On Wednesday (10th February) I head off to New Orleans for a lecture on Thursday to the Biblical Studies Department of New Orleans Baptist Theological Seminary, and then to take part in the Greer-Heard Counterpoint Forum on Friday night and all day Saturday. The Forum web site includes information on speakers, the schedule and registration here.
The focus for this Forum is “How did Jesus Become God?” (not to my mind the most felicitous way to frame the question, but I wasn’t asked). The Forum begins Friday evening with a dialogue between Bart Ehrman and Michael Bird, followed by Q&A.
Then, on Saturday, there will be presentations by Simon Gathercole (Cambridge), Dale Martin (Yale), Jennifer Wright Knust (Boston University), and myself. Ehrman and Bird are invited to respond to each of these presentations. I’m honoured to be included among such a roster of respected scholars, and I look forward to their presentations.
One persistent commenter in response to my earlier posting about Christian preference for the codex has confidently posited things that only illustrate his ignorance of the data about ancient manuscripts. I shall, therefore, neither post his comments nor name him. Instead, I take this opportunity to correct his ignorance. There’s nothing wrong with being ignorant of some specialized subject–we’re all in that situation on this or that one. But it’s passing strange for someone so obviously inadequately informed then to make confident (even arrogant) claims based on his ignorance. That is not acceptable. But now to the corrections.
First, he incorrectly claimed that people must have used the codex much more regularly than the MSS data indicate, for otherwise how would they have made insertions of material into texts? Several errors here. For one thing, the way ancient texts were altered (by omissions, additions, other changes) wasn’t mechanically by physically adding or cutting out bits. Instead, it was in the copying process. Each time a given text was copied, there was the opportunity of making changes, either accidentally or deliberately. Texts on rolls could be changed just as easily as those on codices. The physical book-form had nothing to do with it.
Furthermore, as to codices, the earliest form seems to have been “single gathering” construction, a number of sheets laid on top of one another and then folded and stitched together. You couldn’t remove individual leaves, as each leaf was one half of a folded sheet. And on all the sheets, except the most inside one, one leaf had material from the early part of the text, and the other leaf had material from another, later part of the text. So, if you removed one sheet, you made two deletions, not one. And if you added a sheet or removed one, you would have to take the whole codex apart and then re-sew it together again. As for multiple-gathering codices, there also removing or adding leaves wasn’t an easy task. You see? One really needs to study the physical items closely before making claims.
Second, he claimed that, because the data on the Leuven Database of Ancient Books was heavily based on papyri from Egypt (where conditions more readily made for the survival of papyri), we can’t apply these data (particularly the obvious preponderance of the bookroll for literary texts all through the first three centuries AD) generally. In Rome (he claimed), things could be different, and he proposed that there the codex was more heavily used.
Well, again, ignorance is the mother of the claim. For we do have data about preferred bookforms in Rome and the West from the early centuries. For example, there is the library found in Herculaneum, which comprised a few hundred papyus bookrolls of literary texts that were carbonized in the eruption of Mt. Visuvius in 79 AD. So, wrong again. All actual data confirm that the bookroll was the preferred bookform for literary texts in this early period, East or West, Greek or Latin.
Martial’s famous Epigrams include mention of what he describes as an experiment of a local bookseller in preparing small, portable leather codices of his poetry for travellers. But it’s clear that this was a rather isolated experiment, and not indicative of any larger pattern. I’ve actually gone through the LDAB listing of all second-century non-Christian codices (there aren’t that many), and confirmed that they largely are workaday collections of recipes, astronomical tables, magical formulas, etc., with a few examples of copies of literary texts. My ill-informed commenter could do well to take the time to do such work before making further claims.
So, bottom line: The bookroll was overwhelmingly the preferred book form for literary texts all through antiquity till sometime in the 4th century AD, and continued to be used heavily even after that. E.g., per the LDAB, about 98% of second-century non-Christian copies of literary texts are bookrolls. By contrast, Christians overwhelmingly preferred the codex, with particular fervency for those literary texts that they treated as scripture.
For further reading:
On the ancient bookroll: William A. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto: University of Toronto Press, 2004).
On the ancient codex: E. G. Turner, The Typology of the Early Codex (Philadelphia: University of Pennsylvania Press, 1977; reprint, Eugene, OR: Wipf & Stock, 2011)
On ancient libraries, including Herculaneum: George W. Houston, Inside Roman Libraries: Book Collections and Their Management in Antiquity (Chapel Hill, NC: University of North Carolina Press, 2014)
On the basics: E. G. Turner, Greek Papyri: An Introduction (1968; reprint, Oxford: Clarendon Press, 1980)
And my own book: The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2005)
It is just a bit tiresome to have to make the same points over and over again about early Christians and the codex, but it seems necessary. I laid out some matters in an earlier posting here, for example. So in this posting I will simply list some key points briefly.
- Christian preference for the codex is readily demonstrable from data that can be obtained open-access from the Leuven Database of Ancient Books (here).
- This Christian preference was especially, notably strong for copies of texts used/read as scripture (copies of Old Testament writings, and Christian texts so used, over 95% codices). For other texts (theological treatises, and other kinds), early Christians were comparatively more ready to use the bookroll (about 1/3 of copies of these sorts of texts are on bookrolls).
- Modern references to the supposedly “obvious” advantages of the codex aren’t matched by any statements of Roman-era writers. If the advantages of the codex were so obvious, why were the early Christians the only ones to perceive them? It seems counter-intuitive to me. The only ancient comments about the usefulness of the codex that I know are in Martial’s Epigrams, where he mentions a local bookseller who produced small codex-copies of his poetry, which he commends for taking on your travels. We have examples of such small/miniature codices, and equivalent bookrolls too. But the early Christian codices aren’t typically such small/miniature ones. So, the Christian preference for the codex doesn’t seem to be explained as deriving from a preference for pocket-sized editions.
- And the supposedly “obvious” superiority of the codex for finding particular passages in larger literary texts seems to me shaped too much by our greater ease with the leaf-book and our lack of ease with the bookroll. Ancient Jews likewise pored over their sacred texts in minute detail and made reference to specific passages, but steadfastly preferred the bookroll.
- Likewise, speculations about the relationship of the codex and the emerging Christian canon are typically misinformed and so erroneous. The NT canon isn’t the product of preference for the codex. In about the same period, Judaism established a canon, all the while firmly preferring bookrolls for their scriptures. And the earliest fragments of codices seems to be from single-text ones, such as P52, a remnant of a copy of GJohn.
- Well into the 3rd century, Christians were experimenting with developing ways to construct codices to accommodate multiple texts, such as P46 (Chester Beatty Library & University of Michigan), a copy of Pauline epistles, or P45 (Chester Beatty Library), a copy of the four Gospels and Acts. That they were still working at how to construct codices to accommodate multiple works shows that the collecting of writings came first and the concern to copy multiple texts in one book came subsequently. In short, the codex didn’t shape the emerging canon; instead, the emerging canon drove and shaped the development of codex technology among Christians. And, by the way, Christians of the 2nd/3rd centuries seem to have been at the “leading edge” in codex technology.
- So, the big question: Why did early Christians so firmly and concertedly opt for the codex? They left us no comments on the matter, so we scholars have to devise the best guesses that we can. Personally, I side with the great papyrologist, Colin H. Roberts, in thinking that it was likely deliberate, to give early Christian copies of texts a marked form that distinguished them from the larger book-culture of the time.
I’ve been asked to give main points from my Peter Craigie Memorial Lecture given earlier this week in the University of Calgary. My lecture gives the gist of one of the chapters in my forthcoming book on distinctives of early Christianity, one of those distinctives being the prominent place of texts in early Christianity, making it a “bookish” religion.
- From a very early point the reading of texts was a typical part of corporate worship gatherings. This was unusual in the Roman-era setting for a religious group. Indeed, the only analogy was the use of texts in synagogue gatherings.
- In the production of new texts, likewise, early Christianity was remarkable and unusual. By my count, there were at least some 200+ texts that we know of composed by ca. 250 AD.
- The efforts at copying and dissemination of texts comprise a further distinguishing feature. This trans-local dissemination of texts reflected and furthered the sense of early Christian circles being connected with other circles in a larger, trans-local fellowship.
In these and related phenomena, early Christianity was unusual and “bookish.” That is, as some others have noted, it was a remarkable and distinctive “textual community.”
I’m currently visiting the University of Calgary to give the Peter Craigie Lecture: “A ‘Bookish’ Religion: Reading, Writing, and Disseminating Texts in Early Christianity.” (Tuesday, 12 January, 7.30pm., if anyone nearby reads this blog site and would like to attend.) The university online advert is here.
The lecture gives the gist of a chapter in my forthcoming book on early Christian distinctives in the Roman world (publication expected late summer 2016).
I knew Peter Craigie for a number of years and was shocked with many others at news of his death in an automobile crash in 1985. For a brief bio on him, consult the Wikipedia article here.
In a recent article, Zachary Cole (Edinburgh PhD student) traces the occurrence of numbers written fully or as alphabetic numeral abbreviations in Codex Washingtonianus (Codex W), showing that the “unique employment of numerical abbreviations in W falls into a remarkable pattern that coincides precisely” with the well-known textual shifts in the manuscript: “Evaluating Scribal Freedom and Fidelity: Number-Writing Techniques in Codex Washingtonianus (W 032),” Bulletin of the American Society of Papyrologists 52 (2015): 225-38.
Codex W is an oft-cited instance of what textual critics refer to as “block mixture.” This = shifts in the complexion of the text, with different blocks aligned with different textual traditions: Matthew 1:1-28:20 (“Byzantine”); John 1:1-5:11 (Alexandrian/Western; John 5:12-21:25 (Alexandrian); Luke 1:1-8:12 (Alexandrian); Luke 8:13-24:53 (Byzantine); Mark 1:1-5:30 (Western, although the textual shift may begin a bit earlier in Mark 5); Mark 5:31-16:20 (in earlier work often labelled “Caesarean/proto-Caesarean,” but this label is now shown dubious. Instead, this part of Codex W is aligned closely with P45, and these two seem to share a somewhat distinctive textual tradition here).*
It isn’t clear why Codex W exhibits these shifts. The original editor, H.A. Sanders, proposed that Codex W may derive from a manuscript of the Gospels that had been prepared in the aftermath of the terrible persecution under Diocletian, which included the destruction of Christian books. In this scenario, Sanders suggested, the copyist had to make do with partially extant copies that derived from varying textual traditions, the copyist shifting from one to the next to put together a complete text of the Gospels.
Whatever the situation behind this “block mixture,” Cole’s study shows that the fascinating shifts from writing out numbers fully or as alphabetic symbols (e.g., α=1; β=2, etc.) conforms precisely with the block mixture shifts. This is remarkable.
The import, Cole urges cogently, is that Codex W reflects a copyist carefully copying his/her exemplars, even at the level of how they rendered numbers. In short, Codex W reflects a high level of fidelity in the copying process. It’s one of a number of recent studies that justify re-examining some oft-repeated stereotypes about ancient Christian copyists and how they handled their texts. There is, in fact, little evidence of a “wild” or “free” attitude in which copyists readily manhandled texts to suit their own purposes. The manuscript often cited, of course, is Codex Bezae, with its many, often-unique, variants. But Bezae now seems more like a bizarre exception rather than as indicative of some supposedly wider practice.
*On the text of Mark in Codex W, see Larry W. Hurtado, Text-Critical Methodology and the Pre-Caesarean Text: Codex W in the Gospel of Mark (Grand Rapids: Eerdmans, 1981).
In his most recent book, seasoned and respected New Testament scholar, Robert H. Gundry, presents the bold thesis that the Gospel of Matthew presents the Apostle Peter as an apostate who is irredeemably damned: Peter: False Disciple and Apostate according to Saint Matthew (Eerdmans, 2015; publisher’s online description here).
Most readers of Matthew by far (to put it mildly) have judged that the text presents Peter prominently and, largely, favourably. Indeed, Matthew 16:13-20 has typically been seen as bestowing upon Peter a particularly positive significance and role. In Roman Catholic tradition, this text has been proffered as a biblical basis for the primacy of Peter and, by extension, the Pope as Peter’s successor.
Protestants, on the other hand, have tended to see the “rock” upon which Jesus says he will build his church as Peter’s confession that Jesus is “the Christ, the Son of the living God” (v. 16). In influential studies by Protestant scholars such as Oscar Cullmann or Martin Hengel, however, the text is taken as giving Peter a leadership role that was confined to his own time, giving no direct basis for a direct “succession” in the Papal office.
For readers of any persuasion, however, the thesis advanced in Gundry’s book will come as something of a shock. Gundry insists that, just like Judas Iscariot, the Peter of the Gospel of Matthew is presented in a very negative light as a total and final failure. Other scholars might hesitate to defend such a view, given that it appears that no one previously in the 1900 years of reading of Matthew has advocated it. But Gundry is undeterred, giving scant quarter to any objection, and defending his thesis at every turn in the discussion.
His approach is first to examine every reference to Peter in Matthew in chapters 1-5, and then discuss places where he alleges that Matthew deliberately omitted reference to Peter (chapter 6). Then, Gundry sets this view of Peter in the context of Matthew’s well-known emphasis on true and false discipleship (chapter 7) and on persecution as a threat to disciples (chapter 8). In Gundry’s argument, Peter in Matthew is the prime example of the “tare” that is to be uprooted from the true plants, and is the poster-boy of the disciples who fail under opposition. A blog-posting doesn’t permit the space to engage Gundry’s discussion of all the many passages he addresses. Suffice it to say that in a good many instances his discussion fails to convince.
In method, Gundry’s analysis is a application of “redaction criticism,” in this case examining what are often very small differences in wording of passages shared with Mark in particular, thereby to contend that these differences signal an implicitly unfavourable picture of Peter. Gundry is relentless in pressing for a negative treatment of Peter in practically every one of the numerous texts discussed. So, for example, Matthew’s form of the angelic command to the women at Jesus’ empty tomb (28:7) has them sent to Jesus’ “disciples,” whereas Mark (16:7) has the women sent to “his disciples and to Peter.” Gundry contends that this omission of Peter reflects Matthew’s view of him as a failed disciple who isn’t among the disciples who are restored. If you’re already inclined to Gundry’s view, you could take this “omission” this way, I suppose. But if you require more overt evidence, Gundry’s claim will likely seem dubious, or at least requiring more basis.
Similarly, Gundry reads the reference to Peter as “first” in the list of the twelve apostles in Matthew 10:2 in light of the sayings elsewhere in Matthew that “the first shall be last, and the last shall be first” (19:30; cf. also 20:16). So, Gundry contends, Matthew’s reference to Peter as “first” isn’t positive, but is simply another hint that he is to be rejected.
But in this latter text isn’t it worth noting that Matthew specifically identifies Judas Iscariot as the one who handed Jesus over (10:4), with no equivalent statement about Peter. So, if Matthew similarly wanted to make Peter an apostate, why didn’t he make it as explicit? Why all the supposedly veiled and subtle character assassination that Gundry has to explain for us in this passage and others, and that has eluded previous readers for so long, when it is clear that the author of Matthew knew how to label someone overtly as a “baddy” when he wanted to do so? One might even judge that, if Gundry is correct, the author of Matthew is one of the most spectacularly misunderstood and unsuccessful authors of all time.
Again, Matthew 16:13-23 is obviously the crucial text for any view of Peter in Matthew. Uniquely, Matthew seems to most readers to have Jesus congratulate Peter over the divine revelation given to him about Jesus (v. 17). Gundry, however, strives to to downplay this by urging that Jesus’ statement is a criticism of Peter, that he required divine revelation. But how is “makarios” (“blessed”) a rebuke? In the same passage Peter is also personally given “the keys of the kingdom of heaven” and authority to “bind” and “loose” upon the earth (v. 18, note the second-person singulars in these statements). Gundry attempts to sidestep this by erroneously claiming that the Matthean “Great Commission” (28:16-20) extends this binding/loosing authority to “all the other apostles . . . Judas Iscariot included.” Matthew 28:16, however, refers to the “eleven” as given the Great Commission, reflecting Judas’ prior betrayal and suicide (27:3-10), and so implicitly including Peter among those remaining apostles who are restored and given the Commission.
But the only potential difficulty with his thesis that Gundry seems to grant is the early tradition that Peter suffered martyrdom as a follower of Jesus in the 60s. This leads Gundry to propose that Matthew was written “prior to the mid-60s” (101). If, as most scholars hold, Matthew knew and used Mark, this would require an astonishingly early date for the latter, yet there is no reference to this matter. This could seem a classic example of allowing a hypothesis that itself requires substantiation to serve as a basis for a major reconstruction of the literary history of these writings. As I say, “bold” if nothing else!
But surely the potential problem with Matthew supposedly constructing a picture of Peter as an irredeemably damned apostate didn’t commence with his martyrdom. By all indications, Peter had acquired a widely-known stature as a leader in the young Jesus-movement much earlier than that. For example, Paul’s oft-cited statement that he spent a fortnight with “Kephas”(a.k.a. Peter) only a few years after the revelatory experience that changed him from opponent to proponent of the gospel about Jesus (Galatians 1:18) surely reflects Peter’s early and wide recognition as a prominent figure, as do the other references to Peter in this epistle.
So, how did the odd notion supposedly occur to the author of Matthew that it was credible, useful and appropriate to portray Peter (albeit implicitly) as a damned apostate? To be sure, the Gospel of Mark presents Peter as collapsing in shameful denial (14:66-72), contrasting this with Jesus’ exemplary steadfastness in the interrogation by the temple authorities (14:55-65). Matthew (26:69-75) follows Mark’s lead, echoing the scene where Peter denies Jesus in the courtyard, while Jesus is arraigned. In Mark and Matthew, Peter’s failure functions as a stark warning for the original readers who might face arraignment for their faith. Moreover, Mark (14:26-31) and Matthew as well (26:30-35) present Peter’s collapse, and the failure of the other apostles, as foretold by Jesus. But both writers (including Matthew, at least to most readers) also picture the failed apostles, except for Judas Iscariot, as restored collectively in encounters with the risen Jesus.
In short, it was apparently “safe” for these authors to portray the failure of the apostles because they were restored. And it was safe to portray in such explicit terms Peter’s failure in particular because he was well known as having been included among those to whom the risen Jesus appeared, and so among those charged with leadership in the Jesus-movement. That is, a portrayal of Peter’s failure wouldn’t have been taken as the last word about him. And to judge from Matthew’s “reception history,” that is how it was understood from as far back as we have any evidence.
Gundry urges “an unblinking exegesis of the Petrine passages in Matthew” to “overcome interpretive and ecclesiastical traditions and the attractiveness of a Peter who offers us a mirror image of our flawed but redeemable selves” (108). Who could object to “unblinking exegesis”? But, in fact, as clearly is the case at every point in Gundry’s discussion, his thesis requires him to make judgements and posit things that aren’t actually explicit. That is, he has to urge an interpretation, an inference, in every instance. And much more is required, in my view, to make plausible the inferences that he urges. For, as I’ve indicated, the favourable “ecclesiastical traditions” about Peter actually seem to have begun quite early, and spread quite quickly and widely. So, I repeat: Under what plausible circumstances would the author of Matthew have hoped to make credible a picture of Peter as a damned apostate? Gundry’s very brief register of what he calls “possibilities” (102-3) hardly suffices.
It is interesting that among the assumptions that Gundry itemizes as undergirding his study, there is no reference to the kind of readers to whom Matthew was likely directed, who they likely were and what their previous knowledge about Peter may have been. Granted, initially, they may have been (as he urges) “first-time auditors” of Matthew; but it is most unlikely that they didn’t already have a view of Peter as a prominent figure and early leader in the Jesus-movement. That is, these “first-time auditors” likely knew very well that Peter wasn’t in fact a failed apostate. So, how could Matthew have hoped for any “traction” with such a picture of him?
Also, given that Judas Iscariot explicitly serves in the Gospels as the iconic failed disciple, the unredeemed apostate, what further need was there for Peter to model this fate? It would seem a bit redundant! Moreover, is there any confirmatory evidence of some party in the early Jesus-movement that had it in for Peter, and so would have welcomed trashing him in this manner? None that I know of.
In sum, despite the vigor with which Gundry argues his case, I find that it lacks in historical plausibility and proceeds too heavily on debatable inferences. Gundry seems to me to impute into the treatment of Peter in Matthew an intent that is, to put it gently, hardly obvious. For a recent and wide-ranging collection of studies on Peter, the following: Peter in Early Christianity, eds. Helen K. Bond & Larry W. Hurtado (Eerdmans, 2015; publisher’s online description here).
 Oscar Cullmann, Peter: Disciple-Apostle-Martyr. A Historical and Theological Study, 2nd ed.,trans. F. V. Filson (London: SCM, 1962); Martin Hengel, Saint Peter: The Underestimated Apostle, trans. Thomas H. Trapp (Grand Rapids: Eerdmans, 2010).
Once again, I advertise the programme of the Vacation Term for Biblical Study offered each summer. Again this year, it is run as “Summer Biblical Study in Cambridge,” meeting this time in Westminster College. As in previous years, there are two one-week terms: 31 July – 05 August, and 7 -12 August, each running from the Sunday through the following Friday.
This is an unusual opportunity to enjoy a short and concentrated term of study in biblical subjects informed by lectures from established scholars in their fields. It’s an excellent option for ministers to get an academic “re-charging,” and also for any others who would like a week (or two) of high-quality and fully accessible lectures.
One of the unique additional features of the VTBS programme is the opportunity to commence or further your study in New Testament Greek and/or biblical Hebrew. Instruction is offered at all levels.
There are a limited number of substantial bursaries available.
You also have time during the week to take in the scenic town of Cambridge and environs!
For further information on this year’s programme, costs and how to register for the programme, go to the VTBS website here.
There is a recently-launched online and open-access journal that looks very promising, to judge from the most recent issue: Journal of the Jesus Movement in its Jewish Setting, From the First to the Seventh Century. The online link is here.
This issue (issue 2, 2015) includes an informative article by Brent Nongbri, “The Concept of Religion and the Study of the Apostle Paul,” which draws upon Nongbri’s book: Before Religion: A History of a Modern Concept (New Haven: Yale University Press, 2013), in which he shows that the modern notion of “religion” is of relatively recent vintage. In his recent article, Nongbri applies this to the study of Paul, offering criticism of what appear to be insufficiently considered applications of this concept to Paul.
Richard Ascough’s article, “Paul, Synagogues and Associations: Reframing the Question of Models for Pauline Christ Groups,” draws upon his own extensive work on “voluntary associations” in the ancient Roman period. Ascough proposes that “associations” should serve as the umbrella cateagory, under which various sub-types can be placed, one of those sub-types being Pauline congregations.
Ralph J. Korner (“Ekklesia as a Jewish Synagogue Term: Some Implications for Paul’s Socio-Religious Location”) shows that “ekklesia” was sometimes used to designate a “synagogue” of Jews, and proposes that this affects how we understand Paul’s use of the term. It wasn’t simply a term used for the assembly of a city, but also had the possibility of this more specifically Jewish connotation.
William S. Campbell (“‘A Remnant of Them Will be Saved’ (Rom 9:27): Understanding Paul’s Conception of the Faithfulness of God to Israel”) lays out an extended case for the view that Paul retained a strong belief in the ultimate salvation of his ancestral people.
Thomas Wayment & Matthew J. Grey (“Jesus Followers in Pompeii: The Christianos Graffito and the ‘Hotel of the Christians’ Reconsidered”) offer a fresh analysis of a graffito reportedly found in the mid-19th century in Pompeii, and critically assess scholarship on it and the building in which it was found. The graffito has been widely doubted as evidence of Christians in Pompeii (which means earlier than the destruction of the town in 79 AD). Wayment and Grey contend that the graffito is a valid reference to Christians, however, and note that it is entirely plausible that Christians were in Pompeii, given references in Acts to Christians in Rome and Puteoli by the 60s.
Among the other contributions to this issue is Miriam DeCock’s review-essay on Daniel Boyarin’s book, The Jewish Gospels: The Story of the Jewish Christ. DeCock lodges what I judge to be telling criticisms of the book.
The editor, Anders Runesson, and his editorial board are to be congratulated for launching and managing this new journal. I have long contended that online journals were the wave of the future, and I hope that this venture succeeds.
I note with sadness the death of Professor I. Howard Marshall, Emeritus Professor of New Testament in the University of Aberdeen, earlier this month. Professor Marshall was a very productive scholar, with many doctoral students, and a reputation for cordiality.
He was particularly encouraging to his PhD students and other younger scholars. His written output includes commentaries, monographs, and many articles.
A memorial service will be held on 16 January 2016 in Crown Terrace Methodist Church (8 Crown Terrance), Aberdeen, at 2 pm. Anyone intending to attend the service is asked to email his widow, Maureen: email@example.com.