Fwd: Re: Workshop on Transcription of Medieval MSS at Berkeley April 26 (fwd)

[M. J. Driscoll]

>  From Charles Faulhaber:
> 
> >In the teiheader, it seems to me that there needs to be a "repository
> >statement," a place where one can record explicitly the name and location
> >of the repository that owns the manuscript being transcribed. We've used
> ><bibl> as a stopgap, but it really isn't a bibliographical entry.
> >
> >Any suggestions?

This is precisely what the <msIdentifier> element, which goes in the  
<sourceDesc> in the <teiHeader> and can contain the subelements 
<country>, <region>, <settlement>, <institution>, <repository>, 
<collection> and <idno>, is meant to do.   Of course, it hasn't 
actually been adopted by the TEI yet, but hey, you can't have 
everything.  There is, incidentally, no difference of opinion on this 
point, so far as I am aware, between the MASTERites and the 
Dutschke/Proffitt camp.

> >
> >I'm still struggling with the correct coding of words so that they can be
> >extracted correctly for the purposes of concordances, word lists, and the
> >like. Thus a word can be broken across two lines, but there needs to be
> >some way to bring the two parts back together. It can have all sorts of
> >abbreviations and addition-deletion combinations; but you still have to be
> >able to extract the word with the addition but minus the deletion.
> 
> and later:
> 
> >I've seen no solutions on either the repository or the word problems.
> >David Seaman suggested using <orig> as a workaround, and it does work, but
> >in fact that actually flips the relationship, since what one is making is
> >_not_ the original reading but the cleaned up editorial reading. In some
> >sense it's the same relationship as between <sic> and <cor>.
> >
> >I think that TEI needs a <w> with a lot fewer constraints on what can go
> >inside.

We've run into exactly the same problem here in the Old Norse 
speaking.world.  In order to do the sorts of things manuscript people 
(at least) want to do, <w> needs both more attributes, REG, for 
example, and something for the "cleaned up editorial reading"(if 
different from the regularised form - we've been using REND, but 
I'm not completely happy with it) and, as Charles says, fewer 
constraints on what can go inside.  At the moment, one cannot have 
<abbr> or <expan>, <orig> or <reg>, <add> or <del> and so on 
inside <w>, which renders it fairly useless for transcription purposes.   
The alternative would be a new element, called, say, <word>, but 
why bother?

Matthew