Hi Stephanie,
 
This is somewhat of a problem for us in Houston as well.  Occasionally I raise the issue with our staff, but inevitably it dies down as a not terribly pressing issue.
 
Essentially, we are inconsistent (surprise!) with our approach.  Most of the time, though, we end up making constituent records for these unknown, but regional makers.  This seems to work fairly well, and a systematic approach seems to be evolving on its own to try and determine the central aspects of the "constituent" (New England, for instance) and adding in the other portions of that phrase as prefixes or suffixes

prefix: Coastal
Constituent display name:  New England
suffix:  , probably Salem, Massachusetts, or Portsmouth, New Hampshire
 
This presumes that our users are more likely to search by a constituent as "New England" than to limit it by "coastal new england".  
 
In terms of place, then, we use a place type of "possible place" fairly often, making separate place records for each possible place.  Like you, we try to avoid putting in "probably Salem, Massachusetts" as we have our query assistant for place terms set to show the list of all available terms used, and "Probablys" sort separate from "Salem". 
 
Did that make sense?  Let me know if it didn't.
 
David Aylsworth
Museum of Fine Arts, Houston

________________________________

From: The Museum System (TMS) Users on behalf of Stephanie Hansen
Sent: Mon 4/20/2009 5:25 PM
To: [log in to unmask]
Subject: Re: 'Unknown' Constituent Records



Dear TMS list-serv,

I'm hoping we can pick up on this old topic again to help me make a decision
on how which field(s) to use for our decorative arts/sculpture "unknown"
constituents. We wouldn't normally display "unknown" for dec arts-but
rather "Costal New England, probably Salem, Massachusetts, or Portsmouth,
New Hampshire" for example. Does anyone else list a place instead of
an "unknown" constituent like this?

The problem is-that most of them are more than 48 charaters
(prefix/suffix/culture) and most of them have qualifiers
like "possibly", "probably", "or", etc.-so they don't work in the geography fields
either. I'm also afraid it will get too "out of sight/out of mind" in text entries or
a udf.

Do any of you have constituents that refer to a place-or has anyone come up
with a good solution to listings such as the above?

Many thanks for any advice!!

Stephanie Hansen
Milwaukee Art Museum