Unity and Scrutiny

Our students presented their editorial work on the letters from Chamisso to de La Foye last Wednesday. My overall impression was extremely positive concerning the encoding. Concerning the approach of the manuscript (especially the transcription and the commentaries), the results varied a lot. I think this reflects the challenge of teaching students several competences in only one class. We taught:

– German and French literary history of the early 19th century;

– Archival and philological approach to handwritten letters (which elements are relevant, what can be considered a reliable information, how do you construct reliable information on the basis of a handwritten letter);

– Comparison between a 19th/early 20th century prints and the original manuscript (levels of censorship);

– XML-encoding after the TEI-P5 Guidelines as we implemented them in our own project guidelines (which are not completely stable and have had to be modified on the fly to deal with specific problems of the corpus in the process).

All of this mostly from scratch. I did not realize at the beginning of the semester that this meant we would not be able to really work on commentaries. We have a special feature that allows some commentaries to appear only in one of both versions. Those that appear in the diplomatical version only are meant to comment on manuscript-relevant phenomena (elements that are corrected or striked through and don’t appear on the reading version). The commentaries for the reading version are meant to give more information on the general context for readers who might not know much about it. We did not get as deep into the differentiation of those two levels as I had expected we would.

Some students worked on only one letter during the whole semester though, so they had the opportunity to focus on very specific aspects. That’s the point that surprised me most. Their scrutiny did not apply to the elucidation of the context, or to understanding the relationships between the persons that are mentioned, or to the knowledge horizon that can be reconstructed, but was mostly focused on the rendering of the manuscript details. I mentioned in an earlier post that we were asked to introduce the category “diagonal”. I granted that and regret it now. It has been implemented not to tag elements that were written by the author, but something that was added by an archivist or by someone who tried to put order in these papers. Which means that 1) it does not contribute to define the text and 2) it would be a huge work to treat that kind of addition in the same way on each and every letter. This is one point where we will have to get rid of some fat.

There has also been some debate concerning the big “6” that appears on the envelope (more seldom a “10” or a “9” or a glyph that some refuse to interpret as a number). This “6” has been generally interpreted as a note made by a postman. I think this is as far as we should get on that question. The fact that the 6 is written all over the rest of the envelope text can be commented, but it will not be rendered in our visualization. Also, the question was asked of how we should deal with single letters that are overwritten within a word. I remember encoding them in my Beausobre and Tieck letters, but there were not so many of them. Apart from the fact that the visualization of that kind of phenomena, when you find them in a larger scale in a text, does not look good (as I said, we will have to reconsider our visualization eventually), this seems to be a huge time investment to be done properly for that corpus.

Two things seem necessary to me now. The first one is to clarify for everyone in the team the relationship between text and image, because we obviously have different notions of it. It felt to me like I spent the whole presentation repeating “You don’t have to render it on the text’s side, because the picture shows it, the reader can see it”. I have to think about a way to define this more clearly, especially because we will be giving a follow-up class on that corpus (to be precise, Anna and Sabine will) in the next semester.

The second thing is that, in order to maintain some unity in the whole edition, it is necessary to think of a space where to explain the specifics of each corpus. Sometimes you can’t really tell if Chamisso writes ss or ß. I have had this problem with Tieck too, but since Tieck is generally coherent in his use of German grammar, I have chosen, when I couldn’t read it properly, to render it with the sign he generally uses in that context. For Chamisso, it is more complicated to define his personal grammar – mostly because his grammar is not coherent. So the decision was to interpret the ambiguous letters (ss/ß or capital letters at the beginning of a word where you can’t really be sure they are capital letters) after the grammatical logic. This should be explained in the small editorial notice that will now accompany the corpora individually, together with a basic bibliography.

Also, I will be opening a section “women writers” over the summer. And we are on our way to a cool webpage name. We already have gotten rid of the numbers and are now called: http://tei.ibi.hu-berlin.de/berliner-intellektuelle/ . hu-berlin.de stands for Humboldt University. ibi stands for Institut für Bibliothekswissenschaft (the institute hosting the server). And tei… well, obviously, the server is conceived to host projects based on the TEI…

Be intuitive! (or not?)

As the complexity of the phenomena we want to edit grows (something that seems to me hard to avoid as new corpora get incorporated), we are confronted with new problems regarding the way we present them on our web interface. We initially aimed at developing a fully intuitive interface and to offer an edition that would be readable and understandable without having to first get acquainted with the specifics through an editorial notice. This dream of editorial intuitiveness seems less and less realistic. The closer we want to get to the materiality of the text – e.g. the closer to critical editing we are coming -, the harder it is to keep the interface intuitive. Anna was certainly right when she said last week that we would have to re-think  the web interface concerning the way it renders aspects of the texts’ materiality. I had started to think in that direction too. But how should that go?

I still think that it is the vocation of our edition to be not only a scholarly one, but also one that should make people who usually do not  have much to do with scholarly editing interested in 18th-19th century-texts and, most of all, in manuscripts and the way the relate to us. So if we want to really work towards a more critical editorial work, I can only think of one solution for the time being, and it is still a fairly abstract one. The only way I can imagine these two aspirations cohabiting in the same edition would be to separate more radically the diplomatic transcript (the transcription true to the letter) and the reading version, especially in the visualization we offer of both of them.

Even with that option in mind, I am still worried about the turn towards critical edition that the latest contributions to the edition have been taking. I have already written it, I don’t think I have enough money to support that kind of detailed work. And I am not sure that this is what digital edition should be about. It is one thing to provide for a clean text. It is another one to want to be able to show on the web interface something that would be similar to a historical-critical print. These concerns have  to do especially with the parts of text that are not to be found in the text body or in line. If you take the different positions of the text elements in the line, we only work with left (which is the default version), center and right. Concerning the text bits we find in the margin, we have only been working with horizontal and vertical sofar and just introduced a “diagonal” category. This is the most extreme I can get, anything that goes beyond that would mean starting measuring angles, as Alex said very rightly on Friday. Beyond the encoding itself, we still have no solution as how to visualize either vertical or diagonal – because contrarily to a book, you can’t really turn around a monitor. And because it is – again – not what a digital edition like ours is there for.

I have been considering trying to see what help could TextGrid be with that issue. When we started the digital edition, we wanted to use TextGrid’s tools in order to connect image and text on a more precise level than just a page. It has turned out not to be so crucial after all, so I had left it at that. But when it comes to connecting text parts that are disseminated at different places of the page with the corresponding transcription, it might come in handy.

Maybe part of the solution I am looking for  is actually to look more closely into the existing tools, hoping not to get lost there. Anna, Sabine and Alex will certainly get some inspiration in Hamburg where they will present our edition at the DHD. I am looking forward to hearing them tell me about it (maybe I can convince at least one of them to start tweeting until then…).