Wikipedia talk:Disambiguation

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Disambiguation
WikiProject icon This page is within the scope of WikiProject Disambiguation, an attempt to structure and organize all disambiguation pages on Wikipedia. If you wish to help, you can edit the page attached to this talk page, or visit the project page, where you can join the project and/or contribute to the discussion.
 

Two words on usage[edit]

Here I added two words to the usage criterion, making clearer that we are discussing usage on Wikipedia, which of course is what we have always meant as our criterion here. I was summarily reverted, of course, but I think it's a worthwhile clarification (as per WP:NOTDICT and other reasons). I reiterate that it's only a clarification; we obviously are only talking about Wikipedia usage here, right? Red Slash 02:51, 18 September 2014 (UTC)

  • Oppose "on Wikipdia". Readers, at least the vast majority of the target audience, do not live "on Wikipedia", and editors should get out of the "on Wikipedia" mindset. Editors should know what a proper encyclopedia is, but readers should not. Whatever a reader searches with, the result should be meaningful and helpful. --SmokeyJoe (talk) 03:39, 18 September 2014 (UTC)
  • Support. Dab pages are Wikipedia navigation tools. If the reader is searching the Oxford English Dictionary or the NYT microfiche, the dab page is irrelevant. Editors of Wikipedia nav pages (such as redirects and dab pages) should get into the "on Wikipedia" mindset. The readers don't have to, of course, any more than they need to get into the "OED" mindset to use the tools that its editors (with their "OED" mindsets) put there for their navigation. -- JHunterJ (talk) 11:42, 18 September 2014 (UTC)
  • Wikipedia doesn't exist in isolation; a search on Wikipedia is just one way among many of accessing an article. The general formulation is a better test to use.--Trystan (talk) 13:51, 18 September 2014 (UTC)
    And this discussion isn't about how articles appear in searches not on Wikipedia. Disambiguation pages aren't destinations; they're navigational tools to get readers to their destination. If a Google search gets the reader to the sought article, the dab, primary topic considerations, and any redirects are irrelevant. Disambiguation pages exist as the on-Wikipedia way of accessing an article. -- JHunterJ (talk) 14:09, 18 September 2014 (UTC)
    How articles are titled is inextricably linked to how search results outside of Wikipedia appear, and what people will see when arriving at a specific article title from outside of Wikipeda. For example, the most prominent feature of a Google search result for a Wikipedia article is the the title of the article, including the disambiguation term if that is part of the article. It therefore makes a significant difference whether we have selected a primary topic or not. Similarly, if I use my phone's operating system features to pull up the Wikipedia article with a given name (not really what I would think of when thinking of searches "on Wikipedia"), where I land depends on how we have titled the articles.--Trystan (talk) 18:53, 18 September 2014 (UTC)
    See response to Dicklyon below. -- JHunterJ (talk) 11:04, 19 September 2014 (UTC)
  • But navigation hints should be based on what the user knows, not what the project helds. Requiring that readers . If you draft the disambiguation pages exclusively using the topics covered, and not the meaning, sound and overall usage of the term, you'll be forcing readers to "get into the on Wikipedia mindset" as well. Diego (talk) 17:03, 30 September 2014 (UTC)
  • Oppose "on Wikipdia". Ambiguity is ambiguity. Readers don't much care which other articles Wikipedia has. When an article title shows up in a search, having a title the unambiguously specifies the topic is always a good thing. Dicklyon (talk) 19:15, 18 September 2014 (UTC)
    How the article is titled is a different question than what the primary topic for a title is. The primary topic can always be the target of a redirect from the ambiguous title if the topic is better titled with something else. -- JHunterJ (talk) 11:04, 19 September 2014 (UTC)
  • Oppose "on Wikipedia", per the above. Ambiguity is, indeed, ambiguity; it doesn't have to be limited to WP, arising from WP, or even obviously existing on WP (we do have some articles that are "pre-emptively disambiguated". The ongoing rash of animal breed name RMs raises some of these issues.  — SMcCandlish ¢ ≽ʌⱷ҅ʌ≼  16:36, 30 September 2014 (UTC)
  • Oppose per Dicklyon and SMcCandlish. Both DAB pages and primary topics should be mostly decided based on the common knowledge about the ambiguous term, not the subset of that knowledge that happens to be included in our incomplete project at any particular time. Diego (talk) 17:03, 30 September 2014 (UTC)
  • Support, if only to stop, or at least to slow down, the proliferation of endless, repetitive, time-wasting, and ultimately useless discussions about the ever so subtle degrees of "ambiguity" any given term can have "in real life". Adding these two simple words confines the definition of "ambiguity" to Wikipedia articles, and ambiguity "on Wikipedia" is the foundation on which the whole structure of disambiguation is built (with only a handful of exceptions, some of which were made for good reasons and others not). Wikipedia's scope is enormous, but not limitless. If a title of one article would be ambiguous with a title of an article describing some other concept, but that latter article does not exist, either create (or request, or at least demonstrate the encyclopedic value of) that article or think whether that topic is missing for a valid reason (such as being non-notable, for example). There's no need to expand the definition of ambiguity to the infinite sundry of all possible subjects; doing so will serve no one well.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 30, 2014; 17:24 (UTC)
  • Support (Double Support if nailing this down to en) as disambiguation is not about solving Google's search, LinkedIn, Wiktionary, Facebook or other info silos. Widefox; talk 22:57, 15 October 2014 (UTC)
It is about helping users find the relevant information within the Wikipedia/Wiktionary silo, though. That involves taking into account the real-world knowledge that readers from outside the Wikipedia project may have. Diego (talk) 17:13, 22 October 2014 (UTC)
No, it is not our job to help users find dictionary words on Wiktionary, certainly not at the cost of making it harder to find articles on WP with names that match the term at issue. --В²C 22:05, 4 November 2014 (UTC)

Further on usage[edit]

I do, however, agree that the language is contorted and begs copy-editing. I suggest the following change:

From
A topic is primary for a term, with respect to usage, if it is highly likely—much more likely than any other topic, and more likely than all the other topics combined, to be the topic sought when a reader searches fora reasonable user would expect to find under that term.
to
A topic is primary for a term, with respect to usage, if it is much more likely than all the other topics combined, to be the topic a reasonable user would expect to find under that term.

to get away from tying "primary topic" and article titles to searches. Articles should not be altered to satisfy poor search strategy. Wikipedia now, long since, has an excellent search engine with learning algorithms, and this search engine far better serves search queries than the occasional title fiddling. If the topic is primary, readers should get what they were expecting, or at least in hindsight not be astonished, regardless of how they came to be reading the article. --SmokeyJoe (talk) 03:39, 18 September 2014 (UTC)

I think we need to change "more likely" to "much more likely", in recognition of the situation that getting to an article you didn't expect is a much worse experience than getting to a disambig page that helps you find what you want. Dicklyon (talk) 03:44, 18 September 2014 (UTC)
"much more likely than all the other topics combined". I slipped this "much" quickly in.[1] It is more restrictive than the previous language (only much more likely than a single other topic, only "more likely" than all combined). I agree that "much more likely" is much more sensible. "more likely" is kind of weak. --SmokeyJoe (talk) 03:58, 18 September 2014 (UTC)
I don't think that's necessarily true. I have asked the Foundation to conduct some empirical research about the experience that users have when searching for ambiguous or potentially ambiguous topics, but I don't think we can honestly say that we know for sure that a reader arriving at a usage of the term that is "more likely than all the other topics combined" to be the target of the search experiences any great distress at the fact. For more distressful, I think, is arriving at a disambiguation page containing a fairly large number of arcane and closely related terms, like Congo (where, at least, we have broken our usual rule and included a helpful image). bd2412 T 04:04, 18 September 2014 (UTC)
Your logic, which I don't disagree with, applies equally to the existing text, doesn't it? We don't *know* "the topic sought when a reader searches".
I don't think a disambiguation page is ever more distressing than astonishment. Astonishment, as when you think you know what you are downloading, you get something you find to be unrelated, and there is not a logical route apparent to what you did want.
Is there a problem with a disambiguation page containing a fairly large number of arcane and closely related terms. Probably. Surely though that is a different problem than having primary topic defined in terms of unknowable hypothetical search behaviour and expectations. On Congo, yes it is a big information dump if you are expecting to arrive straight at the page covering a nation/place called "Congo". Perhaps the top three most likely wanted pages should be listed in the dab page lede. Maybe "other possibilities" could be in a collapsed box. Why is there a rule against a small helpful image? --SmokeyJoe (talk) 04:30, 18 September 2014 (UTC)
Oppose. This whole concept of "search engine strategies" has no bearing on primary topic determination. Primary topic determination is about enabling users using the basic WP search mechanism, which doesn't use a strategy, but just takes the user to the article matching the name of the entered term. When a search algorithm is used, primary topic is irrelevant, because they take article content into account and criteria such as how often users pick a certain article out of the results, etc. Titles are hardly relevant when search algorithms are used. --В²C 22:11, 4 November 2014 (UTC)

Disambiguating the Measure B article[edit]

Opinions are needed on the following matter: Talk:Measure B (Los Angeles, 2012)#Article move. A WP:Permalink to that discussion is here. Flyer22 (talk) 01:57, 3 November 2014 (UTC)

PROPOSAL: delete the historical significance consideration from PRIMARYTOPIC[edit]

Once again WP:PRIMARYTOPIC is being used to support and oppose the same proposal. This time see Talk:Worcester#Requested move. As I argue there, this is the quintessential example for why the historical significance criteria should be removed from WP:PRIMARYTOPIC. It is not helpful to send people to an article about the English city when they're seeking an article about the US city, or a sauce. ALL that should matter is what people are most likely seeking when entering the search term in question into the search field and clicking on GO. The minute you allow for any other consideration, including the vague and meaningless "historical significance", you're throwing the user under the bus and just creating a contentious situation for no good reason. This is about numbers, period. Do we serve 40% and hinder 60%, or vice versa?

When people cite the same policy/guideline as supporting both sides, something is very wrong.

Therefore, I hereby propose that the historical significance consideration be removed from WP:PRIMARYTOPIC. --В²C 00:08, 4 November 2014 (UTC)

UPDATE: Specific proposal. Current wording:

There is no single criterion for defining a primary topic. However, there are two major aspects that are commonly discussed in connection with primary topics:

A topic is primary for a term, with respect to usage, if it is highly likely—much more likely than any other topic, and more likely than all the other topics combined—to be the topic sought when a reader searches for that term.
A topic is primary for a term, with respect to long-term significance, if it has substantially greater enduring notability and educational value than any other topic associated with that term.
In other words, the proposed changed wording:

A topic is primary for a term if it is highly likely—much more likely than any other topic, and more likely than all the other topics combined—to be the topic sought when a reader searches for that term.

--В²C 02:28, 5 November 2014 (UTC)
  • Diametrically disagree. Historical significance fits the core purpose of an encylopedia, to cover significant things in an historiographical, scholarly manner . Search likelihoods and click rates are matters for search algorithms. Way better would be to remove the loaded, confusing, "PRIMARYTOPIC" term and replace it with something that can be easily explained. --SmokeyJoe (talk) 00:29, 4 November 2014 (UTC)
    • Well, you've got it totally backwards. Do you not realize that when our users type a term like Worcester into the search box and click Go, there is no "search algorithm" that applies? It just Goes to the page that matches that term. Users who are going through a search mechanism that use search algorithms, like Google, are shown the search results based on all kinds of criteria independent of the article title. That's why Googling for Nixon results in our article for Richard Nixon, not any of the redirects to the article, including Nixon. We could title that article X451Qsujk&vvv and the Google search for "Nixon" would still show it. The title matters to the WP search/Go scenario; not to the search algorithm scenario. That's what PRIMARYTOPIC has always been about - to make sure that the most likely topic to be Go searched with a given term is the one associated with that term. Maybe the name "primary topic" is confusing, but the underlying meaning is definitely not "the most historically important topic associated with that term"; it's "the topic most likely to be sought with that term". --В²C 00:49, 4 November 2014 (UTC)
  • Strong Oppose - both likelihood and long-term significance seem sensible to counter WP:RECENTISM / WP:NOT#NEWS Widefox; talk 01:12, 4 November 2014 (UTC)
    • You did not address the problem I presented, did you? --В²C 18:52, 4 November 2014 (UTC)
      • To clarify your proposal, you've made a proposal based on what you see as a problem. The proposal is flawed as it doesn't address my (and other editors) concerns that WP is an encyclopaedia WP:NOT#NEWS, NOTGOOGLE, and dabs aren't search engines so a consideration of encyclopaedic factors (RECENTISM / longevity / educational worth / WP:SYSTEMIC) is prudent and in-line with that, despite how less quantitative / objective those factors are compared to simple viewcounts. I've now changed to Strong oppose, as I oppose this in principle and practice and per the flawed example (which appears to currently have no consensus). In summary, the problem you've presented here is oversimplification, with a bad example. Both the example discussion and here is flawed by weak proposals, and as PRIMARYTOPIC and that example may need looking at, there should be no prejudice for a better worded proposal at a later time. For instance, the English town name is the source of the US, and sauce (if you pardon the pun), so there may be encyclopaedic merit in having the minority seeing it instead of a dab for US / sauce readers. There's little merit the other way around. As for having the dab located at the basename, the US article is still a click away (so it doesn't help them) and by ignoring the UK city readers in your proposal (and any merits of that) they are disadvantaged by adding a click. More pain, no gain. I personally think a clear discussion of pageviews alone for it at a later date may indicate there's no clear primary topic, but that debate has been muddied. There's a lot of merit in BD2412's view of historical significance trumping pageviews. Widefox; talk 12:49, 5 November 2014 (UTC)
  • Oppose pretty much as strongly as possible. Ephemeral pop groups, songs, albums, films, and other such topics have long been known to adopt the names of existing topics of historical importance, enjoy a spike in popularity while they are being heavily advertised, and then subside into the ether from which they came. Even if they enjoy popularity that is unusually long for their media, they still do not and can not play as important a role in history as the things whose names they take. No matter how popular an album titled "Parachutes" may be, it will never be as important a topic from a historical perspective as the fabric device that carried troops into enemy territory to make it possible for World War II to turn out the way it did. The problem with the discussion at Worcester is that it really is weighing apples and apples, rather than the usual apples and oranges situation to which this debate applies. If this were between any city named Worcester on the one hand, and recently popular album titled Worcester on the other hand, there would be no contest. It is because we are discussing two old cities, one older, one larger, that we have an absence of a primary topic. bd2412 T 01:13, 4 November 2014 (UTC)
    • You're not addressing the problem that exists any time the two criteria indicate different answers to the same question ("Is there a primary topic, and, if so, what is it?"). What kind of guideline does that? --В²C 18:52, 4 November 2014 (UTC)
      • Hey, if it was up to me, historical significance would always prevail over recent pageviews. bd2412 T 18:56, 4 November 2014 (UTC)
        • Great. Let's send not just 60%, but 99% of the users to an article they're not searching for. --В²C
          • What makes you think that going by historical significance would tend toward 99% (or even 60%)? Over the long term, the term of greater historical significance is also likely to draw the most reader interest. bd2412 T 20:01, 4 November 2014 (UTC)
      • The dilemma you present is a false dichotomy, and the guideline offers the right solution for that case: "if there is no primary topic, the term should be the title of a disambiguation page". I.e. if none of those criteria dominates over the other, let the user choose what they're looking for, instead of forcing the wrong article upon them. The guideline does not need any change. Diego (talk) 20:20, 4 November 2014 (UTC)
  • Oppose. A real problem has been cited, but this isn't a solution. As BD2412 noted, the "Worcester" case contains incorrect applications of the guideline. Perhaps we could clarify its wording to reduce the likelihood of confusion. Let's not throw out the baby with the bathwater by failing to recognize the criterion's important role in minimizing recentism. —David Levy 02:36, 4 November 2014 (UTC)
    • Worcester is just one example. Don't get hung up on it. There are countless others, both actual and potential. This "guideline" gives two answers to the same question not just once in a while, but perhaps even in the majority of applicable cases. --В²C 18:52, 4 November 2014 (UTC)
      • I'm aware that Worcester is just one example. I've personally encountered others. As I noted, a real problem exists, but I disagree that the proposed change is a viable solution. The issue stems from the guideline's misinterpretation, so we should seek to clarify its wording. Abandoning the criterion in question would eliminate valid and invalid applications alike, thereby replacing the current problem with a worse one. (See Yaksar's and Vegaswikian's comments.) —David Levy 22:36, 4 November 2014 (UTC)
  • Comment: It would probably be best to not advocate a policy change until after the debate where you’re citing that policy has concluded. —174.141.182.82 (talk) 07:17, 4 November 2014 (UTC)
  • Oppose - massively disruptive and shortsighted; why don't we just change the url to www.shitpedia and be done with it. In ictu oculi (talk) 07:40, 4 November 2014 (UTC)
  • Oppose. In my experience it is not unusual for people on both sides of an issue to cite the same policy. That doesn't mean there is anything wrong with the policy. Omnedon (talk) 14:32, 4 November 2014 (UTC)
    • Facepalm. --В²C 18:52, 4 November 2014 (UTC)
  • Oppose - if you're looking to serve the highest number, then you're best off having a disambiguation at the main name. When it's something relatively close (i.e. pageviews on one topic doesn't consistently outnumber the other by large factors), you're best off just disambiguating and letting people making their own choice, rather than forcing a large number of people still to the wrong page under a false dichotomy. The numbers you give seem to be based on the pageviews of 47,000 and 39,000 for the US and UK cities respectively, making 55% and 45%, which is closer than you suggest (only 1.2 times as many people for the US city). So you can either serve 55% and hinder 45% by swapping them over based purely on pageviews for one period, or in cases that close, you can serve 100% of the readers by giving them an informed choice with a main name disambiguation page. — Sasuke Sarutobi (talk) 16:10, 4 November 2014 (UTC)
    I disagree somewhat with that interpretation of the outcome. If 100% of the readers are looking for a specific article (i.e. not a disambiguation page), then taking them to a disambiguation page disappoints 100% of the readers. It is important to remember that a disambiguation page is merely a tool to assist in navigation. In a situation where, say, 55% are looking for Option A, 40% are looking for Option B, and 5% are looking for one of a group of obscure options, then having Option A at that title with a hatnote pointing to Option B immediately satisfies 55% of readers while providing no more inconvenience to those seeking Option B than going to a disambiguation page would (since that option is right there in the hatnote, one click away). bd2412 T 16:42, 4 November 2014 (UTC)
    • Why would showing a list of all the content that Wikipedia has to offer under the searched title be a disappointment to readers? If that was true, Google would be such a failure. Assisting in navigation is the right choice when there are several navigation targets. And you assume that showing the wrong article has zero cost, which is false - noticing that one has arrived to the wrong article has a significant recovery penalty if the page looks like the expected result, but is not it. Diego (talk) 20:26, 4 November 2014 (UTC)
      • I have yet to see any evidence of just how much of a "recovery penalty" there is for landing on the wrong page. Presumably, this is not the same for all readers or for all topics. For example, if a reader is looking for the company, Apple, and arrives at a page about a fruit, they should very quickly be able to find that Apple, Inc. is in the hatnote and go right to it. For a situation like two cities named Worcester, it might take a few moments of additional reading to realize that the wrong one has been reached, but that again is a problem of comparing Apples to Apples rather than Apples to Oranges. bd2412 T 21:35, 4 November 2014 (UTC)
    I agree that the penalty will largely vary between topics and readers; that's why sometimes having a primary topic makes sense, instead of allways showing a disambiguation page. Apple is a very good example of the reason why the long-term significance criterion is needed. Apple Inc. gets three times as much visits as the fruit, yet no one would suggest placing the company article as the primary topic even if its WP:COMMONNAME is "Apple", not "Apple Inc.", and no one would be surprised to find the fruit at the base name.
    As for the penalty of landing into the wrong page, you need to look no further than at loading times. A typical disambiguation page is between 1Kb to 5Kb long, while the average PRIMARYTOPIC candidate can be 30 to 60Kb in size, plus images. There's people accessing Wikipedia on a data plan on mobile or through third-world networks, and size matters a lot there if you're loading useless information that will be immediately discarded. Diego (talk) 16:51, 5 November 2014 (UTC)
  • Support. Whether a subject is a primary topic or not should be decided on a case-by-case basis anyway. And while there are undoubtedly situations where historical significance will play a role in determining which topic can or should be considered "primary", to legislate it as a guideline is unnecessary and redundant. We don't enumerate all other possible factors that may play a role in a primary topic debate, so singling out "historical significance" makes little sense. Removing this bit will simplify the guideline with no ill effects (apart from reducing the pomposity of the "encyclopedicity" mavens), and when we have a chance to chew even a tiny bit off our bureaucracy, we should take it.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); November 4, 2014; 16:25 (UTC)
    • Exactly. When "historical significance" is truly significant, it is reflected in usage in reliable sources and page view counts anyway. Calling it out separately only allows us to give "historical significance" more significance than reliable sources do. It's a form of original research, actually. --В²C 19:07, 4 November 2014 (UTC)
    • That would make sense if we also remove the bit about "primary with respect to usage", for those very same reasons. Diego (talk) 20:30, 4 November 2014 (UTC)
      • For the record, I have no problem with removing that bit as well.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); November 4, 2014; 21:07 (UTC)
        • Primary with respect to usage is based on the assumption that how frequently a given term is used to refer to a given topic in reliable sources aligns and predicts well how likely that term is to be used to search for an article about that topic, relative to other uses of that term. It's not a perfect, but it's a true guideline that works well in the vast majority of cases. We also use page view counts to strengthen or weaken a case for primary topic. So the reasons to exclude long-term significance do not apply to removing with respect to usage.
          • I didn't say usage shouldn't be considered when discussing which topic (if any) should be primary. There are situations where it needs to be taken into account, just like there are situations where historical significance needs to be taken into account. There are also situations where a whole range of other factors may need to be taken into account. We don't need to enumerate each possible one in the guideline, though. Acknowledging that from time to time situations arise, which warrant a consideration of primary topic, is the only useful purpose of WP:PRIMARY. Codifying specific guidance is nothing but redundant wikilegalism.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); November 4, 2014; 21:47 (UTC)
  • Oppose per [[User:BD2412 and his "apples to apples" comment. This consideration really is necessary to prevent issues of recentism, which lead to cases where a subject which is newer or popular in the moment will get more views even though in the long term these can be expected not to last. Of course, what it's not meant to do is be used as an argument from preventing articles which have received higher views for the entire lifespan of Wikipedia and have no indication of ever dropping in relative interest or significance from being moved. There are still obvious exceptions -- for example we can comfortably say that the musician Prince should not be the primary topic over the concept Prince regardless of how page views turn out in the long term. But when comparing apples to apples, as in two cities, neither of which has recently experienced some dramatic upswing in coverage or interest, recentism isn't really a concern. The case in question, Worcester, is not one of these "Prince" exceptions. Similar faulty arguments were made in the move discussions for Boston, for example, where long-term significance was interpreted by some to mean "The UK city came first" when one can say that the significance, both historically and for a long term view which shows no sign of changing the MA city being the primary is pretty much common sense. This guideline has been misinterpreted, either unintentionally or in some cases deliberately for narrow-minded nationalistic reasons, but that does not mean it is not still necessary.--Yaksar (let's chat) 16:56, 4 November 2014 (UTC)
    • But basically, the guideline does not say to base it on historical significance, but rather long-term significance, a concept which takes both historical, current, and future significance into account.--Yaksar (let's chat) 16:58, 4 November 2014 (UTC)
    • How do you propose to assess future significance? Diego (talk) 20:32, 4 November 2014 (UTC)
    • We can assess likely future significance by examining the dropoff in significance over time of other topics of the same genre, and projecting a similar pattern with respect to current topics. I suspect that if Wikipedia had started in the 1970s, some people would be arguing that the primary topic of Bananas should be a Woody Allen movie. bd2412 T 21:39, 4 November 2014 (UTC)
  • Neutral. I think the problem is that historical significance is being simplified to age in discussions. Clearly that should be clarified since how much older one place is then another is not the determining factor. Vegaswikian (talk) 18:10, 4 November 2014 (UTC)
  • Question - What is the actual wording of the line we want deleted or changed here?--Yaksar (let's chat) 19:40, 4 November 2014 (UTC)
    • I added an update to the proposal with the specific wording. --В²C 02:28, 5 November 2014 (UTC)
  • Oppose because of the example given by B2C. Particularly as we have WP:USPLACE to complement problems of primacy over issues like Worcester (as the US city is already naturally disambiguated using the state under WP:USPLACE it reduces the need to disambiguate Worcester, Worcestershire--which is not usually done in British English). Indeed B2C is this proposal a round about way to work towards the ending of WP:USPLACE? -- PBS (talk) 20:36, 4 November 2014 (UTC)
  • Well, the disambiguation standard set by WP:USPLACE doesn't really concern primary topic. It's more to deal with issues of identifiability. If a city in the UK (or anywhere) was named Sacramento, it would still need to be disambiguated, even though the article on the US city is at Sacramento, California.--Yaksar (let's chat) 20:41, 4 November 2014 (UTC)
  • To answer PBS (talk · contribs), no, this has nothing to do with USPLACE. I've given up on that fustercluck. --В²C 21:34, 4 November 2014 (UTC)
  • Suggestion to open new discussion. I am more fond of the new proposal than outright removal of 'historical significance', but I think that it is substantial enough a change in wording to warrant a new discussion, lest any chance of consensus gets lost in debating two different things. I propose that we draw a line under this discussion, and open a new one for the new wording, as I believe that, while the new wording still seems vulnerable to recentism, it makes a strong and valuable suggestion to help clarify the existing criteria for PRIMARYTOPIC.Sasuke Sarutobi (talk) 14:08, 5 November 2014 (UTC)
    • How is the "new wording" any different than the original proposal, which was to remove everything not in this new wording? bd2412 T 14:35, 5 November 2014 (UTC)
      • I was under the mistaken impression that the likelihood test was a replacement proposal, rather than just the one not to be removed. So even after the update, I still oppose the proposal. — Sasuke Sarutobi (talk) 18:03, 5 November 2014 (UTC)
  • Procedural close - I'm not quite sure what the procedure could be for something which just starts as "PROPOSAL" without a RFC template to attract a wider selection of editors, but it's already clear that without approaching WP:SNOW this propsal isn't wanted and is sucking bytes out of article space edit time. Propose someone closes it. In ictu oculi (talk) 00:27, 6 November 2014 (UTC)
2nd that. (Commenting similarly in the example move request). Widefox; talk 00:39, 6 November 2014 (UTC)
  • Support if only to eliminate the idea that being the oldest thing named "XYZ" automatically makes something the primary topic for "XYZ." -- Calidum 03:46, 6 November 2014 (UTC)
    • I’m not sure if it’s a matter of personal interpretation, but I thought age alone was irrelevant here. If something came along later, years after the original XYZ, and became more historically significant than the older thing, wouldn’t the guideline as is favor the more recent XYZ? —174.141.182.82 (talk) 06:35, 6 November 2014 (UTC)
      • Theoretically, yes. But that's not always the case in practice. I don't want to get into specific examples, but there is a fairly major going on right now as we speak where editors are arguing that older=prime topic in the face of other evidence. -- Calidum 17:03, 8 November 2014 (UTC)
  • Comment/Question. Several have suggested the problem is that long-term or historical significance is sometimes simplified to age. That may be a problem but it's not the problem. Even if you look at long-term significance properly, it can still indicate a different title than usage in reliable sources indicates what is most likely to be sought. That is the problem. And again, the extent to which long-term significance is, well, significant, this is also reflected in usage and likelihood of being sought - it's intrinsic in that criteria - so why call it out separately? Why give it priority when it's not significant enough in terms of likelihood to be sought? --В²C 05:38, 7 November 2014 (UTC)
    • To counter-balance naked "likelihood to be sought". Likelihood to be sought is not obviously a bad concern, but if taken too far it leads to editors psycho-speculating readers intent, performing original research using non-scholarly tools, and generally thinking at odds to how an encyclopedia editor should be thinking. --SmokeyJoe (talk) 10:59, 7 November 2014 (UTC)
      • But is "likelihood to be sought" naked? Isn't the long-term significance of a given topic very important and often the most significant factor that determines how likely that topic will be sought? Further, Wikipedia is an encyclopedia but it is different from traditional encyclopedias, and one of those differences is the way titles are chosen. From the beginning commonly used titles have been preferred overly "scholarly" or "official" titles, and the way "commonly used" is determined is by looking at usage in, well, commonly used sources like newspapers and popular magazines, not just the scholarly sources by which titles in more traditional encyclopedias might be chosen. And use of ambiguous names as titles in particular, the realm of "primary topic", therefore best reflects likelihood of being sought by that name by our users. --В²C 17:16, 7 November 2014 (UTC)
        • Suppose we were just to go with "likelihood to be sought", but were to qualify that by saying, "likelihood to be sought (on average, over the next thousand years)"? When we discuss pageviews, we usually look at current pageviews generated by Wikipedia's own stats generator, which gives us the pageviews for up to 90 days. That is too short a span to say anything about the place of a topic in history. However, we can't gauge the likely number of pageviews that a page will receive over the next thousand years by any means but by weighing the relative historical importance of the topics, and seeing how those have tended to stack up over the last thousand years. bd2412 T 18:18, 7 November 2014 (UTC)
          • Do any of these proposals clarify being sought? Lets say the subject is Foo and we have two articles, Foo1 and Foo2. When people talk about Foo, they really mean Foo2, but they think they are talking about Foo1. Vegaswikian (talk) 19:26, 7 November 2014 (UTC)
            • BD2412, doesn't WP:CRYSTAL tell us not to go there? I think all that matters is what usage today indicates, ignoring obviously short (a few months) and temporary bursts of popularity per WP:RECENT.

              Vegaswikian, like dictionary writers, our job is to reflect usage, not correct it. If people are using Foo1 to refer to Foo, then Foo1 should be the title of, or redirect to, the article about Foo, which probably should encompass the topic in the article titled Foo2. --В²C 21:44, 7 November 2014 (UTC)

              • Have you read WP:CRYSTAL? It pertains to nothing more than article content. There’s nothing in there about any other kinds of editorial decisions. —174.141.182.82 (talk) 16:47, 8 November 2014 (UTC)
    Wikipedia should not be make titling discussions in a attempt to emulate search engine functionality. --SmokeyJoe (talk) 04:02, 8 November 2014 (UTC)
    Just because CRYSTAL is primarily about content does not mean it does not apply to titles. Anyway, how does deciding on titles based on our guessing about future usage improve the encyclopedia? Smokey, it's not about emulating search engine functionality. It's about making it easier for users to find the articles they are seeking - in fewer clicks. --В²C 22:57, 8 November 2014 (UTC)
    It means we’re less likely to have to rename things based on current events and popularity. I thought you were always in favor of title stability? And by your same logic, just because CRYSTAL isn’t explicitly limited to article content does not mean it philosophically applies to titles. —174.141.182.82 (talk) 23:43, 8 November 2014 (UTC)
  • Oppose because the "historical significance" for ephemeral subjects is typically brief. One of the best reflections could be made for the September 11th attacks which refer to 2001 and not the embassy attack in 2012. Perspective on such things is not always shared, but things work themselves out typically. ChrisGualtieri (talk) 19:33, 7 November 2014 (UTC)
    • But September 11th attacks being about 2001 and not 2012 demonstrates the point of this proposal - when long-term significance is sufficiently important, it's reflected in usage and "most likely to be sought" determination. The topic most likely to be sought by anyone searching for "September 11th attacks" is the 2001 attacks, not the 2012. There is no need to explicitly invoke the "long term significance" criteria to get that right. The problem occurs when the more likely to be sought topic is perceived to be "less important" by some that view the other topic to have "long term significance". That just creates unnecessary urinating matches that would be entirely avoided if we went by "most likely to be sought", period. --В²C 21:50, 7 November 2014 (UTC)
    September 11 attacks is now so old that school children in most of the world don't know about them. Dates without year imply the current year. The article should be disambiguated with the year, per recognizability concerns. --SmokeyJoe (talk) 04:02, 8 November 2014 (UTC)
    Yeah if "Benghazi attacks" wasn't the convention quickly adopted due to the that issue.... though even still, your case further elevates the fact that said discernment issues typically resolve themselves. ChrisGualtieri (talk) 05:50, 8 November 2014 (UTC)
  • Support. As to the repeated concern about recentism: Our setup isn't carved in stone. If someone named Frank Porter Graham gets in the news as a defendant in a highly publicized murder case, and a lot of people want to read his bio, then we can put his bio at that name with a redirect to the former Senator who's there now. Then, maybe after a year or two, the murder fades from the headlines and the Senator's historical significance re-asserts itself in terms of what readers are looking for, so we move the article back. During the time that the murder trial is prominent, and lots of people are coming to Wikipedia for information, there's no reason to send them first to an irrelevant article. JamesMLane t c 02:39, 9 November 2014 (UTC)
    In a case like that, I don’t believe it would be possible to say which individual would ultimately have more historical significance. If this trial were on the level of the O. J. Simpson trial, I would argue the defendant trumps the senator. Or the defendant could be acquitted on the first day and fade into obscurity. It’s just impossible to judge. —174.141.182.82 (talk) 04:55, 9 November 2014 (UTC)
    That's precisely the point. Since it's impossible to judge the future, what we should go by is the current situation. If a use is consistently getting significantly more traffic than all other uses of a name, then it is the primary topic, regardless of the "long-term significance" of any of the other uses. What happens in another year or two or five or ten is a separate matter, to be evaluated then. In almost all cases usage distribution will stay the same, but of course in some cases adjustments will have to be made, to serve our readers best. --В²C 17:55, 10 November 2014 (UTC)
  • Oppose - In my opinion, the purpose of Primary Topic is less about taking the reader to the most popular topic, and more about making sure that when searching for apple, the reader is not taken to an article about the corporation (no matter how popular); and from that point of view, historical significance is infinitely more important than page hits, for example. It is a safeguard to ensure we remain encyclopaedic, and looking at earlier versions of the guidelines, it appears that this was foremost in the minds of those editors too, with phrases like, “most important” and “central meaning”, and no mention at all of how likely the topic is to be sought. Even the word 'primary' means, first in importance. It does not mean most popular. This whole infatuation with page traffic stats and Google hits, has its roots in this edit, [[2]] on the 28 July 2008 at 05:13, where there is an attempt to link usage to what people search for on Wikipedia. This concept is flawed because a significant amount of English speakers don’t use Wikipedia or even have access to the internet. I would much rather see a proposal to remove the ‘likelihood to be sought’ clause, which is unknown, has very little to do with common or established usage and (assuming that the primary term is not the one sought) only saves the effort of a couple of mouse clicks at the most. --Ykraps (talk) 14:14, 9 November 2014 (UTC)
An excellent point. Primary topic is not about web traffic (or search engine) optimization. The project is about writing an encyclopedia, not about making a web site that helps people find the trivia of the moment. olderwiser 14:42, 9 November 2014 (UTC)
  • Oppose the proposal as currently written. The Apple example has been brought up. There was a a 2012 move discussion related to this, where it was suggested to make the computer company the primary topic based on usage. If only the "usage" factor was in place back then, the computer company would have mostly likely been made the primary topic. IMO, that would have been unacceptable in trying to maintain some sort of legitimacy here on Wikipedia -- relegating a centuries-old, core vital encyclopaedic topic like that. And as User:bd2412 has mentioned, pop groups, songs, albums, films, and other such topics have long been known to adopt the names of existing topics of historical importance. What happens if Culture (band), Language (album), Life (film), or any other topic with the same name as a WP:VA generates far more traffic? Under the proposal as written, they would then be made the primary topic over these more vital core encyclopaedic ones. Wikipedia already has a bad reputation for having a systemic bias toward pop culture topics over more encyclopaedic ones. Let's not make it worse. Zzyzx11 (talk) 17:29, 9 November 2014 (UTC)
    The cultural topic bias is more from looking at FAs on easy as pie stuff in comparison to major and very complex topics that simply confound even expert editors and field experts on how to properly construct. A fair point, but a vital topic should definitely elevate it for reasons other than purely historical. ChrisGualtieri (talk) 18:14, 9 November 2014 (UTC)
  • Oppose. Figured I’d weigh in since I’ve participated in discussion here. Wikipedia is not a news source. Some people do look to Wikipedia for news and updates on current events, but we should not cater to people who don’t treat this as an encyclopedia. That is not Wikipedia’s job, and that is not what ought to be expected. As an encyclopedia, the primary topic for any given title should be one with long-term significance. Some here claim that we would best serve the readers by helping them find articles relevant to today’s news, and the significance clause gets in the way of that goal; I say that would improve a news wiki, but not an encyclopedia. —174.141.182.82 (talk) 23:39, 9 November 2014 (UTC)
    To reiterate: The fundamental issue, the foundation of what Wikipedia is WP:NOT, is this: Wikipedia is an encyclopedia. That is why the long-term significance consideration is important. It’s not an encyclopedia of “now.” As best as possible, it should be as relevant next Tuesday or next year as it is today. Choosing titles based on what’s popular right now conflicts with that basic principle. —174.141.182.82 (talk) 20:41, 11 November 2014 (UTC)
  • Oppose. I think the two different criteria, when applied together, generally result in the best results. I think we want to capture not only what readers happen to access the most, but what they might reasonably expect to find at an article title, and historical significance is a part of that. It's not simply a question of getting to your destination article the quickest, but of having the encyclopedia organized in an intuitive way.--Trystan (talk) 04:45, 11 November 2014 (UTC)

Should we create a {{surnamedis}}[edit]

We have thousands of essentially disambiguation pages masquerading as articles, many containing assertions about WP:BLPs without sourcing at all. The {{surname}} is often used on pages where there's little but links to two or more persons with the same surname. In essence, this is a disambiguation page in all but name. The documentation at {{surname}} states that there should be discussion of the origin of the surname. Presumably, as all material at WP must be sourced, this must also be. However, that is the exception rather than the rule. However, {{hndis}} is available for ambiguous names, where it can be placed on such pages as Michael Granger where people with that name are listed out dab style. I think rather than having a bunch of unreferenced "surname" pages, we ought to implement a {{surnamedis}} to hold those pages until someone can bother to write and source something about the surname and source each person on the list as having the surname as thus explained - for instance, someone with the surname Lee may or may not have something to do with a Chinese etymology of the name. I think this will keep surname pages that are essentially disambiguation pages (they say nothing about the surname, just who has it) are properly categorized and maintained, rather than being unsourced articles containing assertions of living people. Carlossuarez46 (talk) 20:59, 4 November 2014 (UTC)

  • I think the more helpful thing would be to tag all of those surname pages as unsourced, and where possible, to merge together pages for surnames having a common origin. bd2412 T 03:31, 5 November 2014 (UTC)
  • Perhaps, but there are perhaps tens of thousands of them. Carlossuarez46 (talk) 05:42, 5 November 2014 (UTC)
I agree that these should generally be dab pages - i.e. anyone linking to such a page should be alerted that the link should be fixed. Edits like this[3] are not improving wp. However, I'm not sure a surnamedis template is needed; it is better to use {{disambig|surname}} as otherwise it could be confusing for an editor who wants to add an entry for something that is not a surname (a place, a ship, a company etc). DexDor (talk) 06:53, 5 November 2014 (UTC)
They are not disambiguation pages. If you subscribe to this nonsense that every individual entry in a list needs to be referenced or it's "violating BLP", why does that magically stop being the case if you pretend the article is a disambiguation page? —Xezbeth (talk) 08:24, 5 November 2014 (UTC)
  • Oppose They are WP:SIA and as list articles they still need sources. Tag unsourced. Right problem, wrong fix. Widefox; talk 11:53, 5 November 2014 (UTC)
  • Oppose. Surname list articles are not essentially dab pages masquerading as articles. Since they are partial title matches and the topics are not ambiguous, they are not disambiguation pages, not in name, and not in function. They are list articles, and some of them are crappy list articles. Unencyclopedic list articles should be deleted, not shielded behind a mask pretending to be disambiguation pages. See also Wikipedia:WikiProject Anthroponymy/Home backup#Background reading -- JHunterJ (talk) 13:25, 5 November 2014 (UTC)
  • JHunterJ's comment is an interesting one. Given that many have some BLP assertions, perhaps new ones get the ole blp-prod treatment if they lack sources. Perhaps for the best, since we do have a search engine that will find all the Jane Does and John Does, when you search for "Doe", the page is unnecessary unless you need to find which John Doe, and then we have {{hndis}}. As for the older ones that cannot be blp-prod'ed, just the old prod as unnecessary but I think that someone will find it "useful" enough to make drama.... Carlossuarez46 (talk) 09:00, 6 November 2014 (UTC)
    • Do you even know what BLP stands for? —Xezbeth (talk) 09:51, 6 November 2014 (UTC)
      • BLP states "any material challenged or likely to be challenged must be attributed to a reliable, published source", bolding in the original. What on Michael Granger is likely to be challenged? Is someone going to say, "how dare you refer to Mike Granger as an athlete, I demand a source for that proposition"? bd2412 T 16:19, 6 November 2014 (UTC)
        • Coming back to typical SIAs, any intro should be sourced but that's not a BLP issue. There's sometimes contentious claims about BLPs on dabs and SIAs - think criminals and suspects etc, uncommon though. Widefox; talk 18:45, 6 November 2014 (UTC)
          • Exactly, and editors who seem to misunderstand BLP should understand that it's a policy that trumps editing guidelines like DAB. Carlossuarez46 (talk) 18:51, 6 November 2014 (UTC)
            • I'm not sure what your point is. A BLP issue on a dab or SIA (or anywhere) can be tackled per policy, with style guidelines like WP:MOSDAB or SIAs being unrelated. The conflation of dabs and SIAs in this proposal isn't helpful as per JHunterJ's clarification of the differences. No need to delete new SIAs or dabs just because they may contain BLP issues, just fix any BLP issues as per usual. Is this theory, or are there examples of an issue? Widefox; talk 21:21, 6 November 2014 (UTC)
            • @Carlossuarez46, for an SIA page like Farquhar, the propositions that need to be cited relate to the ancient origin of the name. I don't see where a BLP issue can arise there at all. I suppose we might need to source the proposition that "Regan Farquhar" is the birth name of rapper Busdriver, but even that doesn't seem particularly contentious. bd2412 T 22:53, 6 November 2014 (UTC)

WP:Primary topic for light bulb redirect[edit]

Opinions are needed on the following matter: Talk:Light bulb (disambiguation)#What is a light bulb?. A WP:Permalink to that discussion, which now has a WP:RfC, is here. Flyer22 (talk) 19:02, 9 November 2014 (UTC)

Links between dab pages[edit]

Regarding the avoidance of links to dab pages, I'm wondering if it wouldn't be best to allow them from other dab pages, if this is not already the case, and update relevant bots and guidelines in accordance. Otherwise, when a dab page is at the primary meaning (e.g. Rock, ROCK, Rok, etc), we need to link between them with either a piped link or through a redirect, each of which have their own downsides. It would seem like a simple algorithm of testing whether the page with the link is a dab page would resolve the matter, for bot and human. ENeville (talk) 00:27, 10 November 2014 (UTC)

  • Unfortunately, this does not avoid the problem that sometime disambiguation links on disambiguation pages are genuinely errors - for example, the page Good Intentions currently contains a link to the disambiguation page, Red Garters; obviously, the intent there is not to link to the disambiguation page, but to link to a particular album that shares that title and appears on that page. Exempting disambiguation pages from the bot reports would mean that errors like that would not be found and fixed. bd2412 T 03:37, 10 November 2014 (UTC)

Dab parameters[edit]

There is a discussion at Template_talk:Disambiguation#Dab_parameters about a recent change to the template documentation affecting how dab parameters are used. DexDor (talk) 07:30, 11 November 2014 (UTC)

MOSDAB documentation[edit]

Hello, came across this one under the Set index articles section of WP:MOSDAB: A set index article (SIA) is a list article about a set of items of a specific type that share the same (or similar) name. For example, Dodge Charger describes a set of cars

Seems to me the Dodge Charger article has gone astray, has become a broad concept article, and no longer qualifies for the SIA tag. Can someone confirm please? Thanks, --Midas02 (talk) 13:59, 15 November 2014 (UTC)