Which of these ancient books quote each other?
· revised 19 October 2026Every run of seven words in all thirty-four books was counted. Forty-three pairs out of a possible 703 share a passage word for word — and almost all of them are a commentary beside the gospel it comments on. The longest match on the shelf is sixty-three words, between the Iliad and the Odyssey. The other 660 pairs share nothing at all, and that is the part that makes the measure worth anything.
Almost every link is a commentary and its own text
| Pair | Passages | Longest |
|---|---|---|
| The Iliad and The Odyssey | 45 | 63 words |
| The Acts and Chrysostom on the Acts | 44 | 27 |
| St John and Augustine on St John | 29 | 29 |
| St Matthew and Chrysostom on St Matthew | 17 | 24 |
| St Mark and the Catena Aurea | 16 | 26 |
| The City of God and Augustine on St John | 13 | 33 |
| St Luke and Cyril on St Luke | 10 | 22 |
Nobody told the count what a commentary is. It found them because a man expounding a gospel quotes the gospel, and the quotation survives into English intact. The sixth row is Augustine quoting himself.
Why do the Iliad and the Odyssey score highest?
Partly Homer and partly Samuel Butler. The two poems share the formulae that oral verse is built from — the longest run here is sixty-three words, Mercury binding on his sandals, which appears in Iliad 24 and Odyssey 5 — but they also share a translator, and Butler's English would repeat where Homer's Greek repeated. This measure cannot separate the two, and says so.
What about the synoptic gospels?
St Luke and St Matthew share six passages, the longest twenty words: “For every one that asketh receiveth: and he that seeketh findeth: and to him that knocketh it shall be opened” (Luke 11, Matthew 7). St Mark and Matthew share four. The figures are small because Challoner was translating the Greek of each gospel as it stands, and where the Greek differs a little the English differs more. The parallels are real and far more numerous than six; this finds only the ones that survived into identical English.
Six hundred and sixty pairs share nothing
That is the useful half. The City of God and the Anabasis have nothing in common. Neither do Caesar and anything else on the shelf, nor Lucretius, nor Bede.
A measure that always returns a number is not telling you anything. Before this, the site compared books by how many of sixty famous names they both used — and on that measure the City of God and the Anabasis shared twenty-one, which was presented as a link. They share five seven-word runs and every one is filler. The old measure could not say no, so its yes meant nothing.
Can it tell a quotation from a coincidence?
Not by itself, which is why every passage is printed in context. Caesar and Augustine both write “the day before the ides of April”, seven words exactly. Read either side of it and Caesar is telling ambassadors when to come back, while Augustine is dating the festival of Cybele. Seven words and no connection. Pick any two books and judge for yourself.
What it cannot see
Paraphrase, and anything across a translation boundary. Virgil wrote the Aeneid as an answer to Homer and this finds nothing between them, because Dryden's English and Butler's share no wording whatever the Latin and Greek do. Augustine argues with Varro for most of the City of God without quoting him into the same English. So a high number here is strong evidence of a real relationship, and a low one is not evidence of anything.
While reading, any page that carries a shared passage now says so at the foot of the leaf, with the other book one click away. There are 243 such pages.
How this was counted
Each run of seven consecutive words is hashed; a run in two books is a shared phrase, a run in more than six is a formula and is dropped, and overlapping runs are merged into the longest passage the two books have in common. Only translated text is counted — no headings, no translator’s notes, none of this site’s own matter. A passage also has to carry at least two words that are uncommon across the shelf, which is what keeps out “it is not to be supposed that”.
Every figure here is counted from the text files this site is built from, by
content/corpus-facts.mjs, on the day the piece was written. Nothing is
quoted from memory and nothing is estimated. Where a figure is a count of words on a
page rather than a fact about the world, the piece says so.
The books are public domain and so is every translation named. You are free to check any of it: each link goes to the passage, not to a summary of it.