TEI: Summary of TEI Manuscript SIG Meeting

<date value="2004-10-23">23 Oct 04</date>Susan SchreibmanFor TEI website

Created by Susan Schreibman from a Template file made by Lou

Converted to P5various and sundry minor updates for placing on web

Summary of Manuscript SIG Meeting Baltimore, 23 October 2004

Outcomes: SIG members will undertake come up with solutions for:

  • time based encoding;
  • will revise Chapter 18, Transcription of Primary Sources.

Summary of the Task Force on Manuscript Description

Matthew Driscoll spoke about status of the new chapter Manuscript Description. The goal of the Task Force was to harmonize the work of two groups who had worked independently on manuscript description, the TEI MMSS and MASTER. This chapter has been written, and will be released with P5.

The draft of the chapter is available online. The element set is also available by going to Roma and generating the element set. Beta testers wanted!

The members of the Task Force feel they have gone as far as they can go at present. However, they are aware that some items have not been addressed. The Task Force recognizes that there is a need for more elements for physical description — orthography, decoration, paleography, binding, etc. The Task Force toyed with the idea of having separate plug-ins rather than elements for general description. But as there is a physical bibliography work group looking at these issues, they will probably come up with an alternative solution.

Those reading the chapter now will notice that there is still a need for examples. The Task Force welcomes their submission. If you have one that would complement what already exists, please send it to Matthew Driscoll and David Birnbaum.

Matthew addressed the differences that might arise with what the taskforce did, which had a bias towards mediaeval manuscripts, and situations that might arise when encoding documents from other time periods. The Task Force, however, tried to generalize their work enough so that it could work with modern manuscripts as well. Ultimately, they would like to see the MS Description chapter be general enough to be used with any text-bearing object — tombstones, urns, etc, and they would be particularly interested in feedback on this aspect.

Issues Raised During the Discussion on Manuscript Description

  • As with the options currently available in the TEI for unstructured and structured bibliographies, the MS Description DTD allows for more structured manuscript description using child elements, or less structured description by simply using p tags with PCDATA for each of the major elements.
  • What does not exist at present is a converter which will convert Master records to this new scheme.
  • Patrick Durusau asked if there is a way of handling a manuscript which exists in multiple parts in multiple locations (that once existed as one unit). Matthew and David explained that this MS Description describes the physical object, and in this way, adopts a library-centric view of the manuscript. Thus there would be four records (for example) which could be linked together. Part of this view is historical — it came from the work done by the previous work groups. An alternative solution would be to use the altIdentifier to identify where the main part is, and then supply information regarding the other parts.
General Discussion on Issues Regarding Manuscript Transcription

Matthew suggested that two of the primary tasks would be the revision of two chapters: Chapter 18, Transcription of Primary Sources, and Chapter 19, Critical Apparatus. Patricia Bart suggested that she would like to see included in the Critical Apparatus tagset an element which does not require adding in additional readings, but allows a user to see patterns in readings. The group asked her to put together a proposal with some examples.

Dot Porter said that at University of Kentucky they are now including descriptive information in the transcription — including codicological markup.She wondered if the descriptive description in the header could be linked to description in the transcription. David Birnbaum said they were doing this in his project by using the standard TEI reference system through the decl which links an element in the header to an element in the body.

Dot also suggested that the TEI might want to explore the visualization tool for finding quires that has been developed at U of Kentucky. Dot explained that one could include a manuscript and an image in which areas of the transcription are keyed to the same location in the manuscript. This tool is part of their Eclipse suite, which will be released under open source license, probably sometime next year. David added that he has not seen much interest in the TEI community in linking images to text and he welcomes this initiative. In a follow-up email to the TEI-L, Dot added that the software is called Edition Production Technology (EPT), which is built and deployed using the Eclipse platform. For a description of the EPT, see “Edition Production Technology (EPT) and the ARCHway Project,” p. 36-38, DigiCULT Newsletter, August 2004. A more extensive article is forthcoming in the next issue of Literary and Linguistic Computing, which is based on the 2003 ALLC/ACH meeting.

Matthew brought up the issue of substitutions. On Friday Lou Bernard demonstrated a new combine tag, which is the opposite the choice. Matthew felt that it might not be a combine element which is needed, but rather a substitution element. However, this element on its own would not deal well with different levels of substitution. David brought of the issue of transpositions. Possibly we need a different way of representing transpositions and substitutions. Daniel Apollon added that they are looking at some of these issues at Bergen.

Syd Bauman returned to the issue of time-based encoding which had been raised earlier in the discussion. He asked if the TEI should find a simple solution to the problem, or look for a more generalized solution.

To wrap up the discussion, Susan Schreibman suggested that the members of the SIG concentrate on two issues over the coming year which seemed most pressing as a result of the discussion: i.e. time based encoding, and revision of the transcription chapter. Patrick seconded this idea adding that he felt that this chapter could be rewritten in a more readable format.

Susan also mentioned that it seemed clear from the discussion that members of the SIG would find it useful to have a page which links to documentation generated by various projects. She said that this could be done under the TEI Training (now TEI Education) SIG. Tone Merete Bruvik added that the Ibsen project will be translating their encoding practice into English and would make it available to the TEI community.


  • Daniel Apollon, University of Bergen
  • Patricia Bart, University of Virginia
  • Syd Bauman, Brown University
  • David Birnbaum, University of Pittsburgh
  • Tone Merete Bruvik, University of Bergen
  • Matthew Driscoll, University of Copenhagen
  • Patrick Durusau, Society of Biblical Literature
  • Ying-chun Hsieh, National Digital Archives Program, Taiwan
  • Michael Popham, Oxford University
  • Dorothy Porter, University of Kentucky
  • Perry Roland, University Virginia
  • Susan Schreibman, University of Maryland