Out-Copied

An editor’s ending, added after the fact — which then appeared in almost every copy anyone ever made

R846’s Paris Liberty edited her big sister: the tablet in her arm pairs two revolutionary dates where New York’s carries one, so a mould taken from Paris transmits the revision as faithfully as the figure. Its closing turn asked for a kept thing whose copies inherit an editor’s addition until the addition is simply what the thing says. The clearest case is not a statue. It is twelve verses at the end of a Gospel.

The abrupt source. The Gospel of Mark, in its earliest surviving complete Greek copies, stops at chapter 16, verse 8: the women flee the empty tomb in fear and say nothing to anyone. No appearance of the risen Jesus, no commission, no ascension — the book ends, in the usual translation of its last words, “for they were afraid.” The two great fourth-century codices, Sinaiticus and Vaticanus, both end there.

The addition. Verses 9–20 — the so-called longer ending, a compact summary of resurrection appearances, a commission to preach, promised signs, and an ascension — are, on the prevailing scholarly view, a second-century addition, composed from material circulating in the other Gospels and Acts. The evidence is of several kinds at once: the earliest complete codices lack the verses; Eusebius, writing around 325, reported that the accurate copies ended at verse 8 and that the longer ending was absent from almost all copies known to him, and Jerome later said much the same; the vocabulary and style diverge from the rest of Mark; and a rival, much shorter ending also survives in a handful of manuscripts — in some, copied back to back with the longer one, as if a scribe who met both refused to choose. Modern critical editions print the twelve verses in double brackets, the editors’ highest-confidence mark that a passage is secondary.

And yet it won the count. Here is the turn that answers R846. The addition is old: Irenaeus was already quoting its verse 19 as Mark’s by about 180, well before Sinaiticus and Vaticanus were copied. And once it was in circulation it out-reproduced the ending it replaced almost completely. Of the surviving Greek manuscripts of Mark, over 1,600 contain the longer ending; only a handful clearly end at verse 8. Quoted in the textual literature, the Alands put the longer ending in 99 percent of the Greek manuscripts, enjoying for centuries what they call practically official ecclesiastical sanction. For most readers across most of Christian history, Mark did not end in frightened silence. It ended with the addition — because the addition is what their copy contained, and a copyist copies the book in front of them, not the book behind it.

Two kinds of verdict, both kept. So the thing now carries its dispute inside itself, permanently. By descent, the verdict is not close: count the copies and the variant is the text, at roughly ninety-nine to one. By age and weight of witness, the verdict is also not close, in the opposite direction: the earliest complete copies, the reported testimony of Eusebius and Jerome, and the internal evidence all point to a book that ended at verse 8. Scholarship did not resolve this by deleting the winner. It resolved it by printing the winner inside brackets and a footnote — the addition kept, labelled as an addition, in the same book. The tablet Paris added to Liberty can at least be seen. This addition is seen mainly by people who read footnotes.

The toy. Schematic only — a tally, not a simulation. Walk the witnesses in the order the evidence arrives and watch the share held by the longer ending flip. The first count is the two earliest complete Greek codices (0 of 2 contain it). The last is the reported tally of surviving Greek manuscripts (over 1,600 contain it, a handful clearly do not; the figure below uses 1,643 against 3, which computes to about 99.8 percent). Nothing between those counts is a census — the middle steps are attestations, not tallies — and that is the point the toy exists to make: a variant’s share of copies is a fact about its descendants, not about its birth.

Before any witness: two endings in circulation, one of them written by the author, one of them not yet distinguishable by count alone.

What is actually kept. Paris’s revision travelled by mould: faithful, visible, one statue at a time. This revision travelled by the copying rule itself. Every scribe who received a book ending at verse 20 handed on a book ending at verse 20, and the addition needed no argument, no permission and no occasion — only exemplars. That is how an editor’s afterthought becomes, for a millennium of readers, simply what the thing says: not by convincing anyone, but by being present in the copies from which all later copies were made. The scholars’ brackets are the first widespread attempt to keep the thing and its history in one object at once — the text as received, with the seam still showing.