User talk:Joe Roe/Notability and new articles

Draft RfC on notability and new articles
Disagreement on how much responsibility NPP has to check notability has come up across a few discussions recently. and I discussed it at User_talk:Usedtobecool and both thought that the time might have come to establish what the broader community's expectations are.

, if you have time, your thoughts on how to phrase such an RfC would be appreciated. I've started a draft at User:Joe Roe/Notability and new articles. –&#8239;Joe (talk) 16:34, 22 March 2024 (UTC)
 * To me only the community can decide if an article is notable and the way that happens, in reality, is either by no one nominating it for deletion or it being nominated for deletion and by being kept when at AfD So no single NPP can establish notability. Perhaps something like "should always confirm that an article appears to be notable under the notability guideline" and similar permutations. This is imperfect but is the best wording I can come up with in the time I have right now. Barkeep49 (talk) 17:55, 22 March 2024 (UTC)
 * I think more clear concise language to this effect would be to swap establish for check in the reviewer prompts. That way it's centered on the question of "what actions should new page reviewers be taking" rather than implying that reviewers have decisive authority. signed,Rosguill talk 18:56, 22 March 2024 (UTC)
 * Thanks both. I agree this is better wording. –&#8239;Joe (talk) 08:04, 24 March 2024 (UTC)
 * Is RFCing this a good idea and a good use of time? Seems like this is likely to have an outcome of question 1 option 1 (NPPs need to make a good-faith attempt to research and estimate the notability of every article) and question 2 option 3 (article creators do not need to place enough citations in the article to help NPPs easily verify notability). In my opinion, this has been the status quo for a very long time.
 * Note that question 2 at this point is becoming a bit of a perennial proposal. For example, Village pump (proposals) is currently in progress and is likely to have the same outcome as Requests for comment/Deletion of uncited articles, which was closed only a month ago as no changes to current policy. – Novem Linguae (talk) 18:50, 23 March 2024 (UTC)
 * On question one, I know that is your opinion, but others—myself included—disagree. We could go on forever making assertions and counter-assertions on what is the status quo based on our respective experiences, or we could put it to the wider community and get a definitive answer – hence the need for an RfC. I do not know what the outcome will be, and I don't think anyone else does either.
 * On question two, this is indeed similar to the proposals in those discussions (which I participated in), but different in that it only seeks to establish community expectations, not create a new process or rule. Many of those in favour of a hard minimum citation requirement have argued that there is already an "unspoken" expectation that all articles contain sources, or that responsible article creators point to GNG-level sources – those against say there is no such thing. So making the unspoken spoken here could help make progress on that dispute. But that's a side point, my main reason for including the second question is to nudge respondents into thinking of NPP as a two-sided process. I hope that way we can avoid a worst-of-all-worlds scenario where the community asks NPP to become the notability police but does not put any expectations on article creators that would make that feasible. –&#8239;Joe (talk) 08:19, 24 March 2024 (UTC)


 * I don't think this is ready for an RFC.
 * The first question
 * Words like "establish" and "check" are unclear. If you mean "does a BEFORE search" instead of "looks at the sources already cited on the page", then you have to actually say "does a BEFORE search", or something very close to that.
 * @Usedtobecool, when you said on your talk page that NPPers tell you that they always check notability, do they mean "I always use my favorite web search engine" or do they mean "I always look at the ==References== section to see what's already been cited"?
 * Perhaps this would be a little clearer: Some years back, some experienced editors did a sting operation on new article reviewing.  They picked out notable subjects (mostly BLPs), created brand-new accounts, and wrote "ugly" articles (e.g., with bad formatting or other simple mistakes that you would expect a newcomer to make).  The results were not encouraging; many notable subjects were recommended for deletion through one process or another.
 * If I did that today for a BLP whom I'm certain is notable, and I left it in a barely cited state (one URL, just enough to escape BLPPROD), how many of NPPers do you think would actually do a WP:BEFORE search to check for uncited-but-WP:NEXIST sources?
 * The second question
 * Here, I think the word should is not our friend. Different people have different ideas about what that word means.  We could ask:
 * Shall we create a new rule, similar to WP:BLPPROD, that requires all new encyclopedia articles (but not disambiguation pages and pages, such as the List of lists of lists, whose purpose is primarily navigational) to cite at least one source?
 * If we want this rule, we should just have this rule.
 * The bigger question
 * At some level, NPPers are WP:VOLUNTEERS, and we can't require them to take positive actions. We can require them not to take actions (e.g., "Do not mark a page as patrolled unless you have done a BEFORE search"), but we can't require them to actually do that work (e.g., "If you look at an article, you have to do the whole reviewing checklist, and we'll magically know if you only checked for CSD-worthy problems and abandoned the boring parts to be done by other editors!").  So I think that it might make sense to ask the NPPers if they want to do this work, or if they'd rather than some other part(s) of the community took on that work. WhatamIdoing (talk) 00:03, 24 March 2024 (UTC)
 * Thanks, this is helpful. But is it just wording that makes it not ready? Or something deeper?
 * The difficulty with the first question is that it depends on the outcome of the second. If article creators aren't expected to 'show their sources' (the status quo), then basically check/establish = WP:BEFORE. But if we start asking people to list the case for notability up front, NPP could potentially get away with a quicker check. I did mention BEFORE after the options, but perhaps that's not clear enough?
 * I think asking to create a hard rule requiring sources is a non-starter at this point and not quite what I'm trying to get at. The question is more does, as a guideline subject to the usual exceptions etc., the community expect article creators to make a positive case for notability from the start, or is that only something that comes into play relevant when it is challenged at AfD? In other words, do we give creators the benefit of the doubt when it comes to notability? My impression is that the most old hands, and the written-policy status quo, says yes; but that many younger, maintenance-focused editors say no, there is something similar to WP:BURDEN for notability. This seems to me to be the source of many of the recent disagreements regarding notability, sourcing, and new article creation.
 * It's a good point that this is really a negative ("don't review if...") rather than a positive instruction, though I suppose the end result is the same? I'll see if I can improve the phrasing in that regard.
 * I also thought about a sort of pre-RfC amongst current reviewers. Two things gave me pause on that. The first is that I'm quite sure there won't be a consensus; we've already discussed it quite a bit, and we're deadlocked, because only a tiny proportion of the c. 1700 editors that can mark new pages as reviewed follow WP:NPPR/WP:NPPC. The second is that NPP is not an independent part of the community; its composition reflects what is expected of it. If the community says NPP should be about patrolling for CSDs, it'll attract people who find zapping spam and vandalism rewarding. If the community says NPP is about policing notability, it'll attract people who think notability is important. And importantly, that decision does not just affect patrollers themselves, but everybody who writes new articles. So I think it's something that needs broad input: it could be that the community ends up asking for something infeasible (e.g. option 1 + option 3), in which case we'll just have to come back to them in six months and show that it didn't work. But I hope it opts for one of the more practicable combinations (e.g. option 2/3 + option 3 or option 1 + option 1/2). –&#8239;Joe (talk) 09:34, 24 March 2024 (UTC)
 * (The wording needs help, but I think it goes deeper than that.)
 * Notability checks
 * I think you're making a lot of assumptions. For example,  is not necessary in all cases.  In some cases, the NPPer would have enough prior knowledge, without having to check or search for anything, to be able to say "That's the athlete who scored the winning goal in yesterday's big game; of course he's notable" or "Oh, look, another boring article on some plant species.  Well, species are basically always notable...".  There are a lot of subjects that are basically always notable:  species, diseases, heads of state, natural disasters with multiple fatalities, this year's edition of major sporting and cultural events.  You don't need to find sources to determine whether the 2024 Academy Awards is notable; you just need to know that we have articles on every single one of them since they started in 1929.
 * I think it is feasible at this time to expand the WP:BLPPROD rules to all subjects, for all newly created articles. That means naming a minimum of one source, but not necessarily fully demonstrating notability.  It would be very quick for NPPers to assess (either it already has a source, or you tag it with sticky prod), and it would not require any subjective and therefore disputable judgement calls (the newspaper article is 200 words long; is that SIGCOV or not?).
 * Requiring authors to add one source doesn't mean that they're adding a source that demonstrates notability. (For one thing, most of us think it usually takes at least two sources to do that.)  It might make notability checks a little more feasible, and it might slightly reduce the number of insta-rejected encyclopedia articles (because, in reality, NPP basically doesn't accept unsourced articles (your own track record)), but it wouldn't have a significant impact on that process.  It would, however, make our written rules more closely align with our actual practice, and it could clarify expectations for the depth of work being done by NPP.  If we say "New encyclopedia articles must cite at least one source", and NPP says "Okay, we'll check for one source", then we'll have clarity about what NPP does (e.g., checks for one source in the current version of the article) and what NPP doesn't commit to doing (e.g., perfect review of the subject's notability).
 * Path forward
 * I don't think the proposed questions would address your question about does...the community expect article creators to make a positive case for notability from the start, or is that only something that comes into play relevant when it is challenged at AfD? I agree with you that clarity would be desirable; I don't think the first question addresses this at all.  The second question, however, has a problem of who decides that, followed by quis custodiet ipsos custodes?  Determining notability is complex and not really something that a single editor can always do (especially if the decision is against notability).  It could create a system ripe for bias abuses (I'm pretty sure no article about that political party/school/genre/country are notable), and then implicitly dumps it all on NPP.  Someone would guess wrong about the notability of a subject, and we'd end up with another round of outrage.  Previous rounds of outrage are why we've had all these restrictions, instruction creep, and attempts to "school" or otherwise professionalize NPP.
 * I think officially adopting a rule that new encyclopedia articles must cite at least one source would take a baby step towards having article creators make a positive case for notability from the start, without overburdening anyone. (I suppose that if the community clearly rejected such a rule, then that would also provide some information around that question.)
 * If you remember our discussion last month about splitting NPP reviews into two separate buttons (an urgent check for CSD followed by a more leisurely check for notability, etc.), then I think that's another way to reach the goal of providing clarity to NPPers and the broader community about what they can realistically expect from NPP work. This might attract (more) anti-spam/anti-vandal NPPers, and some of them might even breathe a sigh of relief to know that they are officially permitted to do just the part that they care about.  It might also create an opportunity to recruit subject-area specialists for the second pass (e.g., to recruit people from WikiProject Pharmacology to assess notability for drugs and drug manufacturers, without them feeling like they had to learn the whole NPP process).
 * The bottom line is that I think the next big RFC in this area should sound more like this:
 * [RFC question:] New articles about BLPs are required to cite at least one source.  Should we extend that requirement to all new encyclopedia articles?
 * [Explanation:] When the Proposed deletion of biographies of living people policy was adopted in 2010, it required that all new encyclopedia articles about living people cite at least one (1) source.  New articles without any sources whatsoever were tagged with {{subst:Prod blp}}, and if no sources were added, they could be deleted a week later.
 * I now propose that this process be expanded to cover all new encyclopedia articles, regardless of subject. This would mean that new encyclopedia articles about any subject (but not, e.g., Disambiguation pages, redirects, and lists such as the List of lists of lists that are primarily navigational in purpose) could be proposed for deletion if they remain completely unsourced for more than one week after creation.
 * Current state: Encyclopedia articles about living people that remain completely unsourced for 7 days can be deleted.  Encyclopedia articles about other subjects have no such deletion pathway.
 * Proposed state: New encyclopedia articles about all subjects that remain completely unsourced for 7 days can be deleted.
 * Note: About seven years after its original adoption, BLPPROD was expanded to cover older articles about BLPs.  In the future, the community would obviously be free to expand this in any way that it chooses, but the present proposal is solely for new encyclopedia articles and would impose exactly the same requirements as are already in place for BLP articles.
 * I think this would be likely to pass, and I think that NPP could then adopt that quick and easy assessment as their commitment. WhatamIdoing (talk) 20:22, 24 March 2024 (UTC)
 * I've just read over the current discussion at Village pump (proposals) and I'm no longer certain that it would be adopted. It seems that many editors would rather run unsourced → draftspace → delete than unsourced → tag for a week → delete. WhatamIdoing (talk) 21:11, 24 March 2024 (UTC)
 * "One source" thing is a non-starter for the community and for NPP as well because it does not help with the problem NPP has, which is having articles that look borderline notable but do not contain positive proof of notability cited in the article. If we were to ask for a requirement on article creation, it would be "cite at least one source that meets WP:SIGCOV or one source that verifies a particular SNG". We can't ask that of new editors but I wish we would of editors with, say, 5000+ edits or 20+ article creations. We don't and they keep churning out borderline crap that's not good enough to happily approve and not bad enough to take them to ANI.So, to address the current form of the RFC proposal, For the first one, it is the question I want to ask the community. But I would clarify that we're asking about articles where sources in the articles give insufficient proof of notability, and I would clarify that the question is regarding approval of articles, not nomination for deletion. For the second one, I do not think it has a chance in its current form because the community will never support putting that amount of responsibility on new editors. Isn't the current minimum, like 10 dummy edits and 4 days? The community likely won't support adding that requirement for experienced editors as I mentioned above, but I think that's the way to go if we want an affirmative answer to such a proposal, with as much workshopping beforehand as it takes. If on the other hand, all we are seeking with the second question is to set the backdrop for the first question, it's fine but I think we'd have to discuss presentation further. The current proposal likely does not drive home the connection between two requirements. — Usedtobecool ☎️ 17:41, 29 March 2024 (UTC)
 * I don't think the community would be adopting this for the sake of the NPP backlog. I think they would be adopting this for the sake of limiting the (small) inflow of articles into Category:Articles lacking sources.
 * Adding one source – any source, so long as there's something – doesn't really make the work of determining notability faster (we could agree that NPP need not do anything more than glance at the ref list to make sure that one source was cited) but would:
 * give you a written rule that was obviously and indisputably "violated" by those "editors with, say, 5000+ edits or 20+ article creations" when they really did create 100% unsourced articles, which means you could have discussions at ANI about individual behavioral problems, and
 * would more closely align our written regulations with the actual practice (because NPP currently doesn't appear to accept 100% unsourced articles in practice).
 * BTW, I looked at the first five articles in the current month for that cat, and four of them were mistagged: they had unref (which qualifies BLPs for sticky prod) and should have had no footnotes (which does not).  The actual practical change for extending the existing BLP rule to all subjects would probably be small. WhatamIdoing (talk) 18:47, 29 March 2024 (UTC)
 * Just to be clear, my aim here is certainly not to get NPPers to do a WP:BEFORE on every article. I don't think it is necessary to do that and I don't think anyone is asking them to. I've been strongly critical of the tendency to think that one person can determine notability, and in my own approach to reviewing I recommend disregarding notability aside from the kind of 'common sense' screening you describe.
 * That said, a lot of the latest generation of NPPers do feel like it's their job to determine notability one way or the other. I think that has little to do with policy or broad consensus and a lot to do with the past expansive visions of NPP of certain dominating individuals, which you've also alluded to, but it's too late to put that genie back in the bottle. So the system you are worried about creating is the one I am worried already exists, at least for a substantial subset of reviewers: they do (relatively shallow) checks for notability, and if they don't find it immediately the article gets dumped in the draftspace black hole. But of course it's hard to do notability assessments, so as Usedtobecool describes a lot of articles just get left behind, clogging up the queue and causing (IMO misplaced) resentment against 'lazy' article creators – until someone with a more pragmatic approach comes and clears them out.
 * So yes, I want and hope the community will say that no, don't be silly, you don't have to do detailed notability checks on every article. That would let us get everyone back to a more streamlined and pragmatic approach to NPP. But if it goes the other way, at least we will be able to proceed with a clear goal and try our best to make it manageable.
 * On sourcing, I actually do usually 'pass' unreferenced articles in the sense that, unless there are other problems, I tag them with unreferenced and mark them as reviewed. But I'm old fashioned in that regard, and I'm willing to concede that the vast majority of NPPers now draftify unreferenced articles on sight. Which is reasonable enough but a bit unfair on new editors who are never told about the 'unwritten rule'. I go back and forth on whether the best solution to this is to try and get NPPers to stop doing it (probably via WP:DRAFTIFY, which IMO currently does forbids it but is terribly written) or to codify the new status quo as you suggest. It was probably a mistake to conflate this with the issue of notability at NPP. –&#8239;Joe (talk) 09:03, 30 March 2024 (UTC)
 * "@Usedtobecool, when you said on your talk page that NPPers tell you that they always check notability, do they mean "I always use my favorite web search engine" or do they mean "I always look at the ==References== section to see what's already been cited"? [and the rest of the discussion]" I don't want to give the impression that I have more data than everyone else. But I mean the former, based on anecdotes I've read, and because, it's easy for a NPP reviewer to do every one of the rest of things they are supposed to do, and relatively quick too. NPP backlog is almost entirely articles that got stuck on the last step: reviewers are not comfortable doing either of nominating them for deletion or marking reviewed. You may be right that WP:BEFORE doesn't actually happen most of the time, but the trouble with NPP backlog is articles don't get marked reviewed without a BEFORE, they just get skipped, dozens of times by perhaps dozens of editors. We couldn't tell reviewers to do a BEFORE; BEFORE is hard and uninviting. And there is still no appetite in the community for making BEFORE the responsibility of any party other than the one seeking deletion, including page creators. But we could tell reviewers they can mark articles reviewed without doing a BEFORE if there's a reasonable claim of significance/indication of notability in the article that they've verified. I don't know if that would be a good or a bad thing. But I think it would take care of the backlog. Another argument in favour is once it's determined that there is nothing urgent about whether the article should be kept or removed, the decision to BEFORE and AFD can be postponed and is also best left to editors of the applicable topic areas, who are likely to know better than a random editor with the NPP button.Poorly written articles on notable topics may get deleted by various pathways for the same reason that very few people do a BEFORE. It's unfortunate, but the current proposal does not affect that part either way. It's a different issue. The question isn't BEFORE and deletes, it's BEFORE and keeps. — Usedtobecool ☎️ 17:20, 29 March 2024 (UTC)
 * I think you've hit the crux of the issue with reviewers are not comfortable doing either of nominating them for deletion or marking reviewed. For old hands like me, that's a category that doesn't exist: if you're not prepared to nominate an article for deletion, you mark it as reviewed, end of. But I think you are absolutely right that a large number of articles are getting passed over because people feel that they cannot hit the tick without making sure the topic is notable. What I would dearly, truly like to understand is why they have this barrier. Because [you] can mark articles reviewed without doing a BEFORE if there's a reasonable claim of significance/indication of notability in the article that they've verified is exactly how I learned to review and, as far as I can tell, what the guidelines still say. –&#8239;Joe (talk) 09:09, 30 March 2024 (UTC)
 * It sounds like we have two possible futures, which look approximately like this:
 * Defence in depth: NPP is our first line of defense against CSD-worthy problems, but they aren't burdened with everything.  Their job is to check for copyvios and big, obvious, common-sense problems.  If the individual NPPer doesn't see a (reasonably clear) reason to send off for deletion, then they should get it out of the NPP queue.  "Mark as reviewed" means "does not obviously need to be deleted".  This means that they'll tag an attack page for CSD, and send a wholly unsourced BLP off to WP:BLPPROD, but it is specifically not their duty to (a) determine whether Wikipedia should have an article about a subject with borderline notability (e.g., a new musician, a smaller business, a minor politician) or (b) prevent readers from finding ugly new articles (e.g., articles that are badly formatted, uncited, poorly translated, or otherwise feel "low-quality" to some individuals).  If there isn't a reason to delete it today, then it's marked as reviewed today.
 * One-stop shopping: NPP does it all.  NPP will do its traditional tasks of CSD-worthy problems and alerting editors to problems, but it will also take responsibility for determining the likely notability of all subjects, making sure that notability is demonstrated by an adequate, and protecting readers from "low-quality" articles.  They will only accept articles that they believe is about a notable subject and whose current form is (in the reviewer's opinion) reasonably decent.  Articles that have problems that aren't deletion-worthy will get tagged and draftified, but they will also be kept in the queue for future re-review.  "Mark as reviewed" means that they endorse the article.
 * Does that sound about right? WhatamIdoing (talk) 18:09, 30 March 2024 (UTC)
 * WhatamIdoing, yes, that's about it. Those questions arise in the following context:We don't actually need NPP to perform any of the individual steps, except for marking reviewed. So, the purpose of NPP is to make sure at least one experienced eye looks at every single article that gets added. There is no guarantee that any patroller, NPP or any other, will catch what needs catching. Copyvio, for example, is something we can't be sure we've caught all even if multiple experienced editors check. In fact, we know we're not catching copyvios like we ought to if everyone had the competence and diligence for every step. Physics, Mathematics or Medicine articles, many reviewers will look at and still not know whether they're looking at utter garbage or a high quality article. Instead of having a workflow for passing those articles to WikiProjects, we have one where we do nothing hoping someone competent will eventually find it if it remains in the queue long enough. All borderline articles are not the same. While I don't think sporting season articles or articles of the type "X sport in Y event" should be making up the backlog instead of approved to be handled by normal editing processes, I think NPP had better handle living persons, organisation and product articles. I know we've got inclusionists so extreme that they'll argue even blatant paid editing is fine as long as it adds content to Wikipedia, but we really can't afford to let spamming lose. Many borderline articles are also borderline on promotion, UPE or COI. The amount of work NPP along with non-hatted spam fighters and sock hunters put in is precisely the reason that community at large gets the sense that COI/UPE isn't a big problem. If we stopped that, it would be. I think that's where the biggest opposition is going to come from. On the other hand, if we started carving out exceptions, WP:CREEP opposes may dominate. — Usedtobecool ☎️ 12:33, 31 March 2024 (UTC)
 * If we had an RFC with those two options (in more polished language), do you think you could easily pick one or the other? If the RFC ended with a clear consensus for your side, would you be satisfied with that outcome, or would you be wishing that there had been a third option in the mix? WhatamIdoing (talk) 05:43, 1 April 2024 (UTC)
 * I've think you've summed up the two NPP philosophies perfectly there. It would be interesting to pose the question at such a high level. –&#8239;Joe (talk) 11:29, 1 April 2024 (UTC)
 * I think those are the right questions to be asking. For me, back when I was doing NPP, the answer would have dictated how much I did rather than if I did it at all. For me, as someone who doesn't do as much anymore, I can say that if the "One-stop shopping" became the norm it would make the barrier to my re-entry higher because of the commitment that entails. But if that's what the community wants that's what it should get. Best, Barkeep49 (talk) 15:33, 1 April 2024 (UTC)
 * With minor adjustments, I would support the first one. With basis in community consensus (there apparently isn't one currently, either way, and has caused some friction), we would assure reviewers they would not get into trouble for passing articles that could not be deleted without an AFD. We still tell them that we should try and remove spam best we can, via AFD if necessary, but that we need not waste our time and energy looking for sources for articles that can neither be CSDed/draftified nor could any party conceivably be paying money to add to Wikipedia. The concern would first be our core content policies, and notability guidelines, second. But that'd make it easier for sloppy, high-volume article creators, so we would have to start doing something about editors who don't listen to feedback. I do not know if we'd find consensus for that. Significant chunk of the community would pick having one million more poorly written articles over 200K well-written ones, I think. — Usedtobecool ☎️ 09:13, 4 April 2024 (UTC)
 * I have been wondering whether it would be good to add something to the first about NPPers also being editors, and as such being permitted to do anything any other editor would like to do – but that normal editing is not technically part of the NPP process. For example:
 * If you 'mark as reviewed' a copyvio article, that's an NPP problem. If you make a habit of it, we'll remove the NPP rights.
 * If you 'mark as reviewed' an article with spelling errors, that's not an NPP problem. If you make a habit of it, an individual or two might grumble about "Wikipedia's reputation" and their preference for gatekeeping, but you'll still be welcome to do NPP work.
 * If you fix the spelling errors and 'mark as reviewed', then fixing the spelling errors is the kind of normal collegial editing that we want to see all good editors doing, and only marking the page as reviewed is an NPP action.
 * Do you think that would help? IMO we don't want to discourage NPPers from improving new articles; the first version is supposed to mean that marking an article as reviewed doesn't need to wait for fixable problems to be fixed. WhatamIdoing (talk) 20:17, 4 April 2024 (UTC)