The story behind “The Story Behind Who Changed the Bible and Why”
In Misquoting Jesus, the New York Times bestseller subtitled The Story Behind Who Changed the Bible and Why, author Bart Ehrman fires a shot meant to sink the ship of any Christian who thinks the New Testament documents can be trusted. Here it is:
What good is it to say that the autographs (i.e., the originals) were inspired? We don’t have the originals! We have only error-ridden copies, and the vast majority of these are centuries removed from the originals and different from them, evidently, in thousands of ways.... There are more variations among our manuscripts than there are words in the New Testament1 [Emphasis in the original].
Ehrman is right on the facts, as far as they go. There are 130,000 words in the New Testament, yet the surviving manuscripts (the handwritten copies) reveal something like 400,000 individual times the wording disagrees between them.2 Indeed, Ehrman points out, the manuscripts “differ from one another in so many places that we don’t even know how many differences there are.”3
Further, Bart Ehrman is an accomplished scholar with impeccable bona fides. He co-authored The Text of the New Testament (4th Edition)—an academic standard in the field—with Bruce Metzger, arguably the greatest New Testament manuscript scholar alive at the time.4
The Washington Post says Misquoting Jesus “casts doubt on any number of New Testament episodes that most Christians take as, well gospel.” Publishers Weekly promises that Ehrman’s arguments “ensure that readers might never read the gospels or Paul’s letters the same way again.”5
Which, of course, is exactly what Ehrman wants. Misquoting is the kind of what-they-don’t-want-you-to-know exposé that has become popular in recent years. Ehrman “exposes” discoveries that sabotaged his own “born-again” faith while a graduate student at Princeton, leaving him with the agnosticism about God he now embraces.6
Has the Bible been changed over 2,000 years of copying and recopying? Ehrman answers, “Yes, significantly.” Worse, the massive number of alterations make it virtually impossible to have any confidence of reconstructing the autographs.
Without the original renderings, there is no inspired text. Without inspired Scripture, there is no orthodox Christianity, only a jumble of spiritual ideas about Jesus expressed in a diverse body of conflicting texts that have tumbled down to us through the corridors of time.
Is this skepticism justified? Simply put, no. In spite of Ehrman’s credentials, his who-knows-what-the-original-text-said view is not the majority opinion of textual scholars. This includes Bruce Metzger, Ehrman’s mentor, to whom he dedicated the book. The reasons for this confidence are based in the nature of the reconstructive task itself.
Reconstructing Aunt Sally’s Recipe
A manuscript is a hand-copied text. For the first 1500 years after Christ, all copies of the Bible were reproduced by scribes who did the best they could—in most cases—to faithfully transmit the text. Inevitably, mistakes happened which were then compounded geometrically when the flaw was copied, spawning multiple copies with the same error in subsequent generations of texts.7 Some changes, it seems clear, were intentional and even theologically motivated.
Given that history, it’s hard to imagine how an original can be restored. The uncertainty, though, is based on two misconceptions by the rank and file about the history of the communication of ancient material like that found in the New Testament.
The first assumption is that the transmission is more or less linear—one person passing the message on to a second who gives it to a third, etc., leaving a single message many generations removed from the original. Second, the objection assumes oral transmission which is more easily distorted and misconstrued than something written.
Neither assumption applies to the text of the New Testament. First, the transmission was done in writing, and written manuscripts can be tested in a way oral communications cannot. Second, the transmission was not linear, but geometric—e.g., one letter birthed 10 copies which generated 100 and so on.
Let me illustrate how such a test can be made. It will help you see how scholars confidently reconstruct an original from conflicting manuscripts that are centuries removed from the autograph.
Pretend your Aunt Sally learns in a dream the recipe for an elixir that preserves her youth. When she awakes, she scribbles the complex directions on a sheet of paper, then runs to the kitchen to mix up her first batch of “Sally’s Secret Sauce.” In a few days, she is transformed into a picture of radiant youth.
Aunt Sally is so excited she sends detailed, handwritten instructions to her three bridge partners (Aunt Sally is still in the technological dark ages—no photocopier or email). They, in turn, make copies for ten of their own friends.
All goes well until one day Aunt Sally’s schnauzer eats the original script. In a panic she contacts her friends who have mysteriously suffered similar mishaps. The alarm goes out to the others who received copies from her card-playing trio in an attempt to recover the original wording.
Sally rounds up all the surviving handwritten copies, 26 in all. When she spreads them out on the kitchen table, she immediately notices differences. Twenty-three of the copies are virtually the same save for misspelled words and abbreviations littering the text. Of the remaining three, however, one lists ingredients in a different order, another has two phrases inverted (“mix then chop” instead of “chop then mix”), and one includes an ingredient not mentioned in any other list.
Do you think Aunt Sally can accurately reconstruct her original recipe from this evidence? Of course she can. The misspellings and abbreviations are inconsequential, as is the order of ingredients in the list (those variations all mean the same thing). The single inverted phrase stands out and can easily be repaired because one can’t mix something that hasn’t been chopped. Sally would then strike the extra ingredient reasoning it’s more plausible one person would mistakenly add an item than 25 people would accidentally omit it.
Even if the variations were more numerous and diverse, the original could still be reconstructed with a high level of confidence with enough copies and a little common sense.
This, in simplified form (very simplified, but you get the point), is how scholars do “textual criticism,” an academic enterprise used to reconstitute all documents of antiquity, not just religious texts. It is not a haphazard effort based on guesses and religious faith. It is a careful analytical process allowing an alert critic to determine the extent of possible corruption of any work and, given certain conditions, reconstruct the original with a high degree of certainty.
This last point raises the key question of this entire discussion: Regardless of the raw number of variants, can we recover the original reading with confidence? The answer to that pivotal question depends on three factors. First, how many copies exist? Second, how old are the manuscripts? Third, what is the exact nature of the differences (the variants)?
How Many and How Old?
If the number of manuscripts available for comparison are few and the time gap between the original and the oldest copy is wide, then the autograph is harder to reconstruct. However, if there are many copies and the oldest ones are closer in time to the original, the scholar can be more certain she has pinpointed the exact wording of the initial text, for all practical purposes.8
To get an idea of the significance of the New Testament manuscript evidence, note for a moment the record for non-biblical texts. These are secular writings historians rely on for all their data from antiquity that have been restored with a high level of confidence based on available textual evidence.9
Josephus’ first century document The Jewish War survives in only nine complete manuscripts dating from the 5th century—four centuries after they were written.10 Tacitus’ Annals of Imperial Rome is one of the chief historical sources for the Roman world of New Testament times, yet, surprisingly, it survives in only two manuscripts dating from the Middle Ages.11 Thucydides’ History survives in eight copies. There are ten copies of Caesar’s Gallic Wars, eight copies of Herodotus’ History, and seven copies of Plato, all dated over a millennium from the original. Homer’s Iliad has the most impressive manuscript evidence for any secular work with 647 existing copies.12
Note that for most ancient documents only a handful of manuscripts exist, some facing a time gap of 800-1500 years or more. Yet scholars are confident they have reconstructed the originals with a high degree of accuracy. In fact, virtually all of our knowledge of ancient history depends on documents like these.
The Biblical Manuscript Evidence
The manuscript evidence for the New Testament is stunning by comparison. A recent count shows 5,500 separate Greek manuscripts.13 These are represented by early fragments, uncial codices (manuscripts in capital Greek letters bound together in book form), and minuscules (small Greek letters in cursive style).
Among the 2,795 minuscule fragments dating from the 9th to the 15th centuries are 34 complete New Testaments.14 Uncial manuscripts providing virtually complete New Testaments date back to the 4th century and earlier. Codex Vaticanus is likely the oldest, dated c. 325-350.15 The magnificent Codex Sinaiticus, dated c. 34016, contains half the Old Testament and virtually all of the New. Codex Alexandrinus contains the whole Old Testament and a nearly complete New Testament and dates from the mid-5th century.17
The most fascinating evidence comes from the fragments. The Chester Beatty Papyri contains most of the New Testament and is dated mid-third century.18 The Bodmer Papyri II collection, whose discovery was announced in 1956, includes most of the first fourteen chapters of the Gospel of John and much of the last seven chapters. It dates from A.D. 200 or earlier.19
The most amazing find of all, however, is a small portion of John 18:31–33, discovered in Egypt. Known as the John Rylands Papyri and barely three inches square, it represents the earliest known copy of any part of the New Testament. The papyri is dated at A.D. 117-138 (though it may even be earlier),20 showing that the Gospel of John was circulated as far away as Egypt within 40 years of its composition.
Keep in mind that most papyri are fragmentary. Only about 50 manuscripts contain the entire New Testament. Even so, the textual evidence is exceedingly rich, especially when compared to other works of antiquity.
Two other cross-checks on the accuracy of the manuscripts remain: ancient versions (translations) and citations by early church Fathers known as “patristic quotations.”
Early in the history of the Church, the Scriptures were translated into Latin (10,000 copies exist21). By the 3rd and 4th centuries the New Testament had been translated and reproduced in Coptic and Syriac, and soon after in Armenian, and Georgian, among others. 22 These texts helped missionaries reach new cultures in their own language as the Gospel spread and the church grew. Translations help modern-day scholars answer questions about the underlying Greek manuscripts.
In addition, there are ancient extra-biblical sources—catechisms, lectionaries, and quotes from the church fathers—that cite Scripture at great length. Indeed, the patristic quotations themselves include virtually every verse in the New Testament.23
I want you to notice something here. The chief concern Bart Ehrman raises regarding the biblical texts—the massive number of variants—can only arise with a massive number of manuscripts. Scholars universally consider this a virtue, not a vice—good news, not bad—because the condition causing the problem is the very condition providing the solution. The more manuscripts available for comparison, the more changes that will likely appear, but also the more raw material to use for comparison to fix the problem the variants pose.
This mountain of manuscripts gives us every reason to believe the originals have been preserved in the aggregate. No missing parts need be replaced. We have 110% of the text, not 90%.24 The real question is this: Do we know how to separate the wheat from the chaff to recover the original reading? That depends entirely on our last question: What is the nature of the variants themselves?
Those Pesky Variants
According to manuscript expert Daniel Wallace, “A textual variant is simply any difference from a standard text (e.g., a printed text, a particular manuscript, etc.) that involves spelling, word order, omission, addition, substitution, or a total rewrite of the text.”25 Note that any difference, no matter how slight, is added to the total count.
What exactly are those differences? They can be divided into two categories: significant variants and insignificant ones. An insignificant variant has absolutely no bearing on our ability to reconstruct the original text. The meaning remains the same, regardless of which reading is the original.
For example, well over half the variants (yes, more than 200,000) are spelling errors,26 due either to accident (the ie/ei mistake is as common in Scripture as it is in our own writing), or different choices of phonetic spelling (kreinai vs. krinai). A host of others are immaterial differences in abbreviation or style (a definite article appearing before a name—“the James”—omitted in another because it adds nothing to the meaning).27
Clearly, some insignificant variations are theologically important. The rendering in the KJV of 1 John 5 (the Comma Johanneum) appearing to echo the Trinity is about a significant doctrinal issue, but clearly this variant is not in the original so it creates no textual concern. It appears in only a four manuscripts, the earliest dating from the 10th century (four others have it penciled into the margin by a scribe),28 and is almost universally acknowledged to be a corruption. Further, the doctrine of the Trinity does not rely on this text, but is verified by many other passages not in question.
A similar problem occurs with thousands of other variants that appear in only one manuscript (“singular readings”). These obvious mistakes are easily corrected.
Here’s how Wallace29 sums up the variations:
- Spelling differences or nonsense readings (e.g., a skipped line)
- Inconsequential word order (“Christ Jesus” vs. “Jesus Christ”) and synonyms
- Meaningful, though non-viable variants (e.g., the Comma Johanneum)
- Variants that are both meaningful and viable
Wallace’s last category constitutes “much less than” 1% of all variations.30 In other words, more than 396,000 of the variants have no bearing on our ability to reconstruct the original. Even with the textually viable differences that remain, the vast majority are so theologically insignificant they are “relatively boring.”31 These facts Ehrman himself freely admits:
Most of the changes found in our early Christian manuscripts have nothing to do with theology or ideology. Far and away the most changes are the result of mistakes, pure and simple—slips of the pen, accidental omissions, inadvertent additions, misspelled words, blunders of one sort of another.32
Wallace’s fourth category—those variants both meaningful and viable (in a textual sense)—is the only one of any consequence. “We are talking here,” write Kostenberger and Kruger, “about a situation where there are two (or more) possible readings, and the evidence for each reading...is relatively equal.”33
Here the analytical skills of the professional textual critic are applied to weed out the most unlikely variants. She has at her disposal a specific set of rules—the accepted canons of textual analysis—that enable her to resolve the vast majority of conflicts to recover the original with a high degree of confidence.
Ironically, this is precisely the point Ehrman unwittingly demonstrates as he closes out his case against the New Testament documents.
Ehrman’s “Top Ten”
On the final page of the paperback edition of Misquoting Jesus, Ehrman lists the “Top Ten Verses That Were Not Originally in the New Testament.” It serves as his parting salvo, but in reality proves his entire thesis false.
First, I immediately recognized six of the ten citations, and in every case my own Bible translation (NASB) makes a marginal note that these verses are not in the earliest manuscripts. No surprises here.
Second, one third of Ehrman’s “Top Ten” list actually is in the New Testament, after all. Luke 22:20, 24:12, and 24:51b are, in fact, questionable in Luke. They do appear, however, almost word for word in uncontested passages (respectively, Matthew 26:28 and Mark 14:24; John 20:3–7; Acts 1:9, 11).
Third, nothing of theological consequence is lost by striking any of the variants Ehrman lists, even the long ending in Mark (16:9–20) or the engaging but likely non-canonical account of Jesus and the woman caught in adultery (John 7:53–8:11).
Finally (and most damaging), Ehrman’s list proves just the opposite of what he intends. For all his hand wringing that the original text is lost forever, his list itself demonstrates it’s possible to recognize the most important spurious renderings and eliminate them.
Ehrman’s own works (Misquoting and also The Orthodox Corruption of Scripture) prove that the text-critical methods mentioned above—the very methods he uses to critique the New Testament—are adequate to restore the original reading. It is proof that the massive number of variants do not interfere with our ability to recapture the original, but instead the rich manuscript evidence we possess allows us to weed out the vast percentage of variants. Otherwise Ehrman would not be able to say with confidence his “Top Ten”—or any other verses—are not in the New Testament.
This is a fact he acknowledges (again, ironically) in another work. Compare the pessimism of Misquoting Jesus with the optimism expressed in Metzger and Ehrman’s The Text of the New Testament:34
Besides textual evidence derived from New Testament Greek manuscripts and from early versions, the textual critic compares numerous scriptural quotations used in commentaries, sermons, and other treatises written by early church fathers. Indeed, so extensive are these citations that if all other sources for our knowledge of the text of the New Testament were destroyed, they would be sufficient alone for the reconstruction of practically the entire New Testament [emphasis added].
Bart Ehrman has two books with his name on them that give the exact opposite impression.35 And both were published the same year (2005).
What can we conclude from the evidence? Virtually all of the 400,000 differences in the New Testament documents—spelling errors, inverted words, non-viable variants and the like—are completely inconsequential to the task of reconstructing the original. Of the remaining differences, virtually all yield to a vigorous application of the accepted canons of textual criticism.
This means that our New Testament is over 99% pure. In the entire text of 20,000 lines, only 40 lines are in doubt (about 400 words),36 and none affects any significant doctrine.
Scholar D.A. Carson sums it up this way: “What is at stake is a purity of text of such a substantial nature that nothing we believe to be doctrinally true, and nothing we are commanded to do, is in any way jeopardized by the variants.”37
Our chief question has been, “Can we reproduce the original New Testament to a high degree of certainty?” Even Bart Ehrman, in spite of himself, demonstrates we can.