On 27/06/2012 19:02, Brian Briscoe wrote:
While I agree with Jim that our current catalog interfaces have many problems, I don’t believe that all of the logic behind them is, to use his term, “broken.” The case for controlled vocabularies is very solid and the weaknesses in keyword have been convincingly proven. As a matter of fact, not all library users are keyword-searching literate. There are still many who continue to use the old-fashioned library OPAC interface that they learned in the past. That is why, IMO, it is important to continue to keep access to the library’s “old” catalog interface available for users as an alternative search on the landing page.
I have no problem with updating the terms used in LCSH and our other taxonomies/folksonomies/thesauri to more currently-familiar terms, but then those “old” terms must be retained as “See” references.
There are some pretty revolutionary things going on in library information access systems. Unfortunately, the RDA debacle has overshadowed them for attention. But there are good things happening. This is not the time to throw up our hands and cry that all is lost. This is the time to focus our attention and our message on things like controlled vocabularies that allow for real collocation of information. Machine matching and relevancy algorithms have not shown sufficent progress in this area to this point.
Librarians will never provide the fastest information. We never have. But we can provide the most accurate information. And that is the niche that we should be focusing on.
I am not making myself understood. I am 100% in favor of controlled vocabulary, and just as much in favor of the subject headings with subdivisions. What I say is broken is the dictionary part of the catalog, or expecting everyone to look things up by following arrangements based on left-anchored alphabetical order, which just doesn’t happen anymore. What are the alternatives? Classed arrangements should certainly be tried, as Mac suggested. I personally haven’t seen anything I have liked very much, and in any case, there has been relatively little done with authority files, at least that I am aware of. The best that I know of is the AAT, e.g. the heading for Adirondack chairs http://www.getty.edu/vow/AATFullDisplay?find=chairs&logic=AND¬e=&english=N&prev_page=1&subjectid=300171702, showing the hierarchies, use fors, all kinds of notes, and so on. It’s nice. But whatever is selected for the authority file, it all needs to be incorporated into the actual catalogs.
Concerning the very idea of an “authorized form” (i.e. single) to the exclusion of others, that can be reconsidered as well. I have mentioned in an earlier posting (http://blog.jweinheimer.net/2011/09/re-objection-to-authors-birth-year.html) Thomas Hyde’s catalog at the Bodleian and the unique method he used for “authorized forms”, which included the Use Fors. I could see the same thing being tried today, so Dostoyevsky would display something as:
Dostoyevsky, Fyodor, 1821-1881, or, Dostoievski, Fédor Mikhailovitch, 1821-1881, Dostoievski, Fiodor, 1821-1881, Dostojevski, F. M., 1821-1881, Dostojewskij, Fjodor M., 1821-1881, Tʻo-ssu-tʻo-yeh-fu-ssu-chi, 1821-1881 more…
“more…” would link to the entire authority record.
Like I said, there are lots of ideas from the present and from the past.