Never say never: I didn’t hold them as a reference but I was evidently wrong. A colleague has just pointed me to a recent (July 2007) paper published by Oracle, and which can be found here, that discusses the concepts of tagging and folksonomies. It bridges (quite well) the gap with pre-web 2.0 taxonomy and ontology concepts, at the same time explaining why the new methods, properly used, can be an improvement.
The document includes almost granular details on how to build them with Oracle’s ECM suite, but (unless you’re actually interested) skipping those parts doesn’t detract from a solid explanation of a very relevant concept.
Kudos where it’s due: it’s a hard bridge to explain, and they’ve done it. IMHO.