Lost in the copying

The Bible is the most translated, most copied, and most argued-over text in existence, which makes it the best case study we have in what translation and transmission actually do to words. The famous stories of biblical mistranslation are told constantly. What is less often said is that several of the famous stories are themselves mistold; the field's best lesson is how carefully it corrects its own legends.

Begin with the horns. Michelangelo's Moses sits in Rome with two horns on his head, and the standard explanation is a blunder: Jerome, translating Exodus into Latin, met the rare Hebrew verb qaran, formed from the noun for "horn" but describing Moses' face shining, and wrote cornuta, horned. The trouble with the blunder story is that Jerome knew exactly what he was doing. A Jewish Greek translation had rendered the word "horned" before him; and Jerome, in his own commentary elsewhere, glosses the verse as "glorified, or, as the Hebrew has it, horned". He chose the concrete, literal rendering with the metaphor in full view. The horns on the statue descend from a translator's deliberate literalism, filtered through centuries of readers who no longer saw the choice behind it; which is a subtler and truer story about translation than the schoolboy error usually alleged.

The most consequential single word is in Isaiah. The Hebrew of Isaiah 7:14 promises a sign: an almah, a young woman, is with child. The Greek Septuagint rendered her parthenos, which came to mean virgin; the Gospel of Matthew, quoting the Greek, found there the virgin birth foretold. Hebrew has a more specific word for virgin, and mainstream scholarship is in little doubt that Isaiah's oracle concerned a birth in the prophet's own political moment. Even here the mistranslation story wants care: parthenos in the translators' Greek could mean simply "girl", so the doctrinal weight was loaded onto the word later. The honest summary is that a theological edifice came to rest on the Greek rendering of one Hebrew noun; when the Revised Standard Version printed "young woman" in 1952, at least one American pastor burned a copy publicly.

Then there are the words that were never written but crept in. The clearest is the Comma Johanneum, the only verse in the New Testament that states the Trinity outright: "the Father, the Word, and the Holy Spirit, and these three are one." It appears in no Greek manuscript before the fourteenth century; it grew up in the Latin tradition and was absent from Erasmus's first printed Greek New Testament in 1516. A Greek manuscript containing it then surfaced with suspicious convenience, apparently written to order in Oxford around 1520, and Erasmus admitted the verse to his third edition while doubting it in print. From there it flowed into the King James Version and three centuries of memory. The famous version of this story, that Erasmus had rashly promised to insert the verse if a single Greek manuscript could be produced, is itself a legend; the historian who checked, H. J. de Jonge, found no trace of any promise. A story about an interpolation acquired an interpolation of its own.

Two of those passages are long enough to matter to anyone. The Gospel of Mark, in the two oldest complete manuscripts, ends at verse 16:8, with the women fleeing the empty tomb, afraid, telling no one; the twelve verses of resurrection appearances that follow in most Bibles are a second-century addition, differing in style, stitched on at a visible seam. And the story of the woman taken in adultery, with the most quoted line Jesus never certainly said, "let him who is without sin cast the first stone", is missing from every early manuscript of John; it wanders between locations in the ones that carry it, a beloved orphan of the tradition. Neither passage is a secret. Modern Bibles print both, bracketed and footnoted; translation committees, faced with text they judge unoriginal and readers who love it, have generally chosen honesty in the margin over surgery on the page.

Some legends soften rather than add. Everyone has heard that the "eye of the needle" was a low gate in Jerusalem a camel could squeeze through kneeling; there was no such gate, and the earliest trace of the story is a medieval gloss a thousand years after the Gospels. A Greek variant reading "rope" for "camel", one vowel apart, appears late and thinly, an early copyist's or reader's flinch at the saying's severity. The hyperbole is the point, and it has ancient company; the Talmud speaks of an elephant through a needle's eye. Both domestications are rejected by mainstream scholarship, and both keep being retold, because the original is uncomfortable, and translation by anecdote always prefers comfort.

Sometimes the puzzle is genuinely open. The sea the Israelites cross is, in Hebrew, yam suph, the Sea of Reeds; the Red Sea arrived via the Greek translators and has held the title ever since. It is regularly cited as a plain mistranslation, but elsewhere in the Hebrew Bible the same phrase unambiguously names the Gulf of Aqaba, which is part of the Red Sea system; so the term's own usage cuts both ways, and the crossing's geography remains an honest scholarly dispute rather than a settled correction.

And sometimes the change was quieter than a word: an accent, a vowel, a name. In Romans, Paul greets Junia, "outstanding among the apostles". Junia is an ordinary Roman woman's name, and the church fathers read it so; John Chrysostom praised her as a woman and an apostle in the same breath. The male name "Junias", by contrast, is attested nowhere in the ancient world. Yet from the medieval period, and in printed Greek texts from 1927 until 1998, she appeared as a man, on no manuscript authority at all; the earliest copies carry no accents to decide the matter, and editors resolved the ambiguity in the direction they expected. Her restoration in modern editions was argued sign by sign, most fully by Eldon Epp. One woman's name, masculinised for centuries by expectation alone, is as clean a demonstration as exists of how transmission bends toward the assumptions of the transmitters.

That is the pattern Bart Ehrman's popular work has pressed hardest: scribes were human, and some changes lean doctrinal, a heightened verse here, a softened reading there. His examples are real and mostly uncontested as textual facts; whether Mark's Jesus met the leper with compassion or with anger turns on one word, and the angrier reading is likely the original. But the conclusion most readers take from the popular telling outruns what its author defends. Ehrman himself, asked directly, wrote that his position "does not actually stand at odds" with his teacher Bruce Metzger's: essential Christian beliefs do not hinge on any single disputed variant, because no doctrine rests on one verse alone. The manuscripts vary; the variations are overwhelmingly minor; the major ones are catalogued, bracketed, and argued in the open. The scandal, if there is one, is not that the text changed in copying. It is how long the honest footnotes have been there, and how few of the famous stories survive contact with them.