Book People Archive

Re: LC Convenes Working Group on the Future of Bibliographic Control

Bowerbird@[redacted] wrote:
>>   I'd hate to see a potentially useful subject description mechanism 
>>   (and one that libraries have put over 100 years of work into)
>>   thrown overboard prematurely just because the tools that 
>>   libraries have provided to work with it have been so bad.
> 
> i won't comment either way about subject indexing, but
> aren't there a bundle of contradictions in that sentence?

Fair cop; I was in a hurry, and made a complete mess of that sentence.
What I was trying to express, though, is in fact a bit of a paradox;
we in libraries have spent over 100 years building up a detailed
vocabulary of subject terms that has a whole host of rich relationships
expressed and expressible in it: the Library of Congress Subject
Headings.  Yet, in those 100 years, we have
not gotten far at all at giving users (or librarians) good tools for
*using* this vocabulary to easily find things they're looking for.
(Try doing a subject browse on a popular topic in most online catalogs,
and you'll quickly see what I mean.)

It's as if you were trying to get to, say, a nice apple orchard north
of Philadelphia, and instead of being given a map with roads, towns, and
orchards all nicely laid out on it, what you had to work with was a
telephone directory (which did at least let you look up apple orchards),
an alphabetical list of towns, and another document listing the roads that
go between different towns, arranged alphabetically by road name.

Okay, maybe I exaggerate a bit in my analogy.   But not by much.

At the same time, folks have come along with other tools that *are*
easier to use (think Google, tagging networks, recommender systems, etc.)
These are useful too for many things-- I don't intend to deny that--
but the underlying information structures that they use have not yet
been shown to scale up as well to let researchers find the most relevant
resources for their particular specialized inquiries in large collections.
They may find decent-quality popular resources on a more general topic
area, or lots and lots of books that happen to mention a particular
term which the reader is curious about.  But that isn't good enough
for many kinds of research.  Still, most librarians can't help noticing
that not many of their users are using their subject search and browse
tools, and a lot of them are using Google et al.

Going back to our orchard analogy, I could probably use Google Earth and
mouse around the satellite views until I spotted a characteristic pattern
of trees that signifies an apple orchard.  Then I could trace the roads back
to figure out how to get there.  That would be pretty intuitive, and kind
of fun to do.  But it might not be as effective as using a map painstakingly
drawn by someone ahead of time that had orchards and roads nicely
marked on it.  We could draw such maps if we did a little more work--
we've already painstakingly compiled the orchard and road information,
after all, but we've presented it in a way that makes most users throw up
their hands and try their luck with Google Earth instead.

And, who knows?  Maybe Google Earth *is* the best way to go.  But we won't
know that for sure if we never bother to draw our maps and give them to
people.  And if we throw out our phone books and street directories before
we make the maps based on them, we might never know whether we might
have provided our users better maps after all.  (It'd be especially
tragic to do this if Google Earth hasn't actually covered the ground
we've documented-- here I'm thinking of the many items in libraries that
have *not* yet been the subject of digitization or otherwise been looked
at by new kinds of searching mechanisms.)

Perhaps all I've done here is replace a sloppy sentence with a strained
analogy.  But maybe I've piqued enough interest to entice people to
have a look at what I'm working on.

John ( http://labs.library.upenn.edu/subjectmaps/ ) Mark Ockerbloom