Tagging bubbles in a sea of spam
I think the debate is partly hampered by the fact that everyone tends to conflate two very similar but distinct issues. Clay Shirky says the most important characteristic of cheap metadata is that “it’s made by someone else,” but a lot of the recent discussion has focussed on self-classification.
In either case my own thinking is that tagging and self-classification will both survive breaking onto the open web and the spamming that will ensue. I do not believe that the practice is intrinsically resistant against malicious intent, and I do believe that it will not be of much use on the open web. However, it is provenly useful in closed spaces. And because everyone will carve their own closed space out of the open web, tagging will remain useful.
The trend already exists: there is spam in my email inbox, and there is spam in my Google search results, but there is no spam in my aggregator. My blogroll is a closed space under my control and contains a self-selecting sample of content producers who have no interest in spamming their audience.