Bruce Sterling: We used to rely on philosophers to put the world in order. Now we’ve got information architects. But they’re not doing the work – we are.

There’s a revolution going on in the art and science of categorization, and its name is folksonomy, a term invented by information architect Thomas Vander Wal. Folksonomy is like taxonomy, the traditional way to impose structure on the blooming confusion of raw reality. For instance, the human being known as Thomas Vander Wal might be taxonomized as kingdom Animalia, phylum Chordata, subphylum Vertebrata, class Mammalia, subclass Eutheria, order Primata, suborder Haplorhini, family Hominidae, genus Homo, species sapiens. That’s accurate, and experts agree it’s true. Unfortunately, to make full use of this scheme, you need to be trained in com­parative anatomy, as well as Latin.

A folksonomy, on the other hand, arises spontaneously as Net users encounter information, think about what it means, and tag it with descriptive words. Then software makes the information accessible via a simple keyword search. The results aren’t definitive or scientific, but they can be very useful.

Consider Google: Type in “Thomas Vander Wal” and bingo! The search engine deftly tracks Web pages that link to the name, so in no time flat you can learn not just about him but about people who care enough to praise or condemn him online. Who did the heavy lifting? Certainly it wasn’t clever taxonomists navigating the Net’s white-water rapids of ephemera. It was a mob of interested people – folks – and the machines working behind the scenes that tossed in some technological onomy.

Folksonomy emerges from a combination of two inventions: (1) machines that can automate at least some of what it takes to classify information and (b) social software that makes users willing to do at least some of the work for nothing. You’ll notice that 1 and b don’t really go together. Folksonomy is like that. A pinch of free work and a peck of mechanical sorting will get you from 1 to b. Examples, which include the social bookmarking Web sites,, and, are proliferating.

The Flickr photo-sharing service harnesses the power of folksonomy to organize a mighty torrent of images flowing from the world’s digital cameras, phones, and PDAs. The principle is simple: It’s a drag to name or describe the zillions of private photographs you shoot each year, but that labor is a lot less onerous to people who like to surf snaps online.

Thus, Flickr breaks up the world into folksy categories that genuinely interest the online audience. In Flickrland, the world is composed of Architecture, Beaches, Cameraphones, Dogs, Europe, Friends, Graffiti, Honeymoons, and on and on. Nobody invented this scheme, and best of all, it’s an ongoing, democratic process. It’s a product of group interaction, like footpaths trampled across a virgin wilderness by a herd of bison.

A folksonomy is nearly useless for searching out specific, accurate information, but that’s beside the point. It offers dirt-cheap, machine-assisted herd behavior; common wisdom squared; a stampede toward the water holes of semantics. There’s room for scholarly smarts in this approach – for instance, you might invent a really cool term like folksonomy – but mostly, it’s a new way to crowd-surf. It’s as though you threw a kayak into a mosh pit and glided not just through Web pages but through labels, concepts, and ideas, too.

That won’t lead you to a specific piece of information, but it’s new and fascinating. What’s more, it’s native to the Web. Flickr has invented an eyeball-stickiness engine – a colossal photoweb slide show where you can glom your eyes onto whatever other people might call “ice” or “fire” or “sexy.”

It remains to be seen whether folksonomies will implode under the weight of immense numbers of users, or flame or spam out under the malignant attacks of free riders and rip-off artists. If so, the creation of unstructured taxonomies will likely sidle away from human users and toward machines; software will crawl over every image on the Web, count the tint and intensity of every pixel, pull in vague cues from the traffic statistics, and sort the mess. Computers don’t “know” what images mean, but they never give up, they work around the clock, and they’re not as crooked as people.

Ultimately no human brain, no planet full of human brains, can possibly catalog the dark, expanding ocean of data we spew. In a future of information auto-organized by folksonomy, we may not even have words for the kinds of sorting that will be going on; like mathematical proofs with 30,000 steps, they may be beyond comprehension. But they’ll enable searches that are vast and eerily powerful. We won’t be surfing with search engines any more. We’ll be trawling with engines of meaning.

More here.