Jump to content

Wikipedia talk:Set index articles/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

Criteria

The article now says in the second The criteria for creating, adding to, or deleting a set index article should be the same as for a stand-alone list. This might be true of the selection criteria, whether X should be in the list or not; however, it does not seem to be true of the creation criteria which are clearly stated in the first paragraph. I am not even sure that it works for the selection criteria for the same reason. The main criteria is sharing a common characteristic as well as a common name. Lists such as List of people from South Carolina have three possible characteristics that provide for inclusion, a set index page should have one. I think that it might be better to change this to read The criteria that are used for stand-alone lists may provide some guidance. --Bejnar (talk) 00:02, 9 June 2014 (UTC)

Just to be clear (re "now"), that wording hasn't changed recently. What aspect of MOS:SAL do you think doesn't apply to SIAs? Changing from "should" to "may provide some guidance" makes the instructions much weaker and hence harder to enforce consistency. DexDor (talk) 04:38, 9 June 2014 (UTC)

Notability for inclusion

Many lists, especially those of people, require notability for inclusion. Set index articles should only do so explicitly. Set index pages, such as those on notable families, will often include non-notable members to show the relationships between notable ones. Similarly some set index pages, such as Dodge Charger are more effective if they are complete (i.e. list the universe of Dodge Chargers) which often means listing a non-notable version. I suggest adding the following sentence: Set index articles that restrict inclusion to notable members only, shall do so explicitly on the talk page. --Bejnar (talk) 00:02, 9 June 2014 (UTC)

I'm very not keen on which MOS applies to a page being dependent on a notice on the talk page - it would make wikignoming much slower. How about saying something like "Every entry in a SIA should be referenced - either on the SIA itself or on a linked article" ? - that way totally non-notable people (or whatever) can be removed quickly. DexDor (talk) 04:52, 9 June 2014 (UTC)

Applicable WikiProjects?

Out of interest, are there any particular WikiProjects under which set indices would fall? — Sasuke Sarutobi (talk) 15:27, 9 June 2014 (UTC)

I've added several project tags. DexDor (talk) 21:41, 15 June 2014 (UTC)

SIA example used here is nominated for deletion

You may consider reviewing Wikipedia:Articles for deletion/List of peaks named Signal Mountain. The article, List of peaks named Signal Mountain is used as an example of good practice in this guidance. Informed views on the role of the nominated article vs. the role of Signal Mountain, a dab page, would be appreciated. --doncram 01:40, 13 August 2015 (UTC)

Contradiction

Hello.

I recently read the page's lead and I see contradictions in the first paragraph:

  • The first sentence indicates two criteria: Name and type. It reads: "A set index article (SIA) is a list article about a set of items of a specific type that share the same (or similar) name."
  • This half-sentence says name alone is the criterion: "A list is only a SIA if inclusion of an item in the list is due to the name of the item."
  • The other half of the sentence implies name alone is not the criterion: "e.g. every entry in a list of earthquakes may include the word "earthquake", but that does not mean the list is a SIA."

Even the editors of this page do not seem to agree with the criteria.

  • I thought that perhaps because two out of the three indicate name alone is not sufficient, then the second item must be a typo. My edit: [1]
  • Hike395 thinks that I am wrong and name alone is sufficient; he tried to eliminate the third item outright, and forgot to address the contradiction with the first item. His revert: [2] And his edit: [3]
  • DexDor disagreed with Hike395 but the article returned to original contradictory state. His edit: [4]

Alright people. What is the criteria? Name alone? or name and type?

Best regards,
Codename Lisa (talk) 16:34, 25 August 2015 (UTC)

I don't see any contradiction here, but perhaps there are opportunities to make it clearer (note: I added the earthquakes example). Here's a more detailed example - consider the following hypothetical page
List of tall towers in Foo City

The tallest towers in Foo City are:

  • Foobar Tower, a 300m high office building
  • Barfoo Tower, a 200m high apartment block
Is that page a SIA (because every entry uses the word "Tower") ? IMO it's not. If you say it is a SIA then when another entry is added (e.g. "* Foo City TV mast, a 100m high antenna") which doesn't use the word "Tower" then presumably it's no longer a SIA. IMO a page should only be a SIA if inclusion on the list is based on the names of things. If you agree with me that pages like the towers example are not SIAs then how do you think the guideline could be made clearer? DexDor (talk) 21:59, 28 August 2015 (UTC)
@DexDor: The following sentence in the page disagrees with you and says the opposite:

A list is only a SIA if inclusion of an item in the list is due to the name of the item.

In other words, the towers listed do not need to be tall or even a tower, as long as their name includes "tower"; i.e. "Ivory Tower", a local nightclub can be in the list. Seeing the contradiction yet?
As for my opinion, I am pretty much in your camp. Still, my edit is reverted, so, I believe there is a dispute.
Best regards,
Codename Lisa (talk) 19:57, 29 August 2015 (UTC)
No, I don't see a contradiction; what do you think the contradiction is? DexDor (talk) 20:04, 29 August 2015 (UTC)
@DexDor: Well, please answer the question then: Does my example "Ivory Tower" (a nightclub) merit inclusion or not?
Best regards,
Codename Lisa (talk) 20:06, 29 August 2015 (UTC)
The nightclub (assuming it isn't a tower) doesn't belong in a list of towers. If the list page was titled "List of things in Foo City with 'Tower' in their name" (or similar) then the nightclub would belong in the list (and the TV mast wouldn't). That page would be a SIA (as inclusion in the list is based on the names of things) - although whether it would survive a deletion discussion is another matter. DexDor (talk) 20:18, 29 August 2015 (UTC)
Then you accept that type is also a criterion, don't you? In that case, the sentence must look like this: "A list is only a SIA if inclusion of an item in the list is due to both the name and the typeof the item." Best regards, Codename Lisa (talk) 21:09, 29 August 2015 (UTC)
Am I right in thinking that the discussion has moved on from whether name is relevant (we both think it is) to whether type is relevant? I don't think the change you propose is necessary (that sentence is specifically talking about name being a criterion), but it would be consistent with "of a specific type" in the first of your bullet points, but that would appear to mean that List of things named Daedalus would not be a SIA so maybe "of a specific type" should be changed. Note: I think the intent of that wording is to distinguish SIAs from dabs. DexDor (talk) 22:03, 29 August 2015 (UTC)
I wouldn't say the discussion is moved on; I clearly asked "What is the criteria? Name alone? or name and type?" in the opening thread. But the discussion remains confusing. Best regards, Codename Lisa (talk) 22:53, 29 August 2015 (UTC)

As Lisa says, the very first sentence of WP:SIA summarizes the concept nicely: "A set index article (SIA) is a list article about a set of items of a specific type that share the same (or similar) name." So, it's clear that it's both name and type, which was the original intent from the discussion that set up SIA. I agree that the sentence about earthquakes is confusing as it stands (which I why I wanted to delete it from the guideline).

List of things in Foo City named Tower should not be an SIA, but is a dab. It's a list of things with the same name but different type. Most importantly -- there is no specialized information that is shared between the items, so it should have a very simple format per WP:MOSDAB. List of towers in Foo City is neither a dab, nor an SIA, but is simply a list article. List of things named Daedalus would redirect to Daedalus (disambiguation), and is clearly a dab, not an SIA --- same name, different type.

Talking about a list of earthquakes named earthquake seems to be giving people odd advice. Theoretically it's an SIA (because the type is earthquake and the name is earthquake), but the fact that the name of an item is also the type of the item seems like an extreme corner case. Talking about this in the guideline will confuse readers. It certainly seemed to have confused Lisa. List of earthquakes is clearly neither an SIA, nor a dab.

Can we simply drop the sentence about earthquakes? It doesn't add anything to the guideline. —hike395 (talk) 02:49, 30 August 2015 (UTC)

Later --- I hadn't realized that List of things named Daedalus is actually a separate list article. To me, that's a clear misuse of WP:SIA, and should be merged into a dab (I proposed the merge). If people are using the earthquake sentence to justify the existence of List of things named Daedalus, then the sentence really needs to go. It was never supported by consensus. —hike395 (talk) 03:03, 30 August 2015 (UTC)
Thanks, Hike. That was useful, especially the link to the original discussion.
Now, can I ask both of you two something to better clarify things? What about a page called List of server types in computing? It would contain web server, mail server, file server, application server, catalog server, directory server and print server. Would it count as a SIA?
Best regards,
Codename Lisa (talk) 00:01, 31 August 2015 (UTC)
I think what you're asking about is analogous to List of earthquakes named earthquake, i.e., List of servers named server. The fact that you're asking makes me appreciate why DexDor added the sentence to begin with --- perhaps it isn't such an extreme corner case after all and I was speaking too rashly, above.
In theory, List of server types in computing would be an SIA, because the type is "server" and the name is "server". However, I think that is not what was intended for SIAs. I think that List of server types in computing is a plain list article. Should we explicitly carve out an exception if the name and the type are the same, e.g., List of mountains named Mount? Perhaps we can substitute the following for the earthquake sentence:
Articles where items names are shared with their types, e.g., List of earthquakes named earthquake or List of mountains named Mount should not be considered set index articles. Please consider carefully whether such articles should have such titles, or simply be standard list articles such as List of earthquakes.
In practice, I wonder if this really makes any difference to editors' workflow. SIAs are list articles, these self-referential list articles are also list articles. Whether the latter are SIA or not doesn't seem to make much practical difference, except for the {{dmbox}} label at the bottom of the article. Now, Doncram over at WP:DPL is proposing changing some of the dab tools to also handle links to SIAs. In that case, we may wish to be careful about what gets labeled as a SIA. —hike395 (talk) 08:23, 31 August 2015 (UTC)

How bots handle SIAs?

I know that if I link an ambiguous term, I will get a notification from a bot with the suggestion to disambiguate? Does this happen with SIA pages? I.e., if I wikilink [[Schumacher]] in an article, will the bot bother me? If yes, then this guideline may want to mention this similarity with dab pages, just as in mentions the similarity with list pages. Staszek Lem (talk) 17:54, 10 August 2015 (UTC)

Currently, I do not think SIA pages are included in the bot disambiguation notifications. olderwiser 19:38, 10 August 2015 (UTC)
Seconded. There is no systematic review of links to SIA pages like there is for disambiguation pages, although indeed perhaps there should be. Thanks for commenting about this guideline. Should the guideline provide more coverage of gray areas between disambiguation pages and SIAs? The guideline seems to suggest SIAs are better, because they are not subject to formatting requirements of DABs, without providing balance going the other way. --doncram 01:55, 13 August 2015 (UTC)
I'd rather not get an alert from a bot about a link to an SIA. I have often hatnoted to a set index article from articles listed within it, especially handy when then SIA is large. Also I've done quite a lot of work on the comics SIAs and the SIAs will often be legacy superhero aliases, which it is handy to link into from the articles about the individual character - Hank Pym is probably the most famous one I can think of. Ant-Man has expanded into a larger article (so doesn't count as an SIA) and Giant-Man has recently been expanded [5] so may no longer been an SIA any more, but Goliath (comics) and Yellowjacket (comics) are, and I wonder if Wasp (comics) should be (but it probably isn't worth the effort moving it and tidying up incoming links).
It might be a robot could drop a note into the SIA talk page, that way the creator of the page would get informed (they've presumably gone through the pages that link to it to ensure the incoming links are right. Or perhaps if they are a newer editor a robotic nudge might be handy. However, I don't think it is a big enough problem to warrant a robot checking it.
What I would like is an alert when someone redirects the SIA because they mistake it for an incomplete disambiguation (I had to do that with Yellowjacket just now and run across dozens of examples - just did Vision (comics) too) although perhaps the person doing the redirect should get an alert. Actually, I was saying that in jest but that would be handy, as I'm pretty sure most of the people doing that either don't know about set indices or have misunderstood what they are. (Emperor (talk) 04:20, 20 September 2015 (UTC))

Title of a set index article when the primary topic is already used?

Hello. How the title of a set index article should be disambiguated when the primary topic is already used? with (disambiguation)? Ip Man (disambiguation) and Kuala (disambiguation) are correct? Saeidpourbabak (talk) 08:14, 19 April 2017 (UTC)

I don't see any reason Ip Man (disambiguation) should be considered an SIA instead of a disambiguation page. Kuala itself might not be a WP:PRIMARYTOPIC, in which case it could be moved. --BDD (talk) 18:13, 20 April 2017 (UTC)
Ip Man (disambiguation) changed to dab, Kuala AfDed. Widefox; talk 16:00, 29 April 2017 (UTC)

Why?

I don't mean to be cheeky, but I've never understood: why do "set index articles" exist? As far as I can tell, it's a completely invented concept that, in practice, seems to be a way for us to make a disambiguation page without following MOS:DAB. Why is this a valuable distinction? --BDD (talk) 14:37, 16 March 2017 (UTC)

SIAs mainly exist because the Ships wikiproject wanted (and may still want) pages that fulfill the same role (or a _very_ similar role) as dab pages, but can have redlink-only entries (e.g. A-class submarine). See: Wikipedia:History of SIAs. DexDor (talk) 20:32, 16 March 2017 (UTC)
Thanks, that's a very helpful overview. Sounds like my initial assessment is not wrong, though... --BDD (talk) 21:21, 16 March 2017 (UTC)
I don't see anything of value in A-class submarine that couldn't be in A-class. Since MOS:DABRED allows redlinks in DABs as long as there is a bluelinked article with that redlink, the mere addition of red links doesn't seem that useful. For the A-class submarines, the Spanish ones are mentioned at List_of_retired_Spanish_Navy_ships#Submarines, where they could be redlinked. ---- Patar knight - chat/contributions 22:34, 16 March 2017 (UTC)
Good point. This reminds me of the debate over empty sections in mainspace articles. I'm generally of the opinion that it's unseemly to show readers what should be internal processes when there isn't a clear benefit to them. But I can understand the argument in signaling, either via a redlink or an empty section, "Yes, we're aware of this, even though we don't have anything more for you right now." --BDD (talk) 23:55, 16 March 2017 (UTC)
I mean we could do that by having something like * A-class submarines (Spain), a retired class of Spanish submarines. in the main DAB page, which would meet MOS:DABRED, or just not have the redlink and only the second link, which would meet WP:DABMENTION. Having the separate seems like doubling the number of pages to maintain, while not providing any additional benefit to readers.---- Patar knight - chat/contributions 01:16, 17 March 2017 (UTC)
I was involved in the original definition of SIAs, so I can give you my perspective. Indeed, the original started with Ships, but WP:WikiProject Mountains made us generalize the concept. Readers may want to know statistics about a list of mountains with the same name: either so they can identify the relevant one ("Oh, yes, this is the really high one in Wyoming"), or just out of curiosity ("Wow, there are 37 peaks named Silver!"). Ships with the same name often have a shared history that is useful to know. Now, some editors wanted to force all lists of objects of the same type with the same name into the DAB format, but that was artificially limiting perfectly valid list articles. So, we broke out the SIA concept to allow people to build these list articles. I want to emphasize these should be valid list articles, not dab pages. —hike395 (talk) 06:42, 17 March 2017 (UTC)
Thanks for that perspective hike395. I was involved then too, or was at least observing from the sidelines, and your description matches what I recall. I think a crucial detail that is often overlooked is that SIAs should be valid list articles, rather than disambiguation page look-alikes with a different template (although I admit I have on occasion swapped templates without doing anything to improve the SIA). I fear that some editors who patrol Wikipedia:Disambiguation pages with links, sometimes choose to transform disambiguation pages with persistent or problematic links into SIAs, or some other non-disambiguation type simply to get it off the list. I also think there has been a proliferation of SIA subtypes without any real input from a sponsoring project willing to maintain and improve the pages, and in many cases, these would be better off remaining as disambiguation pages (IMO). olderwiser 10:20, 17 March 2017 (UTC)
Thanks both for your perspectives. I definitely agree that tighter "regulation" of SIAs could be helpful for all. The thing for me with mountains is that just about every country, every US state, and I assume many other jurisdictions, has its own "List of mountains in" article. Where those articles don't exist, it just seems like it would be so easy to make them and then use MOS:DABMENTION on the disambiguation page and call it a day. Put more concisely, a list entry is going to be much more helpful than a dab-like line with no functional link. Which brings us back to the idea that SIAs should be true lists, rather than just wild-west disambiguation pages. --BDD (talk) 14:12, 17 March 2017 (UTC)
IMHO unless the SIA provides some tangible benefit to the reader (e.g. statistics in Mountain pages, picture galleries for plant SIAs, additional content for context), they should be reconverted into DABs and/or merged back into the parent DAB page. WP:SIANOTDAB says: "A set index lists things of only one type and is meant to provide information as well as navigation". If the SIA does not do this, then it's not a SIA, but a miscategorized DAB. This in turn hinders reader navigation when people turn them into SIA and delete the resulting "(disambiguation)" redirects to them on the pedantic reason that the target is now a SIA. I would propose replacing "An SIA need not follow the formatting rules for disambiguation pages, although many SIAs do." with An SIA should not follow the formatting rules for disambiguation pages. SIAs that do follow the disambiguation rules should either be tagged as a disambiguation page and linked from the main disambiguation page or merged back to the parent disambiguation page.---- Patar knight - chat/contributions 03:30, 18 March 2017 (UTC)
I don't think we want to force DAB-conforming SIAs back to DABs. For example, look at Ironwood (the second-most-visited plant SIA). Right now, it's clearly in DAB format. But, because it belongs to Category:Set indices on plant common names, I found it and now I can add a gallery and species distributions. If this were forcibly DABed, then I wouldn't have found it to fix it.
How about if we make a warning template, {{Not a set index}}, and a tracking category Category:Set index articles without extra information? That way, they can still be classified as SIAs, but then either improved into a real SIA or converted back to a DAB. Would that satisfy you, Patar? —hike395 (talk) 04:18, 18 March 2017 (UTC)
That sounds like a workable idea. There are certainly SIAs that do add value, and people should be able to find them. We could also have a category for DABs that could be turned SIAs. Maybe after X time spent tagged as {{NOTSIA}}, it should be reconverted to a DAb and the page added to Category:Set index candidates (hopefully with the tagging/categorization done in a semi-automated fashion if possible). ---- Patar knight - chat/contributions 05:16, 18 March 2017 (UTC)
The emphasis should be on retagging some SIAs as dabs so that inlinks get fixed (rather than deliberately making them less like dabs). For example, a page such as Pied flycatcher should be a dab page (not be adorned with photos, map etc). (I've just fixed a load of inlinks to that page). DexDor (talk) 07:33, 19 March 2017 (UTC)
@DexDor: Why do you think that, DexDor? I would think that telling the readers the family of the species (Muscicapidae vs. Monarchidae), and supplying a photo of each would be informative and help guide them to the right bird. Is it because a list of two items is too short? Or because it's too obscure? —hike395 (talk) 13:37, 19 March 2017 (UTC)
The main advantage of a page like that being tagged as a dab (rather than a SIA) is that inlinks are likely to be fixed quickly (the person making the link gets a DPLBot notification) and/or efficiently (because there are editors/tools that specialise in fixing inlinks to dabs). Adding the family would be fine (whether the page is tagged as a SIA or as a dab). Photos aren't (afaics) totally prohibited on dab pages (see MOS:DABICON), but I don't think it would help disambiguation in this case (the birds are found in different parts of the world) and for many dab/SIA pages (e.g. listing lots of people with the same name) having a photo for each entry would make the page bigger. If you think pied flycatcher should be tagged as a SIA rather than as a dab then what about, for example orange emperor? Note: Your ping didn't work because you didn't add a signature in the same edit. DexDor (talk) 07:01, 22 March 2017 (UTC)
While we're revisiting SIA/dab split, aren't there mechanisms already for dealing with list articles that have limited or no difference from a dab.. I notice that many SIAs are unsourced, e.g. List of people with surname Smith (which also has the article Smith (surname)). Why is it that SIAs, as fully-fledged list articles are not challenged on notability or some other current content reason per WP:SAL? Existing content levers (unsourced, notability etc) can be to put to use converting unhelpful ones to dabs, and also references would justify any redlinks that are not existing in articles. I think Patar knight is on to something - they can't be compliant fully-fledged articles and comply with MOSDAB. From my experience, it's a complex value judgement for converting based on preserving valid SIA entries which would be lost as a dab, project knowledge. Any clarity would be welcome. Widefox; talk 10:53, 20 March 2017 (UTC)
I think the two Smith articles actually illustrate the purpose of the list/DAB/SIA split quite well; List of people with surname Smith is just that — a list — while Smith (surname) is a prose article detailing the history, etymology, and distribution of "Smith" as a surname.
To me, is the point of SIA is that it allows scope for a sectioned, prose treatment of subjects that share not just a name, but a type (i.e. they are all members of the same set, like with a set of ships), especially in the case where it is unlikely to generate enough content for more than a collection of stubbed articles (or while such content is being compiled). Disambiguation, on the other hand, is simply to differentiate between similarly- or identically-named subjects of different types (or not all necessarily the same type). One of the key differences is that information on differing objects is likely to be structured differently, and therefore any attempts to treat them together in a prose article will become messy. For example, an article on a person and a ship named after them will likely have a biography of the person, with their birth, life, and what they are known for; none of these sections would have relevance in the structure of the ship's information, especially beyond the reason for the ship's naming.
So the point of SIA is not that it is a list format, like DAB articles are, but precisely that it isn't; it's an article format, so it gives scope to expand a list with prose about the listed items. I agree with Widefox that set index articles should be expected to adhere to the same standards as other types of articles, but I don't believe that we should begin converting SIAs that don't currently adhere to article standards. If anything, I think that would risk muddying the purpose of SIAs even further, as you would start building a collection of articles that have the same scope as SIAs, but have been changed from being SIAs. I think that would be especially confusing to any future editors who are trying to understand SIAs. — Sasuke Sarutobi (talk) 08:56, 22 March 2017 (UTC)
User:Sasuke Sarutobi just to say the obvious Smith (surname) isn't an SIA, and SIAs are list articles. Widefox; talk 11:19, 22 March 2017 (UTC)

Take two random examples

Burton (name) had the wrong template on it, and Dylan (name) was missing the template. Is there a difference between a set index article and an article containing a set index? —Xezbeth (talk) 13:05, 22 March 2017 (UTC)
Thanks. I put the cats in the template but nothing changed. They're both still SIAs. Widefox; talk 10:38, 23 March 2017 (UTC)

Isn't the fundamental problem with SIAs that they are dual use - both navigation (dab by another name, "!dab", solely navigation), and project use (not solely navigation). SIAs allow projects their own freedom away from the strictness of MOSDAB, and the more that projects exercise that freedom, the less useful the SIA is as solely navigation. Put another way, we don't link to dabs, do we link to SIAs? As a "!dab", that's bad, but as a fully fledged list article it's good (else it's an orphan). It can't be both good and bad! Take Pied flycatcher, if it was a dab, I'd put Australian pied flycatcher as the link for the second entry, but as an SIA I'd leave for the project. Wouldn't it be simpler and serve readers better to get feedback from projects about revisit/tweak/deprecate SIAs? I see it more as an internal maintenance split ("pedantic" has been said by others) but there's no doubt project views not yet voiced. It's possible projects would be happy to create redlinks in (list) articles, and then they'd be valid dab entries. Has anyone asked recently? Any list needs to justify itself including WP:V and I really don't see SIAs going that way of fully-fledged list articles (which is how they are defined), so they remain in no man's land between dabs and lists allowing entries that fail WP:V, hindering disambiguating incoming links whilst being, by definition, ambiguous about the validity of incoming links themselves. They're a compromise, but are they still a useful one? Widefox; talk 11:19, 22 March 2017 (UTC)

Although I reserve some rebuttal of your comments directly above, I generally agree with your conclusion that SIAs came of compromise, and say Yes, I believe their purposed existence is quite useful.--John Cline (talk) 11:22, 23 March 2017 (UTC)
Yes, they're in the "let sleeping dogs lie" category for me. Projects desire for unallowed dab redlinks can be worked around easier than creating SIAs in between lists and dabs. I've just discovered we have {{Siadn}} - the equivalent of {{Disambiguation needed}}. The SIA definition currently is broken as they can't both be inappropriate to link to, and necessary per their definition as fully-fledged articles (above). Similar for references: as a !dab, they don't need references as they mirror referenced content in articles, but as an article they do, else SIA entries fail WP:V, for example. Projects creation todo lists are better handled either in articles (lists and others) where WP:V clearly applies or internally on project pages. If we clarify their purpose, we may be able to eliminate the paradoxes. I suspect that the currently internally inconsistent definition is the root cause of current maintenance issues causing division at RfD, CSD, AN, WP:suppress. We could fix the cause or put a plaster on the symptoms. Fundamentally, dictionaries organise by the same spelling, encyclopedias don't. There's a default preference IMHO not to have articles based on the same name per se, unless notable. That's surely the default position? Widefox; talk 09:51, 24 March 2017 (UTC)
I think you've put your finger on it, Widefox. If SIAs are valid list articles (which I believe that they should be, e.g., Dodge Charger, List of peaks named Signal), then they should be linkable. Should we change this in the guideline? —hike395 (talk) 12:17, 24 March 2017 (UTC)
I suggest getting wider consensus before proceeding either way "forwards" (keeping SIAs, and narrowing the definition as linkable, full list articles, essentially a subset of list articles with entries of the same name, and removing the {{Siadn}} etc) or "backwards" to say they are like dabs in some way. Maybe some SIAs need to go forwards, and some converting to dabs, I don't know. While I'm here, if there's a reason for a grouping - like a "brand" e.g. Dodge Charger, what's the difference between that and a WP:DABCONCEPT e.g. the Nokia Lumia example? If the only reason for the grouping is the same name and type, which I presume List of peaks named Signal is, then it should fend for itself as a cohesive list inclusion requirement/notability as a topic (per encyclopedia not dictionary). I'm being very precise to highlight, I don't have an opinion if SIAs still have a use, but if they can be in theory divided into DABCONCEPTS, lists, and dabs then they may not be needed, or at least we can codify their raison d'être. Widefox; talk 17:27, 24 March 2017 (UTC)
I believe that if we didn't explicitly allow SIA list articles in a guideline, that editors who prefer dabs would convert them all to dab format. IMO, this would make WP less useful. —hike395 (talk) 10:47, 28 March 2017 (UTC)
Can you help me understand the use? e.g. what's the difference between the SIA and dab concept examples? Widefox; talk 10:56, 28 March 2017 (UTC)
A link to a "Dodge Charger" may be valid (a broad concept), maybe sometimes not specific enough, but a link to "List of peaks named Signal" needs disambiguating as it is never valid as a broad concept unless there's an underlying reason for them to be in a group above their name (ie a broad concept), so unless there's a reason for the list to exist per se, it is a WP:CHIMERA. We don't generally link to dab pages, so why should we link to !dabs? It's not obvious to me what circumstances a link to the peaks would be valid. That may be intuitive given the title List of peaks named Signal but not so obvious if it was titled Signal Peak or Signal Mountain, as is common for SIAs. On first impression, the lede of the peaks SIA seems to fit within the scope of the broad concept Beacon, and be US centric. Widefox; talk 11:02, 25 March 2017 (UTC)
See also:

Moving on

I came late to the party, and it's unfortunate that the discussion got stale without any action produced. I share the opinion of many editors above that SIAs are an unnecessary distraction and should be deprecated, and I'm contemplating starting a RfC on the issue. However, structuring a successful RfC is a challenge, as they often got sidetracked or rejected for a variety of reasons, procedural or substantial. I would appreciate help of all involved in the discussion in formulating the RfC and preparing answers for questions that might ensue. I feel it would be worthy of listing it in Wikipedia:Centralized discussion.
I'm thinking of something like the following format:

Proposal: Wikipedia:Set index articles should be deprecated and gradually converted to disambiguation pages, lists or articles, as appropriate. Policy page Wikipedia:Set index articles should be marked as historical.
Arguments in favor
  • SIAs can be and often are inadvertently linked, which is seldom an appropriate target. If converted to disambiguation pages, a bot will notify editors about mis-links and ask them to fix them.
  • <More points from the discussion here>
Arguments against
  • Since set index articles allow for wider formatting latitude than disambiguation pages, particularly allowing red links, references and coordinates, MOS:DABRL will need adjustment to accommodate reasonable exceptions (for example, red links and coordinates of geographic features, and red links for presumably notable people, possibly with {{Interlanguage link}})
  • Category:All set index articles has 69,000 entries, so it is vast amount of work. However, it is populated automatically; the majority comes from {{Surname}} (41393) and {{Given name}} 11718, so that leaves only a few thousand pages to review.
  • <More arguments to be defined>
Process outline
Support
Oppose
Discussion

If anyone's interested, let us set up a subpage Wikipedia talk:Set index articles/RfC draft with above contents, to work on ironing it out? Obviously, the proposed process is just my draft idea. No such user (talk) 14:40, 21 July 2017 (UTC)

I think that multiple editors (above) see the value in SIAs, so I predict you're not going to get consensus. If you do decide to move forward, please poll the WikiProjects that have specialized SIAs: Mountains, Ships, Comics, etc. —hike395 (talk) 05:16, 22 July 2017 (UTC)
I don't have too high hopes for a consensus, but I'd like to test where the community stands on the issue nonetheless. I did have in mind those wikiprojects indeed.
Shamelessly pinging BDD, Patar knight, Bkonrad, DexDor, Sasuke Sarutobi, Widefox, John Cline, Xezbeth (hope I got'em all) for an opinion whether it's worth pursuing. No such user (talk) 14:26, 24 July 2017 (UTC)
I don't support wholesale conversion of SIA into disambiguation pages. Some, like surname and given name pages are nothing more than partial title matches which are generally only allowed on disambiguation pages in short doses, if there are many such names a separate list is preferred (and for the record, I think the case for including given names on disambiguation pages is far, far weaker than surnames with the relatively unusual exceptions where a person is commonly known by the given name only; in contrast, it is not unusual for a person to be referenced by surname). Other SIAs which aim for comprehensiveness, such a lists of peaks can provide some additional information over a simple disambiguation page, but in general I'd like to see such SIAs have a sponsoring project that actively maintains them with defined standards and can articulate a clear rationale for having a particular type of SIA. So I could support a narrower proposal that focused on specific types of problematic SIAs. olderwiser 14:39, 24 July 2017 (UTC)
Agree with Bkonrad. Maybe we can document the different uses of SIAs.
My (latest) SIA understanding is that with notable exceptions like surnames and clearly given names, SIAs need a dab covering the (bulk of the) items anyhow. Is that duplication useful for readers? We could convert those. Just because I don't get why we have SIAs outside of the name lists etc, doesn't mean they aren't useless. SIAs escape both WP:LISTN, and WP:MOSDAB but per my reasoning above they aren't defined well enough currently and fall down define (in an internally inconsistent way) the crack between lists, dabs and WP:DABCONCEPTs. I'd want to hear from projects that find SIAs useful about what they actually need - e.g. surname and given name list articles yes for finding an article which doesn't generally overlap with dabs, but can't be normal lists - those shouldn't even be called "set" IMHO as they aren't attempting to be a complete set of those topics with redlinks, but a finding mechanism. We wouldn't want to create redlink items in those. I'm struggling to understand how SIAs satisfy WP:LISTN - even the example List of peaks named Signal doesn't seem to have a source detailing the set? I can't see anyone questioning the notability either. Beacon details how "Beacon Hill" is common, but Beacon Hill is a dab and surely both together would make a much more useful dabconcept for readers which wouldn't be UK or US centric? Wouldn't it be more useful for readers to have examples of signal mountains without the constraint that they have to be also called "Signal Mountain".
We have exceptions to MOSDAB to aid readers selecting their article, but does that need to extend as far as a sortable table? Widefox; talk 15:50, 24 July 2017 (UTC)
I would support such a proposal, but fear it would not be successful. If it resulted in a reduction of SIAs, though, I would count that as progress. --BDD (talk) 15:57, 24 July 2017 (UTC)

Counterterrorism Center (disambiguation)

I converted the dab Counterterrorism Center (disambiguation) to an SIA. Suggest moving to "List of..." but leaving for others as no confidence in current state of SIAs/titling with (disambiguation) currently. Widefox; talk 16:03, 29 April 2017 (UTC)

SIA to evaluate

Further to the above SIAs, Order (mathematics) seems like it would be better as a dab, or putting back in the dab it was split from. Widefox; talk 13:17, 17 August 2017 (UTC)

Does WP:INTDAB apply to set index and other list articles or does it even matter?

There are discussions in several venues concerning this question.

Additional input welcome. olderwiser 12:55, 12 March 2017 (UTC)

 You are invited to join the discussion at Wikipedia talk:Disambiguation#RfC: INTDAB links to non-dab pages. Shhhnotsoloud (talk) 07:58, 26 June 2018 (UTC)

SIA request: Simoncini (surname)

Anyone willing to create an SIA for this Italian/Sammarinese (not sure which, really) surname, preferably from the Anthroponymy WikiProject? I happen to notice a lot of persons on this Wikipedia with that surname, including:

Thanks! 99.203.31.12 (talk) 19:56, 9 October 2018 (UTC)

Done. Staszek Lem (talk) 20:56, 10 October 2018 (UTC)

SIA article naming, linking, templates, bots

Relevant portion of discussion about SIAs for train lines copied from WT:D

For a SIA, there's no need for an underlying reason why the different train lines need to share the same name. We have SIAs for mountains that share the same name, and plants that share the same common name, and those don't necessarily have underlying reasons. I think that Shhhnotsoloud's suggestion is a good one. —hike395 (talk) 17:06, 12 January 2019 (UTC)

There is also a pile of SIAs on places in Russia which coincidentally share a name, and they're an infernal nuisance. Just about every one I've fallen across has had one or several bad links in which needed disambiguating. Example: Vostok (inhabited locality) lists 16 places in Russia whose name means 'east'. There was a bad link to that page from Total S.A., which I've just removed: Total Vostok isn't based in a place called Vostok; Total Vostok is the name of the company (with offices in Moscow, St. Petersburg, Rostov-on-don, Yekaterinburg and Kazan). At one time, the link was to Vostok and someone had tagged it {{dn}}. Another editor (who has been around for six years, has made several thousand edits, and has a dozen WP:GAs to their credit) had 'solved' that problem by linking to the SIA and removing the tag. The useless link had therefore lain unnoticed for over a year.
Apart from country and name, those places have nothing in common. From WP:SIA, "Being a set of a specific type means that the members of the set have some characteristic in common, in addition to their similarity of name" (emphasis in the original). For me, merely being somewhere in Russia doesn't cut it. An SIA could be written listing all the places in USA called Springfield, and it would be equally dangerous.
I do not think identity of train line names is enough to make them "of a specific type". Quoting WP:SIA again, "every entry in a list of earthquakes might include the word "earthquake," but that alone does not mean that the list is an SIA".
I'd wager that placename, mountain, plant, ship, album title, and other SIAs all suffer from exactly the same problem of bad links. It's bad for the encyclopaedia. Narky Blert (talk) 18:49, 12 January 2019 (UTC)

<snip>

The original intent of SIAs was to allow list articles be about items of the same type (ships, cars, mountains, comics) with the same name, without being forced to follow the strict rules at MOS:DAB. The 2007 consensus (supported by editors from various WikiProjects) was that adding additional type-related facts and metadata was helpful to our readers and encyclopedic. For example, SIAs about mountains show their location and elevation (see List of peaks named Signal).
The point of SIAs being about the same type of object is that a list article can objectively compare them. All mountains have elevations, for example. All train lines have length and locations and ridership. Items in a DAB can be of any time as long as they share the same name, so DAB articles don't make sense as list articles.
To me, train lines are objects of "a specific type", are comparable, and are perfectly legitimate subjects of list articles. I believe that a SIA named Red Line (transportation) (or List of transportation lines known as Red Line) would be legitimate. I would note that the SIA guidelines, as currently written, would recommend List of train lines named Red, which is a bit more WP:CONCISE and WP:PRECISE.
Hope this helps. —hike395 (talk) 12:27, 16 January 2019 (UTC)
@Hike395: I like the idea that because SIAs are always list articles, 'List of' titles are ideal. It works well if Fizbuz is a DAB page which includes a link to List of widgets called Fizbuz. It works badly if there is no DAB page, and Fizbut redirects to List of widgets called Fizbuz.
My principal (and principled) objection to SIAs is on practical not aesthetic grounds. It is that they, exactly like WP:PTOPICs, accumulate bad links-in which cannot easily be found and are rarely fixed. That is bad for the encyclopaedia.
Vostok (inhabited locality), which I posted above, is a case in point. When I first fell across it, I tried to clean it up – but, working by eyeball alone, I missed one.
I have an idea which might flag those bad links-in for correction. I think it looks straightforward and practical. However, its implementation would need the services of one or more editors experienced in bots and templates, and I lack the energy to learn those skills.
I recently set out a closely-related proposal at Wikipedia talk:Disambiguation pages with links#Where next? part 1 of 2 – intentional links to DAB pages (SIAs would be #3 of 2...). If any editor would be willing to do the dirty implementation work, I would be happy to lay out my idea for finding bad links-in to SIAs (and to name pages, my original #2 of 2) as a formal detailed specification. Narky Blert (talk) 22:29, 22 January 2019 (UTC)

@JHunterJ, Shhhnotsoloud, Dekimasu, Narky Blert, and R'n'B: SIA discussion copied here, in case you're interested.

Incorrect linking to SIAs has been a persistent problem for years. It would be nice to solve it. Some of the proposed methods of solving it have been:
  1. Redefining (or eliminating) SIAs, see discussion from 2017, above, especially comments by Widefox
  2. Eliminating SIAs by modifying MOS:DAB to allow more useful data
  3. Forcing SIAs to start with "List of..."
  4. Narky Blert's proposal of a combination of a bot and template, in order to manually check in-links to SIAs and mark appropriate ones.
To me, expanding MOS:DAB could be the easiest to accomplish, but might interfere with dab navigation. Forcing SIAs to start with "List of..." is appealing, but contravenes WP:CONCISE (e.g., do we really want to rename Cypress to List of plants named Cypress?). I think Narky's proposal might work --- if there's someone who wants to write the bot.
It also seems to me that most of the SIA in-link pain is caused by {{surname}} or similar articles. Is that right? It seems that the vast majority of SIAs are such articles: can we fix most of the problem by creating a fix just for those "name" articles?
What do other editors think? —hike395 (talk) 09:58, 23 January 2019 (UTC)
SIAs are as easy as we want to make them - as lists they must be sourced. How stringent we are on requiring sources for the set of the items is debatable. The tension comes from the desire of projects to create lists evading the stricter requirements of MOSDAB without satisfying the necessary extra requirement of providing references. Result = mess. Apart from specific types - names, ships etc - most should be challenged and dealt with IMHO. Widefox; talk 19:38, 23 January 2019 (UTC)
@Widefox: two related (non-rhetorical) questions for you.
  1. When you're saying lists must be sourced, are you following WP:V+WP:RS (for the SIA items themselves), or WP:LISTN (for notability of the list criterion)?
  2. Let's assume that we add and enforce requiring sources in SIAs. Does that help the in-linking problem? Are you expecting most SIAs get deleted via gradual cleanup by either conversion to dabs or AfD?
hike395 (talk) 10:30, 24 January 2019 (UTC)
Yes:
  • WP:V All material in Wikipedia mainspace, including everything in articles, lists and captions, must be verifiable...any material whose verifiability has been challenged or is likely to be challenged, must include an inline citation that directly supports the material (emphasis own) so yes SIAs as lists must have sources themselves. For example redlinks with no mention/use in WP which are invalid per MOSDAB are valid per SIA and can only be verified with a citation.
  • As SIAs inclusion criteria is the name and type and acts predominantly as a navigational aid (some/many SIAs such as name articles often transcend that), then WP:LISTN is the debatable part and SIAs could be explicitly added to WP:PURPLIST#Navigation as an exception to satisfy WP:N.
My general point is that the strictness of dabs/MOSDAB serves readers well (by definition, if not it should be changed) while allowing us to keep dabs in line with policy and ease of maintenance. The definition of SIAs falls in the gap between dab and list and thus evades both, which must be fixed per WP:V. As usual, projects don't get to override wider consensus such as WP:V per WP:LOCALCONSENSUS a WikiProject cannot decide that some generally accepted policy or guideline does not apply to articles within its scope. . These problems are a directly result of the internally inconsistent (presumably intentional) SIA definition - a duality of dab-ish/list-ish which evades necessary but opposite aspects of each which are mutually exclusive (e.g. dabs shouldn't have incoming links, SIAs should, dabs should not have sources, SIAs should). Widefox; talk 12:22, 24 January 2019 (UTC)
Apart from country and name, those places have nothing in common - but the same logic applias to ship SIA: apart from being a ship they have nothing in common. As someone mentioned, ships mey be compared by various parameters - same with localities: population, where, when established, etc. Staszek Lem (talk) 17:53, 24 January 2019 (UTC)
The above ping does not seem to have worked, which probably accounts for the lack of responses here to what was a fairly lively discussion on the other page. Dekimasuよ! 18:56, 5 February 2019 (UTC)

Feedback sought for proposed article-to-DBA/SIA conversion at Allopathic medicine

Your feedback would be appreciated at Talk:Allopathic medicine#Make this page a disambiguation page. Thanks, Mathglot (talk) 06:34, 17 May 2019 (UTC)

Apologize for the ping if it's unwanted, but we could still use opinions at this discussion. Pinging top ten editors at WT:SIA: @Widefox, DexDor, Renamed user 2560613081, BDD, Patar knight, Bkonrad, Staszek Lem, No such user, and Mr. Guye: Thanks, Mathglot (talk) 22:59, 20 May 2019 (UTC)

Maybe someone watching this page can chime in at Wikipedia:Articles for deletion/List of lakes named McArthur with an explanation of why set indexes are completely different from disambiguation pages. I do not seem to be having any luck. Aymatth2 (talk) 16:07, 2 September 2019 (UTC)

That discussion was closed as "no consensus", but it was clear that there was confusion over where these lists are appropriate. See below for a proposal. Aymatth2 (talk) 15:40, 14 September 2019 (UTC)

New section: Suitable topics

This is to propose adding a new section on "Suitable topics" that would say something like the below:

Suitable topics

  • Like disambiguation pages, a set index list is not expected to cover a notable topic, and usually will not. Tropical Storm Bill is a valid set index, although it is unlikely that anyone will write a scholarly paper devoted to tropical storms named Bill.
  • A set index list should be potentially useful to readers who are looking for information. "List of men named John" does not pass this test.
  • If the list is likely to grow beyond the size where a reader can easily find the entry they are looking for, it should be broken into smaller lists with narrower scope. "List of lakes called Mud Lake" would be too large to be useful. "Mud Lake (Alaska)" has a more practical scope. But in some cases subheadings may be an alternative to breaking up a list.
  • If some of the entries have articles, or may have articles in the future, the list would normally be considered useful. That is, the list contains things for which readers would expect Wikipedia might have information.
  • Conversely, if it is very unlikely that any of the entries will ever have articles the value of the list as a set index may be questioned. "Cottages named Mon Repos" probably falls into this category.

Any comments, suggestions for improvement, would be welcome. If there are no strong objections, I will go ahead and add the new section. Aymatth2 (talk) 15:40, 14 September 2019 (UTC)

I don't think this is consistent with WP:D. There are two questions that have to be answered for a suitable SIA:
  1. If this SIA were converted to a DAB page, would it be a proper DAB page?
  2. Is this SIA a proper list article?
The problem with the proposed suitability criteria, above, is that it fails the first question. If you had a DAB page that only has redlinks which "may have articles in the future", then it would be deleted.
Another problem is that the suitability criteria, above, aren't much more precise than the tangle of rules at WP:LISTN+WP:LISTPURP+WP:CSC.
Let's start with some concrete examples, and see if we can come up with some guidelines that fit them. There seem to be (at least) five kinds of SIAs:
  1. Thorough list articles, like Dodge Charger or List of ships of the United States Navy named Enterprise. I don't think anyone has a problem with these kind of SIAs, where every entry is notable and corresponds to an WP article. This corresponds to the "every entry is notable" part of WP:CSC.
  2. Dab articles with added statistics or information to assist in disambiguation. For example, List of peaks named Signal or Cypress. These seem somewhat controversial.
  3. Long SIAs with few or no blue links to their topics, e.g., List of lakes named Fish Lake. I suspect these fail WP:INDISCRIMINATE.
  4. Mislabeled DABs, such as Aditus or Baker Hill, with no extra information beyond a dab page.
  5. Mislabeled broad-concept articles, such as Activation.
The question is: what new SIA criteria would definitely include (1), maybe include some or all of (2), exclude (3), and maybe exclude (4) and (5)? I suspect the answer lies in WP:CSC --- either all entries in a SIA should be notable and have blue links, or a SIA should be a short complete list supported by a reliable source.
@Shhhnotsoloud, Bkonrad, RedWolf, DexDor, Narky Blert, Widefox, Doncram, DGG, and Cullen328: What do other editors think? —hike395 (talk)
User:Hike395 forgot to sign the above post so redoing pings @Shhhnotsoloud, Bkonrad, RedWolf, Narky Blert, Widefox, Doncram, DGG, and Cullen328:. DexDor (talk) 21:06, 15 September 2019 (UTC)
The proposal would seem to allow a set-index article to be created about any group of objects, and is not acceptable at all. There is no way that consensus will be changed to allow any editor to create a Wikipedia page about just any collection of mostly or all non-notable items, where the notability of that type of collection, as a collection, is not established. The existing standard requires that any SIA must meet standards for notability of all standalone lists, which is fine as it is. --Doncram (talk) 21:47, 15 September 2019 (UTC)
(1)The criterion for navigational artcles is not notability , but usefulness to the reader. If it's something that someone might rationally look for, it can be justified -- this is the same sort of justification we use for redirects. This is particularly true for geography, because according to the basic policy, WP:PILLARS Wikipedia combines many features of ... gazzeteers. .Since WP is presented in written language, names are usually a rational basis for combining things--people may know a name, but be unsure of any further characteristic to look up. One could make a similar argument about dates, because for anything in the past, even the recent past, dayes are the usual waywe organize information. I. Someone might, for example, know of a Fish Lake, and see if they could find something about it here. The Fish Lake example used here could be converted from class 3 to Class 2 by including at least geographic cooordinates. and perhaps the size.
(2)'Another way of looking at topics in class 3 is that they could link to sections within articles. For the various Fish Lakes, the current articles does give the county , and thus does provide some minimal useful information. It could link to a section on the geography of the country. Unfortunately, few if any of the county articles go into such geographic detail. This is a problem with those articles--they, not individual articles, are the place to include geographic detail. Political subdivision articles in WP are often written much too superficially and mechanically. . DGG ( talk ) 23:56, 15 September 2019 (UTC)
(ec) I think enumerating the different valid types may be useful, with examples yes. Some of those types currently satisfy "Like disambiguation pages, a set index list is not expected to cover a notable topic" and others don't. The bottom line is an SIA sits on the spectrum between dabs and lists (including the extremes), as shown by the simple test - should it be linked to or not? Widefox; talk 00:11, 16 September 2019 (UTC)
In the second bullet, "List of men named John" is used as an example of a set index list that would not be potentially useful to readers who are looking for information. Such an article would be ridiculously long because "John" is such a common name, but presumably that's not the only reason the article wouldn't be useful - because excessive length is covered separately in the fourth bullet. So it would probably be clearer if you used a name that doesn't have this problem. e.g. List of men named Archibald, or List of men named Kalle. However, note that we do have WP:Anthroponymy articles like Archibald (name) and Kalle, which do act as set index articles for people having a particular given name or surname. So... I guess I'm confused about this whole bullet point. Colin M (talk) 17:24, 17 September 2019 (UTC)
Based on the above, a fresh and simplified attempt is given below, with numbered bullets for easier discussion:

Suitable topics

  1. Like a disambiguation page, a set index list is not expected to cover a notable topic, and usually will not. Tropical Storm Bill is a valid set index, although it is unlikely that anyone will write a scholarly paper devoted to tropical storms named Bill.
  2. The list should contain a type of thing for which Wikipedia might have information such as geographic features, vehicles or other types of SIAs as listed above.
  3. A set index list should be potentially useful to readers who are looking for information.
  4. Some of the list entries may link to articles, some may link to sections within articles, and some may just give basic information such as the coordinates and size of a geographical feature.
To DGG's comment, I agree that a list of lakes does not have to link to full articles for each lake, could link to sections in broader articles, and can usefully have entries that do not link to anything, but just give gazetteer-type information like size and location. To say that all the entries must have articles, or else the list should be notable in itself, would rule out a lot of useful lists.
Anyway, any comments on the simpler set of criteria above? Aymatth2 (talk) 21:11, 23 September 2019 (UTC)
@Aymatth2: The question is --- how do allow lists that are useful to our users, per the gazetteer section of the first pillar, while not allowing indiscriminate lists of information? I'm concerned that your second proposal still errs on the side of allowing indiscriminate lists. Notice that WP:NGEO doesn't allow articles on arbitrary geographic features, while your proposal will allow lists of such features.
I'm thinking that we should not set out brand-new criteria of notability for SIAs, but instead put together existing pieces such as WP:LISTN, WP:GNG, and WP:CSC to try to strike a balance between being a gazetteer and forbidding indiscriminant lists. —hike395 (talk) 13:34, 24 September 2019 (UTC)

Combining existing guidelines

Here is a proposal that takes existing guidelines and definitions, combines then to form a subset that I think might satisfy most editors and many readers:

Notability guidelines (v.2)

Per Wikipedia's notability guidelines for lists and established purposes of lists, set index articles should provide both navigational and informational value to our readers. Without navigational value, set index articles risk becoming indiscriminate lists of information. Without informational value, a set index article should be converted to a disambiguation page or be deleted.

A set index article has navigational value if it fulfills one of the following positive criteria:

checkY It lists three or more items that have links to articles or sections
checkY It lists two items that have links, but neither of them is a primary topic, following the hatnote guideline
☒N A set index article with zero or one item with links to article or sections is not allowed

A set index article has informational value when it has a suitable topic (see below) and when it provides information beyond links to wikipedia articles, such as background information, geographical data, images of items, or references to reliable sources.

A topic of a set index article is suitable if it fulfills one of the following positive criteria:

checkY The topic of the set index article has received significant coverage in reliable sources that are independent of the subject, per the general notability guideline
checkY The set index article includes only notable items, per the first common list selection criterion. This criterion can be fulfilled if every item has a Wikipedia article.
checkY The set index article is a short, complete list of every item that is verifiably a member of a group, per the third common list selection criterion. If this criterion is used, the SIA should not be longer than 32K (again, per the third selection criterion).
☒N The second common list selection criterion should not be used to justify a set index article. A set index article should be not an incomplete or large collection of non-notable items.

I think this will allow SIAs that are helpful to users, while suppressing indiscriminate lists and dabs that are misclassified as SIAs.

Please note the interaction between the navigational value of a SIA and the suitable topic criteria. The criterion that prevents an editor from splitting a huge list of items (10,000 lakes named X) into smaller SIAs (100 lakes named X in Scotland) is that each SIA has to link to 2 or 3 WP articles, to be navigational. But if each smaller SIA has enough blue links (e.g., Summit Lake (Alaska)), then it's ok for that to standalone.

Even if you don't agree with my proposal, I think it would be helpful to define SIA notability/suitability based on existing guidelines and terminology.

@Aymatth2, Shhhnotsoloud, Bkonrad, RedWolf, DexDor, Narky Blert, Widefox, Doncram, DGG, and Cullen328: What do other editors think? —hike395 (talk) 14:43, 24 September 2019 (UTC)

@Hike395: I think the above is moving towards a viable definition, but could be simplified as follows (the wording can be tweaked to include links to other guidelines):

Notability guidelines (v3)

  1. A set index is valid if it meets the criteria for a disambiguation page or for a standard list; that is if each list entry links to an article about the entry, or the topic of the list is itself notable
  2. A set index is also valid if it provides both navigational and informational value
    A. Several entries in the list give links to articles or article sections that provide information related to the entry, and
    B. The list provides information beyond simple links to articles, such as background information, geographical data, images of items, or references to reliable sources, and
    C. It is a short (less than 32k), complete list of every item that is verifiably a member of a group
Aymatth2 (talk) 17:05, 24 September 2019 (UTC)
I'm sorry, but these proposals do not adequately address notability of lists, and notability of items, and definition of what information is allowed for each item. Any serious proposal for SIAs must not simply abandon all notability standards for Wikipedia. It is not okay for a few editors here to make up a local consensus that "anything goes", "there is no longer any notability requirement for Wikipedia" or the like, which is essentially going on here. What prompts this discussion right now is a couple editors, me included, being involved in debate about some lists of lakes which, in my opinion:
a) are indiscriminate collections not meeting standards required of standalone lists (i.e. there is not any source found, nor conceivably would any source exist, that would talk about that group as a collection). Where it is inconceivable to some editors, at least including me, that any reader is interested in the collection per se, or would expect that an encyclopedia would have such a collection. Where the list is not a natural split (by some significant/defining/notable characteristic) of any bigger legitimate list. A serious proposal must put forward some standard for what is acceptable as topic of an SIA; these do not. "Having three notable items" is not adequate for that; I am sure there are three notable persons named Jefferson Davis but we do not want to then allow a list of many thousands of persons of that name, even if it is verifiably a complete list of every such occurence in a given U.S. census, say. Nor do we want a list of lakes at elevation between exactly 300 to 301 meters above sea level that are included in a huge lakes database, nor do we want a list of all lakes verifiable within GNIS having some three-character string in their names.
b) have no item-level notability standard at all, i.e. specifically allowing any pond or puddle, no matter how truly insignificant, to be included into a list-article (unlike the names SIAs which seem to require that there be an article for a person to be included, i.e. following the most common standard for list-item notability, that all items are themselves Wikipedia-notable). A serious proposal must set some standard distinguishing themselves from saying "anything goes". About lakes, and about people, there exist millions of them which are not notable and we do not want to include them at all in any lists in Wikipedia. Note it is pretty well established that lakes are generally NOT expected to be notable; we are not a Gazetteer about lakes. For example, what in the standard precludes allowing a list of all persons named in the 1910 U.S. census (broken up any way you want, into hundreds or thousands of smaller lists)?
c) have no standard at all for the material that is allowed into the list-article, for any or all members. Like, for a lake, to include manufactured pseudo-information like "this lake is in a certain climate zone", and "this lake is composed of water" and this lake is in a certain watershed or at a certain elevation, or that "there was a community picnic once at this lake", information which is trivial and non-defining and/or directory-like. Equivalent to opening the floodgates to allow any article about a school, say, to include list of names of teachers, and lists of classroom numbers, and other utterly non-encyclopedic information to be added. Or equivalent to allowing anything at all about any person to be put into a huge section about them, as long as it is contained in a list/SIA along with similar huge sections about other persons.
On a larger level, the proposal must not contradict Wikipedia consensus for schools, say, where it is well-established that Wikipedia is not a Gazetteer about them, and does not want individual articles about every high school in the world, or every middle school, or every elementary school. And it will not be acceptable to a wider community for a small group here to say, "never mind, we now allow anyone to write as much as they want about any elementary school, and put it into Wikipedia, as long as it is packaged in a list/SIA along with the same for several other elementary schools sharing some trivial characteristic". A serious proposal must not allow that.
Another test that a proposal must pass is: how does this not allow general directories of things. List of items in the 1920 Sears-Roebuck catalog? Lists of travel-type information, e.g. places to stay when you visit Mexico?
The existing SIAs about ships, and about names, do not fail all of these standards. These proposals would allow literally anything to be put into wikipedia as far as I can tell. I oppose these proposals because I perceive some to be trying to allow anything and everything in, exercising no judgment. I appreciate Hike395 trying to compose some compromise, but their proposal is not discriminating enough. --Doncram (talk) 19:02, 24 September 2019 (UTC)
The v3 version is actually a tighter standard than the third Wikipedia:Stand-alone lists#Common selection criteria (which it includes as 2C and should link to) since it also requires that at least some of the list entries have links to articles or article sections that give related information, and that the entries give more information than just links. Any list that meets the v3 criteria meets the existing standards for lists. Aymatth2 (talk) 19:49, 24 September 2019 (UTC)
Simply not so. This version v3 (along with Hike395's V2) both make a basic mistake in interpreting Wikipedia:Stand-alone lists: y'all skip all discussion of notability of lists there, and incorrectly assume that any example of any type of list described by "Common selection criteria" is okay. That is not what the guideline says. V2 and V3 both would allow "List of elementary schools with Washington in their name", or "List of elementary schools in Osage County, Oklahoma" or "list of persons named Jefferson Davis as of 1920" or "list of places to stay in Osage County" (split out further if necessary to meet arbitrary 32k size), as long as the list included tons of non-encyclopedic information about each one (so they carried "information") and as long as they included every item in some specific directory (well-defined set) of all hotels, motels, airbnbs, or whatever. Because every elementary school can be mentioned somewhere else, and there exist three historic hotels in Osage county, and so on. It might allow practically any existing directory of anything to be included, fully in wikipedia, though perhaps divided into small pieces. This is non-discriminant about topics, about membership, about content. It is a non-starter as a proposal, and this reply did not even attempt to respond to my post. --Doncram (talk) 20:54, 24 September 2019 (UTC)
While notability is often a criterion for inclusion in overview lists of a broad subject, it may be too stringent for narrower lists; one of the functions of many lists on Wikipedia is providing an avenue for the retention of encyclopedic information that does not warrant separate articles, so common sense is required in establishing criteria for a list. Aymatth2 (talk) 22:20, 24 September 2019 (UTC)

I'd like to make the following points:

  1. I think Doncram is analyzing these criteria outside the definition of a set index article, which is a list article about items of the same type with the same name. Many of those examples would fail to be SIAs under this definition:
    • ☒N List of high schools in the world
    • ☒N List of teachers
    • ☒N List of items in the 1920 Sears-Roebuck catalog
    • ☒N List of places to stay in Mexico
    • ☒N List of lakes with the same three-character string in their name
    • ☒N List of all persons in the 1910 U.S. Census
    • ☒N List of places to stay in Osage County
  2. Similarly, some of the negative examples proposed by Doncram would fail the informational criteria under v2:
    • ☒N List of people named Jefferson Davis --- I would be utterly shocked if you could ever form a complete list of all people named Jefferson Davis. If you could, I think such a list would be much longer than 32k and be against guidelines.
    • ☒N List of people named Jefferson Davis in one census --- I would be surprised if there were 2 or 3 blue links in this case.
  3. Notice that, under v2, every SIA would require one or more reliable sources to create or edit:
  4. the list notability criterion is currently a mess, and doesn't have specificity to exclude inappropriate lists. It doesn't require a list to fulfill the general notability guideline. It refers to WP:LISTPURP, which requires a list to be informational or navigational. The common selection criterion are part of WP:SALAT, which talks about appropriate topics for lists. Thus, building the SIA criteria out of the common selection criteria seems like a good idea.
  5. I'm open to a discussion about what information is appropriate to add to a SIA. But the concern that the v2 proposal will allow junk lists seems misplaced to me.
  6. Let's say this discussion doesn't come to a consensus, and we are going to stick with the current SIA guidelines. Then we are going to have a series of frustrating subjective discussions for every SIA, similar to the McArthur Lake deletion discussion. Right now, the list guidelines are not prescriptive enough to avoid a lengthy discussion that would end with "no consensus to delete". If editors don't like these proposals, how can we strengthen them while still allowing "good" SIAs (like mountains or ships)?

I would love to hear opinions from more editors (if people are interested). —hike395 (talk) 03:19, 25 September 2019 (UTC)

To be clear, I am o.k. with v2, and agree with referring to existing guidelines, but think it can be compressed to something like v3, which could be further compressed to: "Must meet the DAB criteria, or meet GNG for the list topic, or be short, complete and have both navigational and informational value." Aymatth2 (talk) 10:31, 25 September 2019 (UTC)
Hike395, your post seems unfair about me, and perhaps is sarcastic(?), and seems to include some mischaracterizations. For one, your proposal V2 and the other proposal V3 do not limit SIAs to only being about same-named items. I sincerely did think you were proposing that an SIA can be any set of things, not limited in that way. I believed this in part because others have blandly asserted that an SIA is a set of similar things, and in part based of what you wrote in your proposal, and in part because of assertions about "cross-categorization lists" in recent discussions in which you've been involved. "Cross-categorization lists" are generated by the intersections of two bigger lists, e.g. "list of architects" vs. "list of regions" or vs. "list of year-ranges" or whatever. Note any SIA of same-named items of some type of thing can be defined as a "cross-categorization" of a "list of all things of that type" (typically a valid list-article) crossed against "list of things named X" type lists (not usually or not ever a valid list-article). I view "being named X" as a trivial characteristic, similar to elevation, say. This may or may not be your view, but at least some noted that wikipedians have been uncomfortable about cross-categorization lists in general, so asserted that means any cross-categorization list is okay even if they are defined as "things of a type" crossed against any trivial characteristic, and so asserted that means any SIA is okay (e.g. where the SIAs are defined as same-named items or same-elevation items or whatever of that type). It was reasonable for me to think these proposals were allowing for lousy lists such as "List of lakes at elevation 300 meters", "list of lakes at elevation 400 meters" etc. Which happen to seem as trivial to me as "list of lakes with McArthur or similar in their name", "list of lakes with Christine in their name", etc. It is an advancement of discussion, I guess, for you to say you want to limit/revise your proposal. And I guess that change in your proposal would probably take care of the "List of Sears-Roebuck catalog items".
Also it is unfair to suggest I put forth lists of teachers for being SIAs. I did not. I was pointing out that new SIAs about schools might be vehicles for local editors to put in whatever trivial info they wanted in "development" about their school, like including names of all the teachers, and the school colors, and school mascot, and so on. User:Hike395, I have to run, can reply more later. --Doncram (talk) 23:58, 25 September 2019 (UTC)
(edit conflict) Doncram I certainly didn't intend sarcasm: I apologize if it sounded like I was sarcastic. It looks like it's a misunderstanding. I meant what I said --- I thought you may have analyzed these criteria outside the context of the SIA. I can see why you thought that, since it wasn't explicit from the criteria. Both Aymatth2 and I assumed that this was going to be added to the existing WP:SIA guideline, so that the context was implicit.

V.2.2

Suitability guidelines (v.2.2)

Set index articles, as defined above, are list articles whose items are the same type (e.g., ships) and have the same or similar name. As list articles, they should obey Wikipedia's notability guidelines for lists. Set index articles have a stricter set of notability requirements than than general list articles. These stricter requirements are defined in this section.

Per established purposes of lists, a set index articles must serve both a navigational and informational purpose. If a set index article were not navigational, it would risk becoming an indiscriminate list of information. Without an informational purpose, a set index article should be converted to a disambiguation page or be deleted.

A set index article is presumed to have a navigational purpose if it fulfills one of the following positive criteria:

checkY It lists three or more items that have links to articles or sections
checkY It lists two items that have links, but neither of them is a primary topic, following the hatnote guideline
☒N A set index article with zero or one item with links to article or sections is not allowed

A set index article is presumed to have an informational purpose when it has a suitable topic (see below) and when it provides information beyond links to wikipedia articles, such as background information, geographical data, images of items, or references to reliable sources.

A topic of a set index article is suitable if it fulfills one of the following positive criteria:

checkY The topic of the set index article has received significant coverage in reliable sources that are independent of the subject, per the general notability guideline
checkY The set index article includes only notable items, per the first common list selection criterion. This criterion can be fulfilled if every item has a Wikipedia article.
checkY The set index article is a short, complete list of every item that is verifiably a member of a group, per the third common list selection criterion. If this criterion is used, the SIA should not be longer than 32K (again, per the third selection criterion).
☒N The second common list selection criterion should not be used to justify a set index article. A set index article should be not an incomplete or large collection of non-notable items.
Can we start the discussion over? Given v2.1 these guidelines, is it still too forgiving? Or too strict? Or should they be thrown out and replaced with something else?
(Aymatth2: please forgive me not creating a v3.1 --- I'm happy to try to reduce the complexity of v2.1 down to a 3.1, but I'd like to see whether 2.1 is anywhere near acceptable. Feel free to create a 3.1, if you wish) —hike395 (talk) 03:33, 26 September 2019 (UTC)

A few general comments:

  1. We cannot say set indexes have a stricter set of notability requirements than general list articles. If that were the case, editors would just follow the general list rules. But a set index will rarely have been discussed as a group or set by independent reliable sources. It is an example of a "more complex list", where there is a lack of consensus about notability requirements. This is trying to define criteria for a certain type of complex list where the existing guidelines are lacking, and to get consensus.
  2. If a set index has been discussed as a group, then it covers a notable topic and can be kept, but we should be clear that a set index will rarely have been discussed as a group. If it meets the criteria for a disambiguation page, then it is just a DAB on steroids and can be kept. We should make those points, then focus on the SIAs that are not about notable topics and are not DABs.
  3. "Less than 32k" is reasonable, maybe too high.
  4. "Complete" is reasonable. It may be worth spelling out that a source must be given for the list, such as a database. A Google search is not enough. It would be useful to separately give examples of sources for lists of geographical features, ships, trains, insects etc. Sources like geonames.org do not confer notability on the list, but show that it is more than an arbitrary collection.
  5. We do not expect all the Mud Lakes to have articles, or even sections in articles, but some of them must at least have been mentioned in other articles. If an entry on one of the Mud Lakes leads to an article on a municipality or a river basin, that has value even if the target article says little about the lake itself. The reader learns where it is, and learns a bit about that location. But if none of the Mud Lakes has even a section in a broader article, we must question whether the list is valid. I doubt that the existing guidelines are much help here, but think it is crucial. Something like "at least one of the entries must be the topic of a separate article, or discussed in some detail in another article."
  6. It probably goes without saying that all the entries must give more than just the name, since all the names are the same. Still, we can say it, and give examples of the types of information that may be given. Possibly we should say that we require a citation for an entry if it gives more information than is given by the cited source for the list as a whole.
  7. I would prefer to say only what qualifies for a set index, not what does not qualify. If we say the list must be A, B and C, that is clear. If we add that it must not be X, Y or Z, then we introduce a gray area of lists that are not A, B and C, but are not X, Y or Z either.

Aymatth2 (talk) 13:52, 26 September 2019 (UTC)

These are very good points. I fear we have a deeper problem brought up by Doncram. See "Heart of the Matter", below. —hike395 (talk) 11:16, 27 September 2019 (UTC)

about some terms

  • 1. How about SOSNIs? How about dropping the term "set index article" or SIA altogether, and how about SOSNI or "set of same-named items" instead? I think that might be a lot clearer if these proposals are indeed being limited to being about sets of same-named items. SIAs in general have a bad name; there is baggage which maybe could be dropped in a new proposal about SOSNIs now.
Doncram Can we possibly defer renaming until we get consensus on notability? I'm happy to rename if that makes editors happier. —hike395 (talk) 03:49, 26 September 2019 (UTC)
Sure, fine. --Doncram (talk) 19:12, 26 September 2019 (UTC)
  • 2. "Has navigational value" is nebulous. Any wikipedia article or list-article "has navigational value", obviously, because all articles have at least some outgoing wikilinks, and those links help readers about a topic navigate to other topics that may be of interest to them. It is meaningless as a term in defining anything here. Hmm, it is just occuring to me that in proposal V2, maybe the proposer meant to convey that an SIA is a list of same-named things by saying merely that a SIA "has navigational value" (because we're supposed to understand that means it is a list of same-named things?). Is that what was meant? Then say so, please.
  • 3. "Has information value" is also pretty useless as a term for defining what a SIA (or SOSNI) is. I suppose that "has information value" means that there is something, anything at all beyond names of the items, for at least some of the items. E.g. coordinates for all the places in a list, or random factoids about one or another item in a list. Note that every disambiguation page "has information value" because while it lists same-named items (e.g. Smith House includes a number of places named exactly that, plus similarly named others), it also includes additional information (such as city-state locations). There are disambiguation-rules-enforcing editors who seem to be devoted to stripping out any extra information, beyond some minimum which they deem sufficient for "navigational purposes", but even they must allow some "information" beyond the names of items to be kept, to distinguish between the items. Maybe you want to say that a SIA or SOSNI is different than a DAB, because a SIA/SOSNI is given permission to include additional information beyond the minimum allowed in a DAB? And then you need to say something about exactly what type of additional information is allowed and why (and hopefully this will rule out trivial or non-encyclopedic "information" like "this lake is composed of water" or "this lake is in the Mediterranean climate zone" or "this school's principal is Hosmer Lemeshow"). --Doncram (talk) 02:43, 26 September 2019 (UTC)
"value" was my terminology, I should have used "purpose", which is the exact wording from WP:LISTN and WP:LISTPURP. My intent was to define whether a SIA or SOSNI had informational or navigational value by establishing criteria for those. Let me rewrite v2.1 into v2.2 (above) to use "purpose".
I'm not sure how to define what per-item information is acceptable in a SIA, especially in terms of existing guidelines. Let me poke around and think about it. —hike395 (talk) 03:49, 26 September 2019 (UTC)
I don't see any way to define per-item information in a general sense. The appropriate information depends on what kind of "set" it is. For mountains, that information would be location (lat/long, political subdivisions). For ships it could include type of ship (frigate/battleship/etc.), country, various dates (comission/decomission, ordered, laid down, launched), notable battles/events the ship is associated with. For a common name shared by several plants, information could include geographic range, taxonomic family, flower color or photos. Information appropriate for one set isn't going to be appropriate for another (and information which may be relevant for a set type overall should be omitted if doesn't differentiate between members of a particular SIA/SOSNI (i.e., no point in stating that ships are frigates or plants have white flowers if that is true of all listed entries on one page)).Plantdrew (talk) 19:31, 26 September 2019 (UTC)
Maybe I accidentally stumbled over a general sense of appropriate per-item information above. One point is that included information should differentiate at least one item in the SIA from other items. Second point (not sure how to phrase it well) is that the information should have the potential to help a reader with a passing familiarity with one of the items who wishes to learn more about that item identify the item they are interested in (such a reader might know that a ship served in WWII, or that a plant grows near where they live, but won't know the ships pennant number or the scientific name of the plant (where pennant number/scientific name serve to provide unique titles that somebody more familiar with the subject might recognize)). Plantdrew (talk) 21:20, 27 September 2019 (UTC)
I am moving towards the idea that we should have three new sections in this guideline, which can be debated and evolve separately:
==Criteria==
Properties of a list that are necessary or sufficient. I see three types of list: DAB+, Notable topic and Other (there must be a better name).
==Appropriate information==
General suggestions on what does and does not belong, and specific examples for common types: ships, lakes, plants
==Relevant guidelines==
List of existing guidelines that are directly relevant, with a link, name and short outline of what it is about, but not the words, which may change
Does anyone have views on how to discuss three different threads? Maybe one by one? Aymatth2 (talk) 22:19, 27 September 2019 (UTC)

List of places named Fire Station Number 10

Hike395 asked elsewhere for me to give examples of list-articles that met criteria proposed here, yet would be bad list-articles. Here is one: Talk:Fire Station No. 10/Proposed set index article. It was composed by running a report out of the U.S. GNIS system, finding all occurences of "Fire Station 10", and then dropping out some which had the number "100" or "101" or otherwise did not seem to be closely enough named to be part of a "same-named set". Perhaps "in the United States" should be added to its name, because all members are. It meets (or could be adjusted slightly to meet) proposed criteria:

  • It is a set of same-named items. About 418 items are in the set. People could quibble, leading to a few deletions. Some clarifying edits would be needed to fix up some items more properly, e.g. for item #227, "Beaver Springs Fire Company" is the first/primary GNIS name; the article should clarify GNIS reports it is also known as "Beaver Springs Volunteer Fire Company Station 10".
  • There are three or more items of that name having articles. It seems a total of 5 of these are in fact notable places; 413 are obscure (418 numbered ones less 5 probable duplicates) and are not likely to ever have Wikipedia articles.
  • It is a purportedly complete set, in that it is all the hits in GNIS. If someone can point out another "fire station 10" exists, then that can be added, and it will again be a purportedly complete set.
  • It is "informative" in that it has city-state locations. And has GNIS numbers, though not properly labelled (and not that GNIS numbers should be reported). It easily could have coordinates for every one, but i stripped the coordinates out of the download. Some random facts could easily be added about any one of the five notable members, and probably some random facts could be added for several/many of the non-notable ones.
  • It is less than 32k in size. Or it is very close. Note some top disclaimer stuff could be dropped, and several more items could/should probably be dropped for not being close enough name-matches.

Note if coordinates were added back in, for all, or if there was much other development, then it would be over 32K. In which case it could be split into two or more list-articles, say "List of places named Fire Station 10 east of the Mississippi" vs. same "west of the Mississippi". If it were split, then at least one part would temporarily not have 3 members with separate articles. Easily remedied: just manufacture a few separate articles explicitly just for the purpose of defending the list, but pretend otherwise, and plan to engage in battleground behavior if anyone else seeks to delete those. Under the proposed standards, it would take extraordinary effort and still probably fail, for anyone to dispute the list-articles. --Doncram (talk) 19:12, 26 September 2019 (UTC)

That is a useful example. So far not many lists like that have been started, and It is hard to see how this meets any of the standard purposes of lists. Common sense can apply in a deletion debate, and this list would be unlikely to survive a challenge. Still, the guidance here could say that there should generally be no more than 100 entries, and a reasonable number of the entries should link to descriptions in other articles. "Generally" and "reasonable" leave some room for debate. We could add that lists over 10 entries should be organized so a reader can readily find an entry they are looking for. Aymatth2 (talk) 19:58, 26 September 2019 (UTC)

The heart of the matter

What information is OK to add to a DAB to turn it into a SIA?

Doncram made an insightful comment (above): "because a SIA [...] is given permission to include additional information beyond the minimum allowed in a DABs [...] exactly what type of additional information is allowed and why (and hopefully this will rule out trivial or non-encyclopedic 'information')", and made an example of a SIA that would follow the v2 criteria ("List of fire stations named 10"), but clearly is a junk list. Plantdrew also makes a good point that we can't specify, in general, what information is beneficial to add to a DAB, because it depends on the type of object.

I browsed the guidelines, and the only relevant one I found was the guideline against indiscriminate lists. Doncram's example of "List of fire stations named 10" clearly and obviously violates WP:INDISCRIMINATE. However, applying WP:INDISCRIMINATE is going to be a matter of judgement and discussion per case. My hope of finding an objective set of criteria for acceptable SIAs may be a fool's errand --- it's always going to be a matter of discussion.

I have some suggestions that may make things a bit better:

  1. Let's explicitly add WP:INDISCRIMINATE to the SIA guidelines, saying that SIAs should not be indiscriminate lists of statistics or data.
  2. In the future, at AfD or at article talk pages, it could be helpful to use WP:INDISCRIMINATE to argue against bad SIAs, rather than notability (because list notability is such a mess).

I'm going on a short wikibreak. Maybe other editors can think of something better while I'm gone? —hike395 (talk) 11:16, 27 September 2019 (UTC)

  • Let's not give up. I am confident that it is possible to define objective criteria that will help with most set index discussions – most articles will clearly meet the criteria or will clearly not. There will always be borderline cases. We should reference and be consistent with existing guidelines, but we need to be bold and add more guidance. The set indexes are examples of complex lists for which the present guidelines are insufficient.
Notability is rarely relevant to set indexes, which resemble disambiguation lists in that respect. It is very unlikely that anyone will ever write about "Lakes in Quebec named Lac Bourgeois", but it is quite likely that a reader will look for information on a lake with that name. A set index can give information such as municipality, river basin and coordinates, and where possible link to articles that hold other relevant information. There is value to the list even though the overall list subject is not notable.
We can mention WP:INDISCRIMINATE, but it is vague and none of the examples resemble set indexes. I see nothing wrong with a set index table having a column called "Miscellaneous" that holds more information about each entry where available, citing sources. Perhaps it says "Named after Fred McIntyre, a surveyor who measured it in 1854". Some would say that is indiscriminate trivia, and others would say it is relevant information. It may be impossible to get consensus.
I suggest we add not one but two new sections: "Relevant guidelines" and "Set index criteria". The first can briefly say how each guideline applies (or does not apply) and the second can summarize rules for the three types of set index: disambiguation, notable topic and hybrid. I will give drafts below. They can be discussed and implemented independently. Aymatth2 (talk) 13:15, 27 September 2019 (UTC)
Summary: they are either 1. dab+ or 2. list-. The + is superfluous additional info/entries evading the strictness and maintainability of MOSDAB so arguably bad, and the - is a weakness such as a) notability, b) inclusion criteria/scope incorrectly based on dictionary rather than encyclopedia viewpoint or c) a catchall for useful long lists such as names. The WP:SIA example List of peaks named Signal is 2.b as shown by the existence of Signal Hill (disambiguation) (where entries such as Signal Hill, St. John's aren't included in the SIA) (and Signal Peak (disambiguation) -> Signal Mountain). Widefox; talk 09:06, 29 September 2019 (UTC)
If set indexes were either disambiguation pages or simple lists, there would be no need for this guideline. But set indexes are not disambiguation pages: they only include items of the one type, may include items that do not have articles, and may provide links to articles other than the main list item topics. And set indexes are not simple lists: the list topic will rarely be notable, since all the items have in common is that they are the same kind of thing with the same or similar name. Set indexes are complex lists that help readers find encyclopedic information on topics that may or may not have stand-alone articles. Aymatth2 (talk) 01:38, 3 October 2019 (UTC)
@Aymatth2: I think you've described it correctly: SIAs are like super-DABs with extra information to help the user navigate. That was the original intent of Mountain SIAs from RedWolf and myself. But other editors have raised good points: "When can editors add more information to a DAB to turn it into a SIA?" Right now, the only effective limit is that a DAB can be turned into a SIA if all of the items are of the same type. "What sort of information is acceptable to convert a DAB to a SIA?" remains unanswered, except for applying WP:INDISCRIMINATE.
In the original discussion that established SIAs, there was a tacit understanding that individual WikiProjects would define limits and guidelines to their corresponding SIAs. In the 12 years that have followed, this seems to have broken down. In retrospect, that wasn't a good idea. But is there anything else to do other than apply WP:INDISCRIMINATE? I still don't have any good ideas.hike395 (talk) 13:38, 3 October 2019 (UTC)
Later --- one possible idea is further restrict v2.* to have SIAs only contain items that are notable in themselves (e.g., obeying WP:GEOLAND) or have blue links. This would be similar to WP:LISTOFLISTS. This would restrict the number of rows of a SIA and prevent arbitrary database dumps. It wouldn't control the content of SIAs, but if the number of rows is restricted, then the potential for junk SIAs goes way down. —hike395 (talk) 13:47, 3 October 2019 (UTC)
An SIA only holds information about one kind of thing, can include entries that do not have stand-alone articles, and can contain more information and links than a DAB. It serves one of the common purposes of lists, which is to provide encyclopedic information on topics that may not be notable in themselves and do not warrant stand-alone articles. It also allows links to related articles, so an entry for a lake that has no stand-alone article could link to articles on the county or municipality that holds the lake, and the river that drains it. This is helpful to readers. It is related to a DAB, but I see it as a different animal.
We may be over-concerned about a problem that rarely comes up: a junk set index. If a set index is obviously useless, there should be consensus in an AfD that it should go whatever the guidelines say. Anyway, I propose to break this proposal down into simpler units, which can be discussed separately:
==Related policies and guidelines== Shortcut, name, brief description of what each policy or guideline is about. Not the full text, which may change
==Typical information== Examples of what might be included for different types of thing. User:Plantdrew gave a good start for this. Should also mention WP:INDISCRIMINATE and include a 32k size limit. I would prefer a much smaller recommended limit, but that can be discussed later.
==Criteria== This the tough one, and can be left to last. A basic criterion would be something like "useful to readers looking for information", which should eliminate junk lists. It should say that the overall list topic does not have to be notable, since it is a complex list, but the list items should be the kind of thing a user would think Wikipedia might have information about. A good test of that is some of the items should have articles.
The first two do not seem controversial. If there are no objections, I will add first versions to the guideline. Aymatth2 (talk) 14:18, 3 October 2019 (UTC)

Could you create a draft of the new proposal(s), perhaps at WT:SIA/Draft, so that editors can have something concrete to discuss? —hike395 (talk) 07:53, 4 October 2019 (UTC)

See below. I would prefer to try for consensus on these two before revisiting the question of criteria. Aymatth2 (talk) 14:38, 4 October 2019 (UTC)
My reaction: No way in hell. Yes there are objections. Sorry, I don't mean to be disrespectful, but some assertions here seem to be claiming victory and acting as if the general policies and guidelines and consensus of Wikipedia don't matter. And trying to proceed in a boring, bureaucratic, banal kind of way, to try to change everything. In fact, the existing guideline and consensus is that SIAs are not exempt, that they must meet notability guidelines for all standalone lists. You are allowed to make some other proposal, but IMHO it is disrespectful and unproductive to just assert victory and proceed into wordsmithing, when there is NOT agreement.
I think it was Ronald Reagan, in debate prior to his election as U.S. president, who used the term There you go again... vs. whatever the Democratic candidate had just said. Reagan was being patronizing and dismissive in a way which was infuriating to some and delightful to others. Here, I experience the combo of feeling mildly "infuriated" that the general discussion is going completely backwards, and/or I feeling amused or bored or whatever. And I want to comment "there u go again" about repetition going on now.
It is asserted just above, that An SIA only holds information about one kind of thing, can include entries that do not have stand-alone articles, and can contain more information and links than a DAB. It serves one of the common purposes of lists, which is to provide encyclopedic information on topics that may not be notable in themselves and do not warrant stand-alone articles. It also allows links to related articles, so an entry for a lake that has no stand-alone article could link to articles on the county or municipality that holds the lake, and the river that drains it. This is helpful to readers. NO WAY IN HELL. This is bland assertion that, in effect, there is no standard. That anyone can put any list of items into Wikipedia, with no control. Whether or not they are same-named items. It is assertion, yet again, that the entire list of items in the Sears-Roebuck catalog can be put into Wikipedia. That yany horrible set, including "lakes named McCarthur" or "places named Fire Station 10", or the set of ornaments on my neighbors lawn, or anything at all, can be put into Wikipedia.
It is also asserted just above that "We may be over-concerned about a problem that rarely comes up: a junk set index." Not so. In fact we are here these days exactly because one or a few editors recently started creating junk set index articles and putting them into Wikipedia, one salient example being "list of lakes named mcArthur". And asserting privilege to create more junk set index articles like "list of lakes named Jumbo". Along with junk articles about obviously non-notable lakes. Along with junk content like "this lake is in the sub-arctic climate zone", "this lake is made of water", as if that established the legitimacy of one pond among hundreds of thousands in its general region. Along with obtuse, obstinate arguing in a wp:Battleground way, not making any concessions, etc. And they temporarily "won" one or a couple AFDs because of confusion among editors about what the SIA guideline actually does say, or because of bad judgment by the closers of the AFDs, or whatever. Another one or two related AFDs were closed correctly, and the SIA guideline does in fact remain in place, outlawing the junk set index articles that have been pushed forth. --Doncram (talk) 15:51, 4 October 2019 (UTC)

Renewed or new proposal by Aymatth2

In this edit i create this new header and subindent two sections created by Aymatth2 below. Please allow this formatting, which I make with intention of furthering discussion. It is otherwise getting too unwieldy. I am not supporting this proposal or set of two proposals or whatever, I am just trying to set them aside. It is premature in the general discussion going on, on this page, to be writing out that kind of detail, as if you won some battle and are just going to proceed. NO WAY IN HELL. Note, Hike395 suggested you put your proposal in a separate page, and I think they and others would think the stuff below is too much detail, here, and not reflective of any new consensus, and premature. In fact I would prefer, and suggest, that the now-indented subsections be closed/hidden, too. --Doncram (talk) 15:51, 4 October 2019 (UTC)

@Doncram:. The two proposals below are independent of the discussion on notability, and independent of each other.

  • The first is to add a "see also" type of section pointing to related policies and guidelines.
  • The second is to add a "how to" section, giving advice on what should and should not be included in lists of this nature.

Please leave them as separate items that can be discussed, resolved and closed independently, and do not disrupt rational discussion. They can probably use wordsmithing, but do not seem controversial. Why would we not help editors find relevant guidelines? Why would we not give them advice on the contents of a set index? Aymatth2 (talk) 12:33, 6 October 2019 (UTC)

A draft new section follows, listing all the guidelines etc. mentioned in the previous discussion. It may help editors navigate through the maze of concepts.

Related policies and guidelines

A summary of related policies and guidelines is given below. Editors should ensure that any set index article is compatible with these policies and guidelines. Refer to the current versions of the policies and guidelines for details.

Disambiguation
WP:DAB Wikipedia:Disambiguation How to give articles unique names, link to the right article, ensure readers can quickly find what they are looking for
WP:PRIMARYTOPIC Wikipedia:Disambiguation#Is there a primary topic? How to choose whether a term should be the title of an article or of a disambiguation page
WP:MULTIDABS Wikipedia:Disambiguation#Disambiguation page or hatnotes? When to use a disambiguation page versus a hatnote in the article on the primary topic
Lists
WP:LISTPURP Wikipedia:Manual of Style/Lists#Purposes of lists Lists may provide information, support navigation or support Wikipedia development
WP:SAL Wikipedia:Stand-alone lists Guidelines for content, style and names of list articles
WP:SALAT Wikipedia:Stand-alone lists#Appropriate topics for lists Lists should not be too general, too broad, too specific, or contain only trivial, non-encyclopedic content.
Notability
WP:GNG Wikipedia:Notability#General notability guideline If a topic has received significant coverage in reliable independent sources it is suitable for an article or list.
WP:LISTN Wikipedia:Notability#Stand-alone lists A list is notable if it has been discussed as a group by reliable sources. Lists that provide information or help navigation are also often kept.
WP:CSC Wikipedia:Stand-alone lists#Common selection criteria Common criteria are that every entry is notable, or the list is short and lists every member of the group. Lists where no entry is notable are rarely appropriate.
WP:NGEO Wikipedia:Notability (geographic features) Gives guidance on notability and holding information about geographic features in Wikipedia.
General
WP:5P1 Wikipedia:Five pillars 1. Wikipedia is an encyclopedia Wikipedia combines features of encyclopedias, almanacs, and gazetteers.
WP:INDISCRIMINATE Wikipedia:What_Wikipedia is not#Wikipedia is not an indiscriminate collection of information Data should be put in context and explained with references to independent sources. Gives examples of inappropriate articles.
WP:MOSLINK Wikipedia:Manual of Style/Linking Guidelines as to when links should and should not be used, and how to format links.
WP:RS Wikipedia:Reliable sources Articles should be based on reliable, published sources.

All comments welcome. Aymatth2 (talk) 13:47, 4 October 2019 (UTC)

This passage seems to me to be premature and unhelpful. Looks like putting forth a panoply of guidelines, for arriving editors to pick and choose among as they proceed in doing whatever the hell they want. Not helpful. What is needed on this Talk page is general discussion about role of SIAs, how they are to be defined or perhaps abolished because there is no way to define what is allowed different than the existing guideline that notability as for Standalone lists is required. Not helpful to go into wordsmithing for what is presented in the guideline. There is no consensus to change the guideline towards allowing everything and anything into Wikipedia. I think this section should be closed. --Doncram (talk) 17:05, 4 October 2019 (UTC)
@Doncram: Based on the complexity of the discussions on this talk page, an index of relevant guidelines is sorely needed. If you feel that guidelines for set indexes in general are not appropriate, you could raise that in a deletion debate, perhaps at Wikipedia:Miscellany for deletion. Aymatth2 (talk) 12:33, 6 October 2019 (UTC)

I think having links to relevant existing guidelines would help editors a lot. Note that the third relevant guideline at WP:CSC sadly doesn't currently say "Avoid lists where no entry is notable.", but the opposite. I originally thought that was the source of the disagreement (but now I understand better). Unless you are thinking of making the third column containing material of how these other guidelines apply to SIAs? If so, that should be explicitly labeled, and I will re-read this section with that understanding. —hike395 (talk) 14:32, 6 October 2019 (UTC)

The second point in WP:CSC strongly discourages lists where all the entries fail the notability criteria, summarized as "Avoid lists where no entry is notable." We can take that as applying to short, complete lists as defined in the third point: they should contain some notable entries. At this stage I prefer to keep the third column as just a very short description of the guideline. The wording of the guidelines will evolve, so anything given here should be so basic it is very unlikely to change. Aymatth2 (talk) 15:03, 6 October 2019 (UTC)
@Aymatth2: Interesting. I typed "third" but meant "second", above. But your response is interesting, because you interpreted the second criterion as "avoid this", while I thought everything in WP:CSC was explicitly permitted, despite caveats. To me, the list notability and topic guidelines seem frustratingly vague and self-contradictory --- "well, you could do that, but that doesn't seem like a good idea, but it's up to you, but don't do it for lists of people". I guess my point is that the second WP:CSC might not mean "avoid lists with no notable items". At this point, I'm terribly confused by WP:SALAT and WP:LSC. —hike395 (talk) 15:27, 6 October 2019 (UTC)
WP:CSC does not absolutely prohibit lists with no notable items, but strongly recommends putting them into articles on the parent. I have the tweaked the wording. When we get round to developing notability criteria for set indexes, we can be more forceful. That is a separate discussion. Aymatth2 (talk) 16:10, 6 October 2019 (UTC)

Typical information

A draft new section follows discussing the type of information that may be held in a set index article, based on suggestions by User:Plantdrew and others.

Typical information

  1. The information given by a set index article will depend on the specific type of items listed. For example:
    • For mountains, that information may include latitude and longitude, height, range and political subdivisions.
    • For ships it could include type of ship, country, various dates (commission/decommission, ordered, laid down, launched), notable battles or events the ship is associated with.
    • For a common name shared by several plants, information could include geographic range, taxonomic family, flower color or photos.
  2. Refer to the categories listed above (e.g. Category:Set indices on storms) for examples of the type of information that may be appropriate.
  3. Link to the article or article section on the subject of each entry, if there is one, and to related articles, e.g. political subdivisions, battles, taxonomic families. E.g
  4. Information need not be repeated if it is the same for all entries. There is no point in stating that ships are frigates or plants have white flowers if that is clearly true of all the entries.
  5. There should be enough information on each item to differentiate it from the other items.
  6. The information should have the potential to help a reader with a passing familiarity with an item to identify the one they are interested in. Such a reader might know that a ship served in World War II, or that a plant grows near where they live, but will not know the ship's pennant number or the scientific name of the plant.
  7. As discussed in Wikipedia:What Wikipedia is not#Wikipedia is not an indiscriminate collection of information, data should be put in context and explained with references to independent sources. The set index should not contain a mass of minute and unexplained detail.
  8. As discussed in Wikipedia:Stand-alone lists#Common selection criteria, if the list includes every member of the group, it should be less than 32K in length.
  9. Long lists should be structured so that a reader can readily find the item they are interested in. Alphabetical sequence, subheadings and sortable tables may be useful
  10. The introduction to a list that contains every member of the group should identify the source(s) for the complete list, which may be online databases, gazetteers, etc. Results of a general web search are not adequate.
  11. List items do not require citations if they only give information provided by the source(s) cited in the introduction to the list. If an item gives more information, that should be backed up by citations.

All comments welcome. Aymatth2 (talk) 14:36, 4 October 2019 (UTC)

A couple comments:
  • I don't find number 3 (Link to the article on the subject of the entry, if there is one) clear. I think an example would help.
  • Numbers 10 and 11 seem, at best, overly narrow. 10 says The introduction to the list should identify the source(s) for the list. It seems plausible that a SIA might have a wide range of sources, in which case it would be preferable to attach one or more inline citations to each individual row (possibly in a dedicated "Source" column), and there would be no canonical source to cite in the introduction to the list. Instead of saying "you must use this citation style", it might be better to frame it as an optional recommendation, i.e. "The information must be verifiable. Here's a citation style that's commonly used in set index articles that rely on a small number of sources (e.g. online databses, gezetteers)..."
Colin M (talk) 15:26, 4 October 2019 (UTC)
@Colin M: I have given an example for #3, and have clarified #10 to say it applies to complete lists. WP:CSC is a bit ambiguous on this, but seems to imply that a list must either be notable in itself or complete as proved by a source. "Lakes in Quebec called Bourgeois" is unlikely to be a notable topic, but this source can be taken as a reliable basis for a complete list. The idea, I think, is to prevent random selections of entries. With most types of thing there probably are reasonably convincing source databases. Aymatth2 (talk) 16:04, 4 October 2019 (UTC)
Re #3, Ah, I think I get it now. I thought it was talking about the set index article as a whole, rather than each individual list entry. I think just replacing "the entry" with "each entry" would make it a lot clearer (and remove the need for the example).
Re #10: So taking an arbitrary example, would you say List of physicians named Apollonius should not exist based on the current sourcing? For scientific topics (geographic features, astronomical objects, storms), it may be easy to find a database that provides a canonical list of all appropriate entries, but I think this is rarely true for other domains for which SIAs currently exist, e.g. Category:Set indices on comics, Category:Set indices on sports, Category:Set indices on mathematics, etc. Would you delete all or most of the articles in these categories (assuming no single source exists that provides an exhaustive list)? Colin M (talk) 17:13, 4 October 2019 (UTC)
Re #3, I have tweaked the wording but left the example.
Re #10, This discussion probably belongs in a yet-to-be-written section on notability criteria... A list is o.k. if all the entries are notable, whether or not they have articles yet, or if the list is complete, or if the list is notable as a whole. List of physicians named Apollonius may just squeeze in under the first rule, and some medical encyclopedia may have made a list that could be cited for the second or third rule. I doubt Apollonius is at risk of deletion. There are sports databases, databases of comics and mathematical encyclopedias that can be cited. What is ruled out are lists where a few items are notable and the rest is a random sample of items that are not notable. Aymatth2 (talk) 18:45, 4 October 2019 (UTC)
Sounds reasonable. Colin M (talk) 21:29, 4 October 2019 (UTC)
This detailed discussion about wording is not helpful IMO, is way out of line with existing established consensus, and is asserting stuff which would effectively allow any shite at all into Wikpedia. Premature to be discussing exact wording. See my last comment in general discussion which was going on above. The consensus has not changed. IMO this section should be closed/hidden/stopped. --Doncram (talk) 15:55, 4 October 2019 (UTC)
You raise the concern of "allowing any shite at all into Wikipedia", but the draft in this subsection isn't really about what topics are appropriate for set index articles (i.e. notability). Rather the intention seems to be "if you've identified a topic that's appropriate for a set index article, here's how to go about writing it". The advice also seems to be in line with established practice in existing SIAs (except perhaps #10, which I commented on above), and consistent with core content policies. Also, in the previous discussion, hike395 specifically requested that Aymatth2 create a draft of their proposal(s) "so that editors can have something concrete to discuss", so I don't see this as premature or out-of-line. Colin M (talk) 18:07, 4 October 2019 (UTC)
@Doncram: I hear your concern about junk SIAs. I think a discussion here will help us create a clearer guideline. Conversely. shutting down the discussion will leave us with the current SIA guideline, which you've convinced me is inadequate. Judging from your responses, you seem angry about this topic. I'm concern that this anger may discourage editors from expressing their ideas. Let's stay WP:CALM and open up the discussion to many involved editors. —hike395 (talk) 15:09, 6 October 2019 (UTC)

@Aymatth2: I suspect that this section may not be as uncontroversial as you may think. Think of a SIA like a table, where there are guidelines for acceptable rows (which items should be covered?) and columns (what data should be added to each item?). I used to think that the controversy with SIAs revolved around acceptable rows, but now I'm thinking it revolves around acceptable columns (see above). What you've written is good guidance (IMO), but I'm still struggling with the DAB vs. SIA distinction. DABs are good because they are minimal, and let readers get to articles they want to quickly. SIAs might be good, because they still have DAB-like functionality, but also give more data to help users decide which articles to read and also give more context by comparing items of the same type. But how do we capture that as guidance? —hike395 (talk)

To me the key difference is that a set index may be a complete list with some items that may never have articles. That needs discussion, and belongs in the to-be-written section on notability criteria. This section gives guidance on what information may be given, sourcing etc., assuming the list is legitimate. Aymatth2 (talk) 16:01, 6 October 2019 (UTC)
While SIAs can contain redlinks, I think we need to address the fundamental question: when is a SIA better than a DAB for a given list of items?. Otherwise, the guideline seems woefully incomplete. And that is related to what information may be given, because the difference between a DAB and a SIA is precisely that extra information. That's why I'm bringing it up here.
Part of the problem is a tension in Wikipedia. Most articles (including lists) are encouraged to be very complete, with many details given. (If you look at featured articles or lists, they are judged harshly if they leave out non-trivial information.) In contrast to other pages, DABs are strictly regulated to be minimal. SIAs are halfway between DABs and list articles --- and so they are pulled in two contradictory directions: should they be minimal to make navigation very easy; or should they have a lot of details, to be good articles? Or should there be some other guidance (e.g, information that helps disambiguation?). I think this proposal should address this somehow: I'm hoping we will come up with a good way to answer this. —hike395 (talk) 16:25, 6 October 2019 (UTC)
I would say that a set index is better than a DAB when some of the entries do not have articles but may be relevant to readers. That can be discussed in the section on notability criteria. Point #9 (above) says a set index should be organized for easy navigation, which is true of any list but bears repeating. Points #5 and #6 also talk about navigation within the list. Most existing set indexes are compact, both horizontally and vertically, but there would be nothing wrong with a more verbose set index that becomes a featured list.
Let's focus on getting some basic points accepted and published. Perfect is the enemy of good. We can always add refinements later. Aymatth2 (talk) 19:21, 6 October 2019 (UTC)

Simple solution?

I have hesitated jumping into this discussion as it gets harder to keep in mind the majority of points made on either side of the debate. I think there are two primary issues at play here.

  1. The first is when to use dab vs SIA and what criteria should be used to make that decision. I like the SIA format when it comes to listing a modest number (perhaps 10-20?) of mountains that have the same names but where the majority of the links would be red (and probably for a long time). Nobody (myself included) seems too keen on creating the articles for the red-linked mountains listed in Granite Peak which I had created back in 2004 although it was one of my first attempts at a dab page which would later become an SIA. Since tables and references are not supposed to appear on dab pages, the SIA was a way to organize some source information so that it would help future editors in expunging those red links and eventually moving 'Granite Peak' back to a dab. Then there's the List of peaks named Bear Mountain with 144 peaks by that name listed in GNIS. That would make a rather awkward dab page. However, if we only list those mountains that have existing articles, the list would be down to just a handful.
  2. The second is the need for SIA in the first place as it pertains to notability and article linking. Docncram makes some good points, especially with the schools example. Notability can be highly subjective and as such, editor disagreements are not uncommon as a consequence. Going back to List of peaks named Bear Mountain, does this page meet current notability criteria as defined by the policy pages listed by others here? If criteria met, does this now permit others to create list articles on other geographical entities with the same name? If not, why do we still allow the List of peaks named Bear Mountain to remain? While we can easily create all the current red-linked articles based on GNIS information, only a few of these new articles would probably ever be expanded beyond the basic GNIS information. Are each of the 144 Bear Mountains notable enough to have their own individual article based simply on a reformatting of information contained in GNIS? I say no to this and to the other red-linked Granite Peaks.

If Wikipedia really wants to remove SIA (although I'm not of that opinion as yet), one "simple solution" is that the existing SIA information (e.g. tables and references) could be moved to the talk page and a note made on the converted dab page informing editors to first review the talk page if they would like to make changes such as linking to a new article page for a place listed on the talk page. If people want to make lists of mountains, hills, rivers, lakes, etc. with the same name they can simply create a separate talk sub-page for that list (e.g. Talk:List of mountains/List of mountains named Bear Mountain) and link to it from the main talk page. I don't believe any of the existing policies would prevent that. Do not take this as me being in full agreement with one side of the debate, I'm just offering a solution that might get us through the apparent impasse. On occasion, some people have reminded me of KISS. RedWolf (talk) 04:19, 8 October 2019 (UTC)

  • Let's not rush into anything radical. The discussion so far has been extremely productive, raising many useful points. I see progress towards adding sections that discuss the contents of a set index article and that list the relevant guidelines. Those guidelines, as User:Hike395 pointed out quite early in this discussion, may be largely sufficient. When we get round to a section on "Selection criteria" I think we should define three classes of set index article:
    • Extended disambiguation: All the entries are notable, but some do not yet have complete articles. That is, WP:CSC type 1
    • Short and complete: The list verifiably holds all items of a given type and name, and at least some are notable. That is, WP:CSC type 3
    • Notable: The list itself is notable. That is, WP:GNG
I can put together a draft for a section like that, which I think should be fairly simple and solidly rooted in widely accepted guidelines, but would prefer to get agreement on the contents and standards sections first. Aymatth2 (talk) 12:56, 8 October 2019 (UTC)
@Aymatth2: This comment is really helpful for me! Now I think there are two different types of SIA, perhaps with two different forms of selection/notability criteria and also maybe two different criteria of which information to add.
The first kind of SIA is exemplified by Dodge Charger or List of ships of the United States Navy named Enterprise. This is a list article that happens to be about items of the same name and same type. In general, the more information we add to the article, the better the article becomes (acknowledging WP:SUMMARYSTYLE, where a very long article can be split into sub-articles, which is what happened at Dodge Charger). This kind of SIA would be notable because of WP:GNG or WP:CSC1 (all items are notable/bluelinked). These are "wide" SIAs, where there is a lot of information about not-very-many items.
The second kind of SIA is exemplified by Cypress or Granite Peak. These are much more navigational in purpose. They may have redlinks. They should only have extra information that is helpful for navigation. That is, images or geographical data to help distinguish the items. Adding more information beyond what's helpful for navigation is not a good idea here. We need to avoid WP:INDISCRIMINATE lists. If the ratio of redlinks to bluelinks gets too high (e.g, more than 10:1 ?), then the SIA loses its navigational value. These are "tall" SIAS, where the is a small amount of navigationally-relevant information about a lot of items.
I think the guidelines will become much easier to think about if we don't try to mash both of these styles into one overall guideline.
It seems to me that "wide" informational SIAs are generally uncontroversial. They're normal list articles, using WP:GNG or WP:CSC1 notability guidelines. They can potentially become featured lists.
Much of the concern seems to centered around navigational "tall" SIAs. Here's one test: a reader is thinking of an item X (e.g., a peak named "Granite Peak"). They want to know whether WP has an article about X, and if so, which one it is. Would such a reader find a specific tall SIA more useful than a DAB? I think for Cypress, the answer is "yes". For Doncram's list of fire stations named 10, I think not. The answer for Granite Peak is "maybe". The answer for List of peaks named Bear Mountain: probably not. We should only be creating tall SIAs instead of DABs if they are definitely more helpful for navigation. Otherwise, they should be converted back to DABs.
Does this make sense? What do other editors think? — hike395 (talk) 07:24, 10 October 2019 (UTC)
I don't want to be negative, but think it will be impossible to draw the boundaries. A set index that belongs to all of the classes defined above: all items notable, short & complete, list notable in itself, like Dodge Charger or Cypress, is uncontroversial. A set index that belongs to one class (short and complete) like List of peaks named Bear Mountain, may be more controversial, but does meet the WP:CSC3 guideline so should be kept. It may be useful to someone trying to locate their particular Bear Mountain. Perhaps an editor will start articles on more of these Bear Mountains, most of which have probably been described by various sources. I would prefer to keep this guideline, add the sections on typical information and related guidelines, and add a section describing the three overlapping classes: all items notable (CSC1), short & complete (CSC3), list notable in itself (GNG). That, I think, will greatly reduce argument about these lists. The new sections can be tweaked as needed. But is there a real problem? Do editors keep adding ridiculous set index articles that lead to lengthy deletion debates? I think not. Aymatth2 (talk) 12:25, 10 October 2019 (UTC)

@Aymatth2: I think the boundaries that you defined, above, were clearly defined and non-overlapping. That's what triggered this line of thinking for me. In other words:

  • SIAs based on WP:GNG or WP:CSC1 are "informational" SIAs. Either the overall topic is notable or every list item is notable. Informational SIAs should strive to be complete articles, ideally with as much non-trivial information as possible. Canonical example: Dodge Charger. If only minimal information is in an informational SIA, then the SIA should be converted back to a DAB. Informational SIAs are acceptable wikilink targets.
  • SIAs based on WP:CSC3 are "navigational" SIAs. These are short complete list of items, not necessarily all blue-linked or notable. The extra information in a navigational SIA beyond a DAB should only be added if it is helpful for a reader to find the correct article or help them realize that the article is not written. Navigational SIAs should not be WP:INDISCRIMINATE lists or data dumps. There should be a minimum number or fraction of blue links in a navigational SIA. Wikilinks to navigational SIAs should be fixed to refer to correct articles. Canonical example: Granite Peak

I'm proposing this split based on a number of problems that I see in the current SIAs (see Category:Set indices):

What do other editors think of the material in green, above?hike395 (talk) 18:24, 12 October 2019 (UTC)

I like the sound of the two (green) categories. Could they be restated as a) lists and b) dabs? For the latter, how does the example Granite Peak satisfy WP:V, specifically the strictness of MOSDAB means info on a dab can be verified by the internal inks, but if Granite Peak had no references the entries with red links (but no blue) would fail WP:V (it actually has inline EXT rather than no source per entry). As they are project aids, why aren't articles that fail WP:V moved out of mainspace where they may progress (with compliance) quicker? Widefox; talk 01:13, 13 October 2019 (UTC)
I'm a little reluctant to call navigational SIA "dabs", because then people will say that they must obey MOS:DAB and only that. The third list common selection criterion (what we're calling WP:CSC3) says "The inclusion of items must be supported by reliable sources," which fulfills WP:V. I think this implies that Navigational SIAs with no references should be deleted or copied out of mainspace (as you say). Navigational SIAs are not meant to be project aids, but actively useful for readers. They are DAB++ pages. — hike395 (talk) 04:05, 13 October 2019 (UTC)
My point being that there's either navigation aids (with exemptions for a. WP:V and b. desirability of incoming links) or articles (list or otherwise). I've never seen an argument for a logical, consistent position for SIAs which sit in between those two completely mutually exclusive sets, which by definition are irreconcilable. MOSDAB is strict, yes, but does allow enough information to differentiate topics, so any tables, images etc (of say mountain heights etc) would correctly need justification for navigation only, but be currently allowed as per the examples of them. If extended info is needed solely for navigation then MOSDAB should be clarified, if it is currently unclear about this. The only insurmountable I know is entries with redlinks without a blue, which can be fixed by adding the redlink to articles (especially lists) and adding the blue to the list, done. Any leftover entries where there's no obvious article/list could be created by projects. What sort of articles wouldn't be used as links in other articles? It is much more useful for readers and writers to not blur the distinction between the two irreconcilable types, however tempting for utility. Having the structure of WP based on a logically consistent foundation is the only way forward. Widefox; talk 11:42, 13 October 2019 (UTC)
I think we are getting over-complicated. Forget about DABs. Set indexes are lists, and must conform to one or more of WP:GNG, WP:CSC1 or WP:CSC3.
Lets take a typical geographical set index. A sortable table lists all Canadian rivers with a given name, citing a government source for the list, which shows it is complete. Each entry gives name (always the same), basin size, mouth coordinates, province and municipality. Half the names are blue linked. It meets WP:CSC3. All the entries are named after a given 17th century European explorer, so it is possible that the list also meets WP:GNG, if unlikely. Most rivers are notable, at least in WP:GEOLAND terms, so it may well meet WP:CSC1 too.
Why would we agonize over it? There is solid precedent for lists like this, which are useful to readers. It may take them to an article on their river, but if there is none it gives them some information and lets them find it on the map. WP:GNG, WP:CSC1 or WP:CSC3. are good enough. We cannot say that set indexes must meet more rigorous criteria than lists in general. We could say they do not need to meet such rigorous criteria, but I see no reason for that. If we want to change WP:CSC, the change should be discussed on the talk page for that guideline, not here. All that is needed here is an outline saying set indexes are lists, and must conform to one or more of WP:GNG, WP:CSC1 or WP:CSC3. Aymatth2 (talk) 13:03, 13 October 2019 (UTC)
Yup, that's an entirely useful and consistent way forward - they are a subset of lists and as such are nothing to do with dabs. Widefox; talk 14:34, 13 October 2019 (UTC)

Distinguishing between "informational SIAs" vs. "navigational SIAs" seems helpful.

  • Basically, things labelled SIAs are exceptions to regular rules for list-articles or for disambiguation pages. In the past, there were disambiguation-focused editors removing redlinks, for example, from some worklists of same-named ships which nowadays would be considered fine. Either all the redlinks have since turned blue, so a minimalist worklist became a valid disambiguation page. Or there was extra info included and the worklists have become valid as regular, standalone list-articles. Since then, standards for list-articles have changed, and old issues no longer apply. (Note, Wikipedia:OkayVsNotOkayListsOfPlaces is an essay drafted by me in part about how standards for list-articles are wider now, by the way.)
  • About "informational SIAs", the resolution is that we should not have those, at all, any more, I think. Consider examples:
    • Dodge Charger, a supposed "canonical example" mentioned on wp:SIA, is fine as an article/list-article. It does not need to be called an SIA.
    • List of ships of the United States Navy named Enterprise, another wp:SIA example, is clearly a list-article. Note there is a source, this one, cited in the list-article, which explicitly considers the set of U.S. navy ships named USS Enterprise. It does not need to be "protected" from deletion by labelling it an SIA.
    • USS Constellation is fine as list-article IMHO, and/or is fine as a disambiguation page.
    • List of peaks named Signal: This became an "informational"-type list-article in editing by me and others adding information. However, the basic idea is that a "Signal mountain" or "Signal peak" is a basic kind of thing (which can have an article), and that we can have a list of examples, either combined with the basic article or split out as a separate article. In this case, the current list is restricted to places named exactly "Signal Mountain" or "Signal Peak", but that should be changed: it should be a list of mountains historically used as signal mountains, no matter what their name. Presumably all the ones literally named "Signal Mountain" can be kept in the list.
    • Are there any examples of "informational" list-articles of same-named-items that "need" to be labelled "SIAs" in order to protect them from attacking editors?
The history of SIAs started with "ship index" articles having their informational content removed by editors who insisted that MOS:DAB covered such articles. That was back in 2007, but I bet that if we removed the protection of WP:SIA we would be back in the same situation. I would strongly recommend not taking away this guideline. — hike395 (talk) 15:34, 15 October 2019 (UTC)
  • About "navigational SIAs", these should only be articles of same-named items that have the "same name" as the name of the article. E.g. U.S.S. Constellation is potentially one of these, but something named List of ships named U.S.S. Constellation would not be. A reader might search for "U.S.S. Constellation", and arrive there, but no reader would type "List of ships named U.S.S. Constellation". The first serves the navigational need for the reader (and can sensibly have the SIA tag stating "This article includes a list of ships with the same or similar names. If an internal link for a specific ship led you here, you may wish to change the link to point directly to the intended ship article, if one exists."); the second does not, and cannot sensibly hold any tag like that.
    • It may be okay to allow a "set of same-named items" with some extra stuff not allowed in a disambiguation page, if it is at the "same name", and if the extra information is limited. Extra info can include redlink items (calling for an article to be created, probably with an inline citation to a source supporting notability), blacklink items (where no future article is expected), general and item-specific sources, some limited data like coordinates (and if coordinates are given, then also allow {{GeoGroup}} template so readers can see where the places are).
    • In exchange for allowing that to exist, the set of same-named items should accomplish something or more than one thing. For example, it must serve as the sole or main place where disambiguation is done for all items of that name. Note there has been confusion in the past where if there are a set of same-named things, such as "USS Constellation", where all the items are of the same "type", then that is allowed to exist as a SIA at title "USS Constellation". But if and when some other notable thing of the same name comes into existence, such as a movie or a novel named "USS Constellation", then SIA-focused editors have come in and enforced a rule (why this rule came to be, I do not know) that the same-type ones must be moved to "List of ships named USS Constellation" instead, and a separate disambiguation page must be created at "USS Constellation". I think that is confusing and duplicative and unhelpful. Instead, just one "USS Constellation" should exist, with a sensible introduction like "USS Constellation may refer to any one of several ships listed below, or it may refer to: USS Constellation (movie) or USS Constellation (novel)" and then give the list of ships. Or maybe give the list of ships first, but mention up top something like "USS Constellation usually refers to one of the following ships, but there are some other uses (see further below)". I think it is important to get rid of the awkward duplication of SIAs overlapping with DAB pages, which is awkward and certainly is really difficult to explain to anyone else (the only explanation I have ever heard is that this is done according to the rule, begging the question of why there is such a rule).
    • Or another way the same-named set accomplishes something, is that it may allow us to head off creation of separate articles and should accomplish what is polite to readers in terms of acknowledging existence of things that are only marginally different, if different at all, from accepted same-name items. This is a pretty big deal. For example, an expanded set-article for the topic of "Lincoln Middle School" could contain all of the 15 examples in current Lincoln Middle School disambiguation page, plus other 47 more examples (as can be seen in Talk:Lincoln Middle School/LincolnMiddleSchool-draft-SIA, a draft showing complete list of such places; please forgive its current formatting). Note that currently, anyone who graduated from a Lincoln Middle School is insulted by the fact that Wikipedia obviously allows articles about many such places (even though many of these are very weak articles), and even allows redlink mentions about others (which perhaps are weakly supported by redlink mentions from corresponding school district articles technically satisfying MOS:DABRL, an obscure rule/loophole which most readers will not grok at all, whether or not the place is at all likely to be notable), but inexplicably keeps on deleting mention of their one unless they create an article for it. Which they may try to do, only to have that deleted rudely by SPEEDY, PROD or AFD, probably unfairly given poor state of other accepted articles, and the negative cycle continues. We can simply show their one in the set of same-named items, and a lot of grief and churning can be avoided.
--Doncram (talk) 01:25, 15 October 2019 (UTC)

Simple solution? break

@Aymatth2: I'd like to make three points:
  1. You've adopted something very close to my interpretation of WP:LISTN -- that a list must fall under WP:GNG, WP:CSC1, or WP:CSC3 to be suitable for inclusion in WP. But this is neither the consensus view (see Doncram's objections), nor is fully inclusive of all kinds of lists (see, e.g., WP:LISTOFLISTS). I think we can't make broad statements that this interpretation is correct. We can, however, establish it as a guideline for SIAs, if we gain consensus.
  2. We can say that SIAs must follow stricter standards that the ones you proposed, because SIAs have an unusual position of being between DABs and full list articles, which could impact notability. As a thought experiment, pretend that there was consensus to remove WP:SIA. Then, SIAs would fall under MOS:DAB and almost all SIAs would be turned from list articles into DABs. That's a much higher stricter standard (i.e., none of them are acceptable list articles) than what you're proposing. I think we're trying to get to a consensus on what guideline we should settle on.
  3. I think we can't propose any guideline weaker than the one you've proposed, without violating the spirit of WP:LISTN or WP:N and permit junk SIAs.
hike395 (talk) 15:34, 15 October 2019 (UTC)
The general rule is that guidelines can get weaker as we move from the general to the more specific, not stronger. WP:GNG is the most general guideline, and any article that meets it is acceptable. WP:CSC, WP:GEOLAND etc. are more specific guidelines and allow some articles that would not pass WP:GNG. WP:SIA, applying to a subset of lists, could be weaker than WP:CSC (I see no reason for that) but cannot be stronger. That is the way Wikipedia works and must work.
The selection guidelines are self contained, so if an article meets one of them, it is acceptable. For example, someone may have started out as a soccer player, moved into academia, then found their niche as a TV pundit. Their article is acceptable if it meets WP:FOOTY, WP:ACADEMIC or WP:GNG. It does not have to meet all three. If a set index article meets WP:CSC or WP:DABSTYLE or WP:GNG it is acceptable. It does not have to meet all three. This is true whatever WP:SIA has to say.
Many different types of list give links to articles about some or all of the list entries. They are not DABs unless they conform to WP:DABSTYLE.
Many lists tagged as set indexes should be DABs. Those are the ones where all the entries link to articles, and only minimal information is provided about each entry. Part of this guideline should state that lists that qualify as DABs should be formatted and tagged as DABs. We are left with lists that match the definition of set indexes, meet one or more of WP:GNG, WP:CSC1, or WP:CSC3, and clearly are not DABs. These are the ones this guideline should focus on.
If we ignore Doncram's angry outbursts, I would say we have tentative agreement on the proposed #Related policies and guidelines and #Typical information, which can always be tweaked later. I will start a proposal for Common selection criteria below, trying to reflect what has been said so far. Aymatth2 (talk) 11:37, 16 October 2019 (UTC)
@Hike395: I have learned a lot in this long discussion, and have come to think that stating the obvious about what kind of content an SIA could have and what other guidelines apply may be the best that can be achieved in the short term. We can always refine wording and add more later. I suppose I could claim that silence means consent and just push through the three proposed new sections, but would like your views first. Thanks, Aymatth2 (talk) 19:21, 22 October 2019 (UTC)
@Aymatth2: Sorry for the long silence, been busy IRL. Will respond more fully later today. — hike395 (talk) 20:54, 22 October 2019 (UTC)
Thanks. I look back at my first proposal and see it was very naive. If anyone could object to mentioning any of the guidelines, or to any descriptive text, or to any of the points on typical information, I will trash it and move forward rather than argue the point. What should be scrapped from the proposed text? Aymatth2 (talk) 23:25, 22 October 2019 (UTC)

@Aymatth2: I've re-read the three sections you've written (guidelines, typical information, types), and I don't object to anything you've written. A number of other editors have been unhappy with SIAs over the years --- I suspect we're missing an opportunity to make a guideline that both more closely covers SIAs that are commonly accepted to be "good", yet would fix the cloud of dissatisfaction around SIAs. But that may not be an outcome we'll reach today. So, I support what you've proposed. — hike395 (talk) 11:28, 23 October 2019 (UTC)

@Hike395: Thank you for that. I was also hoping it would be possible to give more in the way of guidance, but I think there will be some value in the new sections. Maybe adding some more examples of good / acceptable / bad might be useful. But not now. Aymatth2 (talk) 11:38, 23 October 2019 (UTC)
@Hike395: I have made the changes. There is inconsistency between the statement in "Common selection criteria" that a set index may be a disambiguation page and WP:SETNOTDAB, which says a set index article is not a disambiguation page. I am not sure the best way to resolve that. Maybe tweak WP:SETNOTDAB to say "is not necessarily"? Aymatth2 (talk) 13:03, 24 October 2019 (UTC)

Common selection criteria

A draft new section follows listing common reasons for considering a set index valid.

Common selection criteria

A set index article, a list of items of a specific type that share the same (or similar) name, may be one or more of the following:

  • Disambiguation page: all the entries in the list have articles, and the information given for each entry serves mainly to distinguish that entry from the others. Should conform to the manual of style for disambiguation pages. See MOS:DAB.
  • Notable list: the list topic has been discussed as a group or set by independent reliable sources. The individual entries do not have to be notable in themselves. See WP:LISTN.
  • List of notable items: the list topic may not be notable in itself, but all the items are notable. Red-linked entries are acceptable if the entry is verifiably a member of the listed group and it is reasonable to expect an article could be forthcoming in the future. See Wikipedia:CSC.
  • Short, complete list: includes every item that is verifiably a member of the group. It is reasonably short and could be useful or interesting to readers. Inclusion of items must be supported by reliable sources. Lists where no entry is notable are rarely appropriate. See Wikipedia:CSC.

Refer to the relevant guidelines for further details.

The above should be considered in conjunction with the proposed #Related policies and guidelines and #Typical information. Any suggestions for improving the wording welcome. We should keep it as short and focused as possible to avoid duplicating content in the related guidelines. Aymatth2 (talk) 10:52, 17 October 2019 (UTC)

As Aymatth2 says above, there is a problem with saying that SIAs include DABs. They're meant to be different. I think the first item in this box should be removed. I removed it from the live guideline (for now), because it rendered the guideline self-contradictory. — hike395 (talk) 08:08, 25 October 2019 (UTC)
I am o.k. with that. If a set of items with the same type and same or similar name all have articles, the editor has a choice of formatting and tagging it as a DAB or as a SIA list of notable items. When I read "Set indexes and disambiguation" followed by "Common selection criteria" that seems clear enough without having to spell it out. Aymatth2 (talk) 12:18, 25 October 2019 (UTC)

Can lists of lists be SIAs?

I noticed that Lists of German politicians is categorized as a set index article. Its name was previously List of German politicians, so I'm guessing someone decided it was a SIA because it was a list of things sharing the same name, with that name being "List of German politicians"? But that seems like a tenuous interpretation to me. It seems there are a fair number of similar examples. Here are just a few:

Does anyone object to my removing the Set indices category from WP:LISTOFLISTS articles such as these? Colin M (talk) 18:35, 1 November 2019 (UTC)

I would say that lists of lists are not set indexes because they are not lists of items. See WP:LISTOFLISTS. They should be in Category:Lists of lists. Aymatth2 (talk) 20:18, 1 November 2019 (UTC)
Petscan currently finds 58 pages that are in Lists_of_lists and in Set_indices. I've looked at some of the edits that have caused thus (example) and (imo) they don't make much sense. DexDor (talk) 21:51, 1 November 2019 (UTC)
I agree: it may be moves that have created these anomalies. Lists of lists shouldn't be SIAs. Shhhnotsoloud (talk) 08:28, 2 November 2019 (UTC)
Thanks to all for chiming in. I've removed the SIA category from the examples above, and recategorized many other similar articles from SIA -> Lists of lists. Colin M (talk) 20:41, 4 November 2019 (UTC)
Since we're still on this subject, would it also be feasible to create a footer template indicating lists of lists that works in a similar fashion as the set index footer? Jalen Folf (talk) 21:28, 4 November 2019 (UTC)
That seems like a good idea, not hard to do. See Wikipedia talk:Stand-alone lists#Proposed "List of lists" template. Aymatth2 (talk) 22:33, 4 November 2019 (UTC)

 You are invited to join the discussion at Template talk:Set index article#Short description. ‑‑Trialpears (talk) 19:35, 13 February 2020 (UTC)

Members of and visitors to this WikiProject are invited to comment on and to contribute to a new WikiProject - Wikipedia:WikiProject Bluelink patrol. The founders think that it complements but in no way infringes on any existing WikiProject - and also, that we cannot have got everything right. Narky Blert (talk) 22:03, 27 January 2021 (UTC)

SIA or List?

Seaboard Air Line Depot is currently classified as a disambiguation page, but does not fit the criteria. I am not sure if it should be an SIA or a regular list article. Any input at Talk:Seaboard Air Line Depot will be appreciated. - Donald Albury 18:05, 20 October 2022 (UTC)