The Ring Closes at the Sheep Gate

• Anchor Bible Dictionary — “Nehemiah,” “Cupbearer,” “Susa,” “Artaxerxes,” “Jerusalem (Walls and Gates),” “Sanballat,” “Tobiah,” “Geshem,” “Wadi Daliyeh,” “Esdras, Second Book of.”
• Dictionary of the OT: Historical Books (IVP) — “Ezra and Nehemiah, Books of,” “Nehemiah,” “Prayer,” “Jerusalem,” “Persia, Persians,” “Fortifications,” “Lists.”
• NIDOTTE / TDOT — חָזַק (ḥāzaq / hiphil heḥəzîq, take hold, reinforce, repair), פָּרַץ (pāraṣ, breach), פָּלַל (hiṯpallēl, pray), לֵב (lēḇ, heart), מַשְׁקֶה (mašqeh, cupbearer).
• On the Greek — the NETS introduction to 2 Esdras (Esdras B), for the continuous numbering across Ezra and Nehemiah and the character of the translation. Anyone looking up “Nehemiah” in a Greek concordance and finding nothing is looking under the wrong name.

• H. G. M. Williamson, Ezra, Nehemiah (WBC); Joseph Blenkinsopp, Ezra–Nehemiah (OTL); Lester L. Grabbe, Ezra–Nehemiah. Critical scholarship: the philological and topographical work is used and the historical conclusions are weighed, not adopted wholesale.
• On the topography of chapter 3, any recent survey of Persian-period Jerusalem — read knowing that the line of the wall and the extent of the city are genuinely disputed. See Stage 5.

Cupbearer is an access position, not a serving job. Across the ancient Near East the man who tasted the king’s wine had to be trusted absolutely and stood physically closest to the monarch. Tobit 1:22 has Ahikar holding the office together with the signet and the accounts. 1:11 therefore explains the mechanics of the entire book — and it also explains 2:2. A man whose face the king reads daily cannot hide four months of grief.

Chapter 3 is shaped like an administrative document, and the Hebrew shows it. Its spine is a repeated formula — next to him / next to them… repaired… — with a name, a patronymic or a town, and a stretch of wall. The verb הֶחֱזִיק carries thirty-four of those entries and does nothing else in the book. This is what a survey looks like when it is written down: labour assigned by household and guild against measured sections, which is how large works were organised under the Persian administration and how the Chronicler records temple duty by course.

The neighbours are documented officials. Sanballat appears as governor of Samaria in the Elephantine papyri, and the family recurs across generations in the Wadi Daliyeh documents. Tobiah’s name and family are plausibly connected with the later Tobiads in Ammon; Geshem fits the Qedarite chieftains named in Arabian inscriptions. That is why the first attack in 2:19 is a legal charge and not a raid. The external evidence establishes that these were real provincial administrators. It does not corroborate the events of the narrative, and this page does not claim that it does.

• The Elephantine papyri (Aramaic documents from Persian Egypt — witnesses, not Scripture) — including the reference to the sons of Sanballat, governor of Samaria.
• The Wadi Daliyeh papyri (Samarian administrative documents — witnesses, not Scripture) — the Sanballat family in the Samarian governorship over several generations.
• Arabian and Assyrian inscriptions on Qedar (royal and votive texts — witnesses, not Scripture) — for Geshem the Arab and what an Arab chieftain’s standing amounted to.
• Tobit 1:22 (Second-Temple; Apocrypha — a witness, not Scripture in our confession) — Ahikar as royal cupbearer, keeper of the signet, administrator. Cited only for the standing of the office.
• Josephus, Antiquities 11.5.6–8 (§§159–183) (Second-Temple — a witness, not Scripture) — his retelling of the mission. He gives the wall two years and four months. A retelling, and a reminder that ancient readers also found the fifty-two days of 6:15 striking.

• The Greek does not vary its rendering of the chapter’s verb nearly as much as one might expect. Counted in 2 Esdras 13:1–32: ἐκράτησεν/ἐκράτησαν twenty-eight times and κατέσχεν/κατέσχοσαν four — thirty-two occurrences of two verbs that both mean took hold of, against the Hebrew’s thirty-four. The shortfall is accounted for: 3:7 is absent from the Greek altogether, and the columns differ over the verb at 3:8, 3:15 and 3:19. The drumbeat survives translation.
• Three column differences in this passage that a reader should see rather than be spared. 2:1 — Hebrew “I had not been sad in his presence,” Greek “there was no other in his presence.” 2:12 — Hebrew “to do for Jerusalem,” Greek “to do with Israel.” 3:14 — Hebrew “he rebuilt it,” Greek “he and his sons, and they roofed it.” All three are printed in the codex with a translation note. None of them is adjudicated here, and no claim is made about which text is prior.
• A ketiv/qere at 1:9. The word rendered “and bring them” is written והבואתים and read וַהֲבִיאוֹתִים. The read form is printed. The whole Hebrew corpus on this shelf is screened on the qere, and the difference here is orthographic.

• Nehemiah 1:1 ↔ Nehemiah 1:11 — the chapter frames itself. Both verses end וַאֲנִי הָיִיתִי… The phrase occurs seven times in the whole Hebrew Bible: Psalm 109:25; Daniel 8:2, 8:5, 10:4, 10:9; and Nehemiah 1:1 and 1:11. Two of the seven are the first and last verses of this chapter, and the Greek shows the same frame (καὶ ἐγὼ ἤμην, six occurrences in the 30,348 Greek Old Testament verse units screened here, two of them 2 Esdras 11:1 and 11:11). This is the page’s structural claim about chapter 1, and it is a count, not an impression.
• Nehemiah 1:5 ↔ Daniel 9:4 — five consecutive words, tie band 1. הָאֵל הַגָּדוֹל וְהַנּוֹרָא שֹׁמֵר הַבְּרִית, “the great and awesome God who keeps the covenant.” The deepest contact anywhere in Nehemiah 1, and it is with the other great exilic penitential prayer. Band 1 means nothing else in the Hebrew Bible reaches that depth against this verse. Direction of influence is not asserted and cannot be measured.
• Nehemiah 1:7 ↔ Deuteronomy 7:11 — five words, tie band 1. “…and the statutes and the rules that…” The confession is built out of Deuteronomy’s own vocabulary, which is what 1:8 says it is doing: “remember the word that you commanded your servant Moses.”
• Nehemiah 1:9 ↔ Deuteronomy 30:4 — three words. בִּקְצֵה הַשָּׁמַיִם מִשָּׁם, “in the uttermost parts of heaven, from there.” Nehemiah is praying that promise back to God in Susa — about as far east as an Israelite had been taken.
• Nehemiah 1:11 ↔ 1 Kings 8:28 and Daniel 9:17 — four words, tie band 3. “…to the prayer of your servant, and to…” Solomon’s temple dedication language, used a thousand miles from the temple. The band of 3 is the honest reading: this is shared liturgical idiom, not a quotation.
• Nehemiah 2:5 ↔ Esther 8:5 — five words, tie band 1; Nehemiah 2:7 ↔ Esther 1:19 and 3:9 — four words, tie band 5. The court language of chapter 2 measures against the Persian court book, not against prophecy.

Nothing in Nehemiah 3:1–32 reaches four. Against 23,172 other verse units of the Masoretic corpus, the deepest run anywhere in the builders’ list is three consecutive words — seven verses reach it, all of them on personal names (Eliashib the high priest, Meremoth son of Uriah, Meshullam son of Berechiah). Chapter 1 reaches five twice; chapter 2 reaches five once. The building list is the flattest block in the passage.

And four of its verses share no two consecutive words with any other verse in the entire Hebrew Bible: 3:22, 3:24, 3:27 and 3:32. On the Hebrew-side convention used here (unigram_fallback=False) a zero means exactly that and no more — no two consecutive words. All four verses do share individual words with plenty of other verses. Two of the four (3:24, thirteen words; 3:27, twelve) are of ordinary length for the chapter, whose mean is about sixteen; 3:22 is a five-word verse and the zero there is worth less.

Now the counter-example, printed so the zeros are not read as more than they are. Twenty-one verses of Nehemiah 3:1–32 max out at a run of two, and the tie bands on those twos are enormous: 3:9 has a band of 23, 3:8 a band of 19, 3:29 a band of 15. The shared text is וְעַל־יָדָם and עַל־יָדוֹ — “and next to them,” “next to him.” That is a formula, not a connection. The same screen that produces the four striking zeros also produces twenty-one meaningless twos, and both come out of the same run of the same tool. This is why the shelf prints the band.

Across the 67 CATSS rows of 2 Esdras 11–13 (= Nehemiah 1–3) against all 7,927 verses of the Greek New Testament: all 67 reach two, 36 reach three, 11 reach four, one reaches five, and nothing reaches six. For scale, the whole of Greek Ezra–Nehemiah — 672 rows — tops out at six, once. The New Testament does not quote Nehemiah 1–3 anywhere, and the screen is consistent with that.

The single five is 2 Esdras 13:14 (Nehemiah 3:14) against John 4:12, tie band 1 — καὶ οἱ υἱοὶ αὐτοῦ καί, “and his sons, and.” Band 1 makes it look like a signal. It is not, and there are two independent reasons to say so.

First, three of those five words are not in the Hebrew. Masoretic Nehemiah 3:14 reads “he rebuilt it”; the Greek reads “he and his sons, and they roofed it.” The page’s deepest measured New Testament contact rests entirely on a Greek plus. Second, the words themselves are the commonest possible Greek. John 4:12 is the Samaritan woman at the well — “our father Jacob… drank from it himself,” and the Greek runs on καὶ οἱ υἱοὶ αὐτοῦ καὶ τὰ θρέμματα αὐτοῦ — and his sons and his livestock. There is no thematic contact with a man repairing the Dung Gate. It is a formula colliding with a formula, and it is worth printing precisely because it looks like more than it is. (The irony is free: the deepest verbal contact between the Nehemiah material and the New Testament falls in the mouth of a Samaritan, in a book whose chief antagonist is the governor of Samaria. That is an observation, not an argument.)

The next band down shows the same thing. Eleven rows reach four. Some have band 1 — 2 Esdras 11:10 against Revelation 14:14 (καὶ ἐν τῇ χειρί), 11:11 against John 17:26 (τὸ ὄνομά σου καί), 12:8 against Revelation 21:14 (τὸ τεῖχος τῆς πόλεως, “the wall of the city”). And one has a band of 16: 2 Esdras 11:4 shares καὶ ἐγένετο ἐν τῷ with sixteen verses, fifteen of them in Luke. Same depth, and one of them is Lukan style and nothing else.

• Nehemiah 3:1, 3:32 → John 5:2 — the Sheep Gate. προβατικός occurs three times in the 30,348 Greek Old Testament verse units screened here, all in 2 Esdras, two of them the frame of this chapter; and once in the Greek New Testament, at the pool where a man had lain thirty-eight years. Shared word, different inflections, and a measured run of zero — no two consecutive words in common. The link is lexical and thematic; this page claims no allusion and no citation. What it will say is that the gate the sheep came through for sacrifice is the first thing the high priest rebuilds and consecrates, and the last thing the list names.
• Nehemiah 1:5–11 → Daniel 9:4–19; Ezra 9:6–15; Deuteronomy 30:1–5. The family of long confessional prayers that quote the covenant back to God. Measured contact: five words with Daniel 9:4 (band 1), five with Deuteronomy 7:11 (band 1), three with Deuteronomy 30:4. A shared prayer tradition, demonstrated by measurement rather than asserted.
• Nehemiah 1:11 → Genesis 40:1–23; Tobit 1:22. The cupbearer’s office. Joseph read a cupbearer’s dream in a prison; Ahikar held the post in Assyria (Apocrypha, a witness). Proximity to the king is the mechanism of this book.
• Nehemiah 2:5, 2:7 → Esther 1:19; 3:9; 8:5. “If it pleases the king.” Five words with Esther 8:5 at band 1; four with Esther 1:19 and 3:9 at band 5. Two Judeans in two Persian palaces using the same submitted formula — and both of them asking for the life of their people.
• Nehemiah 3:12 → Numbers 27:1–11; Judges 5:7. Shallum’s daughters on the wall; Zelophehad’s daughters winning an inheritance ruling; Deborah. The Greek keeps the daughters (καὶ αἱ θυγατέρες αὐτοῦ) even where, two verses later, it supplies sons the Hebrew does not have.
• Nehemiah 3:5 → Matthew 21:28–31; Philippians 2:5–8. “Their nobles would not stoop to serve their Lord” — the one criticism in the whole list, and it is about stooping. Named as a resonance and nothing more: there is no verbal contact between these texts, and none is claimed.

• H. G. M. Williamson, Ezra, Nehemiah (WBC) — the standard treatment; especially careful on chapter 3 and on the relation of the Nehemiah memoir to the rest of the book.
• Joseph Blenkinsopp, Ezra–Nehemiah (OTL) — strong on the Persian provincial setting and on Sanballat, Tobiah and Geshem.
• Lester L. Grabbe, Ezra–Nehemiah — on the historical questions, including the size and population of Persian-period Jerusalem.
• The NETS introduction to 2 Esdras — on the character of the Greek translation, which is notoriously literal and which is why the Hebrew drumbeat of chapter 3 survives into Greek almost intact.

• ⚠️ No Hebrew manuscript evidence was consulted beyond the Masoretic text held here. The three column differences reported at 2:1, 2:12 and 3:14 are differences between one Hebrew text and one Greek text on this disk. Whether a Qumran fragment, a Vorlage reconstruction, or another Greek witness bears on any of them was not checked and cannot be checked from these corpora. The plus at 3:14 — which carries this page’s deepest New Testament run — may be a translator’s expansion, a different Hebrew text, or a harmonisation toward 3:12. This page does not know which, and says so rather than choosing.
• ⚠️ The Greek is CATSS/Rahlfs and nothing else. No apparatus was read; no Lucianic or Origenic recension was compared; the Alexandrinus and Vaticanus readings behind Rahlfs’s text were not weighed. “Absent from Rahlfs” is a statement about one printed edition, not about the Greek tradition. 3:7 may well stand in other Greek witnesses.
• ⚠️ The run screen finds verbatim contact, not allusion. It cannot see a thought reused in different words, and it cannot see word order that has shifted. Everything reported above as a “connection” with no run attached — the Sheep Gate and John 5, the nobles who would not stoop and Philippians 2 — is an editorial judgement, and is labelled as one.
• ⚠️ The four zeros in chapter 3 are a property of the corpus screened, not of the universe. They are measured against the 39-book Masoretic corpus only. They say nothing about Ben Sira, Qumran, or any other Hebrew outside that canon.
• ⚠️ The topography of chapter 3 is genuinely disputed and this page does not attempt it. Which gate stood where, how much of the Persian-period city the circuit enclosed, and how the list maps onto excavated remains are all contested. Estimates of the population have moved by an order of magnitude in living memory. The page reports the shape of the list, not the shape of the wall.
• ⚠️ The relative order of Ezra and Nehemiah is unresolved — it turns on which Artaxerxes is meant. See the Ezra 7–10 study. The canonical order is followed here without any claim that it settles the history.
• ⚠️ The identification of the Nehemiah memoir is a critical reconstruction. That chapters 1–7 and parts of 12–13 derive from a first-person source is widely held; its original genre — royal report, dedicatory memorial, self-defence — is argued. Chapter 3 is in the third person and reads like an archival document, which has led many to treat it as a source incorporated rather than composed. That is an observation about the material and not a judgement on its accuracy, and this page takes no position.
• ⚠️ ESV footnotes and textual apparatus were not used to typeset the English. The English column is the ESV as quoted under Crossway’s standing permission; where the Hebrew and Greek differ, the codex prints both columns and a translation note rather than following either version silently.

• He prayed for four months before he asked for anything. The two dates are the argument: Chislev in 1:1, Nisan in 2:1, the same twentieth year. And when the moment came, the request had a duration, named officials, and a specific quantity of timber. Prayer and preparation were not two activities.
• The most important sentence in chapter 1 is the last four words. “Now I was cupbearer to the king.” He does not lead with his access; he leads with his grief. But the access is real, and he uses it. Where you already stand is not an accident to be apologised for.
• The chapter people skip is the chapter that is measurably unlike anything else. Nothing in Nehemiah 3:1–32 shares more than three consecutive words with any other verse in the Hebrew Bible; four of its verses share not even two. That is what a list of actual named people doing actual specific work looks like when you measure it. It cannot be generalised because it is not general.
• One concrete response. The ring closes because party after party each took one section — thirty-four entries between 3:4 and 3:32, and the gates before them. A goldsmith did not become a mason; he took a stretch of wall. Somebody’s daughters took theirs. The only people criticised in the whole list are the ones who would not stoop. Find the section next to you this week and take hold of that one — which is what the verb in this chapter actually means.