Monday, July 30, 2018

David Smith: Why GA 1411 Should Not be GA 1411: On Classifying Catenae Manuscripts

The following is a guest post from David C Smith. David and I worked together a few years ago in Athens with CSNTM. He is completing a ThM at Dallas Theological Seminary and is a Deacon in the Anglican Church in North America. His masters thesis, entitled “A Study of the Text and Paratext of the Catena on Luke in GA 1441,” should be completed this summer. Here is some of the fruit of that work.


The question of Catenae Manuscript classification has been raised by some as the details and purpose for composition have been more clearly understood. I wish to contribute some data to the conversation that has come from my work on GA 1411, a relatively unexplored and unanalyzed MS from the National Library of Greece. I was fortunate enough to digitally preserve this artifact with my good friend Andrew Bobo as we worked for CSNTM in the summer of 2015.

The purpose and use of catenae manuscripts has been explained more fully in the NTTC world by scholars such as William Lamb, Hugh Houghton, and David Parker. Lamb in particular has argued that these manuscripts were not primarily written as witnesses to the New Testament, but were written as educational textbooks in the medieval Byzantine scholastic context. Much of his argument is based on situating these manuscripts in their historical context, and in turn showing how this understanding of the compositional purpose of catenae manuscripts explains the high amount of variance in not only their biblical text but in their patristic commentary. In addition, this explanation of their purpose calls into question their use as witnesses of the New Testament alongside other continuous-text manuscripts, which are clearly written to transmit the NT text.

For many reasons, not much work has been done on the Greek catenae manuscripts as artifacts to see if the actual details of the manuscripts support this conclusion from Lamb and others. That is, if we were to examine the textual and paratextual details of a given catenae manuscript, would the phenomena of the artifact be best explained by the theory that they are educational textbooks/anthologies?

My master’s thesis at Dallas Theological Seminary is a study of GA 1411 to attempt exactly this. I am trying to observe the details and phenomena of this manuscript to judge whether or not Lamb’s explanation best fits.

Having done some work on transcribing the manuscript I want to share some early data that demonstrates what I am trying to do. I hope this accomplishes two things. First, I think it will give an idea of what the outcome of the study will be, since the data I give below seem to be consistent throughout the MS. Second, I hope this provides a model as scholars work to think more proactively about categorization (as I know recently posted Postdoc opportunities as Birmingham are aimed at doing).


GA 1411 contains the Catenae on John and Luke and is an Alternating Catena. (Note: Parker suggests “Frame” and “Alternating Catena” as the names for the two types, whereas Meade suggests “Marginal” and “Text.” I will use Parker’s terminology.) I am specifically studying the Catena on Luke 1–3 in GA 1411, starting on page 93b.

The first graphic shows how much biblical text is present in the manuscript for Luke 3:1–5. The Robinson-Pierpont Byzantine Textform is the base-text.

text = Majuscule Biblical Text Present in GA1411
text = Minuscule Biblical Text Present in GA1411
text = Variant Present in GA1411

Ref. GA 1411 Text
Luke 3:1 Εν ετει δε1 πεντεκαιδεκατω της1 ηγεμονιας Τιβεριου Καισαρος, ηγεμονευοντος Ποντιου Πιλατου της2 Ιουδαιας, και1 τετραρχουντος1 της3 Γαλιλαιας Ηρωδου, Φιλιππου δε2 του αδελφου αυτου τετραρχουντος2 της4 Ιτουραιας και2 Τραχωνιτιδος χωρας, και3 Λυσανιου της5 Αβιληνης τετραρχουντος3,
Luke 3:2 επι1 αρχιερεων Αννα και Καιαφα, εγενετο ρημα θεου επι2 Ιωαννην τον Ζαχαριου υιον εν τη ερημω.
Luke 3:3 Και ηλθεν εις1 πασαν την περιχωρον του Ιορδανου, κηρυσσων βαπτισμα μετανοιας εις2 αφεσιν αμαρτιων·
Luke 3:4 ως γεγραπται εν1 βιβλω λογων Ησαιου του προφητου, λεγοντος, Φωνη βοωντος εν2 τη ερημω, Ετοιμασατε την οδον κυριου· ευθειας ποιειτε τας τριβους αυτου.
Luke 3:5 Πασα φαραγξ πληρωθησεται, και1 παν ορος και2 βουνος ταπεινωθησεται· και3 εσται τα σκολια εις1 ευθειαν, και4 αι τραχειαι εις2 οδους λειας·

At a glance, you can see that much of the text from these verses in Luke’s gospel are missing. This is not an anomaly. It is normative for GA 1411 that much of the biblical text of Luke is missing. It cannot be rightly called continuous-text. If we were to look at a larger sample of text, you would see that GA 1411 omits entire pericopes from Luke’s first 3 chapters. Focusing on these five verses, what explains the large gaps?

In the shortened occurrences of v. 1 and v. 5. In both cases, the highlighted portions are written and end with the phrase “και τα εξης,” (line 17 below) which means “and so forth.” Here, the scribe has clearly and purposefully left out the remainder of the verse and opted for a filler phrase. Presumably, this saves both space and time, allowing the reader to get to the meat of the manuscript, the patristic commentary.

γραφαι = biblical text in majuscule script
[γραφαι] = biblical text in minuscule script
← = hanging indent
v.1 = Scripture reference

104a Column 1

Line Verse Text
6 v.4b παροντος του κυ̅ [ετοιμα]
v.4b [σατε] φησι [την οδον αυ-]
v.4b [του] τουτεστιν ευτρεπισθη-
-τε προς παραδοχην
10 ων αν βουλοιτο νομοθετειν
των του νομου σκιων α-
v.4c -φισταμηενοι ευθειας
v.4c δε ποιει τας τριβους του θυ̅
ο δια δικαιοςυνης αυτας ο-
15 -δευων σκολια γαρ η κακια
v.5a ← το δε πασα φαραξ πληρωθη
v.5a σεται και τα εξης ως προς
αντιθεςιν τινος αφιβαλ-
λοντος η προφητεια
20 αποκρινεται επειδη γαρ
v.4c ειπεν [ετοιμαςατε την]
v.4c οδον κυ̅ ευθειας ποει-
v.4c ται τας τριβους αυτου ε-
ναντης δε η τραχεια 

There are two phenomena that stand out from this transcription. The first is the double occurrence of text from v. 4c with v. 5 in the middle. The second is the alternating of majuscule and minuscule text.

First, let’s look at the double occurrence of v. 4c. What could explain this? As discussed above, the “τα εχης” of v. 5 in line 17 stands in for the rest of the verse. This means the following text is commentary on that whole verse. Additionally, line 16 is set off by an ekthesis, or hanging indent, to show it is the start of a new section of commentary on a new passage of Scripture, which is a typical feature for alternating catenae. This then makes it clear that the text of v. 4c in lines 21–23 are not functioning as scriptural references preceding commentary but are an allusion within the commentary being copied by our scribe. Likewise, the text of v. 4b and v. 4c in lines 6-8 and 12–13, respectively, are also allusion coming at the end of a section of commentary by Cyril.  Importantly, these two fragments of v. 4 are the only text of Luke 3:4 in our manuscript, leaving the first half omitted. This means GA1411’s witness to Luke 3:4 is only scriptural allusion in patristic commentary and is fragmentary.

Notice also the variants in lines 13 and 23. Line 13 reads, “Make straight the paths of God,” whereas line 23 says, “make straight his paths,” not to mention the addition of δε of L13. The former follows the LXX in Isaiah 40:3, except that it omits the pronoun ἠμῶν at the end. The latter follows the text of Luke as we believe it originally existed. So here we have two versions of a verse in the same column, one that aligns with the LXX and the other with the NA28.

Our second observation is related to the script. As a sort of “control,” let’s look at the way verse 5, and a new section of commentary is presented. At Line 16, v. 5a (shown below) is set off with a hanging indent and is very clearly set apart as biblical text with uncial script and a phrase break, even though the form of the script written immediately before and after the clause marker is debatable.

Lines 16–19
Now contrast the presentation of v. 4c. In both occurrences of v. 4c, the text is a mixture of majuscule text and ligatures characteristic of the scribe’s cursive script, making it unclear if the scribe is setting this off as quoted scripture or not. Neither instance of the text is clearly superior to the other, as far as whether the scribe considered it the primary occurrence of v. 4c. See the images below.

Lines 12–13

Lines 21–24
In the first occurrence on line 12, the scribe begins in an informal majuscule and then sort of resorts back to minuscule. In the second occurrence beginning in line 21, he starts in minuscule and then quickly jumps to majuscule, apparently once he realizes this is scriptural allusion. But then, he begins to write εναντης in majuscule and shifts back to a cursive script in the middle of the word. The switch is abrupt. Either he is being imprecise or he realized in the middle of the word that it was not found in Luke 3:4.

It is unclear then, since neither occurrence of v. 4c bears the distinguishing marks of v. 5a, which one (if either) is the primary quotation. This results in two dilemmas. First, how would one transcribe Luke 3:4 from GA1411? Which version would be the text witnessed therein? Second, should one even say that Luke 3:4 is witnessed to—copied, in the traditional sense—in GA1411? A case could be made that Luke 3:4 is only alluded to in this manuscript, via the commentary, and not actually being quoted in either case by the scribe or the original commentator, as stated above.

Some Conclusions

If William Lamb in his recent work is correct, catenae MSS have an intended use different than faithfully transmitting the biblical text. Their use as interpretive training manuals required a particular process of composition that categorically jeopardized their transmission of the biblical text.

It seems to me that GA1411 bears the marks of exactly that sort of document: one written for passing on the tradition of interpretation and training new monks and priests in that tradition. It was not written primarily to faithfully pass on the text of the New Testament. This should call its classification within the Gregory Aland system into question. The question isn’t about the quality of the scribe, the character of the script, or the quality of the manuscript as such, but about the intentions of the scribe.

And if the purpose for composing GA 1411 jeopardizes the biblical text transmitted in it, shouldn’t we categorize it differently?


  1. Great work, David! This seems like a fruitful area of study.

    I remember back when I was trying to locate a passage in Luke in this manuscript, it was a real challenge. I'm pretty sure that the passage I was looking for (Luke 9:35) occurred on folio 162v, but since the quote I wanted was isolated within the commentary rather than in a distinct block of continuous text, I ran into another problem: the commentary often seems to quote from a more Matthean text than a Lukan one, and it does so here (the commentary mentions parts of the Matthean reading ο αγαπητος εν ω ευδοκησα at different points) so it wasn't clear to me if the text being cited was that of Luke or Matthew.

    The bottom line here is that I agree with your point on reclassification for another reason: manuscripts that primarily contain commentary without text can be very ambiguous in parallel passages.

    1. Thanks Joey! What project got you looking at this manuscript? I'm glad the images were used by more than one person.

      And yes, you are right. The commentary/sermons being passed along in these manuscripts is very frequently referencing parallel passages in the synoptics. It looks like a researcher (Gregory?) penciled in the parallel passage references in GA1411 in some of those later pages, so it is quite clear this is happening and the "Sermons on Luke" heading from page 93b is probably not a very good one. Just the other day I came across an piece of exposition on a verse from John 1, in the midst of commentary on Luke 1. It's clear at least to me what this manuscript is and what it is not, but I think it's just about doing the work to demonstrate that with the details of the artifact itself.

      What is important to note is the variance of style even within Catenae MSS. It's going to take work to sift them out. Some truly do pass down scripture in the traditional sense and perhaps should not be recategorized. Some others, like GA1411, do not. I hope the folks at Birmingham sort all that out for us. :)

    2. I've been working on collating every manuscript I can access at a single variant, so I've gotten to encounter several unusual manuscripts. For instance, I've noted the following manuscripts as other potential "commentaries with text": 055 (Titus of Bostra mentions the pericope in Luke 9:28-36 on 205v, but does not include the text), 599 (the passage should occur on 398r, but the text is only cited in pieces), 1027 (it's possible that the entire pericope occurs in this one, but I haven't found it yet because the commentary covers passages from different gospels completely out of order), 1366 (Titus of Bostra mentions the pericope on 101v, but does not include the text), 1419 (another catena from the National Library of Greece; it alludes to the pericope on 131v, but the citations of the text skip to Luke 10 on 132r), 1507 (another one where the scriptural citations are out of order), 1814 (part of the pericope is covered on 289r, but not all of it), 1821 (the citations of the text skip from Luke 9:21-22 on folio 136r to Luke 10:21 on folio 137r), 2453 (Titus of Bostra mentions the pericope on 239v, but does not include the text), and 2456 (I haven't been able to identify the text in this one at all). All of these might be candidates for recategorization.

    3. Based on what you're saying, I would imagine making the recategorization decisions on these manuscripts will not be as involved as I had originally thought. I think the powers that be will be able to identify a few characteristic phenomena they can just look for in each MS and find relatively quickly.

      I'm not familiar with 1419 from the NLG. Do you mean 1412?

    4. No, it's 1419; the INTF VMR has microfilm images, but CSNTM has much nicer color photographs ( If you look at folio 131v, you can see a reference to the pericope περι της μεταμορφωσε(ως) in faded gold ink, but none of this passage is cited. One page later, on 132r, the commentary quotes Luke 10:1-3. The closest preceding quotation is from Matt 21:10-11 on 130v.

    5. Ah. Rob and Jacob photographed that one. I didn't realize there was another MS with Titus in it at NLG. Do you happen to know what page the Catena on Luke begins on?

    6. The closest content I could find to the beginning of Luke is on 56r, where the commentary alludes to several bits from Luke 1 and offers the Greek meanings of several proper names. It's possible that the very beginning of the commentary is lost.

      Before this page, the manuscript no longer marks quotes from the text in an easily-identifiable manner. Interestingly, the first pages seem to cover the beginning of Genesis.

  2. Typo 1441 for 1411 in first paragraph?

  3. DS: “It is normative for GA 1411 that much of the biblical text . . . is missing. It cannot be rightly called continuous-text”

    My 20-year-old notes regarding 1411 in the PA region of John attest to the same:

    "MS 1411 is commentary with only partial text interspersed. The text per se is often isolated or choppy. This should not be considered a continuous-text MS. The only extant text portions are 7:51 προτερον; 7:52 μη και...εγηγερται (in three separate segments); 8:12 ειπεν αυτοις εγω...κοσμου. Nothing more in the collation area."

  4. Speaking of CSNTM...they just announced where they have traveled this summer and what they have digitized: