Philip Sandifer wrote:
On Apr 10, 2008, at 1:48 AM, Jussi-Ville Heiskanen wrote:
A sad case in point being the scientologist driven campaign to shut down The anon.penet.fi server. The Observer was spoofed by them into reporting that 90% of child porn on the internet is trafficed by the server. Even though The Observers own readers wrote in in huge volume that that was quite preposterous, espescially as Julf had put severe restrictions on the size of emails that could be sent through penet, and yet, The Observer never admitted it had erred, but merely pedaled down the story by noting that "Johan Helsingius had consistently denied the allegations." Which is a very poor form of apology for getting the story wrong.
To my mind, this settles the issue. There are clearly instances where accuracy and truth are pre-requisites for material. We simply do not report absurd slanders like this. Anybody who does not understand this should find a different project.
That still leads us to the initial question---how do we determine accuracy and truth? Generally, we determine it by consulting and citing sources, rather than doing original research ourselves. For example, if through consulting archival documents I determine that the standard attribution of some ancient Greek poet to the 4th century BC is actually incorrect, and the truth is that he lived in the 2nd century BC, it would still be appropriate for Wikipedia to report that he lived in the 4th century BC, unless I get my new estimate published in a classics journal first.
The problem with this seems to arise mainly with recent things where good sources don't actually exist. In that case, I'd argue it's not an issue of us doing insufficient original research, but of placing too much trust in marginally reliable sources. A single report in a newspaper is only a very marginally reliable source, so if a particularly surprising allegation has only the support of one newspaper article, maybe we shouldn't report it at all, unless it itself becomes a newsworthy allegation where other sources start reporting on the controversy, or is corroborated by other sources.
This is distinguished from sources like a full-length biography of a person written by a respected author---especially in the case where a person has multiple full-length biographies available, I would argue that the article *must* always be based exclusively on such sources, not on a Wikipedian's original research into "accuracy and truth". But where no such sources exist, some judgment in excluding probably-wrong allegations with marginal sourcing can work.
-Mark