Pensieri di un lunatico minore

Tagging versus categories

I have reached an odd crossroads. I am contemplating moving to tagging (a’la folksonomies), rather than the traditional categories approach of marking entries on my blog. Originally, I had started using hierarchical categories, and found them, like hierarchical folders in email, a limiting exercise in needless fussiness. They didn’t really add anything worthwhile to what I was doing. Then, after the creation of things like del.icio.us and Technorati, and my work with things like the Dublin Core I started to think more around the ideas of tagging and metadata, rather than structure and organization.

Tagging seems to me to be more about inter-relations, and not organization. It is a way to find similarity and a path through the maze. This appeals to me, and so I’ve started looking at migrating to a tagging-focus, rather than categories. Categories, however, at least as they are implemented in WordPress are little more than metadata tags, so the first thing I did was start tagging things with multiple categories and flattening the categories so they were one dimensional in nature. This got me part of the way there.

Now I’m thinking about using something like WP-Keywords to begin more extensive tagging of what I write. The trouble is, as with so many things written by other people, it doesn’t work the way I think. First, unlike Flickr, it uses commas to separate tags, rather than spaces¹, and seems to store everything as simply a string in the database. This means it’s harder to get things back out of it, and instead requires post-processing in one of my least favorite languages. This has forced me to rewrite chunks of the system, and will continue to do so until I get to the point of having a system that works the way I want.

For example, one of the functions in WP-Keywords allows you to specify the way the list is assembled and displayed. None of the others do, so I’m moving that functionality around so that things are more consistent. Also, in the mean time, I do not want my category choices to impact my tagging choices, even though they are comparable metadata.

Finally, I’m at a point of confusion on the use of case and capitalization in tags as well as how multi-word tags should be created. There seems to be no standardization, and everyone seems to do “whatever they like,” which is great, except it makes it difficult to fall into a pattern. Do you use quotes to distinguish multi-word tags (e.g., "Dick Cheney")? Do you slam the words together (e.g., dickcheney), or do you use CamelCase (e.g., DickCheney) instead²?

¹ This is something that annoys the hell out of me. I wish people would come to some conclusion about the de-facto method by which we enter multiple items onto a single line. Commas are my favorite as a delimiter, but seem to lose out to spaces most of the time.

² I’ve also seen people use it as two tags, which seems absolutely mystifying from a metadata perspective.