Suzanne, in a sense I think you are asking cataloging questions. One of the
reasons for leaning on cataloging rules is that there are often many
different ways that you can represent the same item (look at how many
different citation rules there are). None of them do everything, and none
are "perfect," but you choose one and go with it. The cataloging rules then
inform your electronic record structure. I think it's hard to answer these
questions in a vacuum because there is no single answer.

That said, it sounds like you want to maintain the relationships between
the parts, which leads me to think that you might do better looking at
archival "cataloging" rather than library cataloging. The archival
community is less interested in pure identification and retrieval and is
more interested in defining structure and context. Libraries are willing to
finesse structural relationships as long as retrieval mechanisms (i.e.
keyword searches) will deliver the records to the user.

That said, I know that archivists use the EAD ( as
their data format, but I don't know what they use as rules. METS can give
you structure, but it doesn't provide cataloging -- so the cataloging has
to come from elsewhere. We store our EADs in METS, since METS works well as
a wrapper around metadata. We also store MARC-XML in METS, MODS in METS,
and DDI (Data Document Initiative) in METS. So METS could wrap around a set
of records and give them structure, but it doesn't help you know what
should be in your records in the first place.


>Record 1 is for the part 4 of volume 4
>Record 2 is for volume 4 of the 7 volume set
>Record 3 is for the 7 volume set ( with a relatedItem type="host" to the
>over all multivolume set).
>Do you all think that next I should start looking at METS to figure out how
>to make a structure map to house these separate MODS records????

Karen Coyle           [log in to unmask]