Wikipedia talk:Notability (geographic features)

Discussion of proposed RFC questions

This is essentially the nine proposals that were made in the workshop, with HEB's proposal struck, and the two proposals Giulio wanted to withdraw withdrawn. FOARP (talk) 10:22, 31 March 2026 (UTC)

I like most of these proposals, I particularly like Proposal 3. Great work to everyone in the workshop!

My main thoughts are on Proposal 2:

I'm not sure if people will object to the line beeing drawing between hamlets and villages, but I do agree that essentially village or village equivalents should be the extent to GEOLAND. I'm just imagining some technicality in some country's system regarding how settlements are named in sources (exceptions to this probably would fall under GNG). I think the administrative areas should also, have a similar "size" caveat. In Myanmar, for example, there are technically some "100 household" service management units within the actual smallest real unit of local governance.

For Proposals 5 and 6, I am similarly unsure of "upper-level" in the qualifier for administrative division. To me, that would read as excluding second-level administrative divisions or below. I agree that things like irrigation districts should not be notable, but I would prefer it to be using something like Proposal 2's definition with public services or a more clear wording on what is "upper-level" EmeraldRange (talk/contribs) 12:27, 31 March 2026 (UTC)

A suggestion ... upper-level, legally recognized administrative divisions "such as Cities, towns, and villages," possibly? - Enos733 (talk) 23:10, 1 April 2026 (UTC)

To me, villages are very much not examples of "upper level administrative division", indeed to the extent they have any administrative function they are about lowest level. Thryduulf (talk) 23:44, 1 April 2026 (UTC)

Where I am at, other than title, there is no legal distinction between a city and a town or village. - Enos733 (talk) 00:33, 2 April 2026 (UTC)

I'll restate my support for this moving forward here. I think these proposals are an improvement on the current phrasing of the guideline and hopefully we can clean this up. Giulio 13:11, 31 March 2026 (UTC)

Especially (but not exclusively) if lines are drawn on the basis on nomenclature, it would probably be useful to have a page describing the terms used for different settlement types in each country and what they typically look like on the ground. The aim would be purely informational so that when people are talking about an e.g. village in e.g. Myanmar they are all talking about the same thing. The page would explicitly be independent of any notability (or other) guidelines (although such could reference this page) and be purely descriptive, with explicit notes that it described typical ranges/examples and there are exceptions. Thryduulf (talk) 13:17, 31 March 2026 (UTC)

Looks promising. Donald Albury 13:30, 31 March 2026 (UTC)

• Looking at it, I think we can merge Prop 1 and Prop 2. That would read:

This would bring the number of proposals down to five which I think would be welcome (particularly to any closer!). Prop 3 and Prop 6 probably need some word-smithing to turn them into actual proposals for implementation, though I think the intent of the proposals is clear. FOARP (talk) 13:50, 31 March 2026 (UTC)

I think that is the easiest proposal to understand. Legally recognised just doesn't cut it without a huge checklist for each country, and even then some dont even have them! Davidstewartharvey (talk) 14:34, 31 March 2026 (UTC)

1+2 is the best, no need to argue over the definition of one, you must provide an RS to do so. And if a place has no RS confirming such a detail, well can it be notable then?. Although: 'Subdivisions, housing developments' should say this 'and' or is there something else that is meant to be included?

'Former villages that are now suburbs of Cities or towns do not lose their notability but show their former status by reliable sources.' should probably say 'former villages and towns' and decapitalise city. Traumnovelle (talk) 19:27, 31 March 2026 (UTC)

done.

Additionally I've tried to rewrite Prop 3 and Prop 6 as proposals for implementation per the above. FOARP (talk) 07:55, 1 April 2026 (UTC)

re Proposal 6 - when I first proposed this (see Workshop: Proposal 9) it was as an alternative opening to the original Proposal 7, now renumbered and reduced/rewritten as Proposal 5. FOARP's rewritten Propl 6, although it could stand, of course, as a separate proposal in its own right, does not really reflect my intention, which is to include an express option to define, country by country, the units that can be presumed notable. This is because there are so many variables that no single definition is going to cover them adequately, and because it may reduce otherwise endless further discussion. Ingratis (talk) 07:20, 26 April 2026 (UTC)

Proposal 6: Populated settlements constituting cities, towns and villages (including abandoned or depopulated settlements) as further defined in [Document A] and legally established administrative divisions, also as further defined in [Document A], are presumed to be notable.

"Range of services" probably needs clarification / less vagueness. Otherwise it could be argued as SNG green-lighting abstract sets of lines on a map. One could say an irrigation district or a library district provides a range of services. Maybe "broad range of services" or "diverse range of services". North8000 (talk) 17:07, 1 April 2026 (UTC)

"Range of services" could probably benefit from some examples. A very small town may not be able to do much, but in the US, I'd expect them to repair the streets and make some land-use decisions (Zoning), and also to have a government entity that can pass a Local ordinance (e.g., setting the speed limit or declaring today to be Smallville Day). In much of the US, water and schools are handled by separate agencies, and emergency services are often outsourced to the county.

Alternatively, it might be possible to say something like "This should be interpreted as including places such as Smallville in England and Ruralton in Newfoundland." WhatamIdoing (talk) 00:55, 2 April 2026 (UTC)

I'd say adding the word "broad" does the least injury to the origin intent of the drafter here. FOARP (talk) 07:50, 2 April 2026 (UTC)

• I like proposal 5 as a concept, but as it reads currently, it feels more like an essay (such as in WP:OUTCOMES). Is there a way to simplify it. --Enos733 (talk) 23:02, 1 April 2026 (UTC)

  Ignore this - The proposal rewrites all of WP:GEOLAND, rather than amends any one section. I do still think it reads more of an essay. - Enos733 (talk) 23:06, 1 April 2026 (UTC)

  I'm not sure how to redraft without hurting the original intent. Perhaps you can make some suggestions as to what it is you want to change? FOARP (talk) 08:30, 2 April 2026 (UTC)

  I could probably go something like:

  1. Legally Recognized Entities Populated, legally recognized settlements and primary administrative divisions, specifically counties, municipalities, and other principal sub-national jurisdictions, are likely to be notable regardless of their historical status or current population size. Settlements must be verifiable. Current entities must possess formal legal status within their respective nation. Other recognized entities or settlements, such as census tracts, census-designated places, or special district governments, do not meet this specific criteria and must establish notability in accordance with WP:GNG.

  2. Unrecognized and Informal Places Populated places lacking formal legal recognition require evidence of notability in accordance with WP:GNG. This category includes residential developments, commercial parks, and local neighbourhoods. If sources are insufficient for a stand-alone article, the information is merged into the article for the encompassing legally recognized entity or administrative subdivision. The presence of individual notable facilities like schools or post offices does not make a settlement notable (see WP:INHERENT).

  3. Disputed Regions and Mass Creation Notability for disputed regions is determined based on the depth of source coverage independent of the validity of territorial claims. Content regarding disputed areas or competing names is merged into broader political or conflict articles when a stand-alone page is not supported. Geography articles are not to be bulk created without prior community consensus. -Enos733 (talk) 21:17, 3 April 2026 (UTC)

  In general, I like that, but there is that pesky "legally recognized" thing. Where I live we have things like "soil and water conservation districts" and port or inlet districts that have publicly elected governing bodies and taxing powers, which I do not believe should be exempted from the GNG. The port and inlet districts are generally not a problem as the ports and inlets themselves usually meet the GNG and the taxing districts can be mentioned in the articles about the port or inlet. I really do wonder if we need this at all. The justification for having an SNG for populated places is to predict which places are likely to meet the GNG. Why not just fall back to using the GNG? Donald Albury 14:56, 4 April 2026 (UTC)

  @Donald Albury, a couple of things. First, I read my proposed rewrite as establishing a general principle and then being clear(er) about how it should be defined (in this case the limiting factor s "other recognized entities....do not meet this specific criteria"). The second thing is that I don't necessarily share the idea that WP:GEOLAND is as broken as other editors do. The challenge has been mass creation of entities that are virtually nothing but a census-type listing. - Enos733 (talk) 15:31, 6 April 2026 (UTC)

  As I said, I do not believe that "legally recognized" is a useful criteria. I don't think that we can come up with any criteria that we could apply even-handedly to populated places in every country. We could go with something like the country-specific guidelines at Wikipedia:Naming conventions (geographic names)#Region-specific guidance or, my preference, just rely on the GNG. Donald Albury 17:23, 6 April 2026 (UTC)

  I just took a look at the RFC that got us here. The first response was that the problem is mass creation. I think we can solve that by being specific with our examples in the prose and by specifically frowning on mass creation. The second question is about what legal recognized means. While not completely the same, an editor did create WP:NSUBPOL to provide clarity of which subnational politicians do and do not meet NPOL. I think a similar effort can be made here to provide the necessary country by country guidance. - Enos733 (talk) 22:55, 6 April 2026 (UTC)

• As someone who mostly writes about places that no longer exist, a "legal recognition" criterion is a non-starter. Such information about an extinct settlement will hardly ever be available. Zero^talk 08:48, 2 April 2026 (UTC)

  Possibly you could expand on this a bit. "legal recognition" is in the present WP:GEOLAND standard, some of the proposals for change include it, some don't. In that context, what is it you want to see? FOARP (talk) 09:57, 2 April 2026 (UTC)

  Per the above comment, it would likely be helpful if the RfC quoted the present standard before the proposals to change it and also included an explicit option to retain the status quo. Thryduulf (talk) 10:01, 2 April 2026 (UTC)

  That's prop 4. The proposals are ordered in the order they were proposed in - i.e., neutrally. FOARP (talk) 10:09, 2 April 2026 (UTC)

  While numbering the proposals for change in chronological order makes sense, I don't think including the status quo in the middle of the list of proposals that change the status quo is helpful, ideally it should be first or last but wherever it is it should be quoted. Thryduulf (talk) 10:17, 2 April 2026 (UTC)

  Anywhere we chose to put it - especially first or last - is going to draw accusations of bias. In this way we avoid that. People can read all the proposals before !voting. Crazy, I know. FOARP (talk) 11:39, 2 April 2026 (UTC)

  I'd actually say it was more biased to position the status quo as if it were a proposal for change when it, by definition, is not. People need to be able to easily compare proposed changes with the status quo in order to form a meaningful and relevant opinion on the change, the unbiased way to do that is to clearly present the status quo as the status quo in an easily referenced position. Thryduulf (talk) 11:44, 2 April 2026 (UTC)

  I'm sure at least a link to WP:GEOLAND will be provided in any preamble to the discussion. FOARP (talk) 12:22, 2 April 2026 (UTC)

Further to the exchange above, my house is in about 20 different legally recognized places. 4 of them (Town, County, State, Country) are what an average person would consider a "place". The other 16 are legally defined zones defined by a set of lines on a map, each with some specialized purpose, sometimes with a specialized set of services. For example a school district, a park district, a fire protection district, a forest preserve district, a precinct etc.. Regarding the "range of services" SNG criteria, one could claim that the forest preserve district offers a range of services. But a town (even a small one) or county offers a diverse range of types of functions and services. North8000 (talk) 19:26, 2 April 2026 (UTC)

For that matter, many homeowner associations have legal recognition and provide some range of services (e.g., snow removal, landscaping), but by and large these are not considered as places under GEOLAND. Similar for condo coops. By itself, provision of services is not a distinguishing factor. older ≠ wiser 20:03, 2 April 2026 (UTC)

I don't know how to put it into words, but what I think that what that specific criteria is reaching for is that it is governmental, and in the primary hierarchy. E.G in the US that would be city/town/village, county, State, Country (and in some exceptional cases townships). And, amongst governmental areas, these have a higher diversity of functions that those more abstract sets of lines on a map. North8000 (talk) 20:31, 2 April 2026 (UTC)

I'm going to guess that few of those are legally defined and that most of them are administered under laws which allow whatever agency to administer them, or even establish them in the first place. That's the issue with census CDPs: they are defined by the census strictly for counting, and have no authority behind them. They only exist so that statistics can be produced for places that don't have defined boundaries, legally or otherwise. Mangoe (talk) 19:36, 2 May 2026 (UTC)

• Say a village existed in the 18th century and ceased to exist in the 19th century. It wasn't famous for anything in particular but there is enough known about it that an article of a few sentences and a couple of sources could be written. How are we to interpret the "legally recognized" criterion? The government isn't around any more and we don't know how they saw this village. (This is a common situation in historical work.) Why should that matter? Why should a government, even an extinct one, decide what is notable for us anyway? I'd like to see the "legally recognized" thing replaced altogether by a guideline that places with a distinct identity and name, having a continuing population for an extended period of time, qualify without the need for government approval. Zero^talk 14:08, 5 April 2026 (UTC)

  I'm not certain that "extended period of time" is always required either. For example some ghost towns and construction camps are notable despite only existing for a few years. Thryduulf (talk) 15:57, 5 April 2026 (UTC)

  But isn't the whole point of this exercise to determine how to characterize types of populated places that can be presumed to be notable and bypass GNG? For any places such as archeological sites, ghost towns, housing developments that can pass GNG, this guideline is not relevant. older ≠ wiser 16:21, 5 April 2026 (UTC)

Not implying an answer to your broader question. But remember that the SNG criteria is just an alternate to GNG. I think that the situation that you described would meet GNG as typically applied to GEO, especially if you created a real (albeit small) article as you describe vs. a stub.

I agree that "legally recognized" is not a very good criteria. It's so vague that has little correlation with wp:notability and excludes some wp:notable ones and can be interpreted to include lots of non-notable ones. North8000 (talk) 21:25, 5 April 2026 (UTC)

I hate to say it but to have all of the effort that has been put into this (sorry I missed that it was even happening) come to fruition this is going to need more work. More work is a burden, but not as much of a burden as having all of the work that has been done on this go to waste. Also having 3 or more mutually exclusive choices means that mathematically any one of them is unlikely to get the super majority needed to be accepted as a consensus. (yeah, I know it's not a vote). My suggestion is to get the folks who have been actively involved on this come up with one version / proposal that is most aligned with the group's thoughts. And make that the proposed change. And then the group that developed it should all support it as their product even if it wasn't exactly what each individual preferred. Then a well-written, clear concise RFC. And the proponents should also be ready to put in a well-written rationale for the changes in as early responses. (but not in the RFC itself which needs to be neutral.) Sincerely, North8000 (talk) 20:44, 6 April 2026 (UTC)

This is modelled on WP:NSPORTS2022, where a change did ultimately come about because support rallied around one option. We've had numerous "just try one option" proposals in the past that never got anywhere. FOARP (talk) 07:43, 7 April 2026 (UTC)

I gave several different suggestions in my post. Another way to approach the math problem is to ask respondents to respond on every alternative. Sincerely, North8000 (talk) 20:46, 8 April 2026 (UTC)

IDK, I get the feeling that not enough people want this to happen. I'm not going to bother with something that for sure is going to be a lot of trouble if not enough people want it to happen. Am I wrong? FOARP (talk) 09:37, 27 April 2026 (UTC)

I don't think you're wrong. I'm not seeing much enthusiasm for changing the status quo from, and certainly no general enthusiasm for a specific change. Thryduulf (talk) 19:56, 27 April 2026 (UTC)

I personally don't have the time now to participate, sorry. I would like this to change very much, as when I get started again on reviewing US articles it's going to continue to affect my submissions. I would suggest one thing: the biggest continuing dispute I see is what constitutes legal recognition. I feel I must be very hard-nosed about this: postal service naming is not law, and census counts are not law, and other administrative actions do not create law as far as this is concerned. But I'm sure I'm going to be shouted down about that, and I really, as I said, don't have the time to mount an extended defense of that. Mangoe (talk) 02:33, 28 April 2026 (UTC)

Totally agree. However, I feel that there is too many editors who do not want to change the Status Quo, even if it is clearly impossible to manage legally recognised, because they don't like change. The same attitude is being seen at the RFC on the renaming of AFD - the process has changed to include the merger process but most don't think the name should change as it will confuse new users! Davidstewartharvey (talk) 05:16, 28 April 2026 (UTC)

Yes, I'm all for clarifying the vagueness, but discussions to change the status quo are an exercise in banging one's head against the wall when there is a significant number of editors for whom a bare mention of a population and location in any sort of quasi-governmental database or publication is sufficient for an article. older ≠ wiser 10:55, 28 April 2026 (UTC)

Sigh! I have to agree with that. I really wish we could tighten up the minimum standards for articles about populated places, but then, I seem to be out step with much of the community about a lot of things. Donald Albury 15:26, 28 April 2026 (UTC)

The Nikolai Kurbatov case (see below) really is a quite convincing argument that GEOLAND needs to change. It's yet another mass-creating case caused by GEOLAND. Indeed, its an example of exactly the thing that some people have been claiming isn't or even can't happen. Having thought about it, I think we still need to go ahead with something here. FOARP (talk) 20:55, 29 April 2026 (UTC)

Yet again, that's a mass creation problem not a GEOLAND problem (or a sportsperson notability problem, or a redirect problem, or a dictionary words problem, etc). I'm not saying there isn't a problem, I'm saying that it's a different problem. Changing GEOLAND will not (because it cannot) fix a mass creation problem. Thryduulf (talk) 22:36, 29 April 2026 (UTC)

Not to rehash the massive amount of discussion here, but it cannot be merely a coincidence that mass-creators have chosen the topic-areas where we have ridiculously low standards for inclusion in which to mass-create. It can also be seen that what was done in NSPORTS2022 did put a stop to mass-creation in that area. FOARP (talk) 07:39, 30 April 2026 (UTC)

It is a GEOLAND problem because GEOLAND is what makes these mass-creation cases so hard to clean up. Take for example this discussion Wikipedia:Articles for deletion/Sandogo.

A user mass-created about 80 stubs on Burkina Faso villages using bad sources, with incorrect coordinates, outdated names sometimes duplicating existing pages, things being listed as villages despite not being villages etc. (I have to add that currently there are articles on ~25% of the villages of Burkina Faso, mostly mass-created about 20 years ago but at least they were using a good source so I've been gradually working on improving them). The end result was keep on the basis of GEOLAND because users found that some of them were real places even though all the pages had a number of inaccuracies or outright mistakes and bad sources. Of course no-one who voted to keep on that basis then bothered to systematically go through the articles and sort them out and the creator of the articles refused to acknowledge there was any problem with them and stopped editing soon after, so the end result was that all those error ridden articles stayed up for years until I stumbled on the mess, salvaged the pages I could and sent the rest through AfD.

So because the current version of GEOLAND prevented mass-deletion of a set of flawed mass-created articles, mis-information stayed up for years. And this was only a small and fairly recent case, this isn't even scratching the surface on the tens of thousands of such articles created particularly between 2006 and 2010 that remain online and unaddressed because it is such a slow and tedious task.

You are welcome to propose a change to the policy on mass-creation if you think it can be fixed that way, but given that there are already supposed to be limits on mass-creation and this keeps happening and we don't have a good way to deal with it then clearly something needs to change. Giulio 11:40, 30 April 2026 (UTC)

Agree with everything said here. Also have to point out that when in the past we discussed changing our guidance around mass-creation, typically the same people who oppose doing anything about notability issues also oppose doing anything about mass-creation. The allegations of bad faith and "hounding" the people who mass-created tens of thousands of PAG-failing articles onto the encyclopaedia are rarely far behind. FOARP (talk) 14:31, 30 April 2026 (UTC)

Yeah, no, there's not really a difference between the mass creation issue and the GEOLAND issue; mass creation happens because people look at the guideline and see that it implies that any "official" list of places is "legal recognition" therefore that they can go ahead and make a settlement article for each entry. These "don't have to satisfy GNG" subject guidelines are all problems in this wise because they are interpreted to mean "I don't have to have enough info to write more than a stub," but this one of the worst offenders and the level of WP:IDIDNTHEARTHAT in the face of many times this has come up is really disheartening. Mangoe (talk) 15:54, 2 May 2026 (UTC)

Yes, achieving any sort of meaningful change in situations like this requires a Herculean effort and the patience of Job. older ≠ wiser 16:25, 2 May 2026 (UTC)

100% agree @Mangoe. It is *ASTOUNDING* to have been told over and over in last year's discussion that this wasn't an issue, that we were parading "straw goats", that any example was essentially cherry-picked, that the whole thing was an issue of the past and that no-one was doing it any more, that we were simply making it up when we said that GEOLAND was qualifying single buildings as "legally-recognised populated places" and that "only advanced by people seeking to discredit" the policy were saying this.... and then to have a case that demonstrates the validity of everything we said happen again.

It's very disheartening as well. You know that any attempt to do anything about this is likely to lead to a lot of heat. You know that there's going to be people trying to shut it down and throwing around all sorts of accusations. I have to be honest and say I was tempted just to let the whole thing lapse because that would be easier.

But the Russian selo case really clarifies that we have to do something here because even the idea of slowly working on the existing articles - a task that would take decades - is a non-starter whilst these articles continue to be created at a rate far faster than any clean-up effort can deal with them. FOARP (talk) 18:52, 2 May 2026 (UTC)

Mass creation is a behavioral problem, and needs behavioral solutions rather than notability solutions. WhatamIdoing (talk) 19:39, 2 May 2026 (UTC)

Notability guides behavior; that is its only purpose, as is, for that matter, anything else we discuss about writing and editing an encyclopedia. This isn't even a distinction without a difference. Mangoe (talk) 02:24, 3 May 2026 (UTC)

The distinction is this: When the problem is notability, we address it with the Wikipedia:Deletion policy (including Wikipedia:Deletion policy#Alternatives to deletion). When the problem is behavioral, we address it with the Wikipedia:Blocking policy.

Compare:

• Nobody is allowed to create an article on _____ because Wikipedia does not want an article about it.
• This editor specifically is not allowed to create any more articles about _____ because this editor specifically has a history of screwing up in this area (e.g., struggling to distinguish between notable and non-notable topics due to personally misunderstanding certain sources).

WhatamIdoing (talk) 03:39, 3 May 2026 (UTC)

1) OK, so the problem is behaviour. Then you’re ok with dealing with the product of that behaviour? That’s where this always falls down: the insistence that mass-created articles should be dealt with as if they were special, individually-crafted snow-flakes, rather than as production-line articles, results in them being impossible to deal with.

2) Acting as if the guideline isn’t causing the behaviour is just denying the facts. The mass-creators in this field all, when challenged, cite the guideline as a no-further-explanation-needed justification for what they are doing. It is not simply a coincidence that when people want to turn out a large amount of articles with minimum effort (say, to be the person who created WP’s 7 millionth article) they choose this topic-area to do it in.

3) Even just focusing on behaviour, I’m sorry but do you want to be the one to take the latest mass-creator in this field to ANI? Knowing that they admit struggling with a whole range of mental illnesses? I’m an admin so I suppose it’s my job, but I honestly don’t want to, and find it hard to blame them specifically for just following what to them (and numerous other people) appears a straight-forward interpretation of the rules. After being involved in dealing with C46 and Lugnuts I would rather just change the rules to prevent this from happening again. FOARP (talk) 07:21, 3 May 2026 (UTC)

Yeah, from my perspective the behavioral issue here is that each time this issue comes up, which in reality it has been coming up for a decade now at least, the response is invariably to wring some hands, declare that there's nothing wrong with the current text of the guideline by alternately pettifogging any attempt to alter it and then simply saying there's nothing wrong; and when a new mass creation is discovered, no matter how recently it happened, the response invariably is to ignore that we agreed that we were not going to permit it, and to force yet another article-by-article review as the only recourse. And to put that in perspective: we never finished dealing with Somalia, where there were maybe a few hundred articles to review. We never finished with California. I think we did completely go over Arizona, and I made at least a first pass over North Dakota and I think Montana and Idaho. I'm not sure if I ever finished Utah. Most recently I went over every "unincorporated community" in Indiana, and it took well over two years to finish, with some 450 nominations; in that time I estimate I looked at somewhere between 3000 and 4000 Indiana articles. Of the various people who worked on these in the past, I am the only one left doing it, and at the moment Real Life and a degree of burnout has brought the effort to a halt in Illinois, where I expect to have to look at as many articles as for Indiana. So the solution is, by intransigence and a great deal of WP:IDIDNTHEARTHAT, to stymie all attempts to deal with the issue by making a hopelessly long task the only permitted approach, and at that, it keeps getting longer because in practice more mass creations are allowed to stand. And it also bugs the hell out of me that the people who keep showing up to thwart doing anything about the issue never participate in those AfD discussions, even to try to save articles. And the most fundamental point of all, that a large percentage of these mass-created articles are inaccurate due to misinterpretation of the sources or because they aren't reliable is always blown off, even though I bring it up every time this subject comes around again. But the guideline, as it is now worded, encourages people to trust these sources and assume that they are authoritative because they assume that every listing from a governmental source constitutes "legal recognition"; and when they do it, the same group of people always step in to prevent anything being done about it. That's the behavioral problem here. Mangoe (talk) 11:42, 3 May 2026 (UTC)

Why are you "nominating" these instead of "boldly merging" them (e.g., to a list inside a county article)? WhatamIdoing (talk) 22:20, 3 May 2026 (UTC)

I vaguely recollect that in one case in the recent past, the mass-creating editor has simply reverted the bold merges and carried on. Then we just get back to the current topic of discussion ("legally recognized" can be interpreted in an extremely open-ended fashion). Choess (talk) 22:53, 3 May 2026 (UTC)

1. IMO step one is to stop the behavior; step two is to see whether the editor is willing to help clean up the mess. This works best when we intervene early.
2. I agree that improvements to this guideline could reduce the undesirable behavior. I don't think that "change" the guideline will be as useful as "clarifying" it by providing examples ("Laketon has had a population of zero since it was flooded to create the water reservoir in 1974, but it still counts as a formerly populated place because it was a proper functioning city for 167 years before then") and information ("Do not use any of the following databases, because they're census tracts instead of actual populated places, unreliable, or otherwise prone to misinterpretation:"). I could also see some value in saying "If you plan to create more than 25–50 per day, then you must have WP:MASSCREATE permission; if you plan to create more than 100 from the same census or database source, then you should first present your sources at the ____ page and ask editors to confirm that it's considered reliable for this purpose."
3. I'm willing to address editors who are engaging in inappropriate mass creation, but as a practical matter, I don't patrol new articles, so my willingness to help probably wouldn't actually make a difference.

WhatamIdoing (talk) 22:19, 3 May 2026 (UTC)

@Mangoe, I'm looking at your comment that These "don't have to satisfy GNG" subject guidelines are all problems in this wise because they are interpreted to mean "I don't have to have enough info to write more than a stub".

I'm concerned about people conflating article content with subject notability. The GNG does not, and has never, required anyone to write any amount of article content. The GNG doesn't officially even require that it even be possible to exceed stub length. Meeting the GNG's SIGCOV requirement (assuming editors could agree on what that means...) sometimes results in a good infobox and a few sentences, which is stub-length. The GNG also doesn't, and has never, required anyone to actually cite sources or "demonstrate" that the subject meets the GNG. I would not want anything in this guideline to suggest that article content determines the subject's notability. WhatamIdoing (talk) 19:38, 2 May 2026 (UTC)

If there is significant coverage, then as a rule it is almost always the case that that it would eventually be possible to write more than a stub. But fair enough. We're still left with the same problem: that the "legal recognition" clause, because nobody really agrees on its meaning, is routinely interpreted as saying that any listing of of place names of any sort from any governmental source (a) is legal recognition, and (b) is to be presumed accurate as interpreted by someone creating articles for each entry in it. The second clause, experience shows, is false: some of them are accurate as far as they go (e.g. census counts), but many of them are error-ridden (e.g. GNIS, as we have documented in WP:GNIS, and GNS/GEONAMES has the same problems except far worse) and a lot of them do not mean what people assume they mean (e.g. the Iranian abadis and the Philippine baragays; it looks as the the Russian/Soviet sources have similar problems). And the first clause isn't true either, as I've pointed out over and over about US CDPs.

And I'd be more willing to worry about such "conflation" if the solution weren't always to ignore all the people who have complained for well over a decade about dealing with these articles and the arguments about the guideline that show up over and over and over and over and over in AfD discussion. Every little quibble always seems to revert very quickly into "oh, we have to stick with the status quo". But this has come up time and again; the very first doubt about legal recognition shows up in Archive 2 from 2013, before this was even elevated to a guideline. Mangoe (talk) 03:06, 3 May 2026 (UTC)

The problem of "nobody really agrees on its meaning" is why I think we need a country-by-country list of sources, the many and varied ways people have misunderstood them in the past, and lots of examples. For example, w:ru:Село suggests to me that a selo is probably a valid "legally recognized" populated place, but that the editor writing such an article needs to do significantly more work than just "_____ is a selo (village) in Russia", because the factual statement might well be "_____ was designated a selo (a word for a larger village, with a church that rang a bell) in pre-Soviet Russia; its boundaries differ from the modern rural census tract named after it". WhatamIdoing (talk) 03:54, 3 May 2026 (UTC)

Maybe there is scope to expand Wikipedia:Geographic references to do this. List of national and international statistical services would probably be a good thing to weave into it. Cheers, SunloungerFrog (talk) 06:09, 3 May 2026 (UTC)

@WhatamIdoing - the issue with that idea of what a Selo is, is it wasn’t what was applied later by Soviet and Russian authorities. We’ve had Russian Selo with populations of 0, 1, or 2, and no evidence that more people ever lived there. Even Abadi has, as one of its possible translations, “village”, it just wasn’t used by Iranian authorities in that way. FOARP (talk) 06:44, 3 May 2026 (UTC)

Yes, that's my point: "the editor writing such an article needs to do significantly more work than just "_____ is a selo (village) in Russia", because the factual statement might well be" one (notable) thing and it might well be another (non-notable) thing, and there is no way to tell the difference from certain sources. WhatamIdoing (talk) 20:41, 3 May 2026 (UTC)

I think it's worth trying, especially with the mass creation case below. Something needs to be done. BilledMammal (talk) 07:52, 30 April 2026 (UTC)

Right. This is EXACTLY what some people kept saying wasn't an issue, or wasn't happening, or wasn't happening any more, or couldn't happen, or we were just making up. *EXACTLY*.

These are mass-created articles, created because they give a notional pass of the WP:GEOLAND standard due to local law (in this case Russian) granting "legal recognition", but which in reality has led to the mass-creation of thousands of stub articles about single buildings with a population of 1 or 2. This is yet another case of this, after the Californian, Iranian, Azeri, Emirati, Burmese, Sri Lankan, Filipino, Turkish, Polish, and Belarusian cases.

This is not just us making things up, or trying to carry on a grudge against an already-indef'd editor, or being crazy deletionists. This is an actual real problem. FOARP (talk) 10:06, 30 April 2026 (UTC)

FYI, I should also make you aware of Central African Republic to add to that list, I already sent the most egregious examples through AfD (Wikipedia:Articles for deletion/Abou-Ndoulaf) however there are plenty more articles that were part of the same mass-creation effort that had coordinates pointing vaguely near something that looks like a village so someone would likely need to go through them more thoroughly for any AfD nomination to be successful, its on my to do list, but behind another of other things and I'm taking a bit of a break from geographical article right now. Giulio 13:48, 3 May 2026 (UTC)

I've sort of been mulling on the options here and it seems to me that there are broadly four choices:

1. The status quo
2. Remove the SNG totally, use GNG only
3. Some revised words that are stricter in some way about the definition of "legally recognised", maybe with per-country subpages that set out in detail how that works on a country by country basis
4. In addition to "legally recognised and populated", require at least one reliable source with significant coverage to be present in the article, as per WP:SPORTSIGCOV.

Of those, I imagine that few people will like #2, though it should solve mass creation. #3 would be a lot of work to get in place. #4 should also solve mass creation, and is somewhat simple to understand, as it only adds one thing to the existing SNG, and we can point out the parallel to SPORTSIGCOV. Just my two centimes, anyway. Cheers, SunloungerFrog (talk) 14:56, 30 April 2026 (UTC)

Your #2 would be my preference. #4 would be a minimally acceptable alternative. #3 is impractical as there is such variability around the world it would basically be moving the 'vagueness' from a global problem to a country/regional problem. It is extremely unlikely such specific guidelines would be forthcoming on a wide scale anytime soon and without defining any sort of default, this would essentially promulgate the status quo (or worse, when one country/region decides an article on every verifiable cluster of houses is acceptable, other counties/regions will want to emulate that). older ≠ wiser 15:12, 30 April 2026 (UTC)

FWIW, a possible compromise might be to set #2 as the default with some provision to allow countries/regions to explicitly define what types of localities can be presumed notable for that area (which would be subject to community review before implementation). older ≠ wiser 15:23, 30 April 2026 (UTC)

As the proposed of 2, I would of course say yay. The issue is administrative areas and how you word it. Davidstewartharvey (talk) 15:39, 30 April 2026 (UTC)

I am not a geographer, so I don't know, but is there a formal geographical word or phrase that encompasses administrative areas and likewise one for cities/towns/villages, so that we can use those formal phrases in the wording? Cheers, SunloungerFrog (talk) 16:23, 30 April 2026 (UTC)

Maybe "Administrative units" or "Special-purpose districts" for the further and "settlements" for the latter? And we are clear that any SNG pertains only to the latter? Cheers, SunloungerFrog (talk) 16:49, 30 April 2026 (UTC)

The issue is the differences in different countries. In the UK, we have Parish, District/Borough, Unitary, County councils, plus national parliaments (Scotland, Wales, Northern Island). They manage different parts of local services. However, take the US, schools are managed by the local school boards, which in the UK is the councils responsibility. In India we have Community Development blocks, which manage planning and development which in the UK is the councils responsibility. One of the points put previously is that the administrative areas provide a variety of services. Davidstewartharvey (talk) 17:22, 30 April 2026 (UTC)

I also prefer #2, with #4 at a minimum if #2 does not reach consensus. Defining "legally recognized" for every country in an even-handed way would be a Herculean task, and one I do not think would be worth the effort. Donald Albury 18:32, 30 April 2026 (UTC)

I also prefer #2 with #4 as a good alternative. Does anyone have an idea how much more empirically restrictive #2 is than #4? If #4 gets most of the job done then might be worth going with as the proposal. A binary choice RFC (adopt proposal or keep status quo) may be best to avoid !vote splitting. Agree that something needs to be done here. NicheSports (talk) 05:43, 1 May 2026 (UTC)

I don't think vote-splitting is a concern if the proposals are separate, support/oppose polls in an over-arching survey. People can support both or oppose both, or !vote only in one poll and not in the other in that context. That, anyway, was how WP:NSPORTS2022 worked out.

Country-by-country guidance is something that can be an option in any event, even with keeping the present GEOLAND standard - it would just get a little bit pointless if we switch to GNG though not entirely so - so should be kept detached from any other change.

I think proposals around requiring significant coverage need to engage with the issue of what that significant coverage could be (cf. people arguing that listing in a census is significant coverage). If people want to exclude purely statistical listings, they need to say that explicitly. FOARP (talk) 08:19, 1 May 2026 (UTC)

I'd suggest something along the lines of WP:SPORTSIGCOV, for example:

All articles about settlements must include at least one reference to a source providing significant coverage of the subject, excluding database and statistics sources.

Cheers, SunloungerFrog (talk) 08:39, 1 May 2026 (UTC)

That could work in place of/as a re-write of existing prop 3 which had similar intent. FOARP (talk) 08:46, 1 May 2026 (UTC)

I like that. It is simple and enforceable. We would need to discuss how it would apply retroactively, however. Donald Albury 14:59, 1 May 2026 (UTC)

Re retroactiveness, I would not propose mass- or bot- anything (tagging, nominating for deletion, searching for offending articles etc.), at least in the first instance. But rather a simple maintenance template akin to {{No significant coverage (sports)}} that editors can add as and when they come across an article that merits it, or the usual merge, redirect, delete processes. (That's what I do when I have a go at poorly-sourced sportsperson bios).

One of the things we do on WP:URA is run a regular bot sweep of article space and write a list of "probably unreferenced" articles to some user space pages, which human editors then go through and consider what, if any, maintenance templates might be added, or other disposal methods. We could consider something similar, and over time, deficient articles will get weeded out or tagged and eventually fixed. Cheers, SunloungerFrog (talk) 15:37, 1 May 2026 (UTC)

I think including something along those lines for pre-existing articles in an RfC will increase the odds of it passing. I would prefer something more stringent for articles created after the adoption of a new notability standard, and maybe something intermediate for recently mass-produced articles. But then, that would complicate the RfC, so probably should be put off for future discussion. Donald Albury 15:56, 1 May 2026 (UTC)

@SportingFlyer @Newimpartial @Giuliotf @EmeraldRange @James500 @Silverseren @Alexandermcnabb @Seraphimblade @Kingsacrificer @Passengerpigeon Katzrockso (talk) 20:13, 1 May 2026 (UTC)

@WhatamIdoing @Enos733 @Joseph2302 @Sirfurboy @David Eppstein @Andrew Davidson @Stepwise Continuous Dysfunction @Brickguy276 @Seav @N1TH Music Katzrockso (talk) 20:13, 1 May 2026 (UTC)

@GothicGolem29 @Arutoria @Aquillion @Dantheanimator @Zero0000

Editors that were involved in the last discussion last October or the workshop phase that I haven't seen comment here yet that might want to opine. Please add anyone I may have missed. Katzrockso (talk) 20:14, 1 May 2026 (UTC)

• I'm unhappy with all of these. First, unless I misunderstand, all of them contain the presumption of notability in some form; is this the result of a prior RFC? Second, few of them (and few of the arguments for them) reference sources at all. Sources are always required per WP:V and we cannot have criteria that go around this. The reason independent criteria other than the GNG exist is because, per WP:NOTABILITY, The subject-specific notability guidelines generally include verifiable criteria about a topic which show that appropriate sourcing likely exists for that topic. That is to say that subject notability criteria don't really bypass the GNG; they must rely on an implicit assertion that "if something passes this criteria, then sufficient sources to pass the GNG almost certainly exist, even if we don't have them on-hand." And GEOLAND in particular has had problems with this, producing endless permastubs for which the sources manifestly do not exist. Based on this I feel that one option should be to delete GEOLAND entirely (ie. subject everything under it to the red line of the GNG; the sources must actually be demonstrated, because history has shown that we don't have a good way to demonstrate that their existence can be inferred.) But beyond offering that option, in terms of wordings for GEOLAND that I'd find acceptable... I strenuously object to how option 2 (the sole one to reference sources and therefore the sole acceptable option to me) removes the vital typically presumed to be caveat; contextual situations always exist, and in this case in particular it's important. Fixing that would give us:

• Cities, towns and villages are typically presumed to be are notable, but their status must be backed up a reliable sources.

I still am not really satisfied with this, though, because the first part (the "typically presumed" part) has no real meaning; only the sources matter. And the second part has the problem that some people assert that a mere line in a database is enough to base an article on, which is completely unacceptable. A more precise wording would perhaps be:

• Cities, towns and villages are typically presumed to be are notable when their status is backed up by significant coverage in independent reliable sources. Databases, lists, and governmental records that do not qualify as significant coverage of the specific location in question are never acceptable sources for this purpose.

The problem is that, fundimentially, databases and lists simply do not show that appropriate sourcing likely exists for that topic; therefore, we must have an option that strictly bars them from being used. Personally, I would also consider separating off the databases and lists that do not pass WP:SIGCOV are never acceptable sources for the purposes of notability under WP:GEOLAND into its own RFC question, since I feel it's one of the core disputes historically and could be applied to any formulation for the rest. --Aquillion (talk) 20:51, 1 May 2026 (UTC)

Which takes us to: How do you define SIGCOV? WhatamIdoing (talk) 20:55, 1 May 2026 (UTC)

Well, we do have WP:SIGCOV for that (...addresses the topic directly and in detail, so that no original research is needed to extract the content. Significant coverage is more than a trivial mention, but it does not need to be the main topic of the source material.) but honestly I don't think we need a hard-and-fast definition beyond that; editors can reasonably argue over it if something comes to AFD, and IMHO that's a good thing - our policies should provide room for contextual discussions. The important part is that someone who creates an article, and who wants to argue that it should exist, needs to be able to argue in good faith that the sources they've found are significant coverage. If we need something specific to GEOLAND and problems we've run into surrounding it I'd say that, aside from the base requirements of SIGCOV, another key point is that it shouldn't be something indiscriminate - ie. if you can find a thousand entries or pages that are basically identical aside from a few numbers, it's obviously not SIGCOV. But I don't think we need to spell that out, do we? More hazy situations can be worked out on a case by case basis if necessary (but I doubt it will be; the fundamental dispute here is over sources that I suspect everyone agrees do not and could never pass SIGCOV.) I mean SIGCOV is a fundamental part of the GNG, it's not something obscure or poorly-defined. --Aquillion (talk) 21:17, 1 May 2026 (UTC)

Looking at the four options outlined above:

• The status quo
• Remove the SNG totally, use GNG only
• Some revised words that are stricter in some way about the definition of "legally recognised", maybe with per-country subpages that set out in detail how that works on a country by country basis
• In addition to "legally recognised and populated", require at least one reliable source with significant coverage to be present in the article, as per WP:SPORTSIGCOV.

I think that #4 is a bad idea, because it directly conflicts with WP:NEXIST, because editors don't agree on what constitutes WP:SIGCOV (we even have some editors who believe that "significant coverage" means that the source makes a Wikipedia:Credible claim of significance, rather than being a significant amount of (usable) coverage), and because it will inevitably lead to proposals, a few short years hence, that all non-compliant articles be deleted or draftified. (See also, BTW, the current draft proposal to do that for thousands of sports articles.)

I think that #3 has the most possibility, but I think that "stricter" is the wrong concept. It mostly needs to be "clearer", which largely means that it needs to use language and concepts that translate easily between various jurisdictions. For example, give examples that include Incorporated town in the US and say things like "obviously a water district doesn't count, even though it may be a legally recognized entity". WhatamIdoing (talk) 20:54, 1 May 2026 (UTC)

This isn't really the place to argue over individual proposals (that can wait for the RFC), but I do want to say that I feel that your arguments against #4 contradict themselves. On one hand you say that you think there's confusion over WP:SIGCOV, which I strenuously disagree with; but then you say that it will lead to proposals to delete large swaths of articles - proposals you obviously think might succeed, otherwise why be worried about them? And for those to succeed they would obviously have to reach a consensus on SIGCOV, which suggests, to me, that you do recognize that there's at least the broad outline of a community agreement on SIGCOV (sufficient that it could, if applied to existing articles, result in many of them being deleted), you just don't like the implications of applying that in this topic area. --Aquillion (talk) 21:24, 1 May 2026 (UTC)

you say that it will lead to proposals to delete large swaths of articles - proposals you obviously think might succeed, otherwise why be worried about them? Because proposals like that take very significant amounts of time and effort to push back against, often because the proponents keep demanding more and more evidence and sometimes shrubberies. Time and effort that cannot be spent on improving the articles that some people have taken a dislike to and insist are fixed before their deadline. Without this time and energy spend defending against the proposals, then there is a good chance they would pass - to the very significant detriment of the project. Thryduulf (talk) 21:34, 1 May 2026 (UTC)

Thryduulf understands the problem. And so do you, except that you after you admit that "editors can reasonably argue over" whether a reliable source constitutes SIGCOV, you say "that's a good thing". I do not agree with you that an increased amount of arguing is "a good thing".

Here's a simple and predictable scenario:

• Fact: We have something around 10,000 articles for US census-designated places, and thousands of similar articles for many other countries.
• Fact: A non-trivial fraction of those articles, particularly about small towns, only cite their national census.
• Fact: Wikipedia editors do not agree on whether the US census reports constitute SIGCOV. For example, one AFD regular has declared that only sources containing multiple paragraphs of prose can constitute SIGCOV. Some editors take the WP:WHYN approach and say that if a Wikipedia editor can write multiple paragraphs of prose from the source, then that's a significant amount of coverage. Others follow the Wikipedia:One hundred words standard.
• Prediction: If we adopt a "must have SIGCOV" rule, someone who objects to the existence of stubs, bot-created articles, articles that don't say why the subject is 'significant', and/or articles that have not been manually expanded will try to get them deleted or draftified en masse.
• Prediction: We will spend (some would say waste) many, many, many man-hours discussing the proposal.

Ignore whether the proposal will succeed. (I don't think it will succeed in the US. It might succeed if the proposal focuses on a low-income, non-English-speaking country and the proponent first puts in some time to cultivate a belief on wiki that the national census or other main source of data for that country is unreliable; that's what finally worked for the so-called Iranian "villages", after all.) Success for a proposal doesn't actually require an agreement about SIGCOV; it only requires that enough editors say "yes" or "no" to get a decision made. But from where I'm sitting, the fact that we will have l-o-n-g discussions about this IMO means the community loses. We say that the most valuable thing is editors' time, and a fight over whether to delete thousands of articles that contain sources that some people dislike and that other people would accept except that some other people haven't already expanded is a time sink of the first order.

Let me be particularly clear about the disagreement. This is a realistic outcome:

• Delete because it only cites a census database that does not contain even a single complete sentence in prose. Only prose counts (talk) 12:00, 32 Octember 2026 (UTC)
• Keep because the census database contains literally 114 different facts about this place. Count the facts (talk) 12:30, 32 Octember 2026 (UTC)
• Delete because the article is too short. IHateStubs 12:13, 32 Octember 2026 (UTC)
• Keep because the article could be expanded using the sources already cited in the article. ISeePotential 12:43, 32 Octember 2026 (UTC)
• Delete because Wikipedia's reputation is hurt by articles with only one source, and we should blow this up per WP:TNT. Make it shine 14:56, 32 Octember 2026 (UTC)
• Keep because Wikipedia should be a place where people can find out about every part of the world, and it could be expanded if someone wanted to put in the effort. It's not FA material at the moment, but deletion is not cleanup. Bright future 15:26, 32 Octember 2026 (UTC)
• Delete because only one source is cited. Multiplerequired 16:07, 32 Octember 2026 (UTC)
• Keep because local sources should be available for any place in this census. I smell sources 16:37, 32 Octember 2026 (UTC)
• Delete because SIGCOV is required, but the source doesn't have any coverage explaining why this little town is significant to anyone who doesn't live there. Itsinsignificant 16:42, 32 Octember 2026 (UTC)
• Keep because SIGCOV is required, and the census proves that this little town was significant enough to the whole country to be included in the census. Itsimportant 17:12, 32 Octember 2026 (UTC)
• Delete because it was created by bot, so it wasn't even significant enough for a human to write about. Ban the bot 17:49, 32 Octember 2026 (UTC)
• Keep because there was a community consensus to create this back in the day, as proven by it being part of an authorized bot-based WP:MASSCREATION. Support your local bot op 18:19, 32 Octember 2026 (UTC)

None of them agree with each other, and yet they're all ideas that we've seen at AFD and in discussions about these articles.

If we are particularly unlucky, we won't only waste time and energy debating this, but we'll also end up with another round of Wikipedia:Arbitration/Requests/Case/Conduct in deletion-related editing and Wikipedia:Arbitration Committee/Requests for comment/AfD at scale will actually happen. (That one so far has only produced 524 comments on its talk page, if we don't count the existence of Wikipedia:Arbitration Committee/Requests for comment/Article creation at scale, which was created on the same day by the same editor to deal with the same problem, against it.)

I want to avoid having fights over what counts as a "reliable source with significant coverage". I particularly want to avoid mass-deletion efforts that start the week after such a rule gets adopted. Basically, I think that option #3 #4 would be non-functional for our community at this point in time. WhatamIdoing (talk) 23:00, 1 May 2026 (UTC)

I want to address this point specifically: But from where I'm sitting, the fact that we will have l-o-n-g discussions about this IMO means the community loses. We say that the most valuable thing is editors' time, and a fight over whether to delete thousands of articles that contain sources that some people dislike and that other people would accept except that some other people haven't already expanded is a time sink of the first order.

To give some concrete examples on what the problem is: above I mentioned this discussion Wikipedia:Articles for deletion/Sandogo a new editor rapidly pumped out ~80 articles on Burkina Faso villages, these articles were based on faulty sources (i.e. the coordinates they gave for the "villages" often pointed to scrubland with no nearby houses, they used a template and didn't update the administrative divisions when creating each new article, the sources used were websites which scraped the data from old US gazetteers which used spellings are not used today). Even the people who voted keep, upon inspecting some of those articles noticed these problems. This isn't a case of me not liking the sources, the articles were factually incorrect, but because a couple of editors managed to locate a handful of these villages, this clear misinformation stayed online for years. I eventually stumbled on this mess, and it took literal days of my editing time to go through these systematically one by one, fix the articles that could be fixed and send the remaining articles one by one through WP:AfD, taking up more editors time when it would have been far quicker to WP:TNT the lot of them and re-start them from scratch with proper sourcing once it is found.

And I'm not saying that we should blanket delete all geographical permastubs, often times these articles are created en-masse based on one database source. For example the bulk of the existing Burkina Faso village stubs were created using what appears to be a government source that labels these places clearly as villages that send representatives to do local government stuff, I personally believe that the information would have been better presented as a table on the page for the administrative divisions, but either for most people they clearly pass the current barrier for WP:GEOLAND and so I have spent most of the last 6 months of my editing time to bring them up to a better standard, and I still have a fair way to go. But then we have countries like Myanmar where the same editors that created the bulk of the Burkina Faso geostubs created another 1000+ stubs, but this time using sources that get their data from 80+ year old British gazetteers which have all the same problems with inaccurate locations and old spellings that are no longer used, and someone is going to have to go through those at some point and sort them out. Once cases like this are found where articles have been created with systematic problems the way to deal with these while wasting the least possible editor time is to WP:TNT the lot and let them be re-created once good sources become available.

What your statement essentially demands is that editors who want to get this mis-information off wikipedia should spend literal years of their editing time going through these articles one by one and fixing them or sending them through AfD individually rather than trouble the community with the occasional request to bulk delete a couple thousand badly sourced articles riddled with inaccurate information. I will assume good faith and that you are just not aware of how bad the situation is for some country's stubs, and I'll invite you to look through, for example, Category:Populated places in Hkamti District and figure out how long it would take to correct the coordinates of each village, AfD the ones that fail WP:V provide at least one reliable (not even in depth) source for each one. Giulio 00:44, 2 May 2026 (UTC)

I don't want to have fights over whether a tiny town in the US, sourced only to the US Census, is a notable topic.

I believe that "require at least one reliable source with significant coverage [according to the personal and subjective opinion of the AFD nom] to be present in the article, as per WP:SPORTSIGCOV" will result in certain editors starting such fights.

If "require at least one reliable source with significant coverage to be present in the article, and oh-by-the-way, the US Census and any equivalently reputable national census definitely counts as SIGCOV" would solve your concern, then it would likely address my concern, too. WhatamIdoing (talk) 01:06, 2 May 2026 (UTC)

" It might succeed if the proposal focuses on a low-income, non-English-speaking country and the proponent first puts in some time to cultivate a belief on wiki that the national census or other main source of data for that country is unreliable; that's what finally worked for the so-called Iranian "villages", after all" - I have always been clear that the Iranian census is reasonably accurate: I'm sure they counted the number of Iranian people in each general area correctly on the day of the census. It just isn't accurate for the purpose that people on WP have tried to use it for, which is as a list of "villages". It was never intended as this. It was not designed to be this. It is only the insistence of WP editors that some authoritative "list of villages" must exist because they would like one to exist for their policies to make sense that ever indicated otherwise. And yes, the same issue affects many of our census-generated article-sets, as we are discovering with the Russian rural localities with low/no population. FOARP (talk) 17:53, 2 May 2026 (UTC)

If a tiny town in the US has a count in the census, it is either because (a) it has a local government and borders which are defined in legislation, or (B) because a CDP has been defined for it, in which case the census has established the borders for the purpose of counting. In the first case there is no dispute that it satisfies the precise case that the guideline intends to cover. If it's a CDP, that usually means that it is a town for which someone wants statistics, but in some cases CDPs have been established for subdivisions or neighborhoods within cities, which are places which we would normally force to satisfy GNG. However, most small towns fit neither case, and the census doesn't provide data on them. Mangoe (talk) 02:12, 4 May 2026 (UTC)

Minor nit, but CDPs are not defined for neighborhoods within cities (or other incorporated places). CDPs are defined for urbanized areas outside of an incorporated municipality (although they will be within one or more administrative subdivisions). CDPs are intended to provide data for such unincorporated urbanized areas to provide a basis for comparison with incorporated places). older ≠ wiser 02:36, 4 May 2026 (UTC)

Basically, I think that option #3 would be non-functional for our community at this point in time. Just to clear up my own confusion, did you mean option #4 would be non-functional? As you'd previously suggested that I think that #3 has the most possibility. Cheers, SunloungerFrog (talk) 05:31, 2 May 2026 (UTC)

Yes, sorry! WhatamIdoing (talk) 19:42, 2 May 2026 (UTC)

@Thryduulf and @WhatamIdoing as you both mentioned this. For option #4, if we were to include in the RFC proposal a statement along the lines of

If WP:GEOLAND were to be amended in this way, future proposals for mass deletion, redirection or draftification of articles that existed before the amendment may not be made solely on the basis of the revised WP:GEOLAND; all such articles must be considered individually on their merits

would that suffice to help reduce the risk of future lengthy bitter arguments? If not, is there other wording or another measure that you think would help? Cheers, SunloungerFrog (talk) 05:46, 2 May 2026 (UTC)

I think this could just be a stand-alone proposal in case of any amendment. FOARP (talk) 17:36, 2 May 2026 (UTC)

Maybe. I wonder whether editors such as @Giuliotf would find the restriction bearable, though, and in the case of a terrible mess (which has happened, and which will unfortunately probably happen again), it might turn out to be a bad idea.

This is IMO a not-unreasonable path:

• Community: "Here is the new rule. All existing articles are grandfathered under the old rule and must not be proposed for mass anything."
• Editor, a couple of years later: "Ugh, the Ruritanian articles are a total disaster! A hundred GEOLAND articles are citing only HoaxSite.com! And that stupid grandfather clause means I have to individually review each one and nominate them one at a time!"
• Community: "In the face of this misinformation disaster, we hereby repeal the grandfather clause."
• Another editor: "Finally! I've been waiting years for this. I hereby mass-nominate the following mass-created articles about indisputably real US towns for removal..."

Consequently, I think that identifying sources that definitely do indicate notability and some that are not necessarily valid for GEOLAND purposes (e.g., Wikipedia:Reliability of GNIS data) would be a stronger approach. WhatamIdoing (talk) 19:24, 2 May 2026 (UTC)

I am strongly opposed to option 2, which would degrade the quality, usability and value of the encyclopedia, as well as its function as a Wikipedia:Gazetteer. I would be most likely to support option 1, since I think that the case that there is a problem with the status quo has not been proven. I wouldn't be opposed to option 3, which, while this has been dismissed as requiring too much work, would be the most fruitful in resolving some of the disputes. Katzrockso (talk) 21:01, 1 May 2026 (UTC)

Can you please explain how "degrade the quality, usability and value of the encyclopedia". Currently we have the masses creation stubs everywhere based on the presumption that if its in a census its legally recognised. 2 outlines what settlements are acceptable, and at least has a standard to mark it against, when legally recognised is not a realistic achievement, as previously discussed, as lots of countries dont have legally recognised settlements. Davidstewartharvey (talk) 11:54, 2 May 2026 (UTC)

…or extend “legal recognition” to individual buildings. FOARP (talk) 16:16, 2 May 2026 (UTC)

With regards to whether vast amounts of geostubs improve or degrade the quality of the encyclopaedia, I would like to make point that I haven't seen anyone else make:

Geostubs, unlike stubs about, for example, sports people who competed in the Olympics decades ago, require regular maintenance. If we take the Burkina Faso geostubs that were mass-created in 2008, the information they contained was a 2005 population estimate of each village and maybe some inaccurate coordinates. Since then the census results from 2006 we made public, another census was conducted in 2019, and the results were later made public and in 2025 the administrative divisions were changed impacting, every single village , these articles need a lot of work to be kept up to date and while an individual may be able to take it upon themselves to maintain the stubs for a small country like Burkina Faso (which is what I've spent several months trying to do), there are plenty of other countries and simply not enough people who are willing to do this tedious work. This creates enormous amounts of WP:WIKIDEBT as these articles become more and more out of date as time passes, and if the entire content of the article is a name, administrative division, population at an arbitrary point in time and coordinate location, then the information would perhaps be better presented and easier to be kept up to date as part of a list of, for example, villages in each administrative division. This would likely improve the quality of the encyclopaedia without losing any information. Giulio 16:38, 2 May 2026 (UTC)

Right. Our Iranian "village" articles are overwhelmingly based on the 2006 survey also. 20 years and two census later, and no-one is updating them. Even when C46 - who created the overwhelming majority of them - was still on here they weren't bothering to update them. And honestly, why would anyone want to? It's a task best performed just by running a bot - which should instantly give people pause: is an article-set that can only reasonably be written and maintained by bots really what we want to maintain?

These articles were in large part created on the now-obviously-wrong theory that if article-sets were made then other editors would be inspired to come and work on them. That's not what's happened. FOARP (talk) 17:47, 2 May 2026 (UTC)

I'm not concerned about outdated contents. "Here's the population recorded in the 2005 census" is not misinformation. I'm much more concerned about calling something "a village" or "a town" when it's just a census tract in a rural region.

I wonder if we could put together a list of national censuses with practical information for editors. The US census corresponds to the goals well enough that we can (and do) have a bot run the updates. The Iranian census needs warnings about the difference between an actual village and the census tracts. How many others could we include in a list/table/helpful information page? WhatamIdoing (talk) 19:50, 2 May 2026 (UTC)

Having outdated census data, while sub-optimal is at least acceptable if properly labelled, the bigger issue is when administrative boundaries move as then that becomes misinformation when a place is listen in Province X when it hasn't been in that province since the province boundaries changed in 20XX.

I think the fundamental question is when we can have pages like Banfora Department what value is there in adding a separate page for each of the listed villages when the only information presented in those articles can be contained in a simple table like that when the downside of such a system is that it makes maintenance so much harder and more tedious? Giulio 21:08, 2 May 2026 (UTC)

Do you know of any editors who genuinely believe that editing a wikitext table is easier than editing a paragraph of plain old text? I don't. "Just" adding a new column (e.g., to make a space for the newest census data) is considered one of the most difficult tasks. WhatamIdoing (talk) 23:26, 2 May 2026 (UTC)

It's not editing a table vs editing a paragraph of plain text, its editing a table on a single page vs editing paragraphs of text and info boxes on 22 different pages, and as the person who created that table and updated the population on those 22 different pages, I can tell you that working on a single page is significantly easier. Giulio 23:31, 2 May 2026 (UTC)

One final thing on how these mass-produced articles pollute the information space: just zoom in on any stretch of empty, unpopulated desert in Iran on Google Maps and you'll see place names in it. Click of them and there's a reasonable chance that Google Maps will serve you a link to one of Carlossuarez46's Iranian "village" articles. I just did it and literally the first place I clicked on was this. And yes, this was a C46 article. Even when it doesn't serve you such a link, there's a good chance that Google Maps originally scrapped WP to get the location and we've since deleted the article.

WP is a data-set that is scrapped by other other data-users. Systematic errors on WP end up metastasising all over the place. For example, I wouldn't be surprised if C46's "village" articles have polluted military targeting data-sets. FOARP (talk) 09:50, 3 May 2026 (UTC)

Other websites also scrape the databases that C46 used. WhatamIdoing (talk) 20:47, 3 May 2026 (UTC)

I actually agree completely that in many cases it makes sense to merge up the smallest localities to a larger administrative geography. Fram gave an example of what it might look like for small Russian localities somewhere and it seemed eminently reasonable. Katzrockso (talk) 00:22, 3 May 2026 (UTC)

FRAM also explained why this isn't working - the task is just immense, and GEOLAND is a no-further-excuses-needed justification for blocking or undoing such work. FOARP (talk) 08:12, 3 May 2026 (UTC)

Wikipedia is not a gazetteer, and in any case real gazetteers limit themselves to listings of names and at most very summary characterizations and statistical data. We do not need to write articles on any place in order to satisfy the supposed "functions of a gazetteer". Mangoe (talk) 03:10, 3 May 2026 (UTC)

Right. We are not a gazetteer (literally a kind of dictionary that accompanies a map or atlas) any more than we are an almanac.

Most gazetteers are just lists of names and corresponding map-locations. Exactly the kind of content that WP:NOT rejects and has rejected since very early in the project.

Wikipedia is an encyclopaedia. This means it is a summary of what secondary sources say about a topic. Notably, censuses are not secondary sources, but are primary sources. FOARP (talk) 06:36, 3 May 2026 (UTC)

That's a nice essay you wrote to clarify the distinction between "being" a gazetteer and having some "features" as described in Wikipedia:Five pillars. I'd like to see more help content like what you've put in Wikipedia:Wikipedia is not a gazetteer#Commonly-used gazetteer or gazetteer-like sources. WhatamIdoing (talk) 20:47, 3 May 2026 (UTC)

My thoughts would remain the same. Populated places would have a list under an appropriate administrative division (in the US that would be by county; other countries would have different "parent" administrative divisions to use.) The normal notability guidelines would apply to whether there should be an article about one, but for those wanting a "gazetteer", the lists would serve to have the few factoids available about the non-notable places (location, population, etc.). Existing "articles" on non-notable places could then be redirected to the appropriate place. Seraphimblade ^{Talk to me} 11:31, 3 May 2026 (UTC)

I like #3, sort of. IMO, the purpose of SNGs is to provide easy heuristics for editors. If members of a class of articles almost always meet the GNG, but particular sources are difficult to reach or esoteric, the SNG can bypass a lot of unproductive inclusion/deletion fights. Problems arise when the heuristic in the SNG fails to deliver, and most of the members of that class are not demonstrably notable. I think we have spent way too much time tied up in interpreting "legally recognized". In the narrow sense (an entity with distinct borders and multi-functional government, not just a census district) this worke reasonably well as a heuristic in the Western Anglosphere. It obviously fails when extended to census districts; and it's not clear to me that it works well in other parts of the world, where there may be a much lower density of secondary sourcing. I think Donald Albury's objection that this constitutes a "Herculean amount of work" is best answered by developing the SNG piecemeal, as we get a better picture of what sources of information exist for a given country, rather than trying to to it for the whole world at once. Choess (talk) 15:05, 3 May 2026 (UTC)

I am going to start updating and expanding Wikipedia:Geographic references in the hope that it can form the nucleus of an informational annex to the SNG, or as supporting material for a revised version as per my option #3 about. I agree that it will be a lot of work, but I think what we do should be worthwhile anyway even if we decide not to go with an RFC. Cheers, SunloungerFrog (talk) 16:11, 3 May 2026 (UTC)

• I've been thinking about this discussion and the issues being addressed. I fall toward the line of mass creation is a behavioral problem. But, if we are going to try to be more precise, I propose "Primary administrative divisions and municipalities holding a charter or formal grant of authority from their government..." --Enos733 (talk) 20:47, 3 May 2026 (UTC)

  Offhand, I doubt that editors will know what that means. For example, does Swindon count? Almost a quarter million people live there, but it apparently has no royal charter. WhatamIdoing (talk) 20:51, 3 May 2026 (UTC)

  In the UK basically only cities and towns that had markets in the medieval period have charters, and some settlements that have been cities since time immemorial (pre 1189 in this context) might not even have that. It's also arguable whether the charters are from the government or from the monarch. Modern grants of authority are to councils rather than settlements, and the two frequently differ. As ever though, everything to do with UK administrative geography is almost fractally complicated!

  Many small settlements particularly will never have had a grant of authority of any description as they have never had any administrative responsibility. Thryduulf (talk) 22:59, 3 May 2026 (UTC)