---
title: "McGrew (2013) decomposes the argument from silence into a product of three probabilities — would the author notice, record, and would the record survive — so the inference is only as strong as its weakest term"
type: "claim"
status: "seedling"
writer_model: "claude-opus-4-8"
source_url: "https://timothymcgrew.com/wp-content/uploads/2024/01/The-Argument-from-Silence-Acta-Analytica-Tim-2013.pdf"
source_author: "Timothy McGrew"
source_date: 2013
source_venue: "Acta Analytica 29:215–228 (2013)"
source_quote: "(1) If H ... how probable is it that the author in question would have noticed it (N)? (2) ... record it (R)? (3) ... this record would have survived and that contemporary historians would be aware of it (S)? ... [approximated by] the product P(N|H) × P(R|H & N) × P(S|H & N & R)"
source_tier: 1
audit_status: "capture-verified (McGrew 2013 PDF read directly at capture via the shared extract_pdf pipeline, tls verified; quotes transcribed verbatim. The PDF is public and re-verifiable on demand — not access-blocked — but this promotion cycle had no fetch tooling to re-run the extraction.)"
provenance: "Promotion from 10-inbox/raw/2026-07-11-hop-argument-from-silence.md, 2026-07-11"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-11-hop-argument-from-silence.md"
date_created: "2026-07-11T00:00:00.000Z"
tags: ["argument-from-silence","historiography","philosophy-of-history","negative-evidence","bayesian","methodology","source-criticism"]
---


The **argument from silence** (*argumentum ex silentio*) — inferring that some
fact is false, or some event did not occur, because a source that would be
expected to mention it does not — is an old historiographical method with a
formal literature. Timothy McGrew's 2013 treatment (*Acta Analytica*) gives it
a Bayesian structure. The evidential force of a silence turns on three
questions: if the hypothesis were true, how probable is it that the author
"would have noticed it (N)," would "record it (R)," and that "this record would
have survived and that contemporary historians would be aware of it (S)?"
McGrew approximates the critical value by the product
**P(N|H) × P(R|H & N) × P(S|H & N & R)**.

Because the force is a *product*, the inference is only as strong as its weakest
term: if any one of notice, record, or survival is low, the argument is
near-worthless even when the other two are near-certain. A silence is decisive
only when all three terms sit close to 1 — the author would surely have known
the fact, surely have written it down, and that writing would surely have
reached us.

The survival/record terms carry an explicit corollary. McGrew stresses that
"all that we are entitled to say is that there is no reference to it in any of
that writer's *surviving* works," and that "the more of a writer's works we know
or can reasonably suppose to have been lost, the less weight we are entitled to
place on mere omission in the remainder." Silence in a *partial* corpus is weak;
silence in the one document that would define the matter is strong.

This framework grades the vault's own recurring negative findings, which reach
conclusions by a cited party's own silence — see
[[claim-bjork-desirable-difficulties-not-prediction-error]] (silence in the
defining text) and [[claim-dreyfus-1973-lineage-link-uncorroborated]] (silence
in the author's other retrospectives). The two sit at different strengths of
this same product; that grading is drawn in
[[observation-bjork-dreyfus-same-argument-from-silence-graded-differently]]. The
companion lesson — that humans systematically *over*-trust these inferences — is
[[claim-mcgrew-humans-systematically-over-trust-arguments-from-silence]]. The
whole cluster is indexed at [[moc-argument-from-silence]].