In article <[log in to unmask]>, "Robert A.
Baron" <[log in to unmask]> wrote:

> I've run into situations in which a multi-part item (portfolio, teaset,
> etc.) is donated in parts -- a year at a time.  For these items the
> numbering system proposed above does not work well because the accession
> number scheme attempts to track two potentially dispirate concepts: date of
> acquisition and acquisition group.  For this reason, in one collection
> management system I built, I provided for a separate group number.  The
> accession number was left to record the logic of attribution and items were
> united by reference to a group number that could be attached to any number
> of items.  If the group number is recorded in a multi-valued field, it can
> also be used to track groups within groups.  When using the group number,
> the first entry would catalogue the group name and successive entries, e.g.
> 1996.50.1, 1996.50.2 would define the parts.  In addition, accessories
> associated with the group, a case or a frame can also be catalogued in this
> way.  A query for Ben Shahn Portfolio would produce the top number
> (1996.50) and a subsequent query for 1996.50.* would bring up every item in
> the group.

I may be missing the point.  If you wanted to find your Ben Shahn
material, wouldn't you perform a find in the Artist field.  And if you
wanted to find material donated by a certain person, wouldn't you perform
a find in the Donor field?  I think there are enough numbers in the world
already.  Finding aids in your catalog and donor cards and on your
computer (country, maker, object type, etc.) should eliminate the need for
layers of numbers.  Your computer can also find multiple fields (ex. donor
and object type).

And don't forget the power of the mind.  A certain institution transferred
material from one donor to us at four different times.  After being here a
year, I knew by heart three of those numbers (and we have 2600+ accession
groupings.)  After three years, I can list many of the objects in those
transfers (not all 1000 though).

Sally Baulch
Collections Manager
Anthropology and History Division
Texas Memorial Museum

PS I favor having a collection number, an individual number denoting a
group, and then a part designation.  We do employ the alphabet to list
individual parts and usually don't end up with alphabet soup (with a few
exceptions.)  My closest example would be Curtis prints that were grouped
by the artist into 20 portfolios.  Each portfolio got a number and then
each photo was assigned a letter (1192-69 A: collection number, object
number, part letter).  We've kept the same accessioning system as was
decided upon when the Museum opened.  It's not my favorite, but it works.
I cringe when someone numbers one shoe 1996.1.1 and the other 1996.1.2.