Ten Words That Belong to a Psalm

• Anchor Bible Dictionary — “Luke, Gospel of,” “Jericho,” “Tax Collector,” “Olives, Mount of,” “Temple, Jerusalem,” “Sadducees,” “Levirate Marriage,” “Son of Man.”
• Dictionary of Jesus and the Gospels (IVP) — “Triumphal Entry,” “Temple Cleansing,” “Apocalyptic Teaching,” “Destruction of Jerusalem,” “Taxes,” “Resurrection,” “Son of David.”
• New International Dictionary of NT Theology & Exegesis — ἀρχιτελώνης, ἐπισκοπή, ἐδαφίζω, κεφαλὴ γωνίας, ἰσάγγελος, ὑστέρημα, καιροὶ ἐθνῶν.
• Commentary on the New Testament Use of the Old Testament, ed. G. K. Beale and D. A. Carson (Baker, 2007) — the Luke chapter, where 19:38, 19:46, 20:17, 20:42–43 and 21:24–27 are treated verse by verse.
• None of these were consulted for the measurements below. The measurements were run on the text files named in the footer.

Every page on this shelf states its versification offset, including where the answer is that there is none. On a New Testament book there is no second column to be offset from. But on this passage the offset decides how two of the page’s headline references must be written.

Counted off disk. Unlike Luke 17 and Luke 23, this passage has no missing verse number in any column — every reference on this page means the same verse in the Greek and in the ESV. (SBLGNT Luke prints no 17:36 and no 23:17, which is why the book holds 1,149 verses under a numbering that runs to 1,151. Neither gap falls here.)

The offset that does bite is the Psalter’s.

The Greek Psalter runs one number behind the Hebrew from Psalm 9/10 through Psalm 147, because it joins Hebrew Psalms 9 and 10 into one. Both of this page’s deepest Old Testament runs fall inside that range, so every citation here is written twice — once in the numbering of the text that was actually searched, once in the numbering a reader will find in an English Bible. The house rule on this shelf is that the Septuagint chapter and verse is shown alongside the English wherever they differ, and here they differ.

Luke 19:2 calls Zacchaeus an ἀρχιτελώνης — a word that occurs once in the Greek New Testament and is never explained. Jericho sat on the balsam and date trade and on the road up from the Jordan, and the chief of a tax district there would have been a wealthy man; the text says so directly (καὶ αὐτὸς πλούσιος) and explains nothing further.

At 19:8 he offers half his goods to the poor and fourfold restitution to anyone he has defrauded. The verb is συκοφαντέω, which occurs twice in the Greek New Testament and both times in Luke (3:14; 19:8). ⚠️ What could not be checked here. The fourfold figure is regularly connected to Exodus 22:1 and to Roman law on furtum manifestum, and to rabbinic discussion of a tax collector’s repentance. No papyri, no Roman legal text, no rabbinic tractate and no inscription were consulted for this page. Those readings are reported from the reference works named in Stage 1 and none of them was verified from a primary document on this disk. What the measurement adds is negative and firm: Luke 19:1–10 tops out at a run of 5 against everything outside Luke, and at 5 against the whole Greek Old Testament — and both fives are narrative formulae, not contacts. The 5 against the Old Testament is at 19:9 and the tied verse is Genesis 24:5, because the five words are and he said to him.

Eight New Testament verses call it τὸ ὄρος τῶν ἐλαιῶν, the mount of the olives, in one case or another. Two call it τὸ ὄρος τὸ καλούμενον Ἐλαιῶν, the mount called Olive-grove. Both of those two are in this passage.

This is a lexical and distributional fact, and the page does not turn it into a theory of sources. Luke uses both forms, three verses apart at 19:29 and 19:37. What is countable is that the proper-noun form brackets the temple section exactly, and that the pair run between 19:29 and 21:37 is 5.

Luke 19:43–44 puts three words into two sentences that the rest of the New Testament never uses. χάραξ, a palisade or siege-rampart, occurs once; περικυκλόω, to encircle, once; and one verse later ἐδαφίζω, to level to the ground, once. Three siege terms, three hapax legomena, four verses that measure 3 against everything outside Luke.

21:5 has admirers of the temple pointing at its ἀναθήματα. The word is a votive offering hung up in a sanctuary; it occurs once in the Greek New Testament, and it is not the word ἀνάθεμα, an accursed thing, which occurs six times and is spelled with an epsilon. ⚠️ The two nouns are one letter apart and manuscripts do confuse them; no critical apparatus was consulted for this page and the reading printed here is simply the SBL edition’s.

21:2 gives the widow two λεπτά. The coin is named three times in the Greek New Testament: Mark 12:42, Luke 12:59 and here. Mark adds a gloss for a Roman reader — ὅ ἐστιν κοδράντης — and Luke does not. The temple treasury chest, γαζοφυλάκιον, is named five times in the Greek New Testament and four of the five are this scene in the two Gospels. No numismatic source, no Mishnah tractate on the temple chests, and no archaeological report was consulted here.

133 verses; 2,038 words in the MorphGNT text of Luke 19–21.

⚠️ A hapax count is not an argument. Twelve of the words checked for this page occur once in the Greek New Testament. That is a normal figure for 2,038 words of Greek narrative, no exhaustive hapax census was run, and does not on its own mean anything about the writer. It is printed because several of the twelve cluster: three siege terms in two verses (19:43–44), and two commercial terms in the Zacchaeus scene. Clustering is worth noticing; it is not evidence, and this page does not treat it as evidence.

All 133 verses of Luke 19:1–21:38 were screened twice: once against the 6,778 non-Lukan verses of the Greek New Testament, once against the 30,637 verse units of the Greek Old Testament held on this disk. Contiguous runs of whole word forms, never lemmas, with the tie band beside every run.

Against the non-Lukan New Testament, across the 133 rows: 132 reach two, 101 reach three, 69 reach four, 46 reach five, 31 reach six, 26 reach seven, 19 reach eight, 17 reach nine, 10 reach ten, 5 reach eleven, 3 reach thirteen, and nothing reaches fourteen. Against the Greek Old Testament: all 133 reach two, 97 reach three, 38 reach four, 9 reach five, 3 reach eight, 2 reach ten, and nothing reaches eleven.

Luke 20:17 reaches 10 against the non-Lukan New Testament and 10 against the Greek Old Testament. Luke 20:43 does the same. In both cases the run string on the two screens is byte-for-byte identical.

When a row’s Old Testament run and its New Testament run are the same length and the same string, the two numbers are not two facts. They are one fact seen twice: the words belong to the Psalm, and everything that carries them — Luke, Matthew, Mark, Acts, Hebrews — is carrying the Psalm.

The census, screened across all 7,927 Greek New Testament verses.

⚠️ Why Luke’s two rows are shallower than Matthew’s and Mark’s, and why that is not a fact about Luke. Matthew, Mark and Hebrews carry the whole of Psalm 109:1 inside a single verse; Luke splits it across 20:42 and 20:43. This screen measures verse by verse, so a quotation split by a verse division is measured twice at half length. The 14 and the 10 are therefore not comparable, and the page says so rather than letting the table imply that Luke quotes less exactly. Joined, Luke’s two verses carry the same quotation the others do.

1. The verse that names Moses and measures two against him. Luke 20:28 opens Μωϋσῆς ἔγραψεν ἡμῖν and then states the levirate law. Against CATSS Deuteronomy 25:5, the law being named, the run is 2. Against the whole Greek Old Testament it is 4, tie band 1, and the tied verse is Genesis 38:9 — the death of Onan — because Genesis 38 uses σπέρμα τῷ ἀδελφῷ where Deuteronomy 25:5 does not. Against Mark 12:19 the same row runs 13. A citation formula is not a verbal contact, and this row proves it in one line.

2. The Son of Man in a cloud, and Daniel 7:13. Luke 21:27 measures 9 with Mark 13:26 and 1 with Daniel 7:13 — in both Greek Daniels held here, the Old Greek and Theodotion. The shared word is καί. Every commentary names Daniel 7 at this verse and the screen cannot see it. The reason is visible: cloud singular against clouds plural, and articles present against articles absent.

3. The house of prayer, and Obadiah. Luke 19:46 says it is written and quotes Isaiah 56:7 and Jeremiah 7:11. It measures 4 against the whole Greek Old Testament, tie band 2, and one of the two tied verses is Obadiah 1:18, which has no relation to it at all. The Obadiah page prints this same pair as its own standing counter-example; here it appears from the other end, and it says the same thing. A run of four with a band of two is not evidence.

4. The verse that touches nothing. Luke 19:6 — so he hurried and came down and received him joyfully — is the one row in all 133 that does not reach even two consecutive words anywhere in the 6,778 non-Lukan verses of the Greek New Testament. Its tie band is 4,443. A band of four thousand is the screen saying: this shares single ordinary words with more than half the corpus and two consecutive words with none of it. The most famous sentence in the Zacchaeus story is, verbally, the emptiest row on this page.

All four Gospels quote CATSS Psalm 117:26 (ESV 118:26) at the entry. Three reach six. Luke reaches three.

Luke 13:35 is the control that turns this from a curiosity into a result. It is not that Luke quotes the Psalm less closely; it is that at 19:38 he puts one word inside it, and a contiguous-run screen cuts at an insertion. The pair run between Luke 19:38 and Luke 13:35 is 3, for the same reason. And a second Lukan fingerprint is in the same verse: the six New Testament occurrences of ὡσαννά are all in Matthew, Mark and John; Luke has none, anywhere in the Gospel, and writes peace in heaven and glory in the highest instead — three words of which (δόξα ἐν ὑψίστοις) stand at Luke 2:14.

The gradient is the plainest structural fact the numbers produce. Chapter 19 is Luke’s own: one row in forty-eight reaches nine, and the material that nobody else has — Zacchaeus, the minas, the tears over the city — sits at three, four and five. From 20:1 the runs climb and stay high, and they stay high because from 20:1 onward Luke is telling the same temple confrontations that Matthew and Mark tell. ⚠️ That is a description of the distribution, not an explanation of it. The screen cannot say why the material Luke shares is worded alike and the material he alone has is worded differently; it can only say that it is. Direction of influence is not asserted here or anywhere on this page.

No critical apparatus, on any side. The Greek printed here is the SBL edition as digitized by MorphGNT and nothing else. Luke 19–21 carries variant readings that matter — among them whether βασιλεύς stands at 19:38 and in what position, and whether 21:5 reads ἀναθήμασιν or ἀναθέμασιν. This page’s central finding about 19:38 depends on a word whose text is disputed, and it says so here rather than burying it. If a different edition were screened, that row could move. The two rows at 20:17 and 20:43, which are the page’s headline, are far more stable — but they were not checked against an apparatus either.

No Hebrew was screened for this page. The Psalms behind 19:38, 20:17 and 20:42–43 were measured only in Greek, against CATSS. Nothing here says anything about how the Hebrew of Psalm 118 or Psalm 110 relates to the Greek, and a reader should not take the Greek agreement as a claim about the Hebrew.

No Second Temple, rabbinic, Roman or documentary material. No Qumran text, no Josephus on the siege — which is the obvious comparison for 19:43–44 and 21:20–24 and was not made — no Mishnah, no papyri on tax farming, no numismatics for the λεπτόν, no inscriptions, no archaeological report on the Herodian temple. Everything this page says about culture in Stage 2 is either a count taken off the text files or a reading reported from the reference works, and each is labelled.

The Greek Old Testament held here is not the whole Septuagint. It is 64 CATSS files, 30,637 verse units. Isaiah, Jeremiah, Lamentations, Ezekiel and Daniel are present on this disk and were screened, which is what makes the Daniel 7:13 and Isaiah 56:7 measurements above possible. But a null result against this corpus is a null result against this corpus.

The claim that ‘the run belongs to the Psalm’ is an inference, not a measurement. What was measured is that three verses share the same ten words with each other and with Psalm 117:22, and that the strings are identical. That is consistent with three evangelists independently reproducing a Greek Psalter they all knew, and it is equally consistent with two of them copying the third. The measurement cannot distinguish those, and the page does not pretend to. What it rules out is only the reading that the ten words are evidence of literary contact between the Gospels specifically, since the same ten stand in a text older than all three.

The 19:38 finding rests on a single inserted word. It is a real measurement — 3 against 6, with Luke 13:35 as an internal control at 6 — but it is one word wide, and the word is text-critically contested (above). A page that presented it as a settled fact about Luke’s theology of kingship would be overreaching, and this one presents it as a fact about a run.

The chapter-by-chapter gradient is partly an artefact of what the other Gospels contain. Luke 19 measures low against the non-Lukan New Testament in large part because 19:1–27 and 19:41–44 have no parallel anywhere — so there is nothing for them to be deep against. That is close to circular, and the table is printed with the caveat attached.

And one thing the low numbers do NOT mean. Luke 21:20–24 is saturated with Old Testament language — the trampling of Jerusalem, the sword, the captivity among the nations — and its five rows top out at 4 against the whole Greek Old Testament (21:21 and 21:23, and both fours are ordinary vocabulary). The screen sees no contact because there is no verbatim contact, and that is not evidence that the language is not scriptural. It is evidence that Luke is not quoting. The same applies to 19:41–44, whose Old Testament ceiling is 5 — and that 5, at 19:42, is the phrase on this day tied with four unrelated verses. The absence of a run is not the absence of a background, and this page would be badly misread if its numbers were taken that way.

• Joseph A. Fitzmyer, The Gospel according to Luke X–XXIV (Anchor Bible 28A, 1985) — the standard technical treatment of these chapters.
• François Bovon, Luke 3: A Commentary on the Gospel of Luke 19:28–24:53 (Hermeneia, 2012), and Luke 2 (2013) for 19:1–27.
• John Nolland, Luke 18:35–24:53 (WBC 35C, 1993).
• Nicholas Perrin, Jesus the Temple (2010) — on the entry and the temple act, which this page measures and does not interpret.
• Steve Moyise and Maarten J. J. Menken, eds., The Psalms in the New Testament (2004) — the volume that treats Psalms 110 and 118 across the corpus, which is exactly the question this page’s two headline rows raise and cannot settle.
• Mark Goodacre, The Case Against Q (2002), and Christopher M. Tuckett, Q and the History of Early Christianity (1996) — the two sides of the question the thirteen-word runs bear on and do not decide.
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer.

Run a verse against the New Testament and you learn who else says it. Run it against the Greek Old Testament and you learn what it is quoting. Twice in these three chapters the two questions came back with the same answer — the same number and the same ten words — and that coincidence is the whole finding. Luke 20:17 does not agree with Matthew and Mark about a stone; all three of them are saying Psalm 118:22, and the agreement is what is left when the Psalm is subtracted. Luke 20:43 is the same again with Psalm 110:1.

And the mirror image sits six chapters back in the same book. At the entry, three Gospels carry six words of Psalm 118:26 and Luke carries three — not because he knows it less well, but because he sets a word down in the middle of it. The King. He had already quoted the line whole at 13:35. Here, coming down the Mount of Olives with the city in front of him, he breaks it open to put a title inside.

Which is the honest summary of what this page is for. The engine found the doubled ten and the broken six and would never have found the frame; a reader sees at once that the passage begins and ends on the same hill and could not have found the doubled ten. Both are in the text. The page prints the runs, the tie bands, and the four places where the numbers mislead — the citation that measures two against the law it names, the Son of Man who measures one against Daniel, the house of prayer tied with Obadiah, and the verse that touches nothing at all.

The passage opens with a short man in a tree who is told to come down because the visitor is staying at his house tonight, and it closes with the whole people coming early in the morning to the temple to hear him. In between he weeps over a city for not knowing the time of its visitation. Today salvation has come to this house.