At 09:46 AM 5/29/97 +0300, Alexandra Apostolidou wrote:
>Dear colleagues,
>We are in the process of cataloguing the collections of a new art
>museum. We would appreciate any feedback we can get. Here are some
>questions:
>When we have a set of objects that relate to each other (such as a
>triptych) how do you deal with it? Do you accession it as a whole by
>attributing a number like x.x.1-3? Do you catalogue it first as a set
>and then each part of it separately?
>How do you catalogue installations?

The best way to catalogue multi-part items is to identify the whole as a
single entity and then to catalogue each piece separately.  This works well
for triptychs and items that have individually identifiable parts, but less
well for sets of the same items repeated, such as tea-sets or chess-sets.

If you are cataloging on paper, it is easy to identify 12 saucers as
x.x.1-12 and 12 cups as x.x.13-24, and to number them accordingly. But if
you are cataloging in a computerized collection management system it is
best to have one specifically designed to handle part/whole relationships.
Most of these permit parts to be catalogued as separate items and yet be
understood as part of a singular entire.  For example a portfolio of prints
by different artists should exist as a published item and as individual
items.  A locomotive may be catalogued as an entire, and, depending upon
the needs of the museum, its component parts may or not be catalogued
separately.

It is sometimes difficult to preserve the identity of the whole in paper
cataloging when the parts are not accessioned at once.  For example, if a
triptych is received one-panel-at-a-time from different sources, it will be
impossible to identify each piece using the x.x.1-3 formula.  What is
needed is a means of identifying intellectual constructs that are not
necessarily physical entities.  A poor system may identify the triptych as
one item and its component parts as three more.  Being able to decide which
records stand for physical items and which stand for descriptive entities,
therefore, is also important.

One good way to learn about part/whole relationships is to discuss the
issue with the various vendors of museum object management software.  Each
vendor will have his/her own formula, a reflection of their understanding
of the real-life problems.  They can provide considerable insight.


===========================
Robert A. Baron
Museum Computer Consultant
P.O. Box 93
Larchmont, NY 10538 (USA)
mailto:[log in to unmask]