SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 7

This is my last post of the series on the SBL 2008 book review session SBL24-129 in which James Royse's recent monograph Scribal Habits in Early Greek New Testament Papyri (NTTSD 36; Brill, 2008) was reviewed. The series concludes with Royse's response to Dirk Jongkind. For earlier posts in this series, see below.




Royse responds to Jongkind

Finally, Dirk Jongkind:

Jongkind has pointed to some specific features that could bear improvement. Some of his points are matters of style and presentation, and I am probably not the best judge of what is good or bad there. But I would like to respond briefly on a couple of points.

First, let me confess that I like footnotes. In the works of others, footnotes are often among my favorite parts. And, for what it is worth, some of my favorite parts of this book are in the footnotes. During the writing, as I pursued various paths of enquiry, I often thought of the comment of Herman Melville in Billy Budd (Chapter 4): “In this matter of writing, resolve as one may to keep to the main road, some bypaths have an enticement not readily to be withstood.” Of course, I may have given in too often to such enticement, and perhaps for the sake of clarity and explanation I should have cited Melville’s comment in a footnote.

Second, as already noted, Jongkind wonders about the value of the summaries. But let me say something about Jongkind’s particular example of a scribal habit: “that a transposition can be explained best by assuming that initially the scribe forgot a word, noticed this, and inserted it somewhat belatedly at the first possible opportunity.” This does seem to me to be an important point (although I would replace “forgot” by “omitted”), and I believe that one can often see this habit at work in various manuscripts. Perhaps I should have emphasized it more. But it does not seem to me to be “hidden in the commentary” (as Jongkind says). It is mentioned in the text in the discussion of the transpositions of each of the six papyri. And at each of those places there are—dare I say?—footnotes that will lead the reader down a bypath to find in Supplementary Note 9 (pp. 755–56) that this observation was made by Colwell and even earlier by Havet, Hoskier, and others. Of course, more could have been said. Perhaps some day I will say more on this topic, but in the meantime I encourage others to investigate this tendency in these six papyri and in other manuscripts. Many other tendencies deserve more extended treatment. After all, I attempted only a “partial fulfillment” (as I say on p. 101) of the goal of a commentary on the singular readings of these papyri.

Third, I appreciate Jongkind’s comments on my use (or lack thereof) of the interesting and important essay by Junack. And I will concede that more could have been done with this issue of whether a scribe is copying by letters, by syllables, by words, or by some larger units. I appealed to the distinction as made by Colwell on occasion, but did not collect systematic data or attempt to draw the data together into some overall perspective. I would encourage others to do so. However, I would say that (a) my chief interest was “about the type of errors we find,” and that (b) trying to say anything more seems to me to run the risk of attempting to gain insight into the psychology of the scribe. And this latter task is very difficult, to say the least. I believe that we can see that a scribe tends to omit syllables, let us say, and that that “type of error” can inform our evaluation of the scribe’s readings. I would hesitate to say much more. However, at least one could attempt to determine what general patterns there are with respect to letters, syllables, and so on, in a scribe’s errors. And my own attempts there were not systematic.

Fourth, with respect to Jongkind’s discussion of isolated textual traditions, let me say that, in my opinion, the primary purpose of studying the scribal habits of manuscripts is to sharpen or revise our analysis of readings. This happens on different levels. At the first level, we can hope to find that a particular manuscript displays specific tendencies in its errors, and we can then use those tendencies in our assessment of the value of that manuscript at some particular variation unit. For example, if we find (as I believe we do) that P46 tends to omit portions of the text by a leap from the same to the same, then P46’s support for a reading that can be so explained may be, to that extent, discounted. That is, we will find it more likely that P46 created that shorter reading, and thus less likely that that shorter reading goes back to the exemplar of P46. For such an inquiry the singular readings provide, I believe, the best evidence for the scribe’s tendencies. And then at the second level, we can hope to generalize on the tendencies found in specific witnesses. That is what happens in the canons of internal criticism. If we find (as I believe we do) that scribes in general tend to omit portions of the text by a leap from the same to the same, then we may reject readings that could have arisen in such a way. Now, my general point about isolated traditions was that, for such purposes, whether we have the work of one scribe or the combined work of several scribes is irrelevant for the assessment of readings.

However, I completely agree with Jongkind’s summary comment that it is much more likely that we do not have such “complex scribes” in the New Testament tradition. In fact, though, Jongkind’s work on Codex Sinaiticus (pp. 144–47) provides one apparent example in the Septuagint: we have there a passage from 1 Chronicles, namely 9:27–19:17, inserted into 2 Esdras. Jongkind notes that this insertion is unique to Codex Sinaiticus, but that there is no sign of correction. What he then (reasonably enough, as it seems to me) infers is that the exemplar of Codex Sinaiticus (i.e., the exemplar at this point of the Septuagint) had the same insertion, and was used for both the initial transcription and the earliest stage of correcting activity. If this is so, we can see a little isolated tradition, consisting of that exemplar and Codex Sinaiticus. But, as Jongkind notes, the textual evidence for 1 Chronicles is comparatively weak, as Brooke and McLean cite only 25 manuscripts. And certainly the breadth and complexity of the manuscript tradition of the New Testament make such isolated traditions much less likely.

Fifth, I turn to Jongkind’s discussion of the shorter reading. Here of course I did attempt to integrate the results for the six papyri, and to say something about the implications those results have for the canon of preferring the shorter reading. Jongkind is correct in reminding us that Griesbach’s first canon is a much more nuanced, and much more complicated, piece of advice than the principle of simply preferring the shorter reading. Indeed, I suspect that the nuances and the complications are precisely what have caused it to be replaced in many subsequent lists of canons by simpler and more direct principles. That is, perhaps everyone will agree with Hort that “scribes were moved by a much greater variety of impulse than is usually supposed” (cited in chapter 1, footnote 35). Nevertheless, having a canon of criticism that tells us that scribes do this, and also that, and then sometimes something else except when they are doing some other thing, true as it may be, may not really provide much guidance in choosing among readings.

Ultimately, editors of the text and most critics of the text want to make choices of some kind or other. And for that purpose simple, direct principles are the most useful. For example: “Prefer the reading of Vaticanus and Sinaiticus.” “Prefer the reading that is not harmonized.” “Prefer the shorter reading.” Those are the sorts of principles that inform most modern texts. Indeed, we have on record in Metzger’s Textual Commentary the principles used to construct, or at least to justify, our current “standard” text in Nestle-Aland. And we see there nothing like Griesbach’s first canon. Moreover, it is not all that often that one sees even in the totality of discussion of a variation unit such disparate points that Griesbach combines. Of course, most variation units receive no discussion at all. But the ones that do often involve the conflict of two or three principles. The shorter reading is not supported by B and א and friends, or the reading of B and א agrees with a parallel. Under those circumstances the editors have to balance principles. But if one has to balance the sorts of things that Griesbach tries to balance in his first canon, one may give up in despair at ever reaching a decision. (Of course, such considerations do not show that Griesbach’s canon is false.)

Finally, I should note that Jongkind’s own study of Codex Sinaiticus has provided yet further evidence that early scribes tended to omit rather than to add. This adds to my conviction that the preference for the shorter reading is fundamentally mistaken. And I wonder if there is, or really ever was, any evidence at all that scribes tended to add. In any case, there is increasing evidence, from the work of Hernández on Revelation, of Head on the early less extensive papyri, and of Jongkind on Codex Sinaiticus, that omission was more common than addition, and thus that the scribal tendency underlying the preference for the shorter reading is illusory.

Of course, we have here an overall tendency. Within that tendency there may still be plenty of additions, as we see in these six papyri, which could account for Jongkind’s observation that “traditions tend to grow over the course of centuries” (p. 4). Also, the text may have been affected not only by the changes introduced by scribes but also, as Jongkind notes, by the efforts of revisers, redactors, and editors. As extreme examples, we may think of the ways that Matthew and Luke, at least on the two-document hypothesis, handled the text of Mark. Clearly, their overall tendency was to expand. But that tendency to expand did not prevent Matthew and Luke from omitting on occasion, as at Mark 1:32, where each adopts one clause of a redundancy in Mark, or Mark 4:26–29, the parable of the seed growing secretly, where each of Matthew and Luke chooses to omit the passage entirely. Now, I would not wish to define precisely the difference between copying activity and editorial activity, although, like other things, we usually know them when we see them. And thus we distinguish the scribe of P46 or P66 from Matthew or Luke or Origen. But within the transmission of the New Testament (or the Septuagint or the Masoretic Text) there are many factors at work, and surely we cannot expect to explain the complications that we find by appeal to anything other than complex, sometimes conflicting, tendencies. However, despite such qualifications, I believe the evidence strongly supports the view that early scribes of the New Testament tended to shorten the text. And that is, if not the entire story, at least an important part of the story.

Again, I thank all of the panelists for their insightful and stimulating remarks.

Earlier posts in this series:

Part 1: Juan Hernández' presentation

Part 2: Royse responds to Hernández

Parts 3-4: Haines-Eitzen's review and Royse's response

Part 5: Peter Head's review and Royse's response

Part 6: Dirk Jongkind's review

SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 6

Dirk Jongkind
SBL24-129 New Testament Textual Criticism



In the series of summaries from SBL in Boston, we have arrived at the last reviewer who is our co-blogger, Dirk Jongkind.

The dutch people are known for their cheese, tulips and chocolates, but I have also read somewhere that they are known for being direct, open and honest. This is of course a cliché, but, in this case, the description (save the cheese, tulips and chocolates) did fit well with Jongkind’s presentation.

J. started with an apology for maybe being too harsh (I didn't think he was; and the friendly tone characterized all of the four presentations as well as responses). Then he proeceeded, first, with the many good things to be said of this book. After briefly mentioning a number of positive aspects, from which he has benefitted a lot he proceeded with the criticism. “Royse’s book,” he said, “is easy to criticize not because it has errors, but because, given its sheer volume it must contain some inconsistencies.” Examples of negative features that J. mentioned include the rather prolonged genesis of the book; too many footnotes commenting on too much, using too heavy prose. Too extended discussions and too many references to modern authors. After all these footnotes and the extensive treatment, the conclusions of a chapter amount to a single page. At the same time, many good observations are hidden in the footnotes. Royse could have provided better access to these observations.

Then J. turned to an example on p. 51ff where Royse defends Colwell’s approach to study singular readings. J asked a critical question whether we study the scribe or the manuscript? What does Royse say of those who study a group of readings that include non-singulars too, as Barbara Aland does? [Royse reviews her method on p. 60ff]. J. said that if studying the scribe, then it is warranted to study the singular readings, but if we study the tendency of the manuscript, there is no need to exclude non-singular (one studies the accumulated errors). Even when studying non-singular readings, one can often discern that a reading derives from a scribe’s mistake (regardless whether it is an earlier scribe or not).

Junack’s Abschreibspraktiken also appeared in 1981, and is an extremely important essay. There are nine references to this study in Royse. In none of the nine places is there a reference to Junack’s point where he attacks some of Colwell’s conclusions about how the scribes of P45, P66 and P75 had copied. Junack completely disagrees with Colwell and J. thinks rightly so, whereas Royse repeats Colwell’s conclusions. Scribal habits, J. said, is also about how the scribe copies and Royse fails here.

[Colwell had concluded that “In general, P75 copies letters one by one; P66 copies syllables, usually two letters in length. P45 copies phrases and clauses” (Colwell, “Scribal Habits,” 380). Junack pointed out that this position, that the scribes of P66 and P75 were copying text without attention to the mening, is untenable. Junack thought it more plausible to ascribe the particular features and errors of these scribes, not to the way they copied the letters, but to the action of the hands; cf. D. Jongkind, “Singular Readings in Sinaiticus” in Textual Variation: Theological and Social Tendencies? (eds. H.A.G. Houghton & D.C. Parker; Gorgias Press, 2008), 48-49.]

The next point that J. brought up concerned Griesbach’s canon of preferring the shorter reading. Royse does not gives attention to Griesbach’s several exceptions to the rule. Most of the omissions fall under Griesbach’s exceptions. Griesbach was talking about the substantial readings.

J. pointed out that in a correction phase much more is added than removed, so in Sinaiticus.

[TW: One must not forget the “additional reading”-aspect; perhaps some corrections in Sinaiticus, for example, are to be seen not as alternative readings but additional readings...]

James Royse´s reponse will follow...

SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 4-5

SBL24-129 New Testament Textual Criticism

The next reviewer was no other than Peter Head

First he explained why he had accepted the honour of an invitation to review the book:

(1) "I had interacted with James Royse’s ThD in my first academic publication (Head, 1990) as well as subsequently (Head, 2004, 2008) – and these are actually discussed on pp. 720ff" [I missed the first reason (Pete might fill in) - Pete just did]
(2) "I met Dr. Royse at SBL meetings;"
(3) "I saw the price of the book and saw the chance for a free copy."

Then Head excused for his own delay sending his review to Royse. But he noted to his comfort Royse's delay of 27 years between the dissertation and the monograph (see also the apology in the preface).

Head had done a "redactional study" of the two works. He concluded that there is three times as much material in the monograph. Where, then, have things stayed the same? – Most significantly, the summary conclusions are practically identical. The general conclusions stand. Especially the challenge to the traditional canon to prefer the shorter reading.

What has been expanded? Head notes that the papyri are treated a bit more like artefacts (not just like reservoir of readings), relating to Haines-Eitzen’s urge.

Royse has written a small monograph of each one of the papyri! Head focused the rest of his discussion on the study on P66. There are 129 singular readings indicated in the 1981 work. In the book 128 singular readings are listed. Hence, we lost one singular reading, which is found on p. 487 (John 12:12; the evidence of Codex Koridethi has been revised). On p. 408 Royse discusses John 18:47 which could potentially provide an additional singular, but after discussing the ink etc Royse decides to follow the editio princeps.

Colwell concluded that the scribe was careless (wildness in copying). By contrast Royse thinks the scribe was rather careful to render a literal version of his Vorlage (the scribe was responsible for the changes!). Head is not convinced that we must choose either to follow Colwell who studied P66* or Royse who based his characterization on P66c. It is a matter of who corrected, and against what Vorlage. [PMH: both phases are very important as illustrating scribal habits/behaviour.] Royse thinks the scribe has a tendency to omit.

Royse is rather critical of the treatment of corrections in the International Greek New Testament Project volume on the papyri of John. In no case does the IGNTP make clear what happened! The editors only offer their opinion of what was the original and what is the correction. Royse thinks it is certain that the scribe of P66 was Christian (harmonization to other parts of NT; curious use of nomina sacra and staurogram - this could be extended to other papyri as well; e.g., P45, P46, P47, P72 and P75 with harmonizations to other books).

[Head also noted an important general issue: "Royse is attempting to revise the traditional canons – developed on the basis of medieval manuscripts and generalizations about scribal habits from them – on the basis of singular readings in the early papyri; which by definition made no impact on the wider scribal and textual tradition – on this question I think we still need to do some more thinking. Scribal habits determined on the basis of singular readings do I think (cf. Jongkind) reveal something about scribal behaviour, but may be no so clearly about the general tendency of the textual tradition."]

Finally, Head appreciates the transparency, which is easy to follow, i.e., what exactly did the scribe do? The book has been produced on a PC. Head congratulated the author, the series editors and the publishers on such a splendid piece of book production. [TW: should we interpret this ironically; is Pete a Mac freak like me? PMH: not at all. I was commenting on the interesting fact that this beautifully produced book, produced using Nota Bene on a PC, was a revision of a dissertation written on an IBM Selectric typewriter]. Head made it his mission to find a typo, but could not detect one until 401 note 14 “P46” should have been “P66.” [I also noted "one moment of unclarity (page XVII para 2)"]

Royse responds

Next, Peter M. Head:
All of the speakers have remarked upon the history of this work, its first appearance as a dissertation, and then its second appearance in the present form. But Head has given most attention to the continuities and changes, and seems better able than I am to describe what has happened. Certainly, as I revised and expanded the dissertation I was well aware that there were some dangers in the expansion, and that perceptive readers (such as our panelists), trained to detect layers of textual accretion, would be able to see that the material from 1981 did not always flow smoothly into the material from 2006, as both Jongkind and Head on occasion note. I am honored that Head has given such careful attention to the development of this work.

Also, his decision to look at the work on P66 in more detail reflects my own understanding of the importance of that manuscript. Particularly the study of the some 465 corrections seems to me to shed much light on the nature of copying a New Testament book around the year 200. Although I devoted a great deal of time to an analysis of those corrections, crucially aided by the studies of Gordon Fee and Errol Rhodes and many other scholars, I suspect that there is yet much to be discovered. What is especially interesting is that in P66, as also in P46 and again some 150 years later in Codex Sinaiticus, the extensive corrections preserve in one manuscript several layers of textual change that can, at least in theory, reveal much about what was happening to the text during the early period. Of course, the challenge is to organize what lies in such layers in a perspicuous manner.

I mention a few points very briefly. In his note 5 Head wonders about my decision to confine attention to singulars in A occurring in the first two chapters in Revelation; I chose the first two chapters for examination simply in order not to digress too much, and my point there was a very limited one. Naturally, a wider and more thorough investigation, as we now have in Hernández’s work, Scribal Habits and Theological Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus, and Ephraemi, is much to be welcomed.

With respect to Head’s finding “one moment of unclarity” at p. xvii, second paragraph, line 2: let me note that one should delete the phrase “and asterisked”— I honestly don’t know what happened there, but it is an unwarranted addition. And at p. 401 n. 14, line 2 from the end: yes, “P46” should be “P66.”

In addition to this, Royse actually ended his response to Peter with this off-the-record statement: "I did find another typo, but I will let you find it for yourself."

I can now proudly announce that I have found it on p. 453, n. 105, “a construction that indeeds [sic] makes perfectly good sense.” I can't believe Peter missed this one!

After this came the final review by a certain Dutch scholar. Don't forget to tune in!

Earlier posts in this series:

Part 1: Juan Hernández' presentation

Part 2: Royse responds to Hernández

Parts 3-4: Haines-Eitzen's review and Royse's response

SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 3-4

The next reviewer was Kim Haines-Eitzen (KHE).

In 1991 first year in graduate school, KHE discovered Royse’s work. She was not alone in her fascination for Greek New Testament manuscripts! If you want to pay attention to Hort’s dictum that knowledge of documents should precede judgment on readings, you must study Royse’s work which focuses on the starting point of textual criticism. To many of us Royse’s book is widely cited. His list of singular readings indicates those that are supposed to have been introduced by the scribe we study.

Then KHE turned to some specific papyri. After examining P45 on over 90 pages, Royse arrives at the “basic features of the copying by the scribe” (p. 197) which are as follows (here abbreviated): The scribe is concerned to produce a readable text, and is successful in reading this goal with virtually no correction. Thus, there are few nonsense readings, few corrections, and few obvious errrors. Secondly, the scribe has a marked tendencey to omit portions of the text. Thirdly, harmonization is a frequent cause of error, with harmonization to the immediate context being the most important type. Fourthly, stylistic and grammatical improvements are sometimes attempted. Finally, the scribe is rather rarely subject to certain errors of sight, perhaps also of hearing, of influences of similar forms, and simply of oversight. KHE then continued with P46. However, she questioned whether these brief summaries “help us get a sense of a scribe.”

In any case, P72 is most interesting to KHE. This scribe stands out as is evident in the described main features of the scribe (614): “Alone among our papyri P72 gives evidence of some theological purpose in the scribe’s creation of singulars: two singular readings and one asterisked reading appear to reflect the scribe’s belief that Christ was fully God.” KHE adds that the handwriting of this scribe is easily worst. She adds that this is perhaps why she is most sympathetic to this scribe (laughter). Royse thinks that it is tempting to conclude that theological changes came about not until 300 C.E.

[Additional comments by TW: In his monograph, Royse refers to Barbara Aland who relates the changes in P72 to the fact that it is very late among the early papyri and that it is part of a manuscript containing a Christian’s personal collection of writings. He also refers to other studies, including my own, that have shown interesting connections with the non-New Testament writings contained in the whole Bodmer Codex of which P72 is a part.

In any case, he says, “the contrast between these three readings and what we find elsewhere in our six papyri is striking and indicates how unusual such alteration must have been in the early period.” Royse further refers to Min who found no theologically intentional readings in the early Matthean papyri. Finally, he points to Hernández who examined Sinaiticus, Alexandrinus and Ephraemi Rescriptus in Revelation. He reports a few readings in the two former that are of Christological import, but no such readings in the latter. Interestingly, Hernandez himself told us in his presentation (now available for download in the sidebar) that his original plan was to write about this subject, before he found Royse’s dissertation!]

Anyway, KHE disagrees with Royse on this particular point, i.e., she thinks theological changes were introduced earlier on, but, as one of her concluding question (see below) implied, she thinks more research is necessary to bring clarity on this point. She did not develop this further but instead concluded by posing the following questions, as a kind of “where do we go from here”:

1) To what extent does the focus on singular readings push towards the text from the form, ie. the physical features. Could we not include the physical features (codicology, handwriting, etc) to situate the scribes.

2) Royse hints that the theological variants come in 300CE. but most of the textual variants in the tradition were already introduced. Where are we with theological variations?

3) Text-types: Royse refers to text-types, e.g., P72 is Alexandrian. KHE is disappointed with this terminology. How does it help us to understand the early papyri? [Here I may recommend reading the review of D.C. Parker’s book by Epp which focused on this issue].

Final conclusion: “We have still a lot to do.” See also Royse’s own suggestions for future research (737ff).

Royse responds:

Next, Kim Haines-Eitzen:

Both Haines-Eitzen and Jongkind draw attention to the brief summaries of scribal habits for each of the six papyri. Haines-Eitzen questions whether the summaries “help us get a sense of a scribe,” and Jongkind remarks that a one-page summary of scores of pages of analysis is not all that helpful. Of course, to provide extended “integrated reflection” (as Jongkind puts it) on the scribal habits would have been to risk expanding the book even more. But perhaps in such extended discussion I could have avoided some of the tensions (shall we say) in the summaries that Haines-Eitzen points out. Alternatively, perhaps it would have better simply to forego such summaries and let the analyses speak for themselves. At least that would have avoided the shortcomings, and omission is always so much more tempting than addition.

I turn to Haines-Eitzen’s concluding three points (or “questions”).

Point 1: I would concur that the physical features of manuscripts can be crucial. Haines-Eitzen, of course, has given much attention (in her provocative study, Guardians of Letters) to the peculiarities of P72 and of the codex of which it forms a part (or perhaps two parts, 1–2 Peter and Jude), giving careful attention to the curious features(scribal and otherwise) of the disparate texts joined there into this “third-century miscellany” (as she calls it), and arguing that it was the product of an early scribal network.

And Jongkind, in his recent work, The Scribal Habits of Codex Sinaiticus, creatively combines many aspects of that very complex manuscript, such as the arrangement of the quires, the nature of the paragraphing, the use of nomina sacra, and the scribal tendencies to produce certain sorts of variations, all sorted out among the three scribes and the various correctors. Here again, I would not wish to be seen as in any way claiming completeness or finality in my analysis, and I would welcome further and more comprehensive discussions.

Especially the codex to which P72 belongs seems, as Haines-Eitzen well observes, to be a different sort of physical object than the usual New Testament manuscript, and it would be natural that its unique properties should have implications for our understanding of its text and for our evaluation of its readings for purposes of textual criticism. I might refer further to the fascinating analysis of the “Bodmer Miscellaneous Codex” by Tommy Wasserman in chapter two of his recent The Epistle of Jude: Its Text and Transmission, which integrates codicological and textual considerations of this compilation.

Point 2: I would not wish to appear at all certain on such a complex topic as the theological corruptions of the text. But I would observe that it is (I believe)perfectly consistent to hold that dogmatic changes began to occur around 300 and that the majority of textual variants arose during the first three Christian centuries (that is, before 300). We have to keep in mind that the vast majority of textual variants do not involve (as it seems) theological corruption. So, while most textual variants may have arisen early, the comparatively few theological corruptions could have been late on the scene. Of course, others have thought to find theologically motivated readings in, say, P46. I have not been inclined to agree, but in any case the numbers of such readings would be, I believe, comparatively small; but that doesn’t mean that they didn’t exist.

Point 3: I used the term “textual type” in connection with the papyri with some hesitation (see, e.g., p. 15 n. 52). Of course, in calling, for example, P72Alexandrian, I was simply repeating what others have said, and while such terminology may be anachronistic for the early papyri or otherwise problematic, it seems to me to be useful shorthand for describing the textual relations. But I would hope that my investigation does not depend to any great degree on such characterizations.

In his response to KHE, Royse also mentioned some related points that Jongkind had made. However, I have not yet posted Jonkind's presentation. On the other hand, Royse had a separate more lucid response to Jongkind, so therefore I hope our readers will have forbearance with this little anachronism.

Earlier posts in this series:

Part 1: Juan Hernández' presentation

Part 2: Royse responds to Hernández

SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 2

In the first part of my report on this book review session, I summarized Juan Hernandez presentation. He kindly offered the full version, which is now available for download in the right sidebar (TC Files). In this second part I was initially going to post a summary of the next presenter, Kim Haines-Eitzen's response to Royse's monograph. However, earlier this week I received a message from James Royse himself who, in this spirit of sharing, offered his complete response, which he has re-edited since the meeting. I will therefore now post the first part of this response. After this series is complete, the full version (PDF) will be made available among the TC Files.


James Royse responds (pt. 1):

First, let me thank AnneMarie Luijendijk for organizing this session, and my fellow panelists for their perceptive and generous remarks. It is, of course, an honor for an author to have one’s work reviewed by such distinguished scholars of the text of the Bible. And I am pleased, and more than a little embarrassed, that my work, begun when we were all very much younger (if alive at all), has received such positive reception. Time does not permit a full discussion of all the points that they raise. Very generally I would wish to second the calls for further investigation of various issues; there is much to be done in the study of these six papyri and other important manuscripts, and in the study of scribal habits, and there are too few of us to give adequate attention to all the readings in all the manuscripts. Nevertheless, the panelists have raised some points to which I would like to reply, in the spirit, well illustrated by their remarks, of scholarly cooperation in advancing our common goal of shedding light on the process of the transmission of the text of the New Testament. I will basically follow the order of the speakers, but with some cross-references.

First, there is Juan Hernández:
I appreciate Hernández’s comments, and am pleased to have been associated in some small degree with his doctoral work. I confess that some of the details of my dissertation have faded from my mind with time, and have been overwritten with the revisions and expansions found in the current book. Indeed, hearing Hernández’s description was in some ways like hearing of someone else’s work. But I am very grateful for his kind words, and for his comments about the role that my dissertation played in his own study of the New Testament text. As a very small footnote to all this, the appendix in my dissertation on P46 and the Ethiopic, to which Hernández refers, was not included in the revised book. Rather it is cited as a forthcoming article. At some point I decided that separate publication would be more appropriate, and would save some space in the current book. However, I never seem to be able to put the finishing touches on the article. Perhaps the enthusiasm of the current session will enable me to conclude that little work as well, although for now it exists only as an appendix in the dissertation.

SBL Boston, Book Review of James Royse Scribal Habits in Early Greek NT Papyri, pt. 1

I will now try to finish my maraton report from the SBL in Boston – I still have some months to do this before the next meeting. You may think it comes a bit late, but it is too important to be passed over. It is time for the final, the review of James Royse, Scribal Habits. Again I want to state that this summary is in my own words, sometimes omitting, sometimes elucidating what was said in the session.


SBL24-129 New Testament Textual Criticism



Theme: Review of James Royse, Scribal Habits in Early Greek New Testament Papyri (Brill)

AnneMarie Luijendijk, Princeton University, Presiding
Juan Hernandez, Bethel College, Panelist (25 min)
Kim Haines-Eitzen, Cornell University, Panelist (25 min)
Peter M Head, Tyndale House, Panelist (25 min)
Dirk Jongkind, Tyndale House, Panelist (25 min)
James Royse, San Francisco, CA, Respondent (25 min)
Discussion (25 min)

Juan Hernandez
The first reviewer was Juan Hernandez. He offered a personal reflection of what Royse’s work has meant for him. In his witty introduction he stated, “This is a work of ‘singular importance’. It is an exemplar for all who want to study scribal habits.” What then did Royse’s work mean for a Ph.D. student specifically? In 2003 Hernandez stood as a crossroads. He had to come up with a dissertation proposal. His interest was in textual criticism, but there was no text-critical scholar around. How was he to embark without specialists? He contacted a lot of scholars about what to do. He was interested in the Book of Revelation. The path eventually led to James R. Royse. The approach of Epp (The Theological Tendency of Codex Bezae) and Ehrman (The Orthodox Corruption) on theological variation aroused his interest. Could this perspective be applied to Revelation? One problem was the fact that there is no Western text of Revelation. Could Ehrman’s method be applied? No. Christological controversies and their bearing upon the transmission seemed rather marginal. Then Hernandez found the work of Royse (i.e., his dissertation).

The undertext was: Scrutinize every claim!; Check everything for yourself! Even Royse’s 1981 dissertation was considered groundbreaking and remained a standard work for twenty-seven years (until it was superseded by the monograph). To put this work into perspective, E.C. Colwell had urged that someone someday would publish a commentary on singular readings (and he was talking about the three papyri he had studied). Royse came along, corrected and updated Colwell’s original work. Royse checked the readings against all available editions. His study was much more nuanced than an attempt to classify the text of these witnesses in broad terms, e.g., according to text-types. Royse goal was to cast light of each scribe’s habits. This was necessary to be able to arrive at canons of criticism applicable to papyri. Hernandez said he was “electrified” by the 1981 study. He was given “access.” It was a gigantic how-to-do manual. It was permeated of transparency. Everything was available for scrutiny. His prior interest was in theological variation. Royse, however, was not a friend of “theologizing.” Royse applied an extreme caution. He singled out only three variants that were theologically motivated (in P72). It was cristall clear for Hernandez that he would have to learn from Royse. He now had to put behind him notions of theological variation and start with the mundane, the facsimiles! Everything that he needed was there in Royse’s work, being the model.

Then Hernandez said some words about the monograph: The current publication superseeds the prior study in a number of ways. It has more chapters, more nuances in classification of singular readings, many new topics, and exhaustive appendices. The only missing thing was P46 in relation to the Ethiopic (a joke!). In the new work Royse has identified sixty-four additional singular readings (I think). The thesis stands concerning the implications for the canons of criticism. The burden of proof still rests with those who will prefer the short reading (in the early papyri). In sum, the monograph is a veritable encyclopedia of scribal habits (surpassing Colwell’s original urge for a commentary).

More to come.

Update: Juan Hernandez has kindly offered his full presentation to our readers. It can be requested from tomwas[at]spray[dot]se. (See also next post.)

SBL Boston, Book Review of David Parker, An Introduction to the NT MSS and Their Texts, pt. 5

We have come at last to the final part of the book review session (SBL 24-32) of David Parker's An Introduction to the NT MSS and Their Texts (CUP), featuring David Parker's response. Again, this is a recapitulation from my notes and memory.

Part 5

David Parker’s response
Parker began by expressing his gratitude for the kind things that have been said about his book. [Rather humurous in light of the serious critique that had just been raised.] He now wanted to share some of his experiences writing the book, and this also answered some of the critique. On the one hand, he did not make the new order of this introduction up. On the other hand, he did not use the "proto-type model" of Scrivener’s introduction. Parker thinks this old way is not the best way to go about today. Instead, he wanted to create a story in a narrative framework. The ordering of the book simply represents the story Parker wanted to tell. He decided to tell two stories. The first is the “knowledge of documents”-story. It locates the book in a philological tradition. Part one is about the documents. Part two describes how the documents have been studied, why the text changed, and how to edit the text. Part three shows how this applies to the sections of the New Testament.

Parker did not expect to write this book when he did. But he felt there was a gap between those that apply textual criticism, and those who are editing the NT. The former did not know about the major recent changes that are taking place. This is not the fault of the discipline at large. The concept of editing needed explanation. The books that were out there were completely out of date in this regard. The second part of the book, “textual criticism and editions,” in particular offers a “what is happening today-perspective.”

Finally, concerning the third part of the book, “the sections of the New Testament,” it is usual to provide a section termed something like “practice.” Parker had found these parts boring in other books; they want the reader to imitate. If these rules are so useful why does none agree on their application? Housman said “every textual problem is potentially unique." In his response, as in the book, Parker continued to show his disinterest for the canons of criticism. He does discuss thoroughoing ecclecticism, conjectural emendation, and the like, but such is admittedly woven into the overall framework of the book.

Then some words about indexing and navigating: Here Parker largely blamed the publisher (Cambridge University Press), since he himself wanted to put everything in there. The publisher did not. They wanted a “target”-perspective (focused on those issues that get a paragraph). Parker will discuss this again with the press for future editions.

The issue of readership is important. The book is intended to share contemporary scholarship within the guild. There is no point in writing an introduction for anybody. The discussion has to take place. Already Parker’s earlier monograph, The Living Text of the Gospels, was a kind of introduction for a more general audience. All four of the panelists are North-American males over 50 years. Parker says the book is also aimed at the new generation. We need more voices.

About the choice to put the images on a website, it is because they come out too badly in the printed book. All you see is an outline of a manuscript. You see so more on the website. There is only this way to go. You can always access the internet on you mobile phone but the display may be a bit small [laughter].

Parker then responsed to some of Ehrman’s critique. He had suggested that Parker is someone who knows everything about manuscripts but not texts. Ehrman kept using the phrase “idiosyncratic” about the book, e.g., about the use of terms. Parker was not problematizing the terms but tried to make them clear. The issues are tendentious. Ehrman had complained that Parker did not discuss Royse’s work on scribal habits, but he does and he is even mentioned in the index. Many of Ehrman’s examples neglect a different part of the book. For example, palaeography is indeed discussed on five pages; there is a discussion of Latin and of Coptic palaeography in the book. What Ehrman is revealing in his critique is his own views on various matters. For example, Parker emphasized that the quantitative analysis is past. [Here I should point out to Parker’s defense that various versions and developments of quantitative analysis, including Ehrman’s Comprehensive Profile Method, is indeed treated in the book, albeit briefly but with ample bibliography (pp. 163-65). Parker’s conclusion (p. 165) is that one common problem with these methods is the definition of group relationships of the basis of the whole text as it stands in a MS, and on the basis of a pre-determined sample. But in recent years a different technique has been developed – the ECM implied – which “has been applied with worthwhile results.” In this part of the book, the quantitative methods actually get more coverage than the CBGM!]

Then Parker proceeded with Epp’s defense of text-types, i.e., Epp’s objection against Parker’s rejection of the concept: Parker said he accepted everything in Epp’s response [Did I misunderstand this!?] But then he took a hyphothetical example of 50 copied MSS. The bottom-line of the example was that the concept of text-types is very static. Parker wants to know how the forms developed, a family tree, how the text changed. The division into text-types does not tell you that, and the contamination between text-types is a hinder. The CBGM gets around the problem with contamination. It is now possible to build a family tree of the textual flow. Text-types are no longer particularly important, we are now able to move beyond them.

[To some extent I agree with Parker here. Although I still find it difficult to talk about text-critical problems in practice without being able to generalize in the traditional way, e.g., “the Alexandrian witnesses attest to reading x” or “the Byzantine tradition is split between x and y.”]

How do we fit the versions into this? (Epp suggested that a consideration of the “Western” text-type must take the versions into account). The point Parker made is that you cannot reconstruct the Greek Vorlage of e.g., the Syriac version. In the Catholic Epistles it happens to be possible to pinpoint the Harklensis group. But often otherwise it is not possible in detail. [Hence, there are serious methodological problems with the identification of the “Western” text-type to include versional witnesses.”]

As for the definition of textual history, brought up by Holmes, textual history, said Parker, is the history of the whole text after the variants have been described. Parker then separated two issues in Holmes’ response:

1) Authorial fallacy
There is a real gap between the textual tradition and the author’s intention. Here Parker managed effectively to use the reference to his associate, Peter Robinson (to whom Holmes had referred), in detail to his own defense. In the example of Chaucer, the textual tradition of which Robinson has worked with, we can get back to just a little step from Chaucer, though there is still a gap between the “initial text” and the “authograph.” [It does not surprise me that Parker managed to use this particular objection to his own defense – he must have had opportunity to discuss these issues with Robinson over hundreds of cups of coffee/tea in that nice project room they have in Birmingham.]

2) The second-century text
Parker now brought up that they are reconstructing Paul’s collection. The goal is to edit the corpus. There is again a gap between the “initial text” and the author’(s’) text. In the question of exegesis, there will always be a gap. It is positivistic to say that NA27 is what Matthew, Mark, Luke and John, etc., wrote. Just read what the editors of NA27 say in their introduction.

Finally, Parker emphasized, supposedly concerning what method he uses, that every place in the tradition may be potentially unique as to what factors (call them canons if you like) are applicable.

In the time for questions I wanted to bring up several issues, but my main comment was this:

What exactly is reconstructed by the CBGM, the editors of the the ECM themselves are actually not agreed on. The easiest hypothesis, says Gerd Mink [Occam and I agree], is that the “initial text” of a given book is the original text, not introducing any other explanation. Particular factors like the affect of a collection have to be addressed. But, interestingly, the results of the CBGM in the Catholic Epistles show that the individual books have their distinct transmission history, suggesting that we can reconstruct beyond the point when the books were collected. Moreover, I have already pointed to the use of intrinsic evidence to move beyong the “archetype,” and this also applies to an archetype of a collection. I am now eagerly awaiting the results for the Pauline corpus (Parker nodded his head – he is too).

What I really appreciated with Parker’s response was that he managed to keep it in a good tone, in spite of the "attacks," especially by Bart Ehrman. And, finally, the presiding Amy Anderson gave Parker the last word of the session, which he actually ended with a positive appeal to the panelist and to all present, "Times are exciting, let us move on!" This in the same enthusiastic note on which he ends his book. Let me therefore close this review by citing the conclusion of the book (p. 349):

“The curent generation is experiencing one of the most momentous periods in the transmission of biblical texts and of much in textual scholarship. It will be equally exciting to see what happens in the next few years.”

SBL Boston, Book Review of David Parker, An Introduction to the NT MSS and Their Texts, pt. 3-4

Okay, I know this is a bit late, but for completeness sake I now continue my "SBL sessions review marathon."

This is the third and fourth part of the book review session (SBL 24-32) of David Parker's An Introduction to the NT MSS and Their Texts (CUP), featuring Michael Holmes and Larry Hurtado.

Disclaimer: Note that this is a summary in my own words. I took notes very quickly, and there may be things that I have misunderstood, but, nevertheless, I hope to have captured the essence of the reviews.

Part 3

Michael Holmes
Holmes started with some appreciative words about the book and the good intention behind it, as expressed by Parker in the introduction. Holmes likened the book to “a fantastic map, a catalogue and an instructional manual.”

He then gave some concrete examples of its extensive treatment of subjects like corrections in MSS, editing text and editions including the challenge of electronic editing, tools for manuscript studies, collating MSS, information about subgroups of MSS, etc.

Then Holmes brought up some less satisfying issues, some of which I do not repeat (e.g., those that were brought up in Ehrman's review). For example, as a result of the organization of the book, which does not promote usability, and the incomplete index it took Holmes very long time to find the indication of how many Greek and Latin bibles that have been preserved.

He then proceeded with more substantial critique, specifically relating to Parker’s description of his method, the history of the text, and the goals of textual criticism, factors which form the overall understanding of the discipline:

1) Method
Parker does not mention the term "reasoned ecclecticism," but Parker apparently uses this method himself. He focuses on the documents, then the texts. But there is no discussion of method or criteria. The canons are woven into the discussion of examples. The method is scattered and woven in.

2) The history of the text
Parker says: “Textual history is the history of the changes in wording of a text” (p. 179). This involves collection of material and examination of the evidence "in order to relate the different forms of text chronologically” (ibid.). Parker rejects the concept of text-types (cf. Epp’s review that focused on this question), but Holmes is dissatisfied with the terminology Parker uses himself to describe the history of the text. What does he mean by “forms of text”? Presumably the text as a whole, as reflected in Parker’s definition of textual variant: “the entire text as it is present in a particular copy” (p. 4). At other times, however, Parker seems to speak about individual variation units, or the distinct textual history of individual books. In sum, Parker’s description of the concept of textual history is ambiguous and the terminology is unclear.

3) Goals and objectives
After all, Parker’s view of text-critical methodology and textual history reflects a consensus to a great extent. However, not his view of the goals and objectives. Parker has based his whole book on Hort’s dictum that knowledge of documents should precede judgments of readings, which actually forms an inclusio of the book. Whereas Hort’s goal was ultimately to reconstruct the authorial text, it has now been complemented by secondary goals and different terminology (e.g., “initial text,” etc). Parker thinks the initial text is from the end of the second century. And the gap between the initial and authorial text is a big problem.

Parker suggests that variants should be treated as early forms of commentary (not authorial). His formative vision is a move from the first traditional goal to the second goal, the reconstruction of the history of the text. Here Holmes finds himself in strong disagreement with Parker. Take the statement that contemporary editors do not claim to reconstruct the author’s text. They may have been wrong, but have they not expressed that purpose? This is still an open question. Interestingl, Holmes here refers to Peter Robinson (Parker’s close associate at the ITSEE in Birmingham), who has indeed expressed his desire to reconstruct the authorial text of works he edited!

Holmes continued to emphasize the continuity between the text at the end of the second century and the text that preceded it. He brought up some critique of William Petersen’s use of patristic evidence (e.g., Justin’s text), to which he appealed in order to prove that the text was unstable during the earliest era. Holmes concluded that the challenge of showing continuity between the text before and after the second century is not greater than of showing discontinuity. Parker is correct that we should not underestimate the challenge, but we should neither exaggerate it.

Part 4

Larry Hurtado
Hurtado was impressed by the width, and cover of languages, tools and resources, and more. Parker’s own enthusiasm is apparent. He promotes the study of manuscripts in the whole first part of the book. Hurtado felt gratitude and appreciation for very much. He was a bit surprised of Parker’s attempts to offer theological advice in the book (I assume he primarily had in mind the section 5.3.3 “Textual criticism and theology” on pp. 185-90).

Then Hurtado proceeded to mention some critical matters:

1) Who is the primary reader?
Hurtado suggested that the primary readers among students will only be a small number of post-graduate students. There are little incentives for less familiar readers to become enthusiastic themselves.

2) The lack of explicit criteria
As several other reviewers, Hurtado complained about the lack of an explanation of criteria. This is curious since the CBGM is ultimately based upon such criteria. The dicussion of textual criticism is particularly lacking.

3) The organization of the book
The order of presentation is warranted in one way (focusing on “knowledge of documents”), but the decision is a misjudgment for an introductory book, where one must particularly take into account the interest of the readers (who are primarily interested in the New Testament). Hurtado thinks the publishers must be interested in offering an attractive introduction to the field of textual criticism too, as the title suggests. The undifferentiated enthusiasm with all MSS is not effective (cf. Hurtado’s own particular interest in the earliest MSS pertaining to the early Christians). Parker seems to presume that MSS of any period of time will be interesting for the student. Moreover, Parker should have made better connections between the parts of the books (I presume Hurtado especially referred to the first and second parts, “the documents” and “textual criticism and editions”).

It was not a good decision to put the images on a web-site. The reader must have an internet connection open at the same time. There should at least have been some key samples included in the book. It will be impossible to study the book e.g., on the bus, or at the airport. [This of course is an overstatement in light of the development of “wire-less” connections.]

4) Other issues
Hurtado continued with some specific problems e.g., with some terms. The definition of reading seems to be better applied to a variation-unit (including “place”). Especially in an introduction these labels and the consistent use of them are very important. Moreover, there are also some unfortunate errors concerning MSS, e.g., P12is not a copy of Hebrews in particular, and not a roll.

Then Hurtado mentioned some words about the CBGM method, mainly suggesting that what is generated depends on the criteria used in individual units. He did not think the CBGM is an advance in terms of how we evaluate the evidence in individual passages, it just offers more transparency.

In the time for questions I brought up this last remark about the CBGM. I actually think the CBGM is an advance not only in transparency, because it offers a new and more controlled way of moving from internal to external evidence, a kind of “calibration” of the external evidence.

In the final post on this book review session I will attempt to summarize David Parker’s response.

SBL Boston, Cosaert on Clement's Text

My report from the SBL in Boston continues. The following is a summary from the first paper of the New Testament Textual Criticism session on Saturday afternoon. This was the session in which I presented a paper with Jennifer Wright Knust, which probably caused some distraction on my part.

Carl P. Cosaert, “The Text of the Gospels in the Clement of Alexandria”

Since Ehrman’s dissertation on Didymus the Blind, there has been a steady flow of studies on church father’s text especially those located in Alexandria (Origen, Athanasius, Didymus, Cyril). Now the turn has come to Clement of Alexandria. What form of the text existed in Alexandria in the 2th century? Did his text influence later fathers? These questions led Cosaert to research the subject. In this paper he summarizes his study of the text of Clement.

The study is based on testpassages in Matt, Luke, and John. The methodology was developed by Ehrman (including Comprehensive Group Profile Analysis). The first step, however, involves the identification of citation and allusions. These are then collated against control witnesses representing various textual strands, and also including other Alexandrian fathers.

The first table displayed data based on 118 significant variations in Matthew. At the top, closest to Clement were Athanasius and Origen, then Sinaiticus. After that came a puzzling combination of witnesses. It has been suggested that Clement’s text represents an “early stage of Alexandrian text.” The conclusion in Matthew, however, is that Clement has no clearcut agreement with any one group. It is split between the Alexandrian and Byzantine groups. Cosaert had hoped that the group profile analysis would clear things up, but it did not. Clement does not preserve any distinct readings for a goup. The results have been more definitive in Didymus and Athanasius.

In Luke, there were 143 units of variation (more data). Here Origen comes at the top, then D, 33 and Didymus. Cosaert pointed out that Clement’s text were closer to the Western text in Luke. (Clement had 58% agreements with D i Matt and 66,4% in Luke). [TW: Note that the difference is ca. 10 readings.]

The text of Clemens in John has much clearer affiliation. Alexandrian witnesses dominate the top of the list (7 of the first 10 are Alexandrian). According to Cosaert, it is not surprising that Didymus is at the bottom of the list, since Didymus according to Ehrman shifts textual character in John. Hence, Alexandrian is probably the classfication to apply in John.

But in Matthew and Luke the text is mixed (with Byzantine influence in Matthew, Western in Luke). The results are different from other Alexandrian fathers.

Cosaert suggested two possibie explanations:
1. Clement’s textual relationships may not be so clear because the readings are not distinct or exclusive for any one group (shared by more text types). [This is a good reason!] Then he looked at variants. For example in Matt 12:36 it was apparent that the testpassage was not so useful for classification purposes, since the tradition is split. Same in other examples in Luke 10:21 and 10:42. There is some need of adjustment.

2. There were two major text types, Alexandrian and Western, without any one dominating. [TW: Or, Clement used different sources.] There was no clear situation in Alexandria. The text is more mixed (cf. also the Coptic tradition).

Cosaert also looked at how Clement used the text. For example, whenever he is using Paul he seems to clearly follow a manuscript. Furthermore, he often he attempts to harmonize.

Comments and questions:
Hurtado: The suggestion that there are different transmissional dynamics is compatible with analysis of papyri too. Some are diificult to classify, and other like P75 align with later witnesses (continuity).

Jongkind: What do we mean with a “fluid text”. Does not your conclusion undermine your own analysis (there was no clear text types at this time, they develop later, why include them?).

[I did not record a reply to this question, if there was one.]

Read more about Cosaert’s study as reported previously on this blog here.