Which New Paradigm?
In a recent blog posting, Brent Nongbri takes issue with my postings advocating a new paradigm in the study of earliest textual transmission of Christian texts such as the Gospels: here. Brent promises further postings, which I’ll look forward to with interest. But, essentially, he seems to me to advocate a version of the one that I think needs replacing.
I should also note that, contrary to what Brent seems to assume, the case laid out in the recently published monograph by Lonnie Bell, The Early Textual Transmission of John: Stability and Fluidity in Its Second and Third Century Greek Manuscripts, NTTSD, no. 54 (Leiden: Brill, 2018) doesn’t actually stand or fall on the palaeographical dating of P66 or P75. Bell’s study instead bases itself on the full number of extant portions of 2nd and 3rd century manuscripts of GJohn (which number about 17). His aim was to assess the extent and nature of textual variation in this pool of data. Colleagues will have to read the study before making judgments about it.
Nongbri refers to a recent journal article by Matthew D. C. Larsen, “Accidental Publication, Unfinished Texts and the Traditional Goals of New Testament Textual Criticism,” Journal for the Study of the New Testament 39.4 (2017): 362-87. Nongbri seems to think that it points persuasively in the direction of a very different paradigm, but I’m not so sure. The article is an advance notice on Larsen’s forthcoming monograph, and so I look forward to studying it in due course. But, as for the journal article itself, I find it more ambitious than persuasive. I’ll sketch briefly my reasons.
The first (and major) portions of the article are given over to a survey of textual practices that I should have thought were already familiar to scholars in ancient textual studies. Some Roman-era authors alleged unauthorized publication of their works by others (which may or may not be a literary topos used by some authors of the day). We also knew that some authors and groups produced revised versions of their literary works.
Indeed, in NT studies these ideas have been drawn upon, for example, in proposals about a “proto-Mark” or “proto-Luke.” Likewise, it is pretty clear that two or three editions of Paul’s epistle to the Romans circulated (Harry Y. Gamble, The Textual History of the Letter to the Romans, SD 42 [Grand Rapids: William B. Eerdmans Publishing Company, 1977]). In short, I don’t think that NT textual critics are quite as naive in these matters as Larsen seems to allege.
Larsen contends, however, that these phenomena make the traditional practice and aims of textual criticism invalid; but I fail to see why. Some of the phenomena he cites seem to me somewhat less than relevant, such as the passage in 4 Ezra (14:23-26, 37-48) where Ezra is instructed to have 94 books written, 24 of which are to be made public and the others kept secret. Unless I miss something, the text doesn’t seem to present us with an example of the sort of “accidental” publication that Larsen claims., or any reason to call NT textual criticism to a halt.
Larsen makes much of early references to the Gospel of Mark as hypomnemata and apomnemoneumata, which he urges should be taken as “disorderly or unpolished notes” (377). Larsen proposes that Matthew be seen, not as “a separate piece of literature from Mark,” but, instead, as giving “alterations of Mark” that are “fairly minor” (378). I leave it for others to judge whether this characterization of either writing fits. But, given that GMatthew is some 65% again larger than GMark, it seems to me a bit of a stretch to characterize GMatthew as a “fairly minor” alteration.
Moreover, Justin Martyr refers to all the Gospels as apomnemoneumata (1 Apology 66.3; 67.3), and in contexts that hardly were intended to represent the Gospels as simply “disorderly or unpolished notes.” So I think Larsen may make too much rest on this term.
To be sure, I agree that GMatthew is what I would call a “friendly” appropriation of the GMark, incorporating about 90% of GMark. But the extended birth narrative, the large body of sayings material (so carefully crafted into five discourse-blocks), the extended resurrection-appearance narrative, and the rather well-known Matthean emphases and vocabulary all seem to most of us to amount to a new work in its own right, not simply a revised version of GMark.
And it appears that ancient Christians took Mark and Matthew as quite distinguishable works as well. Perhaps the most obvious indication of this is that both works were preserved and copied. GMatthew didn’t eclipse GMark (although, to be sure, GMark was cited and copied far less frequently than GMatthew in the earliest period when the Gospels circulated individually). Contrast this with the disappearance of “Q” and the other sources often thought to have been used by the authors of the Gospels.
Also, in my view, another major problem in Larsen’s case (at least as put forth in the article) is precisely his blurring of the actions of composition, editing, “publication,” and copying texts. Granted, authors of texts often produced successively revised editions of their works. But it is also important to note that copyists didn’t perform revisions, but copied, albeit with varying levels of skill and varying practices (e.g., James R. Royse, , Scribal Habits in Early Greek New Testament Papyri, NTTS 36 [Leiden: Brill, 2007]). The copying process should not be confused with the editing or revising of texts. And it is the copying process that is the focus of textual criticism, not the process of composition or revision of texts.
It is, I think, noteworthy that all of our earliest extant copies of the Gospels, even the fragmentary remains, are readily recognizable as such. That is, the extant manuscript evidence from the second and third centuries doesn’t show multiple versions of Matthew or John, for example. To be sure, the evidence does show the various kinds of variants that can characterize a text that is frequently copied, under various circumstances, and by copyists of varying skill and attention to the task. But the point is that what we seem to have are copies of this or that given text, not copies of variant editions of that text. Contrast this with the evidence that the Shepherd of Hermas was transmitted in more than one edition (Malcolm Choat and Rachel Yuen-Collingridge, “The Egyptian Hermas: The Shepherd in Egypt Before Constantine,” in Early Christian Manuscripts: Examples of Applied Method and Approach, ed. Thomas J. Krause and Tobias Nicklas (Leiden/Boston: Brill, 2010), 191-211.
We think that GMark likely acquired various endings perhaps sometime in the early second century. But this likely reflects the comparison of GMark with the other Gospels, i.e., another indication of the continuing use of GMark as a text in its own right alongside the other Gospels.
So, with all due appreciation for the fluidity of textual composition and “publication” in the ancient Roman era, I think that NT textual criticism still has a validity and a future. And I remain persuaded that the new paradigm that we need in our approach to the textual history of the Gospels (and other texts) is closer to what I’ve advocated.
Comments are closed.
Larry Hurtado's Blog 
Great post. I do wonder, though, on what evidence rests the statement that “to be sure, GMark was cited and copied far less frequently than GMatthew in the earliest period when the Gospels circulated individually,” particularly the term “copied” and the expression “when the Gospels circulated individually.” Do we even have any evidence that reaches back to “when the Gospels circulated individually?”
Jonathan: My statements about “copies/copying” and the Gospels circulating “individually” draw upon mainly the evidence of our earliest copies, papyrus manuscripts (typically fragmentary). That we have remnants of a number of manuscripts/copies of Matthew and John, and our only early copy of Mark is in P45 (mid-3rd century) indicates (1) that the Gospels circulated in codices individually, particularly in the 2nd century, and (2) that Mark was not copied as often.
I was not aware that p45 was from a single-gospel codex. I would like to know, though, just what fragments are judged certainly to have been parts of single-gospel codices. Are there even any?
Jonathan: P45 isn’t a single-gospel codex. It is the four NT gospels plus Acts in one codex, which makes it remarkable. P52 was a copy of GJohn (only). P1 Matthew. P5 John. etc. Just check the listing of papyri in any Nestle-Aland edition of the Greek NT. There are a number of single-gospel mss listed.
Thanks for the detailed response, Larry. I’ve posted a follow-up on my blog and also (in a separate post) some further thoughts from Mike Holmes and Matthew Larsen.
Thanks for keeping me updated on your postings, Brent. I’ll continue our discussion in postings myself in due course.
Dr Hurtado, thank you for you post. Does Acts provide an example of different extant editions? I realise we’re talking about the Gospels but since Luke also composed a Gospel then it may have some relevance?
Benjamin: Yes, in the case of Acts we do have what can be taken as two “editions” I suppose. But (1) the amount of difference is only about 12/-14% more in the text of Codex Bezae,(2) this material is a whole bunch of small additional material here and there, which doesn’t look like a systematic “edition”, but more like an accumulation of editorial insertions made across perhaps a few centuries, for (3) Codex Bezae (our only Greek witness of this enlarged form of Acts) is 5th century, and we really don’t have earlier manuscript witnesses to these distinguishing variants.
Dr. Hurtado. I did not suggest that aLuke was a scribe, in fact that is my point. Although he did copy from Mark, he did NOT do so with the aim of copying as accurately as he could. This is exactly why I think that what happened to the NT during the later period when the majority of the changes were introduced by scribes tasked with making accurate copies of texts should not be taken to apply to the earlier period when authors were creating those same texts, over unknown time periods, and with unknown numbers of drafts / versions.
… [edited for conciseness–LWH]
Finally, because authors did copy, but scribes did not author (or should not have done when copying!) I would like to suggest that the term ‘copyist’ should not be used when referring to a scribe performing his job.
David, you simply affirm what you presuppose. That’s called circular reasoning. You presume that copyists of texts did what the author of GLuke did with Mark, but you have no basis for this and everything we know about ancient copyists tells us you’re wrong. I repeat: Unless you’ve actually studied the matter of ancient copying and ancient authorial work, you’re only positing wishful thinking.
Oh, and based on the sort of close and extended study that I refer to, I can say (with pretty much all other papyrologists) that “copyist” is a perfectly good term for those who . . . copied texts. By contrast, the author of GLuke didn’t copy GMark; he drew upon GMark to write his own new text. It is the difference between making a copy by hand of something like an encyclopedia article and incorporating quotations from it in your own essay. The two actions aren’t at all the same. I think now, however, that we’ve handled this matter sufficiently. Another restatement of your inadequately informed belief will not be helpful.
Dr. Hurtado, I’m afraid I must reply one more time, because I need to set the record straight. I am agreeing with you that the author of Luke did NOT act like like a scribe. As a result I believe that what we can deduce about scribal habits from the extant mss (which are all copies) CANNOT be used as a model for determining how what we see now in the gospels may have been created by the gospel authors in the process of them creating the text we see now, . . . . .
It is perhaps my use of the term ‘copying’ with regard to what the gospel authors did in places that misled you, . . . .
DAvid: If you agree with me that (1) copyists did one thing, which (2) isn’t what the authors of GMatthew or GLuke did (which was composing fresh texts, using GMark as one of their sources), then what in the world have you been cavelling about all these exchanges?? Let it go.
Dr. Hurtado, I will not re-hash my previous arguments, but do want to raise a couple of related issues prompted by this post. The first is that this is where I think the CBGM is really helpful, by adding rigor to the process of uncovering the earliest variant readings (at least, based on the extant evidence). The second is the distinction you make between authors and copyists, as if they were always different people with different tasks to perform. For example In my view of the world the authors of Matthew and Luke were both, and I see no reason to believe that they were the only people at that time blending existing written text with new material. Basically, I believe that the convenient division of the production of the NT texts we see today into the work of authors and copyists, each effectively working in their own ‘silo,’ in not tenable when applied to the very beginning of Christianity, and I would argue that Lk 1:1-4 is evidence of that.
David: Pointed question (but necessary, given your stout assertion of an opinion): What experience do you personally have in examining ancient manuscripts and in studying the practices of copyists? Your “belief” about this or that is your privilege. But does your “view of the world” of the authors of Matthew and Luke rest on anything substantial other than . . . your own preferred beliefs?
Nothing in Luke 1:1-4 addresses the question. The author there refers to various other writings. Not to the copying of writings. And the author doesn’t pose himself as copying, but as taking in hand to compose his own text.
Dr. Hurtado, having just read Brent Nongbri’s recent blog post on this subject it has crystallized for me (I think) the basis of our disagreement. I simply don’t agree with the sharp distinction you make between authors and copyists in the early years of Christianity. For example, the author of Luke both copies (not necessarily exactly) existing material (e.g. Mark) and adds his own, thus what you refer to as “his own text” is only partially so.
It is the existence of the synoptic gospels that to me says this is the kind of production paradigm we should be considering for at least the earliest part of the period prior to the production of the extant evidence that we do have, none of which is likely to have been written when the authors were still alive.
To your first point, if you are referring to studying variants then I’ve been doing that for years, because it’s part and parcel of understanding the thing I am most interested in, which is the synoptic problem.
David: Your response shows that you don’t really grasp the activities of ancient authors or copyists. The author of GLuke drew upon sources. That’s not being a copyist. That’s being an author. Copyists (often called “scribes”) copy. they don’t combine sources, or create new texts. They copy, sometimes well, conscientiously, skillfully, and sometimes otherwise.
And my question wasn’t about studying “variants”, but studying the actual manuscripts and the processes that produced them. I don’t wish to be mean to you. But opinion is one thing, and competence to form one another.
David,
When I was in college, a standard assignment was to colour-code the synoptics. A different colour highlighter would be used to identify material common to Luke and Matthew versus common to Luke and mark, then Mark and Matthew, Matthew singular, Luke singular, etc. We did this in Greek, of course, and every little difference – even just a difference in tense or number with the same verb – was highlighted in the appropriate colour. When we were done, we would sit and discuss all the differences we’d identified, theorizing possible explanations, looking for patterns or trends in the types of differences in each case author.
Due respect to your experience studying variants, but have you ever done an assignment like that? Because if you had – if you had looked at the resulting rainbow of a page or discussed the coherence and consistency of so many of the differences – I think you’d have trouble categorizing Luke’s use of Mark as “copying” .
Ryan, I actually have my own home-grown set of synoptic parallels on my laptop, so yes, I have done that. As I commented to Dr. Hurtado, it’s my use of the word ‘copying’ that is the problem . . . . [edited for conciseness–LWH]
So, I don’t think I anywhere have suggested that aLuke was a copyist, but he did draw on sources, copying exactly in some places, paraphrasing/re-writing in others, as well as creating original material. Whatever he did, whatever sources he used and how he used then, and however many drafts/stages Luke went through before he finished his gospel, he was not doing what scribes/copyist later did when copying it. The two processes are different, which is the point I was trying (unsuccessfully it seems) to make.