Eight Words in Jeremiah

• Anchor Bible Dictionary — “Revelation, Book of,” “Seal,” “Hades,” “Tribes of Israel,” “Trumpet,” “Apocalypticism.”
• Dictionary of the Later NT and Its Developments (IVP) — “Revelation, Book of,” “Martyrdom,” “Persecution,” “Apocalyptic, Apocalypticism.”
• NIDNTT / TDNT — σφραγίς, θλῖψις, ῥομφαία against μάχαιρα, δεσπότης.
• Commentary on the NT Use of the OT (Beale & McDonough on Revelation) — the standard catalogue of this passage’s alleged Old Testament sources, against which the measurements in stage 4 can be read.

Revelation 6, 7 and 8 carry the same verse numbers in the Greek file this page measures and in the ESV.

The one verse of difference is at 12:18, outside this study: the Greek file numbers as its own verse a clause (and he stood on the sand of the sea) that English Bibles print at the end of 12:17. Nothing in chapters 6–8 moves.

The offsets that do matter on this page are on the other side of the measurement. Greek Jeremiah is numbered quite differently from Hebrew Jeremiah, and this page’s headline finding cites it: Greek Jeremiah 51:13 = Jeremiah 44:13 in a Hebrew or English Bible. Greek Psalm 41:2 = Psalm 42:1 English. Greek and Hebrew 2 Chronicles 2:14 = 2 Chronicles 2:15 English. Both numbers appear every time.

The fourth seal names four ways to die: sword, famine, pestilence, and wild beasts of the earth. That set of four is Ezekiel’s. The wording of three of them is Jeremiah’s. The measurement can tell the two apart.

Ezekiel 14:21 has the whole set — ῥομφαίαν καὶ λιμὸν καὶ θηρία πονηρὰ καὶ θάνατον, sword and famine and evil beasts and death — but in the accusative and without the repeated ἐν, so not one two-word sequence survives the comparison. Revelation 6:8 has Ezekiel’s list and Jeremiah’s grammar. The page reports that and asserts nothing about how it happened.

How rare is the eight-word phrase? Here is the whole gradient.

So it is a Jeremianic formula that thins out fast, and only one row in 30,348 carries all eight. In the Greek New Testament the same eight words — and even the first three of them — stand in Revelation 6:8 and nowhere else.

A denarius was about a day’s wage; a χοῖνιξ was roughly a day’s grain ration for one man. So the third seal prices a day’s bread at a day’s pay for wheat, or three days’ worth of the cheaper barley. It is famine pricing without famine language: a man can eat, or a man can feed his household, but not both.

And then the voice protects the luxuries: do not harm the oil and the wine. Which is the strangest clause in the seal, and the page will not pretend to settle it. It has been read as the deep-rooted olive and vine surviving a drought that kills the shallow-rooted grain; as a rich man’s table untouched while the poor queue; and as an edict of Domitian on vineyards. None of those can be checked from the corpora held here, and the page names them as readings, not findings.

The measurement offers one thing, and it is a warning. This row’s deepest Old Testament run is six words, tie band one: καὶ τὸ ἔλαιον καὶ τὸν οἶνον, at 2 Chronicles 2:14 (2:15 in English) — Hiram writing back to Solomon about payment for the Lebanon woodcutters, the wheat and the barley, the oil and the wine. Wheat, barley, oil and wine in that order is simply how the ancient world listed staples. Six shared words with a band of one, and no relationship at all. Print it beside the eight-word run at 6:8 and the difference between a signal and an accident stops being a matter of opinion.

Revelation 7:5–8 names twelve tribes. Read straight off the Greek, in order:

Dan is absent. Ephraim is absent. Manasseh and Joseph both stand as tribes in their own right. Levi is counted. Judah is first. Those five facts are readable off Revelation 7 itself and none of them is disputed. That the Manasseh-and-Joseph pairing is unlike the Old Testament twelve-tribe lists — where Joseph either stands for both his sons or gives way to Ephraim and Manasseh — is a claim about the Old Testament rather than a reading of this passage, and it has not been screened here.

What the page will not do is explain the omission. Dan’s absence has been tied to the idolatry at Dan in Judges 18 and 1 Kings 12, to a patristic tradition that the antichrist would come from Dan, and to simple scribal loss. None of that can be measured from the corpora held here, and one of the three would explain the Ephraim omission and the other two would not. The page reports the list. It does not supply a motive.

What the measurement does turn up is small and pleasant. The phrase ἐκ φυλῆς Ἀσήρ, of the tribe of Asher, occurs in the Greek New Testament in exactly one other verse: Luke 2:36, of Anna the prophetess in the temple. And ἐκ φυλῆς Βενιαμίν occurs in exactly one other: Acts 13:21, of Saul the son of Kish.

The seventh seal is the loudest thing in the book and it opens onto silence. ἡμιώριον occurs once in the Greek New Testament and not once in the 30,348 verse units of Greek Old Testament held here.

Readers have connected the silence to the hush in the temple while the incense was offered, and to be silent, all flesh, before the Lord in Zechariah 2:13 and Habakkuk 2:20 and Zephaniah 1:7. Measured, none of those three reaches this row: Revelation 8:1’s deepest Greek Old Testament run is three words — ἐν τῷ οὐρανῷ — with a tie band of twenty-five, which is to say, a formula. The thematic connection may well be right; there is no verbatim contact to support it, and the page says both halves of that.

Structurally the silence does something plain: it is the seam. The seventh seal contains the seven trumpets, so the scroll does not finish being opened — it opens into the next cycle. The half hour is the space between the last seal and the first trumpet.

The image of wormwood is thoroughly Old Testament. The word is not.

So a reader working only from the Greek columns would never meet ἄψινθος before Revelation 8:11. What John does is use the ordinary Greek name of the plant, and then make it a proper name — ὁ Ἄψινθος, with the article, as the name of a star. Direction of influence is not asserted: the page does not say John rejected the Septuagint’s renderings, only that his word is not one of them.

Revelation has a signature sword and does not use it for the second horseman. Five of the book’s six occurrences of ῥομφαία are the sword from the mouth of Christ. The sixth is 6:8, in the fourth rider’s licence — and there it stands inside the eight words shared with Greek Jeremiah, where the Greek Old Testament regularly renders the Hebrew חֶרֶב with ῥομφαία.

The second horseman gets μάχαιρα μεγάλη instead — the shorter blade, the ordinary word for a soldier’s or a butcher’s knife, and the word Revelation uses again only at 13:10 and 13:14. The page notes the two words and does not build on the difference; Greek writers of this period use the two with considerable overlap, and one verse is not enough to prove a distinction.

Revelation 6:16 and Luke 23:30 both echo Hosea 10:8. Only one of them measures as a use of the Greek Hosea.

Read the three rows themselves and the arithmetic stops being surprising.

Greek Hosea 10:8 has the mountains told to cover and the hills told to fall. Luke 23:30 has both nouns and both verbs, with the two imperatives the other way round — mountains fall, hills cover — and the four words it shares run straight across the swap. Revelation 6:16 has no hills at all, puts rocks beside the mountains instead, and where Hosea and Luke both say καλύψατε, cover, Revelation says κρύψατε, hide.

So: two words, a different verb, and a missing noun. That is a common picture, not a common text. And the picture is a common one — the same terror stands in Isaiah 2:10, 2:19 and 2:21, and Revelation’s previous verse reaches Isaiah 2:19 at six. Nothing here says Revelation is quoting Luke, or Hosea, or Isaiah, and nothing says which way any borrowing would run. The page reports numbers.

And the deepest run this verse has is meaningless. Five words with a tie band of one, and the one verse in 30,348 that carries them is Exodus 12:29, where τοῦ καθημένου ἐπὶ τοῦ θρόνου, who sat on the throne, describes Pharaoh on the night the firstborn died. This page prints that against itself. A run of five with a band of one sounds decisive until you read what it landed on.

Revelation 6:2 and Revelation 19:11 share nine consecutive words, and no other pair of verses in the Greek New Testament shares them.

Whether the first horseman is Christ or a counterfeit of him is one of the oldest arguments about this book. The measurement does not settle it, and this page is going to say so plainly rather than let a nine-word run do work it cannot do.

Both readings survive the number. If the two riders are the same, the nine words are the reason. If they are not, the nine words are exactly what a deliberate counterfeit would look like — and the tenth word is where the two part: a bow against a name. A run measures contact, not identity. Notice also that the first horseman is described almost entirely in the same passive as the other three: a crown was given to him.

All 47 rows against 30,348 verse units of Greek Old Testament. The distribution first, so the deep rows can be read in proportion.

And the rows that reach five or more, every one of them, with the band:

Four of these thirteen are certainly accidents and the page says which. Solomon’s supply list (6:6), the storm that hit Jonah’s ship (8:8), the Persian governor’s colleagues (6:11) and the deer panting for the water-springs (8:10) have nothing to do with the seals or the trumpets. Every one of them has a tie band of one — the same band as the Jeremiah run at 6:8. A tie band of one tells you a phrase is rare. It does not tell you a text is being used. Two of the thirteen (7:11, 7:12) land in books this shelf does not treat as Scripture.

The same 47 rows against the other 7,880 verses of the Greek New Testament. Almost everything deep is Revelation talking to itself.

The one that goes outside the book is the instructive one. Revelation 6:15 and Acts 4:26 share six consecutive words with a tie band of one, and there is no contact between them whatever. Acts 4:26 is a quotation of Psalm 2:2; Revelation 6:15 reaches the same six words because it too stands on Psalm 2:2. Two verses can share a long, rare, exact run because they are both looking at a third text. That single row is the best short argument on this page for why a run is evidence of contact and not of dependence.

Compare the last row, too. The doxology at 7:12 reaches six words in seven different New Testament verses. Same run, band of seven instead of one, and it means nothing at all — it is how Greek-speaking Christians ended a prayer.

The screen finds verbatim contact. It does not find allusion, and on this passage that gap is unusually wide. Revelation 6:12–14 reads like Isaiah 34:4 and Joel 2:31 to any reader — sun black as sackcloth, moon like blood, stars falling, the sky rolled up like a scroll. Measured, 6:13 reaches Isaiah 34:4 at one word and 6:14 at two. The pictures are shared; the sentences are not. A page that reported only the runs would have to say Revelation 6:12–14 has no Old Testament background, and that would be false.

So this page has been careful to run the measurement in both directions. Where a famous connection measures low (Hosea 10:8, Isaiah 34:4) it says so and does not delete the connection. Where an unfamous one measures high (Jeremiah 51:13, 2 Chronicles 2:14) it says so and does not manufacture a connection. Neither number decides anything by itself.

A specific weakness, named. The headline finding — eight words, band of one — is a strong result about the Greek Old Testament as this shelf holds it. It is not a demonstration that Revelation used Greek Jeremiah. The phrase sword, famine and pestilence is a Jeremianic commonplace in Hebrew as well; the eight-word Greek form could be reached by a writer who knew the formula in either language, and the page cannot distinguish those. Direction of influence is never asserted here.

No critical apparatus was consulted, for either testament. This matters more on Revelation than on most books, and the page has three known exposures. (1) At 8:1 the file measured here reads ὅταν where many editions read ὅτε, and the editor flags the word. (2) At 6:1 the living creature says Ἔρχου alone; older English Bibles read come and see, from a longer Greek text. (3) At 8:13 the creature crying in mid-heaven is an ἀετός, an eagle; a body of manuscripts reads ἀγγέλου, angel, which is where the King James “angel flying” comes from. The page reports the file it measures and does not adjudicate any of the three.

No Latin, Syriac or Coptic witnesses. No papyri. No Second Temple apocalyptic sources — 4 Ezra, 2 Baruch and 1 Enoch are the obvious comparanda for a seal-opening vision, a fixed number of the righteous and a delay until the count is full, and none of them was consulted. No Roman-period evidence on grain prices, on Domitian’s vine edict, or on the Parthian mounted archer sometimes read behind the first horseman.

The measurement corpus is not a canon, and the page has flagged where that bites. The 30,348 verse units include the deuterocanonical books and the Psalms of Solomon. Two of this passage’s thirteen deepest rows land in 4 Maccabees and Judith, and both are labelled in the table above. A hit in those files is a fact about Greek, not about Scripture.

⚠️ THE CORPUS THIS PAGE IS MEASURED AGAINST GOT SMALLER ON 31 AUGUST 2026, AND NOT ONE NUMBER ON THIS PAGE MOVED. THE PAGE PRINTS THE HISTORY RATHER THAN SWAPPING THE FIGURE IN SILENCE. The Odes — the fourteen liturgical canticles bound after the Greek Psalter, which arrived with the last twenty CATSS files — have been taken out of everything this shelf calls “the Greek Old Testament” for screening purposes. They are still held on the disk. They are no longer screened. Two reasons, either of which would be enough: Ode 9 is Luke 1 (the Magnificat and the Benedictus) and Ode 13 is Luke 2 (the Nunc Dimittis), so a screen that includes them can report a Gospel’s deepest Old Testament contact when what it has found is Luke matching Luke; and ten more of the canticles are second copies of passages already in the corpus (Ode 1 = Exodus 15, 2 = Deuteronomy 32, 3 = 1 Samuel 2, 4 = Habakkuk 3, 5 = Isaiah 26, 6 = Jonah 2, 7 and 8 = Daniel (Th) 3, 10 = Isaiah 5, 11 = Isaiah 38), so any tie band touching one of them was doubled. HELD is 30,637 verse units in 64 files; SCREENED is 30,348. The difference is exactly the 289 Odes units, and every screening figure on this page now names the second number.

⚠️ And here is the whole effect of that on this page, which is almost nothing — which is worth printing too. Re-run over 30,348 units, every run and every tie band published above is unchanged: 6:8 is still eight at Jeremiah 51:13 in a band of one, 6:6 still six at 2 Chronicles 2:14, 6:15 still six across Isaiah 2:19 and Psalm 2:2, 7:12 still six at 4 Maccabees 18:24, 6:16 still five at Exodus 12:29, 8:3 still five at Numbers 4:11, and the whole distribution table — 1, 1, 4, 13, 28, 45, 47 — stands exactly as printed. Not one hit on this page was in Ode 9 or Ode 13, so nothing here was Luke measured against itself. Three tie bands narrowed, and none of the three is a figure this page prints: 6:10’s band of 258 is now 253, 6:12’s 18 is now 17, 7:10’s 26 is now 25. A page whose findings survive a corpus getting smaller is a page whose findings were not resting on the corpus’s edges, and that is the only thing this correction has to report.

A note on the corpus, dated. The Greek Old Testament on this shelf went from 44 CATSS files to 64 — the complete upstream lxxmorph set — on 30 August 2026, hours before this page was written. Four of this page’s findings come from files that were not here the day before: Exodus 12:29 at 6:16, Numbers 4:11 at 8:3, Judith 5:8 at 7:11 and 4 Maccabees 18:24 at 7:12. Any figure on any Deep Research page that counts occurrences “across the Greek Old Testament held here” is a figure about a corpus of a particular size on a particular date, and this one names both.

It does not tell you who the first horseman is. Nine words with chapter 19 is the strongest datum anyone can put on that table, and it cuts both ways. The page prints it and stops.

It does not tell you who the 144,000 are, or whether they are the great multitude. It shows the hear/see structure and the absence of Dan, and leaves the identification to the reader and the church.

It does not periodise the seals. Whether these are first-century events, a recurring pattern of history, or a strictly future sequence is the central question readers bring to Revelation 6, and nothing in a run screen touches it. Pleasant Springs holds a historic premillennial position; this page does not argue it and does not assume it, and a reader from another position will find nothing here that has been decided against them by measurement.

• G. K. Beale, The Book of Revelation (NIGTC) — the fullest treatment of the Old Testament in Revelation.
• David E. Aune, Revelation (Word Biblical Commentary, 3 vols.).
• Craig R. Koester, Revelation (Anchor Yale Bible 38A).
• Richard Bauckham, The Climax of Prophecy — on the seals, the trumpets and the interludes as a structure.
• Beale & McDonough, “Revelation,” in Commentary on the New Testament Use of the Old Testament.
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer. Where this page disagrees with a commentary it is disagreeing about a number, not about a judgement.

There are three questions in these three chapters. The martyrs ask the first from underneath the altar — how long? The kings and the generals and the slaves ask the second from inside the caves — who can stand? The third is the elder’s, and it is the only one answered in words: who are these, clothed in white robes?

The second question is answered too, but not with a sentence. The verb of 6:17 is σταθῆναι, from ἵστημι. It stands 21 times in Revelation, and after 6:17 the next five occurrences are 7:1, 7:9, 7:11, 8:2 and 8:3 — every one inside this passage, and nothing further until chapter 10. Four angels standing at the corners of the earth. A multitude no one could number standing before the throne. All the angels standing around it. Seven angels who stand before God. Another angel who stood at the altar.

The kings ask who can stand, and the book answers by showing five sets of people who are standing — and one of them is a crowd out of every nation, in white, holding palms. They are not standing because they escaped anything. The elder says they came out of the great tribulation, and that their robes are white because they were washed in blood. The answer to “who can stand” is not “the strong.”

Two chapters of thunder, hoofbeats, earthquake, a great multitude crying out with a loud voice, myriads of angels. Then the seventh seal, the last one, the one the whole book has been waiting for — and silence in heaven for about half an hour.

Then the seven angels are given seven trumpets, and an angel comes with a golden censer and stands at the altar — the same altar the souls were under at 6:9 — and the smoke of the incense goes up with the prayers of all the saints. And then he fills the censer with fire from that altar and throws it down.

What goes up from the altar is the prayers. What comes down from the altar is the answer. “How long?” was asked at 6:10 and is not replied to. Two chapters later it comes back as thunder and lightning and an earthquake. The half hour of silence sits between the question and the answer, and it is the only measured length of time in the whole sequence.

One of nine Revelation studies on this shelf. It follows 1:9–18, 2:1–3:22 and 5:1–14, and is followed by 9:1–11:19, 12:1–14:20, 15:1–18:24, 19:1–20:15 and 21:1–22:5. And the tear that God wipes away at 7:17 is wiped away again, in six of the same words, at 21:4 — the first time in the middle of the trouble, the second time after it is over.