Actually it keeps what you do dirty and unreliable. We have an article, "Damned knowledge" which partially addresses this question. Political and ethnic considerations should play a minimum role is what is considered knowledge. That is one reason I defended User:Deeceevoice so strongly (although she undoubtedly considers me her worst enemy). I continue to maintain that should a village elder (or high school student) in Africa manage somehow to get on-line and write an article about their village we should be happy to have the input. Consider our article about Bonanza, Colorado, a decrepit collection of shacks in a barren gully in Colorado, who if you would believe Wikipedia, has an per capita income of $66,857. If I should add information from my personal knowledge that the reason for this astounding statistic is that one of the fourteen souls who was counted in the 2000 census has a nice income, but while he does not live in the town, other than brief periods in the summer, maintains his legal residence there so that he and his friends can vote there, in one of the tiniest incorporated towns in the United States, I would be adding unverified information. If I went further and speculated that the reason for the struggle for control of the town government was related to hopes regarding legalized gambling, I might be even further out of line. Yes, the line must be drawn somewhere, but an image of this desolate wasteland might prove very illuminating, despite being very much original research.
Fred
On Mar 2, 2006, at 9:11 PM, Fastfission wrote:
The philosopher Paul Feyerabend was known for arguing that the scientific method, as a method, was by definitive restrictive to what sorts of knowledge could be integrated into the scientific corpus. He was right, but I think he erred when he implied that this was not, in fact, the entire point of having a scientific method: it's a necessarily non-holistic form of knowledge, but it is reliable form of knowledge. You trade away some of your possibilities, but you get reliability and robustness in return.
I think it is an apt analogy in this case as well. A verfiability policy of our sort might lose the elders, but it keeps what we do include clean.
FF