10/18/96
To: USMARC
From: Ed Glazier, RLG
Subj: Notes fields in holdings records

In regards to the following question posed by Robin Wendler:

    Yet another question for us MARCophiles.  Several notes in the
    bibliographic format seem to be by their very nature specific to a
    given holding, for example:

            541     Immediate Source of Acquisition
            561     Provenance
            562     Copy and Version Identification Note

    While I realize that the issue of expanding the holdings format
    is a controversial one, I would be happier if these notes could be
    communicated at the holdings level.  What are the pros and cons of
    permitting these fields to be used in USMARC Holdings as well as in
    USMARC Bibliographic?  There is certainly a precedent in, for example,
    the dual citizenship of the 583 field.  What do you think?

My own belief from implementing the display of USMARC holdings
records in RLIN is that we would be better off if notes fields that
are by their very nature copy-specific were unambiguously moved to
the Holdings format or at least, given parallels in the Holdings
format that did not duplicate the tags used in the Bib format.  In
RLIN, holdings data, whether for a single copy or for multiple
copies, is always contained in a holdings segment attached to the
bibliographic segment of the record.  The fields that don't exist in
the holdings format must always be displayed with the bib segment,
even if there are multiple copies.  Those fields like 583 that exist
in both formats cause special problems, because it is not always
possible to determine unambiguously whether they are intended to go
with the bibliographic data or the holdings data.

The difficulties stem from the provisions of the Holdings format
that allow holdings for a single copy to be communicated as data
imbedded in a bib record but that require holdings to be in separate
holdings records when information about more than one copy is to be
communicated.  As Robin suggests, you can include such notes in a
bib record for a single copy and can't when you have multiple
copies.  For an institution attempting to communicate holdings data
in accordance with UFHD, this may mean that some local records must
be manipulated (including stripping out data) before the records can
be "legally" communicated.  An approach that would be more friendly
to local systems would be to allow them to communicate holdings data
in more or less the same way it is stored in a local system that
supports UFHD.  The consistency of requiring holdings data always to
be in holdings records whether for single or multiple copies strikes
me as a more logical way to communicate this data and to attach note
fields unambiguously to the copies to which they relate.


To:  [log in to unmask]