User:Wwwwolf/DelEssay

This essay really needs work, it's not really ready yet. Needs more meat and all. Plus, you probably also want to see Wikipedia:Attribution meanwhile. --wwwwolf (barks/growls) 14:02, 21 October 2006 (UTC)

This essay explains when and how Verifiability, when used alone, is a dangerous argument to use for deletion of articles.

Verifiability policy is harsh... or is it?

When most people read the verifiability policy, they notice quite a few bits:

only contain material... published by reputable sources... cite a reputable source... challenged or removed by any editor... ...obligation... lies with the editors wishing to include the material...

Some would describe the verifiability policy as one of the harshest policies Wikipedia has. There's absolutely no room for any unsourced comments? There's no room for any even little dubious sources? Any meanie can remove anything they want that doesn't have an inline source?

However, what people ignore are the little words:

should only contain material... should cite a... may be challenged...

See? The policy has those little words. When you think of them, they're actually not that bad. Notably, the policy doesn't say:

must at all times only contain material... must, unconditionally, cite a.... Every user's most fundamental obligation, under pain of sanctions from the Arbitration Commitee, is to challenge or remove... The editor removing the unsourced material for any reason is always right no matter what, not the person wishing to include the material...

Here's a healthy suggestion: This policy is not meant to be a harsh policy. Some leniency may be tolerated, as long as it's justified, rational, or sane. The facts should still somehow check up, however, and in the unlikely case it doesn't, then it can be removed. Outright rubbish that fails all source checking can be deleted. However, plausible material that may very well have a source somewhere should be left for the time being, even if it's suspicious.

Verifiability is usually for Cleanup, not Deletion

A big point for everyone is to remember that Deletion is not Cleanup.

Deletion is a last-resort option. It's meant for subjects of which there are severe doubts that the subject isn't notable at all.

Lack of sources may mean lack of verifiable notability. However, many deletion debates have sprung up because the article doesn't have an assertion of notability, much less one that has source - yet a lot of them, in the end, end up being recognised as notable. Why? Often because independent sources claiming notability often do materialise, even when they don't happen to be in articles right at the time.

The Verifiability policy states it's the responsibility of the editor who wants to keep the information to find sources. However, it's also the responsibility of the deletion nominator to do some background work to establish whether something else can be done to the article, or if we're talking of a hopeless case. "The article doesn't have sources, I'm getting 80,000 Google hits and I can't separate the wheat from chaff myself, so this article has to be deleted, now" is not a constructive way to propose a deletion - this is the "Deleter of unreferenced facts is always right" approach, which is not the goal of the verifiability policy.

What sole "verify or else" leads to, at worst

Here's a thought experiment: An article gets nominated for deletion solely for not being verifiable. There's a good chance the article subject meets the notability criteria, but due to big arguments and a lot of shouting about verifiability, the article gets deleted.

Without starting a long cry for justice and endless battle to change the Wikipedia policies, or starting a giant big anti-Wikipedia troll site (both are likely waste of time, the latter even more so), the only logical procedure the people who want the article kept is to recreate the article, with proper sourcing. The previous AfD only concerned the verifiability; if the new article is properly sourced, that discussion is moot. What remains, for second AfD, is the notability discussion, which may be on a lot shakier ground when sources are provided.

So, the big question is: what did the first AfD accomplish?

The obvious answer is: Absolutely nothing, except sweep the history of the article under carpet - and even that may be undone with history-only undeletion.

So what would have been different, if instead of nominating the article for deletion, the nominator would have tagged the article with {{unreferenced}} and sprinkled some {{citation needed}}s around at the troublesome spots?

Nothing, really. Aside of the fact that the article might have gotten a lot better a lot earlier, and there would have been no reason to bring the article up for deletion, and no reason to annoy people who like to keep the article.

Therefore...

Here are some simple rules that may help making deletion debates involving Verifiability a lot easier.

1. Does this really need to go? Someone thought this fact was worth adding. Someone thought this topic was worth discussing. There are cases when the fact is wrong, but hey, it can't be that way all the time.
2. Prefer tagging to deletion. If the only problem is verifiability, tag stuff with {{unreferenced}} and {{citation needed}} rather than bringing the thing for deletion. While we're at it, if there's only a doubt that notability may be a problem, try {{notability}}, rather than bringing the thing straight away to PROD or AfD.
3. Identify problems, don't delete outright. Be constructive, give benefit of doubt... but flag things that you find suspicious and, without a doubt, need a source. Wikipedia is a living document. Full disclosure - "an editor thinks this is a widely known fact, we don't yet have a source for this" - is better than sweeping other people's well-intentioned work under the carpet.
4. In deletion nominations, only throw Verifiability around if a cursory search for additional sources brings up very, very, very little. 300 Google hits is somewhat bad. 10 Google hits is abysmal. I can't even begin to describe the lack of reliable sources in typical case if there's two Google hits. But if you get 5,000 Google hits, there should be some doubt a few of them may be reliable. 10,000 Google hits, and something is clearly not right. 1,000,000 Google hits is... well, saying something about verifiability with straight face may need some explaining. But my point is, if you search any conventional source and there seems to be a large number of references, consider the possibility that something is probably up.