[Clay Shirky][1] was the first to [come up][2] with the idea of “augmented tags”.

It always takes time to fully grasp the content of Clay Shirky articles, especially when they try to summarize a complex debate as ontologies vs. folksonomies. In his response to Gene Smith, Clay shows no doubts:

Tags are a useful, even vital, tool for anything that can be referred to with a URI. Which is a pretty considerable subset of everything.

He is also quite certain about the final result of this match:

People who believe that tagging will co-exist peacefully with classification schemes have underestimated tagging.

And the reason to make such a strong statement is not ideological but economic.

Categorization systems favor stable categories not because the world is stable but because categorizers are busy and their time is expensive.

For all the many differences between tagging and classification, the key one is cost. It is simply too expensive to hire professionals to do the work once a system that uses peer production is also available.

I think this is true and applies both to professional and amateurs categorizers. Rearrange an ontology in order to be coherent with the reality it tries to describe is a a big hassle not just for the librarians of big institutions but also and especially for amateurs involved in organize their weblog posts or their pictures. Unless they use a tag enabled blogging software or Flickr!

In the final part of the post Clay is even a bit cruel with Gene, easily dismantling all of his three examples (Amazon, Wikipedia, Yahoo) aimed to demonstrate how classification is sometimes the only right answer.

I was completely new to the XRI (Extensible Resource Identifier) syntax, but then I came across this post from Drummond Reed’s weblog on “open tagging”.

During the last 3 weeks I kept pondering about the tag molecule idea and ended up disagreeing with most of my previous post about it. The core idea was to enrich every tag with an explicit declaration of a “dictionary authority”, i.e a third party tag space. That would improve the semantic of the tag assigned by an author and the effectiveness of tag queries.

Besides the problem of consistently coding all the components of the tag molecules both in the HTML page and in the RSS feed, I was struck by my own example about the Ferrari owner. To set up that example, I had an hard time trying to find a good tag space focused on the automotive sector. No tools are available today for searching among tag spaces. Moving quite randomly I found the autoblog.nl, but to be honest I was partially satisfied because of the language: how possibly can our english speaking Ferrrari owner perform a search in the english blogosphere choosing a dutch dictionary?

Tag spaces are the results of the collective tagging efforts of individuals (the tag spaces of weblog) or groups (e.g. the del.icio.us tag space). Some of them are very large and fast evolving, others are small and focused, etc. But which is the right one to refer to for every one of my searching or tagging needs? I’m afraid that this approach could be theoretically sound but actually impractical …