Yet another mass creation problem[edit]

We are now seeing a bunch of articles at AfD of Sri Lankan villages which may have been entered from the government statistical site. At any rate, that is the only citation, and it is a link to the whole site, so finding the supposedly cited data is already a major chore.

Are we ever going to do anything about these? Mangoe (talk) 13:37, 1 November 2023 (UTC)Reply[reply]

According to this SNG this and another 3,000,000 villages that don't have articles are wp:notable. But I think that it's been shown that the crowd doesn't want a strict GNG to apply, so as to preclude an editor building a real article on one of them when they don't have GNG sources. So I think that the solution is to have a policy or guideline against mass creation of stubs. While one could argue that that is a wiki-wide job, it is really most applicable to NGeo because it is the "loosest" SNG. Maybe we should create a geo-speciic one here. The main tenets would be:
IF you are relying on this SNG for notability, you need to create real articles, not stubs. Multiple sources and several sentences of text including some info that didn't come from database entries. This applies (only) to new articles.
So technically, this is merely a requirement to use the Geo SNG
North8000 (talk) 14:04, 1 November 2023 (UTC)Reply[reply]
You can see the articles created by Ser Amantio di Nicolao on 18 January 2011 here. As you can see by scrolling down, the run of articles ran straight from Sri Lanka to Hungary, and the speed of creation makes it very unlikely that data from Sri Lanka was being used - more likely an alphabetical database of names was being used, most likely GNS.
For the avoidance of doubt obviously 2011 is a long time ago and I am not attaching any particular amount of blame to the creators of these articles. However, the articles they created clearly do not pass GEOLAND as created, and most of them cannot be improved due to lack of verification that would allow anyone to even understand what the real topic of the article is.
And yes, this does show why GEOLAND has acted as a loop-hole for effectively spamming Wikipedia with articles about "villages" whose existence as such cannot be confirmed. FOARP (talk) 16:30, 1 November 2023 (UTC)Reply[reply]

Unordered lists[edit]

Does anyone mind if I were to change the unordered lists (i.e. bulleted lists) to ordered lists (i.e. numbered lists) under the "Settlements and administrative regions" and the "Engineered constructs" sections? I feel like this would make referencing the individual lines easier in AfD discussions. — Red-tailed hawk (nest) 03:10, 8 November 2023 (UTC)Reply[reply]

On the notability of places sourced solely from census data[edit]

For your edification:

older ≠ wiser 15:56, 22 November 2023 (UTC)Reply[reply]

Census tracts, at that. The creator has been around for almost 6 years, and was just given a barnstar for creating a lot of start-class articles! Donald Albury 22:29, 22 November 2023 (UTC)Reply[reply]
I have dropped a reminder about the requirements of GEOLAND on the creator's talk page. Donald Albury 22:41, 22 November 2023 (UTC)Reply[reply]

If it is populated and generally recognized as a place (not some abstract entity like an irrigation district) this guideline mostly gives it a green light regardless of db type sourcing. This is an inherent problem with this guideline as it mostly green-lights about 2,000,000 yet-to-be-created village stubs. We should probably tweak this. North8000 (talk) 23:25, 22 November 2023 (UTC)Reply[reply]

But, GEOLAND specifically says that a census tract is not recognized as a place and thus, not presumed to be notable. Donald Albury 00:11, 23 November 2023 (UTC)Reply[reply]
Rampur, Tundla (census code 125462) is listed as a village in the census,[1] so the prohibition against census tracts is presumably not applicable. Whether this village is recognized by any government, who knows? — hike395 (talk) 01:23, 23 November 2023 (UTC)Reply[reply]
"this guideline mostly gives it a green light regardless of db type sourcing" I'm sorry but maybe I'm missing something, this guideline doesn't appear to give any green light. It doesn't appear to modify which sources count towards notability at all. Horse Eye's Back (talk) 04:01, 26 November 2023 (UTC)Reply[reply]
  • GEOLAND is broken. That's been known for a very long time. The article-sets that we have already are problematic, the ones coming down the pipeline will be worse.
  • If people are going to write articles based on super-long documents, they need to give page numbers to show that they did actually check the source. Asking readers to check PDFs that are hundreds or thousands of pages long is asking too much. FOARP (talk) 23:25, 24 November 2023 (UTC)Reply[reply]
@FOARP: Do you think we should emphasize that census data is not evidence of legal recognition? Or is it? (as you say, the guideline is not fit for purpose). — hike395 (talk) 23:49, 24 November 2023 (UTC)Reply[reply]
@Hike395 I think the RFC that was raised only a couple of months ago show the mixed feelings about the SNG. GEOLAND is broke, as there is only such a thing as a "legally recognised place" in the USA. How To fix it? Think that is impossible without the whole community actually acknowledging that it is. Davidstewartharvey (talk) 08:43, 25 November 2023 (UTC)Reply[reply]
@Hike395 - I think "legal recognition" was always just a wrong turn. Editors are not lawyers. Especially they are not Indian public law lawyers (or, to take another recent case, North Carolina constitutional lawyers from the 19th century). Even speaking as a lawyer (but not specialising in anything like the relevant area) I have no insight at all into what exactly Iranian law says about the status of any specific community. But the GEOLAND requires us to know what forms of "legal recognition" a place has - something even the average inhabitant of that country won't know.
Just as much to the point, law in this area varies wildly between countries so we end up with hyper-granular coverage of what are essentially individual houses in some countries, and broad-brush coverage in others. I actually don't object to articles about Indian communities of several hundred people nearly as much as I object to Russian "villages" of 2 people, which in reality are railway stations in the middle of nowhere and for which we have no evidence have ever been anything else. With the Indian communities it is credible that there's some more coverage out there, for a place of 2 people this just isn't likely at all.
Finally we are effectively breaking NPOV by predicating our coverage of a country not on any objective criteria, but on what the local authorities in that country have decided should have a particular status. FOARP (talk) 09:58, 25 November 2023 (UTC)Reply[reply]
How does GEOLAND require it? I just double checked and it doesn't appear to. Horse Eye's Back (talk) 15:32, 26 November 2023 (UTC)Reply[reply]
Require what? Obviously knowing what "legal recognised" is for a particular country means knowing what the law in that country says. FOARP (talk) 12:06, 27 November 2023 (UTC)Reply[reply]
Sorry, though it would have been completely obvious what was being referred to: "But the GEOLAND requires us to know what forms of "legal recognition" a place has - something even the average inhabitant of that country won't know." Horse Eye's Back (talk) 15:10, 3 December 2023 (UTC)Reply[reply]

We need to formulate a fix to this SNG in a way that would pass and not be a radical change from what is currently widely in articles. Sincerely, North8000 (talk) 15:57, 25 November 2023 (UTC)Reply[reply]

References

  1. ^ "Firozabad" (PDF). Census of India. p. 81. Archived from the original (PDF) on 2023-03-27.
The issue with "legally recognized" was that it was a backhanded way of saying "incorporated", that is, established in law. Instead, people keep trying to push it as meaning "has an entry in some governmental source", which really isn't the same thing. My personal stance is that just having a entry in a census database isn't good enough: at the very least I want a location, and even then it's reasonable to argue that such places can be listed in a table, and don't need to be padded out into an article by slapping on an infobox.
I'm beginning to think that the whole notion of notability guidelines has gone off the rails. they've turned into a way to protect articles on things that will never satisfy GNG when they are mass-created from listings. Mangoe (talk) 16:05, 25 November 2023 (UTC)Reply[reply]
Exactly this. I had always understood "legally recognized" as being an entity that has legal standing in a court of law (it can be a party to lawsuits). Unfortunately, this is based on US-centric legal distinctions (or at least may not accurately represent situation in large portions of the world). older ≠ wiser 16:16, 25 November 2023 (UTC)Reply[reply]
Precisely. This concept of "incorporated" is entirely US-centric. All countries work differently. -- Necrothesp (talk) 11:32, 27 November 2023 (UTC)Reply[reply]
I've got to say I get particularly frustrated when people say that a potential GNG or at least a coverage-based notability standard for inhabited places is "anglocentric". The present standard isn't even "anglocentric", it's specific to a single country: the United States. It definitely was not written for the countries of CEE. FOARP (talk) 12:04, 27 November 2023 (UTC)Reply[reply]
Yes, place them in a list or table. If sources sufficient to meet GNG are found, then a stand-alone article can be created. We shouldn't be creating stub articles with inadequate sourcing on the assumption that sources may exist "out there". Donald Albury 18:23, 25 November 2023 (UTC)Reply[reply]
Maybe an easy way to ease into this and solve the problem is to make articles on regions which have towns, settlements as sections, or table entries. They would still need sourcing to be in the article but not GNG sourcing. This would also allow a graceful way to handle geostubs via moving the info leaving a redirect. These could start out as being for largerish areas (like provinces) and then get divided later if needed. North8000 (talk) 18:49, 25 November 2023 (UTC)Reply[reply]
To combat the issue of one user making one-line stubs of any named locality in the Prairie Provinces, this was what was done in Alberta -> Big Lakes County#Communities and localities. List approach used instead of a table. Redirects were then created for all localities to this article heading. Because I created the redirects I was able to see on my watchlist when someone tried to created an article on them. My concern at the time wasn’t GEOLAND but more so unreferenced one or more lines of garbage, often using incorrect assumptions like that these places were hamlets when they were not. Hwy43 (talk) 20:16, 25 November 2023 (UTC)Reply[reply]
Thats a behavioral issue not an issue with policy or guideline, if an editor is being disruptive then bring it up with them and if they don't stop bring it to ANI. WP:COMPETENCE after all. Horse Eye's Back (talk) 15:34, 26 November 2023 (UTC)Reply[reply]
I did something similar to what User:Hwy43 points to in Historic communities of Alachua County. I didn't create redirects to places on the list, but I have placed anchors on some of the entries to allow linking in to the small bit of content that is in the list. Donald Albury 02:34, 26 November 2023 (UTC)Reply[reply]
Might be a good idea to promote this. The folks who like to make a lot of geostubs might switch over to creating such articles for all of the states and provinces. And then we have less of a potential issue for the 2,000,000 settlements which meet the SNG requirement but don't yet have a separate article. North8000 (talk) 15:29, 26 November 2023 (UTC)Reply[reply]
Why can't we just get rid of those folks? Horse Eye's Back (talk) 15:34, 26 November 2023 (UTC)Reply[reply]
The vast majority won't have the sort of coverage we would need for the stand-alone list to be notable... Making a list of non-notable things isn't a solution if the list itself isn't notable. Horse Eye's Back (talk) 15:36, 26 November 2023 (UTC)Reply[reply]
If the list is in the article about the containing entity, and that entity meets GNG, then there is no problem. The question then is whether such a list in a subarticle would fail GNG because it is a seperate list article rather than a section in an article. If it is determined that Historic communities of Alachua County should not exist because it does not meet the criteria of GNG, the list could be merged into the Alachua County, Florida article. By all means, though, let us discuss how GNG applies to list articles. Donald Albury 17:50, 26 November 2023 (UTC)Reply[reply]
I was thinking more of something with substantive entries (such as a section or substantial table entry) for each item rather than a list article. On a different note, I doubt that such an article would get deleted for notability. Or maybe we could explicitly cover it here. It's hard to imagine that this SNG greenlights each of the 1,500,000 future articles on cities and settlements in China and India, but not an article on the cities and settlements of the Hunan province of China. North8000 (talk) 17:36, 27 November 2023 (UTC)Reply[reply]
Maybe something like: "Articles collectively on cities and settlements within one of top two levels of the primary administrative divisions of a country are presumed to be notable." North8000 (talk) 17:52, 27 November 2023 (UTC)Reply[reply]
We've been over the idea of a level-based notability and it's never found consensus and basically just repeats the problems of the legal-recognition standard (i.e., you need to understand the legal statuses of communities in every country in the world). FOARP (talk) 08:46, 30 November 2023 (UTC)Reply[reply]
If the list is in the article about the containing entity then it is not a stand-alone list. Horse Eye's Back (talk) 15:12, 3 December 2023 (UTC)Reply[reply]

List articles (and similar) do need a lot of clarification. What makes it difficult is that it is a topic that is completely scattered amongst several policies and guidelines, and none of them really handles it.North8000 (talk) 17:40, 27 November 2023 (UTC)Reply[reply]

I could try a test example, a province in China. Going in it would be "eyes wide open" regarding these challenges.

Do you think I should try starting one? North8000 (talk) 15:44, 2 December 2023 (UTC)Reply[reply]

I think what you are talking about is quite different from something like List of township-level divisions of Hebei. That type of list currently exists for some administrative divisions of China. I think what you are proposing sounds much like what I did at Historic communities of Alachua County. That is a list of 47 items of two to six sentences each (plus a couple of pointers to an alternate name for a place). Some of the items link to a stand-alone article. For places that do not meet GNG, the item includes what can be verified from reliable sources. The trick will be choosing the administrative level in each case that will produce a list of reasonable length. Donald Albury 23:10, 2 December 2023 (UTC)Reply[reply]
Yes, I agree. But I think that one difference between the historic communities item and my idea is that your historic communities article is a more or less complete list whereas I'm thinking about something which typically would have only have a tiny fraction of it's stated scope. Basically to handle this case: An editor is considering making an article on one of the approx 1,000,000 villages in China that don't currently have an article. They have a few factoids on it, maybe a thumbnail-suitable picture. Likely to become a permastub. It complies with GEOLAND. The article I'm think of would be good place for them to put that entry into a row in a table or as a section. North8000 (talk) 15:03, 3 December 2023 (UTC)Reply[reply]
It might work. I suspect that some editors get a kick out of starting thousands of new articles. Even if you create a nice framework for someone to place their information, I predict that there'll still be editors who make geo-permastubs. — hike395 (talk) 17:14, 3 December 2023 (UTC)Reply[reply]
I phrased it that way to clarify the difference. It could also be a place for others to merge articles into. North8000 (talk) 17:53, 4 December 2023 (UTC)Reply[reply]

Do my articles pass notability guidelines?[edit]

I have created numerous articles for Sri Lankan villages using a single source: an 1896 gazetter written by a local British judge:

The above are just a few. (Note that I actually didn't create the above articles, only revamped a bunch of old unsourced village stubs created in 2011 using this source, but I plan to nonetheless create new articles using the same template for the ones without articles.)

Do these articles pass notability guidelines with the one 1896 source? The opening of the book states "I confess to innumerable errors. Those who use the book will, I hope, pardon these errors on finding much that is accurate and interesting."

So I do not know how to go about this. What I have done and can do is state "according to such and such gazetter..." before any historical facts about each village. I don't think we need to assume the quote from the book applies to the population data, since this is obviously just taken from officially published government data. But if this is a real issue I can remove this demographic info if you deem this source unreliable.

I don't want to go ahead and create 300 of these articles only for them to be AFD'd. I would like an opinion on this. Thanks! —  Melofors  TC 09:45, 11 December 2023 (UTC)Reply[reply]

Thanks for your work and for posting the question. I think that it reflects recognition that the there are millions of settlements that could pass wp:ngeo where I don't think that most wikipedians would figure is appropriate for our encyclopedia to have millions of new articles on. May I suggest creating a broader article on "municipalities of xxxx" and put the content of those potential article into substantial table entries or article sections which have basic info on them including general info and a small image) While current policy/guidelines do not give guidance towards this, IMO it is a route to handle these types of entries with the objective of being an encyclopedia of articles. . Sincerely, North8000 (talk) 00:33, 12 December 2023 (UTC)Reply[reply]
In the case of the articles you've listed above, they seem to be expansions of articles created by Dr. Blofeld that were based off what appears to be the Sri Lankan Census. Those are fine; they meet the threshold of legal recognition and they're populated, so they easily meet the first bullet in WP:GEOLAND. Continuing to expand those sorts of articles makes plenty of sense and improves the encyclopedia.
As for creating new articles that are solely based on the old work, I would advise trying to pair them up with at least some other source before making a standalone article. There's a good chance (if the village is still around) that the transcription/name may have changed in the past 125 years, and we want to try to avoid duplicate articles on the exact same subject. If the sole source is the colonial-era Gazetteer, it's not clear to me that the threshold of "legal recognition" would apply; the work appears to be a description of remote villages more than a document with legal authority, and you would need to meet GNG absent evidence of legal recognition even though it's a populated place. In that case, it may be better to add the villages with attribution to a relevant list article/article on a broader geographical area as North8000 has suggested above. If you can find other sources that talk about a particular village (such as censuses or other similar works), then we may well have a case for a standalone article while reducing the likelihood of accidentally duplicating an existing article. — Red-tailed hawk (nest) 20:55, 12 December 2023 (UTC)Reply[reply]
Sorry but they dont meet WP:GEOLAND. In the US, census data states if the settlement is recorded it has legal recognition. However as stated in the last RFC this rule is not the same for all census across the world. In fact looking at the Sri Lanka census rules it looks like the only legal recognition is for the district, not the settlement. Davidstewartharvey (talk) 06:30, 13 December 2023 (UTC)Reply[reply]

We need people to expand the existing stubs rather than create more!♦ Dr. Blofeld 09:48, 13 December 2023 (UTC)Reply[reply]

A way to fix geoland?[edit]

On one hand, this SNG greenlights millions of future permastubs which don't have sourcing to make a real article and which aren't a real article. On the other hand, Wikipedia (IMO rightly so) tilts the scale towards geographic articles because it is a highly enclyclopedic area and there is no consensus for and resistance to totally eliminating this tilt. Elsewhere I proposed a new type of article where these could be grouped, but this post is about having separate articles on these.

IMO one can often find an answer by putting common sense / general consensus into words. For example, regarding the millions of settlements in China and India which don't have an article. More specifically, I'm talking about ones which meet this SNG but can't demonstrably meet GNG. If someone started a pattern of creating permastubs on all of these (even if not mass creation per se), I don't think that that is wanted. But if somebody created a more substantial article with several sources and substantial unique content on one of these settlements, but which does not have full GNG sources, I think that most people would be happy that the SNG allowed it in. They also don't want to do anything that would enable a deletion-fest on existing articles.

What if we put this into words in this SNG? That using this SNG for new article requires meeting all of the current SNG criteria plus having several sources and substantial unique prose content?

Sincerely, North8000 (talk) 16:37, 15 January 2024 (UTC)Reply[reply]

But the SNG is flawed. Legally recognised is not standard that can be used. As per last year's attempt to clean up the SNG, it is different per country (even if they have one). The original SNG was US biased, based on the fact that settlements in a census are Legally recognised, but most other places are not. A question from an editor before Christmas was should they create Sri Lankan villages based on a census, but Sri Lankan government census only legally recognise the region. Davidstewartharvey (talk) 16:48, 15 January 2024 (UTC)Reply[reply]
IMO this modification of the SNG would solve that. If they want to use the SNG way in, besides meeting the existing SNG criteria, a new article would need to have multiple sources and substantial content. North8000 (talk) 18:39, 15 January 2024 (UTC)Reply[reply]
But was the failure of the RFC last time being editors worrying that multiple sources and substantial content was too much like GNG. Davidstewartharvey (talk) 21:10, 15 January 2024 (UTC)Reply[reply]
The last time IMO the problem was that the proposal essentially WAS GNG and also applied to existing articles. My idea is more lenient than that in both areas. A major GEO editor (who creates high quality substantial articles and whose opinion I consider valuable) flatly opposed it. IMO many of their articles would not pass GNG, but all of them would pass under my idea. North8000 (talk) 21:39, 15 January 2024 (UTC)Reply[reply]
The problem is that NGEO simply requires WP:V that it is or was a populated place. Many permastubs don't even get there. I don't think anything really needs fixing. SportingFlyer T·C 15:53, 17 January 2024 (UTC)Reply[reply]
There lies the problem. Because a census has a "place", editors take GEOLAND and state its a legalised populated place, and is verified by the census. Against GNG it would fail. Census data can only be legally recognised place in the US where it is stated in law. As per the thread above "Do my articles pass notability guidelines?" Sri Lanka only legally recognises the district not the village. That's why GEOLAND is flawed. Davidstewartharvey (talk) 16:12, 17 January 2024 (UTC)Reply[reply]
I mostly agree with the first half of your post. It's something that I basically already stated in my OP. But there is a problem or at least an ambiguity. There are millions of settlements and entities that qualify under this SNG which don't have articles which do not demonstratively qualify under GNG and many of which couldn't qualify under GNG. Again, setting aside the possibility of mass creation...should we welcome creation these as articles even if stubs or likely permastubs?
  1. Maybe the answer is as you imply, "yes" and we should just clarify that.
  2. Or maybe it is as I proposed here to add a criteria that there be multiple references with content beyond "it exists"
Sincerely, North8000 (talk) 16:03, 17 January 2024 (UTC)Reply[reply]
Is there enough information in the article to verify the settlement as a "legally recognised" populated place? That's all we need. (Legally recognised is in quotes because it's doing the job of "distinct hamlet, not neighbourhood" here.) SportingFlyer T·C 16:09, 17 January 2024 (UTC)Reply[reply]
To answer the question in your first sentence, for my example/question, the answer is "yes", there is enough information in the article to verify the settlement as a "legally recognized" populated place, and meet the other SNG criteria (not a neighborhood or abstract district) . North8000 (talk) 16:25, 17 January 2024 (UTC)Reply[reply]
As stated, "millions" referred to ones that meet NGEO in China and India combined. North8000 (talk) 18:47, 26 January 2024 (UTC)Reply[reply]

And here we go again: another huge DB dump[edit]

I have discovered in looking at an AFD for a spot in Myanmar that essentially all the populated place articles for that country are sourced to GMaps, Bing maps, and/or Maplandia. They are all just dots on maps, not even as good as GNS (which itself is bad, particularly in a country like this one) because we have no idea at all where the names next to the dots are coming from. At least they are in Latin characters, but picking one at random, I couldn't find it at the supposed (somewhat vague) location on GMaps.

Whatinthehell are we going to do about these? Nobody should have to take these hundreds if not a few thousand articles through AfD; it's not even clear that there's anything we could rigorously verify these against, and even simply checking them against the maps is a massive effort. Can someone please tell me why we shouldn't just delete the lot and start over with something more like a reliable listing (e.g. a census)? Mangoe (talk) 00:14, 18 January 2024 (UTC)Reply[reply]

Notability of latitude / longitude[edit]

There's a discussion at WP:VPP#Latitude and longitude articles which may be of interest. RoySmith (talk) 14:52, 26 January 2024 (UTC)Reply[reply]

Categories