Wikipedia sourcing rules

Thomas A. Fine · June 15

Watching the WAN show, @LinusTech reviewed his own web page, and a couple of times commented that he was the only source of certain pieces of information, and therefore it shouldn't be in there.

This misunderstands how wikipedia rules work (and just how broken they actually are).

In theory, Wikipedia relies solely on secondary sources. Primary sources are not allowed, nor are subject matter experts. To clarify, a primary source would be, for example, a scientist's published research. A secondary source would be an article about the scientists research. So this would seem to include Linus as a sole source for a piece of information. But what happens is that Linus says a thing, and then a secondary source covers that thing. That is when Wikipedia can source it. Perhaps ideally the article should say "Linus said" or "Linus claimed", but realistically, only if that's how the secondary source portrayed it. If the secondary source offered Linus' claim as a plain fact, then that is how Wikipedia will report it.

This presents the obvious way to game Wikipedia — if you control any secondary source, you can say whatever the hell you want and Wikipedia can rely on it. But it's worse than that. If the claim is refuted by a primary source, it can NOT be removed from Wikipedia on those grounds because they don't accept primary sources. I have literally had this happen to me more than once as an editor, where a claim was made that was provably false, but because I was using primary source to make my edit, my edit was rejected.

Wikpedia also does not allow editors to do their own research or reach their own conclusions. Again, this creates situations where provably false things can not be changed in Wikipedia because secondary sources make claims that are wrong. This is particularly problematic with topics that have been strongly politicized, as well as more generally any situation where a myth is widely accepted and repeated. The vast number of secondary sources repeating the false (possibly politicized) myth will always win against any primary source on Wikipedia that refutes the myth.

I too find Wikipedia generally reliable, and depend on it regularly (as Linus and Luke said, dependent on the topic). And I understand the purpose of this primary/secondary sourcing rule (to prevent reliance on individual biased sources). But at the same time, this primary/secondary distinction is Wikipedia's greatest flaw, and i wish more people understood that this is how its sourcing works.

WhitetailAni · June 15

2 hours ago, Thomas A. Fine said:

Primary sources are not allowed

That's... not true. They have restrictions, but they're allowed? Outright banning primary sources would be insane???

https://en.wikipedia.org/wiki/Wikipedia:No_original_research#Primary,_secondary_and_tertiary_sources

Kisai · June 15

16 hours ago, Thomas A. Fine said:

I too find Wikipedia generally reliable, and depend on it regularly (as Linus and Luke said, dependent on the topic). And I understand the purpose of this primary/secondary sourcing rule (to prevent reliance on individual biased sources). But at the same time, this primary/secondary distinction is Wikipedia's greatest flaw, and i wish more people understood that this is how its sourcing works.

Wikipedia is not a reliable source. USE the primary source, eg a paper, a journal, when the thing you need has to be correct and you believe the primary source to be honest. Use the secondary source when the thing needs to be verified as true. The reason you see all these companies (eg nvidia) trying to control their narrative in reviews is because they want those secondary sources to be biased towards them. Hence you can not trust a secondary source who is also a sponsored by the primary source at ANY point in it's history.

The main problem with wikipedia being reliable or not is really a question of how clean the hands are when dealing with data. Information about small companies, politicians, or even people who are notable in a community can not be reliable at any point, because the internet has this culture of trolling people for ***** and giggles. So people will routinely vandalize people who are not popular enough to have someone care about their wikipedia entry, but the wikipedia editors themselves will also routinely delete entire pages that people had spent thousands of hours researching because they deemed the subject "not notable".

Wikipedia's Achilles heel is it's notability standards. Either everything is of notable value, or nothing is. There is no in-between. There are more vanity pages on wikipedia because these editors have a big ego and have "About me" pages longer than their contributions to the site. These pages should not exist. Period. If you are an editor on Wikipedia, your bio should be 200 characters about your qualifications about your language writing and nothing else, because nothing else matters as an editor. You can be a subject matter expert, but YOU are not THE subject matter expert of every page on that subject.

Anyway I stopped adding anything to wikipedia after some big ego over there decided that comics were deservedly lost media, so that site has forever lost my support. Use it if you need a source, but skip the article.

YoungBlade · June 16

The politics of Wikipedia are the worst part of the platform. They are a necessary evil, but still an evil. I was a Wikipedian 20 years ago, and fighting over whether something qualified as a legitimate source was just one of the many, many areas that drove me away.

However, in the fights I had, I never once saw someone claim that a primary source was off limits for a factual dispute about whether something is actually true. I'm sure it can happen, and maybe the politics of the platform today are radically different compared to the early 2000s, but those sorts of disputes usually ended with the claim being removed from the article altogether. Not with the secondary source being retained in spite of the primary one.

With that said, using secondary sources is good practice for an encyclopedia, but it does then rely on choosing good sources. The secondary sources used by Wikipedia do seem to be more lax these days - I'll see citations to random websites sometimes, which was heavily discouraged 20 years ago. I got some sources removed for not being of sufficient quality. Citing a YouTube video or podcast would have been met with pushback for sure, yet I see such things cited regularly. Perhaps the desire to cite everything won out over the desire to only cite with quality sources.