToKnow_Inference+LackOfCompleteKnowledge_Conflict), distinguished by the organized-campaign flavor (vs single anonymous indiscretion).Native palette expansion (4 new tokens)
First chantier use of
ExpertOpinion_Inference,ExpertiseInconsistency_Conflict,PracticalReasoning_Inference,Bias_Inference— all confirmed native (1x/1x/2x/2x). 21 distinct native tokens demonstrated across 9 PR.§7 I/RA/CA decomposition (jsboige enriched DoD)
All 6 leaves are undercuts — deception/interrogation defeats the inference the audience/target is manipulated to draw (source ⟹ reliability, false premise ⟹ genuine reaction), not the surface claim. No undermine, no rebut.
ExpertiseInconsistency_Conflictundercut onExpertOpinion_Inference(presumed source inconsistent with actual)LackOfCompleteKnowledge_Conflictundercut onPositionToKnow_Inference(anonymity hides position-to-know)ConflictingGoals_Conflictundercut onPracticalReasoning_Inference(friendly-surface premise false)Bias_Inferenceself-undermining)ConflictingGoals_Conflictundercut (multiple objectives conflict)Decomposition recorded but not serialized — pending jsboige's schema decision (PR #707 §4). This PR is in-sub-sub anchor and independent of that decision.
DoD status (#498)
AIF_skosMappingTypecoherentGate boundaries (HARD)
aif:assertion emitted.Chantier #498 cumulative (post-PR-9, projected)
baseline 70 → 100 strict / 101 inclusive (+30 across 9 PR) / 104 DirectRef-loose. Ambiguïté subfamily COMPLETE 3/3 (18/18 leaves) — 2nd fully-complete subfamily after Fallacious comparison (PR-1/2/3). 4 subfamilies touched overall. Remaining Misleading-language: Equivoque (2 mapped + 18 unmapped, large — split by depth), Arbitrary definition (HELD pending I/RA/CA schema ratification).
Relates: dispatch
ih617l(tertiary), #498 (chantier), PR-1 #699, PR-2 #701, PR-3 #703, PR-4 #705, PR-6 #708, PR-7 #711, PR-8 #713, coverage-status #707, verification-audit #709, family-gap #712, #133/#130 (existing OWL), #499 (inverse), #677 (0 fabrication), #192, #458.🤖 Generated with Claude Code