I am in the early stages of developing a tagging protocal of EAD elements
and and an overall design structure for EAD prototypes at the University
of Illinois archives. I am assuming that someday it may become possible
to run our entire finding-aid system using files marked up in EAD--which
is a big assumption, but I would like to be prepared in case it comes to
pass. In any case, it is my belief that the EAD markup protocol should be
designed so that no changes to the marked-up files will be necessary in
the future. In other words, a good structure should be designed early on,
which would allow, if we later think feasible, marked up documents to
supplant our current system of paper and database finding aids.
I have thought about this issue for a while, and am wondering if anyone
has advice on which approach (or possibly a different approach) would be
the best way to ensure a stable enough structure to make future
compatibility more likely:
1) mark up each discrete "collection" in separate EAD files. In
some cases the "collection" would be an entire record group, in some cases
a separate sub group, a portion of a sub group, or even a single record
series. So the <archdesc> level in each finding aid would vary. This
process would obviously simplify the creation of finding aids, but I have
a concern: if in one record series archdesc level="record group" and in
another "series," the latter will not include record group and sub group
information. Also, it would seem to compilicate things for the
programmers who have to design the search engines, etc lying behind the
system, to have the higher components shifting levels all the time--or is
this not an issue since the "level" attribue will take care of it.
2) I could design a system so that each record group is coded
as a single EAD document; in other words the higher levels of the
components would invariably correspond to a particular level in our
current classification system (i.e. <archdesc level="Record Group"> <c01
level="Record Group"> <c02 level="subgroup" c03 level="record series".)
It seems to me this offers a very structured environment so that all
record series could some day be included in a master system. But the
files would be huge and difficult to maintain. Also, I am not sure how to
deal with some related record series within the subgroups. These are part
of associational archives that we took on, but were forced into a
classfication system created for University records. I thought
about making <c03>'s level attribute variable--either "collection" (in
which case c04's level would become "record series" for everything
thereunder or "record series" but am not sure this would work--again this
is probably my ignorance of the supporting programming paramaters.
My questions:
1) Is it possible to combine my two methods mentioned above by
creating an hierachy of linked EAD documents? In other words have a
"master" EAD document for each record group, under which subgroup
documents are linked, under which "collection" or individual record
series documents are linked. If so, how would I go about this, since I
don't seem to see it mentioned in the tag library?
2) Could I use <eadgrp> and its related <...grp> elements. I know
this is not recommended for record series held by a single archives or
adminstrative unit, but don't understand why.
3) Is there some other way to accomplish what I am attempting, or
would I just be better off handling all files on a case by case basis and
forgoing my dreams/fantasies of an EAD based cataloging/finding-aid
system?
Thanks for any help, and sorry if this question is too naive from an EAD
novice!
Chris Prom
Visiting Assistant Archivist
University of Illinois Archives
|