Jim, this is something I question, and I'm going to do a blog post on
it. There are some definite problems with using the "dash-dash" form if
you want to link these headings with the name headings that will
supposedly be created. Also, the loss of the meaning of the subheadings
(topical, geographic, etc.) will in general make the headings less
meaningful. I don't know why the headings were given the form they were
(although it is obviously easier to encapsulate them in skos without the
complication of subheadings or facets).
I've also been reading through the introductory parts of Lois Mai Chan's
book on LCSH trying to get a better feel for what the LCSH file actually
represents. Given that what is in the subject authority file is not
subject headings but *patterns* for subject headings, it isn't clear to
me how one would make use of the ID service in a linked environment. I
am particularly struck by the emphasis in Chan on making choices --
having to choose one term or one form. Now that we have identifiers, we
shouldn't be limited to any one natural language form for a subject
heading, but should be able to associate any number of strings with the
same identifier as long as they are semantically equivalent. It seems to
me that this could change the act of subject cataloging from one that
reduces to one that expands. I think that's worth exploring.
kc
Weinheimer Jim wrote:
> This probably belongs better on the id.loc.gov list, but I hesitate to subscribe to something else....
>
> But this may be the correct list anyway to get a more general opinion: Is there general agreement that the LCSH headings should definitely be in the form that a heading with lots of subheadings, e.g.:
> Aeronautics, Military--Accidents--Italy, Northern--History
>
> should have only one URI? http://id.loc.gov/authorities/sh2007009179.
>
> In reality, this heading is a a big, mixed up conglomeration of concepts and it could be argued that the heading above came together as it did only after a series of historical accidents. By this I mean that some cataloger in the past decided that "Accidents" can be a valid subdivision of a specific type of subject while e.g. "Architecture" cannot.
>
> And, if the idea is to cooperate with other terminologies, wouldn't it be better to break this heading into its facets? Here, there are several facets (essentially one for each word) but at least each subdivision could get a different URI. The URIs would be collected and displayed as each community would want.
>
> In this way, I don't think I am arguing for descriptors since libraries could continue with traditional LCSH strings. It's just that each subdivision would get a separate URI, and massive duplication could be avoided (e.g. Accidents as a subdivision would not have separate headings 1000 times, and History 100,000 times). So we could retain our practices and I think we could interact more easily with descriptors and other thesauri.
>
> Do others have any thoughts on this? Now that people can take the LCSH, maybe somebody can try splitting them apart and see what happens.
>
> Jim Weinheimer
>
>
>
--
-----------------------------------
Karen Coyle / Digital Library Consultant
kcoyle@kcoyle.net http://www.kcoyle.net
ph.: 510-540-7596 skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
Received on Mon May 18 2009 - 11:05:17 EDT