On “Extant Evidence” and Inferences
In my previous postings on the subject, I’ve referred to what we can infer from “extant” manuscript evidence. One reader expressed doubts about being able to make any valid inferences, given that only a small amount of manuscript evidence remains extant. So, I’ll explain my reasoning.
First, although only a small amount of what was originally produced, the extant remains of early Christian manuscripts do amount to a body of evidence. That’s where we start. That we have only a small amount of the original body of evidence still means that we have some evidence. That shouldn’t be devalued.
Secondly, the extant manuscript evidence is random, not pre-selected for preservation. So, if anything, this actually enhances the heuristic value of the evidence. That is, the random nature of the evidence means that it wasn’t filtered to achieve a particular result.
Thirdly, all our information about early Christianity in the first three centuries shows a remarkably lively “networking” and interchange of texts, manuscripts, beliefs, practices, etc. I’ve documented this in previous publications such as this one: Larry W. Hurtado, “Interactive Diversity: A Proposed Model of Christian Origins,” Journal of Theological Studies 64 (2013): 445-62. (The pre-publication text of this essay is available on this blog site here.) This means that, although nearly all of our early Christian papyri come from sites in Egypt, they are likely to be representative of how manuscripts looked in the wider Christian network.
Finally, as with any historical matter, conclusions are susceptible to revision in the light of new evidence or corrections to our approach to what we have. So, when I write that in light of extant evidence the dominant approach to the copying of the Gospels in the first observable centuries was basically careful and stable, that is a conclusion subject to correction . . . but only by the introduction of new evidence or demonstration of evidence overlooked. Abstract references to what might or might not have been there are hardly of any use. At the very least, it is totally fair to say that the extant manuscript evidence doesn’t support the widespread assumption that in the early centuries the copying of the Gospels was “wild” and “chaotic” etc. So, we have to ask why that assumption seems so attractive, despite it having so little basis in the extant evidence.
Comments are closed.
Larry Hurtado's Blog 
As far as I can make out, especially after reading Nongbri, there are no New Testament manuscripts, except from possibly a few scraps, that can confidently be dated earlier than 200 BCE. So strictly speaking, on the basis of extant evidence alone, nothing can be said about the period of textual transmission before 200 BCE. Yet there is good internal evidence that the text was fluid in this period, including editorial additions such as John 21, and rearrangements such as 2 Corinthians. Plus there is evidence of second century editorship in the titles used and nomina sacra notation. And isn’t it valid to draw comparisons between different historical phenomena to infer that the earliest period of New Testament transmission is prima facie likely to have been a period of greater textual variation?
Donald: Your comment illustrates the dogged determination to posit some kind of major fluidity to the copying of Gospels in the earliest period, despite the evidence of our earliest manuscripts. Yes, mss that can be dated to the (late) second century are only a handful of remnants. But as I wrote, they are evidence and shouldn’t be dismissed. They are random which enhances their importance.
Moreover, they are consistent with the copying evidenced in our numerous 3rd century mss too. So, there is no basis for thinking that things suddenly changed after midnight 200 CE (NB: “BCE” = BC). Otherwise, you have to posit a phantom late second century recension of the Gospels (a la Koester).
Finally, the data that you cite has nothing to do with COPYING. They (e.g., John 21, a putative “creation” of 2 Cor) are all LITERARY productions, in the one case a new edition of John and in the other case (perhaps) a pasting together of letters to form 2 Cor. Not the same thing as copying a text. How often must I make this point before you get it??
So, in sum, the best and most economical inferences from our RELEVANT extant evidence (i.e,, copies of texts) are (1) what we see in earliest evidence is indicative of what went before, (2) and what we see is a relatively and dominant concern simply to copy the text before the copyist. No wild liberties taken. I think this should be sufficient to settle this exchange.
All great points. I would also add that I don’t think that people can fully appreciate the nature and quality of the data pertaining to early Christian manuscripts until they compare it to other fields of study in antiquity. For instance, the manuscript evidence for the works of Homer, Plato, Aristotle, Plutarch or any other Greco-Roman writer is notably inferior to that of most NT texts. Yet we don’t find scholars of such texts/writers throwing up their hands in surrender, exclaiming, “We just can’t make any inferences from this data.”
Dear Larry, thank you for pointing us back to the extant evidence and to draw our conclusion from what they support. We have a fairly new tool from the INTF data in Münster which gives an at-a-glance view of the extant evident at any point in history, showing the number of manuscripts which witness to any chapter of the NT, the oldest witness, hover-over to show the full witness list, and links to the images at the INTF if available. We hope it is useful. The history slider, along with interesting events at distinct points in time, is on the right. It slides from both the top and the bottom.
http://crosswire.org/study/rb/
Thanks for this tip! Perhaps you’d like to write a guest blog on the INTF and its latest facilities for online usage?
Prof. Hurtado, on your second point, I can see that ms survival is relatively random, overall, based on climate and so on. But weren’t some texts selected to be preserved, whether by recopying or by placing them somewhere that might protect them, such as in jars or human burials? Even Oxyrhynchus texts were selected by some people (including non-Christians) to be trashed. And the latter may help the randomness, but may not exclude some selection factors applying. And might some selection have included texts presumed to be well copied? And some selected to be destroyed or reused for other purposes? I know you are referring to Christian mss, but, for comparison, say Qumran Cave manuscripts (TaNaK plus) may have been selected by those at Qumran (and at allied camps whose people brought mss to Qumran caves) as texts they hoped would be available when, as they saw it, the Jerusalem Temple would be run according to their understanding of proper purity. (Comparing Qumran to Masada: Khirbet Qumran was usually dry but susceptible to occasional flooding of rain torrents draining from higher ground to the west {and maybe the aqueduct overflowing], whereas Masada was higher than its surroundings and had casemate walls–hence perhaps more ms survival writing the Masada lived space than in the Qumran lived space.) I’m not arguing here either for more careful or more chaotic tradents than anyone claims, but merely to ask whether the remains are absolutely random.
Stephen: Your comments about the Qumran material vividly illustrate my point! The Qumran material was placed in those jars and caves, and reflect the choices and concerns of those who did so to preserve those items. But the Christian mss from Egypt include a good many fragments that come from trash heaps, discarded, not place there to be found at some time. So, we have to assess the evidence for each site in its own terms.