Thursday, February 02, 2012

Mark 1:41 and Ehrman

Just published: Peter J. Williams, 'An Examination of Ehrman’s Case for ὀργισθείς in Mark 1:41', Novum Testamentum 54 (2012) 1-12.

I argue that ὀργισθείς is actually the lectio facilior, and that the internal evidence is strongly in favour of σπλαγχνισθείς (please excuse vertical accents!)

Abstract
Bart D. Ehrman has made one of the most detailed cases for the reading ὀργισθείς in Mark 1:41. This critical examination of his arguments and of the text seeks to demonstrate that his presentation, like that of other scholars, has not adequately explored ways by which an original σπλαγχνισθείς could be changed, largely by accident, into ὀργισθείς. It also highlights some of the methodological tensions in Ehrman’s arguments for ὀργισθείς and argues that advocates of ὀργισθείς need to give a plausible explanation for the felicitous fit of σπλαγχνισθείς as both a contextually appropriate and graphically similar word.

26 Comments:

maurice a robinson said...

Not yet having seen the article, I wonder whether Dr Williams considers the matter to have originated within the Latin with back-translational effect on the Greek of Codex Bezae.

As it stands, it-d has iratus while at least the Vulgate-based Latin reading is misertus preceded by autem.

If in an Old Latin predecessor the M beginning misertus were dropped by haplography from the autem hypothetically preceding, a reasonable sequence of Latin-based error and attempted correction of what appeared to be a nonsense reading would be plain:

AUTEM MISERTUS -> AUTEM ISERTUS -> AUTEM IRATUS

Even if such seems too remote a possibility, the alternative of simple scribal omission of the initial m from misertus would still produce an equivalent scenario. Even more to the point, Wordsworth-White cite one Latin MS that reads miseratus which, assuming a dropped m, could result in iseratus, leading even more naturally to the iratus reading.

Had either possibility occurred in some Latin ancestor, the alteration to iratus in a predecessor of it-d and some other Old Latin MSS (a ff2 r1) would be understandable, even if in paraphrasing, the scribe of it-d ended up replacing an intial Iesus autem with a simple et.

If so, the same scribe could easily have translationally adjusted the Greek of D to match the Latin, and thus no reason need be sought to explain two otherwise widely differing Greek words.

Just my thoughts....

P.J. Williams said...

Thanks. I've certainly considered that in the past, but have preferred the simpler solution of a purely Greek corruption. ὀργισθείς is actually the most common in 'biblical Greek' of all forms ending in ισθεις. It and σπλαγχνισθείς also both begin with the same shape C and share gamma. They are therefore graphically similar.

For the very earliest scribes of the NT σπλαγχνισθείς is likely to have been an unfamiliar term as it is extremely rare outside the NT. It would certainly have been a very difficult reading, especially for any scribe whose first attempt to copy some of the NT began with Mark. Almost certainly they would never have seen the word before.

So what we most likely have is the replacement of an unfamiliar term by a familiar term. Ironically, this is something Ehrman appeals to elsewhere.

I am not suggesting this is deliberate. Rather I make the rather obvious point that to prove that something is the most likely of all explanations is not in any way to show that it is likely. There are numerous routes, some of which are variations on a theme, as to how one word could be corrupted into another. Arguments for deliberate change (such as Ehrman's) must be weighed against all of the variegated scenarios for accidental change. Thus considered deliberate change is relatively unlikely in this instance.

Finally, a significant problem with proposing ὀργισθείς as original is that we have to suppose that at just the very point where the author made one of his most difficult statements, just by sheer chance there happened to be a contextually perfect, graphically similar, word which would not only solve the problem, but also enhance Jesus' character. Another coincidence is that the word which the troubled scribe happened to choose (σπλαγχνισθείς) was that it was extremely rare, but happened to occur in other bits of the NT, so that by chance the corpus with the highest concentration of that word got an even higher concentration of that word.

Too many 'coincidences'.

P.S. Huff said...

"For the very earliest scribes of the NT σπλαγχνισθείς is likely to have been an unfamiliar term as it is extremely rare outside the NT."

How confident are you about that? It seems plausible to think written and oral vocabulary diverged a bit; and even if that is ruled out, the fact remains that only a select portion of Greek literature survives.

I certainly don't pretend to have any expertise in this area. It just seems odd to think Mark, Matthew, and Luke would repeatedly use a word that their audience had difficulty understanding.

Tommy Wasserman said...

I remember that Jeff Cate's paper at the SBL in London last year was very convincing. Pehaps Jeff can share his thoughts.

Timo Flink said...

"AUTEM MISERTUS -> AUTEM ISERTUS -> AUTEM IRATUS"

Pardon me for saying it, Dr Robinson, but this suggestion seems problematic to me. If a scribe would see AUTEM ISERTUS, it stands to reason that (s)he is likely to write AUTEM MISERTUS rather than AUTEM IRATUS, because it is easier to double the 'M', and thus correct the misspelling, than drop out 'SE' and add 'A', which requires creative thinking.

My two cents ...

Jonathan C. Borland said...

Dear Dr. Williams,

I do think your argument is quite ingenious.

A few questions:

1. Is there any evidence that other Greek copyists here or elsewhere ever had a problem with σπλαγχνίζομαι?

2. Is it not a bit coincidental that the only Greek manuscript that is known not to read σπλαγχνισθείς in Mark 1:41 is also one that is known to be corrupted from the Latin in other places?

3. Is it easier to believe that the Latin side of Bezae influenced the Greek than vice versa, considering point #2 above and especially that the Latin side has other important and early supporters (a ff2 r*). In other words, do you think the ancestors of OL a ff2 r* are genetically related to the single ancestor of D? (This was Wettstein's question when arguing Latinization of a single Greek manuscript versus the improbability of many various Latin manuscripts being caused by the reading in a single Greek manuscript.)

Sincerely,

Jonathan C. Borland

James E. Snapp, Jr. said...

I look forward to reading the article. I've juggled the possibilities -- at one time, I thought that ORGISQEIS probably originated in a Syriac translation in which similar words were confused, and that this Syriac rendering influenced the Old Latin; I've also considered the possibility of corruption in Latin that Dr. Robinson has mentioned.

But after further consideration, it seems more cogent to posit that an early Western editor of a Greek copy of Mark removed a term which he thought that readers would have trouble with, and replaced it with a more familiar term which, to him, seemed to fit the context. (See Mk. 7:19 in D for a similar phenomenon.)

I also note that the theory that ORGISQEIS is original would imply an unlikely scenario in which the insertion of SPLANGCHNISQEIS must have happened to occur to copyists in several channels of transmission, even though it is not in harmony with the parallel accounts in Matthew and Luke. (Wouldn't scribes have simply removed the word instead?)

But I don't consider ORGISQEIS and SPLANGCHNISQEIS to be all that graphically similar. Just look at them on the page. The endings are identical but the beginnings are quite distinct. Unless one resorts to positing an exemplar in which the first part of the word was illegible, istm that the replacement must have been intentional, not accidental. The most one might say is that the ending of SPLANGCHNISQEIS might have brought ORGISQEIS to the editor's mind.

Yours in Christ,

James Snapp, Jr.

maurice a robinson said...

Dr Flink's observation certainly could be the case in relation to my first suggestion. However, given that the Old Latin MSS don't actually seem to reproduce the Vulgate AUTEM MISERTUS, my other suggestion about an erroneous dropping of the M with attempted correction of the resultant nonsense would seem more likely (this particularly if the result might relate to the possibility of W-W's ISERATUS with a dropped M).

Regarding Dr Williams' comment: is ORGISQEIS really "the most common in 'biblical Greek' of all forms ending in ισθεις"? Perhaps "most common" is relative, given that forms ending with -ISQEIS are already quite rare.

I see the canonical LXX with only 4 forms ending in -ISQEIS, 2 of these being ORGISQEIS. The apocryphal LXX books have 15 forms ending in -ISQEIS, 3 of these being ORGISQEIS.

In the NT, of the 15 forms ending in -ISQEIS, ORGISQEIS itself occurs only twice.

So at best, when including LXX data, one has only 7 Greek biblical occurrences of ORGISQEIS. Yet in the NT itself, SPLAGCNISQEIS outnumbers ORGISQEIS two to one (4 occurrences to 2); even BAPTISQEIS occurs 3x, which would seem to make these forms "more common" to a NT scribe than ORGISQEIS.

I think a better case could be made from your remaining observation, namely that ORGH and its related forms are far more common in biblical (LXX and NT) Greek than SPLAGCNON and its related forms, which (if one is arguing for a particularly Greek-based alteration rather than a Latin-based) would then tend to attract a single Greek scribe over to a more commonly familiar word (by my electronic count: ORG-/WRG- = 47x NT, 317x LXX, 102x Apoc; SPLAGCN-, ESPLAGC- = 26x NT, 4x LXX, 21x Apoc).

Jeff Cate said...

I think the strength of Pete’s argument is that it’s not coincidental that so many of the variations in NT Mss are so similar in spelling... which I think indicates we must always consider accidental occurrences in transmission that end up making sense.

But like Maurice, I think the variant arose in Latin. And I considered the same argument that Maurice states because it’s striking that the reading IRATUS is always proceeded by ET (instead of AUTEM) in the Old Latin Mss, which seems to make the reading IRATUS appear genealogically related to a Latin exemplar, not a Greek one. And we have tangible evidence of such a change occurring in Codex Usserianius Primus, it-r1 (see the middle of the image http://tinyurl.com/7gwmb9b where IRATUS is changed to MISIRTUS).

But I argued in London this summer that this variant arose in Latin because there is no verbal equivalent of SPLAGXNIZOMAI in Latin (like there is with the Latin noun VISCERA for the Greek noun SPLAGXNA). And that the Latin readings IRATUS (anger) and MISERTUS (compassion) represent scribes struggling to convey the emotion of SPLAGNISQEIS into Latin... and that the Greek side of Bezae ends up with ORGISQEIS due to cross-pollenization with the Latin side of the manuscript.

My apologies for not getting my theory published yet so it can be considered and critiqued. I have a few loose ends I need to tie up using the TLG, but I’ve been swamped with large classes and other research projects. Thanks, Tommy, for your kind words.

P.J. Williams said...

Jim,
Since omicron and sigma both begin with the shape C I would have thought that the beginnings of the two words are similar. We also know that there is a tendency in reading simply to look at the beginning and ending of a word.

Of course, I agree that the variant might have arisen in Latin or (far less likely) in another complicated scenario involving some form of Aramaic. My point is that appeal to a further language is not necessary. The occurrences of σπλαγχνίζομαι in other books of the NT needn't mean that a scribe wouldn't have found the occurrence in Mark unfamiliar.

Based on current evidence it is reasonable to suppose that any scribe not familiar with other parts of the NT would find this word hard to analyse.

If I may be permitted to quote myself:

"Now in biblical Greek there is only a limited number of words ending *ισθεις: there are 20 different forms and only 34 occurrences of all such forms show up in Old Testament, New Testament, and Apocrypha. Of these ὀργισθείς is the most common with five occurrences besides the text in question. Three other forms, including σπλαγχνισθείς have three occurrences each, three have two occurrences each, and the rest only occur once."

Add to that the overwhelming frequency of ὀργισθείς outside Greek related to the Bible.

Mark Goodacre said...

I enjoyed reading the article, Pete. Some great points made. Disappointed to see you ignoring the Farrer theory, though, while mentioning all manner of other theories, especially as consideration of the Farrer theory would actually support your case -- Luke's omission of the Marcan SPLAGXNISQEIS is one of many minor agreements in this passage, part of an 18 word verbatim agreement between Matthew and Luke, one of the longest strings of verbatim agreement anywhere in the Synoptics. You are in good company, though. Non-engagement with the Farrer theory remains endemic in scholarship and it does not look likely to change any time soon.

A couple of typos:

p. 6, "СΠΛΑΓΧΝΙӨЄΙС" is missing a C. The typo is here important because it makes the word look shorter than it is. One of the questions I would have about the alleged mistaking of СΠΛΑΓΧΝΙСӨЄΙС for ОРГӀСӨЄӀС is that the latter looks so much shorter than the latter, ОРГ for СΠΛΑΓΧΝ.

p. 11, "At one moment Jesus is so oriented towards the helping others": presumably this should be "the helping of" or "helping others".

One other thought about the scribe who is alleged to be unfamiliar with СΠΛΑΓΧΝΙСӨЄΙС. You mention that he may have been copying Mark first, but I'd have thought it worth at least pointing explicitly to the fact that if he has previously copied Matthew, he has already written СΠΛΑΓΧΝΙСӨЄΙС twice, in Matt. 18.27 and 20.34.

P.J. Williams said...

Thanks, Mark, for these comments. Sorry for ignoring Farrer. Of course I'm giving Q more probability than I myself would hold to, for the sake of not making my argument depend on a particular synoptic model.

I think the likelihood that a scribe had not copied Matthew before copying Mark is considerable:

At the early phases of transmission gospels would be copied individually.

A four gospel manuscript might be copied by more than one scribe.

As Bezae is the only Greek witness for ὀργισθείς we should note that its gospel order is Matthew, John, Luke, Mark. So Matthew, at least, is not fresh in the memory. But the error is likely to have occurred deep in the textual pre-history. We have no way of knowing whether or not Matthew was known to the scribe.

My argument does not really depend on the scribe having no familiarity with σπλαγχνισθείς. I'm saying that there is a range of natural processes whereby a less-known term can be replaced by a better known one, without appeal to deliberate intention, and that scholars should not assume (as they always do) that σπλαγχνισθείς is a lectio facilior and ὀργισθείς the lectio difficilior. It is just as easy (arguably easier) to reverse these descriptions.

Timo Flink said...

"Even more to the point, Wordsworth-White cite one Latin MS that reads miseratus which, assuming a dropped m, could result in iseratus, leading even more naturally to the iratus reading."

I admit, this is possible, yet I would think it more natural for a scribe to correct MISERATUS to MISERTUS, since it requires dropping only one letter. IRATUS is again more difficult for a scribe to produce. It requires dropping three letters.

Yes, my two cents may seem mechanical, but I think that in the mind of scribe the process of coping a manuscript was just that: a mechanical labour with little thinking involved.

Timo Flink said...

If the reading indeed was ISERATUS, the probability for correcting it to IRATUS is increased. Then it is a case of add 'M' or drop 'SE'. On the basis of the fact that AUTEM has M, I think the scribe is more likely to add the second 'M' to correct the haplography.

All this is to say that I don't think it is so easy to count IRATUS as a creative correct-the-misspelling phenomenon, even though it is possible it did take place.

Simply put, whether we read СΠΛΑΓΧΝΙСӨЄΙС or ОРГӀСӨЄӀС, the change is more likely to have taken place in the Greek manuscripts, which Latin ones just reproduce (IMO), even if ms. D got it from its Latin side d.

Is this now three cents?-)

Stephen C. Carlson said...

If I'm understanding the section on the synoptic relations right, it seems that the argument is that since there is a dispute over Markan priority, one shouldn't appeal to synoptic interrelationships in textual criticism.

I have two questions:

1. Metzger claims that "Markan priority" is a consideration for evaluating intrinsic evidence. Is Metzger wrong? Should we abandon this consideration?

2. Is there really a major dispute over Markan priority? My sense is that ever since the death of William R. Farmer, the Griesbach hypothesis and other forms of Matthean priority have been decidedly on the wane. Why can't we rely on a broad agreement within the field about Markan priority (see point 1)?

Bob Relyea said...

Hi Stephen,

I believe Mark G. said "Farrer" hypothesis, not "Farmer". Farrer assumes Markian priority, but claims Luke used Matthew, thus dispensing with Q.

bob

Stephen C. Carlson said...

Hi Bob,

I'm very familiar with Austin M. Farrer and the theory that bears his name.

I'm also familiar with William R. Farmer and the work he's done in spreading the Griesbach Hypothesis (or as he liked to call it, the Two-Gospel Hypothesis).

These two, with similar names but different ideas, should not be confused.

P.J. Williams said...

Stephen,
I'm not saying that Markan Priority cannot be used as a criterion, merely that its probability must be factored into the ultimate assessment of the likelihood of a conclusion.

So a 70% likely hypothesis building on another 70% likely hypothesis is 49% likely, i.e. more likely wrong than right.

One might assign Markan Priority different levels of probability, be it 99%, 80%, etc. Any textual explanation which assumes Markan Priority and then makes a judgement about priority of a particular reading has to multiply the probability of that judgement by the (assumed) probability of Markan Priority.

Very often in Biblical Studies conclusions which are themselves judged to be 80% likely or more are simply treated as 'solid foundation' on which to build further hypotheses. I suggest that Ehrman should not do this, but of course his practice is very common.

Mark Goodacre said...

Thanks for your comments, Pete. I suppose that what I want to underline is that when one thinks synoptically, it can help one to appreciate the richer possibilities for weighing the textual evidence than when one thinks solely along the restrictive lines of the Two-Source Theory. So when weighing Bart's argument about Matthew and Luke independently agreeing to omit ORGISQEIS, it's worth calling him on his failure to notice that this is just one of so many minor agreements in the verse that it becomes quite implausible that Matthew and Luke are independent. In other words, that strengthens your point that his solution relies on a particular solution to the Synoptic Problem -- he has not considered that the live alternative actually explains the data (at least in the verse in question) better. Matthew and Luke cannot be used as evidence for the text of Mark if it is at least possible (no need to be any more committal than that) that Luke is here copying from Matthew, something that the extended verbatim agreement between the two may suggest. Bart habitually assumes that Matthew and Luke are independent, and that causes him to miss the key evidence against his own hypothesis in this verse.

On the broader issue of Marcan Priority that arises later in this comment thread, it's worth noting also that this pericope, the Leper, has one of the most obvious signs of Matthew's dependence on Mark, the fatigue in 8.4 (cf. my "Fatigue in the Synoptics", NTS 1998). Let's just make sure that when we talk about Marcan Priority we remember that this is not the same thing as the Two-Source Theory. The difficulties in Ehrman's thinking here are related to Matthew's and Luke's independence of one another, not to their dependence on Mark.

Interesting points, by the way, on the issue of a given scribe's familiarity with SPLAGXNIZOMAI, though I wondered whether it would have been pointing out in the article that this word is pretty common in the Synoptics -- 5/4/3, with the participle in question at 2/2/0.

P.J. Williams said...

Mark,
An excellent and obvious point about how Farrer's hypothesis would truly undermine Ehrman's argumentation. It shows a blindspot in my synoptic thinking that I didn't see it!

I think that the frequency of SPLAGXNIZOMAI in the Synoptics is precisely why it might occur to so few scholars (especially ones who began with NT Greek rather than Classical or other Koine) that it would probably have been a rare word at the earliest stage of transmission.

Timo Flink said...

I am just wondering is it possible both variant readings go back to the author of Mark? This is speculative of course, but sometimes I think we forget that he could have written a draft, edit it, release the final text, and keep the draft to himself. And then the draft would have a limited circulation. Has anyone ever done a study, which gathers all of these problematic variations, and try to project them to two different "autographs"? I haven't found such a work (yet), though some essays come close to suggesting this. Then, how would we know if that is the case?

P.J. Williams said...

Timo,
Obviously it's difficult to disprove such an elaborate hypothesis, though it's scarcely necessary to explain the data. I think I would judge the originality of ὀργισθείς (!) to be more probable than a two edition theory.

Also, one would never be able to demonstrate (or even show it to be likely) that both readings came from the same person, and there could be plenty of counterarguments.

Peter M. Head said...

"Has anyone ever done a study, which gathers all of these problematic variations, and try to project them to two different "autographs"? I haven't found such a work (yet), though some essays come close to suggesting this."
This is suggested often enough for Acts (e.g. Strange but with plenty of predecessors); and on occasion for Paul (either on micro level, e.g. at Rom 5.1; or at macro level, e.g. Trobisch on editing original corpus). I'm not familiar with anyone constructing this for Mark. There have been Deutero-Mark theories, some of which take account of textual variation (as well as minor agreements). E.g. O. Linton, ‘Evidences of a Second-Century Revised Edition of St Mark’s Gospel’ NTS 14 (1968), 321-355 (I wasn't myself persuaded that the evidences established a single reviser as Linton proposed).

Timo Flink said...

Thx Peter, I was thinking of Mark :) I am aware of the proposals for Acts and some Pauline epistles. Linton's theory isn't example what I had in mind, though it's close.

P.J.: yes, I would think them unnecessary too in case of Mark (for now :)), but that does not mean they cannot be possibilities. My two cents isn't such an elaborate hypothesis, really. I don't think ancient authors are that much different from their modern counterparts. How many of us write and rewrite our works more than once?-)

I just think that the variation probably goes back to a pre-history of our known manuscript tradition, and I just ask myself a question how far back can it go? I have no answers, not satisfactory ones anyway.

maurice a robinson said...

Writing in the DTS blog cited in the comment section of the other thread ("Is the original New Testament lost?"), Dan Wallace stated:

"Suppose a papyrus had the word 'the Lord' in one verse while all other manuscripts had the word 'Jesus.' New Testament scholars would not adopt, and have not adopted, such a reading as authentic, precisely because we have such abundant evidence for the original wording in other manuscripts."

I just wonder then how this plays out in relation to Dan's advocacy of the reading found in but one Greek MS (D) that happens to read ORGISQEIS in Mk 1:41...

Doc Shaphan said...

Bruce Metzger (1971) summarises the two possibilities but after siding with the "iratus" equivalent mentions 3 important aspects of the SPLANX * reading. Maybe NT scholars should for non-dogmatic but text-critical considerations should renew attention to pay "scriptura scripturae interpres". This (as I have indicated elsewhere is related to the neglected but socio-scientific concept of "Biblical holism".
It may be true that SPLANX* is not a popular word in a synchronistic approach to semantics - as advocated by Prof Jan Louw (cf his 1988 Lexicon] - but the meticulous rabbi that penned Hebrews (with one of the greatest NT vocabs and words difficult to find elsewhere uses SPLANX* with respect to the High Priest I.X.
This strong word ("feelings that wrench at your inner-stomoch"!) is also used in all Synoptics in differing ways of Jesus and by Jesus - 66% are individualistically utilised.
Maybe Jeremias could inquire as it possibly being an "authentic Jesus-word" translated into koine!! haere@xtra.co.nz