Thursday, November 27, 2008

SBL Boston, Payne on the "Umlauts" in Vaticanus

This is the last paper summarized in this first NTTC session on Saturday. A paper by Michael Theophilos was announced in the program book, but that paper had been registered twice by accident (dittography?), so it was presented in a different session.

Philip B. Payne, "'Umlauts' Matching the Original Ink of Codex Vaticanus: Do They Mark the Location of Textual Variants?"

Payne began by advertising the facsimile which costs "just $5990" or something like it. (Payne's company, Linguist Software, is a retailer of this edition.)

As many of you know by now, Canart/Payne have earlier identified eleven places where the Umlauts in Codex Vaticanus have the apricot-colored original ink. The news is that the number of such Umlauts have increased with 40 new places! So now we have 51 instances. Another thing Payne wanted to demonstrate was that the scribe who penned them had the goal of systematically going through the codex to compare to another/other MSS. Payne presented some statistical counts which showed a very, very high "chi-square value" (statistics is not my thing, but Stephen Carlson helped me get that right in the comments section). Apparently, Payne had been checking all places against Nestle-Aland to find out if there was variation at those places where Umlauts exist, in order to calculate what the chances are that this reflects textual variation. The problems with this particular procedure (using Nestle-Aland 27) came up in the questions session.

Another important observation related to the "mirror impression umlauts" (see Wieland Willker's description here: These imprints show that the Umlauts were added *after* the text was written, and when there was already a "provisional binding" (so that the wet ink would make a mirror on the opposite page).

Payne made some points in order to prove that their were multiple exemplars for the Umlauts:
(a)Some Umlauts did not affect a mirror-Umlaut (even on the same page where there are such imprints). Therefore they were probably written at another stage, and, that the scribe compared the MS to multiple exemplars several times.
(b) Multiple exemplars also explains how the Umlauts seem to reflect so many different strands. Not one MS can account for all the variants.
(c) It also explains why some pages have more Umlauts than other (there is a page in Romans with a very high number). Perhaps the exemplars then were papyrus exemplars with just Romans ...

Payne thinks this also points to an early date; While the MS was in the scriptorium the scribe(s) had access to many other MSS to compare with ... The more time that elapsed from the writing of the MS, the less likely that new Umlauts would have been added. No other MS has been shown to contain Umlauts indicating textual variants, according to Payne.

Further, he made the suggestion that Umlauts be added in critical apparatuses. Swanson has begun to add Umlaut-data. (I pointed out to him afterwards that I included the information in my own edition of Jude, although only in the textual commentary).

Then Payne went on to some Umlauts in particularly discussed places, namely the PA (John 7:53-8:11) and the passage in 1 Cor 14 (about women). According to Payne, the presence of an Umlaut on the page where the PA is omitted is the earliest evidence for its existence at this location (earlier than Jerome). Payne thinks the evidence is important for its antiquity. Concerning the passage in 1 Cor 14, about the silence of women, the MS includes the verse, but, according to Payne, is the earliest evidence that it was omitted in another MS. [Hence, this evidence can be appealed to by those who suspect the passage is a later addition and/or do not like the passage.]

In sum, Payne proposed that his observations have four key implications:

1) the scribe was aware of variation, and thought that they were important

2) he or she wanted to preserve the most original text

3) the Umlauts provide windows into the text (even for passages where papyri has not survived. Twenty of the fifty-one original-ink Umlauts occur where no variants occur in the papyri.

4) the high proportion of Umlauts where there are known variation (the statistics) proves that the variants have survived in our extant MSS to a very high degree [my remark: it points to the tenacity of the textual tradition.]

5) this proves that the most ancient text has been preserved

Q: "What makes NA27 an appropriate standard to compare with? There are 25 Umlauts in James in Vaticanus. It is 540 lines of text. The NA apparatus has 220 places of variation, more than every third line has variation. If I switch to the Editio Critica Maior in James, it has 761 places of variation, not a single line without variation. If my son marked lines, he would have a very good chance to hit a line where there is variation [100% implied]."

Answer: "It is the variants with significance that matters (and that is what is in the NA), not the minor variants... Payne thought Schmid's calculation was a confirmation (I didn't quite understand why).

Q: "Was there a companion collection where the reader could check the alternative reading?"

A: No, I don't think so. The scribe was a professional calligrapher working in a scriptorium (perhaps the best calligraphy of all extant MSS). There must have been many MSS, especially in a scriptorium in Alexandria. Anyone could compare with MSS in the Scriptorium. A companion volume was not necessary and there are no indications of chapter and verse numbers to facilitate such a use."

Q: Jongkind questioned the reference of the Umlaut in 1 Cor 14:34-35, and suggested that it may refer to the known transposition there (vv. 34-35 follows after 14:40 in D F G and other Western witnesses), it need not indicate an omission of the whole.

A: Payne thinks it is not about transposition, then it would have two sets of Umaluts, but Dirk then asked if there are other examples of this. Payne has not checked. Payne thinks that if it refers to a transposition, the Umlaut would be at v. 40 (since only one Umlaut was used). Or he/she would have added two Umlauts.

Q: "Was it the original scribe who added the Umlauts?"

A: Payne thinks there are some reasons to believe so: It is generally agreed that it has preserved a very good text. We know that in the case of Luke of John (P75) it is ancient. The quality of Vaticanus' text suggests that the scribe copied ancient exemplars or those that had the reputation of being very good texts. It is this type of person that would take time and trouble to do this kind of examination (i.e., to add Umlauts in comparison to multiple exemplars).

Q: "But could they not have been added by someone else, who thought there should be another variant?"

A: Payne answers that then the actual readings would probably be indicated in the margins, but if it was the scribe it would suffice with just the Umlauts.

Q (from a papyrologist): "What is an 'Umlaut'? Why do you use this term?"

A: Payne explained what an Umlaut referred to (textual variation), and that he had coined the term because it referred to a unique sign and therefore not known until recently in papyrology, and because of its convenience without having to make longer explanations like "two dots placed next to each other..." since most will know what an Umlaut usually stands for, i.e., the two dots places over o in some alphabets like the German and Swedish).

My own concluding comment (made in retrospect): I think the statistics Payne referred to, and the method of comparing with Nestle-Aland was unconvincing. Otherwise the paper was good, some of the subsequent answers to questions a bit speculative. The case of 1 Cor 14:34-35 is ambiguous, more data is required. Moreover it would be important to check for Umlauts without MS support (and here we should compare with all available data). A high degree of Umlauts where there is no known variation would probably imply an early date (more likely for the textual data to have been lost).

Update: I have updated the text slightly thanks to Stephen Carlson (see comment). I should also add that it was unfortunate for the presentation that Payne did not show images. That would have been most helpful in this case, especially for those who were unfamiliar with Umlauts, but also in order to see the nature of the ink, the imprints, the position of the Umlauts, etc.


  1. Hi Tommy, just a couple of comments:

    1. I thought the price was $5990 or something like it.

    2. It's a "chi-square" analysis. The validity of it depends on the assumptions in setting up the test. I'd have to see the write-up to know whether the test was set up properly.

    3. I really wish that there was some name other than "umlaut," and it would have been very nice if he called attention to any other MS with this notation.

    4. As for whether the umlauts show awareness of the two major variants that Payne focused on, the umlauts at 1 Cor. 14:33 and John 7:52 could indicate two variants listed in Tischendorf but not in Nestle-Aland. In other words, there may well be a variant at those places, but not the ones in N-A.

  2. Stephen,

    thanks for your comment. I will correct the reference to chi-square, and the price.

    It was very nice to finally meet you there in Boston, and to meet the "real person" behind the name, which I have known for almost a decade.

  3. Tommy: According to Payne, the presence of an Umlaut on the page where the PA is omitted is the earliest evidence for its existence at this location .... Payne thinks the evidence is important for its antiquity.

    Except that I have long maintained (contra Payne) that the supposed PA umlaut at 7:52 actually points to the very real EGHGERTAI/EGEIRETAI variant, and not to the presence of the PA in the comparing MS.

    On the other hand, the (very faint) umlaut that appears following the end of John's gospel seems clearly to indicate the presence of the PA at that location in the comparing MS.Payne either ignores or is unaware of that particular umlaut.

  4. Thanks Tommy for these very helpful notes and comments.

  5. It was nice meeting you too, Tommy. I had to ask Larry Hurtado to identify you when you were asking a question.

  6. Good summary Tommy,

    My impression: I liked Philip's enthusiasm and genuiness. It was an interesting paper with a useful handout listing all 51 of these supposed examples. Lacking any images made it difficult to actually address. I remain fairly unpersuaded that these are as ancient as has been proposed (the use of statistics is generally a sign of forthcoming implausibilities). I would put them in the XVIth century myself.

  7. Oh, and I thought he said $5,999.

  8. For earlier essays by Payne on this subject see

  9. Good to hear that, Peter!
    That the umlauts are that old is against all we know of these times. I, too, remain very skeptical.

    My pet theory is that one day a prospective member of a fraternity decided to create some new parchment by washing an old worn-out codex.
    When the abbot learned what the novice had done he ordered him to carefully retrace everything during the rest of his life.

  10. "No other MS has been shown to contain Umlauts indicating textual variants, according to Payne. . . he had coined the term because it referred to a unique sign and therefore not known until recently in papyrology"

    In the first place, marginal notations of doubtful text are nothing new; we posess the precise information that 276 copies of John contain them in the margin alongside the text of the PA. What isn't often pointed out is that several of the oldest mss, Aleph and B among them, contain similar markings at the insertion-point where the ms OMITS the PA.

    More to the point, however, is that Origen used umlauts in his Hexapla to show words that were in his source text (LXX) but not in the text he used for comparison (the MT). Either Payne is unaware of the plates of Codex Colberto-Sarravianus, umlauts and all, in Metzger's "Manuscripts of the Greek Bible," or he must have been carefully parsing his words, a la Ehrman, to the effect that he doesn't consider these to be text-critical notations.

  11. Thanks for calling attention to Codex Colberto-Sarravianus in Metzger's "Manuscripts of the Greek Bible." However, looking at the plate, I can't see anything like the "umlauts" that Payne has identified in Codex Vaticanus.

  12. Google Books has it online.
    Page 80 unfortunately is currently in lacunate status but page 81 is extant. I now see that many of what looked like interlinear umlauts are actually the overlines of NS (apparently referring to Joshua), but check out the ends of lines 9 and 11 in the first column, and the middle of line 4 in the second column.

  13. Dear Daniel, those are certainly not Umlauts. The horizontal lines at the end of lines 9 and 11 in the first column indicate final nu, which is quite common. The middle of line 4 in column 2 is a hexaplaric obelus sign that Origen used to mark words (or paragraphs) in the LXX which were not represented in the Hebrew.

  14. "hexaplaric obelus sign"

    In other words, an indicator of a textual variant between the Hebrew and Greek--in this case, a combined umlaut and metobelus.

    Might not the umlauts in Vaticanus reference textual differences in an extinct Syriac text?

    There is a website with a plate containing unmistakable umlauts in this codex but I have a feeling it's about to be debunked. It's halfway down this page:

  15. Please note that there are also umlauts and other signs in the OT of Vaticanus. I find it rather strange that nobody is analyzing them.
    They may be crucial for our understanding.

  16. Daniel, we know that Origen used various signs relating to textual criticism, but the question here was whether the Umlaut sign, two parallell horizontal dots, in Vaticanus was unique (and, hence, the new term coined by Payne). You suggested that Origen used Umlauts, but he apparently did not. He used other signs. (In fact I have seen a number of different signs in MSS to mark the presence of textual variants, but that was not the point, was it?)