On 3/14/07, Andrew Gray shimgray@gmail.com wrote:
talk-page was that I'd turned up a letter to a newspaper from a relative of the subject which mentioned, in passing, the correct deathdate - it seemed fair to accept that where sources differed, going to a source as close as possible to the subject seemed the most accurate.
If you've got two sources saying different things, NPOV generally encourages you to cite both. I did something like this in [[Kate McTell]] where half the sources I used claimed that Ruby Glaze was a pseudonym, and half said they were two distinct people. IMHO it's much easier to convince the reader that you're right if you cite both versions and explain why one is more likely.
I really don't see anything wrong with me footnoting a) as "was born in Such-and-Such<ref>Personal correspondence with the Wikimedia Foundation, June 17th, reference ABC1234567</ref>. Yes, we could ask them to issue a rather dull press release, or write a blog post, or (in one case I recall) update the details on their myspace page. But no reasonable academic or reporter objects to incorporating corrections of trivial, non-contentious details from those who know about the article; why should we?
Well, the obvious problem is that a future editor is not going to know the background to the problem, and, following policy, will probably reinstate the "public record" version. That's why "private correspondance" just isn't good enough in most cases. It's just not durable enough.
Steve