Wednesday, November 17, 2010

SBL GNT: Three Questions and Replies

Earlier this week I hosted a chat session on Facebook (here) about the SBL GNT (occasionally an old dog can learn a new trick or two). A few of the questions raised and answered may be of wider interest, so I am taking the liberty of posting them here as well.
Question: one person wondered why the selected critical editions where chosen, and why others (such as the earlier editions of the GNT by Eberhard Nestle) were not.
Reply: The selection of four editions to serve as primary resources for establishing the new SBL edition—WH, Tregelles, Robinson-Pierpont, and the Greek text behind the NIV translation—is based on a combination of reasons. Westcott-Hort—the starting point for both the early Nestle and the UBS editions—is a classic, and an obvious choice (and it was available in e-format). Re the Byzantine tradition: unlike WH, I’m convinced that at points, only the Byzantine tradition preserves the original reading, and so it had to be represented in the mix. Also, it is the textual tradition of choice for the Orthodox tradition, and so deserved to be represented. The Robinson-Pierpont edition (available in e-format) is an excellent representative of the Byzantine tradition, and thus was an easy choice in this regard. As for Tregelles, I’ve long been impressed by his independent and insightful judgment, an impression confirmed by David Parker’s analysis of the text of James in the Editio Critica Maior in comparison to earlier editions, which validated a high percentage of Tregelles’ choices. With Tyndale House having recently made it available in e-form, it commended itself as a good counterpoint to WH.
The choice of a modern critical edition was more complicated. First, there aren’t many to choose from. Second, the publishers were rightly concerned that the new edition have a clear and uncontestable copyright—both for the sake of the new edition, and out of respect for existing editions. Thus in light of their concerns it seemed best not to use NA26-7/UBS3-4 in the initial stages of creating the SBL edition. Once this decision was made, then the reconstructed Greek text behind the NIV translation commended itself as an alternative: it records, where it differs from UBS/NA, the independent textual choices of the original NIV translation committee (an international group of respected scholars), yet (because the NIV translation committee used the UBS/NA text as its starting point) it reflected, indirectly, the NA/UBS textual tradition as well. In effect, using the Greek text behind the NIV allowed the project to maintain a clear copyright, to acknowledge (indirectly) the well-earned excellence and stature of the NA/UBS editions, and to include (where the NIV differs from the NA/UBS text) a contemporary alternative perspective to the textual decisions of the committee responsible for the NA/UBS editions.
In short, the choice of these four editions to serve as initial resources (but certainly not the only resources) for building the new SBL text reflects a combination of historical considerations, text-critical theory, publishers’ concerns, and pragmatic considerations.
What about other editions? Each has benefits, but also drawbacks. For example, Tischendorf relied too heavily on Sinaiticus. Von Soden (still valuable as a source of information) built his text on flawed theoretical assumptions; and Vogels, Merk, and Bover are heavily influenced by von Soden. The original Nestle tradition was built on the editions of WH, Tischendorf, and (after Weymouth) B. Weiss, who like WH was heavily influenced by Vaticanus, so the Nestle text (up through the 25th edition) is very similar to WH. Either one would provide a good starting point, but both together add little to what one gains from either one alone. Similar considerations apply to other potential candidates (such as the Greek text, edited by R. V. G. Tasker, putatively behind the New English Bible).

Question #2: “Why is the NIV considered more standard than the NA in the Apparatus?”
Reply: I don’t consider the Greek text behind the NIV translation “more standard” than the NA text. The fact that “NIV” is visually more present in the apparatus than “NA” is largely a matter of pragmatism. Let me explain in two steps. First, once the Greek text behind the NIV was selected as one of the four primary resources for the new SBL edition (see the answer to the first question above for the reasons behind this choice), it had to be included in the apparatus alongside the other three editions. Second, from the start I wanted the apparatus to include the evidence of the NA/UBS text, from which the Greek text behind the NIV differs at about 235 or so places. So now the question was a practical one: how best to incorporate the NA testimony into an apparatus that already includes the NIV testimony? Option 1: insert the NA evidence at every one of the 6928 variants, or option 2: insert it explicitly only at those 235 or so places where it reads differently than the NIV, and for the other nearly 6700 places let the NIV stand for both NA and NIV. With either option, the full testimony of NA is present in the apparatus. However, option 2 results in a more compact apparatus, and reduces the chances for error in constructing the apparatus, so the choice was made to adopt option 2 rather than option 1. Note that the choice is entirely a pragmatic one, and does not imply any claim that the “NIV” is somehow “more standard” than NA. Given the sequence in which the apparatus was built, it was simply a pragmatic choice to cite NA only where it differs from NIV. In any case, the full testimony of NA is presented, either explicitly or implicitly, in the SBL GNT apparatus for every one of its 6928 variation units.

Question #3: Would the end product be different if I had started with the manuscripts rather than editions? That is, did the use of existing editions rather than the manuscripts as the starting point for the SBL edition bias the outcome?
Reply: I don’t think that the end result would have been any different if I had started with the manuscripts rather than existing editions. To explain why, let me quote a section from the Introduction to the edition:
Where all four editions agreed, the text was tentatively accepted as the text of the SBL edition; points of disagreement were marked for further consideration. The editor then worked systematically through the entire text, giving particular attention to the points of disagreement but examining as well the text where all four editions were in agreement. Where there was disagreement among the four editions, the editor determined which variant to print as the text; occasionally a reading not found in any of the four editions commended itself as the most probable representative of the text and therefore was adopted. Similarly, where all four texts were in agreement, the editor determined whether to accept that reading or to adopt an alternative variant as the text.
That is, once the four editions were compared, I then systematically went through the entire text of the NT in light of the manuscript evidence, not only where one or more of the four differed, but also where all four were in agreement (and at several places, I did adopt a reading not found in any one of the four). In short, the entire SBL text was decided on in light of the manuscript evidence (not on the basis of the editions); in addition, in view of the fact that there are places throughout where I either agree with each one of the four against the other three (SBL/WH vs. Treg NIV RP: 98x; SBL/RP vs. WH Treg NIV: 66x; SBL/NIV vs. WH Treg RP: 59x; SBL/Treg vs. WH NIV RP: 28x) or reject all four, I don’t think the use of existing critical editions biased the outcome.


  1. Thanks for posting this Mike, really helpful.

  2. Thanks.

    Here is a little question/problem that I am curious about. Perhaps some readers of the SBLGNT might be able to assist me. Working from the premise that the SBLGNT is always correct, how many corrections to the text have been made which were not in the text of Trg, or WH, or Byz?

    Just in case this can be easily done using some ingenious search-software to sift through the apparatus: the question may be rephrased so as to ask how many times in the SBLGNT does Holmes adopt a variant that is either unique to NIV, or unique to NA, or unique to Holmes?

    A shiny ruble to whoever is first with the correct answer!

    Yours in Christ,

    James Snapp, Jr.

  3. James,
    some of the information you seek can be found in the Introduction. It includes, for example, the following data:
    SBL + WH vs. Treg NIV RP: 98
    SBL + Treg vs. WH NIV RP: 28
    SBL + NIV vs. WH Treg RP: 59
    SBL + RP vs. WH Treg NIV: 66

    And footnote 9 reports that "In all, there are fifty-six variation units in the SBLGNT where the editor preferred a reading not found in any of the four primary editions. In thirty-eight of those instances, the editor’s preferred reading is also read by WHmarg (30x) and/or Tregmarg (2x) and/or NA (10x)."

    I realize this is not all the information you are seeking, but it may include some of it.

  4. FYI, I tried to install the SBLGNT eSword module on my computer and it failed, due to the fact that my computer is 64-bit Windows 7.

    David Robert Palmer

  5. This comment has been removed by the author.

  6. Dear Mike:

    You stated -

    "the entire SBL text was decided on in light of the manuscript evidence (not on the basis of the editions); ... I don’t think the use of existing critical editions biased the outcome. "

    My question is not whether editions per se biased the outcome, but whether or not methodology skewed the outcome.

    In particular, we have noted some 70 rather obvious cases of homoioteleuton in the Alexandrian text, and discussed them here:

    70 Homoioteleuton errors < - - Click here.

    Right now we are composing reconstructions of the physical layout of the master-copy behind the ancestor of Aleph/B, which gave occasion for the errors to arise.

    Since your methodology obviously led you to reject most of these readings as cases of homoioteleuton, are you willing to at least describe the methodology you used which led you to that conclusion?

    I am not expecting you to change your editorial decisions, or defend in detail each of the 70 cases.

    I am only interested in the methodology, i.e., the set of steps, and/or list of 'canons' or rules by which you arrived fundamentally at an entirely different point of view than us regarding these 70 variation units as a group.

    thank you in advance,

  7. to Nazaroo: a quick look at your proposed errors are not always (IMO) necessarily cases of homoioteleuton, though some cases probably are. There are other explanations as well, such as harmonizations that can produce longer texts.

  8. > My question is not whether editions
    > per se biased the outcome, but
    > whether or not methodology skewed
    > the outcome.

    > In particular, we have noted some 70
    > rather obvious cases of homoioteleuton
    > in the Alexandrian text, and discussed
    > them here:

    If you mean by methodology 'textual critical theories', then yes it would bias the outcome, just like the textual critical theory that assigns all 70 variants you mention to homoioteleuton biases the outcome in a different direction.

    If by methodology you mean the method of comparing the versions, then I would say it doesn't. If these are truly alexandrian homoioteleuton and the verses present in the majority of manuscripts, then the comparison with RP would have shown them up. Clearly Mike Holmes gave RP readings a fair hearing, and that each of the 70 places you discuss would have been examined in detail.

    The kind of thing that would be skewed from this examination would be some reading in which WH, Treg, NIV and RP are in agreement, and is not (yet) a "well known" textual issue.

    The lack of a Western witness is the obvious hole, but clearly those Western readings that may have some claim to originality were examined since some those readings were adopted.


  9. What's most interesting to me is that SBLGNT elevates the Byz reading from 'late corruption' to 'original text' a whole 66 times. I can think of a couple of possible causes for this:

    1) Publication of the Byz text has caused textual scholars, who long since dismissed the TR as 'dethroned,' to take a look at those variant units (seldom even mentioned by NA) where the baby may have been tossed out with the bathwater.

    2) A change in methodology, since the evidence being brought forward was known to previous generations of textual scholars who nonetheless rejected it.

  10. Permit a combined response to the posts by Nazaroo and Daniel Buck: One can add to Dan's two possibilities (neither of which is at all a factor with respect to the SBGNT) a third possibility: a different perspective or understanding of the history of the transmission of the text. Methodology does not work alone; it works only in conjunction with a view of the history of the text. Westcott-Hort and Zuntz work utilize a very similar methodology, but arrive at rather different results because they work with very different perceptions of the history of the transmission of the text.
    My own understandings of methodology and textual history have been deeply (but certainly not exclusively) shaped by Zuntz. I have written a fair amount on both aspects (method and history); for an overview, please take a look at my bibliography page ( For those who want or require something in electronic rather than printed form, I have just posted an edited version of a paper I gave at the SBL meetings in 2006 (see: I think the three illustrations of different ways of viewing the textual history of the NT (on pp. 9, 10, and 14) will give some idea of why the SBLGNT gives more attention to Byzantine readings than do, e.g., either WH or the NA/UBS editorial committee.
    (If the above links don't work, you may access both via; the link to the paper is on the "SBL Greek New Testament" page.)

  11. Dear Mike:

    You state in your 2006 paper:

    "This idea of the tradition as a broad stream, with the surviving evidence representing only small scattered subportions of it, is, in my opinion, an extraordinarily fruitful aspect of Zuntz‘s historical reconstruction."

    If you still hold to this reasonable point of view, I invite you to look at our reconstruction of the stemma for John 7:53-8:11 here, which takes into account a more modern "stream-view" of the history of textual transmission, and allows for the paucity of early MS evidence.

    PA Stemma <- - Click here.

    Of similar interest should be our identification of a significant "hole" in the MS evidence for centuries 6 to 11, discussed in our article here:

    The Missing MSS < - - Click.

    We would be delighted to have your comments on these issues updated.


  12. Dr. Holmes,

    Thanks for the link to your lucidly written 2006 SBL paper.

    You say:

    "Thus one‘s view of the history of the text is no less important than one‘s basic approach to methodology.

    "For example: Maurice Robinson works with the same toolbox of transcriptional and intrinsic considerations as virtually all the rest of us. Yet he has produced a quite different text. A fundamental reason for this different outcome from the use of the same set of tools is that he works with a much different conception of the history of the text. On his interpretation of the history of the text, it is impossible for a reading supported only by 'Western' or Alexandrian witnesses (or a combination thereof) to be an original reading, and this conviction deeply shapes his use of internal considerations."

    Has Dr. Robinson ever stated it quite this way? I have thought that he, just as Hort with regard to the Alexandrian manuscripts, has always allowed for the theoretical possibility that a purely Byzantine reading (rare as they are!) could be inauthentic or have the appearance of being inauthentic. It is in such cases, just as Hort held with regard to his favorite group of manuscripts, that we would do well to trust the type of text that we are morally certain has habitually preserved the original text on a sequential basis in the vast majority of other places. No doubt Dr. Robinson will chime in as soon as he is feeling better after recovering from quadruple-bypass heart surgery, but as one of his former graduate students I think he would probably word it more like this: It is certainly possible that a purely Byzantine reading is wrong, but more probable that it is right!

    Jonathan C. Borland

  13. Hello, Jonathan,
    I probably first proposed this analysis (and used the Powerpoint diagram) of Dr. Robinson's view of the textual history of the NT at a conference in 2005 at which he and I were the two main presenters. At that time, he heartily affirmed both my analysis and the diagram as an accurate representation. Even earlier, at the conference on textual criticism at SEBTS in 2001, I asked Dr. Robinson whether there are, in his estimation, any instances where other textual traditions preserve an original reading that is not found in the Byzantine tradition. His answer was a clear "No."
    So, as far as I know, this was an accurate description of Dr. Robinson's position at the time I wrote it.

  14. Dear Dr. Holmes,

    Thank you for recounting the occasion where Dr. Robinson said that in his estimation he did not think a Byzantine consensus reading was ever corrupt. But obviously, even under Hort's theoretical history of the text, the clearly Byzantine reading almost always has either 'Western' or Alexandrian support of one kind or another. This is one reason I think your statement needs clarification.

    As I'm sure you have done, I have read all of Dr. Robinson's publications and always find that he deals with textual issues in probabilities and likelihoods. For this reason I also found your claim that Dr. Robinson says that ". . . it is impossible [emphasis mine] for a reading supported only by 'Western' or Alexandrian witnesses (or a combination thereof) to be an original reading" to paint an impression of his view that is generally unacceptable in academic circles, especially if he has not specifically stated such in publication. (At least a footnote after such a statement would have been appreciated!)

    Of course Dr. Robinson may support your statement as is, but I am certain (from his publications and private conversations) that he would fully support your statement with the expression "not likely" in place of "impossible."

    Jonathan C. Borland

  15. Forcing me out of bypass recuperation (which is no fun at all)....

    Borland: Has Dr. Robinson ever stated it quite this way?

    Holmes: He heartily affirmed both my analysis and the diagram as an accurate representation. Even earlier ...I asked Dr. Robinson whether there are, in his estimation, any instances where other textual traditions preserve an original reading that is not found in the Byzantine tradition. His answer was a clear "No."

    And in fact Dr Holmes is quite correct. There is a caveat, however, and that is that one should not confuse the conclusion and consequent application of the theory as developed with the initial inquries that led to the development of the theory itself.

    Just as Westcott and Hort should not be faulted for following strictly a methodology and application based on their theoretical inquiry that led to the development of a position favoring Aleph-B except where "western non-interpolations" occur, so also my Byzantine-priority position should not be misconstrued as having begun with an a priori assumption that the Byzantine reading was always correct.

    Once a further detailed study of transmissional considerations was added to the mix, I became ever more certain that Kenneth W. Clark's original suggestion to me was more likely correct than not; namely:

    That unless all texttypes can be shown to be later disparate developments stemming from a now-non-extant "original texttype", transmissional considerations strongly favor a single, reconstructable, existing textype as normative, so long as a reading of that texttype is overwhelmingly supported by its constituent witnesses (thus the principle does not apply when the primary Byzantine witnesses are seriously divided; but elsewhere, then yes, the term "impossible" would be appropriate, so long as all the preliminary exploration and inquiry is not turned into a mere unscholarly a priori).

    I do differ from one of Dr Holmes' statements, however: "this conviction deeply shapes his use of internal considerations." On the contrary, I continue to use and apply all relevant internal considerations to all readings within a variant unit, and teach my students to go and do likewise.

  16. This comment has been removed by the author.

  17. Dear Sirs,

    I stand corrected and sorry to have aroused Dr. Robinson from his recuperative stasis!

    Jonathan C. Borland

  18. Okay; I feel a little punchy from a long drive earlier today, but I wanted to kick around this question before bedtime:

    How many textual changes to the NT text have resulted from the discoveries and analyses made between 1881 and 2010?

    Here is the answer that I get, using SBLGNT as a standard for comparison:

    SBLGNT has 94 readings accepted by Trg or by RP which were not adopted by WH or by NA. And SBLGNT has 59 readings adopted in NA which were not accepted in WH. Plus SBLGNT has 56 special readings.

    So the changes between the 1881 WH text and the 2010 SBLGNT text = 94+59+56 = 209.

    And since the total number of variant-units involved = 6,928, this means that the changes between the 1881 WH text and the 2010 SBLGNT = 3.017% of the variant-units.

    There's a small adjustment to consider; on 10 occassions, SBLGNT disagrees with the NIV-base-text but agrees with NA27; I didn't check to see if all ten of these readings agree or disagree with WH, but they won't significantly change the results.

    Does this sound about right, or not?

    Yours in Christ,

    James Snapp, Jr.

  19. An interesting question arising from Snapp's calculation would be this: is our text 97% certain and we disagree for the rest 3%?-) If this is yes, what would that say about the preservation of the NT text (the history of transmission) overall?

  20. Timo Fink:
    "is our text 97% certain and we disagree for the [other] 3%?"

    I guess it depends on what your definition of 'we' is. But it does appear that publication of the SBLGNT has slightly lowered the first number and raised the latter.

  21. Timo,

    I'm still not sure I've got the numbers right. But assuming that they do imply that 97% of the readings adopted in WH were also adopted in SBLGNT, I'd adjust the nomenclature: instead of saying that the text is 97% certain and we disagree for the remaining 3%, it would be better th say that 97% of the variants adopted in 1881 were adopted in 2010. The percentages of the text as a whole and the percentages of the variant-units won't be the same.

    Yours in Christ,

    James Snapp, Jr.

  22. Thx, James.

    Indeed it depends on the "we" :) I'm sure Dr Robinson would not say "amen" to SBLGNT, at least in terms of the actual text.

    But it seems to me that the more things change the more they stay the same. It is as if WH is the default text "we" (whoever) are trying to improve and end up with WH-lookalike. My point is this. If the text of 1881 and the text of 2010 are so close to each other despite all the manuscript findings of the 20th century, would it be correct to say that almost all the readings are already known and whatever new manuscripts "we" find, they only add their support to the already known pool of variants. Seldom we find anything new.

    It seems to me that basically we have three streams of texts (Alexandrian, Syro-Latin, Byzantine) and this has not changed, even though more than a century of research has been conducted. Are "we" any closer to writing the history of transmission than "we" were a century ago?

    Something to ponder.


    P.S. I did not include CBGM as it's still an ongoing project.

  23. Flink: I'm sure Dr Robinson would not say "amen" to SBLGNT, at least in terms of the actual text.

    This is readily granted, since I neither would be in agreement with any other Greek NT text based upon reasoned or rigorous eclectic principles, nor with any text that would reflect essential texttypical difference.

    The fact is that I would concur far more in terms of theory, method, and transmissional considerations with anyone whose goal would be to argue a "pure Alexandrian" originality or "pure Western" originality, based upon a legitimate reconstruction of the presumed archetypes of either texttype, than with any current form of eclectic praxis (whether such archetypal reconstruction can be accomplished for minority texttypes is another matter entirely).

    As Clark and Epp both had noted (and with which I concur), eclecticism -- rigorous, reasoned, local-genealogical (= CBGM) or otherwise -- is at best a "holding action", still awaiting a valid replacement theory that will supplant current eclectic critical text methodology: one that is primarily externally based and historically sound, based on reasonable transmissional considerations. Aside from Byzantine-priority, no such theory seems as yet to exist (pace the claims of either Holmes or Zuntz).

    Flink: [We] end up with WH-lookalike....The text of 1881 and the text of 2010 are so close to each other despite all the manuscript findings of the 20th century

    This was in fact K. W. Clark's primary thrust some 40+ years ago in his "Today's Problems" article. Dr Flink in fact asks the same question as did Clark:

    Flink: Are "we" any closer to writing the history of transmission than "we" were a century ago?

    I would suggest that for those who practice within the various eclectic positions, the answer really has to be no.

    And by the way, though I did not mention it in my previous recuperation comment, my and Pierpont's original starting position before any theory or analysis was that of rigorous eclecticism, allowing any variant reading potentially to be correct, and evaluating all candidate readings on that initial basis. Only later did we develop in various stages what ultimately became the Byzantine-priority position.

  24. This may be as good a place as any to bring up the implications of Hikmat Kachouh's study of the Arabic gospels corpus on transmission theory. Having assumed a Byzantine recension at the end of the 4th century, Kachouh finds in the Arabic mss clear evidence for a gradual takeover by the Byzantine text, culminating about the 9th or 10th century in Southern Palestine and much later in Egypt, with some Coptic readings persisting in diglots into the 19th or 20th century.

    Since a great number of Arabic mss are dated by colophon, and the Islamic conquest permanently changed the Arabic language, Arabic mss can be dated with great precision as far back as the late 8th century. At that time, 'pre-Byzantine,' Diatessaronic, Old Syriac, and Caesarean readings dominate the Arabic corpus. Inasmuch as these oldest Arabic manuscripts are from a climate equally dry as Egypt's, it's interesting to see them exhibiting a quite independent text from any of the Eygptian papyri, yet much less Byzantine than the manuscripts which came to dominate in the 9th to 13th centuries, even in Egypt, and even to some extent in Coptic diglots (in which it was clear that the Arabic side had an independent history).

    Most Arabic versions had their source in Syriac, and a Peshitta vorlage dominates. But many readings can be traced back to Old Syriac, even in versions that were heavily edited toward the Peshitta. Only one version was corrected toward the Harklean, and that incompletely.

    Another interesting note is that some readings found only in a cross-section of ancient versions--but particularly Old Latin--are found in an Arabic ms from a strictly Greek vorlage, possibly indicating a common vorlage with readings now extinct in Greek (this has implications for the theory of an Eastern origin for 'Western' readings).

    One last contribution this study makes to textual transmission is the realization that many gospel books circulated independently, especially John. Furthermore, John was usually the least affected of the 4 gospels by correction toward a different vorlage.

    In conclusion, although the Arabic evidence fits a 4th-century existence of the Byzantine text in the dry climate of the Mideast, it shows it taking 400 years just to begin displacing older text-types, and another 400 years to more or less complete the process: definitely not in line with an Orthodox Recension Theory.

    Also, as a side note--there are two multilingual palimpsests with Arabic overlying, inter alia, Greek texts that don't appear to be Scripture--rather the opposite of how it usually works with palimpsests. They are Sin. Ar. 514 and Sin. Ar. N.F. 8/28.

  25. Dr. Hurtado (see his recent blog posts)

    has made an observation that demands serious consideration, whether one is looking for better methodology to handle existing materials, or simply better materials (i.e., more archaeological finds).

    He suggests that the future of NT TC really will be found in evidence and theories concerning the history of the text //before// 200 A.D., the critical period.

    This is the clear watershed and mystery-barrier where all the exciting things happened. It can only be penetrated by better methodology and more evidence.

    If the key to the NT text is in this critical period of development, then all things point farther and farther //away// from the W/H text, which after all is simply the developed and standardized 3rd century text.

    Its time we put that lame toy away, and focused not only on extant papyri evidence, which significantly differs from W/H, but also on finding more MSS from this period, as well as supporting and corroborating evidences such as the earliest NT writers.


  26. You have all made the news:

    Mr. Scrivener has quoted each of you in his own discussion of Eclecticism, here:

    Eclecticism. If anyone wants to add anything to their positions, I'm told he will accomodate you.


  27. Not that I am thrilled to have "made the news" (which particular "news" I normally don't read or interact with for reasons that should be obvious).... I do object to one statement cited via that link:

    "...acknowledged experts like Dr. Maurice Robinson are clearly Eclecticists"

    Most assuredly that is the one position that I do not hold and therefore do not accept such term being applied to my position.

    This was apparently due to the misunderstanding of what I stated respecting the order of inquiry when seeking to frame and establish a consistent methodological approach.

    The point is much the same as Westcott-Hort's initial and preliminary examination of readings on internal grounds prior to their establishment of particular external criteria that then becomes the basis for the establishment of the text without "eclecticism" per se playing a determinative role in the final result.

    Also, tell "Mr Scrivener" that there is no such word as "irregardless".

  28. A quick correction regarding the number of differences between the SBLGNT and the Westcott-Hort text: the difference is not 209, as suggested in an earlier post, but in fact 879 (see the Introduction to the edition, p. xii)--ca 12.6% of the total.

  29. Thx Mike for the correction. I should have checked it :) 12.6% is a rather significant step away from WH, but I think we all agree that WH still lingers in the background vis-a-vis RP.

    Let me throw in another thing I ponder sometimes. Were would "we" end up if the original authors wrote more than one version of their own text(s)? If the current textual streams are in some way mutated-by-the-scribes versions of the different revisions by the original authors of their own texts, which one (if any or all) would "we" call the original text? How would "we" know this?

    I know these questions too have been asked before, but the point is that if "we" need to ask them again, how much progress has really been made in NTTC and how far do "we" still have to go?

    Mike knows (and agrees with me) that in orthographical issues we still have a lot of work to do. But before we can get there, more fundamental issues still seem to linger...


    P.S. I think SBLGNT is a much needed improvement over NA

  30. Dr. Robinson: Yes, its obvious you are not an eclecticist. I can't imagine what Mr. Scrivener was thinking. - he has taken correction though, which is encouraging.

    To reply to Bob Relyea, who said:

    "If these are truly alexandrian homoioteleuton and the verses present in the majority of manuscripts, then the comparison with RP would have shown them up. Clearly Mike Holmes gave RP readings a fair hearing, and that each of the 70 places you discuss would have been examined in detail."

    I doubt they got a fair hearing, or were even seriously considered as homoioteleuton, but Mike Holmes will have to defend that claim.

    On the same front, Mr. Scrivener has been a very busy beaver, and has produced some 50 reconstructed physical layouts that would easily generate those errors. These diagrams, using special fonts and typical letter and column layout make the features of each Variation Unit visible at a glance.

    He has chosen only the most probable cases to display, and the result is actually very plausible, even overwhelming. While the odd case might have alternate causes, it is hardly credible that these 50 cases would have homoioteleuton features yet have other origins.

    Homoio in Aleph/B

    The diagrams are outstanding.


  31. Thanks, Dr. Holmes; I had a sense that the 209-number was not the stat I'm looking for, but couldn't quite see why not. If only I had reconsulted your ETC entry of Oct. 26, where you listed the number of differences between SBLGNT and each edition cited in its apparatus: out of 6,928 variant-units, there are . . .

    "SBLGNT—WH: 6,049 agreements, 879 disagreements

    SBLGNT—Treg: 5,701 agreements, 1,227 disagreements

    SBLGNT—NIV: 6,312 agreements 616 disagreements

    SBLGNT—RP: 969 agreements 5,959 disagreements

    Also, the SBLGNT differs from NA27/UBS4 at 542 places, and thus the two will agree at 6386 places."

    So, returning to the question about how many readings are in SBLGNT that are not in Trg, WH, or Byz (but which are in either NIV, NA, or SBLGNT), maybe this will work:

    Out of 6,928 variant-units, in 337 cases (i.e., 879-542) SBLGNT agrees with NA while disagreeing with WH. Plus, there are 56 cases in which SBLGNT disagrees with NA and with WH. So is the answer 393?

    Is there a mathematician in the house?

    Yours in Christ,

    James Snapp, Jr.

  32. The remnants of a primitive h.t. error in the ancestor of the Alexandrian archetype is preserved in Matt 5:39, where the original was:


    Due to h.t. error (PI...PI) the nonsensical RAPI THN resulted.

    RAPI THN was then corrected to RAPIZEI EIS in the Alexandrian archetype now preserved in Aleph B W. Therefore this should be treated as a single variation unit and not two separate ones as now occurs in NA27.

    Jonathan C. Borland

  33. JB:
    "RAPI THN was then corrected to RAPIZEI EIS in the Alexandrian archetype now preserved in Aleph B W."

    Sinaiticus reads RAPIZI EIS THN, corrected to RAPIZI EPI THN.

    Vaticanus and Washingtonius read RAPIZEI EIS THN. But B has SOU later in the verse, and Aleph and W don't. So you come across with more confidence in your reconstruction than I think is warranted.

  34. I was talking about the most primitive error that gave rise to to the others, namely, the accidental omission of SEI EPI from RAPISEI EPI (PI...PI). I don't deny that of course there were various reconstructions and cross-contaminations after the primitive transcription error that affected primarily Alexandrian manuscripts.

    Jonathan C. Borland

  35. Matt 5:13

    I also think h.t. error may have contributed to the variation we see here, with the proposed original


    altered by h.t. error to


    and then corrected to


    Jonathan C. Borland