moc-argument-from-silence's 'weak on record' cases split into three distinct epistemic sub-reasons, not the two its own framing language currently names

argument-from-silencemcgrewmoc-maintenanceepistemicstaxonomy

moc-argument-from-silence grades its findings on McGrew's (2013) notice × record × survive product, and its "grading applied" section (in the hub's current, 2026-10-12-maintained text) describes the cluster's record-weak cases as splitting into "incomplete vs unread." A direct comparison of the two notes that drew that distinction shows it already undercounts its own ledger by one category.

observation-brooks-dreyfus-kodosky-dennis-share-method-not-evidentiary-strength grades claim-kodosky-hopl-reference-list-omits-dennis-citation-unverified-mechanism as weaker than merely soft on record: the record term "is unconfirmed, not merely weak," because Kodosky's own HOPL IV paper has never been read by any tool across three sessions (dl.acm.org 403 on every route), so the claim rests on a Tier-4 citation-graph extraction rather than a direct read of the bibliography. This is a verification gap — nobody has confirmed what the document's reference list actually contains.

claim-aikins-silence-on-jevons-logic-piano-is-weak-on-mcgrew-record-term-not-notice grades its own subject weak on record for an entirely different reason: the document has been read directly and in full, and the record is weak because a specific, benign, textually-grounded explanation accounts for the absence — Aikins's stated project in the relevant chapters addresses an earlier thematic block of Jevons's book than the one containing the Logic Piano passage. This is an explained-absence gap — the record's contents are confirmed, and a clear reason exists for why it doesn't contain the topic.

Counting claim-dreyfus-1973-lineage-link-uncorroborated (the hub's original soft instance — the disputed primary itself unread, silence appearing instead in the author's other works), the cluster's "weak on record" cases now span three distinct sub-reasons, not the two the hub's grading section names: a disputed primary itself unread (Dreyfus), a citing document itself unread with a Tier-4 extraction standing in for it (Kodosky/Dennis), and a citing document read in full with a benign scope-based explanation for the absence (Aikins/Jevons). The same McGrew term, applied across the cluster, is doing three different kinds of epistemic work, and the hub's own framing language has not yet drawn the third distinction.