The Source He Names, and the Source He Doesn’t

• Anchor Bible Dictionary — “Jude, Epistle of,” “Enoch, First Book of,” “Moses, Assumption of,” “Michael,” “Watchers.”
• Dictionary of the Later New Testament & Its Developments (IVP) — “Jude,” “Apocryphal and Pseudepigraphal Writings,” “Adversaries,” “Canon.”
• NIDNTTE — τηρέω, ἀσέβεια, ἔλεος, ζόφος, προφητεύω.
• Dictionary of NT Use of the OT (Beale & Carson) — the Jude chapter, which spends most of its length on texts that are not in the Old Testament.

Jude itself has no offset. This was derived from the corpora rather than copied from a table.

Identical. There is no Hebrew column and no Septuagint column for a New Testament book, so the usual MT/LXX/ESV question does not arise for the passage itself. The only wrinkle is a filing convention: because Jude has one chapter, printed Bibles cite Jude 7 while the MorphGNT files label the same verse Jud 1:7. This page prints both on every row.

One textual note that is not a versification note: at verses 22–23 the manuscripts divide over whether there are two groups of people or three. The SBLGNT prints three; some witnesses have two. The verse numbers are the same either way, so nothing on this page shifts — but a reader comparing translations will see the difference.

The single Greek Old Testament partner of this page’s deepest run sits inside the hardest renumbering on the shelf.

Greek Jeremiah is about an eighth shorter than the Hebrew and moves the Oracles Against the Nations from the back of the book (Hebrew 46–51) into the middle, reordering the nations inside the block. The shelf’s versification table marks Jeremiah PROVISIONAL for exactly that reason and forbids copying a printed table. So this one block was aligned here by content, verse by verse:

Within the Edom oracle the Greek number is the Hebrew number minus six, and the block runs Greek 30:1–16 = Hebrew 49:7–22. Four independent anchors agree, and the fifth row shows where the block stops. That is a claim about this block only — nothing here should be carried over to the rest of Jeremiah, where the shelf’s own tooling still says PROVISIONAL.

Practical consequence: a reader chasing Jude 7 into a printed Septuagint and turning to Jeremiah 49 will find a different chapter entirely. And a second consequence for this shelf — Jeremiah 49:14 is the verse Obadiah 1 also carries, which is the subject of the Obadiah study. Jude’s deepest Old Testament contact lands four verses further down the same oracle.

Jude 14–15 names a figure out of Genesis, says he prophesied, and then quotes him — and the words quoted are not in the Old Testament.

The page will not call this the only such passage, because it is not. John 11:50–51 quotes Caiaphas and then says he prophesied, and those words are not in the Old Testament either. What is distinctive about Jude 14 is the combination: a patriarch out of Genesis 5, named with his generation, quoted as a prophet, from a book outside the canon.

The sentence corresponds to 1 Enoch 1:9, part of the Book of the Watchers, a Jewish work of roughly the third to second century BC that survives complete in Ge’ez, in Greek fragments, and in Aramaic pieces from Qumran.

⚠️ 1 Enoch is not held on this shelf in any language, and no part of it was measured for this page. The identification above is the standard one and is reported as background, not as a measurement. What was measured is the other side of the question: how the verse behaves against the corpus that is held.

Jude 14 shares no three consecutive words with any verse in the Greek Old Testament held here. Which is what a page ought to expect if the quotation really does come from somewhere else — but the page cannot verify the somewhere else, and says so.

Two genuine MT/LXX divergences sit in the background verses, and both are worth printing. At Deuteronomy 33:2 the Hebrew reads מֵרִבְבֹת קֹדֶשׁ, from myriads of holiness, and the Greek reads σὺν μυριάσιν Καδης — where the Hebrew has the word for holiness the Greek column has the place-name Kadesh. Which column stands behind which is not asserted here. At Zechariah 14:5 the Hebrew ends all the holy ones with you and the Greek ends all the holy ones with him. Jude’s phrase, with his holy myriads, matches the sense of the Hebrew of Deuteronomy and the pronoun of the Greek of Zechariah — and reproduces the running wording of neither, though the single word μυριάσιν does stand in the Greek of Deuteronomy 33:2. Nothing here says which text anyone was reading.

Jude 9 reports a dispute between Michael and the devil over the body of Moses. That episode is not in the Old Testament.

Origen, Clement of Alexandria and later writers attribute the story to a work they call the Assumption of Moses. A related work, usually printed as the Testament of Moses, survives in a single damaged sixth-century Latin palimpsest and breaks off before any such scene. Whether the two are one book is a live question in the scholarship.

⚠️ None of that was read for this page. Neither the Assumption nor the Testament of Moses is on this shelf in any language. The paragraph above is a summary of the standard account and is labelled as such. It is background, not evidence.

What could be measured is the one clause of verse 9 that does have a canonical neighbour — and the measurement is small. Michael’s reply, ἐπιτιμήσαι σοι κύριος, is the sentence the angel of the Lord speaks to the Adversary at Zechariah 3:2: ἐπιτιμήσαι κύριος ἐν σοί. Same rare optative, same speaker-and-adversary setting, different word order — so the contiguous-run screen scores the pair at one. See stage 4: this is the clearest counter-example on the page.

Verse 6 tells the story of Genesis 6:1–4 in the form it had taken by the first century.

Genesis gives four verses to the sons of God and the daughters of men. In the Second Temple retellings — 1 Enoch 6–21 above all — the Watchers leave their station, are bound, and are kept in darkness until the great day of judgement. Jude 6 uses all four of those elements in one sentence: they did not keep their ἀρχή, they left their dwelling, they are kept in chains under gloom, for the judgement of the great day. Not measured — 1 Enoch is not on this shelf. Read as a witness to how Genesis was being read, not as Scripture.

Verse 11 is a woe, and the form is a prophet’s. οὐαί with a dative and a ὅτι clause is how Isaiah, Amos and Habakkuk open an indictment. The three names inside it — Cain, Balaam, Korah — had by the first century become a standing trio for greed and rebellion; that grouping is a commonplace of the period’s literature and was not measured here, because the sources for it are not on this shelf.

The house position on these books: 1 Enoch, the Assumption of Moses and Jubilees are read here as historical witnesses to how Second-Temple Jews were reading Genesis and Numbers, not as Scripture. That Jude quotes 1 Enoch is a fact about Jude’s rhetoric. It is not, by itself, a canonical judgement, and this page does not make one.

Counted by lemma across all 27 books: 16 of Jude’s 226 distinct lemmas occur once in the New Testament and that once is here.

ἐπαγωνίζομαι (3) · παρεισδύω (4) · ἐκπορνεύω (7) · δεῖγμα (7) · ὑπέχω (7) · φυσικῶς (10) · Κόρε (11) · σπιλάς (12) · φθινοπωρινός (12) · ἐπαφρίζω (13) · πλανήτης (13) · ἀσεβέω (15) · γογγυστής (16) · μεμψίμοιρος (16) · ἀποδιορίζω (19) · ἄπταιστος (24)

⚠️ And here is the counter-example this figure needs. Sixteen in 25 verses sounds extraordinary until it is normalised, and then Jude is fourth, not first:

Jude’s vocabulary is unusual, and 2 Peter’s is more unusual still. What the table shows is a family: the short late letters cluster at the top and the Gospels and the Johannine letters at the bottom. 1 John has one hapax legomenon in 105 verses. That is the real spread, and any claim about Jude’s vocabulary has to be made inside it.

This is the sharpest thing the page can say about the two letters, and it is not a run-length finding at all. Three words in the whole Greek New Testament have exactly two occurrences, and in all three cases one is in Jude and one is in 2 Peter. The contiguous-run screen scores those three pairs at two, one and three — which is to say the screen barely notices them. Rarity and run length are different measurements, and a page that reports only one of them will mislead.

Every one of Jude’s 25 verses was screened against every verse unit of the Greek Old Testament held on this shelf.

⚠️ Re-run 24 September 2026 over the 30,348 verse units now screened (63 CATSS files, of 64 held; first run over 23,061 units in 44 files). The top of the table held — the same five at Jude 7, the same single partner. Two rows rose from two to three on the new books, each with a tie band of one: Jude 10, the unreasoning animals, against 4 Maccabees 14:14, and Jude 20, in the Holy Spirit, against Psalms of Solomon 17:37. Neither is claimed as a source. And the band at Jude 25 rose from 2 to 9: seven verses of Tobit carry the same doxological for all the ages.

The two rows that go deepest, with their bands:

Why the band at Jude 7 is one, checked by hand. Seven verse units in the screened Greek Old Testament (re-counted 24 September 2026 on the enlarged corpus: still seven, none of them in the newly added books) carry Σόδομα καὶ Γόμορρα as a sequence: Genesis 13:10, Genesis 19:24, Deuteronomy 29:22, Isaiah 13:19, Amos 4:11, Jeremiah 27:40 and Jeremiah 30:12. Only one of them continues with καὶ αἱ.

Two of the seven reach four, and only one reaches five — which is exactly the sort of margin a reader should be shown rather than told about. The fifth word is αἱ, the feminine plural article, and it is the whole difference between a band of one and a band of three.

And here is the detail worth the whole row. Hebrew Jeremiah 49:18 and 50:40 are a near-doublet, and both end the clause with the same Hebrew noun שְׁכֵנֶיהָ, her neighbours — 49:18 prefixing it with a plain waw (וּשְׁכֵנֶיהָ) and 50:40 with waw plus the object marker (וְאֶת־שְׁכֵנֶיהָ). The Greek renders that one Hebrew noun two different ways in the two places — αἱ πάροικοι αὐτῆς at Greek 30:12 and τὰς ὁμορούσας αὐταῖς at Greek 27:40. Jude 7 lines up with one of them and not the other, and that is why the tie band is one rather than two. No claim is made about which text Jude was reading, or about whether he was reading one at all.

All 25 rows against all 7,927 Greek New Testament verses, with Jude’s own rows removed from the target.

The counter-examples, printed here so the five at verse 13 is not read as more than it is. Verses 17 and 21 both reach five and both have a tie band of thirty-four. A run of five with a band of thirty-four tells a reader nothing at all about where it came from. The deepest run in the letter, six words at verse 25, is a doxological formula shared with Romans. Depth without a band is not evidence.

A second counter-example, and this one is instructive. Jude 7 — the verse that reaches five into the Greek Old Testament — reaches only three into the New Testament, and its partner is Mark 4:19, the interpretation of the parable of the sower. The three shared words are καὶ αἱ περί, and in Mark they belong to the desires for other things. Same three words, entirely unrelated sentences. That is what the instrument does when left to itself.

The relationship between Jude and 2 Peter 2 is the most discussed literary question in the New Testament letters. Measured as contiguous shared word-forms it is thinner than its reputation, and it is unmistakably there.

First the thinness — the famous paragraph-by-paragraph parallels, measured:

Two of the closest thematic pairs on that list measure one. Whatever the relationship is, at sentence level it is not copying.

Now the reality of it, because “thin” means nothing without a baseline. All 25 Jude rows were screened against every New Testament book in turn:

2 Peter is 61 verses — eight tenths of one per cent of the Greek New Testament — and more Jude rows reach three words into it than into Matthew, Luke, Acts or Romans, each of which is between seven and nineteen times longer. It is also the only book that takes four Jude rows to a depth of four or more and holds the one five with a tie band of one. The gap is the relationship.

What the measurement will not do. It will not say which letter was written first. The shelf’s standing rule applies here without exception: direction of influence is never asserted. A run screen counts shared sequences and has no access to chronology. Every reconstruction — Jude used 2 Peter, 2 Peter used Jude, both used a third source, both drew on a shared body of preaching — produces exactly the numbers above. The numbers are reported. The inference is the reader’s.

Jude 9 against Zechariah 3:2 measures one word. Every commentary connects them, and the commentaries are right.

Same verb, same mood, same subject, same addressee, same scene — an accuser being rebuked rather than answered. Two shared word-forms out of three, and not one of them in the same order, so the longest contiguous run is one.

This is the page’s own argument against itself, and it belongs here rather than in a footnote. A contiguous-run screen finds verbatim contact. It does not find allusion, it does not find shared idiom, and it cannot see a writer reusing a sentence with the words rearranged. Jude 7 reaching five and Jude 9 reaching one does not mean verse 7 is a stronger connection than verse 9. It means the instrument can see one of them and not the other.

The claim is that Jude’s deepest contiguous run against the Greek Old Testament is five words, once, at verse 7. That claim is true of the corpus screened here — every CATSS book, 30,348 verse units in 63 files (of 64 held; the Odes are held but withdrawn from screening) — and that corpus is not the whole Septuagint, and not every book Jude draws on.

So the honest statement of the finding is this: across the 30,348 Greek Old Testament verse units screened on this shelf, Jude’s deepest contiguous run is five words with a tie band of one, at verse 7, partnered with Greek Jeremiah 30:12. The rows where a deeper run was plausibly hiding were 11 and 14–15; Numbers has now been measured and hides nothing, and the book still missing is 1 Enoch. Genesis 6, Genesis 19 and Zechariah 3 are in books that are held, were measured, and come out low. The exodus of verse 5 is no longer a gap: Greek Exodus is held, and that verse’s own wording ἐκ γῆς Αἰγύπτου comes back three words, with a tie band of one hundred and four across the screened corpus — twenty-four of them in Exodus itself. A formula, not a quotation. Until 24 September 2026 this paragraph read “across the 23,061 Greek Old Testament verse units this shelf holds”, named “Numbers and 1 Enoch” as the missing books, and called the exodus “a third gap — the book of Exodus is not held either”, with a tie band of sixty-one.

No critical apparatus for the Greek New Testament. This matters twice in 25 verses. At verse 5 the SBLGNT prints Ἰησοῦς — Jesus saved a people out of Egypt — where other witnesses read the Lord or God, and the doctrinal weight of that variant is obvious. At verses 22–23 the witnesses divide over two groups of people or three. The page prints the text it holds and names the alternatives, and has weighed the evidence for neither.

No 1 Enoch in Ge’ez, Greek or Aramaic. No Assumption or Testament of Moses. No Qumran material. No Testaments of the Twelve Patriarchs. No patristic sources read in the original: Eusebius’s classing of Jude among the disputed books, Jerome’s report that some rejected it because of the Enoch citation, and Tertullian’s defence of Enoch are all named here from the standard summaries.

No stylistic or syntactic analysis. Jude’s Greek is good and its sentences are long; whether it is closer to 2 Peter’s register than the run screen can see was not tested, because the instrument on this shelf measures shared sequences and not style.

And a limit on the method itself, which is worth more than any of the above. A contiguous-run screen counts shared word forms in order. It cannot see paraphrase, it cannot see shared structure, and it cannot see an argument built out of an Old Testament story in the writer’s own words. Jude 5–7 and 11 are exactly the kind of writing this screen is blind to — the substance is Exodus, Genesis and Numbers, and the vocabulary is Jude’s. The finding of this page should therefore be read narrowly: the letter does not copy, which is not at all the same as saying the letter does not use. Jude 9 against Zechariah 3:2, printed in stage 4 at a run of one, is the proof of that inside the page’s own numbers.

Jude was disputed in the early church, and the reason given was the Enoch citation. Eusebius lists it among the antilegomena; the Muratorian Fragment includes it; Jerome reports that it was rejected by some because it cites an apocryphal book. None of those sources was read here in the original.

The measurements on this page are compatible with more than one account of who wrote the letter and when. The three-word overlap with James 1:1 can be read as one family’s habit or as a convention two writers shared. The relationship with 2 Peter can be read in either direction or in neither. The screen cannot choose between those, and neither will this page.

What can be said flatly is that quoting a book is not the same as canonising it, and that this page has no way to measure the difference. Jude names Enoch as a prophet and uses his words. What that implies about the status of 1 Enoch is a theological judgement, made by readers and by councils, and not by a run screen.

• Richard J. Bauckham, Jude, 2 Peter (Word Biblical Commentary 50) — the standard technical treatment, and the source of the widely-followed reading of Jude as a structured midrash.
• Richard J. Bauckham, Jude and the Relatives of Jesus in the Early Church.
• Peter H. Davids, The Letters of 2 Peter and Jude (Pillar NTC).
• Gene L. Green, Jude and 2 Peter (BECNT).
• Thomas R. Schreiner, 1, 2 Peter, Jude (New American Commentary).
• George W. E. Nickelsburg, 1 Enoch 1 (Hermeneia) — for the text Jude 14–15 quotes and this page could not check.
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer. They are named as the places to check this page against.

The thing that came out of the measurement was not what went into it. The expectation was that a page on Jude would be a page about the Enoch quotation. The Enoch quotation is the one line here that no corpus on this shelf can check — and when the verse is screened against everything that is held, it comes back with a maximum of two words and those two words are and also.

Meanwhile the letter’s deepest verbal contact with the Greek Old Testament, five words with a tie band of one, is at verse 7 — a verse that names nobody, cites nothing, and simply says as Sodom and Gomorrah and the surrounding cities. The clause it matches is in an oracle against Edom, and a reader with a printed Septuagint has to know that English Jeremiah 49:18 is Greek Jeremiah 30:12 to find it at all.

Which is a fair description of how anybody actually carries a text. The things we announce we are quoting are the things we have looked up. The things that come out word for word are the things that were already in us. Jude announces Enoch and reproduces Jeremiah.

And the letter knows what it is doing with all of it. Five times in 25 verses a piece of remembered material is followed by οὗτοί εἰσιν — these are the ones. The exodus, the angels, Sodom; Michael; Cain, Balaam, Korah; Enoch; the apostles. Each time the old word is laid down and then aimed. Jude is not writing history. He is naming people at his readers’ own table.

The frame of the letter is one verb. Verse 1: kept for Jesus Christ, passive, done to them. Verse 6: the angels who did not keep are the ones he has kept. Verse 21: keep yourselves in the love of God, imperative. Verse 24: to him who is able to keep you from stumbling — and there the verb changes, which the page notes rather than smooths over. Kept, and keep, and kept again. The letter that opens by naming a danger closes by naming who holds the door.

This page closes a gap: Jude had no study on this shelf at all. Its nearest neighbours are the 2 Peter study, which measured the same relationship from the other side, and the Obadiah study, which works the same Jeremiah oracle four verses further up.