The Case as It Stands

What the charge says

Read the charge slowly, because it has three parts, and each one has to be proved. Invented — not copied from an older book, not translated from one, but made up. In the Middle Ages — the framework of the book, not just its surface, belongs to the fourteenth century or later. To serve Muslim teaching — the book was shaped to agree with the Qur’an. If any one of the three fails, the verdict as written fails with it.

That sets the bar honestly. This site argues elsewhere that the text may well descend from the book found in Barnabas’s tomb; it does not have to prove that here. On this page the only question is whether the charge, as it stands, is proved. That is a lower bar than you may have expected, and the site does not pretend it is higher.

What the evidence showed

The name is older than Islam. The Decretum Gelasianum, a Roman list of accepted and rejected books traditionally dated to 492–496, rejects the Gospel in the name of Barnabas. The List of Sixty Books, a catalogue of the seventh century, carries the name again. Prophet Muhammad was born around 570. A Gospel under this name existed, and was condemned, before there was a Qur’an for any forger to agree with.^[1] What this can carry: a name and a date. The decree describes none of the books it rejects, so it says nothing about the contents. But a name and a date are enough to leave the verdict with a coincidence it has never explained.

A Gospel manuscript lay on the apostle’s chest. In the reign of the emperor Zeno — on the traditional dating, the same few years as the decree — the tomb of Barnabas was opened on Cyprus, and a Gospel manuscript was found on his body. The book went to the palace in Constantinople. Later tradition called it Matthew. The Barnabas Tomb Gospel tests that label. A readable Greek book kept in the emperor’s library fits it far better than an untranslated Aramaic original does — so the label is where the inquiry begins, not where it ends.^[2] What this can carry: a Gospel text physically tied to Barnabas, within a few years of the title’s condemnation. It also carries a real possibility, which the site’s pages take seriously: that a tradition going back to that book lies beneath the Italian text. What it cannot yet show is a documented chain between the two, and no page here pretends otherwise.

The objections, tested one by one. If Fabricated? took the verdict’s own grounds in turn. The Dante comparison was examined there and did not survive; it plays no part in what follows. The Morisco connection — the Spanish copy that turned up among Muslims forced to convert in Spain — proves opportunity, not authorship: that copy was made from the Italian, not the other way around. The jubilee, the Church’s holy year, which the text says now cometh every hundred years, is one sentence, and it dates one sentence. The banks, gold pieces and fool’s white robe show where the translation was made, as the vocabulary of any translation does; they say nothing about where the text came from. And the Gospel readings that match Italian harmonies — books that blend the four Gospels into one running story — are the strongest of the grounds, and they turn out to point past Italy, as the next paragraphs show.^[3] What survives of the grounds is this: the copy is sixteenth-century, and the Italian is Italian. Nobody disputes either. Neither touches the first word of the charge.

The forger could not have read these books. Since 1709 the ground has given up books that no medieval Italian could have read: the Dead Sea Scrolls, Enoch, the Old Syriac Gospels, the Gospel of Peter. Witnesses Found After the Manuscript ran every agreement between Barnabas and those books through the same test. First result: nothing in the text requires a book found after 1709 — the verdict passes that question, as it should. Second result, the one nobody had asked for: wherever an earlier explanation exists, it is almost always ancient. It is an early Christian story about the apostle Peter, the Clementine Recognitions, from the fourth century; or a Christian-Jewish poem, the Sibylline Oracles, from the second or third; or a Latin Gospel harmony that was already on a shelf in 546. It is almost never a book of the supposed forger’s own century.^[4]

The scenes keep ancient company. Nine whole scenes of Barnabas match a known book closely enough to lay the two side by side, line by line — and a handful of smaller things match as well. The Company Barnabas Keeps sorted them by date. Every one of the whole scenes — the monk who sold himself for a widow, the pinnacle of the Temple, the Senate’s decree, the flight to Damascus, the magician’s transformation, Enoch and Elijah, the painless birth, a hell with seven floors built on the seven sins that Gregory the Great listed around 590 — matches a book written before 600. The medieval shelf holds the Golden Legend twice, the legend the Carmelite order told about Elijah’s mountain, one jubilee sentence, and the coins and casks: two books, one sentence, and a vocabulary. And four of the books on the ancient shelf stand on the Decretum’s own list, a few lines from the Gospel in the name of Barnabas.^[5] Which text drew on which, that page leaves open. The age of the company is not open.

The Gospel quotations lean on Luke and John, and passed through Latin. Of the sentences in Barnabas with a close Gospel parallel, Luke and John account for about seven in ten, Matthew for little more than two in ten. Whatever the tomb manuscript was, the book in our hands is not built on Matthew. And where its quotations can be checked against the Latin, they side with Jerome’s Latin Bible, the Vulgate: Job’s as the bird is born to fly is Jerome’s phrase. Six of its Gospel readings, meanwhile, have no Italian harmony behind them at all; the nearest shelf for them is a Latin harmony tradition older than any Italian one.^[6] So the harmony readings do not lead to Venice. The readiest route for them runs through Latin — a route, not a proof — to a stage no earlier than about 400 and possibly as early as the sixth century.

The name of the Prophet is a translator’s choice, and the content shows no Muslim hand. The name of Prophet Muhammad stands in the text fifteen times. In eleven of them a title — the Messenger of God, the prophet who is to come — can stand in the name’s place without breaking the sentence. The other four are sentences about a name itself: the writing Adam sees on his thumbnails and above the gate of paradise, and the line Mohammed is his blessed name. But scripture has its own way of foretelling a coming figure, and it would have put a name in those places too — either the name itself, or a word that carries its meaning. The Praised One is what Ahmad and Muhammad mean, and The Paraclete weighs whether that word once sat only a sound or two away from the Greek Periklytos. A translator who already knew whom the word meant would have written the name out.^[7] That is what this site has argued from the start: a later translator, reading the text’s descriptions of the coming messenger, sometimes wrote the name in place of the description. And the content around the name shows that the later touch went no further. A hand that knew the Qur’an well enough to rework the book would have corrected what the book gets different. The book counts nine heavens; the Qur’an counts seven, sura after sura. The homepage reads that difference as one of idiom rather than doctrine — and for that very reason it is one a forger had every reason to avoid. The book gives the title Messiah to Prophet Muhammad; the Qur’an gives al-Masih, the Messiah, to Jesus by name in four chapters. The book has a birth without pain; the Qur’an describes Mary’s labour. Two of the Qur’an’s most repeated points, left different rather than copied.^[8] That is the third part of the charge — to serve Muslim teaching — and it fails on the book’s own pages.

The quarrel with Paul is an old one. Barnabas treats Paul as the man who led the believers away from the Law. That is not a medieval idea. It is the memory kept in the Clementine Recognitions, where Paul walks on stage unnamed as that enemy, and in the Jewish texts that Paul in Jewish Texts collects. Barnabas reads in that direction with a consistency that looks like a habit of mind rather than a plan: where a desert father’s story has the Gospel, Barnabas has the book of Moses; where the Gospels have my Son, it has my servant.^[9] A forger would have had to reinvent, from the inside, a way of thinking that belonged to the second century.

The keepers of the manuscript were not Moriscos. The Italian manuscript surfaced in Amsterdam in 1709 in the hands of John Toland. In 1713 John Frederick Cramer, a counsellor to the King of Prussia, presented it to Prince Eugene of Savoy, and it entered the imperial library in Vienna with the rest of the Prince’s books. The trail runs through the top of Europe, and A Historical Hypothesis shows how such a book could pass between collectors for two centuries and leave almost no public trace.^[10] What this can carry: interest, custody, and a silence that fits its period. It does not date the text. It only removes the Moriscos from the author’s chair.

What you would have to accept

Each paragraph above can be answered on its own, and for a century that is how the verdict has answered them — one at a time, never together. Read together, they set a price. To keep the verdict, you have to accept five things at once.

That a forger passed over the books of his own century, again and again, to reach for Palladius, Tertullian, and the Recognitions — books that were already old when the Decretum was drawn up.

That he copied his Gospel quotations not from the Italian harmonies on sale in Venice but from a Latin harmony older than any of them.

That he wrote to serve Muslim teaching and let his book differ from the Qur’an on the heavens, on the Messiah’s title, and on Mary’s labour.

That his title happened to match a Gospel condemned in Rome nine hundred years before he was born.

And that he rebuilt the Jewish-Christian quarrel with Paul in the exact shape of a fourth-century story about Peter.

Each of the five can be explained away. Each needs its own story, and none of the five stories helps with the other four. Against them stands one story, the one A Historical Hypothesis sets out: a framework older than the Middle Ages, a Latin stage, an Italian translation, and translators who knew the name of the Prophet and the price of a gold coin in Italy. That story does more than fit the evidence. It predicts it — old company for the scenes, Latin under the Italian, a name written, in all likelihood, where a title or a name-word once stood, a surface that smells of the sixteenth century. A story that predicts the evidence is stronger than a story that has to explain the evidence away five separate times.

What this site concedes

A summing-up that hides its weak points is not worth a juror’s time, so here are this site’s, in the open. The copy is sixteenth-century. The Italian is Italian. The jubilee sentence is a fourteenth-century sentence, and the Carmel chapters are not older than the Carmelites. The name of the Prophet stands in the text fifteen times, and the site’s account of how it got there is a hypothesis about transmission, not a recovered history. The site cannot show a documented chain from the tomb manuscript to the manuscript in Vienna; it holds a tradition descending from the tomb book to be a real possibility — the evidence above is why — but a possibility it argues for, not a fact it has proved. And it does not claim to know which sentences of the Italian text still keep their earliest wording. Translation changes words, and the site’s own pages show how much.

It also says what would change the picture. An Italian or Latin source, dated before 1500, that tells the flight to Damascus or the seven-floored hell the way Barnabas tells them. A demonstration that the six harmony readings with no Italian witness arose on their own. A manuscript of the text older than the sixteenth century that showed the framework to be late. Any of these would move the case. In a hundred years, none has been produced.

The verdict this site asks for

The Gospel of Barnabas is a medieval copy. It is, in all likelihood, a medieval translation. It is not a medieval invention.

The grounds that were supposed to prove invention — the jubilee, the harmonies, the money, the Moriscos — prove copying and translation, and nothing beyond that. The evidence that was never weighed points the other way. A title condemned before Islam. A Gospel on the apostle’s chest. Scenes that keep company with the first six centuries. Latin under the Italian. A text that differs from the Qur’an on its most repeated points. All of it points to a framework older than the hands that carried it — old enough that a tradition going back to the book found in Barnabas’s tomb is a real candidate for its source.

A medieval forgery is not a finding. It is a label that has outlived the evidence it was built on, kept alive by repetition. What the text behaves like is something else: a document whose roots run back into the early centuries, which reached us through one or more layers of translation, each layer leaving the fingerprints of its own time — a Latin phrase here, an Italian coin there, a name written, in all likelihood, where a title or a name-word once stood.

Your decision

A juror is not asked whether the site’s account is certainly true. A juror is asked whether the charge was proved. On that question the file is clear: the copy is medieval, the company it keeps is not, and the one-line verdict that stands in the standard reference works is the one line the evidence does not support.

You have read the file. The pages behind it are open for anyone who wants to check a paragraph against its source. When you are ready, turn to the text itself and read it knowing what it is — and what it is not.