[tei-council] SV: Witness proposal

[Matthew James Driscoll]

In terms of revamping <witness>, which was never a terribly sophisticated
element, this proposal makes perfect sense. It is not clear to me, however,
whether the newly redefined <witness> element is intended as an alternative
to <msDescription> or whether there are situations in which one would use
both. Neither appeals much to me, I must say, especially as I can't see that
what is being proposed can't be done with <msDescription> as it is. The
original idea behind <msDescription>, in any case, was very much that it
should be able to be used for more than just cataloguing (our insistence on
this was what led to the bulk of the disagreements with "TEI-MMSS"), and it
was indeed my understanding that <witList> and <witness> were now to be
deprecated as redundant. If there are things which Gautier, representing
German, Italian and French practice, genuinely can't do using the current
<msDescription> mechanism it would be far more useful, it seems to me, for
us to try and adapt the existing mechanism, rather than coming up with a
parallel one.

Looking at his proposal, the chief lack seems to be a means of indicating
the sigla, for which he proposes a new <sigil> element; these can easily
appear within <msIdentifier> as <altIdentifier>, which takes a type
attribute. Otherwise the elements are largely from the <msDescription>
module, suggesting that perhaps we aren't so far off the mark after all.

Matthew

 

 
-----Oprindelig meddelelse-----
Fra: Lou Burnard [mailto:lou.burnard at computing-services.oxford.ac.uk] 
Sendt: 22 March 2007 00:22
Til: Matthew James Driscoll; TEI Council; gotpoupeau at infonie.fr
Emne: Witness proposal

In early november last year, Gautier Poupeau whom several of you will 
recall from his presentation at the Members meeting, sent me a document 
outlining a proposal to address the inconsistency currently subsisting 
in the Guidelines about how manuscript (or print) witnesses should be 
documented.  I owe him, and the Council, a major apology for having not 
done anything with it -- it simply got mislaid and then forgotten. 
Fortunately, he's a persevering kind of chap, and I have now recovered 
the draft and placed it somewhere you can see it, viz:

 http://www.tei-c.org/Drafts/witnessPart/proposal-witness.xml

It's in French, so I will summarize briefly the proposal here (I will 
also translate the document fully if so requested)

* create a new class called model.witnessPart. which is a subclass of 
model.phrasel; use this class in the content model of <witness>
* add to this class a number of elements currently specific to 
msDescription, specifically:
   origDate, material, physDesc, seal, dimensions, 
filiation,msIdentifier, msContents, rubric;
* re-introduce the "sigil" attribute as a new <sigil> element
* add a "class" attribute to <witness>

The suggestion is based on Gautier's extensive experience in adapting 
TEI to the description of cartularies and similar documents, with due 
regard to practice in German, Italian, and French collections.

I think the work is sufficiently straightforward to do quite quickly 
(i.e. it can be done for P5); I also suspect that it would be a very 
smart political move to be seen to be doing something in this area.

Initial comments?

(It occurs to me also that this might be something the newly revived mss 
SIG would like to kick around -- but that could happen independently of 
plumbing his suggestion into P5)