Wikipedia talk:Disambiguation

"Covered by"

I just noticed the comment from B2C on Talk:On My Way (Charlie Brown song) that "ambiguity [on WP] has no meaning other than WP article name conflict", as stated in the first sentence of WP:D. --B2C 06:27, 12 August 2013 (UTC)"

Disambiguation in Wikipedia is the process of resolving the conflicts that arise when a single term is ambiguous—when it refers to more than one topic covered by Wikipedia articles.

Perhaps we should expand "covered by Wikipedia articles" to "covered by the text of Wikipedia articles" since B2C (a self proclaimed "title expert") is reading "covered by Wikipedia articles" as "covered by the title of Wikipedia articles" - ? Seriously. Is this sentence open to this misunderstanding? In ictu oculi (talk) 07:26, 19 August 2013 (UTC)

I am sorry if that sounds catty against B2C, he's not the only one misciting this by any means In ictu oculi (talk) 01:40, 21 August 2013 (UTC)

• Many of us would dispute that it is indeed a misunderstanding. On the specific subject of songs, it just doesn't make sense to displace an actual article in favor of a dab that lists song titles linking to, at best, an album page. While there are occasional exceptions, standalone articles generally get preference over list items (cf. the semi-recent Larry the Lobster RM). And indeed, "list item" is probably the best we can call most album tracks. If essentially all we can say about a song is that it exists, again, why make it harder for readers to find an actual article on a song of the same name? If I want to find out about, say, DJ Khaled's "On My Way", I'm going to find out there isn't an article, and I'll be disappointed. Short of creating an article (if it meets notability standards), there's nothing really we can do about that. But we can choose to streamline the experience for readers searching for songs that we do cover.

What I don't know is where MOS:DABMENTION comes into play. Should a song only be mentioned on a dab if there's a reasonable expectation that it will have an article? That seems extreme, but might actually be an area of compromise on this issue. Personally, I think such listing are still useful as navigational aids. --BDD (talk) 03:25, 21 August 2013 (UTC)

I think the way we usually do it - having a line for the song with a link to the album where it is covered - is signal enough that we don't have an article on the song. Anyone who is going to be disappointed by that will be disappointed whether they learn of it on the disambiguation page, or in further fruitless searching. I support having those links on a disambiguation page if one exists, but would not bump a song out of the primary topic spot to create one, nor would I bother to create one if we only had one article whose title was the song. bd2412 T 04:00, 21 August 2013 (UTC)

BDD, hi, if you think that B2C is understanding the guideline correctly, shall we put it to an RfC?

Re songs it's a particular problem with songs because we have two forms of article handling songs, single articles and album articles; constantly notable album tracks by notable artists will have no article. Then we get left with Example (unknown artist single) vs. Example (Rolling Stones album song). WP:DAB ... an alternative to an RfC would be to enshrine a Example (unknown artist single) vs. Example (Rolling Stones album song) in WP:DAB to show we don't require standalone articles for songs to be notable. In ictu oculi (talk) 06:34, 21 August 2013 (UTC)

• I don't know the original intention of the text or whether many people believe it, but I think the language use is a very bad idea. It is a bad idea to redefine common words for Wikipedia purposes. Doing so hurts the accessibility of the project to newcomers. Ambiguity is most likely to be understood as real-world ambiguity to a newcomer. If the intent is to describe technical article titling ambiguity, then describe it as technical article titling ambiguity and don't redefine dictionary words. --SmokeyJoe (talk) 06:50, 21 August 2013 (UTC)
  • There is no redefining here... Only an understanding of context.

    Contrary to what In ictu oculi thinks I think, I know "covered by WP" does not mean "covered by the title of Wikipedia articles". Some topics are covered in subsections of articles. WP's process of disambiguation applies when a term can be used to refer to more than one topic covered by WP, whether that topic is the main topic of an article or a sub-topic of an article.

    For example, we don't have an article about the kidnapper named James Dimaggio, but he is a sub-topic of Kidnapping of Hannah Anderson. Therefore, if we we ever need to cover another James DiMaggio, that name will be subject to our process of disambiguation. But, for now, despite the known existence of many other James DiMaggio's, as long as none have sufficient notability for coverage on notability, this James DiMaggio remains the primary topic, and James DiMaggio redirects to the article that covers him as a sub-topic.

    To clarify this, we could change the current wording to say the following:

  Disambiguation in Wikipedia is the process of resolving the conflicts that arise when a single term is ambiguous—when it refers to more than one topic covered by Wikipedia articles (either as the main or a secondary topic of an article). Terms that are ambiguous only with uses not sufficiently notable to have coverage on Wikipedia are not subject to Wikipedia's process of disambiguation.

--B2C 17:01, 21 August 2013 (UTC)

I agree with SmokeyJoe that redefining "ambiguity" to refer only to a restricted subset of its English meaning is a bad idea. I like the addition in parentheses, that clarifies that ambiguity is not something restricted to article titles. The second sentence, though, is contrary to common (and quite recent) practice at DAB discussions, and B2C knows that this line of reasoning has been strongly opposed by several editors. So let's keep definitions short and allow editors to apply guidelines a they see fit for each particular case, and reduce WP:CREEP by making guidelines that are descriptive instead of prescriptive. Diego Moya (talk) 19:16, 21 August 2013 (UTC)

Wikipedia redefined "ambiguity" long ago, much like Wikipedia redefined "consensus" long ago. Nothing inherently good or bad about the ideas. And we can (and do, I think) avoid the (English) ambiguity by normally referring to the (Wikipedia) ambiguity as "Wikipedia ambiguity". -- JHunterJ (talk) 20:01, 21 August 2013 (UTC)

I'm not sure about the phrase "the main or a secondary topic of an article"; what is a secondary topic? How about "as the subject of an article, or discussed in an article"? I think the second sentence is fine. As a matter of practice and good policy, we do not list terms on disambiguation pages unless they are in Wikipedia, or clearly belong in Wikipedia. bd2412 T 20:13, 21 August 2013 (UTC)

I gave an example of a secondary topic above.... James DiMaggio is covered as a secondary topic of Kidnapping of Hannah Anderson (of which the main topic is the kidnapping itself). But your wording is clearer. --B2C 20:51, 21 August 2013 (UTC)

Yes, the ships have sailed on redefining "consensus" and "ambiguity" long ago. There are countless words written with these redefined meanings implied by the uses of these terms, in comments as well as in policy and guidelines. I don't know where or how we would even begin to change this.

I should note that almost every endeavor, from science to politics to law to engineering, requires specialized terminology that includes "redefining" commonly used words to have specific narrow meanings in the relevant contexts. Remember, this is not terminology that readers encounter - it's terminology for efficient communication among the editors. Insisting on using the dictionary definitions of these words rather than the specialized WP meanings in WP editorial discussions is not helpful. --B2C 21:00, 21 August 2013 (UTC)

Redefining common words is inherently bad in terms of accessibility of Wikipedia to new editors. Requiring enculturation is bad, as should be minimised. Ambiguity and consensus are not lost, B2C in particular is particularly wrong about WP:Consensus. The ship may have sailed for WP:Notability, but the precedent is no justification. --SmokeyJoe (talk) 21:34, 21 August 2013 (UTC)

It stills seems there is a problem here MOS:DABMENTION allows mention of e.g. album tracks, but then when it comes to titles we count those MOS:DABMENTION as if they don't exist? In ictu oculi (talk) 12:16, 23 August 2013 (UTC)

It's not about pretending they don't exist; it's about not giving them primacy. Once we resort to parenthetical disambiguation, naming becomes a matter of real estate, constrained by the inability to have two articles with the same name. So when deciding those titles, it's entirely right and proper to only compare against other titles. We shouldn't have an article Foo (Bar song) without an article Foo (song), any more than we should have an article Foo (topic) without an article Foo. --BDD (talk) 17:25, 23 August 2013 (UTC)

BDD says "We shouldn't have an article Foo (Bar song) without an article Foo (song), ...." So if I understand this correctly, you are saying at least one song must remain ambiguous before you can disambiguate other songs? That doesn't make sense. --Richhoncho (talk) 10:29, 24 August 2013 (UTC)

That's just plainly wrong, BDD. Any topic may be the primary topic for a number of different terms. Only one of those terms can be the title for the article about that topic - all the other terms for which that topic is the primary topic must redirect to that article. For example, New York City is the title, but the primary topic of "New York, New York" is also that article, and so New York, New York redirects to New York City despite the existence of many other uses of "New York, New York". If it were true that "We shouldn't have an article Foo (topic) without an article Foo", then we couldn't have New York, New York (film) since we don't have an article at New York, New York. --B2C 05:59, 26 August 2013 (UTC)

But Why? Doesn't the first line of WP:DAB say exactly that that we don't disambiguate by titles. We could have Eric Blair (Ontario politician) (in fact it's also a redirect) with no Eric Blair article, for obvious reasons. So why can't we have Under the Sun (Cheryl Cole song) (before moved by Unreal7) when the other 9 songs are only mentioned on albums? We can and do have articles Foo (topic) without an article Foo. In ictu oculi (talk) 04:31, 24 August 2013 (UTC)

Petitio principii, my friend. I don't read the first line of WP:DAB that way, nor am I alone in not doing so. With the Eric Blair example, I half feel like you've tricked me, but really it's my fault for making too sweeping a statement. Of course as a matter of WP:PRIMARYTOPIC, we can have a Foo (bar) if the primary topic for Foo is the article Bar (I believe there are editors out there who may disagree with that, though they are decidedly a minority). I don't see the relevance of this sort of case to song articles, however. --BDD (talk) 05:01, 26 August 2013 (UTC)

I think I see what you mean now. We should not have an article at Foo (disambiguation) if we do not have at least one other use on WP for 'Foo'" (whether that use is represented as an article title, one of several on the dab page for "Foo", or as a redirect, does not matter). Similarly, we should not have an article with title Foo (Bar song) without an article or redirect at Foo (song). If that's what you meant, I agree.

By the way, the relevance of your sweeping statement error to song articles is that we have New York, New York (Moby song) even though we have no article at New York, New York (song) (it's a redirect to the Music section of New York, New York (disambiguation)). --B2C 06:21, 26 August 2013 (UTC)

I think part of what In ictu oculi may be getting at is there appears to be a bias among some editors that if there is only one "article" about a song, that by default becomes the "primary topic" for the title Foo (song) regardless of whether there are other songs with that title on the Foo dab page (and without any discussion about primacy). older ≠ wiser 11:02, 26 August 2013 (UTC)

Yes, we can and do have articles titled disambiguated as Foo (disambiguation) without an article at Foo, as long as Foo is a redirect to the primary topic for the term "Foo", or "Foo" has no primary topic and so the Foo dab page is at Foo. --B2C 05:51, 26 August 2013 (UTC)

• In my view, if a subject is listed on a dab page (e.g., the title of a song) with a link to where an article is found that discusses that subject in any depth whatsoever (even just one sentence within a larger article that provides a context for the subject, and perhaps even just identifying the subject explicitly without further elaborating on it), or (similarly) if a redirect exists for a subject that points to an article that discusses that subject in any depth whatsoever, then that subject is "covered" on Wikipedia. I'm not sure if that's what B2C means by "a secondary topic of an article" or not. The sentence "Terms that are ambiguous only with uses not sufficiently ..." seems wordy and unnecessary. I really hate the phenomenon where someone claims that one particular song or album should be considered the primary song (or primary album) with a particular title and should therefore be moved to the ambiguous title "My Favorite (song)" instead of residing permanently at "My Favorite (Particular Artist's Name song)", when there are several other songs or albums listed somewhere on Wikipedia that have the same title (especially when the candidate primary article has little depth or was recently released). That just creates an unending churn of arguments about whose song is the most important and is why it is a generally good idea to avoid partial disambiguation. —BarrelProof (talk) 12:52, 23 August 2013 (UTC)
  • Yes, if a subject is discussed "in any depth whatsoever (even just one sentence within a larger article that provides a context for the subject, and perhaps even just identifying the subject explicitly without further elaborating on it)", as demonstrated by a dab page link, redirect, or hatlink to that discussion... that's what I mean by "a secondary topic of an article". The only reasonable interpretation of "topic covered by WP" I can fathom is any topic which is either a main or secondary topic of an article (using "main" instead of "primary" here to avoid confusion of "primary topic" of an article with "primary topic" of a term). --B2C 23:30, 25 August 2013 (UTC)
    • I have added a foot note clarification regarding what "topic covered by Wikipedia" means. It follows the first sentence of the page. The text of the note "hovers" over the cursor when the cursor is moved over the foot note. I hope that addresses the issue initially raised in this section --B2C 06:48, 27 August 2013 (UTC)

Yes, this should fix the main problem. Thank you. In ictu oculi (talk) 03:00, 31 August 2013 (UTC)

So let's get on and add it? In ictu oculi (talk) 23:26, 10 September 2013 (UTC)

The footnote's already there. To me, it seems fine. —BarrelProof (talk) 23:39, 10 September 2013 (UTC)

Sorry, I knew that but forgot, I guess that illustrates User:Cuchullain's comment in relation to Talk:Good for Me (song) that the footnote is both recent and not very visible. It seems to me that songs represent a special challenge for WP:DAB in that for few other prolific subjects we have the (standalone song) (song on a collection of songs) divide. We don't have to disambiguate [Village Z (standalone)] from [Village Z (mentioned in a album of villages)] for example. Doesn't happen for bios much, though in theory I guess it could : [John Smith (standalone musician)] vs [John Smith (musician in a band article)], [John Smith (standalone hockey player)] vs [John Smith (hockey player in a team article)]. Less of a perennial problem than for songs though, given that most songs, by far, don't have articles, but still have Google Book and iTunes hits. In ictu oculi (talk) 00:52, 12 September 2013 (UTC)

Now it is part of the page it seems reasonable to add into the main text, rather than hiding away as a footnote. After all it's addition has already been debated above. --Richhoncho (talk) 21:46, 12 September 2013 (UTC)

@Richhoncho: please go ahead. In ictu oculi (talk) 04:13, 18 September 2013 (UTC)

(entertainer)

Moved to Wikipedia talk:Naming conventions (people) as suggested.
The following discussion has been closed. Please do not modify it.
The questions for this RfC are: 1) should we continue to encourage/allow the use of "(entertainer)" as a disambiguator in entertainment-related articles, and 2) if so, under what circumstances. Dohn joe (talk) 06:12, 31 August 2013 (UTC) Recent months have seen a number of requested moves to replace the "(entertainer)" disambiguator. Results have been mixed. Each individual RM has had relatively limited participation, so it would be nice to see if we can come to wider consensus on the issue, or if going case-by-case is necessary. A sampling of recently concluded RMs Talk:Craig_Owens_(singer)#Requested_move Talk:Mario_(American_singer)#Requested_move_1 Talk:Psy/Archive_1#Requested_move_2 Talk:Ashanti_(singer)#Requested_move_21:28.2C_21_July_2013_.28UTC.29 Talk:Usher_(entertainer)#Requested_move_.28August_2013.29 Talk:Chris_Brown#Requested_move_3 Why drop (entertainer)? It can be a vague, ambiguous catchall that in itself is not very descriptive. It's often only used sparingly in sources to describe a given person. One editor has suggested that the term only rightly applies to jesters, vaudevillians, burlesque acts and the like. Why keep (entertainer)? In the case where a person is notable across multiple areas of entertainment (i.e., as a singer and a dancer), choosing one of those areas to disambiguate may be misleading or less accurate than using an umbrella term. There are two fundamental questions to answer first. One is the definition of the word itself. One definition of "entertainer" is "a singer, comedian, dancer, reciter, or the like, especially a professional one". This would seem to at least allow for the possibility of its use on WP. The second question is how important WP:AT is in choosing a disambiguator. We rely heavily on sources for the actual name of a subject. Is that reliance lessened when we choose disambiguation? Does "entertainer" itself need to be found in sources, or can it be used to disambiguate when it makes sense otherwise? What are the options? Eliminate (or strongly discourage) "(entertainer)". There is (almost) always a better option. Use "(entertainer)" only in the rare case where a person is most often referred to as an entertainer, or where no other reasonable alternative exists. Allow "(entertainer)" only in cases where no particular area of entertainment predominates over another. Allow "(entertainer)" whenever a person is notable for more than one area of entertainment. Allow "(entertainer)" even when a person is notable for only one area of entertainment. Disambiguate using each area for which a person is notable (i.e., "(singer/actress)" or "(comedian/musician)"). Please feel free to alert the relevant WikiProjects or others. Dohn joe (talk) 06:12, 31 August 2013 (UTC) Survey Please indicate the number(s) of your preferred options here, along with a brief explanation if you wish. Leave it alone - per WP:CREEP - for the time being at least. The majority of RMs (around 20 of them) have already gone against use of "entertainer" as a neologism and/or imprecise per Talk:Drake (rapper), Chris Brown, Talk:Sinbad (comedian) etc. The only 2 exceptions as far as I know are Talk:Madonna (entertainer) and Talk:Usher (entertainer). Although for certain pop-star articles there is strong WP:LOCALCONSENSUS in favour both of the FOO (entertainer) title dab, and "the entertainer FOO" in leads", fundamentally (entertainer) is an umbrella Category:American entertainers which per WP:CRITERIA for recognizability and sources can normally be entrusted to WP:RS without legislation. If editors recognise the principle of using language found in sources they hopefully won't hold out for "the entertainer FOO" against "the actor FOO" "the singer FOO" forever on the basis of 1% or 2% of sources. In ictu oculi (talk) 08:43, 31 August 2013 (UTC) Incidentally we do have some "and" dabs already on en.wp Junaid Khan (actor and singer). In ictu oculi (talk) 08:56, 31 August 2013 (UTC) Wrong project. The naming conventions for articles determine the titles (including qualifiers, if the desired title is ambiguous and not primary). The disambiguation guidelines determine the titles of disambiguation pages. This should go to WP:NCP or similar. -- JHunterJ (talk) 14:19, 31 August 2013 (UTC) Discussion

Needed: a simple "how to create a disambiguation page" section

I spent a few minutes googling "how to create a disambiguation page" and variations thereof. There's no simple, easy-to-follow guide for how to figure out how to change an old redirect https://en.wikipedia.org/w/index.php?title=Britten&redirect=no to a simple disambiguation page that also includes https://en.wikipedia.org/wiki/John_Britten and https://en.wikipedia.org/wiki/Britten_Motorcycle_Company and https://en.wikipedia.org/wiki/Britten_V1000 If we on Wikipedia want more and new users, these arcane and complicated information pages really need to be streamlined and user-friendly and convenient. Spending half an hour learning intricate syntax is fine if you want an exclusive boys club, but not terribly convenient for the amateurs. Could also start with automatic signing of posts like this one, lest I forget the four tildes. Pär Larsson (talk) 14:38, 3 September 2013 (UTC)

That is an excellent idea. We might be able to set up a "preloaded" disambiguation page template, along the lines of:

Foo may refer to:

• [[Foo (bar)]], a bar
• [[Foo (yang)]], a yang in Ruritaria
• [[Foo (album)|''Foo'' (album)]], an album by the Ruritarians

{{disambig}}

I have asked whether this is possible at Wikipedia:Village pump (technical)#Creating a form to make a new disambiguation page. Cheers! bd2412 T 15:23, 3 September 2013 (UTC)

Have you asked WT:WPDAB? They are likely to have considered such matters before. --Redrose64 (talk) 07:47, 4 September 2013 (UTC)

@BD2412: excellent, do it. In ictu oculi (talk) 04:15, 18 September 2013 (UTC)

Yunak (disambiguation)

Second opinion and action requested on this from experts in MOSDAB. Staszek Lem (talk) 16:24, 19 September 2013 (UTC)

I think you're probably right for the time being. Since the term isn't mentioned anywhere at Slavic mythology, the Slavic word isn't a very useful inclusion there, and since dabs don't have reference sections, information included there should be backed up by article content. I can't read Bulgarian, so for all I know, the whole thing is a hoax. Perhaps Bigzteve should create a sourced stub for Yunak (mythology), which could be listed much more elegantly on the dab anyway. --BDD (talk) 16:37, 19 September 2013 (UTC)

WP:PDAB

An attempt to modify, keep, or remove the partial guideline would have resulted no consensus, especially since no one closed it. Also, I tried to rename The Price Is Right (U.S. game show), but the proposal failed. This calls for further discussion before we can make a survey. --George Ho (talk) 04:15, 21 September 2013 (UTC)

It was archived automatically due to nobody wrote in the section for a few weeks, but that doesn't mean we can't re-post it here. Also, it was requested to be formally closed but no one has done so. Tbhotch.^™ Grammatically incorrect? Correct it! See terms and conditions. 04:33, 21 September 2013 (UTC)

Per Wikipedia:CONSENSUS#Consensus: "In discussions of proposals to add, modify or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit". There is no differing guideline for the addition of policy without consensus; therefore, a lack of consensus as to this policy means that it gets removed. I will do so now. bd2412 T 01:09, 23 September 2013 (UTC)

• User:BD2412, there was a large majority supporting saying clearly what the WP:DAB guideline already said anyway. Removing it is an invitation to the same old nonsense about WP:PRIMARYCHEESE, WP:PRIMARYICEHOCKEYPLAYER starting again... except it won't start because it is only pop music where the agitation for ambiguous disambiguation is happening. Talk:Thriller_(Michael_Jackson_album)#Requested_move_4. This nonsense could be killed quickly by someone submitting a RM for selecting one of dozen footballer at John Smith to be WP:PRIMARYFOOTBALLER. In ictu oculi (talk) 03:30, 24 September 2013 (UTC)
  • Removing the name of an artist from the title of an article about the artist's song or album seems exceptionally unhelpful to me. Most readers would probably actually rather see the title "Arbitrary Phrase (Dohn Joe song)" than "Arbitrary Phrase (song)". Not including the artist's name makes the whole thing rather confusing and generally less recognizable, and dooms us to perpetual future discussions, instability, and broken cross-references when different artists have songs of the same name and when new songs appear as time moves forward. The ebbs and flows of popularity guarantee that any identification of a "primary song" will be impermanent and will be challenged with each new release. (The "primary cheese" example gave me a bit of a giggle – thanks for that.) —BarrelProof (talk) 04:50, 24 September 2013 (UTC)
  • @In ictu oculi, I would be delighted if you could give me a headcount showing this "large majority". bd2412 T 11:22, 24 September 2013 (UTC)

This discussion has moved around all over the place, I'd be delighted also if I knew where it was. In ictu oculi (talk) 11:26, 24 September 2013 (UTC)

Nevertheless, if you claim to have support for the proposition, it is your burden to demonstrate this support. I have not seen it. Perhaps a new discussion is in order. bd2412 T 14:11, 24 September 2013 (UTC)

I thought this discussion at the village pump inserted PDAB and although there is much disagreement there is no consensus to remove. Cheers. --Richhoncho (talk) 15:43, 24 September 2013 (UTC)

It is, quite frankly, not at all clear what there is consensus for from that discussion, since the parameters of the discussion changed while it was ongoing. For example, it remains unclear from that discussion whether Kiss (band) would need to be renamed, and if so whether that title would have to be redirected to Kiss (disambiguation), or to the renamed American band. It is impossible to say that consensus was achieved if the question can not be answered from reading the entire discussion as to what consensus was achieved. Furthermore, in the follow-up discussion now at Wikipedia talk:Disambiguation/Archive 40#WP:PDAB, which received much broader participation and is therefore a fairer reading of community consensus, only three editors supported keeping PDAB as it was at the time (Dicklyon, In ictu oculi, and 76.65.128.222); nine editors supported removal of PDAB (Tbhotch, BDD, B2C, Obi-Wan Kenobi, Zanhe, Lil-unique1, Enric Naval, David Eppstein, and Amakuru); another editor (Neelix) stated, "I would be happy for WP:PDAB to be removed so long as WP:INCDAB remains in tact"; and six editors proposed that PDAB be revise, clarified, or modified to allow titles like Kiss (band) and Thriller (album) to remain the titles of the best known users of those names, which would effectively reverse what PDAB said at the time - JHunterJ wrote "For disambiguation, either allowing or permitting a "primary song" will work"; I, BD2412, supported replacing it with a policy that would allow Kiss (band) to stay where it is; WikiRedactor propose that "this situation should be handled case-by-case", and that "If it remains, WP:PDAB should only be a guideline to suggest a possible solution"; Jackson Peebles suggested that we could use IAR, but that the policy needs clarification as to "what to do for partially disambiguated pages", and jc37 stated "this is a subjective determination that needs to be decided on a case-by-case basis". If anything, there is a clear consensus from the last discussion to remove PDAB. Again, a new and clearer discussion is needed before we make so sweeping a policy change, one for which proper notice is given to affected projects. bd2412 T 16:24, 24 September 2013 (UTC)

←

Where should the WP:PDAB shortcut be retargeted now that the section no longer exists in the guideline? Should it point to the archived discussion? --SoledadKabocha (talk) 23:19, 24 September 2013 (UTC)

I have retargeted it to the section on partial title matches for now, since I think that is a reasonable target for a link titled "PDAB". I am not philosophically opposed to having a policy on partial disambiguation, but we should hammer out something that does not immediately draw howls of protest and get overriden in move discussions. There is a fairly obvious difference between truly ambiguous song title like Mother, where no one song can claim to be primary, and no one song should have the title, and Thriller, for which one song is overwhelmingly famous, but for which another song merely exists. bd2412 T 23:57, 24 September 2013 (UTC)

User:BD2412, sorry, but please see the last lengthy discussion, and the last half dozen times Kiss (band) was cited as justification for WP:PRIMARYBAND, WP:PRIMARYSONG and WP:PRIMARYALBUM. It's an outlier, an exception, an extreme case: 키스 Kisu were active for 18 months in South Korea, the article doesn't even have a single English source. In cases like this we expect editors to use their brains. The cases we are discussing are not mega-notable English topics vs obscure non-English topics but for every single band, album, and song picking the [Nirvana (TOP band)] vs [Nirvana (English band)], [Thriller (TOP album)] vs 4 other albums, [Run for cover (TOP song)] vs 6 other songs.. and so on. This is already disruptive and potentially massively disruptive. What if all the other WikiProjects were to start doing what Tbhotch and Wikiredactor want and we have (TOP river) (TOP cricketer) (TOP film)? ...unlikely I know since Billboard doesn't have a hit parade of rivers or cricketers.

What is the point of all the previous discussion if it's going to be ignored simply because it was dispersed. In ictu oculi (talk) 01:12, 25 September 2013 (UTC)

The Kiss (band) question is the most important question, because the policy as written does not allow for it. We should not implement policies without exceptions, but with the idea that people will ignore them where they don't make sense. We should instead implement policies that have those exceptions built in, because if Kiss (band) is allowed despite the policy, then why wouldn't Nirvana (band) be allowed despite the policy? bd2412 T 01:22, 25 September 2013 (UTC)

Just my two cents: both Kiss (band) and Nirvana (band) can very well be allowed per WP:IAR (since the benefits are self-evident). Doing so, however, does not mean that a "primary band" absolutely must be established for all eponymous bands or that bureaucratic nonsense of the likes of WP:PRIMARYBAND, WP:PRIMARYSONG, WP:PRIMARYALBUM, and WP:PRIMARYCRICKETPLAYERFROMANONCOMMONWEALTHCOUNTRY should be developed and enforced; User:In ictu oculi's position has my full support in that.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 25, 2013; 01:45 (UTC)

(edit conflict) I gree with BD2412. Kiss (band) is an important example. It's no outlier. It's typical of exactly what PDAB is about. When one use of a common word is very well known, then it's almost certainly the primary topic within its topic area. --B2C 01:48, 25 September 2013 (UTC)

@bd2412 - it is an outlier, it outlies so far out that it's lying out in non-English sources hangul websites. Nirvana isn't an outlier because both Nirvana bands exist in English sources and the US Nirvana band had to pay the UK Nirvana band's lawyers for use of the name. That is different from Kiss vs 키스 Kisu a Korean girl band with no English sources. If an argument is made let it be from Nirvana, comparing Kiss with 키스 Kisu isn't the argument the editors pushing for "Primary band" are making, and if Kiss is a concern a simple footnote in the guideline not to count non-Latin alphabet bands with 1 single can cover it. In ictu oculi (talk) 01:22, 27 September 2013 (UTC)

Relevant to the ongoing saga of partial disambiguation, there is a discussion at Talk:Wrecking Ball (Miley Cyrus song)#Requested move which may be of interest. older ≠ wiser 16:36, 29 September 2013 (UTC)

TWODABS needs an example

• Someone familiar (musician), American jazz singer
• Someone familiar (footballer), English footballer

This would help Users read and understand what TWODABS is saying. In ictu oculi (talk) 00:53, 23 September 2013 (UTC)

Maybe. But then I think that the advice against dab pages in this case is too strong. Why should we not want dab pages with only two entries? How many 2 entry dab pages will never be expanded? And what are we hurting? If we change that section, maybe we can clarify this comment, roughly, if the hatnote would extend well over one line on a standard page. What is a standard page and what is a line? Do we have a standard line? Do I test this full screen on my 27 inch display or my 10 inch display? Vegaswikian (talk) 06:23, 24 September 2013 (UTC)

One of the guiding principles behind titles and dabbing is getting readers to the articles they seek in the fewest number of clicks. In the two dab situation, if you go with a dab page 100% of those searching with the base name will need 2 clicks to get to the article they seek. Without a dab page, as long as the more likely target is at the base name, more than 50% will get to the article they seek in 1 click, and the remainder will get to their article in 2 clicks. Hence the preference to avoid a dab page when there are only two uses.

Personally, I disagree with the primary topic requirement. If there is no primary topic, and even if we put the wrong topic at the base name, it's still no worse than the dab page situation (2 clicks).

Now, if a third use arises, that can change everything, but we should wait until we actually have a third use on WP to address that in each case. --B2C 00:43, 25 September 2013 (UTC)

I agree with that entirely. If one use is likely to get 51% of the traffic, then we basically move the navigational function otherwise served by the disambiguation page to the top of the article that gets 51% of the traffic. In a sense, the "disambiguation page" is still there, it's just handily condensed into a hatnote. bd2412 T 01:04, 25 September 2013 (UTC)

So let's change TWODABS to say that! --B2C 01:42, 25 September 2013 (UTC)

I think you need more than just the two of you agreeing on this. I think the importance of the number of clicks argument is vastly overrated. IMO, it's vastly less confusing to have a disambiguation page than flipping a coin. For one, mistaken links to the wrong page are vastly easier to identify. older ≠ wiser 02:04, 25 September 2013 (UTC)

That is a legitimate concern, but it is as much of a concern for Mouse and Apfel, and even George Washington, which have many possible uses but (to varying degrees) clear primary topics. bd2412 T 02:19, 25 September 2013 (UTC)

Yes, that is a problem with primary topics (and is one reason I prefer having a relatively high bar for a primary topic). But I don't agree it is as much of a concern. When there is a relatively clear primary topic, most links will be for the primary topic, and even for those that are not, the element of surprise should be minimal. When it is a toss of the coin, a significant number will arrive at the wrong page and scratch their heads. Some might even be ticked off that "their" topic has been dissed. older ≠ wiser 02:27, 25 September 2013 (UTC)

For a coin flip, I don't imagine there's much more head-scratching at landing on the "wrong" page than at landing on a disambiguation page. If the reader is looking for Joe Blogs (janitor) and ends up on Joe Blogs (plumber) with a hatnote saying they might be looking for the janitor, it is no less wrong for that reader than looking for Joe Blogs (janitor) and ending up at Joe Blogs (disambiguation), with pretty much the same note saying you may mean (among other things) the janitor. The disambiguation page still tells the reader their guy is not the number one topic. However, it is telling everyone that, including the people for whom their guy really is the number one topic. bd2412 T 02:32, 25 September 2013 (UTC)

Except that many readers will see one marginal topic being primary as an insult to the other marginal topic (despite repeated remonstrations by various disambiguation regulars that this isn't actually what being primary topic signifies). While some might still be miffed that "their" topic is not primary, at least a disambiguation page indicates neutral treatment. older ≠ wiser 03:05, 25 September 2013 (UTC)

Do you believe capitulating to misunderstanding is good policy? That's what you seem to be advocating here. Or do you believe that being at a base name signifies greater importance over other uses? --B2C 18:15, 25 September 2013 (UTC)

I think that in cases where there is no basis for determining a primary topic, we should not pretend that there is one by flipping a coin. older ≠ wiser 18:48, 25 September 2013 (UTC)

@Bkonrad, what about the Shropshire cricket fan who is incensed that some American judge is given precedence over John Roberts (Shropshire cricketer)? What about the Bob's Burgers fan who thinks that John Roberts (actor) should be primary? Granted, a WP:TWODABS situation is a much smaller universe of choices, and usually a much lower threshold of notability, but we can't let the devotion of fans of person "A" overrule evidence that between the two possibilities, person "B" is more likely to be the subject readers are searching for. bd2412 T 18:45, 25 September 2013 (UTC)

There is an clear and obvious case for a siting Chief Justice of the United States to be the primary topic, so I'm not sure what your examples are supposed to illustrate. Some editors are upset that Americans is about residents of the United States. Nothing that we can do about that since discussion has repeatedly established that there is a primary topic for the term. TWODABS only applies in cases where there is not a primary topic. older ≠ wiser 18:53, 25 September 2013 (UTC)

Actually, in re-reading WP:TWODABS, I find the last paragraph to be almost unintelligible.

If a disambiguation page does not appear to be needed because there are only two articles with the same title (one of them a primary topic), but there could reasonably be other topics ambiguous with the title on Wikipedia now or in the future, the {{about}} hatnote should be used to link to the disambiguation page. At the same time, the {{Only-two-dabs}} template should be added to the top of the disambiguation page, which will inform users that the page has only two ambiguous terms, and may be deleted if, after a period of time to allow readers and editors the opportunity to expand the disambiguation page, additional disambiguating terms are not found. The {{Only-two-dabs}} template will also list the article in Category:Disambiguation pages containing one non-primary topic, allowing other editors to locate these pages and help in expanding them.

So, to paraphrase, if a disambiguation page is not needed ... the {{about}} hatnote should be used to link to the disambiguation page? Huh? older ≠ wiser 19:06, 25 September 2013 (UTC)

Do we really need {{Only-two-dabs}}? I'd vote to remove it given how little usage of it we have and the fact that it probably does not serve a useful purpose. I guess using {{about}} to list the dab page when it has only 2 entries could be confusing. But the flip side is that once you set up about it works no matter how many entries there are on the dab page. Also I could argue that for many of these two dab cases, we have not really determined if there is really and truly is a primary topic, in which case the dab page should be at the main namespace. Vegaswikian (talk) 19:29, 25 September 2013 (UTC)

{{Only-two-dabs}} was a compromise arrived at after a long discussion; on one side, you had those arguing if there are only two topics, you MUST use hatnotes; on the other side were those arguing that Sexual harassment should not have a hatnote pointing to Sexual Harassment (The Office). {{Only-two-dabs}} was a solution that lets us create a DAB page in these (rare) cases, which will hopefully attract additional/useful DABs allowing the DAB page to be kept; if none are found, then you go back to hatnotes. If the explanation in the policy is confusing I'd be happy to help redraft it.--Obi-Wan Kenobi (talk) 20:00, 25 September 2013 (UTC)

OK, so I think I confused myself and perhaps others in misremembering or conflating previous discussions about TWODABS. It appears that in fact it only applies when there IS a primary topic and only one other non-primary topic. I think I was thrown off by the mention of 51% -- which is at best an extremely minimalist criteria for a primary topic when there are only two topics. If there is not a clear primary topic, I think a disambiguation page is the best option. Of course, editors are completely free to be BOLD and make a claim that one topic is primary, but if challenged, I think only 51% more traffic without any other supporting evidence would be viewed as rather flimsy. older ≠ wiser 19:18, 25 September 2013 (UTC)

I've already said too much, but just to clarify -- much of what I wrote seems more relevant for determining whether there is a primary topic rather than what to do if there is a primary topic AND only one other topic. For the latter, I agree that in most cases a disambiguation page is unnecessary and a simple hatnote to the other topic is all that is needed. There are some cases where it might be of some benefit to a disambiguation page, such as cognates, homophones, or to facilitate searches for partial title matches. older ≠ wiser 19:25, 25 September 2013 (UTC)

Bkonrad, you say: "in cases where there is no basis for determining a primary topic, we should not pretend that there is one by flipping a coin". May I reword that without using the term "primary topic"?

In cases where we cannot determine which of two topics is most likely to be sought by readers searching with the name the two topics have in common, we should not place either at the base name.

This is your position, yes? If so, why? Why should we not place either topic at the base name in such a situation? Why not allow roughly half to land on their desired topic without having to go through a dab page, while still leaving the other rough half one more click away from their destination, just as they would all be with a dab page at the base name? --B2C 19:28, 25 September 2013 (UTC)

Yes, that is. I think the reason why is self-explanatory. If there is no primary topic, then why pretend that there is one? It simply makes a mess of things. Page traffic for the pseudo-primary topic becomes meaningless as there is no way to tell how many were actually intending that page or just landed there by mistake. This can make subsequent page move more complicated as well. older ≠ wiser 19:39, 25 September 2013 (UTC)

What if it is not "a coin flip"? What if, based on Google hits and page views while the pages are both at disambiguated titles and other measures like that, we can discern a clear 55%-45% split in favor of one topic? Or a 60%-40% split? Or a 67%-33% split? At what point does it make more sense to have a hatnote rather than a separate page? Surely we can not require the same dominance for a two-way split as we would for a ten-way split? bd2412 T 19:52, 25 September 2013 (UTC)

If it is not a coin flip, then it seems like there is a primary topic. Like I said, editors are free to be BOLD and assert a primary topic, but if challenged they need to back up such claims. And if consensus determines there is not a primary topic we shouldn't let the non-primary remain simply because it happened to be the first topic or because some editor long ago thought there was a primary topic. older ≠ wiser 20:25, 25 September 2013 (UTC)

Bkonrad, I know this might sound pedantic, but can you explain your reasoning without using the term, "primary topic"? So instead of saying, "if there is no primary topic, why pretend there is one?", you would ask: "if there is no topic that is most likely to be sought, why pretend there is one?" We can answer that: Because then seekers of that topic get to their desired article right away, without any cost to others.

I don't buy the page traffic complication argument. That assumes significant numbers of people seeking topic B give up when they reach the article for topic A, rather than click on the hatlink to the article they are seeking. --B2C 19:53, 25 September 2013 (UTC)

Yes indeed, traffic statistics are not very reliable even in the best of circumstances. When a non-primary topic occupies an undisambiguated name, their value approaches nil. As for picking a primary topic by chance, I'm sorry but IMO it is simply nonsensical to do so. We are an encyclopedia and we should not be choosing article titles based on chance. And I don't agree that there is no cost to having a non-primary topic at the undisambiguated title. older ≠ wiser 20:25, 25 September 2013 (UTC)

"One of the guiding principles behind titles and dabbing is getting readers to the articles they seek in the fewest number of clicks"

It is foolish for this to be a "guiding principle". It amounts to adjusting titles to optimise search engine success rates on a simple numbers basis at the expense of logical structure. It sound like people don't know that search engine algorithms are clever and continually learning. The search engines (google/yahoo/Wikipedia) learns what people are searching for, and if out of (1)Joe Blogs (janitor), (2) Joe Blogs (plumber), & (3) Joe Blogs (disambiguation), a large majority want the plumber, then a dumb "Joe Blogs" search will take them to the plumber's pages.

In an encyclopedia, people expect a logical structure. Finding Joe Blogs (plumber) at Joe Blogs implies that there is only one on-esoteric subject titleable as Joe Blogs. If that is true, than that's fine, but if there are many reasons people might search "Joe Blogs", and if these people aren't familiar with the plumber, then the non-use of a disambiguation is a confusing disservice.--SmokeyJoe (talk) 22:38, 25 September 2013 (UTC)

Suppose that rather than saying "the fewest number of clicks" it said "with the shortest search process"? bd2412 T 22:53, 25 September 2013 (UTC)

"Shorten the search processes" is a far superior guiding principle than "reduce the number of on-Wikipedia clicks".

Less ambiguous titles mean that search engines work better, and that when a search engine returns a list of page titles, more descriptive titles greatly facilitate a shorter search process. --SmokeyJoe (talk) 23:57, 25 September 2013 (UTC)

Or take the other approach, and say that the guiding principal is to not send a significant number of readers to the wrong page. The number of clicks does not matter if too many readers are getting to the wrong article. Vegaswikian (talk) 23:44, 25 September 2013 (UTC)

The dab page is always the wrong article. --B2C 23:57, 25 September 2013 (UTC)

No, it isn't. The DAB pages contains a very nice summary. I often find DAB pages useful to the point that I don't need the articles. --SmokeyJoe (talk) 00:02, 26 September 2013 (UTC)

Then what does it matter if that "very nice summary" is tucked into the hatnote of the more likely search target? bd2412 T 00:16, 26 September 2013 (UTC)

There are many reasons why a DAB page is better than showing a wrong article when there's not a primary topic, but two are directly relevant to this discussion:

• DAB pages have shorter load and render times than whole articles. This is invaluable for people accessing from older computers with limited resources, from mobile, or from countries with bad internet connections.
• DAB pages are optimized for, well, disambiguation. It's much easier to locate the article you're looking from from a short, well-classified navigation list that compares all existing topics that use the name than from a cramped hatnote that is displayed as secondary content in an article. Diego (talk) 09:28, 26 September 2013 (UTC)

We are, however, talking about WP:TWODABS situations here. In my experience, where two topics share the same name, if both are obscure (even if one is clearly better known), both articles will be short and have fast load times. If one is obscure and the other is the clear primary topic, the primary topic will tend to be much longer, and to have a longer load time (and to be the page that the reader was searching for). Also, since the hatnote is the first thing the reader will see on the page, if there is only one other topic, it will be easy to find, as it will be the only thing there. The proposition that it's easier to locate what you are looking for from a dab page is false if the article you are looking for is, in fact, the primary topic. To give a very clear example, there are two articles on men named Barack Obama; one is the President of the United States; the other is his father, Barack Obama, Sr. Under your formula, Barack Obama would be a disambiguation page, so that editors looking for the Sr. Barack Obama could find that article after a shorter page load time. bd2412 T 13:48, 26 September 2013 (UTC)

"if the article you are looking for is, in fact, the primary topic" - Nobody is suggesting to put a DAB at the base name when there's a primary topic. TWODABS is about having DAB pages in addition to the hatnote in the primary topic article (instead of replacing the hatnote). "since the hatnote is the first thing the reader will see on the page" - That's a very strong assumption to make. Hatnotes are small, italicized/low contrast, separate from the main article body of text, shifted to the right, and located right under the huge title in bold. All those visual make them hard to spot; the natural starting reading point after scanning the page is the beginning of the article text, as it should be; i.e. hatnotes are made deliberately hard to read, so that they don't distract from the content. Having a DAB is all about the principle of least surprise: if there's a primary topic, the reader should be expecting it (by definition) even if it's not the one they're looking for. Though if the article in the base name is not a primary topic, readers still would have to read the first sentences to notice that they are not at the article they're not looking for, which is worse than noticing they've arrived to a navigation DAB page. Diego (talk) 14:07, 26 September 2013 (UTC)

That returns us to the question of what constitutes a primary topic. Is it the article that 51% of readers are likely to be looking for? 60%? 70%? bd2412 T 14:12, 26 September 2013 (UTC)

Page view stats are only one criteria (and the quality is far from precise enough to make decisions based on small differences). Even if you had a 60/40 split, if other criteria indicated there were no primary topic, then we would need consensus to place one ambiguous topic as primary over another. older ≠ wiser 15:01, 26 September 2013 (UTC)

Smokey, titles are irrelevant to searches done via search engines. We could move an article to a random string and Google will find it. If you Google for "Samuel Clemens" the first search result is Mark Twain (as it should be). In terms of choosing titles to improve user search experience, the only searches that matter are those done via the WP Find/Go box. While modern browsers display the early search results, this is not always the case, especially if using a mobile device or any browser without java script. That's the user we need to have in mind when we are trying to help get readers to the articles they seek in the fewest number of clicks. And that has nothing to do with "adjusting titles to optimise search engine success rates". --B2C 23:57, 25 September 2013 (UTC)

"titles are irrelevant to searches done" What? --SmokeyJoe (talk) 00:02, 26 September 2013 (UTC)

If you search via a search engine like Google, the search engine will find the article without regard to what its title is. Therefore, it is pointless to consider search engines when deciding titles. Titles are irrelevant to searches done via search engines. In fact, dab pages are largely irrelevant to such searches. For example, if you Google for "Washington", the search results will show all relevant WP articles, and you can click on which you want. You'd never get to the dab page (which explains why dab page view counts are so law, even when they are at a base name).

However, when searching via the WP Find/GO feature, especially without autosearch enabled for some reason, then titles are critical to helping users find the articles they seek. Those are the users we want to help find the articles they seek in the fewest number of clicks. This is absolutely critical to understanding and appreciating how we choose titles, dab pages and redirects on WP. --B2C 00:36, 26 September 2013 (UTC)

B2C, I actually agree with you that it is pointless to consider external search engines when deciding titles, as whatever process they use to produce results are essentially a black box and we are not in the search engine optimization business. However, the internal search is good reason why it is a bad practice for an non-primary topic to sit on an undisambiguated title. older ≠ wiser 01:33, 26 September 2013 (UTC)

To be clear, search engine results can be useful for determining relative likelihood of anyone searching for a given topic with a certain term. It's just that a person using a search engine is not the theoretical user whom we consider when deciding how to title, link, dab, etc.

As to whether it's a "bad" practice for a non-primary topic to sit on an undisambiguated title, let's remember that it's common practice for WP articles to be at ambiguous titles. That is, anyone using internal search to find a non-primary topic, and searches with that topic's ambiguous name, is going to end up at the wrong article. So when this happens, we don't consider it "bad". At worst, it's non-optimal for those users, but for everyone involved, it's the most optimal configuration. We consider it better than sending everyone to the dab page at the ambiguous base name, because if we have the primary topic article at the base name, at least most users will get to their article with one click, while others will be two clicks from the dab page, and three clicks from their desired article. We consider that acceptable, not bad. But a TWODAB situation is different, because there the worst case is everyone being two clicks from their desired article (one click to dab page; second click to their article). So it's already better than the acceptable situation for the non-primary topics where an ambiguous title has more than two uses and a primary topic. What we're talking about is whether this already better-than-acceptable situation can be even better. What if we cause everyone to go to one of the two articles? At worst, that's identical to landing on the wrong article when searching for a non-primary topic. In fact, it's better, because the person landing on the wrong article is one click (on a hatnote link) from being at their desired article; no worse than they would be if they landed on a dab page. How is that "bad"? --B2C 22:32, 26 September 2013 (UTC)

anyone using internal search to find a non-primary topic, and searches with that topic's ambiguous name, is going to end up at the wrong article. Well, not necessarily. Unless they have the drop-down listing disabled for some reason, they will see a listing of articles with the title. Now if there is an obvious primary topic, most people will realize this and won't expect some other article at the base name. The parenthetical disambiguating terms will help in selecting the desired article. Also if I understand correctly, these articles are displayed in a ranked order by popularity, not by simple alpha sorting. In most cases, this means the disambiguation page appears lower in the listings, even when it is at the base name. Counting clicks, is in my opinion a completely misleading criteria for titling articles in an encyclopedia. I remain convinced that if there is no basis for choosing a primary topic we are doing readers a disservice by pretending that there is one. older ≠ wiser 00:03, 27 September 2013 (UTC)

We can ignore the user who sees the drop down too, especially in the context of TWODABS. Either way he'll see two choices, at least one of which is or is not clearly the one being sought.

I just don't see how it's any more of a disservice to take someone to an article with a hat note link to their desired article than it is to take them to a dab page with a link to their desired article. Fundamentally, why is the latter preferred at all, much less enough to warrant taking everyone to the dab page? --B2C 00:43, 27 September 2013 (UTC)

Because by choosing one as the primary topic we are in effect violating neutrality. older ≠ wiser 02:02, 27 September 2013 (UTC)

Well here's a real example stumbled on by chance two days ago:

• Donald Pippin (Broadway director)
• Donald Pippin (opera director)

There's no way one of these can claim to be primary, primary depends on whether you like Broadway or opera, and I thought (and judging by edits so did other editors) that they were the same person. Why shouldn't this, or something like this, be in the guideline to show that we do use twodabs. In ictu oculi (talk) 01:09, 27 September 2013 (UTC)

• I agree that this is a perfect example of an ambiguous name that can not have a primary topic, and should be a two-link disambiguation page. bd2412 T 01:24, 27 September 2013 (UTC)
  • This was a really good example, until an third entry was found. But it could still illustrate why Donald Pippin (director) is not a good disambiguating term. older ≠ wiser 02:02, 27 September 2013 (UTC)
    • We should have a category for base pagename disambiguation pages containing only two links, to indicate which pages this structure is appropriately applied. bd2412 T 14:34, 27 September 2013 (UTC)

Aha! John Quested.

John Quested may refer to:

• John Quested (producer) (born 1935), film producer and owner and chairman of Goldcrest Films
• John Quested (aviator) (1893–1948), English World War I flying ace

Both are moderately substantial articles on people of limited notability confined to a single field. There is no likely primary topic between them. bd2412 T 15:43, 27 September 2013 (UTC)

Good coin toss example. So, currently, anyone searching for "John Quested" will not be taken to the article they seek, but instead to the TWODAB dab page, from where they have to select which one they want, and click on it, before they get to the article they seek. If we move either article to John Quested, then at least the users seeking that John Quested will get there immediately, while the others will still be one hatnote link click away from their article. Seems like an improvement, regardless of which one is moved to the base name, over the current situation.

Bkonrad, I don't understand how putting either at the base name is a violation of neutrality, since putting an article at a base name is not a statement conveying relative importance or anything else we need to be neutral about. Besides, we can do the "coin toss" alphabetically (aviator comes before producer), which of course would be totally neutral. --B2C 21:01, 27 September 2013 (UTC)

What if they were changed to "filmmaker" and "pilot"? I actually think that this is an example of TWODABS situation where the base pagename should indeed be the disambiguation page. However, I think that these are relatively rare. bd2412 T 21:12, 27 September 2013 (UTC)

Well, going alphabetical is just one suggestion. Chronological is another. Anyway, putting aside how we decide which one is at the base name, isn't it better to have one at the base name, rather than none? --B2C 21:51, 27 September 2013 (UTC)

(after ec) Tentatively agree with BD2412 and disagree with B2C. Yes, you can claim that putting an article at a base name is not a statement conveying relative importance or anything else we need to be neutral about but that is not how many people see it. I think it is inappropriate to simply pick one as the primary topic with no basis whatsoever. Persons looking for this name with the internal search will see the entries for the disambiguated names listed before the disambiguation page. Most external searches will display a snippet of the article to provide context as to what the article contains. I see no reason not to leave the disambiguation page where it is. It may be that better disambiguating terms could be chosen, but that is a different matter. older ≠ wiser 22:00, 27 September 2013 (UTC)

We agree some people (not sure about "many") see "putting an article at a base name as a statement conveying relative importance". What I'm not clear about is whether you, Bkonrad, are one of those people. Assuming you're not, you're essentially advocating capitulation to a view based on a misunderstanding. Isn't capitulating to, and choosing titles based on, that misunderstanding perpetuate that misunderstanding? --B2C 22:08, 27 September 2013 (UTC)

I think placing one article at a name without disambiguation and another at a title with a parenthetical term very strongly suggests there is some relative difference in importance between them. older ≠ wiser 22:23, 27 September 2013 (UTC)

Nevertheless, I would also say that if there is even a discernible 55/45 split favoring the importance of one subject over another, that subject should be the primary topic of the title. bd2412 T 23:07, 27 September 2013 (UTC)

Wait, what? Your precision of measurement is only as precise as the markings on your stick. In this case,, we don't have such precision available to us, by any means. I think you can only determine a "primary" topic when one is overwhelmingly more likely to be #1. Eking it out shouldn't count.--Obi-Wan Kenobi (talk) 23:13, 27 September 2013 (UTC)

Which is kind of why I way that page hit/counts should be used to show why something should not be the primary topic rather then the other way around. Using raw numbers to show that they are is simply full of problems. Vegaswikian (talk) 23:24, 27 September 2013 (UTC)

(edit conflict)Bkonrad, oh! No wonder you hold the view that you do. Okay, let me ask you this...

Since you think placing one article at a name without disambiguation and another at a title with a parenthetical term very strongly suggests there is some relative difference in importance between them...

• ...do you think we very strongly suggest that Welland, Ontario is more important than Harley, Ontario?
• That Nicholas Campbell (Canadian actor) is more important than Douglas Campbell (actor), also a Canadian?
• That Happy Sad is a more important album by Tim Buckley than is Lorca (album)?
• That Lorca, Murcia is more important than Cieza, Murcia?
• That Limerick is more important than Cork (city)?
• That New Providence High School is more important than Union High School (New Jersey), both in Union County, New Jersey?
• That the rural locality of Tatarstan named Koshlauch is more important than these other Tatarstan rural localities: Alat, Russia, Shali, Republic of Tatarstan, Gruzinsky (settlement), and Murali, Arsky District, Republic of Tatarstan?
• That the Saône-et-Loire commune of Saint-Edmond is more important than the Saône-et-Loire commune of Sainte-Foy, Saône-et-Loire?
• That the youth magazine Putsj is more important than the youth magazine Bravo (magazine)?
• That the Ebro river of Spain is more important than other rivers of Spain that start with E, like Ega (river) and Esla (river)?
• That the Britz locality of Berlin is more important than Spandau (locality) and Mitte (locality)?

Because if you do really think that placing one article at a name without disambiguation and another at a title with a parenthetical term very strongly suggests there is some relative difference in importance between them, these examples are just the tip of the ice berg of countless instances in our titling that, according to your belief, suggest "some relative difference in importance between [two topics]" when, actually, there is none. How shall we address this? --B2C 23:14, 27 September 2013 (UTC)

Not at all. You're comparing apples to oranges. If you need the obvious to be stated, then to be precise, I'll amend my comment that placing one article at a name without disambiguation and another with the same base name at a title with a parenthetical term very strongly suggests there is some relative difference in importance between them. older ≠ wiser 23:44, 27 September 2013 (UTC)

I thought you might limit your general statement to this special case. But that begs the question: why does this strong suggestion about importance apply only to topics that share the same base name?

I mean, if the lack of parenthetic disambiguation in John Quested would strongly suggest that that John Quested is more important than John Quested (aviator), why doesn't the lack of parenthetic disambigution in Nicholas Campbell strongly suggest that that Canadian actor with surname Campbell is more important than Douglas Campbell (actor)? Why doesn't the lack of parenthetic disambiguation in Britz strongly suggest that that Berlin locality is more important than Spandau (locality)? --B2C 00:13, 28 September 2013 (UTC)

Well it seems pretty self-evident, IMO. The lack of parenthetical disambguation Nicholas Campbell has no relevance whatsoever in comparison to a person with a different name, such as Douglas Campbell (actor). That Douglas Campbell (actor) has a parenthetical disambiguation only implies that there is at least one other person with that name of equal or greater relative importance. The same applies for localities and most any other case. older ≠ wiser 00:24, 28 September 2013 (UTC)

Let me give you a concrete example. How would you resolve Talk:Jerome Frank (lawyer)‎#Requested move? bd2412 T 00:48, 28 September 2013 (UTC)

(after ec) That is a pretty good case for consideration. I had been watching the discussion but hadn't !voted yet. From what I've seen so far, the balance is tipped in favor of the lawyer being the primary topic, but it is not a slam-dunk. That is a case where personally I would be fine with the either the lawyer or the disambiguation page being at the base name. While the lawyer is marginally more notable than the psychiatrist, it is not so overwhelming that the lawyer can be presumed to be THE Jerome Frank. Leaving the "(lawyer)" appended to the name would help searchers to distinguish between the two. older ≠ wiser 01:50, 28 September 2013 (UTC)

This is what I don't understand about your reasoning. The argument - Leaving the "(lawyer)" appended to the name would help searchers to distinguish between the two - (i.e., making the topic more recognizable from the title alone) - applies just the same whether the lawyer is actually the primary topic or not. In fact, it applies, and is relevant, to titles of the vast majority of our articles - articles about topics that are not widely known - whether the titles are ambiguous or not. Why do you feel it's somehow relevant and important in the TWODABS case, but not, apparently, in other cases? --B2C 04:21, 28 September 2013 (UTC)

When there is an actual primary topic, we can expect that most readers are not unduly surprised to see that particular topic is located at the base name. Same applies in distinguishing between a primary and non-primary topic. Where there are ONLY non-primary topics, the parentheticals help readers to distinguish the topics. older ≠ wiser 11:27, 28 September 2013 (UTC)

When there is an actual primary topic, we can expect that most readers are not unduly surprised to see that particular topic is located at the base name. That is true in cases where the primary topic is broadly known. But there are countless relatively obscure "primary topics" like Lorca, a Spanish town with fewer than 100,000 people. So people searching for Lorca (album) will not be "unduly surprised" to find the town at Lorca (the album is named after Federico Garcia Lorca, not the town), but they will be "unduly surprised" to find one of the two Jerome Frank's at Jerome Frank? I don't get it. --B2C 23:51, 29 September 2013 (UTC)

There are undoubtedly many such marginal primary topics. Whether they are truly a primary topic or simply an example of WP:OTHERSTUFFEXISTS might need some community discussion to determine. That there may be some inappropriately placed marginal primary topics does not mean we shouldn't care in other marginal cases if a non-primary topic is posing as a primary topic. Lorca is a moderately sized municipality with a long history in a western European country. Hard to say whether readers would be unduly surprised that it is the primary topic rather than the poet. That would be a matter for discussion. How commonly is the poet referred to as simply "Lorca"? How commonly does the municipality turn up in a variety of reliable sources? older ≠ wiser 00:55, 30 September 2013 (UTC)

The point is that it doesn't matter much, in terms of the problem caused by undue surprise, in any of the countless arguably marginal cases like Lorca, whether the topic at the base name is bonafide primary or not, and it matters even less, a negligible amount, in TWODABS cases, because the one and only other topic is as easily accessed via the hatnote link as it would be if the dab page was at the base name. --B2C 17:04, 30 September 2013 (UTC)

Well, no. If Lorca is not the primary topic, it should be moved. Similarly, if there is no basis for placing one article of two as primary, we should not. older ≠ wiser 17:25, 30 September 2013 (UTC)

But in a marginal case (like perhaps Lorca is), it may or may not be the primary topic, by definition. We simply don't know whether people are more likely to search for that topic than the others. If it is more likely (to a sufficient degree), or if it meets the historical significance criteria, then it should be at the base name. So, in such a case, on a scale of 1 to 10, where 1 is "totally inconsequential" and 10 is BLP-violation critical, how much does it matter if we get it wrong? In particular, in a marginal case, where does it fall on the "matter scale" if it's not the primary topic but is at the base name anyway? Whatever it is, I suggest it's pretty far down the scale, much closer to 1 than to 10, and, more importantly, in a TWODABS case, how much "getting it wrong" (non primary topic is at base name) matters is even lower on that scale... pretty much a 1. --B2C 18:00, 30 September 2013 (UTC)

Once again, it is a matter of principle. Where there is no basis for choosing a primary topic, we should not pretend that there is one and present a bias to readers. older ≠ wiser 18:53, 30 September 2013 (UTC)

This is where my concern comes in. I agree that in a TWODABS situation "where there is no basis for choosing a primary topic", we should not assert one. However, I would say that where there is just about any for choosing one, then we should choose that one. I would contend, for example, that a 57/43 split in average page views or Google hits favoring one over the other is sufficient. bd2412 T 19:23, 30 September 2013 (UTC)

I agree, although personally I would opt for a higher threshold. But I have no real problem if an editor is bold and adjusts pages on such a basis. However, if that arrangement is questioned, there may be factors to consider other than simply traffic stats and consensus may determine there is no primary topic. older ≠ wiser

(edit conflict)BD2412, a 57/43 split arguably does not support a claim of primary topic per our primary topic criterion. However, I agree that that does not mean the dab page should be at the base name. Why? Because it's helpful if one of the articles is at the base name, and it doesn't matter if people land on a dab page or another article when either one leaves them one click from where they would like to go. Putting one of the two articles at the base name, instead of the dab page at the base name, even if neither is the primary topic, might not be win/win, but it's win/tie... a plus overall.

So, more generally, in a case of TWODABS, either article, not the dab page, should be at the base name, regardless of whether either actually meets the primary topic criteria.

A while ago we seem to achieve consensus that an article at a base name does not necessarily mean the article's topic is the "primary topic" of that base name, most notably in the "only topic" case. For example, the topic of Thomism is not the primary topic of "Thomism" even though its article is at Thomism - because that use is the only use of "Thomism". So we already have countless precedents of articles about topics that are not primary being at the base name. So let's not get hung up on that. Being at a base name title does not necessarily means that article's topic is the primary topic. It might be the only topic, or it might be the winner of a 57/43 TWODABS situation, or it might be the winner of a coin toss 50/50 TWODABS situation.

Bkonrad, saying "it's a matter of principle" is evading the question. How much, quantified on the 1-10 "matter scale" I presented above, does it matter if we present this "bias" to readers? How much does it matter? --B2C 20:01, 30 September 2013 (UTC)

First, I stand by the principle that if there is no basis for choosing a primary topic, we should not do so. You seem to feel that there is something inherently bad about having a disambiguation page at the base name where there is no primary topic. I don't agree. older ≠ wiser 20:07, 30 September 2013 (UTC)

You're still evading the question. What exactly is the principle, and how much does it matter if we break it?

I don't feel "that there is something inherently bad about having a disambiguation page at the base name where there is no primary topic"! I'm saying that when there is just one other topic, it's better to have either topic at the base name, then a TWODAB dab page at the base name. It's better for two reasons:

1. Mysterious principles notwithstanding, practically speaking, it's no worse for anyone to any significant degree to have either article, rather than the dab page, at the base name. Even in a 10/90 case, those seeking the much more likely-to-be-sought article are not worse off if the unlikely-to-be-sought article is at the base name, than if the dab page is at the base name.
2. It's definitely better by a significant degree for those seeking either of the two articles to have it at the base name.

Of course, it's even better to have the one of the two that is more likely to be sought at the base name, but the law of diminishing returns applies with respect to which one of the two is at the base name as the likelihood of each being sought approaches the other. The main point is that either at the base name is better than the dab page at the base name, for the reasons just stated. --B2C 20:45, 30 September 2013 (UTC)

Well, I didn't want to be too blunt, but I didn't answer your question because I have no interest in following you down a rabbit-hole of phantasmagorical thought exercises. The principle, I've already said, is that if there is no basis for choosing a primary topic, then we should not do so. We are not in the business of page traffic optimization. To only consider the imagined inconvenience of those arriving at a disambiguation page is missing the point. older ≠ wiser 21:47, 30 September 2013 (UTC)

Well, if we're going to be blunt, what you're avoiding is not going "down a rabbit-hole of phantasmagorical thought exercises", but discovering the vacuity of your argument. You're presenting your principle as a Golden Principle which Shall Not Be Violated, and the only reasoning to support it that you've provided is a vague reference to supposedly avoiding the subjecting of readers to undue surprise. What you're unwilling to do is quantify the cost of violating "the principle" that "if there is no basis for choosing a primary topic, then we should not do so". Assigning a 1-10 value to how much it matters if we violate that principle would clarify your position - which is what you are apparently choosing to avoid. And that's understandable. Keeping a position vague makes it easier to defend. --B2C 22:53, 30 September 2013 (UTC)

Well, show that you have a consensus that agrees with you and I'll defer, but I won't apologize for refusing to dance to the tune you choose. The value is that we are not capriciously selecting article titles. If you cannot understand that, then then I don't know what else to say. older ≠ wiser 00:34, 1 October 2013 (UTC)

I get the value of pointlessly capriciously selecting article titles. But I don't agree capriciously choosing one of two articles to be at the base name in order to avoid putting a dab page at the base is pointless.

Do you get the value of landing directly on the article which you seek rather than on a dab page with a link to the article you seek? Rather than on an article with a hatnote link to the article you seek? These are the values at conflict here. Now, unless these values are quantified somehow (hence the Matter Scale and question), how can we make a reasoned decision? --B2C 00:48, 1 October 2013 (UTC)

Do you get that there is more to an encyclopedia that helping half of the persons looking for a page? Do you understand that including the disambiguating term in such marginal cases does help readers avoid landing at the dab page and get to the article they want? Overemphasis on quantification of values that have no actual numerical correlate is fallacy, much like precision bias. Reasoned decisions are made all the time without resorting to quantification of qualitative data in ways that completely miss the point. older ≠ wiser 01:17, 1 October 2013 (UTC)

You're still not answering my questions, and you're characterizing a modicum of attention to quantification of values as "overemphasis". The point is to estimate value to WP on 1-10 "Matter Scale" for the purpose of relative comparisons. It's easy to say "A is important", and "No, B is important". But until you introduce a method by which to quantify "importance" you have no reasonable way to ascertain which is more important.

As to your questions, yes, of course I get "that there is more to an encyclopedia that helping half of the persons looking for a page?". That's why I want to make sure those other things are not significantly compromised.

And, of course I "understand that including the disambiguating term in such marginal cases does help readers avoid landing at the dab page and get to the article they want?" But that's irrelevant because the disambiguated term can exist as a redirect and will thus be just as helpful as when it's part of a title. --B2C 19:41, 1 October 2013 (UTC)

Yes, Bkonrad, that distinction is indeed self-evident to us WP title wonks. But not to anyone else. To others, the distinction is artificial. Either parenthetical disambiguation diminishes importance, or it doesn't. You can't have it both ways. --B2C 00:58, 28 September 2013 (UTC)

No, I think you're making unreasonable artificial assumptions. older ≠ wiser 01:50, 28 September 2013 (UTC)

I like the usage of the term WP:COINTOSS and that would make a good shortcut to the section of WP:TWODABS dealing with 60/40 situations. A real example in a grey box John Quested would still be visually helpful. I don't see any substantive reason above for not giving an example, nor against this example. In ictu oculi (talk) 23:18, 30 September 2013 (UTC)

• Born2cycle, landing on the article I want, despite it being imprecisely titled and my search terms being imprecise, on the statistical basis that most people searching imprecise things want popular articles, is a good way to sell junk food to customers who don't know what they want, but a bad justification for stacking the popular things at the front if it confuses the existence of the full, comprehensive set of products. If a popular article is disambiguated, people wanting the popular article will still land on it. The disambiguated title will be explicitly linked. The search engines, external and internal, will continue to do a good job, statistically, guess what is wanted from imprecise search terms. This drive to shorten titles, at the expense of recognizability to the unfamiliar, and the occasional cost to unusual readers, has no supposed benefit with any reasonable justification (and that's having read your userpages). --SmokeyJoe (talk) 01:10, 1 October 2013 (UTC)
  • The reasonable justification is that there is no upper limit on making titles recognizable. Once you accept the argument that a title could be "improved" by making the title more recognizable, there is practically no end to where that argument could be applied. Just to pick a few examples from SPECIAL:RANDOM...
    • Prohner Stausee could be "improved" as Prohner Stausee (lake in the Vorpommern-Rügen district in Mecklenburg-Vorpommern, Germany)
    • Acrochaetiales could be "improved" Acrochaetiales (order of red algae)
    • Bada la Djandza could be "improved" as Bada la Djandza (village on the island of Anjouan in the Comoros).

    If you don't see how that is a reasonable justification for opposing lengthening titles that already meet our criteria requirements (including recognizability - to those familiar with the topic), I can't help you. --B2C 01:21, 1 October 2013 (UTC)

• These are facetious arguments, along the line of the straw man fallacy. --SmokeyJoe (talk) 01:37, 1 October 2013 (UTC)

Strawman? I'm not saying you're intentionally arguing for changes to titles such as those I've listed above. I'm saying such changes are a logical outcome of your argument.

You referred to improving "recognizability to the unfamiliar" as a good thing, yes? If that's not acceptance and outright advocacy for "the argument that a title could be 'improved' by making the title more recognizable" (which would lead to title changes such as those listed above), then please explain... for I misunderstood. I don't see where or how you draw the line. --B2C 01:54, 1 October 2013 (UTC)

Recognizability to the unfamiliar. I am thinking of readers familiar with a less popular topic, who are not familiar with the more popular topic, and where the two are ambiguous at their natural titles. Recognizability is good. More recognizability is not necessarily better as per your strawmen. A title that is mistakenly recognized as something else by a reasonable set of readers familiar with the something else, is bad.

You like lines. But trying to artificially invent bright lines creates trouble. I suggest one line for you: Your "a better recognizable title is better title argument leads to ridiculously long titles" argument is never a justification for shortening a reasonable title. It is like "straw men exist that way so we have to run in the other direction". I am not frequently advocating changes to lengthen titles, but you are frequently advocating changes to shorten titles. --SmokeyJoe (talk) 02:35, 1 October 2013 (UTC)

SmokeyJoe, I was going to say much the same. Glad someone else did. It looks to me like a rather improbable slippery slope. older ≠ wiser 02:09, 1 October 2013 (UTC)

"A title that is mistakenly recognized as something else by a reasonable set of readers familiar with the something else, is bad." But, then, aren't most primary topic titles, excluding the ones about widely known topics like Paris, "bad"? I mean, given that the term is ambiguous, won't it be recognized as one of the other uses by a "reasonable set of readers familiar with" one of the other uses?

Yes, I like bright lines. Bright lines are the antidote to bickering about issues that ultimately don't matter. When a bright line is possible, I prefer that to vague and ambiguous guidance, for that reason. --B2C 19:41, 1 October 2013 (UTC)

An undisambiguated title for a topic A, unrecognized by most readers, that is mistakenly recognized as topic B by a reasonable set of readers unfamiliar with topic A but familiar topic B, is bad.

In other words, a primary topic should be recognizable to most readers, or at least recognizable to most readers who recognize the similarly titleable other topics.

Paris is (or should be) recognizable to all educated people in an international context, so there is no astonishment in someone searching an international encyclopedia for "paris" to upload the article Paris.

Welland is a better example of bad. The Ontario city is obscure and unrecognized to an international audience. Many readers will be familiar with a competing Welland, yet have no knowledge or interest in the Ontario city. The people of Adelaide may very well assume that their Welland is unique, and will be astonished to upload the link en.wikipedia.org/wiki/Welland and be astonished that they have uploaded a large article on a completely unrelated place. Possibly, they wouldn't be astonished, due to familiarity with Wikipedia's North America bias, but this bias is bad.

Neither Someone familiar (musician), American jazz singer, nor Someone familiar (footballer), English footballer, are presumably broadly internationally recognized, so neither should be presumed to be primary, regardless of the ratio of their respective negligible non-local recognizabilities. --SmokeyJoe (talk) 00:28, 2 October 2013 (UTC)

First, thank you for being clear about your position. I wish Bkonrad would follow suit.

Welland is not atypical. It is but one in a myriad of such examples. Your position is that primary topics should not be at base names unless they are "recognizable to most readers". If adopted in policy, this position would require renaming most titles of primary topic articles (since most primary topics are like Welland... not "recognizable to most readers". Thankfully, there is no consensus support for your view. --B2C 01:34, 2 October 2013 (UTC)

No. Not "recognizable to most readers", but "recognizable to most readers who recognize the title". That is, it is, a topic is not a primary topic if it is prone to be misrecognized as something unrelated. This is a direct application of the Principle of least astonishment. It will have no effect on advanced search engines because they are very good and qucikly adaptable at finding maximum likelihood solutions. "recognizable to most readers" is an absurdity and it looks to be another case of you exaggerating opponent positions into strawmen.

Welland is indeed a common inocuous example, looking just at Welland, Ontario and Welland, South Australia, ignoring the other significant related and unrelated Wellands. Millions of South Australians are likely to encounter mention of Welland in reference to Welland SA, and should they have reason, if they go to Wikipedia, they'll be astonished to find themselves taken to Welland ON, a largish page, downloading for them at backwater speed. --SmokeyJoe (talk) 02:29, 2 October 2013 (UTC)

So if you like bright lines, why do you like muddying matters by creating a pretense of a primary topic where there is no basis for one? If you really prefer bright lines, it is by far simpler have a disambiguation page at the base name where there is no basis for choosing a primary topic. older ≠ wiser 20:24, 1 October 2013 (UTC)

Just because there is no basis for a primary topic does not mean there is no basis for putting either article at the base name. Here are the three bright lines for a TWODABS situation I suggest:

1. If either use is primary, that use is at the base name.
2. If neither use is primary, but one use is more likely to be sought, that use is at the base name.
3. If neither is more likely to be sought, whichever was created first is at the base name.

No mud. Very clear. --B2C 21:48, 1 October 2013 (UTC)

Seems unnecessarily complicated and open to various misinterpretations and potential for dissension. Even simpler, if there is no primary topic, a disambiguation page should be at the base name. older ≠ wiser 22:41, 1 October 2013 (UTC)

Fine. Even simpler: If there is no primary topic, whichever was created first is at the base name. --B2C 23:14, 1 October 2013 (UTC)

Or simpler and more accurate, if there is no primary topic the disambiguation page goes at the base name. Vegaswikian (talk) 23:57, 1 October 2013 (UTC)

It's not any simpler. The wording is no simpler, and it requires three entities (dab page + 2 articles) instead of just two entities. That's more complex.

How is it more accurate? Accurate with respect to what? --B2C 01:34, 2 October 2013 (UTC)

It's simpler because it minimizes contentiousness. Under your simplistic alternative, an article on some backwater hamlet would remain at the primary topic even if another article came along of equal significance. That does not make sense. It is not accurate to suggest that there is a primary topic when there is none. older ≠ wiser 11:30, 2 October 2013 (UTC)

What does "remain at the primary topic" mean? Let's not conflate a characteristic of a topic with an article place or title. Articles cannot be at topics, primary or not.

Contentiousness about what, exactly? The original use would remain at the base name unless and until another use was the primary topic, or there was a total of three or more uses. Let's take a closer look at #A hypothetical TWODABS example. --B2C 15:50, 2 October 2013 (UTC)

I think you know perfectly well what I meant by "remain at the primary topic". Perhaps it would be more precise to say "remain at the base title", but your objection is obtuse pedantry. Contentiousness is a very possible outcome if the two topics are of comparatively equal status. Why should one be allowed to hold the undisambiguated title when there is no basis for choosing between the two? older ≠ wiser 16:26, 2 October 2013 (UTC)

Indeed, I knew what you meant, and I knew you knew I knew what you meant. But I still had a valid point: we should not conflate the terms, for that conflates the concepts, concepts for which the distinction is quite relevant to this discussion. Yes, "remain at the base title" is more accurate. Or, perhaps, "remain at the undisambiguated title".

No basis for choosing between the two? Excuse me? "Whichever was created first" is the basis. You may not like it, but it is a valid basis for choosing which one of two articles should be located at the undisambiguated title. Not only is it valid, but it is contention-proof, since only one of the two can possibly meet the created first criterion. --B2C 20:26, 2 October 2013 (UTC)

It may be your opinion that it being first created is a valid basis. Fortunately, it appears to be solely your opinion at present. older ≠ wiser 21:07, 2 October 2013 (UTC)

Why does everything have to be WP:IDONTLIKEIT? It's simpler and more accurate since it does not force a primary topic when none exists. Bottom line is that DAB PAGES ARE GOOD. Your goal to bring every decision to a back or white choice does not work in our grey world. There are a number of two dab pages out there for technical terms that the average reader runs into. How do we know which one applies? Generally I have to tag those links for disambiguation needed because the only editors who can figure these our are individuals with a medical degree or a science degree in that area. I will oppose any efforts to force readers to the wrong page simply to satisfy a solution that is not the best for the encyclopedia or its readers. And yes, I consider forcing 49% too high of a number! Just remember that most topics are ambiguous and that is the real world we should acknowledge. Vegaswikian (talk) 17:04, 2 October 2013 (UTC)

I disagree with the proposition that "DAB PAGES ARE GOOD". They're more like an appendectomy - a good thing to have if your appendix is about to burst, but you wouldn't want to have one if it wasn't necessary. bd2412 T 17:55, 2 October 2013 (UTC)

Agreed, and I would add to that: "LANDING ON AN ARTICLE PAGE WITH A HAT LINK TO THE SOUGHT ARTICLE IS NO WORSE THAN LANDING ON A DAB PAGE WITH A LINK TO THE SOUGHT ARTICLE". --B2C 20:26, 2 October 2013 (UTC)

That I definitely don't agree with. Hatnotes have often been criticized as being too easy to miss. Further, when you have to wait for a big page to load instead of a comparatively lightweight dab page, the dab page is preferable. older ≠ wiser 21:07, 2 October 2013 (UTC)

Not all links are created equal. Articles are optimized for reading its content, not for finding the hatnote; readers will naturally start reading the content, and only after discovering that they're reading the wrong target they'll look after the link. DAB pages are optimized for navigation, so it's clear from the onset that the right action is scan and find the link. So no, they're not comparable; the hatnote is objectively worse even if it contains the same information. As people are not computers, presentation matters. Diego (talk) 22:10, 2 October 2013 (UTC)

Your arguments are all based on your opinion that we should always be able to determine a primary topic, even if we do so by rule or by assumption. That is simply not reasonable for readers since it produces too many bad links and as Diego points out and was raised in previous discussions, linking to a dab page gets more eyes on the link and it gets disambiguated to the correct link. Bad links to a primary article especially when a primary topic is hazy do not get the review needed to have them fixed to point to the correct article. So what is wrong when we have a target that tells editors that they are not linked to the correct article? Vegaswikian (talk) 23:07, 2 October 2013 (UTC)

Vegaswikian wrote: Your arguments are all based on your opinion that we should always be able to determine a primary topic,

How anyone could think that it is my opinion that "we should always be able to determine a primary topic" when I've been abundantly and repeatedly clear that I'm talking about TWODABS cases where there is no primary topic is beyond me. If that is truly your understanding of my opinion after all this, then reasonable discussion is not possible.

But perhaps you too are conflating "primary topic" with "title is base name" (why do people do that?).

My position is this: we are certainly able to determine which of two articles that share the same common name was created first. There can be no dispute about our ability to determine this. We can also agree that we can determine whether we have consensus about either of TWODABS topics being the primary topic, and, that if consensus about that is in the affirmative, then that topic should be at the base name.

The only dispute is about what to do if neither topic is primary. In that case your position is that both titles should be disambiguated and there should be a dab page. And my position is that the article that was created first should be at the base name, and the article created later should be the one and only one that is at a disambiguated title (and that there is no need for a DAB page). We also disagree on the "cost" of landing on the "wrong" article, the "cost" of landing on a dab page, the benefit of landing on the sought article instead of on a dab page, and the benefit of landing on the sought article instead of landing on the "wrong" article. At any rate, I certainly do not hold that opinion "that we should always be able to determine a primary topic". --B2C 06:24, 3 October 2013 (UTC)

• DAB pages are good. They are a navigation aid, and a very successful one. People only arrive at DAB pages for one of these reasons: (1) They want the DAB page, wheter because they are looking for something and are not exactly what what it is or some other reason; (2) They entered inadequate information as a search query; (3) It has been mislinked, and this needs fixing.

  If people would do search queries with reasonable information input, the search engines (internal or external) wouldn't send them to DAB pages. This practice, of assuming that most inadequate searchers want the most popular page serves as bad confirmation bias. Inadequate searchers should learn to search a little better. Placing the ambiguous but most popular page at the undisambiguated title defeats the search engine's cleverer algoriths because our default search box will take you to the title of exact match by default. It confuses external searchers because they look first at the list of returned page titles. If the most popular topic were at a disambiguated title, a search for that title would take the searcher there if the search algorithms and current data supported it, and if the decision is a mistake, modern search engines can detect that, and learn. --SmokeyJoe (talk) 01:19, 3 October 2013 (UTC)

The original TWODABS

Before there was a TWODABS shortcut this is what the relevant paragraph said[1]:

If there are two topics for a term but neither is considered the primary topic, then a disambiguation page may be used; an alternative is to set up a redirect from the term to one of the topics, and use disambiguation links only.

I'm not sure when that got usurped by the current much more restrictive language, by I am not aware of a single case where anyone would be better served by a dab page than by a variant on the alternative described here (the variant is: put either of the two articles at the term; a hatnote link to the other at a disambiguated title). --B2C 01:21, 1 October 2013 (UTC)

Since you bring it up, I looked a bit further into the history. It was added 09:21, 21 August 2008 by Kotniski as part of a massive rewrite attempting to harmonize WP:DAB and WP:MOSDAB. The edit summary indicated additional section on an important subject which was hidden away on MOSDAB; however, at that point in time, I could find no exact correlate on MOSDAB. The closest I could find was here, though the content is quite different. A short time after Kotniski's edits, there was a very relevant discussion of that specific text. That discussion referenced an earlier, inconclusive discussion here (which by the way has an interesting statistical analysis based on some informal usability testing). Also, precipitating the former discussion, the text was modified 14:13, 22 September 2008 by JHunterJ to require a disambiguation page if there is no primary topic. Shortly after, Kotniski reverted and flagged the section as disputed. A few days later, Kotniski revised the text to However if there are two topics for a term but neither is considered the primary topic, then a disambiguation page is used. older ≠ wiser 02:05, 1 October 2013 (UTC)

Right. And my edit simply brought it into line with the rest of the disambiguation guidelines: if there's no primary topic (for two or two hundred ambiguous topics), the dab goes at the base name; if there's a primary topic (for two or two hundred topics), the primary topic goes at the base name. Making it consistent didn't make it more restrictive; the restriction was always there in the guidelines. -- JHunterJ (talk) 19:47, 1 October 2013 (UTC)

I would dispute, however, that a topic must be well known to be primary. I would say that it merely must be significantly better known than the other, or significantly more important. That is part of the reason why I would say that if there happens to be a plant sharing a name with a little-known song, the plant should always be primary in that pair. bd2412 T 19:58, 4 October 2013 (UTC)

I kind of agree with you except for your comment about the plant and song. These cases need to be determined based on fact. So there will be cases where if we do have a primary topic it would be the song. Of course in most cases like this there will not be a topic that is significantly more important or better know to the point what we can prove it. Vegaswikian (talk) 20:05, 4 October 2013 (UTC)

DAB pages and primary topics

I essentially concur with SmokeyJoe's reasoning, and I find it a very reasonable interpretation of "highly likely to be the topic sought when a reader searches for that term": if there are different significant audiences around the world and one of them is likely to have never heard about the primary topic, it can't be a primary topic - as those readers are not looking for it with any probability. I think distinguishing between "primary topics" (those recognizable by almost anyone that knows the name) and "high volume topics" (those looked for by many people, but that a significant number of readers would recognize as a different topic). The first could safely be placed at the base name; for the latter, I think the optimum would be achieved by having disambiguation pages ordered by number of visitors - thus the high volume topic is easy to spot, but those looking for other topics still get a helpful navigation page (instead of the wrong article plus a tiny hat note). Diego (talk) 12:01, 2 October 2013 (UTC)

• Yes, audiences should not be sent to a topic when they were expecting to find a different a different topic. However, where all competing topics derive from a parent topic, such as Avatar, it is fine for the parent topic to be undisambiguated. --SmokeyJoe (talk) 01:02, 3 October 2013 (UTC)

A hypothetical TWODABS example

Building on the discussion above, let's consider the effect of the two TWODABS approaches being discussed,

• If there is no primary topic, a disambiguation page should be at the base name.
• If there is no primary topic, whichever was created first is at the base name.

on the following hypothetical but realistic situation:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation), as it's so obscure the plant remains the primary topic.
3. Some time later, as the size and notability of the corporation grows, there is no primary topic.
4. Eventually the corporation meets the primary topic criteria

Under If there is no primary topic, a disambiguation page should be at the base name, we would have:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation), as it's so obscure the plant remains the primary topic.
3. Some time later, as the size and notability of the corporation grows, there is no primary topic.
   • The plant article is moved to Climpaton (plant)
   • A new TWODAB dab page is created at Climpaton
   • The corporation article remains at Climpaton (corporation).
4. Eventually the corporation achieves primary topic status
   • The plant article remains at Climpaton (plant)
   • The TWODAB dab page is moved to Climpaton (disambiguation)
   • The corporation article is moved to Climpaton.

Under If there is no primary topic, whichever was created first is at the base name, we would have:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation), as it's so obscure the plant remains the primary topic.
3. Some time later, as the size and notability of the corporation grows, there is no primary topic.
   • As the original use the plant article remains at Climpaton.
   • The corporation article remains at Climpaton (corporation).
4. Eventually the corporation achieves primary topic status
   • The plant article is moved to Climpaton (plant)
   • The corporation article is moved to Climpaton.

Now, which scenario is simple, more stable and less contentious? Which is more complex, less stable and more likely to foster contention?

Consider that under If there is no primary topic, whichever was created first is at the base name, nothing changes between stages 2 and 3, so there is no need for the community to even evaluate when that happens. But under If there is no primary topic, a disambiguation page should be at the base name, the community is required to ascertain that point, to decide when to disambiguate the title of the plant article to make room for the dab page at the base name.

--B2C 15:57, 2 October 2013 (UTC)

You've demonstrated you're quite adept at constructing hypotheticals that favor your position. Consider the equally likely scenario in which the second Climpaton is out of the box of comparable notability to the first. Does it really make sense to leave the first created at the base name instead of a disambiguation page? Or consider the possibility that the first article might have some long-standing and stable, though minor, position in its knowledge domain. Then suppose the newly created second article has some flash-in-the-pan spike. Some editors would want to immediately move the second article to be the primary topic. Other editors would want to see if it passes WP:ONEEVENT or WP:RECENTISM. In any case, your hypotheticals depend on an eventual outcome which might elapse over a period of days, months or years. If the two topics are of comparable notability for more than a few weeks or months, I think placing the disambiguation page at the base name would be a more encyclopedic approach than inertia. older ≠ wiser

(edit conflict) I fail to see anything undesirable in the first scenario; in fact, I regard it as providing the optimum balance for each of the intermediate states of notoriety for both the plant and the corporation. (Also I don't see why "stability" should be a value to consider; this is a wiki, it's assumed that things change).

At the third step (when the size and notability of the corporation grows), you're assuming that the corporation has grown so popular that it's a contender against the long term significance previously held by the plant (otherwise, it wouldn't dethrone it as primary topic). This means that more visitors will be interested in the company than in the plant, but not enough to make its article primary. In that situation, the disambiguation page is a better compromise than sending many visitors to the plant article.

Most likely, the progression wouldn't cause any contention - for low traffic articles, both the company article and the DAB will likely be created by respective editors that wanted to learn about the corporation, and found the plant article at the base name - the first editor at step 2 placing a hat note to the new company article, and the second editor at step 3 creating a DAB when noticing that the plant shouldn't be primary. If any of those steps create any controversy at all, then the default (placing a DAB at the base name) should be triggered, as controversy means that neither topic is primary, unless a consensus can be achieved.. Diego (talk) 16:25, 2 October 2013 (UTC)

Diego, well put. The scenario you describe seems quite common to me. And, as I've said elsewhere, I think it would also be quite OK if the editor(s) creating or editing the second article were BOLD and moved the second article to be primary, if they had some reasonable basis to do so. If other editors questioned that move, then discussion might result in determining there is no primary topic and a disambiguation page is placed at the base name. older ≠ wiser 16:34, 2 October 2013 (UTC)

I think the proposal is too haphazard. Rather than going solely by what is made first, I would prefer some predetermined order of encyclopedic precedence. For example, if there were two things having the same name, but neither of them being particularly notable (and therefore having no primary topic), then biological genera would always take precedence over place names, which would always take precedence over company names. If there are two people or companies with the same name, always go with the earliest in time (first born or first founded). If there are two biological genera with the same name, always go with mammals over birds over reptiles over fish over arthropods over plants over fungi over protazoa. That said, I am fine with having a TWODABS disambiguation page unless there is some indication of a more important topic. bd2412 T 16:28, 2 October 2013 (UTC)

• I agree with this notion of order of encyclopedic precedence. It is already at play in places, it is natural, and it is expected by new readers. Examples include counting numbers a presumed to be calendar years. Natural world topics are usually taken to have encyclopedic precedence over commercial products. Parent topics (like Avatar) having precedence over conceptual spinouts feels logical. The idea fits well with the category system. --SmokeyJoe (talk) 01:02, 3 October 2013 (UTC)

Have fun trying to get agreement about such an order of precedence. older ≠ wiser 16:34, 2 October 2013 (UTC)

Yes, I know. Still, the project has ironed out disputes of that nature before, and I think the kinds of thing I specified above are generally common sense. bd2412 T 16:45, 2 October 2013 (UTC)

It's an interesting idea, and my first thought was the same: "good luck trying to get consensus". It may be easier for some categories than others, though; I can see it applied to topics from living things and natural sciences taking precedence upon company names, but not between topics in the same group, or for people from different epochs. Diego (talk) 16:55, 2 October 2013 (UTC)

I concede, it wouldn't work so well for individual people. There are clearer contrasts, I think, between something like the title of a song of borderline notability on the one hand, and a biological species or a molecular formula on the other hand. bd2412 T 17:04, 2 October 2013 (UTC)

It is practically impossible to create two articles simultaneously. I submit that for every single term that is an undisambiguated title of an article or dab page, there was an original article at that term when no other article with that name existed on WP. When an article about a second use is created, there needs to be consensus that the first use is not the primary topic before its title is disambiguated. This isn't haphazard, BD2412 (talk · contribs); the default claim on the base name for its title by the original article is implicit in WP:TITLECHANGES and the status quo reigns rule. --B2C 16:58, 2 October 2013 (UTC)

The rule isn't haphazard, but the result is. It should be the other way around, a consistent result, no matter which article happens to have been first created. bd2412 T 17:24, 2 October 2013 (UTC)

(ec) It isn't quite accurate to say there needs to be consensus that the first use is not the primary topic before its title is disambiguated. Any admin can make such an evaluation and BOLDly move the articles (or non-admins can as well, if the new name doesn't already exist). Or non-admins might make a technical request if it appears uncontroversial. Unless you're suggesting that any change of primary topic is automatically controversial and therefore must be discussed before moving. Of course, such BOLDness might be challenged (or if it happened some time ago, someone might question whether circumstances have changed. Point is, that whomever creates the second page has to consider what sort of disambiguation to use, and in that process is very likely to consider whether the second topic is equal to or surpasses the first article. It need not be a big production. Consideration of the applicability of TWODABS by the community only really arises when the existing arrangement is questioned. WP:STATUSQUO (which is an essay, BTW, only really applies in the context of reversions) and it DOES NOT mean change does not happen -- it quite explicitly calls for discussion to establish consensus. Similarly, WP:TITLECHANGES primarily applies to controversial title changes. Many would consider that the need to disambiguate a title because there is another article that could have the same title and no obvious primary topic would be a good reason to change the title. I think the clearest, most readily comprehensible guidance to editors is that if there is no primary topic, use a disambiguation page at the base name. older ≠ wiser 17:38, 2 October 2013 (UTC)

A formal discussion is not the (crucial missing words added in red. Sorry! --B2C 04:39, 3 October 2013 (UTC)) only way to reach WP:CONSENSUS, Bkonrad, and it's not even the most common way. BOLD editing (and renaming) is also consensus supported, and much more common than consensus reached or established through discussion. --B2C 20:56, 2 October 2013 (UTC)

*Absolute nonsense. Formal discussion in practice means constrained discussion is suited for dispute resolution. Articles are not built and improved by formal discussion. One of our most useful and consensus-reflecting policy pages, WP:5P, was not built by formal discussion, go review it history. You appear to again promoting confusion between consensus and closing discussions. Formal discussion places like the US legislature are a current example of how formal process does make for consensus. You don't sort out marital problems in a court room. AfD and RM produce respected results because consensus is the the results of the processes are to be respected. Very rarely have I seen anything close to a contested formal RM discussion lead to compromise, negotiation and consensus building. No. Formal discussions are much more likely to entrench initial positions, contrary to consensus building. --SmokeyJoe (talk) 01:00, 3 October 2013 (UTC)

• • Indeed the initial statements was nonsense... due to a typo (now fixed... in red). I presume that's what you're referring to as "nonsense". Sorry about that! --B2C 04:39, 3 October 2013 (UTC)

• sorry for missing the typo and believing that you meant that. --SmokeyJoe (talk) 04:57, 3 October 2013 (UTC)

Unclear what your point is. Your comment made the need for consensus out to be a big deal. I was pointing out that title changes happen all the time based on judgement of individual editors. older ≠ wiser 21:20, 2 October 2013 (UTC)

You wrote: It isn't quite accurate to say there needs to be consensus that the first use is not the primary topic before its title is disambiguated.

My point is that it is accurate. Just because the title is disambiguated by a bold move based on the judgement of an individual editor doesn't mean consensus does not support it. Consensus is still needed before the title is disambiguated; whether the title change is made unilaterally or after discussion is irrelevant. If consensus support is unclear (the change is potentially controversial), then it's not supposed to be changed without discussion. There is nothing inaccurate about "there needs to be consensus that the first use is not the primary topic before its title is disambiguated. That's the point. Is that clear? --B2C 04:34, 3 October 2013 (UTC)

No, it is not accurate to say that consensus is required before a title is disambiguated. Consensus may support actions of an individual, but any individual editor can disambiguate a title. It is a perversion of language to call the actions of a individual editor consensus. Consensus can be assumed if there are no objections, but such acquiescence may last only until there is awareness of the changes. older ≠ wiser 11:47, 3 October 2013 (UTC)

Semantics, but, the quote is not accurate. If disambiguation is required, any editor can do it. This especially applies to new articles written by the editor who is now adjusting the titles. If you were to write a handful of articles today, and then submit requests for renames, you would deserve a trout, or at least a link to Help:Move. Where your wrong is that you imply that consensus exists before anyone has even looked. WP:Silence does not stretch that far (or short). WP:Silence requires reasonable time. You do not need specific consensus to add or improve content, unless maybe you have a bad history of poor judgement. --SmokeyJoe (talk) 05:30, 3 October 2013 (UTC)

Yes, if disambiguation is required, any editor can do it, but only with good reason to believe that consensus agrees that disambiguation is required. Lacking sufficient good reason to believe that means a discussion is required. In any case, there needs to be consensus that the first use is not the primary topic before its title is disambiguated. So the quote is accurate.

And for consensus to exist in a given situation does not mean some significant number of editors more than one have to look at the situation and agree. If policy supported by consensus clearly indicates a certain decision in a given situation, then consensus supports that decision, by definition, even if only one person (like the creator of an article) is even aware of the situation. --B2C 08:06, 3 October 2013 (UTC)

Diego (talk · contribs) makes the following claim above:

At the third step (when the size and notability of the corporation grows), you're assuming that the corporation has grown so popular that it's a contender against the long term significance previously held by the plant (otherwise, it wouldn't dethrone it as primary topic). This means that more visitors will be interested in the company than in the plant, but not enough to make its article primary. In that situation, the disambiguation page is a better compromise than sending many visitors to the plant article.

The third step is: Some time later, as the size and notability of the corporation grows, there is no primary topic. That doesn't necessarily mean "more visitors will be interested in the company than in the plant, but not enough to make its article primary", but let's assume that is the case, because it is possible. Let's say the distribution of interest is 60% for the company, and 40% for the plant, and consider what happens to those entering "Climpaton" in the WP search box and clicking on Go in each case.

Under If there is no primary topic, a disambiguation page should be at the base name, we would have:

1. 100% arrive at the dab page at Climpaton.
2. 40% click on the dab page link to Climpaton (plant)
3. 60% click on the dab page link to Climpaton (corporation).

Under If there is no primary topic, whichever was created first is at the base name, we would have:

1. 100% arrive at the plant article at Climpaton which, for 40%, is the desired destination.
2. 60% click on the hatnote link to Climpaton (corporation).

I fail to see why the dab page scenario is a "better compromise". --B2C 17:28, 2 October 2013 (UTC)

Because there are a number of benefits that DAB pages provide, and that you're not taking into account by only counting page clicks. For a start, in the second scenario 60% of users are being shown a wrong article, while in the first they see a navigation page. You consider arriving to the DAB a failure, disregarding the extra time that readers need to recover from landing at a target article that it's not the one they're looking for. Instead I consider that a DAB page is a hand-tailored search results page, no worse than typing "Climpaton" at Google - in fact much better, as it only shows relevant links that belong to Wikipedia articles, presented in a nice format optimized for navigation.

Second, DAB pages not only help at navigation from searches. As Vegaswikian points out, linking to a non-primary topic produces problems also for internal Wikilinks, when editors create the link to Climpaton expecting that the company will be at the other side. If there is no primary topic, readers following the link will arrive to a DAB page instead than to the plant (which would be totally absurd if the context for the link was about the company). Second, the editor that created the link will get a warning from a bot, prompting them to fix it with the right meaning; and other editors will also be able to fix it with the link disambiguation tool. Diego (talk) 17:49, 2 October 2013 (UTC)

This is a very incomplete analysis. First, there is no reason to think that 100% of readers looking for Climpaton would arrive at the disambiguation page. Many, perhaps most, would choose the desired target based on the parenthetical term from the search box drop-down. Remember, the disambiguation page typically appears after the other uses in the dropdown list. Similarly, incoming links to the disambiguation page have a much higher probability of being fixed than mistaken links to a incorrectly placed non-primary topic, so readers are significantly less likely to arrive at the "wrong" page from internal links if a disambiguation page is at the base name. older ≠ wiser 17:45, 2 October 2013 (UTC)

The analysis is complete with respect to all that matters. Users using the drop-down list are just as irrelevant as Google searchers - the article will be selected through the search mechanism and the sought article will be reached directly. In a TWODAB case, at least one of the two titles will be disambiguated, and will either be recognized as the one being sought, or the one not being sought. In the latter case, if the one being sought is undisambiguated, it will be known to be the desired article. So all that matters is the scenario in which a person enters the search term and clicks on GO, and the analysis is complete with respect to that. --B2C 20:26, 2 October 2013 (UTC)

The analysis is complete with respect to all that matters Strongly disagree for the reasons already stated. Users using the drop-down list are just as irrelevant as Google searchers - the article will be selected through the search mechanism and the sought article will be reached directly -- unclear what this is supposed to mean. For the record I strongly disagree that internal search is as irrelevant as external searches. In a TWODAB case, at least one of the two titles will be disambiguated, and will either be recognized as the one being sought, or the one not being sought. In the latter case, if the one being sought is undisambiguated, it will be known to be the desired article Another fine example of shortsighted binary black/white reductionism resting on a dubious presumption that an undisambiguated title can be recognized as correct equally easily as a parenthetically disambiguated title. I suggest it is far easier for a reader to recognize the desired title in such situations when it is disambiguated rather than having to do a mental calculus -- not X, therefore it must be Y (although maybe Y is something else...I guess I'll just have to click on it to find out). older ≠ wiser 21:27, 2 October 2013 (UTC)

I wrote: Users using the drop-down list are just as irrelevant as Google searchers - the article will be selected through the search mechanism and the sought article will be reached directly

Bkonrad replied: unclear what this is supposed to mean.

Sorry for being unclear. This is crucial to understand. What this means is that regardless of which of the following scenarios are shown in the drop-down list are:

• Climpaton and Climpaton (corporation),
• Climpaton (plant) and Climpaton, or
• Climpaton (plant) and Climpaton (corporation),

the people searching for the plant and the people searching for the corporation will both know which in the drop-down list to click (if nothing else because they'll know which not to click), so, "Users using the drop-down list are ... irrelevant" when it comes to deciding which of the three naming possibilities to use. They are just as irrelevant as those using Google (which also takes you directly to the article you seek, regardless of which of the three naming possibilities are used). So the only situation relevant to choosing which of the three naming scenarios to select is internal search without drop-down list, which is exactly the case considered in my analysis, especially those simply entering the search term (e.g., "Climpaton") and hitting Go.

You wrote: an undisambiguated title can be recognized as correct equally easily as a parenthetically disambiguated title

That's out of context. I'm not suggesting anyone will recognize the title from the undisambiguated title all alone. The context is choosing between the undisambiguated title and the title disambiguated with something that either matches or doesn't match the desired topic. If it matches, the disambiguated title is it; otherwise the undisambiguated is it.

For example, consider Alunite. You may be looking for the mineral, or the town in Utah. Your choices are Alunite and Alunite, Utah. If you're looking for the mineral, even if you don't know about the town, you'll know that Alunite, Utah is not about the mineral, so Alunite must be your article. Similarly, if you're looking for the town and you don't know about the mineral, you'll know Alunite, Utah is your article, not Alunite. There is no need to move Alunite to Alunite (mineral). And there is no more benefit to anyone in disambiguating both titles in any other TWODABS situation, including when neither is primary. --B2C 04:29, 3 October 2013 (UTC)

Yes, I understood that already. But you do not address the point that comparatively it takes longer to process the scenario in which one option is clearly marked and one is unmarked than the scenario in which both options are marked. There is benefit for all to clearly mark the options when there is no primary topic. older ≠ wiser 11:34, 3 October 2013 (UTC)

Why the final clause, when there is no primary topic? Isn't the benefit there even when there is a primary topic? In this very case, for example, Alunite and Alunite, Utah, "there is benefit for all to clearly mark the options"... no? I mean, Alunite is a relatively obscure mineral. I, for one, never heard of it. So if I encounter the town in my travels or on a map or in some citation somewhere, and choose to look it up, there would be benefit if Alunite displayed as Alunite (mineral), not as Alunite.

I can't deny the benefit, just as you cannot deny that the benefit would apply for primary topics too, especially for any primary topic that is not broadly recognized. Like Allunite.

So now we're back to the Matter Scale, because this is really about how much it matters. Yes, it might take a bit longer "to process the scenario in which one option is clearly marked and one is unmarked than the scenario in which both options are marked", but how much longer, and, more importantly, how much does that bit more matter? If you're still unwilling to try to quantify the answers to these qualitative questions in a manner that allows for relative comparisons and weighings, I don't see how the discussion can reasonably proceed. --B2C 17:00, 3 October 2013 (UTC)

It is part of the nature of a primary topic that it is unmarked. Alunite is clearly the primary topic and no one looking for the ghost town in Utah that was formerly a mine for the mineral should be surprised that the mineral is the primary topic. Even if one might not know precisely what "alunite" is, no one familiar with what the town in Utah is would be surprised that something else is at the unmarked base name. There perhaps might be some marginal benefit in creating a redirect from Alunite (mineral), but that would seem superfluous in this case. Such a redirect might be appropriate where the primary topic has multiple facets that could potentially be considered as separate sub-topics. No, I really don't think there is any point to discussing how much it matters that a disambiguation page should be used when there is no encyclopedic basis for choosing a primary topic. older ≠ wiser 17:18, 3 October 2013 (UTC)

You claim: It is part of the nature of a primary topic that it is unmarked.

I apologize for sounding pedantic again, but topics, primary or not, are not marked (disambiguated) or unmarked (undisambiguated). Titles are marked or unmarked. This probably seems like nitpicking but I suggest there is a correlation between clarity in terminology usage and clarity in thinking about the concepts associated with that terminology. So let's try to be more careful, okay? It is part of the nature of the title of a primary topic article that it is unmarked.

Anyway... you claim that "no one familiar with what the town in Utah is would be surprised that something else is at the unmarked base name". Fair enough. But check it out. What you're saying is that anyone familiar with a (non primary) use in a TWODABS situation (like the town of Alunite) would not be surprised to find another use at the base name, even if they're not familiar with it (like the mineral Alunite). I agree again. Let's call this the Bkonrad Principle, okay?

Now, shouldn't this Bkonrad Principle apply to a TWODABS case where neither use is primary? I mean, whether the use at the base name is primary or not can't matter to someone who is unfamiliar with that use. Therefore, whether the unfamiliar use at the base name is primary can't be a factor in whether they are surprised by it being at the base name. And, so, per this principle, anyone familiar with one of the non primary uses should be no more surprised to find the other (unfamiliar) non-primary use at the base name than he or she would be if that unfamiliar use was primary.

This is exactly my point, and one of the main reasons I support, If there is no primary topic, whichever was created first is at the base name for TWODABS cases. It shouldn't be any more surprising, nor any more problematic in any way, for anyone to find the other (unfamiliar) use at the base name in a TWODABS case if it is not primary, than it is for someone looking to find the town of Alunite to find the (unfamiliar) mineral at Alunite, which you agree is acceptable. --B2C 04:38, 4 October 2013 (UTC) added "which you agree is acceptable" for emphasis. --B2C 06:36, 4 October 2013 (UTC)

B2C, the existence of benefit is a very weak argument. Many things have benefit in a narrow context but are a bad idea in a wider context. Communism, for example.

Alunite is not a relatively obscure mineral. We know little of your geological knowledge, so " I, for one, never heard of it" means little. How long is your list of recognized minerals? List of minerals reports that there are only 4700 of them. Alunite is mined commercially around the world.

Alunite is the primary topic for the term "alunite", discounting the town, for multiple reasons. Ambiguous non-primary-topics, where the competing topics are unrelated and individually significant, should be disambiguated, per the principle of least astonishment. The town is not unrelated.

No. The Born2cycle-Bkonrad Principle is not looking very worthy. It is introduced on a false premise (that someone could reasonably be familiar the town of Alunite but not familiar with the mineral Alunite). Familiarity implies holding some non-negligible amount of information. If a term evokes recognition of a familiar topic, and a competing topic is unfamiliar, then this competing topic should not be ascribe "Primary Topic" status. Given that the convoluted argument begins with a dubious invented principle, its not worth further study. It amounts to an arbitrary bright.

Arbitrary bright lines are bad. They would mean that Wikipedia would be run according unnatural rules. This hurts accessibility of the project to new editors, and remaining accessible to newcomers is an even more important principle than anything mentioned so far. --SmokeyJoe (talk) 06:51, 4 October 2013 (UTC)

Smokey, I agree "the existence of benefit is a very weak argument", though I'm not sure why you're pointing this out. "The existence of a benefit" is not my argument.

Bkonrad contends that someone who knows of the town but not the mineral should not be surprised to find the mineral at the base name. I agree. You seem to believe anyone reasonably familiar with the town would also have to be familiar with the mineral. What if it's someone driving through the area with a passing interest in ghost towns? We may disagree on how likely something like that may be, but that's irrelevant. The point is that being familiar with only one of two topics named X does not mean one would be surprised to find the unknown topic at the base name X. Finding an unfamiliar topic at the base name of the non-primary topic being sought is simply not problematic in any significant way. That is not my entire argument, but it is one of the underpinnings. In other words, finding an unfamiliar topic at the base name of the non-primary topic being sought is not unnatural, nor does it hurt accessibility of the project to new editors or to anyone else. --B2C 07:23, 4 October 2013 (UTC)

• What if it's someone driving through the area with a passing interest in ghost towns?

  Well, these people are not familiar with the town, they have just stumbled upon it. As soon as they know just the first little bit, they will have encountered mention of the mineral. Alunite is not very interesting an example. A more challenging example is Welland. Consider Welland ON vs Welland SA. I consider both obscure, and that a whole lot of biases are at play. Also consider Welland ON vs Colin Welland, or any other unrelated pair.

  "The point is that being familiar with only one of two topics named X does not mean one would be surprised to find the unknown topic at the base name X."

  The contrary is exactly my point. If someone can reasonably assume that their X is the only X, they will be astonished to find X udisambiguate is something completely unrelated. This can occur very easily for obscure topics both named ofter sometime historically distant, like Welland. (unlike Alunite).

  "Finding an unfamiliar topic at the base name of the non-primary topic being sought is simply not problematic in any significant way."

  This sounds like you don't value the principle of least astonishment.

  I have never said that all of your arguments are wrong. However, you conflate arguments, and some of your arguments are not palatable.

  "nor does it hurt accessibility of the project to new editors or to anyone else"

  Here you are denying what someone else has already said to you from experience. It is also my experience. If you search for something you expect to find, and find something else, unexpected, unrelated, unfamiliar, then you are astonished, and astonishment has a discouraging feedback to the searcher. This discouragement is a negative contribution to accessibility. Have you read much about the principle of least astonishment? --SmokeyJoe (talk) 08:02, 4 October 2013 (UTC)

If you're looking for the town and you don't know about the mineral, you ought to, and you will as soon as you learn the basics about the town. --SmokeyJoe (talk) 05:17, 3 October 2013 (UTC)

Diego Moya (talk · contribs), indeed, arriving at a dab page is a failure, as that is not arriving at the article being sought. Similarly, arriving at the "wrong" article is also a failure. Further, these failures are comparable. Neither is better, or worse, than the other. Both make it obvious that the sought article was not reached, and both make recovery from the failure - clicking on a link to reach the sought article - equally clear and easy.

The claim that there is extra time required to "recover" from landing on a "wrong" article, as compared to landing on a dab page, is pure fiction, as far as I can tell. Do you need this extra time when it happens to you? I know I don't. Do you know anyone who needs the extra time? --B2C 20:56, 2 October 2013 (UTC)

The claim that there is extra time required to "recover" from landing on a "wrong" article, as compared to landing on a dab page, is pure fiction, as far as I can tell Another faulty presumption. Have you ever watched a well-formed usability test? I have and I'd say it is far more likely to disorient a reader to end up at the "wrong" article than to arrive at a disambiguation page. older ≠ wiser 21:20, 2 October 2013 (UTC)

In that case, perhaps the question should be, how can we better reorient the reader who has landed at the "wrong" article? Perhaps if the situation is a TWODABS with one topic being only slightly more notable than the other, we could make the hatnote more prominent. bd2412 T 21:28, 2 October 2013 (UTC)

The claim that there is extra time required to "recover" from landing on a "wrong" article, as compared to landing on a dab page, is pure fiction, as far as I can tell - It is not. As Bkonrad pointed out, this old discussion had an editor who found people requiring up to 30 seconds to find out the hat note, and others not discovering it at all; this mirrors my experience when arriving to the wrong article. If you try to optimize the hatnote, you start acting against the purpose of the article, which is allowing the content of the article to be read. If you make the hatnote more salient, you hurt those readers who have arrived to the right article. The way to help readers is to not force them into the wrong article from the start. As DAB pages are optimized for browsing and are more lightweight than articles, they load faster and finding the right link is easier. If you allow each type of page to be used for its main purpose (disambiguation pages to decide ambiguous titles, articles to read content), instead of mixing their goals, each page can be used to its best. Diego (talk) 22:42, 2 October 2013 (UTC)

By that reasoning, why not make Apfel and Mouse and George Washington disambiguation pages? They are all very big pages, with comparatively long load times, so a person reaching them while searching for something else will be terribly inconvenienced. bd2412 T 00:44, 3 October 2013 (UTC)

I submit that no reasonable reader would search for "Apple" or "Mouse" or "George Washington" and have no recognition of the topics found at these articles. No reasonable reader would be astonished at what is returned. Therefore, they are good at the undisambiguted titles. --SmokeyJoe (talk) 01:23, 3 October 2013 (UTC)

A reader looking for the Apple computer company might not be "astonished" but they will be just as inconvenienced by the page load time as one who is astonished. If the links that people make are any indication, quite a few editors don't give much thought to what will be on the other end of the link when they type it (whether in the edit box or a search window). bd2412 T 02:24, 3 October 2013 (UTC)

A reader looking for apple computer searching "apple" should expect the fruit to appear. They may be inconvenienced if they have difficult access, but they shouldn't be astonished. I consider the "minimize astonishment" argument to be more important than the "minimize inconvenience", although both are valid. B2C appears to have no consideration for astonishment of a minority, weighing consideration of averages only. I'm not sure how you weigh minimizing minority astonishment verses maximizing average ease of access (noting averages will be dominated by typical readers accessing popular pages)?

Quite a few reader don't give much thought to what will be on the other end of the link when they type?. Yes. This is even my behaviour when working on a good computer with fast connection speed. Should we title articles to anticipate common clumsy searching? I think we should weight the principle of least astonishment highly, and rely on search engines to interpret search queries. I suggest that fast and easy internet access encourages clumsy searching, and that we are not working for quality if we optimize for clumsy searching. (google will always beat us in that race)

Quite a few editors don't give much thought to what will be on the other end of the link when they type it? I appreciate that this is a source of irritation, including to people like you who fix misdirected links. I can well imagine that not-so-careful editors dealing with popular topics will often link to the undisambiguated title regardless of whether the undisambiguated title hosts a DAB page. This is an editor issue that requires education, and I think it is bad practice to alter the product (compare Wikipedia:Product, process, policy) to ameliorate editors' bad habits. --SmokeyJoe (talk) 03:49, 3 October 2013 (UTC)

No one is suggesting that there is anything wrong with Apfel.

What about someone searching for the town of "Alunite" and ending up on an article about the mineral at Alunite? Do you believe that article should be moved to Alunite (mineral)? Why or why not? --B2C 04:43, 3 October 2013 (UTC)

I added the hatnote {{See also|Alunite, Utah}} to Alunite.

No, I would not support disambiguating Alunite (unless we decided to disambiguate all such minerals).

Reasons: (1) Alunite, Utah is not a real town, but a company town of no known wider significance, meaning no one will look for it unless they already know about it, nd if they know about the company town they should be expected to already know abou the mineral; (2) The town has no known continuing significance, being a ghost (company) town; (3) the town was named after the mineral mined locally that motivated creation of the town, and so the town name is derivative of the mineral.

There is less reason to move Alunite than to disambiguate Apple (fruit). --SmokeyJoe (talk) 05:13, 3 October 2013 (UTC)

Okay, good. We agree about that, though I take it you meant to say, "... than to disambiguate Apple" (Apple (fruit) is already disambiguated). --B2C 07:57, 3 October 2013 (UTC)

OK, a lot to take in here since I last checked, and a whole lot I disagree with. First, yes, you are being overly pedantic. A primary topic, in the contect of disambiguation, only exists when there is an unmarked title (or redirect). You objections are purely pedantic). You wrote: What you're saying is that anyone familiar with a (non primary) use in a TWODABS situation (like the town of Alunite) would not be surprised to find another use at the base name, even if they're not familiar with it (like the mineral Alunite). No, this is not my point at all. It is not "any" use -- it is the primary use. It is precisely because there is a very clear primary topic that there is no surprise that the non-primary topic does not have the unmarked title. Turn the example around. Suppose that the ghost town article was created first at the unmarked name (and that the creator ignored the USPLACE naming convention). I expect that readers looking for the mineral would be very surprised and irritated to find the article on the ghost town with the unmarked title.

Similarly, if neither of two topics are primary and one is placed at the unmarked title, there will be a element, if not of surprise, perhaps more accurately of disorientation, for readers looking for the other topic. As SmokeyJoe has pointed out, your proposition completely ignores usability and assumes that hatnotes are perfectly equivalent to dab pages for assisting in navigation. They are not. With two non-primary topics of approximately equal notability (which after all is the essence of a TWODAB situation), first, the likelihood of a reader arriving at the disambiguation page is minimized by first, having both titles marked and secondly, by having the disambiguation page at the marked title -- mistaken links to the dab page are more likely to be fixed. older ≠ wiser 09:21, 4 October 2013 (UTC)

What you're talking about suggests an extension to recognizability in the context of ambiguous names. If I understand correctly, you guys believe that recognizability is more important in titles of topics with ambiguous names than in titles of topics with unique names. I mean, if Slagox is the name of only one topic, say an obscure plant, then you're fine with its title being at Slagox, right? Because it needs to be recognizable to only people who are familiar with the plant. But if it's also the title of some obscure book (just notable enough such that neither the plant nor the book is the primary topic), then the plant article title needs to be recognizable not only to those familiar with the plant, but also to those unfamiliar with the plant but who are familiar with the book. Similarly, the title of the book needs to be recognizable to those familiar with the book and to those unfamiliar with the book but who are familiar with the plant... right? Is that what you're trying to say? If so, this recognizability nuance is not explicitly documented anywhere, so far as I know. Would you favor changing the description of Recognizability to say that somehow? Currently it says this:

The title is a name or description of the subject that someone familiar with, although not necessarily an expert in, the subject, will recognize.

--B2C 18:54, 4 October 2013 (UTC)

I don't think Recognizability is the right place. It is really a matter of disambiguation, which I think already addresses this under WP:TWODABS: As discussed above, if an ambiguous term has no primary topic, then that term needs to lead to a disambiguation page. In other words, where no topic is primary, the disambiguation page is placed at the base name. You seem to be the only one actively disputing that guidance. older ≠ wiser 19:39, 4 October 2013 (UTC)

Another hypothetical TWODABS example

Above, Bkonrad claims the hypothetical scenario I constructed favors my position. He encourages us to consider a situation in which the second use is "out of the box of comparable notability to the first." Fine; let's consider the effect of the two TWODABS approaches being discussed on such a situation:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation). The article is created when the corporation has achieved notability comparable to the plant. So, there is no primary topic.
3. Eventually the corporation meets the primary topic criteria

Under If there is no primary topic, a disambiguation page should be at the base name, we would have:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation). The article is created when the corporation has achieved notability comparable to the plant. So, there is no primary topic.
   • The plant article is moved to Climpaton (plant)
   • A new TWODAB dab page is created at Climpaton
   • The corporation article remains at Climpaton (corporation).
3. Eventually the corporation achieves primary topic status
   • The plant article remains at Climpaton (plant)
   • The TWODAB dab page is moved to Climpaton (disambiguation)
   • The corporation article is moved to Climpaton.

Under If there is no primary topic, whichever was created first is at the base name, we would have:

1. An article about a plant named Climpaton is created at Climpaton; it is the only use of that name.
2. Some time later, a new article about a small new software corporation also named Climpaton is added at Climpaton (corporation). The article is created when the corporation has achieved notability comparable to the plant. So, there is no primary topic.
   • As the original use the plant article remains at Climpaton.
   • The corporation article remains at Climpaton (corporation).
3. Eventually the corporation achieves primary topic status
   • The plant article is moved to Climpaton (plant)
   • The corporation article is moved to Climpaton.

Now, which scenario is simple, more stable and less contentious? Which is more complex, less stable and more likely to foster contention?

Consider that under If there is no primary topic, whichever was created first is at the base name, nothing changes at stage 2, so there is no need for the community to even evaluate anything at that point. But under If there is no primary topic, a disambiguation page should be at the base name, the community is required to ascertain whether the second use is sufficiently notable for the first use to no longer meet primary topic criteria, to decide whether to disambiguate the title of the plant article to make room for the dab page at the base name. --B2C 17:12, 2 October 2013 (UTC)

• You're taking about apple. It is not complicated at all. The natural world topic takes precedence, and everything named after it gets disambiguated. --SmokeyJoe (talk) 00:58, 3 October 2013 (UTC)

What? The topic of apple is unquestionably (per consensus) the primary topic of "apple", so of course I'm not talking about that. I'm talking about when there is no primary topic. --B2C 07:54, 3 October 2013 (UTC)

<aside> Your use of "unquestionably" makes your statement false, unless you believe in steamrolling minority opinion. There is significant minority opinion found in that discussion. You ascription of "consensus" is wrong.</aside>

Your hypothetical construct resembles Apple so much that many of your hypothetical premises are unrealistic. I suggest that you stick to real examples and not ask others to follow you down fantasy garden paths. --SmokeyJoe (talk) 08:12, 4 October 2013 (UTC)

This second Climpaton hypothetical represents any one of countless TWODABS situation in which initially neither of the two topics is primary, and eventually the newer use becomes primary. Apple is not even close to TWODABS (more like FIFTYDABS), nor did it ever lack a primary topic; the original use was always the primary topic. Apple meets neither of the factors that are relevant to this discussion (Apple, Inc. doesn't even meet the irrelevant "software corporation" factor). Are you following this at all? --B2C 08:42, 4 October 2013 (UTC)

Required link to the other non-primary article in TWODABS

As someone who was around when TWODABS was drafted, I know that the intention to write the guideline was to avoid removing the link to the non-primary article from the primary one. Thus inserting in the text that the DAB page can exist "instead of a link directly to the other article" would reverse the very reason why the guideline exist. The link to the other article should not be removed. Diego (talk) 16:12, 26 September 2013 (UTC)

Bkonrad, the text you're writing would have the effect of allowing the link to the other article to be removed even when there are only two articles. This was not the intention of the guideline. In the case of Sexual harassment, the link to Sexual Harassment (The Office) could be removed precisely because Sexual harassment (disambiguation) is not a TWODAB anymore. The text should be rewritten to clarify that removing the link to the other page can be done only after expanding the DAB. Diego (talk) 16:20, 26 September 2013 (UTC)

As someone who was also around and actively participated in discussions that led to TWODABS, I respectfully disagree with your recollection of the intention. One of the initial impetus for TWODABS arose out of discusion of the santorum rule and Trivial hatnote links where a number of editors wanted to avoid linking directly to what they considered a "trivial" topic in the hatnote of a primary topic and instead link to a disambiguation page. Prior to this, I think the guidance was much clearer that such disambiguation pages could be speedily deleted on sight. The {{Only-two-dabs}} template and a since deleted {{Consider disambiguation}} were intended to allow some flexibility in the delete on sight rule. And current practice shows considerable variation. On the one hand, if there is no link to a twodab disambiguation page from the primary topic hatnote, the disambiguation page is effectively useless and should be speedily deleted (IMO). Yes, the Sexual Harassment disambiguation has since been updated, but it was recently given as an example by another editor, which is why I mentioned it. There are variations in current practice.

• Technology contains links to both Technology (band) and the TWODAB Technology (disambiguation).
• Poet, which is the target of redirect at Poetess, contains a link the American rapper The Poetess as well as to the TWODAB Poetess (disambiguation).
• Aphrodisias contains only a link to Aphrodisias of Cilicia and Aphrodisias (disambiguation) is not linked in the hatnote and is not linked from any other article.
• Celsius, which is target of redirect at Celcius, contains only a link to Celcius (album) and not to Celcius (disambiguation). The disambiguation page is only linked from the album article.
• Retrograde and prograde motion, which is target of redirect at Retrograde motion only contains link to Apparent retrograde motion and not to the Retrograde motion (disambiguation), which is only linked from a redirect at Retrograde motion (astronomy), which is flagged as an incomplete disambiguation.
• Jalt is a redirect to Japan Association for Language Teaching, which doesn't have any hatnote. Jalt (disambiguation) is not linked from any other article.
• Chorwon County, which is target of redirect at Chorwon, only has direct link to Cheorwon County. Chorwon (disambiguation) is not linked from any article.
• Hwadae County, which is target of redirect at Hwadae, does not have any hatnote. Hwadae (disambiguation) is not linked from any article.

Now granted, some of these obviously need to be fixed and are not necessarily illustrations of best practices, but in looking over the archived discussions, it seems pretty clear that a significant number of editors wanted the option to in effect obfuscate what they considered as trivial hatnote links behind a disambiguation page, even if the page contained only two links. I didn't agree with the reasoning, and it was pretty well shouted down at Wikipedia talk:Hatnote/Archive 3#Trivial hatnote links, but the notion appears to have some residual viability such as in discussions about hatnotes for Santorum, or Sexual Harassment, or Johnny Cash -- all of which ultimately resulted in expanding the disambiguation page beyond the TWODAB threshold (although TBH, Santorum (disambiguation) is pretty marginal as Santorum Amendment is really an unambiguous partial title match that is already addressed in the main article. Johnny Cash (disambiguation) is also pretty marginal.

Now, if I'm wrong and there is consensus that a hatnote at the primary topic for a TWODAB does not need to link to the disambiguation page, then I'm confused as to why these TWODABS should not simply be speedily deleted as completely useless.

As a side note, there is at least one pretty remarkable SINGLE TOPIC disambiguation page at Season 4 which has been kept by consensus a couple of times now. older ≠ wiser 17:48, 26 September 2013 (UTC)

Some thoughts. First, for pages with a next-most-likely topic, I like the presentation of Apfel and Mouse, each of which has a link to the number two topic, and then a link to the disambiguation page. Obviously these don't apply to a TWODABS situation, but they might were a third link is added to the disambig page. Second, Apparent retrograde motion is not ambiguous to Retrograde motion, at least no more than Penrose stairs is to Stairs. Celcius (disambiguation) should be deleted or redirected to Celsius (disambiguation). We have no indication that the redlink is notable. bd2412 T 18:01, 26 September 2013 (UTC)

Bkonrad, please read what I'm actually saying because we're not in disagreement, and you're misunderstanding my position. Of course you should always have a direct link to the "trivial" non-primary article when the DAB has only two links. And I'm also one of the guys to defended keeping a link to the DAB instead of deleting it, allowing it to grow instead of just deleting it in place. My reverting of your edit is because I agree with everything you've said in the talk page and the edit summary, but you're actually writing the opposite in policy of what you defend here.

Your sentence "an about hatnote can be used to link to a disambiguation page (either in addition to or instead of a link directly to the other article)" means that it's OK to have the link to the two-dab page ONLY, and remove the direct link to the other article. If you mean that both the direct link and the link to the DAB need to exist, you should write just that. Instead what you've written allows for all the following, elevating them to common possibilities:

• Technology can use a link to the TWODAB Technology (disambiguation) [only], instead of the link to Technology (band) [which could be removed].
• Poet, can contain a link to the TWODAB Poetess (disambiguation) [only], instead of the link to The Poetess [which could be removed].
• Retrograde and prograde motion could contain a link to Retrograde motion (disambiguation) instead of the current link to Apparent retrograde motion.

See the pattern? This is what you say is allowed; none of those possibilities are current community practices.

(Jalt is a weird case, as the DAB contains only one "ambiguous" article - the other link is to a section; I've added a "redirects here" note, although it could be the rare exception to the "include a direct link" rule).

If you really want to say that the above changes should be commonly allowed, then say so, and present arguments to defend that; but I still think it's a change of the previous policy, and certainly is not what you're defending in the talk page and summary edits (heck, with this edit you say that readers should not be forced to the DAB page to reach the other article, but that's just what the text you included allows). Diego (talk) 19:28, 26 September 2013 (UTC)

This language an {{about}} hatnote can be used to link to a disambiguation page, in addition to a link directly to the other article. suggests that is is OK if the hatnote does not link to the disambiguation page. It seems completely pointless to me to have a disambiguation page that is not linked from the primary topic. We'd be better off suggesting that these be speedily deleted. While my preference would be to only have a direct link to the alternate topic in the hatnote, if there is a disambiguation page and there is consensus (for whatever reasons), I don't think it is so terrible to have the hatnote link only to the disambiguation page. Of course, if after some time the dab page is still only two dabs, it should IMO be deleted and a direct link to the alternate article placed in the hatnote. older ≠ wiser 20:10, 26 September 2013 (UTC)

I think it's contextual. In some cases a link to topic #2 + the dab should be in hatnote; in others just the dab suffices - indeed it was the hiding of 'trivial' links that inspired the compromise of two dabs template. If the dab can't be expanded after a certain time, then it can be deleted. We should however always have at least a link to the dab page from the primary topic.--Obi-Wan Kenobi (talk) 11:31, 27 September 2013 (UTC)

Here is another case: Nagina Masjid (disambiguation). Nagina Masjid only links the disambiguation page in the hatnote, which seems pointless to me. bd2412 T 13:54, 27 September 2013 (UTC)

Yup; that would be a case where you could just remove the dab page entirely.--Obi-Wan Kenobi (talk) 14:09, 27 September 2013 (UTC)

Actually, examining the history, I saw there was another entry that had been deleted, although there it is mentioned in another article, so TWODABS might not apply in this case either. older ≠ wiser 14:25, 27 September 2013 (UTC)

Some TWODABS statistics.

Thanks to User:R'n'B, we now have a more precise grasp of how much of this project involved WP:TWODABS pages. Per R'n'B, "User:RussBot/Two-link disambiguation pages/001 is the beginning of a list of over 43,000 pages. That is, out of 235,400 total disambiguation pages, almost 20% contain only two (or fewer) links". That is quite a lot, in my opinion. bd2412 T 16:46, 4 October 2013 (UTC)

Interesting fact. The discussion (above) is pretty intense and too much for a newcomer to take in. Thanks keeping it brief! Nwjerseyliz (talk) 23:57, 4 October 2013 (UTC)