> Miraculously, one of those books is about what coots eat: _Food habits of
> the American coot with notes on distribution_ by John C. Jones (1940). It
> has the subjects:
> Coots
> Birds--Food
>
> Which I think nicely illustrates one problem with LCSH. Coots are birds, so
> by logic Coots--Food would also apply. But that's not logic you can encode
> easily.
Actually, there are ways of doing this, some of which I am sure are
already being used by people on this list.
It is nuts to retain precoordinated LCSH entries verbatim in an online
environment. Cataloging practice has long been to express the
"aboutness" of works over multiple headings. LCSH is an archaic
committee-determined vocabulary lacking in both structure and
completeness. It makes little sense to librarians, let alone normal
humans. The rules for constructing subject strings are so convoluted
that even catalogers with 20 years of experience must consult the list
of headings and the Subject Cataloging Manual to produce them. Because
of the complexity and other factors, cataloging is much less
consistent than librarians like to imagine. This is why practically no
patron uses subject searches.
It is unreasonable to expect anyone to guess LCSH headings or to take
the trouble to find out which ones apply to the need at hand. We don't
like software written by people who don't understand what staff do
because it imposes screwball workflows. Likewise, our patrons will not
appreciate it if we force them to use a search process they find
unnatural. Only in libraryland is it necessary to train people to use
a web page -- even kids seem to find what they want without problems
in google, kazaa, flickr, wikipedia, youtube, and a zillion other
information sources.
Despite their shortcomings, LCSH headings and call numbers are useful
for giving context to search results and grouping them. But this needs
to be done in the background without the user needing to know how our
tools work or how information is structured.
Catalog entries, tags, and full text all have their place. Even
presuming infinite labor and financial resources, the optimum search
strategy for surrogates (i.e. catalog records) and the materials
themselves (i.e. Google) are not necessarily the same. Also, the
efficacy of searching surrogates rather than actual resources depends
on what you are looking for.
kyle
Received on Tue Jan 29 2008 - 15:23:35 EST