The size of the manuscript evidence
The New Testament survives in roughly 5,800 Greek manuscripts, plus around 10,000 in Latin and thousands more in Syriac, Coptic, Armenian, and other languages. Nothing else from the ancient world comes close. Most classical works survive in a handful to a few hundred copies, and the earliest of those is often about a thousand years younger than the original composition.
The time gap is just as striking. Fragments like P52, a scrap of John's Gospel, date to within a generation or two of the originals, and substantial copies survive from the second and third centuries. For most ancient texts, no one expects anything like that.
Where "400,000 variants" comes from
The number is real, and it's often quoted to unsettle people. It stops being alarming once you know what a "variant" is: any place where any manuscript differs from any other — a misspelled word, a swapped word order, a missing article. If a single slip gets copied into two thousand later manuscripts, it counts as one variant attested by two thousand witnesses.
That means the total scales with the number of copies, not with the amount of corruption. A text preserved in only three manuscripts would have almost no variants — and almost no way to check itself. The huge count is a side effect of having far more evidence than any other ancient book.
What the differences actually are
- Spelling and sound-alike errors. Vowels that came to be pronounced alike, so a scribe taking dictation wrote the wrong letter. By one standard breakdown (Daniel Wallace and others), around 70% of all variants are spelling differences.
- Eye slips. Writing something once that should be written twice, or the eye jumping from one line-ending to an identical one further down and skipping the text between.
- Deliberate changes. Scribes sometimes harmonized one Gospel to its parallel, smoothed rough grammar, or added a clarifying phrase. These are the variants that matter, and they're the ones scholars examine hardest.
The variants that are both meaningful and genuinely undecided are well under 1% of the total.
The places where the text really is in doubt
It would be dishonest to say nothing is uncertain, so here are the biggest examples:
- The longer ending of Mark (16:9–20) and the woman caught in adultery (John 7:53–8:11). These are the two largest disputed passages, and every serious modern Bible flags them in a margin or footnote.
- 1 John 5:7 (the "Comma Johanneum"). This is the clearest case of a doctrinally motivated addition, and it is the kind of change the discipline is built to catch — which is why most modern translations leave it out.
Notice what that tells you. A hidden rewrite would not be flagged in the margin of the Bibles people buy at the store. The fact that these passages are openly discussed is evidence that the process works.
How scholars decide between readings
Textual critics weigh external evidence — how old a reading's witnesses are, how widely spread they are, which family of manuscripts they belong to — against internal evidence, meaning which reading best explains how the others arose. A useful rule of thumb: the harder reading is usually earlier, because scribes tend to smooth things out rather than make them rougher. The result is printed in the standard critical editions (NA28 and UBS5), with the alternatives noted underneath.
"But it's like the telephone game"
That image assumes one chain: person A tells B, B tells C, and the message degrades. Manuscript copying isn't one chain — it's a branching tree. Copies were made in many places, and surviving manuscripts descend from different lines. An error introduced in one line doesn't appear in the others, so comparing independent witnesses is exactly how the error gets spotted. A text that had been quietly rewritten to a single approved version would not leave thousands of independent manuscripts that still disagree in small ways while agreeing on everything that matters.
What even skeptical scholars say
Bart Ehrman, an agnostic whose book Misquoting Jesus helped popularize the "400,000" figure, writes in that same book's appendix that his argument "does not actually stand at odds with Prof. Metzger's position that the essential Christian beliefs are not affected by textual variants." His real scholarly claim is narrower: that we can't always be certain of the exact wording of a particular verse. That's a fair point, and it's worth taking seriously — but it's very different from "the Bible was changed to say something else."
The honest bottom line
The New Testament text has not been hidden or quietly edited. It's the best-attested text of the ancient world, the uncertain spots are known and labeled, and the earliest recoverable wording can be reconstructed with high confidence. Every major Christian teaching rests on multiple passages across multiple books, none of which depends on a disputed reading.
If you want to go deeper, see How We Know the New Testament Text, the method behind it in How Textual Criticism Actually Works, the broader charges answered in The New Testament on Trial, and the manuscript explorer. Have a specific verse you're worried about? Ask the research assistant — it answers from the same sourced library, with citations you can check.