Wikipedia talk:Disambiguation

 You are invited to join the discussion at Talk:Luigi (disambiguation) § Requested move 30 July 2025. -- Joy (talk) 08:49, 25 August 2025 (UTC)

BTW, there's now also Talk:Luigi#post-move.

The outcome in the statistics seems to have been fairly moot. --Joy (talk) 07:27, 15 November 2025 (UTC)

I noticed this discussion at Talk:Se7en (disambiguation) where no less than two out of four people in support of the proposal said it was surprising to see that term disambiguated. I thought it was a reference to the movie as well, but what surprised me in turn is that nobody cared for decades.

We have history that goes only one decade back and over 600k views happened in that time. In the more recent low-intensity period, usually over a thousand views each month still.

When we have RM discussions about disambiguation, and see a thousand hatnote clicks a month, that sounds notable. We often discuss much lower volumes in these discussions.

I'm not saying the editors aren't being genuine when they say they are surprised, but the genuine surprise of a handful of us editing enthusiasts (counting myself in!) probably pales in comparison with a lack of a surprise on the part of a huge body of readers.

This reminds me of the many arguments over the years where it wasn't clear why it would be an impediment for readers to have to click once on top of the list to get to the most popular topic. Like at e.g. Talk:AOC (disambiguation). In summary, it really isn't. The average reader uses the Internet through search engines of all sorts, and is very well aware of the concept of the top item being the top item and they just have to follow through by clicking on that.

I think Wikipedia is the one being the odd one out here, by having a concept of a primary topic, encouraged by a rather open-ended guideline.

I can't help but think that this grew organically from how articles start - a volunteer writes something at "Foo", and then it takes extra effort from another volunteer to assess the title precision and decide on "Foo (disambiguator)" or "Foo Natural Disambiguator" etc and then write a list at "Foo".

We don't want this extra effort to be the default because it's an impediment to editors - but it's largely irrelevant to the readers in the long run.

I think we should revise the text of the primary topic guideline to make it so that we only choose one if there's a clear consensus that there's both one by usage and by long-term significance. Only in that sort of a case, where showing a list has a risk of astonishing readers, is the short-circuiting really worth it. In all other borderline cases, we can't really claim the readers will be anywhere close to astonished.

Trying to gauge efficiency of e.g. 45% of clicks being skipped if we promote a borderline topic to primary is just a weird form of micromanagement. It's especially jarring as we don't ever come back to assess the impact of these promotions. We don't track the readers, we don't poll them to verify if they liked the new navigation, we do none of the due diligence. If someone doesn't like it, they have to become a bit of an editor and complain on a talk page. Which may or may not be monitored by the people who !voted for the change.

Having simpler criteria would probably benefit us as a community, by tempering the constant churn of disambiguation-related RMs of unclear value. Simply empowering editors to be bold and disambiguate, while requiring a more sturdy rationale up front in order to pick a primary topic, would probably be a net benefit for the encyclopedia. --Joy (talk) 08:05, 15 November 2025 (UTC)

Oh, another reason I called this micro'ing is because the same day I had checked on Talk:Luigi#post-move, only to find that the impact of our RMs was not really detectable in our usage statistics. --Joy (talk) 12:16, 21 November 2025 (UTC)

There’s so much misunderstanding here it’s difficult to address. I suspect that’s why no one else has responded.

For now I’ll just say this. Usage stats are often not affected by RMs because most page views come from direct links from external search engine results. That vast majority of users is served by the external search engines. The purpose of primary topic is to best serve the small minority using WP internal search. But the effects of RMs can be seen, over time, in some cases. Sometimes dramatic effects. See Talk:The Americans#Effect on page view counts and the underlying stats in the subsequent section for a classic example.

Some believe that we should not have primary topics; that all ambiguous titles should be, or redirect to, dab pages. After all, searching for “Paris” on WP internal search would be more like searching on Google if the user was taken to Paris (disambiguation) rather being taken directly to the article about Paris, France.

But if that’s what users want, they can search on Google. Identifying primary topics and naming articles accordingly is a valuable service in itself. Someone may not know that the French city is the most likely sought topic for “Paris” searches. Or an article about a TV series is the most likely desired destination for “The Americans”. Or that a borough in Manchester is the most likely sought topic for Reddish. (I certainly didn’t know that one, and I find it interesting to learn that).

Since the historical significance criteria was added I’ve always thought it was redundant and added unnecessary complexity. After all, any topic with true historical significance should be reflected in page view statistics. Thankfully, the singer-actress was established at Madonna before the historical significance criteria was added here. If a star with that name had risen to such popularity only after PT was modified, they’d probably be relegated to a disambiguated title.

So just having the historical significance criteria in there is bad enough, but I’ve accepted it and even use it in my arguments when applicable. But to make meeting it a requirement for PT? That would be terrible. That would justify disambiguating even Paris, as well as Madonna, and countless other well-established primary topic articles. Strong oppose.

— В²C ☎ 13:05, 6 February 2026 (UTC)

@Born2cycle, perhaps you missed Talk:Madonna/Archive 22#Requested move 18 July 2020. The entertainer's article was not always at that title (and it took a LOT of discussions to get there). older ≠ wiser 13:19, 6 February 2026 (UTC)

Wow! Indeed I did, or forgot. But it still reinforces the strong consensus for primary topic. Surely many searching for the Madonna with “Madonna” might be surprised if not momentarily astonished to land upon an article about the pop star, but that’s no reason to disambiguate it. Per consensus. — В²C ☎ 13:46, 6 February 2026 (UTC)

I have no idea why you would say this about Paris and Madonna. Both of the current primary topics there have long-term significance according to our existing criteria, and would not automagically move away from their current destinations just because we would require an examinination of that. Surely if they are in fact primary topics, it would be reasonably easy to demonstrate that they have long-term significance?

Overall, the apparent ease with which you just ended up at this sort of a negative conclusion reinforces the idea that we should have clearer criteria and force people to think about things beyond just !voting.

I haven't seen the discussion about "The Americans" before, and a quick skim indicates it's a whole lot of it, so I'll reserve judgement on that for now. --Joy (talk) 15:32, 6 February 2026 (UTC)

I believe you're misapplying the Principle of Least Astonishment. If it were correct to apply it here, then person looking for information on Paris, France, would be equally astonished to browse to Paris only to find that it's a disambiguation and they have to click one more time to get the info they're looking for. Conversely, if they're looking for a different Paris, whatever reaction they have to finding an article on Paris, France, topped with a hatnote isn't likely to rise to the level of astonishment, confusion, perplexity. It isn't like an Easter egg link, or like a redirect that points to an article that doesn't, in fact, have any information on the redirected title. It will take them where they want to go. It seems to me that their reaction is more likely "Oh! I see" followed by clicking the disambiguation link. Largoplazo (talk) 13:43, 6 February 2026 (UTC)

If we can make a reasonable argument that Paris, France is so significant that people would be astonished if they saw a disambiguation page, and the usage also matches, then there is a clear primary topic and I don't really know what you mean. --Joy (talk) 15:34, 6 February 2026 (UTC)

I'm saying they would be no less astonished than people who aren't looking for the French capital are when they're taken to the article that is about the French capital, see that that's what it's about, and also see the hatnote. To the extent that any of this is astonishing at all ("astonishment" being an overwrought characterization of the reaction people have to disambiguation pages and hatnotes) the suggested change wouldn't be an improvement. It would be replacing astonishment with astonishment.

Have there been floods of messages at the village pump and the teahouse from readers moaning about their astonishment? Largoplazo (talk) 16:32, 6 February 2026 (UTC)

Well, the key distinction is that in the case of a primary topic, that contingent of people is just too small, and therefore it doesn't matter. I wholeheartedly agree with you that very few people would ever be astonished. Most of the time, nobody cares, and the issues are mundane, niche. We need a standard that works well for those cases. The current one leaves too much to interpretation and encourages senseless loss of volunteer time on flamewars. --Joy (talk) 19:18, 6 February 2026 (UTC)

Yes, change might be needed; however, and this is just a thought: I haven't heard the term "flamewars" for a long time. I respect this revolutionary reference work, this Wikipedia, and have respected it since about 2006 when I first started editing it as an IP. I have found that the building of an encyclopedia largely hinges on "disagreement". That is the origin of much of the improvement that has taken place here. There have been countless times when editors disagreed and were able to eventually come to some meeting of the minds. This situation is one in which there has long been implicit and explicit consensus, which has always been subject to change. In short, disagreement is actually a good thing that has led to this encyclopedia as it exists today. We should never discount the value of arguing editors, even if fire is ignited and ice melts and flows. P.I. Ellsworth , ed. – welcome! – 19:56, 6 February 2026 (UTC)

Well said! — В²C ☎ 20:20, 6 February 2026 (UTC)

There's a possibility of healthy disagreements, sure. And then there's a possibility of a very small number of editors anonymously arguing on the internet for sport, without actually helping readers. --Joy (talk) 20:46, 6 February 2026 (UTC)

There are also those who fail to assume good faith. —В²C ☎ 07:07, 7 February 2026 (UTC)

The assumption of good faith only goes so far when we have to spend years having to correct the same data misinterpretations over and over again. We should make sure our standards reflect what we've learned and the best current practice, and enable that, not vibe-based disputes. --Joy (talk) 08:52, 7 February 2026 (UTC)

Oh Joy! you are both correct and a tiny bit incorrigible. WP needs more of you. You seem to think "the key distinction is that in the case of a primary topic, that contingent of people is just too small, and therefore it doesn't matter." Then you go on to say, "there's a possibility of a very small number of editors anonymously arguing on the internet for sport," and they matter very little as well. When building a great reference work, what matters are the five pillars and the struggle to achieve consensus when any disputes arise. My own weakness is that I'm still working on the building of consensus. Many parts of it still mystify me. As a closer I've become fairly adept at recognizing consensus; however, I still fall short when it comes to actually building and achieving it. We all have our strengths and our weaknesses. Fortunately, I've found that when my weakness shows, there's always another editor who comes in and improves my work. And that's really what WP's all about deep down. We let others improve our work, and that's not an easy thing to get used to. I do like your style; I like it as much as I like WP's style. P.I. Ellsworth , ed. – welcome! – 19:22, 7 February 2026 (UTC)

Well, that is an issue of context. The encyclopedia is huge and the standards apply to everything, so small amounts of incongruity don't really matter. However, the community of volunteers who maintain that is not huge, and even relatively small amounts of wasted work do matter. --Joy (talk) 09:20, 12 February 2026 (UTC)

GBR -> GBR (disambiguation)

GBR will then re-direct to Great British Railways

Most articles in the disambiguation page aren't really referred to that term as much as Great British Railways. Maybe it might be too early as the scheme hasn't been fully established, but it is something that can be considered in the future, hopefully now. The page is over four years old anyway now. Fortek67 (talk) 19:24, 9 December 2025 (UTC)

I would like to propose moving the article Heidi to Heidi (novel), since the animated television series Heidi, Girl of the Alps has comparable notability and cultural relevance. ~2025-40709-46 (talk) 14:37, 15 December 2025 (UTC)

It would be best if you posted to Talk:Heidi instead. --Joy (talk) 14:40, 15 December 2025 (UTC)

Done, thanks! ~2025-40802-25 (talk) 15:10, 15 December 2025 (UTC)

I stumbled upon an oddity:

• Time travel is a primary topic page (lower case t)
• Time Travel is a disambiguation page (upper case T)
• Time Travel (disambiguation) redirects to Time Travel (upper case T)
• Time travel (disambiguation) redirects to Time Travel (lower case t)

Is this reasonable? I guess that the justification could be that "Time Travel" implies the name of a work but "time travel" is the concept? Johnjbarton (talk) 17:08, 2 February 2026 (UTC)

That is standard WP:DIFFCAPS procedure. People never type "Time Travel" to refer to the concept, unless they have input a typo. It is far more likely to refer to the title of a work, therefore it is a DAB page. ᴢxᴄᴠʙɴᴍ (ᴛ) 09:48, 5 February 2026 (UTC)

After engaging in the umpteenth discussion about WP:NAMELIST, this time at Weizmann (disambiguation), I went and reviewed the guideline text again. It currently says:

To prevent disambiguation pages from getting too long, articles on people should be listed at the disambiguation page for their given name or surname only if they are reasonably well known by it.

There's a few problems with this:

In the case a disambiguation list exists but no separate given name or surname list exists, i.e. with a list tagged {{disambig|given name|surname}}, it could imply that some people should be de-listed altogether. This doesn't usually happen, but what does happen is that the entire list of people is split out. This in turn leads to several bad outcomes:

• we necessarily bury biographies behind another click
• people start thinking of the two lists as separate topics as opposed to an implementation detail
• people start thinking that all the entries in the split-out list are inherently partial title matches
• people stop thinking of the possibility that we should maybe just not have separate lists

Furthermore, it's not really clear what the standard is for reasonably well known by it. The guideline tries to clarify that by continuing with this:

We reasonably expect to see Abraham Lincoln at Lincoln (disambiguation), but very few sources would refer to the waltz composer Harry J. Lincoln by an unqualified "Lincoln", so he is listed only at the Lincoln (surname) anthroponymy article.

The example of one of the best known American presidents tilts the scale here. Most people are not known by their unqualified name.

The guideline can't just cover the e.g. 1% of most popular topics. It has to be helpful for the bland, low-popularity cases, too, which are inherently the vast majority of ambiguous topics in the encyclopedia.

To clarify - this is the comparison of page views of those items:

• Weizmann Institute of Science - 168 views/day
• Chaim Weizmann - 430 views/day
• Ezer Weizman - 172/day
• Martin Weitzman - 23/day

Whereas for the Lincolns we have:

• Abraham Lincoln 16,166 / day
• Harry J. Lincoln 3 / day
• ...and another 4 screenfuls of other Lincolns inbetween those two, including 3,241 / day, 1,999 / day, etc
• ...and a bit of a long tail after Harry too

So not only is Abraham so high up that it bears no resemblence to the Wiezmann situation, Harry is so low that it doesn't match it, either.

Furthermore, using a comparison of 16k and 3 to set a standard is just plain weird, because that's a difference of about four orders of magnitude. If it needs to be said, that's huge.

And it doesn't even help us really discern what to do in the example of the list of Lincolns, let alone elsewhere. --Joy (talk) 06:32, 5 February 2026 (UTC)

"people start thinking that all the entries in the split-out list are inherently partial title matches" - because they are? DAB pages are meant to resolve ambiguity caused by 100% completely identical names. Someone named "Harry Lincoln" is not confused with "Lincoln". Set list articles of names are intended as navigation guides, but can't be a competing primary topic in a DAB page unless the name itself is notable, nor can the names be competing primary topics if they aren't known mononymously as that name. ᴢxᴄᴠʙɴᴍ (ᴛ) 09:44, 5 February 2026 (UTC)

No, these pages are not "meant to resolve ambiguity caused by 100% completely identical names". We don't even have to talk about human names in that regard. If this was so, much more fundamental things would be different:

• We would never have WP:PRECISION and WP:DABCOMBINE discuss variations
• WP:RPURPOSE would not include half as many purposes
• We would never have the {{distinguish}} hatnote in over 100,000 pages
• Talk:Courtney Cox (disambiguation)#Requested move 18 March 2023 would never have been decided in favor of a typo redirect

The idea that readers approach the names of topics with anything approaching 100% strictness would be completely detached from the practical reality of Wikipedia.

It's really frustrating trying to reason when there is this big of a disconnect in the premise. Do you see how this is problematic? --Joy (talk) 11:18, 5 February 2026 (UTC)

Actually, the main reason disambiguation is required on Wikipedia is precisely because only one article can be located at a specific title. We generally do not include partial title matches. The vast majority of persons are not commonly known solely by their surname alone (let alone by given name alone). The inclusion of such name lists, IMO, has always been a bit of a compromise in that there is some value in providing readers a way to search for people by surname like a directory. Because of the similarity in function with disambiguation pages, these lists are allowed on disambiguation pages, but as the guidance indicates, they may be split off to separate index pages. IMO, given name lists should never appear on dab pages except for those persons such as royalty or performers who are commonly known by their given name alone. But I agree, the criteria for inclusion of any individual or splitting is fuzzy, but I'm not sure how it can be clarified to the satisfaction of all interested parties. older ≠ wiser 12:34, 5 February 2026 (UTC)

I think the fundamental problem is that we like to think about a narrow scope of disambiguation, one that we organically defined ourselves through editorial consensus rather than a more general consensus of editors and readers, as a solution to reader navigation between article titles.

There is no proof that this is so. Rather, the statistics that we have point to readers coming in through external search engines a whole lot instead. That is, they do not arrive through our disambiguation mechanisms, but pre-filtered. The steady churn of requested moves likewise indicates that the levels of organic consensus are not great, esp. when we discuss titles untouched for two decades and editors are surprised how something is so 'wrong'.

Speaking of which statistics we have - we do not seem to have public statistics about reader navigation through our own search engine, nor do we have overall stats about the disambiguation mechanisms like total views of disambiguation pages compared to other views, or clicks on hatnotes compared to views or other clicks. The clickstream data is anonymized to the extent that it's often too vague to be useful on anything but the most popular pages. There is so much that we don't know within the scope of things we know about, let alone if there's things we don't even think about.

The conclusions that we draw from trying to observe our disambiguation mechanisms in this vacuum are necessarily biased. Judging by the amount of discussions where I've had to correct wrong interpretations of WikiNav output, most of the time our conclusions tend to be simply wrong.

We think we know what we're doing with all of our editorializing, but we don't. The sooner we disabuse ourselves of this harmful notion, the better. --Joy (talk) 09:19, 6 February 2026 (UTC)

And just as a tangentially related aside, I just came across Nagele (disambiguation) which is a surname list using disambiguation naming convention. It had been a disambiguation page (and does have a non-person primary topic and at least one non-person partial title match place name), but apparently because it is mostly a list of names, it was recategorized as a surname page. Should this be renamed to Nagele (surname)? Should this page give some guidance about such cases (while not common, they are not exactly unusual either)?

I just boldly moved the page to (surname) as it is clearly marked as a name list rather than a DAB page. I think that policy implies that something only be a dab when there are other non-names to disambiguate. ᴢxᴄᴠʙɴᴍ (ᴛ) 09:39, 6 February 2026 (UTC)

Yeah, this is a good illustration of how we just assign too much meaning to details that probably don't affect most readers. That list started by saying it's a surname, and listed the biographies, but then mentioned another use of the Nägele Palace, which is its own topic that was named after the surname of a person called that way.

Fundamentally, the properties of these elements don't really matter as much as the fact that we need to serve e.g. all of these people:

• someone who reads a citation to Nagele in a book and wants to look up who that author was
• a reader who is traveling the Netherlands and comes across a sign saying Nagele and wants to look up that place
• someone walking in Timisoara observing a palace marked Nägele and wants to learn more about that

None of these reader contingents seem terribly more numerous or important than the others, and there's no real value in making it easier or harder for any of these people to navigate. --Joy (talk) 09:52, 6 February 2026 (UTC)

This is a followup to these earlier examples:

• Wikipedia talk:Disambiguation/Archive 56#a change in page views between primary topic and primary redirect
• Wikipedia talk:Disambiguation/Archive 57#followup to how page views can change between having and not having a primary topic or primary redirect

This time I've observed it at Talk:Telegram (disambiguation)#followup to move discussion. When telegram was moved from being a primary redirect to a section in a larger article, into being a standalone article, the views of the article instantly jumped > 2x, and the views of the statistical redirect in the hatnote - that is, clicks on the hatnote - jumped > 4x.

I theorized before that a section hatnote is not as visible as the top-of-the-article hatnote, because there's less spacing around it, and the surrounding layout is somewhat different (the heading is smaller and has an extra blue edit link). But I don't think this sort of an instant and consistent change of over 100% on this data point can be easily attributed to just this difference in formatting.

It's got to be coming from input pre-filtering. It really feels like we're just indirectly pushing the buttons on the statistical engine / algorithms within Google Search with our navigation changes. It's like our different formatting merely 'competes for its attention' differently. And from this perspective it's hard to say if we're actually doing something well, or if it's just observing some vaguely random effects.

If any Google Search engineers or product owners would ever happen to be seeing this, your input would be most welcome. (If you're not comfortable doing it in a public forum like this, you can also email me through my user page.) --Joy (talk) 15:23, 24 February 2026 (UTC)