The Prayer Cut in Two

• Chapter 29 — 28 and 28, but not the same division: Greek 29:16 = KJV 29:16–17; Greek 29:17–18 = KJV 29:18.
• Chapters 30, 31, 32, 34 — one to one.
• Chapter 33 — Greek 33:16–17 = KJV 33:16; Greek 33:18–31 = KJV 33:17–30; Greek 33:32–33 = KJV 33:31.
• Chapter 35 — Greek 35:4 = KJV 35:4–5; Greek 35:5–10 = KJV 35:6–11; Greek 35:11–12 = KJV 35:12; Greek 35:13–16 = KJV 35:13–16; Greek 35:17–18 = KJV 35:17; Greek 35:19–21 = KJV 35:18; Greek 35:22–23 = KJV 35:19; Greek 35:24 = KJV 35:20.
• Chapter 36 — Greek 36:1 = KJV 36:1–2; Greek 36:2–15 = KJV 36:3–16; Greek 36:16–17 = KJV 36:17; Greek 36:18–25 = KJV 36:18–25; Greek 36:26–27 = KJV 36:26.

• Anchor Bible Dictionary — the article on Ben Sira (Sirach, Ecclesiasticus); the articles on the Apocrypha and on the Septuagint.
• Eerdmans Dictionary of Early Judaism — Ben Sira; Wisdom literature; Cairo Genizah.
• Dictionary of New Testament Background (IVP) — Sirach; Apocrypha and Pseudepigrapha.
• On the text — any article that says plainly which numbering it uses in chapters 30–36. An article that cites ‘Sirach 33’ without saying whether it means the order printed here or the Greek manuscripts’ order cannot be followed for this range.

• Patrick W. Skehan and Alexander A. Di Lella, The Wisdom of Ben Sira (Anchor Bible 39) — a full English commentary.
• Joseph Ziegler, Sapientia Iesu Filii Sirach (Göttingen Septuagint XII/2) — the critical Greek edition, and the source of the verse numbers the Hebrew file uses.
• Pancratius C. Beentjes, The Book of Ben Sira in Hebrew (Brill, 1997) — all the Hebrew manuscripts with a synopsis; the Hebrew file uses its sigla A–F.
• Solomon Schechter and Charles Taylor, The Wisdom of Ben Sira (Cambridge, 1899) — the first edition of the Genizah leaves; page scans of it are held on this disk and were used by the Hebrew file’s builder for spot-checks.
• M. Z. Segal’s complete Hebrew edition — whose own numbering the Hebrew file keeps as a second number in chapters 30–36.
• NETS, Sirach (Benjamin G. Wright) — a modern English of the Greek. Not quoted here; named because it is the obvious next step.

The grandson’s prologue has no Hebrew in the file used here, and the Clementine Latin on this disk folds a Latin form of it into its 1:1 — but the King James file on this disk does not carry it at all: its Sirach begins at 1:1, “All wisdom cometh from the Lord”, and all 1,393 of its lines are numbered verses. So the prologue is quoted here from the CATSS rows, with our own gloss:

οὐ γὰρ ἰσοδυναμεῖ αὐτὰ ἐν ἑαυτοῖς Εβραϊστὶ λεγόμενα καὶ ὅταν μεταχθῇ εἰς ἑτέραν γλῶσσαν
for they do not have equal force — the things said in Hebrew in themselves, and when they are carried over into another tongue (our gloss)Sirach, Prologue, CATSS rows 21–22

The Hebrew file acquired for this series lets these pages set his Hebrew beside his Greek, and in this range it can do so for 119 of 218 rows. Where the two differ, the page reports what each column reads and does not say which one moved.

Reported, not measured: the Hebrew of Ben Sira was lost to the West for centuries and recovered in pieces from the Cairo Genizah from 1896, with further fragments from Qumran and from Masada. Measured, on the file used here: in this range the Hebrew is almost all MS B (159 main rows and 28 margin rows), with MS F (43), MS E (10) and MS D (3) — 243 rows over 118 verse numbers. Chapters 29 and 34 have no row at all.

What the file is and is not. It is a set of diplomatic-style transcriptions on Hebrew Wikisource, made almost entirely by one editor known only by a user name, pinned to fixed revisions; it is not a critical edition. Wikisource marks several of its pages unfinished — in this range the MS B pages for chapters 31, 32 and 35, whose margin readings are partly untyped, and the MS E pages. “No row in the file” is therefore not the same as “not in the manuscript”, and the page says “in this file” every time.

Reported, not measured: the account given in the commentaries and critical editions (not consulted here) is that every surviving Greek manuscript of Sirach carries two long blocks of chapters 30–36 in exchanged order, that the Latin, the Syriac and the Hebrew carry the other order, and that the modern critical editions — Rahlfs and Ziegler — print the latter. The usual explanation is that two runs of leaves changed places in an ancestor copy.

Measured here, by content: the CATSS file, the KJV and the Clementine Vulgate all carry one order. Brenton’s English — commonly described as a translation of the Vatican manuscript’s text — carries the other. Aligned to the KJV line by line (its English is, verse after verse, the KJV’s own wording: 203 of its 212 lines in these chapters share ten or more consecutive words with a KJV verse, median 17), Brenton’s chapters run:

• Brenton 30:1–24 = KJV and CATSS 30:1–24.
• Brenton 30:25–40, 31, 32 and 33:1–11 (73 lines) = KJV 33:16–36:11a — block Y, the household, the dreams, the sacrifices and the prayer, up to “gather all the tribes of Jacob”.
• Brenton 33:12, 34, 35 and 36:1–16 (72 lines) = KJV 30:25–33:15, plus a last line, Brenton 36:16, that matches no whole KJV verse — block X, the cheerful heart, the gold, the banquet and the law — ending with that line, which joins the watcher to the inheritance.
• Brenton 36:17–31 = KJV 36:12–26.

A Greek column in the same order — held only as OCR. The disk also holds the Internet Archive text of a Bagster printing of Brenton’s edition, catalogued 1900, which sets Greek and English together. At the three joins its English column runs as the eBible.org file does. Its Greek column is poorly OCR’d and is not measured here, but at the three joins it can be read. Printed exactly as the OCR has it (OCR errors kept — am is the scan’s misreading of two Greek letters; a slash marks a line-break; margin verse numbers removed), with the OCR line numbers:

• After 30:24 (OCR lines 230809, 230810 and 230812; line 230811 is OCR noise): καὶ θυμὸς ἐλαττοῦσιν ἡμέρας, καὶ πρὸ καιροῦ γῆρας ἄγει / μέριμνα. / Ὡς καλαμώμενος ὀπίσω τρυγητῶν, ἐν εὐλογίᾳ Κυρίου — the end of 30:24, then the second half of the row CATSS numbers 33:16, without κἀγὼ ἔσχατος ἠγρύπνησα.
• The petition’s first half (OCR lines 231199 and 231201; line 231200 is OCR noise): Σύναγε πάσας φυλὰς Ἰακώβ. / Λαμπρὰ καρδία καὶ ἀγαθὴ ἐπὶ ἐδέσμασιν τῶν βρωμάτων — then the cheerful heart of CATSS 30:25.
• The far end (OCR lines 231578–231579, after 33:15’s two and two): Καγὼ ἔσχατος ἠγρύπνησα, καὶ κατεκληρονόμησα αὑτοὺς / καθὼς am ἀρχῆς. — the watcher’s first clause and the petition’s second half, joined by and.

Read beside CATSS, that column has the petition’s first word as Σύναγε where CATSS has συνάγαγε, and the verb of its second half as κατεκληρονόμησα, a first-person form, where CATSS has the imperative κατακληρονόμησον. At the far end, then, its two first-person verbs make one sentence of the watcher’s clause and the inheritance; and the watcher’s clause, which Brenton’s English prints at both joins, is absent from this Greek column’s first join and stands at the far end. These are readings of one poor scan, not a collation; the page uses them only to say that a Greek text in the other order is on this disk, and what it reads at the joins.

• Deuteronomy 28:8 → Sirach 29:12. “Shut up alms in thy storehouses”: the Greek row ἐν τοῖς ταμιείοις σου shares five consecutive words with Deuteronomy 28:8, where the Lord commands the blessing on Israel’s storehouses — band 1 across the 30,348 screened Greek Old Testament units.
• Proverbs on surety → Sirach 29:14–20. The surety vocabulary stands 15 times in the screened Greek Old Testament: 7 in Sirach and 6 in Proverbs. Proverbs warns against standing surety; Sirach 29:14 says a good man will do it, and then counts the cost.
• The Clementine Latin — reported to be the Old Latin rather than a fresh translation by Jerome — is divided more finely: 250 rows here against the Greek’s 218. At its 36:13 it carries a clause between the two halves of the petition — ut cognoscant quia non est deus nisi tu — that also stands in its 36:2.

All 218 reach one; 196 reach two; 79 reach three; 11 reach four; 1 reaches five; nothing reaches six. For the whole book, 1,402 rows, the ceiling is the same: 4 rows reach five and nothing reaches six — and one of the four is in this range.

Bold rows are band 1 from both sides: the Sirach row has one New Testament partner at that depth, and that New Testament verse has no deeper or equal partner anywhere else in the Greek Old Testament screened here.

Sirach 33:13 → Ephesians 1:9: five words, band 1 both ways. The core, κατὰ τὴν εὐδοκίαν αὐτοῦ, occurs once in the 30,348 screened Greek Old Testament units and once in the 7,927 New Testament verses. In Sirach it is the potter’s pleasure over the clay; in Ephesians it is the pleasure by which God made known the mystery of his will. Nothing here says either writer knew the other.

Sirach 29:24 → Luke 10:7: four words, band 1 both ways. ἐξ οἰκίας εἰς οἰκίαν — from house to house — once in each corpus. Sirach: the misery of the lodger who moves from house to house and dare not open his mouth. Luke: the instruction to the disciples sent out in Luke 10 not to keep moving from house to house (our gloss). Two guests, two sets of instructions, one phrase.

• The potter is not the run. Sirach 33:13 is a potter-and-clay verse, ὡς πηλὸς κεραμέως, and the New Testament’s potter and clay are in Romans 9:20–21. Measured: one word against each. Even the two nouns of ὡς πηλὸς κεραμέως stand side by side only twice in the screened Greek Old Testament (here and Isaiah 41:25; Isaiah 29:16 and Jeremiah 18:6 have them with the article between) and never in the New Testament, where Romans has them in other cases. The five-word run is elsewhere in the same row.
• Two runs of four that belong to someone else. 36:13 → 2 Thessalonians 1:9 is four words — but that New Testament verse runs eight words into Isaiah 2:10, 2:19 and 2:21. 32:23 → Matthew 22:37 and Luke 10:27 is four — and Matthew 22:37 runs five into Deuteronomy 6:5 and seven more units. Both New Testament verses run deeper into other Greek Old Testament units than into Sirach; Sirach is standing near the same words.
• Lay up treasure → Matthew 6:19–20: one word. 29:10, 29:11 and 29:12 each measure one against the Matthew verses. The idea is near; the wording is not. Sirach’s storehouse is spelled ἐν τοῖς ταμιείοις σου; Luke 12:3 has its first three words, the noun in another spelling, which this screen reads as a different word.
• The tribes of Jacob → Matthew 19:28 and Luke 22:30: one word, tribes. The New Testament speaks of the twelve tribes of Israel; the Greek phrase φυλὰς Ιακωβ is not in it.
• The prayer of the humble → Revelation 8:4: one word. → Luke 18:7: two. And 35:13, that the Lord will show no partiality against a poor man (our gloss), against Luke 20:21: one.

Each Hebrew main-text line of the range (215 lines; the 28 margin rows are left out, because their square brackets carry main-text words the transcriber supplied) against all 23,204 Masoretic verse units, single words not counted; a damaged, restored or queried word breaks a run. 110 lines share no two consecutive words with any verse of the Hebrew Bible; 92 reach two; 13 reach three; none reaches four. A line reported at 0 may still share single words. The manuscripts write the divine name differently from the Masoretic text; reading it as the Masoretic spelling lifts 2 lines from 0 to 2 and changes nothing at the top.

The table shows seven of the 13 lines that reach three. The other six reach three only with stock sequences of particles and common words — the nations which… not (30:19, band 6), who is this (31:9 and 31:10, band 3 each), upon all the nations (36:2, band 8), a man who… not (36:31 in MS B, band 2), and a man who has no and which has no (36:31 in MS D, band 5), our glosses — and are left out of the table.

Read across, not down. The last column is not a cross-language run; it is the Greek row screened against the Greek of the same verse. Six of the seven three-word Hebrew contacts do not recur on the Greek side — at Proverbs 19:11 the Hebrew line shares three words and the Greek row shares none. The seventh goes the other way: at 36:5 the Hebrew reaches three with Naaman’s confession in 2 Kings 5:15, while the Greek row 36:4 reaches four with it and five with 2 Samuel 7:22 and Isaiah 45:14. Against 2 Samuel 7:22 the MS B line (אין אלהים זולתך) stops at two, because the Masoretic text joins and to its word for there is no and the screen strips no prefixes. Two screens, two languages, two different pictures, and the page offers no explanation of the difference.

Hebrew: שבטי יעקב stands once in the Masoretic text (Isaiah 49:6) and in MS B’s 36:13; the line’s other half, בימי קדם, stands once too (Psalm 44:2, Masoretic numbering; 43:2 in the Greek). The whole line’s deepest run into the Masoretic text is two, reached by six verses: Isaiah 49:6, Psalm 44:2, and four that share כל שבטי, all the tribes. Greek: φυλὰς Ιακωβ stands at Isaiah 49:6, Sirach 36:10 and Sirach 48:10, and nowhere else screened; the 36:10 row runs three words into Isaiah 49:6 — one of five Greek Old Testament units it reaches three with, the other four sharing καθὼς ἀπ' ἀρχῆς, as from the beginning. Sirach 48:10, where Elijah is to restore the tribes, belongs to the page after next; its MS B line breaks off before the tribes are named.

• Any Greek manuscript. The displacement is established on this page only through Brenton’s edition: its English, and the Greek column of the Bagster printing, held only as poor OCR and read at the three joins. Which Greek text Bagster set beside the English, and whether it is the text Brenton translated, cannot be told from local files. The claim that every surviving Greek manuscript carries the displacement is the commentaries’, and is not verified here.
• Ziegler’s Göttingen edition, Beentjes, Skehan–Di Lella, Segal and the NETS Sirach were not consulted for this page. Every statement about them is reported.
• The Syriac. Not on this disk for Sirach. Its order is reported, not checked.
• Whether the Hebrew lines missing from the file are missing from the manuscripts. The file has no rows for chapters 29 and 34, none for Heb. 36:10–12 (CATSS 36:7–9), and Wikisource marks pages in this range unfinished. “Absent from the file” is all the page claims.
• The Masada scroll has no lines in this range and, in the file’s own words, no second source to check it against; it is not used here.
• Why the KJV and the Clementine carry the order they do. Measured: they do. Why: not something this disk can show.

• The thesis rests on one edition. Brenton’s edition is the only witness on disk to the other order: its English in two digitisations, and a Greek column read only through a poor OCR, at three places. If that edition’s arrangement were its editors’ and not the manuscripts’, the page would be describing a printed edition’s order — which is why every sentence about the manuscripts is marked reported.
• The alignments are by content. KJV to Greek: 218 rows onto 211 verses, with every merge and split listed in the versification note and anchored by the self-check on proper names and rare words. Hebrew to Greek in chapters 35–36: by content, anchored the same way, and the coverage figure of 119 rows depends on those alignments.
• The Hebrew is a wiki transcription, one editor’s, unpointed, with lacunae; a page in another range was vandalised in 2025 and is pinned to the revision before. Every Hebrew line printed here is compared character for character with the file, but the file has not been compared with the manuscripts by us.
• Every gloss of the Hebrew and of the prologue is ours and marked so; where a line was too obscure or too damaged to gloss honestly (30:25, most of 35:21) the page says so and does not.
• The run screens find verbatim contact only. Luke 18 is the page’s own demonstration of what that misses.
• A digital Greek Swete text sits in the corpus directory and was deliberately NOT used. It prints these chapters in the order used here, but its labels carry the other order’s numbers: after 33.16a it runs 33.25 to 33.40 — the sixteen numbers Brenton gives the same material in his chapter 30 — and it labels the cheerful heart 30.13b. A text re-ordered by someone, with the old numbers left in, is not a witness to either order.

• Skehan and Di Lella, The Wisdom of Ben Sira (Anchor Bible 39).
• Ziegler, Sapientia Iesu Filii Sirach (Göttingen Septuagint XII/2).
• Beentjes, The Book of Ben Sira in Hebrew (Brill, 1997).
• Schechter and Taylor, The Wisdom of Ben Sira (Cambridge, 1899).
• Brenton, The Septuagint Version of the Old Testament (1851 is the date eBible.org gives; the file also carries an Apocrypha introduction from a later Bagster printing of about 1900, so which printing it reproduces for Sirach cannot be told from local files) — held here as the public-domain verse-per-line edition from eBible.org; used for its order only.
• The Septuagint Version of the Old Testament and Apocrypha, with an English Translation (London: Samuel Bagster, catalogued 1900) — Greek and English; held here only as the Internet Archive OCR text; read at the three joins, not measured.

• Ask which order. “Sirach 33” is two different passages depending on whose numbering you are reading. Before you repeat a reference to these chapters, find out whose it is.
• The one sentence CATSS, the KJV, the Latin and MS B all keep whole is a prayer that the scattered be gathered. “Gather all the tribes of Jacob together”. In the other order it is broken off mid-petition; in Hebrew it fits on one line.
• The obvious links measure one word; the measured ones are quiet. The potter and Romans 9: one. Treasure and Matthew 6: one. The two runs unique on both sides are a phrase about good pleasure and a phrase about moving from house to house. Measure before you repeat.
• The Greek reads the lowly; the Hebrew reads the poor. Two witnesses to one line, and the page lets both stand.
• One concrete response. Sirach 29:9 does not name its commandment. The Torah’s own command to open the hand and lend to a poor brother is Deuteronomy 15:7–11, which is Scripture for us; “Help the poor for the commandment’s sake” is a witness to how that kind of command was being heard before Christ. Lend to someone in need this week without setting a day to be repaid.