On 12/19/06, jayjg jayjg99@gmail.com wrote:
Actually, that's making things worse; we're now using weasel words ("appears that", "probably is") to cover our original research ("the consensus is"). If we only have primary sources, then we say what the primary sources say. If we have a secondary source that says "the consensus of legal opinion is", then great. If not, then *we* cannot become that secondary source that draws that conclusion, partly because we're not experts in this field, but mostly because it's original research. Remember, the second you say "the consensus is", you need to be able to state exactly who considers there to be a "consensus". Otherwise you are "introducing an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or synthesis to a reputable source."
Are you trying to define the pinnacle of perfection that every article should strive for but ultimately very few will achieve? I'm trying to find a workable compromise that will reduce the worst problems of OR and alert the reader to weaknesses in the article.
There's a hard line and a soft line on a lot of this stuff. Pick a random article on Wikipedia. Does it meet WP:V? Probably not. Do you rip it to shreds, moving all material to the talk page pending sources? Or do you use common sense and only remove material for which lack of sources is going to be harmful to someone?
If we actually have the tertiary source that tells us that there is consensus among secondary sources, then great, we'll obviously cite it. But if we don't, we're doing our readers a disservice by avoiding any mention of the general trend that we've found in our research. By contrast, we help our readers if we can convey "It looks to us that most researchers agree with X, but don't take our word for it".
Steve