The chat at indieweb.org has been quite busy lately with people brainstorming about taxonomies, specifically about how to classify places that serve coffee. This isn’t something I am interested in, although the broader topic of classifications and taxonomies I do wrestle with from time to time. What triggered this post is the discussion over there that focused on emergent versus deliberately designed taxonomies.
My own preference is definitely for emergent taxonomies, where the words you use to classify things are free to grow and change. My problem is that I can never remember previous tags or categories that I have applied to things, which means that often the tag I apply is almost but not quite a similar tag. And that leads to a profusion of tags that is less than helpful. For example, over the years my bookmarks have accumulated, at the last count, 587 different, unique tags. And of those, 213 are attached to only a single entry. Pull in those attached to two items and the total climbs to 317, more than half the total number of tags.
That’s pretty absurd, and for a brief moment I entertained the fantasy that one of the large language models might be able to bring some sense to the problem by suggesting ways to merge adjacent-meaning tags. Pure fantasy. The best it could do was to suggest sticking to either singular or plural, which is very sound advice and which I generally fail to do. Standardising on capitalisation would be a good idea too.
I’m left not much further forward than when I began. I have a list of all the tags I have used. I could focus on the scarce ones and try to find a better fit. I could start with the most popular — far and away the one I use to save items for my newsletter — and try to be add more useful tags. I could bite the bullet and admit to myself the fact that I seldom search by tag, that most of my bookmarks are the result of the collector’s fallacy, and that my problem results from that fallacy coming home to roost.
Whatever I decide, the task seems overwhelming; I have around 10,000 bookmarks. Editing the tags online racks up costs in Pika. Editing them offline risks quickly getting out of sync but may be the best way to go. What I really want, of course, is for the bookmarking app (the wonderful linkding) to show me something like a dropdown list of tags when I start saving a new bookmark. That is probably really difficult. In the end, I’ll probably just keep muddling along.
As far as tagging photos goes, I don’t have nearly as much of a problem because Photo Mechanic makes it relatively easy to stick to a deliberately designed hierarchical taxonomy that is also reasonably easy to extend. Best of both worlds.
Two ways to respond: webmentions and comments
Webmentions
Webmentions allow conversations across the web, based on a web standard. They are a powerful building block for the decentralized social web.
“Ordinary” comments