Friday, January 14, 2022

An Interview between Kenneth Clark and Maurice Robinson from 1977


The following interview from the 1970s is courtesy of Maurice Robinson shared here with his permission.

During the time I was studying NT textual criticism under Kenneth W. Clark, and just before beginning my doctoral studies in Fort Worth, I asked Clark if he would answer a number of questions in an interview format. He agreed, under condition that such would not be for general publication until long after he had died. As a result, my manually typed transcript of the taped interview (3 May 1977) lay buried among my papers for the past 43 years, and was frankly forgotten until it was rediscovered in a long-unpacked box this Fall. Sufficient time now having passed (Clark died in 1979, at age 81), it seems wise to electronically retype the transcript for full release. 

The approximately two-hour interview occurred at Clark’s home in Durham, North Carolina, on 3 May 1977, when he was 79 years old. He had been quite ill that previous winter, but was in reasonably good health at the point of the interview, although at times talking about other subjects and often repeating previous statements. However, the following transcribed excerpts are interesting and perhaps pertinent to NTTC theory and method even today:

MAR:        It seems that in your earlier articles you basically accepted the Westcott-Hort theory, but that this view had modified as time went by; first, to the status of “questioning” its validity, and most recently of “doubting” its general correctness.


CLARK:   My views really have never changed. I never had been quite convinced of the acceptability of the Westcott-Hort theory; there are too many unproven historical claims, and it relies too much upon subjective factors in its basic reliance upon internal evidence. As you know, I have always been strongly opposed to eclecticism; yet the idea that we are capable of picking and choosing the readings which best suit the context, and are therefore textual critics (whether or not we need utilize the documents which contain that very text) — this is our current “critical” stance, and we are much the worse for it. Again, Westcott-Hort were far more than the eclectics of today: they were document partisans — the nemesis of all poor text-critical theory. Far too attached to Vaticanus as an “infallible” standard.


As to why an increasing criticism of the Westcott-Hort theory seems to develop in my writings, I believe I was just further developing that which I have always held. It is true that much of my critique had been delayed, but that was for an entirely different reason: every new discovery of papyri had to be analyzed, because many of my criticisms would be affected thereby; in fact, many of the building-blocks of the Westcott-Hort theory were severely weakened, without a word from me; the papyri had toppled their theoretical building-blocks. However, had the papyri been known to Westcott and Hort, their text would still have been essentially the same. In fact, had they been in possession of Papyrus 66, the Bodmer MS, and knew nothing whatsoever about Papyrus 75, the close relative of Vaticanus, they would have rejected the evidence of P66 out of hand — and why? Because the text of P66 did not sufficiently parallel B. On the other hand, had they been in possession of P75, without P66, they would have praised it out of hand. And why? Because its text was so like B. You recall that Colwell wrote about Hort having put “blinders on our eyes”? Well, Hort had them on as well!


MAR:        You have stated that we are now working with and are in fact bound to a new “Textus Receptus”, in the form of an Alexandrian text rather than the old TR of the Byzantine type. It has been quite disturbing to Eldon J. Epp and Gordon Fee that you have so characterized our current critical texts as though they were somehow thereby “inferior”. Is that what you intend by your statements, or have Epp and Fee misinterpreted your point?


CLARK:   I should not say we should call the current critical texts “inferior” by any means. However, I have made it quite clear that all current critical texts have not moved far from the Westcott-Hort text, despite all the new discoveries such as the Koridethi Gospels (Θ), the papyri, and the increased studies into the lectionaries, versions, and fathers — none of which had been accomplished in the days of Westcott-Hort. Yet it should be clear to any unprejudiced mind that the Alexandrian texttype — though excellent in many respects — is not and cannot be regarded as the original text of the autograph MSS. Yet what do we see? In every critical edition since Westcott-Hort we have a reproduction, more or less, of their Alexandrian-based text — an exception being the work of A. C. Clark in Acts, who deliberately followed the Western text (if such can in fact be called a “text”), thinking it to be original.


MAR:        In regard to Souter, a while back you mentioned to me something to the effect that you felt Souter’s text to in fact be the closest we currently have available to the autograph text: does this not conflict somewhat with your statement regarding the new “Textus Receptus” of the Alexandrian type, since Souter’s text was basically a reprint of Palmer’s reconstruction of the Greek text presumed to underlie the Revised Version of 1881, and closely followed Westcott-Hort?


CLARK:   Let it be understood that I do not consider Souter an “edition” in the critical sense, since he basically only attempts to present the presumed Greek text underlying the Revision of 1881, and does not pretend to provide for us the “original” text.[1] Yet, I still say Souter’s is “nearer the autograph”, as well as we can now define that statement. And why? Because, although the 1881 Revision Committee did in fact utilize Westcott and Hort’s page proofs, they in no way implicitly followed it. Certainly, many of the more obvious Westcott-Hort readings were followed, primarily in areas of lengthy omission (the Revised Version did present a “shorter” text, of course). However, the Revisers were not blind followers of Westcott-Hort: they themselves possessed sufficient critical acumen to think for themselves on many items, and in so doing, rejected the Westcott-Hort proposals. But when they so rejected Westcott-Hort, they did not gravitate away from the Textus Receptus toward the Western text, but instead retained the TR readings — yet today’s modern critics would call this a “regression” or a “catering” to public opinion as regards the then-forthcoming [ERV] translation. But I think not: I feel they were truly attempting to fulfill the critical task assigned them.


Now, in so saying this, let me add some qualifications to my statement that I feel Souter to be “nearer the original” than other current critical texts. I have noted previously that we seem to be bound to the Alexandrian texttype in all modern critical editions — not the texttype pure, as Westcott-Hort had followed, but still a text sufficiently Alexandrian that would have been recognized as such by, say, any Alexandrian copyist of 300 AD. Now I have also said that we really should not and cannot believe today, with all the new resources available to us, that the Alexandrian texttype can in fact be the autograph text; there are too many factors against it. Therefore, almost any text that moved away from the Alexandrian-pure would thereby become closer to the autographs (in saying so, I intentionally exclude the aberrancies of the Western witnesses, so-called; I do not consider them a “text” or “texttype” insofar as our critical task remains the restoration of the autograph).


Therefore, since Souter does move away from the Alexandrian texttype to a degree — and that greater than, say, Vogels or Merk or Bover — and Nestle[25] and the UBS[3] text are really still basically Alexandrian, and cannot be considered a move “away”, since Souter presents a more decidedly different text than those based upon the Alexandrian witnesses, it is more likely to being closer to the original text than would be any of those which are still too closely bound to Westcott-Hort and their text.


Now, let me further qualify this: when I say “Souter”, I must add the further reservation, “plus his apparatus”; for I must reserve to myself at least the opportunity and privilege to evaluate the readings — and this at times possibly quite different from Souter since, as I have stated, he was not in fact interested in actually determining for himself the “original” text. And this is why I say Souter over any of the other critical texts: between his actual text and a judicious use of his apparatus (which is quite smaller than, say, that of Nestle; but by that the readings are thereby reduced to those of real importance for criticism — and there is nothing wrong with eliminating relatively unimportant readings from the apparatus: Nestle does so continually, and the UBS text almost entirely so, and Westcott-Hort in fact more so), Souter simply has the better apparatus for my purpose. It may be that, say, Vogel’s text may in fact be “closer to the originals” than the text of Souter, but it is not my purpose, nor am I an any position to compare and evaluate the various editors’ texts as a whole — I have enough problems in attempting to determine the true reading in a single variant [unit]!


MAR:        In regard to that: I would like to know precisely how you are able to make a determination of the true reading in any single variant [unit] without having a developed conception of the history of the text. Both you and Colwell have agreed that “Textual criticism without a history of the text is impossible”; and you have even despaired whether we will ever be able to write that history, or whether we are just “pursuing the retreating mirage of the original text.”


CLARK:   Let me note that I have indeed long called for in-depth research which might lead to the determination of the history of the text; in fact, I have made it quite evident. But all my calls have gone unheeded. As for myself, I must work from some principles of textual history; these as such I do not consider absolute, but only as working hypotheses. In fact, most of these have been long held by one scholar or another, though possibly not all together in the way I combine them. I have not been able to make this my life’s work, nor am I so equipped to do so. Thus, what I tentatively hold, and thereby proceed upon, is a “skeleton history”, sufficient to be of some service as a preliminary guide.


Now what are these principles? They include:


(1) the recognition that nearly all variant readings (say 99%) likely had their origin in the period before 200 AD; and whenever a reading is widespread, over many texttypes and MSS, this must be taken as an absolute certainty.


(2) However, singular or near-singular readings are to be immediately eschewed. In 99% of the cases, they are likely the creation of that very scribe, whether intentionally or quite by accident — and readings which have so little support likewise have little claim to originality.


(3) The Western readings (I cannot call them a “text”) have little claim to be a part of the original autographs; their support is too limited, and is almost entirely concentrated within the Latin MSS, the exceptions being but few, Bezae, Coptic MS G67 and other scattered versional readings come to mind, but that is all. However, within the Western readings, there is likely much authentic data, factual, which can be illuminating indeed. Even the additional agrapha of Christ may be authentic, and may be useful for interpretation, but they are not part of the original text just because they might be factual.

(4) The Caesarean texttype, as you have shown [MAR, ThM thesis on Textual Interrelationships in Acts], it has no really unique group readings; this still seems strange to me: it should have them, yet in this case we apparently have the one exception. Anyhow, since it has no unique readings, it appears in fact to be a purely eclectic text, constructed out of a combination of Alexandrian and Byzantine readings. How that pattern was determined, I am at a loss to say; but, because it appears to be thus purely an eclectic text, we should be able to dismiss it from consideration as a texttype in its own right; that is, as regards its readings. In doing so, we are thus left with but two real claimants to originality — and here I refer to individual variant readings, and not to the texttypes as a whole — that is, the Alexandrian and Byzantine readings. I had to qualify my statement, else you might think I felt either the one texttype or the other to be original. In fact, since I have criticized the Alexandrian texttype, you might think I believe the Byzantine texttype (in all its glory) to represent the autograph, which I most certainly do not. But neither do I believe the Alexandrian texttype to be original, as we have noted. To my mind, the true text lies somewhere in between the two; precisely where, I cannot say.


MAR:        But if the true text “lies in between” the Alexandrian and Byzantine texttypes, this can only come about by a combination of their readings. Yet you just rejected the Caesarean texttype precisely because it was composed of a combination of readings from both the Byzantine and Alexandrian MSS. Aren’t you here in contradiction with yourself? And if so, how can you proceed in the evaluation of variants?


CLARK:   The answer, to me, is as simple as it might seem puzzling: I believe the original text today cannot be found in any presently identifiable texttype or recension, as all texttypes are in fact mixed in character, just as we now know that all MSS are mixed. Thus, the Caesarean texttype, by virtue of its being identifiable (whether or not it possesses any “unique group readings”) must represent a later or further stage of development from wherever it began — whether from a text closely approximating the autograph or perhaps approximating the Alexandrian, or perhaps the Byzantine (I feel certain it could not have derived from the Western). In any case, it soon became purely eclectic, being produced by the mixture of Alexandrian and Byzantine readings in their definite pattern (whether the Byzantine and Alexandrian texttypes existed as such at this early time, I cannot say; it is quite possible that both existed in rudimentary form by 200 AD).


In the same vein as my rejection of the Caesarean texttype also lies my rejection of the Alexandrian and Byzantine texttypes: I cannot consider the Alexandrian to be original (as current opinion would have it), for the Alexandrian is a well-developed and basically consistent texttype — but perfection and consistency is a hallmark of result, not of a beginning. And so with the Byzantine texttype: von Soden’s Kr (if it really exists, and this has been questioned [by Lake]) represents the final, polished form of the Byzantine type — but it would be absurd for anyone to claim that the Kr represents the pure autograph. Now, whether von Soden’s K1 was in fact the earliest stage of the Byzantine, as he claimed, or whether the Ka type was the Byzantine archetype (as Lake claimed) is immaterial; but the point is that, in either case, the K1 or Ka type must necessarily represent more closely the autograph. Call it genealogy if you will — though it is not in fact genealogy; that method was rejected even by Hort, since it cannot be implemented in practice — but we must have some way of tracing out the earlier stages of a texttype in order to move back towards the autograph. Thus, the Ka or K1 represents the “proto-Byzantine” text, as such. We should likewise be able to note, say P45 as the “proto-Caesarean” text; and there are certainly “proto-Alexandrian” witnesses to be found. By setting out in order the text of all these proto-witnesses, we will in fact have taken a large step toward the resolution of the original text.


However, let me note that in proposing this method, I do not concur with [Kurt] Aland as such, for he has supposed that the papyri are the Key to the original text, and the papyri alone! But we cannot progress on that basis, for, if we did not have P75, then many readings of Vaticanus would be excluded from consideration. And similarly, if readings of the proto-Byzantine text are not generally found in the papyri (though they in fact are found in P66), this should not automatically exclude them from consideration as quite possibly original, as Aland would have it. We must remember that in the papyri we are dealing with a very limited geographical area (though quite differing forms of text), and Aland’s method provides a very one-sided view of what probably was the true state of affairs.


We simply must have a more balanced methodology than any currently being practiced; else, we will continue to make no real progress in the research into the true history of the text.


MAR:        Let me move to another area: what value do you think lectionary research has into the problem of the history of the text?


CLARK:   It is possible they might hold the key. This because they are still relatively unknown. I must note with regret that Allen P. Wikgren has now left the Chicago Divinity School, and thus their Lectionary Project, begun by Colwell, is for all practical purposes now dead. Likewise, Irving Sparks has left Claremont, and thus the International Greek New Testament project on Luke is stalled. As for the lectionaries, it is to their commendation that they have managed to preserve a stable text form through the centuries, even more so than the Byzantine MSS generally. Indeed it is the most stable texttype we are acquainted with, and quite conservative in preserving many ancient readings, particularly since its readings were  not corrected to the Byzantine standard, as so many of the later minuscules; rather, the Lectionary Text in fact must have been Byzantine in character from its inception, circa the fourth century. The presence of Caesarean readings in some of the pericopes therein must also be attributed to its early date, since these [readings] were neither revised in, nor altered once in; this also indicated that the Caesarean type was current before the lectionary system was developed, as the readings adopted follow an early form of the Caesarean text without question.


MAR:        You have called for a “new system” of determining among variants that would move away from eclecticism as currently practiced. Are you then calling for a purely documentary theory, or would you still call for a judicious use of internal evidence as a determining criterion?


CLARK:   What we need most of all is to get back to a basically documentary theory, using internal evidence solely to decide when either (1) the documentary theory leads us to a reading which on all principles of internal evidence seems not in any way to be original; or (2) when the documentary evidence is hopelessly divided. I would, however, issue a few caveats: (1) No one MS or group of MSS is infallible; thus, a new documentary theory must have some way of explaining this and setting up a methodology which would not proceed blindly in following any one texttype or manuscript; (2) The best supported evidence must have diversity throughout a number of text-groups.


MAR:        Would locations that reflect diversity of origin also help?


CLARK:   Most assuredly it would; however, we have so few MSS which can be definitely localized that the aid they would proffer would be inconsequential.


MAR:        So do you frankly feel that we were ever going to recover the original text? Or do you still consider as you have written that we are “pursuing the retreating mirage of the original text”?


CLARK:   If you mean absolute certainty, then I can most emphatically assure you that the original text will always be such a “mirage”, an unattainable goal. But if you mean relative certainty, well, I think we in fact have that even now, at least, say, 95%. However, in textual criticism, as in all sciences, perfection is always the goal aimed at, though most often never realized in fact. So with the NT text. Mind you, this 95% certainty that I say we now have is also relative: relative to our methodological principles and to our available evidence. I do not think that either any change of method or of evidence henceforth will seriously alter this 95% certainty; however, the 5% remainder, where all is fluid, is where new discoveries and a new methodology must alter our conclusions whenever they occur. It is significant that the pure eclectic school, with its “hunt and peck” methodology, has basically concurred with the 95% certainty portion.


In developing new methods and theories, we presently are left in a quandary: obviously Westcott and Hort’s method must be ruled out if we are to proceed. We cannot first judge among readings, and then follow those MSS which appear to provide the “best” readings, else would we not again return to the polished Alexandrian as the best text? But if we return to judge only the documents themselves, how shall we proceed? Shall we return to Burgon and defend pure majority text? Or shall we continue to follow the MSS which comprise the “best texttype”? Here is where I confess that I am not able to judge or project.


MAR:        Since you bring up Burgon, what do you think about his “seven canons” of criticism for establishing the original text, particularly as a basis for comparison in establishing a documentary theory?


CLARK:   Of Burgon’s seven canons, I react favorably to all but “Number”, as you might imagine. Westcott and Hort’s “Genealogy” — though never put into practice — did successfully put to rest the claims of “Number”; however, I do not then classify myself with those who thereby eliminate the [numerous] Byzantine readings from consideration: some are certainly original, though we tend to view them with skepticism when they are not supported by MSS of other texttypes.


As for the matter of scribal reliability and “character” of witnesses [Burgon’s “Respectability” principle], here is where we are sadly deficient: we blithely accept the readings of a MS without evaluating the peculiar propensities of its scribe. What if he is prone toward omissions — accidental, I mean — by homoioteleuton? Any omission he makes of that type which makes a “better” reading cannot be bandied about at will as an “authentic” witness to such a reading; it may be purely accidental. What then if a scribe is given to italicism [itacism]? When an Aorist is present in such a MS, it may well be in fact a Present that has been italicized [itacized]. We simply are burdened down with the fact that we have too many witnesses to the text to make us willing to devote the needed time for complete examination. Just as MSS need to be grouped and classified, so must each scribe in every MS be classified and evaluated. This practice alone would help us in getting back to the exemplar which lay before the scribe.


MAR:        In regard to “Number”, what then do you make of Hort’s “theoretical presumption” [WH, Introduction, 45] that “a majority of extant documents is more likely to represent a majority of ancestral documents at each stage of transmission than vice versa”, unless one postulated (like Hort) a “recension” that sharply altered the autograph and thus changed the numerical direction of transmission?


CLARK:   First, let me say that I do not believe there were any great recensions that took place which resulted in an official promulgation of that specific recension as authoritative, I tend to concur with Colwell’s thesis that the “recension” as such was the result of a long process, with numerous accretions ever being added to “build up”, as it were, the structure of the “recension”. By this, I am thereby saying that I disagree with the “logical presumption” of Hort that, in lieu of an official “recension” the autograph would be represented by the greater number of descendants — why, that would logically mean that we should expect the Byzantine MSS all to agree 100%!


I still feel, correctly, that Hort’s genealogical “proof” has adequately demonstrated, Q.E.D., that no presumption can be made on the mere basis of “Number” as to the originality of a given text — and this apart from any need to postulate an official “recension”. Moreover, let us presume that you do make mere “Number” the criterion: what you then have is not, as would be presumed, a majority rule of many venerable witnesses, but a mob-rule of a motley crew — some of which may be respectable, but most of which are “lower class”, who force you to accept a reading solely because of their say-so. This is dictation of readings chosen, and not science.


MAR:        What then do you think of the “Majority Text” writings coming from writers such as Hodges or Pickering?


CLARK:   I know of them, but am not all that familiar with their writings. I am more familiar with Edward F. Hills and his KJV Defended volume.


MAR:        Hills’ work was not  promoting the Majority Text, but as its title implies, the KJV position. Hodges, a professor at Dallas Seminary, is currently working on a Majority Text edition of the Greek NT, with a limited apparatus.


CLARK:   If that is the case, I truly welcome such a project. It was long needed, even though the time is too long past for its most useful purpose, that of a more accurate collation base. We now have too many MSS collated against the TR [Stephens 1550 TR, Oxford 1873 reprint] to change in mid-stream to what we should have been using from the beginning. Had we possessed a true Majority Text back in the last [19th] century, its service would have been invaluable in immediately determining that all variants were non-Byzantine; as it is, we now must identify in all collations the variants from the TR which are in fact Byzantine — they should have been in the collating base, and not be variants from it.


MAR:        Since I don’t yet know the format of the Hodges Majority Text edition [not published until 1982], would it be preferable to have the Majority Text as the main text, or the Stephens 1550 as the main, with Majority variants in the margin?


CLARK:   By all means below the text!  This would provide an instant key to check our collations with, so that, even in collations almost a century old, we could instantly identify the Byzantine readings among our [collated] “variants”, and exclude them from our consideration in “[text]typing” a MS.


MAR:        Do you think that a Majority Text edition would be closer to the original text than the TR currently is?


CLARK:   Certainly so, whenever the changes thus introduced in fact reflected the make-up of the autograph; and, as a whole, as I have said, we have about 95% certainty in this regard. However, the Byzantine text — whether as found in the “majority text” or in the TR — is in fact an entity which exists in the fluid 5%, so I cannot really make judgments there until we get back to a basic textual theory to guide us. We are not finding the original text where we ought to find it, so let’s look where we don’t expect it.


MAR:        Subjectively then, do you feel that the original text lies closer to the Alexandrian or Byzantine, or right in the middle?


CLARK:   I cannot say. Incidentally, I must hedge on terminology, since I do not want to call the Byzantine the “majority text” nor the “majority text” the Byzantine; I wish rather to call it a “Byzantine consensus text,” so that I can equally speak of an “Alexandrian consensus text” (another reason why I can rule out any claim for the Western to at all represent a “text”, since it has no “consensus” or “majority” form within those witnesses which make it up).


MAR:        Fee has written critiques of the Majority adherents Hodges and Pickering, have you seen them or have an opinion of his main points?


CLARK:   Fee is a good scholar, and a careful scholar. However, he has always been all too willing to be guided by the Hortian blinders. His excellent work on P75 and B is unequalled, yet to thus proclaim their resultant text to be the original seems far too premature. Obviously, his criticism of Hodges stems directly from his prejudices in this matter.


MAR:        According to Fee, your own position would also come under the same indictment as Hodges, since you have de facto sided against Fee by suggesting a move away from the Alexandrian texttype when determining the original text.


CLARK:   This is true. Yet I believe it is more a question of close-mindedness versus openness. I certainly appreciate Hodges’ work, though I cannot concur with his contention that the “majority text” is in fact the original. Yet I can in no way concur with Fee’s contention that the Alexandrian text must needs be original either. I remain open to the question, whereas it seems all around me are already convinced — I remain to be convinced!


MAR:        [Kurt] Aland has suggested that if the original text were to be recovered, it would appear to be a “mixed” text in relation to our existing witnesses, classifying it in hindsight from our presently identified texttypes. Would you agree with that assessment?


CLARK:   Well, Aland, on the basis of what he has chosen to see in the early papyri, has determined that the original text was so mixed in quality that it bears no relation to any currently known texttype. I tend to think this fallacious. First of all, these early papyri all derive from one locality, Egypt, and would thereby tend to reflect the local traditions in the various places in Egypt where these texts were current. Secondly, the papyri themselves are almost entirely “mixed” between the Alexandrian and Western types — yet on all hands it is generally believed the Byzantine readings were somewhere, somehow in existence during this period before 200 AD. If so, where are they? To be sure, there are some distinctive Byzantine readings in p66 and p45, but hardly enough to warrant the 200 AD claim — unless they were current in other localities besides Egypt; which localities, of course, would not have been conducive to preservation of papyrus. Thereby, I would say that Aland’s position, as it now stands, is in error.


For myself, I do not have any real idea whether or not the original text was “thoroughly mixed”. Of course, all texts are “mixed” to some degree, if you define mixture as the presence in a MS of a given texttype of readings which are generally not found in that texttype but in another well-defined textual entity. Some readings, of course, overlap and are shared by various texttypes, such as the Alexandrian-Western to the exclusion of the Caesarean-Byzantine. These also represent early mixture. Yes, I do think the original text was “mixed”, but not to the great degree Aland postulated.


MAR:        But if by definition there was no actual mixture present in the autograph text, would you be saying that the original would be a precursor of a single particular now-known texttype, or would it be thoroughly mixed from our perspective?


CLARK:   It may well have been the former rather than the latter. As P75 has shown, the Alexandrian is the earliest known texttype to reach full glory, and of course Fee has declared that it must thereby be original. However, it may well be that another texttype was equally current back then, of which no early remains have yet been found. Were such another early texttype  in fact discovered, Fee’s position would need considerable revision — and such may indeed come to light one day.


MAR:        So, what then would happen if a clearly early Byzantine MS papyrus fragment from 200 AD or before were discovered? Or do you think this would be beyond the realm of probability?


CLARK:   No, it is quite possible indeed. After all, Codex Alexandrinus is of the fifth century, and likely could have been copied from a third or even second century exemplar. In fact, unless there actually was a recension (which I tend not to accept), I see no reason why a substantially Byzantine MS as old as p66 could not have existed. Perhaps someday we will discover such. Will it shake up textual theory? I believe so for those who have pinned their hopes on the Alexandrian text as being original, or on those who desire a “mixed” eclectic text such as Fee or Aland — but it would not disturb me; I expect new surprises ever to come forth in textual matters. The small effect it would have on me would be to cause me to re-evaluate a few difficult readings — this is one of the benefits of not having a fully integrated theory of the text to force your hand at every turn: I still long for such, though I fear it will never be fully realized.


[What remains of my typed transcript breaks off at this point, although the interview continued, with other matters also being discussed — MAR].

[1] Alexander Souter, Novum Testamentum Graece. Editio Altera. Oxford: Clarendon, 1947, p. v: (“Prefatio”) “editionis Noui Testamenti Graeci lex haec constituta est, ut is textus, qui Anglae recensioni anno 1881 editae subesse uidetur quique Oxoniensium manibus teritur, denuo sed nitidore forma typis imprimeretur.”


  1. Absolutely fascinating. Thanks very much for sharing this, PG and MAR!

  2. I am struck by Clark's claim that most singulars are the creations of the scribe of the manuscript in question. He said "However, singular or near-singular readings are to be immediately eschewed. In 99% of the cases, they are likely the creation of that very scribe, whether intentionally or quite by accident”. How can this be? Wouldn’t most of the singulars in the Vorlage be reproduced by the scribe? How would he correct singular omissions, for example? Are we to assume that copyists always worked from multiple exemplars? Royse also seems to believe that singulars must betray the habits of the scribe rather than his predecessors. Am I missing something?

    1. No, most singulars would not be reproduced by the scribe. Think of it this way. If every scribe included all the text of his exemplar, plus errors he introduced, the manuscript chain would get more and more error-ridden at every step, withe the ones at the tail end of the chain resembling the mangled phrase at the end of a game of Telephone (aka Chinese Whispers). Istead, what James Snapp has shown in his "Head to Head Contests," a manuscript 800-1000 years old can be much closer to the standard than one 1600-1800 years old, apparently because the transmission stream of the former was of the sort that weeded out errors as they arose.
      Thus, if we find an error that appears in only one manuscript, it can most likely be attributed to that scribe; if on only two closely related mss, probably the scribe of the earlier one. Manuscripts with accumulated errors tended to either become "corrected manuscripts," or were replaced in the copying stream by "authentic manuscripts," both of which we read of as being preferred by ancient scholars.

    2. Stop it, the whole’ Head to Head Contest’ is a sham, Snapp chooses the manuscripts and the text to be compared, If your argument is based on this, it is also false.

    3. Timothy Joseph,
      << Snapp chooses the manuscripts and the text to be compared >>

      Ahem. Granting that I select two manuscripts for comparison in my Head-to-Head contests, and granting that I select the passage, how does either point add up to a "sham"?
      How would you go about making a comparison of the accuracy of the scribes of the texts of two manuscripts at a specific passage?

    4. Daniel, I think you are assuming that copyists made many errors and that they did not tend to accumulate with each generation of copying, while I assume that each copyist made few errors and that they did accumulate. Is that right? If so, you need to explain how a copyist would avoid duplicating the errors in his exemplar.

      If we did not have F(012) you and Clark (and many others it seems) would falsely accuse the scribe of its sibling, G(010) of making many singulars which we know go back to at least to their common ancestor.

      Royse, who is normally very thorough, discusses the origin of singulars (p92-93) but simply assumes his conclusion that they can be attributed to the copyist himself. What is going on? Are text critics reading each other too much and falling victim to group think? Perhaps Peter Head could chime in?

    5. Richard, see Royer’s discussion of the “complex scribe.” He is well aware that not every singular is the work of that manuscript’s scribe.

    6. Thanks, Peter. That is good to know. I have been reading Farnes. If I have understood him correctly his test samples of minuscule 821 has 23 variants that would be singulars if we did not happen to have its vorlage, 0141. However, only 5 of these "singulars" are not also in 0141. This would suggest that the majority of singulars are not the creation of the scribe.

    7. Which would seem to be in keeping with what Royse says about the “complex scribe.”

    8. I think there's a book on purple manuscripts that addresses this as well. You can take N022 O023 and Σ042, all copies of the same exemplar, and look at how many readings in any one of them would be singular if we didn't have the other two. It shows that many of the unique readings in those three manuscripts were already in the exemplar but only survived to this day in the three of them.

    9. Thanks, Elijah. Your work on N022 O023 and Σ042 is ground breaking. You conclude that only about a third of singulars can be attributed to the scribe, if these three manuscripts are representative. I am going to read your thesis with interest.

      All, Elijah's work was discussed on this blog in March 2018 and his thesis is online.

    10. Richard, thank you so much for your kind words. I do have to say that although the published version is mostly unchanged from the unpublished version that is available (apart from that I added a conclusion section, took out two appendices and published them as articles and I might have added an appendix not in the thesis version [can't remember]), I made two significant (in my opinion) mistakes—I didn't take recent history into consideration and identified the wrong city of where N022 was 'discovered' in modern times, and in the concluding chapter, I somehow forgot how maths work and left out one of the manuscripts (I think O023) in my totals in the conclusion chapter. I did fail Calculus 2 at university though the first time I took it, so perhaps forgetting to add a whole manuscript is not all that surprising.

    11. [two significant mistakes that are corrected in the published version]

    12. Elijah might have failed calculus on the first try, but the medical degree that allowed him to remove two appendixes and insert a third came through ok.

  3. I'm rather surprised by the claim that the Caesarean text has nothing original in it. I would think that Jesus Barabbas would be a primary example to the contrary.

  4. As always, another superb article by Dr. Robinson. Thanks!

  5. "The Lectionary Text in fact must have been Byzantine in character from its inception, circa the fourth century."

    That's quite a claim from Clark.

    1. Which part: the type of text or the date, or both?

    2. The assigned date, and the implication: that the lectionaries echo a Byzantine text from the 300s.

    3. The reigning paradigm at the time was clearly that the lectionary text (a) was itself distinct, though of a generally Byzantine character; and (b) had numerous readings of (presumed) earlier date than what appeared among continuous-text Byzantine MSS of the 5th and later centuries.

      This apparently is what drove Clark's statement, as I interpret it.

  6. Professor Robinson, could you comment on what Clark believed constituted a text type? I am not too surprised that he refrained from counting the so-called Western Text as one. But given what he said about that, would applying the same criteria consistently not also undermine both the Alexandrian and Caesarean text types (both of which this interview makes apparent that he believed were deserving of the label, in contrast with the Western)?

    I also am a bit confused about Clark's position. It seems like he is still stuck with an eclectic text when all is said and done. Am I missing something here?

  7. Clark would consider a texttype as a particular cluster of related MSS which normally would have some "distinctive" readings peculiar to most MSS related to that cluster at any given point (the terminology "cluster" was not then being used, so far as I know, but that would be what he was describing as opposed to any product of a formal recension).

    Clark was still eclectic in one sense, in that he would not advocate following a single texttype throughout; but he did not lean to reasoned or thoroughgoing eclecticism in individual variant units if such meant disregarding the overall pattern of readings involved in such (at least that's how I understood him at the time).

  8. "(2) However, singular or near-singular readings are to be immediately eschewed...and readings which have so little support likewise have little claim to originality."

    And thus Burgon's second "Note of Truth" — "2. Consent of Witnesses, or Number" "Trad. Txt." p.29

    I believe Clark probably agreed more with Burgon in regards to "Number" than he may have realized. For even Hort acknowledged/believed that the combination of B/א was weightier than either B or א (would be) without the "consent" of the other (generally speaking). So we can like it or lump it, but "Number" certainly stands as a "Note if Truth." Or else, why do men on every front uniformly reject the poorly attested readings found within the TR corpus? e.g. Luke 2:22, Acts 9:5-6, Eph. 3:9, I Jo. 5:7-8, Rev. 16:5 etc.

  9. Dr. Robinson, in your judgement, do you believe that Clark was being loose, casual and/or utilizing hyperbole (in regards to his "99%") when he states(?):

    "(1) the recognition that nearly all variant readings (say 99%) likely had their origin in the period before 200 AD; and whenever a reading is widespread, over many texttypes and MSS, this must be taken as an absolute certainty."

    This statement would seem to clash with an earlier statement of his:

    "It is of particular interest to realize that many textual alterations first appeared in Byzantine copies of the NT. It has been widely held and often repeated that the important alteration of text occurred before A. D. 200, but this view is considerably modified by the panoramic research of the IGNT project. It is true that every additional copy collated yields new readings of exegetical consequence."

    It seems to me that Clark didn't necessarily have every single variant reading, (especially late non-viable readings,) in mind when explaining principle "(1)." Perhaps you can shed some more light on this point. Thanks!

  10. I'm just a simple layman, but I can't find the text of Alexander Souter. Might any of y'all know if his GNT is available online?