Part of what I'm asserting though is that it would be generally useful to people, because they may be filtering
for what others are filtering
against. Say that you have a phobia of frogs. You can filter out frogs, but someone researching them can filter for them .
I mistook how AO3 works. I guess I'm thinking more like Tumblr or Flickr. Say it isn't centrally organized. We're just tagging them likes we sees 'em. So someone creates the frog tag because they want to use it and it doesn't exist. Then they tag something because it has a frog. It never crosses their minds that someone may have a frog phobia; the filtering is done by the frog fearer themselves, so it isn't disclosed to bullies who might use it against them, or anyone else they want to keep it from for reasons of their own privacy.
There's still problems here, one person may read `frog` and another `frogs`, mistakes can be made, things can be mistagged deliberately, etc. but I think there's promise to the approach.
If you have resources to recommend on library science, I'd be keenly interested.
