Re: Discussion of id.loc.gov

From: Patrick Cates <Cates_at_nyob>
Date: Tue, 19 May 2009 09:27:26 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
James Weinheimer:

"I don't believer that is what I am saying. It's that the heading
"Aeronautics, Military--Accidents--Italy, Northern--History" is a
conglomeration of lots of concepts, and giving them all one URI doesn't
seem
to be the best solution.

It would seem to me that the better solution is to give each subdivision
a
separate URI and then the record can bring the URIs together in
different
ways. in this case:
http://id.loc.gov/authorities/sh85001371
http://id.loc.gov/authorities/sh99002078
http://id.loc.gov/authorities/sh85069034
http://id.loc.gov/authorities/sh99005024"

I assume I'm pointing out something that everyone already knows, but
many of the long LCSH strings like the one above were created only
because Voyager, LC's ILS, can't recognize that a subject string is
valid, even if it recognizes all the parts as valid, unless the string
has its own heading (geographical subdivisions are a particular problem
because Voyager doesn't seem to know what to do with the 781 in
geographical headings).  If you go to the authority record for these
headings, you will see a 667 note that says "Record generated for
validation purposes."  Starting in 2007 LC has generated thousands of
these headings, because, basically, Voyager has crappy authority
control.  I think that a system derived from LCSH should strip out these
records in favor of their constituent parts.

Patrick Cates
Technical Services Librarian
St. Mark's Library
General Theological Seminary
175 Ninth Avenue
New York, NY 10011
212-243-5150 x354


This e-mail was sent from the GTS and Desmond Tutu Center email service. If you are not the intended recipient of this email, please notify the sender immediately or call us at 212.243.5150 x376.
Received on Tue May 19 2009 - 09:29:10 EDT