An interesting range of perspectives here in the article on tags as cognitive maps as well as in the comments posted to the article. One of the themes that stands out in the comments is the importance of seeing tags as a social phenomenon. While the cognitive or semantic model of tagging is a good description of how one mind might produce tag associations, even online there’s a social dimension to tagging cultures.
Tagging really is a good example of a dynamic system, a social system, even. As some comments note, we update our tags as we find others that are more useful/accurate/popular. So the tagging process involves 1) association and submission 2) cultural context 3) review, update, resubmit.
Tagging and tag folksonomies should each demonstrate a tendency, a refinement and closing in on a vernacular of sorts that repsresents a combination of 1) most effective associations and 2) most common associations. What you get is an effective and efficient form of speech. One that gives up poetic/aesthetic dimensions for precision. But because tags are also navigation (links to online content), the sacrifice makes sense. Navigation systems ought to be consistent. But who knew they could sustain so many labels.
(i’m reminded of a friend who once joked of making an actual scale map of the world. tags: an full scale map of language?)