TEI Manuscript SIG Meeting Report 01
The first meeting of the Manuscript Special Interest Group held in Nancy was extremely productive. While the group did not come to any conclusions regarding how encode manuscripts according to the TEI Guidelines, many issues were raised. Participants found that many of the difficulties in encoding manuscripts using the TEI were common across disciples, languages, genres and time periods. A summary of the meeting is below.
Pat Bart, David Birnbaum, Ivan Boserup, Anh Bui, M.J. Driscoll, Patrick Durusau, Ahn Jaewon, Jindrich Marek, Alberta Nodar, Elena Pierazzo, Dorothy Porter, Gautier Poupeau, Peter Robinson, Susan Schreibman, Harold Short, Mirko Tavosanis, Adriaan van der Weel, Edward Vanhoutte, John A. Walsh.
Summary Report: Six Issues
- Modern manuscripts and mediaeval/classical manuscripts are different things, or are they?
- There may be several reasons for encoding manuscripts, and the end result may not necessarily be the production of a critical edition. Non-critical editing, or the transcription itself can be a goal. Therefore, it would be semantically incorrect to use the Critical Apparatus subset to encode several layers of additions, deletions and variation in the transcription of one document.
- The transcription of a manuscript interprets the writing process of an author/scribe. Since writing takes place in time, techniques are needed to encode absolute and relative time in manuscript transcriptions.
- Writing is also a topological process. This is an issue, especially when dealing with modern manuscript (draft) material and correspondence. Therefore the encoding needs also to take account of the position of the text on/in the manuscript.
- Particularly in the case of modern manuscripts and correspondence, an encoder must often decide between two encoding philosophies: i.e., whether the encoding is capturing the documentary appearance of the page (i.e. encoding the text according to what is physically present on the page regardless of the text's logical sequence), or whether it is to record the textual/ontological/teleological flow (regardless of the text's physical appearance on the page). dif
- A more detailed set of tags is needed to describe the accidentials of the text: for example, a change in ink, or a change in writing utensil (pen, pencil, etc). The element <handShift> is not expressive enough to encompass the recording of these changes.
Edward Vanhoutte kicked off by proposing that the SIG be divided into two distinct SIGS, one to deal with problems and issues in encoding modern manuscripts, and the other to deal with those in encoding mediaeval manuscripts. The majority of people present at the meeting felt that there should be one SIG as the commonalities were greater than the differences. The group thus voted to stay together.
Vanhoutte went on to discuss the problems he has in encoding the internal dynamics of a text, such as using speech transcription tags to encode time-based events. Peter Robinson said that there are similar problems when encoding mediaeval manuscripts. Vanhoutte argued that the encoding strategies suggested by the TEI for critical apparatus are not suitable for the encoding of variants present in a single text. Robinson said again, that similar variants occur in mediaeval manuscripts.
The group agreed that the recording of time and sequence are problems that the SIG should address. Matthew Driscoll agreed with Vanhoutte that the <app> structure should not be used for recording variants within one text. Robinson reported in his projects he does not use <add> and <del> but <app> .
Driscoll suggested that we create a <subst> tag within which to put a <del> tag to show that a substitution is being made.
Markus Beigenheimer asked if anybody had worked with fragments.
David Birnbaum pointed out that the Guidelines describe a manuscript as a physical object, which is a problem when manuscript parts exist in separate locations, or when a manuscript has been torn and exists as two or more separate objects. Gautier Poupeau said that he used the rend attribute with <seg> to create coordinates of a manuscripts which can join fragments. Birnbaum asked that the SIG consider additional solutions to this problem
A discussion ensued which considered how to record time-based events. Elena Pierazzo suggested that a rend attribute may not be enough to describe time. Pierazzo suggested that the textual fact and the bibliographical fact need to be recorded separately while creating one whole text that is readable as a separate witness. Burnard suggested that the a <sync> tag might be used to deal with this problem.
A discussion of whether milestones are good for this purpose followed.
Jindrich Marek raised the problem of deciding what the object is (i.e., items vs the text). This occurs, for example, when there are several sermons in one text. He also suggested that a best practice be developed which can help encoders display TEI-encoded texts in PDF, LaTEx, etc.
Pierazzo brought up the issue of the explosion of the page in modern manuscripts: that is, what is the text when the author begins a letter, for example, on a sheet of paper, then continues it on an envelope, or on the back of a piece of paper stuck in the letter, etc. She asked how we should record this type of textual condition. Mirko Tavosanis raised the point that again, that this raises the problem of what theory of encoding is the encoder following: that of transcription or of codicology. He feels that the two methods are entirely different. Connected to this is the issue of how to encode the documentary aspect of the text from the logical flow of the text. Encoders tend to adopt one philosophy of encoding. In the case of capturing the documentary aspects of the page, the encoder may then use anchors, pointers and links to indicate the logical flow of the text. Conversely, encoders capturing the logical flow may use notes to explain where the text appeared originally. In either case, the Guidelines should more thoroughly document these philosophies of encoding, providing examples and alternative methodologies.
Towards the end of the discussion, several suggestions were made as to how to proceed with future work. Patrick Durseau felt that much of the discussion revolved around how to encode time-based events in manuscripts. He suggested that we think of encoding in terms of the following:
- authorial intention
- the text itself
- the analysis of the text
- the physical object/the medium on which written
- the interaction between text and physical object: sometimes one can't read the text because of damage to the physical object
Robinson suggested that the SIG undertake a survey to ascertain how different projects handle particular problems associated with the encoding of manuscripts, which would result in a list of problems that the SIG could then address. He also pointed out that we don't have enough examples of best practice, and suggested that the SIG undertake to create this. Harold Short added that the SIG take account of the work going on already in terms of best practice, such as the work of the research group at UNC at Chapel Hill. Driscoll suggested that it might be valuable to have a conference dedicated manuscript transcription where people can discuss particular issues.
- how to handle time based encoding
- how to record place based encoding
- how to encode fragments
- how to record codicology (the substance of the medium, ink stints, etc); Robinson noted that the TEI already has mechanisms to record this, but it needs to be better documented
- issues of substitutions
- issues of variation
- to clarify the role of using the critical apparatus tagset in manuscript transcription (which is dependent on whether one is encoding an edition or encoding a manuscript transcription)