Nineteen Numbers, Fourteen Sayings

• Anchor Bible Dictionary — “Proverbs, Book of,” “Septuagint,” “Wisdom in the OT.”
• Dictionary of the OT: Wisdom, Poetry & Writings (IVP) — “Proverb, Genre of,” “Translations, Ancient,” “Fear of the Lord.”
• NIDOTTE — יהוה in Proverbs, גאון, חנך.

Every treatment of Proverbs says the Greek text differs from the Hebrew in its arrangement, most visibly after 24:22. That is visible here: the Greek chapter labels in this file run 1–24 and then 30–36, with none numbered 25–29.

What the files cannot tell you is why. Whether a saying stands at a different number because of the history of the Greek text, or the Hebrew, or the numbering applied to either, is not something a count can reach. This page reports which numbers each file carries and whether a saying can be found elsewhere in it, and stops there.

Nineteen of twenty-four in a fifth of the book.

How the last row of that table is measurable at all. The CATSS chapter labels run 1–24 and then 30–36; there are no labels 25–29. The rows labelled 32 to 36 stand against Masoretic chapters 25 to 29 verse for verse — 28, 28, 27, 28 and 27 rows against 28, 28, 27, 28 and 27 verses — and the content agrees at every point checked, from the Hezekiah heading at 32:1 through to 36:27. Read at those labels, Masoretic 25:1–29:27 has no absent number at all. A scan that reads the labels literally reports 138 absent numbers there. That figure is a fact about the numbering and it is not used on this page.

The nineteen fall in five stretches:

And the asymmetry runs one way only: no Greek base row in chapters 16–22 lacks a Masoretic verse of the same number.

Three of the nine consecutive numbers are among them, standing together in a lettered block eleven rows before the gap begins. So the nine-number run is a fact about numbering. It is not nine missing sayings, and this page does not call it that.

The remaining fourteen were each screened against all 939 rows of the file on their own decisive vocabulary. Nothing was found for any of them. Six are consecutive — Hebrew 20:14–19 — and that six-verse stretch is the sharpest single result on this page.

“No counterpart that could be found” is exactly as strong as the screen behind it, and the screen is printed in stage 5. A saying can be present in wording so different that a vocabulary probe will not see it. This page therefore says could not be found and never is not there, and it never says that anything was removed from anything.

The obvious way to do it produces false positives, and one nearly reached this page.

A draft matched Hebrew 15:31 — the ear that listens to life-giving reproof — to the clause ὁ δὲ φυλάσσων ἐλέγχους σοφισθήσεται inside the row at 16:17. But παιδεία with ἔλεγχος is a stock pair in this file, standing in eight rows — 3:11, 5:12, 6:23, 12:1, 13:18, 15:10, 15:32 and 16:17 — and the row at 13:18 carries the identical participle phrase one word from it. Meanwhile Hebrew 15:31’s own decisive words find nothing: no row in the file carries a lodging-word at all, and of the sixteen rows carrying an ear-word, 16:17 is not one. The match was to a formula, not to the verse.

2 Corinthians 9:7 ends ἱλαρὸν γὰρ δότην ἀγαπᾷ ὁ θεός — God loves a cheerful giver.

The row numbered 22:8a in this file reads ἄνδρα ἱλαρὸν καὶ δότην εὐλογεῖ ὁ θεός — God blesses a cheerful man and a giver. There is no Masoretic verse numbered 22:8a.

The two share ἱλαρόν, δότην and ὁ θεός, and differ in the verb. The page reports the overlap and says nothing about which text is drawing on which, because nothing here can settle that.

Every one of the 181 Greek rows in 16:1–22:16 was swept against all 7,927 verses of the SBL Greek New Testament. 23.8% reach a three-word run; 0.6% reach four; none reaches five. The maximum is four, at the row numbered 21:23, and the band has one member on each side — Acts 8:35.

For comparison on the same measure: Proverbs 1–9 reaches a maximum of eleven, at 3:12 with Hebrews 12:6; Proverbs 10–15 reaches nine, at 11:31 with 1 Peter 4:18. This section is the shallowest of the three.

Proverbs has 915 verses. The 458th is 16:18, with 457 before it and 457 after.

That is arithmetic on the verse divisions in the edition used here. The page does not attribute it to the Masoretes and cites no source for a traditional mid-book note, because none was consulted. The middle word of the book falls elsewhere, inside 17:2.

Each search was run over all 939 rows, accent-stripped, matching on word-initial stems. For Hebrew 20:14–19: gold-word with lips-word together — none; pearl-word — none; garment-word with surety-word together — none; pledge-word — none; bread-word with falsehood-word together — none; gravel-word — one row, at 34:3, which answers the Masoretic verse of its own number. For Hebrew 22:6, no row pairs a child-word with a training-word. For Hebrew 18:23, no row carries a supplication-word.

Every one of those is a negative result about a probe, scoped to this file and to these stems.

1. That the Greek runs terser here. By label the Greek-to-Hebrew word ratio falls across the three sections — 1.786, 1.525, 1.460 — and that is an artefact of the nineteen absent rows. This section has 181 Greek rows for 191 Hebrew verses, while the previous one has 192 for 184. Counted on number-matched pairs only, which is the only measure where the two sides being divided actually answer each other, the ratio is 1.677, then 1.454, then 1.527 here — the section is above its neighbour, not below. Per row the Greek here is the longer one: 11.81 words against 10.66.

2. That the section is unusually isolated in the canon. No verse here shares a four-word run with any other book — but Proverbs is the lowest book in the thirty-nine on that measure, at 0.3%, and three of its six collections reach four words zero times. Measured at the three-word grain the section is on the high side of its own book, at 4.2%. The isolation belongs to Proverbs, not to this section, and the corpus-wide figure that made it look remarkable is produced by Kings, Chronicles, Deuteronomy, Jeremiah and Numbers.

3. That any saying was removed or added. Nothing on this page says that, and nothing here could establish 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 nineteen numbers or on the nine lettered rows. And no published English rendering of the lettered rows was consulted — the glosses here are working translations.

The first figure on this page took one line of code: nineteen Masoretic numbers with no Greek row to match, against five in all the rest of the book. The second figure took the rest of the work, and it is the one that is true.

Five of the nineteen were in the file the whole time. Three of them sat together, eleven rows before the gap that supposedly swallowed them. A count of labels would have called them missing and been wrong.

What is left is smaller and firmer: fourteen sayings that could not be found, six of them in a row. That is worth knowing, and it is worth knowing that it is fourteen and not nineteen.