Monday, August 14, 2017

What is a Catena Manuscript and Why should we Care?


In the course of my research on the hexaplaric fragments of Job, I became immersed in its catena tradition. I also became aware that most handbooks and introductions to the Septuagint that mention these MSS did not describe them well, and they usually assumed too much knowledge on the part of the reader, especially the beginner, or worse, the specialist did not understand catena MSS either. A picture is worth a thousand words, and a visual of these MSS allows one to understand commentary on them and what the “C“ symbolizes in a critical text’s apparatus. Knowledge of their material layout aids in understanding their contents.

“Catena” is the Latin word for “chain,” and it will become evident below why these MSS were named as such. The details of the textual tradition of the Job catena need not detain us here. See my article on this topic for details, but one does not need to wade through it to appreciate the content of this post. There are two types of catenae MSS: Marginal and Text.

Marginal Catena

The layout of the marginal catena MS is probably how the name “catena” obtained. The comments of church fathers are chained around the bible text.

Tyrnavos 25 (10th)

A 10th century MS, Tyrnavos 25 represents the earliest text tradition of the catena of Job. This may not mean that it represents the earliest material layout. Our earliest MS artifacts of the Job catena are from the eighth century. They have a similar layout but their text has already been revised and updated. This MS has some of the features of their layout but an earlier text.
Part of Job 2
Prominently displayed in the center of the folio is part of the bible text of Job 2. One can then see how various notes and comments are chained around this text. There is an intricate system of notation next to the left of the bible text with corresponding notation to the notes and comments on that part of the text in the margins. The larger comments at the top, left, and bottom are usually from church fathers (Olympiodorus, Chrysostom, Didymus the Blind etc.). In the right margin, there is a hexaplaric note (unattributed) to Symmachus as well as other exegetical notes (many times anonymous though there is a note to Iulian [the Arian] in the right margin of this folio). In this catena MS, these four elements are usually found: (1) bible text, (2) catena/patristic comments on the text, (3) hexaplaric notes (places where one or more of the Three’s readings contrasted interestingly with the bible text of the Seventy), and (4) anonymous exegetical scholia on the text.

Kopenhagen, Kgl. Bibl., Gamle Kgl. Saml. 6 (10th –11th)

The layout of a marginal catena MS could become quite elaborate as this example shows. 
Part of Job 28
A similar system of notation is used here to indicate on which part of the text is being commented. There are far fewer marginal notes in this MS, though the readings of the Three Jewish revisers were either omitted or sometimes included in the commentary of church fathers (accidentally) or simply included anonymously within the chain.

Text Catena

Although these manuscripts that included bible text with commentary are all usually referred to as catena MSS, it would be a mistake to think they all had the same layout. Many of these MSS had a continuous, linear layout with bible text written first and then immediately below the relevant text the comments were added. Genua, Durazzo-Giustiniani A I 10 (9th–10th) is a good example:

Same part of Job 2 as in Tyrnavos 25 above
The bible text is indented slightly and marked with a straight obelos (at χρονου 10 lines from the top) or lance like index near the middle of the column of text in this folio. There are two lines marked with an ÷. These lines are the beginning of Job’s wife’s extended soliloquy in the Greek tradition, an addition to the Hebrew text probably first noted by Origen. In the right margin is the same anonymous hexaplaric note we observed above that probably should be attributed to Symmachus. In the left margin just below the bible text is ολυμ marking an excerpt from Olympiodorus of Alexandria, who wrote a full length commentary on Job sometime in the first half of the sixth century or late fifth.


So what is a catena MS and why does it matter? Short, marginal notes were added to bible MSS early on, and probably, from the early sixth century (at least for Job), more substantive excerpts of commentaries were added at the relevant places in these MSS. That is, the catenist probably used an existing MS, which already contained the bible text and the shorter scholia, and added the comments from church fathers. Thus these MSS contain (1) a valuable witness to the biblical text in Greek, (2) a fragmentary record of other Greek versions such Aquila, Symmachus, and Theodotion, (3) early exegetical scholia on the text, and (4) excerpts from the more widely read commentaries on a given book. What layer of redaction each of these elements belongs is an interesting and open question.

Alongside the biblical text, they preserve a very important part of its history of interpretation. I have tried to give a basic, elementary description of these important MSS here. I hope the images of them help orient one to their material layout so that one might be able to imagine them when one encounters commentary on them or when the “C“ is spotted in a critical apparatus. For text criticism it is important to realize that catena MSS are very ‘’busy,” and therefore, scribal mistakes between text and margin occurred and often times different readings between commentary and text were preserved side by side as it were.

Credits: The first image was sent to me by Zisis Melissakis, who prepared the digital images of the MS at Tyrnavos. The second and third images were shared with me by Dieter Hagedorn.


  1. Thanks so much for this accessible introduction to catena manuscripts!
    I wonder, though, if the gradual accretion of scholia is a good model for the development of catenae. We don't have manuscript evidence (either OG or NT) for the gradual development of scholia into full-fledged catenae, and the sorts of marginalia that we do find in earlier manuscripts don't include exegetical extracts from earlier authors—with the sole exception of the sort of marginal collation that happens with the Three (and other early revisions of the OG in some books), a quite different phenomenon. Marginal catena appears all at once, not as an evolution from previous sorts of marginalia.
    Our first evidence for the 'chain' of extracts is instead something like what you've called the 'text catena', although the origin of this scholarly practice probably should not be attributed to Procopius of Gaza (rather, probably originates somewhat earlier). The preserved evidence suggests that the move into the margins was a secondary phenomenon in the C7/C8—coincidentally the same point that a number of other sorts of extensive marginal apparatus first appear (not least, the Masorah for the Hebrew Bible and the marginal scholia for various Greek classical authors).
    While it's unlikely that catena as a way of organizing exegetical knowledge begins in the margins of biblical manuscripts, it's quite plausible that the Job catenae start there.

  2. Thanks for this comment, Jeremiah. I wondered whether I should have added that little bit at the end without being able to say more. Let me flesh it out some more.

    I don't think scholia evolved into full fledged catenae, as if we could show snapshot by snapshot how shorter comments grew into larger ones. What I'm suggesting about the oldest Job catena is that the catenist, who compiled the patristic comments sometime after Olympiodorus, probably started with a base layer that contained a bible text with Aristarchian signs, readings from the Three, and probably other anonymous scholia. To this text, he added the patristic excerpts or catenae. For Job, the earliest MSS (both marginal catenae) such as Ra 612 (8C) and Ra 740 (8/9C) already attest the revision or the Γ redaction of the Hagedorns' stemma, and therefore the date of the oldest Job catena must be pushed back before this time. The Hagedorns speculate shortly after Olympiodorus or around the middle of the sixth century.

    Tyrnavos 25 (Ra 788) or its close sister, Ra 250 (Hagedorns' G), is the most conservative representative of the oldest Job catena. Is it suggestive that these MSS which preserve the earliest form of the text may also preserve the earliest layout of the oldest Job catena? I suppose a later scribe could have transformed an earlier text catena into a later marginal catena about a century later and preserved the material conservatively. But it seems equally plausible to me that the scribe of Tyrnavos 25 and the scribes before him are simply copying the text and its layout in front of them.

    I'm a little confused by your last sentence. If the Job catenae began organizing its exegetical information in the margin, then wouldn't that mean that catena also began that way? Or are you suggesting that the Job catena is an outlier or exception within the stream of catenae?

    Thanks for chiming in, Jeremiah. I value your expertise in this area.

    1. Yes, thanks for this blog. I too am struck by the apparent similarities to Venetus A, the 10th century ms of Homer’s Iliad, which is furnished with a wealth of scholia of various types, including marginal, inter-marginal, interlinear, some with lemmata and some without, as well as critical signs, some referencing Aristarchus and some Zenodotus, and obeloi to indicate lines that were “athetized” (not the same as “deleted”). Also some diagrams scattered throughout, including one in book 8 illustrating the relationship between Tartarus, Hades, and the upper levels of the cosmos. Now I am curious to compare the ways in which LXX and Homeric pages are structured. Thanks again. Graeme Bird, Gordon College.

    2. That's very interesting, Graeme. Scribal activity didn't develop or occur in a vacuum. Does Venetus A also collect more extended commentary or comments on the text of Homer? I'm aware of the critical signs matter you referenced. Scholia were probably used as reading aids. But do you think scribes commented on this text as they did say on the text of Job or Genesis?
      Thanks for your comment, Graeme. I'm eager to learn more.

    3. Dear John,
      Thanks so much for your further comments, and apologies for the confusing way I framed the end of my last comment.
      I was attempting to leave open the possibility that while catena *as a way of organizing exegetical knowledge* seems to develop first in the main-text format before moving into the margins as catena marginalis, it might be possible that the Job catena *as a specific manifestation of catena* was created directly in the margins, by analogy to other catenae that had already made the jump from main-text to marginal format.
      I don't think there's evidence for gradual development from scholia and marginalia into full-fledged catena marginalis, but once marginal catena exists, there's no reason why other catenae couldn't be created directly in the margins. Not having done a lot of work on the Job catena, I didn't want to prematurely exclude that possibility in this instance.


    4. Thanks, Jeremiah. So what is the earliest evidence for catenae generally? Does it precede the mid-sixth century? It sounds like you are saying it does.

      Again, you are the expert on catenae in general but I'm a little skeptical that marginal notes of the Three and other exegetical scholia don't in some way become precedent for marginal catenae. But again, I want to see your evidence and fuller argumentation some time.


    5. Gilles Dorival makes quite a strong argument for catena (text-/Breitcatena) to at least the Psalms and Pentateuch as early as the early/mid-fifth century, and my own research tends to confirm that hypothesis, although we have only limited and debatable physical evidence prior to the C7.
      In a sense, marginalia are a precedent for other sorts of marginalia; the Three (+friends) are certainly important. But I'm skeptical that one can draw a more-or-less straight line from isolated marginalia and variant readings to catena as it actually appears in the manuscripts.
      Part of the problem is whether we have any significant number of 'exegetical scholia' prior to the appearance of full-fledged catena marginalis. We *might* have one at Jeremiah 17.27 in Rahlfs 817 (C4), and there are sparse annotations in Rahlfs 928 (=P. Ant. 8 + 210, C3). In neither case do these seem to be extracts or citations, such as we find in the catena tradition, but rather isolated notes made by readers.
      That's it until we get to the C7 with Rahlfs M (=Paris, BN, Coisl. 1) and, in the C8, Rahlfs 406 (=Turin, Bibl. Naz., B. VII. 30)—and both of these are full-fledged catena marginalis.
      Complicating the question somewhat is whether one should include the Euthaliana to the Corpus Paulinum, Acts, and the Catholic Epistles as a predecessor in format to catena marginalis. I'm still thinking about that issue.

    6. Very helpful, Jeremiah. It sounds like the period we can't give an account materially is C4/5-C6/7. Were more marginalia such as scholia being added to MSS during this time? Or are you happier with the conclusion that catenae and scholia were added all at once in the seventh century?

      This is a good dialogue. It's helpful for me to think through this matter again. Thanks for indulging.

    7. I'm happier with the latter conclusion, although there are limitations to the evidence. We have more marginalia of other sorts appearing in those centuries—lectionary marks, secondary paratextual headings, Eusebian apparatus, Euthaliana, variant readings, glosses in another language, etc.—but virtually nothing that looks like the content that characterizes catena.

      For comparison, I think it's helpful to look outside the biblical tradition at other manuscripts from the same period. The movement from collections of running-text excerpts to the margins of a primary text, which I argue is what creates marginal catena, is also almost certainly what happens in the case of the Homeric scholia. (At least initially, although later on, multiple marginal catenae get combined, which makes everything messier.)

    8. And likewise thanks for the dialogue! It's always great to have conversation partners on catena.

    9. Kristin De Troyer6/05/2018 7:26 pm

      Thanks, John, for this very instructive description. I will use, of course with mentioning your name, a couple of your examples. Fernandez Marcos also puts the beginning of them in the 6th century. He follows the description of Deconinck who differentiates between four forms: catenae with columns, marginal catenae with biblical text in the middle of the page as you have them in the first section, text catenae, as the one you have in the section on text catenae, and margin catenae, in which the ms originally only had the biblical text and the margins were later filled up with catenae. I think the catenae of Thomas Aquinas (the catena aurea) is a good example of the first category. Now, in which ones of these do we find clear assignments of authorship? Thanks again for your work! Greetings from Salzburg, Kristin!

  3. It looks like a page of Talmud.