My friend Jon Dron has finally nailed his own (and no doubt others) ideas about the collective nature of Wikipedia. His recent post notes:

  • the individual actions that create most of the articles,
  • the groups of administrative types who manage the overall infrastructure and set in place the algorithms that manage the look, feel and performance of the system
  • the networks of mostly regular users responsible for maintenance and collaborative development of the articles and finally the way we mine
  • the wiki as a collective resource.

I realize that some folks think this task of dividing and allocating ideas into categories is an arbitrary function that just gives rise to arguments (see for example Dave Snowden’s diatribe and his focus on ‘crews’.)

However I think our ‘Taxonomy of the Many’ classification system has value and defend it and classification systems in general in the rest of this post.

The development of a model such as the one we have just described demands a thoughtful rationale. The emergent Net has been referred to by a number of scholars as defying distinct classification. Dave Weinberger (2007) in a thought provoking history of classifications systems – from botanists. to chemists, librarians to biologists, argues that “everything is miscellaneous”. By this he means that information, ideas and objects can be classified in almost an infinite number of ways, by an infinite number of user-organizers in infinite number of contexts. Thus, the good classification systems (if necessary) emerge both from bottom up folksonomies and authoritative, top down classification systems. Each organization schema can be operationalized and continuously be adapted by agents, creating an infinite set of useful and personalized organizational schemes. The power of machine sorting and aggregating can then be used.

But these forms of highly granulated and personalized searches, are dependent upon a conceptual model. Searchers must know the shape and dimensions of information if they are to be able to see its application in their contexts and then know how to retrieve and apply that information thereby creating knowledge. Visual and verbal models can be used to provide a graphical ‘big picture’ useful in providing mental models of new applications. These then provide guiding heuristics that can be used both independently and integrated into the other applications within the personal learning environment.

When developing a classification system it is often difficult to know how finely to divide categories. Wikipedia defines a “lumper” as “an individual who takes a gestalt view of a definition, and assigns examples broadly, assuming that differences are not as important as signature similarities. A “splitter” is an individual who takes precise definitions, and creates new categories to classify samples that differ in key ways.” So I guess Jon and I could be described as lumpers, though we resist lumping everything on the net into the miscellaneous category.

Weinberger (2007) also points out that Carl Linnaeus, the famous Swedish inventor of the Linnaean taxonomy for biological organisms “seemed comfortable with the idea that he had spent his life devising an order that was useful, if not true”. P. 73. And of course the Linnaean system was developed before Darwinian notions of evolutionary provenance and even later genetic classification of species.

A continuing problem with taxonomies in general and especially of socially defined entities has been labeled as Sorites Paradox It notes the challenges of deciding just when an item moves from one category to another. At what height does a man become a “tall man” or even more contentious at what weight do I become fat?

We’ve tried not to get too hung up on these details nor requiring that categories in the taxonomy be rigidly exclusive. So Jon’s description of Wikipedia as a Net resource that operates at all four levels of our Taxonmomy scheme, doesn’t particularly both me, and I still argue (and hope that you) find the taxonomy of some conceptual and practical value.