Wikipedia talk:When sources are wrong

Soft launch

So this is an idea that's been bouncing around my head for months. I don't think someone's done an essay on this before? If so, oops, there goes 4 hours of my life. But if not, well, I'm hoping this is of some use. It's a whole lot of thoughts and I'm not sure it's where I want it yet. @EEng, Tryptofish, Epicgenius, Vaticidalprophet, theleekycauldron, and Guerillero: Y'all are some people I thought of while writing this, either from involvement in examples mentioned or from past discussions. Ideas for improvement and further examples very welcome. -- Tamzin^{[cetacean needed]} (she|they|xe) 20:42, 30 July 2023 (UTC)[reply]

Have skimmed to about ~70% comprehension; will look in greater depth, but exciting essay! Biddenden Maids does the "footnoting" variant, in its first footnote. I've used it as an example a few times, as an FA and therefore as theoretically "best practice" for its version of the issue. Vaticidalprophet 20:45, 30 July 2023 (UTC)[reply]
More than worthy. Some time ago No such user pointed me to Talk:Bijeljina massacre/Archive 4 as a useful parallel to the Grant dilemma, and I've been meaning to bring it to your attention maybe you can use it here. I'm still mired in some IRL stuff but will keep an eye here in case I can inject some superficial, paternalistic comments from time to time. E Eng 20:55, 30 July 2023 (UTC)[reply]
Another interesting case was Talk:Eubie_Blake#Blake's_Correct_Birth_Year_-_1883_-_is_Provided_by_Blake_Himself_on_"The_Tonight_Show_with_Johnny_Carson": subject had been giving his own birthyear as 1883 for at least 50 years; countless sources during that time repeated 1883 as his birthyear; between 2000 and 2015, new high-quality sources showed that 1887 was unquestionably the correct year; nonetheless, even today plenty of sources keep unthinkingly publishing the defunct 1883 date. E Eng 23:43, 30 July 2023 (UTC)[reply]
Thanks for the ping, and I think the essay will be very useful. I've also done only an initial read so far, and will give it more thought in time. My first thought goes to the idea of "preponderance of reliable sources", as the way that we decide which source is wrong. --Tryptofish (talk) 20:57, 30 July 2023 (UTC)[reply]
@Tryptofish: I would go with "preponderance of the highest-quality reliable sources" myself. -- Guerillero ^{Parlez Moi} 21:55, 30 July 2023 (UTC)[reply]
I agree with the concept of that, although I used that wording because it's what I remember from language used in discussing NPOV and RS. --Tryptofish (talk) 22:05, 30 July 2023 (UTC)[reply]

I remember that when I wrote Ora Nichols#Work with Orson Welles, I had to decide what to do with sources that contradicted each other about an incident in which Nichols and Welles had argued. I dealt with it in footnote #19. --Tryptofish (talk) 22:08, 30 July 2023 (UTC)[reply]
Thanks for mentioning me Tamzin. I've also read this essay and think it's pretty good.
The New Amsterdam Theatre article contains several instances of conflicting sources where I took another approach. The basic form of this approach is that several reliable sources would agree on something, but another, normally-reliable source gives a conflicting statistic, so I'd put the conflicting statistic in a footnote. It's a variant of approach #3 (get it right and add a footnote), where most of the sources agree on a given statistic, but there's not enough info to definitively say that the conflicting source is wrong, either - at least not at a certain point in time. For instance, it may be true that the dome in the New Amsterdam Theatre was called "A Song of Flowers" at one point and that the dome's original name was "The Song of the Flowers", but the oddball source says that the dome's original name was "A Song of Flowers". – Epicgenius (talk) 21:00, 30 July 2023 (UTC)[reply]
@Epicgenius (& @Guerillero in re Minute Man, which I actually did look at while writing this based on a past conversation we had): I thought about expanding this to be Wikipedia:Conflicts of sources, and maybe that should be written yet, but ultimately decided it was best to frame this about cases where editors can agree that a particular claim is either false or highly suspect. I've written one article with so many source conflicts that I created a separate ref group just for disagreements on numerical details, and that's not even all the disagreements. (Some number of Rangers, reenforced by some other number of Rangers, either harassed or were harassed by some number of people who were probably dealing drugs and may or may not have been Crips, due to a security camera that was installed either shortly before or some time before, and then either fired upon them or were fired upon, resulting in a gunfight of some number of minutes, in which some number of bullets were fired, which struck either zero, one, or more than one people, leading to either some or all of the Rangers' guns being confiscated—i.e., basically every detail is disputed.) And that's definitely related to this kind of thing, sharing a lot of the same solutions, but I think there's an important difference between "the AP says it was 12 Rangers and the local paper says it was 15 and they can't both be right" and "we literally have articles from both the Times and the Post confirming that the Times didn't run the Unabomber manifesto". Not shooting down those suggestions, just not sure where to go with them. -- Tamzin^{[cetacean needed]} (she|they|xe) 22:00, 30 July 2023 (UTC)[reply]
Actually wait, strike that, think I have an idea. -- Tamzin^{[cetacean needed]} (she|they|xe) 22:05, 30 July 2023 (UTC)[reply]
Okay, I think this works out pretty well. -- Tamzin^{[cetacean needed]} (she|they|xe) 22:39, 30 July 2023 (UTC)[reply]
The Minute Man extensively uses notes to do a mixture of Approach 3 and Approach 4. talk:Marriage License#Comment is an example of Approach 6. Saint Vincent Beer uses Approach 4 due to the dirth of sources --Guerillero ^{Parlez Moi} 21:50, 30 July 2023 (UTC)[reply]