Thursday, December 15, 2005

Thoughts on Folksonomy

Daniel H. Pink recently wrote a New York Times article on Folksonomy that has many a librarian/cataloger panties in a bind. Instead of seeing it for what it is, a useful way to organize collections of data for personal and public use, they see it as a breakdown of the system they hold so dear.

The most important part of that article is the following:
"People aren't really categorizing information," Vander Wal says. "They're throwing words out there for their own use." But the cumulative force of all the individual tags can produce a bottom-up, self-organized system for classifying mountains of digital material.

I say, power to the people.

Its easier than ever for people to share and organize information. I don't think I'm alone in the belief that the way libraries organize their stuff is confusing (Library of Congress classification in particular). I still have no clue why computer books are sprinkled throughout the stacks for no 'apparent' rhyme or reason. I'm not saying there isn't one, I'm sure there are a million rules about what goes where. I'm saying its not clear to a non-librarian/cataloger. I get that its not meant for browsing, but browsing is what most people want to do. Online and off. How many instances have we had recently where the cataloging rules posed an issue? That should tell you something. Libraries, the places that are built on the organization of information, are difficult for the average person to use.

Why not let the people that use the place advise you about the best way to organize it so that they can find what the need? People want easy and convenient. Now, for the first time, they are beginning to have a choice. That doesn't mean there shouldn't be any rules at all. I think you can have the best of both worlds if you just let people organize to suit their needs on the surface and figure out how to keep your rules alive and well behind the scenes.

That's just my two cents.

P.S. Library Thing

1 comment:

LibraryThing said...

Thanks for the PS!

I'm continually amazed of the disconnect between the way tags are presented--are they like, better than or worse than library classification schemes? how do normal people differ from professionals?--and the real-world examples cited--anything but books, and not always real people.

A number of art museums will not let me add tags to objects in their collection? This is exciting!? Who but museologists "used" museological classifications before this? Who but an exceptionally bored altruist would spend their time tagging items in an art museum? What would those tags really mean, compared to tags on or LibraryThing? Or take Amazon's new feature to allow users to tag the stuff Amazon sells. Why would anyone want to do that? It's like volunteering to fluff pillows at the local Sheraton. That and some tag-spam potential.

So, in this connection, your PS was much appreciated. LibraryThing isn't just the idea of a folksonomy, but the very thing. 17,000 book lovers have entered over 1 million books and 1.6 million tags, with minimal press coverage.

Do folksonomies work? This is no longer an academic question. There's some data. I'd love to find a Library Science grad student to look at it all. But I can see some early patterns. Tags fall in power distribution. People tag in different ways and for (seemingly different purposes). Even without "suggested tags," people's tagging drifts together. Tags "cluster" meaningfull, allowing good "related tag" browsing.

But do folksonomies improve finability? I think so. Clay Shirky predicted one way. The Library of Congress has no classification for "queer." But to a large number of people that term has significance and is useful for findability. Look for queer in LibraryThing and you books like More Tales of the City by Armistead Maupin. In the LC it's "Fiction," "City and town life fiction" and "humorous stories." ("Pardon me, Mr. Librarian, do you have any humorous stories about city and town life?" "Aesop? It's about mice? Thank you.")

The same goes for tags like "Christian living," "Cthulhu" and "dark fantasy." They mean something. They allow you to connect similar books and people who own books. And in each case, these connections would be largely opaque by Library of Congress classifications. Does it always work? No. I doubt that single-user tags like "vampire smut" will catch on and help others. But they work for that user, and they don't get in anyone else's way.

Lastly, LibraryThing pushes user cataloging way past tags. Regular people make authority control decisions! Are Clive Staples Lewis and C. S. Lewis the same author or different? Even the best libraries struggle with those issues. LibraryThing puts that decision in users' hands with a simple "combine" and "separate" link. Soon it will be letting users decide if "Huckleberry Finn" and "Mark Twain's Huckleberry Finn" are really the same. The value is obvious, and because LibraryThing users own their books and the whole social system is based on these agreements, they see real value in helping out.

I'm not sure we're heading to a user-controlled book utopia. LibraryThing still picks up LC numbers when it can, and I at least want them. But something's certainly happening...

Thanks for the post and sorry for the long reply.