The Gospel of Mary: What Is on the Missing Pages?

How much of the Gospel of Mary is missing?

Ten of nineteen pages. The only substantially complete copy of the Gospel of Mary is the Coptic text that opens the Berlin Codex — Papyrus Berolinensis 8502, tractate 1, a book of about the fifth century — and in that copy pages 1 through 6 and pages 11 through 14 are gone. Nine pages remain. In Mark M. Mattison's public-domain English translation, the one this library prints, those nine pages run to roughly 1,100 words; at the same density the missing ten would hold something near 1,200 more. The book we can read is a little under half of the book that was written.

That is a strange kind of damage, and worth noticing: not scattered holes across every leaf, but two clean blocks — one at the front, one through the middle. Quires of a codex come apart in blocks. What was lost was lost as physical bundles of leaves, probably long before the codex was found, which is why the surviving text starts and stops so abruptly rather than fading at the edges.

Where exactly does the text break off?

In two places, both mid-thought. The surviving text opens in the middle of a conversation already running, with a question whose asker is never named, and it breaks a second time in the middle of the Savior's answer about how visions are seen. Everything printed below is verbatim from the Gospel of Mary as this library prints it; the numerals inside the lines — 7 and 15 — are the manuscript's own page numbers, marking where a new page begins.

The first surviving words, at the top of page 7:

7 “Then will [matter] be [destroyed], or not?”

Six pages of book stood before that question mark, and the answer that follows — that every nature dissolves into its own root — is a reply to a discussion we cannot hear. The second break comes roughly 570 words further on, in the middle of Mary's report of what the Savior told her about the mind:

“In response the Savior said, 'They don't see in the soul or in the spirit, but the mind which [exists] between the two is [what] sees the vision [and] it [that …]

The sentence has no end and the quotation marks never close, because page 10 does not have a page 11 after it. When the text resumes, four pages later, the scene has moved entirely: no more dialogue about the mind, but a soul under interrogation, already past its first challenger and answering its second:

15 “And Desire said, 'I didn't see you going down, but now I see you're going up. So why are you lying, since you belong to me?'

What was on the lost first six pages?

The setting, almost certainly, and the first third of the dialogue. Ancient revelation dialogues open by staging a scene: who is speaking, where, and when. The Gospel of Mary as we have it never does this — page 7 gives us disciples already questioning a Savior already answering, with no introduction of place, occasion, or cast. That material has to have stood in pages 1 through 6.

Beyond the frame, the safest inference is more of the same kind of thing: question and answer on matter, nature, sin and the body, since that is exactly what page 7 is in the middle of. Karen King, who has written the standard English study of the text, reads the whole first movement as a post-resurrection dialogue of that sort, which fits the genre and fits the fragment. What nobody can say is whether Mary was introduced there, whether the crucifixion was mentioned, or whether the text ever explained why the disciples are so frightened when the Savior leaves them a page later. The honest sentence is short: the opening of the Gospel of Mary is unknown.

What was on the lost pages 11 through 14?

The end of the vision teaching, and the start of the soul's ascent. Page 10 stops with the Savior distinguishing soul, spirit and the mind between them; page 15 begins with the soul disputing with Desire, which the text will shortly identify as the second in a sequence — the third power it meets is named outright: “Again, it came to the third power, which is called 'Ignorance.'” So the four missing pages contained the remainder of that teaching, the transition into the ascent, and the first power's interrogation in full.

There is even a reasonable guess at the first power's name. When the soul reaches the fourth power, the text lists the seven forms it takes, and the list opens Darkness, Desire, Ignorance — recapitulating the powers already passed in order. That symmetry is why the first power is generally reconstructed as Darkness. It is a strong inference from a list we do have, not a text we possess, and we mark it as such. The forms themselves, and the question of whether they have anything to do with the seven demons of Luke 8:2, are a separate matter.

Do the Greek fragments fill the gaps?

No — and this is the fact that surprises people most. Two Greek fragments of the Gospel of Mary are known, both from Egypt, both older than the Coptic codex, and both landing squarely on material that already survives. They confirm what we have. They recover nothing we lack.

What they do prove is worth more than it sounds. The Rylands fragment was published by C. H. Roberts in 1938, the Oxyrhynchus fragment by P. J. Parsons in 1983; together they show a Greek Gospel of Mary being copied in Egypt in the third century, in at least two different books, two hundred years before the Coptic copy was made. They also show that the wording moved. In the Greek of P.Ryl. 463 the ending differs from the Coptic: Levi alone goes off to preach, where the Berlin Codex has the whole group go out to teach. Small, but it tells you that no single copy is the Gospel of Mary — only the copy we have.

Could the missing pages still turn up?

Possible; not likely by any route you can plan. The leaves were already missing when the codex was bought in Cairo in 1896, so whatever happened to them happened in antiquity or in the antiquities trade, unrecorded either way. The codex then waited fifty-nine years for publication: the printed sheets of the first edition were destroyed by water in Leipzig in 1912, two world wars followed, and the editio princeps finally appeared under Walter Till in 1955 — by which time an entirely different find had rewritten the field.

Which is the moment to kill the most common error about this book. The Gospel of Mary is not a Nag Hammadi text. No copy of it was in the sealed jar a farmer struck at the base of the Jabal al-Tarif cliff in December 1945, and it appears nowhere among those thirteen codices; it reached Berlin half a century before that discovery. The reason it is grouped with the Nag Hammadi library in bookshop editions is publication history, not archaeology: Till held his edition back for years hoping to use the newly surfaced Nag Hammadi copies of another work bound in the same Berlin codex, then gave up waiting. As for hope: Parsons published P.Oxy. 3525 in 1983, eighty-seven years after the codex surfaced. The witness list is still capable of growing.

Why we print the gaps as gaps

Because a hole is evidence and a patch is not. Where the manuscript stops, this library prints the note — Pages 1 through 6 are missing, Pages 11 through 14 are missing — and where a word is damaged, the translator's square brackets stay in the line. The alternative is the reconstruction that reads smoothly and quietly tells you something no ancient page says, which is how a gospel with ten lost pages ends up carrying a secret marriage or a suppressed doctrine in popular retellings. Nothing in this text is suppressed. It is torn, and the tear is the most honest thing about it. Our full rules for that live in the editorial standards; the nine surviving pages, about fifteen minutes of reading, are here in full — including the scene where Peter challenges Mary's right to have been taught at all, which happens to have survived intact.