I would like to start by putting my response into context. I have a background in Classics but my main research interests lie with the Byzantine and Medieval period, with manuscript and texts created after or around 1100 AD and what they can tell us about both the development of the Greek language and about the written discourse in Greek from this time on. I am currently working at the Grammar of Medieval Greek project (1100-1700) at the University of Cambridge where we try to produce a Grammar of "Medieval Greek" (the vernacular Greek of this period) in book-form.
As a Medievalist in this position I became increasingly aware of the fact that we need to approach Greek language and literature as a continuum without introducing separating lines where historically there were none and ignoring the continuity of written discourse in Greek across the millenia. It is therefore that I am very much intereted editing both Classical, Byzantine, Medieval and Early Modern Greek texts in new ways that allow us to take into account exactly this issue of continuity.
I will structure my response in the form of short statements arising from Gabriel's positioning paper.
I find it important to stress the fact (as I think Gabriel does in his paper) that the notion of "markup" is not something that relies only to the use of digital methods for the edition / presentation of texts. Conventions relating to different types of editions and different frameworks for editions need to be adjusted to the digital medium; in fact I strongly believe that textual criticism needs to be (re)viewed in the light of the possibilities that digital technologies offer to us today.
Is every edition that’s available in machine readable form automatically a “digital critical” edition? The obvious answer is no, in the same way that every conventional edition of a classical or postclassical text is not automatically critical. I would be inclined to pose the same minimal requirements that apply to conventional critical editions to digital ones: use and analysis of all available witnesses, explicit recording of editorial practice in the introduction and the apparatus and critical notes. The digital medium thought offers additional possibilities as far as editorial practices are concerned which I feel that need to be addressed in this workshop.
I would like to make a distinction between a digital representation of a “conventional” critical edition and an edition that’s been “born” digital. In my understanding this distinction is crucial, as it affects the methodological steps that lead both to the constitution of the critical text and its presentation to the reader / end-user of the edition. Since the introduction of digital transcriptions of witnesses and the possibilities of automatic or semiautomatic collation of digital transcriptions the editor has more choices than the traditional “Lachmanian” method. The editor might choose to present one text to his end-users and record textual variants in a critical apparatus or equally decide to make all individual manuscripts available, dynamically linked and aligned to each other together with digital facsimiles of the actual witnesses. The choice of markup goes thus hand in hand with his choice of editorial method, the tools that are available for each method and the desired output. I strongly feel that the notion of a “single critical” text needs to be challenged, at least for texts available in multiple witnesses that exhibit an extensive amount of textual variation.
Texts available in a single witness pose different questions. I generally agree with Gabriel’s description of the issues involved. I can say from my own experience that in such cases the distance between a digital transcription and a critical edition seem to be quite close, also because it is now possible to automatically create (through a processing application) an edition from a richly encoded transcription of a primary witness.
Do we understand the critical apparatus of a digital edition simply as an on-screen visual layout of variants following more or less conventional conventions (Latin terminology and the like) or are we in favour of a non-traditional approach (as described by Gabriel in his paper) thus enabling the user of the edition to interpret the text in ways that are not possible now? I don’t think that there is a definite answer to this question, especially for classical texts; the situation is different for Byzantine and Post-Byzantine texts. I am pretty convinced that there must one “basic” text available to the end-user as a starting point of his exploration of the text (especially in texts with a very complex tradition). I would be in favour of a markup technique for textual variation that allows the full text as presented by the witness to be reconstructed and presented on screen, thus for a more egalitarian approach, although I can understand that this might lead to “nonsense” readings to be presented to the reader.
I agree with Gabriel that the only reasonable approach for an Open Source Critical edition would be to use TEI for all annotation layers applied to a given text. The second choice involving the use of a TEI schema declared in RelaxNG and documented in an ODD file would give more freedom to editors: I can hardly imagine the existence of Guidelines that would cover all different textual types and genres. Having said that I believe that editors should be encouraged to follow guidelines as far as the markup of text critical issues are concerned. I would in other words distinguish between the actual critical text (structure, textual variation, critical reconstruction of the text) and its analysis in the form of commentary, data mining etc.