Wikipedia talk:Manual of Style/Disambiguation pages

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
WikiProject iconManual of Style
WikiProject iconThis page falls within the scope of the Wikipedia:Manual of Style, a collaborative effort focused on enhancing clarity, consistency, and cohesiveness across the Manual of Style (MoS) guidelines by addressing inconsistencies, refining language, and integrating guidance effectively.
Note icon
This page falls under the contentious topics procedure and is given additional attention, as it closely associated to the English Wikipedia Manual of Style, and the article titles policy. Both areas are known to be subjects of debate.
Contributors are urged to review the awareness criteria carefully and exercise caution when editing.
Note icon
For information on Wikipedia's approach to the establishment of new policies and guidelines, refer to WP:PROPOSAL. Additionally, guidance on how to contribute to the development and revision of Wikipedia policies of Wikipedia's policy and guideline documents is available, offering valuable insights and recommendations.
WikiProject iconDisambiguation
WikiProject iconThis page is within the scope of WikiProject Disambiguation, an attempt to structure and organize all disambiguation pages on Wikipedia. If you wish to help, you can edit the page attached to this talk page, or visit the project page, where you can join the project or contribute to the discussion.

Should the existence of an article for the entry topic be a factor in WP:DABORDER #4?[edit]

Right now, we recommend sorting by similarity to the base title (with some detail on what that means), then alpha or chrono as appropriate. There's no mention of putting MOS:DABMENTION entries below ones where there's an actual article for the ambiguously named topic. But in the wild, that's actually done quite often (e.g. Difference), and is probably to readers' benefit: people are more likely to be looking for topics that are notable enough for an article than those that aren't. The downside is maintenance when new articles are created; one hopes that whoever updates a dab with a new link also moves said entry up, but that doesn't always happen.

Bottom line: Should we suggest sorting this way, as is often done, or leave it alone? And if yes, what priority should that have with respect to the other sorting criteria? —swpbT • beyond • mutual 18:51, 3 November 2023 (UTC)[reply]

I think we should just suggest that the person editing the disambiguation page makes up a redirect for the mentioned topic. That's the best context to be able to decide which disambiguation marker to use. If a newbie sees that e.g. The Difference (Matchbox Twenty song) exists, they will naturally try to edit that, and won't be tempted to start editing The Difference (song) or Difference by Matchbox Twenty or The Difference (2002 song) etc, which would lead to more maintenance work. --Joy (talk) 19:07, 3 November 2023 (UTC)[reply]
That's a great solution for e.g. songs redirecting to albums or artists, and probably many other situations, but there are some situations where the relationship between the mentioned topic and the article that mentions it is less obvious, and a redirect could leave readers wondering "why did I end up here?" In that case, I'd probably leave a red link or no link to a disambiguated title, and keep the blue link later in the entry, but that leaves the question (to me anyway) of where such entries should be placed. —swpbT • beyond • mutual 19:38, 3 November 2023 (UTC)[reply]
I'm not sure I really recognize that use case, can you post an example of some sort so it becomes clearer? The closest I can think of is e.g. airport codes, where we might not want the first link in the entry "XYZW (ICAO code), the code for Foo International Airport" at Xyzw because it's... too contrived? --Joy (talk) 02:47, 4 November 2023 (UTC)[reply]
I would generally write that entry as "Foo International Airport (ICAO code XYZW)". But anyway, I think it's generally recognized, as Bkonrad notes below, that it's not always appropriate to create a redirect when the real target is in the description. —swpbT • beyond • mutual 12:49, 4 November 2023 (UTC)[reply]
I think this page should not make any recommendations about whether to create redirects in such cases. Several projects have naming conventions and notability guidelines and have routine issues with editors who mechanistically create redirects for unnotable topics (for example songs, minor characters or episodes). olderwiser 11:28, 4 November 2023 (UTC)[reply]
But if the redirect is for a non-notable topic, why should it really be listed in the disambiguation page? Either something is a legitimate search term and a source of ambiguity, or it is not - what's the middle ground? --Joy (talk) 11:35, 4 November 2023 (UTC)[reply]
Notability is generally not considered in whether to include a item on a disambiguation page. There is some vague agreement that subjects mentioned once only in passing in an article that is mostly unrelated to the subject don't need disambiguation (for example, a person mentioned once in an article about an event or another person altogether). But where to draw the line is unclear as entries that satisfy WP:DABMENTION are generally accepted as entries on dab pages.
But specifically regarding redirects, I think it can be confusing and even misleading to have a redirect from a very specific subject to some other article unless that subject has some independent notability and might have potential to exist a standalone article (or at least be a topic that readers are likely to search for). In such cases, I think it is better for the link to be in the description to make it clear 1) that there is no separate article on the topic and 2) what the actual topic of the linked article is. olderwiser 12:41, 4 November 2023 (UTC)[reply]
I see what you're saying but I still think it leads to too little clarity in composing disambiguation pages. Redirects are navigational aids, they don't need to have independent notability - we actually have a tag {{R with possibilities}} for such redirects and otherwise it's clearly optional. If a topic is coherent and worthy of a mention in navigation, it's worthy of a redirect; if it's not, if it's not even worthy of a red link, then it shouldn't be in navigation at all, we can just let full text search handle any such inquiries. --Joy (talk) 13:40, 4 November 2023 (UTC)[reply]
That has been the subject of considerable discussions in the past. However, dab guidelines currently say very little regarding the role of notability. Red links advises to avoid adding red links to articles that are unlikely ever to be written, or are likely to be removed as insufficiently notable topics. WP:DAB mentions notability in regards to determining a primary topic. WP:DABMENTION is the main guidance entries for topics that do not have their own article. Using a redirect is offered as an option there -- but I'm not sure it is appropriate for this page to provide recommendations about when to create such redirects -- especially where subject-matter projects may have their own guidelines for such redirects.
BTW, a redirect tagged as {{R with possibilities}} would be a good candidate for using the redirect on a dab page. However, this would generally not be applicable for many redirects concerning episodes of barely notable series or minor characters or songs that have no significance other than that they were part of an album. olderwiser 14:08, 4 November 2023 (UTC)[reply]
What's the problem with redirects from songs, minor characters, or episodes, if they're properly named and mentioned in the target? Redirects are dirt cheap and don't really need maintenance. —swpbT • beyond • mutual 12:44, 4 November 2023 (UTC)[reply]
These do often require maintenance as the 'container' target of the redirects are frequently the subject of repeated moves and or deletions due to notability concerns. olderwiser 12:49, 4 November 2023 (UTC)[reply]
But double redirects and orphaned redirects are taken care of by bots. Is there any human workload there? —swpbT • beyond • mutual 12:53, 4 November 2023 (UTC)[reply]
Is there a bot that removes dab entries for an orphaned redirects? I think there might be one that de-links such redirects, but that subsequently does require some additional manual update to the dab page to fix or remove the now unlinked entry. Particularly for minor characters, less so for episodes or songs, the actual content may often be removed from the container article and as such, while the redirect might still exist, it now fails WP:DABMENTION. olderwiser 13:11, 4 November 2023 (UTC)[reply]
We should always have these redirects point to named anchors, that way bots can easily detect if the destination goes missing. I mean it's not like it's so much harder than for the bot to grep the destination article text for the exact same string and see it missing, but still. --Joy (talk) 13:42, 4 November 2023 (UTC)[reply]
I'm not aware of any bots that function in the way you describe. And in any case, I don't see how such a bot would be able to remediate the disambiguation pages on which the redirect appears. olderwiser 14:11, 4 November 2023 (UTC)[reply]
A bot can traverse the entire list of links on a disambiguation page, recurse into each of the destination pages and look for the named anchor. If it fails to find it, it can post a message on the talk page of the disambiguation page to say it's gone. I think I've seen one bot do something like that, I think it's the one linked from {{broken anchors}}. --Joy (talk) 14:20, 4 November 2023 (UTC)[reply]
Yes, it certainly might be possible, but even so, there would still be some "human workload" as result which is what swpb had asked about. olderwiser 14:55, 4 November 2023 (UTC)[reply]

To editors Joy and Bkonrad: This thread has gotten a bit side-tracked. To get back to the central question: assuming there are some entries where a redirect or redlink is not appropriate, so the only link in the entry is in the description, should we say anything about where to place such entries with respect to ones with a link up front? —swpbT • beyond • mutual 14:24, 4 November 2023 (UTC)[reply]

No, because the existence of the link is not actually necessarily indicative of anything, since you can see the apparent inconsistency in how we add or don't add links. --Joy (talk) 14:26, 4 November 2023 (UTC)[reply]
For this specific question, in the past I think I had advocated for not making a distinction between whether an article exists or not with regards to order. Others, I think for example, JHunterJ, had advocated for making such distinctions. Personally, I no longer have any strong opinion on that particular aspect of organization. Or to be clearer, I don't generally make changes in how lists are organize based on that criteria -- if the list has separated the non-articles to the bottom, I generally leave it that way, and the same if the list is fully organized by alphabetical order or chronological order. I think I still have a preference for using alpha or chrono order without regard to article/redirect existence as I think such a list is easier to parse. olderwiser 14:48, 4 November 2023 (UTC)[reply]
I typically assume that the existence of an article corresponds to a higher likelihood of being the article sought, so falls under "should be ordered to best assist the reader in finding their intended article", in addition to being easier to scan leading blue links at the front then scan descriptions with link for entries that don't have a blue link at the front. But if the reader would be better assisted in some cases with another arrangement, that should be the priority. -- JHunterJ (talk) 19:10, 22 January 2024 (UTC)[reply]

This discussion reminded me that I had proposed changing the wording in this section a few months ago. The only response at that time was supportive, so I have boldly done what the supporter suggested (which was not quite the same as my proposal.) If this change is acceptable, it may make the question in this section moot.

For reference, the wording I proposed was "Within each group or section, entries should be ordered to best assist the reader in finding their intended article. Articles that are more likely to be the reader's target should be listed above those on more obscure topics. Other methods such as chronological ordering may be used where appropriate. Alphabetical ordering should be used when no other method is suitable. Entries may also be organized by similarity to the ambiguous title."

--Srleffler (talk) 17:33, 4 November 2023 (UTC)[reply]

I think the issue here is that these aren't really some known standards, but just using the terms like intended, likely and obscure, which are harder to measure and decide upon and could lead to disputes. The similarity part is also vague - as all entries on a disambiguation page actually have to be ambiguous with the term described, and multiple terms can redirect to the same disambiguation list, it's often impossible to decide what are these levels of similarity, while anything that is merely similar needs to go to the see also section anyway. --Joy (talk) 08:33, 5 November 2023 (UTC)[reply]
Well, most of our guidelines use fairly subjective "leave room for editorial discretion" language like this instead of being written in absolutes, because they are guidelines not policies, and editors tend to get up in arms when guidelines are worded very restrictively ("WP:CREEP!", etc.)  — SMcCandlish ¢ 😼  12:50, 5 November 2023 (UTC)[reply]
Ironically WP:CREEP argues against bloated pages, yet here we are arguing how to sort disambiguation lists that are inherently bloated when sorting is such a concern :) --Joy (talk) 13:30, 5 November 2023 (UTC)[reply]
Yes, I am proposing that the ordering of items on dab pages needs to be somewhat subjective, because there is no benefit in using an absolute, objective ordering that does not serve the purpose of the page, which is to help the reader find the article they were looking for, when they do not know the correct title of the article. Putting obviously more common topics ahead of obviously obscure ones improves the utility of the page. That there are entries in between with no obvious ordering is not a problem—any ordering will be fine. Editors can disagree over the exact ordering; it's OK. We will find consensus, as usual.--Srleffler (talk) 16:42, 5 November 2023 (UTC)[reply]
But that's an idealistic view of how readers approach navigation. Why would we assume that e.g. people aren't more comfortable with alphabetic sorting for all but the most popular of items? When we get into the weeds, one person's common easily becomes another person's borderline obscure. Besides, consensus would be built by the average editor, which may or may not match what the average reader wants. --Joy (talk) 19:06, 5 November 2023 (UTC)[reply]
Further, the reader has generally no idea at all which subject is a bit more popular than another one (only in cases of a really overwhelming difference, as at Michael Jackson (disambiguation), so ordering them by popularity will just appear as completely random ordering to anyone not privy to the popularity stats (which may change radically over time anyway). And "there is no benefit in using an absolute, objective ordering" isn't accurate, since a fixed order would at least become predictable to longer-term readers, just as our fairly consistent parenthetical disambiguation strings become predictable, our article titling patterns become predictable, our section ordering mostly becomes preditable, etc.  — SMcCandlish ¢ 😼  19:51, 5 November 2023 (UTC)[reply]
For a dab page, random ordering would be just as good as alphabetization. If a reader has ended up on a dab page, they do not know the title of the article they are looking for. Ordering the entries alphabetically by title is sorting the list by the one thing you can be absolutely sure that the reader does not know. Even an approximate ordering by likelihood will be much better than alphabetization. Put obviously more common articles at the top and obscure ones at the bottom. Put bigger things before smaller ones (countries > states > cities > towns; bands > albums > songs). Anything that doesn't have a clear ordering can be put in any order; it doesn't matter. Absent any better ordering, even ordering entries by when they were added to the list will be better than alphabetization.--Srleffler (talk) 21:12, 5 November 2023 (UTC)[reply]
I don't know that's useful for most use cases. Imagine you're a person who read something about a small town of Foobarbaz on the border of two US states, and want to look that up. You encounter a list of a dozen items sorted by population size, but it doesn't actually say that size in the caption. Do you really benefit from the sorting? Or, you know you're looking for a person with the common surname of Foobarbaz who's a less well-known musician, but can't remember the exact given name, but you'll know it when you see it again - how does sorting by popularity help - how do you gauge how much of the top of the list can you skip over? You're going to try Ctrl+F or just look over everything until you find the right person anyway. --Joy (talk) 21:36, 5 November 2023 (UTC)[reply]
You have to consider the bigger picture. Of all of the users who end up on the Foobarbaz dab page, more of them will be looking for the country of Foobarbaz than for the small town of Foobarbaz. By listing the former above the latter you reduce the search time for the larger population of users. The presumption is exactly what is in your last sentence: most readers will need to look over everything in the section to find what they are looking for. The best we can do is put the more likely choices first, so that the larger number of users find what they are looking for quicker.--Srleffler (talk) 03:43, 6 November 2023 (UTC)[reply]
The problem here is that you're guessing on what constitutes "more" and "larger". It's easy enough for the top of the list, but how to decide between what's #4 and what's #6 on the list without some of the readers just becoming frustrated by a seemingly random list? --Joy (talk) 14:06, 6 November 2023 (UTC)[reply]
No, not guessing. If there's ten items on the list, in most cases it will be clear which are the top three and which are the bottom three or four. How to decide what's #4 and what's #6 is it doesn't matter. Any ordering you like is fine there. A random list is better than a list ordered by the one thing the reader does not know. --Srleffler (talk) 05:01, 7 November 2023 (UTC)[reply]
"A random list is better than a list ordered by the one thing the reader does not know" clearly is not correct, since for the end reader there is no discernable difference, but for editors the latter is more maintainable.  — SMcCandlish ¢ 😼  17:18, 7 November 2023 (UTC)[reply]
"Maintainable" is a pointless fixation here. It bothers some editors not to have a fixed, objective order to impose; it worries them that some other editor will change the order. If the order is not helpful, there is no reason to impose it. If different editors order the list differently, so what? Whether the list's order is "maintainable" is irrelevant.--Srleffler (talk) 06:30, 8 November 2023 (UTC)[reply]
There are advantages in having a list where an editor can see easily where a particular entry would be found, if present. PamD 13:53, 8 November 2023 (UTC)[reply]
I agree. It will help the editor to see if an entry they want to add is already on the list, for example. I am not adverse to alphabetization when there is not a better ordering. My argument is that in many cases there is a better order, for at least part of a list. --Srleffler (talk) 18:07, 11 November 2023 (UTC)[reply]
I'm sorry, I'm going to have to see some evidence for such bold assertions, that it will be clear which is which, or that any order is fine for readers. An alphabetically sorted list, or a list sorted by e.g. listed years of birth or something else that is in the caption, is at the very least apparent and predictable to the average reader. I would not underestimate the amount of reader frustration if we started intentionally randomizing lists, or in turn editor frustration over edit wars over "I think this criterion makes my topic go 2 spots higher!". --Joy (talk) 09:46, 8 November 2023 (UTC)[reply]
Joy: Not idealistic, but pragmatic. It's not about assuming that readers aren't comfortable with alphabetic ordering. Rather, we know for certain that the reader does not know the title of the article they want, so sorting alphabetically by that title is pretty much the least useful way possible for ordering the list. Yes, the ordering editors arrive at by consensus may not be ideal for the readers, but it will be better than what you get by blindly alphabetizing the list. Just moving the most likely 10% of the entries to the top and the most obscure 10% to the bottom and leaving the rest of the list unsorted will be better than alphabetizing for nearly every reader.--Srleffler (talk) 21:12, 5 November 2023 (UTC)[reply]
Think of the use cases where the reader remembers vaguely what they're looking for but doesn't have it exactly, for example you know you're looking for a person who you know the surname for, but can't remember if it's a Kirsten or a Kristen or a Kirstie or a Katherine or a Karen or a Corinne or something like that - alphabetic sorting actually helps there because it shortens the search time. --Joy (talk) 21:42, 5 November 2023 (UTC)[reply]
In most such cases, alphabetizing by the title the Wikipedia article happens to have is not going to help. You have a point, though, with half-remembered names. Alphabetizing by people's first name on a dab page for the family name is not a bad idea.--Srleffler (talk) 03:43, 6 November 2023 (UTC)[reply]
@SrlefflerSrleffler I think the statement f a reader has ended up on a dab page, they do not know the title of the article they are looking for is not accurate. Entries on a disambiguation page are ambiguous and they could all have the same title. If a reader did not know the title of an article, they would not even know to look at the disambiguation page. What you seem to be assuming is that readers know nothing else about the article other than the title which has very different implications and I think rather unlikely. For the most part, the parenthetical portions (or other disambiguating aspects) of the title generally should not be surprising and are meant to help readers distinguish one article from another. Granted, not all titles are equally effective at this -- but in the abstract, this is a very bizarre discussion. There needs to be some illustrative examples of where one type of sorting might be better than another. olderwiser 22:32, 5 November 2023 (UTC)[reply]
The reader may not know the title in advance but may guess it from a glance at the dab. In such cases, alphabetical order is useful. For example, someone landing on Springfield#United States will soon notice that the options are called "Springfield, state" and ordered by state, and can then pick the relevant town quickly from a couple of screenfuls of alternatives. Certes (talk) 23:01, 5 November 2023 (UTC)[reply]
@Certes: Note that the entries at Springfield#United States are not ordered alphabetically by the Wikipedia article title. They are sorted alphabetically by the name of the state, regardless of the actual title of the article. So, for example, Springfield, Albemarle County, Virginia follows Springfield, Virginia near the end of the list, rather than coming at the start of the list, before Springfield, Arkansas. This is a good example of what I am proposing: Someone thought through how to order this list rather than just blindly alphabetizing by whatever the article title happened to be. Alphabetizing by state makes sense, because someone looking for a town named Springfield likely knows that.--Srleffler (talk) 03:50, 6 November 2023 (UTC)[reply]
@older: My thinking comes from exactly what you wrote: "Entries on a disambiguation page are ambiguous and they could all have the same title." The reader knows what they are looking for. What they do not know is the title of the actual Wikipedia article for the specific ambiguous meaning they want. I presume the reader will scan down the list looking for the correct entry based on the parenthetical portion and the description. Sorting the list alphabetically doesn't help, because the reader doesn't know the title until they see it. Putting the more likely options first does help, because more users will find what they are looking for early in the list, without having to look over all the options.--Srleffler (talk) 03:43, 6 November 2023 (UTC)[reply]
"Putting the more likely options first does help" is only true when the most likely options or a couple of more likely options are much, much more likely. If one option is, say, 43% likely and another is 38% likely (how would we determine that?), the average reader has no way of knowing that, so the order will just seem jumbled.  — SMcCandlish ¢ 😼  15:46, 6 November 2023 (UTC)[reply]
In most cases it's pretty clear. On a dab page there will usually be some entries that are clearly more likely and some that are clearly more obscure topics. If not, the list should probably be ordered some other way. I'm not arguing for anything so fine as distinguishing between 43% likely and 38% likely. I'm arguing that in most cases there are a small number of entries that are clearly more common than the rest and some that are clearly more obscure. Separating the former from the latter is not rocket science.--Srleffler (talk) 05:01, 7 November 2023 (UTC)[reply]
This discussion is another sidebar from the original question, but looking closely again at MOS:DABORDER, I find a couple of points of confusion in item #4: That point purportedly concerns ordering within each section -- but the examples given would not typically be part of the same section -- how is it the order implied based on similarity to the ambiguous title supposed to be applied within a section when the examples provided are not illustrative of meaningful differences that would occur within a single section?
Somewhat less significantly, Moss, Monterey County, California has been renamed to Moss Landing, California and the entry on the dab page no longer illustrates the principle of ordering here. olderwiser 14:14, 5 November 2023 (UTC)[reply]

The redirect MoS:dab has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2023 December 2 § MoS:dab until a consensus is reached. Utopes (talk / cont) 21:32, 2 December 2023 (UTC)[reply]

Proposal to resolve nomenclatural confusion between split long lists and parenthentically disambiguated page names[edit]

 – Pointer to relevant discussion elsewhere.

Please see Wikipedia talk:Naming conventions (lists)#Fixing disambiguation confusion.  — SMcCandlish ¢ 😼  22:49, 25 January 2024 (UTC)[reply]

There's a discussion about band names as parenthetical disambiguators[edit]

See Wikipedia talk:WikiProject Albums#WP:THEBAND in article titles, where the original poster wrote:

So the possibilities seem to be:

  • Option A: Strength (The Alarm album)
  • Option B: Strength (the Alarm album)
  • Option C: Strength (Alarm album)

What do we do?

The wise users of this page may have some ideas. SchreiberBike | ⌨  03:12, 26 January 2024 (UTC)[reply]

IAR restatement does not belong on this page[edit]

For some reason this page devotes a section to a restatement of WP:IAR, which I removed yesterday since it has nothing to do with DAB pages. Editor PamD put it back with the explanation that "it's a useful reminder that dab pages are there to help readers".

This is not at all true. Quite the contrary, it wastes the time of every reader who gets that far in the document. A conscientious reader will go through the section, then re-read it trying to figure out what the relevance is to DAB pages, and finally realize that there is none. Just because IAR applies to all Wiki pages, it doesn't belong on the every Help page. Dan Bloch (talk) 22:08, 26 January 2024 (UTC)[reply]

I disagree. It's helpful in the specific context of dab pages to remind editors that the detailed rules laid out here will sometimes be overridden by the need to make the page serve it's intended purpose.--Srleffler (talk) 20:59, 27 January 2024 (UTC)[reply]
Why is it helpful here, but not helpful in every other Wikipedia: page? Dan Bloch (talk) 21:40, 27 January 2024 (UTC)[reply]
Because dab pages are far more rule-bound than most pages are. Dabs have a highly constrained style. Editors sometimes forget that even in such a tightly constrained case, sometimes you have to break the rules to accomplish the goal.--Srleffler (talk) 00:03, 28 January 2024 (UTC)[reply]
As IAR applies to any guideline, I would expect this section to have a few representative examples to illustrate its point and justify its existence. —Bagumba (talk) 01:12, 28 January 2024 (UTC)[reply]
At bare minimum. And they would have to be essentially incontrovertible examples nearly no one could disagree with. I'm skeptical that's possible, and tend to agree with the OP. Guidelines should never pre-emptively cite IAR; it just invites unjustified "exception" making on an "I like it better" basis.  — SMcCandlish ¢ 😼  07:04, 3 February 2024 (UTC)[reply]

MOS:DABCOMMON and pseudo-primary / almost-primary topics[edit]

MOS:DABCOMMON says that if there is a single WP:primary topic, to put that by itself at the top of the dab page. But what about when no single topic is primary but one topic is the most common among the various topics? See, for example, Twister and This version of Jump. I was under the impression we should only format a dab page that way if there is a dominant primary topic (i.e., if the dab page is not at the base name). So I changed the format of Jump as shown in the current version [converted to a permalink for posterity].

Should we change item 1 of the instruction to say something like "The primary or single most common topic, if there is one, should be placed at the top."

Should my change to the format of Jump be reverted? Should the format of Twister be changed?

—⁠ ⁠BarrelProof (talk) 05:13, 21 February 2024 (UTC)[reply]

There's a threshold for making a topic primary, putting it at the base name and calling any dab Foo (disambiguation). There's another threshold for hoisting a topic to the top of a dab. Is the second lower than the first? In other words, is there any case where a topic is close enough to primary to appear at the top of the dab but not primary enough to live at the base name? If not then we should either move the topic to the base name or demote its dab entry to the ranks. The two cases here are complicated by the fact that they would be a primary redirect rather than an article title. Does Twister (tornado) or Jump (leap) qualify as meeting the second threshold but not the first?
I don't support the current wording of Jump, where an artificial topic of "several songs" has been created as a secondary entry for WP:DABCOMMON. If we are going to hoist Jumping, better to leave it as the only entry at the top. Certes (talk) 11:24, 21 February 2024 (UTC)[reply]
I'm not sure I quite understand the base question, when it's covered by the existing phrasing of MOS:DABCOMMON, just below the mention of the primary topic: In cases where a small number of main topics are significantly more likely to be the reader's target, several of the most common meanings may be placed at the top, with other meanings below. See Mojave or Mercury for examples of this.
The exact details of the formatting of the top part is actually an area where we don't know enough about. I don't think I've ever seen any research done, such as trying to figure out what's the most common way our average reader parses a disambiguation list. The guideline text has mostly grown organically, with bold edits by editors, not necessarily based on hard science.
I recall seeing at least one case - echoes - where we had the single common link formated as a link inside the top sentence. This didn't seem to have the desired effect, so I reformatted it to be part of a top list with a longer caption. The next month the clickstreams showed more readers noticed it - though still not very many.
At Charlotte, I remember we had added a top list of common meanings, and then someone quickly added an explicit section heading. Stats showed a lot of interest in the top items of that common meaning section, though not as much for all of them (do note that we don't have too much of a sample here so as time passes this may change).
In the specific case of "several songs", that seems unlikely to be useful because it's not a link - we'd be forcing the reader to keep reading / scrolling / searching instead of just clicking, which seems suboptimal. I recently formatted such items with links to their sections at Major, you might want to have a look at that. --Joy (talk) 14:41, 21 February 2024 (UTC)[reply]
One interpretation of the base question is: New York usually means one of two topics, and they are listed prominently at the top rather than being buried in #Places. Is it ever good to do this with a single topic, or would such an important topic necessarily be primary? Certes (talk) 15:02, 21 February 2024 (UTC)[reply]
Yes, I think so, because we have numerous examples of a single topic being more popular than others, but not to the level of being a primary topic, because there's some sort of level of ambiguity that keeps it in a plurality as opposed to a supermajority, so to speak. In the RM for the aforementioned echoes, a lot of people thought echo was like that with regard to long-term significance. Something similar happened in/after the RMs for rumours, hurts, erasure. There's probably many more examples, these are just some of the recent ones that I could think of off the top of my head. --Joy (talk) 17:10, 21 February 2024 (UTC)[reply]
I agree with the OP that we should say that there can be a single popular topic at the top, even though it isn't the primary topic. This currently isn't clear. The MOS:DABCOMMON instructions called out, In cases where a small number of main topics are significantly more likely to be the reader's target, several of the most common meanings may be placed at the top... implies that the rule is only for cases with more than one such topic.
But the devil is in the details. What would it look like? Adding The primary or single most common topic, if there is one,..., as the OP suggests, implies that the new text would look the same as the primary topic case ("A school is an institution for learning"), which fails on two grounds: the wikilink (school) only succeeds because it's the primary topic; and additionally we would typically want the wording to be "most commonly refers to", not "is".
I think the new text should be added in the MOS:DABCOMMON section. Revised wording (new last sentence) might be something like:
In cases where a small number of main topics are significantly more likely to be the reader's target, several of the most common meanings may be placed at the top, with other meanings below. See Mojave or Mercury for examples of this. In the uncommon case where there is only one such topic and it is not the primary topic, it can be likewise be added at the top, e.g., "Twister most commonly refers to a tornado."
Dan Bloch (talk) 23:51, 21 February 2024 (UTC)[reply]
Jumping is almost certainly primary for the word Jump, so the base title should redirect to it, and the dab should be at Jump (disambiguation). We should definitely not add "or single most common" to WP:DABORDER #1, as primary topics and common meanings are two different things that are handled differently: primary topics are linked in a non-bulleted, complete sentence at the very top, followed by "X may also refer to:", whereas common meanings, even if there is only one, are regular, bulleted, sentence-fragment entries, with some sort of introductory line or header. I don't really think it's necessary to clarify that WP:DABORDER #2 applies to a single common meaning; after all, one is a "small number". But if such clarification is necessary, it should just be a short parenthetical: "In cases where a small number of main topics (or just one)"; the whole extra sentence proposed by Dan Bloch is really excessive IMO. —swpbT • beyond • mutual 19:15, 22 February 2024 (UTC)[reply]
An RM discussion at Talk:Jump#Requested move 18 April 2017 did not reach a consensus on the primary topic question, so Jump is a disambiguation page at the base name. —⁠ ⁠BarrelProof (talk) 19:24, 22 February 2024 (UTC)[reply]
Well, if consensus is that Jumping is really a common meaning and not a primary topic, then it should follow the common meaning formatting. To me, it seems like the primary topic, but I'm only one editor. The important point is that it should be one thing or the other; there are no "pseudo-primary" topics. —swpbT • beyond • mutual 19:28, 22 February 2024 (UTC)[reply]
I suspect there are a lot of cases where there is only one "almost primary" / "highest priority" topic, i.e., one topic that has much higher long-term notability and gets much more readership interest than the others, but not enough to meet WP:PRIMARYTOPIC criteria. For example, consider a hypothetical case where one topic has 35% of page views and the rest of the interest is scattered amongst 5+ other topics, and the dab page has 50 topics on it, and no other special considerations apply. In such a case, we cannot create a reasonable short list at the top that has more than one topic in it, because it is too difficult to figure out which and how many entries should be in that short list. I like the idea of using "most commonly" rather than "is" for such a situation. I agree that we should avoid letting the instructions get too complicated, but I think it is probably pretty common for the short list to only have one good candidate. —⁠ ⁠BarrelProof (talk) 19:46, 22 February 2024 (UTC)[reply]
This is an interesting situation. Wikinav for Jump shows 41% of outgoing views in Jan-2024 go to jumping and 32% to Jump (Van Halen song); in Dec-2023 the percentages were nearly equal (a slight edge to the Van Halen song). Two other topics had >5% as well. However, given that there are multiple songs by the name, I think it would be disservice to readers to pull out the Val Halen song into a common uses section. olderwiser 20:14, 22 February 2024 (UTC)[reply]
I'm not committed to that sentence, but without it we still need an example. So how about:
In cases where a small number of main topics (or just one) are significantly more likely to be the reader's target, several of the most common meanings may be placed at the top, with other meanings below. See Mojave, Mercury, or Twister for examples of this.
Dan Bloch (talk) 19:47, 22 February 2024 (UTC)[reply]
checkY —⁠ ⁠BarrelProof (talk) 19:59, 22 February 2024 (UTC)[reply]
I've made the change discussed above ("a small number of main topics (or just one)..."). Dan Bloch (talk) 19:40, 24 February 2024 (UTC)[reply]
The words "several of" don't make sense with that change, so I deleted those words. —⁠ ⁠BarrelProof (talk) 01:20, 25 February 2024 (UTC)[reply]

This issue also affects Foo (disambiguation) pages. For example, Apple (disambiguation) sensibly has a short list at the top because, when the reader types "Apple", we feed them the fruit which is good for them rather than Apple Inc. which most of them were actually looking for. Although three common topics are listed in this case, it's quite feasible that there might be just one common topic, even when a different topic is primary. Certes (talk) 20:03, 22 February 2024 (UTC)[reply]

Interesting case. For that one we have not just a short list, but also a regular primary topic identifier at the top saying "An apple is an edible fruit." That is followed by "may also refer to" and the short list. That seems sensible to me. —⁠ ⁠BarrelProof (talk) 20:41, 22 February 2024 (UTC)[reply]
There are plenty of similar cases. In the tech field we have Alphabet (disambiguation)mainly for Alphabet Inc. – and Telegram (disambiguation), where Telegram (software) has 66 times more views than anything else including the primary topic (though Telegraphy gets almost as many). Elsewhere, we have Model (disambiguation) with Model (person) being five times as popular as the primary topic, and Baler (disambiguation) listing Baler, Aurora with nearly twice the following of the farm machine. Sports are a popular place for alternative meanings, with Arsenal (disambiguation) being dominated by Arsenal F.C. rather than the ammo dump. Certes (talk) 22:20, 22 February 2024 (UTC)[reply]
Let's note for the record that these aren't all identical, though - I don't think there's been serious proposals in recent memory to change most of these, because the long-term significance arguments have been inherently clear, but the discussions at Talk:Telegram (disambiguation) did show we do have a bit of a navigation problem there. --Joy (talk) 21:07, 24 February 2024 (UTC)[reply]

Switching to an -ing form in the middle of a disambiguation list[edit]

So, some dispute with an revert-warring anon has broken out, at the DAB page Hustling Hustler. There's a line item there that has long read:

  • Hustler, someone who deceives others by hustling, usually in sports

following our normal practice of using the term that is the disambiguation page as the line-opening word. The anon absolutely insists on changing this to:

  • Hustling, deceiving people into betting on a game at their disadvantage

though this does not have clear pertinence to the disambiguation Hustling (and if kept in this form would be moved to the "See also" section as a related word). The anon is badly misinterpreting MOS:DAB, but will not take no for an answer, at least from one party (and acuses of "vandalism" if challenged), so it will probably require multiple parties' involvement to settle this silliness. (PS: Their version also has the problem that it changes the definition to refer to games when the term is primarily used with regard to the related but distinct sphere of sports).  — SMcCandlish ¢ 😼  21:23, 21 February 2024 (UTC)[reply]

Seems like a basic rule of list-construction to use parallel forms. I'm not sure where in the MOS that might be mentioned (if it is at all). olderwiser 21:40, 21 February 2024 (UTC)[reply]
@SMcCandlish:, what disambiguation page are you referring to? You linked four times to the article Hustling, but zero times to disambiguation pages. --R'n'B (call me Russ) 21:42, 21 February 2024 (UTC)[reply]
Hustler. olderwiser 21:43, 21 February 2024 (UTC)[reply]
D'oh! Yes, I meant Hustler, which just goes to show that switching to "hustling" in mid-list is confusing.  — SMcCandlish ¢ 😼  23:48, 21 February 2024 (UTC)[reply]
But to be honest, aside from the form, I think the updated description is actually a bit clearer. The former presumes a reader knows what "hustling" is. olderwiser 21:43, 21 February 2024 (UTC)[reply]
I tend to agree with you. The insistence on starting every line with the disambiguated term can lead to awkward formulations when carried to an extreme. It's a good style practice where it flows naturally, but style shouldn't trump comprehension. --R'n'B (call me Russ) 21:47, 21 February 2024 (UTC)[reply]
There is no comprehension problem of any kind with "Hustler, someone who deceives others by hustling, usually in sports" appearing in the disambiguation list of Hustler.  — SMcCandlish ¢ 😼  23:48, 21 February 2024 (UTC)[reply]
Really? Seems a pretty pedestrian error to use a variation of the same term to explain that term. olderwiser 23:56, 21 February 2024 (UTC)[reply]
They could be combined:
  • Hustler, someone who engages in hustling (deceiving people into betting at a disadvantage), usually in sports
That's just a matter of copyediting. The issue here is the attempt to force the -ing form of the word into a list in which is does not belong, and abuse of guideline wording that does not mean what the anon things it means, to try to make that happen.  — SMcCandlish ¢ 😼  23:51, 21 February 2024 (UTC)[reply]
Yes, this is better. olderwiser 23:57, 21 February 2024 (UTC)[reply]
Isn't this a clear case for a one-off exception to MOS:DABPIPE? "Hustler, someone who deceives people into betting at a disadvantage, usually in sports". Link and term are at the left where expected, and there's no circular definition. —swpbT • beyond • mutual 19:46, 22 February 2024 (UTC)[reply]
That's the right approach. The piped link makes it clearer while preserving the preferred form. --Srleffler (talk) 19:10, 24 February 2024 (UTC)[reply]

Sorting entries by clickstream[edit]

At IPA, User:Joy is insisting on sorting the "commonly refers to" list according to the clickstream data rather than alphabetical. Is there a guideline that supports doing this? Has there been discussion on sorting order on DAB pages? Nardog (talk) 02:03, 10 March 2024 (UTC)[reply]

The order for the commonly-refers-to list isn't documented specifically, but the spirit of the text suggests that MOS:DABORDER #4, which allows either way of sorting, should apply to it.
Also FWIW, of the two commonly-refers-to examples, Mercury is in frequency order and Mojave is near-random but closer to frequency than alphabetic. Dan Bloch (talk) 02:55, 10 March 2024 (UTC)[reply]
I've elaborated further at Talk:IPA#clickstreams data.
On a more general note, we know that alphabetical lists work well as a safe option when there's no other way for the reader to predict their traversal of a long list, but in short lists it's unlikely that readers spend a lot of time figuring out sorting.
The key question here would be what constitutes a short list. In the case of top 4 MOS:DABCOMMON items right at the top of the page, it seems easy enough, but we also have various different cases. For example, the WP:NAMELIST guideline has completely decimated lists of people, where practically none get listed up front in short lists, and practically all get listed in long lists. This has had a negative effect on navigation that I've complained about before - wherever we happen to have a vaguely popular item in the short list, it gets way more traffic than a otherwise very popular items in the long list.
Random Brazilian footballer lists, or for example various entertainers like singers or rappers, or in turn minor toponyms and whatnot, effectively spam the top-level disambiguation lists just because they are single-word matches, while biographies with vastly different long-term significance are relegated to long lists, where they're actually much harder to find. A recent egregious example of something like this is Talk:Boyle#post-move.
We'd really have to do some research about what actually makes the most sense for readers (Wikipedia:Readers first). From other web user research it's become generally know that the contents of the first page of text is the most relevant (i.e. scrolling is not great), and that people read typical websites in an L-shaped fashion, from the top left towards the top right and then from the top left to the bottom left (i.e. there's a huge difference in where something is). Obviously another question is if and how the average web user changes their behavior as they become the average Wikipedia reader. The difference between the observed Wikipedia reader clickstreams from the list at IPA and the general Google Search Trends for IPA are an interesting case for this - apparently people at the encyclopedia act differently in this regard than people at the general search engine.
In an earlier thread, I also noted differences in how the existence of a section heading and the length of caption may apparently affect the outcomes of our MOS:DABCOMMON lists. --Joy (talk) 10:52, 10 March 2024 (UTC)[reply]

Generic issue about dab page items[edit]

Please express an expert opinion in Talk:Qasr#Agreed: ONLY places called solely Qasr to be listed. The issue is whether two-part names, such as "proper_name + object_type", e.g., [[<something> Prison]] and the likes, belong to [[<something> (disambiguation)]]. - Altenmann >talk 21:00, 15 March 2024 (UTC)[reply]

The guidance for this is at WP:PTM. --Joy (talk) 22:30, 15 March 2024 (UTC)[reply]