The Last Word Is Children

• Anchor Bible Dictionary — “Ezra, Book of,” “Esdras, First Book of,” “Marriage (OT),” “Foreigner.”
• Dictionary of the OT: Historical Books (IVP) — “Ezra and Nehemiah, Books of,” “Marriage and Divorce,” “Ethnicity,” “Persian Period.”
• NIDOTTE — נָכְרִי, מָעַל, יָשַׁב (hiphil), בָּדַל, גֶּשֶׁם.
• Commentary on the New Testament Use of the Old Testament (Beale & Carson) — on 1 Corinthians 7:12–16, which is the New Testament passage most often set against this chapter.

Within the book, the three columns agree verse for verse.

The Hebrew and the Greek columns were counted off the files on this disk, not copied from a table; no ESV file of Ezra is held here, and the ESV row is the standard Masoretic division. So every reference on this page means the same verse in all three columns.

Outside the book, two conversions are needed and both catch people out.

The mapping was checked verse by verse against both Greek files: 1 Esdras 8:88 is Ezra 10:1 (while Ezra prayed… there gathered to him) and 1 Esdras 9:36 is Ezra 10:44 (all these…). Thirty-nine verses of Ezra 10 correspond to thirty-six of 1 Esdras 9, because 1 Esdras compresses the roll of names.

1 Esdras is not Scripture on this shelf. It is quoted here the way Josephus or Sirach is quoted — as a witness to how the story was told in Greek, and because it is the only other continuous Greek account of this chapter held on this disk. Josephus retells the same events in Antiquities 11 and was not consulted (stage 5).

All 44 rows of 2 Esdras 10 were screened against all 55 rows of 1 Esdras 9. This is Greek against Greek, both from CATSS, with the same normalisation on both sides.

★ And the deepest agreement in the chapter breaks off on the word foreign.

2 Esdras 10:11: … ἀπὸ λαῶν τῆς γῆς καὶ ἀπὸ τῶν γυναικῶν τῶν ἀλλοτρίων
1 Esdras 9:9: … ἀπὸ τῶν ἐθνῶν τῆς γῆς καὶ ἀπὸ τῶν γυναικῶν τῶν ἀλλογενῶν

Seven words identical — τῆς γῆς καὶ ἀπὸ τῶν γυναικῶν τῶν — and the run is fenced at both ends by a lexical choice: peoples against nations at the front, another’s against of another race at the back. The one place the two Greek Ezras say the same thing at length is the place where they stop being able to say the same thing at all.

What this does not show. Nothing here says which Greek version is earlier, or that either translator saw the other, or that either is closer to the Hebrew. Two translations of related Semitic material will diverge in vocabulary as a matter of course. What is reportable is where they diverge, and it is at the word the whole chapter turns on.

The split is total, and it runs across whole books.

All seven occurrences of ἀλλότριος in the Greek Ezra fall in this one chapter (10:2, 10, 11, 14, 17, 18, 44). 1 Esdras never uses that word for the women; it uses ἀλλογενής, which is a word about descent, eleven times across 8:66–9:36.

★ And ἀλλογενής occurs exactly once in the Greek New Testament. Luke 17:18 — οὐχ εὑρέθησαν ὑποστρέψαντες δοῦναι δόξαν τῷ θεῷ εἰ μὴ ὁ ἀλλογενὴς οὗτος, Was no one found to return and give praise to God except this foreigner? The word 1 Esdras uses eleven times for the women who were sent away is the word the New Testament uses once, for the one leper who came back. Screened over all 7,927 verses of the Greek New Testament: one hit, and that is it. This is a word count, not a citation, and the page does not make it one. Neither writer is quoting the other and no direction of influence is claimed.

⚠️ Not measured here. The same word ἀλλογενῆ is reported to stand on the stone barrier inscriptions from the Jerusalem temple, warning outsiders not to pass. Those inscriptions are not among the texts held on this disk and were not checked. They are named because a reader will meet the claim elsewhere, and flagged because this page did not verify it.

Ezra 10:44 is the last verse of the book, and it is difficult in the Hebrew — difficult enough that the Masoretes flag the first verb with a ketiv and a qere.

Measured: the two Greek endings agree on two consecutive words — πάντες οὗτοι, all these — and on nothing else. Different verb for the marrying, different word for the women, and a completely different last clause. One text ends with children born; the other ends with children sent away.

What can and cannot be said about that. What can be said: the Hebrew does not report the outcome, and 1 Esdras does. What cannot be said from these files: which reading is older, whether 1 Esdras is interpreting a difficult Hebrew clause or preserving a different one, or what actually happened to the women and children. No critical apparatus was consulted for either text.

The one thing all three do. Every column ends the book on the children. The last word of the Hebrew is בָּנִים; of the Greek row, υἱούς; of 1 Esdras, τέκνοις; of the ESV, children. They disagree about what happened to them and agree about who the book is left looking at.

Ezra 10:6 is the fasting verse. Both Greek versions have it:

2 Esdras 10:6: ἄρτον οὐκ ἔφαγεν καὶ ὕδωρ οὐκ ἔπιεν
1 Esdras 9:2: ἄρτου οὐκ ἐγεύσατο οὐδὲ ὕδωρ ἔπιεν

Longest shared run: one word. Different case for the bread, a different verb for the not-eating, a different negation. The two verses do share οὐκ, ὕδωρ and ἔπιεν — but no two of them stand together, in that order, in both. This is the same sentence about the same man on the same night, and as word forms the two have no two consecutive words in common. It is the clearest demonstration on this page of what a run measurement can and cannot see: the screen finds verbatim contact, not meaning.

Counted on the consonantal text of both chapters:

The two chapters are one continuous scene — 10:1 picks up the prayer of 9:6–15 mid-motion — and the narrator changes person across the join. The page reports the seam and does not theorise about it; whether it marks a source, a compiler, or a stylistic convention is not a question a word count can answer.

Ezra 10:14 contains הַהֹשִׁיב and לְהָשִׁיב. Strip the vowels and they are the same four consonants.

The first is the hiphil of ישב, to settle — everyone in our cities who has settled foreign women. The second is the hiphil of שוב, to turn back — until the fierce wrath of our God is turned back from us. Nehemiah 13:27 uses the identical consonantal form להשיב in the first sense.

The Greek row has to choose and does: ἐκάθισεν for the one, ἀποστρέψαι for the other. Whether the Hebrew is playing on it is not something a screen can decide, and this page does not claim it is. What is checkable is that the forms coincide, in this verse, once.

In Ezra 9:3 and 9:5 Ezra the priest tears his בֶּגֶד, his garment. In Malachi 2 — the same century, the same complaint — the verb used five times for the faithlessness that chapter is about — once of a man against his brother (2:10), then of Judah (2:11) and of the wife of a man’s youth (2:14–16) — is בָּגַד, to deal treacherously (Malachi 2:10, 11, 14, 15, 16). Same three consonants.

They are treated as separate roots by the lexicons, and this page treats them that way. It is noted because a reader running the same search will find it, and will want to be told plainly that a shared consonantal skeleton is not an etymology.

Ezra 9:1–2 names five of the seven nations of Deuteronomy 7:1 — and three more besides — and speaks of a holy seed. Chapter 10 is the chapter that acts on it. Screened Hebrew against Hebrew, verse by verse:

Ezra 10 does not quote Deuteronomy 7. It does not share a single two-word sequence with it. Deuteronomy 7:3 says you shall not intermarry with them; you shall not give your daughter to their son, nor take their daughter for your son. Ezra 10:3 says let us make a covenant with our God to put away all these wives and their children. The two verses share no word at all. The chapter invokes the Law — let it be done according to the Law, 10:3 — without reproducing any of it.

And the Nehemiah result is the one that should surprise a reader most. Ezra 10 and Nehemiah 13:23–27 are the pair every commentary sets side by side. Screened verse against verse across the whole of Nehemiah 13, they share twenty-one two-word sequences and one three-word run, and nothing longer. Only eight of the twenty-one are about the marriages: נָשִׁים נָכְרִיּוֹת six times, the definite הַנָּשִׁים הַנָּכְרִיּוֹת once (10:11 / 13:26) and הֹשִׁיבוּ נָשִׁים once (10:18 / 13:23). The other thirteen are formulas — בֵּית הָאֱלֹהִים nine times, עַל־זֹאת twice, בֶּן־אֶלְיָשִׁיב and אֲשֶׁר לֹא once each — and the three-word run is about something else. Two accounts of one crisis, and no shared sentence between them. They also do not do the same thing: Nehemiah 13:25 records a beating and an oath, and no wives being sent away.

The deepest Hebrew run in the chapter is at 10:6, against 23,160 other verse units of the Masoretic corpus.

Exodus 34:28 is Moses on the mountain for forty days, neither eating bread nor drinking water, writing the words of the covenant on the tablets — the covenant renewal, after the golden calf. Ezra 10:3 proposes to cut a covenant, and the six words Ezra 10:6 uses for its fast are the six Exodus 34:28 uses for the man who wrote one. Which way that runs, if it runs either way, a screen cannot say and this page does not.

⚠️ And here is the counter-example this page owes the reader. A tie band of one makes that run look unique. It is unique as word forms and not as a sentence. Searched for the idea rather than the forms, five verses in the Hebrew Bible report that a man ate no bread and drank no water:

Three of the four others are Moses; the fourth is an Egyptian slave abandoned sick in open country and found by David’s men. Two are outside the band only because Deuteronomy tells it in the first person, and 1 Samuel 30:12 only because it says the same thing in a different word order, and moves the waw — it reads 2 against Ezra 10:6. (The man of God of 1 Kings 13 is under a standing command not to eat bread or drink water, repeated through that chapter; it shares one word with Ezra 10:6.) The band of one is a fact about grammar, not about rarity of thought — and a page that printed the six-word run without this table would be overselling it.

All 44 CATSS rows of 2 Esdras 10 against all 7,927 verses of the Greek New Testament.

Across the 44 rows: 41 reach one, 37 reach two, 19 reach three, 3 reach four, and nothing reaches five.

Both of the chapter’s tie-band-of-one runs land on the same New Testament verse, and both are formulas. Luke 9:45 reads οἱ δὲ ἠγνόουν τὸ ῥῆμα τοῦτο… καὶ ἐφοβοῦντο ἐρωτῆσαι αὐτὸν περὶ τοῦ ῥήματος τούτου — it says this saying twice, in two different cases, which is exactly why it catches two different rows of a chapter whose own refrain is this matter. A formula meeting a formula. There is no connection here and the page asserts none.

⚠️ The counter-example is in the third line of the table. Ezra 10:10 reaches the same depth of four, and its band is seven — every one of them Luke, because καὶ εἶπεν πρὸς αὐτούς is Lukan boilerplate. Depth without a band is not evidence, which is why this shelf prints the band beside every run.

Paul’s instruction to a mixed marriage is the passage most readers reach for against this chapter, so this page measured the contact rather than assuming it.

1 Corinthians 7:12: εἴ τις ἀδελφὸς γυναῖκα ἔχει ἄπιστον, καὶ αὕτη συνευδοκεῖ οἰκεῖν μετ’ αὐτοῦ, μὴ ἀφιέτω αὐτήν — if any brother has a wife who is an unbeliever, and she consents to live with him, he should not divorce her. And 7:14 ends: ἐπεὶ ἄρα τὰ τέκνα ὑμῶν ἀκάθαρτά ἐστιν, νῦν δὲ ἅγιά ἐστιν — otherwise your children would be unclean, but as it is they are holy.

There is no verbal contact. None. The pairing is thematic, and the page says so plainly rather than dressing it as a link. What is true is that both passages end on the children of a mixed marriage, and say opposite things about them. What is not shown here is that Paul had this chapter in view. A run of one word is the same result you would get from any two Greek sentences picked at random.

Malachi 2:10–16 belongs in the same conversation and does not help either side tidily. It condemns marrying the daughter of a foreign god (2:11) and dealing faithlessly with the wife of your youth (2:14–15), in the same breath, in the same generation. Its verb for the offence is בָּגַד; Ezra 10’s is מָעַל. Two books facing one crisis with two different vocabularies for it.

Twenty-six of the chapter’s forty-four verses are the roll of names, and three places in it can be seen to differ with no apparatus at all:

The roll names 111 men — counted off the consonantal text, taking the name after each of the sons of formula as the family head, treating Kelaiah (that is, Kelita) in 10:23 as one man, and reading 10:38 as of the sons of Binnui: Shimei. Read 10:38 the other way — Bani, Binnui and Shimei as three men — and the total is 113. The page prints both because the verse is a known crux and a single number would be a false precision.

And eighteen of the forty-four rows — every one of them inside the roll — share no two consecutive words with any other verse in the Hebrew Bible. (On the Hebrew screen a zero means exactly that; the individual names do occur elsewhere.) Most of this chapter is text that exists nowhere else.

No critical apparatus was consulted, for any of the three texts. Not for the Hebrew, not for 2 Esdras, not for 1 Esdras. Where this page reports that the columns differ — 10:15, 10:25, 10:37–38, 10:44 — it is reporting what the printed editions on this disk say, and nothing about manuscript support, age or probability.

The relationship between 1 Esdras and 2 Esdras is a live scholarly question and this page does not enter it. Whether 1 Esdras preserves an older form of the material, or is a later reworking, or is a fragment of a lost longer work, has been argued in every direction for a century. The measurements here are indifferent to the answer: they say how much of the wording the two share, not which came first.

The 111 is soft in the third digit. The count depends on how 10:38 is read, on treating Kelaiah and Kelita as one man, and on the two Malchijahs of 10:25. The page gives the rule it counted by and the alternative figure. A reader who counts differently will get 112 or 113 and will not be wrong.

Ezra 10:15 is genuinely undecided. עָמְדוּ עַל־זֹאת can be read as opposition or as taking charge; the ESV takes it as opposition, and both Greek witnesses take it the other way. This page cannot settle it and does not pretend the Greek settles it either — a translator faced with an ambiguous idiom makes a choice, and a choice is not evidence about the Hebrew.

And the hardest limit is the one that is not textual at all. What happened to those women and those children after the first day of the first month is not in the Hebrew, is not in the Greek row, and is asserted in exactly one sentence of 1 Esdras. The book of Ezra does not tell us, and this page will not fill the gap.

No Qumran material — no published Dead Sea manuscript preserves any part of Ezra 10, and none was consulted. No Targum (there is no classical Targum to Ezra–Nehemiah), no Syriac, no Vulgate, no Old Latin of 1 Esdras. No Josephus, whose Antiquities 11 follows 1 Esdras rather than the Hebrew at several points and which would be the obvious next witness to read. No Elephantine papyri, which are the closest documentary evidence for how a Judean community outside Jerusalem in this same century actually handled marriage.

No temple-barrier inscription (see stage 2). No rabbinic material. No archaeology of Persian-period Yehud, whose population estimates bear directly on the question of what 111 households meant.

And nothing on this page is an ethical judgement. The shelf’s job here was to measure what three texts say and where they stop saying the same thing. The chapter is reported without endorsement and without explaining it away, which is the same position the companion page on Ezra 7–10 takes.

• H. G. M. Williamson, Ezra, Nehemiah (Word Biblical Commentary 16).
• Joseph Blenkinsopp, Ezra–Nehemiah: A Commentary (Old Testament Library).
• Lester L. Grabbe, Ezra–Nehemiah (Old Testament Readings).
• Zipora Talshir, I Esdras: From Origin to Translation (SBL Septuagint and Cognate Studies 47).
• Tamara Cohn Eskenazi, “Out from the Shadows: Biblical Women in the Postexilic Era,” JSOT 54 (1992).
• Katherine E. Southwood, Ethnicity and the Mixed Marriage Crisis in Ezra 9–10 (Oxford).
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer.

The chapter opens with a crowd — men, women and children — gathering around a man lying face down in front of the house of God, and it ends twenty-six verses of names later with a sentence about children. In between it is dated four times and it rains through all of it.

Three things this page found that it did not expect to. One: the chapter shares no two-word sequence with Deuteronomy 7, the law it says it is obeying — and the deepest run it does have with the Torah is a fasting formula, six words long, from the covenant renewal after the golden calf. Two: where the two Greek versions of this chapter agree most deeply, seven words at verse 11, the run breaks off exactly on the word foreign — one says another’s, the other says of another race. Three: the last verse ends the book three different ways, and the two Greek endings have two words in common.

And then the thing they all do agree on. The Hebrew’s last word is בָּנִים. The Greek row’s is υἱούς. 1 Esdras ends σὺν τέκνοις. The ESV ends children. Whatever else the texts do with this chapter, not one of them lets you close the book without looking at the children.

This page closes the book of Ezra on this shelf. Chapters 1–6 are at Ezra 1:1–6:22 and 7–10 at Ezra 7:1–10:44; this study takes the last chapter on its own, because it needed three texts side by side and one of them is not in the Bible.