!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: Clay Shirky on Sorting Content

Monday, November 16, 2009

Clay Shirky on Sorting Content

An earlier post discussed what the Internet guru Clay Shirky thinks about the future of newspapers. Now I've come upon another essay of Shirky's that provides food for thought for anyone interested in how tags are used to characterize and sort digital information.

Shirky's essay is based on a pair of talks he gave back in the spring of 2005, so when you read it, you'll find many of the points he makes are old news. What I found particularly useful was his analysis of the conditions under which a traditional pre-set taxonomy, such as the Dewey decimal system, is appropriate and, on the other hand, when we are well-advised to adopt a user-generated, free-form taxonomy.

Shirky argues that a traditional hierarchical taxonomy works well if the domain of information to be organized has these characteristics:
  • Small corpus

  • Formal categories

  • Stable entities

  • Restricted entities

  • Clear boundaries (i.e., no items that can arguably fit more than one category)
Shirky offers the Periodic Table of Elements as a prime example of a domain with the above characteristics.

Shirky goes on to list the characteristics that those participating in the catorizing must have in order for a pre-set taxonomy to be appropriate:
  • Expert catalogers

  • Authoritative source of judgment

  • Coordinated users

  • Expert users
Not surprisingly, Shirky argues for a user-generated taxonomy when the converse sets of characteristics hold:
  • Large corpus

  • No formal categories

  • Unstable entities

  • Unrestricted entities

  • No clear boundaries
  • Naive catalogers

  • No Authority

  • Uncoordinated users

  • Amateur users
The above characteristics fit the Web.

Shirky enumerates the advantages he sees in adopting a tagging system for information stored on Web pages. For example, one is able to develop an informed opinion of the likely value of a particular set of tags by paying attention to who did the tagging and how long ago the tagging was done.

Shirky's concluding thought is that "by letting users tag URLs and then aggregating those tags [across users], we're going to be able to build alternate organizational systems, systems that, like the Web itself, do a better job of letting individuals create value for one another, often without realizing it."