[tei-council] bibliography

[Lou Burnard]

Sebastian and I had some discussion of this this morning, and I have 
subsequently done a bit of poking around. Here are my conclusions and 
some recommendations.

There are three different kinds of bibliography we might ideally want to 
  see included in P5.

(1) A reading list, giving key works on text encoding theory and 
practice, and also possibly enumerating key TEI-related publications

(2) A reference list, giving standard form bibliographic info for all 
the reference works actually mentioned or linked to in the running text

(3) A citations list, giving at least title/author details for texts 
from which examples in the text have been taken

Some items in (2) will definitely also appear in (1); some equally 
definitely won't. Almost none of those in (3) will appear elsewhere.

At present, the Guidelines have their own version of (1) which is sorely 
out of date. Some items in it (but very few) are linked to from 
footnotes in the text; for the most part, however, bibls which would 
ideally appear in (2) are either missing, or embedded in a <note> in the 
text. As for (3) there is a very small number of <bibl type="cit">s 
which might possibly be extended semi-automatically, located within the 
body of the text.


Sebastian has drawn my attention to the recently-announced bibliography 
produced by the Education SIG, which is now under editorship of Kevin 
Hawkins (see 
http://www.tei-c.org/Activities/SIG/Education/tei_bibliography.xml). I 
think this is potentially a suitable replacement for the current 
Bibliography and have written to Kevin and Susan to ask whether they are 
willing for me to incorporate a version of it in the Guidelines. If they 
are I think that solves the major headache of producing (1)

Turning to (2), there are 154 <bibl> elements within the body of the 
Guidelines. Extracting these automatically into a separate file for 
updating wouldn't be hard. I need someone to volunteer to  structure 
that file consistently with the format adopted for (1), and of course 
eliminate duplicates, check validity etc. I could then replace all those 
<bibl>s with <ref>s, the target of which would either be in (1) or a new 
(2), made for the purpose, containing referenced works not related to 
markup or TEI.

Producing proper bibliographic references for (3) would be much harder 
however. There are 1872 <egXML> elements in the Guidelines. Even 
discounting the 398 of these which are empty, and the unknown number 
citing from the same source, that still leaves an awful lot of checking.
Do we really think it would be worth the effort? I think the best we can 
hope to do is provide a separate list of "works exemplified" giving what 
little information we have, which needs to be pitched not as a 
bibliographic tool but as a kind of "spot the source" crib.


All in favour?


Lou

Sebastian Rahtz wrote:
> It struck me that normalizing a <bibl> for as many <exemplum>s as we 
> possibly
> can would be very useful for legal purposes; if we are challenged on our
> right to take bits of published works, indicating that something *is*
> a quotation would show (at least) that we were not being deliberately 
> dishonest.
> 
>