Four chapters of straight narrative in the middle of a book of poems: an Assyrian field commander preaching at the wall of Jerusalem, a king who turns his face to the wall and is given fifteen more years, and an embassy from Babylon that gets shown the whole treasury. Screen all ninety verses against the rest of the Hebrew Bible and seventy-six of them come back pointing at the same place: 2 Kings 18–20, verse for verse and almost always in order — two of the seventy-six point outside the sequence — each with a tie band of one. The median run is ten consecutive words, the maximum is twenty-one, and twelve verse pairs are identical from the first word to the last. And the overlap does not fade — it has two edges. 2 Kings 18:14–16, where Hezekiah surrenders and pays, has no counterpart in Isaiah. And Isaiah 38:9–20, the Writing of Hezekiah, has no counterpart in Kings: a heading and eleven verses of psalm, two of which share no two consecutive words with anything in the Hebrew Bible — and those eleven are exactly what a Greek manuscript tradition copied out a second time, as Ode 11. Against the whole Greek New Testament, meanwhile, the deepest run in these ninety verses is six words, and it is an accident. See the method →
Isaiah 36:1–39:8 — sixteen rows across all four chapters (the chapter and verse numbers agree in the Hebrew, the Greek and the ESV throughout these four chapters — 22, 38, 22 and 8 verses in every column; there is no offset here)
Hebrew (Masoretic) · Greek (CATSS/Rahlfs) · ESVThe Shape of the Passage
Sequence — twelve scribal sections, and one of them has no twinNothing on this page bears on the authorship of the book of Isaiah, or on the authorship of Kings. The question of how many hands stand behind Isaiah’s sixty-six chapters is real, it is old, and it is not answerable from the files on this disk. The question of which of these two accounts was written first is the oldest question anyone asks about this passage, and this page does not answer it either — the house rule forbids asserting a direction of influence, and a run length cannot carry one. What follows reports who the text says is speaking and who it says is listening, and stops there.
Author
These four chapters carry no superscription, and they are the only sustained narrative in the book. Isaiah 1–35 is oracle and poem; 36–39 is a story told in prose, with dates, names, troop numbers and a chain of custody for a letter. The book’s headings stand at 1:1, 2:1 and 13:1, none of them here, and the only internal dating here is the fourteenth year of King Hezekiah at 36:1.
Isaiah is a character in his own book here, and always in the third person. יְשַׁעְיָהוּ בֶן־אָמוֹץ is named at 37:2, 37:5, 37:6, 37:21, 38:1, 38:4 (as the recipient of the word), 38:21, 39:3, 39:5 and 39:8. Not once in ninety verses does the prophet say “I” — except inside the psalm of 38:9–20, where the “I” is Hezekiah’s and the heading says so.
The same ninety verses stand, in nearly the same Hebrew words, in 2 Kings 18:13–20:19. That is the central measurable fact about this passage and stage 4 gives the whole profile: median run ten, maximum twenty-one, twelve pairs identical first word to last. The measurement does not say which text is prior, whether either writer saw the other, or whether both drew on a third thing. Commentators have argued all three for two centuries. The screen is silent on all three and so is this page.
The Greek column is used here as a witness to a second ancient text, never as evidence about who wrote the Hebrew one. Where the two disagree — and on this page they disagree at 36:7, 37:36, 37:38, 38:8, 38:9, 38:14, 38:17, 38:22, 39:6 and 39:8 — the page prints both and names neither prior.
Audience
The most striking thing about the audience is that the enemy addresses it directly, in its own language, over the heads of its officials. 36:11 is a request to negotiate in Aramaic so the men on the wall will not understand; 36:13 is the Rabshakeh raising his voice and shouting יְהוּדִית — in Judean — on purpose. The audience inside the story is a city listening to a sermon preached by the Assyrian empire, and the sermon’s thesis is that trusting God is the same mistake as trusting Egypt.
Then the audience shrinks to one man in a bedroom, and then to one man showing off a treasury. Chapter 38 has two people in it and one of them is dying; chapter 39 has an embassy from a Babylon that is not yet an empire. The book’s last narrative word is Babylon, and the chapter after it opens Comfort, comfort my people.
How it has been carried. Hezekiah’s psalm has been sung as a canticle in Christian daily prayer for many centuries, on the strength of the Greek heading Προσευχὴ Εζεκιου and its place in the Odes. That is older than any of us and is not invented here. What this page adds is only the measurement: the eleven verses that became a canticle are exactly the eleven verses 2 Kings does not carry, and the second Greek copy of them agrees with the first in three rows out of eleven. The page does not claim the measurement is the meaning.
Orient on Sennacherib, the Rabshakeh, the Hezekiah narratives and the Song of Hezekiah.
Which articles are relevant?
- Anchor Bible Dictionary — “Isaiah, Book of,” “Hezekiah,” “Sennacherib,” “Rabshakeh,” “Merodach-baladan,” “Esarhaddon,” “Ararat,” “Odes, Book of.”
- Dictionary of the OT: Prophets (IVP) — “Isaiah: Book of,” “Hezekiah,” “Assyria, Assyrians,” “Zion Traditions,” “Prophecy and Historiography.”
- Dictionary of the OT: Historical Books (IVP) — “Kings, Books of,” “Chronicles, Books of,” “Synoptic Problem in the Historical Books,” “Hezekiah.”
- NIDOTTE / TDOT — בָּטַח (bāṭaḥ, trust), מִכְתָּב (miḵtāḇ, a writing), שְׁאֵרִית (šʽērîṯ, remnant), מַעֲלוֹת (maʿəlōṯ, steps).
- TDNT / EDNT — προσευχή, σημεῖον, ἄγγελος κυρίου — and for stage 4, ἰδοὺ ἡμέραι ἔρχονται as a formula rather than a citation.
The verse numbers — and there is no offset here
Isaiah 36, 37, 38 and 39 have twenty-two, thirty-eight, twenty-two and eight verses in the Masoretic text, in the CATSS Greek and in the ESV alike. Every reference on this page therefore means the same verse in every column. That is stated because this shelf states it even when the answer is nothing: several of its books — Joel, Jonah, Malachi, the Psalter, Jeremiah — are numbered differently in Greek and English, and a reader who has met those pages is entitled to know that this one has no such trap in it.
There IS an offset against 2 Kings, and it starts inside this passage. Isaiah 37 has thirty-eight verses; 2 Kings 19 has thirty-seven. Isaiah 37:15 and 37:16 are together what 2 Kings 19:15 is on its own — the Hebrew of Isaiah 37:15 is a five-word introduction (and Hezekiah prayed to the Lord, saying) that Kings folds into the prayer verse. From that point to the end of the chapter Isaiah’s number runs one ahead of Kings’: Isaiah 37:38 is 2 Kings 19:37. This page prints the Kings reference beside every run so no reader has to hold the arithmetic.
Across the whole book of Isaiah, the CATSS file and the Masoretic text differ in exactly two places. The Greek file holds 1,289 rows where the Hebrew holds 1,291 verses. It has no row for Isaiah 2:22 and none for Isaiah 56:12; the chapter maxima agree everywhere else, all sixty-six chapters. Neither gap touches this passage.
What that does and does not mean. It is a statement about the files this shelf holds, not about ‘the Septuagint’. Whether the two omissions belong to the Greek tradition or only to this edition is a question about manuscripts, and no manuscript evidence was consulted for this page.
Bibliography surfaced:
- John N. Oswalt (NICOT), Brevard S. Childs (OTL), Joseph Blenkinsopp (Anchor Yale), J. Alec Motyer, H. G. M. Williamson, Hans Wildberger (Continental) and Otto Kaiser on Isaiah 13–39.
- Brevard S. Childs, Isaiah and the Assyrian Crisis (1967) — the study that set the terms of the modern discussion of Isaiah 36–39 and 2 Kings 18–20.
- Mordechai Cogan and Hayim Tadmor, II Kings (Anchor); Marvin Sweeney; Christopher Seitz, Zion’s Final Destiny — on the relation of the two accounts.
- William Gallagher, Sennacherib’s Campaign to Judah; Paul Evans, The Invasion of Sennacherib in the Book of Kings; Lester Grabbe (ed.), ‘Like a Bird in a Cage’.
- Ronald L. Troxel, LXX-Isaiah as Translation and Interpretation; Arie van der Kooij; Isac Seeligmann — cited from the secondary literature and not consulted directly for this page.
Four tellings of one campaign, and how far apart they are when you count.
The Rabshakeh is not a name
רַבְשָׁקֵה stands in exactly sixteen verses of the 39-book Masoretic corpus held here, and every one of them is in this story.
| where | verses |
|---|---|
| 2 Kings | 18:17, 18:19, 18:26, 18:27, 18:28, 18:37; 19:4, 19:8 |
| Isaiah | 36:2, 36:4, 36:11, 36:12, 36:13, 36:22; 37:4, 37:8 |
Eight and eight, and nowhere else in the Bible. It is an Assyrian court title — the rab šāqê, chief cupbearer — carried into Hebrew as a word and treated by the English tradition as a name. Two other Assyrian officers are named in the Kings account and dropped by Isaiah: the Tartan and the Rabsaris of 2 Kings 18:17 stand in Isaiah 36:2 only as the Rabshakeh, alone.
On the Greek side Ραψακης also stands in sixteen rows — and one of them is not a twin. They are not quite the same sixteen. Eight are the same eight 2 Kings verses; seven are Isaiah verses; and the last is Sirach 48:18. CATSS Isaiah 36:11 does not name him at all — where the Hebrew has they said to the Rabshakeh, the Greek row reads only καὶ εἶπεν πρὸς αὐτόν, and he said to him, which is also why that row’s deepest New Testament run is a four-word formula. The Sirach verse is measured below.
Ben Sira tells the whole story in nine verses — and shares three words with it
Sirach 48:17–25 is a Second Temple retelling of exactly this passage, and this shelf holds it in Greek. Ben Sira wrote in Hebrew around 180 BC and his grandson put the book into Greek later in the same century, so what is measured below is the grandson’s Greek. Ben Sira has Hezekiah fortifying the city and digging for water (48:17); Sennacherib coming up and sending the Rabshakeh (48:18); the people crying out and the Holy One answering ἐν χειρὶ Ησαιου (48:20); the angel wiping out the Assyrian camp (48:21); the sun going backward and life added to the king (48:23).
Measured word-form against word-form, the deepest contact between those nine Greek verses and the ninety Greek verses of Isaiah 36–39 is THREE consecutive words — Δαυιδ τοῦ πατρός, Sirach 48:22 against Isaiah 38:5.
That is the point of putting Ben Sira here. A writer who plainly knows the story, names its characters, and reproduces its plot beat for beat shares almost no verbatim wording with it. Verbatim contact and dependence are different things, and this is the cleanest demonstration of it on the page. It is also the calibration for the New Testament figure in stage 4: a run of six is not much, but a run of three is what a real, admitted retelling looks like on this screen.
The Assyrian side of the story, and what this shelf cannot check
Sennacherib’s own account of the campaign survives — the Rassam Cylinder and the Taylor Prism — whose dates were not checked for this page — in which the king says he shut Hezekiah up in Jerusalem like a bird in a cage, took forty-six walled cities, and received a heavy tribute including thirty talents of gold. The gold figure agrees with 2 Kings 18:14; the silver figure does not (the prism says eight hundred talents, Kings says three hundred). Neither figure is in Isaiah, which does not carry the tribute at all.
Flagged, not blended in. No Akkadian corpus is held on this disk. The prism is cited here from the standard published translations (Luckenbill; Pritchard’s ANET) and was not read in the original for this page. Every measurement printed anywhere else on this page was made against files this project holds; this paragraph and the Lachish relief below are the exceptions, and they are flagged rather than folded in.
What can be verified from the files is the shape of the biblical accounts. Four tellings exist and three of them are measurable here: Isaiah 36–39 (90 verses), 2 Kings 18:13–20:19 (81 verses), 2 Chronicles 32 (33 verses) and Sirach 48:17–25 (9 verses). Stage 4 prints all three comparisons in one table, and the spread is from twenty-one consecutive words down to three.
- 2 Kings 18:13–20:19 — the parallel account; the whole of stage 4 is about it.
- 2 Chronicles 32:1–33 — the third telling, with the tunnel and the letters, and a very different verdict on Hezekiah’s envoys.
- Sirach 48:17–25 — the Second Temple retelling, held here in Greek.
- Psalm 46; Psalm 48; Psalm 76 — the Zion psalms often read against the deliverance of 37:36. Not measured on this page.
- The Lachish reliefs (British Museum, from Nineveh) — the Assyrian picture of the siege 2 Kings 18:14 names. Cited, not examined.
Four words, and what the screen says about each.
Eight verses in the whole Hebrew Bible, four of them with the noun unprefixed: Exodus 32:16, Exodus 39:30, 2 Chronicles 21:12 and Isaiah 38:9; the other four prefix it (Deuteronomy 10:4, Ezra 1:1, 2 Chronicles 35:4, 2 Chronicles 36:22). Elsewhere it is engraving on stone or gold, or a letter. Here it heads a psalm — and the CATSS Greek calls it προσευχή instead.
Two verses in the whole Hebrew Bible — Isaiah 38:14 and Jeremiah 8:7, where the same pair of birds stands together. Nobody is certain which bird it is. The CATSS Greek has only one bird here, χελιδών, where the Hebrew has two words.
Sixteen verses, eight in 2 Kings and eight in Isaiah, and nowhere else. An Assyrian office, not a personal name. The Greek Ραψακης adds one more: Sirach 48:18.
Eight forms in Isaiah 36 alone, in the Rabshakeh’s two speeches — 36:4 (twice), 36:5, 36:6 (twice), 36:7, 36:9, 36:15. The word the enemy uses more than anyone in the chapter is the word for faith.
בָּטַח — the Assyrian preaches a sermon on trust
The Rabshakeh’s Hebrew is built on one root, and he uses it eight times in the twelve verses from 36:4 to 36:15, across both of his speeches. On what do you rest this trust of yours? … you are trusting in Egypt … if you say we trust in the Lord … do not let Hezekiah make you trust in the Lord. בָּטַח at 36:4 (twice, as verb and noun), 36:5, 36:6 (twice), 36:7, 36:9 and 36:15 — eight word forms in seven verses.
And the Greek does not carry the pun. CATSS renders it with πείθω / πέποιθα in most places — be persuaded, be confident — and with ἐλπίζω, hope, in others. A Greek reader hears a mixed vocabulary where a Hebrew reader hears the same word hammered.
Nothing in this panel is a run measurement. It is a count of a root in a chapter, made against the corpus files, and it is labelled as such so that it is not read as one of the screens.
מַעֲלוֹת — steps, and a word that does five jobs
Isaiah 38:8 uses מַעֲלוֹת five times in one verse. צֵל הַמַּעֲלוֹת… בְמַעֲלוֹת אָחָז… עֶשֶׂר מַעֲלוֹת… עֶשֶׂר מַעֲלוֹת… בַּמַּעֲלוֹת. The English tradition splits it into dial and steps; the Hebrew has one word, and the same word heads the fifteen Psalms of Ascents, Psalms 120–134 (שִׁיר הַמַּעֲלוֹת in fourteen of them, שִׁיר לַמַּעֲלוֹת in Psalm 121).
The CATSS Greek keeps most of the repetition and changes the owner. ἀναβαθμός stands four times in the Greek row, once fewer than the Hebrew. And where the Hebrew says the shadow had gone down on the steps of Ahaz, the Greek says τοῦ οἴκου τοῦ πατρός σου — of the house of your father. Ahaz is not named in the Greek of this verse. Whether that is a translator declining to name the idolatrous king, or a different text in front of him, this page cannot say and does not.
Kings and Isaiah do not tell this sign the same way either. In 2 Kings 20:9–10 Hezekiah is offered a choice — shall the shadow go forward ten steps or back ten? — and he chooses backward as the harder thing. Isaiah 38 has no choice and no dialogue. That is why 38:8 measures only four shared words against Kings, on a page whose median is ten, and why its best match is 2 Kings 20:11 rather than the neighbouring verse.
Everything measured, with its tie band, and the counter-examples printed beside it.
Screen 1 — the Hebrew of Isaiah 36–39 against the rest of the Hebrew Bible
Method: each of the 90 Masoretic verses of Isaiah 36–39 against the 23,114 verse units of the 39-book corpus outside those four chapters. Consonantal word forms; cantillation and vowel points removed; qere read; final letters folded to medial; maqqef a separator and never a weld; the scribal section markers {ס} and {פ} stripped before measurement. The unigram fallback is OFF, so a row reported as 0 shares no two consecutive words with any other verse — it does not mean the verse shares no words.
Result: 76 of the 90 rows have exactly ONE deepest match in the whole corpus, and in all 76 that one verse is inside 2 Kings 18:13–20:19. All fourteen exceptions are in Isaiah 38, and twelve of the fourteen are 38:9–20. The distribution of deepest runs: 3 rows reach twenty-one, 2 reach twenty, 2 reach nineteen, 2 reach eighteen, 2 reach seventeen, 2 reach sixteen, 2 reach fifteen, 4 reach fourteen, 7 reach thirteen, 5 reach twelve, 5 reach eleven, 5 reach ten, 6 reach nine, 5 reach eight, 6 reach seven, 8 reach six, 4 reach five, 6 reach four, 5 reach three, 7 reach two, and 2 reach nothing at all.
| Isaiah | run | tie band | the other end | what is shared |
|---|---|---|---|---|
| 36:6 | 21 | 1 | 2 Kings 18:21 | the broken reed of Egypt, all but the first two words of the verse |
| 36:7 | 21 | 1 | 2 Kings 18:22 | the high places Hezekiah removed, all but the first two words |
| 37:4 | 21 | 1 | 2 Kings 19:4 | the words of the Rabshakeh … the remnant that is left |
| 37:38 | 20 | 1 | 2 Kings 19:37 | the whole verse — twenty words against twenty, identical |
| 39:3 | 20 | 1 | 2 Kings 20:14 | what did these men say, and where did they come from? |
| 37:3 | 19 | 1 | 2 Kings 19:3 | a day of distress … no strength to bring them forth (whole verse) |
| 38:1 | 17 | 1 | 2 Kings 20:1 | set your house in order, for you shall die |
| 37:16 | 17 | 1 | 2 Kings 19:15 | God of Israel, enthroned above the cherubim … you made heaven and earth |
| 38:14 | 0 | — | nothing | shares no two consecutive words with any verse in the Hebrew Bible |
| 38:16 | 0 | — | nothing | the same — and both do share individual words |
Read the top three rows against the bottom two. Twenty-one with a band of one is as strong a verbatim result as this shelf has ever printed. Zero with the fallback off is its opposite. Both are in the same four chapters, and they are eleven verses apart.
The two edges — where the twins stop being twins
Aligning the two accounts verse by verse gives seventy-seven pairs. Median run ten, mean 10.5, maximum twenty-one, and forty-one of the seventy-seven reach ten or more. Twelve pairs are identical from the first word to the last: 36:4, 36:9, 37:1, 37:3, 37:5, 37:8, 37:22, 37:31, 37:32, 37:37, 37:38 and 39:7, against 2 Kings 18:19, 18:24, 19:1, 19:3, 19:5, 19:8, 19:21, 19:30, 19:31, 19:36, 19:37 and 20:18.
Then there are two holes, and they are not the same size or the same kind.
| the hole | size | what is in it | measured |
|---|---|---|---|
| In Isaiah: 2 Kings 18:14–16 | 3 verses | Hezekiah sends to Lachish — I have done wrong; withdraw from me — pays 300 talents of silver and 30 of gold, and strips the gold from the doors and doorposts of the temple | against all 90 verses of Isaiah 36–39 these three reach runs of 3, 2 and 2, tie bands 2, 7 and 3 |
| In Kings: Isaiah 38:9–20 | 12 verses (heading + 11 of psalm) | the Writing of Hezekiah — I said, in the middle of my days I must depart… you have cast all my sins behind your back | against the whole of 2 Kings 20 the eleven psalm verses reach a maximum of 2, and 38:10 shares not one word |
Verse by verse against the whole of 2 Kings chapter 20: 38:10 → 0; 38:11, 12, 13, 14, 15, 16, 17, 18, 19 → 1 apiece (single common words — not, until, all, to me, upon, for); 38:20 → 2, בֵּית יְהוָה. That is an edge, not a fade. The verse before the psalm, 38:8, has a Kings twin. The verses after it, 38:21–22, have Kings twins. The eleven in between have nothing.
And the page does not say why. Whether Isaiah’s editor added a psalm, or Kings’ editor dropped one, or both inherited different collections, is a question about direction of influence. The screen cannot answer it, and the house rule forbids guessing. What is measurable is the shape of the hole and the sharpness of its edges.
Screen 2 — the same passage against the whole Greek New Testament
Method: the 90 CATSS rows of Isaiah 36–39 against all 7,927 verses of the Greek New Testament (SBLGNT via MorphGNT). Contiguous runs of whole word forms, never lemmas. Accents, breathings and case dropped on both sides; final sigma folded; MorphGNT’s own spelling equivalences applied — οὐ/οὐκ/οὐχ, ἐκ/ἐξ and the movable nu — so that a shared run is not cut at a word the two editions merely spell differently. Exhaustive and deterministic.
Across the 90 rows: 89 reach two, 68 reach three, 23 reach four, 2 reach five, 1 reaches six, and nothing reaches seven.
| Isaiah | run | tie band | the other end | what is shared |
|---|---|---|---|---|
| 39:6 | 6 | 1 | Hebrews 8:8 | ἰδοὺ ἡμέραι ἔρχονται λέγει κύριος καί |
| 37:16 | 5 | 4 | Lk 16:17, Ac 4:24, Ac 14:15, Re 14:7 | τὸν οὐρανὸν καὶ τὴν γῆν |
| 37:1 | 4 | 16 | Mk 4:4, Lk 5:12, Lk 7:11 and 13 more | καὶ ἐγένετο ἐν τῷ |
| 38:1 | 4 | 11 | Lk 1:8, Lk 2:6, Lk 3:21 and 8 more | ἐγένετο δὲ ἐν τῷ (8 of the 11) and καὶ εἶπεν πρὸς αὐτόν (3) |
| 37:4 | 4 | 7 | Mt 4:7, Mt 4:10, Mt 22:37, Mk 12:30 and 3 more | κύριον τὸν θεόν σου |
| 38:12 | 4 | 4 | Mt 13:1, Lk 10:12, Ac 2:41, 2 Th 1:10 | ἐν τῇ ἡμέρᾳ ἐκείνῃ |
| 37:21 and 38:5 | 4 | 1 | Revelation 1:8 | λέγει κύριος ὁ θεός |
| 38:11 | 4 | 2 | Luke 3:6, Acts 28:28 | τὸ σωτήριον τοῦ θεοῦ |
| 36:2 | 4 | 1 | Mark 13:26 | μετὰ δυνάμεως πολλῆς καί |
Calibration against the rest of the book. Of the 1,289 CATSS rows in Isaiah, 41 reach six or more with the Greek New Testament and 18 reach ten or more (Isaiah 6:10 is deepest at twenty-four, then 54:1 at twenty-three, 53:8 at twenty). These four chapters contribute exactly one of the 41 and none of the 18. The most narrated chapters in Isaiah are the least quoted ones.
The deepest New Testament run on this page is an accident, and here is the arithmetic
Isaiah 39:6 and Hebrews 8:8 share six consecutive Greek word forms with a tie band of one — and Hebrews 8:8 is not quoting Isaiah.
| the six shared words | what comes next | |
|---|---|---|
| CATSS Isaiah 39:6 | ἰδοὺ ἡμέραι ἔρχονται λέγει κύριος καί | λήμψονται πάντα τὰ ἐν τῷ οἴκῳ σου — they will take all that is in your house |
| Hebrews 8:8 | ἰδοὺ ἡμέραι ἔρχονται λέγει κύριος καί | συντελέσω ἐπὶ τὸν οἶκον Ἰσραήλ — I will conclude a new covenant with the house of Israel |
Hebrews is quoting Jeremiah. Hebrews 8:8–12 is the longest Old Testament citation in the New Testament, and it is the new-covenant passage — Jeremiah 38:31–34 in the Greek numbering, 31:31–34 in the Hebrew. The six words are a prophetic formula, and they stand in ten rows of the Greek Old Testament held here: Amos 8:11, Amos 9:13, Isaiah 39:6, and Jeremiah 7:32, 9:24, 16:14, 19:6, 23:5, 23:7 and 28:52. The tie band of one is a fact about the New Testament side only; on the Old Testament side the band is ten.
And the run only exists because of where the Greek puts a phrase. The Masoretic Isaiah 39:6 opens הִנֵּה יָמִים בָּאִים — behold, days are coming — with no says the Lord attached, and closes with אָמַר יְהוָה. The CATSS Greek opens ἰδοὺ ἡμέραι ἔρχονται λέγει κύριος and closes εἶπεν δὲ ὁ θεός. The twin verse, 2 Kings 20:17, has ἰδοὺ ἡμέραι ἔρχονται and does NOT have λέγει κύριος after it — which is why 2 Kings 20:17 is one of the twenty Greek Old Testament rows carrying the three-word form and not one of the ten carrying the five-word one.
This page does not say that Isaiah’s translator ‘added’ the phrase or that Kings’ translator ‘dropped’ it. Both are claims about a direction, and the house rule forbids them. What is printable is that the six-word run between Isaiah and Hebrews exists in one column of this page and not in the other, and that the New Testament verse making it is demonstrably about something else.
Screen 3 — the same story told four times, and how far apart the tellings are
Four accounts, three of them measurable against the files this shelf holds.
| the two texts compared | deepest run | typical run | what that means |
|---|---|---|---|
| Isaiah 36–39 vs 2 Kings 18:13–20:19 Hebrew against Hebrew | 21 | median 10, mean 10.5 | verbatim twins; 12 of 77 pairs identical end to end |
| CATSS Isaiah 36–39 vs CATSS 4 Kingdoms 18–20 Greek against Greek, the same verse pairs | 18 | median 6, mean 6.4 | only 13 of the 77 pairs reach ten, against 41 in Hebrew |
| Isaiah 36–39 vs 2 Chronicles 32 Hebrew against Hebrew | 7 | 46 of the 90 rows reach one; 32 reach two; one reaches zero (this comparison alone runs with the unigram fallback ON, so a one is a single shared word and the zero means no shared word at all) | the same events, told independently |
| CATSS Isaiah 36–39 vs Sirach 48:17–25 Greek against Greek | 3 | — | an admitted retelling, and almost no verbatim contact |
The Greek row is the one worth sitting with. Isaiah 36:7 and 2 Kings 18:22 share twenty-one consecutive Hebrew words. The Greek of the same two verses shares two — because CATSS Isaiah 36:7 is nine words long and CATSS 4 Kingdoms 18:22 is thirty-six. Whatever produced two near-identical Hebrew accounts, the Greek versions of them are two separate pieces of translation work, and the screen shows it in the numbers rather than asserting it.
2 Chronicles 32 is the control. Its deepest contact with any of these ninety verses is seven consecutive words: 2 Chronicles 32:12 against Isaiah 36:7 — את במתיו ואת מזבחתיו ויאמר ליהודה ולירושלם, the high-places taunt again. Everything else is one or two words. Chronicles knows the story, tells it, and does not share its sentences.
Screen 4 — the psalm, copied twice in the same Greek tradition
The Greek Old Testament held here is 64 CATSS files and 30,637 verse units, screened down to 30,348 with the Odes excluded. The Odes are a Greek manuscript appendix: 289 rows in this file — fourteen numbered pieces plus thirteen title lines. Two of them are New Testament — Ode 9 is Luke 1 and Ode 13 is Luke 2 — and ten more are second copies of passages already in the Greek Old Testament. Three of those ten are Isaiah: Ode 5 is Isaiah 26:9–20, Ode 10 is the Song of the Vineyard (Isaiah 5:1–9), and Ode 11 is Isaiah 38:10–20. The other seven are Exodus 15 (Ode 1), Deuteronomy 32 (Ode 2), 1 Samuel 2 (Ode 3), Habakkuk 3 (Ode 4), Jonah 2 (Ode 6) and Daniel 3 twice over (Odes 7 and 8). Odes 12 and 14 — the Prayer of Manasseh and the Gloria — are neither.
Ode 11 is Isaiah 38:10–20, and the Odes file keeps Isaiah’s own verse numbers for it — eleven rows, labelled 10 to 20, under the heading Προσευχὴ Εζεκιου. These are the eleven verses 2 Kings does not have. Measured against CATSS Isaiah 38 row by row:
| row | Ode 11 words | CATSS Isaiah words | shared run |
|---|---|---|---|
| 38:10 | 17 | 16 | 8 |
| 38:11 | 18 | 15 | 9 |
| 38:12 | 32 | 36 | 31 |
| 38:13 | 23 | 17 | 8 |
| 38:14 | 24 | 27 | 11 |
| 38:15 | 10 | 7 | 7 |
| 38:16 | 14 | 14 | 14 — identical |
| 38:17 | 16 | 16 | 16 — identical |
| 38:18 | 20 | 20 | 20 — identical |
| 38:19 | 22 | 18 | 16 |
| 38:20 | 18 | 22 | 18 |
Three rows of eleven come back word for word. Eight do not. ‘Second copy’ does not mean identical copy. That is worth stating on a page that reports run lengths: the same passage carried twice inside one manuscript tradition does not come back the same, which is a useful measure of how much a run of six across two different corpora is actually worth.
Why the Odes are excluded from the screens. If they were left in, Isaiah 38 would show up as its own doublet and every run figure in this passage would be quietly inflated — a page about the psalm would report the psalm agreeing with itself at thirty-one words and call it a finding. The figure this page states as its Greek Old Testament scope is the screened one, 30,348, not the held one, 30,637.
What the scribal section markers cost, and the qere
The Masoretic file this shelf holds prints the scribal section markers inside the verse text, and the frozen normaliser tokenises them as words. {ס} and {פ} stand at the end of 12 of this passage’s 90 verses, and of 3,447 of the 23,204 verses in the Masoretic corpus held here. A run can therefore be extended by one word that is not a word.
The screen was run both ways, and here is the whole difference on this page. Exactly one row changes its run: Isaiah 37:32 reads as thirteen with the markers counted and twelve with them stripped. Exactly one other row keeps its run and loses tie-band entries: 38:22 ties with 1 Chronicles 26:22, 2 Chronicles 8:16 and 2 Chronicles 35:2 only because of a marker, and with them stripped its band falls from four to one. Nothing else on the page moves at all. Every Hebrew figure printed here is the marker-stripped one. normalize.py is frozen and was not edited; the strip was applied downstream of it.
The qere is read — and these four chapters carry exactly three ketiv/qere pairs, none of them in a codex row above. Two stand in 36:12, where the written text is coarse and the read text is polite ((חראיהם) read [צוֹאָתָם], and (שיניהם) read [מֵימֵי רַגְלֵיהֶם]); one stands in 37:30 ((ואכול) read [וְאִכְלוּ]). A fourth, at 39:2, the edition marks as trivial. Every Hebrew measurement on this page reads the qere, and the raw Masoretic markup was checked verse by verse rather than assumed.
Where the argument is, where this page stands, and where it is weakest.
What is actually argued about here
The relation of Isaiah 36–39 to 2 Kings 18–20 is one of the longest running arguments in Old Testament study, and it has three standing positions. (1) Isaiah’s version is the borrower — the traditional and still common view, and Kings’ extra tribute paragraph is often cited for it. (2) Kings is the borrower — argued from the way the narrative sits inside the Isaiah scroll and from the psalm. (3) Both draw on a shared source, and the differences are editorial. Childs, Cogan and Tadmor, Seitz, Sweeney, Williamson and Evans have all written at length. Nothing measurable on this shelf touches the question. A run length is symmetrical: it says the same thing read from either end. This page therefore takes no side and says so plainly, rather than letting a silence read as agreement.
The two-campaign question is likewise untouched here. Whether 2 Kings 18:13–16 and 18:17–19:37 report one Assyrian campaign or two is a historical argument built on Assyrian annals and archaeology. This page notes only that the three verses in question have no Isaiah counterpart, at runs of three, two and two.
Greek Isaiah is itself a scholarly subject. Seeligmann, van der Kooij and Troxel have argued at length about how free its translator is and whether he contemporises. Several divergences printed on this page — the missing high-places clause at 36:7, Armenia for Ararat at 37:38, a prayer for a writing at 38:9, the sins thrown behind me rather than behind your back at 38:17, and the wish rather than the statement at 39:8 — are exactly the sort of thing that literature is about. This page reports them and does not explain them, because explaining them means asserting a direction.
What this page did NOT check
No manuscript evidence of any kind was consulted. Not the Great Isaiah Scroll (1QIsaᵃ), which covers all sixty-six chapters and is the single most obvious thing to check against a page about Isaiah; not 4QIsa fragments; not Codex Vaticanus, Sinaiticus or Marchalianus for the Greek; no critical apparatus for either column; and no manuscript of the Odes, whose contents and order vary between the great codices — which matters, because Ode 11 is a load-bearing part of this page. The Hebrew is Sefaria’s Miqra according to the Masorah as this project holds it, and the Greek is Rahlfs as digitized by CATSS. Where this page says two texts differ, it means these files differ.
No Akkadian text was read. Sennacherib’s prisms are cited in stage 2 from published translations. There is no Akkadian corpus on this disk and the screen cannot reach one. The same applies to the Lachish reliefs, which are named and not examined.
The ESV column was not verified against a held ESV text. No ESV file for Isaiah exists in this project and no Crossway API key has ever been issued to it. The English above was typed, and its verse alignment rests on the verse counts and chapter openings, which do agree across all three columns here. The Hebrew and Greek columns were pulled programmatically out of the corpus files and were not typed.
The Hebrew of Ben Sira was not consulted. Parts of Sirach survive in Hebrew in the Cairo Geniza and Masada manuscripts, and no manuscript of either was checked for this page. This shelf holds only the Greek, so the three-word figure in stage 2 is a Greek-to-Greek measurement and nothing more.
Targum, Peshitta, Vulgate and the Qumran pesharim were not consulted — nor was 2 Chronicles 32 screened against 2 Kings, which would have completed the triangle in stage 4 and was left undone.
Rabbinic and patristic reception was not surveyed. Hezekiah’s psalm has a long life in both, and in Christian daily prayer as a canticle; none of it was read for this page.
Where this page is weakest
1. The verse alignment between Isaiah and Kings was set by hand. Seventy-seven pairs, built from the best-match table and then corrected at the one seam the two books have — Isaiah 37:15/37:16 against 2 Kings 19:15, after which Isaiah’s numbering runs one ahead. If that seam is placed wrongly, the median and mean in stage 4 move. The 76-of-90 figure does not depend on it: that one comes from an unaligned screen against the whole corpus.
2. The structure diagram rests on one edition’s paragraph marks. The twelve sections are the {פ} and {ס} of the Sefaria file. Those markers are a real feature of the Masoretic tradition, but manuscripts differ over them, and none was collated. Note in particular that the scribes make 38:9–22 a single section, psalm and fig-cake together; the measurement splits them at verse 20, and the diagram prints both facts rather than choosing.
3. The passage boundary is literary AND administrative. 36:1 does begin a new thing and 39:8 does end one — 40:1 is Comfort, comfort my people. But this study also covers exactly these four chapters because they were the last chapters of Isaiah 30–39 not yet quoted anywhere on this shelf, and the page says so rather than pretending the two reasons are one.
4. Two claims were withdrawn while this page was built. (a) A draft called Isaiah 39:6 → Hebrews 8:8 ‘the New Testament’s use of this passage’. It is not: Hebrews is quoting Jeremiah, the shared words are a formula standing in ten Greek Old Testament rows, and the claim was rewritten as this page’s main counter-example. (b) A draft said the Odes ‘preserve’ the Greek text of Hezekiah’s psalm; they carry a second copy of it that agrees with the first in three rows out of eleven, and preserve asserts a priority the measurement does not carry.
5. Nothing here settles what happened at Jerusalem in 701 BC. The angel, the 185,000, the plague theories, the withdrawal, the tribute — the page counts words and stops. The history of the campaign is not a measurement and is not offered as one.
Bibliography, and what was and was not read
- Oswalt (NICOT), Childs (OTL), Blenkinsopp (Anchor Yale), Williamson, Wildberger, Motyer, Kaiser — consulted as the standing frame for Isaiah 13–39; no commentary supplied a number printed on this page.
- Brevard S. Childs, Isaiah and the Assyrian Crisis; Christopher Seitz, Zion’s Final Destiny; Cogan and Tadmor, II Kings (Anchor) — on the relation of the two accounts.
- Gallagher, Sennacherib’s Campaign to Judah; Evans, The Invasion of Sennacherib in the Book of Kings; Grabbe (ed.), ‘Like a Bird in a Cage’ — the source of the Assyrian material in stage 2, which was not read in the original.
- Luckenbill, The Annals of Sennacherib; Pritchard, ANET — the prism translations, cited and not collated.
- Seeligmann, The Septuagint Version of Isaiah; van der Kooij; Troxel, LXX-Isaiah as Translation and Interpretation — named, not consulted.
- Text: Sefaria, Miqra according to the Masorah; Rahlfs’ Septuaginta as digitized by CATSS (including the Odes file); SBLGNT as digitized by MorphGNT. These files are the only evidence behind every number on this page.
What the measurements leave you holding.
The thing worth sitting with
Two books tell the same story for ninety verses, and each one keeps something the other does not. Kings keeps the surrender: the message to Lachish, the three hundred talents of silver, the gold cut off the temple doors. Isaiah keeps the song: eleven verses of a man who thought he was dying, in which he compares himself to a bird and to a shepherd’s tent pulled up, and then says that God threw all his sins behind his back. Neither hole is a fade. Both are edges. The verse before the psalm has a twin; the verses after it have twins; the eleven in between share, at most, two words with the whole of 2 Kings 20.
And the loneliest verses in the passage are inside the song. Isaiah 38:14 and 38:16 share no two consecutive words with any other verse in the Hebrew Bible — not one adjacent pair in 23,114 verses. 38:14 contains a bird nobody can identify — a word that stands in only one other verse, Jeremiah 8:7 — and it ends I am oppressed; be my surety. A man on his deathbed does not talk like the rest of the Bible, and the measurement can see it.
Meanwhile the New Testament, which quotes Isaiah more than any other prophet, does not quote these four chapters at all. Eighteen rows of Isaiah reach ten or more shared words with the Greek New Testament. None of them is here. The deepest run these ninety verses make is six — behold, days are coming, says the Lord, and — and the verse at the other end of it, Hebrews 8:8, is quoting Jeremiah about a new covenant. The most dramatic story in the book is the least cited part of it.
What the numbers are for. Not to prove the connections — the twinning of Isaiah and Kings was obvious to anyone who ever read both, and Ben Sira knew the story more than two thousand years before anyone counted. The numbers are for keeping the page honest about which claims rest on the text and which rest on the reader. Twenty-one words with a band of one is the text. Zero with the fallback off is the text. Hebrews 8:8 sounding like Isaiah 39:6 is the reader, and this page has said so.
One last thing the text does without any help from the measurement. Hezekiah’s last recorded words in this book are the word of the Lord that you have spoken is good — said about a prophecy that his sons will be carried to Babylon. The Hebrew adds his reason: for there will be peace and truth in my days. The Greek turns it into a prayer: let there only be peace and righteousness in my days. Two ancient texts, two very different men, and the next verse in the book is Comfort, comfort my people.
Every Deep Research page traces its passage across the whole sweep — the Old Testament, the Second-Temple world between the testaments (historical witnesses, not Scripture), and the New Testament. Follow the thread:
Pleasant Springs Church — Deep Research
Research method after Michael S. Heiser’s Bible Research Method (examine dictionaries → ancient literature → important words → textual connections → academic publications → share), with the Pleasant Springs “Author & Audience” foundation. Scripture set in the Hebrew, the Greek editions named on each page and the ESV (Old Testament), and the Greek New Testament and the ESV (New Testament), per the practice of Pleasant Springs Church. Isaiah research grounded in Oswalt (NICOT), Childs (OTL and Isaiah and the Assyrian Crisis), Blenkinsopp (Anchor Yale), Williamson, Wildberger, Motyer and Kaiser; on the two accounts in Cogan and Tadmor, Seitz and Evans; on the Assyrian side in Gallagher, Luckenbill and Pritchard. Text: Hebrew from the Masoretic text as printed in the edition consulted — Sefaria’s Miqra according to the Masorah; no critical apparatus was consulted for the Hebrew and no Dead Sea Scroll was collated, 1QIsaᵃ included. Greek from Rahlfs’ Septuaginta as digitized by CATSS; Greek New Testament from the SBL edition as digitized by MorphGNT. The standing caution for this shelf applies: these pages say “the CATSS Greek text reads…” and never call a Greek reading early or late, and never say which text acted on which. ⚠️ Verse numbering. There is no offset in this passage. Masoretic, CATSS and ESV all give Isaiah 36, 37, 38 and 39 as 22, 38, 22 and 8 verses, in the same order, with the chapter breaks in the same places, so every reference on this page means the same verse in every column. There IS an offset against 2 Kings, and it begins here: Isaiah 37:15 and 37:16 are together what 2 Kings 19:15 is alone, so from that point Isaiah’s verse number runs one ahead of Kings’ to the end of the chapter. Across the whole book the CATSS file holds 1,289 rows where the Hebrew holds 1,291 verses — it has no row for Isaiah 2:22 and none for Isaiah 56:12, and the chapter maxima agree in all sixty-six chapters otherwise. Whether those omissions belong to the Greek tradition or only to this edition is a question about manuscripts, and no manuscript evidence was consulted. ⚠️ The ESV column was NOT verified against a held ESV text. No ESV file for Isaiah exists on this disk and no Crossway API key exists in this project; the English was typed and its verse alignment rests on verse counts and chapter openings. The Hebrew and Greek columns were pulled programmatically from the corpus files. ⚠️ Every cross-corpus run figure on this page was measured on 31 August 2026 and applies MorphGNT’s own spelling equivalences — οὐ/οὐκ/οὐχ, ἐκ/ἐξ and the movable nu — so that a shared run is not cut at a word the two editions merely spell differently. Figures published on this shelf before 30 August 2026 are lower for that reason; nothing was copied from a sibling page and every number above was re-measured. ⚠️ A run of zero on the Hebrew screen means NO TWO CONSECUTIVE WORDS. The Hebrew-side screen runs with the unigram fallback off, so max = 0 means the verse shares no adjacent word pair with any verse outside Isaiah 36–39 — it does not mean the verse shares no words. Both zeros on this page, 38:14 and 38:16, share individual words freely: with the fallback switched on they return 1, with tie bands of 1,804 and 911. The Greek-side screen runs with the fallback ON. ⚠️ The Masoretic text held here counts the scribal section markers as words. {ס} and {פ} stand inside the verse text of the Sefaria file and the frozen normaliser tokenises them — at the end of 12 of this passage’s 90 verses, and of 3,447 of the 23,204 verses in the Masoretic corpus held here. Stage 4 prints the screen both ways: exactly one row moves, 37:32, from thirteen to twelve. Every Hebrew figure on this page is the marker-stripped one. normalize.py is frozen and was not edited. ⚠️ Greek Old Testament figures are scoped to the 64 CATSS files held here, 30,637 verse units, screened to 30,348 with the Odes excluded — which matters on this page more than on any other, because Ode 11 is Isaiah 38:10–20, the very passage this study is about; two more of the fourteen Odes are Isaiah as well (Ode 5 is Isaiah 26 and Ode 10 is Isaiah 5). Leaving the Odes in would have had the psalm agreeing with itself at thirty-one words and called it a finding. ⚠️ This page never says which of Isaiah and Kings is prior. A contiguous-run measurement is symmetrical and carries no direction. The oldest question anyone asks about this passage is one the screen cannot answer, and stage 5 names the three standing positions without choosing among them. Two claims were withdrawn while the page was built and both are named in stage 5. Typology is handled per the house position: what the text says, what the New Testament itself draws, and what we see from this side of the cross — with the seeing left to the reader.
Scripture quotations marked ESV are from the ESV® Bible (The Holy Bible, English Standard Version®), © 2001 by Crossway, a publishing ministry of Good News Publishers. ESV Text Edition: 2016. Used by permission. All rights reserved. The ESV text may not be quoted in any publication made available to the public by a Creative Commons license, and the ESV may not be translated in whole or in part into any other language. www.esv.org
|
Did this study bless you today? 💚 Give to Support PS Church100% of gifts go to the General Fund — thank you. |