[If this is considered too off-topic for this list, please ignore this mail. My main point has still something to do with ConTeXt, but I guess this discussion shouldn't be continued on the list.] On 07/22/2012 08:07 PM, Pablo RodrÃguez wrote:
And to explain that a bit: it's not merely "ugly." If all you want is
the printed book, you don't care about the ugliness and simply code this way to get the desired output. However, we are in the 21st century. We should be beyond the point where a critical edition is the printed text, we should think of the typeset result as just one way of representing the logical structure of the edition. With a syntax as the one Wolfgang shows above, it is difficult for most parsers to understand what is meant. Which means most ways of representing such a structure will fail because it's not a consistent logical construct. And yes, as Pablo pointed out, TEI itself hasn't reached a clear conclusion on such points, and they are specialists who have been working on these problems for quite a while... Thomas, many thanks for your reply.
You are the real expert on this topic. I don't really know why there are so many people from TEI working on textual variants, but my guess is that this might be also related to the different needs each of them might face for each kind of texts. Probably the needs to critically edit an ancient Greek or Latin author might differ with the ones for an early modern (or even contemporary) English or German author.
I'm not an expert at all, but I'm trying to put together a research project that would help us make some progress here. But your assumption is basically correct: there are many different philologies with different habits and norms, and TEI has to take that into account.
There is another issue that I would like to discuss. My question is what changes in a critical edition with no page model. I don't mean that critical editions need to be printed (it isn't a paper-based model), but I'm not so sure they can be properly represented without a page model. So, if I'm not wrong, it isn't only a question of data representation, but it is related to the logic of the text structure itself.
Some have characterized the electronic text as infinite, in opposition to a page-based text that by definition finite. XML is a good example of a human-readable text, but this human-readability is relevant because of a prior machine-readability. XML is meaningful and useful for non-coders as source code to generate a human-understandable representation of text.
Footnotes can be displayed not using a page model, because reference is on both the body and the note texts. A hyperlink is the right way to link each other. So, an infinite text is not a problem. The footnote doesn't need to be on the same page (as in a printed book), because there is a way to go to the note and back to the text (as on the physical book).
But linenotes are different. The reference is on the note, but not on the body. The same line can have many linenotes. And the same word or passage can be referenced in more than one apparatus simultaneously. Linenotes work on a page model, because all relevant information is given at a glance. Looking at a page, one knows which words of text passages have relevant information on the apparatus(es). Using the model of the infinite text, there are some issues, unless one reconstructs the page model on a screen model (I mean, that each portion of body text displayed in the screen has also the apparatus(es) included on that same screen). These issues are: which words or text passages have additional information, how to distinguish between references to different apparatuses and how to access to each of these different apparatuses. Maybe marking the text with different features might be a way to distinguish them (colors, underline or a mixture of both). And enabling contextual information is the way to workaround these issues. But I wonder how this is really helpful in practice.
Sorry, but I'm afraid I'm a bit skeptical about this. Probably I'm wrong, but I think it will take some time before having an ePub file containing the electronic version of a critical edition.
I think you misunderstand. I'm absolutely not interested in epub. I'm not arguing against printed output per se (or electronic representations of printed output such as pdf). But it has severe limitations that we need to transcend. I take as an example the text that I'm currently re-reading, Ovid's Metamorphoses in the new edition by R. Tarrant, published in 2004. There are more than 400 complete medieval manuscripts of this text. As things stand now, with a printed edition, no editor can investigate all of them. No editor can record in his apparatus the readings of all the manuscripts he has consulted. No editor can record all the data of the secondary transmission (quotations, allusions, translations etc.) But Tarrant has certainly much more information available than he can include in his edition. Some of it is published in his articles and books, some of it is included in the private collations of manuscripts which he produced in order to produce his edition. When he finishes his career and (I hesitate to write this since he's a friend of mine) his life, most of this will be lost forever. Now I describe the way I would like to see a critical edition of Ovid's Metamorphoses organized: there is a single huge database (as things stand now, most probably an xml file) which contains every reading in every manuscript. This file resides in some sort of version control repository, let's say it uses git. Scholars have write access to this repository and are constantly filling in new information. Some sort of peer review process decides which changes will be pushed to the master repository. Proper markup allows them to fill in anything they consider important: questions of transmission and textual criticism, grammatical, metrical, linguistic points. Every single passage is linked against images of all the manuscripts, so if you're reading this on a device with access to the web, one click will bring up a picture of the corresponding passage in every manuscript you want. If you want to see the text of manuscript X, one click will give you that text. Or the text and apparatus of all 15th-century manuscripts in French libraries. Or all lines displaying a certain metrical structure. (That's why the proper xml structure is so important.) I'm absolutely convinced that having this information in this sort of format will allow us to make huge progress in textual criticism (think of the programs that simulate evolutionary development in biomedicine and apply those to our database!) This does not mean that printed editions will become useless. If an accomplished scholar with years of experience in Ovid wants to publish his edition, he will simply pull the repository at a given moment, he will then decide what will go into his text and his apparatus (i don't think that machines will ever be any better than human beings with regard to critical judgment), and he will then have the result printed. But again: the printed book will only be a representation of the information, it will not be identical to the information. With the proper tools, an editor can decide what this representation should look like, and that's where considerations about several apparatuses, note types, indexes and the organization of information come into play. And that's the area where I see ConTeXt playing an important role because of its powerful potential to process xml, to produce interactive documents, to combine text and illustrations. For the time being, this is a dream, but I'm optimistic we will approach this dream, slowly, but steadily. In order to do this, however, we have to understand that the printed book is just a tool, not the end in itself. All best Thomas