Twenty-Three Words, and One

• Anchor Bible Dictionary — “Luke, Gospel of,” “Centurion,” “Nain,” “Widow,” “Anointing,” “Pharisees,” “Parable,” “Gerasa,” “Legion,” “Demon.”
• Dictionary of Jesus and the Gospels (IVP) — “Faith,” “Forgiveness of Sins,” “Women,” “Table Fellowship,” “Rich and Poor,” “Parable,” “Elijah and Elisha.”
• New International Dictionary of NT Theology and Exegesis — πίστις, σῴζω, ἁμαρτωλός, ἀγαπάω, ὑπομονή, ἐπισκέπτομαι, μυστήριον.
• Dictionary of NT Use of the OT (Beale & Carson) — on Luke 7:22, 7:27 and 8:10, which are the three places in this passage where the question of an Old Testament source is live.

This shelf states the versification offset on every page, including the pages where there is none. This is one of them.

Chapters 7 and 8 are identical in both columns, so every Luke reference on this page is a single number. Across the Gospel as a whole, two numbers carry no text in the SBLGNT — 17:36 and 23:17 — and the ESV prints both in footnotes rather than in the text. The two agree on which verses exist; they differ only in whether an empty number is printed. Neither is in this passage.

The offsets that do bear on this page are on the Old Testament side, and they are all aligned. Exodus 23:20, Genesis 37:10, Isaiah 6:9, Isaiah 49:13 and Malachi 3:1 carry the same number in the Masoretic text, in the CATSS Greek and in the ESV. The one book-name difference is 3 Kingdoms 17:23 = 1 Kings 17:23: the Greek names the book 3 Kingdoms where the English names it 1 Kings, and chapter 17 runs to twenty-four verses in both. A warning about Malachi, because it is a real trap: the Hebrew and the Greek both end at chapter 3, and the ESV prints a chapter 4. ESV Malachi 4:1–6 = MT and Greek Malachi 3:19–24. The verse this page cares about, Malachi 3:1, is 3:1 in all three columns.

Luke 7:27 says περὶ οὗ γέγραπται, of whom it is written, and then quotes. Here is what the Greek columns held on this shelf actually say, measured rather than assumed.

The two Old Testament verses as this shelf holds them:

Exodus 23:20 — καὶ ἰδοὺ ἐγὼ ἀποστέλλω τὸν ἄγγελόν μου πρὸ προσώπου σου ἵνα φυλάξῃ σε ἐν τῇ ὁδῷ…
Malachi 3:1 — ἰδοὺ ἐγὼ ἐξαποστέλλω τὸν ἄγγελόν μου καὶ ἐπιβλέψεται ὁδὸν πρὸ προσώπου μου…

Three differences do the whole job, and all three can be pointed at. (1) Malachi has the compound ἐξαποστέλλω where Exodus and all three Gospels have the simple ἀποστέλλω. (2) Malachi reads πρὸ προσώπου μου, before my face; Exodus and all three Gospels read πρὸ προσώπου σου, before your face. (3) Malachi puts καὶ ἐπιβλέψεται ὁδόν between the messenger and the face, which breaks the sequence outright.

And the second half of Luke 7:27 belongs to neither column. ὃς κατασκευάσει τὴν ὁδόν σου ἔμπροσθέν σου — who will prepare your way before you — matches Exodus 23:20 at zero and Malachi 3:1 at zero as a sequence. It has Malachi’s way and Exodus’s your, and the verb is in neither.

⚠️ What this page does not do. It does not call Luke’s citation wrong, does not say the quotation “is really” Exodus, does not decide whether Luke inherited a composite text or made one, and does not argue from the fact that Mark 1:2 attributes the same words to Isaiah while Luke names no prophet at all. Sorting that out needs a critical apparatus and a study of pre-Christian testimonia collections, and neither was consulted. The page prints the runs and stops.

Luke 7:15 ends καὶ ἔδωκεν αὐτὸν τῇ μητρὶ αὐτοῦ. So does 3 Kingdoms 17:23.

Searched as an exact word sequence rather than as a run, that six-word phrase stands once in the 30,637 Greek Old Testament verse units held here — 3 Kingdoms 17:23, Elijah bringing the widow of Zarephath’s son down from the upper room — and once in the 7,927 verses of the Greek New Testament, here. One hit on each side, and nothing else in either corpus.

Around it the scene keeps company with the same story without repeating its words. Both are widows; both have lost a son; both stories end with a verdict about the man who raised him — but the two verdicts do not use the same word and do not share a phrase. The widow of Zarephath says now I know that you are a man of God (ἄνθρωπος θεοῦ, 3 Kingdoms 17:24); Luke’s crowd says a great prophet has arisen among us (προφήτης μέγας, 7:16). Measured against each other, those two verses run to 1. One phrase in the scene is verbatim; the rest of the resemblance is thematic, and the screen cannot see it. Luke 4:25–26 names the widow of Zarephath explicitly, three chapters earlier, and that passage is on this shelf. This page reports the overlap; it does not claim Luke built the scene from 3 Kingdoms 17.

What can be counted. ἀλάβαστρον, the alabaster flask, stands in two New Testament verses — Matthew 26:7 and Luke 7:37. μύρον stands in thirteen, three of them in this scene (7:37, 7:38, 7:46). καταφιλέω, the intensive kiss, stands in six, and two are here (7:38, 7:45); a third is Judas in Mark 14:45.

What could not be checked, and the page says so plainly. This shelf holds Hebrew, Greek Old Testament and Greek New Testament texts. It holds no Tosefta, no Talmud and no Greco-Roman symposium literature at all, and the only Mishnah and Dead Sea Scrolls files on the disk are eight isolated passages and the 11Q5 Psalms scroll, fetched for other studies and no use here. So the ordinary Stage 2 questions here — what a host owed a guest by way of water, oil and a kiss; whether an uninvited person could stand at the edge of a formal meal; what a woman’s loosed hair signified in a first-century dining room; how a Roman officer in Galilee would have related to a local synagogue he had paid for — are real questions that this page has no primary text on disk to answer. They are named here rather than answered from memory. Anything the page said about them would be a commentary’s claim wearing a measurement’s clothes.

Luke 7:50 and Luke 8:48 carry the identical Greek clause, word for word: ἡ πίστις σου σέσωκέν σε· πορεύου εἰς εἰρήνην. The English splits it.

The verb σῴζω genuinely covers both rescue and cure, and a translator has to choose. The ESV chose one way at 7:50 and the other here, and so drew a distinction the Greek does not draw. The page is not calling the rendering wrong. It is pointing out that a reader working from English alone cannot see that Luke wrote the same five words twice, forty-eight verses apart, to a woman who was forgiven and a woman who was cured — and that the repetition is the tightest measurable bracket in these two chapters.

Greek against Greek. MorphGNT field 6 — the normalised word, the column that already resolves elision and the movable nu; CATSS FORM column, beta-code transliterated first. Both sides lowercased, accents and breathings dropped, final sigma folded, then filtered to Greek letters — the same filter applied to both corpora, together with the editions’ own spelling equivalences. Contiguous runs of whole word forms, never lemmas. Exhaustive and deterministic; nothing sampled, so no seed applies.

⚠️ What a low number means here. The screens use a unigram fallback, so a row sharing no two-word sequence still reports 1 if any single word matches. On this page 1 means “shares no two consecutive words”, never “shares nothing”. That is why Luke 8:10 measuring 1 against Isaiah 6:9 is a statement about sequence and not about relationship.

⚠️ And the tie band is printed beside every run, because a run of 7 with a band of 1 is a signal and a run of 3 with a band of 43 is a formula, and this page contains both.

Distribution over all 106 rows: every row reaches at least 2; 96 reach 3; 58 reach 4; 35 reach 5; 21 reach 6; 16 reach 7; 12 reach 8; 7 reach 10; 4 reach 12; and one reaches 23.

Every one of the 7,927 Greek New Testament verses was screened against every verse outside its own book. Here is the entire top of that list.

Twelve verses in the New Testament reach 23 or more outside their own book; Luke 7:8 is one of them. Only four verse-pairs reach deeper, and one other pair ties. Read the first row of that table beside Luke 8:10. The deepest cross-book contact in the New Testament is two writers quoting Isaiah 6:10 in identical words — and one of those two writers, in his other book, condenses Isaiah 6:9 to a clause that shares no two consecutive words with it.

Distribution over all 106 rows: 14 rows reach exactly 2, 49 reach 3, 31 reach 4, 9 reach 5, 1 reaches 6 and 2 reach 7. Nothing in these two chapters reaches eight words of the Greek Old Testament.

⚠️ The counter-example this page prints against itself. Luke 8:20 is not a quotation, is not an allusion, and is almost certainly not thinking about Joseph. It is seven words of ordinary family speech that happen to sit in the same order in Genesis 37:10 — and it ties with the passage’s only citation formula. If the reader takes the 7 at Luke 7:27 as evidence of a source, the 7 at Luke 8:20 must be taken the same way, and it cannot be. The number never makes the judgement.

The highlighted row is the one that surprised this page. The widow of Nain is the only movement in these two chapters whose deepest Old Testament run is deeper than its deepest New Testament run. It is also one of the movements with no Synoptic parallel — 8:1–3, the women who paid, is another, and whether 7:36–50 counts as a third depends on whether it is the same event as Mark 14:3–9, which this page does not decide. Where Luke has no Gospel parallel, the New Testament numbers collapse; that is what the screen is measuring, and it should not be read as a judgement about the material.

The same effect explains 7:36–50. Fifteen verses, and the deepest thing in the block is 5. Set beside it 8:26–39, where fourteen verses reach 13 — because Mark 5 tells the same story. The screen measures whether a story has a verbal twin. It does not measure whether a story matters.

And the comparison that makes the point. Luke 4:18 — where the scroll is physically open in his hands — reaches twelve consecutive words with Greek Isaiah 61:1. Luke 7:22, answering John from the same prophet, reaches one. The difference is not that one is Scripture and the other is not. It is that one is read aloud and the other is summarised, and a contiguous-run screen can see the first and is blind to the second. Saying so is the whole reason this table exists.

Every number on this page came from three text files and nothing else: the MorphGNT digitisation of the SBL Greek New Testament, the CATSS digitisation of Rahlfs’ Septuaginta, the Masoretic text as printed in Sefaria’s Miqra according to the Masorah. No commentary, no monograph, no journal article and no critical apparatus was consulted for any figure printed here. The books listed below are where a reader should go next; they are not sources for anything above.

⚠️ Specific things this page could not check.

(1) No critical apparatus, on any side. The SBLGNT prints Γερασηνῶν at 8:26; the manuscripts offer three place-names. The SBLGNT marks the clause ἰατροῖς προσαναλώσασα ὅλον τὸν βίον at 8:43 as contested, and the ESV prints it with a footnote saying some manuscripts omit it. This page reports that those variants exist and adjudicates none of them. Every run measured here is a run in one printed edition.

(2) No Second Temple Jewish literature beyond the Greek Old Testament files held here. There is no Tosefta and no Talmud on this disk at all, and the only rabbinic, Qumran and Josephus files on it are isolated fragments fetched for other studies — eight Mishnah passages, the 11Q5 Psalms scroll, and Josephus’ Antiquities I in Whiston’s English — and not one of them touches Luke or a Galilean table. So the questions a Stage 2 would normally answer — table etiquette, purity and touch, a woman’s hair in public, what a synagogue-building benefactor was to a Galilean town — are named in Stage 2 and left unanswered rather than filled in from memory.

(3) No Greco-Roman military or legal texts. The centurion’s speech is the deepest measured thing on this page, and this page cannot tell you how a first-century officer’s authority actually worked, only how many of his words two Gospels share.

(4) The screen finds verbatim contact, not allusion — and Luke is the worst possible book for that limitation. Luke 7:22 gathers Isaiah’s vocabulary from at least four chapters and measures 1 against every one of them. Luke 7:11–17 is shaped like 3 Kingdoms 17 and shares six words with it. A reader who took this page’s low numbers as evidence that Luke is not using Isaiah would be misusing the measurement, and the page says so in three separate places on purpose.

(5) Nothing here settles the Synoptic question, and nothing here tries. That Luke 7:8 and Matthew 8:9 share twenty-three words is a fact. That one of them read the other, or that both read a third thing, or that both preserve one memorable speech, is not a fact and is not asserted anywhere on this page.

(6) The structure diagram is a sequence, and the pairing pattern in its footnote is an observation, not a proof. The only part of the diagram that rests on a measurement is the bracket linking movements 4 and 10, at 7:50 and 8:48.

(7) The Old Testament corpus has a shape. The 30,637 Greek verse units screened here include the deuterocanonical books, which is why 1 Maccabees keeps appearing in the tie bands. That is a fact about the corpus, not about Luke. Where a run lands in 1 Maccabees or 1 Esdras with a band of eleven, it is a formula.

• Joseph A. Fitzmyer, The Gospel According to Luke I–IX (Anchor Bible, 1981).
• John Nolland, Luke 1–9:20 (Word Biblical Commentary, 1989).
• François Bovon, Luke 1: A Commentary on the Gospel of Luke 1:1–9:50 (Hermeneia, 2002).
• I. Howard Marshall, The Gospel of Luke (NIGTC, 1978).
• Richard Bauckham, Gospel Women (2002) — on Joanna, Luke 8:3 and Luke 24:10.
• Kenneth E. Bailey, Jesus Through Middle Eastern Eyes (2008) — on Luke 7:36–50 as a banquet scene.
• C. K. Barrett, The Acts of the Apostles (ICC) — on Acts 28:26–27 and the Isaiah 6 citation at the end of the second volume.
• None of these were consulted for any number on this page. The measurements were run on the text files named in the footer, and the structure diagram is labelled a sequence because that is what it is.

The deepest verbatim contact in these one hundred and six verses is not a saying of Jesus, not a beatitude, not a parable and not a quotation of Scripture. It is a Roman officer, speaking through messengers, explaining that he has never had to be present to have his orders obeyed.

Counted, not asserted: twenty-three consecutive Greek words shared with Matthew 8:9, tie band one, and one participle away from the two verses being identical. Only four verse-pairs in the whole New Testament reach deeper across a book boundary. He never appears on stage. He is quoted, at second hand, by friends he sent — and the sentence that made both Gospels keep his exact words is a sentence about not needing to be there.

And in the same block the prophetic quotation goes the other way. Luke 8:10 puts Isaiah 6:9 into a subordinate clause and keeps not one pair of consecutive words from the Greek Isaiah. The same author, writing his second volume, quotes the same verse at thirteen consecutive words (Acts 28:26) and the next verse at twenty-four (Acts 28:27, against Greek Isaiah 6:10); and Acts 28:27 shares thirty-four consecutive words with Matthew 13:15, the deepest cross-book pair in the New Testament. Twenty-three words for a centurion; one for a prophet. This page reports the asymmetry and does not explain it.

At the end of chapter 7 a woman with a reputation is told ἡ πίστις σου σέσωκέν σε· πορεύου εἰς εἰρήνην. Forty-eight verses later a woman who has been bleeding for twelve years is told the same eight words, in the same order, with one word of address added.

Searched as an exact sequence over all 7,927 verses of the Greek New Testament, πορεύου εἰς εἰρήνην occurs twice, and both are on this page. Mark 5:34 stands at exactly the same moment of exactly the same healing and writes a different verb. The repetition is Luke’s, and it brackets everything between the two women: the sower, the lamp, the family redefined, the storm, the tombs.

One is forgiven and one is cured, and the Greek does not distinguish them. The English printed on this page does — the ESV reads saved at 7:50 and made you well at 8:48. That is a defensible choice, and it is also why this shelf prints the Greek beside the English rather than instead of it.

One of ten Luke studies on this shelf, and the one that covers chapters 7 and 8. Two of the other nine bear directly on what is measured here: Luke 4:16–30, three chapters earlier, where the same Gospel reaches twelve words with Isaiah 61:1 because the scroll is open; and Luke 24:13–35, where Joanna — named here at 8:3 — has just come back from the tomb. The storm in 8:22–25 is studied from Mark’s side at Mark 4:35–41, and the seven-word run at 8:20 is reported from Mark’s side at Mark 2:1–3:35.