Thanks for this thoughtful comment. I agree with you that folksonomies and taxonomies will coexist; that there are areas where one is more useful or relevant than the other. Thanks also for articulating some of what I was trying to get at, especially when you wrote about the discussion in favor of tagging has taken place in an intellectual vacuum.

As I noted to you separately in an email, I would like to take you up on your offer to somehow utilize what you’ve built for academic research.

I think Web 2.0 boosters—I am excluding myself—too often pretend to a superiority over “traditional cataloging” that is under-informed an unearned. The web is, of course, full of boosterism and ignorance, and the din of this can, I imagine, set a librarian’s teeth on edge.

But these tendencies are hardly restricted to the worst stuff. Take, for example, Clay Shirky’s “Ontology is Overrated,” in my mind the most intelligent and persuasive argument for tagging against “traditional” classification. About half the talk is devoted to the shortcomings of the Dewey Decimal and LC Call Number systems: They are limited to one subject because a book can only live one place on a shelf; they are inflexible because renumbering millions of books would be impracticable; they are as non-objective and impermanent as tags because Dewey allocated more numbering space to Christianity than to Islam, and so on and on. It’s all true to some degree, but too full of cheap shots and lacking in context. Your point about the Web 2.0 assertion that librarians never considered users’ needs is well taken. Nor do things come out quite the same way with LC Subject Headings. Worse, the whole argument for tagging takes places in an intellectual vacuum, as if librarians hadn’t been talking about these issues for centuries, and as if librarians and Del.icio.us users were trying to solve the same problem, and librarians were just worse at it.

As you note, presuming your point and proving it are not the same thing. SOMEBODY has to test this stuff. On balance, I’m excited about folksonomies. Even without rigorous testing, I think it’s clear there are SOME domains in which they make more sense than traditional cataloging, or, for example, full text searching. But is that true for libraries? I don’t know. I’d like to see some data before making up my mind whether, as you put it, tags are a “better way of organizing information broadly.”

In my gut, I think the answer is that the question is misframed. There is no “best” or “better” way of organizing information. Organization is not a end but a means. “Better” for what? For whom? Under what circumstances? When the dust settles I expect tags to take their place alongside other ways of finding stuff—good for some things, worse for others.* Other new ways, like full text searches or “people who own X also own Y” statistics of LibraryThing, will also be there, how prominently we can only guess. In any case, professional cataloging will not go away.

Lastly, let me put in a personal plea. Won’t SOMEONE take me up on my offer to put LibraryThing data at the service of academia? LibraryThing has 2.6 million tags—FAR more than Amazon or any other book service. And they’re about BOOKS, not websites, restaurants or snapshots. Isn’t there some enterprising library science grad student who wants what I got?

*Footnote on useful and un-useful tags:

As Shirky argues, no library has a “queer” classification, yet “queer” means a lot to a lot of people. Armistead Maupin’s _Tales of the City_ is one of LibraryThing’s top “queer” books (http://www.librarything.com/catalog/24748). The LC subject headings are “City and town life” and “Humorous stories.” In this case, I think tags come out on top. Nobody searches for “city and town life” “humorous stories”—not unless they want the City Mouse and the Country Mouse. Users’ tagging here also produces some good reading suggestions.

But there are just as many cases where tags fall down. For example, I use the tag “magic” to apply to my collection of academic monographs on Greek and Roman magic, with some cross-cultural comparanda. On LibraryThing, once I leave my own catalog, magic is entirely taken over by fantasy books.

There might be some algorithmic way to mitigate the context-less nature of tags—for a program to see that magic “clusters” in different ways—but the core problem will never go away. The advantages of tagging—brevity, spontaneity, low barriers to entry, etc.—are also its disadvantages.

That’s certainly the conclusion I arrived at:
http://cites.boisestate.edu/v6i4a.htm