Timothy McGrew

Philosopher whose 2013 paper "The Argument from Silence" (Acta Analytica 29:215-228) gives the old historiographical inference — that a source's failure to mention something can itself be evidence — a formal Bayesian structure: the evidential force of a silence as the product of three probabilities, that the author would have noticed the fact (N), recorded it (R), and that the record would have survived for a later reader to find (S). Matters to this vault as the named source behind the framework grading every negative-citation finding the vault has reached by a cited party's own silence — Brooks/Dreyfus, Speelpenning/Linnainmaa, Griewank/Amari, Kodosky/Dennis, and others — indexed at moc-argument-from-silence. The same paper carries the vault's own standing caution about the method: humans systematically over-trust exactly this kind of inference, so the framework exists to grade silences, not to bless them uniformly.

References

• claim-mcgrew-argument-from-silence-is-notice-record-survive-product

• claim-mcgrew-humans-systematically-over-trust-arguments-from-silence

• observation-bjork-dreyfus-same-argument-from-silence-graded-differently

• observation-brooks-dreyfus-kodosky-dennis-share-method-not-evidentiary-strength

• claim-lipton-construes-argument-from-silence-as-inference-to-best-explanation

• claim-walton-classifies-argument-from-silence-as-argument-from-ignorance

• Captures: 2026-07-11-hop-argument-from-silence, 2026-08-15-what-genuinely-connects-brookss-1991-intelligence-without-representation

• 2026-09-19: the 2013 paper carries a footnote beyond its notice/record/survive apparatus, previously unpulled into the vault — McGrew reports two rival scholarly classifications of the argument from silence he does not himself adopt: Lipton (with Sinnott-Armstrong & Fogelin) as inference to the best explanation, Walton as a species of argument from ignorance. See claim-lipton-construes-argument-from-silence-as-inference-to-best-explanation and claim-walton-classifies-argument-from-silence-as-argument-from-ignorance (observation-ach-abduction-lucky-widrow-notes-share-argument-from-silence-method-classified-as-abductive).

• 2026-10-04 (promotion of a recovered capture): the product is applied to Isaacs's 1951 RAND colleague-silence for the first time, finding it weak on a term the hub's existing soft instances (Dreyfus, Kodosky/Dennis) don't represent — notice, not record: a document that never had occasion to generate the kind of record a silence argument needs, as distinct from one whose record is merely incomplete or unread (claim-isaacs-1951-rand-silence-is-weak-on-mcgrew-notice-term-not-record-or-survival).

• 2026-10-11: A third application, and the mirror image of the Isaacs case — Aikins's silence on Jevons's Logic Piano grades weak on record, not notice: Aikins demonstrably read the book closely enough to cite it to the page, but his own stated project in those chapters never gave him occasion to invoke the machine's chapter two thematic blocks later (claim-aikins-silence-on-jevons-logic-piano-is-weak-on-mcgrew-record-term-not-notice).

• 2026-10-17: the framework also marks its own boundary — the prerequisite that there is a silence to grade at all. The vault's Warden used that boundary to keep Werbos's obscurity testimony OUT of moc-argument-from-silence while folding Isaacs's P-257 silence in: Werbos's is authorial self-explanation of a reception gap, not a silence read against a corpus, so N/R/S never engages it. Surfaced by a bridge check on the two notes' MOC-gap observations (observation-werbos-and-isaacs-maintenance-gap-notes-are-a-hedged-bridge-split-by-hub-epistemic-type).