The Little Apocalypse, Measured

• Dictionary of Jesus and the Gospels (IVP) — “Apocalyptic Teaching,” “Destruction of Jerusalem,” “Eschatology,” “Olivet Discourse,” “Son of Man,” “Temple,” “Persecution.”
• Anchor Bible Dictionary — “Abomination of Desolation,” “Temple, Jerusalem,” “Son of Man,” “Apocalypticism,” “Daniel, Book of” (including the Old Greek / Theodotion problem).
• NIDNTTE — βλέπω, γρηγορέω, ἐκλεκτός, θλῖψις, βδέλυγμα, ἐρήμωσις, ὠδίν, γενεά.
• Dictionary of New Testament Background (IVP) — “Apocalyptic Literature,” “Antiochus IV Epiphanes,” “Destruction of Jerusalem,” “Testament Literature” (for the farewell-discourse genre).

• R. T. France, The Gospel of Mark (NIGTC) — the fullest defence of reading most of the chapter of AD 70; Adela Yarbro Collins, Mark (Hermeneia); Joel Marcus, Mark 8–16 (AYB); Robert Guelich / Craig Evans (WBC); William Lane (NICNT).
• George R. Beasley-Murray, Jesus and the Last Days: The Interpretation of the Olivet Discourse — a full history of interpretation of this chapter alone.
• Timothy J. Geddert, Watchwords: Mark 13 in Markan Eschatology; Morna Hooker, The Son of Man in Mark.
• On the Greek Daniel: Alexander Di Lella and Louis Hartman, Daniel (AB); T. McLay, The OG and Th Versions of Daniel.

The disciple in v.1 is not being naive. Herod’s temple platform was one of the largest building projects in the ancient world. Josephus describes stones “forty-five cubits in length, five in height, six in breadth” (War 5.184–227; cf. Antiquities 15.392–402, which gives smaller figures for the sanctuary stones), and the rabbinic saying ran that whoever had not seen Herod’s building had never seen a beautiful building (b. Bava Batra 4a). The Greek ποταποὶ λίθοι — “what sort of stones!” — is the reaction the architecture was built to produce.

The “abomination of desolation” had a history before Mark, and it was not a prediction — it was a memory. In 167 BC Antiochus IV Epiphanes had a pagan altar set on the altar of burnt offering. 1 Maccabees 1:54 reports it in Greek that our screen can read: ᾠκοδόμησεν βδέλυγμα ἐρημώσεως ἐπὶ τὸ θυσιαστήριον — “he built an abomination of desolation upon the altar.” The book of Daniel names the same thing in three places (9:27; 11:31; 12:11). So when Mark 13:14 says the abomination of desolation, the phrase is already a technical term with a date attached to it.

Josephus supplies two candidates for a first-century recurrence, and neither settles the verse. In AD 40 Caligula ordered his statue erected in the temple; only his death prevented it (Antiquities 18.261–309; Philo, Legatio 188, 207–348). And during the revolt Josephus says the Zealots installed a clown of a high priest and filled the sanctuary with bloodshed, which he himself calls the fulfilment of an ancient oracle about the city being taken when its own people defiled the sanctuary (War 4.147–157, 388). Mark says only ὅπου οὐ δεῖ, “where it ought not to be,” and uses a masculine participle for a neuter noun. He is being less specific than his sources, not more.

The farewell discourse was a recognised literary form. Jacob in Genesis 49, Moses in Deuteronomy 31–33, Joshua in Joshua 23–24, and the whole genre of the Testaments of the Twelve Patriarchs: a leader about to die gathers his own, predicts a time of trouble, warns of deceivers, and issues commands. Mark 13 sits in that form exactly — which is worth saying, because it means the chapter’s first job is not to inform but to fortify.

The four watches in v.35 are Roman, not Jewish. Jewish reckoning divided the night into three watches (Judg 7:19; Lam 2:19); the Roman army divided it into four. Mark names all four — ὀψέ, μεσονύκτιον, ἀλεκτοροφωνία, πρωΐ — and elsewhere uses the Roman scheme too (6:48, “about the fourth watch”). For a Gentile congregation that is simply the way you say “at any hour of the night.”

The chapter is organised by a verb of perception, and it never becomes a verb of calculation. Four times — vv.5, 9, 23, 33 — the discourse stops and says βλέπετε. Not reckon, not watch the sky: look out for yourselves. The one indicative use, v.2, is Jesus asking whether they see the buildings; the four imperatives are all about not being taken in. In a chapter that everyone reads for a timetable, that is a striking distribution.

Three New Testament hapax legomena in thirty-seven verses is a high density, and all three sit in the same semantic field: προμεριμνάω (do not rehearse your defence), ἀπόδημος (the master is away, indefinitely), ἀλεκτοροφωνία (one of four unpredictable hours). Mark reached for unusual words precisely where the subject is time you do not control.

And ἐκλεκτός is this chapter’s private property. Mark uses it three times and nowhere else. It matters that all three uses are protective rather than triumphal: the days are shortened for them, the deceivers aim at them, and they are collected from the four winds. Mark never defines who they are and the chapter never tells anyone to check whether they are on the list.

The distribution first, because it is the finding. Across the 37 verses of Mark 13 screened against the whole Greek Old Testament: 36 of 37 reach two, 28 reach three, 17 reach four, 2 reach five, 1 reaches six — and nothing reaches seven. For comparison, on this shelf’s Matthew 22 page a single verse reached fourteen; on the Matthew 13 page, twenty-four. Mark 13 has no citation formula anywhere in it, and the numbers agree: the chapter’s ceiling is six.

Now the part that is genuinely surprising. Mark 13 is called the little apocalypse and is read everywhere as the Gospel’s Daniel chapter. But the deepest verbal contact any verse of it makes with the Greek Old Testament is Isaiah 13:10, at six words, tie band 1 — καὶ ἡ σελήνη οὐ δώσει τὸ — and after that the chapter simply stops going deeper.

Put the whole Gospel behind it and the contrast sharpens. We ran the same screen over all 673 verses of Mark. Its deepest Old Testament contacts are 12:30 at 21 words (Deuteronomy 6:5), 12:36 at 14 (Psalm 109:1 LXX = ESV/MT 110:1), 1:3 at 13 (Isaiah 40:3), then 12:10–11 at 10 apiece (Psalm 117:22–23 LXX = ESV 118:22–23) and 10:7 at 10 (Genesis 2:24). Not one of Mark’s fifteen deepest Old Testament contacts is in chapter 13. Chapter 13’s ceiling of six is beaten by chapters 1, 3, 7, 10, 11 and 12, matched by 2, 5 and 14, and beaten downward by 4, 6, 8, 9, 15 and 16 — so it is not the shallowest chapter in Mark. What the number says is that a chapter everyone reads as saturated with scripture measures like Mark’s ordinary narrative, because it alludes constantly and quotes never.

Which is exactly the limit this method has, and it is worth stating in the strongest form. The screen finds verbatim contact. It does not find allusion. Nobody reading Mark 13:26 doubts that it is Daniel 7:13. The measurement says 1.

Daniel is the one Old Testament book that survives in two substantially different ancient Greek translations: the Old Greek and the version attributed to Theodotion, which is the form that displaced the Old Greek in Christian manuscripts almost completely. We hold both, as separate files. That makes a question askable here that a single-text screen cannot ask.

• Mark 13:14 matches the Old Greek and not Theodotion. The articled phrase τὸ βδέλυγμα τῆς ἐρημώσεως occurs in exactly one verse of the entire Greek Old Testament held here — Daniel 12:11, Old Greek. Theodotion at 12:11 reads βδέλυγμα ἐρημώσεως without the articles, and the measured run from Mark drops from 4 to 1. The articleless form does occur elsewhere — Old Greek Daniel 11:31, Theodotion Daniel 12:11, and 1 Maccabees 1:54 — and Daniel 9:27 in both versions has the plural, βδέλυγμα τῶν ἐρημώσεων.
• Mark 13:19 matches Theodotion and not the Old Greek. θλῖψις οἵα οὐ γέγονεν runs 4 against Theodotion Daniel 12:1, whose Greek reads ἔσται καιρὸς θλίψεως θλῖψις οἵα οὐ γέγονεν. The Old Greek of the same verse reads ἐκείνη ἡ ἡμέρα θλίψεως οἵα οὐκ ἐγενήθη, and the run drops to 2.
• Mark 13:26 matches neither. Against Daniel 7:13 the run is 1 in the Old Greek and 1 in Theodotion. Mark writes τὸν υἱὸν τοῦ ἀνθρώπου ἐρχόμενον ἐν νεφέλαις; both Greek Daniels write ὡς υἱὸς ἀνθρώπου without articles and τῶν νεφελῶν τοῦ οὐρανοῦ in the genitive plural. Every content word is cognate. Not two of them are consecutive and identical.

A page that prints only the numbers supporting its thesis is not measuring anything. Here are two results from the same screen that cut the other way.

• The Son of Man verse’s deepest match is a battle report. Mark 13:26 — the most obviously Danielic sentence in the chapter — returns a maximum of 4 with a tie band of 3, and the three verses are 1 Maccabees 9:60, 1 Maccabees 12:42 and Isaiah 36:2. The shared four words are μετὰ δυνάμεως πολλῆς καί, “with a great force and…” — in 1 Maccabees, Bacchides marching on Israel and Tryphon marching on Scythopolis; in Isaiah, the Rabshakeh arriving at Jerusalem. Nobody thinks Mark is quoting a Seleucid troop movement. The phrase is ordinary military Greek that happens to sit next to a genuine allusion the screen cannot see.
• A tie band of one does not mean a connection. Mark 13:17 — “alas for those who are pregnant and nursing in those days” — returns a run of 4 with a band of 1, which looks like a find until you read the match. It is Judith 8:1, and the shared words are ἐν ἐκείναις ταῖς ἡμέραις, the commonest time-marker in Greek scripture, which happens to be uniquely spelled that way in that one verse of the corpus. Uniqueness is a property of the corpus searched, not of an author.

And one observation offered as nothing more than an observation. Mark 13:3 sets the discourse εἰς τὸ Ὄρος τῶν Ἐλαιῶν κατέναντι τοῦ ἱεροῦ. The phrase τὸ ὄρος τῶν ἐλαιῶν occurs in exactly one verse of the whole Greek Old Testament held here — Zechariah 14:4, where the LORD’s feet stand on it and it splits in two — and Zechariah’s next word is τὸ κατέναντι, “the one opposite.” Mark’s next word is κατέναντι. The run stops at four because of one definite article. We are not claiming Mark is alluding to Zechariah 14 here; the Mount of Olives is a real hill east of a real temple and there is nowhere else to sit and see it, and the same phrase occurs in eight New Testament verses as plain geography. It is noted so a reader can weigh it, not lean on it.

The same 37 verses were then screened against the whole Greek New Testament outside Mark — 7,254 verses. Here the numbers are much larger, as they must be, because the synoptic parallels are not allusions.

And the two deepest are not where anyone would look for them.

• Mark 13:13 — run of 16, tie band 1 — against Matthew 10:22. That is the entire verse. καὶ ἔσεσθε μισούμενοι ὑπὸ πάντων διὰ τὸ ὄνομά μου. ὁ δὲ ὑπομείνας εἰς τέλος οὗτος σωθήσεται stands word for word in both, and Matthew 10 is the mission discourse, not the Olivet discourse.
• Mark 13:12 — run of 15, tie band 1 — against Matthew 10:21. Again the whole verse, differing only in the opening conjunction (καὶ παραδώσει against παραδώσει δέ).
• Matthew has the same material in his own chapter 24, but split and modified: 24:9 carries the first half of Mark 13:13 with τῶν ἐθνῶν added, and 24:13 carries the second half verbatim. So the two deepest matches Mark 13 makes in the New Testament are with Matthew 10, and its Matthew 24 counterparts are shallower.
• The rest of the top band is where one would expect: 13:17 = 13 (Matt 24:19); 13:31 = 13 (Luke 21:33); 13:28 = 12 (Matt 24:32); 13:24 = 11 (Matt 24:29); 13:29 = 11 (Luke 21:31); 13:30 = 10, band 2 (Matt 24:34 and Luke 21:32). All tie bands as stated.

• R. T. France, The Gospel of Mark (NIGTC); Joel Marcus, Mark 8–16 (Anchor Yale); Adela Yarbro Collins, Mark (Hermeneia); William Lane (NICNT); Craig Evans, Mark 8:27–16:20 (WBC); Morna Hooker (BNTC).
• George R. Beasley-Murray, Jesus and the Last Days — a full history of the interpretation of Mark 13, including the “little apocalypse” source hypothesis of Timothée Colani and his successors, and its critics.
• Timothy J. Geddert, Watchwords: Mark 13 in Markan Eschatology; Rikki Watts, Isaiah’s New Exodus in Mark — relevant to this page’s Isaiah result, though Watts’ argument is thematic rather than verbatim.
• Josephus, Jewish War 4.147–157, 388 and 6.249–270; Antiquities 15.392–402 and 18.261–309; Philo, Legatio ad Gaium 188, 207–348 — the primary sources behind Stage 2.
• Classic: Chrysostom, Homilies on Matthew 75–77 (on the parallel); Augustine, Letter 199 (De fine saeculi), which turns on the saying in Mark 13:32 and on Acts 1:7; Calvin, Harmony of the Evangelists.

• What is “this generation” (v.30)? The major options are the contemporaries of Jesus (with the temple’s fall in AD 70 as the referent of ταῦτα πάντα), the Jewish people as a people, the human race, or the generation that sees vv.14–23 begin. Preterist and futurist readings divide here, and so do serious commentators. Stage 3 gives the only measured datum we have — two of Mark’s four earlier uses of γενεά carry an explicit pejorative, and all four stand in a rebuke — and it does not settle the question.
• Does the discourse answer one question or two? The disciples ask when and what sign; France argues vv.5–31 concern the temple and vv.32–37 the parousia; others read a single fused horizon throughout. The chapter marks a break at v.32 with περὶ δέ, which is real Greek evidence but not decisive — the same construction elsewhere marks a new topic without changing the referent.
• “Nor the Son” (v.32). The clause is textually secure in Mark and has been theologically uncomfortable since at least the fourth century. It says the Son does not know the day. It does not say why, and this page will not manufacture an answer; the classical discussions (Athanasius, Orations against the Arians 3.42–50; Augustine, Letter 199) are worth reading in full rather than in summary.
• The masculine participle in v.14. βδέλυγμα is neuter; Mark writes ἑστηκότα, a masculine accusative singular, which many read as pointing to a person rather than an object. Matthew 24:15 uses the neuter ἑστός. This is a real grammatical fact and it does not, by itself, identify anybody. Note also that the SBLGNT flags a variant at that very word.
• The date of Mark. Whether the chapter was written before or after AD 70 changes how one reads “let the reader understand,” and it cannot be settled from the chapter. Notably, the flight in v.14 is to the mountains, which is not what happened in the siege, and Luke 21:20 replaces the abomination with armies encircling the city. Both facts are used on both sides of the argument.

• No Hebrew screen was run for this page. Every measurement above is Greek against Greek. The Hebrew of Isaiah 13:10, Joel 2:10 and Daniel 12:1 is relevant to the Stage 4 argument about φέγγος and we quote it in the next section, but Mark 13 was not screened against the Masoretic corpus and no run length on this page rests on Hebrew wording.
• We hold one Greek Old Testament text per book, not a manuscript tradition. The screen runs against the CATSS/Rahlfs files on disk. Daniel is the one place where that limitation is partly lifted, because we hold two versions — and Stage 4 leans on precisely that. But we cannot check the witnesses behind either. The Old Greek of Daniel survives in a very small number of witnesses (essentially Papyrus 967 and the Syro-Hexapla tradition), which means the single most striking result on this page rests on the thinnest textual base on this page. We flag it rather than burying it.
• We did not consult a critical apparatus for Mark. The SBLGNT on our disk carries its editor’s variant markers — the chapter carries 49 of them, counted here — but we hold the markers, not the apparatus. Two of them bear on things said above: the participle at v.14, and the word order at v.25. Anyone building on those two sentences should open an apparatus, and we did not.
• The Isaiah 13:10 result is a run of six, and six is not a lot. It is the deepest in the chapter, and its tie band is 1, and both of those are real. But this page is not claiming that Mark 13:24 is a citation of Isaiah 13:10, because six words with no citation formula does not establish that. What it establishes is a ranking: within this chapter, the Isaiah contact is the deepest one there is.
• The tie band is a property of the corpus searched, not of the ancient world. A band of one over 30,637 verse units is not a claim about every Greek text that existed. The deuterocanonical books are included in the count; so are both Greek Daniels, both Greek Judges and both Greek Joshuas — and both forms of Susanna, Bel and Tobit — which is why some tie bands here are larger than they would be on a canon-only screen.
• We did not read the rabbinic sources in the original. The reference to b. Bava Batra 4a in Stage 2 comes from standard secondary discussion and postdates Mark by centuries in its compiled form.
• Nothing here establishes direction of influence, and the page never asserts one. That Mark 13:13 stands verbatim in Matthew 10:22 is a measured fact about two texts. Which way the material travelled, or whether both drew on a common form, is a synoptic question a run-length screen cannot decide.
• What we deliberately did not do: we did not date the chapter, we did not identify the abomination with a historical figure, and we did not build a timeline. Those are the three things readers most want from Mark 13 and they are the three things the measurement cannot supply.

• The chapter is not a timetable, and its own vocabulary says so. Four imperatives, and every one of them is βλέπετε — watch out, not work it out. The one thing the disciples asked for, a sign and a date, is the one thing v.32 says nobody has.
• The deepest Old Testament contact in the chapter is six words, and it is Isaiah. That is a measured fact, not a devotional flourish. Mark 13 alludes constantly and quotes never — which is why it feels saturated with scripture and measures shallow. Both things are true at once, and holding them together is what honest reading looks like.
• “Let the reader understand” is the only time the discourse turns and speaks to you. Not to Peter, James, John and Andrew — to whoever is holding the scroll. Mark thought his readers would need to recognise something, and he did not spell out what.
• The elect are never told to identify themselves. All three occurrences of ἐκλεκτός in Mark are here, and all three are protective: the days are shortened for them, the deceivers aim at them, they are gathered from the four winds. No one in this chapter is invited to audit the list.
• The last word is γρηγορεῖτε, and it is broken by morning. Mark uses the verb six times: three at the end of this chapter, three in Gethsemane. The men told to stay awake are the men who fall asleep, and Mark puts the two scenes one chapter apart on purpose. Whatever this discourse is asking for, three of the four men who first heard it could not do it.

Step in response: read Mark 13:32–37 and Mark 14:32–42 back to back, out loud, in one sitting. Then ask what you have been calculating that you were told instead to watch.

Set the two texts side by side. Septuagint Isaiah 13:10 (no offset — Hebrew, Greek and English agree on the number):

… καὶ σκοτισθήσεται τοῦ ἡλίου ἀνατέλλοντος καὶ ἡ σελήνη οὐ δώσει τὸ φῶς αὐτῆς

Mark 13:24:

… ὁ ἥλιος σκοτισθήσεται, καὶ ἡ σελήνη οὐ δώσει τὸ φέγγος αὐτῆς

The shared run is καὶ ἡ σελήνη οὐ δώσει τὸ — six forms, tie band 1 over 30,637 verse units. It is stopped at the front because Isaiah has the sun in a genitive absolute (τοῦ ἡλίου ἀνατέλλοντος, “as the sun rises”) while Mark makes ὁ ἥλιος the subject of the same verb. And it is stopped at the back by one noun.

Isaiah’s Greek reads τὸ φῶς αὐτῆς. Mark reads τὸ φέγγος αὐτῆς. Both mean the moon’s light; they are different words.

Where does φέγγος live? In the Greek Old Testament held here it occurs in 24 verse units, 25 times (Ezekiel 1:4 has it twice). Two of them are in Joel, and both are about the same event:

• Joel 2:10 LXX — ὁ ἥλιος καὶ ἡ σελήνη συσκοτάσουσιν καὶ τὰ ἄστρα δύσουσιν τὸ φέγγος αὐτῶν — “the sun and the moon will be darkened and the stars will withdraw their radiance.” (No offset here: LXX/MT/ESV all number this 2:10.)
• Joel 4:15 LXX = ESV Joel 3:15 — ὁ ἥλιος καὶ ἡ σελήνη συσκοτάσουσιν καὶ οἱ ἀστέρες δύσουσιν φέγγος αὐτῶν. Hebrew and Greek Joel have four chapters, English has three, so this verse is Joel 3:15 in an English Bible.
• The measured runs from Mark 13:24 to those verses are 3 in each case (καὶ ἡ σελήνη). By the numbers, Joel is a shallow contact and Isaiah is the deep one.

So the sentence is Isaiah’s clause carrying Joel’s noun — and we want to be exact about what that is and is not. It is not proof that Mark spliced two scrolls. Two prophets describing the same darkened sky in similar Greek is not a surprising thing to find, and φέγγος is a perfectly ordinary word that also appears in Job, Ezekiel and Habakkuk. What is measurable, and re-derivable, is the shape of the overlap: six consecutive words shared with one verse and no other, broken at precisely the point where Mark uses the other prophet’s vocabulary for the same object. That is the whole finding, and it is worth exactly that much.

A note on the Hebrew, since we did not screen it: Isaiah 13:10 has the moon not causing its ’or to shine, while Joel 2:10 has the stars withdrawing their nogah. The Greek distinction between φῶς and φέγγος tracks a distinction that is already there in the Hebrew. We state that as background, not as a measurement — no Hebrew screen was run for this page.

This is the sharpest thing the screen found, and it is worth its own paragraph.

Mark 13:26 — καὶ τότε ὄψονται τὸν υἱὸν τοῦ ἀνθρώπου ἐρχόμενον ἐν νεφέλαις μετὰ δυνάμεως πολλῆς καὶ δόξης. Against Daniel 7:13 the run is 1, in the Old Greek and in Theodotion alike.

Mark 14:62, the same speaker, one chapter later, answering the high priest — ὄψεσθε τὸν υἱὸν τοῦ ἀνθρώπου ἐκ δεξιῶν καθήμενον τῆς δυνάμεως καὶ ἐρχόμενον μετὰ τῶν νεφελῶν τοῦ οὐρανοῦ. Against Theodotion Daniel 7:13 the run is 5, tie band 1 — μετὰ τῶν νεφελῶν τοῦ οὐρανοῦ — and against the Old Greek it is 4, because the Old Greek reads ἐπὶ τῶν νεφελῶν. We screened Mark 14:62 against the whole Greek Old Testament: Theodotion Daniel 7:13 is that verse’s single deepest contact in all 30,637 verse units.

So: one chapter apart in the same Gospel, on the same subject, one sentence touches the Greek Daniel and the other does not. The difference is one preposition and one article-plus-genitive: ἐν νεφέλαις against μετὰ τῶν νεφελῶν τοῦ οὐρανοῦ.