Suppose you have thousands of manuscripts of a New Testament book and a particular verse is worded in one way in 98% of them but another way in just 2%?  Surely the 98% is right, right?  That was an issue I addressed many years ago on the blog, and to some of you, the answer may be surprising.  Here’s how I said it then.


Early on in my study of textual criticism I came to understand one of the major issues confronting scholars in the field – an issue that scholars have been contending with since the eighteenth century.  For the past hundred years or so the vast majority of experts have been convinced by a solution to the problem, but the solution was slow in coming, for all sorts of reasons.  But when I was first introduced to the problem I learned there were two sides that were being taken, and I wrote a paper about it (my first year in college, at Moody Bible Institute).  I continued to be interested in the problem for a long time, and it ended up being the subject of the Masters’ thesis I wrote under the direction of Bruce Metzger.

The problem is this.   We have thousands of manuscripts of the New Testament – at last count, somewhere around 5600 manuscripts in Greek alone (that includes everything from small fragments the size of a credit card with just a few letters written on them to massive volumes with the entire New Testament from beginning to end).  Over 94% of these manuscripts come to us from after the ninth Christian century – so 800 years or more after the books of the New Testament were first written and placed in circulation.  But some of them are early – with fragments going back to the second century and full manuscripts going back to the fourth.

These various manuscripts – the thousands of them – can be grouped together based on their textual similarities.  That is to say, some of the manuscripts agree a lot with each other but not with other manuscripts.  And so, to use the example from yesterday’s post, some manuscripts include the story in John 8 of the woman taken in adultery, and others lack it; some manuscripts have the last twelve verses of Mark and others do not.   Because some manuscripts are very similar to one another, we can assume they are related to one another and we can, in theory, build a kind of family tree.

And so here is the problem.  When we do that, it is the late medieval manuscripts that cohere together whereas early manuscripts tend to be different, and cohere better with each other than with the later manuscripts.

So why is that a problem?  Suppose you have a passage that is attested in, say 1000 manuscripts.  992 of those manuscripts have a verse worded in one way, and 8 have it worded in another way.  It might seem like common sense to say that the 992 are more likely to be the original and that the 8 represent an aberration.  The problem is that typically those 8 are our earliest manuscripts, the ones closest to the original.  (I’m obviously simplifying things here a bit, but not overly simplistically.)   So which manuscripts contain the original text, the one written by the actual author – the 992 majority or the 8 early?  That’s the problem.

The problem is exacerbated by significant historical, cultural, and religious factors, which in a sense can be boiled down, in the English speaking world, to one phenomenon.   The King James Bible is a translation of the text found in the majority of manuscripts, not in the early ones.  And why does that matter?  Because English-speaking biblical scholars of the seventeenth, eighteenth, and nineteenth centuries – up to the twentieth century – were raised on the King James Bible and from their childhoods assumed that it represented the Word of God.  But what if, in places, it was *wrong*?  There was a strong pull not to admit that.  And so scholars argued that in fact the early manuscripts represented aberrations.  It was the majority text, the one on which the King James was based, that represented the original.  (Similar arguments were advanced in non-English speaking realms, especially Germany, for similar reasons distinct to their own environment.)

There was a certain force to the arguments that these “majority text” scholars (as they are called today) could mount.  If there was a change of the text away from the original, how likely is it that it would infect 992 witnesses?  Isn’t it more likely that the original would survive in the majority and that a mistake would affect only a very small number?   If you had 1000 eyewitnesses to an event, and 99.2% said one thing and only .8% said something else, which would *you* believe?

That raised the obvious question, though, of how to explain that the 8 (in this hypothetical situation – which is not too far off from the reality of the case) happen to be the earliest (or, say, the five earliest and three outliers from later)?   If they’re earlier, aren’t they more likely the original?

There were numerous responses to that question that were (and in some circles still are) offered.  One is that it is almost impossible to imagine how an aberration would take over the entire tradition.  If the different manuscripts were copied from different exemplars (the exemplar is the copy that a scribe is copying), and these exemplars were copied from different exemplars, etc. – how could a mistake infect almost the entire tradition?

OK, that may be a hard question to answer (it was indeed hard – it took a couple of centuries for anyone to come up with the convincing explanation).  But you still have to explain why the earliest surviving manuscripts have a text different from the majority text.   Why does the different text show up early, and consistently early, rather than consistently later?If the different manuscripts were copied from different exemplars (the exemplar is the copy that a scribe is copying), and these exemplars were copied from different exemplars, etc. – how could a mistake infect almost the entire tradition?

The supporters of the majority text came up with a sensible solution: the earliest manuscripts are simply the earliest *surviving* manuscripts.  When these manuscripts were originally made (say, back in the fourth century), they represented the minority reading even in their own day.  The majority text of the Middle Ages was also the majority text of the fourth century.  Then why are the surviving manuscripts of the fourth century – with a different text – the only texts of the fourth century that survive?  For a very good reason (it was argued).  Since they were such aberrant manuscripts, no one read them, and so they were never worn out.  It was the majority text manuscripts that were used everywhere, and these must have disintegrated over time.

Clever argument.  But not convincing to much of anyone today.  I suppose I know every major textual scholar – either personally or through email or just through reading their work – in the world today.  And I don’t know anyone who buys the argument, except for fundamentalist Christians who are convinced that for theological reasons, God would not have allowed the original text to be lost for many, many centuries until the labors of modern scholarship could uncover it based on a study of the earliest manuscripts.  That is not a historical argument, though, but one based on a certain set of conservative religious beliefs.  And scholarship of this sort has to be historical, not based on what you personally happen to believe about God.

I realized that already at Moody Bible Institute.  In my paper I argued that the text found in the earliest manuscripts must be original.

Over $2 Million Donated to Charity!

We have two goals at Ehrman Blog. One is to increase your knowledge of the New Testament and early Christianity. The other is to raise money for charity! In fact, in 2022, we raised over $360,000 for the charities below.

Become a Member Today!

2024-05-10T11:51:50-04:00May 14th, 2024|New Testament Manuscripts|

Share Bart’s Post on These Platforms


  1. Karlpeeter May 14, 2024 at 8:44 am

    A YouTube video describes a Jesus that supposedly is seen while an airplane was flying. Apparently Jesus (or UFO) saves the passengers, his image being recorded in a photo. Is it edited? Is the story reliable?

  2. TomTerrific May 14, 2024 at 9:14 pm

    In many of your blogs you use the term “close reading” and I am seeing it used by your colleagues.

    Could you expand a bit on what you mean by the term?

    • BDEhrman May 15, 2024 at 11:01 am

      It means paying very serious and sustained attention to what a text is saying down to its very details and in relation to its context and the use of the same words and ideas in other parts of the writing and of the author. Few people actually read texts that way, but if you see a detailed explanatoin of, say, the opening scene of Hamlet by a real expert who goes into serious depth about what’s being said there, you’ll realize it aint’ what you were thinking when you breezed over it once… disabledupes{a9f75780e39ee3b75d860742d40ecf1c}disabledupes

  3. aandersnjr May 14, 2024 at 9:30 pm

    Dr. Ehrman- this idea of manuscripts (date vs majority) in determining which text is original also seems to apply to the New Testament and orthodoxy as a whole:

    The reason the New Testament is dominated by Paul’s letters is because Paul’s letters are what survived- he was the literate one, after all. So the new testament is full of Paul’s thoughts and ideas, and therefore people assume that because his ideas dominate the NT (i.e. the majority view), they must have been the original ideas. And so Paul’s teachings dominate modern orthodoxy- not based on date but based on majority… Whereas in reality, something like the book of James was likely the more common orthodoxy in the earliest Christian communities- i.e. the original theology… at least that’s what I took from Dr. Tabor’s book about Paul.

  4. bigalster May 14, 2024 at 10:11 pm

    Bart, isn’t there a further problem, that the 94% of 5600 manuscripts we do have are themselves 800 years removedfrom when they were first written? What scribal errors, mistakes, intentional or unintentional could have transpired in that time period?

  5. luigi May 15, 2024 at 4:36 am

    It would be interesting to see what AI trained on the complete corpus of manuscripts would come up with as the “original” text.

    • BDEhrman May 15, 2024 at 11:06 am

      At this stage of their development they’d come up with garbage. Maybe in some years.

  6. normative May 15, 2024 at 10:49 am

    It’s no great mystery (now) how it happens in biology: Each and every one of us is a living embodiment of mistakes (mutations) that infected the entire tradition (genome) while the earliest hominin manuscripts were left in the dust to be dug up millennia later. The interesting question is what made the scriptural mutations more adaptive—more likely to be copied. The boring answer would be that simply by chance certain mutations ended up at institutions that did a disproportionate amount of the subsequent copying—scribal genetic drift. The interesting answer would be if something about the mutant variants made them more appealing to copyists who had to make a choice between competing versions, though given the expense and labor intensiveness of producing copies, I’m curious how often such scenarios arose.

    • BDEhrman May 20, 2024 at 8:20 pm

      It certainly happened on occasion. Some accidental mistakes make a good deal of sense and so it’s hard to know in which directoin the mistake made, since either form of the text appears to work. (In the opening of Revelation we’re told that Christ “washed” us of our sins; or is it “loosed” us from our sins? The two words look almost exactly and do sound exactly alike)

  7. galah May 15, 2024 at 9:31 pm

    Dr. Ehrman, there’s a lot of study about textual changes over the years made by scribes after the original documents were written, but is there any study about changes caused by censorship in the first century, even before anything was written?

    • BDEhrman May 20, 2024 at 8:31 pm

      I’m afraid there would be no evidence, but also no reason to think there *could* be “cencorship” without a church hierarchy; in the first century churches were not under a centralized authority. (Even later there is no systematic censorship in evidence, even when the Roman Catholic Church was in full force.)

Leave A Comment