Resolving the Mystery of GA 2776

I’m in the process of working through all the manuscripts in the Liste tagged as containing Pauline content. One of these mss is GA 2766. The Liste describes the contents of this ms as “aprK.” It dates to the 17th century, is located in the Petraki Monastery in Athens, and is described as containing “c.a. 250” leaves.

However, the portion in the manuscript workspace only contains James through Jude, with no text of Acts or Paul or Revelation. So I dug a little deeper. Pinakes listed only one additional source of information about this ms—Μορφωτικὸ Ἵδρυμα Ἐθνικῆς Τραπέζης. Ἱστορικὸ καὶ Παλαιογραφικὸ Ἀρχεῖο. Μικροφωτογραφήσεις χειρογράφων και αρχείων, Γʹ (1981-1983). While the link to online version provided in Pinakes itself did not work, with a little digging, I was able to find a copy, which provided the following information about our ms (unfortunately the original scan was low resolution): 

This clearly describes the ms as a combination of a printed edition and a manuscript. In her dissertation, Dora Panella provides a description of this edition. “In 1532 Bernardus Donatus published in a single volume the Oecumenian catenae on Acts, the Pauline Epistles and the Catholic Letters together with Arethas’ catena on Revelation.”  A little more digging turned up a nice scan. It was then, comparing this edition to the images on the VMR, that I realized that the handwritten portion of this ms was also copied from this very edition. 

To begin with, the mise-en-page of the ms is remarkably similar to the edition. To give one example, the double diplai in the margin of the ms look exactly like those in edition. However, there are two features of the ms that serve as “smoking guns” that confirm it has been copied from a printed edition, even without a detailed comparison of the text itself: 

First, following the printed edition, the ms (at least in the opening of James) does not have any nomina sacra. They are spelled out, exactly as they are in the printed edition. This is extremely unusual in a Greek ms. In the right circumstances this is itself enough to suggest that a ms was copied from a printed edition. But there is more. 

Second, and perhaps even more significantly, the ms uses catchwords at the bottom of each page, exactly as does the printed edition. As is well known, early printed works put the last word or clause of each page in a separate line at the bottom of each page. These catchwords were then repeated on the start of the next page, which enabled printers to keep the sheets in the proper order. The ms has fewer words per page than the printed edition, so the catch words in the edition do not line up. So the ms has added catch words of its own. Just as in the printed edition (and very unlike standard practice for NT mss), these are given in a line by themselves at the bottom of each page and then repeated at the beginning of the next page.

Taken together, these two factors seem to me to be sufficient evidence to conclude that GA 2776 has not been copied from another ms of the Catena, but from the editio princeps and should therefore be removed from the Liste. 

As a final check, I used Nico Lamme’s TEI collation of Tommy Wasserman’s Jude transcriptions to check to see if GA 2776 had any distinctive readings in Jude that it might share with the printed edition. Since there is a possibility that there are other mss  that have been copied from a printed edition, I set the search (using Nico’s handy find variation script), for all places where GA 2776 has a reading that is found in five or fewer additional witnesses. This search produced three results: 

In Jude 10_4–8 (note that the locations are derived from Lamme’s collation and in some cases differ from Wasserman’s printed edition), there is a singular transposition of δε and μεν that seems to have been corrected by the copyist in scribendo. This is a simple error that doesn’t provide evidence one way or the other.

In Jude 10_20, Wasserman’s transcription has GA 2776 reading φυσικ<d>ο</d>ς with a handful of other mss (this variation is not present in Wasserman’s printed apparatus, but is in his original transcription). In my judgment, the transcription here should probably be φυσικ<d>ω</d>ς. (The scribe copies “ο” with a heavy dot on the top, while the letter is clearly open. However, the center rise of the ω appears to be absent, justifying marking the letter as dubious.) In any case, even if the letter was ο, this sort of variation is rarely genealogically significant. 

In Jude 23_22–24, Wasserman has 2776 joining 1066 1642 in omitting τον before απο. However, an examination of the ms at this place shows that the scribe has written what looks very like his style of τ on top of the α in απο. While it is far from clear, this could possibly be taken as another in scribendo correction. 

While this textual data provides no additional reasons to support the paratextual information provided above, it also provides no challenge to removing GA 2776 from the Liste.