Where the manuscripts differ · Mark 16:9-20
The Ending of Mark
The last twelve verses of Mark are in the great majority of Greek manuscripts and are quoted by Irenaeus around AD 180 — and they are missing from the two oldest complete Bibles we have.
Open most Bibles to the end of Mark and you will find a line of small type: a bracket, or a note that says some early manuscripts end the Gospel at verse 8. That note is not a modern invention and it is not an attack. It is the honest report of what the copies actually say.
Here is the situation in plain words. Mark 16:8 ends with the women running from the tomb, afraid, saying nothing to anyone. Then, in most manuscripts, twelve more verses follow: Jesus appears to Mary Magdalene, then to two disciples, then to the eleven; he gives the commission; he ascends. In our two oldest complete Greek Bibles, those twelve verses are simply not there.
Both of those facts are real. This page holds them both. It also carries the structural case that has been made for the passage — the argument from patterns of sevens — because the founder's instruction here was to show the structure and to show why critics still date the passage late, and to let the reader see the whole board.
What the manuscripts actually show
| Witness | Date | What it reads here |
|---|---|---|
| Codex Sinaiticus (א) | c. AD 330-360 | Ends the Gospel at 16:8, with no long ending of any kind. |
| Codex Vaticanus (B) | c. AD 300-350 | Ends the Gospel at 16:8 — and the scribe left the rest of the column, and the column after it, blank before starting Luke. Defenders read that blank space as a scribe who knew an ending existed and left room for it; critics read it as a scribe who knew other copies had something more and deliberately declined to write it. The blank is real; what it means is argued both ways. |
| Codex Alexandrinus (A) | 5th century | Contains the long ending, 16:9-20, without any mark of doubt. |
| Codex Bezae (D) | 5th century | Contains the long ending, in both its Greek and its Latin columns. |
| Old Latin Codex Bobiensis (k) | c. AD 400 | Has neither the familiar long ending nor a bare stop at 16:8, but a different, shorter ending — a brief paragraph about the women reporting to Peter and Jesus sending out the message of salvation. It is the only witness that gives the shorter ending alone. |
| The Sinaitic Syriac | 4th-5th century | Ends at 16:8. The passage is absent from this early Syriac translation. |
| Eusebius of Caesarea, in his letter to Marinus | early 4th century | Reports that the accurate copies of his day ended at 16:8, and that the passage was found in some copies but not in most. |
| Jerome, in his letter to Hedibia (Epistle 120) | c. AD 407 | Says nearly all the Greek copies he knew lacked the passage — though Jerome also included it in his Latin Vulgate, which is worth knowing when weighing what he meant. |
| Later Greek manuscripts that carry the passage but flag it | 9th century onward | A number of copies include 16:9-20 and mark it with asterisks or obeli — the scribal signs for material a copyist knew was questioned. The scribes who wrote it down were not always sure of it. |
The case each way
That it belongs in the text
The passage is in the overwhelming majority of Greek manuscripts. Well over a thousand copies contain it; a handful do not. If the count of witnesses carries weight — and in every other kind of historical inquiry it does — the long ending has an enormous majority behind it.
Irenaeus quotes it by around AD 180. In Against Heresies (3.10.6) he cites Mark 16:19 — "So then the Lord Jesus, after he had spoken unto them, was received up into heaven" — and attributes it to Mark. That is a bishop in Gaul, more than a century before Sinaiticus and Vaticanus were copied, treating these verses as Mark's. Whatever else is true, the passage is early.
Ending a book at "for they were afraid" is strange. The Greek is ἐφοβοῦντο γάρ (ephobounto gar), and gar is a connecting word — roughly "for" or "because." It normally looks backward to explain something, not forward to close a book. Ending a Gospel on it is unusual enough that readers have felt the jolt for eighteen centuries.
Beyond the grammar, it is strange as a story. Mark has promised a Galilee appearance (14:28; 16:7). A Gospel that announces the resurrection and then stops before anyone sees the risen Lord leaves its own promise unpaid.
The existence of several competing endings tells us something. Besides the long ending and the shorter ending, some manuscripts have both, and one family adds a further paragraph inside the long ending. The early church clearly did not find 16:8 a satisfying place to stop. That is at least consistent with the idea that something original was missing and people kept trying to supply it.
And a lost last leaf is a real physical possibility, not a convenient guess. The outermost sheet of a codex is the most exposed part of the book; it is the page that tears, frays, and goes missing. If Mark's own ending stood on that leaf in an early copy, its loss would produce exactly the situation we find — an abrupt stop plus a scramble of attempted repairs.
That it came later
The two oldest complete Greek Bibles in existence — Sinaiticus and Vaticanus — both end the Gospel at 16:8. They are independent of each other, and they agree.
They are not alone in their own century. The Sinaitic Syriac ends at 16:8, and so do some early Armenian and Georgian copies, which means the shorter form was not one local quirk but a reading present in several languages and regions.
The vocabulary and style do not read like Mark. Scholars counting the Greek find a cluster of words and phrases in these twelve verses that appear nowhere else in the Gospel, packed into a very short space. Mark's characteristic habits — his headlong "and immediately," his particular connectives — thin out abruptly.
Verse 9 reintroduces Mary Magdalene as though she had not just been on stage: "Mary Magdalene, from whom he had cast out seven demons." She was named in 15:40, 15:47, and 16:1. Introducing a character a second time, with an identifying note, is what a writer does when starting fresh material — not what Mark was doing three verses earlier.
The seam between verse 8 and verse 9 does not hold grammatically. In verse 8 the subject is the women; in verse 9 the subject silently becomes Jesus, who is not named in the Greek. A reader coming straight through has to supply a subject the sentence never gives.
Eusebius and Jerome — two of the most widely read scholars of the ancient church, working with manuscripts we no longer possess — both report that the accurate copies of their day ended at 16:8. That is direct testimony from people who had access we do not.
The competing endings cut both ways, and this is the sharpest point. A text circulating with a long ending, a shorter ending, both stacked together, and an extra insertion inside one of them does not look like a text with one original conclusion that a few copies lost. It looks like a gap that several hands independently tried to fill.
The structural case, and what it can carry
Some readers argue this passage carries a deliberate numeric structure. It is set out here at full strength, beside the methodological cautions that come with any pattern claim — because a passage can be carefully structured and still be judged a later addition. Those are two different questions, and neither one settles the other.
The pattern claim
There is a second kind of argument made for this passage, and MannaFest prints it because it is genuinely made and because the founder asked that the structure be shown. It is the argument from numeric design — the claim that the passage carries a pattern of sevens too dense to be accidental.
The framework comes from the older numeric-design tradition that E. W. Bullinger represents. In Number in Scripture (1894) Bullinger set out the position that the numbers of the Bible are not decoration but signature: seven above all, the number he associated with spiritual completeness, appearing in the structure of the text itself and not merely in what the text says. Bullinger applied the argument to this passage directly — in the Companion Bible he defended the genuineness of Mark's last twelve verses.
The detailed counting is associated with Ivan Panin (1855-1942), a Russian-born literary scholar trained at Harvard who spent roughly fifty years counting features of the Greek and Hebrew text. His published counts for Mark 16:9-20 report that the passage contains 175 words — that is, 7 × 25; that its vocabulary, counting each distinct word once, comes to 98 forms — 7 × 14; and that further counts within that vocabulary, including the letters composing it and the distribution of words used once against words repeated, likewise fall as multiples of seven.
The claim built on those counts is twofold. First, that a stack of independent sevens landing simultaneously in a passage of this length is not what accident produces. Second — and this is the part that bears on the manuscript question — that a later scribe attempting to round off an unfinished Gospel could not have produced such a stack while also writing readable Greek that says what these verses say. If the pattern is real, its proponents argue, the passage came from the same designing hand as the rest of the Gospel, whoever held the pen.
That is the case as its proponents make it, and it deserves to be heard in its own terms rather than waved off. It is an argument about design, made by people who believe the text is authored down to its structure — which is a conviction this site shares, even where the specific method is contested.
What the method can and cannot show
The cautions are serious, and honesty requires them to be laid out with the same care as the case.
First, and most sharply: the counts depend entirely on which Greek text you count. Change a spelling, include or drop a small connecting word, follow one manuscript family rather than another, and the totals move. At this passage the wording of the Greek is precisely what is in dispute — that is the whole question of the page. So a method that requires a fixed text in order to run cannot be used to settle which text is fixed. It assumes the thing it is brought in to prove.
Second, the features counted are chosen after the fact. Nobody fixed in advance the list of things that would count as a hit — total words, distinct forms, letters of the vocabulary, words used once, placement of names, and so on. The list is assembled by looking at the passage and noticing what lands. With enough countable features available, some pattern is available in any text of any length. This is not a hypothetical objection: the same style of counting has been run over non-biblical writing — secular prose, ordinary correspondence — and has turned up comparable clusters.
Third, Panin's counts were made against a Greek text he himself reconstructed, and the reconstruction was guided in part by the numbers. He published his own edition on that basis. His work has not been independently replicated under controlled conditions — that is, with the text fixed in advance by someone else, the features fixed in advance, and the counting done by hands that did not know what result was wanted. Until that is done, the counts remain his counts.
Fourth, a probability calculated after choosing what to measure is a different kind of claim from a probability calculated before. This site already handles that distinction carefully where prophecy and odds are discussed, and the same discipline applies here. Numbers presented as odds carry an authority they have not earned when the target was drawn around the arrow.
So here is the plain statement. This is a pattern claim, not a manuscript reading. MannaFest does not treat it as settling whether Mark wrote these verses, and no honest reader should. But MannaFest does not hide it either, and does not sneer at it. The structure people have found here is worth seeing, and so are the reasons it cannot do the work its warmest defenders want it to do.