Wikipedia talk:Notability

Need clearer consensus interpretation of significant coverage and audience.

Companies and organizations; and biography articles are especially subject to notability argument. Although WP:NCORP currently places stronger emphasis on quality of sources, the guidelines only define the two extremes of "a passing mention" as trivial; and and a reliably published book written entirely on the history of IBM on the very extreme end of "significant coverage". To better resist gaming of the rules by marketing and public relations professionals; and connected contributors, it would be very useful to have the shades in between better defined.

Presently, NORG requires multiple significant, independent, reliable, secondary coverage with one of the sources passing broad circulation, and at the absolute minimum, one of it must be national, regional or international. This is often read as a page or two in daily regional paper, plus a cluster of local alternative weekly coverage. This had lead to proliferation of articles about neighborhood restaurants, event venues, book stores and such being ruled "notable", because of a one page coverage in the regional daily paper about something that happened at the place at one point, plus a cluster of local press. How do we interpret notability building effect of "a page or two about the article subject person/organization/company" in a highly specialized books about the discipline (i.e. graffiti art) as opposed to the same amount of coverage in People's Magazine, or Reader's Digest?

WP:AUD could benefit from building consensus on what's considered "limited interest" coverage; as well as "regional coverage".

This AfD failed.

These were the sources used to argue SIRS and AUD requirements:

• https://doi.org/10.1080/15551393.2013.801277

• https://books.google.ca/books?id=KyNVOzSbTGkC&pg=PA38&dq=%22united+graffiti+artists%22&hl=en&sa=X&ved=2ahUKEwj6vpfM66zrAhXnkHIEHSjXC_EQ6AEwAnoECAQQAg#v=onepage&q=%22united%20graffiti%20artists%22&f=false

One of them was a book specializing in graffiti. I personally define things like this as relatively trivial coverage and "limited interest" because it is something you would only see if you specifically go looking for a book on the subject; and a specialized book that is dedicated to a narrowly defined book is bound to have more details about obscure subjects and organizations. Currently, WP:AUD offers no guidance on how to apply "broad audience" vs "limited audience" on highly specialized academic journals or books.

Graywalls (talk) 10:15, 20 September 2020 (UTC)

Books that narrowly cover a specific topic are still acceptable sources (so the "limited interest" is not a valid argument - they are still available to anyone to access, while local sources are only to a local geographic area), like that graffiti book, but it still required the topic to be covered in depth in that, and that's where there's debate that's hard to define , as any definition will lead to gaming the rules. A brief mention in passing is not sufficient, clearly, but do we need a paragraph? a page? a chapter? It's the combination of what all sources give about the topic, in essence. Unfortunately the book preview does not show me what the text is so I can't judge that here. --Masem (t) 13:35, 20 September 2020 (UTC)

Masem, I don't view the WP:AUD thing as "who it's available to", but more from a Conditional probability point of view: Given that the source covers X, what is the probability that it will include coverage of Y? So, let's say you've got four books: one about visual arts, one about painting, one about graffiti, and one about New York graffiti of the 1980's. Each one is covering a progressively narrower topic, and is thus more likely a-priori, to include any specific topic that fall under its umbrella, and thus a weaker signal that the topic is notable. A similar thing happens with geographic audience. The NY Times is, in general, a good source. But, its coverage of local NY topics is not as strong a signal of their notability as coverage by a similar stature newspaper in Chicago, or London, or Beijing. -- RoySmith (talk) 22:29, 20 September 2020 (UTC)

I understand the concern that you can slice a topic area into more narrow and narrow segments to a point where you have an extremely narrow segment that might have a very limited audience, which IS a fair concern. But, at the same time, some subjects do have a reasonable degree where there is finer and finer level of coverage that we would not question a reasonably narrow-focused book. For example, there are plenty of texts in math that are on extremely limited interest subjects, but if you look at the material as you step up from that into broader fields, you'll field works that segment the field like an inverted pyramid, lots and lots of broad math books, and fewer on more narrower subjects. Whereas I can see the problem that if you have broad coverage at one level, and then the next level down, there's nearly next to nothing but one or two books at that fine resolution, so that inverted pyramid concept is gone and its just a spike, that's more like a AUD problem. But that's really really hard to prove. The other factor to make judgement calls on this is how many times other fields look into this; math here is a good example since many other scientific fields typically hang on complex math problems so their results are important even if they aren't worried about how the narrowest problems are solved, they will still talk about them. But in the case where a really narrow area gets coverage but no other field reaches into that, that might be a problem - that was the case a few years back with professional mixed-martial arts which was found to be a walled-garden type thing. I have no idea if graffiti art is similar but knowing the little of the art world I do, I suspect its not that closed. --Masem (t) 02:49, 21 September 2020 (UTC)

• I doubt this is something that can get to hard black-letter standards. Take the "significant coverage" element of the GNG to which Masem alludes. My own take is that we're talking at least 250 words to satisfy: about a page. But ask ten editors and I wager you get at least five different answers. Heck, I've seen certain editors advocate that a sentence is enough, despite the explicit guidance in WP:N that it is not. A consensus on what these things exactly mean just will not happen. Ravenswing 18:52, 20 September 2020 (UTC)
• I would agree with Ravenswing that it's too difficult, if not impossible, to get any bright-line consensus on this. Wikipedia works reasonably well when people concentrate on the specific rather than the general. There are too many "ifs" and "buts" about specific cases for us to come to any general agreement. My own preference is to concentrate on academic books and papers when it comes to sourcing, rather than news reports or popular "broad audience" web sites, because I believe that they are more reliable, but others seem to disagree. Phil Bridger (talk) 20:08, 20 September 2020 (UTC) And I would say that it's more like fifteen different answers from ten editors. Very few people are consistent about such things.
• While the previous statement of the gudieline was not clear, the current practical consensus at AfD over the interpretation of the present version of NCORP seems clear enough. There will always be isolated disagreements and erratic decisions. Myself I do not look at thephysical extent of the coverage, but what is actually written. promotionalism extended to book length is still promotionalism . DGG ( talk ) 21:59, 20 September 2020 (UTC)
• The difficulty interpreting WP:SIRS is kind of unsurprising that is so if you reflect on how the criterion came about: it is essentially the treaty that brought peace to the great inclusio-deletionist war and it describes one of the borders of the territory that the inclusionists were able to defend. I don't like the criterion - not so much because it is not a bright line but because it allows problematic content to stay and excludes some reference-quality material - but the fact that we can work with it is a huge point in its favour. — Charles Stewart (talk) 05:35, 21 September 2020 (UTC)

• The OP has got it backwards. Our policy WP:NOTLAW explains that "the written rules themselves do not set accepted practice. Rather, they document already existing community consensus regarding what should be accepted and what should be rejected." If an AfD establishes that a topic such as United_Graffiti_Artists is ok, as it did, then our guideline should reflect this rather than being altered in opposition to the established consensus. Andrew🐉(talk) 09:16, 21 September 2020 (UTC)

Guidelines and policies, and the outcome of their discussions over the course of time represent community consensus on a wider scale which shouldn't be overridden by local consensus. In AfD's, decisions are rendered all the time based on local consensus that goes against wider consensus, for example series of coverage by the same publication or same journalists getting counted as "multiple sources", even though they're supposed to be counted as one, and hit counts from Newspapers.com and Google search being used as reason for "keep". Graywalls (talk) 06:55, 23 September 2020 (UTC)

So, the topic is borderline on GNG (does not meet it rigorously, but is borderline on the "common practice" GNG standard) and fails Ncorp. Structurally this makes Ncorp irrelevant in this case. Plus nobody argued that it met Ncorp so it's not pointing to needing a change in ncorp. Sincerely, North8000 (talk) 12:24, 21 September 2020 (UTC)

@Graywalls, the last time I was involved in AUD, the canonical example of "limited interest" was a trade magazine dedicated to a very narrow group – a sort of fan rag for professionals. Imagine a "Blue-green Widgets Weekly" periodical that included everything about the world of blue-green widgets, because all 100 of the blue-green widget engineers in the world really wanted that to know absolutely everything, but which might be rather indiscriminate in their coverage ("Feature: Widgets, Inc. Now Selling Blue-Green Widgets in Boxes of 250! Chart: Size difference between 3mm blue-green widgets and 5mm blue-green widgets. Also, readers weigh in on whether teal widgets are blue-green widgets"). We thought this restriction was worthwhile because when you are looking at sufficiently "niche" sources, you can find sources about individual products that editors realistically do not want to have separate pages ("Widgets" are notable; "3 mm blue-green widgets" are not) and because many of these are about products from single vendors, which means that they can have dubious independence. The sources you're looking at are not even remotely comparable to this. WhatamIdoing (talk) 05:01, 27 October 2020 (UTC)

Separately, I've got my doubts about WP:SIRS. So for those who haven't been watching, in every area except NCORP, we generally say that a subject is notable when there are multiple (aka "at least two") sources that have certain qualities that mean they will be handy for creating an NPOV-compliant encyclopedia article. (We additionally require that the subject be acceptable under WP:NOT and that editors agree that it's okay to handle the subject separately, rather than merging it to a larger topic.) Easy, right?

Now about those qualities. They are:

1. significant coverage
2. independent reliable sources
3. secondary reliable sources

(also Wikipedia:Published, extant, and accessible, because sources that nobody can look at don't count at all, and we probably meant to say non-self-published as well, only we forgot to write that down.)

When I ask on this page (as I have, multiple times) it is generally agreed that you need:

1. A minimum of one source that contains significant coverage (think "300 words directly about the subject", only we are absolutely never going to tell you exactly how many words are necessary, because you might believe us, and then we might be unhappy that your article subject qualified as notable, and besides, 300 words that say next to nothing isn't actually as useful as 250 words that contain a lot of facts).
   1. Alternatively, some editors accept a combination of smaller sources that provides the same effect as one source containing a lot of information. So if we take SIGCOV as meaning 10 facts, and you can find six facts in smaller source A and six different(!) facts in smaller source B, then some editors will accept that as being equivalent to a single source that contains 12 facts.
2. A minimum of two independent sources, each of which may or may not actually have significant coverage but which do contain more than a passing sentence about the subject (or a quotation from the subject, if the subject is a BLP). In the "strict" reading, these two sources need to be independent of the subject and from each other, so two nice articles about Widgets, Inc. from the same journalist is counted as just one independent source. Two different journalists = "attention from the world at large". One journalist: Not "attention from the world at large".
3. A minimum of one secondary source, if we feel like it, except that when you get down to brass tacks, aka "actually read the definitions" (or at least WP:PRIMARYNEWS), you discover that secondary sources are less common than is convenient, especially for current events, so this turns out not to actually be a requirement after all, and to the extent that people insist on it, it usually means that they didn't understand what a secondary source is and how it differs from being independent or having significant coverage.
   1. Unless we're talking about medical subjects, in which case we do know what a secondary source is, and we are overly strict on that point. In that area, we are occasionally unable to find truly independent sources, because of the unfortunate fact that it's basically illegal to do anything with an experimental drug without the would-be manufacturer's permission, which kind of kills off the possibility of truly independent biomedical research right there.

The key point is that the SIGCOV source (minimum of one) and the INDY sources (always plural) and the SECONDARY source (if technically any) could be different sources. But if you go over to NCORP's new-ish SIRS section, you have to put forward at least two sources that meet all of these requirements. A non-self-published, INDY primary book (e.g., almost everything published by O'Reilly Media about computers) means nothing; a tall stack of short, non-self-published INDY secondary articles means nothing; a lengthy secondary analysis by someone with a COI means nothing; etc. You could use all of these sources to build the article, and you could end up with a Featured Article this way, but you couldn't use any of them to justify notability under the unique "each separate source must have all of the following qualities" rules that SIRS set up.

I don't know whether this is something that we should try to fix, but I suspect that the reason that AFD returns such 'surprising' results is that SIRS is out of step with the community's real views about what kinds of sources indicate notability. WhatamIdoing (talk) 05:41, 27 October 2020 (UTC)

A strong analysis (and I'm grinning at your blue-green widget hypothetical), but I think you're overthinking it. From way too many years and way too many edits at AfD, my takeaway is that controversial votes come less from a general community disconnect about the rules than that individual editors strongly push their own hobby horses. People come charging in from various WikiProjects to "defend" "their" articles (or, conversely, come charging in to reflexively reject articles of the type they believe in their souls must be purged for Wikipedia to be truly pure), some of the ARS fanatics are quite happy to disavow or reject any notability criterion/guideline as long as they can count coup against Those Evil Deletionists, a number of folks are deeply reluctant to advocate deleting an article that has well-written prose and (to a superficial, three-second glance) many "sources," and once people get their asses into it, they'll defend their positions to the death, subsequent evidence be damned.

I'd assert that there are many fewer "surprising" results over any dislike of SIRS than from these other factors. Ravenswing 06:15, 27 October 2020 (UTC)

@Ravenswing, I submit to you that there is no practical difference here. One of my core beliefs is that Wikipedia:Nobody reads the directions, which means that most the editors who participate in an AFD don't actually know whether their response conforms with this or that rule. The votes from established editors at AFD (in aggregate) actually are the community's views, and when their views don't line up with the written rules, then it is the rules, and not the community's views, that are wrong.

Additionally, in the SIRS case, we have one SNG that its authors wish to interpret as requiring something different from how the community interprets the same words when they appear in the GNG and in other SNGs. That will always lead to confusion, especially among people who are trying to follow the written rules more than their best judgment. WhatamIdoing (talk) 15:59, 27 October 2020 (UTC)

NORG/SIRS is absolutely needed due to the abuse of commercial entities trying to use WP as a promotional outlet; its not as much a problem elsewhere (like BLPs), and that's why it needs its own departure from the GNG. --Masem (t) 16:03, 27 October 2020 (UTC)

Please pardon my confusion, but I have never been clear what the SIRS requirement contributes to the project, for practical purposes, that could not have been accomplished by a very, very stern injunction that the content of press releases never contribute to Notability no matter how widely they are paraphrased or reproduced. Newimpartial (talk) 18:21, 27 October 2020 (UTC)

I think that banning any fact mentioned in a press release would create more problems than it solves. (And, yes, if we wrote such a rule, then there definitely would be someone saying that just because Big Corp mentioned _____ in a press release once, that any and all subsequent news articles about _____ were mere "paraphrases" of the press release.) WhatamIdoing (talk) 18:37, 27 October 2020 (UTC)

I wasn't talking about "banning any fact mentioned in a press release", I was talking about banning regurgitated press releases (and yes, I think the difference is obvious, whatever the wikilawyers would say). IME at least 90-95% of the positive effect of SIRS at deletion discussions, compared to SIGCOV, consists in clearly disqualifying press releases and regurgitated press releases. I have trouble identifying the other 5-10%, to be honest, if it actually exists. Newimpartial (talk) 19:41, 27 October 2020 (UTC)

You didn't mean to talk about banning any fact mentioned in a press release, but that is how it would be wikilawyered. One of my touchstones for this type of editing is this 2011 conversation. Some of the editors (none active these days) tried to remove any positive statement and to discredit any source that didn't talk about lawsuits against the company. There were regular warnings about Churnalism side-by-side with recommendations to accept press releases from a student group as being equivalent to newspaper articles.

WP:ORG mentions the unacceptability of press releases three times, and WP:N mentions it twice. If that's not enough, then I doubt that SIRS, which does not mention press releases at all, is having any effect. WhatamIdoing (talk) 23:16, 30 October 2020 (UTC)

What WP:N is missing though is some equivalent to NORG's clarificarion that any material that is substantially based on such press releases even if published by independent sources. Honestly, from my perspective the rest of NORG is only useful where it is restating what constitutes independent coverage, while the explanation of WP:NOTINHERITED that it offers does more harm than good IMO. SIRS really just provides a thoroughly spelled out, objective-sounding set of criteria that won't under any circumstances allow regurgitated press releases to count, while the requirement that each source capture all of the relevant criteria at once seems basically feckless in my view. Newimpartial (talk) 23:37, 30 October 2020 (UTC)

@Masem, it's been a while since I looked in on the NPPers, but the last time I looked, they said that they had more problems with pop culture than anything else. Indian actors and musicians were their main sore spot. WhatamIdoing (talk) 18:39, 27 October 2020 (UTC)

That's unfortunately problem that we have a weak CSD towards these groups (show the tiniest shred of notability, and you can't be speedily deleted/you will pass an AFC/NPP check), but at the end of the day, much of these are WP:AUD-type problems with first-party, promotional sources, which is out of the prevue of NPP to really evaluate. I'm all in favor of AUD being a part of WP:N but this has been balked at because it would impact some topics over others, or lead to local sources being stripped/disallowed even though that's not what AUD says (local sources can't just be used as the only sources for an article). It's a difficult issue to work around. --Masem (t) 20:12, 27 October 2020 (UTC)

Neither SIRS nor AUD applies to the main problem areas, then. I agree that we are unlikely to expand AUD's reach until we can get people to actually read the directions, which is probably impossible. (Maybe some sort of test... "AUD says that local sources are okay, but you need to have ONE that isn't completely local. Do you (a) agree that local sources are okay, so long as there's ONE that isn't completely local, or (b) will we ban you from participating in AFD?") WhatamIdoing (talk) 23:21, 30 October 2020 (UTC)

@Newimpartial, "the requirement that each source capture all the relevant criteria at once" is AFAICT the main point of SIRS. WhatamIdoing (talk) 17:29, 8 November 2020 (UTC)

Logically, yes it is, but in practice, I don't think it accomplishes much, because (1) editors don't actually apply it and (2) it runs smack up against the spirit and the letter of the GNG and the rest of the WP:N-related guidelines outside of NCORP, so it just doesn't generate much traction even when applied. (This is why I referred to it as "feckless", above.) Newimpartial (talk) 18:19, 8 November 2020 (UTC)

Er, SIRS is repeating exactly what the GNG states, only that I would argue in the context of NCORP, it puts much more weight on should be clear independent and secondary sourcing (due to the potential for ease of promotional sourcing). We're more relaxed on those for other topics outside NCORP but there's still part of the evaluation of sources and the absence of those five is still failing the GNG. --Masem (t) 18:27, 8 November 2020 (UTC)

But there remains the key difference that SIGCOV says they must be met, while SIRS states that they must be met by each source that counts for Notability. That is a distinction with a difference, IMO. Newimpartial (talk) 18:36, 8 November 2020 (UTC)

Yes, that's the key difference. SIRS requires that you produce a minimum of two sources, each of which contains all of these qualities. But if you look back at conversations such as Wikipedia talk:Notability/Archive 69#SIGCOV is badly explained (just a few months ago), there was no agreement that the necessary level of SIGCOV needed to be contained in a single source. SIRS disagrees and says that SIGCOV must be present in a single source. WhatamIdoing (talk) 18:55, 8 November 2020 (UTC)

I don't necessary read it that way, though do I think outside NCORP its more relaxed in terms in how each is applied to each source. That is, I agree NCORP wants each source that is being consider under "multiple" to have "significant coverage" and be "reliable", "secondary" and "independent". In a general topic that's not CORP, I would still expect multiple sources, and in each source that is being called part of the test for notability, that we're looking for significant coverage, the sources are reliable, secondary, and independent. But for example, whereas the "significant coverage" may be a full article about a company in NCORP, we would accept a two-to-four paragraph section of a larger article for other topics, as long as its better than "in passing" and the other qualities are met. Or, in terms of "secondary", we are generally more tolerant of interviews done with creators and the like about their creative works as being secondary, whereas I know interviews with businesspeople under NCORP are almost always flagged as primary and/or dependent due to possible corporate inference. The categories of criteria and application to each source is the same, but the baseline of what applies is far lower for non-NCORP topics (NCORP setting a higher bar for good reasons). --Masem (t) 19:43, 8 November 2020 (UTC)

@Masem, is it fair to summarize your view here as "the written standards aren't applied consistently"? An interview with the subject (e.g., "interviews done with creators and the like about their creative works") is a primary source. So are nearly all the sources about current events when we create the articles. Articles such as 2017 London Bridge attack get created within an hour, when only primary sources are available.

Maybe if we don't actually require (real) secondary sources, then we should just stop saying that we do. WhatamIdoing (talk) 23:09, 9 November 2020 (UTC)

Hmm, that's raising a different issue, as I don't think we universally discount interviews as primary sources as context very much matters. WP determines secondary sources if they are transforming information from primary and other secondary sources. An eyewitness report reiterating a recent event is clearly not transformation and thus primary. But a creator talking about their work - say 5-10 years after its published so that there's a time factor involved - now that's different. They may be providing insights and the like about the work that in context of the work would be secondary. If the interview was being used for their bio page, it may be primary or secondary, it would be very much context dependent. This is the same way we have to look at things like autobiographies, DVD commentaries, and other good sources of information. I would note these are very-much dependent sources, so no way alone could satisfy notability.

Keeping that in mind, the difference still becomes that a quality bar is set higher at NCORP, not that there is an inconsistent application elsewhere. NCORP I know has discussed extensively the caution about interviews in business trade magazines that these are something you can buy and load up with one's own questions for self-promotion (in contrast with typical interviews done for actors, producers in Hollywood, etc.) But if its WSG or Businessweek conducting the interview as part of a feature story, that's probably reasonable. (see footnote #3 there). In contrast, other places outside NCORP will probably not worry too much about the use of interviews with the same concern as long as an RS is conducting them, because they aren't used in the same promotional manner as the business world does. (eg in business, businesses seek the press to get interviews; in most other fields, the press seek people to get their interviews). But the same end issue is that interviews are dependent sources and an article built up only on interview pieces in non-NCORP areas will still not pass notability. --Masem (t) 01:39, 10 November 2020 (UTC)

Notability is Wikipedia's Achilles heel.

There are so many double standards, contradictions, loopholes and gaming that being "notable" can either be p-easy to gain in the case of sports or Iranian villages or almost impossible when it comes to cryptocurrencies or fictional characters. I have had so many "last straws" for Wikipedia so often that I don't bother in most cases, but AFD has p-ed me off more than usual. Wikipedia's 20 year test is coming up soon, and people and donators will soon write more about the big gaps of knowledge deleted by Wikipedia. Second generation wikis that are more inclusionist will soon overtake Wikipedia (even Wikimedia's own Wikidata has more coverage), and Wikipedia will be considered non notable by future encyclopedists. 94.175.6.205 (talk) 14:58, 23 September 2020 (UTC)

Or they become by words for inanity. And today on EvertyhingPedia our featured article is "Lord of the rings characters featured in discussions down the Stoat and whilst, Waterford".Slatersteven (talk) 15:01, 23 September 2020 (UTC)

It's a confusing kludge that has problems but mostly works. Wikipedia:How Wikipedia notability works right now. Someday we could create a grand unified guideline which would simplify and fix. Sincerely, North8000 (talk) 17:53, 23 September 2020 (UTC)

Riiiight, because the world measures encyclopedias by the number of comprehensive articles about ephemeral soap opera characters, one-shot obscure mangas, rural elementary schools, and bands of which no one besides their 50-person fan base have heard. Seems like a classic example of Ravenswing's Law in action. In any event, what makes you think that your "second generation wikis" will make you any happier? As with Wikipedia, they would certainly operate under rules and guidelines that would majorly piss you off in one way or another ... or if one conforms to your prejudices in every particular, you'd be the one writing a heated defense against the naysayers loudly proclaiming that YourFavoriteWiki has garbage standards and rules. Ravenswing 19:06, 23 September 2020 (UTC)

:-)  :-) North8000 (talk) 20:18, 23 September 2020 (UTC)

(The non-sardonic answer I have is this: Wikipedia operates on consensus. That consensus is imperfectly wrought over the better part of two decades, and is heavily influenced by precedent. In effect, a handful of loudmouths 15 or more years ago set in stone a bunch of shibboleths that will never be overturned. Significant institutional change is extremely difficult, because there's always (a) a large claque who like the status quo just fine, and (b) an equally large claque certain that such change will open the floodgates to all manner of craziness.)

(But what other system could succeed? A professionally-run online outfit like Encarta or Britannica? With a fiftieth of Wikipedia's content, they're all in the dustbin. A strictly moderated site with editorial oversight curated by "experts," as with Citizendium? That effort's been moribund for a decade, with less than a ten-thousandth Wikipedia's content. Wikipedia probably won't be any more eternal than Britannica was, but what's going to replace it isn't going to be Same-As-WP-Only-With-Rules-I-Like.) Ravenswing 05:33, 24 September 2020 (UTC)

One of the beauties of Wikipedia is that the people who create and own the content are not the people who own the web site, so if you want to create an encyclopedia about cryptocurrencies and fictional characters you are welcome to do so, and you can use the content that has been created on Wikipedia as long as you acknowledge the writers. Phil Bridger (talk) 19:50, 23 September 2020 (UTC)

Generally, notability's larger problem right now is how we deal with "modern" topics that haven't had the nature of academic analysis, which is basically most any contemporary topic. It's still something we're learning how to do. We found people and companies tried to play the allowances for businesses and organizations, so we had to tighten down the notability guidelines for businesses. We had to significantly tighten those down for cybercurrency when we found there was very little "independent" coverage of that field. We also know that we probably have excessive coverage of sports that could probably be trimmed, but again, that's a learning process. We know its not perfect and that there's no universal rule here. --Masem (t) 20:25, 23 September 2020 (UTC)

IP, I for one am thrilled that we have so many articles about sports and villages and relatively fewer articles about scam cryptocurrencies and fictional characters usually better covered in articles about the works of fiction. It is exactly as it should be. As for your predictions about the future, let me remind you that you do not own a magic crytal ball, and that false predictions of Wikipedia's demise are almost as old as the project. Cullen³²⁸ Let's discuss it 20:30, 23 September 2020 (UTC)

Are there any non-scam cryptocurrencies? Phil Bridger (talk) 20:48, 23 September 2020 (UTC)

I would think the original Bitcoin is one of the few legit ones, but it is itself one foot in a pile of self-promotional manure of other crypto that taints its reputation. We'd still want more non-crypto-related third-party coverage of Bitcoin to talk about it. --Masem (t) 20:56, 23 September 2020 (UTC)

The excessive number of articles we have in certain areas don't bother our readers because they never see them (except on disam pages). Readers only see what they click on a link for. Something those who obsess about gender %s etc should remember. Equally, given how low the views are for the bottom, say, 15% of articles we do have, it's hard to argue there's a widespread unmet need. Johnbod (talk) 23:44, 23 September 2020 (UTC)

Original poster here, I think my ideas are right. Cryptocurrencies are one of the most controversial topics in the field of finance and have been linked to many crimes and scams even threating to destroy twitter. But they also changed peoples riches and many people have become millionaires and many countries like Venezuela have been improved by having cryptocurrency. The topic deserves to be notable and documented. Fictional characters also influence our lives by giving people common stories and references to share, so should be notable through that way. The notability debate has gone on for too long, too many Wikipedians have left the project after having all their articles destroyed by deletionism. I am an ex Wikipedian my self who left years ago but still sometimes makes a few edits. But no matter what happens to Wikipedia, inclusionism will always win in the end because people want to know and if Wikipedia won't provide it, somewhere else will. 94.175.6.205 (talk) 12:02, 26 September 2020 (UTC)

We do cover Cryptocurrency. We do cover Petro (cryptocurrency). So what is it you think we are missing?Slatersteven (talk) 12:10, 26 September 2020 (UTC)

And come now. Every subject and topic has a constituency whose lives they can claim are thereby enlivened, and have common stories and references to share. The most obscure athletes played for teams that had fans who cared. The most obscure porn actresses had fans who were, well, "enlivened." But ... there are reasons we have notability standards. Include every athlete and every porn performer and every thing that claims to be a cryptocurrency, and we not only have no idea whether the information presented is accurate, but it's impossible to police. Wikipedia seeks to be accurate, so that we're not a joke like Urban Dictionary and the like. So ... unless the situation is that it's not really that you disagree that there should be standards, or that articles need to be cited to reliable sources, then what your real beef is -- because you've been very vague on details -- is that the standards are tighter in your pet topic than you'd like. That's the price of doing business in a consensus-driven environment.

I'd certainly make a bunch of changes if I were appointed Dictator of Wikipedia. But I'm not, nor going to be. Not even Jimbo's that, these days. I put in my two cents' worth, and there are notability guidelines out there I'm proud to have created. And beyond that, I live with the compromises, and accept that a lot of editors don't agree with me on this or that. If I couldn't accept that, I'd walk. But beyond that? I figure on taking the chance that there are only a handful of people butthurt enough to flip Wikipedia the bird because their pet obscure athlete or ephemeral minor porn star or -- say -- fly-by-night cryptocurrency that no one ever wrote a news piece about don't have an article. Ravenswing 05:54, 27 September 2020 (UTC)

I am not really interested in cryptocurrencies and fictional characters, but I must say, that I understand the IP and I agree that the Notability policy is the Achilles hell of Wikipedia. I mean for instance my favourite football club is deemed not notable, my favourite band is deemed not notable etc. (I have no personal connection to the any of them of course) and many local topics I would be interested in to read also. I find it very sad that a huge part of human knowledge is not available to the users. I do not want to sound pathetic but I believe, that most posters here do not have a slightest clue, how it is, when the topics you are most interested in are not allowed to be included in the encyclopedia. Ludost Mlačani (talk) 09:27, 27 September 2020 (UTC)

And I would say you are wrong, I have even had articles deleted as not notable (such as my local games club). Hell I have even voted delete on things I own, because they did not meet any reasonable notability requirement.Slatersteven (talk) 09:30, 27 September 2020 (UTC)

And I'll play devil's advocate. It's more than likely that I don't give a damn about your favorite football club, nor about your favorite band, nor about topics local to you. Nor would I wager that you care much about mine. I've had an article in my user space about a man I revere, a professor of music at my university for nearly half a century. Even with my intense personal bias I can't shake the fact that he just falls short of the relevant notability standards, and he's running out of time to pass them. And likely no one in your country has ever heard of him. So where do we draw the line? Unless you believe that there should be NO standards, and that everyone should have an article, and that reliable sourcing should not be required, then you believe there should be some standards. And what would those standards be? Ravenswing 06:45, 28 September 2020 (UTC)

If he was professor at YOUR university there could be problems with COI and PROMO. Otherwise I would have no problems with it. You are aksing abour standards? As long as editors aren't personally involved or paid and information is true and verifiable and no personal data regulation is violated then I see no problem if there is an article about something. Yeah, I don't give a damn about your favourite band but I allow an article about it. Your favourite band has an article and mine does not, that is the main point here. Ludost Mlačani (talk) 10:57, 28 September 2020 (UTC)

My favorite band is an internationally celebrated group with a couple dozen Grammy awards, nine Top Ten hits, and are in the top twenty best selling musicians of all time. Its Wikipedia article has over four hundred cites. If the article on your favorite band had only two reliable cites providing significant coverage, it would be up here. If it isn't popular enough to have passed that rather pathetically low bar of coverage, and/or neither you nor any other of its fans are motivated enough to find such coverage, well. That is the main point here. Ravenswing 20:20, 28 September 2020 (UTC)

Yes, it is a shame that encyclopedic standards are overshadoved by the popularity contest. The whole point of Wikipedia in the last years is just who has the most hits, fans, awards, supporters etc. Ludost Mlačani (talk) 10:42, 10 October 2020 (UTC)

Where is your evidence for that? If you read deletion discussions you will see that "who has the most hits, fans, awards, supporters" is not the point, but it is who has independent reliable sources writing about them on which we can base an article. Phil Bridger (talk) 10:51, 10 October 2020 (UTC)

In this commercialised world those so called independent reliable sources are mostly writing exactly about those "who have the most hits, fans, awards, supporters". Ludost Mlačani (talk) 10:55, 10 October 2020 (UTC)

• I have said it before and will say it again, if RS are not giving your favoured subject enough coverage to generate an article, contact them and complain. They are the ones at fault.Slatersteven (talk) 10:08, 27 September 2020 (UTC)
• Can anyone show me an article of any substance in any other mainstream general-purpose encyclopedia that has been deemed not notable for coverage in Wikipedia? I don't see how our notability standards present a problem if we are covering vastly more than everyone else covers. BD2412 T 15:17, 27 September 2020 (UTC)

There are numerous such articles just in other language Wikipedias with looser rules that are deemed not notable for the English Wikipedia. And of course you see no problem, you are an admin, so you have an access also to deleted articles so to the full knowledge. If WIkiepdia do not want to cover all topics it should at least be fair enough to change its motto. Ludost Mlačani (talk) 16:38, 27 September 2020 (UTC)

I think you know that I was referring to non-Wikimedia collections. In any case, if there were no standards of notability, I could theoretically write an article on my toes, Toes of BD2412, detailing their length in millimeters, circumference, and relative sensitivity to various stimuli. Does the fact that our current standards prohibit that deprive the world of "full knowledge"? If I wrote such an article and it was deleted, would my ability to see it mean I have greater access to knowledge than others? I don't think so. I think there is a difference between information and knowledge, and that our lines usefully draw that distinction. BD2412 T 16:59, 27 September 2020 (UTC)

Well wikipedia is still subjected to laws and there is General Data Protection Regulation, so the article might be deleted for that. But if you are okay with that information to be public, I certainly would not nominate it for deletion. So in one language something is knowledge and in the other the same thing is only worthless information, interesting... Ludost Mlačani (talk) 17:36, 27 September 2020 (UTC)

Nonsense. What that is is one group of editors who have stricter standards, and one group of editors who have looser standards. Honestly, you and the IP are doing the same thing: complaining about the way things are without telling us what you would do differently, and how you would do it. 06:45, 28 September 2020 (UTC)

I would not delete other people's work, if it is at least somewhat verifiable. I do not care if there is information I do not read. Ludost Mlačani (talk) 10:59, 28 September 2020 (UTC)

Just to let you know i’m setting up my own wiki to cover encyclopedic topics discriminated against by the notability policy. Deletionists will be treated the same as vandals on this wiki. I’ve noticed other wikis do the same. Eventually the inclusionist wikis will win out in terms of coverage and deletionist wikis will be forgotten about. People have been arguing about notability for over 15 years, admit it is a failed and discriminatory policy. 94.175.6.205 (talk) 10:33, 29 September 2020 (UTC)

Your choice.Slatersteven (talk) 10:42, 29 September 2020 (UTC)

Meh, whatever. Are we supposed to feel bothered or threatened? Wikipedia has something like fifty times the traffic and content of all those "other wikis" combined. Anyway, you do you. I'm sure you'll feel comforted by the fact that no one here will lose sleep over it. Ravenswing 12:06, 29 September 2020 (UTC)

The hard part of starting your own wiki is coming up with a good name for it. There have been so many attempts that all the really clever names are already taken. Blueboar (talk) 12:16, 29 September 2020 (UTC)

Not sure this is a subject fit for us to discuss.Slatersteven (talk) 12:17, 29 September 2020 (UTC)

94.175.6.205 The "no inclusion rules" entity that you describe already exists; it's called the internet. North8000 (talk) 14:25, 10 October 2020 (UTC)

It's really my ultimate answer to the people who whine about how Wikipedia c!e!n!s!o!r!s stuff, how horrible, we're suppressing information!!!! No, nitwits, we're not eliminating all mention of those things everywhere. They just don't have Wikipedia articles. Ravenswing 21:55, 10 October 2020 (UTC)

Notability optimization

Basically COI editing combined with notable sources ala SEO but for notability. This will happen as new generations of notability policy aware COI editors work in conjunction with the media to get get their articles notability optimized on Wikipedia to avoid deletion. I could easily become "notable" enough with a few thousand dollars and do multiple events to get round the ONEEVENT policy. 77.96.44.212 (talk) 15:18, 27 September 2020 (UTC)

sadly yes, what do you suggest?Slatersteven (talk) 15:21, 27 September 2020 (UTC)

If some organization spends that much money to find a way to make themselves appear notable that NORG's restrictions on sourcing does not catch (that is, getting coverage beyond trade magazines and over a length of time) just so they have a Wikipedia article, well, congratulations to money well spent. There's a limit to what we can practically catch if an organization is this intent on "defeating" our notability requirements. We're trying to catch the cheap and easy SEO attempts that we can readily judge by the sourcing. --Masem (t) 15:24, 27 September 2020 (UTC)

Well, yes: people have been famous for being famous, and paid large sums to gain the notice of the public, pretty much for the entirety of recorded history. They will continue to do so. I think we all have much better uses of our time than worrying over the bare handful of people who thirst so badly to have Wikipedia articles that they'll go that route ... never mind that there is nothing that would be more fatal to Wikipedia's principles and collegiality than to replace WP:V and WP:N with some amorphous concept to only give articles to the "deserving." Ravenswing 06:52, 28 September 2020 (UTC)

People hire public relations firms for a reason. You might be disappointed to find out how many truly notable things do have massive amounts of money to market them. That's capitalism for you. But the most brazen self-promotion doesn't usually get you more than a single news spike, and WP:ONEEVENT helps weed that out. Shooterwalker (talk) 23:31, 28 September 2020 (UTC)

I've been hoping, for a very long time now, that more editors would see the importance of historical significance. While historical significance is well-respected for POV issues, it gets sidelined due to WP:NOTTEMPORARY. --Hipal/Ronz (talk) 22:17, 27 October 2020 (UTC)

I think that the far more common way to buy their way into Wikipedia is to hire a wiki-savvy undeclared paid editor. Our badly written guidelines that define pretty much every editor as a wp:COI prevent the focus that it will take to fix that, and making life hell for the rare ones who declare makes the undeclared problem much worse. North8000 (talk) 01:08, 28 October 2020 (UTC)

Tidying up redundant language in WP:GNG

The GNG currently contains some redundant language in these two bullet points:

• "Reliable" means that sources need editorial integrity to allow verifiable evaluation of notability, per the reliable source guideline. Sources may encompass published works in all forms and media, and in any language. Availability of secondary sources covering the subject is a good test for notability.
• "Sources" should be secondary sources, as those provide the most objective evidence of notability. There is no fixed number of sources required since sources vary in quality and depth of coverage, but multiple sources are generally expected. Sources do not have to be available online or written in English. Multiple publications from the same author or organization are usually regarded as a single source for the purposes of establishing notability.

The point that sources can be in any language or medium is repeated, as well as the point that multiple publications from one author/organization count as one source (which is in both the footnote and the main text). The quality of prose here really needs to be higher, given that this is the central element of one of our core policies. Can we please rewrite it to remove the repetition? {{u|Sdkb}} ^talk 03:25, 28 October 2020 (UTC)

We usually start off with a link or footnote, and expand it and repeat it until it's finally enough that people get the message. This may be the amount of repetition that is necessary. WhatamIdoing (talk) 23:29, 30 October 2020 (UTC)

Notability of things popular with kids and people too young to edit Wikipedia

There are a lot of topics that are popular with kids but do not have articles as the systemic bias of adults not writing as many sources about them (apart from marketing of course) and that the target audience is not yet old enough to edit Wikipedia competently. Draft:Gacha Life seems to be stuck in this limbo as it really is a phenomenon popular with younger audiences. I'm out of the target demographic but I can see the problems that Wikipedia has in this area. The only way out seems to be if such subjects get an unexpected adult following (My Little Pony and Thomas the Tank Engine being the well known examples), or waiting until the target audience grows up and write about them from a nostalgia perspective. Baby Shark only became notable because it got 7 billion views from repeated watching on apps, it would be just another kids song otherwise. 2A01:4C8:72:A70B:24E0:199A:3079:4118 (talk) 18:57, 4 November 2020 (UTC)

We have to wait for reliable sources to cover these areas, and there are cases where they just don't get the coverage. I will note that when it comes to mobile gaming that itself already tends to have a mark against it: there are some but not a lot of reliable sources that cover that area, as many of the sources that do are just overtly highly promotional materials. (We have the same problem when it comes to cybercurrency). Add that when we are talking children's apps, as Gacha Life seems to be, that's even yet another strike because there's little money in that. And judging what I can on Google News hits, it does appear to be a fad, maybe in passing but I can't for sure yet. Basically, in these situations, we're handling notability correctly. --Masem (t) 19:07, 4 November 2020 (UTC)

• WP:GNG isn't intended to be a notability 'bias creator'. But if we cannot actually satisfy GNG for a topic, especially a toy company or something, it's pretty much impossible to write an encyclopaedic article that meets WP:V, WP:NOR. What do you base the article's content from? The company's website / promo material? Therein lies the issue. Reliable sources for notability could maybe be loosened a bit, as eg Techcrunch covers a lot of apps in this area, but often Techcrunch isn't considered an indicator of notability so that may well be something to look into. But Techcrunch also discusses a lot of startups; it may well quickly turn into excessive promo. So there's two sides to this. ProcrastinatingReader (talk) 19:38, 4 November 2020 (UTC)

(after edit conflict) I don't believe that there are notable topics that don't get coverage. I have an 18-month-old grandson, and his favourite computer games are based on Numberblocks and Peppa Pig, his favourite TV programmes are Hey Duggee and Bing and his favourite film is Frozen (despite his great-grandmother saying that it's for girls), recently taken over from Toy Story. All of these have copious coverage in reliable sources, as I'm sure Gacha Life will get if it's really as influential as you say. Phil Bridger (talk) 19:47, 4 November 2020 (UTC)

• ”Notability” and “Popularity” are not the same metric. While there is overlap, there are notable things that are not popular, and popular things that are not notable. Blueboar (talk) 21:46, 4 November 2020 (UTC)

It has always seemed to me that some things are very notable to a small number of people, like most science articles. Some are slightly notable to a very large number of people, so the ones that might be popular. Both are useful and important. As for kids, until they are old enough to read the article, it would be the notability to older kids and adults. Ones like Phil Bridger indicated are likely notable to older (than four) kids, and to adults wanting to know why the kids are interested in them. In addition, sometimes adults like them, too. Gah4 (talk) 22:52, 4 November 2020 (UTC)

Wikipedia:FAILN is IMO the best response to these situations. If you don't have enough independent sources to write an encyclopedia article (that why we need those sources), then mention the subject somewhere else. Add a section to Short film that talks about DIY and app-based options, and namecheck this app. Write about the company, and name and briefly describe all of their apps in that article. You can put content in Wikipedia without putting it on a separate page. WhatamIdoing (talk) 17:38, 8 November 2020 (UTC)

SNG change

Barkeep49, which discussion is this edit alluding to? I think that the edit to the first sentence is an improvement, but the second change is rather significant, as it would imply that topics with SNGs must meet said SNGs in order to be considered notable. _signed, Rosguill ^talk 22:11, 6 November 2020 (UTC)

• Agree, the second change should be reversed. There are always instances where GNG trumps an SNG, imv Atlantic306 (talk) 22:26, 6 November 2020 (UTC)
• I've undone it, as the edit is not accurate. No SNG is also required to meet the GNG - that is, the SNG is an alternative to the GNG - and except for NCORP, the GNG is an alternative to the SNG. NCORP is the exception due where the GNG defers to it because of possible COI issues with promotion. --Masem (t) 22:57, 6 November 2020 (UTC)
• See the discussion here, I believe Barkeep49 was attempting to clarify the SNG. I disagree with Masem. I have always understood that every article must meet GNG, but there is a presumption an article meeting an SNG will pass GNG. SportingFlyer T·C 23:03, 6 November 2020 (UTC)

  @SportingFlyer, there are editors who will tell you that multiple criteria in WP:NPROF don't require any independent sources at all and are meant to bypass the GNG's requirement for independent sources. According to them, it's just up to the individual editor's judgment whether someone had "a substantial impact" on higher education, and sourcing an entire article to someone's résumé or their employer's website is okay. Efforts to change this have been rejected in the past.[1] WhatamIdoing (talk) 17:47, 8 November 2020 (UTC)

  @WhatamIdoing: Oh, I know. NPROF is the exception to the rule here and I'm not really referencing it. As a result, I don't touch NPROF articles, because I have absolutely no idea when notability has been met, or even when there's a notability grey area. My idea in the past is to expand the WP:GNG to include the types of sources that would normally make a professor notable, but even that's controversial since it might shake the apple cart. SportingFlyer T·C 18:30, 8 November 2020 (UTC)

• Clearly I didn't word it well. I know Masem says that no SNG is required to meet the GNG. That's now how I read the May discussion. Some SNGs are quite clearly alternatives to the GNG - e.g. WP:NPROF. Others are shortcuts - that is a way that we can presume notability without having to have the same discussions challenged over and over again. I do agree with Atlantic306 that something which meets GNG but not the SNG is going to be notable. However, even that is a bit controversial. Best, Barkeep49 (talk) 23:05, 6 November 2020 (UTC)

I originally posted this as a new discussion section, but realized that this discussion is effectively about the same thing. So, I hope you don't mind if I just roll my comments into this discussion instead of starting a redundant one.

Hi there. I've been an admin since 2012, and have been very active in participating in and closing AfDs during that time. I took a wikibreak for several years, and returned recently to being more active. As I returned, I noticed an interesting change to the subject-specific notability guidelines on WP:N. The wording has been tweaked, and in my opinion, it creates ambiguities that muddy the exact meaning of the policy. Perhaps consensus around the nature of notability has shifted while I've been away, which is fine if that's true. But, my understanding of notability was that GNG is the ultimate test. My understanding has always been that if it can be shown that an article meets GNG, then it's notable, full stop, and is presumed to deserve a standalone article as long as it doesn't violate any other WP policies. My understanding was that SNGs are simply a shortcut that can be used to quickly estimate whether a topic is likely to meet GNG after an exhaustive search for sources has been completed.

In my opinion, the current wording of WP:SNG includes several statements that contribute to its ambiguity, and make it ripe for misinterpretation and wikilawyering:

"In some topic areas, consensus-derived subject-specific notability guidelines (SNGs) have been written as an alternative to the general notability guideline to allow for a standalone article."

If SNGs are truly an "alternative" for GNG, then that means they can be used as a substitute for GNG, implying that GNG can be ignored if the notability criteria within an SNG has been satisfied. The word "alternative" implies a mutual exclusivity between GNG and SNGs, suggesting that if you don't want to use GNG to determine notability, SNGs can be used as an alternative.

"A topic is not required to meet both the general notability guideline and a subject-specific notability guideline to qualify for a standalone article."

If a topic is not required to meet both GNG and SNG, then that means that a topic only needs to meet one or the other. Again, this implies that if a topic meets the requirements of an SNG, then GNG can effectively be ignored.

"Note, however, that in cases where GNG has not been met and a subject's claim to meeting an SNG is weak or subjective, the article may still be deleted or merged..."

To me, this sentence implies that an article might only be deleted if it can be shown that it doesn't meet GNG and that it doesn't have a strong claim to meeting an SNG. Again, this suggests that if a topic has a strong claim to meeting an SNG, then we need not bother worrying about whether or not it meets GNG.

Indeed, some SNGs themselves contribute to this ambiguity. At the top of WP:NSPORT, we see:

"The article should provide reliable sources showing that the subject meets the general notability guideline or the sport specific criteria set forth below."

That italicized "or" clearly suggests that a topic is considered notable if it passes the SNG, regardless of whether or not it also passes GNG.

Given all of this, I'd like to start a discussion about two different specific questions regarding the current consensus on the nature of notability:

1. If it can be shown that a topic objectively does not pass GNG (i.e. after an infinitely exhaustive search through every document known to man, no one can find sources that satisfy GNG's criteria for notability), but that same topic clearly satisfies an SNG, what is the policy in this case? Should the article be deleted? Or is satisfying an SNG enough to keep the article around, even if it has been proven beyond the shadow of a doubt and everyone agrees that the topic has not received significant coverage in reliable sources that are independent of the subject?
2. If the answer to question #1 is that GNG is still the ultimate test, and SNGs are only subservient shortcuts to GNG (and that an article should be deleted if it can be shown that it does not and will not ever pass GNG), then what can we do to modify the wording of WP:SNG to make this more clear? ‑Scottywong| [babble] || 23:33, 6 November 2020 (UTC)

To reiterate themes that have been discussed over and over again on these talk pages: both the GNG and the SNGs are meant as presumptions of notability to allow for a topic to have a standalone article, in absence of absolutely assuredness that a topic is notability. That is, a topic like World War II is clearly notable - it's not just presumed notable, it absolutely is notable, it will never be questioned as a topic. This is a goal we'd want to see for all articles. But we recognize that this is not always possible for an article to show, nor necessarily desirable as we want articles to be works in progress in an open wiki to have multiple editors to help contribute, but at the same time, we need to be sure that a topic is likely to have good sourcing to actually get to this point. So we have presumptions of notability. One way is by the GNG - if you can show a handful of sources giving indepth coverage a topic, that's a reasonable step towards larger notability. Alternatively, SNGs are generally merit-based criteria that says in certain fields, if the topic meets a specific criteria generally based on some merit, then from past experience we can expect sources to be there and thus presume notability. This is why it's the GNG OR the SNGs as both get to the presumption of notability.

That presumption is rebuttable however, but this requires the person wishing to challenge to show something effectively meeting the requirements of WP:BEFORE. For example, if a topic has been given an article by the GNG because of two sources that seem to meet it, but a person has done a thorough search and found no other sourcing, then maybe that presumption is wrong, and thus that can go to AFD. Same with a topic that met an SNG. But that BEFORE search has to be appropriate - this typically means that a Google search is not sufficient for any topic that was active pre-2000, and may require more local search for topics not in Western countries due to lack of digital resources. So for example, for an article on an Indian cricket player that played a couple matches in 1960 which may pass NSPORT but has no other immediate sourcing, one would like need to review Indian newspapers of the 1960s to see if there was more content, If one does not do an appropriate BEFORE search before such an AFD nomination, then that is usually called out at the AFD, and the article kept because of the lack of a proper BEFORE search to show that no sourcing does exist (since its impossible to prove a negative). But if that BEFORE search does come up empty, then I'd expect the onus to find any additional sources to be on those wishing to keep it, otherwise deletion is appropriate. Thats teu for either the GNG or SNG, whichever has been claimed as the source to be used.

What tends to be the case in editors' minds is that the GNG is the more desirable case over the SNG, but really, they are generally equal tests since they are both presumptions of notability. --Masem (t) 23:52, 6 November 2020 (UTC)

The reason this is coming up is because of the AfD at Wikipedia:Articles for deletion/Willie Thornton (Canadian football) and my asking Scottywong for clarification on the close, partially to figure out what I could have done better with my argument, partially to try to figure out what the rule is when it applies to AfD closes. I agree with almost everything you're saying above, but I don't view the SNG and GNG as "equivalent." I see SNGs as being very important for the assumption that sources exist, but as soon as you question whether sources exist, the sources you need to find must be GNG-qualifying, or, in other words, show the article meets the GNG. The exception as you note is when a SNG is met but we can't definitively show GNG because of a lack of access to quality sources, which is usually only the case when the sources are in a different language or historical or both. It's the reason we fine-tune SNGs to make sure GNG is met in almost all cases. SportingFlyer T·C 00:08, 7 November 2020 (UTC)

Thanks Masem. That description is slightly different than my understanding of notability. The difference is small and it probably seems pedantic, but keep in mind that this is a policy that affect millions of articles, and is used in dozens of deletion discussions every day, so the details matter. My understanding is that if it can be shown that a topic meets GNG, then it is a notable topic, and it is presumed to be suitable for a standalone article or list, as long as it doesn't violate other WP policies. So, GNG is not a presumption of notability, GNG is the definition of notability. The only presumption it includes is whether or not each notable topic is suitable for a standalone article. On the other hand, SNGs are a presumption of notability. They are a shortcut that presumes that a topic will pass GNG as long as it satisfies the SNG. They are not a guarantee of notability, but they're right most of the time. Am I understanding this incorrectly? ‑Scottywong| [express] || 00:12, 7 November 2020 (UTC)

@SportingFlyer: See also this discussion on my talk page that happened recently. ‑Scottywong| [speak] || 00:17, 7 November 2020 (UTC)

The problem with how we currently present WP:N - it is there but its not clear - is that the GNG itself is still a presumption of notability and can be challenged too, just as an SNG. I've talked to before about adding the idea of this target level of "unquestionable notability" that a topic like WWII has that is the ultimate goal, one that there's no deadline to reach, but one that if you can prove impossible to reach if you've doing the sourcing legwork, then we're going to refute the presumption of notability. This then puts the GNG and the SNGs on par with each other. The only reason we tend to want the GNG to weigh more is that because it is closer to meeting WP:V, WP:NOR, and WP:NPOV implicitly by source requirements compared to the SNG. It's not really good to think "Oh, I reached the GNG, the topic's safe." because that's still just a minimum sourcing level. --Masem (t) 00:27, 7 November 2020 (UTC)

I've re-read WP:N from top to bottom, and I don't see anywhere that it says that passing GNG is a "presumption of notability". I do see it very clearly saying that passing GNG is a presumption that a topic deserves a standalone article. But, it does not say that passing GNG is a presumption of notability. Being a notable topic and deserving a standalone article are two different things. It's possible to have a topic that is notable, but does not deserve a standalone article because it violates other WP policies. By the same token, it's possible to have a topic that passes an SNG, but is not notable because it doesn't pass GNG. That is why SNGs are a presumption of notability, and GNG is the definition of notability, and they are not equivalent or substitutes for one another. Therefore, I think that whoever wrote the SNG portion of this page does not fully understand that GNG is the definition of notability, not a presumption of notability, and in my opinion it should be rewritten to make it more clear that GNG must ultimately be satisfied for all topics. Not immediately, obviously. But, if someone challenges the notability of a topic, and they do all of the BEFORE searches that you outlined above, and no one else can find sources to satisfy GNG, then the topic should not have a standalone article, even if it passes an SNG. ‑Scottywong| [soliloquize] || 00:34, 7 November 2020 (UTC)

I would propose that the first two paragraphs of WP:SNG are rewritten to be something more like this (bolded parts have been changed from the current wording):

In some topic areas, consensus-derived subject-specific notability guidelines (SNGs) have been written as a shortcut to help estimate whether an article is likely to satisfy the general notability guideline (GNG). The currently-accepted subject-specific notability guidelines are listed in the box at the top of this page and listed at Category:Wikipedia notability guidelines. These subject-specific notability guidelines are generally derived based on verifiable criteria due to accomplishment or recognition in that field that either in-depth, independent sourcing likely exists for that topic but may take time and effort to locate (such as print works in libraries local to the topic), or that sourcing will likely be written for the topic in the future due to the strength of accomplishment (such as winning a Nobel prize). Thus, we allow for the standalone article on the presumption that meeting the SNG criteria will guarantee the existence or creation of enough coverage to meet GNG.

These are considered shortcuts to meeting the general notability guideline. A topic is not required to meet a subject-specific notability guideline to qualify for a standalone article. Note, however, that satisfying the criteria of an SNG is only a presumption of notability, not a guarantee. In cases where it can be shown that GNG has not been met, the article may still be deleted or merged: a presumption is neither a guarantee that sources can be found nor a mandate for a separate page. ‑Scottywong| [prattle] || 00:43, 7 November 2020 (UTC)

I strongly agree with this change. Good work. SportingFlyer T·C 00:50, 7 November 2020 (UTC)

Seconded. _signed, Rosguill ^talk 02:45, 7 November 2020 (UTC)

Thirded. Reyk _YO! 11:07, 7 November 2020 (UTC)

Agree. That has always been my understanding – that, effectively, SNGs are "subservient" to GNG, and "passing" an SNG does not necessarily mean that a subject passes GNG. Indeed, I have specifically seen at least one case where a subject probably technically "passed" WP:NACTOR, but clearly did not pass WP:GNG/WP:BASIC and that artice was subsequently deleted at WP:AfD. --IJBall (contribs • talk) 01:02, 8 November 2020 (UTC)

Notice that the "presumed" is linked to rebuttable presumption, which explains exactly what I just described. This has always been the intent of the GNG and reflected by the AFD/BEFORE process - GNG is not an assurance that your article will never be challenged on notability terms ever again. (This has been discussed multiple times on these talk pages). That's why I do think it would be nice to add and stress there's a level where we want articles to get to where no one will challenge it, eliminating the rebuttable presumption, but this is present already. And that's the thing is that the GNG is not the ultimate definition of notability on WP. That ultimate definition is that the topic has been shown "worthy of notice" by sources, but of course that's vague as heck and full of qualification holes, so the GNG and the SNGs are guides to get editors towards that. --Masem (t) 01:11, 7 November 2020 (UTC)

Let me iterate why the GNG is taken as a presumption of notability itself. We are careful not to spell out how many sources or how deep the coverage has to be as this will be games, but I can say that 3-ish sources that have 2-3 para on the topic specifically will generally pass the bar for those looking to keep at AFD, assuming it is the topic's first visit to AFD and no other issues exist with it. That to me is reasonably fair to consider the same aspects that the SNGs are supposed to do: enough sourcing basis that a standalone makes sense to be put into mainspace so that other editors can work on it. This also correlates with the guidelines I've seen at those reviewing Draft articles and ACC for when articles are good to go to main space - just enough sourcing to pass this basic GNG test. But that should be clear that one might be able to pull out a paragraph or two of Wiki text from that and that's it. If the article can't ever expand because no additional sourcing ever comes out, we'd likely delete that at a second AFD as long as the one nominating showed their work. This is one reason why we have 2nd (and 3rd and so on) nominations for topics, because the nominators are often rechallenging the GNG aspect of the topic.

Mind you, I agree that the current way WP:N is presented does not make this clear enough, and in this I've spoken of trying to be clear about making sure that editors understand that there's the "holy grail" of notability that one should always strive for their article that no one will ever challenge, but until then, we presume notability (and thus allowance for a standalone) via the GNG or a relevant SNG. Establishing the GNG or the SNG is fine for creation of that standalone, but these are not conditions that you only need to pass once and never touch again; we expect articles to exceed these minimum standards and get as close to that holy grail of unquestioned notability. This wouldn't change anything with how WP is run or the like but simply make it clearer on the interrelationship between WP:N, WP:GNG and the SNGs and their application at AFD (eg it would not cause a mass AFD rush). I'd love to encorporate that idea but that would be a bit of reworking and I'd need to that off the primetime page. --Masem (t) 01:32, 7 November 2020 (UTC)

• I see Masem added a bit in May [2] which includes "Thus, we allow for the standalone article on the presumption that meeting the SNG criteria will guarantee the existence or creation of enough coverage to meet GNG." That contradicts what is said elsewhere. It was discussed and decided long ago there are multiple ways to determine something notable enough for its own Wikipedia article, either the GNG or a subject specific guideline, never did you have to do both. This is because you can be notable based on your accomplishments not on coverage you got. Not sure why we have to argue about this for years now. A scientist is notable for their scientific achievements, even if there was nothing ever written about them. Writers and musicians are notable for their work, even if they don't do interviews and give out information about themselves to be written about. Dream Focus 01:02, 7 November 2020 (UTC)
  • We want quality articles that go into depth on a topic, not one line stubs about a topic if that is all that is possible to write about them. A scientist that may have made an achievement that can be documented by a source but cannot be further documented at all is a problem for us as a standalone article. Mind you, we presume that certain achievements will lead to more sourcing via the SNGs and thus allow the standalones to exist until one can prove otherwise no additional sourcing ever will come, so the onus is on those seeking deletion to prove that out. But if they can, then that's a very valid reason to eliminate a stub that only contains one sourced fact. Mind you, we should find if there's a place to locate it otherwise via redirects and merging as well so the topic is still searchable. And just because these people may not give out information, others can talk about them and give secondary information to tell us why they are important in their field and notable, so we may lack a lot of personal details but we understand the importance of their work, which is valid encyclopedic content.--Masem (t) 01:11, 7 November 2020 (UTC)

    we presume that certain achievements will lead to more sourcing via the SNGs and thus allow the standalones to exist until one can prove otherwise no additional sourcing ever will come, so the onus is on those seeking deletion to prove that out. How exactly is one meant to do that, in practice? How does one prove no library has a book containing coverage on the person, or know the scientist will not end up curing cancer in the future? Yes, these are perhaps slightly exaggerated posits, but the point remains. Many times it's easily said "but you can rebut the SNG presumption!" but without valid ways on how to do it, with practical examples of it actually being done. And shows that it is actually done (successfully, proportionately) more than a relative handful of times. "Keep: Meets WP:SomeSNGHere" works pretty much every time. Permastubs are constantly kept of dead people with a few moments of fame, like 10 mins on a cricket field, most unlikely to have anything else about them. But how do you know they don't happen to be some prince, like the example we discussed the last time this convo happened, Special:Permalink/937542355 (it had a different title back then, as Bellingham Graham), which turned out to be a nobleman: Sir Bellingham Reginald Graham, 7th Baronet. This "rebuttal" of a SNG seems to be near impossible in practice, and completely impossible in practice for some SNGs (like WP:NPROF). I mean, you've all/read/closed seen more AfDs than me, so it's not like I know more than you folks, but I've never once seen a rebuttal of an SNG play out, and I feel convinced the philosophy reiterated here doesn't lining up with the practice. ProcrastinatingReader (talk) 05:58, 7 November 2020 (UTC)

Wp: Notability is a big complex kluge that operates in the fuzzy Wikipedia ecosystem which works most of the time wp:How_Wikipedia_notability_works_right_now but which is such a kluge that few can understand it. In this thread, you folks are trying find an underlying logical structure in a herd of cats, i.e. in a system which fundamentally does not have it. Of course, we can fix that, but that would take some complex work. Until then, please don't accidentally / inadvertently upset the apple-cart which currently mostly works. Sincerely, North8000 (talk) 02:27, 7 November 2020 (UTC)

It does not work, you have people arguing constantly in AFDs for years now. Someone will insist only GNG matters and the subject specific guidelines don't, and argue nonstop about it. What was added in May will just make people argue more. Dream Focus 03:41, 7 November 2020 (UTC)

Except there hasn't been any change in AFD behavior or the like. There has always been confusion on the intersection of the GNG and SNG, which I believe reapproaching it as I suggested (but would require a rewrite) would clarify, but it doesn't change practice or the like. Keep in mind underlying this all is the standard issue of editors wanting to keep the articles they feel are important, notability or not, which always is an issue at AFD. --Masem (t) 03:49, 7 November 2020 (UTC)

I'm hearing general agreement that the language currently used in WP:SNG is confusing, and is actively causing problems at AfD. However, I'm not hearing unanimous agreement about the relative importance of GNG vs. SNG as it relates to establishing the notability of a topic. Some people (including myself) believe that GNG is the ultimate test, and SNGs are shortcuts to assist with estimating the likelihood of a topic passing GNG. Other editors believe that GNG and SNGs are essentially equivalent: as long as your topic meets criteria of GNG or SNG, then your topic is notable.

I disagree quite strongly with the notion that GNG and SNG are equivalent, unless there is a discussion somewhere along the line where that became the consensus. I don't believe that this was true in the past. Even reading the top of WP:N makes it very clear how this all works. Summarizing the first couple paragraphs of WP:N:

1. Notability is the test that we use to determine if a topic deserves a standalone article.
2. Information must be verifiable; a topic shouldn't have an article if there aren't sufficient reliable, independent sources that cover it (hence, why a scientist is not automatically notable for their scientific achievements even if nothing was ever written about them, and a musician is not automatically notable for their work even if there is no coverage of it). If there are no sources about a topic, then there is no verifiable information to put into an article. This is why an SNG cannot be an independent notability test.
3. If a topic meets the GNG, it is considered notable by WP's standards. When a topic is deemed notable, it is automatically presumed to deserve a standalone article. However, that is a presumption, not a guarantee. A notable topic might not warrant a standalone article if it violates other WP policies (like WP:NOT), or if an editorial decision is made to merge the topic with other topics into a single article.
4. SNGs are shortcuts. They provide us with an accurate (but not perfect) estimate of whether or not a topic is likely to meet GNG. If a topic meets an SNG, it is presumed to meet the GNG as well. However, this is a presumption, not a guarantee. Topics must satisfy the GNG, but they need not satisfy an SNG. If a topic satisfies an SNG but its notability is challenged, and an exhaustive search yields insufficient sources to meet GNG, then the article on that topic might be deleted. While SNGs are usually a highly accurate indicator of whether or not a topic is notable, SNGs are not perfect, and there will always be a small set of topics that satisfy an SNG but not the GNG. In such cases, these topics might be deemed non-notable, despite satisfying an SNG.

Does anyone have major disagreements with any of the above statements? If so, can you explain the basis for your disagreement? ‑Scottywong| [prattle] || 05:07, 7 November 2020 (UTC)

Topics must satisfy the GNG, but they need not satisfy an SNG is not true for WP:NCORP I think.

Also, my understanding of WP:NPROF was that it stems from the observation that many academics have coverage of their work, not of themselves. Such likely do not meet the GNG, and many may never. But we will keep them anyway. So the whole "SNG is a rebuttable presumption of GNG" doesn't really hold up with how it is applied. And although I've read it many times, I'm yet to see any good examples of how it is 'rebutted'. In practice, at least in my experiences, if someone clearly meets an SNG, many times they're kept forever. There may be exceptions, but they are indeed exceptions. But really, at AfD, many of the SNGs are applied in different ways, so some SNGs may have more exceptions than others. It's a philosophical mess. ProcrastinatingReader (talk) 05:40, 7 November 2020 (UTC)

I still point out that the the idea that passing the GNG means that notability is satisfied forever is simply not true or how it is treated in practice, since numerous articles have been challenged repeated on failing to meet the GNG multiple times (and not having a SNG to fall under). The GNG is not a get-out-jail-free card for further notability evaluation if the sourcing is weak to just barely pass the GNG, or as WP's concepts on what reliable or in-depth coverage change. Mind you, once you have gotten to the GNG level, it becomes far more difficult to challenge the idea no other sources are present, but this option still does exist and is usually easier to show for more contemporary (post-2000) topics. WP:N may read a different way, but I'm describing practice, and that might point to a bit of chance in WP:N (namely being clear that the GNG itself is a presumption of notability to allow for a standalone). --Masem (t) 06:06, 7 November 2020 (UTC)

So just thinking this through more, and double checking on some history, what we have is a weird dicotomy of what's on paper and what's practice. Basically there's an unwritten rule at AFD that the GNG is the same as having 2-3 sources that cover the topic. (There's a reason we haven't set numbers to the GNG to avoid this type of gaming). Arguably a read of the GNG right now implies you need much more sourcing that than just 2 or 3 sources to be "worthy of note", but from an open-wiki standpoint, that's good enough to show V/NOR/NPOV being met to drop to mainspace and encourage other editors to help. The 2-3 sources is one of those things that just propagates a poor misunderstanding of polcy/guidelines. What I'd think we need to impress on editors that this implicit 2-3 sources is not sufficient in the long run per GNG/SIGCOV and just meeting that basic minimum threshold is a rebuttable presumption of notability, nothing compared to meeting all of the major points of the GNG with flying colors which all doubts of notability are gone. The easiest way to reflect this without stirring up through the other guideslines may be to add advice at the end of the GNG, like (wording flexible) "Topics where only minimal significant coverage through independent secondary sources has been identified may still be presumed notable and can have a standalone article, but editors should continue to seek out additional coverage through appropriate sources to fully demonstrate broader notability. If no additional sourcing can be found, such articles may be deleted or merged with a larger topic." The SNGs would be refering back to the GNG that has this, so this still makes them alternatives to the GNG with the other suggestions, but emphasis that a weak-GNG meeting article is still not clear of future notability considers. --Masem (t) 06:46, 7 November 2020 (UTC)

• There is no requirement for many or multiple sources. For example, I just commented on Health in Switzerland. I identified a good source for the topic which, by its nature, is reasonably complete and satisfactory for our purpose. No further source is required to demonstrate notability in this case. Other cases include biographies where a single good source, such as an entry in the DNB, is adequate to demonstrate notability.

Trying to make exact and precise rules for everything is not appropriate because there are always corner cases. These are guidelines and so explicitly allow for exceptions. They say that common sense should be used so that's what's wanted rather than more bureaucracy and intricate rules.

Andrew🐉(talk) 08:02, 7 November 2020 (UTC)

There is a need for demonstrating significant coverage coming from sourcing. What exactly are the standards for significant coverage is something we purposely leave undefined but you are right that that could easily come from one extremely high quality reliable third-party source like a biography of a person. But most of the issues when it comes to AFD is that we are looking at topics that are at the very lowest levels of "significant coverage" and editors are trying to justify how articles should be kept due to the number of sources, with the common fallacy that "oh, I have 2-3 sources, this article should be kept" when in fact there's zero point where we've set any number for that, nor implied that its just source count -- but its taken too often that way. But to stress: one can star an article that passes the GNG at very low levels of significant coverage at the initial stages of article development - we are far more tolerant at that point - but as an article persists and grows, if editors fail to improve the demonstration of significant coverage that would be expected for that topic, then its GNG presumption of notability will be challenges due to lack of significant coverage. (Eg commonly editors will do this in areas of fictional characters: they'll add one or two secondary sources then spend all their time expanding the in-unverse facets from the work itself which is not significant coverage, and then those end up being deleted because there's just simply no real significant coverage at the end of the day). --Masem (t) 14:17, 7 November 2020 (UTC)

ScottyWong, number 2, "Information must be verifiable", means you can verify what they have accomplished is notable. A website dedicated to a notable award list them as having won it for their scientific accomplishment. You can verify their work is featured in textbooks. Things like that. You can list information known about them, even if no reliable source has given significant coverage of them, just basic information. Dream Focus 08:37, 7 November 2020 (UTC)

The problem is that a lot of people look at the words "rebuttable presumption of notability" and, through some trick of eyesight or psychology, read "permanent unchallengeable exemption from WP:V and WP:N, and an automatic entitlement to a shrine". These are often also the people who wrote the SNG in question- SNGs that are usually ridiculous everything-is-notable bilge. This encourages people to go crawling through databases to find a statistic or two to bloat into a tiny substub. When the subject is a person, we usually haven't got any biographical information and sometimes not even a full name. That can be a worry from a BLP standpoint. And microstubs are generally useless because they contain so little information, have virtually no context for what little information there is, and send the reader flipping madly between multiple articles of the same type to get anything useful out of any of it. It's what's often termed write-only memory: written for the pleasure of the writer with no intention or hope that a reader will ever find it useful. Reyk _YO! 11:29, 7 November 2020 (UTC)

I see that all the time as AFD's, Ahh it just passes SNG so it passes GNG (or does not have to meet GNG).Slatersteven (talk) 11:33, 7 November 2020 (UTC)

Heck, it's why we're having this conversation now. SportingFlyer T·C 11:44, 7 November 2020 (UTC)

• As Masem notes above, the guideline links to "rebuttable presumption", a presumption that something is true "unless someone comes forward to contest it and prove otherwise." (emphasis added) This is an old and well-established legal concept that shifts the burden of proof. See Heiner v. Donnan, 285 U.S. 312 (1932) ("A rebuttable presumption clearly is a rule of evidence which has the effect of shifting the burden of proof.").

In this context, the party challenging notability of a topic that passes an SNG has the burden of proof. In order to challenge the notability of such a topic, it cannot be sufficient to simply say the topic doesn't meet GNG ... otherwise the presumption is meaningless. The real challenge (and one that has never been fully explored) is what type and level of proof must be presented by the challenging party in order to overcome the presumption. That is a discussion worth having.

WP:BEFORE simply requires a Google search, and such searches are woefully inadequate when it comes to many topics, especially articles about individuals and events occurring in the pre-Internet era. Accordingly, mere Google searches cannot be sufficient to satisfy the challenging party's burden of proof in rebutting the presumption with respect to such pre-Internet persons and events. This is not to say the presumption can't be rebutted, but the burden is necessarily and appropriately high in such cases. Cbl62 (talk) 14:12, 7 November 2020 (UTC)

Yeah, BEFORE really should be expanded to stress that a Google search cannot capture pre-2000 topics well, and physical library searches, sometimes even more local to the topic's area, may be required. --Masem (t) 14:30, 7 November 2020 (UTC)

Wikipedia policies and guidelines have toothless "vague general guidance" items and also statements which have specific operational mechanics/ teeth. In a situation where everybody is just trying to do the best thing, both are followed. In a situation where somebody is trying to work the system towards their desired outcome the toothless "vague general guidance" items are easily ignore-able and only the items with specific operational mechanics/ teeth are followed.

So there are two completely different answers to the question above. Where all are being followed, everything comes down to meeting the sourcing requirements of GNG and the SNG is just a shortcut to that. The beginning of GNG says that you just need to meet either GNG sourcing or the SNG, but SNG's pretty much all say (in toothless vague general guidance) that they are merely predictors of existence of sourcing that meets GNG and give deference to that. This completely changes when you have advocacy / somebody working the system. Then the toothless vague general guidance is always ignore-able leaving the only specific SNG criteria to rule. So the defacto rule in in these situations is that they need only meet the specific SNG criteria. North8000 (talk) 15:24, 7 November 2020 (UTC)

Arbitrary break

I feel like we're getting a little bit off-topic, talking about the permanence of notability, the exact quantity of sources required to satisfy GNG, where the burden of proof lies in an AFD, etc. While those are all interesting topics that probably deserve discussion, I think that the crux of the issue is whether or not GNG and SNGs are equivalent, or if GNG is the controlling principle and SNGs are secondary. In other words, if it can be shown that a topic satisfies an SNG, and simultaneously it can be shown that that same topic does not and will not ever satisfy the GNG, is that topic considered notable? I'd like to see if this is something we can get general agreement on here, or if it needs to be taken to an RFC. To be clear, my view is that GNG is the controlling principle, and SNGs are secondary. ‑Scottywong| [yak] || 15:42, 7 November 2020 (UTC)

• Agree- that the SNGs defer to the GNG, and not vice versa. Reyk _YO! 15:43, 7 November 2020 (UTC)
• Agree that SNG's are secondary to GNG, all they should be is a "work in process do not CSD, or AFD for a while".Slatersteven (talk) 15:46, 7 November 2020 (UTC)
• It's not that anyone's getting off topic. It's simply that the issue is not as simple as you pose. Saying that "GNG is controlling" and "SNGs are secondary" is a gross over-simplification of a topic with more nuance than you seem to acknowledge. Cbl62 (talk) 15:56, 7 November 2020 (UTC)

@Cbl62: I'm intentionally trying to ignore the nuance for now, but we can get to that at some point. I believe it's better to get agreement on a fundamental concept first, and then branch out from there to deal with the nuances, exceptions, and edge cases. If we can't even agree on whether GNG and SNG are equivalent or not, we have no common basis on which to discuss nuance. ‑Scottywong| [babble] || 16:04, 7 November 2020 (UTC)

The real crux of the issue is in the nuance. Making a sweeping proclamation, as you propose, without a concurrent discussion of, and consensus as to the the nuance, is misguided and simply serves to undermine the SNGs. Accordingly, and given that your proposal remains a gross over-simplification, I must Oppose. Cbl62 (talk) 16:10, 7 November 2020 (UTC)

• The problem is that people are looking for a definition of “notability”, and we don’t actually give one. We can’t... because the closest we can come to one is: “there is a consensus that this topic deserves an article”. It all hinges on consensus, and consensus is often messy.

Neither GNG nor the various SNGs are meant to be “definitions”... rather they are meant to be indications that help us REACH consensus. Blueboar (talk) 16:37, 7 November 2020 (UTC)

And this is sorta what I feel is missing from how WP:N is written (as explained above). There's a point that no one will think about deleting an article because the evidence of significant coverage is clearly there (WWII) but for articles that are clearly not at that state - particularly those that are starting out , we try to determine if there's potential (and thus reasonable to have a standalone article) by looking at a minimum amount of significant coverage given by the GNG, or where possible, when an SNG criteria is met. We're missing something in how WP:N describes this with the confusion of the GNG shortcut being treated as both the large scale test and the "minimum significant coverage" test depending on context. --Masem (t) 17:17, 7 November 2020 (UTC)

Right, I agree that we're not talking about the definition of notability, we're talking about indications that help us reach consensus on the notability of individual topics. My point is, what role do GNG and SNGs play in giving us indications of notability? In my opinion, meeting GNG provides a direct indication of notability, and SNGs provide an indication of meeting GNG. SNGs do not provide a direct indication of notability, and WP:SNG should make this more clear. ‑Scottywong| [verbalize] || 18:36, 7 November 2020 (UTC)

I agree on the SNG statement as being an indicator for notability, and that we don't have an exact definition of notability since that's consensus-based but we use indicators to judge a topic's likelihood to reach that (and thus get a standalone). But on the GNG, this is where I believe there's confusion between it being a definition or an indicator of notability, depending on the context of use. That is, at AFD, particularly when dealing with a topic that meets an SNG, "GNG" tends to be treated as a definition of notability, while in broader subjects, the "GNG" is treated as an indicator of notability. When may be why we have this confusion. If we make clear of something being the definition of notability, and something else being the source-based indicator of notability (one of these being assigned "GNG" for clarity). Does this make sense of what the issue/confusion might be? --Masem (t) 18:52, 7 November 2020 (UTC)

• Agree with the principle as stated above. I don't think this is as nuanced as we're making it with presumptions and indicators and definitions and all of that. If you have an SNG, you're presumed to merit an article not because the SNG is met, but because the presumption of GNG is likely met, because the SNGs are almost always tailored to what should meet GNG. There are exceptions, but that is the general principle as I've always understood it, making SNGs not "equal" to the GNG, but rather an indicator that GNG is likely met, especially for topics where source searching will be difficult. SportingFlyer T·C 21:05, 7 November 2020 (UTC)
• Disagree This has been discussed for years in various long RFCs and elsewhere. The fact some want to ignore the subject specific guidelines, doesn't mean you have the right to do so. It has always been A topic is presumed to merit an article if: It meets either the general notability guideline below, or the criteria outlined in a subject-specific guideline. Getting coverage in the media is not the only way to prove you are notable enough to have a Wikipedia article. We handle legitimate scientific topics in this encyclopedia, not just popular culture. Wikipedia:Notability (numbers) for example. We also a lot of articles for species which do not have any significant coverage anywhere, just brief facts mentioned in scientific databases. Hebeloma aestivale Hebeloma arenosum and others I see after briefly examining things in Category:Lists of fungal species and clicking on anything in those list which are 99% red links in many cases. Dream Focus 21:43, 7 November 2020 (UTC)

This RfC from mid-2017 produced several conclusions including: "There is clear consensus that no subject-specific notability guideline, including Notability (sports) is a replacement for or supercedes the General Notability Guideline." _wjemather^{please leave a message...} 11:51, 8 November 2020 (UTC)

That RfC from mid-2017 was devoted to WP:NSPORTS and had only incidental comments about wiki-notability guidelines for anything other than athletics. Hardly sounds like gospel to me. For that matter, it appears that a low threshold for inclusion of sports figures is the complaint people have, perennially, about subject-specific wiki-notability guidelines. The trouble isn't with the abstract relation between these two categories of guidelines that we've thought up; it's with whether that threshold is too permissive. XOR'easter (talk) 19:12, 8 November 2020 (UTC)

NSPORT may be the most glaring deficient in many areas, but several other SNGs (including NGEO, NMUSIC, NBOOK) are also problematic in places. It's clear both the SNGs themselves and the defined relationship between them and GNG need work. _wjemather^{please leave a message...} 23:01, 8 November 2020 (UTC)

• Disagree. I see no point in having subject notability guidelines if the general notability guideline automatically trumps them. Meeting one or the other should be enough to establish notability. -- Calidum 21:56, 7 November 2020 (UTC)

  @Calidum: the point in having them is to give editors some guidance as to what is going to be ok as a standalone article and what might not be. It says "if you're challenging this topic which meets an SNG you better have your research and facts in a row to show that it doesn't meet the GNG". This is, rightly, the dynamic I see at AfD as both a participant and as a closer. Best, Barkeep49 (talk) 22:01, 7 November 2020 (UTC)

• My take has been that some SNGs are equivalent to the GNG (e.g. Numbers or NPROF), some call for the strictest interpretation of the GNG (e.g. NCORP), and some are secondary to GNG (e.g NMUSIC). This makes them harder to understand, because you have to figure out which category a particular SNG falls into, and in some cases some parts of an SNG will be equivalent while other parts will be secondary just to add an extra layer of fun. However, this mixture does serve the encyclopedia on the whole by ensuring a diversity of topics receive appropriate coverage. The reason we can have these varying standards successfully is that any topic must be verifiable so regardless of the type of coverage given we can ensure that information conforms with our 5 pillars and core policies, which despite its prominence Notability is not (it's a guideline and for good reason). Best, Barkeep49 (talk) 21:58, 7 November 2020 (UTC)

• I think this is the best description I've read yet. I've agreed, because I'm coming at this from the general rule that all articles must meet the GNG guideline. There are exceptions to the rule where either simply being verifiable is enough because the information is important enough to keep, there are exceptions which use different sources to meet the notability criteria, there are exceptions which are stricter than just getting coverage, but at the end of the day, the vast majority of the articles on the website are required to meet the GNG presumption in order to avoid being deleted at AfD - this is why I've agreed with Scottywong above, as I think the general principle holds. SportingFlyer T·C 23:54, 7 November 2020 (UTC)

• Agree- that the SNGs defer to the GNG, and not vice versa. Johnbod (talk) 22:14, 7 November 2020 (UTC)
• Disagree. Clearly false for WP:PROF as explicitly written there, but also false for many/most other SNGs as used in practice. And I agree with the point of Calidum above that this attitude would make it pointless to have any SNGs. —David Eppstein (talk) 22:20, 7 November 2020 (UTC)
  • WP:PROF is not subservient to the GNG. Uniquely so, I believe. The evidence comes from AfD outcomes. —SmokeyJoe (talk) 22:28, 7 November 2020 (UTC)
    • It's not unique. WP:SPECIESOUTCOMES is another example, at least. —David Eppstein (talk) 22:31, 7 November 2020 (UTC)
      • Yes, good point. I was assuming the 13 listed SNGs, but there is the interesting question of overlap of the SNGs and AFDOUTCOMES. I’m not sure about the SNG WP:NUMBERS, but I think WP:PROF should be considered an exception to the rule that SNGs are sort of subservient to the GNG. I note that WP:PROF (and WP:CORP) had consensus as guidelines before WP:N. —SmokeyJoe (talk) 06:41, 8 November 2020 (UTC)
• The edited statement These are considered shortcuts to meeting the general notability guideline is not reasonable. What really matters is AfD. An SNG statement does not make the GNG test null. SNGs and the GNG do not “trump” each other, they are different perspectives on the question. The SNGs are predictors of whether the topic will meet the GNG and be kepted at AfD. The GNG is a more direct predictor of whether the topic will be kepted at AfD. —SmokeyJoe (talk) 22:24, 7 November 2020 (UTC)
  • The bit about "considered shortcuts to" was added by Masem back in May. I don't see any real consensus for having that snuck in there, since it contradicts years of consensus otherwise. I'm removing it unless there is proper discussion to add that. Dream Focus 22:36, 7 November 2020 (UTC)
    • My language was "alternatives" per [3]. It was changed to "shortcuts" here, but not my edit, and I disagree that in the stance, that "shortcuts" is correct, but this comes to recognize that the "GNG" is meaning two things here. --Masem (t) 23:10, 7 November 2020 (UTC)

      @Masem: you just undid an edit by Dream Focus referencing the May discussion. Your edit was challenged at the time and I don't think you had consensus from that discussion to keep the edits then either. So maybe Dream's revision (and my revision) aren't the right ones (perhaps Joe Roe could come up with something based on his comment below) but either way I don't think you can claim that discussion as proof that your edits had consensus. Best, Barkeep49 (talk) 23:15, 7 November 2020 (UTC)

      • There was a long discussion about the changes in May about those changes with multiple editors. It is not like that hasn't had multiple eyes and other people that also edited it, so while initial parts of it were in debate, the final version was worked out by multiple editors. --Masem (t) 23:46, 7 November 2020 (UTC)

      • Maybe we should remove the section on SNGs entirely for the time being, since there's not really a consensus on how it should read. SportingFlyer T·C 23:33, 7 November 2020 (UTC)
        • I agree. We got along without a section on SNGs for 14 years and adding it has clearly opened a can of worms. – Joe (talk) 23:48, 7 November 2020 (UTC)

          That might be the right option but it always seemed weird to me that we have this whole category of things that we never actually describe anywhere. I just did an edit where I basically took out all the detail and just went with a basic "what is it?". Does this approach, which still lets us decide, in practice and in the individual SNGs, how they relate to the GNG have consensus? Courtesy ping Masem. Best, Barkeep49 (talk) 00:07, 8 November 2020 (UTC)

• Disagree. I'm not saying what should be, I'm saying what currently is. This is not how notability is determined on wikipedia. E.g. NPROF, NCORP, and the fact that we have 500000 articles on small geographic features and train stations that there is no in-depth coverage of in reliable sources. Natureium (talk) 22:29, 7 November 2020 (UTC)
• I'll echo what North8000 said above. Looking for logic in the notability guidelines is like looking for fortunes in tea leaves. There is no logic, because nobody ever sat down to work out notability as a coherent concept; it's a piecemeal compromise that evolved to contain the deletionism vs. inclusionism wars. That's not necessarily a bad thing. Guidelines are supposed to document consensus, and we will never have a universal consensus on what to include and exclude. So if a few editors decide here that the GNGs trump SNGs, but that's not actually going to change anything. It won't magically make all the SNGs fall into line with the logic of the GNG. It won't stop editors in individual AfDs using whatever rules of thumb they find useful to make a decision about inclusion. It won't stop editors occasionally ignoring the guidelines altogether if that seems the best thing to do. It won't change the fact that there have always been topics that are to some extent exempt from notability, because we're also an almanac and gazetteer and having one-line stubs on some things is a perfectly reasonable thing in a general-purpose reference work.

Maybe in an ideal world we could scrap this mess and come up with a logical and coherent system of inclusion criteria, but that isn't going to happen. Recognising that, I think we could do worse than continuing to make decisions on a case-by-case basis at AfD, referring to guidelines written by subject-specialists where available, and falling back on the GNG otherwise. We don't need to worry about abstractions like this for that system to keep working. – Joe (talk) 22:35, 7 November 2020 (UTC)

I hadn't thought about it that way, but I think Joe is correct. Notability standards are kinda like the U.S. tax code. Little bits get added here and there as compromises on specific issues, and the result is what it is. One might aspire to a simplified system, but herding all the cats to get there is unlikely. Cbl62 (talk) 23:09, 7 November 2020 (UTC)

• Disagree per David Eppstein. I'm most familiar with WP:PROF, but there's no reason in principle why the same difficulties addressed by that specialized guideline can't also occur elsewhere. The scope of the Wikipedia project is human knowledge, and there's not going to be a single page of bullet points that sums up how to organize writing across the whole of that territory. XOR'easter (talk) 23:39, 7 November 2020 (UTC)
  • Addendum The intro of WP:N currently says, A topic is presumed to merit an article if it meets either the GNG or the criteria outlined in a subject-specific guideline listed in the box on the right. Language equivalent to this has existed since May 2007. XOR'easter (talk) 18:33, 11 November 2020 (UTC)
• Agree SNGs suggest the GNG is likely to be met, but they do not exempt topics from coverage whatsoever. They can shape what sort of coverage is expected but they should not be allowing for article which have zero significant coverage of the topic. Reywas92^Talk 00:10, 8 November 2020 (UTC)
• Definitely agree that, with only a few exceptions (and, FTR, I probably don't agree that they should be exceptions), SNPs are "subservient" to GNG. This has generally been the practice of most, and is the correct practice. The alternative is to allow SNGs to be manipulated to short-circuit the notability process adding even more articles on non-notable subjects to this project. --IJBall (contribs • talk) 01:14, 8 November 2020 (UTC)

• Agree Prof is the exception rather than the rule. Meeting GNG is crucial for giving us the info to writing a semi decent article. SNGs are just a presumption that the sources exist for us to do this, nothing more. AIRcorn (talk) 06:27, 8 November 2020 (UTC)
• Agree This RfC from mid-2017 produced several conclusions including: "There is clear consensus that no subject-specific notability guideline, including Notability (sports) is a replacement for or supercedes the General Notability Guideline"; and I can't say that anything has significantly changed with the SNGs since then. Too many SNGs (especially sections of NSPORT, e.g. NCRIC) are ludicrously permissive and only reflect what enthusiasts in that area believe is (or should be) notable, while completely disregarding Wikipedia's definition of notability. The result is 1000s of junk one-line perma-stubs. In most areas, SNGs main purpose should be to provide a good indication that subjects will meet GNG, but ultimately GNG is the test. _wjemather^{please leave a message...} 11:51, 8 November 2020 (UTC)
• Agree that this has been discussed several times. The WP:GNG is the only workable standard because how can you write an article without reliable and independent sources? The SNGs are only useful so far as they suggest things where sources probably exist. But in practice, if you don't have the sources, you have to at least re-organize the content into another notable topic, if not remove the article in question. Jontesta (talk) 21:03, 8 November 2020 (UTC)
• Agree SNGs are secondary to GNG and must defer to GNG. SNGs come into play only if the subject doesn't meet GNG. — Ad Meliora ^Talk∕_Contribs 10:01, 9 November 2020 (UTC)
• Oppose any effort to subvert the SNG's. Plus, like has already been noted above, for cases like professors, small towns, high schools, etc. it doesn't really work that way, anyways. Ejgreen77 (talk) 11:58, 9 November 2020 (UTC)

• Are SNGs intended to be secondary to/in the service of the GNG? Objectively no. Some of the SNGs came first, before there was even a main notability guideline, and the GNG was initially a "here's what the SNGs have in common" which then took on a life of its own.

Should SNGs be secondary to/in the service of the GNG? Tricky question. We've certainly been heading that way for all but a few of the SNGs. And there's an excellent argument as to why: WP:NOT + WP:V + WP:NOR + WP:NPOV means we need reliable independent sources that provide us with both a sense of what has received note by others as well as enough material to write a neutral article. I think that most people agree on this but get hung up on the implications, which of course are that a whole lot of our professors, villages, sportspeople, and some others wouldn't make the cut.

Do we want a rule that says reality tv contestants and internet memes are more notable than a professor or town, and which produces several forms of systemic bias? Or do we want a rule that requires us to have a bunch of material based on official websites, unpublished sources, databases, original research, etc. (or else permastubs that do nothing more than log someone's/something's existence)? If people have another option, I'd love to hear it.

If forced to choose, I'd have to go with the requirements of the GNG, I guess. I don't think it's a good thing that whether something/someone will be kept or deleted is sometimes so unpredictable, often with equal chances at AfD depending on which small subset of Wikipedians participate and which of the rules they decide to apply. It's good to avoid being entirely prescriptive, so yes, there will be deliberation where people have different opinions, but those people need to at least be arguing according to the same basic assumptions. Yes, our fundamental policies have some counterintuitive outcomes and produce systemic biases, but I've long felt conflicted by the idea of saying "here are the rules for most things, but we really like [professors, footballers, villages, et al.] so have different rules for them, even if it means we just ignore the rules for large numbers of articles". — Rhododendrites ^talk \\ 13:36, 9 November 2020 (UTC)

• Agree GNG is the biggest criteria and also the ultimate arbiter. The header of GNG is Wikipedia's meta statement regarding existence of an article. It says that SNG's are an alternate way in. As it should be, SNG's say that they are mere predictors of GNG sourcing, i.e. that they only temporarily bypass the requirement for NG sourcing. Until we can fundamentally tidy up this kluge, that is what the current wording says and we can't upset the apple cart. North8000 (talk) 14:00, 9 November 2020 (UTC)
• Agree that SNG's should be secondary to the GNG. At present the concept of SNG is seriously abused by groups highly interested in a specific subject area to allow articles for their favorite often barely notable subjects through creating exceptions to circumvent the GNG. Making the matters worse is that the bar at these SNG is often set ridiculously low and there is far too little community consensus for the creation and amending of these SNG's. This needs to be finally dealt with.Tvx1 16:46, 9 November 2020 (UTC)
• Agree in general. An SNG should not remove any requirements that are listed in the GNG. The example above mentions articles on Train Stations where there is no notable coverage - these articles would not meet GNG but are allowed under the appropriate SNG. In my opinion, if an article fails GNG, it should not be able to then pass an SNG. But I Disagree if the argument is in the reverse. The purpose of an SNG is to add additional explanations on interpreting GNG in the context of specific topic areas. An SNG should not add unnecessary restrictions but it should provide for how to interpret particular sources in order to establish notability. In general, my position is that an SNG can not and should not be "overturned" because of a lack of clarity or a looser explanation or interpretation of the criteria for establishing notability is available in the GNG. A good example is NCORP which adds further explanations on what is meant by an "Independent" source in the WP:ORGIND section. Many editors believe that the definition of "Independent Content" is a new restriction and not intended by the GNG and is therefore "overruled" by the GNG. This has resulted in many AfDs involving NCORP devolving into a messy and overly long debates on whether the perceived strictness of interpretation contained in the NCORP SNG is correct. HighKing⁺⁺ 18:07, 9 November 2020 (UTC)
• Oppose GNG fundamentalism. My own view, articulated in a subsequent section, is that each SNG should be understood as an interpretive guideline for the GNG in a specific context, and therefore not to be overruled by an appeal to some other reading of the GNG that ignores the guidance of the SNG cited. I also disagree with any interpretation that an article needs to meet both SNG and GNG, as if they could be evaluated separately and juxtaposed. They should always be read together, in my view. Newimpartial (talk) 18:55, 9 November 2020 (UTC)
• I agree in principle with Calidum (talk · contribs) in that I see no point in having subject notability guidelines if the general notability guideline automatically trumps them. Only that I see little point in having SNGs. They are merely advisory. GNG is the test of whether you have sufficient sources to write a policy compliant article. If "so-and-so won a gold medal in the 1904 Olympics" is the only thing you can ever write about them given the sources, then you don't have an encyclopedia article; you have a Wikidata entry. GMG^talk 15:45, 11 November 2020 (UTC)

What does the "GNG" mean

I've tried to note this above but I want to stress this again: I think we have a term , the "GNG", that has two different meanings depending on context, which is causing the problem. In one context, it is being used as the definition of notability , and in another context as an indicator of notability. This is easily causes the confusion of what the relationship should be between the SNGs and the "GNG"; the SNGs should be serving to meet the "definition" of notability (that is, they are presumed indicators of notability), but they should be on equal part with any other "indicator" of notability, which would be a bare minimum amount of coverage/sourcing. Unfortunately, even just reading the comments above, both versions are being implied when the "GNG" is mentioned, showing that we've created confusion on this term.

A means to resolve this is, roughly:

1. Established what is currently under the WP:GNG as a the "Definition of notability" that would be at a shortcut like WP:NOTEDEF (NDEF is taken). SIGCOV would stay where it is as well. This represents where we would like articles to get to where no one would question deletion of a topic.
2. The "General notability guideline" would become a new section that would basically saying that a topic is presumed notable to meet NOTEDEF if there is a demonstrated minimum amount of significant coverage in independent secondary sources, but where the article may be otherwise lacking the expected depth of content. We can't assign any numbers to this, but this is where most people would likely read this as being "3 sources" at AFD. This is now firmly establishing the GNG as an indicator of notability.

This would then have it be the case that most SNGs, where the line usually reads "A topic is presumed notable if it meets the GNG or one of the criteria on this SNG", that now this new GNG as an indicator of notability is on par with an SNG in terms of a AFD test. (NCORP excepted, of course, which would say that this version of the GNG is not good enough and one must use that SNG).

In this framework, the GNG and the SNGs still are presumptions of notability, and thus can be challenged if they cannot be expanded beyond that minimum level of sourcing/significant coverage as to get where NOTEDEF would expect us to be at (as well as other non-notability related factors). Here I think most agree that just having the indicator of notability is not a "never can be challenged for notability again" pass, and that making it clear that this new GNG approach with only minimal sourcing can also lead to notability challenges.

The major difficulty here is that the term "GNG" is engrained across WP as to have a good proportion of editors (based on my read of AFD results) to generally take it as the definition of notability, which if this change was made, would need to be retaught to get editors to ween off that. But in exchange, we can clearly state that you can likely indicate you will meet definition of notability (NOTEDEF) equally with either the GNG or an SNG (barring NCORP). --Masem (t) 23:41, 7 November 2020 (UTC)

I don't think this has any more consensus in the discussion above, or in May, or at actual practice at AfD, than your previous version. And FWIW I also don't support it. I think if we're going to be rewriting it, we should be rewriting it along the lines suggested by Joe and Smokey above which is the actual guideline because it's how it ends up getting implemented. Best, Barkeep49 (talk) 23:45, 7 November 2020 (UTC)

• I agree with Barkeep49. SportingFlyer T·C 23:46, 7 November 2020 (UTC)

Except in practice, the meaning of the GNG is already screwed up between the definition and the indicator of notability, even in this discussion thread; both versions are used, and you have to squint and read carefully to figure out context. Making two separate terns to break apart the dual meaning of the GNG is a needed first step; whether the rest of my steps are right or not I don't know but we're not going to get anywhere if we keep having "GNG" mean both the definition and the indicator of notability. --Masem (t) 00:01, 8 November 2020 (UTC)

I've been going to AFDs for years and two means "multiple sources", you don't need three to pass the general notability guidelines. And the guideline page should be perfectly clear A topic is presumed to merit an article if: It meets either the general notability guideline below, or the criteria outlined in a subject-specific guideline listed in the box on the right. To avoid constant arguments from people who don't seem to understand what the word "or" means for whatever reason, we should add in a sentence "you do not need to pass both the GNG if a SSG is passed, both are equal guidelines to determine notability." Dream Focus 23:49, 7 November 2020 (UTC)

Whether its two or three, that's a minor facet, but that's at least passing that indicator to allow a standalone article on a first pass. But if that leaves a stub article, and someone later does a proper search for additional sources (local library searches) and comes up dry as per BEFORE, that can still be deleted because there is clearly no way to expand it to meet the actual full expectation for notability, the initial presumption was wrong. (Of course, the effort to show that no other sources exists will be extraordinary to get there so its unlikely to be challenged, but that's a possibility). That's why the source-based thing, what I'm suggesting should call the GNG, is a indicator of notability, that is equivalent to an SNG and thus a rebuttable presumption . Which is why if we made the definition of notability clear and not on the same term as "GNG", we can say "you can indicate notability with either the GNG or an SNG, you don't need to show both." (or something like that). --Masem (t) 00:01, 8 November 2020 (UTC)

What is the "actual full expectation for notability"? – Joe (talk) 00:18, 8 November 2020 (UTC)

It's as impossible to say as where the line is drawn where we consider a topic to fail the source-based GNG indicator. But there are topic that are so well developed that we would never question their appropriateness of a standalone article, like World War II because of the combination of the sheer number and quality of sources and the in-depthness of the topic such that we have a fully comprehensive stand-alone article. A stub is not that. Its going to be something 100% determined by consensus, case by case, but the more independent, secondary sources you have that go indepth on the topic, the more likely you will get there. --Masem (t) 00:40, 8 November 2020 (UTC)

Let me reframe the question, then: if you have an AfD where half of the !voters say the article fails WP:GNG and the other half say it passes the relevant SNG, what is the appropriate outcome? I'm specifically thinking of Wikipedia:Articles for deletion/Kellen Gordon, where a rugby league was added to the SNG without verification or discussion (and it's since been removed,) and voters largely voted along SNG lines. SportingFlyer T·C 00:56, 8 November 2020 (UTC)

That's a wholly different situation where the SNG was modified without community overview. In which case, the SNG "criteria" should have been ignored as haven't no current consensus and the "GNG" indicator should have been applied (which given the state of article, likely would fail). --Masem (t) 01:20, 8 November 2020 (UTC)

@Dream Focus: Treating the SNG as equivalent to the GNG makes no sense. The result of such a policy would be millions of articles on topics that have receive little to no coverage in reliable sources, and therefore can have little or nothing (verifiable) written about them, except that they exist. As long as Billy Bob sat on the bench for one game with the Pittsburgh Pirates, he automatically deserves a Wikipedia article, despite the only coverage of the game being a brief documentation of the score. These types of articles could only ever be single-sentence sub-stubs that essentially only define the subject and assert its existence. Anything else would be unverifiable. Don't get me wrong: the SNGs are accurate predictors of notability, and they're right at least 95% of the time, if not more. But they're not perfect. There are going to be Billy Bobs that technically fall into a category in an SNG, but just aren't notable by Wikipedia's standards, and have no sources from which to write any content about them. GNG has to be the controlling principle for any of this to make any sense. ‑Scottywong| [verbalize] || 07:28, 8 November 2020 (UTC)

What SNG mentions anything about something so trivia? They have information information available about them to pass the requirement to verify them, just not enough significant coverage of them to meet the GNG. Dream Focus 08:24, 8 November 2020 (UTC)

@Dream Focus: Many NSPORT guidelines specify a solitary match (the level of which is often also a very low bar) as sufficient; in some sports, e.g. cricket, this may involve as little as sitting in the dressing room for the entire duration. As such they are not equivalent to, or a substitute for, GNG. _wjemather^{please leave a message...} 15:30, 8 November 2020 (UTC)

Then fix that guideline, don't try to eliminate/ignore all subject specific guidelines. Dream Focus 15:32, 8 November 2020 (UTC)

We are clearly not talking about one broken SNG here; if we were, this discussion would be taking place there. The problem is that the majority of SNGs are not functioning as an adequate alternative guideline that is equivalent to GNG, and in places there is strong resistance any change of the individual SNGs that may result in deletions. As such, the obvious solution is to clarify/modify the relationship so that all guidelines become fit for purpose. _wjemather^{please leave a message...} 15:53, 8 November 2020 (UTC)

But this comes to my point that I think we have two different meanings of the GNG running around on WP that creates confusion: are we saying equivalent to the "GNG" being the definition of notability, where there's a clear plethora of sourcing, so there's no question of the topic's worthiness of note, or are we talking the "GNG" being the indicator of notability where just having 2-3 sources of coverage is sufficient? As I've tried to outline, the SNGs would be secondary to the "definition" meaning, or would be equal to the "indicator" meaning. (I do not think anyone reasonably is saying that the SNGs should be treated as equivalent to the "definition".) It causes people to talk past each other because they're referencing different definitions when refering to the "GNG". --Masem (t) 16:00, 8 November 2020 (UTC)

We are clearly not talking about one broken SNG here; if we were, this discussion would be taking place there. But it really does look like we are talking about one broken SNG: WP:NSPORT. That the conversation is happening on this particular Talk page doesn't change that. XOR'easter (talk) 22:53, 8 November 2020 (UTC)

The reason why y'all are having such a hard time is because these fundamentals are missing:

1. Define this "wp:notability" attribute. The definition is not anywhere in Wikipedia.
2. What is the purpose / goal of wikipedia screening for / based on wp:notability?

I think that collectively the wp:notability ecosystem knows the answer. But until we can distill it and write it down I don't think much progress on the above can be made.

My best attempt at an answer to #1&#2 is: After a topic passes the wp:not test, a measure of to what degree is it a good idea for it to be an article in an enclyclopedia that is "only" going to have tens of millions (not billions) of articles, and a gatekeeper to screen for that. The primary chosen metric is GNG type sourcing (partly because it also serves the purpose of enabling building a decent article) Secondary metrics are:how strongly it passed wp:not, and possibly the qualities defined in SNG's.

Whether SNG's are or should be a part of the fundamental definition, or shortcuts or predictors or should even exist is the question that can then be worked on. Also whether the GNG sourcing is provided vs. merely predicted. Sincerely, North8000 (talk) 02:45, 8 November 2020 (UTC)

The answer to your very last point is a difficult one, and is probably different for pre-2000 and post-2000 topics. In the case of the later, I think WP:NEXIST-type arguments should be viewed with suspicion at best, and outright skepticism at worst, as nearly all "notable" post-2000 topics should have at least some "internet" coverage. This of course would certainly not be the case for a subject from 100 years ago. But definitely even in the case of pre-2000 topics, I believe the acid test is GNG, and that the SNPs should be viewed as subsidiary to that. --IJBall (contribs • talk) 03:04, 8 November 2020 (UTC)

I agree, but my post was more to outline a viable process to sort that and other things out. North8000 (talk) 03:51, 8 November 2020 (UTC)

My impression of the main purpose of our notability guidelines is that we are trying to create a distillation of world knowledge, whose filtration process creates greater quality and reliability and less spam and cruft. It is also a useful thing to create an unfiltered accumulation of everything that can be accumulated (see archive.org) but that's a different project than ours. In AfDs, I think there are two common mistakes regarding notability. Newcomers to Wikipedia are often confused between "notability", meaning eligibility for a Wikipedia article, and notability with its colloquial English meaning, roughly synonymous with significance or importance. Enthusiasts for some topic believe that their pet topic is important, and therefore that it is notable, without regard to whether it meets the Wikipedia definition of notability. But among longer-term editors, there is also a belief that notability can only mean in-depth coverage in published sources, that when other people say that a topic is significant but poorly sourced that they are speaking in oxymorons, that topics with multiple in-depth published sources should always be considered notable, and that when spam and cruft gets through it can only be addressed in one-size-fits-all quantitative ways, by increasing the requirements for the number of sources or the global reach of the sources we allow for all topics. I think those are mistakes. I think that we need to filter both for significance of the topic and for our own ability to create a sufficiently-high-quality article on the topic, and that those two types of filtration need to be considered separately rather than lumped together. I think that many spammers can achieve any GNG-like bar we might set on availability of sources, that setting that bar higher keeps out good content while failing to keep out spam, and that the way to keep out spam is to have separate criteria for significance. I think that combining a broad interpretation of GNG that allows interviews and local newspapers to count as sources, together with a more critical eye than we usually take to the credibility of those sources for that specific content, is adequate for the ability to create high-quality content. I don't think that getting your press release picked up by a major newspaper gives it higher quality than a carefully-reported independent story in a local newspaper. It might be correlated with greater significance, but it might also be correlated with a bigger press budget. I think that, in many areas, significance can and should be better measured by subject-specific measures. For instance, the much-lamented "if you're an Olympic athlete then you're notable" is a perfectly good measure of significance. Where it falls down is as a measure of our ability to create a high-quality article. So I think we would benefit from a greater separation of significance from GNG into SNGs, together with a different and more general criterion, closer to verifiability than notability, that allows local and non-independent sources such as interviews, press releases from employers, etc., to be used as long as they would be considered reliable sources by our standards, but that only attempts to judge whether the content is there and not how significant it is. —David Eppstein (talk) 07:03, 8 November 2020 (UTC)

I think there's some wisdom with this. But I don't think we can retroactively apply it to some SNGs that were not developed with this concept in mind. I think your example of NYOLYMPIC still would have community consensus and support in this model. However, other parts of NSPORT, I suspect, would not. And I can't imagine doing this criteria by criteria at all the SNGs that we'd need to. So I think we're instead left with our current less documented, more susceptible to who shows up at a given AfD model as the one that is practical. Best, Barkeep49 (talk) 17:53, 8 November 2020 (UTC)

At least part of #2, to me, is that notability serves as a means to balance the fact anyone can create an article without input (potentially allowing vanity articles to be created) while also managing the fact we have a diverse array of topics that have different levels of importance to different editors, and assuring that articles are given a chance to develop in an open wiki environment. Notability assures that at least some independent sourcing exists to avoid pet topics or worst, promotional situations (what NCORP separately has to deal with). That we have the "general notability guideline" as well as SNGs means that we don't judge topics to a common standard of expected quality of sources (like academic journals) but only to the same relative terms of independent secondary sourcing for that topic's area. And that because we are using presumptions of notability, this gives articles an indefinite amount of time to be worked on by all editors to add more sources as long as you've providing indicators that the topic is notable, until the article gets to a state that its clearly an encyclopedic-quality article. --Masem (t) 15:09, 8 November 2020 (UTC)

It means people can send things to deletion discussions constantly, and argue nonstop over the same things, with random results based on who randomly notices and shows up to participate, and the personal bias of whatever administrator closes it. Geological locations, species articles, articles for those who made it to the Olympics but didn't have a publicist or do interviews so they passed the GNG, etc. Either accept the subject specific guidelines, or just get rid of them, no sense arguing for years in thousands of AFDs over this. Those articles will likely NEVER pass the GNG and you know that. Dream Focus 15:25, 8 November 2020 (UTC)

The SNGs are being accepted, but they remains presumptions of notability, which is very difficult to challenge to prove no additional sourcing exists if it a pre-Internet topic. I think WP:BEFORE needs to be better written to explain what nominators are expected to do before sending an article to AFD in the cases where the article meets an SNG. But beyond that, we can't control how certain AFD discussions happen; the deletion sorting project has done has much as possible to make sure those are sorted to the appropriate topic area categories to draw interested editors. --Masem (t) 15:43, 8 November 2020 (UTC)

@North8000, I think you're wrong. We have already written these things down. The opening sentence provides the definition for your #1 ("a test used by editors to decide whether a given topic warrants its own article") and the answer to your #2 (and an elaboration on your #1) is at WP:WHYN.

What we haven't done is taken the WHYN concept back to the SNGs to evaluate whether they're working. NPROF on "anyone who writes a textbook used at multiple universities gets a BLP, even if your only sources are the textbook's copyright page and the university bookstores' ordering page" is broken. NSPORT on "anyone who sits on a bench" is probably not broken for recent decades, but it is broken for sports from a hundred years ago.

@Scottywong, thanks for trying to fix this. These discussions have always beset by worries about whose ox will be gored, and I think now there is a good deal of pessimism that anything will ever improve because of these parochial efforts to give Alice Expert her very own page, and to never merge her into a broader subject that could be more encyclopedic and more appropriately sourced. After all, who would want an article on 1904 Olympic athletes from Ruritania, when you can have a single-sentence stub and a mostly empty infobox for each of them? Or a larger article on Scientists at Big U, when you could have a bunch of separate articles sourced to their own publications? Or a pile of articles about small high schools that nobody can find any information on, especially if you excluded routine coverage of their sports teams?

Digression: I once asked the promoters of Wikipedia:Notability (schools) about a school that existed for one or two years in the 19th century. They were convinced that if it ever issued a high school diploma, we needed a separate article about it. Problem: What would you title the article when you don't know the name of the school (or even if it technically even had a name)? "Unknown school somewhere in This County in the 1870s"? And what would you say about it that couldn't be put into a single sentence of another article?

One of the things I've been thinking is that we need to get some WP:NOT-level agreement that independent sources are an absolute requirement for every single article. A broad, firm agreement that we must be able to find independent sources (at least one or two) for every separate, normal/non-navigational encyclopedia article, with absolutely no exceptions, would go a long way towards solving some of these problems. AFAICT we don't really have that agreement. At most, we seem to have an agreement that independent sources is required for your articles, but not for mine. WhatamIdoing (talk) 18:37, 8 November 2020 (UTC)

@WhatamIdoing: I wholeheartedly agree with just about everything you say. There seems to be a population of editors that prioritize the quantity of articles, arguing for individual articles on as many things as possible, rather than considering taking the small amount of info you can say about borderline notable things and combining them into one article that is far more useful than 10 perma-stubs. The information will still exist on WP, it's not going anywhere, it's just being organized better. Your reasoning shines a light on why GNG exists, and why it should be the controlling principle when it comes to gauging notability. The one thing I'll disagree on is that every article must have independent sources referenced, or they get nuked. While I can't argue that WP would be a better place if that was true, I think I'm still comfortable with the idea that every article only needs to demonstrate that sources exist and are available, in order to escape deletion on notability concerns. This allows WP to remain a fluid place where the creation of new content is easy and encouraged, and if we're being practical about this attempt to clarify GNG, we'll need to acknowledge the other side's perspective and meet them in the middle. ‑Scottywong| [squeal] || 19:53, 8 November 2020 (UTC)

There is definitely a problem with some editors feeling everything should be a standalone article and not comfortable with some topics being covered in larger topics: the whole mess recently over Theresa Greenfield was an example: a person with no notability until she ran for a congressional seat, and only because she appeared to be taking a lead was there a concern about her not having a standalone rather than covering her in the appropriate election article... but she has since lost her race and all notability guidelines indicate she's not notable for a standalone now. Whether that article should stay now or not is a separate question, but the issue before was the fact that editors were upset that her coverage - otherwise failing all notability guidelines - was relegated to the election article but that's exactly our notability guidelines. Or alternatively, I've seen editors complain that we are "deleting" content when we merge and redirect non-notable topics to larger spaces, even after explaining there's ways to get the old content without any admin help. The idea that every topic must be standalone is problematic and related to notability problems, and something we need to cull down. No printed encyclopedia would do that, and while we're not print, we are not going to segment things down so far too that level. --Masem (t) 15:38, 9 November 2020 (UTC)

I've argued many times at AfD for marginal topics to be merged, e.g., for stubs on journals to be merged to the article on their publisher. But identifying uniquely satisfactory merge targets is not always possible. Do we talk about an expert at the page for the university where they work? (And at which one — people move over their careers.) At the article for their research topic? (Many experts have more than one major contribution.) The natural unit of prose is an article on the person, in almost all cases. XOR'easter (talk) 18:51, 8 November 2020 (UTC)

David and Masem, (or everyone) just to help sort out your answer, lets say that there is a person who has absolutely nothing real-world notable about them in any venue and no accomplishments except everyday life type ones. Just a regular person of the billions out there. But, for some random unknown reason, true solid GNG sources exist on them, sufficient to write a thorough non-promotional article. Should wp:notability be designed to let that article happen or prevent it from happening? This is only to try to develop an answer to #2. North8000 (talk) 18:44, 8 November 2020 (UTC)

For example, the answer could be "yes we're trying to be selective, but as a practical matter, we need to stick with GNG because it usually accomplishes that" North8000 (talk) 18:57, 8 November 2020 (UTC)

I would say no. If an article does not have any claim of significance at all (would not pass our A7 speedy deletion criteria) we should not have it. That's a much lower bar than GNG or the SNGs, though. As for "anyone who writes a textbook gets a BLP": no, that clause of WP:PROF (#4) requires both multiple books and wide use of those books. But it's also almost never used, and could be removed from WP:PROF with essentially no change to almost all outcomes. Most academic AfDs hinge on citations to or reviews of the academic's research publications (that is, independent and reliably-published sources about what the academic has supposedly done to become notable), the really notable academics have thousands of citations or dozens of book reviews, and even the non-notable ones generally have many more than the "multiple" required by GNG. So the claim that PROF is broken, based on an inaccurate description of a clause that is never actually used, seems dubious at best. —David Eppstein (talk) 18:55, 8 November 2020 (UTC)

Yes, WP:PROF#C4 is a significantly higher standard than just having written one book, and it's seldom invoked. Offhand, I can't recall an instance where an AfD was closed as keep based on C4 and C4 alone; it and C7 ("substantial impact outside academia in their academic capacity") are the unloved children of the guideline, it seems. XOR'easter (talk) 23:10, 8 November 2020 (UTC)

Given that notability is an extension of WP:IINFO, no we should not. We are not a Who's Who, and just because there are, for some reason, indepth independent sources about a a everyday joe with no other sign of importance, we shouldn't have an article just because the sources exist. But I cannot imagine such a situation that could occur without failing all the other aspects of WP:N: multiple sources for indepth coverage, enduring coverage over time, and not just a burst of sources, particularly not just for one event (BLP1E), etc. It's just an implausable situation based on how we know reliable sources cover topics that I can't see this happening. (unreliable sources, or promotional ones as NCORP worries about, on the other hand...) --Masem (t) 19:36, 8 November 2020 (UTC)

People are talking as though this were an abstract question, but I'm not sure that it is. We have plenty of articles about figures whose claim to Notability is "has X subscribers on YouTube" or "has Y followers on Instagram". Many of these people have enough RS written about them to meet NBIO and the GNG, but I'm not at all confident that they necessarily meet the "something real-world notable about them" criterion proposed by North8000. So I for one would be interested in hearing what people think about these actual cases. Newimpartial (talk) 19:45, 8 November 2020 (UTC)

This is sorta why we need secondary sources from RSes in the area. I know in following RSes about social media influences that "having X followers" is how an influencer's important is measured, and where having X is generally in the millions or higher is where these draw more attention in the mainstream sources. It's a way that the "social media industry" (as such what that is) has come to identify who are key players. So we know this is a factor that is considered important in that circle. This is why we have to consider what's in that field and what's determined to be signs of importance and notability. BUT at the same time, we don't want these to be overly inclusive - eg years ago we had a problem where fans of mixed-martial arts had created far far too many articles on all sorts of topics in that area because they had deemed it important to themselves, but created this walled garden of information, so we had to stop and drastically trim that back. So part of this is, to some degree, an WP:AUD issue, to make sure when we are looking at the sourcing that covers the field in general that it includes works outside of it so that we know there's understanding how the field connects to the real world. So like looking at social media, there's plenty of articles on CNN, NYTimes, etc on these people that talk about these people and their millions of followers, so we know we're using the right metrics and why some of these people are appropriately notable. --Masem (t) 20:09, 8 November 2020 (UTC)

If an article came to AfD and the only arguments for keeping it are "they have X subscribers on YouTube", I'd !vote to delete, and my guess is that I'd be aligned with the consensus on that. On the other hand, if an article came to AfD and a search turned up multiple WP:RS profiles of the subject's work as a video essayist, showing a sustained interest over a prolonged period of time, I'd !vote to keep, and I don't think that opinion would be too controversial either. I've seen cases of both types come through AfD, and that's how they seem to play out in practice. XOR'easter (talk) 20:15, 8 November 2020 (UTC)

If editors are coming to AFD with just the claim "But they have X followers!" and not providing RSes that are making that stance, that's a problem and doesn't help towards notability. Just pointing to the subscriber count on YouTube means nothing for our purposes. But if they come with reliable sources that have made a point that the person has X followers (which is usually going to accompany more details about that person if coming from quality RSes), that's different. --Masem (t) 20:30, 8 November 2020 (UTC)

From a standpoint of encyclopaedicity, I for one am not convinced that RS that confirm that "the person has X followers" necessarily contribute to Notability if they do not support a claim to Notability that goes beyond "has X followers". Even if these sources give some mundane details about the person's life, I don't think this implies anything real-world notable about them in any venue or any accomplishments except everyday life type ones. Basically I am inclined to see these as BLP1E instances where the 1E is "having subscribers/followers". I don't hold this view for an influencer who has created any kind of cultural work, or who has participated in any political controversy, or who meets something like NORG requirements for their personal brand. But for run of the mill YouTubers or Instagram influencers, my suspicion is that there may not be any there there, even if a few RS exist. Newimpartial (talk) 23:12, 8 November 2020 (UTC)

I basically agree with this. Essentially an article that says "this person has X followers" and "that person has Y followers" is akin to being listed in an "index" (tertiary source?). I was in an AfD once where someone tried to claim a listing about a TV series in a book that was basically an index of "All American TV Series 1946–1996" should count as "independent coverage", and I'm pretty sure that view got no takers. A "this person has X followers" article is basically in the same vein. This is why for WP:BLPs, I usually end up looking for in-depth profiles, or at least an in-depth interview on more than just the subject's latest project. But it should perhaps be clearer that "index"-type sources do not contribute to notability under GNG. --IJBall (contribs • talk) 02:36, 9 November 2020 (UTC)

That probably depends on the index: if it's meant to be exhaustive, then inclusion probably doesn't contribute to wiki-notability, while if it's more selective in scope, it might. We run into this not infrequently in AfD's about academic journals, an area where some indices aim to be comprehensive and others are narrowly curated. XOR'easter (talk) 02:53, 9 November 2020 (UTC)

Certain a source that is only documenting that a person X has so many viewers and no other details about X is not significant coverage in the first place, and again, unless you squint and consider this to fall under WP:ENT #2, this meets no SNG. But certainly a source that explains that X has so many viewers and goes into why (such as what activities they have done, a brief history of how they got there, etc.) is then going to provide the significant coverage and a partial indicator of notability. --Masem (t) 04:14, 9 November 2020 (UTC)

Perhaps, but if "what activities they have done" consists of things like "posted their breakfast to the 'gram", I would not regard that as providing support for a claim to Notability, regardless of how much detail the source offers about said breakfast or similar moments of ordinary life. Newimpartial (talk) 04:30, 9 November 2020 (UTC)

But that's why we look for multiple sources too, and in such a case, if multiple reliable sources all discussed this person in depth because they gained millions of follower because they posted their breakfast to social media each day, it's hard to dismiss that as a sign of "worthy of note" (There's obviously other facts, making sure this isn't a burst of coverage, for example). --Masem (t) 05:40, 9 November 2020 (UTC)

Upon reflection, I think the best way to litigate "real-world notability"/"accomplishments" issues would be in terms of their "credible claim to significance", and leave the GNG (and SIGCOV) out of it. Newimpartial (talk) 20:28, 9 November 2020 (UTC)

Well I think that a partial answer to #2 is clear. The mission of wp notability is to do some type of vetting beyond just seeing if there are sufficient sources to write the article. And it's clear the GNG sourcing, beyond being just being an indicator (and possibly the universally accepted and used indicator) of this quality, is (only)one of the attributes that define this quality ("only one" as we decided above) What are the other ones that go into defining this quality? The qualities that the SNG's define? Degree of encyclopaedicness? Real world famousness? Importance / impact, some other things? Sincerely, North8000 (talk) 21:15, 8 November 2020 (UTC)

Language clarified

Given the apparent 2-to-1 consensus above at #Arbitrary break and the explicit consensus at this RfC, I have boldly made some small changes to a few sentences of the SNG guideline, in an attempt to clarify the role that GNG and SNG play in determining notability, and bring them in line with consensus. I believe that these changes are supported by consensus. I'd appreciate if we could discuss the changes here before considering reverting them. The changes can be seen here. ‑Scottywong| [squeal] || 15:24, 9 November 2020 (UTC)

Wow, ok. I didn't even have time to create this discussion section before User:Newimpartial reverted the edit. I suppose I should have seen that coming. Newimpartial, can you explain how you concluded that the 2017 RfC that I referenced is "clearly an overreach"? It doesn't seem clear to me that it was an overreach. The closure of that RfC notes "clear consensus that no subject-specific notability guideline, including Notability (sports) is a replacement for or supercedes the General Notability Guideline." Yes, that RfC is 3 years old, and consensus can change over time, but can you provide evidence that consensus has changed? And it seems clear to me that twice as many editors above support these changes than oppose them. ‑Scottywong| [soliloquize] || 15:31, 9 November 2020 (UTC)

I agree with Newimpartial that the NSPORT RfC can't be used a justification to make such a significant change to WP:N (affecting every other SNG). There might have been a local consensus about the relationship between SNGs and the GNG there, but many editors will have assumed it was only about sport and not participated.

As for judging the consensus in this discussion, my reading is that there's a distinct lack of one on this issue. But also: the discussion is only three days old and still going; you have participated extensively in it; and headcounts are a poor gauge of consensus. I don't think it was appropriate for you to judge the "apparent consensus" at this point. A technicality it may be, but making a definitive statement on this issue will have wide-reaching effects on dozens of other policies, so there should be at least be a clear consensus and proper uninvolved close, and preferably a widely-advertised RfC. – Joe (talk) 15:48, 9 November 2020 (UTC)

Again, the 2017 RfC found "clear consensus" that SNGs (not just WP:NSPORTS, but all SNGs) are not a replacement for the GNG. Why exactly do we need to have another RfC for this? Can you point to any discussions since that RfC that indicate that consensus may have changed? ‑Scottywong| [confabulate] || 16:03, 9 November 2020 (UTC)

Even if we have that second RfC, and it arrives at the same conclusion, I 100% guarantee it still won't count for whatever weak-sauce reason. Reyk _YO! 16:20, 9 November 2020 (UTC)

I don't believe this to be true at all. An RfC held here at WP:N - the overall project page - would be binding in a quite different sense from one held at a specific SNG page. Newimpartial (talk) 16:29, 9 November 2020 (UTC)

The 2017 RfC was held at the Village Pump, not at a specific SNG page. ‑Scottywong| [babble] || 16:49, 9 November 2020 (UTC)

Worser and worser. In fact, I would go further to say that it wasn't a properly formulated RfC in the first place, that it didn't generate appropriate community notification, and that the NAC that was made was thoroughly out over its skis. Calling it a "local consensus" is actually pretty generous: at best, it should be understood as trimming back ridiculous claims made under NSPORT criteria, which is generally how it has been understood IMO. Newimpartial (talk) 16:52, 9 November 2020 (UTC)

But it was also explicitly framed as a question about NSPORT, and many editors will not have paid much attention to it for that reason. The whole point of WP:LOCALCONSENSUS is that while focused discussions frequently can reach a consensus on wider issues, those conclusions can't override broader community consensus if the participation was limited. I don't necessarily disagree with this change, but an honest reading of this discussion, and many, many others over the years, is that the relationship between the GNG and SNGs is something the community is split on. – Joe (talk) 18:41, 9 November 2020 (UTC)

• Needless to say I support the proposed change as reflecting both the 2017 RfC and the emerging consensus here. And I strongly disagree with the mischaracterisations of the 2017 RfC as somehow being "overreach" or "local consensus only". Reyk _YO! 15:47, 9 November 2020 (UTC)

• We're jumping too fast. Again, I stress the problem at the core: what is the meaning of "GNG" when editors point to this page (the larger definition, or the indicator of that definition), which affects how we define what relationship the SNG have (as an indicator/shortcut, or as an alternative, respectively). And yes, figuring out what we are going to call the GNG is going to have repercussions on other SNG pages to make sure we're all consistent. --Masem (t) 15:52, 9 November 2020 (UTC)

I've read the comments you made about this above, but I don't understand the confusion you're referring to about GNG. GNG is very clearly stated. It is WP's definition of notability. If a topic satisfies the requirements set out in GNG, then it is notable. Once we've established that a topic is notable, then that topic is presumed to warrant a standalone article. SNGs are different: if a topic passes an SNG, then it is presumed that it will also satisfy the requirements of GNG, and therefore it is presumed that a standalone article is warranted on that topic. I don't have any confusion regarding GNG's definition and indication of notability: they are the same thing. If a topic has sufficient indications of notability (i.e. significant coverage in independent reliable sources), then by WP's definition, it is notable. ‑Scottywong| [speak] || 16:17, 9 November 2020 (UTC)

But from countless AFDs, we have the case where have the barest amount of sourcing (eg "3 sources") some consider enough to "pass the GNG", but not sufficient that if more sourcing can't be found via BEFORE or at a later time, the article should be deleted. That it, there does exist in practice a "source-based indicator" of notability, wholly distinct from the definition, that is often referred to as the "GNG". Whether this is correct or not , I don't know, but we absolutely need to be clear on this page and make sure all editors know going forward that the definition should be referred to as one thing, and the indicator as something different. Whether we put "GNG" to the definition or not, that's a question of which would cause the least disruption. But I'm all for saying that the SNGs are indicators of meeting the definition of notability.

The alternative is to accept that the GNG is a continuum, from the barest minimum of sourcing to something as extensively sourced as WWII. But this still remains in practice rebuttable for notability particularly if you are floating on the minimal sourcing end - if you can only pull three sources after a comprehensive BEFORE search and the article remains a stub, we'd likely delete at AFD under current practice. Which is why framing the GNG as the definition of notability as an absolute "you've passed notability, never have to worry about it again" is sorta bad because we also will challenge that and people will game the ability to get away with minimally-sourced articles.

That's why I think its better to establish a wholly separate source-based indicator of notability, and make it clear that , just like the SNGs, articles that may pass that source-based indicator are not in the clear and should continue to work to improve to meet the definition. That's a better match of practice, but it comes to making the terminology for practice a bit more explicit. --Masem (t) 17:32, 9 November 2020 (UTC)

OK, let me see if I follow. WP:GNG says, If a topic has received significant coverage in reliable sources that are independent of the subject, it is presumed to be suitable for a stand-alone article or list. The logical structure is: if (meets these criteria), then (presumed to be suitable). The interpretation in the table below is that this defines "notability" by making "notable" equivalent to "presumed to be suitable". But in practice, "notable" is often equated with "suitable", not "presumed to be suitable". When people in AfD's say "fails the GNG", they mean that the article shouldn't exist. XOR'easter (talk) 17:45, 9 November 2020 (UTC)

The issue that we are having right now is how the SNGs relate to the "GNG", and the problem is that the "GNG" as defined on WP:N is not clear if its meant as a presumed indicator test or what we want articles to strive to be in their best shape. And thus how the SNGs relate to the "GNG" changes depending on which way you read it. In the first scenario: If we say that the "GNG" is still a presumption, then it should be considered no stronger an indicator of notability as the SNGs, and thus the "GNG" and the SNGs should be considered equivalent, or as said "the SNGs are alternatives to the GNG towards notability". In the second scenario, If we are saying that the "GNG" is the definition of notability, such that if an article has reached this "GNG" based on consensus, then the topic is no longer presumed notable, we consider it is notable, never in danger of being at AFD due to notability (other factors like WP:NOT, notwithstanding) And thus, in that situation, the SNGs serve to be indicators or shortcuts that a topic can presumably reach that "GNG" definition. That's why I'm saying we should determine what we want the "GNG" to be, which fixes what the relationship that the SNGs has to it and resolves the wording problem. I know that this is trying to wrap around current practice, now acutely aware that this mixup of what the "GNG" refers to between the SNGs and at AFDs happens all too often. --Masem (t) 19:31, 9 November 2020 (UTC)

@Masem: I'm not sure where you're getting the interpretation where GNG is attempting to define what we want articles to strive to be in their best shape. The concept of notability has nothing to do with the ideal state of an article, and everything to do with deciding whether a topic deserves a standalone article. The very first sentence of WP:N says, "On Wikipedia, notability is a test used by editors to decide whether a given topic warrants its own article." Additionally, I think you're misunderstanding the "presumption" aspect of GNG. If you read the text of GNG carefully, satisfying the criteria of GNG does not result in a presumption of notability, it results in a presumption that an article deserves a standalone article. Passing GNG means the topic is notable, full stop, but there are some notable topics that don't deserve a standalone article for whatever reason. This is why "being notable" and "deserving a standalone article" need to be separate concepts. So, in summary, the singular purpose of GNG is to decide of an article warrants a standalone article. If a topic passes GNG, it is considered notable, and is granted the presumption of deserving a standalone article. ‑Scottywong| [babble] || 20:29, 9 November 2020 (UTC)

Is anyone else bothered by the phrasing notability is a test used by editors? "Notability" is an attribute of a topic, which is not the same as a test for that attribute. (And I can say from years of hanging out at AfD that in practice, "being notable" and "deserving a standalone article" are not treated as separate concepts.) XOR'easter (talk) 20:46, 9 November 2020 (UTC)

I'm going to respond more in depth in a new subsection below because I think this is important to clarify. I mean, yes, what is reads is one thing, but how it is practiced is another and that's creating the issue. --Masem (t) 21:16, 9 November 2020 (UTC)

	Is it notable?	Does it warrant a standalone article?
Passes GNG	Yes.	Presumably yes, as long as other WP policies aren't violated. Editorial decisions can also be made to merge notable topics together.
Passes SNG	Presumably yes. But not guaranteed unless it can be demonstrated that the topic can satisfy the requirements of GNG.	Presumably yes, as long as other WP policies aren't violated. But not guaranteed unless it can be demonstrated that the topic can satisfy the requirements of GNG. Editorial decisions can be made to merge notable topics together.

• It has seemed clear to me for years that the closer of the 2017 NSPORT RfC overreached in using that discussion to propose a ruling that affected the interpretation of all of the other SNGs, without editors interested in any of the other SNGs even being notified of the discussion. The whole idea of a "local consensus" is that it is determined among a subset of the community that is interested in a particular issue, and so the consensus reached should at most be binding within the scope of the issue specified in the discussion, and only when it does not conflict with site-wide community consensus. (I feel exactly the same way about attempts to expand SIRS and AUD outside of NORG, for pretty much the same reason, though in that case I am unaware of overreach in any closure on the NORG pages themselves - in that instance the overreach comes from editors trying to apply NORG principles elsewhere on WP.) Newimpartial (talk) 16:04, 9 November 2020 (UTC)
  • I agree with Newimpartial here. As I mentioned above, the 2017 RfC had no substantial discussion about criteria other than WP:NSPORTS, so treating it as a binding statement about all SNGs is simply unjustifiable, whatever the closer's summary happened to say. XOR'easter (talk) 17:15, 9 November 2020 (UTC)

Please, let's not edit war over this change. Leave it reverted if you truly believe that there isn't consensus for it, and let's continue discussing it here. ‑Scottywong| [spout] || 16:29, 9 November 2020 (UTC)

I support Scottywong's changes and feel that they merely align that section with what GNG and the SNG's already say. But since there is a dispute over that we sure don't want an edit war, the previous version should be retained until we come to a decision. Sincerely, North8000 (talk) 16:35, 9 November 2020 (UTC)

• I support Scotty's changes. I still support my suggested version. I also would support removing the SNG section altogether - I do not think it ever had consensus for inclusion and so the stable version would be none at all. My order of preference is Scott's change, the pared down version, and no SNG section. Best, Barkeep49 (talk) 17:39, 9 November 2020 (UTC)
  • I kind of like the pared-down version, though I'd prefer to keep the point about some SNGs also add additional restrictions on what types of coverage can be considered for notability purposes. For example, the SNG for companies and organizations specifies a very strict set of criteria for sources being considered. Perhaps some of the heat in this discussion stems from the impression that SNGs are always weaker than the general guideline, when in reality they can also be about toughening it up. XOR'easter (talk) 18:12, 9 November 2020 (UTC)
    • If wishes were horses, I would like to see language that states that all SNGs should be treated as "interpretive instructions" for the GNG in a specific context. For example, NCORP offers a high bar for sourcing that is specific to organizations, NBOOK specifies that the GNG is met by two "non-trivial" published secondary works, including newspaper reviews and bestseller lists, NAUTHOR clarifies that NOTINHERITED is a one-way street when it comes to creators and their works, and so on. None of them should be read as contradicting the GNG, but all should be understood as giving domain-specific interpretive guidance, i.e., they can't be overruled through an appeal to a woolier and more abstract interpretation of GNG principles. So, to give a concrete example, NBOOK should not be overruled because someone wants to read SIGCOV as saying that bestseller lists don't count, that a 250-word NYT review isn't long enough, or that an AUD-type restriction should be used to exclude otherwise RS reviews. Newimpartial (talk) 18:41, 9 November 2020 (UTC)
      • The language of GNG is intentionally vague, leaving it up to us what constitutes "significant" coverage. I don't think anyone disagrees that SNGs can help to clarify the bounds of significance as it relates to specific categories of topics. The crux of the issue is whether SNGs can or cannot be used as a substitute for GNG. Some editors believe this should be the case: if a topic satisfies an SNG, then it's notable, even if it cannot be shown to satisfy the GNG. The change in language to WP:SNG that you reverted earlier was intended to clarify just that. ‑Scottywong| [confess] || 18:58, 9 November 2020 (UTC)

    I am fine adding the NCORP language and fine with the changes Dream Focus made before Newimpartial reverted them. Newimpartial given your statement about your preferred version why did you revert? A pared down version means people would have to go to the individual SNGs to know what they mean which seems to be your goal. Best, Barkeep49 (talk) 18:58, 9 November 2020 (UTC)

    • As I said in my edit summary, it was a procedural revert with No prejudice against a future change when there is an actual consensus. I haven't seen any such consensus emerge, so really I was just doing BRD.
    • Substantively, I also think the guidance removed here was more helpful than harmful, and that the changes here say some unfortunate things as well, though I am not entirely opposed to that general direction. I disagree with the "either/or" in the current version of WP:N, as I have noted a couple of times now, but there should be consensus on something else before there is a change (and the 2017 NSPORT close doesn't help at all IMO). Newimpartial (talk) 19:11, 9 November 2020 (UTC)
• I agree with the Newimpartial's revert. Such a material change in Wikipedia's core guideline concerning the interaction of SNGs and the GNG should not have been implemented based on votes cast over a span of two days by a small number of editors (13 agree, 8 disagree/oppose, and 1 agree/disagree). Such a change should be subject to a formal RfC with proper notice to the WikiProjects impacted by the SNGs and after ample time for input by interested parties. Cbl62 (talk) 19:12, 9 November 2020 (UTC)

  BOLD changes are a way that things are changed. And there was no RfC to insert that section in the first place and the accompanying discussion, like this one, had a mix of consensus. This is why removing the section altogether is one of the options I support. Best, Barkeep49 (talk) 19:16, 9 November 2020 (UTC)

The way that it actually works right now is that overall GNG's primacy is acknowledged, but that meeting just the SNG is sufficient. The currently discussed change would just tidy that up, not change it. IMO to go beyond that to really tune up the notability kluge overall, we'd need to start with the fundamental discussion that I was trying to facilitate. North8000 (talk) 19:47, 9 November 2020 (UTC)

The part of this that I do agree with is that presumptive notability, in the sense it is generally explained for SNGs in general, is a kludge. NCORP for example isn't based on "presumptive Notability" in any sense I can reasonably imagine, and I don't see any way in which reading NBOOK 1-3 as "presumptive" would be of any value whatsoever (I can see the point in regarding NBOOK 4 and 5 as presumptive, but that is a discussion for TALK:NBOOK - and NBOOK itself doesn't describe any of them as presumptive). On the other hand, the NSPORT SNGs really are all about presumptive Notability.

So yes, it is a kludge, but I think there is pretty profound disagreement about what would remain if the kludge were removed. Prima facie, we have some SNGs that impose apparently higher than the GNG standard, some that are essentially interpretive, and some that are presumptive and NODEADLINE-oriented (which reminds me to point out that the idea of demonstrating that a topic does not meet the GNG is essentially impossible, which was an issue with one of the recently rewording proposals). And from a pure standpoint of Encyclopaedic treatment, I think there is a good argument that NGEO should actually stand independently from the GNG, since otherwise some skeptics will constantly be wikilawywring what counts as SIGCOV for a place that was demonstrably populated in the past.

I do think the "read together" approach is a better kludge than "presumptive Notability", and I am also skeptical that we will ever see an conceptually distinct standard of what is notable that the community will agree on, besides documentability, but I'd rather see that litigated in terms of an article's claim to significance and WP:NOT rather than making the WP:N do a job for which it was never really designed. Newimpartial (talk) 20:14, 9 November 2020 (UTC)

Would it help to introduce new terms and divide up the SNGs? I'm wary of making up more jargon that newcomers will have to understand, but if the current jargon is unclear, then separating the "SNG" box might be worth the trouble. After all, the existence of the "SNG" label is a fact of our own invention, and so any problems with it are of our own making. XOR'easter (talk) 20:26, 9 November 2020 (UTC)

I suppose an attempt to split up the SNGs into presumptive and non-presumptive baskets, at least, might reward the effort. As I've said, NCORP and NBOOK seem clearly non-presumptive, at least to me. Newimpartial (talk) 20:30, 9 November 2020 (UTC)

Any notion of allowing SNGs to override GNG will result in some number of articles on topics that do not have significant coverage in independent, reliable sources. Some people may want that and be ok with that, but I don't think that's in the best interests of the project, and I think most editors would agree with that. SNGs are highly accurate in predicting the likelihood of notability, but they are not perfect. They are a predictor or an estimator, not an actual test. You could develop 1000 rules for exactly what an author has to do to be considered notable, and there will still be some set of authors in the world that satisfy all of those rules, but still have zero significant coverage in reliable, independent sources. WP has never allowed articles on topics that lack significant coverage in independent, reliable sources, and I don't think we should change that now. ‑Scottywong| [confer] || 20:44, 9 November 2020 (UTC)

This is a slippery-slope argument, as far as I can tell. The one SNG that most clearly overrides the GNG is NORG, which does so to make the criteria stronger. There are plenty of organizations that meet the GNG but not NORG. So the first actual example of allowing SNGs to override GNG does not actually result in some number of articles on topics that do not have significant coverage in independent, reliable sources. Neither does NBOOK do so, when it preempts pointless discussions on whether "routine" book reviews should count for book Notability - they simply do, and SIGCOV wikilawyering is not allowed on this point. There is no inevitable march towards poorly sourced articles based on SNG.

Also, the problem I have with your proposed bright line to exclude topics with zero significant coverage in reliable, independent sources is that, except for "zero" and "sources", all of these terms are susceptible to being gamed. I have seen SIRS arguments about independence - applied outside of the NCORP world - that violate the clear intent of the GNG and that try to exclude perfectly valid ENC content. I have seen AUD arguments similarly used to attempt to exclude sources that are perfectly reliable (for non-NCORP purposes) because the topics in question fit some editor's conception of NOT but it is easier to cloak the argument in Notability. And I have seen, perhaps most often, arguments that defy the clear intent of SIGCOV to impose a much higher standard of significant coverage (often even higher than a plausible reading of CORPDEPTH would require) for similar reasons. What I have not seen in my time at AfD is any article ever retained without reliable sources or without independent sourcing. While I have seen the presumptions invoked, and have seen various SNGs used to good effect, I have never seen any unsourced or COI articles preserved at AfD, though I have seen a number of articles deleted for IDONTLIKEIT arguments disguised as Notability. Newimpartial (talk) 21:14, 9 November 2020 (UTC)

Ok, well I closed an AfD not too long ago (Wikipedia:Articles for deletion/Willie Thornton (Canadian football)) where the overwhelming majority was to Keep because the subject "meets WP:GRIDIRON", with editors openly admitting that the subject most likely has no significant coverage, but it doesn't matter because he meets GRIDIRON and because GNG is a guideline not a policy, and there are other paths to notability than GNG. So, my "slippery slope" argument is not just a theory, it's happening in practice. Additionally, people in this very discussion are directly arguing for the existence of articles on topics that don't have any significant coverage anywhere: [4]. This isn't a theoretical edge case, this stuff is happening today. ‑Scottywong| [comment] || 21:38, 9 November 2020 (UTC)

Biological species are a great example of I think encyclopaedicity and "claim to significance" ought to preempt gaming of SIGCOV. RS documentation that a species exists (or existed) and that its classification has been recognized by relevant authorities ought to take precedence over an editor's desire for SIGCOV to mean the same thing in all circumstances. That you interpret such articles as having zero significant coverage in reliable, independent sources is exactly why I don't accept that as a bright line. The argument that such articles are undesirable, because their sources are too specialized or the mentions are too short, or because the sources aren't as independent as you would like, does not mean that anything is sneaking in without RS or with COI, which is what I see as the more relevant criterion. So from my perspective, your argument is a slippery slope. Newimpartial (talk) 22:03, 9 November 2020 (UTC)

There are definitely a small class of topics that we have decided that we are going to be inclusionary on, and ignore notability, because of WP:5P1 ("Wikipedia combines many features of general and specialized encyclopedias, almanacs, and gazetteers.") That includes cataloging verified biological species, and government-document geographical places. Mind you, with all of our sister projects now, we could talk about moving some of these off (eg Wikispecies should alleviate the need to have all verified species, but leave us with those that are notable, redirecting en.wiki searches for non-notable ones to there). But the number of these areas is very small, I can't even think if there are others beyond those two. And we'd obviously need global consensus to add any additional outright inclusionary topic areas as this is well beyond the scope of notability. --Masem (t) 22:29, 9 November 2020 (UTC)

But what kind of standalone article can we write about a biological species, when the only source we have is one that confirms its existence? Probably something like the examples offered in the diff I posted above: Hebeloma aestivale. It's a mushroom from Denmark. Cool. Sure, it's just a stub, and stubs are fine, but this is a stub that will never be anything other than a stub, because there are no sources that discuss this particular mushroom with any significance. If there are no other sources that discuss it, then all we'll ever be able to say about it is that it's a mushroom from Denmark. Don't get me wrong, I'm not saying that we should wipe all traces of Hebeloma aestivale from Wikipedia and pretend it doesn't exist. I'm just saying that it shouldn't have a standalone article. Standalone articles on topics like this serve no purpose, and help no one. It would be far more useful to the reader to include the 2-sentence description of this mushroom in the article on Hymenogastraceae or List of Hebeloma species, rather than asking readers to bounce around from perma-stub to perma-stub to learn what they want to learn. This is why GNG is such a wonderful, elegant idea. When a topic rises to the level of significant coverage in reliable sources, we'll make an article on it. If there is verifiable information about a topic, but that topic itself isn't covered significantly in reliable sources, then let's find a good place to put that verifiable information, but let's not create a standalone article on it. There are some editors that seem to believe that increasing the quantity of articles on Wikipedia is the ultimate goal, and that every verifiable fact deserves its own article. I agree with them that WP should strive to document every verifiable thing in the universe; I only disagree that that we should also strive to spread that information out among a near-infinite number of articles. We can still be an encyclopedia, almanac, and gazetteer without putting each individual sentence in its own article. ‑Scottywong| [gab] || 22:38, 9 November 2020 (UTC)

This is precisely why I will never subscribe to GNG fundamentalism. The people who say "the GNG should apply in the same way to all articles" are exactly the same ones who are willing to ratchet up from SIGCOV to the equivalent of SIRS and CORPDEPTH criteria for all articles - even in subject matter areas where this has never been discussed - rather than allowing the interpretation of the GNG to vary based on the consensus embodied, for example, in geographic, academic or artistic SNGs.

As I have said before, the assertion that this is a stub that will never be anything other than a stub is fundamentally unprovable, and asserting it involves the hubris of pretending to know what is contained in all print libraries, ever, along with the CRYSTAL of presuming to know what future publications will bring. Neither assumption is compatible with basic WP policies. And in most cases of provably real topics where SIGCOV is an issue, the issue is generally not that no verifiable information exists beyond the subject's existence. The problem is usually either that such information is difficult to find - in print, or in other languages, or behind paywalls - or that it is available in sources that may be verifiable but that are PRIMARY or questionably independent, such as published research findings or the self-published work of experts in the field. And it is for these much more typical cases, especially, that I push for verifiably real things (but not for-profit companies) to be included in the encyclopaedia even when no one source may meet the requirements of SIRS. And while I am generally sympathetic to "but it doesn't need a standalone article"-style arguments, I believe that (1) the place for that is in Merge discussions, not AfD and (2) in cases like GEO and species (more than most), categories matter, and WP's category system, in so far as it works at all, only works when it can classify articles in a domain that have been created at a consistent unit of analysis. And to me, that is a good example of an encyclopaedic criterion. Newimpartial (talk) 23:54, 9 November 2020 (UTC)

As to your close which you used as an example earlier, that article has two definitely RS (one with multiple stories), so from my perspecive the "rebuttable presumption" of GRIDIRON was not in fact rebutted in the actual instance. An argument to the contrary would have to be based in the kind of ratcheting interpretation of SIGCOV I referred to just above - this does speak well to your close, of course :). Newimpartial (talk) 00:08, 10 November 2020 (UTC)

GNG is intentionally vague, it does not define exactly what "significant" means for every topic. Therefore, I agree with you that the definition of what is "significant" coverage should be fluid, and should generally be left up to local consensus amongst editors in who are familiar with that topic. I honestly find no difference between SIRS/CORPDEPTH and SIGCOV; they are an expansion of SIGCOV, an attempt to define SIGCOV more exactly for the topic of organizations and companies. Of course, it would not be appropriate to apply CORPDEPTH to an article on a bacteria or a galaxy, and it should be easy to shoot down that argument at an AfD.

While it might be true that the assertion that "this is a stub that will never be anything other than a stub" is typically unprovable, the converse is also true. In the absence of sources, the statement "this is a stub that will become more than a stub someday" is equally unprovable. Both of these assertions are CRYSTAL. Attempting to "prove" these things beyond the shadow of a doubt is not helpful and not the point; we can only go on the information we have today. If, after an exhaustive search, we can't find sources for a topic, then we should assume that no other sources exist and act accordingly by either deleting the article or merging it elsewhere. If, at a later date, sources come to light that were missed in the previous search, then we act accordingly by either re-creating the article or splitting the topic off from the main article.

I agree that all verifiably real things should be documented in the encyclopedia, we just need to be rational about the quantity of articles we use to document those things, to optimize things for the reader. You might have a valid point about categories in some cases. And I agree that Merge discussions should ideally take place on the article's talk page instead of AfD. But, in practice, merge discussions often spontaneously arise at AfD, and there's no benefit to artificially interrupting those productive conversations to move them to a different venue, especially when the AfD might have attracted the attention of a larger group of editors who can have a better discussion and generate a stronger consensus on how to proceed. ‑Scottywong| [babble] || 00:47, 10 November 2020 (UTC)

Whether it's more useful to have several small articles or one larger one will depend on the situation. I've found myself arguing for merges and against them; I've encountered examples where a merge target was uniquely obvious and others where the opposite was true. For example, suppose we have a biographical stub about a scientist, and it doesn't look too likely that we'll find material to expand it with. Should we merge it to (a) the article on the university where they did the work they're recognized for, (b) the article on the university where they work now, (c) the article on the technical topic they're most known for studying, (d) the article on the learned society that elected them a Fellow in recognition of lifetime achievements, (e) the article on the textbook they wrote that passes WP:NBOOK, (f) the article on the representation of their ethnic group in science, (g) ... Any of these choices will likely clutter up the target article and run the risk of the biographical bit getting removed later ("WP:UNDUE"). And, as Newimpartial pointed out, cavalier merges will break the category system. What is optimal for the reader must be evaluated as the circumstances arise (not that there is only one kind of reader, either). XOR'easter (talk) 01:10, 10 November 2020 (UTC)

To Scottywong: so much in there that I can agree with, but then If, after an exhaustive search, we can't find sources for a topic, then we should assume that no other sources exist and act accordingly by either deleting the article or merging it elsewhere. That position is classic WP:NOWDEADLINE, but there is equally a WP:NODEADLINE position in the community, and no clear adjudication between the two. And once again, as I have stated before, I am by no means in favor of articles "without sources", but I also have no objection to permastubs either depending on the context - so long as they have reliable sources (even if they don't meet a reading of the GNG that can't tell SIGCOV from CORPDEPTH) and contribute to encyclopaedic treatment, whether with respect to the category system or to line up neatly with other articles as in NBOOK point 5.

I have signed off on more than my share of merges over time, including List merges, but I am also quite familiar with the one-two shuffle where an article is Merged at AfD because of "borderline notability" but then the merged content is then dropped from the list article because the subject "is not notable". Which is why I am inclined to draw my own bright line against GNG fundamentalism: I think WP:V, NOT, and even WP:SIGNIF are more helpful in resolving these content questions than the GNG. Newimpartial (talk) 01:53, 10 November 2020 (UTC)

Clarify the burden-shifting element to a rebuttable presumption

I agree that SNGs should not "override" GNG. The key is to give due weight to the SNG so that they are meaningful, but not absolute. As noted above, our current guideline states that the presumption holds "unless someone comes forward to contest it and prove otherwise." (emphasis added) The last part is often overlooked in AfD discussions. As with all rebuttable presumptions, the effect is to shift the burden of proof. See Heiner v. Donnan, 285 U.S. 312 (1932) ("A rebuttable presumption clearly is a rule of evidence which has the effect of shifting the burden of proof."). What we ought to do is codify and clarify this existing consensus along these lines:

"Passing an SNG establishes a rebuttable presumption that a topic is notable. It does not, however, create an irrebuttable or conclusive presumption. The effect of a rebuttable presumption is to shift the burden of proof. Accordingly, a party challenging the notability of a topic that passes an SNG has the burden of proof to demonstrate by appropriate and diligent searches that the topic has not received significant coverage of the type required by the general notability standard. The scope of such searches which will vary depending on the circumstances, including geography (e.g., a Belgian topic should include searches of Belgian sources), language (e.g., a Chinese topic should include searches of Chinese language sources), and time period (e.g, Google searches will not suffice for pre-Internet topics)." Cbl62 (talk) 21:41, 9 November 2020 (UTC)

I think, in both guideline (e.g. WP:NPROF) and practice (at AfD), some SNGs don't really fall under of the rebuttable presumption concept. Most. But not all. Which is part of the problem in trying to describe SNGs in the first place. Best, Barkeep49 (talk) 22:39, 9 November 2020 (UTC)

• That doesn't obviate the need for clarification for those SNGs that do create a rebuttable presumption. SNGs are not created equal, and the continued effort to try to formulate a one-size-fits-all solution is not going anywhere. As NSPORTS is among the most controversial, I continue to think a clarification along these lines is the most sensible. Here is revised language to address Barkeep's point:

"Many SNGs (including WP:NSPORTS, ___, ___, ___) establish a rebuttable presumption that a topic is notable. In those cases, the SNG does not create an irrebuttable or conclusive presumption. The effect of a rebuttable presumption is to shift the burden of proof. Accordingly, a party challenging the notability of a topic that passes such an SNG has the burden of proof to demonstrate by appropriate and diligent searches that the topic has not received significant coverage of the type required by the general notability standard. The scope of such searches which will vary depending on the circumstances, including geography (e.g., a Belgian topic should include searches of Belgian sources), language (e.g., a Chinese topic should include searches of Chinese language sources), and time period (e.g, Google searches will not suffice for pre-Internet topics)." Cbl62 (talk) 00:03, 11 November 2020 (UTC)

Is the GNG a test or a standard?

This is extended from discussion above but splitting out to avoid conflation. But again, I think focusing too much right now on the SNG's relationship to the GNG is the wrong approach as defining what we say the "GNG" is.

A hypothetical question: I have created an article on a topic that otherwise doesn't have any SNG apply to it. I have found three references about the topic from decent sources (let's say like Chicago Tribune, USA Today, and CNN), each which discusses the topic over 3-4 paragraphs out of a 20 paragraph article in each case. Presume the coverage is secondary. Between the three sources I can craft a two paragraph stub article.

Now someone comes along and blindly nominates that for deletion for failing the GNG, failing to do any other BEFORE-like searches. Ignoring that issue, I can tell you that the likely outcome of that AFD will be "Keep per GNG" as we have it right now because at AFD, the "GNG" is taken as a minimum amount of sourcing - it is treated like a test.

A year goes by, no major changes to the article are made, and someone else comes along and after doing a fully comprehensive BEFORE search and finds nothing (and lets assume there really is nothing else beyond these three sources), and again nominates it for failing the GNG. And that becomes the hypothetical question: What is the "proper" result here under the notability guidelines?

I'll leave this an open question but it does relate to how we define what the GNG is and subsequent its relation to the SNG. --Masem (t) 21:46, 9 November 2020 (UTC)

And I will give you my realistic hypothetical answer, which is that this instance shouldn't be at AfD at all. The correct venue would be a Merge discussion on the article Talk page, to determine whether the most Encyclopaedic treatment of this sourced information would be as a standalone article or as part of a broader, related topic. WP:N would not offer any help in framing this question, IMO. Newimpartial (talk) 21:54, 9 November 2020 (UTC)

For all purposes, let's assume its impossible to find a place to merge this topic, and that AFD is the right place for it. I agree that if is clear there was a merge target, that a merge discussion should happen instead, but that's not the test I am trying to figure out here. --Masem (t) 22:23, 9 November 2020 (UTC)

It would probably be kept from year to year, but if the coverage isn't WP:LASTING, the article subject isn't that important, and no one works on it, I think in five or ten years it stands a good chance of deletion. The three sources don't need to be temporal in this case, either. SportingFlyer T·C 22:30, 9 November 2020 (UTC)

WP:LASTING wouldn't apply in this case because it's part of an SNG. The example said that there is no SNG that applies. ‑Scottywong| [spout] || 22:48, 9 November 2020 (UTC)

Fair enough, but WP:RECENT isn't part of the SNG and is a relatively similar principle. Once notable, always notable, but there are some topics which may be considered notable once and not notable later due to limited sourcing ageing poorly. SportingFlyer T·C 22:57, 9 November 2020 (UTC)

Different topics require different lengths in an encyclopedia. New York State Route 373 is an FA with just over 1000 words while Barack Obama, also an FA, has 703 words in the LEAD alone and clocks in at over 13,000 words in total. So this hypothetical article should be kept and the nominator of the second better be ready for a whole bunch of people getting upset that there's a second nomination after it was kept in an identical state a year previous. I don't think this is actually useful, however, in deciding what we should say about SNGs on WP:N which is why I'm here. Best, Barkeep49 (talk) 22:33, 9 November 2020 (UTC

Let's say the hypothetical article I'm speaking of at 2 paragraphs is at most 150 words/1500 characters, with those 3 sources (1500 char is where DYK considers an article "long enough"). In other words, its clearly stub territory when compared to NY 373. And let's further state that the 2 paragraphs is the extent the article could be expanded from the sources without going into excessive quoting or paraphrasing or other NOT issues. --Masem (t) 22:42, 9 November 2020 (UTC)

So this hypothetical article should be kept and the nominator of the second better be ready for a whole bunch of people getting upset that there's a second nomination after it was kept in an identical state a year previous. I don't think this is actually useful, however, in deciding what we should say about SNGs on WP:N which is why I'm here. Best, Barkeep49 (talk) 22:50, 9 November 2020 (UTC)

I agree. So long as there are sources which are acceptable and are deemed to establish notability, the article should not be deleted. HighKing⁺⁺ 15:28, 10 November 2020 (UTC)

The hypothetical article that you're talking about should not be deleted in this case. If it was determined that it satisfied GNG last year, then per WP:NTEMP, that won't change. If the most you can get out of the available sources is a few paragraphs, then look for an appropriate merge target. If there is truly no appropriate merge target, then leave the article alone. Sure, it's not ideal to have permanently short articles. But, if there is borderline signficant coverage of a verifiable topic, and there is truly nowhere else to merge information about this topic, then we just deal with having a short-ish article. By the way, this is a very contrived example. In practice, there is almost always an appropriate merge target. Very few topics are so unrelated to everything else in the universe that there is no other article that is closely related. ‑Scottywong| [gab] || 22:47, 9 November 2020 (UTC)

On this, my two cents: the way you've described how you would handle it, that still makes this a test, that the fact that we can't expand beyond a few sources and should find a non-standalone place for the information is exactly what we'd do if a topic failed an SNG as well. (And yes, I would agree there's almost always a merge target). This is what I'm trying to get across is that the fact that some here agree that such an unexpandable stub should be merged means that we have a source-based indicator of notability separate from the definition. --Masem (t) 01:45, 10 November 2020 (UTC)

Are we conflating notability with quality here? Why should a short article on a topic that is notable even be considered for a merge in the first place? HighKing⁺⁺ 15:28, 10 November 2020 (UTC)

Because we do not want stub/short articles. If it is fundamentally impossible to expand the article any further than a stub due to lack of sourcing (and no reasonable expectation of sourcing to come in the future), then we're going to want to merge if at all possible or delete if there's no other solution. The thing is, there is no other policy or guideline that speaks to the number of sources or amount of significant coverage that is needed to build out an article beyond a stub except for WP:N, which is why this is an issue here. --Masem (t) 15:47, 10 November 2020 (UTC)

IMO it is both a standard and a test/metric. This isn't the usual cop-out "both" answer, I think that it is fundamental to how the kluge operates. I think that the standard of how the kluge operates is the standard is mostly GNG, with smaller allowances for other factors such as how easily it passes the wp:not criteria, SNG criteria, degree of enclyclopedicness, importance. famousness. And it's also a pretty good measuring stick and credible information source for all of those things.North8000 (talk) 02:18, 10 November 2020 (UTC)

There's clearly some implicit "indicator of notability" being called out here that we'll initially keep an article with minimal sourcing that is in the ballpack of the current GNG (but otherwise doesn't meet any SNG), but over time if it simply can't be improved with more sources, we're liking to say that presumptions is wrong and we'll look for a merge target/delete at worst - that's the concept I'm getting from these responses. Only judging by that, what we have as the GNG should stay the definition, that SNGs are indicators/shortcuts to that definition, but we really should add a section on WP:N on soft end of the GNG: that while having a minimal number of sources to just barely pass the GNG can be acceptable in the early days of an article's history, it is expected that editors will expand to better demonstrate more GNG-meeting sources/coverage, and that that presumption for the standalone can still be challenged. We say that about the presumption of the standalone article at the top of the GNG already, but we don't say this presumption can be challenged if the number of sources/quantity of significant coverage is judged to be insufficient by consensus at AFD. --Masem (t) 15:47, 10 November 2020 (UTC)

Also, this seems to confirm that the SNGs are not "alternatives" to the GNG, particularly if they also only produce a short stub article. The SNG is sufficient to keep the article to be worked out and look for more sources, but if proven out no further sources/significant coverage of the topic can be made after a thorough BEFORE source search, we'd be looking to merge/delete the topic. It may be possible to get to where the SNG article has enough sources to just meet the GNG but at which point that still come under this consideration of minimal sourcing/significant coverage idea. --Masem (t) 16:03, 10 November 2020 (UTC)

That reasoning would make sense only if the sources that are allowed to count for GNG-notability are the same as the ones that are allowed to be used to source and expand articles. In practice, our standards for what kinds of sources can be used to expand articles are much broader than our standards for what kinds of sources count towards GNG-notability. It's not just that we allow expansion based on interviews, local newspapers, and the like; it's also that sentences in an article such as stating that the subject won a notable award (MBE, say) can be supported by sources that mention the subject only in passing (among a list of other people getting MBE that year). —David Eppstein (talk) 17:34, 10 November 2020 (UTC)

Good point and this is a common source of confusion. A source, such as an interview, may be well used as a WP:RS for verification purposes and also be inadequate for purposes of establishing notability under GNG. I definitely see new editors struggle with this when creating articles. Best, Barkeep49 (talk) 17:39, 10 November 2020 (UTC)

I would like to emphasise the difference between SNGs that are looser on GNG requirements for establishing notability (in circumstances where the SNG lowers the bar and an article is deemed notable by the SNG requirements but would fail GNG) and SNGs that are (perceived as) tighter on GNG requirements. I'm thinking of NCORP here for example which provides an in-depth explanation on the meaning of an "Independent" source to include "Independent Content". Editors often interpret an "Independent" source to merely reflect the independence of the publisher from the topic and point to GNG as evidence in these situations which doesn't provide the additional clarification/explanation. In this circumstance, it is perceived that the NCORP SNG is going beyond providing additional explanation to the GNG requirement and is, in fact, adding new requirements. No amount of explanation at an AfD has proved effective and it is a serious waste of time repeating the requirements ad nauseum. In my view, it would be a mistake to say that the SNG is "subservient" to GNG whereby if an article meets GNG, then the SNG doesn't matter. Barkeep49's mention above of a source, such as an interview, being used to establish notability, for example, crops up time and time again. I know this is a different point that the one Masem is outlining above, but any proposed solution or changes to wording at WP:N or GNG should also keep the above in mind as it is an opportunity to provide much needed clarity. HighKing⁺⁺ 17:59, 10 November 2020 (UTC)

Yes, SNGs should be able to be met with a minimum of one source that meets WP:V (which at minimum required a third-party source - no press releases for example, but also implies reliability and the like), unless otherwise stated by the SNG. But that's to pass the SNG for presumed notability; over time we would expect better sourcing to be added. --Masem (t) 18:12, 10 November 2020 (UTC)

Depending on how far we're going here with redefining GNGs and SNGs, it may be worth considering the minority of SNGs that impose additional restrictions (mostly just NCORP, although NPOL is sometimes interpreted this way to dismiss election coverage as routine, and the NFF section of NFILM also imposes restrictions) as a separate category of guideline. It may be far more trouble than it's worth, but if we're really concerned with inconsistency about what how SNGs interact with GNG, this would be a way to fix it. _signed, Rosguill ^talk 18:56, 10 November 2020 (UTC)

I agree this is true in practice, but I'd be hardpressed to find any policy or guideline outside of STUB that gets into what we want to see for expansion.

Maybe what we are looking at beyond just meeting a bare minimum of GNG-type sources or the SNG criteria are additional sources that each demonstrate the ability to expand the article in ways that meet WP:NOT, WP:V, WP:NOR, and WP:NPOV. A counterexample is common, the case of fictional characters that get barely a few mentions of critical reception after which editors tend to flood the rest of the article with primary sourcing to document the character. Arguably that just meets the GNG, but we have over the last several years deleted/merged numerous articles in this state because of WP:NOT#PLOT issue. But to go on this point and take David's point on interviews (a point of contention being if they are primary or secondary sources), its also often that an article on a fictional character that have some GNG type sourcing can be expanded from interviews with their creators, talking about their concepts for creating the character, design changes, etc., which is routine part of articles that are kept - regardless if one considers that interview primary or secondary, they would fulfill V/NOR/NPOV and doesn't fail NOT.

Or to take another example, the common SNG issue are those NSPORTS of "played one pro game". The common logic that this is used and that has been demonstrated is that , at least for most American sports, is that to be in a pro game you have had to play in some minor league or a collegiate league and not be a slouch - you don't "accidentally" a pro career. There should be readily-available documentation from that minor league/collegiate period that may be more local but should still fulfill V/NOR/NPOV elements of sourcing and also meet the independent and secondary facets for the GNG definition.

But again, I'm not aware of any distinct policy or guideline to this end. STUB is one part, but even large articles that *just* pass the GNG can be deleted because they further can't meet additional GNG sourcing (at least in AFD statements). Ideally, we want more sourcing that is GNG-criteria-worthy quality, but having even additional sourcing that individually satisfies the other guiding policies seems to be more than sufficient. I think it should also caution that articles that continue to only demonstrate minimal coverage or sourcing even if having passed a prior AFD baed on a GNG or SNG test can still likely be challenged later, and the only way to avoid that completely is to expand with this type of additional sourcing: simply passing the GNG or SNG with minimal sourcing should not be taken as a clear pass of notability forever. --Masem (t) 18:12, 10 November 2020 (UTC)

I am going to say again that I don't think that AfD and WP:N in general are the right location to deal with issues related to permastubs and challenging prior AfDs. In the scenarios we are talking about there is clearly verifiable information, and Merge discussions on article talk pages (often with posted RfC) are the best way to resolve that question, rather than throwing the issue into the cauldron of AfD each time. WP:PRESERVE is the relevant principle here, and the question is how that can best be achieved. The articles that I think should be taken to AfD (under related but not identical circumstances) are those without a clear claim to significance, but those aren't best understood as WP:N issues either. Newimpartial (talk) 18:22, 10 November 2020 (UTC)

I shouldn't mean that all stubs or non-notable articles are AFD targets, only that they are no longer appropriate for standalone pages and something should be done with them, with merge/PRESERVE options being of highest priority, I agree. What I am trying to say that in terms of notability, while barely meeting the GNG, or meeting the SNG, may pass one AFD, this is not a free pass to no further challenges to notability and/or article improvement. (Though clearly there is no DEADLINE here and a whole bunch of other facts that we give leniency on the article creation process) What we expect that improvement to be, and how the lack of improvement can be challenged is something I am not seeing well spelled out, which in part may be the confusion created in all this thread. There's a certain amount of language precision the guideline lacks that we can fix, but also a bit about process that we're missing too (which may be more appropriate in an essay-like page rather than a guideline ) --Masem (t) 19:09, 10 November 2020 (UTC)

The concept that every article should eventually be more than a reasonable stub (say two or three paragraphs) is not relevant to notability. For example the Encyclopedia Brittanica has thousands of stub articles and extending articles needlessly is just padding and boredom for the reader who should be considered in this argument. Short but informative articles - say with five or more facts are perfectly acceptable as a permanent stub if the subject is notable as in covered significantly by multiple reliable sources or accepted per consensus such as professors or villages. However, one or two line sub-stubs are ready for merging in my view. The general consensus of wikipedia content creators who mostly don't take part in these sort of discussions is that stubs are perfectly acceptable otherwise they wouldn't be creating them by the thousands every week, imv Atlantic306 (talk) 01:49, 11 November 2020 (UTC)

Paths towards consensus

We're having a pretty interesting and in depth discussion about notability above that hopefully has been as interesting for others as it has been for me. However, ultimately I would like to see us move towards finding consensus about how our SNG section should read. And I'm not sure we're moving towards that; it appears we've reached a stage where a handful of editors say a lot to each other without new voices entering the discussion and/or we are sprawling beyond the core question (perhaps appropriately, perhaps not) of what the SNG section should say. Here are three ideas I have about how we could move towards reaching consensus:

1. We continue on (perhaps others feel we're moving more towards consensus than I do)
2. We do a more structured straw poll (a few set options with a section for people to state their thinking and a separate discussion section)
3. Or we go whole hog and craft an RfC

My preference is for 2 with a weak second choice for 3 (as 3 always remains an option if 2 doesn't work). Best, Barkeep49 (talk) 18:44, 10 November 2020 (UTC)

For something as important as this, I think we need to enshrine this in an RfC. I've already started drafting one at User:Scottywong/SNG RfC draft. It'd probably be a good idea to have a discussion about how it's structured and worded before making it live (but not a discussion about whether you support or oppose it). ‑Scottywong| [comment] || 19:09, 10 November 2020 (UTC)

I am adamantly opposed to any dogmatic credo like "GNG is ultimately considered the only true test for notability", which I find early on in your draft, and will strongly oppose any RfC framed on that basis. The discussion above makes clear that this is far from an accurate description of the status quo as written, even farther from the current practice, and would significantly change the basis for inclusion of many classes of Wikipedia articles. I believe that any draft framed from that starting point is fundamentally broken to the point where it is such a bad framing of the issue that it should not even be raised as an RFC. —David Eppstein (talk) 19:27, 10 November 2020 (UTC)

GNG is the ultimate indicator of notability, but notability is not the ultimate indicator of whether we keep an article. There are other reasons we have articles that are not based on notability (like geographic locations and species, lists, other content that is in mainspace). Just that for most topics that don't fall into the WP:5P1 concepts, notability is a good judgement to consider keeping a standalone article as a starting point, and to judge notability, the GNG is the right test with the SNGs as indicators of that. --Masem (t) 19:56, 10 November 2020 (UTC)

Word games where in one sentence we use "notability" to mean "can we have a Wikipedia article on this topic", in a second sentence we use "notability" to mean "significance", and in a third sentence we use "notability" to mean "does this article pass GNG regardless of whether other criteria allow us to have a Wikipedia article on it" are not a helpful way of viewing this issue, because they end up leading to the same confusion by editors in deletion discussions. —David Eppstein (talk) 20:44, 10 November 2020 (UTC)

Trying to be helpful here: the draft refers to the situation where it can be shown that the same topic cannot satisfy the GNG, which I believe to be essentially a null set. Some language similar to Masem's concept that a topic does not show a strong GNG pass over a period of time, might be more relevant to an actual set of cases. Just an FYI. Newimpartial (talk) 19:31, 10 November 2020 (UTC)

It's definitely not a null set, that's why we're discussing this. The goal is to make it a null set, but this is only theoretical, and impossible to put into practice. SportingFlyer T·C 23:42, 10 November 2020 (UTC)

Can we at least agree that it is essentially a null set when it comes to topics from the pre-internet era, topics where a substantial proportion of the sources are in other languages, and topics where many, relevant sources are paywalled? Unless, of course, serious efforts have been made to find those inaccessible sources (which is almost never done at AfD, in my experience). Newimpartial (talk) 00:52, 11 November 2020 (UTC)

I'd personally replace the word "shortcut" with "indicator" in the revised version. "Shortcut" cheapens the intent of the SNG even though on practical terms it is the same thing. --Masem (t) 19:43, 10 November 2020 (UTC)

I do not think this RfC can be a binary "yes/no" and so I do not think this draft is a workable RfC. If the answer to the draft RfC were to be "no" we might not be any closer to knowing what we should have than we are today. This RfC is trying to do two distinct things: establish the relationship between GNG and SNG and to say what wording we should have on WP:N.

I think we should give serious thought to just doing 1 of these. If we were to focus on what is written at WP:N, I think we could present 3 versions which would roughly be Masem's May change, Scotty's proposed change in Language Clarified, my "pared down" version, and a fourth "there should be no SNG section" option. There might be 1, maybe 2, other versions with presenting. Masem's and Scotty's versions would, by default, end up presenting consensus on the question of SNG/GNG.

Or we could just try to establish the relationship between SNGs and GNG. Even there I thik we'd need three options: 1. GNG is superior to SNG (SNG as shortcut) 2. SNG and GNG are equivalent (qualifying under either establishes notability) or 3. The relationship between SNGs and GNG differs depending on the SNG. Best, Barkeep49 (talk) 19:58, 10 November 2020 (UTC)

The problem with trying to reach a consensus is that none is apparent, which is not surprising given that consensus doesn't scale up as a group grows larger and no longer has a strong alignment in goals. And ultimately, closers following Wikipedia:Closing discussions § Consensus will see a pointer to Wikipedia:Deletion guidelines for administrators § Rough consensus, which allows closers to discount arguments that contradict policy, but not arguments that are counter to guidelines. Thus closers often will only take into account arguments made during the deletion discussion, or else they will be accused of imposing their own personal viewpoint.

For instance, regarding the deletion discussion about a gridiron football player that triggered this conversation, the sports-specific notability guidelines have always deferred to the general notability guideline since they were created. Yet Scottywong said "there's really no way that this AfD would result in a delete consensus", in essence giving precedence to the arguments made in the AfD over the text of the sports notability guideline being referenced. Many editors hew closely to the idea that guidelines should be descriptive of what occurs during consensus discussions, which also favours ignoring arguments in guidelines if they aren't constantly being followed up during deletion discussions. Unless something in this dynamic changes, deletion discussions are fated to always follow whatever opinions are expressed the most frequently, because the system is designed that way. isaacl (talk) 20:31, 10 November 2020 (UTC)

@Isaacl: FYI - the reason I had to close that AfD as Keep is because of the ambiguity in the current wording of WP:SNG. The wording can reasonably be interpreted as meaning that a topic is notable if it passes either GNG or SNG. ‑Scottywong| [chatter] || 21:05, 10 November 2020 (UTC)

That and the fact that, by the time of the close, the article did pass the GNG. :p Newimpartial (talk) 21:08, 10 November 2020 (UTC)

There are a few sources, but if you look closely at the sources, it still doesn't. I don't want to get into an argument about it, but I also don't want to leave this unchallenged. SportingFlyer T·C 23:46, 10 November 2020 (UTC)

I wasn't part of that AfD and have no intention of litigating it here, particularly since the close was correct. But the article had multiple, independent RS, so the only way to contest the GNG pass would be to get into SIGCOV argumentation, and that is one of the most contested (or, from my point of view, misinterpreted) guidelines associated with WP:N. Newimpartial (talk) 00:52, 11 November 2020 (UTC)

Yes, but the editors who agreed upon the sports notability guidelines agreed that it would not supersede the general notability guideline. They explicitly decided by consensus not to make use of the flexibility offered on the notability page, and to state that the sports notability guidelines do not alleviate the need for an article to eventually provide sources to meet the general notability guideline. Thus when editors choose the "or the sports notability guidelines" option, they are not exempted from demonstrating that the general notability guideline is met, if the presumption of meeting the standards for having an article is challenged. isaacl (talk) 22:06, 10 November 2020 (UTC)

I agree that any RfC on this kind of question can't be a binary yes/no. In fact, I suspect there are underlying issues of which the GNG/SNG arguments are only a symptom. The page Wikipedia:Notability only defines "notability" implicitly, and coming at it with fresh eyes, it is not at all clear what distinguishes a topic being worthy of mention within an article and being worthy of an article unto itself. To illustrate, I took the text of WP:GNG and put it on its own page with a few small edits (like changing "notability" to "significance"). It reads exactly like advice for whether to include a topic within an article. XOR'easter (talk) 20:35, 10 November 2020 (UTC)

WP:V determines what is worthy of mention within an article. WP:GNG determines what is worthy of a standalone article. ‑Scottywong| [spout] || 21:21, 10 November 2020 (UTC)

You've captured what the first paragraph of WP:N says, but not what the second (full) paragraph says. Newimpartial (talk) 21:35, 10 November 2020 (UTC)

Also it is false that WP:V determines what is worthy of mention within an article. For instance it is a verifiable fact that the nearest Starbucks to the Liberty Bell is at the corner of Walnut and 8th; that doesn't make it worthy of mention in the Liberty Bell article. See also WP:NPOV and WP:INDISCRIMINATE. —David Eppstein (talk) 22:15, 10 November 2020 (UTC)

It should be noted this Wikipedia:Notability/RFC:compromise from 2008 is the essence of where "SNGs are indicators of notability but do not supercede the GNG" basically comes from, and where most SNGs are written towards with the clear obvious exception of NCORP for obvious reasons. --Masem (t) 20:37, 10 November 2020 (UTC)

I have drafted what the two options I outlined above could look like in RfC form. Best, Barkeep49 (talk) 20:42, 10 November 2020 (UTC)

Just in terms of your three options about the relationship between GNG and SNGs, I can't see how the answer would not be 3., because the status of NCORP and NSPORTS - to cite the extremes - are so obviously not the same. The proposal made earlier to sort out which SNG currently claims which status - NCORP and NPOL are potentially restrictive, NAUTHOR and NBOOK define standards without being notably inclusionary, NPROF explicitly claims to be an independent standard while NSPORT offers at best a reputable presumption, etc. - looks to me to be a more productive discussion. I am not presuming that everyone agrees on which SNG currently falls - or should fall - in which basket, but I think the fact that the nature of their claims differs should be immediately obvious to a discerning reader (which is one of the reasons the close of the 2017 NSPORTS RfC was so obviously an overreach). It would at least be worth trying to map out that terrain, IMO. Masem seems to think the distinction is NCORP/everything else, and there may be others who think the kludge is supposed to work that way, but I'd rather see that defined clearly than have people vote with their heart or gut on option 1. or 2. while assuming away the differences between, say, how people want NPROF and NSPORT to work - which I suspect are usually not the same at all. Newimpartial (talk) 20:46, 10 November 2020 (UTC)

I'm not sure why we'd want to split these concepts up into two RfCs, or only do one or the other. They're essentially all part of the same thing: we want to clarify the relationship between GNG and SNG, and then correct the wording of WP:N to accurately reflect that relationship. ‑Scottywong| [confer] || 21:08, 10 November 2020 (UTC)

By the way, Scottywong, my view is still that if you want to have one single relationship between GNG and SNGs, the closest thing to that which might actually be possible would be an instruction that they be "read together" or that the SNG be read as specifying the GNG in a specific context. For that to work as intended, though, the ones that actually are a separate, rebuttable standard (like NSPORT) would have to be framed differently than the majority, which offer interpretive guidance for the most part rather than separate standards. Newimpartial (talk) 21:15, 10 November 2020 (UTC)

@Newimpartial: I also think Option 3 is what has consensus. But other editors have voiced support for options 1 & 2 during the discussion. If you have an alternative format feel free to edit to provide an option 2a or 3. Best, Barkeep49 (talk) 22:14, 10 November 2020 (UTC)

• In order to carry out a clear RfC that delivers a clear consensus on the matter, I believe it's better to offer a simple choice to RfC participants: support or oppose. If we give participants 7 different choices of options to choose from, it simply muddies the water and the overall result will be unclear. I'm not trying to push my own preferred point of view into the RfC, I'm happy to modify the proposal to ensure more support for it (and in fact, I just did make some tweaks to it based on comments here) or use a different draft as a basis for the RfC. But, I think that we'd all be better served by putting forth a simple, clear proposal for editors to read and voice their opinions on. And if there ultimately isn't a lot of support for it, then we can come back to the drawing board and find a better way to address the problem. ‑Scottywong| [soliloquize] || 21:30, 10 November 2020 (UTC)

  There is only so much editor time and repeat RfCs draw decreasing participation and commitment. So putting forth a proposal that fails is destructive to the cause of consensus. I am all for simple RfCs but I do not think a binary RfC is possible here. I do think a simple constrained choice RfC is possible which is why I suggested two formats for that. But an RfC where we can have people oppose for many different reasons is an RfC that will likely lead to consensus against anything as opposed to an actual solution to this issue. Best, Barkeep49 (talk) 22:16, 10 November 2020 (UTC)

  My biggest issue is that I don't actually think we have a consensus to keep the current version added in May. I think we need three options: the current one, a proposed one, and "neither." If neither wins, we remove the section and have more work to do, but we've at least identified a consensus, but there may be other options I'm not considering at the moment. SportingFlyer T·C 23:49, 10 November 2020 (UTC)

I'm just really not sure that your current formulation of the proposal captures the actual state of the discussion at present. Newimpartial (talk) 21:38, 10 November 2020 (UTC)

Care to take a stab at it yourself? Feel free to edit the one in my userspace if you want. ‑Scottywong| [confess] || 01:27, 11 November 2020 (UTC)

I think the reason forming consensus on the GNG/SNG question is so difficult, with many editors having vastly different ideas about basic definitions and roles, is because we have reached the intellectual limits of the concept of "notability", and the most productive thing we can do is replace it with a more robust framework that works better for our purpose (writing an encyclopedia).

The problem with "notability" is that it is a construct, a subjective evaluation of whether a topic deserves to be an entry in the encyclopedia, that we treat as an objective property that we can discern, discover, measure, or prove, and one that is fixed (once notable always notable). In reality, topics are notable for some people, places, and times, but not others; fundamentally, whether something is important or of interest to readers is not objective or fixed.

Instead of thinking of the encyclopedia as a collection of topics and trying to filter out topics that are "notable" from those that are not, we should think of the encyclopedia as one work: a book summarizing everything we know about everything. We can't practically write it all on one page, so we have to break it up into subpages. Almost every article is a sub-page of some other article; the question is what information we put on what page. Instead of asking "Is it notable?", we should ask, "Do we include this information and if so, where?".

Whether we include the information in the encyclopedia at all is answered by PAGs like WP:V, WP:NPOV, WP:BLP, etc. Where we put information is answered by PAGs like WP:MERGE, WP:CFORK, and WP:AS. We don't really need "notability"; we don't need most of WP:N. We should delete all of WP:N except WP:PAGEDECIDE, and then the SNGs would become guidelines for how to apply PAGEDECIDE in specific topic areas. How about an RFC question about that? :-) Lev!vich 08:40, 11 November 2020 (UTC)

I applaud noting that the real answer needs a fundamental fix. I was attempting to facilitate a conversation to make steps towards that. But my end goal was based on the presumption that, regarding the end result, the notability kluge 95% works right now but that to clean it up (or even make any substantial tweaks) we need to define the current end result of the Kluge, make that the guidepost/mission statement for the kluge, then more cleanly align the kluge with use that definition. You are proposing a bigger departure....I dunno about that. North8000 (talk) 13:34, 11 November 2020 (UTC)

My reframing would be, that I think much of the work some editors wish Notability would do - but that it does not actually do, since the GNG is basically a sourcing requirement - is actually the claim to significance. If we had some kind of consensus - or even just a more open discussion of that, in relation to WP:NOT - then maybe we could just let the GNG be a sourcing requirement and not try to overlay Notability with burdens it can't carry. Newimpartial (talk) 14:29, 11 November 2020 (UTC)

The claim of significant is almost describing the basic test used by WP:CSD to keep articles - is the topic even something worth documenting, and thus a even a lower bar to meet than notability, so that an article can be created from the start. Notability (at least, the SNGs, in combination with the low end threshold of a GNG ) says in practice that we keep these articles around to be worked on and improved to make the article more encyclopedic but drawing editors to focus on relying on sourcing that is independent, secondary, and providing significant coverage of the topic. One can almost envision these (CSDs, the minimum-meeting GNG + the SNGs, etc.) as gateway checkpoints on deciding to keep an article. There is a fair point from Levivich that we don't have a good guide to describe the process of how articles are reviewed for potential merger/deletion at various stages (from draft to mature stages), of which notability only addresses part of that process, nor is the only reason for doing so. Maybe understanding that (by documenting it as best we can come to agreement on that) will better place what notability should be doing. I dont think its far off but again, I believe its about its wording precision at this point, it meaning different things at different context, which messes up newer editors all the time. --Masem (t) 15:12, 11 November 2020 (UTC)

I know what the claim of significance is and how it is used, but I disagree that it is a lower bar to meet than Notability. From reasonably long experience at AfD, I have the sense that many editors there wish for Notability to do the work of distinguishing important from unimportant topics, as well as documentable from non-documentable ones (and I get that vibe from some of North8000's comments on what is truly Notable, as well). It seems to me that we do have a policy-compliant term for this kind of importance, and it is the "claim to significance" - but people generally aren't keen to make WP:NOT arguments at AfD (except for what are clearly edge cases) so we don't have a policy literature on that to the same extent that we do about the WP:N - which, as I have argued elsewhere, is essentially a sourcing requirement.

It occurs to me now, upon reflection, that what some of the SNGs try to do (often awkwardly) is to define what makes a topic within one of these domains significant, but dressing it up in presumptive Notability to go along with the "kludge" we have been discussing. And maybe it is this that rubs me the wrong way about criteria like WP:ENT 2. when it comes to "celebrities" - not that it isn't a reliable predictor of RS coverage, which it probably is, but that "has a large number of Instagram followers" should not in itself be considered a claim to Encyclopaedic significance, which this criterion seems to endorse (and which ENT 1. and 3. do a good job of capturing, as does ENT 2. when it comes to entertainers whose following actually results from their body of work). Newimpartial (talk) 17:23, 11 November 2020 (UTC)

The "notability kludge" has worked 95% of the time but what I meant about reaching the intellectual limits is that the notability kludge's effectiveness % is dropping. At this point in WP's development, we have articles already about most of the "easy" topics, the ones we all agree should have articles. New articles are coming from more specialized, esoteric, or otherwise "difficult" topic areas. For example, we have all the BLPs about American athletes, and GNG works well for American athletes. No so well for African athletes. Especially not 19th century African athletes.

"Significance", in my view, is just another word for "notability" or "importance" or "interest in" or "fame" or many other words. At bottom, it's about "Does anybody care about this topic?" If people care, there's probably GNG sources, and if there are GNG sources, it should probably have a stand-alone page, the thinking goes. If a TV show has a lot of viewers, if an album sells a lot of copies, if an instagram personality has a lot of followers, that's an indicator that people care, there are probably RS, and it's probably notable. We, editors of Wikipedia, are well positioned to determine the notability/significance/importance/interest in/fame/whatever of, for example, a 21st century American athlete. We are not well-positioned to determine the same about a 19th century African athlete. Most of us don't speak the languages, the sources aren't online, and most of us aren't generally familiar with the landscape or topic area. We're reaching the limits of what we can measure when it comes to the importance of topics.

The kludge is not only becoming ineffective, but it's also counterproductive, because it is based on a positive characteristic of the topic (whether you call that "notability", "significance", "importance", etc.). Every time we evaluate "notability" or "significance", we judge the value of the topic. Thus, if we say something is not notable or not significant, we are insulting it, somebody gets offended, somebody vows to "rescue" the topic (and its dignity), and AFDs become flamewars.

We don't need to, and shouldn't, ask "Can we write a policy-compliant article about this topic?" We should only ask, "Is this a policy-compliant article?" It doesn't matter whether it could be, and it's not a question anyone can answer with specificity, especially for those topics for which we have not yet answered the question (19th c. African athletes, etc.). It only matters whether it is. If someone can write a policy-compliant article about any topic, let them do so. If they can't, redirect.

Instead of AfD and RfD we should just have EfD: Entries for Discussion, where we discuss whether a particular title should be an article, a redirect (and if so, to what target), or not an entry at all (hidden, which we incorrectly call "deleted"). Lev!vich 18:06, 11 November 2020 (UTC)

The above discussion, about a social media influencer having X followers, is a very good demonstration of significance versus notability. If I created an article about influencer on the statement they had 1 million followers based on their YouTube subscriber count, that's definitely a claim of significance, but alone not a claim of notability. If no sourcing outside of that YouTube page talks about this influencer at all, then all we have is a claim of significance but had yet to be shown "worthy of note", and thus we'd look to merge or whatever. On the other hand, as soon as you have an article from an RS that establishes that at a certain point in time, that person had a million followers, now we have a claim of notability. One article alone with only that one fact isn't very much significant coverage, but that's at least the first shift from significance to notability. Lacking further sources still might mean a merge or the like but on the basis of barely meeting the GNG rather than outright failing notability. More articles, including ones going in-depth into the person's streaming history and how they got to a million followers, only help strengthen the meeting of the GNG. That's why we have stressed that notability is not about importance or popularity or significance, which is something easily claimed for CSD, but needs to be proven out with sources for the GNG. --Masem (t) 19:41, 11 November 2020 (UTC)

English Wikipedia has a tradition of allowing stub articles in anticipation of further collaborative efforts expanding them. Thus part of the purpose of standards of having an article is to avoid needless churn, deleting stubs or turning them into redirects only to have them turn back into articles again as the content is expanded.

The underlying problem is that Wikipedia editors try to give comparable amounts of encyclopedic content to topics of similar scope, as befitting a general encyclopedia. This requires, however, some kind of value framework to compare disparate topics, and consensus decision-making in a large group isn't well suited for this. Thus English Wikipedia has chosen to examine coverage by third party sources to try to determine what topics should have similar levels of Wikipedia coverage, breaking down large topic areas into articles. While that introduces more information into discussions than just the opinions of editors, it still eventually succumbs to the weaknesses of consensus decision-making. isaacl (talk) 23:32, 11 November 2020 (UTC)

@Barkeep49: The 2008 RfC received only 19% support on the proposal that "SNGs override GNG". The 2017 RfC found clear consensus that no SNG overrides or is a replacement for GNG. Even the discussion immediately above here in #Arbitrary break showed roughly 2-to-1 support for the idea that GNG is the controlling principle, and SNGs are secondary to it. Can you expand a bit on the reasons behind why you think a binary RfC on the narrow topic of the role of SNGs would fail to gain support?

Scottywong, the 2017 RfC close was obviously an overreach, and those commenting on the NSPORT question were clearly not also intending to decide the relationship between, say, NORG and the GNG. Meanwhile, the 2008 compromise preceded the current version of NORG by several iterations, while other SNGs that were recognized in 2008 have since been deprecated. None of these results suggests that a current RfC would gain consensus that the GNG is the "controlling principle" against NORG or NPROF. Also, I don't think a 13-9 result counts as "roughly 2-to-1" and in any case, it is the policy-relevant arguments (only) that are to be given WEIGHT. Newimpartial (talk) 15:03, 11 November 2020 (UTC)

I think the RfC is unlikely to achieve consensus because editors in this discussion have disagreed with it and 2008 was a long time ago. I think it because if you have something about wording, someone could agree in principle but write Oppose because the wording about XYZ is GHI though I think the rest of it is fine. I think it because I don't think it's as binary as you do. Best, Barkeep49 (talk) 15:34, 11 November 2020 (UTC)

Ok, fair enough. I disagree, but I can see that not too many people are agreeing with me, so I'll back off and allow the process to continue. I'll leave my draft in place in case someone wants to modify it or use it as a starting point for a new draft. ‑Scottywong| [yak] || 15:51, 11 November 2020 (UTC)

RfC Drafting

My goal remains, as it was when I BODLY edited the SNG a few days ago to improve what WP:N says about the topic. To that end it seems like an RfC might be necessary. Scotty proposed one draft. I have created two options which I would welcome people to edit and/or comment on which one might be better to launch. Best, Barkeep49 (talk) 02:06, 12 November 2020 (UTC)

I don't understand the purpose of a poll that is asking a purely factual question that readers of the SNGs can clearly answer. Are you trying to learn how well the population of poll participants understand the SNGs? It's as if you had a poll where the options for what color the sky was were: (1) blue (2) gray (3) black (4) it varies depending on the weather and time of day. What useful information would you get out of it? If the RFC came back with a counterfactual consensus what would that accomplish for our policies? It is clearly the case that the only accurate description of our current SNGs is (3) it varies depending on the SNG (see: NCORP being more restrictive, NSPORT being subsidiary, NPROF being independent, SPECIESOUTCOMES being a predetermined answer regardless of sourcing, etc). If you think that should be changed, then an RFC on what GNG says is the wrong way to accomplish that, because changing GNG to promote a dogmatic one-size-fits-all point of view without changing the SNGs cannot accomplish that; it can only put contradictions into our policies. They all would need to be changed at once, and I very much doubt there would be consensus e.g. for relaxing NCORP to allow anything that passes GNG but not NCORP to become notable. —David Eppstein (talk) 02:21, 12 November 2020 (UTC)

Thanks for the question David Eppstein. My bottom line is that I don't think the current Wikipedia:Notability#Subject-specific_notability_guidelines section reflects community consensus. I would like to have a section that does have consensus. That's why four different choices for that section are part of RfC Option 1. However, since most of the discussion has been about GNG/SNG relationship I thought perhaps people might feel that an RfC around that is more successful. Hopefully that makes my purpose clearer. Best, Barkeep49 (talk) 02:28, 12 November 2020 (UTC)

My !vote is for launching Barkeep RFC draft #1 (which I made some edits to), and that's because I'm a strong supporter of Option 4 of that RFC proposal, which is what should have happened here after the bold addition was reverted back in May. Draft #1 is the one that most follows normal procedure: when there is a dispute over what a section of a page should say, put the various proposed drafts up for vote in an RFC. I think BK's RFC draft #2 is likely to come back for Option 3 ("it depends"), which won't be helpful (and won't address the SNG section of WP:N that was added in May). If Scotty's RFC were launched, I would !vote support, but I think it will likely fail for the reasons BK spelled out, and that failure could be misconstrued will definitely be construed by some as consensus for the existing SNG section of N. Beyond that my thoughts are above. Lev!vich 04:31, 12 November 2020 (UTC)

I'm skeptical that any single RfC is going to resolve the big picture SNG vs. GNG. This is, of course, one of the longest-running, extensive, heated debates on Wikipedia, and just invoking the idea of the GNG puts some people off. It seems like there are a lot of people who don't disagree with the fundamental requirement of the GNG, but have some other objection to supporting it (sometimes it's just a historical/interpersonal/principled objection). If anyone really wanted to resolve it, it would (IMO) take starting on a more basic level, getting rid of the terms SNG and GNG. So, e.g. first asking whether or not there's consensus that ~"for a subject to be notable, it should have received significant coverage in reliable sources independent of the subject". When framed that way, it's hard to reconcile dissent with core policies. If there's agreement on that, it would then probably require parsing the different parts of that statement (significant, coverage, reliable, independent) and considering how they apply differently to different subjects (if at all). I'm not necessarily saying this process would be successful, either (or even necessarily worth it) -- just that I think this sort of thing would be necessary to get at the fundamentals. It is, of course, unrelated to what's here presented as RfC #1. — Rhododendrites ^talk \\ 15:33, 12 November 2020 (UTC)

So it seems like RfC Structure 1 is what others think will be more successful (which is great - it was my preference as well) so I have removed the SNG/GNG RfC structure from the draft page. I continue to welcome other feedback on its word/format. Newimpartial, you had indicated you had some tweaks you thought might be helpful. Best, Barkeep49 (talk) 15:52, 12 November 2020 (UTC)

My initial reaction is that I have some difficulty comparing the first three options, given the way they are presented. Is there any way that they could be shown as diffs contrasting with Option 4, or using bold for modified text, or something? Newimpartial (talk) 17:37, 12 November 2020 (UTC)

Newimpartial, the edits I made yesterday to the RFC were intended to improve the exact issue you've raised, I guess I was not successful. What if the diffs were presented like this ("yet another option"), would it be clearer? Or perhaps my changes should be reverted?

• What it used to be:
  • Option 1
  • Option 2
  • Option 3
  • Option 4 - there should be no SNG section (pre-May 2020 status quo)
• What Levivich changed it to (different diffs):
  • Option 1 - current since May 2020
  • Option 2 - proposed revision
  • Option 3 - alternative proposed revision
  • Option 4 - there should be no SNG section (pre-May 2020 status quo)
• Yet another option (Options 2 and 3 don't jump to the section anchor):
  • Option 1 - current since May 2020
  • Option 2 - proposed revision
  • Option 3 - alternative proposed revision
  • Option 4 - there should be no SNG section (pre-May 2020 status quo) Lev!vich 17:47, 12 November 2020 (UTC)

Well, that is a little bit better, but I have the concerns: (1) that there isn't an obvious path to producing alternative proposals in this format (or editing the current ones, at least not that I can see) and (2) that for this purpose I think comparisons to the pre-May 2020 status quo ante ante might be more helpful than comparisons to the "current version" offered by the "Yet another option" diffs.

What I learned substantively from this review (viz. some tweaks [I] thought might be helpful) was that Option 2 in particular makes the unhelpful assertion that the SNGs are presumptive but the GNG is definitive, when it states Note, however, that satisfying the criteria of an SNG is only a presumption of notability, not a guarantee. In cases where it can be shown that GNG cannot been met, the article may still be deleted or merged: a presumption is neither a guarantee that sources can be found nor a mandate for a separate page. As is clear from the first paragraph of WP:N, however, the GNG itself only offers a presumption, not a guarantee, of a topic having its own article. Even if the GNG is met, the article may still be deleted or merged. So in other words, what is presumptive about most SNGs is the "presumption ... that sources can be found", which is rebuttable. However, even sourced articles can be deleted or merged for reasons of encyclopaedic treatment, WP:NOT, etc., and for these purposes GNG-compliant articles are quite as vulnerable as SNG-compliant ones; the "presumption that a topic receive its own article" is no stronger from a GNG pass than an SNG pass.

I fact, I am surprised that I haven't seen this argument elsewhere (and I'm afraid I'll have to go make it somewhere, now), but a major contribution of the SNGs (all of them) is in helping to define NOT with respect to a certain topic area, which doesn't have anything in particular to do with sourcing-based notability. For example, articles on authors that meet the GNG for some reason - but where none of the NAUTHOR criteria are met - should be more liable to be Merged (for example, into articles about events or trends) than sourced articles that represent a solid NAUTHOR pass. This doesn't mean that all authors who fail NAUTHOR are non-notable, as they may indeed be notable for other reasons, but their articles should not be treated at Merge (or deletion) discussions the same way those are treated who pass NAUTHOR. This concept is quite apart from presumptive sourcing-based Notability, which is one reason why text suggesting that the SNGs offer "an indicator of likely GNG" (sourcing-based) "notability" is so radically incomplete. Newimpartial (talk) 18:13, 12 November 2020 (UTC)

The pre-May 2020 status quo ante is no SNG section at all. The section was boldly added May 25, in substantially the same form its in now (current, Option 1). All removals and edits to this section since May 25 have been reverted (except this). The reason I'd support Option 4 is because I agree with Newimpartial that both Options 2 and 3 need further workshopping; what I think should happen (in my view, this is required by our policies) is for the SNG section to be removed and then those who wish to add such a section should workshop language on the talk page and then make an RFC proposing one or more draft sections to add to WP:N. It's somewhat amazing to see multiple administrators edit warring over a bold addition to WP:N, but if it takes an RFC to end the edit war, so be it, and I guess there is no harm in adding Options 2 and 3 (and 5, 6 and 7, if someone wants to write them) to said RFC. Lev!vich 19:04, 12 November 2020 (UTC)

Well that would be a different path towards consensus. Could we agree to remove the section for now and then take the time necessary to come up with a couple different options to present in an RfC for actual approval? Pinging current people discussing this elsewhere in this topic: @Masem, North8000, Scottywong, David Eppstein, and XOR'easter: Best, Barkeep49 (talk) 19:14, 12 November 2020 (UTC)

If you look at the archives on this talk page in May, you'll find that was added because nowhere we actually said "SNG" (as an abbreviation and what that implied) at WP:N and that was confusing people. We need to keep the section for now because we'll leave this page lacking what an SNG is, but it can be marked as disputed for the time being and then when an RFC launched, tagged to indicate that. --Masem (t) 19:27, 12 November 2020 (UTC)

From 2006 until 25 May 2020, Wikipedia got along without an SNG section in WP:N. Surely we don't need to keep the section for now. Lev!vich 19:31, 12 November 2020 (UTC)

I did read that archive before I did the edit that started this discussion. I agree with you that we need a section. Template:Disputed is only for content and Template:Under discussion does not accurately reflect that the section does not have consensus. Are you aware of a different template that could reflect that the guideline section doesn't have community consensus and is under active discussion? If not I'm back to either we go with the the options we have now - the consensus of which might end up being to remove it which I agree is not ideal (but which I prefer over keeping the section that's there now) - or we remove it to try and come up with draft language (or multiple draft language) that people feel good enough about doing to then launch an RfC. Best, Barkeep49 (talk) 19:32, 12 November 2020 (UTC)

As stated at {{Under discussion}}, adding "|section" will tag just the section being under discussion. The {{Disputed tag}} also works the same way. --Masem (t) 19:50, 12 November 2020 (UTC)

Aha. Disputed tag is actually the answer here, I was only aware of Disputed which is just for content. I have added that tag. Best, Barkeep49 (talk) 20:02, 12 November 2020 (UTC)

@Newimpartial I'm going to set aside your thoughts on the specifics of the proposals or the more general argument because, as you note, that really belongs elsewhere in this discussion. In terms of the formatting I considered some sort of summary/description/collapsable box to try and make the differences clearer. However, I was reluctant to characterize Options 1 & 2 (option 3 I'd characterize as a "pared down section"). If you think it helpful are there summaries you can think of after reading it over that might be helpful for others? Best, Barkeep49 (talk) 19:10, 12 November 2020 (UTC)

Guys, I know what the pre-May status quo ante was (and looked at it before making my last comment). What I was trying to say was, I think the best way to present these three proposals is as text added to or modifying the pre-May status quo. This could be done in diffs, in bolded text, or whatever, but should not be presented simply as complete versions where the reader has to compare manually.

I don't know how anyone else participates in RfCs where specific wording is proposed, but my own participation is always predicated on the how the proposals would impact on a specified reference text, and I've seen pretty general agreement in this discussion that the relevant reference is from May.

P.S.: If I had it to do again, I might last week have reverted the guideline to the pre-May version rather than the October version. However, at the time I thought the discussion accompanying the May revision had achieved more consensus than, reading it now, it seems to have had. Anyway, the point was to have something in place during this discussion rather than a continuing edit war, so at least that was achieved. And my appreciation of the limitations of the May changes is much deeper with another week's hindsight. Newimpartial (talk) 19:42, 12 November 2020 (UTC)

I'm confused Newimpartial, that is what the options do because the pre-may status quo is nothing. So the words that are linked all reflect changes because the section started from scratch in May. Best, Barkeep49 (talk) 19:47, 12 November 2020 (UTC)

Right; I see my conceptual problem (the pre-May version is the logical basis, but that doesn't help cast a spotlight on the differences among the alternatives). Thinking about this some more, from a procedural standpoint, perhaps the alternatives are (1) no section - pre-May status quo ante; (2) status quo section since May 2020; and (3) one proposed revised section. I don't see how we can propose two new alternatives in an RfC, and two status quo versions, without confusing the participating editors. If we do get it down to three versions, then the third can be presented (by diff or bold) as a modification of the second. Newimpartial (talk) 19:58, 12 November 2020 (UTC)

Well I tried to create a variant to do what you're suggesting. However, I still don't know how to characterize the differences succinctly as they're quite nuanced. So that led me to a second variant that compares between the May 2020 version and the pared down version, the differences of which I could summarize nicely. Thoughts? Best, Barkeep49 (talk) 20:25, 12 November 2020 (UTC)

Sorry, Barkeep. While you were doing that, I was putting together my own draft on Scotty's RfC page to answer - at least to myself - what problems I have with both Masem's language from May and Scotty's more recent suggestion. Maybe you could let me know what you think of that :). My intent in that draft is to clarify what the current relationship between SNGs and GNG actually is, not to change it to what it "should be"; as I have suggested elsewhere, I think both May and November versions have only incompletely done so. Newimpartial (talk) 20:54, 12 November 2020 (UTC)

Your new draft continues to ignore the reality that different SNGs have and should have different relations to GNG (again: NCORP is more restrictive; NSPORT is subsidiary, NPROF is independent, and SPECIESOUTCOMES is not even a guideline so much as a fiat that all its topics are automatically notable). —David Eppstein (talk) 21:49, 12 November 2020 (UTC)

That is not quite true, since I added , as well as subject-specific guidelines about the sources and criteria relevant to deciding whether a topic should have a dedicated article, which does cover a good bit of NCORP, NPROF, NFILM and NBOOK, at the very least. I didn't get the sense that there was emerging consensus to give a typology of SNGs, much as I personally would prefer that, so I emphasized the relevance of WP:NOT and gave at least a hand wave to the criteria that make, say NCORP restrictive and NGEOLAND inclusive.

If you have more direct language to propose, on this, I would be delighted to see it. No kidding. Newimpartial (talk) 21:57, 12 November 2020 (UTC)

I view the May addition as a useful distillation of current policy/guidelines. It's not a change. Which also means that it is not essential to stay in there if it is disputed, even though I support keeping it. North8000 (talk) 20:12, 12 November 2020 (UTC)

I responded to the talk page because I would love, if possible, to keep this thread focused on the drafting of an RfC. So to that end Newpartial, what are you suggesting we do with that language in terms of an RfC? Barkeep49 (talk) 22:36, 12 November 2020 (UTC)

Well, if the goal really is to reflect the differential status of different kinds of SNGs explicitly in a section of WP:N - a goal that I support - in the upcoming RfC, then I think we need to do what I said last week, and talk through what SNGs make what kind of claims. So far I see NORG, NPOL and NPROF as setting standards higher than or partly apart from the GNG (GEOLAND and SPECIESOUTCOMES would also fit here if we choose to address them). NBOOK and NFILM mostly offer interpretive rules about what counts as reliable, independent sources for Notability in those domains, and NBIO and NSPORT are largely confined to presumptive criteria. But this list isn't comprehensive, and I'm sure someone will object even to these brief suggestions. Newimpartial (talk) 23:37, 12 November 2020 (UTC)

Well first of all SPECIESOUTCOMES is not an SNG so it would be outside the scope of what we're talking about. I really considered whether having some kind of determinative RfC about the SNGs would be productive. But in reality what that turns into, I think, is 13 mini-RfCs (1 for each SNG) and so I ended up rejecting that in favor of the various approaches I've suggested here. Especially because at the end of the 13 mini-RfCs we have data about what needs to go into an SNG section at Wikipedia:Notability but not actually something approaching language itself and that feels like then a second RfC would be needed to agree on what is actually said. Could I trouble you to look at my Variant 2? Best, Barkeep49 (talk) 23:56, 12 November 2020 (UTC)

Continued notability discussion

We're talking about a tweak in a random herd of cats where there is not even any objective defined and agreed on. And with some of the most knowledgeable and best minds in Wikipedia having a very confusing discussion where not even a few folks agree on anything specific. And somehow going to larger scale RFC at this time is going to help that?!. Sincerely, North8000 (talk) 17:58, 11 November 2020 (UTC)

I honestly think you guys are making it way more complicated than it needs to be, looking at every detail and trying to examine every rare edge case. WP:N is a broad, general, high-level, and intentionally vague guideline. It doesn't get into the weeds on things, it paints in broad strokes, and in my opinion we should be thinking that way as well. No one is trying to rewrite WP:N from the ground up here. The main issue, in my view, is clarifying the relative roles of GNG and SNG with regard to determining the notability of a topic. And I think we'd find strong agreement that GNG is the standard for notability, and SNGs are (without exception) a tool to quickly gauge the likelihood of a topic meeting GNG. Topics that meet GNG are presumed to warrant an article, unless other WP policies are violated. Topics that meet an SNG are presumed to meet GNG, unless notability is challenged at AfD and satisfactory sources cannot be produced. That's it. Everything else is a tangent and a distraction. ‑Scottywong| [express] || 18:34, 11 November 2020 (UTC)

In my view, the most impactful of the SNGs is NORG, which clearly is not a tool to quickly gauge the likelihood of a topic meeting GNG. NORG doesn't do anything "quickly", and the framework it provides quite deliberately IMO ratchets up the SNG requirements to deal with the (perceived) pressures of a specific topic area.

I can't see how NORG represents a rare edge case, so any proposed language for WP:N that studiously ignores that case (as did, e.g., the 2017 NSPORT close) will simply never reflect whatever the community actually believes in this area. Newimpartial (talk) 18:55, 11 November 2020 (UTC)

NORG is specifically establish now because we have to fight off people trying to use WP for SEO purpose. Notability, even with the GNG, can be gamed due to how business publications work. The new framework of NORG was established after a large spat of discovering paid editing and COI and the like by businesses, all getting notable articles but really mostly regurgitating press and paid endorsements masked as "articles" in trade magazines. The editors behind NORG spent a lot of time to recognize they had to go above and beyond the GNG to set out rule that would minimize the amount of COI/SEO type article while still leaving articles on truly notable businesses. This is the only case where the SNG supercedes the GNG (in that the GNG is not sufficient and you need to meet this SNG if you fall into NORG).

The only other SNG to stress here in the same vein is NPROF which was developed on the recognizing that academics are an important class of people to document but which are rarely documented in the mainstream (unless the win the Nobel or the like) and that there is very little self-reflecting within academia about themselves - people let their CVs speak for themselves and that's it. But academics recognize implicitly the work of other academics by the research they publish, just that there's very little secondary sourcing that comes along for these people. As such, NPROF exists to document academics that have recognized bodies of work that are otherwise not notable under the usual GNG. This is the only case where the SNG is serving as an alternative to the presumption of a standalone page to the GNG, rather than as an indicator of meeting the GNG, by design. --Masem (t) 19:33, 11 November 2020 (UTC)

I think it's also important to remember that academics typically have huge numbers of mostly-low-quality sources about their work, in citations by other academics to it. In some areas (mathematics for instance) there are also review journals/databases providing in-depth published analysis of every publication they make. One could easily argue on that basis that all mathematicians with multiple reviewed publications meet GNG, a much lower bar than we want to set. Having a separate notability guideline for academics can be interpreted as being restrictive in the same way as NORG, stating that a mere handful of citations or routine-only reviews isn't good enough and that one needs significantly more than that. Whether or not you buy that argument, it is helpful to have the guideline as a way to shortcut that debate and focus on individual academic notability. —David Eppstein (talk) 19:41, 11 November 2020 (UTC)

Right, and my point being are those two presently are the only irregularities in the GNG/SNG structure of the system given what I explained, and those were by consensus when they were established (NPROF I think even predated the WP:N definition IIRC, hence it was a grandfathering of sorts). This is not saying other SNGs - or even project level article guidelines - can have language that may slightly override the GNG, though I think most of the time, this is in the area of carving out special lists of acceptable and unacceptable sources in the topic area for both GNG and SNG evaluation. EG the relatively recent scrubbing of the notability of pornographic actors from NBIO due to the poor quality of sourcing in that specific field. But for the most part outside of NCORP and NPROF, the other SNGs all seem to follow "indicators of notability" approach based on the 2008 RFC. --Masem (t) 19:47, 11 November 2020 (UTC)

And my point about these two cases is that neither would be helpfully elucidated by adding GNG-fundamentalost text to WP:N. I am also not entirely convinced that these are the only such cases, though they are certainly the main ones. Newimpartial (talk) 19:50, 11 November 2020 (UTC)

I disagree that SNGs like WP:NORG supercede or conflict with the GNG. They expand upon the GNG, and define some terms (mostly "significant coverage" and "reliable sources") more precisely for the narrower category of topics to which they apply. This is good, and helpful. Clarifying that GNG is the primary test for notability doesn't water down the SNGs at all. GNG says you need "significant coverage in reliable, independent sources". One might ask, "what exactly does "significant' mean?" The appropriate answer is often, "it depends on the topic, so take a look at the SNGs for more details."

Would it help if we codified this aspect of SNGs into the language of WP:SNG? Making it clear that another role that SNGs play is to expand the specificity of GNG for their topic areas? ‑Scottywong| [prattle] || 19:58, 11 November 2020 (UTC)

NORG (and NPROF to a degree) are the only SNG that goes into depth into explaining what is significant coverage for the topic. All the other SNG are setting criteria for what is an indicator of notability/the existence of significant coverage, but which is not necessarily significant coverage itself. Winning the Nobel itself is not significant coverage, but we know that this produces a plethora of articles at the time it happens to build a good article, if there aren't article before that point, hence it is a great SNG criteria. Participation in the Olympics is not alone significant coverage but it is argued that nearly all Olympians get some type of coverage at a regional or local level once selected to represent their country, hence the significant coverage that comes from that. That is how most of the other SNGs (and NPROF to some extent) are set up. NORG is the only one that goes "this is exactly what we expect as significant coverage and what does not qualify". --Masem (t) 20:24, 11 November 2020 (UTC)

I do not agree with this "are the only" characterization. For example, NBOOK specifies that two RS reviews (inter alia) establish Notability; they are not an "indicator" of it. Newimpartial (talk) 20:31, 11 November 2020 (UTC)

While there are likely other cases like that in the SNGs that say "if these types of sources exist, it is presumed notable", this is not the extent of what NORG is doing. To take the NBOOK 2 reviews, I again set forth the hypothetical case if that's the only two reviews of the work that exist, and for all purposes no other significant coverage of the work exists. Then depending on how much that provides, it might be kept, it may be merged, etc. Or to phrase it another way: that condition in NBOOK sets a minimum requirement to establish the presumption of notability but not the only way a book can have a standalone article. NORG sets a requirement on the type of sourcing in general that must be present on a business or organization article in the first place to even have an article there, with no other route avaiable. --Masem (t) 20:47, 11 November 2020 (UTC)

I am not at all suggesting that NBOOK has a similar logic to NORG. I do feel that it is somewhat similar to NPROF though, in that I do not think there is any substantial editors that would treat the two-review threshold in NBOOK as a "rebuttable presumption" any more than they would the NPROF criteria. The more I dig into this, the more I find what seem to be rebuttable and non-rebuttable criteria mixed within the same SNG: e.g., the two review threshold seems to me to define an aspect of GNG notability for books, but the best-seller list provision seems more of an indicator to me. Newimpartial (talk) 21:09, 11 November 2020 (UTC)

The two review book criteria is still a rebuttable presumption but it is one that is very unlikely to trip an actual rebuttal because you have to prove that those reviews (which by being reviews should be providing significant coverage) aren't sufficient AND that there are no other sources to add. Again, to try to say why NORG is different, the challenge there is not the rebuttable presumption but to look through the sources and ask, is this meeting the hard requirements expected of the sourcing (which does include evaluation of significant coverage); we don't presume an org is notable until it has met those. --Masem (t) 22:18, 11 November 2020 (UTC)

Ok, maybe this is the source of the disagreement: I think we can all agree that (some) SNGs perform two different functions: one function is to provide criteria for what topics are likely to satisfy GNG (e.g., if an individual wins a Nobel prize, if an athlete competes in the Olympics, etc.). The other function is to expand upon GNG within the context of their topic area, to further define what "significant coverage" means within their narrow topic area, and to provide specific examples of what are or aren't considered reliable sources within their narrow topic area, etc. The former function is the one that can't supercede or replace GNG. Just because someone competed in the Olympics isn't a guarantee that there is significant coverage about them, even though it is extremely likely that there is. The latter function works in conjunction with GNG, expanding and clarifying it for each major topic area. It doesn't make sense to talk about whether the latter function "supercedes" or "replaces" or "conflicts with" GNG. It is essentially an extension of GNG. The latter function doesn't define notability, and isn't a complete notability test, but it clarifies WP community norms surrounding what is considered "significant" and "reliable". ‑Scottywong| [communicate] || 22:21, 11 November 2020 (UTC)

I strongly disagree that any SNG (outside of NORG and NPROF) are defining "significant coverage". The NBOOK "two book" is still an indicator of notability that is suggesting that if two reviews exist, more reviews and other details about the work likely exist to be able to clearly satisfy the GNG. For example, take WP:ANYBIO #3 - thats just a single source mention, clearly not sufficient for GNG, but a good indicator. Most of these were all developed on the 2008 RFC so they are not going to be defining what is "significant coverage" for that field to override the GNG like NPROF clearly does. --Masem (t) 22:37, 11 November 2020 (UTC)

The thing is, Masem, that what you are describing isn't what NBOOK actually says. The summary text is A book is notable, and generally merits an article, if it verifiably meets through reliable sources one of a number of criteria, the first of which is met by having two RS reviews. NBOOK does incorporate the idea of presumptive notability, but definitely not of the form "if it meets NBOOK it probably meets the GNG". Rather, it says - in line with the current WP:N - that A book that meets either the general notability guideline or the criteria outlined in this or any other subject-specific notability guideline, and which is not excluded under the What Wikipedia is not policy, is presumed to merit an article - in other words, the presumptive aspect is the same whether the Notability claim is based on NBOOK or the GNG, and that in either case articles can be merged "at discretion", presumably where there is consensus to do so. So while you are continuing to argue for the indicator interpretation, and I can see how NBOOK 3 or NBOOK 5 could be "rebutted" in cases where the sourcing is poor, NBOOK 1 seems to be clearly definitive of Notability in the sense Scottywong just described. Newimpartial (talk) 23:09, 11 November 2020 (UTC)

But the practicality that if you recasted NBOOK overall as clearly making it presumed notability, the practice of what happens with books that are created based on NBOOK #1 (two in-depth reviews) means that in practice those articles will likely never be challenges or if challenged will survive the notability challenge since two reviews happen to typically be sufficient for the GNG. I know it appears that NBOOK has both indicators of notability and one that states a type of significant coverage, but to try to describe that at a higher level across all SNGs becomes super complicated. It is easier to recognize that that we can just say that's still an indicator of notability like nearly all other SNGs, and recognize that in practice is 99.9999% sufficient to meet the GNG on its own.

Part of what I want to be careful here is that we have had problems - notably with NSPORT - of these SNGs adding criteria without getting global consensus. We don't want SNG editors to be adding cases that supercede the GNG without clearly getting global consensus, and when they do , it should be for a very good reason like the NORG situation (to fight off COI issues). So we should not be presenting SNGs as being alternatives to the GNG definition, but as indicators or presumptions of notability. --Masem (t) 23:29, 11 November 2020 (UTC)

But the fact remains that the reality of the SNGs actually is super complicated, and abstract formulations that don't reflect that reality are misleading or just don't work (viz. the 2017 NSPORT RfC close). We wouldn't gain anything, in my view, by pretending n policy that the SNGs all work the same way when they clearly don't - even in policy much less in practice. NSPORT and NORG are simply different kinds of guidelines, and the rest of the SNGs (or even specific criteria) are strung out between them AFAICT, rather than all being clustered close to NSPORT. So if and when we actually change something, I would like to see policy that describes correctly how it is (supposed to be) working, and if we could take the energy people displace from underlying NOT concerns into Notability at AfD - and give it somewhere more productive to go - that would be the most significant benefit to the project I can imagine coming out of a WP:N policy discussion. Newimpartial (talk) 23:43, 11 November 2020 (UTC)

In actuality, the SNGs are a super simple situation with the exception of NORG and NPROF for reasons given that only need to have wording alignment between WP:N and the SNG to make that clear. They have worked this way since 2008. The problem we have over the years is that when an SNG adds or modifies things without checking community consensus , or when we get cases where editors get upset that their articles are being targeted for deletion and don't understand why. I'm all for figuring out language to make sure that it smoother, but this discussion is artificially making the situation far more complex than it needs to be and how it has been in practice. --Masem (t) 03:32, 12 November 2020 (UTC)

You are the one who said earlier that it was "super complicated", Masem, and you have drawn attention to some of the cases that make it so. I think I'll stick with your earlier assessment. Newimpartial (talk) 14:02, 12 November 2020 (UTC)

No, I said that there was a very simple way to explain it, and then there was a very complicated way to explain it, and there seems to be far too much focus on the complicated way. We should be aiming for KISS here, because the more complicated we try to make notability as a concept, the less helpful this discussion towards fixing the problems it becomes. --Masem (t) 14:45, 12 November 2020 (UTC)

@Masem: what would you say IS that simple was to say it? And, regarding where to clarify it, I would like to point out something. W:notability actually structurally has two parts. The beginning is what is accepted as Wikipedia's meta statement regarding the existence of articles and the relationship between GNG and SNG's. (e.g. it even creates/defines wp:not's unique place in the meta-rule. Amongst the many reasons it is accepted as authoritative is that anything in wp:notability reflects much broader scrutiny and consensus than material at individual SNGs, some of which are dominated by fans in that topic area. The remaining 3/4 of wp:notability is GNG. So the beginning of wp:notability is a good place for any authoritative clarification. North8000 (talk) 15:05, 12 November 2020 (UTC)

At least in terms of the relationship of the SNGs to the GNG, that with specific exceptions of NORG and NPROF, they all serve as indicators and set a rebuttable presumption of notability that can be challenged for a merge or AFD if a thorough and proper BEFORE search reveals no additional sourcing or significant coverage. NORG and NPROF have to be identified as unique in their relationship to notability due to the situation with those topics, and we have to recognize that there are other topics per WP:5P1 and OUTCOMES that notability is sometimes not considered. What's being complicated in this discussion is trying to fit NORG and NPROF and these outlyers into that scheme, or trying to figure out specific criteria separately like NBOOK's 2 reviews as different; that makes it far more complicated than it needs to be.

The only thing that I would add, given all this, is that meeting notability itself is not a black-or-white test, and having only minimal sourcing or significant coverage can still lead to a challenge related to notability; the more you demonstrate sourcing and significant coverage, the less likely it will be challenged. --Masem (t) 15:17, 12 November 2020 (UTC)

But that still isn't true. WP:NFILM and WP:NPOL each contain restrictions on what should be considered notable, similar to (though more limited than) those in NORG, and NFILM and WP:NBOOK each contain guidance about what count as RS in those domains in terms of independence and significance. To reduce the status of NFILM, NPOL or NBOOK to a "rebuttable presumption of Notability" (like NSPORT or NBIO) would be a disservice to readers of the guideline, because it would not be describing what these SNGs actually contribute to defining Notability. Newimpartial (talk) 15:29, 12 November 2020 (UTC)

It's still relatively easy to make an additional broad statement about what the SNG's are doing, rather than trying to make what I think is an overly convoluted stance. Again, excluding NORG/NPROF, they all provide indicators of rebuttable presumptions for notability, and they all can provide cases where standalone articles should not be created even though it may be possible under the GNG or the SNG's criteria, and they all can outline sets of reliable sources that are good for that field in addition to what are universally accepted reliable sources. All these functions are important to notability, but they relate to different parts of it. None of these are overriding the GNG itself (outside NORG/NPROF) with these functions. --Masem (t) 15:57, 12 November 2020 (UTC)

I share North8000's assessment of the situation [5]. Honestly, this entire process has made me less convinced that "notability" is a well-articulated concept. The advice in the GNG doesn't actually make very clear at all when a topic ought to have a stand-alone article. The only thing it contributes beyond baseline policies is the notion of "significant" coverage — but "don't include totally trivial stuff" is equally good guidance for what goes within an article, too. XOR'easter (talk) 18:44, 11 November 2020 (UTC)

"Adding GNG-fundamentalist text to WP:N", as Newimpartial puts it, would also require rewriting the introduction to WP:N, which is GNG-ecumenical. XOR'easter (talk) 20:34, 11 November 2020 (UTC)

Unreliable sources for notability

According to this suggestion, I wanted to bring this discussion to the right place.

WP:GNG requires WP:SIGCOV in WP:RS/WP:IS. This is a clearcut policy, esp. if unreliable sources refers to sources based on user-generated content, paid content, press releases, etc.

However, the way WP:RSP has gone, it appears that many partisan sources or other major media sources that have in the past pushed particular agendas or even conspiracy theories have been listed as unreliable.

My contention is that there needs to be some sort of a differentiation between what's considered a reliable source for factual citation, and what's considered a reliable source for notability-related significant coverage. e.g. while Bild, Daily Kos, The Electronic Intifada, Fox news talk shows, Metro, Telesur, The Onion etc have been deemed unreliable for factual coverage, in my view, a profile in any one of these should count towards WP:SIGCOV (I mean, if The Onion is lampooning you, you are probably notable.). On the other hand, any amount of coverage in unreliable sources such as Blogger, Facebook, LinkedIn, Medium, Twitter, Patheos, PR Newswire etc should not. Perhaps instead of RS/IS, we should require SIGCOV in noteworthy sources/ independent sources, even if the noteworthy sources are unreliable for factual information.

Any thoughts? — Ad Meliora ^Talk∕_Contribs 09:55, 9 November 2020 (UTC)

• While in no way supporting “unreliable sources”, the pointer to “reliable” is slightly miscalibrated. The GNG requires secondary source material, which means creative content of an author in commenting on accepted facts. A better word is reputable sources. The GNG should require comment on the topic by reputable sources. “Factual citations” does not go to notability, facts are cited to primary sources. Is The Onion a reputable source? I think no, not really, two satirical sources of coverage should not be sufficient to merit inclusion as a standalone article. —SmokeyJoe (talk) 11:35, 9 November 2020 (UTC)
• Unsure, notability is not fame. The man who sticks 15 pigs up his bum will get coverage, he is not notable.Slatersteven (talk) 11:38, 9 November 2020 (UTC)
  • Coverage. Coverage, but probably not comment. Does the author of the source use the authors own adjectives to comment? The reliability of the count of pigs does not inform the question of Wikipedia-notability. —SmokeyJoe (talk) 11:57, 9 November 2020 (UTC)

  Is this man related to the old lady who swallowed a fly? Reyk _YO! 15:52, 9 November 2020 (UTC)

• We definitely want an article that ends up with reliable sources. While WP:RSOPINION allows for the use of non-reliable sources to include significant opinions on topics, I think its reading would imply that reliable sources have to be present from the start, and that should be reflected in WP:N - that we're looking for coverage from reliable sources as well, most of the time. So definitely would avoid considering typically non-reliable sources as part of notability indicators. --Masem (t) 14:57, 9 November 2020 (UTC)

Context of "Primary" sources

Some sources are primary for certain kinds of information but independent for other kinds of information. This distinction should be highlighted somewhere.

e.g. YouTube and Twitter and Medium and Patreon are obviously self-published media, and can not be relied on for factual information. I mean, obviously.

However, from these same sources, statistical information such as number of channel/profile subscribers, video views, tweet likes etc are clearly reliable pieces of information coming from independent sources.

This has multiple use-cases, including citation of this information in general, and e.g. WP:ENTERTAINER lists "cult following" as one of the criteria. If an entertainer has an inordinate number of followers/subscribers, that may or may not be considered cult following, but it should be legitimate to use social media stats to make that argument. Remember, the whole notion of "cult" following is that it's not mainstream, and therefore unlikely to be covered in traditional RS. — Ad Meliora ^Talk∕_Contribs 11:29, 9 November 2020 (UTC)

Actually they may not be as they can be manipulated to increase numbers.Slatersteven (talk) 11:31, 9 November 2020 (UTC)

Exactly: social media numbers like follows are too easy to game so we don't use that for notability indicators. (I think they pass enough of a CSD test to be kept based on a discussion there, but that's it). --Masem (t) 15:00, 9 November 2020 (UTC)

Yes, follower counts and the like can be gamed, and moreover, their meaning is context-dependent. What counts as a "viral video" now might not be the same as what "viral" meant ten years ago (see the AfD for Al Gore's Penguin Army). Also, having a "cult following" is compatible with getting RS coverage. There are whole scholarly books on "cult classics of cinema", for example. XOR'easter (talk) 20:35, 9 November 2020 (UTC)

There are problems with WP:ENT, mostly arising from the inclusion of the term "celebrities", but I don't think the notion of "cult following" was ever intended to be measured using "follower" statistics. In context, the point seems to be that either a broad fanbase or a narrower but "deeper" fanbase can count for E.2. I don't really see how a "number of followers" measure would represent either one, TBH. Newimpartial (talk) 15:16, 9 November 2020 (UTC)

@Ad Meliora, please read Wikipedia:Identifying and using primary sources and Wikipedia:Independent does not mean secondary. All sources are primary for something. That is a completely separate question from whether they're independent for anything. WhatamIdoing (talk) 23:10, 9 November 2020 (UTC)

YouTube and Twitter and Medium are media platforms, not sources. They do not provide any significant editorial direction themselves, as a whole; they merely make it possible to connect content-providers to viewers. The source is whoever published the YouTube or Twitter or Medium content. If it's an individual person or fly-by-night group, it's probably not reliable. If it's an established organization that provides the sort of fact-checking and editorial control that we expect of reliable sources, it's probably reliable. So blanket statements that YouTube and Twitter and Medium are unreliable completely miss the point. —David Eppstein (talk) 18:09, 10 November 2020 (UTC)