Reading the draft of Descriptive Cataloging of Ancient, Medieval,
Renaissance and Early-Modern Manuscripts (AMREMM, draft of 11/24/99
available: ftp://165.134.156.3/vatican) by Gregory Pass I noticed that
he includes rules for cataloguing archival items (1B2.3.).

As these are rules for MARC based cataloguing, it seems to use a short
description of the item in the title (what would be in EAD <unittitle>?)
field (rather heavy book influence this?). Is this common practice, and
how does it work if you want a longer item level description? It seems
to me that you will have to repeat most of what is given in the invented
title in the description, or make the title far longer to include enough
detail. This redundancy and repetition seems contrary to the principle
of only describing something once: with a finding aid of tens of
thousands of items it will double the length and amount of description
required.

I am interested to know if anyone else has been using EAD for in depth
item level description, (viz a paragraph or even pages of description
per item), as this area does not seem to be addressed in the Application
Guidelines, and has proven impossible to do in a straightforward
fashion.

Aside from it being nice to fix a realistic way to do item level
descriptions after four years of trying, I am starting to look at
indexing (authority controlled searching rather than "back of the book",
that is), which is another area as yet unclear in purpose. Assuming the
use of a <controlaccess> element containing index terms, it would be
useful to know what the result would be of scoring a hit on one of those
terms. Will the idea be to retrieve the entire <c0?> element in which
the <controlaccess> data is held, or something more subtle (the <unitid>
and <unittitle>: the <scopecontent>)? While Z39.50 may make it possible
to experiment with this with some flexibility, I have already found
different retrieval systems displaying strange combinations of elements,
and would suggest if we had an agreed idea of what it is useful/intended
to retrieve it would help make all our (oh, alright, my) data more
consistent.

--
Richard Higgins
Durham University Library