On 10/7/05, Jimmy Wales <jwales(a)wikia.com> wrote:
I'm a big fan of eventualism. But Bill Gates and Jane Fonda are not new
articles, nor are they difficult or obscure subjects. Nor are the
problems I'm _currently_ concerned about with these articles problems
resulting from a lack of knowledge. They are stylistic problems which
are pretty awful.
I think you've put your finger on the problem there. The mechanisms by which
Wikipedia articles improve and degrade are not fully understood, and there
may well be some complex factors that we don't consider seriously enough.
But we can make a stab at guessing the main factor: entropy. An article that
is edited frequently by many people will probably degrade quickly. An
article that doesn't start out with a firm structure will not improve until
a single person or a tightly organized group imposes such a structure. An
article that is part of a series of similar articles that are edited by the
same group of people will tend to have a structure similar to the others in
the series.
We can help people to improve articles by preserving examples of articles
which are the best of their class: exhibiting superior structure, subject
coverage, and writing style. We can give recognition to editors who perform
extraordinarily good improvements on articles. I think we already do some of
these things.
One idea I suggested a while back, half jokingly, is that it should be
possible to make a group decision to revert to an earlier version of an
article. Articles do degrade and it may sometimes be a good idea to
recognise that the overall effect of recent edits has been to destroy what
was good about the article.
Today I summarily removed the Featured Article status from an article,
[[Iraqi insurgency]], because it's now such an unholy mess, with two
competing versions each pushing a point of view, and full of unreferenced
statements and opinions represented as fact. And yet that article survived a
move to remove its featured status as recently as June, 2005. Assuming that
the earlier version really was better than the mess that exists today,
perhaps it would be as well to revert to that version and to find ways to
incorporate later work on this current event without degrading it.