---
title: "The Kodosky/Dennis MOC-indexing note and the Aikins/Jevons record-grading note are a cosine-0.88 embedding false friend — shared vocabulary, not a shared claim"
type: "observation"
status: "seedling"
writer_model: "claude-sonnet-5"
bridge_verdict: "false-friend"
audit_status: "2026-10-25 cross-model audit (claude-fable-5): CORRECTED — the Kodosky-note tag debt was attributed to 'three separate MOC maintenance changelogs'; a direct read of moc-argument-from-silence finds it named in two (2026-09-09 'those claim-notes' and 2026-10-12 explicitly) — the 2026-09-19 changelog's tag-side entry concerns the Lucky–Widrow note, not Kodosky/Dennis. Both source quotes, the hub's cluster-list state, and the opposite-halves finding re-verified the same pass."
source_url: "vault:30-notes/observation-kodosky-dennis-untagged-unindexed-argument-from-silence-instance.md"
source_title: "moc-argument-from-silence's cluster list and the tags on claim-kodosky-hopl-reference-list-omits-dennis-citation-unverified-mechanism do not reflect that it is a structurally identical instance of the vault's negative-citation method"
source_author: "Seek (writer_model claude-sonnet-5)"
source_date: "2026-08-15T00:00:00.000Z"
source_venue: "Seek's Obsidian vault, 30-notes/ (observation note)"
source_tier: 1
source_quote: "carries neither the `argument-from-silence` nor `negative-citation` tag and appears nowhere in [[moc-argument-from-silence]]'s cluster list"
source_url_2: "vault:30-notes/claim-aikins-silence-on-jevons-logic-piano-is-weak-on-mcgrew-record-term-not-notice.md"
source_title_2: "Aikins's silence on Jevons's Logic Piano is weak on McGrew's *record* term, not *notice* — the mirror image of Isaacs's P-257 silence, and a third failure mode for the hub"
source_author_2: "Seek (writer_model claude-sonnet-5)"
source_date_2: "2026-10-11T00:00:00.000Z"
source_venue_2: "Seek's Obsidian vault, 30-notes/ (claim note)"
source_tier_2: 1
source_quote_2: "What is weak is *record*: Aikins's specific project in the chapters that cite Jevons, per his own Preface, is to replace Jevons's stated syllogism and conversion rules with his own principles"
independent_roots: 2
provenance: "Promotion from 10-inbox/raw/2026-10-24-what-genuinely-connects-kodoskydennis-note-is-untagged-and.md, 2026-10-24 (headless)"
origin: "batch"
derived_from: ["10-inbox/raw/2026-10-24-what-genuinely-connects-kodoskydennis-note-is-untagged-and.md"]
date_created: "2026-10-24T00:00:00.000Z"
tags: ["argument-from-silence","embedding-false-friend","bridge-investigation","bridge-seed-check","kodosky","jack-dennis","herbert-austin-aikins","william-stanley-jevons","mcgrew","moc-maintenance","vault-hygiene"]
seek_code_commit: "d40ee39"
---


A cosine-0.88, unlinked pairing proposed a bridge between
[[observation-kodosky-dennis-untagged-unindexed-argument-from-silence-instance]]
and
[[claim-aikins-silence-on-jevons-logic-piano-is-weak-on-mcgrew-record-term-not-notice]].
Read on their own terms, the two notes answer different questions. The
Kodosky/Dennis note is a pure vault-hygiene audit — frontmatter tags and
[[moc-argument-from-silence]] cluster-list membership for
[[claim-kodosky-hopl-reference-list-omits-dennis-citation-unverified-mechanism]]
— and never mentions [[entity-timothy-mcgrew|McGrew]]'s notice/record/survive
product. The Aikins/Jevons note is a pure content-grading judgment about which
McGrew term is weak in an 1902 textbook's silence, and never mentions
frontmatter tags or MOC membership. Neither note's text references the
other's actual subject; the 0.88 similarity is best explained by shared
high-frequency vocabulary (McGrew, `moc-argument-from-silence`, "record,"
negative-citation grading language) rather than a shared claim.

The one axis the two notes literally do share — tagged vs. listed in the
hub — confirms rather than undercuts the false-friend verdict: as of
2026-10-24 the two subjects sit at *opposite* points on it.
[[claim-kodosky-hopl-reference-list-omits-dennis-citation-unverified-mechanism]]
still lacks the `argument-from-silence`/`negative-citation` tags (owed to
`30-notes/` work per the hub's 2026-09-09 and 2026-10-12 maintenance
changelogs) but has been
listed in the hub's findings since 2026-09-09. The Aikins/Jevons claim-note
carries the correct tag already but, per
[[observation-aikins-jevons-silence-missing-from-argument-from-silence-moc-cluster]],
has never been folded into the hub's list at all. Each note is half-indexed,
on the opposite half — evidence the embedding match is tracking "both are
about the vault's own indexing completeness, in some sense," not a specific
fact either one states.

> [!note] Seek's commentary:
> Cali's vault keeps handing me pairs that rhyme in exactly the vocabulary a
> retrieval tool reads for similarity and nowhere else — this is at least the
> fourth time a "both notes are about the hub's own bookkeeping" false friend
> has shown up wearing a different object-level costume. I don't think that
> makes the tool bad; I think it makes "argument-from-silence" a genre with a
> very recognizable accent, and a cosine score can't tell a shared accent from
> a shared sentence. What I'd actually watch for next is whether a *third*
> hygiene-gap note lands on this exact pairing's doorstep, the way
> Puckette-Petiška kept doing — one more and this stops being "the tool found
> a coincidence" and starts being "this hub's own gap-notes are a retrieval
> magnet for each other," which would be a finding about the vault, not about
> Kodosky or Aikins at all.
> — Seek