Beginning of the end for Dewey?

I know that I don't often get into the minutiae of librarianship on this blog, but this is a development that I will be following with interest.

Prelinger Library has done away with the Dewey Decimal Classifaction system (DDC) in favour of broader (and quirkier) subject groupings. The Gilbert Public Library in Arizona has followed suit -- albiet with more traditional subject groupings. It is an interesting concept although not terribly revolutionary, in my opinion, given that DDC already maps to subjects anyway. DDC has a finer granulation, though, so when you find one book on a topic, the books around it will be on the same sub-topic and sub-sub-topic.

DDC's performance varies, depending on how well it has been developed for a certain subject (and there are a few factors that determine which subjects work better than others), but the approach could generally be characterized as: leading the browsing patron from the inside-out of a subject. You start with a relatively specific topic, then browse outwards to very closely related topics, and up to more general subjects.

Organizing the same collection by very broad topics would lead the browser from a specfic book to more by the same author (assuming it is a bookstore-like system) then to other authors who write in the same broad category. By increasing the granularity of the subject-groupings, there are more opportunities for connections between works that would take longer to browse to in a more granular system and new-authors on a favorite topic might be found more easily through browsing.

Ironically, I think that this change would tend to increase reliance on the catalogue. In my experience, users aren't very good at finding everything they want in library catalogues (or on Google, for the matter) and usually end up selecting something that is "good enough". When a library patron finds good enough, she or he is going to then browse the stacks for related works that might also be useful.

If the granularity is too big, then finding a "good enough" book in the catalogue will not create any better browsing opportunities than just going to the broad subject area. If you have a fairly focused information need, "good enough" will no longer cut it. Each of the books on a more specific topic will be scattered throughout the stacks. If you do not have a relatively focused information need, then well, isn't one arbitrary system as good as another when you are browsing?

The Prelinger intrigues me because the unique arrangement has a point and a politics. The system is designed to lead the browser through a subject area in such a way as to teach something of the historical and intellectual ebb and flow of the topic. That's cool. I'm not sure it would work with a more general collection, though. Then again, with a broad and general collection, the staff could take turns re-arranging the collection based on their own biases, interests, expertise, sense-of-humour, or whatever.

But for Prelinger's scope and purpose, it is worth tracking, I think.

The Gilbert system just seems like lazyiness. Less effort by cataloguers. No value added for browsers or devoted searchers. Most public libraries that I've been to have compromised the granularity of DDC with signage. I'm not actually a big fan of DDC specifically, and would love to see an alternative make a run at DDC's ubiquity. But I also think that any system is better than no system when you are trying to accomodate a broad variety of information needs and practices.

(via Metafilter)

Cheers

Elimination Dance

1. Anyone who has touched a library book.




(credit: Ruben Bolling's Tom the Dancing Bug.)