Plato-cohort coherence gaps — audit v1.9 Phase 1 (2026-06-23)

Run: v1.9 Plato-cohort comprehensive audit · branch claude/audit-2026-06-23 Method: 6 read-only cluster readers (5 thematic Plato clusters + 1 cross-tradition integration reader), opus 1M, returning structured findings; orchestrator-assembled. No files were written by subagents. Inputs: 38 Plato source pages + ~34 Plato-native concept pages + the 7 cross-tradition cousin claims + adjacent existing-corpus pages.

Summary

59 findings — 5 high / 21 medium / 33 low. By kind: missing-crosslink 22 · integration-issue 11 · ok-note 11 · undersupported-claim 8 · distortion 4 · coherence-gap 3.

Headline reads: the Plato cohort is internally sound and well-extracted — accurate Stephanus anchoring, candidate-claim status discipline largely observed, the hard intra-Plato distinctions (recollection vs acquisition; Magnesia-second-best vs law-second-best; Minos/Statesman herdsman contradiction) held open rather than collapsed, and cross-tradition false-friends not collapsed anywhere. The defects are overwhelmingly connective, not substantive: one-directional typed links where the new Plato page reaches into the existing corpus but the established page does not reciprocate; a few under-anchored claim legs; one register-over-strong typed connection; and one genuine dangling citation. These route cleanly to Phase 4 (integration sweep) and Phase 8 (claim anchoring).

High-severity findings (5)

H1 [epistemology-method / missing-crosslink] — the-divided-line is the cluster's most under-linked node. It anchors the Republic's noēsis / two-dialectics epistemology but is not reached from anamnesis, elenchus, or aporia; the divided line and collection-and-division pose the same "two dialectics in Plato" question (ascent-to-the-unconditioned vs taxonomic descent) without acknowledging it is one question. → Phase 4: add to anamnesis a link to [[the-divided-line]] under its periagōgē / non-recollective-routes discussion; mirror the "two senses of dialectic" pointer reciprocally on the-divided-line and collection-and-division Open Questions.

H2 [epistemology-method / undersupported-claim] — plato-two-epistemologies Republic leg is unanchored. The claim's text invokes "the Republic's periagōgē (re-orienting an already-present faculty)" as a non-recollective route, but the-divided-line (the Republic epistemology page anchoring noēsis) is neither in the claim's Sources/Wiki-homes nor cross-linked; the Republic leg rests on allegory-of-the-cave/anamnesis prose only. → Phase 8: add [[the-divided-line]] (and/or [[allegory-of-the-cave]]) to the claim's Wiki homes and cite the periagōgē/divided-line evidence leg explicitly.

H3 [ethics-moral-psychology / missing-crosslink] — the Phaedo's moral-psychology content is orphaned. The soul "rules and opposes" the body (94d), purification = wisdom (69c), soul as the part justice improves — all connect to no cluster psychology concept page; Phaedo is referenced only by Forms-metaphysics pages (theory-of-forms, anamnesis, form-of-the-good, allegory-of-the-cave), exactly where the cluster focus directs attention. → Phase 4: add a Connections bullet on plato-phaedo tying 94d to [[tripartite-soul]] (proto-partition: soul opposing its own affections) and 69c to [[socratic-intellectualism]] (purification=wisdom); reciprocate on tripartite-soul as soul-rules-body-without-yet-partitioning.

H4 [politics / integration-issue] — plato-republic is a one-way hub / sink. It is the philosopher-king source the entire political cluster points to, yet its body is purely epistemology/metaphysics (Forms, Good, Line, Cave, mimesis): it never states the philosopher-king thesis, never names the statesmanship/rule-of-law/Republic→Statesman→Laws arc, and its Connections link to none of the cluster pages. → Phase 4: add a Core-Argument/Key-Finding for the philosopher-king thesis (473c–d) after the Rule-18 raw check below, plus Connections into the political arc.

H5 [politics / undersupported-claim] — DANGLING CITATION: Republic 473c–d "Source-page anchored" does not resolve. plato-letters-philosopher-king-biographical-test cites "[[plato-republic]] — the philosopher-king thesis (473c–d). Source-page anchored", but 473c–d / the philosopher-king passage appear nowhere on plato-republic.md. The same dangling anchor underlies plato-statesman-ideal-ruler-vs-philosopher-king and plato-political-art-staged-aporia-to-law. → Wave B (Rule-18 / Phase 6): targeted raw-source check of Republic 473c–d in the Cooper/Hutchinson Complete Works. If verified, add the passage to plato-republic (Core Arguments + Key Passages) so the citation resolves; if not, downgrade the three claims' Republic evidence bullet to an explicit gap note. This is a citation-traceability item (General Rule 16) — three claims currently cite a passage absent from the page they name.

Medium-severity findings (21) — by cluster

epistemology-method

M [missing-crosslink] Cratylus 438e ("reach the things themselves") ↔ Theaetetus 186d ("knowledge in the process of reasoning") — sibling anti-empiricist route-to-the-things, not cross-linked. → reciprocal Connections bullets.
M [missing-crosslink] aporia omits the Statesman's failed-first-division/self-correcting diairesis (262–275; "limb by limb" 287c) — the cluster's clearest methodological-aporia instance. → add to aporia Key Points/Sources, cross-link [[plato-statesman]]/[[collection-and-division]].
M [missing-crosslink] eristic.md cites [[claims#plato-three-counterfeits-of-logos]] (l.30) and [[claims#plato-political-art-staged-aporia-to-law]] (l.34) without the required (candidate) framing — CLAUDE.md compliance slip. → add (candidate) to both. (mechanical; could be done in Wave A but batched to Phase 4.)
M [coherence-gap] "two dialectics" crux stated in parallel on collection-and-division and the-divided-line without acknowledging it is one shared question (see H1). → reciprocal Open-Questions pointer.

metaphysics-forms

M [missing-crosslink] khora asserts *contrasts with* [[flesh]] and the khora-third-kind claim names flesh-as-element a Wiki home, but flesh-as-element has zero reverse link to khora — the flagship Plato↔MP edge is one-directional. → add *has cross-tradition cousin* [[khora]] to flesh-as-element (under Derrida caution; cite claim as candidate).
M [missing-crosslink] plato-timaeus links to no sibling Plato dialogue (Connections point only to MP/concept pages) — receptacle never tied to plato-phaedo/plato-sophist/plato-parmenides/plato-philebus on the source layer despite the peras-apeiron-khora-cousins claim pairing Timaeus+Philebus. → add same-author cousin link to [[plato-philebus]] + Forms link to [[plato-phaedo]].
M [missing-crosslink] Cooper directs the Philebus fourfold be studied with the Sophist's being/not-being, but plato-sophist never links plato-philebus. → reciprocal typed connection.
M [coherence-gap] Timaeus ingest updated khora.md+motifs.md to "6 sources" for the matrix/khora HUB but matrixed-ontology.md still says "5 sources" — three surfaces out of sync. → update matrixed-ontology motif section to 6 + add Timaeus PRIMARY attestation. (mechanical staleness fix.)

ethics-moral-psychology

M [missing-crosslink] [[claims#plato-apology-philosophy-as-highest-piety]] lists piety-to-hosion a Wiki home but piety-to-hosion never cites it (only socratic-ignorance does) — half-realized homing on the page most about piety. → add (candidate) ref in Stakes/Open-Questions.
M [missing-crosslink] The four definitional siblings under-cross-link: Laches names only Protagoras; Euthyphro names Meno/Protagoras but not co-defendants Laches/Charmides. The shared elenchus→aporia machine grounding plato-elenchus-fails-to-define-the-virtues is thinly wired between source pages. → reciprocal "shares mechanism with" bullets on plato-laches/plato-euthyphro.
M [coherence-gap] Euthyphro's piety-as-part-of-justice mereology (12d) is used on both sides without flagging the tension: socratic-intellectualism treats virtue-division as the picture it collapses, while piety-to-hosion/Euthyphro present it as a positive locating move. → note intra-corpus contest on socratic-intellectualism.
M [missing-crosslink] Gorgias's ordered-soul/self-rule (sōphrosynē, kosmos/taxis 491d–508a) — the corpus's positive treatment of temperance — but plato-gorgias never links [[temperance-sophrosyne]] (one-directional). → add to Gorgias Concepts Referenced.

politics

M [missing-crosslink] plato-letter-7 carries the philosopher-ruler thesis autobiographically and is cited by the biographical-test claim, yet links to none of statesmanship/rule-of-law/plato-letters/plato-statesman/plato-laws, and the claim omits it from Wiki homes. → add political-arc Connections bullet + add [[plato-letter-7]] to claim homes.
M [missing-crosslink] statesmanship asserts the Statesman overturns unity-of-virtue (306b hostile-parts) and links *contrasts with* [[socratic-intellectualism]], but socratic-intellectualism gives no reciprocal back-link. → add "is contested on unity of virtue by [[plato-statesman]]".

eros-beauty-rhetoric-art

M [missing-crosslink] Phaedrus and Symposium are the paired eros/beauty dialogues (Cooper pairs them; Symposium=ascent, Phaedrus=recollection route), plato-symposium links plato-phaedrus but not reciprocally — the cluster's most surprising omission. → add contrast bullet on plato-phaedrus.
M [missing-crosslink] plato-hippias-minor contrasts with plato-gorgias/plato-ion (technē-as-knowledge-standard turned paradox-engine, 376b) but neither links back — one-directional craft/technē thread. → reciprocal bullets citing [[claims#plato-hippias-minor-craft-analogy-reductio]] (candidate).

(Full medium list with Stephanus loci preserved in the run digest; the above is the routing-relevant set.)

Low-severity findings (33) — rollup

Predominantly one-directional typed-link reciprocity gaps (mirror an existing edge from the other page): changement-quantite-qualite→peras-apeiron; khora→peras-apeiron; form-of-the-good→statesmanship; plato-ion/plato-gorgias→plato-hippias-minor; elenchus→collection-and-division (Sophist 230b–231b demotion); plato-crito↔plato-laches (many-vs-one-who-knows); priority-of-soul+Phaedo 105d (soul-brings-life precursor register). Plus minor items: Gorgias Stephanus-boundary drift (503d vs 503e — normalize to 503e–508a); eros Alcibiades-I disputed-authorship caveat should travel to point-of-use (l.58); a few inline candidate-claim refs on socratic-intellectualism Stakes lacking adjacent (candidate) tag.

11 ok-notes record genuinely sound integration to preserve as models: the recollection-vs-acquisition false-friend triangulated across anamnesis/maieutics/plato-theaetetus; the Sophist↔Hegel bidirectional non-being↔bestimmte-negation integration; the Derrida "khôra resists analogy" caution held load-bearing across source/concept/claim; the Laches/Protagoras courage-mirror source↔concept↔claim coherence; the-mean's three-way Statesman/Laws integration; the Crito vertical-Laws-contract false-friend vs Hobbes/Rousseau.

Mis-seed flagged: play-as-political-virtue is a Merleau-Ponty/Chouraqui concept (zero Plato content) — out-of-scope for the Plato politics cluster; do not cross-link it to Plato pages.

§A — Cross-tradition cousin disposition table (PRIMARY Phase-8 input)

The cross-tradition integration reader produced explicit dispositions for all 7 cousin candidates. All 7 HELD at candidate this run (no promotion — per the v1.9 Phase-8 disposition rule + bridge-card-signoff §12e: a Latent-Adjacent cross-tradition cousin relation is below claim-grade by design). Classification feeds the follow-up weave run's retire-and-relocate queue.

#	Cousin claim slug	Disposition	Rationale (strip the relation → what remains?)
1	`anamnesis-stiftung-cross-tradition-cousins`	RELATION-ONLY → flag retire-and-relocate	Payoff (Meno problem shared; MP temporalizes an atemporal structure) re-describes the relation; the contestable residue (aitias-logismos↔sedimentation micro-parallel) is explicitly not yet anchored. Already realized as `has cross-tradition cousin` on `anamnesis`.
2	`mp-writing-and-living-inverts-phaedrus`	RELATION-ONLY (borderline) → flag retire-and-relocate	Single typed relation (inversion) dressed as claim; once the inversion-vs-independent-antipode concession is granted, only "same vocabulary pair, opposite evaluation" survives — already carried by `inverts` on `writing-and-living`. Keep candidate only if MP's literature course is shown to engage Plato directly.
3	`peras-apeiron-khora-cousins`	RELATION-ONLY (cleanest case) → flag retire-and-relocate	Entire content = cousinhood + disanalogy guardrail (quality-range/first-kind vs place/third-kind). Payoff is the relation-with-guardrail. Intra-Plato (same-author) — see §B register note.
4	`republic-cave-aletheia-orthotes`	RELATION-ONLY (attribution/genealogical) → flag retire-and-relocate	Reports a position to Heidegger, not a thesis the wiki advances; wiki's own `aletheia` page records Heidegger's 1964 retraction of the Wesenswandel thesis as "nicht haltbar". Carry as typed connection + BRIDGE with the retraction caveat; do not promote.
5	`sophist-not-being-as-difference-vs-determinate-negation`	CONTESTABLE-THESIS → stay candidate	Beyond the heteron↔bestimmte-Negation relation, carries a corrective: the Sophist is the strongest internal Platonic counter-case to Hegel's "Platonic dialectic stops at the negative" charge (which fits the Parmenides, not the Sophist). Contestable. Promotion blocked pending Hegel-Sophist-reception cross-check.
6	`khora-third-kind-mp-third-domain`	CONTESTABLE-THESIS → stay candidate (speculative)	Carries a genealogical thesis: the Timaeus khōra is the missing primary anchor of the existing matrix/khōra HUB (previously floated on MP-reception alone). Opposable on Derrida's grounds. Promotion blocked until Derrida's Khōra/Dissemination is ingested.
7	`peras-apeiron-hegel-masz`	CONTESTABLE-THESIS (weak) → stay candidate + 8.6 merge-check	Carries an engine-reversal thesis (Plato's limit from-above/static vs Hegel's immanent nodal leap) + Pythagoras→Plato→Hegel genealogy. Hazard: may double-count the existing `collection-and-division`↔`bestimmte-negation` parallel re-described — route through Phase-8.6 merge check.

Net: 4 relation-only (→ follow-up weave retire-and-relocate, user-confirmed) + 3 contestable-thesis (→ stay candidate). False-friend audit PASSED on all three high-risk pairs (anamnesis≠Stiftung; khōra≠flesh; Form-of-the-Good≠Being/aletheia) — no collapse anywhere; no anachronistic phenomenology vocabulary imported into any Plato page's own voice (cross-tradition terms stay quarantined to Stakes/Connections/claim-ref sections).

§B — Register-over-strong typed connections (flag for adjudication; NOT auto-fixed)

Two existing typed connections assert *shares mechanism with* where the cited claim only supports a cousin-under-suspicion:

khora ↔ matrixed-ontology (khora.md l.49, plato-timaeus.md l.111): *shares mechanism with* crosses the tradition boundary (Plato→MP/Kaushik reception) on a speculative-confidence, Derrida-quarantined claim — the single most over-strong link in the cohort. Internally inconsistent on khora.md (flesh gets *contrasts with*, matrixed-ontology gets the stronger verb). → retype to *has cross-tradition cousin* (or *contrasts with* pending the Derrida ingest). This is the same schema-tension class as the v1.8 D6 idole↔fetish fix.
peras-apeiron ↔ khora (peras-apeiron.md l.48, plato-philebus.md l.115): *shares mechanism with* for an intra-Plato same-author pair is registrally permissible, but stronger than the peras-apeiron-khora-cousins candidate (whose own counterpressure invokes Derrida "a fortiori"). → downgrade to *contrasts with*, or keep only if the claim is promoted past candidate.

Routing: these are typed-connection retypes (not removals). Per the integration reader, treat as Phase-8 adjudication, not Phase-4 auto-fix; the cross-tradition leg (#1) is also a natural follow-up-weave item. Surfaced to the maintainer at the Wave-A checkpoint.

§C — Routing to later phases

Wave B / Rule-18: Republic 473c–d raw check (H5) — citation-traceability priority.
Phase 4 (Wave B): the reciprocity/missing-crosslink fixes (H1, H3, M-set, L-set), batched ≤20 files/commit, priority = high/medium first. eristic (candidate)-framing fix + matrixed-ontology "6 sources" staleness fix are mechanical.
Phase 8 (Wave C): H2 + the cousin disposition table (§A) + the register-over-strong adjudication (§B) + the peras-apeiron-hegel-masz 8.6 merge-check.
Follow-up weave (separate run): retire-and-relocate the 4 relation-only cousins to typed connections + motifs.md BRIDGE entries (user-confirmed; retirement HALTS).