Folksonomy is an idea that frames some relatively recent practices in terms of some eminently established ones ("taxonomy"). This association between the recent and established practices can be seen as both a strength and a weakness in relation to the process of information (system) design.

The strength in the association between folksonomy and taxonomy is that folksonomies are an often useful topic of consideration when looking at the creation and maintenance of taxonomies. There are many kinds of information systems where there are at least some needs for an "-onomy of the people" that augments formal, centrally managed, taxonomies. And, in this context, folksonomy provides a nice step out from taxonomy and into the bigger conversation about the social life of information.

But, the weakness in this association is in all the ways that folksonomies are totally different than taxonomies—the ways in which you can't conceptually step out from a taxonomy to a folksonomy, and vice versa.

Here's how I think of the difference—first as a couple little formulae:

taxonomy = framework
+ systematic structuring
= systematic set of connections

folksonomy = framework
+ ad hoc labeling
= ad hoc / sometimes systematic set of connections

Or, spelled-out in a bit more detail:

Taxonomies are the result of establishing a centralized / hierarchical framework to support systematic efforts (to structure information) to achieve a systematic (and structured) set of connections.

Folksonomies are the result of establishing a centralized / hierarchical framework to support ad hoc efforts (to label information) to achieve an ad hoc but sometimes systematic (and sometimes structured) set of connections.

One way to dramatize this difference is: I can ask someone to print-out their taxonomy and show it to me on paper. But, it's impossible to print-out a folksonomy. A list of all tags is not a folksonomy. A list of just my own tags is not a folksonomy. A tag cloud, or any other weighted or sorted list, is not a folksonomy. In this taxonomy-sense, there is no such thing as a folksonomy! (Or, from another perspective: when you try to get something like a taxonomy out of a folksonomy, what you get out is, in fact, just a taxonomy!)

Think about the term "horseless carriage" as an early way to describe what we now call cars, trucks and automobiles. Prior to the horseless carriage, the horse is the defining "system" of transportation. Then, there's this thing with no horse that provides a means of transportation without that defining "system."

In terms of this analogy, folksonomies have no "horse." They work as information structures, but they aren't the result of the work (the "tax-" of taxonomy) structuring information—rather they are the result of "folk." So, from the point of view of information structures, folksonomies can't work because there is, in some practical sense, no "horse" / no "tax."

Thomas Vander Wal's recent post on Understanding Taxonomy and Folksonmy Together got me thinking about this. I'm finding it useful to see the association between taxonomy and folksonomy as something like the association between horse and horseless carriage.

Continuing with this analogy, it's worth noting that, in spite of the success of the horseless carriage, not only has the horse and other pack animals continued to provide a significant system of transportation worldwide, it's easy to imagine that horse-based transportation might long outlive the popularity of the car.

Coincidentally, this ties in well with an article about the history of technology that I highly recommend: What Else Is New?, by Steven Shapin, writing in The New Yorker (via rebecca's pocket). The article has a relevant section about the ongoing importance of the technology of horse-based transportation. Here's a good quote about the use of horses during World War II:

Long past the age of steam—and well into the age of automobiles and aviation—the power of horseflesh remained critical. In the Italian campaign alone, the United States Army's 10th Mountain Division used more than ten thousand horses and mules, and the great tank general George S. Patton wished he'd had many more. . .

One could go further in the application of this analogy, e.g., recognizing how folksonomy is largely dependent on specific web system technology (like the car is largely dependent on the technology of paved roads), while taxonomy enjoys a portability across a diverse array of system technologies (like the horse / pack animal is adaptable to a variety of terrains).

But, the main point I want to make is that it's important to be careful when explaining folksonomy in terms of taxonomy (or, vice versa), because that is something like trying to explain the horseless carriage in terms of the horse (or, vice versa).