On 7/10/06, Sarah slimvirgin@gmail.com wrote:
Erik, the important point about newspapers is that all but the tiniest have processes in place to detect errors, and particularly legal problems, prior to publication. We can only hope they use the processes correctly; if they don't, that's not our fault. But Wikipedia has no such process, which is why we rely on what we call "reliable sources" who do.
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