On 7/9/06, Andrew Gray <shimgray(a)gmail.com> wrote: Letters to the editor are checked for authenticity by the newspapers printing them; in fact, the fundamental thing making a source reliable is fact-checking. Usenet posts do not have fact-checkers. Jay.

On 7/9/06, jayjg <jayjg99(a)gmail.com> wrote: I always find it mildly amusing -- and disturbing -- how conservative and, dare I say it, naive Wikipedia can be when it comes to comparing old media with the Internet. Every printed source is held up as a symbol of journalistic integrity, and a guy in Ohio hacking up a moderate article about some website 2 hours before the deadline is considered, per WP:WEB, to make the thing more notable than a hundred blogs with thousands of readers, because he's a "reliable source." Con artists like Stephen Glass or Jayson Blair must have been evil geniuses to pass through the firewalls of fact checking built by the New Republic or the New York Times. It turns out that when the New York Times actually writes about us, they do a crappier job at it than the average blog. But, they are printed on paper and sold on newsstands. It is almost like the physical manifestation of news on paper carries with it some magic, transcendental moment in the age of rapid, electronic distribution. Those who wikify the 1477th revision of an article about a minor pop celebrity look to the profession of The Journalist with the awe of devout altar boys. Take it from a published author and journalist: the process of actually taking an article from submission to print is much messier than you might think, especially in newspapers. Fact checking is often a luxury; as a writer, I had my articles messed up more than once by an editor who thought he knew what he was doing -- and my byline ended up on them. In one case it was so bad that I completely severed my connection to the newspaper, and had to publish a correction online. So, let's get back to the case of people posting on Usenet. If you have a group regular participating again and again in the same newsgroup, and you want to cite a particular post from that group that was _not contested_, you can pretty safely assume that it was posted by the same regular. Sure, you might be dealing with an elaborate hoax dating back several years. Then again, the newspaper you are citing might be subject to the very same kind of elaborate hoax. At some point, you have to start making _reasonable assumptions_. And if you want to verify that it was really written by person X, then take off your WP:NOR hat for a moment and send the guy an e-mail. Problem solved. The assumption that anything posted on Usenet is not reliable is not reasonable. The assumption that an _isolated_ post that does not match an established pattern needs to be treated with caution is much more reasonable. The assumption that, if an author contests the authenticity of a post, we need to believe them, is generally reasonable. And if you want to say "blog X said Y", then of course "blog X" is an excellent source for that. The question in both cases is more one of notability and relevance than one of reliability. What needs to stop is the blind worshipping of printed paper. Erik

On 7/10/06, Sarah <slimvirgin(a)gmail.com> wrote: Wikipedia has quite a lot of processes to detect errors, though it doesn't have the traditional notion of publication. Wikinews does, by the way. But let's not go there. The question is: a reliable source for what? A reliable source that person A said B? Person A tends to be a reliable source for that, and there are other indicators that he is indeed person A than some imagined process of verification. The more important question is whether the statement is relevant to the article, and whether it speaks for itself or is used to support an original speculation. If not, we need a secondary source that establishes the context -- but very often, a primary source is more than sufficient (and in some ways preferable). Our own history is an example. Virtually all of it is documented through electronic mailing lists and edits to the wiki. It is possible to fake mailing list posts just like it is possible to fake them on Usenet. So, are we going to dispute that Larry Sanger wrote the "Let's make a wiki" post on those grounds? Are we going to argue that a mailing list is not a "reliable source" and we need to wait until some piece of paper picks up the same information from the same source, with the added magic pixie dust of print, to turn it into an eternal truth? No, because if someone had faked Larry's post, Larry himself would have pointed it out quite quickly. The situation needs to be looked at in context, rather than reducing it to a simple formula of "reliable sources." We need to be very careful that dogma does not take the place of common sense. Erik

On 7/10/06, Sarah <slimvirgin(a)gmail.com> wrote: The processes used or not used vary nationally and internationally and on a case-by-case basis, and even large papers do not necessarily follow reliable processes. That is not the argument I made. Erik

On 7/9/06, Matt Brown <morven(a)gmail.com> wrote: Matt, many newspapers do make checks for every letter they're going to publish. The point is that you have to supply your details so they can check in principle. As for fact-checking processes, all decent-sized newspapers will have, first, an editor who assigns the story to a reporter and who may check it periodically while it's being written; then it'll be edited by a copy editor who'll look for obvious factual errors and legal problems; and then finally by a page editor. That's the bare minimum and that's only if no problems are picked up. If there are problems, more people are involved (senior editors, maybe lawyers), and many newspapers and serious magazines have much more elaborate basic processes. In addition, all these people are trained researchers/writers/editors. Mistakes in newspapers cost them money, sometimes a lot of money, so they're highly motivated not to get things wrong. We have none of those things. The only thing we have is a policy that says we have to use halfway decent sources who DO have them. Sarah

On 7/9/06, Matt Brown <morven(a)gmail.com> wrote: The San Francisco Chronicle called me back the first couple of times they printed my letters to the editor, and once after I moved from that address to a new one to verify that I was still me. Someone cares... -- -george william herbert george.herbert(a)gmail.com

On 7/9/06, Erik Moeller <eloquence(a)gmail.com> wrote: One problem with citing "blog X" when saying that "blog X said Y" is that the blog might very well not exist in a few years. A copy of a New York Times story, on the other hand, will certainly exist for many years. The other problem is one of original research. Sure, blog X might have said Y, but if that's an important enough point to be made in an encyclopedia article, then someone else will have discussed it. As I see it the purpose of barring original research is not so much to bar unreliable facts as it is to bar original arguments. It's to bar stuff like this, taken from the current [[George W. Bush]] article: "The Commission found no credible evidence that Saddam Hussein possessed stockpiles of WMD. On December 14, 2005, while discussing the WMD issue, Bush stated that "It is true that much of the intelligence turned out to be wrong."[35] However, Bush would remain unwavered when asked if the war had been worth it, or would he have made the same decision if he had known more." Anthony

On 7/9/06, maru dubshinki <marudubshinki(a)gmail.com> wrote: I agree that this would be useful, to the extent it's possible ("there is a 6-12 month lag between the date a site is crawled and the date it appears in the Wayback Machine"). Also, since IA can and will remove pages from the archive under certain circumstances (one of which is when a robots.txt file appears later), you still have to worry about an archive disappearing later. And of course this doesn't at all address the issue of original research - rather than proving something by referencing original sources Wikipedia generally (if not exclusively) should be referencing facts which have already been analysed by experts in that field. Referencing in Wikipedia is currently a big mess. The incremental tweaks seem to be making things worse in many ways. Anthony

[Apologies for the last, quotes-only, email - hit the wrong button] On 10/07/06, Anthony <wikilegal(a)inbox.org> wrote: Mmm. IA have made it reasonably clear that they're not there to compete with paid-for "real archives", if that makes sense - they're not going to be competing with the NYT for access to NYT backstories, which are our major need for links to transient pages. It *is* better than nothing - especially if we extend it to the national legal-deposit archives that will be slowly appearing - but I'm not sure if it'd be enough benefit to be worth the work. -- - Andrew Gray andrew.gray(a)dunelm.org.uk

On 7/10/06, Fastfission <fastfission(a)gmail.com> wrote: This is the crucial point. Someone can trust an article in the New York Times about as much as they trust the New York Times itself. And most people reading the Times would know how much they trust it based on past experience. That way they can go some way to evaluating the reliability of what they are reading even though they don't have the resources to check it for factual inaccuracy, bias etc. You can't do that with somerandomblog.com. You might be able to do it with blogthatswidelytrusted.com but they're few and far between. -- Stephen Bain stephen.bain(a)gmail.com