User:SMcCandlish/Notability and Deletion policy

Notability and Deletion policy

Note, 8 December 2007: My views on this topic have changed radically in the last 18 months; as a result, this version is a substantial overhaul of the early-mid 2006 original.

Wikipedia's capacity to catalogue everything we collectively know is effectively unlimited, so the more extreme "exclusivist"/"exclusionist" and "deletionist" views don't make much sense to me. I can't even guess how many times I've see some person or place or thing or concept mentioned in an article, wanted to know more on that topic, and found that not only was there no wikilink on the term, there wasn't any article to point one at, and then later found out that there had been one, but that it got AfD'd on highly suspect "non-notable" grounds, often with a so-called consensus of 2–5 editors. The situation is improving a little, but AfD has basically turned into an obsessive bloodsport for many (try this: observe the 10 or so most-active AfD participants over some significant span of time, such as a week. Next check this list of editors' edit contributions.)

Cleaning up the mess

I had historically been very skeptical and vocally critical of "notabilty" concepts of various sorts in Wikipedia, period, because of how poorly they were constructed, especially as to their level of personal-preference interpretability. What alarmed me about the NN ("not notable") meme, aside from the fact that it did not actually represent the consensus its proponents claimed until its vast improvements in late 2006 onward, was and to an extent still remains the frequency with which NN is misunderstood and [ab]used, often without any other rationale (and fervently but incorrectly believed to be Wikipedia Policy when it is actually just a guideline and related subject-specific guidelines), in the AfD process of deleting articles.

Wikipedia:Notability (WP:N) was very nearly "owned" when I arrived, had been blatantly editwarred into alleged guideline status, was subject to sudden archiving of still-active discussions, was dominated by a small handful of boosters of the then-current version of the document's language, and was otherwise simply not very reflective of Wikiculture at all.

Through a lot of stress and just dogged insistence, my flagging of WP:N as a {{Disputedpolicy}} and my unwavering arguments with regard to its flaws and problems at that time, in Wikipedia talk:Notability, mostly in November and December 2006, have had a marked impact on improving WP:N. The guideline (no longer disputed that I'm aware of) has come a long way toward being objective, clear and far less prone to subjective abuse. Notably (no pun intended) these improvements did not start happening until the Disputedpolicy tag attracted notice and drew more people into the debate. Suddenly WP:N could not be be micromanaged any longer, ongoing debates could not be swept under the rug, and the document was radically overhauled very quickly. I think a major lesson can be learned from this.

Yet AfD abuse remains

However, despite December 2006 - February 2007 improvements in the guideline's wording, scope, intent and raison d'etre, problems with WP:N's application in AfD remain, largely stemming from the imprecise, subjective and radically changing nature of the "notability" concept in Wikipedia from 2004-2006. The idea veered wildly from "importance" to "fame" to "actionability" (huh?!) and various other concepts before settling down to the more objective criteria we now have (as of early 2007), namely that an article subject needs to be attestable by multiple, independent, reliable sources (the primary notability criterion, or PNC). Meanwhile many Wikipedians had already latched on to one or more of the older, flawed conceptions of notability, and are still using them in AfD as we speak, blissfully unaware that the PNC exists at all, much less that it has replaced the old "famousness", "popularity", and "importance" kinds of notability concepts.

I agree that most (though by no means all) of the articles successfully AfD'd probably did need to be either removed or improved drastically (and I nominate articles for deletion myself, and have lost an article to that process, without putting up a fight about it since the article did in fact have a lot of problems). But it ought to be for valid reasons! There are some useful essays out there about the nature of notability and what not to say in AfD, but they simply don't go far enough.

A proposed solution

I think it is going to take a concerted editor education campaign, perhaps with some inline warning templates for use in response to malformed "NN" "!votes" in AfD, to change this: Tag them with a warning that they will not be counted as anything but neutral comments (however "Strong" or "Speedy" they may be), unless an actionable reason is given. And the initial proponent of the AfD can also proactively warn against misuse of NN (or other oft-abused deletion arguments), when creating the AfD. Additional templates could also be used to gently correct other abuses; most !voters using "speedy" don't seem to know what it actually means in the Deletion Policy context, and many also make highly PoV "I've never heard of this" arguments or bald assertions of non-notability which are clearly really "I don't like it" statements in disguise, and yet others demand actions that defy piles of precedent, such as deleting school articles instead of merging them with school district ones. AfD chatter could probably be greatly reduced by use of standardized "broken argument" tags in lieu of re-re-re-explaining the same issues time and again in long-winded "Comment" posts.

AfD has other problems, though

Invalid "me too, but I don't really understand policy at all" !votes are not the only AfD issue. A serious one is admins closing debates as delete when there is a "consensus" of only a tiny handful of editors. I see this all the time, and have successfully returned articles to AfD via the Deletion Review process in a couple of egregious cases. Three !votes after a week that say "Delete", or even "Strong, speedy delete" for that matter (which is silly; if the AfD really was a candiate for speedy deletion, it would most likely already have been speedy-tagged and an admin would have already deleted it) does not make a consensus on Wikipedia. Grossly insufficient input like that is evidenciary simply of the fact that the article either has no active editors or that they are on wikibreak or something. This doesn't even start to get into a broader issue, that of deleting articles that should instead be fixed (e.g., by the addition of another reference or two to satisfy the PNC; NPOV edits to fix a bias problem; etc.) That's a different discussion for a different time, as is the case of punitive or simply careless deletions (e.g. the sometimes premature closing of an AfD as delete immediately after it has been merge-tagged, such that the merge is thwarted, and admin begging or Deletion Review has to be invoked to get at the salvageable content. Admins: Look at articles again before deleting them, please.

Loose ends, and a way out

Another lingering problem is that WP:N is as much a guideline on what makes an article worth keeping as it is a deletion tool, but the subject-specific notability criteria enumerated at WP:DEL (which by their incorporation into WP:DEL actually have the force of Policy, a nasty loophole!) and the larger collection of such guidelines and would-be guidelines that WP:DEL does not mention at all, are almost uniformly both largely exclusionary ("delete-me" guides) and highly subjectively prescriptive as to details, regardless of the PNC. Thus, the topical notability guidelines (many of which greatly predate WP:N) are mostly in direct conflict with WP:N. Wording twiddles have been made to WP:N to try to skirt this problem, but they are really simply hiding the issue rather than solving it (as of February 9, 2007; future edits may be more productive).

I am personally in the process of trying to create a subject-specific notability guideline, for cue sports, that does not have such a conflict. We'll see how that experiment goes. It relies (to the extent that it has been properly constructed so far) entirely on WP:N and long-standing policies/guidelines, and then gives some non-AfD-actionable advice to editors about what sorts of articles are and are not likely to survive AfD on WP:N, WP:V, etc. grounds. I.e., the fact that it says that local tournament players are not likely to be notable cannot be used as an AfD argument in favor of deletion; if editors of the player's article can satisfy WP:N's PNC, then the player is self-evidently notable. An almost silly example, since it is highly unlikely that any local league player could in fact satisfy the PNC, but if it happens then it does in fact happen, and extreme deletionist whining in contravention of WP:N should be ignored with equally extreme prejudice by AfD-closing admins. I hope that this guideline-to-be can serve as a model for what to do with the rest of the subject-specific notability criteria, when it is better-developed.

There are also other loose ends. For example, there is also a content guideline against use of neologisms, which is a de facto notability guideline of sorts. WP:NOT and WP:NFT are also in part basically variants of notability guidelines in certain ways, and their interplay needs to be factored in and smoothed out.

2010 update

Reviewing this old essay in January 2010, I find that a lot has changed, especially the development of WP:AADD (back then it was just a userspace essay by User:Daduzi), which addresses much of what I was getting at in #A proposed solution. But a lot also has not changed, with an enormous number of XfD !votes being random b.s. that isn't grounded in anything. Not sure what to do about that, but at least some improvement has happened since I first wrote this in 2006.

2012 update

I've seen a lot of further improvement. WP:N is now a key guideline, and everyone seems to understand well what is called the general notability criteria (formerly GNG, general notability guideline) now (multiple instances of non-trivial coverage in reliable sources) and that it consists of objective not subjective criteria. It has finally become clear to everyone that topical notability guidelines cannot trump the GNC and WP:N. That said, I think that the general understanding level of the average Wikipedian of policies, guidelines, essays and their interaction has actually gone down across the entire system, surely as the result of so many good editors and admins leaving. Despite WP:AADD, I see a lot of really inane arguments in almost every non-trivial deletion debate, mostly gross misunderstandings of what a policy or guideline really says, or what it means in interaction with another. I think this is a strong sign that we have too much "legalese" and need to WP:MFD half of these documents, because they are ultimately causing more trouble than they solve and are very clearly turning Wikipedia into precisely what it isn't supposed to be. — SMcCandlish Talk⇒ ʕ(^Õل^ō)ˀ Contribs. 22:04, 3 February 2012 (UTC)

2017 update – we still have some problems

The GNC is now the WP:GNG, "general notability guideline", but essentially the same. I spoke too soon in 2012; at least two wikiproject-connected groups of editors are asserting that their subject-specific notability guideline is an alternative to the GNG. Despite that hiccup, most of the kinks have been worked out of our notability system. We've come a long way from the "fame and importance" stuff of the 2000s proposals. (But ... see below.) I still think most of the subject-specific notability guidelines (SNGs) as they stand right now should just be nuked (but could be given new life; see below again). They seemed very necessary around 2010, but now just seem to cause drama, and AfD is increasingly relying on WP:GNG – it's rightly pointed out there that "this fails [insert SNG here]" is trumped any time the GNG is actually satisfied by the article, while an interpretation that something that clearly fails GNG but somehow passes an SNG is frequently rejected. The purpose of SNGs was to outline what is and is not, within that topic, likely to pass the GNG. I.e., SNGs at present are a fortune-telling system that people have mistaken for a set of rules to enforce. This has proven to be pretty angsty, time-sucking, and impractical. While not actually addressing the problems.

We clearly do still have at least two problems. Over the last several years, these things have become very clear:

1. It's nearly impossible to create an article here about an important and influential academic, unless by chance they're a charismatic publicity hound. A scientist may have their work cited literally thousands of times, but not be covered in-depth in multiple, independent, reliable sources (GNG). The only ones who are ones who do TV shows, popularized-science books, and other things that make them "celebrities", or those who unfortunately get embroiled in a public controversy.

   This problem is why the proponents of WP:Notability (academics) are trying to hold it out as something that supersedes the GNG. The motivation is well-meaning, but this approach is terrible and dangerous – it inspires others to try this with their pet topics, too, even when there's no similar justification.

2. On the other hand, WP is just damned well drowning in bios of pseudo-celebrities. Virtually anyone who's ever appeared in more than one episode of any TV show or had a single in the pop or R&B charts can have an article here. We have thousands upon thousands of junk stubs that are never going to improve. They're worthless dreck; the average IMDb or Discogs.com entry has way more and better info on these people, whiile being a huge database of minor entertainers is not Wikipedia's job. Someone like, say, Chipo Chung has some movie and TV credits in minor roles, and has been interviewed and had some fluff articles written about her in celeb and entertainment rags. Oh, and she did some charity work. I sent this to AfD, and it was kept as "no consensus". Years later, the page has not improved noticeably and in some ways is worse – it is not better sourced, but it has more trivia. And this is one of the borderline cases. Innumerable articles on actors, bands/musicians, pornstars, writers, local politicians, executives, models, sportspeople, etc., etc., are flooding Wikipedia (along with non-bios on not-really-notable songs, albums, films, companies, etc.). It's a maintenance and quality-control nightmare.

   Let's be very clear: landing some acting jobs in TV shows and movies doesn't make someone actually notable in an encyclopedic sense; it just makes them competent in their profession. The only difference between these people and those who are really competent bartenders and cab drivers is that there's a cult-of-personality-and-scandal marketplace surrounding people connected with the entertainment industry. The coverage they get is not really independent; most of these publishers are owned by the same companies that own the movie studios, record labels, and TV networks; those of the former that are not house organs are still totally dependent on the latter – the vast majority of their income is from entertainment advertising dollars.

What's the solution?

One idea I've been mulling is to augment the GNG with topically specific definitions of what "reliable sourcing" means for notability purposes. If that sounds weird, consider that we're already doing exactly that, just minus the "topical" part: what constitutes reliable for GNG purposes is narrower than that required for in-article sourcing, in that "multiple" and "in-depth" are added. So, it's not much of a stretch to add another layer; e.g. for academics, in-depth coverage need not be required, but frequent citation would be. For entertainment "figures", entertainment press would be unreliable for notability purposes.

This would effectively import the only good part of the original "fame and importance" ideas about notability: level of influentialness (influentiality?) of the subject. The idea that frequent citation [of particular kinds, and it varies by field] in journals proves an academic's notability could be extended: if numerous musicians, for example, cite another as an influence/inspiration, this is evidence of notability. This is interesting and unusual from WP's typical viewpoint, because in both kinds of cases this is a new kind of verification: distributed primary sourcing. It could not be used in articles, since it's a form of original research on Wikipedia's part. In determining notability, it's entirely permissible, because that's an internal matter. We do exactly the same kind of permissible OR when we assess publisher and author reputability and thus source reliability, to begin with. Another example is how we decide article titles: determining WP:COMMONNAME and compliance with the article title criteria (recognizability, naturalness, precision, etc.) is an entirely original-research, internal process.

That's all I'll write about this for now, other than to suggest that the real future of the SNGs could be housing this kind of material, with their gist summarized at WP:Notability. Maybe the "rebellion" of a couple of these pages against the GNG is presently mistargted, a WP:LOCALCONSENSUS power struggle that isn't going well (and in that form shouldn't). If there were instead a new site-wide consensus that GNG reform is necessary as proposed here, or in some other form – there's an idea to change GNG into GVG, the general verifiability guideline, and move it into WP:Verifiability) then the SNGs could repurposed for this, since dumping a bunch of reliable-sourcing definitions on a topic-by-topic basis into WP:Notability itself would make the page unwieldy.

 — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  02:25, 17 December 2017 (UTC); revised a little at 22:23, 7 January 2018 (UTC).