As any good librarian knows, classification is a hell of a tough job. Clay Shirky puts it well - you have to be part mind reader, part fortune teller. No matter how clever a taxonomy of subjects you come up with, second guessing search behaviour let alone the future development of new topic areas makes the job near impossible. Dewey had it easy.

Thats why the growing phenomenon of 'tagging' is so interesting. The latest buzzword in the space is 'Folksonomy', which describes the classifications that are generated not by overeducated editors, but by readers themselves. Folksonomy is a bottom up approach to classification which powers photo colloboration sites such as Flickr, or the social bookmarking service del.icio.us. People classify their pictures/bookmarks/web pages with tags, and then the most popular tags float to the top.

Tagging is a powerful intersection between two of the web's prevailing search philosophies - human editing and organic search results. In the early days of search, Yahoo took upon itself the task of classifying the web with informed human editors. Google's timely riposte was that the only way of dealing with infinite amounts of content was an automated solution based on a number of factors including link popularity.

While effective, Google's search results don't solve the problem of classification. Sure, its great if you can use a search engine to find a picture of what a sun cancer spot is, especially if you are a hypochondriac with a passion for self diagnosis. But was that actually the best search term to use? What if all sun cancer spot photos were tagged allowing you to view them all in one spot?

Using del.icio.us's tagging search function in that respect is a revelation. Its not just about good search results. There is a classification structure to the chaos, and like many of these new technologies changing the face of the web, it is emergent.