Scholia to the *Iliad*, 2018

These notes are a highly personal perspective on what we know about scholia to the Iliad in the summer of 2018. They are prompted by a unique moment in the history of the Homer Multitext project: we now have a complete first text of the Venetus A manuscript’s Iliad text and scholia, and are beginning to edit further early MSS with extensive scholia to the Iliad. It is an opportune both to take stock of what we have learned, and ask how we might best work with a digital Homer Multitext to improve our understanding of this material. Parts of these notes are simply questions we could consider together.

Starting point

The scholia we see in MSS of the Iliad draw on diverse sources of material, some of it originating in work as early as that of the Alexandrian editors. The scholia have been transmitted to medieval scribes via complicated routes that we still don’t fully understand. Some MSS seem to replicate more or less mechanically (but perhaps never purely mechanically) material from a single archetype. The clearest example of this is pair Venetus B and Upsilon 1.1. Other MSS creatively draw on material from multiple sources to create a unique new document (such as the Venetus A). These notes outline what I think we can say in 2018 about the transmission of scholia in a few of the oldest MSS of the Iliad.

For the Venetus A, my conclusions/hypotheses rest largely on

early systematic work analyzing large parts of the Venetus A scholia. For the interlinear scholia, I largely follow Nikolas Churik; on the relation of main, intermarginal and interior scholia, I depend on work by Melody Wauke and Charlie Schuhfrieder. My own reading of Venetus A scholia may have biased me towards some interpretations, but I feel confident that when their ideas are applied to a complete edition of the Venetus A (and, if possible, when further methods might be developed to test these ideas more thoroughly), their main conclusions will, taken broadly, still stand.

For the Venetus B and Upsilon 1.1, my brief suggestions are purely impressionistic, based only on reading a small number of scholia. They may be completely overturned as we begin to edit those MSS.

Venetus A

The scholia of the Venetus A are organized in distinct zones on the page, intentionally distinguished when each sheet was prepared, prior to writing any text. It would be dangerous to assume that all scholia of the Venetus A can be treated as part of a single set of “the Venetus A schoia” when the scribe has so deliberately set these zones apart.

The exterior scholia are related to the editorial process, as T.W. Allen argued (I think, convincingly) more than a century ago. They most often refer to multiforms in the text of the Iliad and can comment on multiforms noted as alternatives in the main text of the Iliad. (I use the term “multiform,” but the scribes, and Allen, too, would have thought of these as variants, and recorded these readings in an effort to recover a single, best text of the Iliad.) I see no need to assume that the material commented on in the exterior scholia is drawn from other scholia: it might be easier to explain these notes as collation directly against (one or more) other Iliads.

I don’t know how we could be certain about this, but perhaps with a complete Venetus A, we could start to compare its multiforms against the main text of other MSS? (I would hesitate to trust even the fullest critical apparatus for such a comparison.)

The interlinear scholia belong to a tradition of lexical commentary that repeatedly inspired Byzantine scholars to create complete parallel “commentaries” that Nik Churik calls “metaphrases.” They are not exactly translations, but they can intelligily be read continuously, so they can work as a kind of trot/translation. There are at least two good reasons why thinking of them as “translations” can be misleading to us. First (as we see most clearly in the Venetus A) the continuous metaphrase can degrade into glossing on phrases and single words. This is not really how a translation in our modern sense functions: the metaphrases instead are aligning vocabulary in two dialects (contemporary, highly classicizing Byzantine Greek, and the Homeric text). This is actually quite similar to what we see in Greek-Latin bilingual texts of various kind, and probably reflects a very different understanding of how languages work and therefore how different languages relate to each other. Second, Nik and I were able to show that the glosses that are aligned with the Homeric text are closely related to what we find in the Homeric Lexicon of Apollonius the Sophist. I think we can make a good argument that this overlap reflects a shared tradition of glossing/lexical commentary: or, in other words, that the correlations between the glosses of Apollonius and the glosses in the Venteus A cannot be dismissed as notes that any Greek speaker annotating a Homeric text would have produced.

I may need help with a good statistician to take this argument further.

The main, intermarginal and interior scholia derive from sources completely unrelated to the interlinear glossing. The most obvious difference among these zones is that the main scholia are on average far longer than those in the intermarginal or interior zones. The content of material in the three zones is not identical, but the relations of the content in these three sets of scholia are more nuanced than the obviously different interlinear scholia. Charlie and Melody used topic modelling to tease out some important points supporting the idea that we should read these three zones, too, as distinct “documents,” most plausibly understood as derived from distinct sources.

  1. The zones are not simply designed for distinct topics: no topics are unique to a single zone.
  2. At the same time, topics are strongly correlated with zones, and the concentration of topics by zone is very different.
  3. Some topics are explicitly about discussion and critique of earlier scholars. Two particularly important topics were (1) a “verbatim Aristarchan” topic, that (on several independent grounds) could be interpreted as language transmitted essentially verbatim from the Hellenistic commentary of Aristarchus, and (2) an “Aristarchan critique” topic. Signficantly, the “verbatim Aristarchus” topic appears almost exclusively in the main zone.

These observations suggest that the main scholia derive, at least in part, from a source known to us nowhere else that preserved extensive, verbatim content from Aristarchus’ commentary. The main scholia, intermarginal and interior scholia all incorporate material that derives from sources later than Aristarchus. If the simplest intepretation of the layout of the Venetus A is that these zones were planned for material from distinct sources, then the source(s) of the main scholia had already mixed the verbatim Aristarchan scholia with content of other kinds before the time of the Venetus A’s composition.

I’m beginning to think of a number of ways we could explore this further. We will have more options to explore this as we get better control of the morphology of the texts.

Some scholia of the “later critique” topic are duplicated in main and intermarginal zones. This would be consistent with the suggestion that the scribe of the Venetus A found these comments in different sources, and replicated them in each zone. (The suggestion that these duplications show that the intermarginal zones are used for “short-hand” or “summarizing” repetition in the of material the scribe picked out from the main scholia is not tenable since it does not account for the vast majority of the intermarginal/interior content.) The overlap in main and intermarginal/interior content would then be a consequence of earlier mingling of Aristarchan and later material in the VA scribe’s source for the main scholia.

The intermarginal and interior zones are the most similar to each other. Charlie looked at this question from several perspectives, and I need to revisit his work. There are clear individual cases proving that the scribe did not view these zones as equivalent. The fact that they have a great deal of similarity in content then probably means that the Venetus A scribe had two sources at hand with very similar material.

The Venetus B/Upsilon 1.1

This pair of manuscripts gives us the clearest example of material derived from a single archetype. The only prior publication available to us for comparing them is the edition by Hartmut Erbse. Because his goal was to reconstruct a hypothesized original for these notes, it can be difficult or impossible to determine when his composite edition gives the reading of an individual MS: his critical appartus highlights points where Erbse sees disagreement between various MSS, but also silently “corrects” readings in one MS to agree with another. Our diplomatic editions will therefore be the first time we will be able to compare these two manuscripts closely.

The HMT project’s “total editing” approach (in which diplomatic editions are explicitly tied to locations on an image and physical surface, and all intentionally written marks are encoded) allows us to think about the relation of these manuscripts in several ways, including:

  • how do the Iliad texts of the manuscripts relate to each other?
  • how do the scholia texts of the manuscripts relate to each other?
  • how do the layout and design of bifolio spreads in the two manuscripts relate to each other?
  • does the editorial process differ between the two MSS? Do we see differences in the collation and correction of the Iliad text?

Some other MSS

There are several other important manuscripts with scholia that have already been digitized. Here are a few we could consider working on in parallel with the Venetus A, the Venetus B, and the Escporial Upsilon 1.1.

The Escorial, Omega 1.12

A member of the b family along with Venetus B and Upsilon 1.1. While its main scholia no doubt have much in common with those of the “twins,” it includes other material not in either of those MSS. See summary information with further links from the HMT website.

Geneva 44

Although the subject of an extended study by J. Nicole in 1891, this manuscript has been largely ignored in the past century. For Erbse, it belonged to an insignificant branch of his stemmatic reconsturction; he therefore did not include it in his comparative edition of the Scholia Vetera. Its scholia might reflect a somewhat different transmission than either the Venetus A’s varied material or the scholia of the b family. See some brief notes and links from the HMT website,and a fuller catalog description from the e-codices project.

The “Townley Homer” (British Library, Burney 86)

This is an intriguing and doubtless important manuscript. From its prior print publication, it seems to have more material overlapping with the Venetus A than any other MS we know. At the same time, many of its scholia fit neatly in with the b family group.

Topic modelling across the Townley, the “twins,” and the Venetus A might be give us a quick triangulation on different points in the different streams of transmission of the Iliadic scholia.

The Leiden Homer (Vossianus graecus F 64)

This MS was considered to be of high interest by several of the greatest nineteenth-century editors of the Iliad. Like the Geneva 44, it has received almost no attention for the past century, and has largely been a victim of editorial eliminatio. Its home in the Leiden University Library’s manuscript archive is one of the most pleasant places in the world to work.


Website © 2019-2020, the Homer Multitext project. For licensing on image collections, see the Image Archive page.

Powered by Hydejack v8.1.1