The problem here is that the way subject is specified in LCSH and DC is
inconsistent. As Mary Larsgard points out, DC elements are (at least
implicitly) closer to being single-concept descriptors than the kind of
compound structures the LCSH strings provide. And while I'd concede that
descriptors are more searchable than LCSH strings, I'd also point out that
LCSH strings properly constructed and sorted in a browse index are much
easier to read through and select from than individual titles in a
descriptor-based result set. The main virtue of card files was their
browsability, and as long as heading browsing is a feature of online
systems, that virtue carries over.

The problem with identifying DC Subject elements as "lcsh" comes when we
attempt to move data in these DC records into MARC/AACR2/LCSH based
systems. If a DC element is labeled "lcsh," but is only compatible with
LCSH's main heading vocabulary and not with LC's "Subject Cataloging
Manual: Subject Heading" rules for heading construction, the DC data will
NOT reliably translate as an accurate LCSH heading. The more that happens,
the less reliable and consistent the receiving MARC/AACR2/LCSH database
will be. It would be nice if there were a way to specify that a given
thesaurus has been used for vocuabulary only, and not for syntax, since
this is an important consideration when we attempt to integrate data across
systems and record formats. Without such considerations it's too easy for
interoperable data to become interoperubble.


Stephen Hearn
Authority Control Coordinator/Classics Bibliographer
University of Minnesota
160 Wilson Library               Voice: 612-625-2328
309 19th Avenue South              Fax: 612-625-3428
Minneapolis, MN 55455      E-mail: [log in to unmask]