While much of Wikipedia's information is correct, some is not. And sometimes people lie or are "vicious."

That's what First Amendment scholar and former USA TODAY editorial page editor John Seigenthaler says he discovered in a biography he found of himself on the site. It was riddled with errors, he says, including a statement that falsely insinuated he was party to Robert Kennedy's assassination. He was Kennedy's administrative assistant in the early 1960s, and his pallbearer.

He did correct the entry, as Wikipedia allows. But he also discovered that once information is put out in cyberspace, it's not so easy to pull it back. As he noted in a USA TODAY column last week, the error, which persisted uncorrected for four months, was repeated in other sites. It's unclear how widely the lie has spread.

Three important things to note from this:

1. He was able to correct a mistake found in a published article instantly. He could also instantly comment on why it was a mistake in the edit notes of the article.
2. If a similar mistake is published in print media, he would have no hope of correcting the mistake, and that mistake will be archived in libraries around the country for all eternity. If the mistake was posted on a blog, or an online newspaper, it could take hours or days before the mistake is fixed, assuming the site is willing to make the change.
3. The mistake was only propagated because people took the text of Wikipedia as researched fact.
4. Edits on Wikipedia are anonymous, and so it is an easy tool for spreading lies (unlike a blog which is personal and at least in some way able to be traced, or ignored)

Point three is what hurts John the most, and it should. It, with point 4, is also Dave Winer's main point. But there is nothing to prevent bad journalism, bad research and bad writing.

What the latest high profile problems have hopefully pointed out is that the internet research, or any research, is not for amateurs. One cannot blame just Wikipedia for it's ability to spread crap. Best have a good look at blogs too.

How is finding some text written on Wikipedia and reporting on it any different from finding text on a blog, or in a newsgroup, or some authoritative looking website? Is it the name? If I create a nice looking website called "authoritativeinfo.com" will I get my lies published?

The enormous benefits of point one and two should not be ignored. It irritates me that they are.

Wikipedia keeps a history of all edits made to an article. John is requesting the removal of the incorrect text from the history as well. I'd strongly argue against that. Leaving the history there, with comments (in the edit) and discussion in the article 'discuss' area, provides an explained edit path and reasons for the edit. It's a history of why that piece of text is wrong. Without the history, we have no record of why it was removed, why it ever existed. It becomes a rumour, and some begin to wonder why it was removed? Maybe it was true? Is there a cover up? The less archived history we have, the more power people have to deny and make it up.

Wikipedia recently made a change to require users to register before making an edit [cbsnews]. It won't stop everyone, but it should prevent a few stupid edits, and allow some hope of tracking serial jokers.