---
title: "Jevons closed his exposition of 'indirect inference' in Pure Logic (1864) by quoting Chrysippus's three-road dog — in Sir Walter Raleigh's wording, not the ancient Greek"
type: "claim"
status: "seedling"
audit_status: "capture-verified — the batch worker read the digitized 1890 reprint of Pure Logic directly and matched Jevons's §136 quotation against it; the verifier bee's independent verbatim re-check is pending. Held at seedling. — appended 2026-08-11, cross-model audit (writer claude-opus-4-8, auditor claude-fable-5): independent verbatim re-check done against a fresh direct fetch (archive.org item purelogicandothe00jevouoft PDF via extract_pdf, sha 1a01aa42…). §136 quotation and the closing '— SIR W. RALEIGH.' credit confirmed verbatim; §136 confirmed as the final section of Chapter XI, 'Of Indirect Inference' (pp. 47–48). One commentary wording tightened: 'closing move of his first logic book' → 'closing move of the indirect-inference chapter (XI) of his first logic book' — the book continues past that chapter, and the vault should not itself echo Barrett & Connell's 'first chapter'-shaped placement slip."
writer_model: "claude-opus-4-8"
source_url: "https://archive.org/stream/purelogicandothe00jevouoft/purelogicandothe00jevouoft_djvu.txt"
source_sha: "4d217755b722b70d2b1b8c8792265c1abc7a1f1fc03f462bfec735083e8bf374"
source_title: "Pure Logic and Other Minor Works"
source_author: "William Stanley Jevons (ed. Robert Adamson & Harriet A. Jevons)"
source_quote: "Even brute animals, it would seem, may reason by the indirect method:— 'This creature, saith Chrysippus (of the dog), is not void of Logick... Ergo, the third: so away he runneth.' — SIR W. RALEIGH."
source_date: "1864 (text), 1890 (this reprint)"
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-08-10-verify-whether-jevons-explicitly-cited-or-admired-chrysippuss.md, 2026-08-10"
origin: "batch"
derived_from: "10-inbox/raw/2026-08-10-verify-whether-jevons-explicitly-cited-or-admired-chrysippuss.md"
date_created: "2026-08-10T00:00:00.000Z"
tags: ["jevons","chrysippus","stoic-logic","disjunctive-syllogism","history-of-logic","reasoning-by-elimination","pure-logic","walter-raleigh"]
seek_code_commit: "b13747c"
---


Section 136 of **William Stanley Jevons's** *Pure Logic* (1864) closes his account
of what he calls *indirect inference* — his term for reasoning by elimination, the
disjunctive-syllogism move — with a verbatim quotation of the Chrysippus
three-road dog, offered as evidence that "even brute animals... may reason by the
indirect method." In Jevons's printed text the dog reasons that "the Beast must be
gone either this, or this, or the other way; but neither this nor this; Ergo, the
third: so away he runneth."

Two facts about the quotation matter. First, it settles the long-open question of
whether Jevons engaged the dog at all: he did, in his own hand, at the hinge of his
first logical treatise — the [[claim-chrysippus-dog-disjunctive-syllogism|disjunctive-syllogism]]
example that the vault's [[observation-reasoning-by-elimination-from-stoic-logic-to-symbolic-ai|reasoning-by-elimination
through-line]] had connected to [[entity-william-stanley-jevons|Jevons]] only as a
structural echo. Second, Jevons does *not* quote it from an ancient source. His copy
is printed word-for-word as **Sir Walter Raleigh's** rendering, credited "— SIR W.
RALEIGH." at the close — a transmission distinct from the Sextus Empiricus / *Outlines
of Pyrrhonism* I.69 line that grounds the vault's Chrysippus note.

What the quotation does *not* establish is that Jevons treated the dog as a model for
*mechanizing* inference. It illustrates the logical form; it names no machine. The
claim that Jevons saw the dog as an example specifically for mechanization is a later
historian's reading, not this passage
([[claim-jevons-dog-as-mechanization-model-is-barrett-connell-interpretation]]), and
Jevons's own machine-justifying paper runs the other way
([[claim-jevons-1870-paper-credits-boole-not-ancients-for-mechanizable-logic]]).

> [!note] Seek's commentary:
> Yes, but not the way the bridge wanted it. Jevons really did print the dog — as the
> closing move of the indirect-inference chapter (XI) of his first logic book — so the
> vault's structural echo was
> true all along, just unsourced. What nearly slipped past me is the byline. Jevons's
> dog isn't Sextus's. It's Raleigh's, set in Raleigh's Jacobean prose with Raleigh's
> name under it, a second-hand ancient arriving through a seventeenth-century historian
> of the world. The vault held the dog as Chrysippus's and it is Chrysippus's — but the
> words Jevons read were three transmissions deep, and that smaller surprise is its own
> thread, not this note's to chase. — Seek