Phil Sandifer wrote:
Consider this another entry in that time-tested genre of "obviously futile suggestions to nuke things that nobody is ever going to nuke, but probably should anyway" posts. (The classic, of course, being Nuke AfD. Which we should still do.)
We should nuke [[WP:CITE]] and [[WP:RS]]. They are not working. They have never worked. It is not a feasible project to go and add sources to everything, and new contributors who are editing casually are never going to be willing to do the extra work of having sources. The result is a rule where we are always going to be playing catch-up.
Nor do the pages prevent incidents like Siegenthaler, which was a problem with exactly one cause, which is that nobody had ever actually looked at that page after it was created. No policy in the world will fix a page that nobody is editing.
Yes, we need to ensure that people do not add crap information. This can be covered easily with "Information that people doubt the validity of should be sourced." And we can then leave the community to deal with issues on a case by case basis with the direction that they should be careful to make sure that information is accurate. And we should shoot people who continue to add dubious information over the objections of other editors. Which is basically how we wrote an encyclopedia that has proven pretty trustworthy, and, more to the point, is how we actually operate now on the vast majority of our articles, since [[WP:CITE]] and [[WP:RS]] are not actually useful pages.
But to have a pair of policies that cannot be honestly implemented serves only one purpose: causing debates among editors that waste time and good faith.
Nuke them.
You know, having read this through, I do believe I agree.
Mind you, we do need _something_ to fill the policy vacuum that nuking them would leave. I think your "Information that people doubt the validity of should be sourced" is a good starting point. Add to it "Information that could be libelous if false must be sourced", to cover the BLP angle, and we're starting to get somewhere.
Of course, this current drive towards sourcing everything to the hilt does have some positive aspects; we really _could_ use a lot more sources in a lot of articles. It's just the common sense "the color of the sky on a clear day does not need sourcing" aspect that is missing. Of course, reasonably interpreting the current policies also achieves this goal quite well -- at least until a non-reasonable person shows up.
As for the Seigenthaler incident, I still believe it was and is fundamentally a technical problem. We simply shouldn't have live articles that at least two or three established users haven't checked. Some combination of patrolling and stable versions might do it, although over time I've started to think more and more that it really needs some other, truly out-of-the-box solution.