(28 votes, average: 4.86 out of 5)
Loading...I have begun to explain the field of “textual criticism,” the academic discipline that tries to establish what an author actually wrote if you don’t have his original but only copies made from later times.
In this post I begin to summarize some of the most important information about the textual “witnesses” to the text of the New Testament. I won’t be going into this information at any serious length. We could have many, many, many posts on virtually every single detail that I mention. You don’t want that. Trust me.
There are three kinds of witnesses to the text of the New Testament, that is to say, three kinds of documents that can help us establish what the authors actually wrote.
- First, obviously, are
the Greek manuscripts of the New Testament. These are copies of the New Testament in the language in which the books were originally written, produced by later scribes, who were copying earlier copies that had been made by scribes who were copying earlier copies that were made by scribes…. well you get the point. We have in excess of 5600 whole or fragmentary copies of these copies. I’ll say more about them in a moment.
Unlock 4,000+ Articles Like This!
Get access to Dr. Ehrman's library of 4,000+ articles plus five new articles per week about the New Testament and early Christianity. It costs as little as $2.99/mth and every cent goes to charity!
Learn More!
- Second, we have ancient “versions” of the New Testament. These are translations of the New Testament from antiquity into languages such as Latin, Syriac, Coptic, Ethiopic, Armenian, Gothic, Old Church Slavonic, and so on. We don’t have the “originals” of any of these translations, but only later manuscripts (in the case of the Latin, possibly as many as 10,000 copies!); but where we can use the surviving copies reconstruct what Greek text these versions were based on, then we would know what the Greek manuscripts copied by the translators looked like in the time and place they were doing their translations. This, obviously requires a good bit of technical linguistic skill.
- Third, we have “Patristic evidence,” that is, the quotations of the New Testament by church fathers who were writing about a variety of things in which they cited the New Testament here and there. If you can see how a church father such as Didymus the Blind, in, say, the year 360 CE, quoted the Gospel of Matthew, then you could reconstruct (partially) what his Greek manuscript(s) of Matthew looked like at that time and place. Part of the problem is that we don’t have the originals of the church fathers’ writings either, but only later copies. This kind of evidence is tricky to use. As I mentioned in a recent post, it is what I wrote my doctoral dissertation on (specifically, on the quotations of the Gospels in the writings of Didymus). I’ll devote a couple of posts to it anon.
Everyone agrees that of all this evidence, the most important is the Greek manuscript tradition of the New Testament, so let me say a few things about it here.
New manuscripts of the New Testament – always fragmentary, as it turns out – are being discovered all the time, a lot of them by my friend and occasional debate opponent Dan Wallace and his Center for the Study of New Testament Manuscripts (see: http://www.csntm.org/home/about ). Dan is a professor of New Testament at Dallas Theological Seminary. His Center is set up to photograph and so digitize all the surviving manuscripts of the New Testament, and in the process of traveling around the world to do so, he and his colleagues have discovered previously unknown manuscripts, some of which have not yet been catalogued.
The official cataloguing agency for New Testament manuscripts is in Münster, Germany and is called the Institute for New Testament Textual Research; it was started by a famous NT textual critic named Kurt Aland. It is the institute that gives all the surviving manuscripts their official “numbers.” For over the past 60 years or so, it has done the most significant work in studying the surviving textual witnesses of the NT (especially the Greek, Syriac, and Coptic witnesses). You can get a sense of their work at their website: http://egora.uni-muenster.de/intf/index_en.shtml
We have over 5600 Greek manuscripts of the New Testament that have been discovered and catalogued. These range in size from tiny fragments with a few words from a verse or two, to enormous full-length manuscripts of the entire New Testament.
These manuscripts are typically listed under four categories (the categories can be somewhat confusing because they are not arranged according to the same consistent criteria: some of them are based on the writing material used for the manuscript; some on the style of Greek handwriting; and some on the use for which the manuscript was designed – as I’ll explain below).
- Papyri. These, as a rule (that is often broken) are our oldest surviving manuscripts. They are Greek manuscripts written on papyrus (rather than parchment/animal skin). We at present have around 150 of these.
- Majuscules (used to be called Uncials). These (some of which are older than some of the papyri) are Greek manuscripts written on parchment/animal skin using a large kind of letter, kind of like English capital letters (which are the kind of letters used for papyri as well!). These are our oldest non-papyri manuscripts At present we have nearly 340 majuscule manuscripts
- Minuscules. These are also written on parchment/animal skin, not with majuscule letters but more with a kind of cursive style of writing that was easier and quicker for scribes to produce (and harder for most people today to read). We start getting minuscule manuscripts around the 7th century CE (some majuscules were produced after that date as well). We currently have over 2900 minuscule manuscripts (actually, there may be a few more: I’m having trouble tracking down the precise number just now).
- Lectionaries. These are also written on parchment/animal skin (usually) and are also almost always written with minuscule script, but they are not manuscripts that give an entire NT book or collection of books per se; they are books that take passages from one NT book or another and combine them together for worship services, so that on one day a passage from one book would be written; another day there would be a passage from another book; and so on. We currently have around 2500 of these.
In my next post I’ll say more about these manuscripts and why they are both invaluable for scholarship and, in some ways, problematic.
It intrigues me how interesting you were able to make yesterday and today’s posts even for those of us not inclined to be academic!
Thanks! I try to write for smart people who aren’t trained in the field. And I LOVE it when people in other fields do that for me!
Is there a mnemonic device for remembering which of “Paul’s” letters are authentic?
There must be, but I’ve never tried one. You should go for an acronym! (Thought it’s hard with *two* letters to Corinth)
Not to mention: five books starting with T, of which only the first is authentic! So an acronym is probably not the right approach.
The first four (as listed in our Bibles) are authentic, as are the two starting with P (Philippians/Philemon), which accounts for six of the seven authentic letters. The one remaining happens to be 1 Thessalonians, which I think is the most challenging to fit into a mnemonic.
How about a musical mnemonic? Using the do re mi notation, if mi represents an authentic letter and do an unauthentic one (an arbitrary choice; feel free to substitute your own), and we assign one note to each of of the thirteen letters with four notes to the bar, we get:
mi mi mi mi
do mi do mi
do do do do
mi
Hum that to yourself a few times and it might sink in!
When I was a grad student teaching assistant for first-year Greek at Princeton Seminary, the professor, Cullen Story, much beloved, used his own grammar, for the poor Presbyterian students who were most often more sincere and dedicated than linguistically gifted, but who had to earn Greek to be ordained. Given the audience, he made the entire grammar based on admittedly corny mnemonic devices. “Oh my, a tie” was how students were taught the present middle indicative endings of verbs (-omai, -ei, -tai). And so on, page after page after page. It actually worked! But, well, philologists would not find it amusing….
Hard to make an acronym with no vowels.
Well, my bank used to be BT&T…. But yup, good point.
Does the Bible, especially the NT, reflect any belief in or purport to give evidence or arguments for the existence of purgatory?
I’m thinking also that a big reason for the “invention” of purgatory was rather benign. People who were not (almost) morally perfect didn’t have to be condemned to hell. It opened things up for sinners who repented.
The development of the doctrine of purgatory is a very complicated and long story. I talk about it in my book Heaven and Hell. Short story: none of the early Xns believed in it, but there begin to be hints of something like it in some second and third century texts (such as the Acts of Thecla and the Martyrdom of Perpetua)
Hi Bart,
There are about 5k Greek manuscripts, and there are 400k differences between them. Most of these 400k are insignificant, but still 400k is really a large number as the NT has about 180k words. So how Scholars count differences in these manuscripts?
For example, let us have a phrase of 5 words in 3 manuscripts:
Manuscript-1: A-B-C-D-E.
Manuscript-2: A-F-C-D-E.
Manuscript-3: A-K-C-D-E.
Do you count here 3 differences (type1counting), or do u count one difference (which is word number 2) that has 3 options (type2counting)?
Also, the above phrase has 4 identical words of 5 which gives an 80% rate of similarity. Do scholars have a criterion for similarity? If so, then what is the rate of similarity for these 5k manuscripts?
I did encounter a similar problem last year on a research in Quranic studies. It wasn’t a complex analysis, but I did need to clarify a method and I used type2counting as it would give the rate for differences and similarities in the same time. So, I know a small “bit” about this matter, and I am really curious of how Scholars do the counting for the NT manuscripts.
It actually gets far more complicated than that, but most scholars wold count your example as three variant readings. (Or one original and two variants; but that presupposes we know the original, which is precisely the qeustsoin! So there are three readings. (But a simple complication, so to speak: what if Manuscript 3 was A-K-D-C-E; that would be four variants; but what if Manuscript 2 was F-A-E-C-D. Gets *tricky*!!)
Hi Bart, first of all, love your content! This question is unrelated but has been bugging me. Repeatedly in debates (such as yours against James White), Christians will concede that the Bible has not been preserved completely but that the errors/deviations/variants/etc do not concern matters of crucial theological importance. Hence, the Bible is preserved – but in a narrower sense.
So, for example, the claim would be that Jesus’s resurrection, the Trinity, the divinity of Jesus, etc are all preserved and not undermined by the manuscripts. I haven’t ever actually heard a good rebuttal to these specific claims.
Do we have good reason to believe that these important doctrines were probabilistically unlikely to be in the original manuscripts (which we don’t have)? Are they even attested to in the earliest manuscripts we do have?
Thank you!
Right! I don’t think there’s any reason to dispute that the texts available to us can all be used to reconstruct the major doctrines that many people care about (whether or not the doctrines are right) from the NT writings. (And to construct other doctrines that are completely contrary, from the same texts). Textual changes NEVER will change someone’s mind about such things. But it seems CRAZY to me to think that therefore they aren’t important. It’s like saying it’s not important if we have the books of Mark, Hebrew, James, and 3 John since having them doesn’t affect any doctrines….
How high in quality manuscripts do you suppose the early church fathers might have had access to? If we can dig up nearly 2000 year old fragments nowadays, could someone in the 300s have had access to 1st century gospels or original epistles?
As a separate but related question, do you expect that copying of the early church fathers has produced much error? Since they would have been copied less frequently than the Bible, and in later years when copying was more likely to be done by the competent, perhaps the chances are good that the copies of the church fathers are relatively free of error.
It completely depends on when they were living, and where — and on the accidents of copying in their locale. And yes, copies of the texts of the church fathers were subject to the same vagaries of copying as Scripture (as can seen simply by comparing surviving copies)
Hello Bart,
Evangelical scholars like to say that the New Testament we have today is 99% accurate? Is this based upon all 5600+ available manuscripts or only those copied from ~400 CE onward? Also, do you believe a 99% accuracy rate is correct? If not, what is a more realistic accuracy percentage?
Thank you in advance,
Jason
I talked about that in my course SCiptural Corruption of Scripture that you can find at bartehrman.com Short answer: ask someone sometime where they get the number 99 from. Why not, say, 98.4 or 99.3 or 96 or…. What’s it even *mean*? I have no idea and I’ve worked in this field of research for 50 years (and I can explain the number 50. IN any event, The Ten commandmenst in Exodus 20 in the King James Bible has 315 words or so. In a Bible pbulished in 1631, the printer inadvertently omitted the word “not” in v. 14, so that it now says “Thou shalt commit adultery.” In that passage, the edition is 99.7% accurate. Is that comforting?
Fascinating, Dr. Ehrman. Are you aware of any AI study of those multiple variants that could lead to the reconstitution of something like Q for the stream of texts and fragments that apply? This could also apply to any other text for which we suppose an original form. I’m saying this without any consistant knowledge of how much data would be required to enable the AI methods to work well enough to be useful. Also, maybe work of that nature has already been achieved by scholars and computing doesn’t seem like a way of getting better results.
There’s certainly been a lot of computer work, but nothing specifically with AI that I’m aware of.
A question not directly related to this post: I just finished re-reading “The Origins of the Gospels” by Robyn Faith Walsh in the Fall 2022 BAR. This is a very short summary of a work she had published by Cambridge last year. She has a very different take on how the Gospels were written. As opposed to the idea I’ve often heard attributing them to one or another Christian community, she argues that they should be studied in the context of “Greco-Roman literary culture”. Are you familiar with her work? What is your take on this?
Yes, she’s making a very big splash. Her views reflect many of those of her professor at Brown, a very fine scholar named Stanley Stowers. Most scholars I know are not at all convinced by this view, but who knows? Galileo wasn’t convincing at first either. 🙂 (But there are excellent reasons for thinking that the Gospels reflect understandings of Jesus as shaped within communities. We don’t know of any loner elite Christian scholars from early Xty, though we know of lots of churches that spread stories aroound)
Two short questions:
Is there an organization that actively searches for Gospel manuscripts? I don’t mean organizations that study already found ones, such as you just introduced us to, but a “Private Detective treasure hunter” type. They would know where to look, I guess. A pin in a haystack? Maybe. But worthwhile, in my view.
I think, for example, of Bach’s Brandenburg Concerti, commissioned in 1721 by the Mangrave of Brandenburg , who died in 1734 ;the manuscript sold for about today’s $24, was found by the curator of the Prussian Royal Library in 1849 and published for the first time in 1850. Bach died in 1750, and never heard his masterpieces.
It’s a very wide gap by 18th-19th century proportions, from composition to printing and performance.
Second question:
Can we assume that at some point in time there were manuscripts in Aramaic or Hebrew, or is all the information in Aramaic/Hebrew assumed to be strictly oral?
1. Nope, not really. The reason is that all the monasteries and libraries of the world have already been thoroughly searched, so far as we know; the only way to find mss is by having them turn up in the sands…. 2. All of the NT writings were originally produced in Greek; we do have early Syriac mss (a dialect of Aramaic); translations into Syriac were made probably in the late 2nd c. One of the most famous discoveries of a Syriac manuscript was by two Victorian women (sisters; self-taught ancient linguists) visiting the monastery of St. CAtherines on Mount Sinai. Fasconating story, retold in the book Sisters of Sinai.
As recently as 2013, the Garima Gospels – two complete manuscripts of the four gospels in the monastery of Abba Garima in Ethiopia – were confirmed through radiocarbon dating to have been written in the 4th-6th centuries. They had been recognised for at least 50 years to be the earliest known witnesses to the Ethiopic version of the Gospels, but had previously been thought to be no older than the 10th or 11th centuries. As we know that the Ethiopic version was translated from Greek manuscripts from Egypt sometime in the late 4th century; this confirms the Ethiopic text as most interesting – especially as it shows early characteristics of the ‘Byzantine Text’ – such as the longer ending of Mark’s gospel.
There are a lot of churches in Ethiopia; and many of them have substantial libraries of religious manuscripts; mostly recent (printing was rare until the 20th century), but remote churches may well have older manuscripts that have never been fully catalogued, And as the Garima Gospels demonstrate, these can be very old indeed.
But Ethiopia at present is, alas, not a safe place.
Dr. Ehrman: Though you are no longer a Christian, do you still ascribe to the fact that there exists HISTORICAL RELIABILITY in the Greek manuscripts? Let’s say for The Resurrection? Dr. Gary Habermas, who I am certain you are familiar with, is recognized as one of the leading experts on this subject; to prove his assessments more concisely, he almost always uses the writings of the New Testament to PROVE the resurrection happened. He mentions you quite often as a source of that reliability based on a few of your views on Historical merit of New testament writings. Here is a link to that lecture: https://www.youtube.com/watch?v=S-uBpMMlgDM&t=138s
Any thoughts?
Ah, well, he’s a smart guy — but if that’s actually what he says (I think you might be misreading him) it’s a bit crazy. If you can establish the original wording of what the NT authors wrote (in many places you can NOT), taht has NO bearing on whether what they said is true or not. They certainly did believe Jesus was raised from the dead. Why would the fact they wrote about it make it true? Is something true when you can show that an author ACTUALLY WROTE it??? Is Mein Kampf “true” becuase you can show that the words you read today were actually the words that Hitler wrote? Or Das Kapital? Or The Prince? Or the Qur’an? Or Sayings of Chairman Mao? Knowning what an author wrote is irrelevant to the question of whether what he wrote was historically, scientifically, logically, or in any other way true.
As far as the Patristic writings go, how good is our surviving textual evidence for those? I’m assuming we don’t actually have the originals or the first copies or copies of copies of any of those books either. And does that create an issue? For example do we have variants within the manuscript tradition in these documents that cause debates as well? Josephus’ Testimonium Flavianum, though not a Patristic document, is of course a famous contested one. So I wonder if the writings of the church fathers have these kinds of issues too.
Yup, it creates a huge issue. Most of the time there aren’t many copies, often just a couple, and from many centuries after the original. This is a major area of Patristic scholarshp, producing critical editkons of the text. It was a huge problem for me in my early scholarshpo on Patristic citations of the NT, since the goal was to determine what form of the text was available to a churchfatehr in one time/place or another; but the copies of the church fathers themselves were changed in places (we know this of course whenever there are multiple copies) and one of the first things a scribe might change, accidentally or on purpose, was a Scriptural quotation that the father made that was expressed differently from how the scribe knew the text. Ouch.
Thanks! That’s interesting, but it isn’t what i meant. It’s too recent for that. I made a search, and what I meant must have been 7Q5, a tiny fragment in Greek found among the Dead Sea Scrolls. Some people believe that it is a part of Mark (6:52-53), and that it was written no later than 50 AD. If that is true, then the Gospel of Mark must be at least 20 years older than most scholars believe it is, and even older than (at least most of) Paul’s letters!
You mention 7Q5 only once in this blog, in passing in a 2018 post. You write that the claim that it is a part of Mark is almost universally rejected. But according to the Wikipedia page you refer to, some scholars disagree.
The problem is that there are only a few letters left in the fragment, only once forming a full word (“kai”, which means “and”). To me as a layman, it seems very far fetched to reconstruct full sentences from that. For example, “nnes” is suggested to be a part of the word “Gennesaret”. But surely, there must be lots of other words in Greek containing “nnes”, for example “Peloponnesos”.
Ah, sorry. That is not an issue of Patristics — I misunderstood the qeustion. There is indeed significant scholarsh on 7Q5 and none of the bona fide scholars of the Dead Sea Scrolls or NT thinks it is a fragment of Mark. You will always find exceptions, of course, because scholars love to promote their views. But even those of us who don’t have a stake in the matter — I personally would be unbelievably THRILLED if it were from Mark!! — when we read the scholarship pro and con, it’s simply no contest. The only way to see what I mean is to avoid the Internet — except to find references to the scholarship itself, and then read the scholarship. Even Wikipedia, as good as it has gotten (and remember not so long ago where it was reliably terrible) does not always represent a balance of views, in the sense that they show what the actual balance is (if 99% of NT scholars think X and 1% Y, then it’s perfectly legitimate to say “But some NT scholars have concluded….” Yes, that’s true. But who are they and what are their arguments/evdence?
When did they change the name from Uncials to Majuscules?
Well, in my training we called the Uncials. But scholars such as David Parker have argued that that term is better reserved for a certain form of *Latin* writing and urged us to change, and since I don’t care one way or the other, I’ve gone with th eflow. And it’s true that in a snese it’s better because majuscules then are contrasted with minuscules.
When did the books of the NT started to have chapters and verses? Any of the Greek texts have it?
Once someone did it, all the later copies had the same numerology, be it Latin, Greek, English, or Russian??
Here you go! https://ehrmanblog.org/when-did-the-bible-get-chapters-and-verses/