Here’s a helpful recent post of Larry Hurtado about our friend and colleague Eldon Epp. As it turns out, there were indeed two textual traditions of the Acts of the Apostles in the second century A.D. as was previously thought by my old mentor Bruce Metzger (see his Textual Commentary on the Western Text of Acts).
A “D-Cluster” in Acts? Epp’s Recent Article
Eldon Epp (my revered former teacher, and a senior figure in NT textual criticism) has a recent (and very large) article presenting the argument that there was a dual stream of textual transmission of the book of Acts in the earliest centuries: “Text-Critical Witnesses and Methodology for Isolating a Disteinctive D-Text in Acts,” Novum Testamentum 59.3 (2017): 225-96.
This is a renewed defense, and significant modification, of a view long held, but challenged in recent decades, that there were two distinguishable textual traditions or ways of handling the text of NT writings, particularly evident in the evidence for Acts. The one textual tradition/stream is that represented in such manuscripts as Codex Vaticanus and Codex Sinaiticus, with a number of other early witnesses, often referred to as the “Alexandrian text-type.” The other textual tradition has often been labelled the “Western text/text-type,” and at least as far back as the classic work of Westcott and Hort, this “Western” textual tradition has been posited as equally early.
With some other scholars today, however, Epp sets aside the term “Western” and also the term “text-type,” granting that the former term is not soundly based (the witnesses in question aren’t particularly Western geographically) and that the latter term suggests a tightness of agreement that is misleading. Instead, Epp refers to textual “clusters,” and contends that he can show in Acts witnesses both a “B” textual cluster ( i.e., a group of early witnesses that reflect the sort of text that we have in Vaticanus), and also a “D” textual cluster. His article focuses on evidence for the latter.
Epp also is careful to point out, however, that Codex Bezae Cantabrigiensis (the four Gospels and Acts in a 5th century Greek-Latin bi-lengual codex), should not be used as the basis for identifying the D-cluster. Codex Bezae is a peculiar and quite distinctive manuscript in Acts, some 14% longer overall than the Vaticanus-type text, with a number of expansionist variant-readings at various points. Also, as Epp contended in his oft-cited study, The Theological Tendency of Codex Bezae Cantabrigiensis in Acts (Cambridge: Cambridge University Press, 1966), Bezae has a number of smaller distinctive variants that seem to reflect an “anti-Judaic” attitude.
So, instead, here is the method that Epp follows for identifying a “D-cluster” in Acts: For each variation-unit found (1) identifying the Primary witnesses [already thought to represent a “D” text] available for a given reading; (2) counting the number supporting a presumptive D-Text reading; (3) counting those that do not; and (4) calculating the percentages of witnesses agreeing and not agreeing to the readings in question. Three or more Primary witnesses must be present in a variation-unit to be included. The quantitative results of his analysis: available Primary D-Text witnesses agree with one another 88% of the time on readings in 425 variation-units, while 97% of the time these readings are opposed by both א and B together.
For readers not familiar with this kind of close textual analysis my summary of the approach may seem a bit difficult to grasp. But I hope that the basis approach is sufficiently clear. The key point is that Epp shows that there is a “cluster” of witnesses who agree in supporting a variant at a few hundred places in Acts, where a cluster of other witnesses agree in supporting an alternate variant. And these two groups of witnesses line up respectively as what Epp calls “D” witnesses and “B” witnesses.
Second, the variants that distinguish the “D-cluster” from the “B-cluster” appear to include a number of “intentional” ones, that is, not variants created by accident of copying, but instead created by users of Acts. Epp’s 1966 study mentioned already was a pioneering probe into this kind of variant reading in Codex Bezae, and if he is right that we have an earlier “D-cluster” that agree on a large number of variant readings, a next step would be to analyze their contents and what they tell us about the interests and viewpoint(s) that they reflect.
Although specifics must be tested, the general direction of Epp’s study seems cogent to me. Acts did have its own transmission history. It was not physically copied as part two of the entity that scholars today refer to as “Luke-Acts”. So, Acts may well exhibit a distinctive pattern of textual variants.
As well, our early papyri of the Gospels suggest to me (and others) a spectrum of transmission practices that might be put into two categories. There are those papyri that seem to reflect an attempt at a very conscientious and careful copying of texts, with few accidental changes and little indication of intentional changes. Then, there are other papyri that exhibit a somewhat less skillful copying, and a somewhat greater number of what look like intentional changes (the latter likely made by users of the writing, these changes then included into subsequent copies made from the copies with these users’ alterations).
Here is Epp’s statement of what is at stake and what could be the potential wider gain for historical analysis of early Christianity:
“Any text, of course, would embody various characteristics and mind-sets. It is clear, I think, that the concept of dual texts in Acts has offered access and insight to dual or multiple viewpoints in early Christianity—as portrayed by the author of Acts. That is, in each of the alternate texts of Acts [reflected in the different “clusters” of witnesses], there are differing stories to be told, varying interpretations and vigorous defenses of or opposition to early leadership, to ritual and ethical practices, and to dissimilar understandings of historical and theological matters. To eliminate dual texts is to blur— seriously, I would say—our grasp of how Acts viewed the events and interactions in the ancient churches. If any of this resonates with New Testament scholars, it would be a disservice to treat everything as if all were in a single box. Dual texts of Acts is a ready-made invitation to think outside the box.” (p. 227).
Epp has always been interested in how textual variants can reflect and inform us about theological developments and diversity in early Christianity. He is surely correct that, whatever the results, these data are important and should not be left to languish in the apparatus of critical editions of the NT.