Responding to the problem posed by Marsha--and thinking that the thread may
be interesting the list as a whole--I'll say that I remain convinced that
the best descriptive solution for dealing with multiple, separately
organized and housed, accessions is to fold them together into the logical
group/subgroup/series structure of the component-level description.  I think
it just makes a much more understandable presentation for the end user, who
is able to then grasp the intellectual structure that defines the
collection.  If it is necessary or advisable for the repository to maintain
administrative control of the materials on an accession basis, the component
descriptions can each bear their accession identifier, and other
administrative information (like accession-specific access provisions) can
be presented in a <scopecontent> at that level.  Better yet, with the
release of EAD 2002, those <admininfo> elements should each be available at
all the component levels, as well.

While this approach may break down in very complicated situations where many
separate accessions exist that all interfile intellectually, I still think
that it's the right thing to shoot for.

Dennis Meissner, Archival Processing Manager
Minnesota Historical Society <>
Voice:  651-296-2496  |  Fax:  651-296-9961
E-mail:  [log in to unmask]

-----Original Message-----
From: Marsha Maguire [mailto:[log in to unmask]]
Sent: Friday, November 09, 2001 1:50 PM
To: [log in to unmask]
Subject: Multiple accessions

Last year there was a thread on the subject of encoding additions to
collections, and reading those messages over was very helpful. But I'd like
to reopen this topic if I might, and ask if any of you have encoded
collections consisting of multiple accessions, some of which may never be
merged into the original accession? I'm tagging the finding aid for a
collection that consists of 27 different accessions, and I'm not sure of the
best way to structure these. I could use a separate <dsc> for each
accession, but when each accession has a scope/content note and possibly a
separate restrictions note, would this be placed in a c01 for the accession?
This doesn't feel quite right, but neither does a solution that nests <dsc>s
hierarchically, since each accession is a sibling of every other accession.
I could set up a top-level <dsc> and then a child <dsc> for each accession,
but if we've described the record group as a whole at the top <archdesc>
level, it seems wrong to do it again in a top-level <dsc>.

Anyone else have this problem? Heaven knows, I don't want to become known as
a tag abuser!  :-)

Thanks, all.


Marsha Maguire
Manuscripts and Special Collections Cataloging Librarian
University of Washington Libraries
P.O. Box 352900
Seattle, WA 98195-2900
(206) 543-1879; fax (206) 543-1931
email: [log in to unmask]