The Emerging Battle in Meta-Data

One of the problems with consulting is that I often have to miss really interesting conferences in order to be responsive to my clients. Of course, my clients are pretty interesting as well – my problem really is that I want to be in too many places at once – which is probably a good segue to the topic at hand.

If you work in the world of ideas or information, one of the biggest challenges is how to organize those ideas or information.  Ideas and information rarely fit into neat categories, yet we have long favored trees and branches as a way to classify and organize ideas and information.

David Weinberger, one of my favorite thinkers on the new worlds we are creating, has written a fascinating essay in the February 2005 issue of Release 1.0 on alternative ways of organizing ideas or information – it’s called "Taxonomies and Tags: From Trees to Piles of Leaves".  Unfortunately, the full essay is only available here for sale, but David was able to get permission to produce an excerpt of the essay here for free.

In this essay, David focuses on the notion of tagging as a particularly promising alternative way of organizing ideas or information. A variety of interesting new businesses are emerging on the Web using tags as a key way to add value for users – see, for example del.icio.us for organizing bookmarks and Flickr for organizing photos. David is especially interested in the opportunity to create self-organizing taxonomies through tagging – these new forms of taxonomies have been described as folksonomies. Although tagging can have enormous individual value, the real power and opportunity may come from the use of tagging in a social context to help each other find ideas or information more easily.

Which brings me back to missing conferences – specifically, O’Reilly’s Emerging Technology conference in San Diego last week where Clay Shirky, another favorite of mine, spoke on "Ontology is Overrated: Links, Tags, and Post-hoc Metadata" and David led an informal evening discussion on "From Trees to Tags".

I am particularly interested in this topic both because of the interplay between the individual and the social and because I see it as one more manifestation of a much broader trend playing out in multiple domains – the shift from push to pull mechanisms in organizing and mobilizing resources.  Traditional taxonomies are designed in advance by experts and work best when there is a lot of material that is relatively unchanging and that has clearly defined edges.  The material itself may be pulled by the user when needed, but the taxonomy is pushed from above.  Tagging operates differently – it enables an emergent taxonomy, developed by users themselves at the time of use, it tolerates enormous ambiguity, accommodates rapid change and provides much more flexible ways of sorting on demand.  In other words, users can "pull" taxonomies – create and modify them on the fly to serve their unique needs, but enhanced and amplified by the efforts of others.

As tagging moves from a purely individual way of organizing meta-data to a more social phenomenon, power becomes a more central issue.  Traditional taxonomies like trees clearly embody power relationships, as eloquently discussed in Geoffrey Bowker and Susan Starr’s book, Sorting Things Out. Matt Locke has a great post extending this notion of power relationships in traditional taxonomies to emergent folksonomies. I have an intuition that the key to scaling folksonomies and making them useful to a much broader set of users will hinge upon sorting through these power issues. I also have an intuition that there are some business opportunities here in providing mediation services . . .

Fortunately, I am in Scottsdale, Arizona and looking forward to PC Forum, where I will hopefully have a chance to pursue some of these topics with David Weinberger.  At least this is one conference I get to go to . . . since I am speaking there, I have a good reason to hold off client meetings until after the conference is over.