Twenty-Four Words, and Zero

• Anchor Bible Dictionary — “Isaiah, Book of,” “Seraph,” “Holy One of Israel,” “Zion Tradition,” “Remnant,” “Woe.”
• Dictionary of the OT: Prophets (IVP) — “Isaiah: Book of,” “Call / Commission Narratives,” “Vineyard, Vine,” “Divine Council,” “Day of the Lord.”
• NIDOTTE — קָדוֹשׁ, כֶּרֶם, צֶמַח, הוֹי, שְׁאָר.
• Dictionary of NT Use of the OT (Beale & Carson) — on Matthew 13:14–15, Mark 4:12, John 12:40, Acts 28:26–27 and Romans 9:29, all of which reach into these six chapters.

Isaiah 1–6 is one of the easiest passages on this shelf to number and one of the easiest to get wrong in a single place.

Every chapter number and every verse number in these six chapters is the same in all three columns. The single exception is that Masoretic and English 2:22 has no CATSS row. This was derived from the files on this disk, verse by verse at the seam, and not copied from a numbering table: Greek 2:19, 2:20 and 2:21 correspond word for word to Hebrew 2:19, 2:20 and 2:21, so nothing after the gap is displaced. The Greek chapter simply ends one verse sooner.

⚠️ The ESV row of that table is the one row not derived from a file on this disk. No ESV text of Isaiah 1–6 is held here — the English in the codex above was typed by hand from the printed edition — so the ESV verse counts are the printed English chapter divisions, which for these six chapters are the same as the Masoretic. The Masoretic and CATSS rows were measured; the ESV row was read.

Over the whole book of Isaiah the corpora disagree in only two chapters — 2 and 56, each one verse — giving Masoretic 1,291 verses and CATSS 1,289. The printed English adds a verse at the 63/64 seam and comes to 1,292, which is read and not measured. The famous Isaiah numbering trap is the 63/64 seam and it is fifty-seven chapters away from this passage.

Greek Isaiah 5:1–9 also stands, as a standalone canticle, in the Odes — the appendix of hymns carried in Greek Old Testament manuscripts. It is Ode 10. The two copies were compared word by word for this page, on the accented forms as the two files print them.

Two of the nine verses are word for word the same; the other seven are not; and Ode 10 stops in the middle of verse 9. Some of the differences are small — ἐγενήθη against ἐγένετο at verse 1, οἶκος τοῦ Ισραηλ against οἶκος Ισραηλ at verse 7. Some are not: the Ode does not have καὶ ἐχαράκωσα, and I fenced it with a palisade, in verse 2, and it reorders verse 4.

What this is and is not evidence of. It is evidence that a single Greek Old Testament tradition transmitted this poem along two routes and that the two routes drifted apart. It is not evidence about which reading is earlier, and this page does not say. The Odes are excluded from every run figure on this shelf — the Greek Old Testament here holds 64 CATSS files and 30,637 verse units, but a run screen sees 30,348, because the Odes are held out. They have to be: Ode 9 is Luke 1 and Ode 13 is Luke 2, copied into a Greek Old Testament manuscript, and ten more Odes are second copies of Exodus 15, Deuteronomy 32, 1 Samuel 2, Habakkuk 3, Isaiah 26, Jonah 2, Daniel 3, Isaiah 5 and Isaiah 38. Left in, they would hand every one of those passages a spurious perfect match with itself.

The interjection הוֹי stands 49 times in the Hebrew Bible. Twenty-two of them are in Isaiah — nearly half the book’s share of a word spread across ten books.

Eight of Isaiah’s twenty-two stand inside these six chapters, and six of those eight are stacked in chapter 5 — 5:8, 5:11, 5:18, 5:20, 5:21 and 5:22 — with 1:4 and 1:24 before them. The Greek column renders every one of the six with οὐαί, which is also the word the Gospels use for the woes on the scribes and Pharisees and the word Revelation uses for its three woes. That is shared vocabulary and nothing more; the page draws no line from these six to any of them.

The exact two-word form קְדוֹשׁ יִשְׂרָאֵל stands 22 times in the Masoretic corpus. Nineteen are in Isaiah. Three are not: 2 Kings 19:22, Jeremiah 50:29 and Psalm 71:22.

And 2 Kings 19:22 is the Sennacherib narrative, which stands again in Isaiah 37:23 — so of the three occurrences outside Isaiah, one is the twin of an Isaiah verse. The first of the nineteen is Isaiah 1:4, in this passage, and two more are here: 5:19 and 5:24. A prefixed variant, וְלִקְדוֹשׁ יִשְׂרָאֵל, adds three more — Isaiah 55:5, 60:9 and Psalm 89:19 — and this page prints the exact-form count and names the variant rather than quietly folding it in.

Chapter 6 never uses the title. It says the thing the title is made of — קָדוֹשׁ קָדוֹשׁ קָדוֹשׁ — and this page notes the juxtaposition without claiming that one produced the other.

Uzziah’s death, Assyrian pressure on Judah, and the agricultural law behind the woe of 5:8 are all outside what this page can measure. No cuneiform source was consulted. No Assyrian annals, no Judean seal or ostracon, no archaeological report on eighth-century estate consolidation. The woe of 5:8 — who join house to house, who add field to field — is often read against land-accumulation practices attested in the period; this page names that as the reading it has heard and not as something checked.

Isaiah 5:7 ends with two pairs of words that differ by one consonant each, and the Greek column keeps the meaning of all four and the sound of none.

English versions have tried — he looked for justice, but saw injustice; for right, but heard a riot — and the ESV does not: it gives justice / bloodshed and righteousness / an outcry plainly. The page reports the sound as a feature of the Hebrew consonants only. A reader in Greek or in English is not being denied a meaning; they are being denied a joke that only works in one alphabet.

Isaiah 2:3 has two plurals in the Hebrew and two singulars in the row.

Hebrew: וְיֹרֵנוּ מִדְּרָכָיו וְנֵלְכָה בְּאֹרְחֹתָיו — that he may teach us of his ways and we may walk in his paths. Row: καὶ ἀναγγελεῖ ἡμῖν τὴν ὁδὸν αὐτοῦ καὶ πορευσόμεθα ἐν αὐτῇ — and he will announce to us his way, and we will walk in it. And the verb has changed too: teach in the Hebrew, announce in the row.

The parallel at Micah 4:2 has the same Hebrew plurals. The page notes the difference between the columns and asserts nothing about how it arose — not a variant, not an interpretation, not a mistake.

All 1,289 CATSS rows of Isaiah were screened against all 7,927 verses of the Greek New Testament for this page. The deepest row in the book is 6:10, and it falls inside this study.

Two of the eight deepest rows in the book of Isaiah are Isaiah 6:9 and 6:10, and a fourth is Isaiah 1:9. Nothing in Isaiah reaches 25.

How close is 24 out of 35? CATSS Isaiah 6:10 is 35 words. Matthew 13:15 is 34 and Acts 28:27 is 34, and Matthew 13:15 and Acts 28:27 are word for word identical to each other. The whole difference from the Isaiah row is one pronoun: the row has τοῖς ὠσὶν αὐτῶν where both New Testament verses have τοῖς ὠσίν. That single dropped word is what stops the run at 24 instead of 34.

And 6:9 shows the other thing a contiguous screen does. CATSS Isaiah 6:9 and Acts 28:26 are both 21 words. Two things differ: the row’s καὶ εἶπεν is Acts’ λέγων, and the phrase to this people stands on opposite sides of the verb — row: πορεύθητι καὶ εἰπὸν τῷ λαῷ τούτῳ; Acts: πορεύθητι πρὸς τὸν λαὸν τοῦτον καὶ εἰπόν. The same three words, moved — and re-inflected: dative in the row, accusative with πρός in Acts. The run is 13, not 20, because reordering does not count.

Three of the clearest echoes in this passage measure 1, 1 and 0 on exactly the same screen that gave Isaiah 6:10 a 24. If the runs above are read as more than they are, these are the correction.

Mark 4:12 is quoting Isaiah 6:9–10 and the screen gives it a one against 6:9 and a two against 6:10. Mark reads ἵνα βλέποντες βλέπωσιν καὶ μὴ ἴδωσιν καὶ ἀκούοντες ἀκούωσιν καὶ μὴ συνιῶσιν — every idea of Isaiah 6:9 in a different order with different verb forms. Luke 8:10 is shorter still. The same two verses of Isaiah reach 24 in one Gospel and 1 and 2 in another. John 12:40 reaches four with 6:10 and uses two verbs — τετύφλωκεν, ἐπώρωσεν — that stand nowhere in this Isaiah row.

The vineyard is the sharpest case, because the reason the screen fails is visible. Greek Isaiah 5:2 and Mark 12:1 share four verb lemmas — plant, put around, build, dig — and not one shared verb form, because Isaiah’s row has them in the first person and the parable has them in the third.

Four verbs, four matching lemmas, zero matching forms. A run screen counts forms and is therefore blind to the single clearest piece of scriptural furniture in the Synoptic parables. And there is a further twist the page prints because it cuts both ways: the hedge, φραγμόν, stands in the planting sequence of the Greek Isaiah 5:2 and of Mark 12:1 and Matthew 21:33 — and the Hebrew has no hedge there at all, only at 5:5, where it is removed. The page does not say which column anyone was reading.

And Isaiah 1:2 against Deuteronomy 32:1 is a clean zero. The Hebrew screen on this shelf uses unigram_fallback=False, so 0 means “no two consecutive words shared” — but here the intersection is empty at the level of single words too. Isaiah has שִׁמְעוּ · שָׁמַיִם · וְהַאֲזִינִי · אֶרֶץ; Deuteronomy has הַאֲזִינוּ · הַשָּׁמַיִם · וְתִשְׁמַע · הָאָרֶץ. The verbs are crossed, the nouns are articulated in one and bare in the other, and not one of the four forms matches.

The mountain oracle stands in two books. It was measured on the Hebrew side, verse against verse, across the whole of Isaiah 2 against the whole of Micah 4.

Three verses of dense overlap, then a cliff. This is the same shape Obadiah’s overlap with Jeremiah 49 has: real for a stretch, and then it simply stops. Isaiah 2:5 goes on to O house of Jacob, come, let us walk in the light of the LORD; Micah 4:4 goes on to every man under his vine and under his fig tree. Neither has any counterpart in the other.

Inside the overlap, the differences are small and one of them is a swap.

Isaiah 2:3 and Micah 4:2 are 24 words each and differ in three, one of which is a spelling (וְיֹרֵנוּ / וְיוֹרֵנוּ) and one of which is a conjunction. Where Isaiah says peoples Micah says nations, and one verse later they trade back. ⚠️ The page reports the overlap, its extent and its edge, and asserts nothing whatever about which book had it first — not that Micah borrowed, not that Isaiah borrowed, not that both drew on a third source. All three are held.

Five rows in this passage reach exactly five consecutive words against a single New Testament verse and no other — 2:2, 2:8, 5:9, 5:14 and 6:8. Two of them are printed here because the theme lines up as well, and neither is offered as a connection.

Acts 2:17 is Peter quoting Joel, and it does not begin the way Greek Joel begins. CATSS Joel 3:1 — English Joel 2:28 — opens καὶ ἔσται μετὰ ταῦτα, and it shall be after these things. Acts 2:17 opens καὶ ἔσται ἐν ταῖς ἐσχάταις ἡμέραις, and it shall be in the last days. That four-word phrase stands in exactly one verse of the 30,348-row Greek Old Testament screening corpus, and it is Isaiah 2:2.

James 5:4 is a woe on rich landowners, and Isaiah 5:9 sits in the middle of one. James: the cries of the harvesters have reached the ears of the Lord of hosts. Isaiah 5:9, three verses into the woe on those who join house to house: these things were heard in the ears of the Lord of hosts. The five-word phrase stands in one verse of the whole screened Greek Old Testament and in one verse of the New Testament.

What is claimed, and it is deliberately small. A run of five with a tie band of one is distinctive. It is not a citation, and neither Acts nor James signals one here. The page prints the two measurements, notes that in each case the theme also lines up, and stops. It does not say that Peter reached for Isaiah’s phrase to open a Joel quotation, or that James had Isaiah’s woe in mind — both are readings, and readings are what the counter-examples above exist to restrain.

Chapters 3 and 4 — 32 verses — reach nothing. Screened against the whole Greek New Testament, the deepest row in those two chapters is 4, and the two rows that reach it do so on a date formula: ἐν τῇ ἡμέρᾳ ἐκείνῃ at 3:7 (tie band 6) and 3:18 (tie band 4). A run of four with a band of six is a formula, and the page labels it one. The only row in those two chapters that reaches four with a tie band of one is 3:24 against Revelation 2:2 on τὰ ἔργα σου καί — four common words, and the page draws nothing from it.

And the refrain of chapter 2 produces a run the page will not attribute. 2 Thessalonians 1:9 shares eight consecutive words with the Greek column — κυρίου καὶ ἀπὸ τῆς δόξης τῆς ἰσχύος αὐτοῦ. But it shares them equally with Isaiah 2:10, 2:19 and 2:21, because the row is identical in all three places. The measurement cannot say which verse, and neither can this page. What it can say is why the run is eight and not twelve: 2 Thessalonians reads ἀπὸ προσώπου τοῦ κυρίου and the Isaiah rows read ἀπὸ προσώπου τοῦ φόβου κυρίου — one extra word, φόβου, breaks the sequence.

Two more runs from this passage, printed so they are not mistaken for signals. Isaiah 2:3 reaches five with Matthew 12:4, Mark 2:26 and Luke 6:4 on εἰς τὸν οἶκον τοῦ θεοῦ — a tie band of three, and all three are David eating the bread of the Presence, which has nothing to do with the mountain of the LORD. Isaiah 5:14 reaches five with Revelation 6:15, tie band one, on καὶ οἱ πλούσιοι καὶ οἱ. Neither is a connection. Both are what four or five common Greek words look like when they happen to line up.

No Qumran material was consulted, and for Isaiah 1–6 that is the single worst gap on this page. The Great Isaiah Scroll (1QIsa^a) is the best-known biblical manuscript in the world and it carries all six of these chapters complete. It is not held on this shelf and nothing here has been checked against it.

Three findings on this page are precisely the kind a Qumran reading could bear on, and a reader should treat each as a two-column observation and not a three-column one:

• Isaiah 2:22 — the verse with no CATSS row. What any Qumran witness reads there is not known to this page.
• Isaiah 6:13 — the holy seed is its stump, present in the Masoretic column and absent from the CATSS row. This is a well-known crux and it is discussed in the literature with the scroll evidence in hand. This page has the two columns and no third.
• Isaiah 4:2 — the Branch that is in one column and not the other.

What is stated on this page about all three is only this: the Masoretic column as printed in the edition used has the words, and the CATSS row as digitized does not. No claim is made about which is earlier, what either witness once read, or what produced the difference.

No Hebrew or Greek critical apparatus — not BHS, not BHQ, not the Göttingen Isaiah. The Greek here is Rahlfs as digitized by CATSS, which is one edition and not a critical reconstruction, and where this page says “the row does not have it” it means that file and no more. No Targum Jonathan, no Peshitta, no Vulgate, no Aquila, Symmachus or Theodotion. No rabbinic material, and none of the patristic reception of Isaiah 6 — which is substantial, and which turns on exactly the imperative-versus-report difference this page prints at 6:10.

The Odes comparison is a comparison of two digitized files, not of two manuscripts. CATSS Isaiah 5 against CATSS Odes 10, word for word, on this disk. It shows that the corpus carries the poem twice and that the two copies differ in 33 of 215 words — 31 of them once accent-only differences are folded, as every run screen on this shelf folds them. It says nothing about which manuscripts stand behind either, and the page does not name a manuscript it has not seen.

The literary structure is a reading, and one bracket in it is a measurement. The כאלה bracket between 1:30 and 6:13 was screened over the whole Masoretic corpus and the count is exact. Everything else in the diagram — the six movements, the pairing of 2:2–5 with 4:2–6 — is a reading of the text, is labelled as one, and is not drawn as a chiasm. No commentary’s outline was copied.

Two questions this page deliberately does not answer. Whether chapter 6 records the beginning of Isaiah’s ministry or a later commissioning, and whether Isaiah 2:2–4 or Micah 4:1–3 came first. Both are live in the literature and neither can be settled from the text files this page measures. A page that answered them would be repeating something it had read.

The run screen finds verbatim contact. It does not find allusion, and this passage is the best demonstration of that on the shelf. It gave Isaiah 6:10 a 24 and Mark 4:12 a 2 — and 6:9 a 1 — for the same prophecy. It gave the Song of the Vineyard a 1 against the parable built on it. It gave Isaiah 1:2 a 0 against Deuteronomy 32:1. Every one of those low numbers is correct as a measurement and misleading as a conclusion, which is why they are printed beside the 24 and not buried under it.

Conversely, a high number is not automatically a citation. It is here — Matthew 13:14 says the prophecy of Isaiah is fulfilled and Romans 9:29 says as Isaiah predicted, so the texts themselves signal it. Where a text does not signal it, this page reports shared words and stops.

• Joseph Blenkinsopp, Isaiah 1–39 (Anchor Yale Bible 19).
• H. G. M. Williamson, Isaiah 1–5 and Isaiah 6–12 (International Critical Commentary).
• John Goldingay on Isaiah — series and volume not verified against any catalogue held on this shelf; the ICC Isaiah volumes named above are Williamson’s.
• Ronald E. Clements, Isaiah 1–39 (New Century Bible).
• Ken M. Penner, Isaiah (Septuagint Commentary Series) — for the Greek column.
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer.

Chapter 1 ends with a tree: you shall be like an oak whose leaf withers, and like a garden without water… and none shall quench them. Six chapters later the passage ends with the same Hebrew word: like a terebinth or an oak, whose stump remains when it is felled. The ESV renders that one word אֵלָה as oak at 1:30 and as terebinth at 6:13, so the bracket stands in the Hebrew and not in the English. Those are the only two places in the Hebrew Bible where those four consonants mean a tree rather than the pronoun these.

Between them the book has burned everything: the sacrifices, the city, the vineyard, the leadership, the pride of chapter 2, and finally the people themselves down to a tenth. And where the first tree is a dying one with fire nobody puts out, the last is a felled one with something still in the ground.

The Greek column does not have the last three words. The Hebrew ends זֶרַע קֹדֶשׁ מַצַּבְתָּהּ — a holy seed is its stump. The CATSS row stops one clause earlier. Whichever column a reader is holding, the passage ends with the tree cut down; only one of them tells you what is left. This page prints both and takes no position on the difference.

Twenty-four consecutive words is the deepest verbatim contact anywhere in the sixty-six chapters of Isaiah, and it is here, at 6:10. Then Mark quotes the same two verses and the same measurement gives him a two against 6:10 and a one against 6:9.

Both numbers are true and neither is the whole account. Matthew and Luke-Acts reproduce the Greek sentence; Mark says the same thing in his own words. A screen that counts word forms in order can see the first and is deaf to the second, and if this shelf printed only the twenty-four it would be teaching people to mistake a copying relationship for a theological one.

This study closes the last dark chapters in the book of Isaiah on this shelf — 1–6, alongside the existing studies at 7:14, 9:6–7, 52:13–53:12 and 61:1–3.