Are folksonomies useful?

The simple question prompt for this post is deceptively complex, because besides defining what a folksonomy is, we have to define what usefulness is. Useful for what purpose? For whom? Why?

If we define folksonomies as systems of categorization created by lay people with no specialized vocabulary, the answer is a little different than if we define folksonomies as systems of categorization created by users of the information. Both definitions are correct.

Obviously, folksonomies have many flaws that critics are quick to point out. With no controlled vocabulary, one must simply guess at keywords and hope someone has the same ideas of vocabulary terms. With anyone allowed to tag, spammers are likely and expected, blurring the clarity of the concept cloud for any given subject. Yes, there are flaws and weaknesses. So do all systems for information aggregation. The real controversy, I feel, is that they are the OPPOSITE flaws of other typical search strategies, which is what makes them scary.

Traditional controlled vocabulary solves the problem of inconsistency, but at the expense of potential elitism, wherein only those selected "in the know" searchers will understand how to utilize the system optimally. This phenomenon can be eased somewhat by the use of tools such as search term redirection and taxonomies of controlled vocabulary terms. However, the knowledge elitism of controlled vocabulary can no more be fully eradicated by these features than the opposite flaw of variability can be fully eradicated from a folksonomy. They are inherent features of the system.

Likewise algorithmic searching has its flaws, as computer controlled functions, no matter how complex, necessarily leave out some of the connections that can only be made using direct human insight and intuition. The opposite flaw of folksonomy is that it is overly human, and thus subject to spam, variability, and the worst flaws of direct human usage.

In other words, different types of organizational schemes can all be useful, given a certain situation or potential usage. For some knowledge gathering, only a scientific search can do, and algorithmic function works particularly well. For other types of organization and searching, an aggregated controlled vocabulary can be the most useful, because some concepts are easily definable and broken down into easily manageable terms and subdivisions. Other searches, however, may be far easier using a method more in line with human intuition and variability. Some types of knowledge simply work better that way.

I believe that the more different types of search options are available on any particular data set, the more likely a searcher will be able to effectively utilize one of them. Likewise, some data sets, by their nature, require different organizational schemes, and folksonomy can be one solution to oddball data sets that do not fit into the more rationalized structures of controlled vocabulary, and are not easily found through algorithmic searching.