Paolo Monella Post-doc scholarship in Digital Humanities Accademia dei Lincei, Rome 2012

Why are there no comprehensively digital scholarly editions of classical texts?

First created: 1 April 2012. Last review of the online version: 10 September 2012. Word count: 2620. Digital Classicist mailing list discussion on this paper. Delivered on 15 September 2012 at the IV Incontro di Filologia Digitale, Verona, Italy. To be published in the conference's proceedings (expected publication: 2015).

Any comments? Check my contact page.

Slideshow

Abstract

Starting point: we have a problem. Yes, we do

There is a problem, that seems to be going unnoticed. There are no (not one) digital scholarly editions of a classical text with a manuscript-based multi-testimonial tradition. So I'd like to ask: why so?

Of course, a number of terminological implications lie behind that wording:

• By digital scholarly edition, I mean here one based on encoding of all primary sources of a multi-testimonial tradition. This leaves out wonderful projects like TLG and Perseus, which however do not give account of textual variance, and laudable efforts like Musisque Deoque, that encodes some variants (in TEI/XML), but is based on a transcription of an already constituted text, or the HTML editions with variants by Michael Hendry (Curculio web site; all links accessed 1 April 2012 - but this is no prank - and then again on 15 September 2012).
• By "classical" text I mean it in the dirtiest of senses. Yes, that one: Greek and Latin texts ("Classics") of the ancient "golden" age. Those of the "canon" of Classical literatures. This culpably cuts off Nestle-Aland, the Pseudoepigrapha and Galen.

These restrictions exclude all projecs in the Digital Critical Editions of Texts in Greek and Latin page of the Digital Classicist Wiki, except for Homer and the Papyri (and Homer Multitext), a great project that, however, goes close to the ream of Papyrology (for which see below).

Appeal to the reader: please, gentle reader, don't close this browser tab dismissing me as just another haughty classical philologist who looks upon Late Antiquity, Christian and Medieval texts as lowly. This is not my point.

All I'm saying is: digital textual philology, in Classics, has just not taken off, yet. Yet? Will it? As a matter of fact, digital philology has been around since at least the mid-90's, and in other fields it has already produced important outcomes, when it has not even hit the mainstream: think of editions of Medieval and Neotestamentary texts, documentary or enlightened manuscripts, authorial variants of modern and contemporary authors, epigraphy, papyrology.

If, by textual philology, we mean accounting for the textual tradition and the resulting textual variance, we must conclude that, in Classics, it has just not taken off. Great projects like TLG and Perseus do not (currently) implement it, and the projects that experimented on it (Musisque Deoque and M. Hendry's editions) are not based on encoding of primary sources.

Let me mention Peter Robinson's principles of digital scholarly editions, and particularly Proposition 2 (the bold highlighting is mine):

A digital edition should be based on full-text transcription of original texts into electronic form, and this transcription should be based on explicit principles.

Now, if I managed in dragging digital classicists into discomfort (the other categories of readers will most certainly have abandoned me at some previous point), let us go back to the point. No digital scholarly editions of classical texts. And back to the question: why so?

Right now, I can think of one (simple) reason: classicists don't feel they need digital scholarly editions. And I'll argue that they don't feel so because of the peculiar process of "canonisation" (and consequent "normalisation") of the classical corpus of texts throughout the centuries (and still ongoing) [1].

Point 1: We don't have classical digital scholarly editions because classicists just don't feel they need them

Disclaimer: in the next few paragraphs I'll make a lot of assumptions implying that who was so devoted a reader as to arrive until this point is a humanist and - alas! - probably a classicist.

Let's start from the Text/texts (Text/documents) opposition. In what follows, I'll call "Text" the abstract text of the Aeneid and "texts" each of the actual versions of the text born by a single document (papyrus, manuscript, print edition etc.).

Digital scholarly editions are very good at two things:

1. at focussing on documents
2. and at accounting for the "plural texts" – textual variance – that these documents bear.

So scholars who focus on documents and/or on textual variance, for a reason or another, are currently finding digital editions attractive for their own research agenda and experimenting with it, including:

1. Focus on the document
   • Codicologists, interested in the document as a cultural object (e. g. an artistically valuable enlightened manuscripts, or one having a specific historical value)
   • Epigraphists, papyrologists and editors of documentary manuscripts who work on texts mosly born by only one textual source
   • Palaeographers, studying the specific graphical encoding conventions of a document
2. Focus on textual variance
   • "Genetic" editors of modern and contemporary texts, for whom textual variants bear a high literary value
   • Historical linguists, who may study the evolution of language and orthography through "errors" in inscriptions, manuscripts and modern print materials throughout the centuries (though linguists normally do not publish texts, so they must rely on lazy philologists, who unfortunately have a different research agenda; see Toufexis 2010 for the resulting hurdle)

Very simply put, Classical literary studies

1. do not focus on documents (and texts) but on the Text, while they
2. consider the textual variance introduced in medieval times as merely instrumental to the goal of the (asymptotic) reconstruction of the "original" text.

Why so?

Point 2: They don't feel so due to the peculiar nature of classical texts' tradition (“canonisation”)

The readers who have painstakingly made their way through my solecisms will think that I'm not aware that "canonisation", in English, refers to the Church declaring a person a Saint. As a matter of fact, I wasn't sure about it. Other than being a dirty classical philologist, I am also an Italian writing in English: I don't know which of the two sins is less forgivable.

But then I checked on a dictionary and now I am aware of it. However, I still want to play with words and use "canonisation" both in the Christian sense and as the process of transforming a set of literary works into a revered "Canon".

It is the case that most of the classical texts that made it through the centuries until us made it because they became "canonical" [2]. Virgil's case would be too easy to make. But it is generally true that later ages have considered classical texts "classical", i. e. first-class, "canonical": both a linguistic model (for Medieval monks, Renaissance humanists or modern students learning Latin or Greek) and an unparalleled peak of literary/cultural achievement (see the whole of Christian and Medieval literature/culture, plus the many literary/cultural Classicisms).

This "canonisation" of "classical" texts determines some specific features in their textual transmission and therefore the peculiar nature of their textual variance.

1. "Ancient" intentional textual variance (i. e. the original authorial variance and the 'active' variance introduced by editors in the classical antiquity, both of which must have been very wide) has almost completely disappeared, mostly because ancient texts... are ancient, that is very old. It is statistically difficult anyway that an authorial variant of a work survives two millennia of textual tradition. It is even more difficult due to the "normalising" tension that any "canonical" tradition carries about.
2. Later (medieval, modern... and contemporary) intentional variance was strongly discouraged by the "reverence" that scribes and philologists felt (and still feel) to be due to "canonical" text. A medieval doctor might feel allowed to add his own recipe against flu to the technical, practical, "non-canonical" text of Galen, but a medieval monk would not dare add a iota neither to Christian canonical texts (like the Gospels) nor to classical "canonical" texts (like Sophocles' tragedies).
3. "Medieval" textual variance was largely unintentional, as it was originated
   • either (more seldom) by 'pure' distraction errors: a scribe is distracted because lunchtime is approaching, so he writes "dii" instead of "diu"
   • or (more often) by unconscious or conscious normalisations of the text: a scribe might unconsciously write "Deus" (a form more familiar to him, lectio facilior) instead of "diu", or he might consciously change a reading that he considers incorrect in order to "correct" the text (i. e. to restore what he believes to be the "original" form).
   Curiously enough, most of the actual medieval innovations originated by attempts to neutralise alleged previous innovations. But, after all, isn't this the way we, contemporary textual philologists, still introduce new variance into ancient texts all the time?

The resulting framework for classical texts is the following:

• the textual variance that classicists might consider meaningful (e. g. the "ancient" one - see #1 above) has faded out and was not replaced by later "creative" innovations on the text (due to "reverence" for the "classical" text #2 above), while
• the variance actually existing in our textual sources is considered of little meaning and cultural interest by classicists as unintentional and therefore "erroneous" – originated as it was in most cases either by distraction errors or by erroneous attempts to correct alleged previous errors.

This is probably why classical philologists seem not to feel the allure of digital scholarly editions. With reference to the two main strong points of such editions (representation of textual variance and focus on the documents):

1. Textual variance in most classical texts is considered of little interest. The "variant readings" – confined in the apparatus criticus – are considered both a hindrance and a tool in view of the main goal of reconstructing a "good" text. A "good" text; yes, there is a lot of debate about what this might mean, but let's be honest: this is what classical text editors aim to (for peace's sake, I avoided "right" and "original text").
   • Variants are a hindrance, in that each variant reading has at least a 50% chance of being an "error" (90% if the variants are 10!).
   • Variants are a tool, in that through the variants philologists can reconstruct the "orig..." - oops: a "good" text. This is the only reason why, after the constitutio textus, the "wrong" variants are not thrown away, but kept in the recycle bin of the apparatus criticus: text editors must expose the process that led them to their choices, so that erudite readers can falsify their work and possibly make different choices, by "recovering" readings from the apparatus recycle bin.
   In any case, "medieval" variant readings of classical texts are considered as merely instrumental towards the goal of the constitutio textus, that is of the reconstruction of a form of the text that be as close as possible to the original.
2. Documents too (mostly Medieval manuscript and early Renaissance print editions) are of little interest in themselves. They are just as instrumental to the "reconstructional" goal as the textual variants they bear. Needless to say, the early print philologists (humanists like Manutius) often threw manuscripts to the (actual) waste bin after using them. Unfortunately, many such bins have been emptied afterwards and many documents are not at hand any more. Thank God most classicists today have a manuscripts fetish.

Let's take Virgil's Aeneid. Who else? This is the kind of texts I am talking about. There is an insanely vast number of manuscripts of the Aeneid around the world. Digitising the text of all - or at least the most relevant - ancient manuscripts would take a very long while. A classical philologist would ask: what for?

Point 3. The missing link

In the panorama of digital scholarly editing of multi-testimonial text traditions (well, it's not only classical texts, after all), it seems to me that a link is missing:

• We have plenty of editions (and even large corpora) of "Texts" (in the sense stated above), with no reference to textual sources – do you remember a day lately without news on a novel large project in this field?
• We have projects digitising primary sources (papyri, inscriptions, manuscripts, early prints). Not plenty, but - well - relatively many, given the esoteric nature of the field. Not to speak of the mass digitisation of modern and contemporary books (by the millions). So primary sources are increasingly covered.
• What we miss is the link between these two layers: the "Text" and the "documents/texts", that is to say digital scholarly editions of multi-testimonial textual traditions.

This is not only true of classical texts, of course: it is the case for any text "ancient" and "canonical" enough to have seen its original, "meaningful" textual variance erased (authorial versions etc.) and substituted by a new, "unmeaningful" one (made of "errors").

Sure, if we cross the boundary of the Classical area, a few more examples come up, including the Nestle-Aland, the Roman de la Rose and the others mentioned above, and above all the mother of all digital scholarly editions, the Canterbury Tales Project. Still, however, remarkably few (especially compared to the much vaster overall scope that we are now considering).

In other words, we seem to have "Text" editions and "documents" editions, but do not have scholarly digital editions of multi-testimonial textual traditions.

OK, nobody misses them – as I just argued. At least, classicists don't. But imagine that at some point anyone became more interested in the "unmeaningful" variants of those traditions. Say it's because the diachronic linguists took over (did I mention Toufexis 2010? but even synchronic linguists might beg for editions reporting more (possibly ancient) MSS "deviant" forms). Or because advocates of the "fluid text" or of "radiant textuality" (those subversive literary theorists!) crossed the Rubicon and invaded the Ancient World.

Imagine that "unmeaningful" manuscript variants of Virgil's text become "meaningful", for whatever reason. Would our modelling of the relationship between the "text" (lower case) of a document (e. g. a manuscript) and the "Text" (upper case) be ready to support this new generation of "document-oriented" editions?

Prof. Tito Orlandi, Raul Mordenti and others of the "Roman school" of Digital Humanities would argue that the TEI P5 modelling of literary texts and primary sources fails in keeping formally and explicitly distinct (though interrelated) two levels of modelling:

• one layer (A) should model the document's graphical representation of text, mirroring its specific encoding system (alphabet, writing conventions etc.). This should constitute our digital representation of the document's graphical representation of the text;
• a second layer (B) should constitute our digital representation of the Text of that document;

The relationship between Text and document in digital editions is now at the center of the interest of TEI (see the TEI Manuscripts SIG and the work on Documents and Genetic Criticism TEI Style). This is also one of the main topics of an ongoing research I'm conducting at the Accademia dei Lincei, Rome, and the specific point of a talk I delivered at the Digital Classicist 2012 seminar series, so I won't expand on it any longer.

I'm not arguing that our possible flaws in digital modelling are the central reason for the blatant shortage of digital editions of multi-testimonial textual traditions – not only in Classics. I rather believe that the main reason for it lies in the epistemology lying behind the classical editorial activity, which still gravitates around the idea of the reconstruction of one authoritative text. However, while waiting for the scholarly world to be ready for digital editions linking the gap between Text and documents, we should keep working on the complex methodological issues that such editions bring about.

A couple of footnotes

[1] If you're a classicist and you know it, clap your hands. If, instead, you're participating in this canonisation and you don't now it, see Andrea Cozzo, La tribù degli antichisti, paragraph 4.1.

[2] Most texts, sure, not all of them... I thank Eveline Rutten for pointing my attention to the counter-example of Greek lyric poets. Hugh A. Cayless, Ktêma es aiei: Digital Permanence from an Ancient Perspective (in Digital Research in the Study of Classical Antiquity, ed. by G. Bodard and S. Mahony, 2010, p. 139) has an interesting discussion on three different ways for ancient texts to survive until today (Virgil: "canonisation"; Sappho: quotation; Res gestae Divi Augusti: dissemination).