Thirteen Twins in Hebrew, Three in Greek

• Anchor Bible Dictionary — “Proverbs, Book of,” “Wisdom in the OT,” “Septuagint.”
• Dictionary of the OT: Wisdom, Poetry & Writings (IVP) — “Proverb, Genre of,” “Sayings Collection,” “Translations, Ancient.”
• NIDOTTE — תועבה, מקור, עד.

That Proverbs repeats itself is in every introduction to the book. The lists of “doublets” or “variant repetitions” are old, and no commentary needed a computer to find 14:12 and 16:25.

What a full screen adds is the negative half, which a list cannot give you. A published list tells you which verses repeat. It does not tell you that nothing else in the section reaches four words with anything, anywhere in the Hebrew Bible, or that every one of those runs has a tie band of exactly one, or how far down the next candidate sits. Those are the figures this page is for.

What was not consulted. No commentary’s doublet list was opened, before or after. The thirteen pairs here were produced by the screen and by nothing else, which means the page cannot tell you whether they agree with Fox’s list or Snell’s. They were found by a threshold of four shared words, and a doublet that shares three would not be in them.

Thirteen pairs, seventeen source rows, every tie band 1, no partner outside Proverbs. Six partners sit in 25–29, four inside this section, three in 10–12.

Proverbs 25:1 carries its own heading: these also are proverbs of Solomon which the men of Hezekiah king of Judah copied.

So six of the thirteen deepest pairs in this section have their other member in a block the book itself labels as a later copying. That is a fact about where the verses stand and about what the heading at 25:1 says.

It is not evidence of direction and this page does not use it as any. A shared run of seven words tells you the two verses are the same saying. It does not tell you which collection had it first, whether either was copied from the other, or whether both drew on something neither of them is. The heading at 25:1 is a claim inside the book, not a measurement, and nothing here tests it.

Screened the same way, every verse of Proverbs against every other verse in the Hebrew Bible, the whole book gives 29 verses at four, 11 at five, 4 at six, 6 at seven and 4 at eight.

Every run of six or more in the entire book — fourteen verses, seven pairs — points at another verse of Proverbs, every one with a tie band of 1: 6:10 with 24:33, 12:11 with 28:19, 14:12 with 16:25, 18:8 with 26:22, 19:5 with 19:9, 20:16 with 27:13, 21:9 with 25:24. Seven of those fourteen verses are in this section.

A run with a band of 1 has exactly one partner in 23,204 verses. A run with a band of 14 has fourteen, and is a stock phrase.

The abomination line is the sharpest case on the page. תּוֹעֲבַת יְהוָה alone stands in 19 verses of the Hebrew Bible, 11 of them in Proverbs; on two words it is a formula. Extended to the four-word colon an abomination to the Lord, both alike, it stands in exactly two, 17:15 and 20:10, and it is a pair. Where the formula stops and the pair begins is at word three.

82 of the 268 verses in this section — 30.6% — share no two consecutive words with any other verse in the Hebrew Bible. They are spread evenly: 11 in chapter 13, 17 in 14, 8 in 15, 4 in 16, 7 in 17, 10 in 18, 6 in 19, 8 in 20, 11 in 21.

Some of the section’s best-known lines are among them — 13:12 hope deferred makes the heart sick, 13:20 whoever walks with the wise becomes wise, 13:24 whoever spares the rod hates his son, 15:1 a soft answer turns away wrath, 14:34 righteousness exalts a nation, and 21:31 the horse is made ready for the day of battle.

What that zero means, exactly. This screen was run with the Hebrew-side convention, no single-word fallback, so 0 means no two consecutive words in common — not no words in common. Every one of the eighty-two shares individual words with plenty of verses. And read stage 5 before treating the figure as a property of this section: it is not.

17:25 and 19:13 share three words and a band of 1 — each is a foolish son who is a grief to his father, and neither reaches the four-word threshold this page uses.

So do 13:2 with 12:14 and 18:20, at three with a band of 2. The thirteen pairs are what a threshold of four produces; they are not the complete set of repetitions in the section, and this page does not claim they are. Lowering the threshold to three adds 25 more source rows and starts admitting formulas, which is why it was not lowered.

For each of the thirteen, the two CATSS rows at the answering numbers were measured against each other on the same rule — longest contiguous run of whole word forms.

Three of the thirteen reach four or more in the Greek; two reach five. Two have no Greek row on the 13–21 side at all — and in both of those the partner outside the section does have one, so in the Greek file the doublet stands once.

Run the same screen inside the Greek file — each of the 262 rows labelled 13–21 against the other 938 rows of CATSS Proverbs — and 26 reach four, 9 reach five, 1 reaches seven. Four of those 26 rows sit at numbers the Hebrew pairs, and they make three of the Hebrew’s thirteen pairs.

So it is not that one file repeats and the other does not. Both files carry sayings twice. They largely do not carry the same sayings twice, and this page reports that as a difference between two documents and says nothing whatever about how either of them came to be as it is.

All 262 CATSS rows in 13:1–21:31 were swept against all 7,927 verses of the SBL Greek New Testament. 217 reach two, 53 reach three, 2 reach four, and nothing reaches five.

Both fours have a tie band of 1, and both are strings of function words:

Neither is a use of Proverbs. They are printed because they are the deepest thing the screen found, and a page that printed only its best number would be lying by arrangement.

Measured across all 939 rows of the Greek Proverbs file, every run of five or more with the Greek New Testament falls outside this section — 3:11 and 3:12 with Hebrews 12:5 and 12:6 at nine and eleven, 32:22 with Romans 12:20 at ten, 11:31 with 1 Peter 4:18 at nine, 3:34 with James 4:6 and 1 Peter 5:5 at six, 24:12 with Romans 2:6 at five, 32:21 with Romans 12:20 at five. Seven rows in the book reach five with the New Testament and not one of them is in chapters 13 to 21.

30.6% of the verses here share no two consecutive words with any other verse in the Hebrew Bible. That sounds like a finding. Run the identical screen on other books and it stops being one.

This section is below its own book and well below Job. The isolation belongs to short poetic lines with no narrative formulas in them, not to Proverbs 13–21. The figure stays on the page because it is true and because it is the honest companion to the twin count — but it is not a finding about this section, and the sentence that would have said so has been removed.

20:16 and 21:9 both reach seven words with their partners on the qere. On the ketiv they reach five and four.

20:16 is written נכרים and read נָכְרִיָּה, which is what 27:13 has; 25:24 is written מדונים and read מִדְיָנִים, which is what 21:9 has.

The whole screen was therefore re-run on the ketiv, and the shape does not move. Still 268 verses, still exactly 17 at four or more, still every tie band 1, still not one partner outside Proverbs, still exactly 82 zeros. Only the depths shift: on the ketiv the distribution is 8 at four, 4 at five, 2 at six, 1 at seven, 2 at eight, against 7, 3, 2, 3, 2 on the qere. The individual depths are qere-dependent; the finding is not.

1. That the section is unusually self-referential. It is not measured against anything that would show that. What is measured is that its deep runs all stay inside Proverbs — and so do the book’s, in every one of the seven pairs at six or more. The behaviour is the book’s; this section holds seven of the fourteen verses involved.

2. That the Greek “lost” ten of the thirteen pairs. Nothing on this page says lost, dropped, removed, harmonised or corrected. What is measured is that at the answering numbers the two Greek rows run to four or more in three cases out of thirteen, and that two of those numbers carry no Greek row. Why the file is as it is — the history of the Greek, of the Hebrew, or of the numbering applied to either — is not something a run screen can reach, and no direction of influence is asserted anywhere on this page.

3. That the six pairs reaching into chapters 25–29 tell you anything about Hezekiah’s men. Six of thirteen is worth printing. It is a count of where verses sit. The heading at 25:1 is a claim the book makes about itself and this page neither confirms nor doubts it.

No Hebrew or Greek critical apparatus. No Qumran manuscript of Proverbs, no Cairo Geniza fragment, no Syriac, no Vulgate, no Aquila, Symmachus or Theodotion. No manuscript evidence bearing on any of the thirteen pairs, on the nineteen absent numbers, or on the thirteen lettered rows in this span. No published doublet list — not Snell, not Fox, not Waltke — was consulted before or after, so the page cannot say whether its thirteen agree with anyone’s. No lemmatiser: every Hebrew and Greek count here is a count of written forms, so two forms of one word do not match and a doublet rewritten in different grammar would be invisible to the screen. And the Greek glosses of rows without an English edition to hand — 13:12, 18:8, 33:22 — are working translations made here.

Where this page is weakest: the threshold. Four shared words is a line drawn by hand. It was chosen because three admits formulas — the abomination colon proves that at word two — and because at four every single hit in the section came back with a band of 1, which is the behaviour of a pair and not of a phrase. That is a good reason and it is still a choice, and a different threshold would give a different thirteen.

Read Proverbs 14:12 and then read Proverbs 16:25 and you have read the same eight words twice, eighty-one verses apart, and the book says nothing about it. No as it is written, no again, no cross-reference. The repetition is simply there.

Thirteen times in these nine chapters a saying stands beside another saying that is word for word or nearly. Six of those partners are in a block the book says was copied out generations later. And four times — 19:5 and 19:9 are four verses apart — the twin is close enough that a reader going straight through will hear it.

What is worth sitting with is the other 82. Nearly a third of this section shares no two consecutive words with anything else in the Bible — hope deferred makes the heart sick, a soft answer turns away wrath, the horse is made ready for the day of battle. The lines you can quote from memory are mostly the ones that stand alone. Which of the two facts is the more interesting is a question the measurement cannot answer, and you can.