Wikipedia talk:Hatnote/Archive 3

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

→

Archive 9

Arriving from search engine results

Regarding the section "Disambiguating article names that are not ambiguous", I am wondering why we cannot have hatnotes linking to disambiguation pages from an article that is disambiguated? I've followed the process as outlined in the guidelines during my time here, but I'm wondering why the guidelines do not address the possibility of readers arriving to a Wikipedia article from search engine results. For example, if we have two film articles a few years apart, they'll be disambiguated by the release year. If someone searches in Google for one of the films, they may get the other (especially if it is more popular) and click on it, not really remembering the specific release year. If they go to that article about the other film, there's not a direct way to access a disambiguation page to find the correct film. The person has to then manually search for the main term within Wikipedia to get to the disambiguation page, perhaps having to go through the primary topic's article in the process. Does that example make sense? I think the same thing can happen if a reader only has a title in mind but does not know anything else about it. I'm wondering if not having hatnotes to disambiguation pages makes corrective navigation difficult. Erik (talk | contribs) 18:26, 28 October 2011 (UTC)

I think this does make sense, and I think WP:NAMB could use a second look in this regard. It says (emphasis mine): "Here, the hatnote is inappropriate because a reader who is following links within Wikipedia or using Wikipedia's own search engine would not have ended up at [the unambiguous article]." But it isn't clear why we would only want to help out readers who arrived at the article in those ways, as opposed to, say, from Google. For example, a google search for "Orion" returns one Wikipedia article in the first page (for me, at least): Orion (constellation). In contravention of this guideline, this article does have a hatnote to Orion (disambiguation), which is extraordinarily helpful for those arriving at Wikipedia on this page.--Trystan (talk) 06:50, 12 February 2012 (UTC)

I'm with you on this: at present I'm law-abidingly removing hatnotes which are proscribed by this rule when I see them, but I would prefer it if every disambiguated article had a hatnote with a link to the relevant disambiguation page (and if there isn't a dab page (yet), then make that a redirect to the article at the base title). I quite agree that we should help the reader who has landed at the wrong article via a search engine and needs our help in finding the right one. PamD 09:56, 12 February 2012 (UTC)

Guess I'll be the naysayer here. Reversing NAMB would result in lots more cluttersome (and arguably unnecessary, hence the current guidance) hatnotes on articles, and I question the search engine rationale: surely a good search engine will have our other same-named articles further down in the results, or if the articles are that closely related (e.g. film remakes) there'll already be links between them in the article's body. --Cybercobra (talk) 10:41, 12 February 2012 (UTC)

At least from my personal experience, it's fairly common to arrive at the wrong disambiguated page via search engine results. Sometimes, you genuinely don't know whether or not the search engine result is the right one; for example, consider an article called "Common Song Name (Some Band song)", where you where searching for a song name but didn't know the band - in this case, you probably won't recognise that the disambiguated link given in the search engine is the wrong one, but may discover your mistake fairly quickly when you get to the page. The other case is where you see a search engine result to a Wikipedia page which you know isn't quite the right one, but assume you will be able to get to the right page once you get to Wikipedia. As I say, IME these are not uncommon occurences. However, it's also true that hatnotes clutter up the top of the page, so I can see the argument for avoiding them if we can. I wonder if a sensible solution would be to put links to the disambiguation page in the "See also" section on pages where this guideline forbids a hatnote. This goes against the guideline at WP:SEEALSO, but I'm not sure what the justification is for the restriction against disambiguation links in See also sections.VoluntarySlave (talk) 23:14, 23 February 2012 (UTC)

I agree that hatnotes could be useful in such cases, despite the added clutter.

How about slightly reducing the typeface size in all hatnotes? (In addition to italics and indentation as at present.) As well as reducing the space taken up by the hatnote, the smaller font would be an immediate visual signifier, at only a glance, that the information is not part of the article content. Hatnotes would then be less distracting for readers who have navigated to the right article. At present, hatnotes are instinctively hard to ignore because italics on other websites signify subheadings or standfirsts.

— Richardguk (talk) 00:43, 24 February 2012 (UTC)

The worst of clutterings is that editors think that a Hatnote is disambiguation. It is not. See, for example, {{main}}. -DePiep (talk) 01:02, 24 February 2012 (UTC)

WP:RELATED

Backstory: I saw this edit to The Hunger Games, in which an editor removed the hatnotes linking to The Hunger Games trilogy and The Hunger Games (film), citing WP:RELATED. Not being familiar with that shortcut, I checked it out and saw how it could be interpreted to support removing those hatnotes, but I'm pretty sure that the intent of the guideline was not to bar hatnotes linking between two separate topics that are nevertheless related in some way, such as a book and a film adapted from the book. (I reverted the edit.)

I then edited the guideline to try to clarify what I interpreted to be its intended meaning. As far as I can tell, my edit did not change the meaning of the guideline. My edit was then reverted, so here I am to discuss. I admit to being puzzled because as far as I understand the reverting editor's edit summary, it would seem to be an objection to the actual guideline (in that the editor seems to feel that hatnotes are, and should be, permitted to link to related topics), and not to the text that I had added. Theoldsparkle (talk) 19:21, 2 November 2011 (UTC)

I'm the person who reverted the addition. It may be that the documentation hasn't kept up with actual practices. Hatnotes in recent discussions (such as the renaming of the base {{dablink}} template to {{hatnote}}) seem to seen as used for purposes beyond merely disambiguation. I think the intent of that section is to avoid linking to unambiguous references to topics that are clearly subtopics of the article. But considering that templates such as {{details}}, {{see}}, {{further}}, and {{main}} are all considered to be hatnotes, I think perhaps that section might need some further clarification about when it applies. I didn't see the edit as helping to clarify this. In particular, removal of the qualification of "highly" related, and the implication that it is improper to use hatnotes to link to an article about a related topic, or a specific aspect of the general topic seemed confusing since some hatnotes appear to be designed for precisely that purpose. Perhaps the qualification that [h]atnotes are sometimes improperly used was intended to deflect from the arbitrary application of the rule, but perhaps the topic needs further attention. I'm not entirely sure what direction to go with it. IMO, the examples here are at best advisory only in a very general sense and are not actionable policies with widespread acceptance (or even understanding or awareness). older ≠ wiser 20:53, 2 November 2011 (UTC)

Your objection makes more sense to me now. I'm thinking in terms of hatnotes being used to disambiguate multiple topics (and I think the vast majority of this page is also written with that type of hatnote in mind). I interpret the guideline to mean that disambiguation hatnotes should be used to disambiguate two topics, not to link to a related aspect of the same topic, which is the point I was attempting to clarify in my edit. I understand your point that there are other types of hatnotes where linking to a related topic would be acceptable; I don't think this page is written in a way that takes that into account, and any method of fully addressing that would likely be far more work than I am personally interested in. For the specific subsection in question, would your concerns be addressed if we used the phrasing in my proposed change, but said "disambiguation hatnote" instead of just "hatnote", both above and below the example text? (I would also continue to support changing "highly related" to just "related" in the heading, as I find the "highly" confusing; it seems to suggest that there's nothing wrong with using disambiguation hatnotes to link to a topic that's only somewhat related, as long as it's not related enough to be highly related.) Theoldsparkle (talk) 21:37, 2 November 2011 (UTC)

If there's no objection here, I'll take that to mean my revised proposal is acceptable. Theoldsparkle (talk) 17:52, 4 November 2011 (UTC)

Just to be clear: a hatnote is not a disambiguation note.

Well, sometimes it is. About nine out of ten are. But still: a hatnote should be seen from the hatnotes task: "Has the reader arrived on the right place?". -DePiep (talk) 22:08, 4 November 2011 (UTC)

Which is why I've suggested the change to make explicit that we are discussing disambiguation hatnotes, and not all hatnotes. Theoldsparkle (talk) 20:04, 7 November 2011 (UTC)

Hmm. disambiguation hatnotes sounds like there is a definition. But I do not know about "this is a disambiguation hatnote". Better is: "this is a diasmbiguating hatnote"! I mean to say: the subcategory (unknown as such) is misleading the mind -DePiep (talk) 20:30, 15 November 2011 (UTC)

• I encounted a similar edit here: [1]; in this case the disambiguation link to Superman Returns (video game) has been removed from Superman Returns. I would like to query the interpretation of WP:RELATED in regards to this edit. In the example given in the guideline, Extraterrestrial life in popular culture clearly does not need to be be disambiguated from Extraterrestrial life i.e. they do not share a base name. I believe the intention of the guideline is to not treat sub-topics as disambiguated topics. In the case of Superman Returns, while the video game is clearly a sub-topic it also shares the base name, so the hatnote was serving its correct function by disambiguating two topics that share the same title. Please can someone confirm if my interpretation is correct? I believe the hatnote should remain and WP:RELATED does not sanction its removal in this particular instance. Betty Logan (talk) 16:12, 29 December 2011 (UTC)

I agree the hatnote should be there. Good example for the lead in WP:HATNOTE: "Hatnotes help readers locate a different article they might be seeking". Since the addition "(video game)" is a disambiguation term (and not part of the title), the disambiguating hatnotee is correct. I reinstalled the hatnote. Note for further reading: WP:DAB is also very well developed into this area. -DePiep (talk) 16:25, 29 December 2011 (UTC)

I've attempted to further clarify the guideline to prevent it being misapplied in this way. Theoldsparkle (talk) 19:47, 29 December 2011 (UTC)

Slogan

Even users contributing for many years with tens thousand of edits sometimes misunderstand the purpose of disambiguation hatnotes and have a wrong intuition about it. For one pair of articles see for example this discussion and this attempt to remove a hatnote in the opposite direction (note that "material implication" was a redirect for 6 years until recently). IMHO WP:Hatnote and WP:DABLINK need some short slogan which will distinguish hatnotes from "See also" and navboxes. I propose:

“ Hatnotes belong to the title, the rest of article – to the topic. ”

Incnis Mrsi (talk) 06:43, 27 February 2012 (UTC)

Nice. Josh Parris 08:28, 28 February 2012 (UTC)

I am one of those contributors, Incnis Mrsi! Even after many edits wrt hatnotes, I still have to look up some. And this is where my trouble in understanding comes from: as a baby, the "hatnote" was highjacked by disambiguation guys! They allways keep mixing these two up. They allways assume that hatnotes are there to disambiguate. The good thing of this post is that it writes "disambiguation hatnotes". That is correct, but not useful because they are not specified anywhere. And too often the adjective is left out uncounciously.

Now all grown up hatnotes are together in Category:Hatnote templates, including the dab ones. Andthere are others too, like: {{main}}, {{see}}. They do hatnote, but they do not disambiguate. And also: they pertain to the topic not just the title. Other non-dasambiguation hatnotes: {{selfref}}, {{Redirect}} (in possible situations). In general, disambiguation in WP is about same-spelling, but very narrowly used through WP:DAB.

Concluding: the slogan is wrong (it is talking about disambiguation hatnotes only, right?). My own slogan would be like:

“ Hatnotes exist to help the reader answering: Am I on the right page? (when there are spelling issues, go to WP:DAB) ”

But it could be improved. -DePiep (talk) 09:30, 28 February 2012 (UTC)

Less is more

There is a design philosophy that I would really like to see mentioned prominently in this policy, and that is the idea that less is more. The hatnotes are part of the interface, rather than part of the article, so the principles of user interface design should really be applied. This includes the simplicity principle. Long lists of low-importance hatnotes, such as those found on the Canal article, actually distract both from the article and from the purpose of the hatnotes.

"Canals" redirects here. For the 14th-century Spanish author, see B. de Canals. For the municipality in Spain, see Canals, Valencia.

For other uses, see Canal (disambiguation).

Not to be confused with Channel (geography).

See also: Lists of canals

The reader has to process each of the hatnotes in turn, see if they are relevant to their goal, then (usually) move on to the next line. This is overwhelming users with alternatives and confusing them with unneeded information. It would be much cleaner to say, for example:

For other uses and alternative meanings, see Canal (disambiguation).

I know you guys are sold on the idea of hat notes because you probably spend so much time supporting them, but to me they are almost always an obstacle. I think I've only had a handful of instances where the hatnote was actually useful, and in each case a single line hatnote sufficed. So please, please, include the concept of less is more. It makes for great art, great photography, and some of the most popular interfaces. Just see Google, for example.

Regards, RJH (talk) 22:02, 23 March 2012 (UTC)

I have spend much time with hatnotes last year, but I am not tied to the current setups. I am not sold. Even better, I support this suggestion. It is a good line of thinking. -DePiep (talk) 22:47, 23 March 2012 (UTC)

I agree completely. The simpler, the better. People go overboard with a desire to "complete" any element of the page they can—headings for every paragraph, every field in every infobox, distantly related articles in navboxes, hatnotes for any possible link. It's something we always need to fight against. Hatnotes should be kept to a minimum. —Designate (talk) 01:46, 24 March 2012 (UTC)

+1. NR, 20:50UTC, 24-Mar-12

Thank you. I added a paragraph to the guideline. Hopefully it meets with everybody's approval. Regards, RJH (talk) 15:26, 4 May 2012 (UTC)

Removal of spurious hatnote

Moved from Talk:Health care § Removal of spurious hatnote

(Discussion was moved here by User:RoadView, 11:31, 23 April 2012‎)

Hi, an editor recently re-added a hatnote, that had previously been removed by Rangoon11, referencing an episode of a TV show that happens to be called 'Health care'. I think per the Hatnote policy WP:HAT, and this in particular: "Hatnotes are meant to reduce confusion and direct readers to another article they might have been looking for", this hatnote does not belong. The idea that a user came to the health care article but was actually looking for an episode of "the office" is rather silly. if we started adding hatnotes for every episode of every TV show that happened to use a common noun, it would be a mess. Using wikipedia's own search engine, typing 'health care the office' brings you to the right article - but I have a hard time believing someone would just type 'health care' and expect anything other than the article on 'health care'. As such I've removed the hatnote, and would like other editors to weigh in and defend the hatnote if you think it is worthy. --Karl.brown (talk) 13:35, 24 March 2012 (UTC)

I'm the one who added the hatnote. I agree with you that it seems unlikely that someone would come to this page and be looking for an episode of "the office". But based on WP:SIMILAR, these 2 articles have the exact same name and at that point it's not about how silly it may seem. If it wasn't the exact same name I would not be in favor of it, however in this situation, it is. When I type health care into the search box, it takes me here, and there is no way for me to get to "the office" page unless I type "the office". Maybe there are some people who are unfamiliar with the whole search process or disambiguations and think that once they get here and this was all they saw, that's it. It doesn't really seem like it's too much of a problem for 1 hatnote to site at the top of the article that you can either choose to use or skip over. I think this makes Wikipedia more versatile. If there were other articles with the exact same name as "health care", then we could create a new disambiguation page, so there would likely never be more than 1 hatnote, which completely limits clutter. RoadView (talk) 14:08, 24 March 2012 (UTC)

Thanks. I went to look for other office episodes, and they all have rather generic names. It's not just about similarity - it is about whether readers may have been looking for a different article - and you admit yourself that it is unlikely someone would come here looking for the office. There are hundreds of TV shows, thousands of episodes, and almost all of them have very generic titles. If we started adding hatnotes to all relevant wikipedia articles then it would cause an immense amount of clutter. Thus, until there is justification for adding a disambiguation page here for health care (and on such a page, i think it would be ok to have the office espisode, but a single office episode is not enough to merit a disambiguation page), I think we should remove it. --Karl.brown (talk) 14:38, 24 March 2012 (UTC)

We are in agreement about how often we think it would be used, but I still think, and I'm not usually one to cite policy, based on WP:SIMILAR it should be added regardless of if we think it will be frequently used. I don't think 1 hatnote automatically makes it cluttered. Even if other articles had 1 hatnote for other identically titled pages, I still think it makes it more versatile for those who actually did come there looking for it. If there are 2 or more articles that could be used as hatnotes, then they would be combined into 1 disambiguation page, therefore substantially limiting clutter in my opinion. So even if we did add all the hatnotes for all tv episode articles and such, it would only ever likely result with either a hatnote to the other article or a hatnote to the disambiguation page. I think the hatnote concept is useful in general and this specific hatnote may very well be useful to some, irrespective of if it will widely be used or if we personally don't have a lot of use for it. I would be completely taking your side if the title was not exactly the same. Then I would put most of the weight on how likely it is to be used. This is an interesting topic and I hope others will weigh in on it. 15:02, 24 March 2012 (UTC) — Preceding unsigned comment added by RoadView (talk • contribs)

Hey, guys, I'm here from the 3O board. I think you're (quite understandably) misusing WP:SIMILAR here, RoadView; SIMILAR is for topics that only have the two similarly-named topics and no disambiguation page. Health care (disambiguation) is a bluelink, with quite a few articles on it; if a hatnote is to be made, WP:HAT says that it should point to the dab page, not the Office episode. I'd say we should insert a hatnote pointing to the dab page (which has the Office episode page listed on it). Thanks! Writ Keeper ⚇♔ 06:43, 7 April 2012 (UTC)

You know, I somehow didn't even know Health care (disambiguation) existed. I'm certainly favor adding that hatnote over just the office episode because there are plenty of other disambig links. That leads me to believe that for the other pages, Conflict resolution and Sexual harassment, it would be acceptable for the hatnotes for just the office to be readded, since I don't believe there is currently a disambiguation page for those. If there is no further discussion I will add the hatnotes back in the near future. RoadView (talk) 09:15, 7 April 2012 (UTC)

I'm fine with linking to the Health care disambig page. As for the other two hatnotes, I would suggest we bring that conversation to the talk page for those two pages. I know other eds on those pages have complained about those hatnotes as well, and I myself still disagree but let's have that discussion over there. --Karl.brown (talk) 11:20, 7 April 2012 (UTC)

I'm fine with more discussion on the others, but there seems like a substantial shortage of input with these talk page discussions. Perhaps it can be taken elsewhere to generate a more thorough consensus. I'm sure this issue goes way beyond just the 2 office episode hatnotes. RoadView (talk) 10:22, 9 April 2012 (UTC)

If you want a broader discussion, I'd suggest starting here: Wikipedia_talk:Hatnote, that's where general policy discussions re: the use of hatnotes could be had. You could bring up the examples of 'Office' espisodes, and whether linking them at serious articles like 'Sexual harassment' that don't yet have disambig pages should have a direct link to a TV show as the first thing people read. My own opinion is no, but I'd say you should bring the question to the talk page and see what other editors think in general.--Karl (talk) 15:36, 9 April 2012 (UTC)

Comments made after discussion was moved to this page

In this case there's a disambiguation page at Health care (disambiguation), and the hatnote can point there. It's more of a problem where there isn't a dab page... and in fact this one looks a bit marginal, though I'm not going to challenge its existence. But in general, if there's no dab page, and there is something at a disambiguated title, then the hatnote needs to be there, even if does seem inappropriate and trivial in relation to the main topic. I remember there was some article relating to a key topic for a religious group where there was a discussion like this... I think someone helpfully found another item to justify creating a dab page to move the offensive (to some) frivolity off the base name page and hide it within a dab page link. But in some cases, it just has to stay in the hatnote. PamD 13:32, 23 April 2012 (UTC)

Thanks. There are other edits in question around Sexual harassment and Conflict resolution, and whether a hatnote linking to Sexual Harassment (The Office) is appropriate there. If you scroll up on this talk page, I can see this general subject has been discussed before, but its not clear if a consensus has been reached. Also, if consensus isn't there, I would put in my !vote to say that these should be decided on a case-by-case basis, and that there should be no guarantee of maintaining a hatnote if it cannot be demonstrated that there is a risk of topic confusion. For example, if you want to find the office tv show, searching for "Sexual harassment the office" or "sexual harassment tv" will get you what you want. My fear is if we allow this more generally, there are hundreds of TV episodes and bands with nouns as their title, and hundreds of articles will then have trivial hatnotes in prime real estate. I also agree, this is only a problem where there isn't a DAB page.--KarlB (talk) 13:40, 23 April 2012 (UTC)

My position again is that if there is no disambiguation page, then according to WP:SIMILAR it must be included if the article has the exact same name irrespective of how trivial or frequently it may be used. If there is more than 1 additional article with the same name then the aforementioned dab hatnote would be added instead. The clutter would be minimal and if you have no use for a particular hatnote, then just scroll down a little. The criteria for inclusion of a hatnote shouldn't be a popularity contest in my opinion, I think they should be there in case someone looking for something with the same title won't hit a dead end. RoadView (talk) 14:04, 23 April 2012 (UTC)

recent guidance posted/reverted

There was some recent guidance posted and then reverted. While I agree that consensus has not been reached on that point, I do in general agree with the approach. I hope the editors in question will join this discussion and see if we can't come up with something reasonable. --KarlB (talk) 03:04, 1 May 2012 (UTC)

The recent guidance that I removed directly contradicted the established guidance at WP:TWODABS. A disambiguation page that disambiguates two or fewer extant Wikipedia pages and whose title ends in "(disambiguation)" (i.e., there is a primary topic) is one of the accepted criteria for speedy deletion under WP:CSD G6 (see Template:Db-disambig). It's one thing to enourage editors to try to create a valid disambiguation page where applicable. But this guidance shouldn't encourage the creation of invalid disambiguation pages on the spurious assumption that additional disambiguation terms can likely be found -- the guidance should instruct to first find the additional terms and then create the disambiguation page. The notion that articles about serious topics should not have hatnotes pointing to articles with little comparable notability is dubious. What precisely are serious topics? How is it determined that an article is a "serious topic". And what is the basis for deciding other articles are unworthy of being mentioned in a hatnote? That smacks of censorship. older ≠ wiser 11:42, 1 May 2012 (UTC)

rather than 'serious topics', another way to think of this is, what is the purpose of a disambiguation page? When someone searches for 'sexual harassment', in how many cases are they actually looking for the episode of 'The Office'? The hatnotes should be there when it can be justified that someone ending up on that page was headed somewhere else. In the case of a TV show, typing "sexual harassment tv show" will get you exactly what you want, without forcing everyone who reads up on sexual harassment to consider whether they really want to read about a TV show. And please don't say things like 'smacks of censorship' - this has nothing to do with censorship. Finally, perhaps the guidance at TWODABS could be reconsidered in light of the fact that hundreds of TV episodes have titles like 'the keyboard' or 'the television' etc, and cluttering the prime real-estate at the top of these articles with these disambigs is a waste of real estate. --KarlB (talk) 14:01, 1 May 2012 (UTC)

The hatnotes should be there when it can be justified that someone ending up on that page was headed somewhere else - which can always be justified, because both articles share the same title. Every reader of the second article is likely to arrive to the primary one first. Both disambiguation pages and hatnotes are there to notify that a page with a similar name exists at Wikipedia, and should be used whenever that's the case. RoadView is right above that they are not, and shouldn't be, popularity contests. As long as the topic is notable enough to have a Wikipedia article, it merits a disambiguation page and/or tophat for people looking for it to find it.

Saying that the tophat is not needed because "you can always search for the page instead" is absurd, as links and search are navigation methods that can't replace each other - tophats are there precisely for people who doesn't know how to make an advanced, contextual search query. Those knowing how to search for "sexual harassment my favorite tv show opening season" wouldn't need disambiguation anyway; the links are for those with small computer literacy that will search for "term1 term2" and arrive at whatever page is titled with those terms.

Hiding the links to one article because of the contents of another is against WP:NOTCENSORED: Wikipedia doesn't care of what people will find offensive or bothersome when it conflicts with providing access to relevant information. Tophats are not part of the article were they're placed, they're part of the navigation structure. Diego (talk) 15:54, 1 May 2012 (UTC)

dynamic hatnotes

Apologies if this has been discussed before. What if we had dynamic hatnotes - i.e. hatnotes that only showed up *if* you arrived via a redirect. That way, you could always show the user exactly the right hatnote. If you didn't arrive via redirect, then no hatnotes would be shown (or only a link to the disambig in question. Think of how that might help an article like Century. --KarlB (talk) 13:49, 23 April 2012 (UTC)

• Hiding or removing complex hatnotes: I agree that when a person requests article "Century" then they do not need to see a long, complex hatnote listing all the redirected titles. However, in many cases, an overly complex hatnote is due to text-stuffing of too many concepts into a long article, where smaller separate articles should be used instead. For example, "Centenary" should go directly to "Centenary (disambiguation)" rather than drag through "Century". Also, "Centennial" should be a separate article, to explain that term in common usage, and also handle "Bicentennial" and "Sesquicentennial" together. -Wikid77 (talk) 04:31, 4 May 2012 (UTC)
• Intriguing idea, although it shouldn't hide all hatnotes; e.g. {{distinguish}} isn't for redirects (that'd be {{redirect-distinguish}}). --Cybercobra (talk) 04:35, 4 May 2012 (UTC)

  Yes of course; it would be a special template, just for 'dynamic' hatnotes. editors could always choose other hatnote templates which are always displayed. --KarlB (talk) 12:22, 4 May 2012 (UTC)

  Actually something like this this can be done fairly easily. Take a look at Centennial, set up for demo purposes, and let me know what you think. Rich Farmbrough, 12:50, 6 May 2012 (UTC).(Using some automation)
  

  It looks good; however I note you've used a transclude - the problem with this is, when the user clicks 'edit' at the top of the screen, they may be confused. In addition, it doesn't show the correct "main" article title. But yes, that's the result I was looking for. --KarlB (talk) 13:04, 6 May 2012 (UTC)

  Indeed, that's in fact a VERY bad way of doing it. By having the reader stay at the centennial article instead of actually redirecting them to the century one it means people won't be able to actually edit the century one unless they click on that. It's also somewhat mentally confusing. As an aside, I for one really don't like the hat note's wording at all. ♫ Melodia Chaconne ♫ (talk) 13:32, 6 May 2012 (UTC)

  Agree complete with Melodia that this would be a disastrously confusing way of doing this. I'd much rather see some way of collapsing the multiple hatnotes into a single line that said something like "Multiple terms redirect here. Expand to see other options." The display of the multiple hatnote would then toggle with an expand/collapse link. older ≠ wiser 14:44, 6 May 2012 (UTC)

  Indeed the example is very confusing (Centennial). Not only that, it is an edit of a #redirect into content. So, the #Redirect page name is required (many issues ahead). While also a major issue is: get many many editors to use hatnotes right. Glad older ≠ wiser =Bkonrad comes up with this simple & covering idea to expand the hatnote-line option. (Already getting technical: this should work through {{hatnote}} & its class, and simple lines (one link?) could be plane). -DePiep (talk) 16:08, 6 May 2012 (UTC))

All good points. This will, then, need either a MediaWiki fix or maybe Lua wil be able to solve it. Or.. hmmm let me think. Rich Farmbrough, 17:26, 6 May 2012 (UTC).(Using some automation)

P.S. I'll revert the changes out. Rich Farmbrough, 17:27, 6 May 2012 (UTC).(Using some automation)

thanks, it was worth a shot but I'm afraid that particular implementation won't work - it will likely require a change to media wiki. I've put this proposal into the village pump sandbox, but haven't gotten many comments. Should we refine it slightly and then propose it to village pump, and get behind it to see if we can get other editors supporting this idea in general? Any advice on next steps from the wiki-wisdom welcome here. --KarlB (talk) 23:40, 6 May 2012 (UTC)

Non-existent articles

I think section WP:Hat#non-existent article might be better renamed "Red links", or at the very least make some reference to red links in its text, since this is really what the section addresses. I've never made policy edits, so I don't totally know the protocol. Scientific29 (talk) 02:04, 27 May 2012 (UTC)

Hatnote carriage returns

See Template_talk:Other_uses#Hatnote_templates_and_carriage_returns where there is a suggestion that hatnotes should run on from one to the next without starting a new line. PamD 08:51, 2 May 2012 (UTC)

Discussion better be here. -DePiep (talk) 21:12, 3 May 2012 (UTC)

I don't mind moving it here. -Stevertigo (t | c) 22:50, 3 May 2012 (UTC)

I think it is about all hatnotes, not just that one. (Oh, btw, I might hav an opiniion ;-)) -DePiep (talk) 03:44, 4 May 2012 (UTC)

I agree; not starting on a new line is a good idea (or if they do start on a new line, less vertical space between them).--KarlB (talk) 05:03, 8 May 2012 (UTC)

Yes it does help, particularly when there are three or more hatnote lines. See Irrigation for example. But you can still end up with hatnote bloat, such as on Entropy and Allies. Regards, RJH (talk) 13:24, 8 May 2012 (UTC)

Actually the examples at Irrigation and Entropy are improper uses as explained in Wikipedia:RELATED, so they should be smaller. The one in Allies makes more sense because of the redirect targets (per Wikipedia:Hatnote#Ambiguous term that redirects to an unambiguously named article), and it also happens to be the less bloated of all. Diego (talk) 13:34, 8 May 2012 (UTC)

Yes, maybe we need a hatnote cleanup template. Regards, RJH (talk) 17:21, 10 May 2012 (UTC)

Disagree. Yes hatnotes can be better, but just stringing them does not solve it. e.g. the sentencing may not flow. And indeed, RJH, Hatnotes may need a cleanup. -DePiep (talk) 00:17, 14 May 2012 (UTC)-DePiep (talk) 00:17, 14 May 2012 (UTC)

Stringing them does solve a certain problem, and does not prevent editors from rewriting them for flow. I agree they should flow. And I disagree that hatnotes are always easier to read when on separate lines - perhaps so if they had less whitespace between them, but as they are a certain number look quite clumsy. -Stevertigo (t | c) 05:51, 17 May 2012 (UTC)

How would you suggest cleaning up the hatnote blizzard on the Weather article? Regards, RJH (talk) 04:41, 27 May 2012 (UTC)

ugh. I made an attempt. I think it's better, but could get better still.--KarlB (talk) 05:06, 27 May 2012 (UTC)

Thanks. There was a proposal on the village pump for dynamic hatnotes, which could be used to conceal redirects until they are needed. I think it is do-able, but we'd need some way to collect information about how the reader arrived at the article. Hmm, I thought just occurred to me that the redirect can include a section identifier. Perhaps there is a way to use that by configuring a special section ID (such as "#hatnote")? Regards, RJH (talk) 04:30, 28 May 2012 (UTC)

Dynamic hatnotes are not the way to go IMHO. We need to avoid the tendency to let technology make our decisions for us. And I've fixed the hatnote issue at Weather - my solution was to remove the "Cold weather" hatnote from Weather, and let Cold weather just redirect to Cold Weather (film), but with a hatnote on the film article that points back to Weather. Regards, -Stevertigo (t | c) 06:09, 11 June 2012 (UTC)

Umm, well we let technology determine our actions every time we stop at a stop light. I'm unclear how this is a significant issue in this instance. Regards, RJH (talk) 20:41, 11 June 2012 (UTC)

Stop lights act in a purely logical way, giving each lane of traffic their own turn at passing through. This is a different concept from trying to give people what they want, whether they know it or not. -Stevertigo (t | c) 08:15, 14 June 2012 (UTC)

That's completely backwards. Either Cold Weather should be the main article or it should redirect to Weather. Per disambig rules (and general logic) the current way is wrong. ♫ Melodia Chaconne ♫ (talk) 13:30, 14 June 2012 (UTC)

Another fine mess

• Disease has a sickeningly long stack of hatnotes.
• Wood is somewhat bloated, although not on the same scale.

Regards, RJH (talk) 19:38, 13 June 2012 (UTC)

Tightened "Disease" hatnotes here and "Wood" here. -- JHunterJ (talk) 20:28, 13 June 2012 (UTC)

Thank you. Regards, RJH (talk) 21:26, 13 June 2012 (UTC)

I fixed up the disease hatnotes. Rather than having "human disease" redirect to "disease" and have a trivial hatnote on "disease" liking to the Slayer album "Soundtrack to the Apocalypse" we redirected "human disease" to "Soundtrack to the Apocalypse" and put a serious hatnote linking to "disease" there. Im also going ahead and converting all medical links to "human disease" to simply "disease." -Stevertigo (t | c) 08:05, 14 June 2012 (UTC)

Ive also compressed the two hatnotes on wood to just one hatnote. -Stevertigo (t | c) 08:10, 14 June 2012 (UTC)

I uncompressed the redirect/other uses combos; Leading with "X redirects here" implies that the hatnote is about those redirections, which is not useful to readers not reaching the page through a hatnote. I've also updated the Human Disease (caps) to go to Soundtrack to the Apocalypse. -- JHunterJ (talk) 12:04, 14 June 2012 (UTC)

Wouldn't it be just as easy to say, "For other uses of Disease, see Disease (disambiguation)", then put it all on one line? Regards, RJH (talk) 15:30, 14 June 2012 (UTC)

I'm trying to put myself in the shoes of the reader who has reached Disease but meant to be somewhere else. If they searched on "Disease", the other uses line "pops". If they searched on one of the redirects, the "redirects here" line pops. Putting them together on one line has the IMO negative side effect of "burying the lede" -- they don't read to the end of the hatnote because the beginning of the hatnote lead them to believe it wasn't relevant. Yes, a minimal effort to read to the end of the line, but balanced against the even more minimal impact of a newline, it seems better. -- JHunterJ (talk) 15:51, 14 June 2012 (UTC)

To me it's sensible to lean heavily in favor of the article reader, which in most cases is probably 95-99% of the viewers. For them, the hatnote is an unnecessary distraction and should be kept to an absolute minimum. If we can get rid of a line we should do so, even if it takes the hatnote viewer an extra second to parse the material. Differing philosophies I guess. Regards, RJH (talk) 18:37, 14 June 2012 (UTC)

Right, the article body is there for the article readers (which should always make up the bulk of the readers getting to any article). The hatnotes are not part of the article and are there for everybody else, no matter how few, and should be kept as short as possible while still serving them. The "absolute minimum" would be a bare line of blue links. -- JHunterJ (talk) 19:06, 14 June 2012 (UTC)

Yeah, kind of like:

Perhaps you wanted: Heartwood (disambiguation), Lumber, Wood (disambiguation), Wooden (disambiguation), or Woodland?

Regards, RJH (talk) 19:48, 14 June 2012 (UTC)

Precisely! -- JHunterJ (talk) 19:56, 14 June 2012 (UTC)

I like how human disease works now. I disagree with JHunterJ: "Leading with "X redirects here" implies that the hatnote is about those redirections, which is not useful to readers not reaching the page through a hatnote." - A compressed hatnote, combining standard disambiguation links with 'redirects here' links, can read quite nicely if done correctly, and is not confusing as you suggest. -Stevertigo (t | c) 21:11, 14 June 2012 (UTC)

And I disagree that the two lines are not compressed enough, as you suggest. As I said, the impact either way is minimal; if read, it wouldn't be confusing, but if left separate, it would be better. -- JHunterJ (talk) 21:48, 14 June 2012 (UTC)

Yep, it looks good to me. Regards, RJH (talk) 00:03, 15 June 2012 (UTC)

Poetess

I think the hatnote at Poet needs fixing. I created Poetess (disambiguation) but someone seems to want to keep the existing hatnote. -Stevertigo (t | c) 00:30, 19 June 2012 (UTC)

The previous hatnote better serves the readers, since The Poetess is the only ambiguous topic. I'll look into creating a redirect version of the "consider disambiguation" hatnote this weekend, time permitting. -- JHunterJ (talk) 11:25, 22 June 2012 (UTC)

Trivial hatnote links

Thanks to Bkonrad for urging discussion of this proposed addition. [2] What follows is a proposed addition to the guideline page:

===Trivial links ===

A previous version of the comparative religion article showed:

For the Community episode, see Community (season 1)#ep12

Comparative religion is a field of religious studies that analyzes the similarities and differences of themes, myths, rituals and concepts among the world's religions.

"The subject matter of the disambiguation hatnote is trivial, while the subject matter of the article itself is not. Such trivial link hatnotes should be changed to a standard otheruses or similar disambiguation link, even if there are only two terms to disambiguate, as additional disambiguation terms can likely be found. In short, articles about serious topics should not have hatnotes pointing to articles with little comparable notability."

We discussed this a bit on Jimbo's talk page [3] - the example there was the Johnny Cash article, which had a trivial link to Johnny Cash (song). There is no comparable notability between the person and the song which bears his name. Such links are relatively less notable than the article subject, therefore they should be called "trivial." In such a case, a two-term disambiguation page is preferable to leaving the trivial link as a hatnote, and I've found in many cases once a two-term disambiguation is created, finding a third or fourth term is often not difficult. Regards, -Stevertigo (t | c) 07:16, 2 May 2012 (UTC)

• Oppose: there is no concept of "comparative notability" in Wikipedia, and your "trivia" is the topic of someone else's search. Hmm: let's declare all sportspeople to be trivial in comparison to any historical characters; all places of population less than x in comparison to places in country y; all popular musicians in comparison to pre-20th-century ones; all videogames... ; endless scope for tedious POV debates. No thanks. On the other hand, the concept of dynamic links for redirects, which someone suggested in one of the several places this discussion has started, does seem useful: if the "X redirects here, for Y see Z" only appears for people who've arrived via X, now that would be really useful and would clear some clutter from some pages. PamD 08:49, 2 May 2012 (UTC)

  comment I proposed the idea of a dynamic hatnote above (look up) - please place any further comments/ideas there. thanks! --KarlB (talk) 17:28, 2 May 2012 (UTC)

• Comment: I'd suggest Stevertigo to learn to see the hatnote as part of Wikipedia's navigation structure, instead of part of the article. In this light, the supposed undue weight that the hatnote gives to the minor topic (with respect to the primary one) disappears. Diego (talk) 09:23, 2 May 2012 (UTC)
• Oppose as currently formulated. See my comments (and others) above. I agree with PamD that the notion of "comparative notability" is without foundation. While I more often than not agree with Jimbo, in this case I think he is wrong. At present a disambiguation page for term with a primary topic (the disambiguation page ends in "(disambiguation)" and that contains two or fewer navigable blue links to wikipedia articles is an invalid disambiguation page and can be speedily deleted. The guidance should not encourage the creation of invalid dab pages. I've no problem with encouraging editors to look for additional entries to create a valid disambiguation page or even to create some stubs and redirects if appropriate. older ≠ wiser 11:03, 2 May 2012 (UTC)
• Oppose as refuted by pageview data: The less-famous hatnote links are not a significant problem to censor them as only "~(disambiguation)". The actual pageview counts for articles linked by hatnotes show there is little "unfair" promoted viewing of hatnote-linked articles. As discussed last week at User_talk:Jimbo, the pageviews of song article "Johnny Cash (song)" were not coming due to imagined "unfair" hatnote promotion in the singer's article, where April pageviews, after April 21 (2 days after delinking the song), remained at 27.4 per day (78%), extremely high, compared to March's average of 35 per day. The imagined fears of that famous song being commonly viewed, only due to mention in the singer article, were soundly refuted by the April pageview data (song-stats-201204). For article "Johnny Cash" the pageview stats (singer-stats-201204) showed that only 1-in-1500 people were clicking the song-hatnote link. It is difficult to prove that people even notice the hatnote, when they view an article to check the singer's famous songs, birthdate, hometown, or career events. In fact, for the new dab page, "Johnny Cash (disambiguation)" the pageview stats (disambig-stats-201204) showed only 1 extra person viewing that page on 30 April, when pageviews of the singer soared by 9,098 that day, from 14,898 to 23,996 views on 30 April. That data confirmed, only 1-in-9098 extra readers (at most) clicked on the hatnote link. The hatnote links are just not a problem, and cross-links between 2 articles should continue to be made by hatnote links, in 2 articles with same-name titles, without creating a 3rd, disambiguation page to list just 2 articles. -Wikid77 (talk) 13:51, 2 May, revised 04:31, 4 May 2012 (UTC)
• support it's not just about page views - what you're not capturing is the reaction of editors and users by seeing those hatnotes. The assertion above that one should "learn to see the hatnote as part of Wikipedia's navigation structure" illustrates a misconception of interface design; i.e. that you know what the users want and are looking for, and if they don't like it they just need to deal with it. I'm quite confident that if you took a random poll of readers of the Sexual harassment article, and compared the version with the hatnote linking to Sexual_Harassment_(Beavis_and_Butt-head_episode) and the one without the hatnote (or a disambig hatnote), the majority of those users would prefer to *not* read about Beavis as the first thing that hits their eye. Clicks are not the only measure of how bothersome these trivial hatnotes can be. I'm confident, but not positive - so the question is, would you be willing to set up a random sample to test actual users, to see if they prefer a disambig vs. a link out on Sexual harassment or Comparative religion etc, and if the users prefer to *not* see the hatnote (or a disambig page), would you change your minds? I'm with Jimbo and Stevertigo on this one; WP:IAR and create a 2-page DAB in these cases; page views could be used as a proxy for comparative notability (e.g. if something has 10x or 50x fewer page views, it doesn't get prime real estate)--KarlB (talk) 14:30, 2 May 2012 (UTC)

A problem with using page views as any sort of proxy for "triviality" is that something may well be very popular but still considered "trivial". And if the link is a redirect, things get very complicated. In the "comparative religion" example above, the TV episode is a link to a page with over 23k hits this month while Comparative religion had [http://stats-classic.grok.se/en/201204/Comparative%20religion just under 7k. PamD 15:42, 2 May 2012 (UTC)

Triviality is perhaps the wrong word. Also not sure if it's complicated: [[4]] shows how many hits that particular redirect had (I think?) - in other words, very few. I agree its not a perfect proxy, but at least its something objective we could use. There could be a rule like: if the hatnote'd term gets 50x fewer hits over a 2 month period, then a 2-page DAB can be created. --KarlB (talk) 15:54, 2 May 2012 (UTC)

Ah, that's clearer - as illustrated above it was a direct link to the tv show page, rather than a redirect which could be counted. But if your figure is 50x, note 44787 hits for Sexual harassment and 1410 for Sexual Harassment (The Office), a ratio of less than 32:1. I've got some sympathy for an addition to WP:TWODABS on the lines of:

A dab page may be created and retained when there is an article at the base name and only one disambiguated page or redirect if the base name page has been viewed at least 50 times as often (as measured by [something specified here]) as the other over the last 12 months (or since the newer page was created if less), if any editor believes this would improve the encyclopedia.

Or make it 100? Either way, it wouldn't be implemented very often, would create a small number of simple 2-entry dab pages (which may well get expanded later), would cause very few readers to have to click an extra time, and would save the recurrent discussions about hatnotes such as that at Northern Ireland (ratio 147:1 for the country v. the Euroconstituency). It's optional, so where a hatnote doesn't upset anyone it can be left as is. In the interests of peace, at the cost of a mild inconsistency which no reader will notice, it might be worthwhile. PamD 17:11, 2 May 2012 (UTC)

Sure that seems reasonable. I threw out the number 50 just as an example; what we should do is gather some of the more egregious examples, and see how they fare in terms of hits, and then come to an agreement on the ratio. As for sexual harassment, there is an office episode and a beavis episode, so that crosses the threshold. In fact, we could frame the whole guideline in a different way - to wit, you can only have a hatnote at the top of the page linking to something with a similar name if a) no editor opposes it and b) it can be demonstrated that it gets >= 1/50 the hits (or whatever # we end up with). Otherwise, it always goes to disambig. --KarlB (talk) 17:32, 2 May 2012 (UTC)

No, I'd oppose your reframing. The current situation works fine, except for the odd exceptions. Leave hatnotes as the default for cases with only one or two disambiguated titles. PamD 17:39, 2 May 2012 (UTC)

• Oppose Codifying that "trivial" links can be excluded, or pushed further from the user by forcing them to an unnecessary disambiguation page, smacks of elitism. I've already seen this problem with the Jennifer Lopez page. I have repeatedly tried to add a hatnote to Jennifer Lopez (meteorologist) but it keeps getting removed. It seems Ms Lopez the singer and actress is, in the view of certain article WP:OWNERS, quite clearly the only article anyone searching for, or linking to, Jennifer Lopez would want. This becomes a self-fulfilling prophecy. 1) Article X (disambig) is less important therefore it doesn't need to be linked from Article X. 2) The page views for Article X (disambig) are low proving that people are not interested in going to it. Tassedethe (talk) 16:44, 2 May 2012 (UTC)

  comment Ok i just looked at this - JLo gets about 200,000 hits per month, JLo (meteorologist) gets about 2000 hits/month, so a 100x difference. Thus, per the rule proposed above, you would place a disambig hatnote at the top of the page, which links to a page containing a link to JLo(meteorologist). You are thus inconveniencing 1 user (out of 100) so that 99 of them don't have to deal with seeing a JLo they don't care about. We're talking about a single extra click here, for a marginal use case. Is anyone willing to actually test this and get real user feedback? Would real user feedback change your mind? I see quotes like this: "Go ahead and search for Johnny Cash. Pretend you are looking for the song as a reader who doesn't know how naming works. Did you find the song you were looking for information about? Isn't this the entire point of having these hatnotes?" that make it seem like our users are hopeless and can't possibly think to type Johnny Cash song in the search box, when I'm quite sure most of our users are familiar with Google, which is exactly what they'd do there as well. When you are designing an interface you don't design it for the edge edge edge case, the user who has never used the interweb before, trembling as they type 'johnny cash' into the search bar, and then leaves wikipedia in a state of deep confusion because it doesn't have the song they wanted. The much more common case, which can easily be demonstrated, is that a user will either click through the disambig hatnote (as illustrated by the numbers above), or just refine their search. Again, let's not have these discussions about what users want in the absense of data - lets find a way to collect data and get real user perspectives on this. --KarlB (talk) 17:01, 2 May 2012 (UTC)

• Actually, given that Wikipedia has universal scope, those uneducated users are precisely what should be designed for. Tophats are an elegant and unobtrusive solution to the problem of multiple articles with the same name; in that it doesn't assumes computer literacy on the reader's part, which is much more common than you think if you address a worldwide audience and not just the elite educated classes. The experiment should be designed to include a population of Wikipedia users selected from all countries, methods of access and social backgrounds to be valid. Diego (talk) 17:57, 2 May 2012 (UTC)

It sounds like you're agreeing to an experiment, if well designed. Does that mean you are willing to change your mind about the "no-2-DAB" rule if users prefer it? --KarlB (talk) 18:23, 2 May 2012 (UTC)

If the experiment shows that there's no harm for those users looking for the secondary page, and it really finds a strong negative reaction in a significant percentage of users of the main topic, then sure, why not. But I don't think those effects will be so strong to merit changing the current well-thought policy. How are you planning to run the experiment once designed, anyway? Diego (talk) 19:17, 2 May 2012 (UTC)

• I still don't understand what problem this proposal is supposed to solve. What's so wrong about a link that basically says "Wikipedia also contains this other article with the same title?" Why should anyone take offense at that? Diego (talk) 17:46, 2 May 2012 (UTC)
• Oppose In the above example of the Community episode, I however would be opposed to adding a single hatnote for something that just takes you to an episode on an episode list. But moreover, I regard the number of pageviews as totally irrelevant for hatnote criteria inclusion. If it has the same name, then I want to see it. If the consensus is after 2 or 3 hatnotes we put in a dab hatnote, then I want to see it. I much more favor the versatility of having hatnotes than trying to polish every article by hopefully being able to not have any hatnotes on it. The hatnote is entirely inconsequential for me. I come across them, as everyone does, frequently, and I take a few seconds to see if they can guide me to something else with the exact same title or not, then I move on to reading the article. The hatnote is the least intrusive concept I can think of to satisfy all parties. The more popular article is shown by default upon a search and just in case someone with little knowledge of WP, or someone who didn't even know that a different identically titled article existed, will all be satisfied with the hatnote. The pageview ratio doesn't matter for me, especially when hatnotes probably go unnoticed many times for people who didn't need them as they were taken to the desired page initially. Hatnotes I think are sometimes at their best when they are rarely used because if they weren't there, then perhaps the other article would never be seen as there would be no way for someone to know they there is another article with the same name. WP:SIMILAR seems to dictate this and the subjective notability or more objective pageviews are not a factor. RoadView (talk) 17:59, 2 May 2012 (UTC)

  Frankly as a user I find it distracting; you can see from other comments and edits that other editors/users also find it distracting (even Jimbo doesn't seem to be a fan). I think a neutral disambig page is less distracting, because it follows a common format (vs having to parse a unique hatnote). But the main thing that annoys is that non-neutral, non-wikipedia notion of comparative neutrality/trivality; wikipedia is full of articles of dubious notability, but its also full of serious, well researched topics; and when you hit one of those topics and the first thing your eye catches is an espisode of an American television program, it irks. In any case, I reiterate again, how many editors here would be willing to change their minds based on feedback from real non-pro users of wikipedia, instead of just expressing what we like and don't like ourselves? --KarlB (talk) 18:18, 2 May 2012 (UTC)

I would rather have there be an option to disable hatnotes rather than remove most of the rarely used ones by default. As for the 2 dab aspect, is there now a consensus that we can use hatnotes for identically titled articles and we are now focusing on discussing if we need 2 or 3 hatnotes before we have a dab page? RoadView (talk) 18:38, 2 May 2012 (UTC)

Correct, I believe that was always the case. It seems this all comes down to a specific edge case: high profile article + low profile article. Some are saying 'make a hatnote', while others are saying 'make a DAB page'. The criteria for making a DAB page was laid out by Pam above, which I agree with. I'll copy it here again, so people can edit more. --KarlB (talk) 19:08, 2 May 2012 (UTC)

• Comment – I strongly agree with the concept, but the guideline is too open to interpretation. In the case of trivial hatnote links, I usually just try my best to create a disambiguation page because that minimizes the amount of distracting drivel. (Example: Cygnus X-1 (disambiguation).) Regards, RJH (talk) 19:35, 3 May 2012 (UTC)

Revised proposal (modifying 2-DAB rule)

A dab page may be created and retained when there is an article at the base name and only one disambiguated page or redirect if the base name page has been viewed at least XX times more frequently (as measured by this tool:last 90 days page view) as the other over the last 12 months (or since the newer page was created if less), if any editor believes this would improve the encyclopedia.

Feel free to directly edit the above / propose changes. XX is still TBD. --KarlB (talk) 19:08, 2 May 2012 (UTC)

Guidelines don't usually provide hard numbers as thresholds, but explain the intended effect instead; this way the is left at the editorial discretion of the people discussing each particular case. I'm also against using Wikipedia pageviews as a criterion; those are unreliable for many reasons, as they could reflect a temporary increase of usage or popularity within only a small subset of Wikipedia readers. If this change to guidelines is approved I would propose something along this lines to be added at WP:SIMILAR and/or WP:TWODABS (which already explains the situation of two articles, one of them primary):

In cases where the primary topic is much more likely to be the one sought for (at least between one and two orders of magnitude), a disambiguation page with only two entries may be created and linked from the article located at the base name. In cases where there is no primary topic, a disambiguation page is needed even if there are only two articles.

This avoids setting a one-size-fits-all numeric limit that could be inadequate at certain cases. Diego (talk) 19:38, 2 May 2012 (UTC)

You say you're against pageviews, but pageviews is one of the primary metrics way of understanding popularity on the web. What metrics (that we have access to) would you suggest otherwise? Also you are correct that there might be a temporary increase in popularity, but we are talking about an average over 90 days or more, so I think this is unlikely to cause an issue. I'm not going to press for a hard number, but I can guarantee if we don't establish one, there will be debates "X is 50x less viewed than Y" "well, 50x is not 2 orders of magnitude, thus it can remain" - just having a simple number, even if it isn't perfect, will remove all possibility of argument and let editors get on with their lives. Remember, the difference is quite minor - it's one click vs two clicks. Also recall that many many many wikipedia articles have disambig pages, and we don't think that 2 clicks in that case is imposing on the user. Why is everyone so anxious to protect the Office episodes of the world? --KarlB (talk) 18:34, 3 May 2012 (UTC)

Reconsidering, I'd prefer to add "exceptionally" as in "A dab page may exceptionally be created....". This is a proposal for extreme cases only. PamD 19:46, 2 May 2012 (UTC)

Here is a revised version:

In cases where the primary topic is much more likely to be the one sought for (i.e. has been viewed at least 25 times more frequently as measured by this tool:last 90 days page view), a disambiguation page may exceptionally be created with only two entries and linked from the article located at the base name. In cases where there is no primary topic, a disambiguation page is needed even if there are only two articles.

Pam does that capture the language you wanted? --KarlB (talk) 18:37, 3 May 2012 (UTC)

No, I think if we're going down this route at all it has to be at least 50x, and over 12 months (or since creation of the more recent article), to avoid recentism and make it difficult to rig by artificially boosting pageviews to make some point or other. I think we also need something to stress that it "may" but need not be created. I quite like my original "if an editor believes it will improve the encyclopedia": the default would be to use a direct hatnote, even with a huge pageview ratio (it saves creating a dab page, and saves clicks), but if someone objects then let them make a dab page if the stats are right. (Though I suppose it leaves things open for someone who feels the entire encyclopedia would be improved if all hatnotes only ever led to dab pages...). PamD 22:17, 3 May 2012 (UTC)

50x may be a hard metric to meet. I think 25-30x is more reasonable. Remember, for *any* situation where there are already 3 other terms, we always created a disambig page and no one complains. The reason I put 90 days is it is easy to get the stats for that - for 12 months you will have to do each month individually and then add them up - kind of a pain. Ok let me try again: --KarlB (talk) 22:53, 3 May 2012 (UTC)

In cases where the primary topic is much more likely to be the one sought for, a disambiguation page may exceptionally be created with only two entries, and linked from the article located at the base name, if an editor believes it will improve the encyclopedia. To determine whether the primary topic is much more likely to be searched for, calculate the ratio of primary page views to secondary page views, over the past six months using this tool. If the ratio is greater than 25, the disambiguation page can be created. Note that 25 is an indicative ratio, and particular circumstances may warrant larger or smaller ratios to be used.

Being a hard metric to meet is good; if approved, this should be only for gruesome cases of inappropriate context for very well known pages against absolutely unseen articles, not for articles of comparable weight; any topic with an article at Wikipedia is notable, after all. Also linking in guidelines to a page that is "beta service and may disappear or change at any time" is not a good idea; the recommendation to use the tool has to go. Instead of comparing the relative pageviews of the two articles, I'd go with an absolute limit of pageviews for the target article; anything with more than 30 pageviews a day should be linked to, no matter what's in the origin article. Diego (talk) 06:06, 4 May 2012 (UTC)

Why 50x? Why not make it 2000x and completely make the issue useless? Supporting an artificially high bar isn't helping matters here, nor is it helping to find a compromise. Even if the ratio is 3x or 2x there may be a need to remove the trivial hatnote from the non-trivial article. -Stevertigo (t | c) 23:10, 4 May 2012 (UTC)

That's why setting a numerical hard limit is useless; at the end, the editors evaluating each case will need to reach a compromise about what makes sense for each article. Diego (talk) 08:10, 5 May 2012 (UTC)

I think we should set a nominal limit. This gives editors who want to be bold permission to do so without dispute. if other editors disagree, then a discussion can be had; it's only a guideline, and it doesn't *have* to be followed at all times (there are many many 2-link-dabs already). But agreeing on a metric will help future editors a lot. My suggestion is we gather some data - we find 10 instances of trivial hatnotes, and calculate the differentials. Then see what the data indicates, and set the metric based on the data. Here are some I know of (with 90 day page hits in parens) Atheist (92924) >> Atheist (band) (20742) (4x diff) --Sexual harassment (147874) >>Sexual Harassment (The Office) (4227) (35x diff) -- Comparative religion (23259) >> Comparative Religion (Community) (45) (516x). Can you find some others? It seems for example from these metrics that there would be a good argument to keep Atheist (band), which only has 4x less hits, but I'd argue to get rid of the others. --KarlB (talk) 11:48, 5 May 2012 (UTC)

Spam?

WikiD, the issue is not that hatnote links get clicked, or don't get clicked - the issue is that a link of little notability should not be the first thing one sees on an article of greater notability. User:Tassedethe's above comments support my premise - people like User:Tassedethe want to promote lesser-notable articles like Jennifer Lopez (meteorologist) on the more notable article Jennifer Lopez. This kind of hatnote should be regarded as a kind of spam, given the vast difference in notability. -Stevertigo (t | c) 00:37, 3 May 2012 (UTC)

No evidence of spamming: The issue of being a "kind of spam" assumes an actual spamming activity, but the evidence refutes that impact. The term "spam" is derived from "spam, spam and more spam" where there must be an abundance, of that text, to be considered spam, rather than having one "short hatnote". By counting the pageviews of the hatnote-linked articles, then the imagined impact can be refuted, because people rarely click the hatnote-links. Hence, there is no evidence that people even notice the hatnote, beyond a casual glance, and so the idea of being an "over-and-over spamming nuisance" is just not proven by the pageview data. People were viewing the article "Johnny Cash (song)" because it is a famous song, performed by famous singers, while only 22% of pageviews were clicked from the hatnote in the singer article "Johnny Cash". A hatnote linking to article "Jennifer Lopez (meteorologist)" is just a short note, not a case of flooding or spamming something into an article. I had followed the train of thought which imagined "unfair" promoting of a "trivial topic" but there is no data proving that most people even notice the hatnote, except when the wrong article is displayed, and people use the hatnote to link an alternative article. If the claim is "spam" then there must be proof of spamming, but the activity and the related pageviews refute claims that a single hatnote is spamming. -Wikid77 (talk) 04:13, 3 May 2012 (UTC)

It's nothing to do with "spam" or "promotion", just navigation. If we have an article on the meteorologist, then someone who types her name (or follows a carelessly-made link from another article) must be given a route to find the article. It may be by a hatnote on the primary usage title, it may be via a dab page: one click or two. The poor woman can't help having a famous namesake, and readers seeking her article must be offered a sensible navigation route. That's how Wikipedia works. PamD 06:56, 3 May 2012 (UTC)

I 100% agree with the above comment from PamD. Navigation and versatility are the focus with hatnotes, not popularity. Calling them spam has no serious validity and only shows a personal disinterest or vendetta for hatnotes in general or at the very minimum for whichever ones are unacceptable based on subjective popularity. If you were looking for the Jennifer Lopez meteorologist and were perhaps unaware of the other Jennifer Lopez or unaware of disambiguation then you would type in her name still be able to reach the desired article. Without the hatnote, it would be a dead end, and for what?; to preserve the top space as if it were untouchable just because the non meteorologist Jennifer Lopez is presumably far more popular in general? That's counterproductive in my opinion. RoadView (talk) 08:20, 3 May 2012 (UTC)

I also agree. There's no reason for reducing the visibility of certain articles or making them more difficult to find simply because they are unpopular. older ≠ wiser 11:38, 3 May 2012 (UTC)

Unfortunately, your statement above disagrees with wikipedia policy on primary articles. The fact that 'Jennifer Lopez' brings you directly to JLo's page is exactly 'reducing the visibility of certain articles or making them more difficult to find simply because they are unpopular' - thats what we do every day. The issue is not whether we make them more difficult to find, it's just a difference of 1 click or two clicks. Please don't make this out to be some sort of wiki-censorship. We regularly prioritize articles, and bring users directly to them, because we believe they are more popular or common. Then, we provide disambig pages for the less common. Remember guys, this whole things boils down to one thing: 1 click or 2 clicks for a much-less notable subject?--KarlB (talk) 13:35, 3 May 2012 (UTC)

Which brings us back to my unanswered question above: what is exactly the problem that you're trying to solve? Given that the tophat wouldn't be eliminated anyway, this whole idea boils down to one thing: link to "article X" vs link to "other uses". What is the supposed improvement that option B brings to the world, at the cost of forcing one more click, page loading and navigation step onto some users? Diego (talk) 14:18, 3 May 2012 (UTC)

I think we just have different perspectives. It annoys me. It annoys Steve. It annoys Jimbo. It annoys lots of other editors, as you can see from various talk page discussions. I'm willing to change my mind though, based on user feedback. How many eds above, who claim to be protecting users, are willing to do the same? Thus far only Diego and myself have signed up... --KarlB (talk) 14:59, 3 May 2012 (UTC)

Excuse me for being dense - then, the gist of the argument is that the link "see Jennifer Lopez (meteorologist)" annoys you, but "see other uses" does not? May I ask you what is the origin of the discomfort, if it can be put into words? Is it because it belittles Jennifer Lopez the singer, or what? If we design an experiment, I want to understand what is it that's going to be tested. Diego (talk) 16:47, 3 May 2012 (UTC)

Actually jennifer lopez bothers me less, since it's a person who has the same exact name. But something like "Conflict resolution" or "Sexual harassment" linking to TV episodes, or Johnny Cash linking to a Johnny Cash (song) does bother me. Something about the switch of context, and the assumption that I'm really in the wrong place when 99 times out of 100 I'm not. --KarlB (talk) 18:40, 3 May 2012 (UTC)

• I'm also uncertain why this is such a problem. I assume it's something related to undue weight. But if the goal is to have a dab page for everything to prevent 1 hatnote from being prominently displayed, then it seems like that will lead to making unnecessary dab pages or padding existing ones. As in Sexual Harassment, there is no dab page and there is Sexual Harassment (The Office) so what do we do? According to WP:SIMILAR it has the same name so the hatnote should be added. So what is your opinion? Do we make a dab page and add additional links to it just so we can have "For other uses, see Sexual Harassment (disambiguation)" instead of "For The Office Episode, see "Sexual Harassment (The Office)"? RoadView (talk) 18:23, 3 May 2012 (UTC)
• Exactly. Actually for sexual harassment there is a beavis/butthead espisode, so we already have enough for a disambig page. --KarlB (talk) 18:40, 3 May 2012 (UTC)

My issue with that is the Beavis and Butthead episode doesn't have it's own article. I support adding it to a dab page, but that's only if there was already a dab page. I'm not sure what everyone else thinks about this facet but I don't think I'd want a dab page for sexual harassment if it only had 1 link to the stand alone office episode article and 1 sentence about the beavis and butthead episode. If there was no separate office episode article, you would still be in favor of a dab page with 2 sentences that link to nothing? I really want at least 2 separate articles on a dab page in order to justify having one. RoadView (talk) 19:00, 3 May 2012 (UTC)

see the proposal above - the idea is modify the 'two article dab' rule. and in this case, the sexual harassment dab would include exactly that as you describe it. Also, it wouldn't link to nothing -beavis and butthead is already linked. --KarlB (talk) 19:24, 3 May 2012 (UTC)

• support I'd make the criteria that we don't hatnote cases in which one topic is named after the other. For example the name of the TV-series episode is clearly named after the concept. Similarly a record such as ethnicity (album) is named after the concept of ethnicity, the song Johnny Cash is named for the man etc. I'd make the guideline that those cases in which a topic is named after another should be linked in the see also section.·ʍaunus·snunɐw· 20:29, 3 May 2012 (UTC)
• Oppose for most of the reasons given above, but also for reasons of WP:ACCESS for users of screen reader software. The reason the links appear first on the page is so that someone who physically cannot glance at the page to see whether he's in the right place doesn't have to listen to a whole bunch of irrelevant stuff when he's on the wrong page. WhatamIdoing (talk) 22:08, 3 May 2012 (UTC)

  comment per your screenreader comment, can you explain how that argument works for disambig pages? Someone arrives with a screenreader on Jesus and sees this: This article is about Jesus of Nazareth. For other uses, see Jesus (disambiguation). Are you suggesting that that is somehow a problem for them? We're not proposing getting rid of hatnotes, we're just proposing making disambig pages in certain cases even if there aren't tons of additional links. --KarlB (talk) 22:34, 3 May 2012 (UTC)

• Yeah, I don't buy this "but I'll have to click my mouse twice instead of just once" argument either. Its not a strong argument. -Stevertigo (t | c) 22:38, 3 May 2012 (UTC)

Wikid77 above said: "As discussed last week at User_talk:Jimbo, the pageviews of song article "Johnny Cash (song)" were not coming due to imagined "unfair" hatnote promotion in the singer's article, where April pageviews remained at 27.4 per day (78%), extremely high, compared to March's average of 35 per day." Wikid77 you present a contradiction - you claim that April's numbers remain high, yet you also admit that our discussion at Talk:Jimbo happened just a week ago! Thus the numbers for April will incorrectly include two or three weeks when the hatnote was still at the top of the Johnny Cash article. Can you recompute your numbers? Regards, -Stevertigo (t | c) 22:46, 3 May 2012 (UTC)

• Honestly, just use {{about}} instead of {{for}} and provide a description for the primary page; it pushes the dab link further to the side and de-emphasizes it:
• This page is about the academic field. For the television episode, see List of Community episodes.
  --Cybercobra (talk) 04:49, 4 May 2012 (UTC)
  • The "about" template just adds redundant clutter, since the reader merely has to lower their eyes one line and read the introduction to the article. Regards, RJH (talk) 16:17, 5 May 2012 (UTC)

• End-of-April pageviews for song were 27.4/day: I was referring to the end-of-April pageviews of article "Johnny Cash (song)", during April 22-30 (song-stats-201204) remaining at average of 27.4 per day (78%), extremely high (after the hatnote was changed to delink the song), compared to March's average of 35 per day (1091/31~=35, song-stats-201203). I apologize for not specifying April 22-30 in the earlier thread above. The imagined fears of that famous song being commonly viewed, only due to mention in the singer article, were soundly refuted by the April pageviews, and now by the May pageviews (song-stats-201205). -Wikid77 04:49, 4 May 2012 (UTC)

I understand your point. There appears to be some dropoff, but not as much as one might think. It still remains the case that the hatnote link to the song does not belong on the article about the person. The difference in notability is sufficent cause to create a disambiguation page. Note also that if the disambiguation page was created before a third disambiguation term was found, the two-term disambiguation would have been a viable stub which would have gained more terms/entries as time went on. Thus creating a two-term disambiguation seems to be the proper move as far as wikification and article creation and development are concerned. This shouldnt be too controversial.-Stevertigo (t | c) 06:07, 4 May 2012 (UTC)

• Oppose interposing a disambiguation page where none is needed. Wikipedia distinguishes between notable and non-notable topics. If a topic is ambiguous with another article's topic, disambiguation is needed. If there are only two such topics, and one is primary, disambiguation can be handled through a hatnote. We do not need to inconvenience readers who are looking for topics that some editors find trivial to avoid embarrassing the editors who are looking for topics that some editors find serious. If the topics are truly trivial, and not notable, their articles should be deleted. -- JHunterJ (talk) 15:27, 5 May 2012 (UTC)
• Opppose I 100% agree with what JHunterJ said above me. ♫ Melodia Chaconne ♫ (talk) 15:29, 5 May 2012 (UTC)
• Oppose both proposals. As others have commented, this is a solution in search of a problem. I see no valid reason to inconvenience readers (even a minority) on the basis that some users (even Jimbo) find certain topics "trivial".
  KarlB has argued that this would merely be an extension of our current setup (in which primary topics receive the base titles, thereby pushing other uses from the forefront). I strongly disagree. The current practices are designed to convenience as many readers as possible. We send them directly to the primary topics' articles because those are what they most likely seek, not because the other articles embarrass us. The inconvenience experienced by users arriving at an unintended page is an unfortunate side effect that we seek to mitigate.
  I was very surprised by Stevertigo's assertion that hatnotes along the lines of a link to Jennifer Lopez (meteorologist) from the Jennifer Lopez article constitute "spam". I don't even know how to respond to such a statement. —David Levy 18:32, 5 May 2012 (UTC)
• comment For all of those who are opposed because this additional click will inconvenience the readers, how many of you would change your position if presented with evidence that the majority of users prefer to not see certain (what I would call "trivial") hatnotes? That is, is your belief fixed, or would you be willing to change your stance if real users said "i don't care" or "i'd rather not see a beavis/butthead episode at the top of "sexual harassment"? If your belief is fixed, then you cannot claim to speak for all users - and we should thus take your opinion as the opinion of a single user. If your belief is not fixed, then please help me in designing an experiment which will reveal real user's preferences, and sign your name that you will abide by the results of the experiment, even if the results disagree with your current beliefs. --KarlB (talk) 19:31, 5 May 2012

  Fixed, based on the Wikipedia guidelines and policies (including WP:D and WP:NOTCENSORED). Users in group A do not get to needlessly hinder users in group B just because group A vastly outnumbers group B. -- JHunterJ (talk) 20:19, 5 May 2012 (UTC)

  That wouldn't affect my position, which is based upon the accessibility of encyclopedic content.
  While not exactly the same issue, this is analogous to our refusal to censor material on the basis that some users find it unpleasant. It isn't about pleasing a majority by catering to their preferences; it's about disseminating knowledge. —David Levy 20:23, 5 May 2012 (UTC)

  Ok, then you haven't actually correctly stated your positions in my opinion. It appears as if this has nothing to do with user convenience overall, and much more to do with user convenience for a minority of users, is that correct? In other words, your approach is, design an interface that helps a small minority, even if it inconveniences a vast majority? That's a fascinating design philosophy. Note this has nothing to do with censorship, no-one is talking about removing info from the encyclopedia, and a disambig page is not considered 'hiding' or 'suppressing' either. --KarlB (talk) 20:31, 5 May 2012 (UTC)

  You are wrong. It has to do with all user convenience. No one is inconvenienced by the existing hatnote over the proposed replacement hatnote; some are inconvenienced by the proposed replacement hatnote over the existing hatnote. The disdain for supposed triviality is not inconvenience. It is really quite a boringly simple position; nothing fascinating about it at all. -- JHunterJ (talk) 20:39, 5 May 2012 (UTC)

  Sorry, I guess I'm not understanding your position. You said "No one is inconvenienced by the existing hatnote...the disdain for supposed triviality is not inconvenience." So you are granting that some users, perhaps even a majority, may have disdain for something, but it does not rise to the level of 'inconvenience' in your mind. On the other hand, asking a user to click to a disambig page *is* an inconvenience that cannot be tolerated? I'm just trying to understand b/c you're drawing a fine line between someone being annoyed, having disdain, not liking something, being irritated by something, and being inconvenienced by something. It seems like you're claiming that reading that first line does not inconvenience users, but clicking twice does? If a disambig page is such an inconvenience, why don't we get rid of them all together - we could just put lots of links at the top - after all, by your metric, we would only be causing disdain for readers, but we are inconveniencing those readers who wanted a different page. --KarlB (talk) 20:47, 5 May 2012 (UTC)

  I think you understand it correctly. Disdain is not inconvenience. A neon purple and orange sidewalk (or one with a picture of Beavis & Butthead) might be viewed with disdain, but is not more inconvenient than a beige sidewalk. Running the sidewalk into one building and around the back of the building before continuing on down the street is an inconvenience to those trying to get down the street, no matter how popular and stately the building is. Disambiguation pages are navigational tools. Used correctly, they are not inconveniences. Used incorrectly, such as in your suggestion or if we use a base-name disambiguation page for all ambiguous topics regardless of the existence of a primary topic, they can be inconveniences. -- JHunterJ (talk) 20:57, 5 May 2012 (UTC)

  (ec)I never said we should use a base-name disambig page btw. More importantly, on what do you base this distinction of 'inconvenience' from "everything else"? i.e., if we did a test of users, and they said they preferred one solution to another, would that change your mind? Or does it not matter what users think? What if some people *do* consider these hatnotes inconvenient, or bothersome, or annoying, or whatever other word you want to use? And what if those who have to click twice *do not* consider it inconvenient. Is inconvenience a state of mind? How is it defined? Is it defined by you? Can users have their own definitions of convenience/inconvenience?--KarlB (talk) 21:10, 5 May 2012 (UTC)

  Again, what users prefer isn't the relevant criterion. Perhaps a majority of readers would prefer that we omit views with which they disagree and concepts that they find disturbing, but that doesn't mean that we should comply. (To be clear, I'm not implying that the proposed change is comparable to that.) That some people prefer not to read certain things isn't a valid reason to hide them from view (even if only in the context of a hatnote).
  How could the need to follow an extra link before reaching one's desired article not be an inconvenience? —David Levy 21:44, 5 May 2012 (UTC)

  "[Putting] lots of links at the top" would consume a large amount of space. The proposed change would result in the replacement of one hatnote with another, thereby saving no one the trouble of "reading that first line" (which might or might not be ignored either way).
  No one has asserted that "a disambig page is an inconvenience that cannot be tolerated". We're saying that there's no valid reason to replace a perfectly functional hatnote with a less convenient alternative.
  And indeed, you're conflating the concepts of "inconvenience" and "displeasure", but really, that's a red herring. Use whatever terminology you like. That someone finds a particular topic "trivial" (even of you do want to refer to this experience as an "inconvenience") isn't a valid reason to increase others' difficulty in accessing its article. Our mission is to propagate knowledge, not to please people by catering to their sensibilities. —David Levy 21:06, 5 May 2012 (UTC)

• comment Also, allow me ask another question - is your opposition to the above absolute? I.e., if I gave you two pages with the same name, one which gets 2500 hits/day (say about a major political topic), and the other that gets < 1 hit/day (say, a mildly-offensive cartoon), then there is no situation in which you would support the creation of a disambig page? I just want to gauge how flexible and willing to reach consensus you are, or is your opposition absolute and complete? --KarlB (talk) 19:47, 5 May 2012 (UTC)

  Honestly I'm not quite sure what you're getting at, Karl, but consider the other way around -- what of people who ARE looking for that Beavis and Butthead episode? They probably know it's not going to be at sexual harassment, but they may not know where else to look for it...so, used to how WP does things they go to that page anyway to see where it is. Obviously yes a generic "for other uses" could be added, and the creation of a two entry disambig page, but should that be really nessesary? IMO it's just as harmful for WP to be full of such pages as having an "embarrassing" topic hatnote on a "serious" one. The Jennifer Lopez one it's even sillier -- sure the actress is far far more known, but she's also a pop-culture person compared to a "serious" job like meteorologist. It just seems to me that it's another one of those things that will lead to edit wars that really don't need to happen by virtue of sticking with what we already do rather than changing it. ♫ Melodia Chaconne ♫ (talk) 19:53, 5 May 2012 (UTC)

  For someone looking for the beavis and butthead espisode called "sexual harassment", there are two very easy solutions: 1) Search with "sexual harassment beavis" (which anyone who uses Google would intuitively understand) and 2) click twice. Why is an extra click such anathema? Hundreds of thousands of pages have disambig pages, they are no more bizarre or confusing to a user than a regular hatnote. And, the purpose of the proposal above is to *eliminate* edit wars - the edit wars are already happening, on pages like Jlo, Sexual harassment, etc - hatnotes added, removed, added, removed, etc. This policy would *remove* the need for such edit wars, by giving a bright line, of when you can have one, and when you need a disambig page. I'm still interested in your answer to my question however - are you willing to change your mind if real users have preferences different than yours? I am. --KarlB (talk) 20:01, 5 May 2012 (UTC)

  No one has asserted that the proposed change would prevent readers from accessing the articles that they seek. We're saying that the difficulty would be increased for a reason that we regard as invalid (a desire to hide "trivial" topics from view).
  As noted above, we don't indiscriminately honor the majority's preferences; we abide by principles intended to serve our readership as a whole in a manner consistent with our mission to propagate knowledge. —David Levy 20:23, 5 May 2012 (UTC)

  Exactly. That is the crux here. You are making a cost/benefit argument; your argument is that the current arrangement is optimal because it "serves our readership as a whole" (although I'm still very confused on disambig pages - if they are so bad, why not do away with them, or require them to have 10 links before they can be created?) The real question is, how are you measuring costs, and how are you measuring benefits? It seems you are measuring benefits in terms of # of clicks needed to get to the right article. How are you measuring costs? Or do you consider the hatnotes with tangential information costless to the bulk of the users? Clearly, you don't - that's why disambig pages were created in the first place - to ease cognitive confusion for readers when they enter a page. KarlB (talk) 21:03, 5 May 2012 (UTC)

  I just noticed the above message. I'm replying to note that I've addressed all of the questions elsewhere in the discussion. —David Levy 06:51, 7 May 2012 (UTC)

  Also, FWIW sexual harassment is no longer an issue, a disambig page could be created because there is an office episode and a beavis episode - but suppose the office episode didn't exist - so according to your preferences, every time anyone on the web types "sexual harassment" and ends up on the wikipedia page (which happens over 1500 times a day, and sexual harassment is the #1 link on google), the first thing they would see is this article is about intimidation of a sexual nature. For the Beavis and Butthead episode, click x. Are you really so concerned about that Beavis-searching user, who wandered into the wrong page, that you want to force everyone else reading the page to think about Beavis and Butthead, in order to save that 0.1% of visitors a single forlorn click?--KarlB (talk) 20:12, 5 May 2012 (UTC)

  Why is one less click such anathema? This is an encyclopedia, not a sexual harassment counseling center. Anyone wanting information on the primary topic can read the article they land on. Anyone looking for the notable Beavis & Butthead episode can click on the link in the hatnote. If you feel that WP needs to protect the readers from "trivial" article titles, see WP:NOTCENSORED: ""being objectionable" is generally not sufficient grounds for removal or inclusion of content.". -- JHunterJ (talk) 20:16, 5 May 2012 (UTC)

  Please stop with the censored arguments - in my opinion they don't hold much weight. That's not what is at play here. There are over 200,000 disambig pages. Are you saying those are all somehow a form of censorship? This is about user interface design, and not distracting the user with information they aren't likely to need at that moment. It has nothing to do with censorship. Perhaps you should read WP:NOTCENSORED: "However, some articles may include text, images, or links which some people may find objectionable, when these materials are relevant to the content. Discussion of potentially objectionable content should not focus on its offensiveness but on whether it is appropriate to include in a given article." That is the crux of the discussion. Is it appropriate to have distracting, content-rich descriptions of terms which happen to share a name but which are otherwise trivial in comparison (as measured by page hits, not some abstract standard), or is it better and more appropriate to have neutral disambig pages? And jHunterJ, you haven't answered my question above - would you change your mind if you heard from real users? --KarlB (talk) 20:24, 5 May 2012 (UTC)

  I am saying that the proposed removal of information and convenient linking from the hatnotes on some articles in order to protect the sensibilities of other readers (or more likely, editors) holds no weight at all. Page hits do not determine seriousness or triviality, but popularity and usage -- and is useful for that. -- JHunterJ (talk) 20:42, 5 May 2012 (UTC)

  All we are proposing is substituting a direct link with a disambiguation tag, just as all articles with three or more disambiguation terms already do. Are people "inconvenienced" by standard otheruses hatnotes with three terms? No, they are not. The argument that it is "more convenient" or else "aids navigation" to have a trivial hatnote at the top of a serious article does not hold water. Regards, -Stevertigo (t | c) 22:35, 5 May 2012 (UTC)

  All we are proposing is substituting a direct link with a disambiguation tag, just as all articles with three or more disambiguation terms already do.
  To borrow JHunterJ's analogy, one could make the same argument about eliminating the recognition of primary topics (and placing all disambiguation pages at base titles): "All we are proposing is substituting a direct landing with a disambiguation page, just as all titles with no primary topics already do."
  In either case, the encyclopedia would continue to function, albeit less efficiently. We oppose your proposal to make navigation less efficient for the sake of hiding subjects that you deem "trivial".
  Hatnotes exist to aid navigation. That something isn't bad doesn't justify replacing something better. (Obviously, we disagree on which approach is better.)
  Are people "inconvenienced" by standard otheruses hatnotes with three terms?
  Compared with the alternatives, such a construct increases convenience. The proposed change would do the opposite.
  The argument that it is "more convenient" or else "aids navigation" to have a trivial hatnote at the top of a serious article does not hold water.
  Setting aside your "trivial" and "serious" labels (which are inconsistent with Wikipedia's basic principles), please explain how following one link instead of two before reaching the desired page is not more convenient and doesn't make navigation easier. —David Levy 23:06, 5 May 2012 (UTC)

  Comment on the assertion "...just as all articles with three or more disambiguation terms already do." That's simply not true. There are many articles (i.e the primary topic) where the hatnote contains 2, and sometimes more, links to other articles. This is because editors have decided that is more convenient to give the people at the "wrong" page a quicker route to the article they want. (I'm looking at Frank Winters and Fred Robbins right now.) Obviously if an article has many hatnote links it's confusing and a disambiguation page is a neater option. But for 1 other link it's just making people jump through extra hoops for no obvious benefit. Tassedethe (talk) 04:19, 6 May 2012 (UTC)

  Tassedethe, what you say is true, but in those cases the extra hatnote terms are almost always of comparable notability. When such hatnotes contain links to "trivial" terms, or else terms of lesser notability, these terms are usually migrated to the disambiguation page. -Stevertigo (t | c) 04:29, 6 May 2012 (UTC)

  David, so are you among those who claim to be "inconvenienced" by a standard disambiguation page that requires you to click twice to get to a secondary article? Would you, based on this view, be inclined to make the radical proposal that we should move terms from standard three+ term disambiguation pages back to the article as hatnotes, just because it would be more "convenient"? -Stevertigo (t | c) 04:33, 6 May 2012 (UTC)

  I've already addressed these points. —David Levy 11:46, 6 May 2012 (UTC)

  My view is that there is a real difference in notablity between such articles as atheist and atheist (band), and all who have common sense can see this difference (even unto quantifying this difference by pageviews and perhaps other means (distance from core topics), etc.). Perhaps the opposing view would like to make a stand and undo my change to atheist and restore the hatnote linking to the band and to third link (a novel)? Since the opposing views appear to outnumber support views, perhaps the opposing people can make a stand for their case somewhere, and report this principled stand back here. Regards! -Stevertigo (t | c) 04:41, 6 May 2012 (UTC)

  No one denies that a significant difference in notability exists between the primary topic and the non-primary topic. (That's why the former is considered the primary topic.) We dispute that this distinction is grounds for declining to include a direct link to the non-primary topic's article.
  Wikipedia has two articles about persons named "Jennifer Lopez". One (an entertainer) is significantly better known than the other (a meteorologist), so her article occupies the base title (Jennifer Lopez). To accommodate readers seeking the meteorologist's article and accidentally arriving at the wrong page, we link to Jennifer Lopez (meteorologist) via a hatnote. This is the simplest and most efficient means of enabling the intended navigation. Your assertion that the link constitutes "spam" is quite perplexing. —David Levy 11:46, 6 May 2012 (UTC)

  PS: The whole point is the basis for what should be a hatnote cannot be decided by the arbitrary 3-term minimum for disambiguation pages. Other factors have bearing, as almost every case will show. -Stevertigo (t | c) 05:04, 6 May 2012 (UTC)

Note: In among all this, Atheism (which is the target of a redirect from Atheist) was left with just an "otheruses" hatnote, pointing to a newly-created dab page at Atheism (disambiguation) ( which was so unnecessary and malformed that I've proposed it for speedy deletion), and no indication to anyone searching for "Atheist" to show where that term was to be found. I have now reinstated the "redirect" hatnote, expanded the existing "The Atheist" disambiguation page, and moved it to Atheist (disambiguation). Please don't let your strong views on hatnotes break the linkages which readers need. If term A redirects to page B but has other uses, then we must have a hatnote at page B to say so, either listing the other one or two uses, or pointing to the dab page. Otherwise the reader who wants A is dumped on page B with no clear indication how to proceed. PamD 09:18, 6 May 2012 (UTC)

• Comment to KarlB - if you're serious about this being about the best user interface design, your approach is not the best one. You're 1) proposing a particular final interface design and 2) trying to design an experiment to prove that your design is better than the current one. The right approach to create a good interface would be to 1) gather requirements about what users will need and 2) design an interface that would cater to those needs as much as possible, weighting all with respect to their relative importance. The proposed design should follow the findings of the experiment, not the other way around.

For the sake of argument let's suppose you're right and a non-negligible percentage of readers are confused by the tophat being displayed at the beginning of the article. The functions that it serves are the following:

1. Notifying that the article name has associated more than one Wikipedia article.
2. In case of a tophat following WP:SIMILAR, notifying the number of articles existing (there are only two articles in such case, since there's no disambiguation page).
3. Providing a direct navigation link to the other article.

Numbers 1 and 2 above would benefit all readers of the page, not just those looking for the minor article. Now with these as requirements I can think of an interface that would provide all of the above requirements without getting in the way, for example this tabbed pane for the "sexual harassment" case:

_______________________________________________________________________________________________
| *Primary topic* | (The Office) | (Beavis and Butt-head episode) | Other uses (4 articles) |v|
|                 ----------------------------------------------------------------------------------------
|                                                                                                        |
|'''SEXUAL HARASSMENT'''                                                                                 |
|                                                                                                        |
| Sexual harassment is intimidation, bullying or coercion of a sexual nature...                          |

This interface shows a tab for each article needing Wikipedia title disambiguation; the "other uses" tab would link to the disambiguation page and list the total number of articles to be found there; a drop down |v| could be used to list a quick menu with direct access to the entries if there are few of them. Since all the disambiguation information is kept into the tabbed pane, readers interested only in the main article can completely ignore the navigation structure and won't be inconvenienced by the links in there. But for those interested in all the information contained within Wikipedia for that title, the tabs would be more informative than just a link to a separate DAB page.

See, I'm not suggesting that this interface is the best one; only that the one KarlB proposes is not really an improvement over the current situation, since the purported improvement for some readers is achieved at some negative cost for others, and by removing some useful information. Given that better solutions exist than the one suggested in this RfC, if you KarlB are serious about providing a better interface and not just about addressing your particular pet peeve, I think you should drop your support for this proposal and instead pursue an interface that keeps the link to the minor "trivial" article and does it with a less obtrusive presentation that you no longer find objectionable. Diego (talk) 16:37, 6 May 2012 (UTC)

It's an interesting idea, just it's not really my job to design a new user interface for wikipedia; I'm just trying to do what can be done within the bounds we have. If someone else wants to take that on, feel free. I'd prefer to work on a solution with what we have, then if someone reskins/redesigns how hatnotes work in general, good for them. I've proposed above a different change to make dynamic hatnotes, but that's not really a major UI change, it would be much more back-end (in some cases, fewer hatnotes would appear as a result). The tab interface you're talking about above is interesting, but designing/testing that is beyond the time I have now. We *can* set up an experiment on the other hand to see what users prefer within the bounds of current technology, but two of the major proponents of the current system don't seem to care what users think, and have professed that they will not change their positions even in the face of preference data from real users. So, it's hard to have a UI discussion with people who feel like that - I mean, if you took their arguments seriously, it would lead you to a dramatically different interface - because if adding additional clicks really is bad, then most disambig pages should disappear, and a small list of links should be added to the top of every page. I'm not sure how they can simultaneously defend disambig pages with 4 links, while arguing forcefully for the removal of disambig pages with 2 links - their arguments are not logically consistent.--KarlB (talk) 22:26, 6 May 2012 (UTC)

For example, David Levy said "That someone finds a particular topic "trivial" (even of you do want to refer to this experience as an "inconvenience") isn't a valid reason to increase others' difficulty in accessing its article." And yet, we do that, every day, when a search for Jesus redirects to Jesus, and not Jesus (disambiguation). Thus, we are forcing thousands of people, every day, to click twice in order to find their article about Jesús,_Peru and Zola_Jesus. Why are we willing to do this? If you apply the logic above, there is no valid reason to add additional clicks to a minority just because a majority doesn't want additional clicks. Or, perhaps there is a majoritarian logic at play here after all? Perhaps, wikipedia does agree that it's better to *not* inconvenience the majority, it is better to inconvenience the minority. So then the argument becomes, can you irritate and bother the majority, or can you inconvenience the minority - and how do you make the tradeoff? It becomes a silly semantic game - inconvenience matters, irritation doesn't, etc., and shows little understanding of user interface design. So, if you're making an interface argument, which I am (not a censorship argument), and you're arguing with people who don't agree with or understand basic tenets of user interface design, it's a bit like bashing your head against a wall...In any case, it doesn't really matter - every issue that has been brought up in this debate has been resolved by a disambig page, which has gone uncontested. So, the solution seems to be, just create disambig pages, find some good links to add, and forget about trying to change policy. --KarlB (talk) 22:36, 6 May 2012 (UTC)

If you want to make an interface argument, then please first establish that there is actually a problem that needs to be addressed other than that some editors seem to be irritated by seeing references to trivial topics on "serious" articles. And yet, we do that, every day, when a search for Jesus redirects to Jesus, and not Jesus (disambiguation). Thus, we are forcing thousands of people, every day, to click twice in order to find their article about Jesús,_Peru and Zola_Jesus. You are comparing apples to oranges and this distinction has been explained here before. As a general rule, we do not have disambiguation pages where there is a primary topic that disambiguate two or fewer topics (i.e., there is exactly one other ambiguous article title). Such invalid disambiguation pages are candidates for speedy deletion (also see Template:db-disambig). In such situations WP:TWODABS applies. Where there are multiple ambiguous terms with a primary topic, the hatnote links to a disambiguation page so as not to list every term in the hatnote. In some cases, there might be one or two very prominent uses that are included in the hatnote for convenience. So far, I don't see that you've demonstrated that there is any actual problem with the current arrangement (other than that some editors seem to find it irritating to see "serious" articles defaced with references to trivial topics. older ≠ wiser 23:04, 6 May 2012 (UTC)

Please explain, why not list every term in the hatnote? it's saving clicks, right? I'm sorry but your arguments about saving clicks are not consistent - if it's *always* better to save clicks, then you should almost *always* do so. However, if there are times when you trade off clicks for something else (like usability and clarity), then just admit it, and we can actually have a real interface discussion. And yes, the fact that thousands of viewers are irritated is an actual problem, whether you think it is or not. I've proposed that we create a test to find out for sure, but as you can see above, no-one except Diego has accepted that they might change their views based on feedback from users, so it's hard to make an interface argument if people don't really care about UI... (the first word of which is 'user')--KarlB (talk) 23:37, 6 May 2012 (UTC)

It has already been explained many times now. Please read carefully and try not to let preconceptions get in the way. And yes, the fact that thousands of viewers are irritated is an actual problem See, that is really the crux of the problem. You assume that there is a problem without presenting any actual evidence, based solely on your preferences and some anecdotes. older ≠ wiser 00:04, 7 May 2012 (UTC)

I've tried to read everything in this thread, but I honestly don't understand why it's acceptable to have a DAB page with 3 terms, but not with 2. I also don't understand the subtleties of the tradeoffs involved - why are a few links on top bad? Seriously, please explain it clearly - why not have 4 or 5 links on top of every page, and get rid of tons of DAB pages? If you *don't* think that's a good idea, explain why, using terms of 'convenience', 'irritation', and 'user interface'. Also I've proposed to gather actual data to see how many readers *are* actually irritated, and measure their preferences in terms of interface choices, but as I mentioned, only a single person opposed to this proposal has hinted that he might change his mind based on that data. If the data doesn't matter to those opposed, why should I bother to collect it? I can give you a list of 6 editors who are annoyed, so it's not imaginary - we're real.--KarlB (talk) 00:17, 7 May 2012 (UTC)

I'm not about to rehash everything that's already been explained when you've indicated an unwillingness or inability to understand. As for your proposed test -- I have yet to see a clear, unbiased, testable statement of what the problem is that you're trying to fix. Without such a statement, I'm not going to sign on to a half-baked survey that asks the hatnote equivalent of When did you stop beating your wife? older ≠ wiser 00:49, 7 May 2012 (UTC)

That's not fair, there's zero evidence that I'm unwilling to understand. I'm seriously asking the question - what is the logic that says "4 hatnotes at the top is bad - make a DAB" but "1 annoying/trivial hatnote is good" - and no-one has answered it. Then, how does that logic gel with the idea that you should not make a user click twice if you can avoid it? Also, calling the experiment half-baked is not fair either - I've proposed above, several times, that I want to work with people to develop the experiment - so if you're in, help me develop it - don't make cheap shots at it. Here is a simple problem statement as a start: "Dear user: given the following two interfaces (with DAB, and without), and the following two scenarios, which would you prefer when:1) searching for information about sexual harassment; 2) trying to find a beavis/butthead episode. If you could only chose one interface for the encyclopedia, how would you rate the tradeoff between the two, which would mean either a) Not seeing Beavis/Butthead mentioned at top of Sexual harassment article and having to click twice to get to Beavis/butthead episode or b) seeings beavis/butthead mentioned at top of article; only needing to click one to get to Beavis/Butthead episode. You could vary the question to have different examples, to see how opinions change as relative triviality/seriousness of subjects change (like Jesus/Jesus (death metal band), or "johnny cash", "Johnny Cash (song)", etc. --KarlB (talk) 01:03, 7 May 2012 (UTC)

I'm seriously asking the question - what is the logic that says "4 hatnotes at the top is bad - make a DAB" but "1 annoying/trivial hatnote is good" - and no-one has answered it. Then, how does that logic gel with the idea that you should not make a user click twice if you can avoid it?

Do you honestly not recognize any difference between maximizing the ease with which readers access articles and minimizing the likelihood that they'll encounter material that they find "irritating"?

Note that I'm not asking you to agree that the former should take precedence over the latter. I'm only asking you to acknowledge that a distinction exists. —David Levy 01:36, 7 May 2012 (UTC)

I recognize that you make a distinction. However, I don't recognize that it is a valid binary distinction from the POV of user interface design. Navigational ease is only one of many metrics on which you might judge an interface; the overall goal of an interface is to be clear and easy for the user, and to make it a pleasure for them to use. This is accomplished through navigation, wording, color, button placement, and many other elements that all tie together. That's why when you said you wouldn't change your position, even if presented with user data, it made me believe that you don't understand the fundamentals of UI design, the first one of which is "serve your users". In any case, the point I made has now been answered - the reason you eliminate 4 hatnotes is because that causes clutter - thus a visual confusion can take precedence over navigational convenience. --KarlB (talk) 03:10, 7 May 2012 (UTC)

I recognize that you make a distinction.

Then stop pretending otherwise. You're entitled to express disagreement with our assertion that said distinction is a sensible basis for the setup that we advocate, but please stop claiming that our argument lacks internal logic/consistency.

Navigational ease is only one of many metrics on which you might judge an interface; the overall goal of an interface is to be clear and easy for the user, and to make it a pleasure for them to use. This is accomplished through navigation, wording, color, button placement, and many other elements that all tie together.

We aren't debating our hatnote templates' style or wording. The proposal stems from a desire to increase readers' "pleasure" by removing article titles from view because some readers might find their inclusion unpleasant.

That's why when you said you wouldn't change your position, even if presented with user data, it made me believe that you don't understand the fundamentals of UI design, the first one of which is "serve your users".

Wikipedia serves its users by disseminating knowledge, not by polling them and tailoring its encyclopedic content to reflect their preferences.

In any case, the point I made has now been answered - the reason you eliminate 4 hatnotes is because that causes clutter - thus a visual confusion can take precedence over navigational convenience.

And if you were to present evidence that a single direct link to a non-primary topic's article causes confusion or otherwise impacts readers in a manner adversely affecting their ability to utilize the encyclopedia as intended, I would gladly reconsider my position. —David Levy 05:43, 7 May 2012 (UTC)

(after ec) I wouldn't call repeating the same questions that have already been addressed zero evidence. First, there is a difference between multiple hatnotes and a single hatnote that contains multiple entries. Century is an example of multiple hatnotes. This situation usually arises when when there are multiple redirects to an articles (sometimes as a result of merging). These hatnotes are often seen as undesirable in that they take up a lot of valuable real estate at the top of an article (and are often rife with needlessly redundant phrasing (i.e., for the single by Johnny Doughboy, see X (Johnny Doughboy single)). There is currently no universally good solution for this problem apart from judiciously editing the hatnotes to minimize space while retaining usability. OTOH, a hatnote for a single term that contains multiple links might arise in a couple of different situations. In one, there might be only two or in rare cases three other ambiguous pages (I can't think of an examples at the moment). Another is when there is a disambiguation page, but there is one other very prominent use for the term (or in rare cases, a couple). Apple and oracle are both examples of this. Multiple entries in a single hatnote can run into some of the same criticisms of taking up too much valuable space in an article. That is why we use disambiguation pages rather than listing all the ambiguous links in the hatnote. The problem with disambiguation page where there is only one ambiguous term is that it is entirely useless and unnecessary. The primary topic page has to have a hatnote, regardless of whether it links to a disambiguation page or links directly to the other ambiguous page. That disambiguation page serves no purpose whatsoever other than to assuage the sensitivities of editors who think the mention of trivialities belittles "serious" articles. That disambiguation page would have no other links to it other than the hatnote on the primary topic (and the inevitable links in discussions about whether to delete it or not). I don't see any benefit in such censorship (and it is a form of censorship). older ≠ wiser 01:42, 7 May 2012 (UTC)

I've tried to read everything in this thread, but I honestly don't understand why it's acceptable to have a DAB page with 3 terms, but not with 2.

This can result in a hatnote with one link instead of two. Whether that's where the line should be drawn is debatable, but how can you not perceive a difference (and the numerical parity between one link to an article and one link to a disambiguation page)? —David Levy 01:36, 7 May 2012 (UTC)

Annoyance to editors is secondary. It is annoyance/helpfulness to readers which should be the primary consideration of hatnotes (or anything else to do with article structure). SpinningSpark 00:26, 7 May 2012 (UTC)

I agree 100%. However, two of the strongest opponents have stated that majority preference should not trump inconvenience for a minority, and that they would not change their positions even if presented with data from users indicating a different UI preference. --KarlB (talk) 00:52, 7 May 2012 (UTC)

Assisting readers in accessing articles is an important element of the project's mission to propagate knowledge. Assisting them in avoiding article titles that they find unpleasant is not. —David Levy 01:36, 7 May 2012 (UTC)

Please explain, why not list every term in the hatnote? it's saving clicks, right? I'm sorry but your arguments about saving clicks are not consistent - if it's *always* better to save clicks, then you should almost *always* do so.

No one has asserted that it's "always better to save clicks".

However, if there are times when you trade off clicks for something else (like usability and clarity), then just admit it, and we can actually have a real interface discussion.

Already acknowledged.

And yes, the fact that thousands of viewers are irritated is an actual problem, whether you think it is or not.

That some readers find certain encyclopedic content "irritating" isn't a valid reason to remove it.

I've proposed that we create a test to find out for sure, but as you can see above, no-one except Diego has accepted that they might change their views based on feedback from users, so it's hard to make an interface argument if people don't really care about UI...

In other words, if someone disagrees with your idea of what constitutes an ideal UI, this means that they "don't really care" about it. —David Levy 01:36, 7 May 2012 (UTC)

We *can* set up an experiment on the other hand to see what users prefer within the bounds of current technology, but two of the major proponents of the current system don't seem to care what users think, and have professed that they will not change their positions even in the face of preference data from real users.

We didn't say that we don't care what users think. We said that their preferences don't override our mission to propagate knowledge.

All else being equal, of course we want to make the site as appealing as possible. But this isn't a proposal in which all else is equal; it's one in which article access is made more difficult.

So, it's hard to have a UI discussion with people who feel like that - I mean, if you took their arguments seriously, it would lead you to a dramatically different interface - because if adding additional clicks really is bad, then most disambig pages should disappear, and a small list of links should be added to the top of every page. I'm not sure how they can simultaneously defend disambig pages with 4 links, while arguing forcefully for the removal of disambig pages with 2 links - their arguments are not logically consistent.

Again, no one has asserted that disambiguation pages are "bad". We're saying that direct links, when feasible, are better.

We all agree that it's possible to include too many links in a hatnote (thereby resulting in problematic clutter). Where to draw the line is debatable, but one link obviously isn't too many. There's no logical inconsistency in the belief that linking directly to one article makes sense but linking directly to [quantity >1] articles doesn't.

For example, David Levy said "That someone finds a particular topic "trivial" (even of you do want to refer to this experience as an "inconvenience") isn't a valid reason to increase others' difficulty in accessing its article." And yet, we do that, every day, when a search for Jesus redirects to Jesus, and not Jesus (disambiguation). Thus, we are forcing thousands of people, every day, to click twice in order to find their article about Jesús,_Peru and Zola_Jesus. Why are we willing to do this?

Because it assists navigation to the greatest extent possible. We send readers directly to the primary topic's article because it's what they most likely seek, not because the other articles embarrass us. Such a setup adds a click for some readers, but it subtracts a click for far more readers. This results in a clear net navigational benefit for our readership as a whole.

Conversely, the proposed change would have the opposite effect; navigation would only be made more difficult.

If you apply the logic above, there is no valid reason to add additional clicks to a minority just because a majority doesn't want additional clicks.

That simply isn't true. Navigation is navigation. A desire to avoid encountering "irritating" content is something else entirely.

Again, this isn't about pleasing a majority by catering to their preferences; it's about disseminating knowledge. We seek to assist readers in accessing articles, not to make them happy by providing whatever they "want".

Or, perhaps there is a majoritarian logic at play here after all? Perhaps, wikipedia does agree that it's better to *not* inconvenience the majority, it is better to inconvenience the minority.

Indeed. And a single direct link to a non-primary topic's article (instead of a link to a disambiguation page) inconveniences no one.

So then the argument becomes, can you irritate and bother the majority, or can you inconvenience the minority - and how do you make the tradeoff?

It's very simple. Enabling access to articles with the greatest ease possible is a fundamental element of the project's mission. Suppressing text on the basis that some people find it unpleasant is not.

It becomes a silly semantic game - inconvenience matters, irritation doesn't, etc.

As I wrote previously, quibbling over the terminology is a red herring. Let's say, for the sake of argument, that readers encountering direct links to articles whose titles they find unpleasant are "inconvenienced". It's the context that's materially different (as explained above). This "inconvenience" is based entirely upon personal dislike of encyclopedic information. We don't make material more difficult to access on the basis that some readers (even a majority) find it objectionable.

and shows little understanding of user interface design.

We have a clear understanding that inserting an additional navigational step makes the interface more complicated and less convenient.

So, if you're making an interface argument, which I am (not a censorship argument), and you're arguing with people who don't agree with or understand basic tenets of user interface design, it's a bit like bashing your head against a wall...

Which basic tenet of user interface design involves moving a single link to a separate page, accompanied only by a link to a page at which the user just arrived accidentally?

In any case, it doesn't really matter - every issue that has been brought up in this debate has been resolved by a disambig page, which has gone uncontested. So, the solution seems to be, just create disambig pages, find some good links to add, and forget about trying to change policy.

The operative word is "good". If valid article links have been overlooked, that's fine. —David Levy 01:36, 7 May 2012 (UTC)

Exactly. This was my primary issue with this proposal from the start. The edit that was made to the guideline explicitly advocated creating invalid disambiguation pages on the dubious supposition that more uses will be found. If the guidance were framed to first find other legitimately ambiguous titles to create a valid disambiguation page, there would be little disagreement from me. older ≠ wiser 01:53, 7 May 2012 (UTC)

Aha! We're starting to see the cracks in the facade, at long last.

• "We're saying that direct links, when feasible, are better."
• "Where to draw the line is debatable, but one link obviously isn't too many. There's no logical inconsistency in the belief that linking directly to one article makes sense but linking directly to [quantity >1] articles doesn't."
• "Enabling access to articles with the greatest ease possible is a fundamental element of the project's mission:"

Note how important navigation is to David. Indeed, he claims that enabling access to articles with the greatest ease possible is fundamental. Unfortunately, this is neither written anywhere, nor supported (it is, in fact, refuted by many policies and rules, for example our policies around external links, notability, see-also sections, and what-can-be-present on disambig pages. We actively delete direct links, all over wikipedia, in the interest of clarity - and there is zero evidence that ease of access to articles is fundamental to the project - we seek clarity and relevance, but not 'ease of access' - otherwise, we would pepper the whole encyclopedia with more links, to tangential things on the off-case that a lost user was looking for something, because it would improve their navigation... But, he also admits that "where to draw the line is debatable", thus illustrating that the choice of 'clutter vs one-click links' is an interface decision, and is not based on any fundamental principle. If such an interface decision can be made in that case, then an interface decision can be made for 2-DAB links, so the arguments about 'inconvenience' vs 'clutter' vs 'irritation' are thus rendered invalid. If one can forgo navigation to avoid clutter in once case, so there's no reason once can't make an argument to slow down navigation to avoid irritation/cognitive dissonance in another.--KarlB (talk) 02:36, 7 May 2012 (UTC)

I think you're exhibiting the same deliberate misunderstanding that has hamstrung the discussion all along. This discussion is not about the choice of 'clutter vs one-click links'. Unless you are seeking to completely remove hatnotes, a primary topic with any other ambiguous topics will have a hatnote. Having a direct link in the hatnote where there is exactly one other topic is no more clutter than linking to a disambiguation page. Forcing a disambiguation page in such a case has no benefit in terms of usability. older ≠ wiser 02:47, 7 May 2012 (UTC)

another cheap shot. Thanks for that. I am not exhibiting misunderstanding. I am trying very hard to understand your logic, so I can dismantle it as I'm doing now. I know that you believe a single link to Jesus (death band) is no more clutter than a link to Jesus (disambiguation). But the reason David's argument has now been undermined, is the admission that clutter can be traded off against navigation. Clutter is "a confused disordered jumble of things". In my opinion, descriptions of tangential/trivial/vastly-less-popular articles as the first thing your eye hits, is the very definition of clutter. That's the problem with all of your arguments - you're basing everything on intent (censorship - which you keep on bringing up but which you haven't yet established), whereas I am basing all of my arguments on how to present the most clean and clear user interface to the majority of users. The majority of users, in my opinion (which I'm willing to test - are you?), would prefer the neutral wording 'for other uses, see this' because they don't have to parse it, read it, try to understand it - they can scan it an instantly and decide, they don't need to click. On the other hand, with a descriptive hatnote, especially in the egregious cases we've mentioned before like Beavis, the first thing the reader does is digest this Beavis episode, and it distracts them in the same way 4 links would. So if clutter is an acceptable word for you guys, we'll start using that, and make the argument all about reducing clutter, as you seem agreed that it's a good idea, and that reducing clutter is an excuse to send people to DAB pages... --KarlB (talk) 02:58, 7 May 2012 (UTC)

I know that you believe a single link to Jesus (death band) is no more clutter than a link to Jesus (disambiguation). But the reason David's argument has now been undermined, is the admission that clutter can be traded off against navigation.

"Admission"? You mean the longstanding acknowledgement entirely consistent with my position from the beginning (which you've continually misconstrued)?

Clutter is "a confused disordered jumble of things". In my opinion, descriptions of tangential/trivial/vastly-less-popular articles as the first thing your eye hits, is the very definition of clutter.

Please explain how a single-link hatnote constitutes "a confused, disordered jumble of things".

That's the problem with all of your arguments - you're basing everything on intent (censorship - which you keep on bringing up but which you haven't yet established), whereas I am basing all of my arguments on how to present the most clean and clear user interface to the majority of users.

Please explain how a hatnote linking to a disambiguation page is cleaner and clearer than one linking to an article is (and why this applies specifically to instances in which x disparity in prominence exists between the primary and non-primary topics).

The majority of users, in my opinion (which I'm willing to test - are you?)

You've proposed polling them on their preference. Readers would prefer that we omit all sorts of "objectionable" material. That isn't how Wikipedia works.

would prefer the neutral wording 'for other uses, see this' because they don't have to parse it, read it, try to understand it - they can scan it an instantly and decide, they don't need to click.

I don't understand that argument at all. How is a link to a disambiguation page less confusing or easier to ignore?

Upon arriving at the page, a reader not in need of the link either will or won't recognize that he/she has reached the intended article before reading the hatnote. In the latter case, a link to disambiguation page would be less helpful, as it wouldn't identify the non-primary topic's nature (leaving open the possibility that the link should be followed to reach the desired information).

On the other hand, with a descriptive hatnote, especially in the egregious cases we've mentioned before like Beavis, the first thing the reader does is digest this Beavis episode, and it distracts them in the same way 4 links would.

How is this distracting? The reader immediately knows that the link pertains to an unrelated subject and can safely be disregarded. —David Levy 05:43, 7 May 2012 (UTC)

"Having a direct link in the hatnote where there is exactly one other topic is no more clutter than linking to a disambiguation page. Forcing a disambiguation page in such a case has no benefit in terms of usability." Read this for evidence of why your statement isn't true: [5] "Layout is the sizing, spacing, and placement of content within a window or page. Effective layout is crucial in helping users find what they are looking for quickly, as well as making the appearance visually appealing. Effective layout can make the difference between designs that users immediately understand and those that leave users feeling puzzled and overwhelmed." There is too much to go into in detail, but the point is, I hope that you will understand that navigation is not the end-all/be-all of interfaces, and usability does not just equal one-click-to-get-where-I-want. Indeed there are many cases where different text on the same button can be worse from a usability perspective, as we are arguing here. But again, are you willing to test it? If so, help refine the problem statement, and help me design an experiment, and sign your name that you will change your position if the experiment shows the DAB page to be more usable and preferred. more wisdom from Redmond: "Comprehensible text is crucial to effective UI." "If you do only five things... 1. Work on text early because text problems often reveal design problems. 2. Design your text for scanning. 3. Eliminate redundant text. 4. Use easy-to-understand text; don't over-communicate. 5. When necessary, provide links to Help content for more detailed information.[6]. I think the key phrase there is 'don't overcommunicate' - since we've agreed that the bulk of the users will be scanning, and not clicking the hatnote, not overcommunicating is important to them and their user experience. --KarlB (talk) 04:04, 7 May 2012 (UTC)

Failing to inform users that a single navigation option exists (and instead forcing them to click through to a separate page on which said path is presented alongside the option to return to a page that they reached accidentally) is undercommunication.

This proposal isn't about improving readers' comprehension. It's about sparing them from seeing article titles that some might find "irritating". No matter how many respondents to a poll express such a preference, a desire to please them won't outweigh the importance of efficient navigation. —David Levy 05:43, 7 May 2012 (UTC)

I wouldn't take my interface advice from Redmond; although they have very competent people in UI, their overall corporate structure makes it really difficult to create pleasant UI experiences, and their interfaces are not usually known for an efficient and stress-free elegant design. Diego (talk) 06:13, 7 May 2012 (UTC)

Note how important navigation is to David. Indeed, he claims that enabling access to articles with the greatest ease possible is fundamental. Unfortunately, this is neither written anywhere, nor supported

Our mission is to write an encyclopedia and make it available to the world. Ensuring that people are able to access these articles with as few encumbrances as possible is an intrinsic element.

(it is, in fact, refuted by many policies and rules, for example our policies around external links, notability, see-also sections, and what-can-be-present on disambig pages. We actively delete direct links, all over wikipedia, in the interest of clarity - and there is zero evidence that ease of access to articles is fundamental to the project - we seek clarity and relevance, but not 'ease of access' - otherwise, we would pepper the whole encyclopedia with more links, to tangential things on the off-case that a lost user was looking for something, because it would improve their navigation...

At no point have I claimed that we seek ease of access above all else. Your out-of-context quotation is part of a comparison between "enabling access to articles with the greatest ease possible" and "suppressing text on the basis that some people find it unpleasant". My point is that the latter is unrelated to our mission, not that everything other than the former is.

As I (and others) have acknowledged in this discussion, while the quantity of clicks required is significant, there are other relevant considerations. "Will readers find this title annoying?" isn't one of them.

But, he also admits that "where to draw the line is debatable", thus illustrating that the choice of 'clutter vs one-click links' is an interface decision, and is not based on any fundamental principle.

That dilemma is irrelevant. What quantity of links constitutes "clutter" in this context is debatable, but whether it's two, five or a hundred, it can't possibly be one; that's the absolute minimum, and it's what we'll have either way.

If one can forgo navigation to avoid clutter in once case, so there's no reason once can't make an argument to slow down navigation to avoid irritation/cognitive dissonance in another.

You can make that argument, but it doesn't reflect Wikipedia's basic principles. Clutter makes an article more difficult to read (I assume that you agree with this), thereby interfering with our mission to disseminate knowledge. A reader's encounter with an article title that he/she finds "irritating" has no such effect. —David Levy 05:43, 7 May 2012 (UTC)

"Clutter makes an article more difficult to read .... A reader's encounter with an article title that he/she finds "irritating" has no such effect." (see WP:SURPRISE, for one example of how wikipedia policy comes into play here. For other uses, see Sexual harassment (disambiguation) is easier to read than For the Beavis and Butthead TV episode, see Sexual harassment (Beavis and Butthead). In any case, it doesn't really matter. 2DAB itself states that such pages are harmless...--KarlB (talk) 06:09, 7 May 2012 (UTC)

Actually you're wrong - if the link is properly styled and de-emphasized so that it can be ignored for people not looking for navigation, both texts will be read the same - which is not at all, as the scanning eye will ignore it. And if the user is looking for a TV episode, the second is easier to read since it contains the clues the user is looking for. When there are not "other uses", the second text is definitely better (assuming the visual design is correctly weighted for scannability). Diego (talk) 06:26, 7 May 2012 (UTC)

(see WP:SURPRISE, for one example of how wikipedia policy comes into play here.

WP:SURPRISE is part of an essay, not a "policy". This, of course, doesn't negate the validity of the advice, which I can only assume you haven't read in its entirety. (See the "Paris Hilton" example.)

For other uses, see Sexual harassment (disambiguation) is easier to read than For the Beavis and Butthead TV episode, see Sexual harassment (Beavis and Butthead).

In what meaningful respect?

In any case, it doesn't really matter. 2DAB itself states that such pages are harmless...

What bearing does that have on the matter at hand (the hatnote's content)? —David Levy 06:32, 7 May 2012 (UTC)

• I'm not quite sure what the discussion is currently !voting on (TLDR) but I'd like to support the principle of moving trivial hatnotes to a dab page. I would also like to comment on an earlier suggestion that there is no such thing as comparative notability. Of course there is. We define notability as discussed in a non-trivial way in reliable sources. The more a topic is discussed, the more notable it is. This is recognised, for instance, in WP:PROF which discusses at length measuring notability in terms of number of citations to the subject's work. I'm a little dubious about the idea of setting concrete pageview rules. They might be a helpful indication, but popularity != notability. SpinningSpark 00:22, 7 May 2012 (UTC)

  Thanks. The discussion is a bit all over the place - we've had a few proposals, but for now it's a bit of a back and forth - I feel like there are some editors who are open to a revision of the 2-DAB rule, and others who are diametrically opposed to any revision of the 2-DAB rule. I also agree that popularity != notability; I guess it's a hard call; if you accept disambiguation pages for 2-DAB articles, what are the criteria? Page views are nice, even if imperfect, because they are somewhat neutral. If you don't put in a figure, even a starting figure, then the edit wars will continue. Setting a bar, and then letting people argue above or below that bar, IMHO is better. But, if consensus moves towards a revision, without a specific # of page views, I would still support.--KarlB (talk) 00:52, 7 May 2012 (UTC)

  If you want a specific criterion that would improve UI design, I'd go with screen real state. Put links to other articles in the hatnote when they fit in one line. This is that would provide a clear logic for when showing article links and when not to do it. But then it wouldn't solve your "hiding links I disapprove" concern, would it? Diego (talk) 06:18, 7 May 2012 (UTC)

  I don't think that PamD meant to suggest that no topic with a Wikipedia article is more or less notable than another (the very basis of establishing primary topics); she meant that we don't deem some "serious" and others "trivial", treating the latter as second-class content unworthy of linkage from the former. —David Levy 01:36, 7 May 2012 (UTC)

• KarlB - your all-or-nothing arguments look like a strawman of an otherwise interesting position. At this point, even if your experiment would show people feeling inconvenienced, I wouldn't support the proposal in this thread since there's an alternate approach that would not need to drop the link to the so-called "trivial" article. If you want to make progress, discuss changes to placement, wording, color and all other elements of the hatnote without removing it. (The questions you defined above would be a terrible way to approach the experiment that would bias the answers, BTW. Users should be given a task to perform with a neutral wording - either evaluating the Sex harassment article or finding one of the TV episodes - and afterwards asked if they found something unusual or bothersome about navigation - without inducing them to notice the hatnote). Diego (talk) 06:02, 7 May 2012 (UTC)

Making progress

Ok, I think we've made a lot of progress since the start of the discussion. To date, I think we've established the following:

1. It is legitimate to send readers to a DAB page when doing so de-clutters the interface
2. It is legitimate to send readers to a DAB page when 3 terms can be found
3. Usability should be the determining factor. Navigational ease is one consideration of usabilitiy; but there are others as well (such as legibility, readability, de-cluttering, etc).
4. There are no absolute rules in deciding between these (i.e. when 1 click can be traded off against an extra link) - instead, these design principles have evolved over time, and can continue to evolve.
5. Some people believe that 1-click hatnote is always better than 2-click DAB. Others believe 2-click DAB is sometimes better than 1-click hatnote. Neither group has a monopoly on truth; in fact, both groups are now arguing based on principles of UI design.

All in all, I think that's good news. What it means is, we're having a discussion based on user interfaces - not on some abstract notion that we must always provide instant access to information in the most unencumbered and uncensored way possible.

Also, a few illustrative quotes with which I disagree:

• "No one is inconvenienced by the existing hatnote over the proposed replacement hatnote; some are inconvenienced by the proposed replacement hatnote over the existing hatnote. The disdain for supposed triviality is not inconvenience."
• "what users prefer isn't the relevant criterion."
• "Our mission is to propagate knowledge, not to please people by catering to their sensibilities."
• "a desire to please them won't outweigh the importance of efficient navigation"

These arguments (and similar) have been repeated ad nauseam above. The point seems to be, there are some reasons (like clutter) on which we are willing to try to help users, and make their lives easier. However there are others where we draw the line - the user's preferences no longer matter. The evidence above shows this is true. So it's almost as if we're deciding, on behalf of users, which of their preferences are relevant, and which are not ; what is distracting; what is not; what is clutter; what is not; what is a usability issue; what is a 'user preference' issue. As should be obvious, I disagree with this; I don't think this is black and white; and I don't think any of the editors here can objectively make that determination - user experience is ultimately a subjective issue. The wikimedia foundation itself [7] accepts that allowances should be made for user sensibilities (in this case, worded as "that the principle of least astonishment for the reader is applied"). Even the 2-DAB rule, which states that 2-DAB pages are harmless, doesn't seem to see a problem with 2-DAB pages (and just below, is the exhortation that one should "Break the rules... Usefulness to the reader is their principal goal.")

Thus, allow me to bring this back to the point I made at the beginning - since we seem to agree that clutter can trump navigation, are you willing to undertake an experiment that would measure user experience feedback in terms of clutter, disambig hatnotes, and the like? We don't seem to be very close to consensus, so perhaps it's time to appeal to a higher power. And again, I ask the question - is there anything you might learn from a panel of users that could possibly change your mind? For example, suppose we polled a thousand users of wikipedia, and 99.999% of users said "I don't care about clicking twice occasionally, if it means I won't have to see articles like Beavis at the top of serious pages" - would you still insist that it must remain? E.g if users decide that the tradeoff was worth it (and this is not censorship, this is simply an interface tradeoff that exists on hundreds of thousands of wikipedia pages already), why do their motives matter? Are your motives somehow more "pure"? Please, before making statements that we are here to... or this project is to recall that Jimbo himself agreed that this was silly, and he has more authority to opine on what wikipedia is or isn't IMHO.
We've already established that usability is the determining factor, not ease of navigation. So, if we could set up a test to measure usability of different interfaces, would you sign on to the results and change your mind accordingly? --KarlB (talk) 22:52, 7 May 2012 (UTC)

Some people believe that 1-click hatnote is always better than 2-click DAB. Others believe 2-click DAB is sometimes better than 1-click hatnote. Neither group has a monopoly on truth; in fact, both groups are now arguing based on principles of UI design.

You're attempting to apply your interpretation of UI principles to content-related elements of an issue that also pertains to our UI.

These arguments (and similar) have been repeated ad nauseam above.

Why do you suppose that is? Do you think that we enjoy repeating ourselves?

The point seems to be, there are some reasons (like clutter) on which we are willing to try to help users, and make their lives easier. However there are others where we draw the line - the user's preferences no longer matter.

The distinction is extremely simple and has been explained repeatedly ("ad nauseam", one might say).

Our mission is to provide access to knowledge, not to indiscriminately please people. The ease with which Wikipedia's readers access articles that they seek is a valid consideration (not the only valid consideration, but a valid consideration). The possibility that they'll encounter an article title that irritates them is not a valid consideration; in no way does it adversely affect their ability to utilize the encyclopedia in the manner intended.

Again, I'm not demanding that you agree with this position. I'm asking you to stop feigning bewilderment as to why we want to "help" users in some situations and not others.

So it's almost as if we're deciding, on behalf of users, which of their preferences are relevant, and which are not

In other words, we don't permit people to vote away encyclopedic content on the basis that they would "prefer" not to encounter it.

As should be obvious, I disagree with this; I don't think this is black and white; and I don't think any of the editors here can objectively make that determination - user experience is ultimately a subjective issue.

Whether an article's title is "annoying" is subjective. The relative efficiency of an indirect link and a direct link (where no others are needed) is not.

The wikimedia foundation itself [8] accepts that allowances should be made for user sensibilities (in this case, worded as "that the principle of least astonishment for the reader is applied").

How is a pointer to an article about a subject with a similar name "astonishing", and how does the this alleged "astonishment" interfere with the reader's ability to utilize the encyclopedia in the manner intended?

This comes across as an attempt to spin the aforementioned "irritation"/"annoyance" into a buzzword on which to hang your hat. You've already tried linking to WP:ASTONISH (apparently without realizing that it recommends precisely the sort of hatnote that you seek to eliminate, cited as a means of preventing reader astonishment).

Even the 2-DAB rule, which states that 2-DAB pages are harmless, doesn't seem to see a problem with 2-DAB pages

Why do you keep mentioning that? The debate is about whether to link to such pages from articles, so this is yet another straw man.

(and just below, is the exhortation that one should "Break the rules... Usefulness to the reader is their principal goal.")

I'm a strong supporter of WP:IAR. But in this context, "usefulness" relates to the project's mission, not to what readers might find useful in general. (Dictionary definitions and recipes are useful, but we don't host them at Wikipedia.)

Thus, allow me to bring this back to the point I made at the beginning - since we seem to agree that clutter can trump navigation, are you willing to undertake an experiment that would measure user experience feedback in terms of clutter, disambig hatnotes, and the like?

You're defining "clutter" as "material that readers find unpleasant". That isn't what the rest of us mean, and I don't know how to make our position that this isn't a valid consideration any clearer.

You're ignoring the crux of our arguments, cherry-picking comments that superficially support your stance when context is omitted and meanings are distorted, and asking the same question over and over (as though you're oblivious to our repeated responses and the basis thereof).

We don't seem to be very close to consensus,

I see clear consensus.

so perhaps it's time to appeal to a higher power.

Meaning what? Please elaborate.

And again, I ask the question - is there anything you might learn from a panel of users that could possibly change your mind? For example, suppose we polled a thousand users of wikipedia, and 99.999% of users said "I don't care about clicking twice occasionally, if it means I won't have to see articles like Beavis at the top of serious pages" - would you still insist that it must remain?

Suppose we polled a thousand users of Wikipedia and 99.9% of them said, "I don't care about clicking twice occasionally, if it means I won't have to see statements inconsistent with my beliefs."*

That some people prefer not to encounter certain material isn't a valid reason to hide it from view (even if only in the context of a hatnote).

*To be clear, I'm not equating the proposed change with that.

E.g if users decide that the tradeoff was worth it (and this is not censorship,

It is censorship. That the content happens to appear within a UI doesn't magically change that. You're clinging to a technicality.

this is simply an interface tradeoff that exists on hundreds of thousands of wikipedia pages already)

Hundreds of thousands of Wikipedia pages contain indirect links substituted for direct links to avoid exposing readers to article titles that they find unpleasant?

why do their motives matter?

Because our mission is to provide access to knowledge, not to indiscriminately please people.

Please, before making statements that we are here to... or this project is to

Do you disagree with the statement that "our mission is to provide access to knowledge, not to indiscriminately please people"?

recall that Jimbo himself agreed that this was silly, and he has more authority to opine on what wikipedia is or isn't IMHO.

"I think that almost any argument, on any topic, which has premises beginning with 'Jimbo said...' is a pretty weak argument. Surely the merits of the proposal should be primary, not what I happen to think." —Jimbo Wales

—David Levy 02:33, 8 May 2012 (UTC)

David, your responses are very difficult to read and reply to. You don't need to challenge every single sentence I write - it's annoying.

A few responses:

"Even the 2-DAB rule, which states that 2-DAB pages are harmless, doesn't seem to see a problem with 2-DAB pages" "Why do you keep mentioning that? The debate is about whether to link to such pages from articles, so this is yet another straw man." No, this is not a straw man. This is a discussion about changing the 2-DAB rule. The argument has been made above that a 2-DAB page is a bad idea. I'm pointing out that the policy says, 2-DAB pages are harmless. Thus, the "inconvenience" touted again and again, or the "silliness" or "pointlessness" of a 2-DAB page is not supported by policy at the current moment. I'm sure you understand that creation of a 2-DAB page is a prerequisite to linking to a 2-DAB page in this case. They are two sides of the same coin, so your argument doesn't obtain.

"You're defining "clutter" as "material that readers find unpleasant". That isn't what the rest of us mean, and I don't know how to make our position that this isn't a valid consideration any clearer."

David, you have a POV about what is a "valid" UI consideration, and what is not. And "the rest of us" is just a few editors on this page; it has nothing to do with the general population of wikipedia, and a significant portion of the editors posting here *want* to get rid of these links. It's abundantly clear that your assessment of UI issues is simply a POV, perhaps you've been stating it for so long it appears to be truth. But in fact, it's just an opinion. No more valuable than mine or any other editor's. I'm not defining "clutter" as "material that readers find unpleasant". I'm saying, there are many reasons someone may react negatively to a given user interface element. It is not your role to judge the validity of those responses, and to decide that negative reactions due to # of clicks is acceptable, but negative reactions due to presence of certain words, or layout of UI elements, or size and spacing, etc is not. You're so focused on the fact that some piece of text may annoy the user, and you seem so anxious to do anything in your power to ensure that the user will see that text - in this one tiny little edge case (where 3-DABs can't be found).

You obviously didn't read the link I attached - the wikipedia foundation specifically states that you should not astonish the user: "content on Wikimedia projects should be presented to readers in such a way as to respect their expectations of what any page or feature might contain" What this means is, in an extreme case, is you cannot shock the user with an offensive picture in a place they don't expect just because you believe the encyclopedia is uncensored - there is a balance to be struck which you seem to forget. Let me propose a reductio-ad-absurdum hatnote case - for example, if there was a sex toy called Jesus, it would be absurd to put This article is about Jesus of Nazareth. For the sex toy, see Jesus (sex toy) - WP:IAR would apply and that hatnote wouldn't last 2 minutes. Why? Because it violates what the user expects to see on that page.

Here is another reductio-ad-absurdum - if I proposed that all hatnotes should have red blinking links, I could then use your same logic to say "well, it doesn't impact their ability to click, they can easily ignore it, and it helps a minority users find the link they want - therefore, even if it annoys the f*** out of the users, we must keep the red flashing links"...

Listen, you guys have tried really hard, I will give you credit, to distinguish between some sort of objective measures of interface utility (like # of clicks to get what you want, navigational metrics), and to separate that from subjective measures (dislike, distaste, annoyance, irritation, bothering, cluttering, confusing, bewildering, befuddling, readability, etc)... but I'm afraid the 'clutter' argument (we to DABs to avoid clutter), and dozens of other policies all over wikipedia (like deletion of useful links in disambig pages - which results in not a 2-click or 3-click solution to get what I want, but a dead end!) - has led me to conclude that there is no objective UI standard being applied here; these are all subjective, no matter how clever and impassioned your arguments in defense of unfettered information access.

What if something doesn't annoy them, what if it just bothers them? What if it doesn't bother them, but just distracts them? What if it doesn't distract them, but renders the page less easy to read? What if it's not less easy to read, but lab experiments show their eyes linger for 0.3s longer when the DAB is not neutral? What if it's not less easy to read, but they find it more cluttered? It's a slippery slope, once you start judging which opinions are valid and which are not, you'll never be able to draw a defensible line. So many of you guys above have stated "this is pointless" or "it's a bad interface" or "it inconveniences users looking for xxx" - but these are simply opinions about the efficacy of an interface where a tradeoff has been made. We have plenty of evidence that there are other editors who DISAGREE with you. Thus, who are you to judge, which UI reactions are worth doing something about, and which ones aren't? Sadly, you have no leg to stand on, there is no absolute truth here. You have an opinion, as do I, and the only higher power we can appeal to is users at this point. I'm sorry, but as soon as you admitted that 'usability' and 'readability' were at play, the argument was lost. --KarlB (talk) 04:29, 8 May 2012 (UTC)

Karl, your arguments at this point are a severe case of "I didn't hear you". Even if we acknowledged that there's a problem of clutter in the current tophats, which we don't, you're basically ignoring everything I said about user interface design so your arguments about reactions to UI fall under their own weight; the sub-optimal design you suggest is simply not the best option in any circumstance. Stop beating the dead horse (I can't believe I had to link there); if there's consensus for something, it's that the proposal to remove hatnotes would objectively inconvenience some users for a subjective benefit that is avoiding negative reactions (even you agree to that). So if you want to advance making consensus then find a proposal acceptable to you that doesn't require hiding the link to the secondary article. Diego (talk) 06:28, 8 May 2012 (UTC)

P.S. You want to talk about design trade offs? What design trade offs are you willing to make to build a consensual design that incorporates the needs of all users? So far your position hasn't moved a bit from the initial proposal. Diego (talk) 06:53, 8 May 2012 (UTC)

I'm very happy to continue discussing UI. That's why I think we're making progress - because at least you are using the language of interface design and tradeoffs. We've all dealt with frustrating user interfaces, and this is frankly one of those little frustrating interface issues - again I think its not helpful to try for an objective vs subjective argument, it doesn't get you anywhere. At some point, we just need to talk to users.

In any case, there are thousands of edits every day that *remove* links from wikipedia, for no other reason than some editor thinks they don't belong, and nobody is crying in their milk about that - so the Alamo-like defense of this 1-click difference is quite stunning. In addition, this guideline Disambiguation_page#Disambiguation_pages and others have lots of rules about what *not* to include. While I agree with most of these, it must also be admitted that some of these rules might exclude links which would possibly be interesting/useful to a reader who happens upon that page; but for now, instead of a 1-click or 2-click way to find what they want, those users get zilch. Again, nobody is crying in their milk over that either.

I think David is the one beating a dead horse with his insistence that if something annoys the user, it is off the table in terms of any possible UI changes. Dogmatism here is just misplaced.

I'm thinking about your proposal below; I have not ignored what you said, and we simply disagree. You claim my solution is not the "best" option for UI design; I don't agree, but I'm willing to listen, as long as you state your position as an opinion and not a statement of fact. It sounds like you are also open to consider that the current system is also not the best UI design. So, we have a point of consensus - we agree that this is a UI issue, and that it can be discussed under those terms. As I mentioned earlier I'm not really interested in going into major UI or mediawiki changes, I'm interested in what can be done with the tools we have. I think in order to understand your proposal, I need some examples to work with - do you have any good ones? For example, how would you change Jennifer Lopez? or Sexual harassment (the old version)--KarlB (talk) 07:20, 8 May 2012 (UTC)

We've all dealt with frustrating user interfaces, and this is frankly one of those little frustrating interface issues - again I think its not helpful to try for an objective vs subjective argument, it doesn't get you anywhere. At some point, we just need to talk to users.

You believe that we should allow majority preferences to dictate certain encyclopedic content, and you deny that an argument to the contrary has any logical basis.

In any case, there are thousands of edits every day that *remove* links from wikipedia, for no other reason than some editor thinks they don't belong, and nobody is crying in their milk about that - so the Alamo-like defense of this 1-click difference is quite stunning.

This, I believe, is an example of the "all-or-nothing" straw man arguments to which Diego referred. Editors express disagreement with a proposal to remove a specific type of link in a particular circumstance, and you attempt to refute our logic by pointing out that lots of other links are disallowed/removed (and asserting that our position is either based on the premise that all of these links should be permitted indiscriminately or lacks internal consistency because we aren't saying that).

In addition, this guideline Disambiguation_page#Disambiguation_pages and others have lots of rules about what *not* to include. While I agree with most of these, it must also be admitted that some of these rules might exclude links which would possibly be interesting/useful to a reader who happens upon that page; but for now, instead of a 1-click or 2-click way to find what they want, those users get zilch. Again, nobody is crying in their milk over that either.

It's been decided that due to various factors, such links do more harm than good. This proposal is based on the theory that the same is true of the hatnotes in question. Some of us disagree. You're arguing that our position doesn't make sense unless we also disagree with the other reasons to omit a link.

I think David is the one beating a dead horse with his insistence that if something annoys the user, it is off the table in terms of any possible UI changes.

I've said no such thing. I've explicitly stated that this is not my position. I don't know how to make myself any clearer. —David Levy 08:15, 8 May 2012 (UTC)

Oh, but it is interesting to evaluate what's objective and what's subjective, because that's how you design an interface that's both efficient and likeable; you measure objective parameters (reading time, number of clicks, page loading time) and compare it to what problems they actually notice. If you drop the objective measurement you keep arguing in circles about what users would prefer, which is just what you're doing till now. Even if you don't want to see it as an objective/subjective decision you still have to contemplate the arguments about measurable efficiency, you can't simply hand-wave them away, since efficiency is a core concern to good UI design.

In the case of navigation and reaction time we don't even need to measure them for this preliminary exploratory phase, because we have accurate models of how much time it will take (GOMS, Fitt's Law) and how the reaction time will feel to users ([9],[10]). In this case, we know that the separate DAB page will increase the time to navigate from one page to the other and will increase the mental workload while doing so (since they have to evaluate the new loaded page even when they really don't have any decision to make in it); on the other side, the discomfort experienced is subjective and can't be precisely measured, only estimated through indirect experiments. This is what I was referring to with an objective inconvenience, and it's not just an opinion - there are models defining it and thus it will be the most likely result from the experiment. Of course the interface you design upon those measures is dependent on how it will impact users, but you can't just defer the design decisions upon a user study; it will also be influenced by the goals that the design must cover, which is the part that should be decided by consensus. The previous consensus is that the link to the other article must exist, and your arguments are not presenting a good case to change it.

I don't think that the current system is not the best UI design; I actually find it pretty good and very close to the optimum for what it tries to accomplish - solving the problem of a title clash between articles with the same name, when one of them is primary. In this case the obvious structural solution is to link them directly, something that doesn't happen when there are three articles (which is what makes the general case different from this one). With three articles, a DAB page provides a value that a two-entries DAB doesn't provide, allowing the user to make a choice. Thus the WP:SIMILAR is a pretty good solution. The changes I'm willing to consider are meant to incorporate your new requirement, to avoid a perceived surprise in some cases; but that can't be done at the cost of spoling the current near-optimal design.

I've already pointed a solution within the constraints of the current software - change the hatnote text to provide more context ("Wikipedia also contains an article about...") and reduce surprise (thus the cases would be "Wikipedia also contains an article about Jennifer Lopez the meteorologist" and "Wikipedia also contains an article about the Beavis and Butt-head episode "Sexual Harassment""). Other changes would involve changing the text style and weight, its placement and its navigation structure (for example DABs could be complemented and tophats replaced by horizontal or vertical navbox-like tables: "[Articles about Jennifer Lopez: | Singer | Meteorologist ]). Now it's your turn to show that you're indeed contemplating design options, not just a commitment to a pre-stated design. I've asked you what other trade offs would you contemplate. What's your answer? Diego (talk) 10:10, 8 May 2012 (UTC)

David, your responses are very difficult to read and reply to. You don't need to challenge every single sentence I write - it's annoying.

I'm sorry that you dislike my posting style. I'm not challenging (or even quoting) every sentence you write. I merely seek to provide context.

No, this is not a straw man. This is a discussion about changing the 2-DAB rule. The argument has been made above that a 2-DAB page is a bad idea. I'm pointing out that the policy says, 2-DAB pages are harmless.

It says that the pages themselves are harmless, not that linking to them instead of linking directly to the non-primary article (the proposed change) is.

I'm sure you understand that creation of a 2-DAB page is a prerequisite to linking to a 2-DAB page in this case.

Of course. Obviously, we can't link to such a page without creating it. We can, however, create such a page without linking to it. This is what's described as "harmless".

David, you have a POV about what is a "valid" UI consideration, and what is not.

Indeed. And as I said, I'm not demanding that you agree with this position. I'm asking you to stop misrepresenting it.

And "the rest of us" is just a few editors on this page; it has nothing to do with the general population of wikipedia,

Agreed. I used the phrase "the rest of us" in reference to participants in this discussion. We used the word "clutter" to mean one thing, and you then quoted these comments to bolster an argument in which it meant something different.

and a significant portion of the editors posting here *want* to get rid of these links.

Agreed.

It's abundantly clear that your assessment of UI issues is simply a POV

I've assessed more than UI issues, but yes, I'm expressing my opinions (just as you're expressing yours).

perhaps you've been stating it for so long it appears to be truth.

I don't know what gave you the idea that I regard my views as sacrosanct, but you're mistaken.

But in fact, it's just an opinion. No more valuable than mine or any other editor's.

Of course.

I'm not defining "clutter" as "material that readers find unpleasant". I'm saying, there are many reasons someone may react negatively to a given user interface element.

And I'm saying that a negative reaction isn't necessarily grounds for change. We don't bow to reader "preferences" by removing encyclopedic content on the basis that they find it unappealing. (And yes, this is encyclopedic content.)

It is not your role to judge the validity of those responses, and to decide that negative reactions due to # of clicks is acceptable, but negative reactions due to presence of certain words, or layout of UI elements, or size and spacing, etc is not.

Firstly, I didn't say (and don't believe) that readers' reactions to layout/UI elements are of no concern. Many significantly impact the site's usability. And when it comes to matters of aesthetics, I wrote that "all else being equal, of course we want to make the site as appealing as possible." But this isn't a proposal in which all else is equal; it's one in which article access is made more difficult.

Secondly, your claim that I've cited "negative reactions due to # of clicks" is inaccurate. I've cited increased difficulty accessing articles, not the resultant emotional response.

You're so focused on the fact that some piece of text may annoy the user,

I am?

and you seem so anxious to do anything in your power to ensure that the user will see that text

No, I want users seeking the non-primary article to see it.

in this one tiny little edge case (where 3-DABs can't be found).

If it's a "tiny little edge case", why are you so concerned about it?

Obviously, all of us believe that the issue is important enough to discuss/debate.

You obviously didn't read the link I attached

I'm familiar with the controversial content resolution.

- the wikipedia foundation specifically states that you should not astonish the user: "content on Wikimedia projects should be presented to readers in such a way as to respect their expectations of what any page or feature might contain"

I asked you to explain how encountering a link to an article about a subject with a similar name is "astonishing" and how this interferes with one's ability to utilize the encyclopedia in the manner intended. I await your response.

You've already cited a page explaining that such a hatnote prevents astonishment on the part of readers accidentally arriving at the wrong article.

What this means is, in an extreme case, is you cannot shock the user with an offensive picture in a place they don't expect just because you believe the encyclopedia is uncensored

The proposal pertains to textual links appearing in the standard location.

Let me propose a reductio-ad-absurdum hatnote case - for example, if there was a sex toy called Jesus, it would be absurd to put This article is about Jesus of Nazareth. For the sex toy, see Jesus (sex toy) - WP:IAR would apply and that hatnote wouldn't last 2 minutes.

I assume that you're referring to a hypothetical scenario in which the sex toy is notable enough to have a Wikipedia article and no other uses of the name "Jesus" (apart from those two) are.

I understand why many users would consider such a hatnote objectionable, but I wouldn't. A notable sex toy is as valid a subject for a Wikipedia article as any other. It isn't our place to pass moral judgement.

However, I realize that my argument wouldn't win the day. As you noted, there surely would be consensus to apply IAR by linking to a two-article disambiguation page instead. (I wouldn't agree with this outcome, but I would respect it.)

In other words, the community already is empowered to make exceptions to the rules when the circumstances dictate. Setting aside the question of whether the proposed solution ever is appropriate (and if so, when), I see no need to create special rules for "this one tiny little edge case".

Here is another reductio-ad-absurdum - if I proposed that all hatnotes should have red blinking links, I could then use your same logic to say "well, it doesn't impact their ability to click, they can easily ignore it, and it helps a minority users find the link they want - therefore, even if it annoys the f*** out of the users, we must keep the red flashing links"...

I would question the premise that such a change would positively impact (and not negatively impact) the link's accessibility.

Listen, you guys have tried really hard, I will give you credit, to distinguish between some sort of objective measures of interface utility (like # of clicks to get what you want, navigational metrics), and to separate that from subjective measures (dislike, distaste, annoyance, irritation, bothering, cluttering, confusing, bewildering, befuddling, readability, etc)...

While where to draw the line is subjective, in this context, "cluttered" describes a condition objectively different from "uncluttered" under anyone's standard. In other words, if a person is of the opinion that x links is the reasonable limit for a hatnote, he/she can point to a quantifiable difference between "uncluttered" (x links) and "cluttered" (> x links). But he/she can't point to such a difference between a single link to a disambiguation page and a single link to an article.

but I'm afraid the 'clutter' argument (we to DABs to avoid clutter)

...by which we mean something different from what you mean.

and dozens of other policies all over wikipedia (like deletion of useful links in disambig pages - which results in not a 2-click or 3-click solution to get what I want, but a dead end!)

Again, at no point have I claimed that Wikipedia seeks to indiscriminately provide direct links to articles or that no other valid considerations exist. I've stressed this over and over, but you continue to attack that straw man.

Indeed, we have rules indicating when it is/isn't appropriate to provide a link to a particular article. This debate isn't about whether to to provide a link to a particular article; it's about whether to present it directly or shift it to a separate page. Either way, the link exists (and the hatnote exists specifically for that reason).

What if something doesn't annoy them, what if it just bothers them? What if it doesn't bother them, but just distracts them? What if it doesn't distract them, but renders the page less easy to read?

Do you assert that the hatnotes you seek to eliminate "render [pages] less easy to read"? Such a possibility could be analyzed via a user study (as opposed to polling readers to determine what they "prefer"), but do you have evidence that this is an anticipated outcome? (We could conduct a study to determine whether sentences beginning with the letter "Z" or photographs of penguins render pages less easy to read, but we probably wouldn't without a fair degree of suspicion to that effect.)

What if it's not less easy to read, but lab experiments show their eyes linger for 0.3s longer when the DAB is not neutral?

See above. (Also, there would be other factors to consider.)

What if it's not less easy to read, but they find it more cluttered?

1. I again ask you to explain how it's possible for a one-link hatnote to be more cluttered than a one-link hatnote is.

2. We seek to prevent clutter because it makes a page less easy to read.

It's a slippery slope, once you start judging which opinions are valid and which are not, you'll never be able to draw a defensible line.

Again, it's extremely simple: "Does the issue adversely affect readers' ability to utilize the encyclopedia as intended?" I (and others) assert that the hatnotes in question don't. I welcome evidence to the contrary, but you instead propose that we poll readers on their "preferences" (i.e. what would please them most).

I don't even assert that it's unreasonable to opine that we should place such concerns on equal footing. I merely dispute your assertion that the proposal's opponents have failed to "draw a defensible line". You needn't agree that the distinction in question should play any role in determining articles' content, but please stop pretending that it doesn't exist. —David Levy 08:15, 8 May 2012 (UTC)

Revised proposal

For 2-DAB:

• original: Some disambiguation pages with "(disambiguation)" in the title list only two meanings, one of them being the primary topic. In such cases, the disambiguation page is not strictly necessary, but is harmless. The recommended practice in these situations is to place a hatnote on the primary topic article to link directly to the secondary topic.
• new: Some disambiguation pages with "(disambiguation)" in the title list only two meanings, one of them being the primary topic. In such cases, the disambiguation page is not strictly necessary, but is harmless. The recommended practice in these situations is to place a hatnote on the primary topic article to link directly to the secondary topic, unless consensus dictates that the readers are better served by having the secondary topic on the disambiguation page, so as to not distract readers of the primary topic.

As before, please feel free to edit the above, or copy/paste. Also, please don't 'support' or 'oppose' (unless you don't want to change the status quo), this is a discussion, so is better to provide a change that you agree with instead of saying you don't like the idea above.--KarlB (talk) 05:54, 7 May 2012 (UTC)

Oppose. This is perfectly appropriate response to a proposal, provided that it's accompanied by a rationale (see below). Your instruction to "provide a change that you agree with instead" is based on an assumption that it's unreasonable to support the status quo.

I oppose this proposal because I disagree with the premise that the relevant scenario (in which direct links to non-primary topics "distract readers of the primary topic") exists. Consensus to that effect certainly hasn't been established. —David Levy 06:32, 7 May 2012 (UTC)

You're right; if you are happy with the status quo, and won't agree to any changes whatsover, oppose is a reasonable thing to say here, and I'm sorry that I didn't capture this appropriately; I've now tried to reword. Now that your opinion has been registered, it would be great if you would allow the others to continue to see if we can come to a consensus on a change, which would then in the future be put up for a discussion to a broader audience. Thanks!--KarlB (talk) 19:24, 7 May 2012 (UTC)

I didn't say that I "won't agree to any changes whatsover", nor would I ever adopt such an attitude. But you're attempting to address a problem that I don't agree exists, so I'm unable to propose an alternative solution.

Of course I'll "allow" the pursuit of a consensus-backed change to continue. I couldn't stop it if I wanted to (which I don't). But I hope that you aren't asking me to cease participating in the discussion. —David Levy 22:00, 7 May 2012 (UTC)

Honestly, it really seems like we're going in circles with this whole thing. I appreciate an in depth discussion, but after trying to process everything that's been said, I'm still wondering why this is such a problem. I personally don't find hatnotes distracting and I still don't feel compelled to devise any type of experiment to see if they are for other readers. From my view, it seems as if the standard was set with WP:SIMILAR and even if there is only 1 other identically titled article, then there is no need for a dab page. There comes a point where if a (small) group of people are not on board with the standard then they will just have to deal with it. There seems to be no need to seemingly go out of our ways just to cater to people to want to view "For other uses see Jennifer Lopez (disambiguation)" instead of "For the American television meteorologist, see Jennifer Lopez (meteorologist)". Either way there will be a hatnote, so I cannot support those situations where there is only 1 other article with the same name and still creating a dab page just for what appears to be an extremely small issue of giving more notoriety to a lesser known article. That's not a issue for me and I can't imagine this causing virtually any readers any distress at all. In regard to the proposal, I can't yet see an alternative to appease the, if any, amount of people who want to see a dab for everything. If we were to change this only because certain users feel distracted, then where does it end? What if they are distracted by infoboxes? What if too many blue links distracted them? What if dab pages distracted them? My answer is you can't please everyone even if there are a few out there who are legitimately distracted. Either provide them with an option if possible, or else they will just have to deal with it. It may sound bleak, but considering how complex and deep as WP is, I'm okay with some people not getting everything they want, especially if it's only for an ostensibly negligible benefit. RoadView (talk) 06:50, 7 May 2012 (UTC)

Those are some pretty poor comparisons, I don't think I am the only one who feels infoboxes have some serious disadvantages, WP:IBX is neutral on whether or not they should be included, and WP:OVERLINK already mandates against excessive blue-linking. SpinningSpark 11:22, 7 May 2012 (UTC)

• Oppose, still censoring (removal of information from one article, even if not the entire encyclopedia, and "being objectionable" is generally not sufficient grounds for removal or inclusion of content). Here's a proposal: "Some disambiguation pages with "(disambiguation)" in the title list only two meanings, one of them being the primary topic. In such cases, the disambiguation page is not strictly necessary, but is harmless. The recommended practice in these situations is to place a hatnote on the primary topic article to link directly to the secondary topic, unless consensus at the secondary topic decides that its readers are better served by clicking through a disambiguation page first to avoid putting the specific title of topic on the primary topic article." I also think that's pointless, though. -- JHunterJ (talk) 14:55, 7 May 2012 (UTC)

  I agree that it would be pointless, given the fact that no one (including proponents of change) has claimed that such a circumstance is possible. —David Levy 22:00, 7 May 2012 (UTC)

• Oppose, and I'll propose an alternate approach. If the problem is that the link distract readers because it doesn't have context, lets provide it with a context: instead of the common tophat "For topic X see Y", write the slightly more elaborated Wikipedia has an article about topic X" that will inform readers of why they're reading about a different subject in the hat note.

Thus the guideline would read: "Some disambiguation pages with "(disambiguation)" in the title list only two meanings, one of them being the primary topic. In such cases, the disambiguation page is not strictly necessary, but is harmless. The recommended practice in these situations is to place a hatnote on the primary topic article to link directly to the secondary topic. If consensus dictates that such stand-alone link would be distracting, provide context for the link with a wording in this form: Wikipedia also contains an article with the same title about Example". Diego (talk) 15:39, 7 May 2012 (UTC)

• comment as I said above, please don't oppose - this is not a vote, it is a discussion - there is no need to say 'oppose' (Unless, as David points out, you like the status quo are are unwilling to consider any modification to it).--KarlB (talk) 19:24, 7 May 2012 (UTC)

This is customary. Bolded words are never intended nor used as votes, but to provide a quick and clear summary of each position in discussions where a change requiring consensus is decided. These are always not-votes, and that's by policy. Diego (talk) 20:36, 7 May 2012 (UTC)

Whether it is customary or not, I'm not sure; there are plenty of discussions I see without 'oppose' etc. In any case, I'm trying to seed a proposal, not propose a final approach - so I feel at this point 'support' and 'oppose' votes will only serve to polarize (this is not AfD or CfD, for example) - if you agree we can move forward on the discussion with out them, and then if we come to some conclusion, then bring it as a more well-formed proposal to a broader group - but let's see if we can find some consensus here first.--KarlB (talk) 21:04, 7 May 2012 (UTC)

I'm not unwilling to consider modification. I simply disagree that the purported problem exists, so I'm unable to propose an alternative solution. —David Levy 22:00, 7 May 2012 (UTC)

It doesn't matter that you tagged your proposal with the request not to oppose it. I oppose the proposal. If you would rather not see opposition, you should avoid making proposals. How about this: please don't oppose leaving the hatnote guideline the way it is, and leaving the navigation to a single ambiguous topic unimpeded. -- JHunterJ (talk) 04:01, 8 May 2012 (UTC)

wow. just... wow. You are so fun to talk to. Thanks for that. Do I sense some anger here? You're not obligated to respond here you know. --KarlB (talk) 04:48, 8 May 2012 (UTC)

You probably do sense anger, but it's not mine. If I point out the absurdity of your requests by spinning them around, it's not out of anger, but out of incredulity (aka "wow. just... wow") at the absurdity. -- JHunterJ (talk) 13:04, 8 May 2012 (UTC)

Ive just gone over most of the recent discussion. Kudos to KarlB for doing a fine job of single-handedly dealing with older/wiser (aka Bkonrad) and David Levy, and IMHO effectively dismantling their arguments. Id like to add a few points, and focus on some examples, cause I think were about to breach the impasse here. Ive mentioned this issue three or maybe four years ago (google:"trivial hatnotes"), but it didnt go very far. In fact I think I was the one to coin the term "hatnote" in this context. The current issue began a couple months back when I emailed Jimbo about the trivial hatnote at the Johnny Cash article, and he agreed with my general idea and suggested I raise it on his talk page. He said it would be "fun." I postponed doing so for a couple months because I had just got back from a year-long ban, and wasn't looking to argue with people. So, finally I brought it up on Jimbo's talk page, and here we all are.

I think we've got some common ground. First of all, the most productive discussion seems to center around examples. Let's look at the Johnny Cash article. It used to have a hatnote pointing to Johnny Cash (song). It now has an otheruses tag and a disambiguation page. There are two things to note here. One, Im sure no one wants to undo the otheruses tag, and re-add the hatnote to the song. Why? Perhaps its because the disambiguation now has more than two terms. Or perhaps people realize that the less-notable song doesn't need linking on the highly-notable bio article. It doesn't really matter which view you subscribe to - in this case, the matter was settled. Which brings us to the second matter: How did certain cases get solved? (Someone created the Johnny Cash disambiguation after searching for more terms to add, so that's not a clear-cut example for my purposes here.) The sexual harrassment article IIRC had a hatnote pointing to some TV show. The sexual harrassment article now has a standard otheruses tag and a disambiguation page. But I did not search for third and fourth terms, first, I simply removed the hatnote and added the otheruses tag, creating a 2-term disambiguation page. I did it wrong! - according to opposers here. And yet the result is the same - some other tinkering editor came along and found more terms to add to the sexual harrassment disambiguation page.

The important thing to note here is that I did it the "wrong" way, and yet the result is a positive one. Hence I'm inclined to think that this mode of process can work well for other articles as well. Some article has a hatnote which doesn't quite work? Add an otheruses tag, create a 2-term disambiguation page! If you are inclined, find a third term, if not, don't worry, there are other helpful and fastidious editors who will come along and add more terms to the new disambiguation page. They can come along sooner or later - it doesn't matter. The 2-term disambiguation page is like a stub - someone eventually will come along and add more material. Thats the way wikipedia works. Every article at one point used to be a stub, like this one. Anyway, I think that covers it. Maybe it would help now if we came up with lists of articles which best represent our case. It may be we all agree more than it it would seem. regards. -Stevertigo (t | c) 04:25, 8 May 2012 (UTC) PS: It was actually someone else who created the sexual harrassment disambig page. I must have been thinking of something else. -Stevertigo (t | c) 04:50, 8 May 2012 (UTC)

Thanks Steve. I fixed a few here: [11] and Statue. The robbery one was a mess - omg - I can't believe how long it had been like that. In fact, I have yet to find a case where a DAB page can't be created. Perhaps someone else will present some good examples...--KarlB (talk) 05:03, 8 May 2012 (UTC)

Some article has a hatnote which doesn't quite work?

There's no consensus for the position that the hatnotes in question "[don't] quite work".

Add an otheruses tag, create a 2-term disambiguation page!

And prepare to have the hatnote reverted if it's spotted by someone familiar/compliant with the relevant rules before a third valid usage (which might or might not exist) is added to the disambiguation page. —David Levy 08:15, 8 May 2012 (UTC)

I've also just gone over the discussion. Kudos to Bkonrad and David Levy for effectively dismantling the ill-conceived hindrances proposed by Karl.brown and Stevertigo. If someone mistakenly creates a two-term disambiguation page when there's a primary topic, fix it by restoring the direct hatnote! If someone else can find a third ambiguous topic, no worries, they can add it to the orphan disambiguation page or recreate it if it was deleted in the meantime. That's the way Wikipedia works. -- JHunterJ (talk) 05:44, 9 May 2012 (UTC)

What you propose is the deletionist view - of judging someone's work (a two-term disambiguation page) as something improper and then deleting it, or at least orphaning it. The deletionist never considers rolling up their sleeves and just being helpful - finding a third term for example. What I propose is not deletionist, its a simple and creative way to deal with a minor editorial problem. -Stevertigo (t | c) 08:49, 9 May 2012 (UTC)

And what do you propose for the case where there is not a valid third term for disambiguation, i.e. a third article with that title? In such cases, creating the DAB page is an artificial structure that impairs navigation. That could only be justified in an ignore-all-rules way after evaluating each specific case, not as the guideline recommendation for all similar cases. Diego (talk) 09:20, 9 May 2012 (UTC)

P.S. "Deletionist" is not (or shouldn't be) an insult; some see it as clean up and organization. Though you're right that it's "judging someone's work as something improper". Given that all editor's work is donated under a free license for everybody to judge and further edit, that is not seen as a problem around here. Diego (talk) 09:24, 9 May 2012 (UTC)

@Stevevertigo, why codify laziness in the guideline? Those who find such "trivial" hatnotes bothersome should make the effort to create a valid disambiguation page rather than making dubious assumptions and expecting that others will do the work for you. older ≠ wiser 11:25, 9 May 2012 (UTC)

Older/wiser, why put unnecessary conditions on disambiguation creation? Disambiguation pages are extremely helpful. -Stevertigo (t | c) 00:19, 10 May 2012 (UTC)

Disambiguation pages are only one form of disambiguation. Hatnotes are preferred where there is only one other ambiguous topic. You don't agree. No problem. No one is stopping you from creating such pages, but be prepared for such pages to be subject to speedy deletion ({{db-disambig}}). older ≠ wiser 00:25, 10 May 2012 (UTC)

That's of course what we are discussing - changing the deletionist status quo to accomodate creation of 2-term disambiguation pages, particularly in cases where the less-notable hatnote is defacing a more-notable artice. -Stevertigo (t | c) 00:37, 10 May 2012 (UTC)

Well, I for one remain thoroughly unconvinced by your suggestion. older ≠ wiser 00:48, 10 May 2012 (UTC)

Diego wrote:"And what do you propose for the case where there is not a valid third term for disambiguation, i.e. a third article with that title?" - The point is that we just don't always know, when creating a disambiguation page, just how many other terms there are out there which can go into that disambiguation page. Keep in mind that editing is best thought of as an eventualist concept - eventually someone with the knowledge will come along and add something previous editors were incapable of adding. We can't put the onus on the creator of the disambiguation page to fill up the page with links. Regards, -Stevertigo (t | c) 00:50, 10 May 2012 (UTC)

But we can put the onus to create the disambiguation page on the eventual editor with the knowledge of a new term. There in no more work involved for anyone (it only shifts the work from the fist editor to the second), and it's best for the readers in the meantime. (Also what's the usefulness of a disambiguation page with two links, one of which was already visited?) Diego (talk) 05:44, 10 May 2012 (UTC)

The disambiguation page is only used in the 2-term case because the hatlink is trivial and thus improper. We've talked about a few examples where this was the case: the Johnny Cash, atheism, or sexual harassment articles etc. -Stevertigo (t | c) 06:51, 11 May 2012 (UTC)

We've talked of several alternatives that wouldn't require a 2-term disambiguation page. Creating a useless dab page when there are better ways to "solve" the perceived improperness is absurd. Diego (talk) 08:25, 11 May 2012 (UTC)

The only other way is to create a 3-term disambiguation page - something which often can be done, but we shouldn't put the onus on editors to add a third term. Sometimes its not easy finding a third term. -Stevertigo (t | c) 17:30, 11 May 2012 (UTC)

So we agree that the page should be created only when there are already three terms available? (And see the discussion above for the other other ways that are not "creating a 3-term disambiguation page"). Diego (talk) 17:37, 11 May 2012 (UTC)

No, I'm proposing that creating a 2-term disambiguation is OK, but given a grace period, after which some editor can revert it. -Stevertigo (t | c) 17:39, 11 May 2012 (UTC)

...in the interest of solving a problem whose very existence lacks consensus. —David Levy 19:46, 11 May 2012 (UTC)

The disambiguation page is only used in the 2-term case because the hatlink is trivial and thus improper.

There's no consensus for the position that "the hatlink is trivial and thus improper". I realize that you're attempting to establish such consensus, but you're begging the question. —David Levy 19:46, 11 May 2012 (UTC)

David wrote: "There's no consensus for the position that the hatnotes in question "[don't] quite work"." - Maybe not, but that's what we are discussing. In several cases, some us us have found agreement that certain hatnotes are improper due to a substantial difference in notability between article and hatnote link. This difference in notability has been characterized in a number of ways, such as "trivial" and "spam." The characterization is not important - the important thing is that most people seem to agree that such hatnotes dont belong - as evidenced by the fact that the examples we've given remain unchanged. No one has reverted Johnny Cash or sexual harassment back to their former states, and there is good reason for that. David wrote: "And prepare to have the hatnote reverted if it's spotted by someone familiar/compliant with the relevant rules before a third valid usage (which might or might not exist) is added to the disambiguation page." - Perhaps the solution to our problem might be to give a certain grace period to 2-term disambiguations. Maybe 3 days should be enough. Regards, -Stevertigo (t | c) 17:37, 11 May 2012 (UTC)

Maybe not, but that's what we are discussing.

And instead of accepting that this discussion hasn't established consensus for your position, you're encouraging editors to ignore the outcome and replace direct links with links to two-article disambiguation pages.

In several cases, some us us have found agreement that certain hatnotes are improper due to a substantial difference in notability between article and hatnote link.

"Some of [you]" ≠ consensus.

That there's "a substantial difference in notability" is why one topic's article resides at the base title and the other topic's article is merely linked from it.

This difference in notability has been characterized in a number of ways, such as "trivial" and "spam." The characterization is not important - the important thing is that most people seem to agree that such hatnotes dont belong

That's an outright falsehood.

- as evidenced by the fact that the examples we've given remain unchanged.

You've given no examples in which two-article disambiguation pages currently exist. That they started out that way proves nothing. I've encountered instances in which vandalism persisted in articles for months or years. That isn't evidence of consensus for the inclusion of vandalism in articles. It just means that it went unnoticed.

Perhaps the solution to our problem might be to give a certain grace period to 2-term disambiguations. Maybe 3 days should be enough.

Firstly, you're again relying on the assumption that the direct links in question are bad and should be replaced. There's no consensus for that position, so why should we provide a grace period for an attempted solution to a problem that we don't agree exists?

Secondly, how do you propose we track the two-term disambiguation pages that you're encouraging editors to create? —David Levy 19:46, 11 May 2012 (UTC)

David wrote: "That's an outright falsehood." - What is false about it? David: "Firstly, you're again relying on the assumption that the direct links in question are bad and should be replaced. There's no consensus for that position, so why should we provide a grace period for an attempted solution to a problem that we don't agree exists?" - You can agree to find a compromise solution that works for both camps. David: "Secondly, how do you propose we track the two-term disambiguation pages that you're encouraging editors to create?" - We can create a 2-dab template with a category link, and a notice to the creator that they have x number of days to turn the 2-dab into a 3-dab. -Stevertigo (t | c) 20:32, 11 May 2012 (UTC) PS: David, note template:gn -Stevertigo (t | c) 20:43, 11 May 2012 (UTC)

What is false about it?

You've asserted that "most people seem to agree that such hatnotes don't belong", which reflects neither our longstanding practices nor any relevant discussion (including this one).

You base this claim on the fact that some direct links were replaced with links to two-article disambiguation pages, which weren't removed prior to the disambiguation pages' expansion. As I noted, this proves nothing. Improper edits frequently go unchallenged simply because they go unnoticed.

You can agree to find a compromise solution that works for both camps.

Under our current policies and guidelines, the direct links in question aren't recognized as problematic. Under the suggested "compromise", they would be. This fundamental shift, in no way reflecting consensus, would invite all sorts of solutions to a "problem" that the community doesn't agree exists (including scavenger hunts resulting in questionable links added solely to justify the disambiguation pages' existence).

We can create a 2-dab template with a category link, and a notice to the creator that they have x number of days to turn the 2-dab into a 3-dab.

How would we go about enforcing this template's use?

PS: David, note template:gn

Thanks! I'd been meaning to create such a template. (A similar one was brought to my attention, but its style differed significantly.) See {{gi}} for an italic version. —David Levy 21:28, 11 May 2012 (UTC)

Just because you don't see there is a problem, doesn't mean there isn't one. We've implemented 2-term disambiguations several times now in the context of this discussion, and none thus far have been reverted. If you feel that trivial hatnotes belong at any of the articles we've discussed, please go ahead and revert back to their previous state. The fact that you don't do so means that you tacitly if not overtly agree with the changes. Hence we already have some consensus, or indication thereof. Hence we have a basis for a workable compromise. -Stevertigo (t | c) 00:41, 12 May 2012 (UTC)

Just because you don't see there is a problem, doesn't mean there isn't one.

Agreed. Neither my opinion nor yours is sacrosanct. That's why we're having a discussion. And thus far, there's no consensus that a problem exists (and even less agreement that linking to a two-article disambiguation page is a valid solution).

We've implemented 2-term disambiguations several times now in the context of this discussion, and none thus far have been reverted. If you feel that trivial hatnotes belong at any of the articles we've discussed, please go ahead and revert back to their previous state.

1. The question of whether I feel "that trivial hatnotes belong" is a loaded one. I dispute your assertion that the hatnotes in question are "trivial".

2. In reading this discussion and following links contained therein, I've encountered only disambiguation pages covering more than two articles. (Some originally covered two articles, but they were expanded before I saw them.) I evidently overlooked the two-term disambiguation pages to which you refer, so please point me to them. Thanks! —David Levy 01:38, 12 May 2012 (UTC)

Stevertigo, you're approaching this discussion from a wrong angle. Since you and KarlB are the only ones wanting to change the guideline, you're the ones who should give up something when suggesting a compromise solution; everybody else are fine with the status quo. So far your proposed solution hasn't changed one bit from the original proposal that was already rejected and that has nothing of compromise. If you really want a compromise, offer something that we can agree upon; this won't work the other way around, because the default position is leaving everything as is. Diego (talk) 09:38, 12 May 2012 (UTC)

David, the point is that what started out as 2-term disambiguations soon enough became 3 and more. KarlB and I are in agreement in that 2-term disambiguations should contain more terms, which can be added once a 2-term disambiguation is made. In practice we find that most 2-term disambiguations can easily be made into 3-term or more. The key here is that though 2-term disambiguations should contain more terms, they don't need to contain more terms for sake of an arbitrary rule. He and I, finding some limited consensus here, will continue to remove trivial hatnotes from non-trivial articles out of a good faith effort to improve the quality and readability of Wikipedia articles.

Diego, I appreciate your point. The compromise we can make is this, that 2-term disambiguations should gather new links within a certain time frame - say 2 weeks. If no new links have been found, the disambiguation can be deleted. This is a kind of grace period against what otherwise would be an automatic deletion based on an arbitrary rule. Hence I (and maybe KarlB will join me on this) compromise on the idea that 2-term disambiguations should be made into 3-term disambiguations - I agree that 2-term disambiguations cannot be left as is. In turn, we would like it if your side would agree that 2-term disambiguations can exist for a limited amount of time, to exist as a kind of disambiguation stub, and to gather more terms if available. -Stevertigo (t | c) 06:05, 13 May 2012 (UTC)

For this to be a compromise, I can accept that the 2-term DAB exists if the link to the other article is not removed from the tophat until the dab page is compliant with WP:SIMILAR and WP:TWODABS guidelines (i.e. either deleted or make into a 3-or-more-term page with only valid links). I don't find any value in creating that page only to be likely deleted, but it is a fair compromise for me; I don't know what other editors will think of it, though. Also I think this compromise should be kept informal and not encoded in the guidelines. Diego (talk) 10:14, 13 May 2012 (UTC)

P.S. There's no consensus here, limited nor otherwise, to remove hat notes when there are only two terms to disambiguate; hiding behind an imaginary consensus to justify deleting the top hat against guidelines could only be seen as bad faith, in special after this discussion. See Free base and Freebase for an example where a 3-term can't be created as there are no other articles ambiguous with the term. In that case creating a disambiguation page for a limited time would be just useless; but removing the hat note meanwhile would be truly disruptive. Diego (talk) 18:31, 13 May 2012 (UTC)

P.P.S. And see, this is what cannot be done in any case. The DAB page you created contains an unsourced entry which is against WP:Verifiability, and the target article didn't discuss . I've reverted the hatnote removal and PRODed the useless page for deletion in accordance with the currently upheld consensus in WP:TWODABS. Diego (talk) 19:38, 13 May 2012 (UTC)

While I agree with you completely that two-term dab pages should not be created -- and that editors who knowingly create them in defiance of established practice deserve criticism for failing to attempt even the most rudimentary checking for ambiguous titles -- editors who come across such pages should also perform such checking before prodding the page or tagging for speedy deletion. I found at least on other exact match for the Sufi (disambiguation) dab page and added {{lookfrom}} and {{intitle}} because the term is commonly used as an adjective, which in some contexts could be referenced elliptically as Sufi. older ≠ wiser 20:36, 13 May 2012 (UTC)

True, I didn't check for other articles (that's why I proded instead of requesting speedy deletion, to allow for a valid link to be found). That doesn't invalidate my point though, which is that the page was created with an incorrect entry and the link in the hatnote should hat not been deleted until the dab page included more than two valid links to articles. Diego (talk) 21:26, 13 May 2012 (UTC)

David, the point is that what started out as 2-term disambiguations soon enough became 3 and more.

You also claimed that I've declined to remove links to two-article disambiguation pages, citing this as evidence that I tacitly condone these links' existence (and that this "consensus" constitutes grounds for compromise). Again, please bring to my attention the relevant articles.

KarlB and I are in agreement in that 2-term disambiguations should contain more terms, which can be added once a 2-term disambiguation is made.

That isn't a safe assumption. In some instances, no valid third topic exists.

If there were consensus that tentatively linking to a two-article disambiguation page solves a problem, it might be justified. There is no such consensus. On the contrary, there's longstanding consensus that it makes the encyclopedia worse.

The key here is that though 2-term disambiguations should contain more terms, they don't need to contain more terms for sake of an arbitrary rule.

The rule isn't arbitrary. It exists because two-article disambiguation pages are unhelpful (and linking to one instead of linking directly to an article is harmful).

He and I, finding some limited consensus here, will continue to remove trivial hatnotes from non-trivial articles out of a good faith effort to improve the quality and readability of Wikipedia articles.

You know perfectly well that no such consensus ("limited" or otherwise) exists. In your original reply, you vowed to continue "regardless of the consensus here". You've plainly stated that you intend to enforce your preference (by removing hatnotes that you've unilaterally deemed "trivial") no matter what the community thinks.

I'm reluctant to discuss compromise with someone issuing such a declaration, but if you agree to honor consensus, I'm willing to discuss the type described by Diego.

However, you'll need to explain why the same task (the search for a third valid topic) can't be carried out just as easily before creating a disambiguation page. —David Levy 16:07, 13 May 2012 (UTC)

Continued..

Thank you both for your thoughtful responses. I appreciate your position, that there are some articles which don't seem to have third terms and just can't be disambiguated. Diego offers up a freebase/free base as an example, but I can think of one other term (with which I will form a 3-dab after I'm done here). The sufi case is interesting as Diego and older/wiser approached that case in different ways: Diego sought to delete the dab, but older/wiser thought it sensible to try see if any more terms could be added. The article currently has a hatnote pointing to the song "Sufi", which I think is improper just for the triviality of the thing. Cmon, this is an article about a major branch of a world religion. If there was a song called "Islam" would we want a hatnote to an "Islam (song)" on the Islam article? I don't think so. Hence I use the term "trivial" to describe a relative difference in notablity and even importance. What if we deleted the Sufi (song) article, would we be so much worse off? What if we deleted the Sufi article? See the difference? If we deleted the Sufi article we almost couldn't be called an encyclopedia anymore. If we delete the Sufi (song) article, almost no one would even notice. That my friends speaks to a serious difference in notability. Hence we make the effort to do a proper disambiguation. And as you will see with the free base/freebase case, there are more often than not other terms to add to the disambiguation. Regards, -Stevertigo (t | c) 23:01, 13 May 2012 (UTC) PS: So the compromise is this - don't PROD 2-dabs without attempting first to create a 3-dab, as its not just the 2-dab creator's responsibility to create proper 3-dabs - some responsibility is shared with the PROD-er. -Stevertigo (t | c) 23:12, 13 May 2012 (UTC)

It is indeed the responsibility of the person wanting to create a disambiguation page to ensure that it is a valid disambiguation page (i.e., that it disambiguation more than two existing page). Also, for the record, I have absolutely no problem with a hatnote on what you consider a "serious" topic to what you deem a trivial topic if that other topic is the only other existing ambiguous article. older ≠ wiser 23:33, 13 May 2012 (UTC)

There currently is no disambiguation for Woody Allen. Perhaps I can write a song called "Woody Allen," fill it full of silly lyrics, and publish it via iTunes or Spotify etc. Suppose it sells a few copies, and someone writes an article. Would you then be obligated to add a hatnote link to my silly song article from the Woody Allen bio article? What if I were the one to add the hatnote , in effect spamming the notable bio article with a link to my less-notable song? -Stevertigo (t | c) 23:47, 13 May 2012 (UTC)

Disambiguation is not concerned with notability. If the article satisfies general standards for a Wikipedia article and it is the only other article with that title, then yes, I would put a hatnote to it on the actor/directory's article. older ≠ wiser 23:52, 13 May 2012 (UTC)

Actually I forgot to mention the alternate title of the song is "So Very Young," so the link looks like "Woody Allen (So Very Young)." What then? Do you link? Do you create a hatnote that says: "This article is about the person. For the song, see Woody Allen (So Very Young)" ? Stevertigo (t | c) 23:59, 13 May 2012 (UTC)

Now you're just being annoying. If the article is commonly known as "Woody Allen", presumably supported by references in the article, then the title is ambiguous and a hatnote would be needed. If it is not ambiguous or only partially so, then no hatnote. And the hatnote would probably be a simple, For the song by a sick mind, see Woody Allen (So Very Young) (well maybe without the commentary on the mental state). older ≠ wiser 00:13, 14 May 2012 (UTC)

You've yet to accept the part of the compromise where the direct link between articles is not deleted until the grace period for the DAB page expires. You have to give something for there to be a compromise, and you still haven't moved an inch from your initial position. Diego (talk) 05:42, 14 May 2012 (UTC)

The sufi case is interesting as Diego and older/wiser approached that case in different ways: Diego sought to delete the dab, but older/wiser thought it sensible to try see if any more terms could be added.

No one disputes that such an act is helpful. But that doesn't mean that an invalid disambiguation page should have been created in the first place, let alone that others are obligated to do the work that the creator failed to.

Of particular concern was your inclusion of an inappropriate item in an attempt to justify the page's existence. (And I seem to recall you doing this in the past.) I assume that you believed it to be acceptable, but you obviously dug it up purely because you wanted to find a third item. As I noted above, adopting the position that "trivial" article links should be replaced with links to disambiguation pages would encourage this behavior.

The article currently has a hatnote pointing to the song "Sufi", which I think is improper just for the triviality of the thing.

Yes, you've made your opinion abundantly clear. It's equally clear that there's no such consensus, but you've declared that you intend to enforce your preference anyway.

In this instance, the article now links to the disambiguation page, which has been made valid. There's no reason why that couldn't have been done before replacing the direct article link.

If there was a song called "Islam" would we want a hatnote to an "Islam (song)" on the Islam article?

If the song were notable enough to have a Wikipedia article and no other subjects called "Islam" (apart from the song and the religion) were, yes. Of course, that isn't the case.

I don't think so.

And you're entitled to your opinion. You simply aren't entitled to override consensus.

What if we deleted the Sufi (song) article, would we be so much worse off? What if we deleted the Sufi article? See the difference? If we deleted the Sufi article we almost couldn't be called an encyclopedia anymore. If we delete the Sufi (song) article, almost no one would even notice. That my friends speaks to a serious difference in notability.

That such a difference exists is why the base title (Sufi) redirects to the article about the religion (while the song's article resides at a title with parenthetical disambiguation).

Hence we make the effort to do a proper disambiguation.

Your idea of "proper disambiguation" differs from that of the community at large.

And as you will see with the free base/freebase case, there are more often than not other terms to add to the disambiguation.

I await your explanation of how it's easier to add a third article link to a two-article disambiguation page than it is to simply create a valid disambiguation page from the start. Keep in mind that anyone visiting the base page will see the same two article titles (the one at which he/she has arrived and the one linked via the hatnote).

I also await your pointers to the hatnotes linking to two-article disambiguation pages that you claim I've declined to remove, citing this as evidence that I tacitly condone these links' existence (and that this "consensus" constitutes grounds for compromise). This is my third request. (You've ignored the first two.) —David Levy 02:44, 15 May 2012 (UTC)

Please see Wikipedia's policies WP:NPA and WP:AGF. -Stevertigo (t | c) 00:22, 14 May 2012 (UTC) PS: The point is that content is relevant. We could not put such a link on the Woody Allen bio article because it references a notable controversy, and hence becomes a kind of BLP issue. So what you say about linking to anything as long as there arent three terms just isn't true in all cases. Links must sometimes be judged for their content. -Stevertigo (t | c) 00:26, 14 May 2012 (UTC)

Sorry, I don't see your point. If the song is notable and ambiguous, disambiguation is needed and I would strongly oppose censoring the link from a hatnote if it is the only other ambiguous article. If the song is in fact commentary on controversies involving the subject, then it is likely more appropriate to reference somewhere within a section of the article dealing with the controversy or with derivative cultural impact or even a see also section. older ≠ wiser 01:28, 14 May 2012 (UTC)

If the song is in fact commentary on controversies involving the subject... - then it may not get such a treatment as you propose. I agree with you in general that hatnotes should be migrated to sections, but in fact if its controversial, there may be a strong impetus to remove such a link altogether. The point is that we can imagine certain kinds of hatnotes which are just not acceptable, and you seem to agree on that point: In the case I described, hatnotes can be unacceptable because of their controversial content. In most cases however, we cover up trivial hatnotes by creating disambiguations with three or more terms. If the song is notable and ambiguous, disambiguation is needed and I would strongly oppose censoring the link from a hatnote if it is the only other ambiguous article. - I understand that you feel that disambiguations are simply navigational. I agree that in part, such links are navigational. But my position is also that links contain words which which have meaning, and we therefore must consider the content of such links sometimes. Hatnote links are only partly navigational - they are partly editorial. We make editorial decisions when editing articles. In fact quite often we use 3-term disambiguations to cover up what would otherwise be what I would consider a virtual defacing of the article. We are first and foremost an encyclopedia - that doesn't mean we put links anywhere because it makes for quicker hyperlinking, it means being encyclopedic and judicious about what links are displayed where. There is no shortage of possible examples, such as my hypothetical, where even a 2-dab would be preferable to displaying an inappropriate link. -Stevertigo (t | c) 01:57, 14 May 2012 (UTC)

The point is that we can imagine certain kinds of hatnotes which are just not acceptable, and you seem to agree on that point. No. In the hypothetical situation you described, it was unclear whether the other title was actually ambiguous (if it was commonly known as simply "Woody Allen"). You overlooked my first statement: If the article is commonly known as "Woody Allen", presumably supported by references in the article, then the title is ambiguous and a hatnote would be needed. My other statement that you interpreted as agreeing with you was predicated on conditional statements. older ≠ wiser 23:35, 14 May 2012 (UTC)

The main problem is, there is an uneven standard being applied. The defenders of the hatnote claim they are doing it to provide navigational convenience for the user, and that the users lose out when they have to click twice. However, at the same time, links are being removed from disambig pages by these same people, because they don't fit a narrowly defined definition of what is allowed there. In any case, I agree with the proposal; allow creation of 2-DAB pages, and give a few weeks or month to find appropriate terms to add. In my experience, it can always be done, but perhaps not by the editor who created the DAB page; the existence of the DAB page encourages other editors to come and add. Thus, those who have an itchy trigger finger could simply agree to wait a few weeks before PROD'ing a 2-DAB page. What will it take to just end this discussion and get back to improving the encyclopedia? --KarlB (talk) 02:25, 14 May 2012 (UTC)

What will it take to just end this discussion and get back to improving the encyclopedia?

Well, those who are easily offended could simply agree to wait to remove the hatnote until there's a valid disambiguation page in place. Diego (talk) 06:03, 14 May 2012 (UTC)

Right. The "standard" (that is, the current consensus and guidelines) is even and being applied evenly. There are no "defenders of the hatnote", unless you are referring to the current consensus to provide navigational convenience to the user. Invalid links are indeed being removed from disambiguation pages, as invalid links have been removed from disambiguation pages long before this discussion. The definition of what's allowed there is defined, but simply (not narrowly). Your preferred definition would be over-broad, and does not have consensus. There is trivial effort needed to add valid links to 2-DAB pages (and when you have found valid entries, they have remained). I would say we could give a few minutes or an hour to find appropriate terms to add. In my experience, it cannot always be done: some titles have 1 topic, some have more than 2, but some topics can indeed have only 2 distinct topics on Wikipedia to disambiguate. Thus, those who have an itchy eye easily irritated by a hatnote to an article about pop culture polluting their serious article could simply try to find those valid entries when they exist and live with the correct direct hatnote until they can be found. It will take nothing more than continuing with the current guidelines and consensus to just end this discussion and get back to improving the encyclopedia. -- JHunterJ (talk) 15:02, 14 May 2012 (UTC)

JHunter wrote: "I would say we could give a few minutes or an hour to find appropriate terms to add." - You agree there should be a grace period then? Can we discuss how long this grace period should be? -Stevertigo (t | c) 19:26, 14 May 2012 (UTC)

I don't think your approach to discussion is consistent with the WP:Consensus policy. Why should JHunter or I want to negotiate a grace period, when you've been so far unwilling to recognize the worries of other editors? What are your proposals of alternate solutions that would satisfy all concerns? So far you've only advanced discussion when the other party has introduced an idea which is entirely consistent with your initial request; that is not a healthy consensus-building. I object any proposal you make until you recognize that removing hat notes, even the ones you find trivial, is problematic and shouldn't be done lightly. Diego (talk) 20:08, 14 May 2012 (UTC)

Also, no. "could" is not "should". -- JHunterJ (talk) 20:42, 14 May 2012 (UTC)

Diego, I have conceded on the matter of 2-dabs - originally I proposed we keep them, I now concede that we don't keep them permanently, rather just for a short time, to give other editors a chance to edit them. Your are characterizing my comments here as being not in good faith, and I don't think your characterization is either fair nor made in good faith. Regards, -Stevertigo (t | c) 21:19, 14 May 2012 (UTC)

I don't think you're acting in bad faith. My comments address only your negotiation tactics, not your motivations (and if you can't recognize the difference, well that explain your name-calling). But that doesn't stop me from depicting your arguments as against policy I still find your aggressive bargaining strategy unacceptable and won't participate in further consensus-building until. You're right that you've modified your initial position to accept deleting invalid 2-dab pages; but it's also true that this deleting was the pre-existing consensus and thus it's the least you were supposed to agree, as all editors are expected by policy to respect the established consensus until it's changed. I stand my point: you've yet not recognized the basic problem that either your initial or modified proposals would introduce, and you're not proposing anything to remedy it and thus having a solution satisfying all concerns. Under these conditions I oppose a grace period for malformed dab pages that would remove the proper hat notes from their legitimate place, even for a limited time. Diego (talk) 23:07, 14 May 2012 (UTC)

All you are doing is stonewalling. That means that you are maintaining a pre-existing position without compromising on anything. What's more is you are criticizing me for not agreeing with "pre-existing consensus" (your term for current policy) when the whole point of this discussion is to change the existing policy slightly. I have conceded that 2-dabs should not be permanent, and other editors such as older/wiser and JHunter have indicated a willingness to create a short grace period for 2-dabs. They are trying to find a compromise solution, and they are showing good faith. I suggest you try doing the same. -Stevertigo (t | c) 23:26, 14 May 2012 (UTC)

When all alternatives are worse than the existing consensus, there is no improvement to be gained by moving to a middle ground between the existing consensus and the worse alternatives. I don't support codifying a grace period; there would be a "de facto" grace period in between the time that the hatnote was implemented and the time that it was restored. If the editor needing a grace period absolutely must change the hatnote first before writing the dab page, they could say so on the talk page and then proceed apace with its creation. If I encountered such a situation, I'd assume good faith and leave it. If the other editor was off working on other improvements to Wikipedia rather than creating a disambiguation page with multiple actually ambiguous entries (not partial title matches or search dumps), I'd go ahead and revert it; it can always be restored once the disambiguation page is created. -- JHunterJ (talk) 02:21, 15 May 2012 (UTC)

Why do you assume that compromise is required? I realize that you seek to change the existing rules, but what if there's no consensus for that? Do you believe that every proposal to change a policy or guideline must either fully succeed or lead to compromise, or do you acknowledge that the status quo sometimes prevails?

Diego's point, I believe, is that you're expected to accept and abide by the pre-existing consensus (i.e. agree to it, even if you don't agree with it) until such time as new consensus is established. You've plainly stated that you have no intention of doing so (regardless of this discussion's outcome), and you're treating an offer to permit something already established as a concession on your part. —David Levy 02:44, 15 May 2012 (UTC)

In case you're interested, what I find unacceptable in your style is that you've been repeatedly asked about a requirement that draws a line in what can be accepted as a compromise; and each time you've just ignored the request - instead of clarifying your position as required. If you find that you can't agree to that irrevocable request just say so, so that we can stop wasting our time. Instead, you will find some words in your interlocutors' replies and twist them into something they didn't really say but that advances your position. This is not the way to build the long-standing consensus required for a guideline. Now I'll ask for the last time: will you agree to respect the direct link to the other article in the hat note for the duration of the grace period, in case we allow it? If the answer is yes, as far as I'm concerned we have a deal. If not, well this discussion has been going on too long. (You can't have a more direct approach to compromise than this. But I'm afraid you'll avoid positioning yourself with respect to this, as you did with the repeated requests by David Levy). Diego (talk) 07:23, 15 May 2012 (UTC)

The main problem is, there is an uneven standard being applied. The defenders of the hatnote claim they are doing it to provide navigational convenience for the user, and that the users lose out when they have to click twice. However, at the same time, links are being removed from disambig pages by these same people, because they don't fit a narrowly defined definition of what is allowed there.

Wow, you's still repeating this. Again, no one asserts that all direct article links are appropriate. The community has determined that some are helpful and others are unhelpful. You're literally arguing that it's unreasonable to hold such a position. (We support the inclusion of a specific type of direct article link, and you claim that "there is an uneven standard being applied" because we don't support the inclusion of certain other types.)

In any case, I agree with the proposal; allow creation of 2-DAB pages, and give a few weeks or month to find appropriate terms to add.

1. With or without leaving the direct article link in place until the disambiguation page is made valid?

2. Is there any justification apart from the non-consensus belief that eliminating "trivial" article links is desirable? We all agree the creation of valid disambiguation pages improves the encyclopedia (regardless of the underlying motive), but no one has explained how it's easier to add a third article link to a two-article disambiguation page than it is to simply create a valid disambiguation page from the start. Keep in mind that anyone visiting the base page will see the same two article titles (the one at which he/she has arrived and the one linked via the hatnote). —David Levy 02:44, 15 May 2012 (UTC)

Please recall, this is not about what you want to do or do not want to do. It is about finding a solution in the best interests of the users and the encyclopedia. No evidence has been provided that users are bothered by clicking twice; plenty of evidence has been provided that many users are distracted by certain hatnotes. I've tried a mathematical approach here, your thoughts welcome: Wikipedia talk:Hatnote/formulas. --KarlB (talk) 21:23, 14 May 2012 (UTC)

Ok, now I'm curious. What could be that evidence about many users' dislikes? I only recall you proposing an experiment to gather information about users, precisely because we couldn't know in advance what are their preferences. When did that thought experiment turn into "plenty of evidence"? Diego (talk) 22:48, 14 May 2012 (UTC)

I mean the anecdotal evidence of many editors who have complained in various pages across the wiki about these hatnotes. But yes, we should still measure generic user preferences across the population, but several eds above said they wouldn't change their mind in any case, so ... --KarlB (talk) 23:07, 14 May 2012 (UTC)

If that count as evidence, I call upon all the move requests to and from DAB pages that have been performed from the beginnings of Wikipedia in order to reduce clicks, and all the disambiguation guidelines created for the same purpose. That should count as more than anecdotal, since it's so common as to have been encoded into hard rule. Diego (talk) 23:11, 14 May 2012 (UTC)

Please see any or all of my earlier replies regarding the distinction between users' "preferences" and the efficiency with which they access encyclopedic content. —David Levy 02:44, 15 May 2012 (UTC)

Continued...

As a consensus compromise solution, I'd propose one of the following: 1) 'neutrally' worded hatnotes that link directly to the second meaning (i.e. turn the DAB page into a redirect). or 2) neutrally worded hatnote, that links to an anchor at the bottom of the page; then various single-link DABs/other refs could be added. I've created two samples pages here: #1 and #2; feel free to edit/tweak these pages to tweak the presentation if its something you might be willing to consider. The advantage is, both these solutions attempt to deal with both camps; 1) having a neutral hatnote at the top that doesn't distract the reader; 2) avoiding linking out /loading a different page - either through a direct link, or just an internal ref to a section lower down on the page. --KarlB (talk) 01:15, 15 May 2012 (UTC)

How is that a consensus solution? How is it "neutrally" worded? The hatnote has been neutered (or censored) and the reader seeking the non-primary topic is ill-served, and the reader of the primary topic gains nothing except possibly the maintenance of preferred ignorance that pop culture exists and can't be deleted from Wikipedia. I would tweak it as invited, except the tweak would simply be to reform the hatnote in the current mode: "for the song, see X (song)". -- JHunterJ (talk) 02:08, 15 May 2012 (UTC)

sorry wrong word choice - compromise was what i meant. the hatnote isn't any more "censored" than a link to a DAB page is censored. Tweaking back to "for the song, see X (song)" is just going to back to status quo; if you have an idea that is not status quo feel free to make the tweak (or create a 3rd, etc) --KarlB (talk) 02:22, 15 May 2012 (UTC)

the hatnote isn't any more "censored" than a link to a DAB page is censored.

A disambiguation page link is presented for what it is. You've proposed that we disguise direct links as a means of hiding material that you've deemed objectionable.

if you have an idea that is not status quo feel free to make the tweak (or create a 3rd, etc)

Do you acknowledge that retaining the status quo is a legitimate option? —David Levy 02:44, 15 May 2012 (UTC)

(ec) But I believe all of the possible changes for this hypothetcial problem are going to be worse than the status quo. I don't see any need to worsen actual encyclopedia usage in pursuit of a possible solution to a hypothetcial problem. -- JHunterJ (talk) 02:49, 15 May 2012 (UTC)

Perhaps to make this a bit more dramatic, we could use the hypothetical jesus and Jesus (sex toy) example from far above - the challenge being, how to you give quick access to the toy article, while not pissing off everyone who comes to learn about Jesus - and without just IAR and deleting the link entirely? Both solutions above actually address that - though I like #2, I don't much like #1. --KarlB (talk) 02:25, 15 May 2012 (UTC)

The above question relies on the assumption that we have an editorial responsibility to avoid "pissing off" readers. This is not so.

I previously acknowledged that we nonetheless would end up applying WP:IAR in such a scenario, but that doesn't justify a rule change. —David Levy 02:44, 15 May 2012 (UTC)

(ec) I will give you a better example: Santorum and Santorum (neologism). This also lead to actual hubbub, and settled when the disambiguation page was expanded. I believe in the meantime a 2-dab was attempted and revert-warred over; I think it may have even had a local WP:IAR consensus to do so. But there's nothing there to codify in the guidelines. If Jesus were unambiguous except for a sex toy, Talk:Jesus and Talk:Jesus (sex toy) might reach a similar conclusion to WP:IAR. I'm hopeful that that won't come up for simple "distractions" such as the acknowledgment that pop culture exists, even in the world populated by scientists. -- JHunterJ (talk) 02:49, 15 May 2012 (UTC)

Your first suggestion obfuscates the link's nature, thereby making it more difficult for readers seeking the non-primary topic's article to find it (and probably causing confusion). As JHunterJ pointed out, that's the opposite of making the link "neutral"; it's the replacement of a straightforward description with wording devised solely to hide material that you've deemed objectionable.

Your second suggestion complicates navigation, still adds an additional click, and forces users with slow connections to wait for the entire article to load before proceeding to the one that they seek.

Both suggestions are based purely on the non-consensus belief that eliminating "trivial" hatnotes is desirable. —David Levy 02:44, 15 May 2012 (UTC)

sigh. I'm trying to find a compromise. Don't throw rocks at it. And please stop pointing out that our proposals are non-consensus - it's getting tiring. We know you don't think these hatnotes should be removed, but unless you have something more to add to the discussion, please let those working on compromise solutions move forward. The point of the Jesus example was to use that extreme example to frame the challenge as a thought experiment (the extremeness might help if we decide to use neutral wording for example) - how could we solve this if a) IAR wasn't an option and b) we want to preserve quick access and c) we didn't want to distract the bulk of the readers of the article? The solutions above address that. Can you think of other solutions that fulfill these criteria? Creating a 2-page-DAB also addresses that, except b) is a bit slower. In any case, the two-clicks argument doesn't hold much weight with me now, with admins deleting useful links in DAB pages just b/c they don't fit some definition of 'valid' (quiz: when you delete a link, how many clicks does it take the user to get there?) --KarlB (talk) 03:33, 15 May 2012 (UTC)

sigh. I'm trying to find a compromise.

I understand that. But as discussed above, a compromise isn't necessarily called for. Thus far, it appears that there is no consensus for change. Do you acknowledge that retaining the status quo is a legitimate option?

Don't throw rocks at it.

I'm expressing my honest opinions. If someone were to suggest an alternative that I regarded as acceptable, I would acknowledge this.

And please stop pointing out that our proposals are non-consensus - it's getting tiring.

I used that term in the context of noting the lack of consensus that the problem the proposals are designed to address exists.

We know you don't think these hatnotes should be removed, but unless you have something more to add to the discussion, please let those working on compromise solutions move forward.

I am contributing to the discussion. Do you intend to demand that everyone who disagrees with these ideas refrain from participating?

The point of the Jesus example was to use that extreme example to frame the challenge as a thought experiment (the extremeness might help if we decide to use neutral wording for example)

As discussed above, your definition of "neutral" is contested.

how could we solve this if a) IAR wasn't an option

But IAR is an option. We don't change the rules (or invent new ones) to address a "tiny little edge case". (That's your description of the "trivial" hatnote issue as a whole.)

and b) we want to preserve quick access

How about continuing to provide a straightforward link directly to the article that the reader intends to reach?

and c) we didn't want to distract the bulk of the readers of the article?

(an uncorroborated assertion)

Can you think of other solutions that fulfill these criteria?

I disagree with your criteria.

In any case, the two-clicks argument doesn't hold much weight with me now, with admins deleting useful links in DAB pages just b/c they don't fit some definition of 'valid'

And you're still repeating this. "Sigh", indeed. —David Levy 04:31, 15 May 2012 (UTC)

Its interesting that David and JHunter are both dodging Karl's above hypotheticals and are now falling back on IAR. I agree that IAR can be sensibly applied, but we are attempting to codify into policy certain sensible procedures for dealing with a certain type of article content, namely hatnotes which for one reason or another don't belong at the top of an important-topic article. I described above how Wikipedia could get away with deleting Sufi (song) because it's just not terribly important, but we couldn't delete Sufism because that would leave a big gaping hole in the encyclopedia. This is an example of a very great difference in article importance - what I previously referred to as "notability." Don't want to call it "notability"? Fine, let's call it "importance." But the opposition tries consistently to be agnostic about hatnote content, describing the issue as purely navigational. But Karl and I have consistently come up with examples of where the hatnotes must be regarded for their content! JHunter brought up "Santorum" - if there wasn't the Santorum Amendment linked as the third disambiguation term, then the opposition would probably argue that Santorum (neologism) should be the first thing one should see at the top of Rick Santorum! Can the opposition at least admit that this would be a BLP issue? -Stevertigo (t | c) 05:51, 15 May 2012 (UTC)

Is it a BLP issue that both the politician and the neologism are listed at Santorum (disambiguation)? The tophat in Rick Santorum is exactly the same, because the hat note belongs to the navigation structure not the article content. (hint - BLP policy applies to all text space. If having both terms in the same page was a BLP violation, we couldn't have the disambiguation page either). Diego (talk) 08:08, 15 May 2012 (UTC)

If having both terms in the same page was a BLP violation, we couldn't have the disambiguation page either Diego, the problem is not having both terms in the disambiguation page, its having a potentially slanderous article link (Santorum neologism) at the top of the biographical article (Rick Santorum). -Stevertigo (t | c) 21:02, 15 May 2012 (UTC)

Diego's point is that if the material in question violated WP:BLP (due to defamation or for any other reason), it would do so regardless of the page on which it appeared. Adding an extra navigational step (by editing Rick Santorum to replace a direct link to Campaign for "santorum" neologism with one to a disambiguation page) wouldn't eliminate the violation.

However, editors did present various arguments as to why a direct link was undesirable. I didn't follow the situation closely, but I recall that one of the main concerns was that Wikipedia was unwittingly participating in the campaign (which is specifically intended to harm Rick Santorum's reputation). But when it comes to the method by which we link to an article (as opposed to its actual content), that argument falls under IAR, not BLP.

Of course, the existence of a third topic renders all of this moot. —David Levy 22:52, 15 May 2012 (UTC)

Its interesting that David and JHunter are both dodging Karl's above hypotheticals and are now falling back on IAR.

We're expressing disagreement with some of Karl's arguments. I don't know why you accuse us of "falling back on IAR", given the fact that Karl explicitly cited it in his original "sex toy" hypothetical ("WP:IAR would apply"). My above comment is reiteration of my earlier agreement with Karl's statement (despite the fact that I personally would prefer a different outcome).

Speaking of "dodging", I still await your pointers to the hatnotes linking to two-article disambiguation pages that you claim I've declined to remove, citing this as evidence that I tacitly condone these links' existence (and that this "consensus" constitutes grounds for compromise). This is my fourth request. (You've ignored the first three.)

Likewise, I await your explanation of how it's easier to add a third article link to a two-article disambiguation page than it is to simply create a valid disambiguation page from the start. All three of my previous requests (two addressed to you and one addressed to Karl) have been ignored as well.

And as Diego noted above (timestamp: 07:23, 15 May 2012), Karl and you have continually ignored the question of whether you're willing to leave the direct article link in place until the disambiguation page is made valid.

I agree that IAR can be sensibly applied, but we are attempting to codify into policy certain sensible procedures for dealing with a certain type of article content, namely hatnotes which for one reason or another don't belong at the top of an important-topic article.

There's no consensus for the premise that the hatnotes in question "don't belong". Karl and you nonetheless state this as fact, rely on the assumption that the only reasonable course of action is to find a way to remove them, and portray anyone who disagrees as an obstructionist.

I described above how Wikipedia could get away with deleting Sufi (song) because it's just not terribly important, but we couldn't delete Sufism because that would leave a big gaping hole in the encyclopedia. This is an example of a very great difference in article importance - what I previously referred to as "notability."

And I noted that this "is why the base title (Sufi) redirects to the article about the religion (while the song's article resides at a title with parenthetical disambiguation)."

Don't want to call it "notability"? Fine, let's call it "importance."

We aren't quibbling over semantics. Call it "notability", "importance" or something different. It doesn't matter. We don't dispute that such a distinction exists. We dispute that it's grounds for removing the links in question (a premise that you take for granted and evidently believe isn't even up for discussion). —David Levy 12:03, 15 May 2012 (UTC)

I still await your pointers to the hatnotes linking to two-article disambiguation pages that you claim I've declined to remove, citing this as evidence that I tacitly condone these links' existence - David I didn't say that you haven't touched any existing 2-dabs, just that we have created 2-dabs which became 3-dabs through some helpful editing by other users. In the Sufi case for example, Diego PRODed the disambiguation page, but older/wiser was helpful and found other terms to add. -Stevertigo (t | c) 21:02, 15 May 2012 (UTC)

And as Diego noted above, Karl and you have continually ignored the question of whether you're willing to leave the direct article link in place until the disambiguation page is made valid. - I'm willing to compromise on this point, and say that we should keep the extra term on the article page, along with the new otheruses hatnote, until the disambiguation page gets 3 terms. -Stevertigo (t | c) 21:07, 15 May 2012 (UTC)

David I didn't say that you haven't touched any existing 2-dabs, just that we have created 2-dabs which became 3-dabs through some helpful editing by other users.

You wrote the following:

"We've implemented 2-term disambiguations several times now in the context of this discussion, and none thus far have been reverted. If you feel that trivial hatnotes belong at any of the articles we've discussed, please go ahead and revert back to their previous state. The fact that you don't do so means that you tacitly if not overtly agree with the changes. Hence we already have some consensus, or indication thereof."

You explicitly referred to "2-term disambiguations" and noted that "none thus far have been reverted". Do you mean to say that you were referring to former two-article disambiguation pages to which additional article links were added? And this doesn't strike you as the least bit misleading?

Why would I (or anyone else) want to remove links to such pages? Your assertion that "the fact that [I] don't do so means that [I] tacitly if not overtly agree with the changes" is analogous to claiming that I support the creation of unsourced articles because I don't take issue with sourced articles originally created without citations.

I'm willing to compromise on this point, and say that we should keep the extra term on the article page, along with the new otheruses hatnote, until the disambiguation page gets 3 terms.

This would result in confusion and wasted time. Readers would be misled to believe that additional articles were linked on the disambiguation page, only to find the same two.

And you appear to have once again overlooked my request for an explanation of how it's easier to add a third article link to a two-article disambiguation page than it is to simply create a valid disambiguation page from the start. —David Levy 22:52, 15 May 2012 (UTC)

Do you mean to say that you were referring to former two-article disambiguation pages to which additional article links were added? And this doesn't strike you as the least bit misleading? - Yes, of course I meant former 2-dabs. No, I had no intention of being misleading. Again, the point is that adding new terms to a 2-dab should be seen as helpful editing, and the right thing to do. Simply PRODing a 2-dab should be seen as unhelpful and the wrong thing to do.

This would result in confusion and wasted time. Readers would be misled to believe that additional articles were linked on the disambiguation page, only to find the same two. - Aren't you the one who wrote:

And as Diego noted above, Karl and you have continually ignored the question of whether you're willing to leave the direct article link in place until the disambiguation page is made valid.

This makes it sound like you are looking for some consensus on this point. I was willing to offer some concession, and indicated as much. But perhaps you were simply being misleading? -Stevertigo (t | c) 04:11, 16 May 2012 (UTC)

Yes, of course I meant former 2-dabs. No, I had no intention of being misleading.

I'm not suggesting that you intentionally misled me, but your wording nonetheless had that effect.

Again, the point is that adding new terms to a 2-dab should be seen as helpful editing, and the right thing to do.

I've explicitly acknowledged that it's helpful. Likewise, adding citations to an unsourced article is helpful. So is translating a non-English article to English.

This doesn't mean that we should encourage the creation of two-article disambiguation pages, unsourced articles and non-English articles.

Simply PRODing a 2-dab should be seen as unhelpful and the wrong thing to do.

It's less helpful than adding a third valid article link is, but that type of improvement isn't necessarily feasible (and certainly isn't required).

You advocate the establishment of a grace period. How is WP:PROD anything other than a more accommodating variant? Not only does it provide a week to improve the page, but a single objection by any user in good standing (including the page's creator) instantly and permanently halts the process (even if a third article link hasn't been added). And if seven days elapse and page is deleted, it can be automatically restored upon request.

Aren't you the one who wrote:
"And as Diego noted above, Karl and you have continually ignored the question of whether you're willing to leave the direct article link in place until the disambiguation page is made valid."
This makes it sound like you are looking for some consensus on this point.

I was seeking a response to the question, which I finally received.

I was willing to offer some concession, and indicated as much. But perhaps you were simply being misleading?

How so? Did you interpret something that I wrote to mean "I hereby agree in advance to accept any relevant proposal you make, even if I believe that its implementation would be harmful"?

I'm attempting to gain a fuller understanding of your position. To this end, I still await an explanation of how it's easier to add a third article link to a two-article disambiguation page than it is to create a valid disambiguation page from the start. (This is my sixth request.)

And again, we've yet to establish consensus that a problem exists, so it's unclear why we're at the proposed solution stage (let alone why Karl and you apparently believe that compromise is mandatory). —David Levy 06:25, 16 May 2012 (UTC)

David, you have repeatedly asserted the same points - that there isn't a problem - and when suggestions were made as to establish the problem in an objective fashion (for example, polling users), you rejected that and said you would not change your mind regardless of user preference. It is very clear that there is very little on this earth that would get you to change your mind, and you seem to believe that since you and two other eds do not agree it is a problem, therefore it isn't a problem. However, this issue has been repeatedly brought up, on talk pages, in edit wars, and so on across wikipedia, so to say there isn't a problem is rather daft - consensus != unanimity. I think the disagreement is, is there a solution that is better than the problem? To answer your question "how it's easier to add a third article link to a two-article disambiguation page than it is to create a valid disambiguation page from the start" it's very easy - it's called "many hands make light work"; a given editor may not have the skills or ability or research knowledge to find that 3rd dab, but having created it, he can signal others, and get them working on it. This is eventualism at its finest, and why wikipedia allows stub articles, because a different editor can come along and fix it. We already have evidence of this happening; several small DABs I've created have been enhanced by others later. No single decent Wikipedia article was written by one person.--KarlB (talk) 17:04, 16 May 2012 (UTC)

I'll assert a variation of the point: there isn't consensus that there's a problem. But I've said before, if you want to poll the readers at the article pointed to by the hatnote and find that there is consensus that those readers would be better served by having to click, load a dab page, read, click, and load the target page instead of click, load the target page, that's the consensus we'd go with. It does indeed seem there is very little on this earth that could lead to such a consensus. Many hands can make light work by the first editor observing the "problem" on the talk page, at allowing some of the many hands there to create the hypothetical 3-dab (if it exists), hindering no one needlessly until then. Perhaps there is a technical solution to the "problem" faced by readers of the page who don't need the hatnote, though. Seriously. We could create a spanning template for "pop culture hatnotes" or "distracting hatnotes" or whatever description is apt, and then users could set a user preference to hide those spans. -- JHunterJ (talk) 17:28, 16 May 2012 (UTC)

David, you have repeatedly asserted the same points - that there isn't a problem

While that is my belief, by point is that there's no consensus that a problem exists. I'm not claiming that my personal opinion is sacrosanct.

and when suggestions were made as to establish the problem in an objective fashion (for example, polling users)

I disagree that the proposed "polling" is a valid method of establishing that a problem exists. I've stated this clearly and repeatedly.

you rejected that and said you would not change your mind regardless of user preference.

Again, please see any or all of my earlier replies regarding the distinction between "user preference" and the efficiency with which readers access encyclopedic content.

We don't "poll" people to determine what material they'd "prefer" not to read.

However, this issue has been repeatedly brought up, on talk pages, in edit wars, and so on across wikipedia, so to say there isn't a problem is rather daft

Oh, there might be an edit waring problem. But that obviously doesn't establish consensus that the hatnotes are problematic.

Users frequently complain about (and attempt to remove) material that they find disgusting, offensive, blasphemous, et cetera. Does that mean that such content should be removed? No. Does that mean that we should invite readers to vote on whether it should be removed (and honor the majority's "preference"). No.

consensus != unanimity.

Agreed. The relevant definition of "consensus" is documented at Wikipedia:Consensus.

To answer your question "how it's easier to add a third article link to a two-article disambiguation page than it is to create a valid disambiguation page from the start" it's very easy - it's called "many hands make light work"; a given editor may not have the skills or ability or research knowledge to find that 3rd dab, but having created it, he can signal others, and get them working on it. This is eventualism at its finest, and why wikipedia allows stub articles, because a different editor can come along and fix it.

The situations aren't analogous. The creation of a stub article presents new material with which other editors can work. Conversely, as I noted previously, anyone visiting an article whose hatnote links directly to another article will see the same two article titles (the one at which he/she has arrived and the one linked via the hatnote). The creation of a two-article disambiguation page presents no new material with which other editors can work.

We already have evidence of this happening; several small DABs I've created have been enhanced by others later.

And those individuals could have created the same valid pages from scratch.

However, I've just come up with an idea for a compromise solution. (See below.) Steve and you have criticised me for failing to offer one, so I hope that you appreciate the fact that I'm making a sincere attempt. —David Levy 21:37, 16 May 2012 (UTC)

Can you think of other solutions that fulfill these criteria? I suggested changing the link style so that it won't look like part of the article and will be recognized unambiguously as navigation. This should satisfy your criterion c); I'd like to hear what you think of how this possibility would affect the perception of the links.

with admins deleting useful links in DAB pages The purpose of disambiguation pages is to provide quick access to pages that match the DAB title, i.e. it's a way to solve a structural problem of namespace collision in titles; other articles don't this purpose. I usually like having links to other articles in the DAB related by partial names, but recognize that they're outside the scope. Diego (talk) 08:08, 15 May 2012 (UTC)

"it's a way to solve a structural problem of namespace collision in titles; other articles don't this purpose"; but that has very little to do with user experience. users don't know what a DAB is supposed to contain, and probably don't care if it contains some useful things that aren't an exact DAB. Making judgements that a given link is not useful in a DAB page b/c it is not an exact DAB without any evidence that this is bothersome to the user is a bit arrogant (and there are tens of thousands of DAB pages that *have* these links, and presumably no-one reading them has bothered enough to remove them) - and as I've pointed out before, it is also hypocritical, to defend to the last the right of users to a single-click, then with the other hand go around slaughtering "invalid" links in DAB pages (presuming that the user could never have a use for those links).

As for your other point about changing the link style, please feel free to edit the Johnny cash examples linked above to show what it might look like.--KarlB (talk) 17:04, 16 May 2012 (UTC)

users don't know what a DAB is supposed to contain, and probably don't care if it contains some useful things that aren't an exact DAB. As I said, I don't mind having extra links at DAB pages for less-than-perfect title matches. But a user who typed an exact title in the URL or search box certainly knows what article they were expecting, and creating an indirect DAB page will certainly increase the work required for that user to reach the desired article. So exact title matching should be given preference - and if there's only one other article with the same exact title, it should be linked directly and not through an intermediate DAB, because that's the most efficient structure for that case. It's OK if you also have a DAB page with other relevant articles, but not at the cost of removing the direct link for the exact title match. Can you understand now why the WP:SIMILAR link is to be treated different than the others? They're accessed through a different flow, so they have different usability requirements. Diego (talk) 21:06, 16 May 2012 (UTC)

New proposal

It's been asserted above that the creation of an invalid disambiguation page serves to "signal others" that a valid one is needed/wanted. So why not simply institute a formal means of accomplishing this — a system through which users can identify disambiguation pages that they believe should exist and flag down editors capable of creating them?

This would sidestep the issue of whether the hatnotes in question are problematic (because users could submit a request for that reason or on any other good-faith basis), address concerns surrounding the existence of invalid disambiguation pages, and promote the creation of helpful disambiguation pages regardless of whether "trivial" hatnotes are involved.

The precise implementation could be discussed, but one possible method would be to create a template to be inserted on the talk page of the relevant article or redirect, resulting in its placement in a category and possibly a bot-populated process page (e.g. Wikipedia:Disambiguation pages for creation), thereby establishing a one-stop list of the disambiguation pages that users want help building.

And if someone does create an invalid disambiguation page (for the purpose of removing a "trivial" hatnote or for any other good-faith reason), instead of deleting it straight away, we could use a sister template to formally request its expansion. And if that doesn't occur before the grace period (whatever duration is agreed upon) expires, the deleted disambiguation page would automatically enter the aforementioned creation request process.

Opinions? —David Levy 21:37, 16 May 2012 (UTC)

At first sight it looks good. It sounds like something that would fall under the supervision of Wikipedia:WikiProject Disambiguation. Diego (talk)

Couple issues, David. Using the talk page to indicate a desired change of this kind seems like extra work (Your side is always talking about the trouble created by disambiguation pages, more clicks and so forth), but more importantly, talk pages already have tons of templating which would make a new template regarding dabs difficult to see or find. Why not just create an article template that has an otheruses tag embedded? This would seed the new disambiguation page, and the template would likely be short lived, given the usual tendency for 2-dabs to collect extra terms. I like your ideas, and appreciate your wanting to find a solution. I think you're on the right track. -Stevertigo (t | c) 21:56, 16 May 2012 (UTC)

Don't mistake more work for the reader with more work for the editors; the latter is manageable as long as the work is distributed and because Wikipedia is a work in progress. Templates at article space usually refer to the quality of content, not to peripheral service pages like DABs, that's why I think it makes more sense having a talk page tag. I agree that David may be onto something, this lateral thinking proposal may be what's needed for unclogging consensus. Diego (talk) 22:15, 16 May 2012 (UTC)

Couple issues, David. Using the talk page to indicate a desired change of this kind seems like extra work (Your side is always talking about the trouble created by disambiguation pages, more clicks and so forth),

As I noted, the precise implementation could be discussed. The above is merely a rough outline of one possible approach.

but more importantly, talk pages already have tons of templating which would make a new template regarding dabs difficult to see or find.

I was focusing on a means of populating the master list(s). I don't regard the actual tag's prominence as particularly important.

Why not just create an article template that has an otheruses tag embedded?

I oppose linking to the disambiguation page from the article namespace before it's ready, as this is unhelpful to readers (a vast majority of whom aren't editors). —David Levy 22:28, 16 May 2012 (UTC)

I now agree with the idea of keeping the existing hatnote in place until a disambiguation page is filled up, so maybe we don't disagree on much else. Certainly you and Diego (and others) don't oppose disambiguation creation just for sake of keeping a hatnote. I think thats clear. And like you say David, we can be agnostic about the reasons for transforming a hatnote into a disambiguation page - even in egregious cases (like Karl's Jesus example). The issue then is the nominal process by which a disambiguation page is created, and I have already agreed that in all but a few cases, a 2-dab is insufficient. In rare cases where a 2-dab may be required, I suppose we can refer to IAR. -Stevertigo (t | c) 23:23, 16 May 2012 (UTC)

Indeed, I see nothing in the above message with which to disagree. What a pleasant turn of events.  :) —David Levy 00:13, 17 May 2012 (UTC)

Cool beans. :). We should get some input from Karl and Diego (and maybe older/wiser) and write up a summary of this discussion. -Stevertigo (t | c) 02:39, 17 May 2012 (UTC)

I've no issue with a process for incubating or proposing new disambiguation pages. I've no suggestions at the moment for the implementation. older ≠ wiser 11:15, 17 May 2012 (UTC)

Sorry I took a break from this discussion for a while. What about a slightly different compromise -

1. If an editor thinks it will possibly improve the encyclopedia, she creates a 2-DAB, and then links to both in the header: the original hatnote, + the 2-DAB page.
2. We *always* make the 2-DAB page more useful than just 2 links, by adding some see also links, like 'in title' or 'look from', partial title matches, or other relevant links to other similar subjects.
3. We create a template that goes on the top of the DAB page. That way, anyone who clicks through to the DAB page will see that it needs help. The template will also add a hidden category of 2-DABs needing expansion/etc. This template can also be put on any 2-DABs one happens to find.
4. Consensus amongst this group is that if a 2-DAB expansion has been there for X months without having been expanded, it will be put up for PROD or AfD, which will give interested editors one last chance to expand it.
5. once deleted, as david proposes, it goes into the archive of DABs-to-eventually-create.

This solution, which is a derivative of David's, walks the line as johnny cash would say:

1. it allows the original hatnote to remain at the top of the article (preserving single-click access) as the first link. If a 3-or-more-DAB ever does get created, then by consensus that part of the hatnote can be removed
2. it also provides an additional DAB link for readers, with seealso/lookin/etc templates that personally I've always found really useful and informative.
3. It gives a way for centralized notification of these pages through the templates/hidden categories
4. And, it doesn't mess up the top of the article page with a template (the DAB page is templated instead); just a single additional link;
5. Finally it gives a little bit more prominence to the DAB-expansion than a talk-page template might; an average user may click through to the DAB and find himself inspired to expand the DAB with something no-one else thought of; indeed I had never really touched DAB pages until I saw a hatnote which annoyed me, leading me here...

In any case, I put together a sample here, as it might be easier to see than talk about: User:Karl.brown/Johnny1 - it also links to a sample DAB page. Please imagine for the purposes of this sample that there is only a single article that matches Johnny Cash, so the DAB itself is written as a 2-DAB. Feel free to tweak it, or the DAB page, to make it look like something that you would agree with. Kudos to David also for the lateral thinking.--KarlB (talk) 04:53, 29 May 2012 (UTC)

I think that this idea might be workable. My main concern, based upon your example, is that it entails linking to a two-article disambiguation page (largely duplicative of the hatnote's direct article link) without warning. Perhaps we could come up with wording that makes this clear to readers (instead of misleadingly promising "other uses"). —David Levy 07:07, 29 May 2012 (UTC)

Thanks. Feel free to change the wording; you can even do so in the Johnny Cash example, so people see what it might look like on a real page. I'm certainly open to any reasonable wording. Also appreciate your help on the wording/formatting of the template at the top of the DAB page - what should it say, how should it be structured - for now it's just a quick hack. --KarlB (talk) 12:53, 29 May 2012 (UTC)

Also FWIW, after my wikibreak from this discussion, I've re-read through some of the above, but I couldn't read all of it...I'm really surprised it went on so long, and I hope we're close to the end... In any case, having been away from it for a while gives a different perspective, so I just wanted to apologize for anything I said above that was not civil or respectful, esp to David, JHunterJ, and Diego. I know everyone here is trying to improve the encyclopedia, and even though we may have different points of view, I'm sorry if I said things that were uncivil or unhelpful. --KarlB (talk) 05:05, 29 May 2012 (UTC)

I accept your apology and apologize for any aggravation caused by my admittedly long-winded replies.  :) —David Levy 07:07, 29 May 2012 (UTC)

Cheers and best regards. --KarlB (talk) 12:53, 29 May 2012 (UTC)

FWIW, I've found a real live example of this in the wild: Sawdust; there was an extant 2-DAB, which had been de-linked, so I linked it at the top, and kept the band album. We could use this one as a live test - try to create a template that encourages others to add to the 2-DAB; then if some time period expires without additional DABs being added, sawdust (disambiguation) will be be put up for deletion. Would be a great experiment to see what happens.--KarlB (talk) 15:56, 3 June 2012 (UTC)

There are four links of potential other uses for "Sawdust". I would say an hour would be a more-than-adequate time period for checking them out to see if anything can be added to the dab. Currently, the hatnote over-promises: there are no "other uses" on the dab page. -- JHunterJ (talk) 21:26, 10 June 2012 (UTC)

OK, I have an idea. How about a new hatnote for this purpose that says "For whatever, see whatever's title. If there is another ambiguous topic, consider adding it to disambiguation link." That would avoid the false advertising problem. -- JHunterJ (talk) 23:08, 10 June 2012 (UTC)

Well I understand Karl's intent here is to formalize a compromise, and its similar to what David, Diego and I talked about earlier: The undesirable 2-dab stays, but so also does the undesirable link to a less-notable or otherwise unrelated article. A few hours to find more links for the 2-dab is not ideal - a few days seems more in keeping with AGF. -Stevertigo (t | c) 05:58, 11 June 2012 (UTC)

To reiterate, there's no consensus that the direct article links in question are "undesirable". The new setup is acceptable because it doesn't rely on that premise. The underlying motive of an editor inserting the new hatnote template might reflect such a belief, but the result is the addition of a fairly harmless link implicitly encouraging the expansion/repair of an invalid disambiguation page. —David Levy 19:28, 26 June 2012 (UTC)

I'd vote to keep it for a month; in some cases, making a 2-DAB into a 3-DAB might require creation of another article, so finding sources and ensuring it is notable/verifiable/etc can take time. In any case JHunterJ (and others), feel free to modify the Johnny Cash article here -> User:Karl.brown/Johnny1 with your proposed adjusted wording, and I'd also appreciate your help on wording of the template at the top of the 2-DAB page. I understand the concerns about over-promising; that's why I proposed that we always add 'intitle' and 'lookfrom' links so we give the user at least something to work with, as well as any particularly notable/useful partial title matches, so the user will never be going to a 2-DAB page with only 2 links on it to click. The purpose in my mind to doing this is

1. to notify expert editors from the projects about the existence of these pages, through the template at the top
2. to give other editors who don't know much about DABS but that may care about the less-notable hatnote a chance to add to the DAB as well

If time runs out (e.g. 1 month), then the 2-DAB would be AfD'd but would somehow remain on a dead-2-dab list for resurrection in the future.--KarlB (talk) 06:13, 11 June 2012 (UTC)

Actually, rather than assuming bad faith, I am happy to leave the time limit undefined. Prods and AFD already have 7-day limits, but discussion at the talk page (and perhaps use of the different hatnote language, if others like it as well) can help defer prods and AFDs until "sufficient" time has passed (whatever that is). With the new language, I personally wouldn't prod or AFD them at all; it falls right back into the "harmless and cheap" category of many esoteric redirects. -- JHunterJ (talk) 10:54, 11 June 2012 (UTC)

Thanks JHunterJ; keeping it undefined is also fine with me; I just don't want the list of 2-DABS-to-expand to become a kill-list - so I hope people will give them a fair shot before PROD'ing. In any case, thanks for the add to User:Karl.brown/Johnny1, I'm quite fine with that language for now. I'd appreciate your input/help on language/formatting at the top of User:Karl.brown/Johnny_Cash_(disambiguation) - I don't really know how to make a template which we'd have to do. In any case, JHunterJ since you seem to be in agreement, could I suggest a couple of next steps?

1. Create a hatnote template, special for this case, that has the language you have proposed in the Johnny Cash article (and since this will be a template, we can centrally change the language in the future if needed).
2. Create a 2-DAB template, special for the 2-DAB page, which tells the user that more DABs are needed;etc and adds it to a category.
3. Decide on the category name, and how the tracking will work (and which projects will be notified?)
4. Decide on what additional wording needs to be added to the 2-DAB guideline as a result of this discussion
5. Once these templates have been created, place them on Sawdust as a 'live' example, and see if people who are part of this discussion (or other eds) want to make other modifications.--KarlB (talk) 13:13, 11 June 2012 (UTC)

I created {{Consider disambiguation}} for #1. And no, no special permission is needed for creating templates, although some (like that one) can be tricky. -- JHunterJ (talk) 00:27, 19 June 2012 (UTC)

#4 and #5 done as well. -- JHunterJ (talk) 00:48, 19 June 2012 (UTC)

Thanks! Ok, I've added this now as a workspace to collaboratively edit the guideline: Wikipedia_talk:Hatnote/2DABRevised; please take a look and make revisions there; once we have consensus on the tweaks to the guideline we can stick it on the main page. Also i note that there are two 2-DAB sections (WP:TWODABS and WP:2DAB)). Why don't we just make a single 2-DAB section? --KarlB (talk) 01:48, 19 June 2012 (UTC)

Finalizing the new proposal

Thanks to the help of JHunterJ, we have now completed most of the steps above. The results are as follows:

• We have a template that can be used to tag 2-dab pages.
  • It currently lives here {{Only-two-dabs}}
  • I welcome ideas on a better name to move it to - I'm not wedded to that at all.
  • Please also look at the language and documentation and make any edits you think will improve it. It uses more or less the language developed by JHunterJ and myself.
  • The template also places the article in Category:Disambiguation pages containing one non-primary topic
  • I've also updated the DAB project here: Wikipedia:WikiProject_Disambiguation#Find_additional_ambiguous_terms_for_2-DAB_pages with instructions. As before, please edit these at will.
• We also have a template that can be used to redirect to a 2-dab.
  • It currently lives here {{Consider disambiguation}}, and the wording was developed by JHunterJ; it encourages the user to add to the DAB, without promising additional terms (because, it's a 2-DAB after all).
• Finally, above I had linked to pages in my userspace, but I've now deleted those as userspace pages don't work well with templates/etc. We now have a real live testcase at Sawdust, so you can see how the whole thing works in the real world wiki.

Final steps - on which I request your assistance

• Refining the guidelines on 2-DAB pages to make this all official. I have placed a draft revised set of guideline changes here: Wikipedia_talk:Hatnote/2DABRevised; please make edits to it and comment on them here.
• Someone should then close this discussion --KarlB (talk) 05:12, 20 June 2012 (UTC)

Added a couple shortcuts ({{o2d}}/{{2d}}/{{2dab}}, {{condis}}/{{cdis}}). -Stevertigo (t | c) 07:15, 20 June 2012 (UTC)

Can the {{Consider disambiguation}} template be edited with something other than ARTICLENAME (disambiguation)? -Stevertigo (t | c) 07:32, 20 June 2012 (UTC)

yes, I think so: Template:For2#.22.E2.80.A6_If_you_are_seeking_another_topic.2C_.E2.80.A6_consider_adding_it_to_.E2.80.A6..22— Preceding unsigned comment added by Karl.brown (talk • contribs) 07:52, 20 June 2012‎ (UTC)

Right. Adding another parameter specifies the dab page to be used instead of ARTICLENAME (disambiguation). -- JHunterJ (talk) 10:54, 22 June 2012 (UTC)

Comments

Sorry, I have not been following this discussion, but I really can't agree with this proposal. I don't much care whether 2-article disambiguation is done through hatnotes, dab pages, or left to editor discretion, but having both is useless clutter. So is a template; if a 2-dab is not wanted then just change it - why clutter it with a template, potentially for ever. This is just generating another excuse for drive-by templating - unnecessary because 2-dabs are not actually harmful. I would not, however, be against a hidden templated note indicating that a 2-dab had been created deliberately so that future editors took this into consideration. SpinningSpark 08:07, 20 June 2012 (UTC)

• The wording of the template, and of the documentation below it, doesn't make clear that a dab page with only two entries is only at risk of deletion where one of those entries is the primary topic. A dab page with two non-primary entries is fine. Having just created Ice Shack (over a redirect and in response to a comment at Talk:Shack), I wouldn't like people to read something which could be interpreted as saying that all two-entry dab pages are problematic. Perhaps the text at Template:Only-two-dabs could be modified to read "Place this template at the top of any disambiguation pages that have only two ambiguous terms, one of which is the primary usage." (My addition in italics) I think that's more important than cluttering the template itself with this extra info. PamD 08:29, 20 June 2012 (UTC)

Pam wrote: I wouldn't like people to read something which could be interpreted as saying that all two-entry dab pages are problematic. - Well thats the point, as things currently stand 2dabs are considered 'problematic.' We are making the case that at the very least they shouldn't be prodded as soon as they are made. Ive seen more than a few sensibly-created 2dabs recently, so it may be that the prohibition against 2dabs altogether needs overturning, but the argument against that is that 2dabs are generally unnecessary. -Stevertigo (t | c) 10:13, 20 June 2012 (UTC)

Sensible 2-dabs, e.g., those that are at the base name, don't need the new hatnote and don't meet prod/db-disambig criteria. -- JHunterJ (talk) 11:39, 20 June 2012 (UTC)

Hi Pam; it's a good point you made, and I agree with the changes; we should make it clear that this tag is for 2-DAB pages which are not at the base name. I think the changes you made capture this.

@SpinningSpark; I agree in principle with your comment, however if you peruse the discussion above, you will see that several editors disagree, and felt that 2-DAB pages should eventually be deleted, and it was better to maintain a direct link to the hatnote. This solution is a compromise; the end result I believe will be better for readers, because in some cases 2-DABS will be turned into 3-DABS (or more), and the encyclopedia will improve; or in some cases, 2-DABS will be deleted, but will have had their chance. In any case, as Diego points out, the 2-DABs will always have "intitle" and "lookfrom", so will always be useful to the reader, even if they end up being deleted in the long run.--KarlB (talk) 14:01, 20 June 2012 (UTC)

That does not make any sense to me at all. If another page comes to light, an editor can recreate the dab or add another hatnote. Dab pages are for people who cannot find the right article, so it is very unlikely that someone coming to the dab page already knows of another article (unless they come there specifically to add an article). In any case, I don't see any pressing need to advertise for additional dabs - it is only likely to generate irrelevant cruft, a large portion of additions to dab pages are made by people who have not understood what a dab page is for. I agree with the "intitle" by the way, that helps to make it even more clear what should be listed. SpinningSpark 19:01, 20 June 2012 (UTC)

I've modified the "Consider disambiguation" to be just informative, not prescriptive. Instructions for editing the DAB page should be kept in the Only-two-dabs template. For the two-DAB page to be useful, it should always contain a See also section with intitle search links; I've added that recommendation to Only-two-dabs. It would be even better if Only-two-dabs actually included the intitle search link, but I don't know how to extract the base title and remove the (disambiguation) part in the template. Diego (talk) 10:13, 20 June 2012 (UTC)

I've reverted that modification, which changed the basic point of the template: to avoid falsely indicating to the reader that additional topics were already listed at the disambiguation page. If the template is just going to collapse to one of the existing {{for}} variations, we should delete it instead. We would then return to the issue of what to do with at-one-moment pointless dabs such as Sawdust (disambiguation), which would again be G6-deletable. -- JHunterJ (talk) 11:39, 20 June 2012 (UTC)

The hatnote template could indicate that the two-DAB may contain links to title searches, but it shouldn't contain instructions to edit the DAB page; that's the responsibility of the maintenance template, and it's out of place at the article top. How about a wording like this one?

If you are seeking another topic, try the search terms in (disambiguation).

Or this:

If you are seeking another topic, try the See also section in (disambiguation).

Until the two-DAB page has additional information to help the reader, it should not be linked from the tophat. Diego (talk) 12:03, 20 June 2012 (UTC)

Hi Diego - either one of those would be ok with me; I also agree, the 2-DAB should always have lookfrom and intitle (indeed, I can't see why every dab doesn't have those), so thanks for that addition. --KarlB (talk) 15:11, 20 June 2012 (UTC)

I've reworded the hatnote to avoid instructing the user and instead identify what help it can render the reader (the search links). -- JHunterJ (talk) 16:20, 20 June 2012 (UTC)

Thanks, now it's more to the point. Diego (talk) 17:37, 20 June 2012 (UTC)

• As for a new template name: I suggest to not use the word hatnote, since it is more like a maintenance tag. -DePiep (talk) 15:10, 20 June 2012 (UTC)

I think {{Only-two-dabs}} could be an appropriate notice at Season 4, but this is a DAB page that only has one ambiguous topic (!) so the current wording won't fit. Thoughts? Diego (talk) 08:48, 22 June 2012 (UTC)

"Season 4" also is not at Season 4 (disambiguation), i.e., the only topic is somehow not the primary one. I think the only salvation for Season 4 is the eventual move of the episode article to the base name, or the discovery of another topic ambiguous with the name. I wouldn't put energy into a parallel set of hatnotes and categories to cover its lonely condition. -- JHunterJ (talk) 11:19, 22 June 2012 (UTC)

Hello! My apologies for taking a break from the discussion, which was purely due to my busyness outside Wikipedia.
In my opinion, the new setup is sensible. The hatnote template introduces a bit of clutter, but this is justified by the addition of a fairly harmless link implicitly encouraging the expansion/repair of an invalid disambiguation page. Nice work, everyone!
I've tweaked the disambiguation page tag's wording and renamed the category for clarity and accuracy. (Disambiguation pages generally aren't considered "articles".)
I was asked to comment, specifically on whether the discussion should be closed. In my opinion, we probably should solicit additional feedback (perhaps in the form of an RfC) to ensure that the change is consensus-backed. —David Levy 19:28, 26 June 2012 (UTC)